old-www/LDP/abs/html/abs-guide.html

147127 lines
2.3 MiB
Raw Permalink Blame History

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML
><HEAD
><TITLE
>Advanced Bash-Scripting Guide</TITLE
><META
NAME="GENERATOR"
CONTENT="Modular DocBook HTML Stylesheet Version 1.7"></HEAD
><BODY
CLASS="BOOK"
BGCOLOR="#FFFFFF"
TEXT="#000000"
LINK="#0000FF"
VLINK="#840084"
ALINK="#0000FF"
><DIV
CLASS="BOOK"
><A
NAME="AEN1"
></A
><DIV
CLASS="TITLEPAGE"
><H1
CLASS="TITLE"
><A
NAME="AEN2"
></A
>Advanced Bash-Scripting Guide</H1
><H2
CLASS="SUBTITLE"
>An in-depth exploration of the art of shell scripting</H2
><H3
CLASS="AUTHOR"
><A
NAME="AEN5"
></A
>Mendel Cooper</H3
><DIV
CLASS="AFFILIATION"
><SPAN
CLASS="ORGNAME"
><BR></SPAN
><DIV
CLASS="ADDRESS"
><P
CLASS="ADDRESS"
><TT
CLASS="EMAIL"
>&#60;<A
HREF="mailto:thegrendel.abs@gmail.com"
>thegrendel.abs@gmail.com</A
>&#62;</TT
></P
></DIV
></DIV
><SPAN
CLASS="RELEASEINFO"
>10<BR></SPAN
><P
CLASS="PUBDATE"
>10 Mar 2014<BR></P
><DIV
CLASS="REVHISTORY"
><TABLE
WIDTH="100%"
BORDER="0"
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
COLSPAN="3"
><B
>Revision History</B
></TH
></TR
><TR
><TD
ALIGN="LEFT"
>Revision 6.5</TD
><TD
ALIGN="LEFT"
>05 Apr 2012</TD
><TD
ALIGN="LEFT"
>Revised by: mc</TD
></TR
><TR
><TD
ALIGN="LEFT"
COLSPAN="3"
>'TUNGSTENBERRY' release</TD
></TR
><TR
><TD
ALIGN="LEFT"
>Revision 6.6</TD
><TD
ALIGN="LEFT"
>27 Nov 2012</TD
><TD
ALIGN="LEFT"
>Revised by: mc</TD
></TR
><TR
><TD
ALIGN="LEFT"
COLSPAN="3"
>'YTTERBIUMBERRY' release</TD
></TR
><TR
><TD
ALIGN="LEFT"
>Revision 10</TD
><TD
ALIGN="LEFT"
>10 Mar 2014</TD
><TD
ALIGN="LEFT"
>Revised by: mc</TD
></TR
><TR
><TD
ALIGN="LEFT"
COLSPAN="3"
>'PUBLICDOMAIN' release</TD
></TR
></TABLE
></DIV
><DIV
><DIV
CLASS="ABSTRACT"
><A
NAME="AEN31"
></A
><P
></P
><P
>This tutorial assumes no previous knowledge of
scripting or programming, yet progresses rapidly toward an
intermediate/advanced level of instruction <EM
>. . . all
the while sneaking in little nuggets of <SPAN
CLASS="TRADEMARK"
>UNIX</SPAN
>&reg; wisdom and lore</EM
>. It
serves as a textbook, a manual for self-study, and as a reference and
source of knowledge on shell scripting techniques. The exercises
and heavily-commented examples invite active reader participation,
under the premise that <TT
CLASS="USERINPUT"
><B
>the only way to really learn
scripting is to write scripts</B
></TT
>.</P
><P
>This book is suitable for classroom use as a
general introduction to programming concepts.</P
><P
>This document is herewith granted to the Public Domain.
<TT
CLASS="USERINPUT"
><B
>No copyright!</B
></TT
></P
><P
></P
></DIV
></DIV
><HR></DIV
><HR><H1
><A
NAME="AEN39"
></A
>Dedication</H1
><P
>For Anita, the source of all the magic</P
><DIV
CLASS="TOC"
><DL
><DT
><B
>Table of Contents</B
></DT
><DT
>Part 1. <A
HREF="#PART1"
>Introduction</A
></DT
><DD
><DL
><DT
>1. <A
HREF="#WHY-SHELL"
>Shell Programming!</A
></DT
><DT
>2. <A
HREF="#SHA-BANG"
>Starting Off With a Sha-Bang</A
></DT
></DL
></DD
><DT
>Part 2. <A
HREF="#PART2"
>Basics</A
></DT
><DD
><DL
><DT
>3. <A
HREF="#SPECIAL-CHARS"
>Special Characters</A
></DT
><DT
>4. <A
HREF="#VARIABLES"
>Introduction to Variables and Parameters</A
></DT
><DT
>5. <A
HREF="#QUOTING"
>Quoting</A
></DT
><DT
>6. <A
HREF="#EXIT-STATUS"
>Exit and Exit Status</A
></DT
><DT
>7. <A
HREF="#TESTS"
>Tests</A
></DT
><DT
>8. <A
HREF="#OPERATIONS"
>Operations and Related Topics</A
></DT
></DL
></DD
><DT
>Part 3. <A
HREF="#PART3"
>Beyond the Basics</A
></DT
><DD
><DL
><DT
>9. <A
HREF="#VARIABLES2"
>Another Look at Variables</A
></DT
><DT
>10. <A
HREF="#MANIPULATINGVARS"
>Manipulating Variables</A
></DT
><DT
>11. <A
HREF="#LOOPS"
>Loops and Branches</A
></DT
><DT
>12. <A
HREF="#COMMANDSUB"
>Command Substitution</A
></DT
><DT
>13. <A
HREF="#ARITHEXP"
>Arithmetic Expansion</A
></DT
><DT
>14. <A
HREF="#RECESS-TIME"
>Recess Time</A
></DT
></DL
></DD
><DT
>Part 4. <A
HREF="#PART4"
>Commands</A
></DT
><DD
><DL
><DT
>15. <A
HREF="#INTERNAL"
>Internal Commands and Builtins</A
></DT
><DT
>16. <A
HREF="#EXTERNAL"
>External Filters, Programs and Commands</A
></DT
><DT
>17. <A
HREF="#SYSTEM"
>System and Administrative Commands</A
></DT
></DL
></DD
><DT
>Part 5. <A
HREF="#PART5"
>Advanced Topics</A
></DT
><DD
><DL
><DT
>18. <A
HREF="#REGEXP"
>Regular Expressions</A
></DT
><DT
>19. <A
HREF="#HERE-DOCS"
>Here Documents</A
></DT
><DT
>20. <A
HREF="#IO-REDIRECTION"
>I/O Redirection</A
></DT
><DT
>21. <A
HREF="#SUBSHELLS"
>Subshells</A
></DT
><DT
>22. <A
HREF="#RESTRICTED-SH"
>Restricted Shells</A
></DT
><DT
>23. <A
HREF="#PROCESS-SUB"
>Process Substitution</A
></DT
><DT
>24. <A
HREF="#FUNCTIONS"
>Functions</A
></DT
><DT
>25. <A
HREF="#ALIASES"
>Aliases</A
></DT
><DT
>26. <A
HREF="#LIST-CONS"
>List Constructs</A
></DT
><DT
>27. <A
HREF="#ARRAYS"
>Arrays</A
></DT
><DT
>28. <A
HREF="#IVR"
>Indirect References</A
></DT
><DT
>29. <A
HREF="#DEVPROC"
><TT
CLASS="FILENAME"
>/dev</TT
> and <TT
CLASS="FILENAME"
>/proc</TT
></A
></DT
><DT
>30. <A
HREF="#NETWORKPROGRAMMING"
>Network Programming</A
></DT
><DT
>31. <A
HREF="#ZEROS"
>Of Zeros and Nulls</A
></DT
><DT
>32. <A
HREF="#DEBUGGING"
>Debugging</A
></DT
><DT
>33. <A
HREF="#OPTIONS"
>Options</A
></DT
><DT
>34. <A
HREF="#GOTCHAS"
>Gotchas</A
></DT
><DT
>35. <A
HREF="#SCRSTYLE"
>Scripting With Style</A
></DT
><DT
>36. <A
HREF="#MISCELLANY"
>Miscellany</A
></DT
><DT
>37. <A
HREF="#BASH2"
>Bash, versions 2, 3, and 4</A
></DT
></DL
></DD
><DT
>38. <A
HREF="#ENDNOTES"
>Endnotes</A
></DT
><DD
><DL
><DT
>38.1. <A
HREF="#AUTHORSNOTE"
>Author's Note</A
></DT
><DT
>38.2. <A
HREF="#ABOUTAUTHOR"
>About the Author</A
></DT
><DT
>38.3. <A
HREF="#WHEREHELP"
>Where to Go For Help</A
></DT
><DT
>38.4. <A
HREF="#TOOLSUSED"
>Tools Used to Produce This Book</A
></DT
><DT
>38.5. <A
HREF="#CREDITS"
>Credits</A
></DT
><DT
>38.6. <A
HREF="#DISCLAIMER"
>Disclaimer</A
></DT
></DL
></DD
><DT
><A
HREF="#BIBLIO"
>Bibliography</A
></DT
><DT
>A. <A
HREF="#CONTRIBUTED-SCRIPTS"
>Contributed Scripts</A
></DT
><DT
>B. <A
HREF="#REFCARDS"
>Reference Cards</A
></DT
><DT
>C. <A
HREF="#SEDAWK"
>A Sed and Awk Micro-Primer</A
></DT
><DD
><DL
><DT
>C.1. <A
HREF="#AEN23170"
>Sed</A
></DT
><DT
>C.2. <A
HREF="#AWK"
>Awk</A
></DT
></DL
></DD
><DT
>D. <A
HREF="#PATHMANAGEMENT"
>Parsing and Managing Pathnames</A
></DT
><DT
>E. <A
HREF="#EXITCODES"
>Exit Codes With Special Meanings</A
></DT
><DT
>F. <A
HREF="#IOREDIRINTRO"
>A Detailed Introduction to I/O and I/O Redirection</A
></DT
><DT
>G. <A
HREF="#COMMAND-LINE-OPTIONS"
>Command-Line Options</A
></DT
><DD
><DL
><DT
>G.1. <A
HREF="#STANDARD-OPTIONS"
>Standard Command-Line Options</A
></DT
><DT
>G.2. <A
HREF="#BASH-OPTIONS"
>Bash Command-Line Options</A
></DT
></DL
></DD
><DT
>H. <A
HREF="#FILES"
>Important Files</A
></DT
><DT
>I. <A
HREF="#SYSTEMDIRS"
>Important System Directories</A
></DT
><DT
>J. <A
HREF="#TABEXPANSION"
>An Introduction to Programmable Completion</A
></DT
><DT
>K. <A
HREF="#LOCALIZATION"
>Localization</A
></DT
><DT
>L. <A
HREF="#HISTCOMMANDS"
>History Commands</A
></DT
><DT
>M. <A
HREF="#SAMPLE-BASHRC"
>Sample <TT
CLASS="FILENAME"
>.bashrc</TT
> and
<TT
CLASS="FILENAME"
>.bash_profile</TT
> Files</A
></DT
><DT
>N. <A
HREF="#DOSBATCH"
>Converting DOS Batch Files to Shell Scripts</A
></DT
><DT
>O. <A
HREF="#EXERCISES"
>Exercises</A
></DT
><DD
><DL
><DT
>O.1. <A
HREF="#SCRIPTANALYSIS"
>Analyzing Scripts</A
></DT
><DT
>O.2. <A
HREF="#WRITINGSCRIPTS"
>Writing Scripts</A
></DT
></DL
></DD
><DT
>P. <A
HREF="#REVISIONHISTORY"
>Revision History</A
></DT
><DT
>Q. <A
HREF="#MIRRORSITES"
>Download and Mirror Sites</A
></DT
><DT
>R. <A
HREF="#TODOLIST"
>To Do List</A
></DT
><DT
>S. <A
HREF="#COPYRIGHT"
>Copyright</A
></DT
><DT
>T. <A
HREF="#ASCIITABLE"
>ASCII Table</A
></DT
><DT
><A
HREF="#XREFINDEX"
>Index</A
></DT
></DL
></DIV
><DIV
CLASS="LOT"
><DL
CLASS="LOT"
><DT
><B
>List of Tables</B
></DT
><DT
>8-1. <A
HREF="#AEN4294"
>Operator Precedence</A
></DT
><DT
>15-1. <A
HREF="#JOBIDTABLE"
>Job identifiers</A
></DT
><DT
>33-1. <A
HREF="#AEN19601"
>Bash options</A
></DT
><DT
>36-1. <A
HREF="#AEN20327"
>Numbers representing colors in Escape Sequences</A
></DT
><DT
>B-1. <A
HREF="#AEN22402"
>Special Shell Variables</A
></DT
><DT
>B-2. <A
HREF="#AEN22473"
>TEST Operators: Binary Comparison</A
></DT
><DT
>B-3. <A
HREF="#AEN22593"
>TEST Operators: Files</A
></DT
><DT
>B-4. <A
HREF="#AEN22728"
>Parameter Substitution and Expansion</A
></DT
><DT
>B-5. <A
HREF="#AEN22828"
>String Operations</A
></DT
><DT
>B-6. <A
HREF="#AEN22979"
>Miscellaneous Constructs</A
></DT
><DT
>C-1. <A
HREF="#AEN23200"
>Basic sed operators</A
></DT
><DT
>C-2. <A
HREF="#AEN23271"
>Examples of sed operators</A
></DT
><DT
>E-1. <A
HREF="#AEN23549"
><I
CLASS="FIRSTTERM"
>Reserved</I
> Exit Codes</A
></DT
><DT
>N-1. <A
HREF="#AEN24336"
>Batch file keywords / variables / operators, and their shell equivalents</A
></DT
><DT
>N-2. <A
HREF="#AEN24545"
>DOS commands and their UNIX equivalents</A
></DT
><DT
>P-1. <A
HREF="#AEN25364"
>Revision History</A
></DT
></DL
></DIV
><DIV
CLASS="LOT"
><DL
CLASS="LOT"
><DT
><B
>List of Examples</B
></DT
><DT
>2-1. <A
HREF="#EX1"
><I
CLASS="FIRSTTERM"
>cleanup</I
>: A script to clean up log
files in /var/log</A
></DT
><DT
>2-2. <A
HREF="#EX1A"
><I
CLASS="FIRSTTERM"
>cleanup</I
>: An improved clean-up
script</A
></DT
><DT
>2-3. <A
HREF="#EX2"
><I
CLASS="FIRSTTERM"
>cleanup</I
>: An enhanced
and generalized version of above scripts.</A
></DT
><DT
>3-1. <A
HREF="#EX8"
>Code blocks and I/O redirection</A
></DT
><DT
>3-2. <A
HREF="#RPMCHECK"
>Saving the output of a code block to a file</A
></DT
><DT
>3-3. <A
HREF="#BGLOOP"
>Running a loop in the background</A
></DT
><DT
>3-4. <A
HREF="#EX58"
>Backup of all files changed in last day</A
></DT
><DT
>4-1. <A
HREF="#EX9"
>Variable assignment and substitution</A
></DT
><DT
>4-2. <A
HREF="#EX15"
>Plain Variable Assignment</A
></DT
><DT
>4-3. <A
HREF="#EX16"
>Variable Assignment, plain and fancy</A
></DT
><DT
>4-4. <A
HREF="#INTORSTRING"
>Integer or string?</A
></DT
><DT
>4-5. <A
HREF="#EX17"
>Positional Parameters</A
></DT
><DT
>4-6. <A
HREF="#EX18"
><I
CLASS="FIRSTTERM"
>wh</I
>, <I
CLASS="FIRSTTERM"
> whois</I
> domain name lookup</A
></DT
><DT
>4-7. <A
HREF="#EX19"
>Using <I
CLASS="FIRSTTERM"
>shift</I
></A
></DT
><DT
>5-1. <A
HREF="#WEIRDVARS"
>Echoing Weird Variables</A
></DT
><DT
>5-2. <A
HREF="#ESCAPED"
>Escaped Characters</A
></DT
><DT
>5-3. <A
HREF="#BASHEK"
>Detecting key-presses</A
></DT
><DT
>6-1. <A
HREF="#EX5"
>exit / exit status</A
></DT
><DT
>6-2. <A
HREF="#NEGCOND"
>Negating a condition using <SPAN
CLASS="TOKEN"
>!</SPAN
></A
></DT
><DT
>7-1. <A
HREF="#EX10"
>What is truth?</A
></DT
><DT
>7-2. <A
HREF="#EX11"
>Equivalence of <I
CLASS="FIRSTTERM"
>test</I
>,
<TT
CLASS="FILENAME"
>/usr/bin/test</TT
>, <SPAN
CLASS="TOKEN"
>[ ]</SPAN
>,
and <TT
CLASS="FILENAME"
>/usr/bin/[</TT
></A
></DT
><DT
>7-3. <A
HREF="#ARITHTESTS"
>Arithmetic Tests using <SPAN
CLASS="TOKEN"
>(( ))</SPAN
></A
></DT
><DT
>7-4. <A
HREF="#BROKENLINK"
>Testing for broken links</A
></DT
><DT
>7-5. <A
HREF="#EX13"
>Arithmetic and string comparisons</A
></DT
><DT
>7-6. <A
HREF="#STRTEST"
>Testing whether a string is <I
CLASS="FIRSTTERM"
>null</I
></A
></DT
><DT
>7-7. <A
HREF="#EX14"
><I
CLASS="FIRSTTERM"
>zmore</I
></A
></DT
><DT
>8-1. <A
HREF="#GCD"
>Greatest common divisor</A
></DT
><DT
>8-2. <A
HREF="#ARITHOPS"
>Using Arithmetic Operations</A
></DT
><DT
>8-3. <A
HREF="#ANDOR"
>Compound Condition Tests Using &#38;&#38; and ||</A
></DT
><DT
>8-4. <A
HREF="#NUMBERS"
>Representation of numerical constants</A
></DT
><DT
>8-5. <A
HREF="#CVARS"
>C-style manipulation of variables</A
></DT
><DT
>9-1. <A
HREF="#IFSH"
>$IFS and whitespace</A
></DT
><DT
>9-2. <A
HREF="#TMDIN"
>Timed Input</A
></DT
><DT
>9-3. <A
HREF="#TIMEOUT"
>Once more, timed input</A
></DT
><DT
>9-4. <A
HREF="#TOUT"
>Timed <I
CLASS="FIRSTTERM"
>read</I
></A
></DT
><DT
>9-5. <A
HREF="#AMIROOT"
>Am I root?</A
></DT
><DT
>9-6. <A
HREF="#ARGLIST"
><I
CLASS="FIRSTTERM"
>arglist</I
>: Listing arguments
with $* and $@</A
></DT
><DT
>9-7. <A
HREF="#INCOMPAT"
>Inconsistent <TT
CLASS="VARNAME"
>$*</TT
> and <TT
CLASS="VARNAME"
>$@</TT
> behavior</A
></DT
><DT
>9-8. <A
HREF="#IFSEMPTY"
><TT
CLASS="VARNAME"
>$*</TT
> and <TT
CLASS="VARNAME"
>$@</TT
> when
<TT
CLASS="VARNAME"
>$IFS</TT
> is empty</A
></DT
><DT
>9-9. <A
HREF="#USCREF"
>Underscore variable</A
></DT
><DT
>9-10. <A
HREF="#EX20"
>Using <I
CLASS="FIRSTTERM"
>declare</I
> to type variables</A
></DT
><DT
>9-11. <A
HREF="#EX21"
>Generating random numbers</A
></DT
><DT
>9-12. <A
HREF="#PICKCARD"
>Picking a random card from a deck</A
></DT
><DT
>9-13. <A
HREF="#BROWNIAN"
>Brownian Motion Simulation</A
></DT
><DT
>9-14. <A
HREF="#RANDOMBETWEEN"
>Random between values</A
></DT
><DT
>9-15. <A
HREF="#RANDOMTEST"
>Rolling a single die with RANDOM</A
></DT
><DT
>9-16. <A
HREF="#SEEDINGRANDOM"
>Reseeding RANDOM</A
></DT
><DT
>9-17. <A
HREF="#RANDOM2"
>Pseudorandom numbers, using <A
HREF="#AWKREF"
>awk</A
></A
></DT
><DT
>10-1. <A
HREF="#PARAGRAPHSPACE"
>Inserting a blank line between paragraphs in a text file</A
></DT
><DT
>10-2. <A
HREF="#RANDSTRING"
>Generating an 8-character <SPAN
CLASS="QUOTE"
>"random"</SPAN
>
string</A
></DT
><DT
>10-3. <A
HREF="#CVT"
>Converting graphic file formats, with filename change</A
></DT
><DT
>10-4. <A
HREF="#RA2OGG"
>Converting streaming audio files to
<I
CLASS="FIRSTTERM"
>ogg</I
></A
></DT
><DT
>10-5. <A
HREF="#GETOPTSIMPLE"
>Emulating <I
CLASS="FIRSTTERM"
>getopt</I
></A
></DT
><DT
>10-6. <A
HREF="#SUBSTRINGEX"
>Alternate ways of extracting and locating substrings</A
></DT
><DT
>10-7. <A
HREF="#EX6"
>Using parameter substitution and error messages</A
></DT
><DT
>10-8. <A
HREF="#USAGEMESSAGE"
>Parameter substitution and <SPAN
CLASS="QUOTE"
>"usage"</SPAN
> messages</A
></DT
><DT
>10-9. <A
HREF="#LENGTH"
>Length of a variable</A
></DT
><DT
>10-10. <A
HREF="#PATTMATCHING"
>Pattern matching in parameter substitution</A
></DT
><DT
>10-11. <A
HREF="#RFE"
>Renaming file extensions<SPAN
CLASS="TOKEN"
>:</SPAN
></A
></DT
><DT
>10-12. <A
HREF="#EX7"
>Using pattern matching to parse arbitrary strings</A
></DT
><DT
>10-13. <A
HREF="#VARMATCH"
>Matching patterns at prefix or suffix of string</A
></DT
><DT
>11-1. <A
HREF="#EX22"
>Simple <I
CLASS="FIRSTTERM"
>for</I
> loops</A
></DT
><DT
>11-2. <A
HREF="#EX22A"
><I
CLASS="FIRSTTERM"
>for</I
> loop with two parameters in each
[list] element</A
></DT
><DT
>11-3. <A
HREF="#FILEINFO"
><EM
>Fileinfo:</EM
> operating on a file list
contained in a variable</A
></DT
><DT
>11-4. <A
HREF="#FILEINFO01"
>Operating on a parameterized file list</A
></DT
><DT
>11-5. <A
HREF="#LISTGLOB"
>Operating on files with a <I
CLASS="FIRSTTERM"
>for</I
> loop</A
></DT
><DT
>11-6. <A
HREF="#EX23"
>Missing <TT
CLASS="USERINPUT"
><B
>in [list]</B
></TT
> in a
<I
CLASS="FIRSTTERM"
>for</I
> loop</A
></DT
><DT
>11-7. <A
HREF="#FORLOOPCMD"
>Generating the <TT
CLASS="USERINPUT"
><B
>[list]</B
></TT
> in
a <I
CLASS="FIRSTTERM"
>for</I
> loop with command substitution</A
></DT
><DT
>11-8. <A
HREF="#BINGREP"
>A <I
CLASS="FIRSTTERM"
>grep</I
> replacement
for binary files</A
></DT
><DT
>11-9. <A
HREF="#USERLIST"
>Listing all users on the system</A
></DT
><DT
>11-10. <A
HREF="#FINDSTRING"
>Checking all the binaries in a directory for
authorship</A
></DT
><DT
>11-11. <A
HREF="#SYMLINKS"
>Listing the <I
CLASS="FIRSTTERM"
>symbolic
links</I
> in a directory</A
></DT
><DT
>11-12. <A
HREF="#SYMLINKS2"
>Symbolic links in a directory, saved to a file</A
></DT
><DT
>11-13. <A
HREF="#FORLOOPC"
>A C-style <I
CLASS="FIRSTTERM"
>for</I
> loop</A
></DT
><DT
>11-14. <A
HREF="#EX24"
>Using <I
CLASS="FIRSTTERM"
>efax</I
> in batch mode</A
></DT
><DT
>11-15. <A
HREF="#EX25"
>Simple <I
CLASS="FIRSTTERM"
>while</I
> loop</A
></DT
><DT
>11-16. <A
HREF="#EX26"
>Another <I
CLASS="FIRSTTERM"
>while</I
> loop</A
></DT
><DT
>11-17. <A
HREF="#EX26A"
><I
CLASS="FIRSTTERM"
>while</I
> loop with multiple conditions</A
></DT
><DT
>11-18. <A
HREF="#WHLOOPC"
>C-style syntax in a <I
CLASS="FIRSTTERM"
>while</I
> loop</A
></DT
><DT
>11-19. <A
HREF="#EX27"
><I
CLASS="FIRSTTERM"
>until</I
> loop</A
></DT
><DT
>11-20. <A
HREF="#NESTEDLOOP"
>Nested Loop</A
></DT
><DT
>11-21. <A
HREF="#EX28"
>Effects of <I
CLASS="FIRSTTERM"
>break</I
> and
<B
CLASS="COMMAND"
>continue</B
> in a loop</A
></DT
><DT
>11-22. <A
HREF="#BREAKLEVELS"
>Breaking out of multiple loop levels</A
></DT
><DT
>11-23. <A
HREF="#CONTINUELEVELS"
>Continuing at a higher loop level</A
></DT
><DT
>11-24. <A
HREF="#CONTINUENEX"
>Using <I
CLASS="FIRSTTERM"
>continue N</I
> in an actual task</A
></DT
><DT
>11-25. <A
HREF="#EX29"
>Using <I
CLASS="FIRSTTERM"
>case</I
></A
></DT
><DT
>11-26. <A
HREF="#EX30"
>Creating menus using <I
CLASS="FIRSTTERM"
>case</I
></A
></DT
><DT
>11-27. <A
HREF="#CASECMD"
>Using <I
CLASS="FIRSTTERM"
>command substitution</I
>
to generate the <I
CLASS="FIRSTTERM"
>case</I
> variable</A
></DT
><DT
>11-28. <A
HREF="#MATCHSTRING"
>Simple string matching</A
></DT
><DT
>11-29. <A
HREF="#ISALPHA"
>Checking for alphabetic input</A
></DT
><DT
>11-30. <A
HREF="#EX31"
>Creating menus using <I
CLASS="FIRSTTERM"
>select</I
></A
></DT
><DT
>11-31. <A
HREF="#EX32"
>Creating menus using <I
CLASS="FIRSTTERM"
>select</I
>
in a function</A
></DT
><DT
>12-1. <A
HREF="#STUPSCR"
>Stupid script tricks</A
></DT
><DT
>12-2. <A
HREF="#CSUBLOOP"
>Generating a variable from a loop</A
></DT
><DT
>12-3. <A
HREF="#AGRAM2"
>Finding anagrams</A
></DT
><DT
>15-1. <A
HREF="#SPAWNSCR"
>A script that spawns multiple instances of itself</A
></DT
><DT
>15-2. <A
HREF="#EX47"
><I
CLASS="FIRSTTERM"
>printf</I
> in action</A
></DT
><DT
>15-3. <A
HREF="#EX36"
>Variable assignment, using <I
CLASS="FIRSTTERM"
>read</I
></A
></DT
><DT
>15-4. <A
HREF="#READNOVAR"
>What happens when <I
CLASS="FIRSTTERM"
>read</I
> has no
variable</A
></DT
><DT
>15-5. <A
HREF="#READR"
>Multi-line input to <I
CLASS="FIRSTTERM"
>read</I
></A
></DT
><DT
>15-6. <A
HREF="#ARROWDETECT"
>Detecting the arrow keys</A
></DT
><DT
>15-7. <A
HREF="#READREDIR"
>Using <I
CLASS="FIRSTTERM"
>read</I
> with
<A
HREF="#IOREDIRREF"
>file redirection</A
></A
></DT
><DT
>15-8. <A
HREF="#READPIPE"
>Problems reading from a pipe</A
></DT
><DT
>15-9. <A
HREF="#EX37"
>Changing the current working directory</A
></DT
><DT
>15-10. <A
HREF="#EX46"
>Letting <I
CLASS="FIRSTTERM"
>let</I
> do arithmetic.</A
></DT
><DT
>15-11. <A
HREF="#EX43"
>Showing the effect of <I
CLASS="FIRSTTERM"
>eval</I
></A
></DT
><DT
>15-12. <A
HREF="#ARRCHOICE"
>Using <I
CLASS="FIRSTTERM"
>eval</I
> to select
among variables</A
></DT
><DT
>15-13. <A
HREF="#ECHOPARAMS"
><I
CLASS="FIRSTTERM"
>Echoing</I
> the
<I
CLASS="FIRSTTERM"
>command-line parameters</I
></A
></DT
><DT
>15-14. <A
HREF="#EX44"
>Forcing a log-off</A
></DT
><DT
>15-15. <A
HREF="#ROT14"
>A version of <I
CLASS="FIRSTTERM"
>rot13</I
></A
></DT
><DT
>15-16. <A
HREF="#EX34"
>Using <I
CLASS="FIRSTTERM"
>set</I
> with positional
parameters</A
></DT
><DT
>15-17. <A
HREF="#REVPOSPARAMS"
>Reversing the positional parameters</A
></DT
><DT
>15-18. <A
HREF="#SETPOS"
>Reassigning the positional parameters</A
></DT
><DT
>15-19. <A
HREF="#UNS"
><SPAN
CLASS="QUOTE"
>"Unsetting"</SPAN
> a variable</A
></DT
><DT
>15-20. <A
HREF="#COLTOTALER3"
>Using <I
CLASS="FIRSTTERM"
>export</I
> to pass a variable to an
embedded <I
CLASS="FIRSTTERM"
>awk</I
> script</A
></DT
><DT
>15-21. <A
HREF="#EX33"
>Using <I
CLASS="FIRSTTERM"
>getopts</I
> to read the
options/arguments passed to a script</A
></DT
><DT
>15-22. <A
HREF="#EX38"
><SPAN
CLASS="QUOTE"
>"Including"</SPAN
> a data file</A
></DT
><DT
>15-23. <A
HREF="#SELFSOURCE"
>A (useless) script that sources itself</A
></DT
><DT
>15-24. <A
HREF="#EX54"
>Effects of <I
CLASS="FIRSTTERM"
>exec</I
></A
></DT
><DT
>15-25. <A
HREF="#SELFEXEC"
>A script that <I
CLASS="FIRSTTERM"
>exec's</I
> itself</A
></DT
><DT
>15-26. <A
HREF="#EX39"
>Waiting for a process to finish before proceeding</A
></DT
><DT
>15-27. <A
HREF="#SELFDESTRUCT"
>A script that kills itself</A
></DT
><DT
>16-1. <A
HREF="#EX40"
>Using <I
CLASS="FIRSTTERM"
>ls</I
> to create a table of contents
for burning a <SPAN
CLASS="ABBREV"
>CDR</SPAN
> disk</A
></DT
><DT
>16-2. <A
HREF="#HELLOL"
>Hello or Good-bye</A
></DT
><DT
>16-3. <A
HREF="#EX57"
><I
CLASS="FIRSTTERM"
>Badname</I
>, eliminate file names
in current directory containing bad characters and <A
HREF="#WHITESPACEREF"
>whitespace</A
>.</A
></DT
><DT
>16-4. <A
HREF="#IDELETE"
>Deleting a file by its <I
CLASS="FIRSTTERM"
>inode</I
>
number</A
></DT
><DT
>16-5. <A
HREF="#EX41"
>Logfile: Using <I
CLASS="FIRSTTERM"
>xargs</I
> to monitor system log</A
></DT
><DT
>16-6. <A
HREF="#EX42"
>Copying files in current directory to another</A
></DT
><DT
>16-7. <A
HREF="#KILLBYNAME"
>Killing processes by name</A
></DT
><DT
>16-8. <A
HREF="#WF2"
>Word frequency analysis using
<I
CLASS="FIRSTTERM"
>xargs</I
></A
></DT
><DT
>16-9. <A
HREF="#EX45"
>Using <I
CLASS="FIRSTTERM"
>expr</I
></A
></DT
><DT
>16-10. <A
HREF="#EX51"
>Using <I
CLASS="FIRSTTERM"
>date</I
></A
></DT
><DT
>16-11. <A
HREF="#DATECALC"
><I
CLASS="FIRSTTERM"
>Date</I
> calculations</A
></DT
><DT
>16-12. <A
HREF="#WF"
>Word Frequency Analysis</A
></DT
><DT
>16-13. <A
HREF="#SCRIPTDETECTOR"
>Which files are scripts?</A
></DT
><DT
>16-14. <A
HREF="#RND"
>Generating 10-digit random numbers</A
></DT
><DT
>16-15. <A
HREF="#EX12"
>Using <I
CLASS="FIRSTTERM"
>tail</I
> to monitor the system log</A
></DT
><DT
>16-16. <A
HREF="#FROMSH"
>Printing out the <I
CLASS="FIRSTTERM"
>From</I
> lines in
stored e-mail messages</A
></DT
><DT
>16-17. <A
HREF="#GRP"
>Emulating <I
CLASS="FIRSTTERM"
>grep</I
> in a script</A
></DT
><DT
>16-18. <A
HREF="#CWSOLVER"
>Crossword puzzle solver</A
></DT
><DT
>16-19. <A
HREF="#DICTLOOKUP"
>Looking up definitions in Webster's 1913 Dictionary</A
></DT
><DT
>16-20. <A
HREF="#LOOKUP"
>Checking words in a list for validity</A
></DT
><DT
>16-21. <A
HREF="#EX49"
><I
CLASS="FIRSTTERM"
>toupper</I
>: Transforms a file
to all uppercase.</A
></DT
><DT
>16-22. <A
HREF="#LOWERCASE"
><I
CLASS="FIRSTTERM"
>lowercase</I
>: Changes all
filenames in working directory to lowercase.</A
></DT
><DT
>16-23. <A
HREF="#DU"
><I
CLASS="FIRSTTERM"
>du</I
>: DOS to UNIX text file conversion.</A
></DT
><DT
>16-24. <A
HREF="#ROT13"
><I
CLASS="FIRSTTERM"
>rot13</I
>: ultra-weak encryption.</A
></DT
><DT
>16-25. <A
HREF="#CRYPTOQUOTE"
>Generating <SPAN
CLASS="QUOTE"
>"Crypto-Quote"</SPAN
> Puzzles</A
></DT
><DT
>16-26. <A
HREF="#EX50"
>Formatted file listing.</A
></DT
><DT
>16-27. <A
HREF="#COL"
>Using <I
CLASS="FIRSTTERM"
>column</I
> to format a directory
listing</A
></DT
><DT
>16-28. <A
HREF="#LNUM"
><I
CLASS="FIRSTTERM"
>nl</I
>: A self-numbering script.</A
></DT
><DT
>16-29. <A
HREF="#MANVIEW"
><I
CLASS="FIRSTTERM"
>manview</I
>: Viewing formatted manpages</A
></DT
><DT
>16-30. <A
HREF="#EX48"
>Using <I
CLASS="FIRSTTERM"
>cpio</I
> to move a directory tree</A
></DT
><DT
>16-31. <A
HREF="#DERPM"
>Unpacking an <I
CLASS="FIRSTTERM"
>rpm</I
> archive</A
></DT
><DT
>16-32. <A
HREF="#STRIPC"
>Stripping comments from C program files</A
></DT
><DT
>16-33. <A
HREF="#WHAT"
>Exploring <TT
CLASS="FILENAME"
>/usr/X11R6/bin</TT
></A
></DT
><DT
>16-34. <A
HREF="#WSTRINGS"
>An <SPAN
CLASS="QUOTE"
>"improved"</SPAN
>
<I
CLASS="FIRSTTERM"
>strings</I
> command</A
></DT
><DT
>16-35. <A
HREF="#FILECOMP"
>Using <I
CLASS="FIRSTTERM"
>cmp</I
> to compare two files
within a script.</A
></DT
><DT
>16-36. <A
HREF="#EX35"
><I
CLASS="FIRSTTERM"
>basename</I
> and
<I
CLASS="FIRSTTERM"
>dirname</I
></A
></DT
><DT
>16-37. <A
HREF="#SPLITCOPY"
>A script that copies itself in sections</A
></DT
><DT
>16-38. <A
HREF="#FILEINTEGRITY"
>Checking file integrity</A
></DT
><DT
>16-39. <A
HREF="#EX52"
>Uudecoding encoded files</A
></DT
><DT
>16-40. <A
HREF="#SPAMLOOKUP"
>Finding out where to report a spammer</A
></DT
><DT
>16-41. <A
HREF="#ISSPAMMER"
>Analyzing a spam domain</A
></DT
><DT
>16-42. <A
HREF="#QUOTEFETCH"
>Getting a stock quote</A
></DT
><DT
>16-43. <A
HREF="#FC4UPD"
>Updating FC4</A
></DT
><DT
>16-44. <A
HREF="#REMOTE"
>Using <I
CLASS="FIRSTTERM"
>ssh</I
></A
></DT
><DT
>16-45. <A
HREF="#SELFMAILER"
>A script that mails itself</A
></DT
><DT
>16-46. <A
HREF="#PRIMES2"
>Generating prime numbers</A
></DT
><DT
>16-47. <A
HREF="#MONTHLYPMT"
>Monthly Payment on a Mortgage</A
></DT
><DT
>16-48. <A
HREF="#BASE"
>Base Conversion</A
></DT
><DT
>16-49. <A
HREF="#ALTBC"
>Invoking <I
CLASS="FIRSTTERM"
>bc</I
> using a <I
CLASS="FIRSTTERM"
>here
document</I
></A
></DT
><DT
>16-50. <A
HREF="#CANNON"
>Calculating PI</A
></DT
><DT
>16-51. <A
HREF="#HEXCONVERT"
>Converting a decimal number to hexadecimal</A
></DT
><DT
>16-52. <A
HREF="#FACTR"
>Factoring</A
></DT
><DT
>16-53. <A
HREF="#HYPOT"
>Calculating the hypotenuse of a triangle</A
></DT
><DT
>16-54. <A
HREF="#EX53"
>Using <I
CLASS="FIRSTTERM"
>seq</I
> to generate loop
arguments</A
></DT
><DT
>16-55. <A
HREF="#LETTERCOUNT"
>Letter Count"</A
></DT
><DT
>16-56. <A
HREF="#EX33A"
>Using <I
CLASS="FIRSTTERM"
>getopt</I
> to parse command-line
options</A
></DT
><DT
>16-57. <A
HREF="#SELFCOPY"
>A script that copies itself</A
></DT
><DT
>16-58. <A
HREF="#EXERCISINGDD"
>Exercising <I
CLASS="FIRSTTERM"
>dd</I
></A
></DT
><DT
>16-59. <A
HREF="#DDKEYPRESS"
>Capturing Keystrokes</A
></DT
><DT
>16-60. <A
HREF="#RPSDCARD"
>Preparing a bootable SD card for the
<EM
>Raspberry Pi</EM
></A
></DT
><DT
>16-61. <A
HREF="#BLOTOUT"
>Securely deleting a file</A
></DT
><DT
>16-62. <A
HREF="#TEMPFILENAME"
>Filename generator</A
></DT
><DT
>16-63. <A
HREF="#UNITCONVERSION"
>Converting meters to miles</A
></DT
><DT
>16-64. <A
HREF="#M4"
>Using <I
CLASS="FIRSTTERM"
>m4</I
></A
></DT
><DT
>17-1. <A
HREF="#SETNEWPW"
>Setting a new password</A
></DT
><DT
>17-2. <A
HREF="#ERASE"
>Setting an <I
CLASS="FIRSTTERM"
>erase</I
> character</A
></DT
><DT
>17-3. <A
HREF="#SECRETPW"
><I
CLASS="FIRSTTERM"
>secret password</I
>:
Turning off terminal echoing</A
></DT
><DT
>17-4. <A
HREF="#KEYPRESS"
>Keypress detection</A
></DT
><DT
>17-5. <A
HREF="#ISCAN"
>Checking a remote server for
<I
CLASS="FIRSTTERM"
>identd</I
></A
></DT
><DT
>17-6. <A
HREF="#KILLPROCESS"
><I
CLASS="FIRSTTERM"
>pidof</I
> helps kill a process</A
></DT
><DT
>17-7. <A
HREF="#ISOMOUNTREF"
>Checking a CD image</A
></DT
><DT
>17-8. <A
HREF="#CREATEFS"
>Creating a filesystem in a file</A
></DT
><DT
>17-9. <A
HREF="#ADDDRV"
>Adding a new hard drive</A
></DT
><DT
>17-10. <A
HREF="#ROT13A"
>Using <I
CLASS="FIRSTTERM"
>umask</I
> to hide an output file
from prying eyes</A
></DT
><DT
>17-11. <A
HREF="#BACKLIGHT"
><I
CLASS="FIRSTTERM"
>Backlight</I
>: changes
the brightness of the (laptop) screen backlight</A
></DT
><DT
>17-12. <A
HREF="#EX55"
><I
CLASS="FIRSTTERM"
>killall</I
>, from <TT
CLASS="FILENAME"
>/etc/rc.d/init.d</TT
></A
></DT
><DT
>19-1. <A
HREF="#EX70"
><I
CLASS="FIRSTTERM"
>broadcast</I
>: Sends message to everyone
logged in</A
></DT
><DT
>19-2. <A
HREF="#EX69"
><I
CLASS="FIRSTTERM"
>dummyfile</I
>: Creates a 2-line dummy
file</A
></DT
><DT
>19-3. <A
HREF="#EX71"
>Multi-line message using <I
CLASS="FIRSTTERM"
>cat</I
></A
></DT
><DT
>19-4. <A
HREF="#EX71A"
>Multi-line message, with tabs suppressed</A
></DT
><DT
>19-5. <A
HREF="#EX71B"
>Here document with replaceable parameters</A
></DT
><DT
>19-6. <A
HREF="#EX72"
>Upload a file pair to <I
CLASS="FIRSTTERM"
>Sunsite</I
> incoming
directory</A
></DT
><DT
>19-7. <A
HREF="#EX71C"
>Parameter substitution turned off</A
></DT
><DT
>19-8. <A
HREF="#GENERATESCRIPT"
>A script that generates another script</A
></DT
><DT
>19-9. <A
HREF="#HF"
>Here documents and functions</A
></DT
><DT
>19-10. <A
HREF="#ANONHEREDOC"
><SPAN
CLASS="QUOTE"
>"Anonymous"</SPAN
> Here Document</A
></DT
><DT
>19-11. <A
HREF="#COMMENTBLOCK"
>Commenting out a block of code</A
></DT
><DT
>19-12. <A
HREF="#SELFDOCUMENT"
>A self-documenting script</A
></DT
><DT
>19-13. <A
HREF="#PREPENDEX"
>Prepending a line to a file</A
></DT
><DT
>19-14. <A
HREF="#MAILBOXGREP"
>Parsing a mailbox</A
></DT
><DT
>20-1. <A
HREF="#REDIR1"
>Redirecting <TT
CLASS="FILENAME"
>stdin</TT
> using
<I
CLASS="FIRSTTERM"
>exec</I
></A
></DT
><DT
>20-2. <A
HREF="#REASSIGNSTDOUT"
>Redirecting <TT
CLASS="FILENAME"
>stdout</TT
> using
<I
CLASS="FIRSTTERM"
>exec</I
></A
></DT
><DT
>20-3. <A
HREF="#UPPERCONV"
>Redirecting both <TT
CLASS="FILENAME"
>stdin</TT
> and
<TT
CLASS="FILENAME"
>stdout</TT
> in the same script with
<I
CLASS="FIRSTTERM"
>exec</I
></A
></DT
><DT
>20-4. <A
HREF="#AVOIDSUBSHELL"
>Avoiding a subshell</A
></DT
><DT
>20-5. <A
HREF="#REDIR2"
>Redirected <I
CLASS="FIRSTTERM"
>while</I
> loop</A
></DT
><DT
>20-6. <A
HREF="#REDIR2A"
>Alternate form of redirected <I
CLASS="FIRSTTERM"
>while</I
> loop</A
></DT
><DT
>20-7. <A
HREF="#REDIR3"
>Redirected <I
CLASS="FIRSTTERM"
>until</I
> loop</A
></DT
><DT
>20-8. <A
HREF="#REDIR4"
>Redirected <I
CLASS="FIRSTTERM"
>for</I
> loop</A
></DT
><DT
>20-9. <A
HREF="#REDIR4A"
>Redirected <I
CLASS="FIRSTTERM"
>for</I
> loop (both
<TT
CLASS="FILENAME"
>stdin</TT
> and <TT
CLASS="FILENAME"
>stdout</TT
>
redirected)</A
></DT
><DT
>20-10. <A
HREF="#REDIR5"
>Redirected <I
CLASS="FIRSTTERM"
>if/then</I
> test</A
></DT
><DT
>20-11. <A
HREF="#NAMESDATA"
>Data file <I
CLASS="FIRSTTERM"
>names.data</I
> for above
examples</A
></DT
><DT
>20-12. <A
HREF="#LOGEVENTS"
>Logging events</A
></DT
><DT
>21-1. <A
HREF="#SUBSHELL"
>Variable scope in a subshell</A
></DT
><DT
>21-2. <A
HREF="#ALLPROFS"
>List User Profiles</A
></DT
><DT
>21-3. <A
HREF="#PARALLEL-PROCESSES"
>Running parallel processes in subshells</A
></DT
><DT
>22-1. <A
HREF="#RESTRICTED"
>Running a script in restricted mode</A
></DT
><DT
>23-1. <A
HREF="#WRPS"
>Code block redirection without forking</A
></DT
><DT
>23-2. <A
HREF="#PSUBP"
>Redirecting the output of <I
CLASS="FIRSTTERM"
>process
substitution</I
> into a loop.</A
></DT
><DT
>24-1. <A
HREF="#EX59"
>Simple functions</A
></DT
><DT
>24-2. <A
HREF="#EX60"
>Function Taking Parameters</A
></DT
><DT
>24-3. <A
HREF="#FUNCCMDLINEARG"
>Functions and command-line args passed to the script</A
></DT
><DT
>24-4. <A
HREF="#INDFUNC"
>Passing an indirect reference to a function</A
></DT
><DT
>24-5. <A
HREF="#DEREFERENCECL"
>Dereferencing a parameter passed to a function</A
></DT
><DT
>24-6. <A
HREF="#REFPARAMS"
>Again, dereferencing a parameter passed to a function</A
></DT
><DT
>24-7. <A
HREF="#MAX"
>Maximum of two numbers</A
></DT
><DT
>24-8. <A
HREF="#EX61"
>Converting numbers to Roman numerals</A
></DT
><DT
>24-9. <A
HREF="#RETURNTEST"
>Testing large return values in a function</A
></DT
><DT
>24-10. <A
HREF="#MAX2"
>Comparing two large integers</A
></DT
><DT
>24-11. <A
HREF="#REALNAME"
>Real name from username</A
></DT
><DT
>24-12. <A
HREF="#EX62"
>Local variable visibility</A
></DT
><DT
>24-13. <A
HREF="#RECURSIONDEMO"
>Demonstration of a simple recursive function</A
></DT
><DT
>24-14. <A
HREF="#RECURSIONDEMO2"
>Another simple demonstration</A
></DT
><DT
>24-15. <A
HREF="#EX63"
>Recursion, using a local variable</A
></DT
><DT
>24-16. <A
HREF="#FIBO"
><I
CLASS="FIRSTTERM"
>The Fibonacci Sequence</I
></A
></DT
><DT
>24-17. <A
HREF="#HANOI"
><I
CLASS="FIRSTTERM"
>The Towers of Hanoi</I
></A
></DT
><DT
>25-1. <A
HREF="#AL"
>Aliases within a script</A
></DT
><DT
>25-2. <A
HREF="#UNAL"
><I
CLASS="FIRSTTERM"
>unalias</I
>: Setting and unsetting
an alias</A
></DT
><DT
>26-1. <A
HREF="#EX64"
>Using an <I
CLASS="FIRSTTERM"
>and list</I
> to test
for command-line arguments</A
></DT
><DT
>26-2. <A
HREF="#ANDLIST2"
>Another command-line arg test using an <I
CLASS="FIRSTTERM"
>and
list</I
></A
></DT
><DT
>26-3. <A
HREF="#EX65"
>Using <I
CLASS="FIRSTTERM"
>or lists</I
> in combination
with an <I
CLASS="FIRSTTERM"
>and list</I
></A
></DT
><DT
>27-1. <A
HREF="#EX66"
>Simple array usage</A
></DT
><DT
>27-2. <A
HREF="#POEM"
>Formatting a poem</A
></DT
><DT
>27-3. <A
HREF="#ARRAYOPS"
>Various array operations</A
></DT
><DT
>27-4. <A
HREF="#ARRAYSTROPS"
>String operations on arrays</A
></DT
><DT
>27-5. <A
HREF="#SCRIPTARRAY"
>Loading the contents of a script into an array</A
></DT
><DT
>27-6. <A
HREF="#EX67"
>Some special properties of arrays</A
></DT
><DT
>27-7. <A
HREF="#EMPTYARRAY"
>Of empty arrays and empty elements</A
></DT
><DT
>27-8. <A
HREF="#ARRAYASSIGN"
>Initializing arrays</A
></DT
><DT
>27-9. <A
HREF="#COPYARRAY"
>Copying and concatenating arrays</A
></DT
><DT
>27-10. <A
HREF="#ARRAYAPPEND"
>More on concatenating arrays</A
></DT
><DT
>27-11. <A
HREF="#BUBBLE"
>The Bubble Sort</A
></DT
><DT
>27-12. <A
HREF="#EMBARR"
>Embedded arrays and indirect references</A
></DT
><DT
>27-13. <A
HREF="#EX68"
>The Sieve of Eratosthenes</A
></DT
><DT
>27-14. <A
HREF="#EX68A"
>The Sieve of Eratosthenes, Optimized</A
></DT
><DT
>27-15. <A
HREF="#STACKEX"
>Emulating a push-down stack</A
></DT
><DT
>27-16. <A
HREF="#QFUNCTION"
>Complex array application:
<EM
>Exploring a weird mathematical series</EM
></A
></DT
><DT
>27-17. <A
HREF="#TWODIM"
>Simulating a two-dimensional array, then tilting it</A
></DT
><DT
>28-1. <A
HREF="#INDREF"
>Indirect Variable References</A
></DT
><DT
>28-2. <A
HREF="#COLTOTALER2"
>Passing an indirect reference to <I
CLASS="FIRSTTERM"
>awk</I
></A
></DT
><DT
>29-1. <A
HREF="#DEVTCP"
>Using <TT
CLASS="FILENAME"
>/dev/tcp</TT
> for
troubleshooting</A
></DT
><DT
>29-2. <A
HREF="#MUSICSCR"
>Playing music</A
></DT
><DT
>29-3. <A
HREF="#PIDID"
>Finding the process associated with a PID</A
></DT
><DT
>29-4. <A
HREF="#CONSTAT"
>On-line connect status</A
></DT
><DT
>30-1. <A
HREF="#TESTCGI"
>Print the server environment</A
></DT
><DT
>30-2. <A
HREF="#IPADDRESSES"
>IP addresses</A
></DT
><DT
>31-1. <A
HREF="#COOKIES"
>Hiding the cookie jar</A
></DT
><DT
>31-2. <A
HREF="#EX73"
>Setting up a swapfile using <TT
CLASS="FILENAME"
>/dev/zero</TT
></A
></DT
><DT
>31-3. <A
HREF="#RAMDISK"
>Creating a ramdisk</A
></DT
><DT
>32-1. <A
HREF="#EX74"
>A buggy script</A
></DT
><DT
>32-2. <A
HREF="#MISSINGKEYWORD"
>Missing <A
HREF="#KEYWORDREF"
>keyword</A
></A
></DT
><DT
>32-3. <A
HREF="#EX75"
><I
CLASS="FIRSTTERM"
>test24</I
>: another buggy script</A
></DT
><DT
>32-4. <A
HREF="#ASSERT"
>Testing a condition with an
<I
CLASS="FIRSTTERM"
>assert</I
></A
></DT
><DT
>32-5. <A
HREF="#EX76"
>Trapping at exit</A
></DT
><DT
>32-6. <A
HREF="#ONLINE"
>Cleaning up after <B
CLASS="KEYCAP"
>Control-C</B
></A
></DT
><DT
>32-7. <A
HREF="#PROGRESSBAR2"
>A Simple Implementation of a Progress Bar</A
></DT
><DT
>32-8. <A
HREF="#VARTRACE"
>Tracing a variable</A
></DT
><DT
>32-9. <A
HREF="#MULTIPLEPROC"
>Running multiple processes (on an SMP box)</A
></DT
><DT
>34-1. <A
HREF="#BADOP"
>Numerical and string comparison are not equivalent</A
></DT
><DT
>34-2. <A
HREF="#SUBPIT"
>Subshell Pitfalls</A
></DT
><DT
>34-3. <A
HREF="#BADREAD"
>Piping the output of <I
CLASS="FIRSTTERM"
>echo</I
> to a
<I
CLASS="FIRSTTERM"
>read</I
></A
></DT
><DT
>36-1. <A
HREF="#EX3"
><I
CLASS="FIRSTTERM"
>shell wrapper</I
></A
></DT
><DT
>36-2. <A
HREF="#EX4"
>A slightly more complex <I
CLASS="FIRSTTERM"
>shell
wrapper</I
></A
></DT
><DT
>36-3. <A
HREF="#LOGGINGWRAPPER"
>A generic <I
CLASS="FIRSTTERM"
>shell wrapper</I
> that
writes to a logfile</A
></DT
><DT
>36-4. <A
HREF="#PRASC"
>A <I
CLASS="FIRSTTERM"
>shell wrapper</I
> around an awk
script</A
></DT
><DT
>36-5. <A
HREF="#COLTOTALER"
>A <I
CLASS="FIRSTTERM"
>shell wrapper</I
> around another
awk script</A
></DT
><DT
>36-6. <A
HREF="#EX56"
>Perl embedded in a <I
CLASS="FIRSTTERM"
>Bash</I
> script</A
></DT
><DT
>36-7. <A
HREF="#BASHANDPERL"
>Bash and Perl scripts combined</A
></DT
><DT
>36-8. <A
HREF="#EX56PY"
>Python embedded in a <I
CLASS="FIRSTTERM"
>Bash</I
> script</A
></DT
><DT
>36-9. <A
HREF="#SPEECH0"
>A script that speaks</A
></DT
><DT
>36-10. <A
HREF="#RECURSE"
>A (useless) script that recursively calls itself</A
></DT
><DT
>36-11. <A
HREF="#PBOOK"
>A (useful) script that recursively calls itself</A
></DT
><DT
>36-12. <A
HREF="#USRMNT"
>Another (useful) script that recursively calls itself</A
></DT
><DT
>36-13. <A
HREF="#EX30A"
>A <SPAN
CLASS="QUOTE"
>"colorized"</SPAN
> address database</A
></DT
><DT
>36-14. <A
HREF="#DRAW-BOX"
>Drawing a box</A
></DT
><DT
>36-15. <A
HREF="#COLORECHO"
>Echoing colored text</A
></DT
><DT
>36-16. <A
HREF="#HORSERACE"
>A <SPAN
CLASS="QUOTE"
>"horserace"</SPAN
> game</A
></DT
><DT
>36-17. <A
HREF="#PROGRESSBAR"
>A Progress Bar</A
></DT
><DT
>36-18. <A
HREF="#MULTIPLICATION"
>Return value trickery</A
></DT
><DT
>36-19. <A
HREF="#SUMPRODUCT"
>Even more return value trickery</A
></DT
><DT
>36-20. <A
HREF="#ARRFUNC"
>Passing and returning arrays</A
></DT
><DT
>36-21. <A
HREF="#AGRAM"
>Fun with anagrams</A
></DT
><DT
>36-22. <A
HREF="#DIALOG"
>Widgets invoked from a shell script</A
></DT
><DT
>36-23. <A
HREF="#TESTSUITE"
>Test Suite</A
></DT
><DT
>37-1. <A
HREF="#EX77"
>String expansion</A
></DT
><DT
>37-2. <A
HREF="#EX78"
>Indirect variable references - the new way</A
></DT
><DT
>37-3. <A
HREF="#RESISTOR"
>Simple database application, using indirect variable
referencing</A
></DT
><DT
>37-4. <A
HREF="#CARDS"
>Using arrays and other miscellaneous trickery
to deal four random hands from a deck of cards</A
></DT
><DT
>37-5. <A
HREF="#FETCHADDRESS"
>A simple address database</A
></DT
><DT
>37-6. <A
HREF="#FETCHADDRESS2"
>A somewhat more elaborate address database</A
></DT
><DT
>37-7. <A
HREF="#CASE4"
>Testing characters</A
></DT
><DT
>37-8. <A
HREF="#READN"
>Reading N characters</A
></DT
><DT
>37-9. <A
HREF="#HERECOMMSUB"
>Using a <I
CLASS="FIRSTTERM"
>here document</I
>
to set a variable</A
></DT
><DT
>37-10. <A
HREF="#LASTPIPEOPT"
>Piping input to a <A
HREF="#READREF"
>read</A
></A
></DT
><DT
>37-11. <A
HREF="#NEGARRAY"
>Negative array indices</A
></DT
><DT
>37-12. <A
HREF="#NEGOFFSET"
>Negative parameter in string-extraction
construct</A
></DT
><DT
>A-1. <A
HREF="#MAILFORMAT"
><I
CLASS="FIRSTTERM"
>mailformat</I
>: Formatting an e-mail
message</A
></DT
><DT
>A-2. <A
HREF="#RN"
><I
CLASS="FIRSTTERM"
>rn</I
>: A simple-minded file renaming
utility</A
></DT
><DT
>A-3. <A
HREF="#BLANKRENAME"
><I
CLASS="FIRSTTERM"
>blank-rename</I
>: Renames filenames containing
blanks</A
></DT
><DT
>A-4. <A
HREF="#ENCRYPTEDPW"
><I
CLASS="FIRSTTERM"
>encryptedpw</I
>: Uploading to an ftp site,
using a locally encrypted password</A
></DT
><DT
>A-5. <A
HREF="#COPYCD"
><I
CLASS="FIRSTTERM"
>copy-cd</I
>: Copying a data CD</A
></DT
><DT
>A-6. <A
HREF="#COLLATZ"
>Collatz series</A
></DT
><DT
>A-7. <A
HREF="#DAYSBETWEEN"
><I
CLASS="FIRSTTERM"
>days-between</I
>: Days between two
dates</A
></DT
><DT
>A-8. <A
HREF="#MAKEDICT"
>Making a <I
CLASS="FIRSTTERM"
>dictionary</I
></A
></DT
><DT
>A-9. <A
HREF="#SOUNDEX"
>Soundex conversion</A
></DT
><DT
>A-10. <A
HREF="#LIFESLOW"
><I
CLASS="FIRSTTERM"
>Game of Life</I
></A
></DT
><DT
>A-11. <A
HREF="#GEN0DATA"
>Data file for <I
CLASS="FIRSTTERM"
>Game of Life</I
></A
></DT
><DT
>A-12. <A
HREF="#BEHEAD"
><I
CLASS="FIRSTTERM"
>behead</I
>: Removing mail and news
message headers</A
></DT
><DT
>A-13. <A
HREF="#PW"
><I
CLASS="FIRSTTERM"
>password</I
>: Generating random
8-character passwords</A
></DT
><DT
>A-14. <A
HREF="#FIFO"
><I
CLASS="FIRSTTERM"
>fifo</I
>: Making daily backups, using
named pipes</A
></DT
><DT
>A-15. <A
HREF="#PRIMES"
>Generating prime numbers using the modulo operator</A
></DT
><DT
>A-16. <A
HREF="#TREE"
><I
CLASS="FIRSTTERM"
>tree</I
>: Displaying a directory tree</A
></DT
><DT
>A-17. <A
HREF="#TREE2"
><I
CLASS="FIRSTTERM"
>tree2</I
>: Alternate directory tree script</A
></DT
><DT
>A-18. <A
HREF="#STRING"
><I
CLASS="FIRSTTERM"
>string functions</I
>: C-style string
functions</A
></DT
><DT
>A-19. <A
HREF="#DIRECTORYINFO"
>Directory information</A
></DT
><DT
>A-20. <A
HREF="#HASHLIB"
>Library of hash functions</A
></DT
><DT
>A-21. <A
HREF="#HASHEXAMPLE"
>Colorizing text using hash functions</A
></DT
><DT
>A-22. <A
HREF="#HASHEX2"
>More on hash functions</A
></DT
><DT
>A-23. <A
HREF="#USBINST"
>Mounting USB keychain storage devices</A
></DT
><DT
>A-24. <A
HREF="#TOHTML"
>Converting to HTML</A
></DT
><DT
>A-25. <A
HREF="#ARCHIVWEBLOGS"
>Preserving weblogs</A
></DT
><DT
>A-26. <A
HREF="#PROTECTLITERAL"
>Protecting literal strings</A
></DT
><DT
>A-27. <A
HREF="#UNPROTECTLITERAL"
>Unprotecting literal strings</A
></DT
><DT
>A-28. <A
HREF="#ISSPAMMER2"
>Spammer Identification</A
></DT
><DT
>A-29. <A
HREF="#WHX"
>Spammer Hunt</A
></DT
><DT
>A-30. <A
HREF="#WGETTER2"
>Making <I
CLASS="FIRSTTERM"
>wget</I
> easier to use</A
></DT
><DT
>A-31. <A
HREF="#BASHPODDER"
>A <I
CLASS="FIRSTTERM"
>podcasting</I
> script</A
></DT
><DT
>A-32. <A
HREF="#NIGHTLYBACKUP"
>Nightly backup to a firewire HD</A
></DT
><DT
>A-33. <A
HREF="#CDLL"
>An expanded <I
CLASS="FIRSTTERM"
>cd</I
> command</A
></DT
><DT
>A-34. <A
HREF="#SOUNDCARDON"
>A soundcard setup script</A
></DT
><DT
>A-35. <A
HREF="#FINDSPLIT"
>Locating split paragraphs in a text file</A
></DT
><DT
>A-36. <A
HREF="#INSERTIONSORT"
>Insertion sort</A
></DT
><DT
>A-37. <A
HREF="#STDDEV"
>Standard Deviation</A
></DT
><DT
>A-38. <A
HREF="#PADSW"
>A <I
CLASS="FIRSTTERM"
>pad</I
> file generator for shareware
authors</A
></DT
><DT
>A-39. <A
HREF="#MANED"
>A <I
CLASS="FIRSTTERM"
>man page</I
> editor</A
></DT
><DT
>A-40. <A
HREF="#PETALS"
>Petals Around the Rose</A
></DT
><DT
>A-41. <A
HREF="#QKY"
>Quacky: a Perquackey-type word game</A
></DT
><DT
>A-42. <A
HREF="#NIM"
>Nim</A
></DT
><DT
>A-43. <A
HREF="#STOPWATCH"
>A command-line stopwatch</A
></DT
><DT
>A-44. <A
HREF="#HOMEWORK"
>An all-purpose shell scripting homework assignment solution</A
></DT
><DT
>A-45. <A
HREF="#KTOUR"
>The Knight's Tour</A
></DT
><DT
>A-46. <A
HREF="#MSQUARE"
>Magic Squares</A
></DT
><DT
>A-47. <A
HREF="#FIFTEEN"
>Fifteen Puzzle</A
></DT
><DT
>A-48. <A
HREF="#HANOI2"
><I
CLASS="FIRSTTERM"
>The Towers of Hanoi, graphic
version</I
></A
></DT
><DT
>A-49. <A
HREF="#HANOI2A"
><I
CLASS="FIRSTTERM"
>The Towers of Hanoi, alternate graphic
version</I
></A
></DT
><DT
>A-50. <A
HREF="#USEGETOPT"
>An alternate version of the
<A
HREF="#GETOPTSIMPLE"
>getopt-simple.sh</A
> script</A
></DT
><DT
>A-51. <A
HREF="#USEGETOPT2"
>The version of the
<I
CLASS="FIRSTTERM"
>UseGetOpt.sh</I
> example used in the <A
HREF="#TABEXPANSION"
>Tab Expansion appendix</A
></A
></DT
><DT
>A-52. <A
HREF="#SHOWALLC"
>Cycling through all the possible color backgrounds</A
></DT
><DT
>A-53. <A
HREF="#SAMORSE"
>Morse Code Practice</A
></DT
><DT
>A-54. <A
HREF="#BASE64"
>Base64 encoding/decoding</A
></DT
><DT
>A-55. <A
HREF="#SEDAPPEND"
>Inserting text in a file using
<I
CLASS="FIRSTTERM"
>sed</I
></A
></DT
><DT
>A-56. <A
HREF="#GRONSFELD"
>The Gronsfeld Cipher</A
></DT
><DT
>A-57. <A
HREF="#BINGO"
>Bingo Number Generator</A
></DT
><DT
>A-58. <A
HREF="#BASICSREVIEWED"
>Basics Reviewed</A
></DT
><DT
>A-59. <A
HREF="#TESTEXECTIME"
>Testing execution times of various commands</A
></DT
><DT
>A-60. <A
HREF="#ASSOCARRTEST"
>Associative arrays vs. conventional arrays (execution
times)</A
></DT
><DT
>C-1. <A
HREF="#LETTERCOUNT2"
>Counting Letter Occurrences</A
></DT
><DT
>J-1. <A
HREF="#USEGETOPTEX"
>Completion script for
<I
CLASS="FIRSTTERM"
>UseGetOpt.sh</I
></A
></DT
><DT
>M-1. <A
HREF="#BASHRC"
>Sample <TT
CLASS="FILENAME"
>.bashrc</TT
> file</A
></DT
><DT
>M-2. <A
HREF="#BASHPROF"
><TT
CLASS="FILENAME"
>.bash_profile</TT
> file</A
></DT
><DT
>N-1. <A
HREF="#VIEWDAT"
>VIEWDATA.BAT: DOS Batch File</A
></DT
><DT
>N-2. <A
HREF="#VIEWDATA"
><I
CLASS="FIRSTTERM"
>viewdata.sh</I
>: Shell Script Conversion
of VIEWDATA.BAT</A
></DT
><DT
>T-1. <A
HREF="#ASCIISH"
>A script that generates an ASCII table</A
></DT
><DT
>T-2. <A
HREF="#ASCII2SH"
>Another ASCII table script</A
></DT
><DT
>T-3. <A
HREF="#ASCII3SH"
>A third ASCII table script, using
<I
CLASS="FIRSTTERM"
>awk</I
></A
></DT
></DL
></DIV
><DIV
CLASS="PART"
><A
NAME="PART1"
></A
><DIV
CLASS="TITLEPAGE"
><H1
CLASS="TITLE"
>Part 1. Introduction</H1
><DIV
CLASS="PARTINTRO"
><A
NAME="AEN43"
></A
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>Script: <EM
>A writing; a written
document. [Obs.]</EM
></I
></P
><P
><I
>--<EM
>Webster's Dictionary</EM
>, 1913 ed.</I
></P
></I
></TD
></TR
></TABLE
><P
><A
NAME="WHATSASCRIPT"
></A
></P
><P
>The shell is a command interpreter. More than just the
insulating layer between the operating system kernel and the user,
it's also a fairly powerful programming language. A shell program,
called a <I
CLASS="FIRSTTERM"
>script</I
>, is an easy-to-use tool for
building applications by <SPAN
CLASS="QUOTE"
>"gluing together"</SPAN
> system
calls, tools, utilities, and compiled binaries. Virtually the
entire repertoire of UNIX commands, utilities, and tools is
available for invocation by a shell script. If that were
not enough, internal shell commands, such as testing and loop
constructs, lend additional power and flexibility to scripts.
Shell scripts are especially well suited for administrative
system tasks and other routine repetitive tasks not requiring the
bells and whistles of a full-blown tightly structured programming
language.</P
></DIV
><DIV
CLASS="TOC"
><DL
><DT
><B
>Table of Contents</B
></DT
><DT
>1. <A
HREF="#WHY-SHELL"
>Shell Programming!</A
></DT
><DT
>2. <A
HREF="#SHA-BANG"
>Starting Off With a Sha-Bang</A
></DT
><DD
><DL
><DT
>2.1. <A
HREF="#INVOKING"
>Invoking the script</A
></DT
><DT
>2.2. <A
HREF="#PRELIMEXER"
>Preliminary Exercises</A
></DT
></DL
></DD
></DL
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="WHY-SHELL"
></A
>Chapter 1. Shell Programming!</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>No programming language is perfect. There is not even a single
best language; there are only languages well suited or perhaps
poorly suited for particular purposes.</I
></P
><P
><I
>--Herbert Mayer</I
></P
></I
></TD
></TR
></TABLE
><P
>A working knowledge of shell scripting is essential to anyone
wishing to become reasonably proficient at system administration,
even if they do not anticipate ever having to actually write a
script. Consider that as a Linux machine boots up, it executes the
shell scripts in <TT
CLASS="FILENAME"
>/etc/rc.d</TT
>
to restore the system configuration and set up services. A detailed
understanding of these startup scripts is important for analyzing
the behavior of a system, and possibly modifying it.</P
><P
>The craft of scripting is not hard to master,
since scripts can be built in bite-sized sections and there
is only a fairly small set of shell-specific operators and options
<A
NAME="AEN62"
HREF="#FTN.AEN62"
><SPAN
CLASS="footnote"
>[1]</SPAN
></A
>
to learn. The syntax is simple -- even austere -- similar to
that of invoking and chaining together utilities at the command
line, and there are only a few <SPAN
CLASS="QUOTE"
>"rules"</SPAN
> governing
their use. Most short scripts work right the first time, and
debugging even the longer ones is straightforward.</P
><P
> <A
NAME="AEN67"
></A
><BLOCKQUOTE
CLASS="BLOCKQUOTE"
><P
CLASS="LITERALLAYOUT"
>&nbsp;&nbsp;&nbsp;&nbsp;In&nbsp;the&nbsp;early&nbsp;days&nbsp;of&nbsp;personal&nbsp;computing,&nbsp;the&nbsp;BASIC&nbsp;language&nbsp;enabled<br>
&nbsp;&nbsp;&nbsp;&nbsp;anyone&nbsp;reasonably&nbsp;computer&nbsp;proficient&nbsp;to&nbsp;write&nbsp;programs&nbsp;on&nbsp;an&nbsp;early<br>
&nbsp;&nbsp;&nbsp;&nbsp;generation&nbsp;of&nbsp;microcomputers.&nbsp;Decades&nbsp;later,&nbsp;the&nbsp;Bash&nbsp;scripting<br>
&nbsp;&nbsp;&nbsp;&nbsp;language&nbsp;enables&nbsp;anyone&nbsp;with&nbsp;a&nbsp;rudimentary&nbsp;knowledge&nbsp;of&nbsp;Linux&nbsp;or<br>
&nbsp;&nbsp;&nbsp;&nbsp;UNIX&nbsp;to&nbsp;do&nbsp;the&nbsp;same&nbsp;on&nbsp;modern&nbsp;machines.<br>
<br>
&nbsp;&nbsp;&nbsp;&nbsp;We&nbsp;now&nbsp;have&nbsp;miniaturized&nbsp;single-board&nbsp;computers&nbsp;with&nbsp;amazing<br>
&nbsp;&nbsp;&nbsp;&nbsp;capabilities,&nbsp;such&nbsp;as&nbsp;the&nbsp;<A
HREF="http://www.raspberrypi.org/"
TARGET="_top"
>Raspberry Pi</A
>.<br>
&nbsp;&nbsp;&nbsp;&nbsp;Bash&nbsp;scripting&nbsp;provides&nbsp;a&nbsp;way&nbsp;to&nbsp;explore&nbsp;the&nbsp;capabilities&nbsp;of&nbsp;these<br>
&nbsp;&nbsp;&nbsp;&nbsp;fascinating&nbsp;devices.<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</P
></BLOCKQUOTE
>
</P
><P
>A shell script is a quick-and-dirty method of prototyping
a complex application. Getting even a limited subset of
the functionality to work in a script is often a useful
first stage in project development. In this way, the structure
of the application can be tested and tinkered with, and the
major pitfalls found before proceeding to the final coding
in <I
CLASS="FIRSTTERM"
>C</I
>, <I
CLASS="FIRSTTERM"
>C++</I
>,
<I
CLASS="FIRSTTERM"
>Java</I
>, <A
HREF="#PERLREF"
>Perl</A
>,
or <I
CLASS="FIRSTTERM"
>Python</I
>.</P
><P
>Shell scripting hearkens back to the classic UNIX philosophy
of breaking complex projects into simpler subtasks, of chaining
together components and utilities. Many consider this a better,
or at least more esthetically pleasing approach to problem solving
than using one of the new generation of high-powered all-in-one
languages, such as <I
CLASS="FIRSTTERM"
>Perl</I
>, which attempt to
be all things to all people, but at the cost of forcing you to
alter your thinking processes to fit the tool.</P
><P
>According to <A
HREF="#MAYERREF"
>Herbert Mayer</A
>,
<SPAN
CLASS="QUOTE"
>"a useful language needs arrays, pointers,
and a generic mechanism for building data structures."</SPAN
>
By these criteria, shell scripting falls somewhat short of being
<SPAN
CLASS="QUOTE"
>"useful."</SPAN
> Or, perhaps not. . . .</P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN82"
></A
><P
></P
><P
>When not to use shell scripts
<P
></P
><UL
><LI
><P
>Resource-intensive tasks, especially where speed is
a factor (sorting, hashing, recursion
<A
NAME="AEN87"
HREF="#FTN.AEN87"
><SPAN
CLASS="footnote"
>[2]</SPAN
></A
>
...)</P
></LI
><LI
><P
>Procedures involving heavy-duty math operations,
especially floating point arithmetic, arbitrary
precision calculations, or complex numbers (use
<I
CLASS="FIRSTTERM"
>C++</I
> or <I
CLASS="FIRSTTERM"
>FORTRAN</I
>
instead)</P
></LI
><LI
><P
>Cross-platform portability required (use
<I
CLASS="FIRSTTERM"
>C</I
> or <I
CLASS="FIRSTTERM"
>Java</I
>
instead)</P
></LI
><LI
><P
>Complex applications, where structured programming is
a necessity (type-checking of variables, function
prototypes, etc.)</P
></LI
><LI
><P
>Mission-critical applications upon which you are betting the
future of the company</P
></LI
><LI
><P
>Situations where <EM
>security</EM
> is
important, where you need to guarantee the integrity of
your system and protect against intrusion, cracking, and
vandalism</P
></LI
><LI
><P
>Project consists of subcomponents with interlocking
dependencies</P
></LI
><LI
><P
>Extensive file operations required
(<I
CLASS="FIRSTTERM"
>Bash</I
> is limited to serial file access,
and that only in a particularly clumsy and inefficient
line-by-line fashion.)</P
></LI
><LI
><P
>Need native support for multi-dimensional arrays</P
></LI
><LI
><P
>Need data structures, such as linked lists or trees</P
></LI
><LI
><P
>Need to generate / manipulate graphics or GUIs</P
></LI
><LI
><P
>Need direct access to system hardware or
external peripherals</P
></LI
><LI
><P
>Need port or <A
HREF="#SOCKETREF"
>socket</A
>
I/O</P
></LI
><LI
><P
>Need to use libraries or interface with legacy code</P
></LI
><LI
><P
>Proprietary, closed-source applications (Shell scripts
put the source code right out in the open for all the world
to see.)</P
></LI
></UL
></P
><P
>If any of the above applies, consider a more powerful scripting
language -- perhaps <I
CLASS="FIRSTTERM"
>Perl</I
>,
<I
CLASS="FIRSTTERM"
>Tcl</I
>, <I
CLASS="FIRSTTERM"
>Python</I
>,
<I
CLASS="FIRSTTERM"
>Ruby</I
> -- or possibly a
compiled language such as <I
CLASS="FIRSTTERM"
>C</I
>,
<I
CLASS="FIRSTTERM"
>C++</I
>, or <I
CLASS="FIRSTTERM"
>Java</I
>. Even
then, prototyping the application as a shell script might still
be a useful development step.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
><A
NAME="BASHDEF"
></A
></P
><P
>We will be using <SPAN
CLASS="ACRONYM"
>Bash</SPAN
>, an acronym
<A
NAME="AEN139"
HREF="#FTN.AEN139"
><SPAN
CLASS="footnote"
>[3]</SPAN
></A
>
for <SPAN
CLASS="QUOTE"
>"Bourne-Again shell"</SPAN
> and a pun on Stephen Bourne's
now classic <I
CLASS="FIRSTTERM"
>Bourne</I
> shell. Bash has become
a <I
CLASS="FOREIGNPHRASE"
>de facto</I
> standard for shell
scripting on most flavors of UNIX. Most of the principles this
book covers apply equally well to scripting with other shells,
such as the <I
CLASS="FIRSTTERM"
>Korn Shell</I
>, from which Bash
derives some of its features,
<A
NAME="AEN147"
HREF="#FTN.AEN147"
><SPAN
CLASS="footnote"
>[4]</SPAN
></A
>
and the <I
CLASS="FIRSTTERM"
>C Shell</I
> and its variants. (Note that
<I
CLASS="FIRSTTERM"
>C Shell</I
> programming is not recommended due to
certain inherent problems, as pointed out in an October, 1993 <A
HREF="http://www.faqs.org/faqs/unix-faq/shell/csh-whynot/"
TARGET="_top"
>Usenet
post</A
> by Tom Christiansen.) </P
><P
>What follows is a tutorial on shell scripting. It relies
heavily on examples to illustrate various features of the shell.
The example scripts work -- they've been tested, insofar as
possible -- and some of them are even useful in real life. The
reader can play with the actual working code of the examples
in the source archive (<TT
CLASS="FILENAME"
>scriptname.sh</TT
> or
<TT
CLASS="FILENAME"
>scriptname.bash</TT
>),
<A
NAME="AEN157"
HREF="#FTN.AEN157"
><SPAN
CLASS="footnote"
>[5]</SPAN
></A
>
give them <I
CLASS="FIRSTTERM"
>execute</I
> permission
(<TT
CLASS="USERINPUT"
><B
>chmod u+rx scriptname</B
></TT
>),
then run them to see what happens. Should the <A
HREF="http://bash.deta.in/abs-guide-latest.tar.bz2"
TARGET="_top"
>source
archive</A
> not be available, then cut-and-paste from the <A
HREF="http://www.tldp.org/LDP/abs/abs-guide.html.tar.gz"
TARGET="_top"
>HTML</A
> or
<A
HREF="http://bash.deta.in/abs-guide.pdf"
TARGET="_top"
>pdf</A
>
rendered versions. Be aware that some of the scripts presented here
introduce features before they are explained, and this may require
the reader to temporarily skip ahead for enlightenment.</P
><P
>Unless otherwise noted, <A
HREF="mailto:thegrendel.abs@gmail.com"
TARGET="_top"
>the author</A
> of this
book wrote the example scripts that follow.</P
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>His countenance was bold and bashed not.</I
></P
><P
><I
>--Edmund Spenser</I
></P
></I
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="SHA-BANG"
></A
>Chapter 2. Starting Off With a Sha-Bang</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>Shell programming is a 1950s juke box . . .</I
></P
><P
><I
>--Larry Wall</I
></P
></I
></TD
></TR
></TABLE
><P
>In the simplest case, a script is nothing more than a list
of system commands stored in a file. At the very least, this saves
the effort of retyping that particular sequence of commands each
time it is invoked.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX1"
></A
><P
><B
>Example 2-1. <I
CLASS="FIRSTTERM"
>cleanup</I
>: A script to clean up log
files in /var/log </B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Cleanup
# Run as root, of course.
cd /var/log
cat /dev/null &#62; messages
cat /dev/null &#62; wtmp
echo "Log files cleaned up."</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>There is nothing unusual here, only a set of commands that
could just as easily have been invoked one by one from the
command-line on the console or in a terminal window.
The advantages of placing the commands in a script go far beyond
not having to retype them time and again. The script becomes a
<I
CLASS="FIRSTTERM"
>program</I
> -- a <EM
>tool</EM
> --
and it can easily be modified or customized for a particular
application.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX1A"
></A
><P
><B
>Example 2-2. <I
CLASS="FIRSTTERM"
>cleanup</I
>: An improved clean-up
script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Proper header for a Bash script.
# Cleanup, version 2
# Run as root, of course.
# Insert code here to print error message and exit if not root.
LOG_DIR=/var/log
# Variables are better than hard-coded values.
cd $LOG_DIR
cat /dev/null &#62; messages
cat /dev/null &#62; wtmp
echo "Logs cleaned up."
exit # The right and proper method of "exiting" from a script.
# A bare "exit" (no parameter) returns the exit status
#+ of the preceding command. </PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Now <EM
>that's</EM
> beginning to look like a real
script. But we can go even farther . . .</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX2"
></A
><P
><B
>Example 2-3. <I
CLASS="FIRSTTERM"
>cleanup</I
>: An enhanced
and generalized version of above scripts.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Cleanup, version 3
# Warning:
# -------
# This script uses quite a number of features that will be explained
#+ later on.
# By the time you've finished the first half of the book,
#+ there should be nothing mysterious about it.
LOG_DIR=/var/log
ROOT_UID=0 # Only users with $UID 0 have root privileges.
LINES=50 # Default number of lines saved.
E_XCD=86 # Can't change directory?
E_NOTROOT=87 # Non-root exit error.
# Run as root, of course.
if [ "$UID" -ne "$ROOT_UID" ]
then
echo "Must be root to run this script."
exit $E_NOTROOT
fi
if [ -n "$1" ]
# Test whether command-line argument is present (non-empty).
then
lines=$1
else
lines=$LINES # Default, if not specified on command-line.
fi
# Stephane Chazelas suggests the following,
#+ as a better way of checking command-line arguments,
#+ but this is still a bit advanced for this stage of the tutorial.
#
# E_WRONGARGS=85 # Non-numerical argument (bad argument format).
#
# case "$1" in
# "" ) lines=50;;
# *[!0-9]*) echo "Usage: `basename $0` lines-to-cleanup";
# exit $E_WRONGARGS;;
# * ) lines=$1;;
# esac
#
#* Skip ahead to "Loops" chapter to decipher all this.
cd $LOG_DIR
if [ `pwd` != "$LOG_DIR" ] # or if [ "$PWD" != "$LOG_DIR" ]
# Not in /var/log?
then
echo "Can't change to $LOG_DIR."
exit $E_XCD
fi # Doublecheck if in right directory before messing with log file.
# Far more efficient is:
#
# cd /var/log || {
# echo "Cannot change to necessary directory." &#62;&#38;2
# exit $E_XCD;
# }
tail -n $lines messages &#62; mesg.temp # Save last section of message log file.
mv mesg.temp messages # Rename it as system log file.
# cat /dev/null &#62; messages
#* No longer needed, as the above method is safer.
cat /dev/null &#62; wtmp # ': &#62; wtmp' and '&#62; wtmp' have the same effect.
echo "Log files cleaned up."
# Note that there are other log files in /var/log not affected
#+ by this script.
exit 0
# A zero return value from the script upon exit indicates success
#+ to the shell.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Since you may not wish to wipe out the entire system log,
this version of the script keeps the last section of the message
log intact. You will constantly discover ways of fine-tuning
previously written scripts for increased effectiveness.</P
><P
><A
NAME="SHABANGREF"
></A
>* * *</P
><P
><A
NAME="MAGNUMREF"
></A
>The
<I
CLASS="FIRSTTERM"
> sha-bang</I
>
(<SPAN
CLASS="TOKEN"
> #!</SPAN
>)
<A
NAME="AEN205"
HREF="#FTN.AEN205"
><SPAN
CLASS="footnote"
>[6]</SPAN
></A
>
at the head of a script tells your system that this file is a set
of commands to be fed to the command interpreter indicated. The
<SPAN
CLASS="TOKEN"
>#!</SPAN
> is actually a two-byte
<A
NAME="AEN214"
HREF="#FTN.AEN214"
><SPAN
CLASS="footnote"
>[7]</SPAN
></A
>
<I
CLASS="FIRSTTERM"
>magic number</I
>, a special marker that
designates a file type, or in this case an executable shell
script (type <TT
CLASS="USERINPUT"
><B
>man magic</B
></TT
> for more
details on this fascinating topic). Immediately following
the <I
CLASS="FIRSTTERM"
>sha-bang</I
> is a <I
CLASS="FIRSTTERM"
>path
name</I
>. This is the path to the program that interprets
the commands in the script, whether it be a shell, a programming
language, or a utility. This command interpreter then executes
the commands in the script, starting at the top (the line
following the <I
CLASS="FIRSTTERM"
>sha-bang</I
> line), and ignoring
comments.
<A
NAME="AEN226"
HREF="#FTN.AEN226"
><SPAN
CLASS="footnote"
>[8]</SPAN
></A
>
</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/sh
#!/bin/bash
#!/usr/bin/perl
#!/usr/bin/tcl
#!/bin/sed -f
#!/bin/awk -f</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Each of the above script header lines calls a different command
interpreter, be it <TT
CLASS="FILENAME"
>/bin/sh</TT
>, the default shell
(<B
CLASS="COMMAND"
>bash</B
> in a Linux system) or otherwise.
<A
NAME="AEN242"
HREF="#FTN.AEN242"
><SPAN
CLASS="footnote"
>[9]</SPAN
></A
>
Using <TT
CLASS="USERINPUT"
><B
>#!/bin/sh</B
></TT
>, the default Bourne shell
in most commercial variants of UNIX, makes the script <A
HREF="#PORTABILITYISSUES"
>portable</A
> to non-Linux machines,
though you <A
HREF="#BINSH"
>sacrifice Bash-specific
features</A
>. The script will, however, conform to the
<SPAN
CLASS="ACRONYM"
>POSIX</SPAN
>
<A
NAME="AEN256"
HREF="#FTN.AEN256"
><SPAN
CLASS="footnote"
>[10]</SPAN
></A
>
<B
CLASS="COMMAND"
>sh</B
> standard.</P
><P
>Note that the path given at the <SPAN
CLASS="QUOTE"
>"sha-bang"</SPAN
> must
be correct, otherwise an error message -- usually <SPAN
CLASS="QUOTE"
>"Command
not found."</SPAN
> -- will be the only result of running the
script.
<A
NAME="AEN269"
HREF="#FTN.AEN269"
><SPAN
CLASS="footnote"
>[11]</SPAN
></A
>
</P
><P
><SPAN
CLASS="TOKEN"
>#!</SPAN
> can be omitted if the script consists only
of a set of generic system commands, using no internal
shell directives. The second example, above, requires the
initial <SPAN
CLASS="TOKEN"
>#!</SPAN
>, since the variable assignment line,
<TT
CLASS="USERINPUT"
><B
>lines=50</B
></TT
>, uses a shell-specific construct.
<A
NAME="AEN279"
HREF="#FTN.AEN279"
><SPAN
CLASS="footnote"
>[12]</SPAN
></A
>
Note again that <TT
CLASS="USERINPUT"
><B
>#!/bin/sh</B
></TT
> invokes the default
shell interpreter, which defaults to <TT
CLASS="FILENAME"
>/bin/bash</TT
>
on a Linux machine.</P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This tutorial encourages a modular approach
to constructing a script. Make note of and collect
<SPAN
CLASS="QUOTE"
>"boilerplate"</SPAN
> code snippets that might be useful
in future scripts. Eventually you will build quite an extensive
library of nifty routines. As an example, the following script
prolog tests whether the script has been invoked with the correct
number of parameters.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>E_WRONG_ARGS=85
script_parameters="-a -h -m -z"
# -a = all, -h = help, etc.
if [ $# -ne $Number_of_expected_args ]
then
echo "Usage: `basename $0` $script_parameters"
# `basename $0` is the script's filename.
exit $E_WRONG_ARGS
fi</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Many times, you will write a script that carries out one
particular task. The first script in this chapter is an
example. Later, it might occur to you to generalize
the script to do other, similar tasks. Replacing the literal
(<SPAN
CLASS="QUOTE"
>"hard-wired"</SPAN
>) constants by variables is a step in
that direction, as is replacing repetitive code blocks by <A
HREF="#FUNCTIONREF"
>functions</A
>.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="INVOKING"
></A
>2.1. Invoking the script</H1
><P
>Having written the script, you can invoke it by <TT
CLASS="USERINPUT"
><B
>sh
scriptname</B
></TT
>,
<A
NAME="AEN300"
HREF="#FTN.AEN300"
><SPAN
CLASS="footnote"
>[13]</SPAN
></A
>
or alternatively <TT
CLASS="USERINPUT"
><B
>bash scriptname</B
></TT
>. (Not
recommended is using <TT
CLASS="USERINPUT"
><B
>sh &#60;scriptname</B
></TT
>,
since this effectively disables reading from
<A
HREF="#STDINOUTDEF"
><TT
CLASS="FILENAME"
>stdin</TT
></A
>
within the script.) Much more convenient is to make
the script itself directly executable with a <A
HREF="#CHMODREF"
>chmod</A
>.
<P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
>Either:</DT
><DD
><P
><TT
CLASS="USERINPUT"
><B
>chmod 555 scriptname</B
></TT
> (gives
everyone read/execute permission)
<A
NAME="AEN315"
HREF="#FTN.AEN315"
><SPAN
CLASS="footnote"
>[14]</SPAN
></A
>
</P
></DD
><DT
>or</DT
><DD
><P
><TT
CLASS="USERINPUT"
><B
>chmod +rx scriptname</B
></TT
> (gives
everyone read/execute permission)</P
><P
><TT
CLASS="USERINPUT"
><B
>chmod
u+rx scriptname</B
></TT
> (gives only the
script owner read/execute permission)</P
></DD
></DL
></DIV
>
</P
><P
>Having made the script executable, you may now test it by
<TT
CLASS="USERINPUT"
><B
>./scriptname</B
></TT
>.
<A
NAME="AEN327"
HREF="#FTN.AEN327"
><SPAN
CLASS="footnote"
>[15]</SPAN
></A
>
If it begins with a <SPAN
CLASS="QUOTE"
>"sha-bang"</SPAN
> line, invoking the
script calls the correct command interpreter to run it.</P
><P
>As a final step, after testing and debugging,
you would likely want to move it to <TT
CLASS="FILENAME"
>/usr/local/bin</TT
> (as
<I
CLASS="FIRSTTERM"
>root</I
>, of course), to make the script
available to yourself and all other users as a systemwide
executable. The script could then be invoked by simply typing
<B
CLASS="COMMAND"
>scriptname</B
> <B
CLASS="KEYCAP"
>[ENTER]</B
> from the
command-line.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="PRELIMEXER"
></A
>2.2. Preliminary Exercises</H1
><P
></P
><OL
TYPE="1"
><LI
><P
>System administrators often write scripts to automate common
tasks. Give several instances where such scripts would be
useful.</P
></LI
><LI
><P
>Write a script that upon invocation shows the
<A
HREF="#DATEREF"
>time and date</A
>, <A
HREF="#WHOREF"
>lists all logged-in users</A
>, and gives
the system <A
HREF="#UPTIMEREF"
>uptime</A
>. The script
then <A
HREF="#IOREDIRREF"
>saves this information</A
>
to a logfile.</P
></LI
></OL
></DIV
></DIV
></DIV
><DIV
CLASS="PART"
><A
NAME="PART2"
></A
><DIV
CLASS="TITLEPAGE"
><H1
CLASS="TITLE"
>Part 2. Basics</H1
><DIV
CLASS="TOC"
><DL
><DT
><B
>Table of Contents</B
></DT
><DT
>3. <A
HREF="#SPECIAL-CHARS"
>Special Characters</A
></DT
><DT
>4. <A
HREF="#VARIABLES"
>Introduction to Variables and Parameters</A
></DT
><DD
><DL
><DT
>4.1. <A
HREF="#VARSUBN"
>Variable Substitution</A
></DT
><DT
>4.2. <A
HREF="#VARASSIGNMENT"
>Variable Assignment</A
></DT
><DT
>4.3. <A
HREF="#UNTYPED"
>Bash Variables Are Untyped</A
></DT
><DT
>4.4. <A
HREF="#OTHERTYPESV"
>Special Variable Types</A
></DT
></DL
></DD
><DT
>5. <A
HREF="#QUOTING"
>Quoting</A
></DT
><DD
><DL
><DT
>5.1. <A
HREF="#QUOTINGVAR"
>Quoting Variables</A
></DT
><DT
>5.2. <A
HREF="#ESCAPINGSECTION"
>Escaping</A
></DT
></DL
></DD
><DT
>6. <A
HREF="#EXIT-STATUS"
>Exit and Exit Status</A
></DT
><DT
>7. <A
HREF="#TESTS"
>Tests</A
></DT
><DD
><DL
><DT
>7.1. <A
HREF="#TESTCONSTRUCTS"
>Test Constructs</A
></DT
><DT
>7.2. <A
HREF="#FTO"
>File test operators</A
></DT
><DT
>7.3. <A
HREF="#COMPARISON-OPS"
>Other Comparison Operators</A
></DT
><DT
>7.4. <A
HREF="#NESTEDIFTHEN"
>Nested <TT
CLASS="REPLACEABLE"
><I
>if/then</I
></TT
> Condition Tests</A
></DT
><DT
>7.5. <A
HREF="#TESTTEST"
>Testing Your Knowledge of Tests</A
></DT
></DL
></DD
><DT
>8. <A
HREF="#OPERATIONS"
>Operations and Related Topics</A
></DT
><DD
><DL
><DT
>8.1. <A
HREF="#OPS"
>Operators</A
></DT
><DT
>8.2. <A
HREF="#NUMERICAL-CONSTANTS"
>Numerical Constants</A
></DT
><DT
>8.3. <A
HREF="#DBLPARENS"
>The Double-Parentheses Construct</A
></DT
><DT
>8.4. <A
HREF="#OPPRECEDENCE"
>Operator Precedence</A
></DT
></DL
></DD
></DL
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="SPECIAL-CHARS"
></A
>Chapter 3. Special Characters</H1
><P
>What makes a character <I
CLASS="FIRSTTERM"
>special</I
>?
If it has a meaning beyond its
<I
CLASS="FIRSTTERM"
>literal meaning</I
>, a <A
HREF="#METAMEANINGREF"
>meta-meaning</A
>, then we refer
to it as a <I
CLASS="FIRSTTERM"
>special character</I
>. Along
with commands and <A
HREF="#KEYWORDREF"
>keywords</A
>,
<I
CLASS="FIRSTTERM"
>special characters</I
> are building blocks
of Bash scripts.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="SCHARLIST1"
></A
>Special Characters Found In
Scripts and Elsewhere</B
></P
><DL
><DT
><A
NAME="HASHMARKREF"
></A
><SPAN
CLASS="TOKEN"
>#</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Comments. </B
>Lines beginning with a <SPAN
CLASS="TOKEN"
>#</SPAN
>
(with the exception of <A
HREF="#MAGNUMREF"
> <SPAN
CLASS="TOKEN"
>#!</SPAN
></A
>) are comments and will
<EM
>not</EM
> be executed.</P
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># This line is a comment.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Comments may also occur following the end of a command.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo "A comment will follow." # Comment here.
# ^ Note whitespace before #</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="WSBCOMM"
></A
> Comments may also follow <A
HREF="#WHITESPACEREF"
>whitespace</A
> at the beginning
of a line.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> # A tab precedes this comment.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="COMMINPIPE"
></A
>Comments may even be embedded
within a <A
HREF="#PIPEREF"
>pipe</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>initial=( `cat "$startfile" | sed -e '/#/d' | tr -d '\n' |\
# Delete lines containing '#' comment character.
sed -e 's/\./\. /g' -e 's/_/_ /g'` )
# Excerpted from life.sh script</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A command may not follow a comment on the
same line. There is no method of terminating the comment,
in order for <SPAN
CLASS="QUOTE"
>"live code"</SPAN
> to begin on the same
line. Use a new line for the next command.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Of course, a <A
HREF="#QUOTINGREF"
>quoted</A
>
or an <A
HREF="#ESCP"
>escaped</A
> <SPAN
CLASS="TOKEN"
>#</SPAN
>
in an <A
HREF="#ECHOREF"
>echo</A
> statement does
<EM
>not</EM
> begin a comment. Likewise, a
<SPAN
CLASS="TOKEN"
>#</SPAN
> appears in <A
HREF="#PSUB2"
>certain
parameter-substitution constructs</A
> and in <A
HREF="#NUMCONSTANTS"
> numerical constant expressions</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo "The # here does not begin a comment."
echo 'The # here does not begin a comment.'
echo The \# here does not begin a comment.
echo The # here begins a comment.
echo ${PATH#*:} # Parameter substitution, not a comment.
echo $(( 2#101011 )) # Base conversion, not a comment.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
The standard <A
HREF="#QUOTINGREF"
>quoting and
escape</A
> characters (" ' \) escape the #.
</P
></TD
></TR
></TABLE
></DIV
><P
>Certain <A
HREF="#PSOREX1"
>pattern matching
operations</A
> also use the <SPAN
CLASS="TOKEN"
>#</SPAN
>.</P
></DD
><DT
><A
NAME="SEMICOLONREF"
></A
><SPAN
CLASS="TOKEN"
>;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Command separator [semicolon]. </B
>Permits putting two or more commands on the same
line.</P
></DIV
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo hello; echo there
if [ -x "$filename" ]; then # Note the space after the semicolon.
#+ ^^
echo "File $filename exists."; cp $filename $filename.bak
else # ^^
echo "File $filename not found."; touch $filename
fi; echo "File test complete."</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Note that the <SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>;</SPAN
>"</SPAN
>
<A
HREF="#FINDREF0"
>sometimes needs to be
<I
CLASS="FIRSTTERM"
>escaped</I
></A
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>;;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Terminator in a <A
HREF="#CASEESAC1"
>case</A
> option [double semicolon]. </B
><A
NAME="DOUBLESEMICOLON"
></A
></P
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>case "$variable" in
abc) echo "\$variable = abc" ;;
xyz) echo "\$variable = xyz" ;;
esac</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>;;&#38;</SPAN
>, <SPAN
CLASS="TOKEN"
>;&#38;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#NCTERM"
>Terminators</A
>
in a <I
CLASS="FIRSTTERM"
>case</I
> option (<A
HREF="#BASH4REF"
>version 4+</A
> of Bash). </B
></P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>.</SPAN
></DT
><DD
><P
><A
NAME="DOTREF"
></A
></P
><DIV
CLASS="FORMALPARA"
><P
><B
><SPAN
CLASS="QUOTE"
>"dot"</SPAN
> command [period]. </B
>Equivalent to <A
HREF="#SOURCEREF"
>source</A
> (see
<A
HREF="#EX38"
>Example 15-22</A
>). This is a bash <A
HREF="#BUILTINREF"
>builtin</A
>.</P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>.</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><SPAN
CLASS="QUOTE"
>"dot"</SPAN
>, as a component of a filename. </B
>When working with filenames, a leading dot is the prefix
of a <SPAN
CLASS="QUOTE"
>"hidden"</SPAN
> file, a file that an
<A
HREF="#LSREF"
>ls</A
> will not normally show.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>touch .hidden-file</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 10
-rw-r--r-- 1 bozo 4034 Jul 18 22:04 data1.addressbook
-rw-r--r-- 1 bozo 4602 May 25 13:58 data1.addressbook.bak
-rw-r--r-- 1 bozo 877 Dec 17 2000 employment.addressbook</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -al</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 14
drwxrwxr-x 2 bozo bozo 1024 Aug 29 20:54 ./
drwx------ 52 bozo bozo 3072 Aug 29 20:51 ../
-rw-r--r-- 1 bozo bozo 4034 Jul 18 22:04 data1.addressbook
-rw-r--r-- 1 bozo bozo 4602 May 25 13:58 data1.addressbook.bak
-rw-r--r-- 1 bozo bozo 877 Dec 17 2000 employment.addressbook
-rw-rw-r-- 1 bozo bozo 0 Aug 29 20:54 .hidden-file</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DIV
><P
><A
NAME="DOTDIRECTORY"
></A
></P
><P
>When considering directory names, <I
CLASS="FIRSTTERM"
>a single
dot</I
> represents the current working directory,
and <I
CLASS="FIRSTTERM"
>two dots</I
> denote the parent
directory.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>pwd</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/home/bozo/projects</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cd .</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>pwd</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/home/bozo/projects</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cd ..</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>pwd</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/home/bozo/</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <I
CLASS="FIRSTTERM"
>dot</I
> often appears as the
destination (directory) of a file movement command,
in this context meaning <I
CLASS="FIRSTTERM"
>current
directory</I
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cp /home/bozo/current_work/junk/* .</B
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
Copy all the <SPAN
CLASS="QUOTE"
>"junk"</SPAN
> files to
<A
HREF="#PWDREF"
>$PWD</A
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>.</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><SPAN
CLASS="QUOTE"
>"dot"</SPAN
> character match. </B
>When <A
HREF="#REGEXDOT"
>matching
characters</A
>, as part of a <A
HREF="#REGEXREF"
>regular expression</A
>, a
<SPAN
CLASS="QUOTE"
>"dot"</SPAN
> <A
HREF="#REGEXDOT"
>matches a
single character</A
>.</P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>"</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#DBLQUO"
>partial
quoting</A
> [double quote]. </B
><EM
>"STRING"</EM
> preserves (from
interpretation) most of the special characters within
<EM
>STRING</EM
>. See <A
HREF="#QUOTING"
>Chapter 5</A
>.</P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>'</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#SNGLQUO"
>full
quoting</A
> [single quote]. </B
><EM
>'STRING'</EM
> preserves all special
characters within <EM
>STRING</EM
>. This is a
stronger form of quoting than <EM
>"STRING"</EM
>.
See <A
HREF="#QUOTING"
>Chapter 5</A
>.</P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>,</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#COMMAOP"
>comma
operator</A
>. </B
>The <I
CLASS="FIRSTTERM"
>comma operator</I
>
<A
NAME="AEN612"
HREF="#FTN.AEN612"
><SPAN
CLASS="footnote"
>[16]</SPAN
></A
>
links together a
series of arithmetic operations. All are evaluated,
but only the last one is returned.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>let "t2 = ((a = 9, 15 / 3))"
# Set "a = 9" and "t2 = 15 / 3"</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DIV
><P
><A
NAME="COMMAOP2"
></A
>The <I
CLASS="FIRSTTERM"
>comma</I
>
operator can also concatenate strings.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>for file in /{,usr/}bin/*calc
# ^ Find all executable files ending in "calc"
#+ in /bin and /usr/bin directories.
do
if [ -x "$file" ]
then
echo $file
fi
done
# /bin/ipcalc
# /usr/bin/kcalc
# /usr/bin/oidcalc
# /usr/bin/oocalc
# Thank you, Rory Winston, for pointing this out.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>,</SPAN
>, <SPAN
CLASS="TOKEN"
>,</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#CASEMODPARAMSUB"
>Lowercase
conversion</A
> in <I
CLASS="FIRSTTERM"
>parameter substitution</I
>
(added in <A
HREF="#BASH4REF"
>version 4</A
> of Bash). </B
></P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>\</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#ESCP"
>escape</A
> [backslash]. </B
>A quoting mechanism for single characters.</P
></DIV
><P
><TT
CLASS="USERINPUT"
><B
>\X</B
></TT
>
<I
CLASS="FIRSTTERM"
>escapes</I
> the character
<EM
>X</EM
>. This has the effect of
<SPAN
CLASS="QUOTE"
>"quoting"</SPAN
> <EM
>X</EM
>, equivalent
to <EM
>'X'</EM
>. The <SPAN
CLASS="TOKEN"
>\</SPAN
> may
be used to quote <SPAN
CLASS="TOKEN"
>"</SPAN
> and <SPAN
CLASS="TOKEN"
>'</SPAN
>,
so they are expressed literally.</P
><P
>See <A
HREF="#QUOTING"
>Chapter 5</A
> for an in-depth explanation
of escaped characters.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>/</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Filename path separator [forward slash]. </B
>Separates the components of a filename (as in
<TT
CLASS="FILENAME"
>/home/bozo/projects/Makefile</TT
>).</P
></DIV
><P
>This is also the division <A
HREF="#AROPS1"
>arithmetic operator</A
>.</P
></DD
><DT
><A
NAME="BACKTICKSREF"
></A
><SPAN
CLASS="TOKEN"
>`</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#COMMANDSUBREF"
>command substitution</A
>. </B
>The <B
CLASS="COMMAND"
>`command`</B
> construct makes
available the output of <B
CLASS="COMMAND"
>command</B
>
for assignment to a variable. This is also known as
<A
HREF="#BACKQUOTESREF"
>backquotes</A
> or
backticks.</P
></DIV
></DD
><DT
><A
NAME="COLON0REF"
></A
><SPAN
CLASS="TOKEN"
>:</SPAN
></DT
><DD
><P
><A
NAME="NULLREF"
></A
></P
><DIV
CLASS="FORMALPARA"
><P
><B
>null command [colon]. </B
>This is the shell equivalent of a
<SPAN
CLASS="QUOTE"
>"NOP"</SPAN
> (<TT
CLASS="REPLACEABLE"
><I
>no op</I
></TT
>, a
do-nothing operation). It may be considered a synonym for
the shell builtin <A
HREF="#TRUEREF"
>true</A
>. The
<SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>:</SPAN
>"</SPAN
> command is itself a
<I
CLASS="FIRSTTERM"
>Bash</I
> <A
HREF="#BUILTINREF"
>builtin</A
>, and its <A
HREF="#EXITSTATUSREF"
>exit status</A
> is
<I
CLASS="FIRSTTERM"
>true</I
>
(<SPAN
CLASS="RETURNVALUE"
>0</SPAN
>).</P
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>:
echo $? # 0</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Endless loop:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>while :
do
operation-1
operation-2
...
operation-n
done
# Same as:
# while true
# do
# ...
# done</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Placeholder in if/then test:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if condition
then : # Do nothing and branch ahead
else # Or else ...
take-some-action
fi</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Provide a placeholder where a binary operation is
expected, see <A
HREF="#ARITHOPS"
>Example 8-2</A
> and <A
HREF="#DEFPARAM"
>default parameters</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>: ${username=`whoami`}
# ${username=`whoami`} Gives an error without the leading :
# unless "username" is a command or builtin...
: ${1?"Usage: $0 ARGUMENT"} # From "usage-message.sh example script.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Provide a placeholder where a command is expected in a
<A
HREF="#HEREDOCREF"
>here document</A
>. See <A
HREF="#ANONHEREDOC"
>Example 19-10</A
>.</P
><P
>Evaluate string of variables using
<A
HREF="#PARAMSUBREF"
>parameter substitution</A
>
(as in <A
HREF="#EX6"
>Example 10-7</A
>).
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>: ${HOSTNAME?} ${USER?} ${MAIL?}
# Prints error message
#+ if one or more of essential environmental variables not set.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><B
CLASS="COMMAND"
><A
HREF="#EXPREPL1"
>Variable expansion / substring
replacement</A
></B
>.</P
><P
>In combination with the <SPAN
CLASS="TOKEN"
>&#62;</SPAN
> <A
HREF="#IOREDIRREF"
>redirection operator</A
>,
truncates a file to zero length, without changing its
permissions. If the file did not previously exist,
creates it.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>: &#62; data.xxx # File "data.xxx" now empty.
# Same effect as cat /dev/null &#62;data.xxx
# However, this does not fork a new process, since ":" is a builtin.</PRE
></FONT
></TD
></TR
></TABLE
>
See also <A
HREF="#EX12"
>Example 16-15</A
>.</P
><P
>In combination with the <SPAN
CLASS="TOKEN"
>&#62;&#62;</SPAN
>
redirection operator, has no effect on a pre-existing
target file (<TT
CLASS="USERINPUT"
><B
>: &#62;&#62; target_file</B
></TT
>).
If the file did not previously exist, creates it.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="REGFILEREF"
></A
>This applies to regular files,
not pipes, symlinks, and certain special files.</P
></TD
></TR
></TABLE
></DIV
><P
>May be used to begin a comment line, although this is not
recommended. Using <SPAN
CLASS="TOKEN"
>#</SPAN
> for a comment turns
off error checking for the remainder of that line, so
almost anything may appear in a comment. However,
this is not the case with
<SPAN
CLASS="TOKEN"
>:</SPAN
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>: This is a comment that generates an error, ( if [ $x -eq 3] ).</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>:</SPAN
>"</SPAN
> serves as a <A
HREF="#FIELDREF"
>field</A
>
separator, in <A
HREF="#DATAFILESREF1"
><TT
CLASS="FILENAME"
>/etc/passwd</TT
></A
>,
and in the <A
HREF="#PATHREF"
>$PATH</A
> variable.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $PATH</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/usr/local/bin:/bin:/usr/bin:/usr/X11R6/bin:/sbin:/usr/sbin:/usr/games</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="COLONFNAME"
></A
></P
><P
>A <I
CLASS="FIRSTTERM"
>colon</I
> is <A
HREF="#FSTRANGEREF"
>acceptable as a function name</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>:()
{
echo "The name of this function is "$FUNCNAME" "
# Why use a colon as a function name?
# It's a way of obfuscating your code.
}
:
# The name of this function is :</PRE
></FONT
></TD
></TR
></TABLE
>
This is not <A
HREF="#PORTABILITYISSUES"
>portable</A
>
behavior, and therefore not a recommended practice.
In fact, more recent releases of Bash do not permit
this usage. An underscore <B
CLASS="COMMAND"
>_</B
> works,
though.</P
><P
><A
NAME="COLONINFUNCTION"
></A
></P
><P
>A <I
CLASS="FIRSTTERM"
>colon</I
> can serve
as a placeholder in an otherwise empty
function.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>not_empty ()
{
:
} # Contains a : (null command), and so is not empty.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="NOTREF"
></A
><SPAN
CLASS="TOKEN"
>!</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>reverse (or negate) the sense of
a test or exit status [bang]. </B
>The <SPAN
CLASS="TOKEN"
>!</SPAN
> operator inverts the <A
HREF="#EXITSTATUSREF"
>exit status</A
>
of the command to which it is applied (see
<A
HREF="#NEGCOND"
>Example 6-2</A
>). It also inverts
the meaning of a test operator. This can, for
example, change the sense of <I
CLASS="FIRSTTERM"
>equal</I
>
( <A
HREF="#EQUALSIGNREF"
>=</A
>
) to <I
CLASS="FIRSTTERM"
>not-equal</I
> ( != ). The
<SPAN
CLASS="TOKEN"
>!</SPAN
> operator is a Bash <A
HREF="#KEYWORDREF"
>keyword</A
>.</P
></DIV
><P
>In a different context, the <SPAN
CLASS="TOKEN"
>!</SPAN
>
also appears in <A
HREF="#IVRREF"
>indirect variable
references</A
>.</P
><P
>In yet another context, from the <I
CLASS="FIRSTTERM"
>command
line</I
>, the <SPAN
CLASS="TOKEN"
>!</SPAN
> invokes the
Bash <I
CLASS="FIRSTTERM"
>history mechanism</I
> (see <A
HREF="#HISTCOMMANDS"
>Appendix L</A
>). Note that within a script,
the history mechanism is disabled.</P
></DD
><DT
><A
NAME="ASTERISKREF"
></A
><SPAN
CLASS="TOKEN"
>*</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>wild card [asterisk]. </B
>The <SPAN
CLASS="TOKEN"
>*</SPAN
> character serves as a <SPAN
CLASS="QUOTE"
>"wild
card"</SPAN
> for filename expansion in
<A
HREF="#GLOBBINGREF"
>globbing</A
>. By itself,
it matches every filename in a given directory.</P
></DIV
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo *</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>abs-book.sgml add-drive.sh agram.sh alias.sh</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="ASTERISKREF2"
></A
></P
><P
>The <SPAN
CLASS="TOKEN"
>*</SPAN
> also represents <A
HREF="#ASTERISKREG"
>any number
(or zero) characters</A
> in a <A
HREF="#REGEXREF"
>regular expression</A
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>*</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#AROPS1"
>arithmetic operator</A
>. </B
>In the context of arithmetic operations, the
<SPAN
CLASS="TOKEN"
>*</SPAN
> denotes multiplication.</P
></DIV
><P
><SPAN
CLASS="TOKEN"
>**</SPAN
> A double asterisk can represent the
<A
HREF="#EXPONENTIATIONREF"
>exponentiation</A
>
operator or <A
HREF="#GLOBSTARREF"
>extended
file-match</A
> <I
CLASS="FIRSTTERM"
>globbing</I
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>?</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>test operator. </B
>Within certain expressions, the <SPAN
CLASS="TOKEN"
>?</SPAN
> indicates
a test for a condition.</P
></DIV
><P
><A
NAME="CSTRINARY"
></A
></P
><P
>In a <A
HREF="#DBLPARENS"
>double-parentheses
construct</A
>, the <SPAN
CLASS="TOKEN"
>?</SPAN
>
can serve as an element of a C-style
<I
CLASS="FIRSTTERM"
>trinary</I
> operator.
<A
NAME="AEN888"
HREF="#FTN.AEN888"
><SPAN
CLASS="footnote"
>[17]</SPAN
></A
>
</P
><P
><TT
CLASS="VARNAME"
>condition</TT
><B
CLASS="COMMAND"
>?</B
><TT
CLASS="VARNAME"
>result-if-true</TT
><B
CLASS="COMMAND"
>:</B
><TT
CLASS="VARNAME"
>result-if-false</TT
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>(( var0 = var1&#60;98?9:21 ))
# ^ ^
# if [ "$var1" -lt 98 ]
# then
# var0=9
# else
# var0=21
# fi</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>In a <A
HREF="#PARAMSUBREF"
>parameter
substitution</A
> expression, the <SPAN
CLASS="TOKEN"
>?</SPAN
>
<A
HREF="#QERRMSG"
>tests whether a variable has been
set</A
>.</P
></DD
><DT
><A
NAME="WILDCARDQU"
></A
><SPAN
CLASS="TOKEN"
>?</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>wild card. </B
><A
NAME="QUEXWC"
></A
>The <SPAN
CLASS="TOKEN"
>?</SPAN
> character
serves as a single-character <SPAN
CLASS="QUOTE"
>"wild card"</SPAN
>
for filename expansion in <A
HREF="#GLOBBINGREF"
>globbing</A
>, as well as <A
HREF="#QUEXREGEX"
>representing one character</A
>
in an <A
HREF="#EXTREGEX"
>extended regular
expression</A
>.</P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>$</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#VARSUBN"
>Variable
substitution</A
> (contents of a variable). </B
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>var1=5
var2=23skidoo
echo $var1 # 5
echo $var2 # 23skidoo</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DIV
><P
><A
NAME="VARPREFIXREF"
></A
></P
><P
>A <SPAN
CLASS="TOKEN"
>$</SPAN
> prefixing a variable name
indicates the <I
CLASS="FIRSTTERM"
>value</I
> the variable
holds.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>$</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>end-of-line. </B
>In a <A
HREF="#REGEXREF"
>regular expression</A
>, a
<SPAN
CLASS="QUOTE"
>"$"</SPAN
> addresses the <A
HREF="#DOLLARSIGNREF"
>end of a line</A
> of
text.</P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>${}</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#PARAMSUBREF"
>Parameter
substitution</A
>. </B
></P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>$' ... '</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#STRQ"
>Quoted string
expansion</A
>. </B
>This construct expands single or multiple
escaped octal or hex values into ASCII
<A
NAME="AEN1001"
HREF="#FTN.AEN1001"
><SPAN
CLASS="footnote"
>[18]</SPAN
></A
>
or <A
HREF="#UNICODEREF"
>Unicode</A
>
characters.</P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>$*</SPAN
>, <SPAN
CLASS="TOKEN"
>$@</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#APPREF"
>positional
parameters</A
>. </B
></P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>$?</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>exit status variable. </B
>The <A
HREF="#EXSREF"
>$? variable</A
>
holds the <A
HREF="#EXITSTATUSREF"
>exit status</A
>
of a command, a <A
HREF="#FUNCTIONREF"
>function</A
>,
or of the script itself.</P
></DIV
></DD
><DT
><A
NAME="PROCESSIDREF"
></A
><SPAN
CLASS="TOKEN"
>$$</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>process ID variable. </B
>The <A
HREF="#PROCCID"
>$$ variable</A
>
holds the <I
CLASS="FIRSTTERM"
>process ID</I
>
<A
NAME="AEN1071"
HREF="#FTN.AEN1071"
><SPAN
CLASS="footnote"
>[19]</SPAN
></A
>
of the script in which it appears.</P
></DIV
></DD
><DT
><A
NAME="PARENSREF"
></A
><SPAN
CLASS="TOKEN"
>()</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>command group. </B
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>(a=hello; echo $a)</PRE
></FONT
></TD
></TR
></TABLE
></P
></DIV
><DIV
CLASS="IMPORTANT"
><P
></P
><TABLE
CLASS="IMPORTANT"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/important.gif"
HSPACE="5"
ALT="Important"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A listing of commands within
<TT
CLASS="REPLACEABLE"
><I
>parentheses</I
></TT
> starts a <A
HREF="#SUBSHELLSREF"
>subshell</A
>.</P
><P
>Variables inside parentheses, within the subshell, are not
visible to the rest of the script. The parent process,
the script, <A
HREF="#PARVIS"
>cannot read variables
created in the child process</A
>, the subshell.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>a=123
( a=321; )
echo "a = $a" # a = 123
# "a" within parentheses acts like a local variable.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="FORMALPARA"
><P
><B
>array initialization. </B
> <A
NAME="ARRAYINIT1"
></A
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>Array=(element1 element2 element3)</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>{xxx,yyy,zzz,...}</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Brace expansion. </B
><A
NAME="BRACEEXPREF"
></A
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo \"{These,words,are,quoted}\" # " prefix and suffix
# "These" "words" "are" "quoted"
cat {file1,file2,file3} &#62; combined_file
# Concatenates the files file1, file2, and file3 into combined_file.
cp file22.{txt,backup}
# Copies "file22.txt" to "file22.backup"</PRE
></FONT
></TD
></TR
></TABLE
></P
></DIV
><P
>A command may act upon a comma-separated list of file specs within
<TT
CLASS="REPLACEABLE"
><I
>braces</I
></TT
>.
<A
NAME="AEN1124"
HREF="#FTN.AEN1124"
><SPAN
CLASS="footnote"
>[20]</SPAN
></A
>
Filename expansion (<A
HREF="#GLOBBINGREF"
>globbing</A
>)
applies to the file specs between the braces.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>No spaces allowed within the braces
<EM
>unless</EM
> the spaces are quoted or escaped.</P
><P
><TT
CLASS="USERINPUT"
><B
>echo {file1,file2}\ :{\ A," B",' C'}</B
></TT
></P
><P
><TT
CLASS="COMPUTEROUTPUT"
>file1 : A file1 : B file1 : C file2 : A file2 : B file2 : C</TT
></P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="BRACEEXPREF33"
></A
><SPAN
CLASS="TOKEN"
>{a..z}</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Extended Brace expansion. </B
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo {a..z} # a b c d e f g h i j k l m n o p q r s t u v w x y z
# Echoes characters between a and z.
echo {0..3} # 0 1 2 3
# Echoes characters between 0 and 3.
base64_charset=( {A..Z} {a..z} {0..9} + / = )
# Initializing an array, using extended brace expansion.
# From vladz's "base64.sh" example script.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DIV
><P
>The <I
CLASS="FIRSTTERM"
>{a..z}</I
>
<A
HREF="#BRACEEXPREF3"
>extended brace
expansion</A
> construction is a feature introduced
in <A
HREF="#BASH3REF"
>version 3</A
> of
<I
CLASS="FIRSTTERM"
>Bash</I
>.</P
></DD
><DT
><A
NAME="CODEBLOCKREF"
></A
><SPAN
CLASS="TOKEN"
>{}</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Block of code [curly brackets]. </B
>Also referred to as an <I
CLASS="FIRSTTERM"
>inline group</I
>,
this construct, in effect, creates an <I
CLASS="FIRSTTERM"
>anonymous
function</I
> (a function without a
name). However, unlike in a <SPAN
CLASS="QUOTE"
>"standard"</SPAN
> <A
HREF="#FUNCTIONREF"
>function</A
>, the variables
inside a code block remain visible to the remainder of the
script.</P
></DIV
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>{ local a;
a=123; }</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bash: local: can only be used in a
function</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
> </P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>a=123
{ a=321; }
echo "a = $a" # a = 321 (value inside code block)
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="BLOCKIO"
></A
></P
><P
>The code block enclosed in braces may have <A
HREF="#IOREDIRREF"
>I/O redirected</A
> to and from
it.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX8"
></A
><P
><B
>Example 3-1. Code blocks and I/O redirection</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Reading lines in /etc/fstab.
File=/etc/fstab
{
read line1
read line2
} &#60; $File
echo "First line in $File is:"
echo "$line1"
echo
echo "Second line in $File is:"
echo "$line2"
exit 0
# Now, how do you parse the separate fields of each line?
# Hint: use awk, or . . .
# . . . Hans-Joerg Diers suggests using the "set" Bash builtin.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="BLOCKIO2"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="RPMCHECK"
></A
><P
><B
>Example 3-2. Saving the output of a code block to a file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# rpm-check.sh
# Queries an rpm file for description, listing,
#+ and whether it can be installed.
# Saves output to a file.
#
# This script illustrates using a code block.
SUCCESS=0
E_NOARGS=65
if [ -z "$1" ]
then
echo "Usage: `basename $0` rpm-file"
exit $E_NOARGS
fi
{ # Begin code block.
echo
echo "Archive Description:"
rpm -qpi $1 # Query description.
echo
echo "Archive Listing:"
rpm -qpl $1 # Query listing.
echo
rpm -i --test $1 # Query whether rpm file can be installed.
if [ "$?" -eq $SUCCESS ]
then
echo "$1 can be installed."
else
echo "$1 cannot be installed."
fi
echo # End code block.
} &#62; "$1.test" # Redirects output of everything in block to file.
echo "Results of rpm test in file $1.test"
# See rpm man page for explanation of options.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Unlike a command group within (parentheses),
as above, a code block enclosed by {braces} will
<EM
>not</EM
> normally launch a <A
HREF="#SUBSHELLSREF"
>subshell</A
>.
<A
NAME="AEN1199"
HREF="#FTN.AEN1199"
><SPAN
CLASS="footnote"
>[21]</SPAN
></A
>
</P
><P
>It is possible to <A
HREF="#ITERATIONREF"
>iterate</A
> a code block
using a <A
HREF="#NODODONE"
>non-standard
<I
CLASS="FIRSTTERM"
>for-loop</I
></A
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>{}</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>placeholder for text. </B
>Used after <A
HREF="#XARGSCURLYREF"
>xargs
<TT
CLASS="OPTION"
>-i</TT
></A
> (<I
CLASS="FIRSTTERM"
>replace
strings</I
> option). The <SPAN
CLASS="TOKEN"
>{}</SPAN
> double
curly brackets are a placeholder for output text.</P
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>ls . | xargs -i -t cp ./{} $1
# ^^ ^^
# From "ex42.sh" (copydir.sh) example.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="SEMICOLONESC"
></A
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>{} \;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>pathname. </B
>Mostly used in <A
HREF="#FINDREF"
>find</A
>
constructs. This is <EM
>not</EM
> a shell
<A
HREF="#BUILTINREF"
>builtin</A
>.</P
></DIV
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN1234"
></A
><P
></P
><P
><A
NAME="PATHNAMEREF"
></A
></P
><P
>Definition: A <I
CLASS="FIRSTTERM"
>pathname</I
>
is a <I
CLASS="FIRSTTERM"
>filename</I
> that includes the
complete <A
HREF="#PATHREF"
>path</A
>. As an example,
<TT
CLASS="FILENAME"
>/home/bozo/Notes/Thursday/schedule.txt</TT
>.
This is sometimes referred to as the <I
CLASS="FIRSTTERM"
>absolute
path</I
>.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>;</SPAN
>"</SPAN
> ends
the <TT
CLASS="OPTION"
>-exec</TT
> option of a
<B
CLASS="COMMAND"
>find</B
> command sequence. It needs
to be escaped to protect it from interpretation by the
shell.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="LEFTBRACKET"
></A
><SPAN
CLASS="TOKEN"
>[ ]</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>test. </B
><A
NAME="BRACKTEST"
></A
></P
></DIV
><P
><A
HREF="#IFTHEN"
>Test</A
> expression between
<B
CLASS="COMMAND"
>[ ]</B
>. Note that <B
CLASS="COMMAND"
>[</B
>
is part of the shell <I
CLASS="FIRSTTERM"
>builtin</I
> <A
HREF="#TTESTREF"
>test</A
> (and a synonym for it),
<EM
>not</EM
> a link to the external command
<TT
CLASS="FILENAME"
>/usr/bin/test</TT
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>[[ ]]</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>test. </B
></P
></DIV
><P
>Test expression between <SPAN
CLASS="TOKEN"
>[[ ]]</SPAN
>. More
flexible than the single-bracket <SPAN
CLASS="TOKEN"
>[ ]</SPAN
> test,
this is a shell <A
HREF="#KEYWORDREF"
>keyword</A
>.</P
><P
>See the
discussion on the <A
HREF="#DBLBRACKETS"
>[[ ... ]]
construct</A
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>[ ]</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>array element. </B
></P
></DIV
><P
>In the context of an <A
HREF="#ARRAYREF"
>array</A
>,
brackets set off the numbering of each element of that array.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>Array[1]=slot_1
echo ${Array[1]}</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>[ ]</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>range of characters. </B
></P
></DIV
><P
>As part of a <A
HREF="#REGEXREF"
>regular
expression</A
>, brackets delineate a <A
HREF="#BRACKETSREF"
>range of characters</A
> to
match.</P
></DD
><DT
><A
NAME="BRACKETARITH"
></A
><SPAN
CLASS="TOKEN"
>$[ ... ]</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>integer expansion. </B
></P
></DIV
><P
>Evaluate integer expression between
<SPAN
CLASS="TOKEN"
>$[ ]</SPAN
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>a=3
b=7
echo $[$a+$b] # 10
echo $[$a*$b] # 21</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Note that this usage is <EM
>deprecated</EM
>,
and has been replaced by the
<A
HREF="#DBLPARENS"
>(( ... ))</A
> construct.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>(( ))</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>integer expansion. </B
></P
></DIV
><P
>Expand and evaluate integer expression between
<SPAN
CLASS="TOKEN"
>(( ))</SPAN
>.</P
><P
>See the discussion on the <A
HREF="#DBLPARENS"
>(( ... )) construct</A
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>&#62;</SPAN
> <SPAN
CLASS="TOKEN"
>&#38;&#62;</SPAN
> <SPAN
CLASS="TOKEN"
>&#62;&#38;</SPAN
> <SPAN
CLASS="TOKEN"
>&#62;&#62;</SPAN
> <SPAN
CLASS="TOKEN"
>&#60;</SPAN
> <SPAN
CLASS="TOKEN"
>&#60;&#62;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#IOREDIRREF"
>redirection</A
>. </B
></P
></DIV
><P
><TT
CLASS="USERINPUT"
><B
>scriptname &#62;filename</B
></TT
> redirects the output of
<TT
CLASS="FILENAME"
>scriptname</TT
> to file
<TT
CLASS="FILENAME"
>filename</TT
>. Overwrite
<TT
CLASS="FILENAME"
>filename</TT
> if it already exists.</P
><P
><A
NAME="REDIROUTERROR"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>command &#38;&#62;filename</B
></TT
> redirects
both the <A
HREF="#STDINOUTDEF"
><TT
CLASS="FILENAME"
>stdout</TT
></A
>
and the
<TT
CLASS="FILENAME"
>stderr</TT
> of <TT
CLASS="FILENAME"
>command</TT
>
to <TT
CLASS="FILENAME"
>filename</TT
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
> <A
NAME="DEVNULLREDIRECT"
></A
>
This is useful for suppressing output when
testing for a condition. For example, let us
test whether a certain command exists.
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>type bogus_command &#38;&#62;/dev/null</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $?</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>1</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Or in a script:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>command_test () { type "$1" &#38;&#62;/dev/null; }
# ^
cmd=rmdir # Legitimate command.
command_test $cmd; echo $? # 0
cmd=bogus_command # Illegitimate command
command_test $cmd; echo $? # 1</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="REDIROUTERROR2"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>command &#62;&#38;2</B
></TT
> redirects
<TT
CLASS="FILENAME"
>stdout</TT
> of <TT
CLASS="FILENAME"
>command</TT
>
to <TT
CLASS="FILENAME"
>stderr</TT
>.</P
><P
><TT
CLASS="USERINPUT"
><B
>scriptname &#62;&#62;filename</B
></TT
> appends
the output of <TT
CLASS="FILENAME"
>scriptname</TT
>
to file <TT
CLASS="FILENAME"
>filename</TT
>. If
<TT
CLASS="FILENAME"
>filename</TT
> does not already exist,
it is created.</P
><P
><A
NAME="REDIRRW"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>[i]&#60;&#62;filename</B
></TT
>
opens file <TT
CLASS="FILENAME"
>filename</TT
> for reading
and writing, and assigns <A
HREF="#FDREF"
>file
descriptor</A
> <SPAN
CLASS="TOKEN"
>i</SPAN
> to it. If
<TT
CLASS="FILENAME"
>filename</TT
> does not exist, it is
created.</P
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#PROCESSSUBREF"
>process substitution</A
>. </B
></P
></DIV
><P
><TT
CLASS="USERINPUT"
><B
>(command)&#62;</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>&#60;(command)</B
></TT
></P
><P
><A
HREF="#LTREF"
>In a different context</A
>,
the <SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>&#60;</SPAN
>"</SPAN
> and
<SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>&#62;</SPAN
>"</SPAN
> characters act
as <A
HREF="#SCOMPARISON1"
>string comparison
operators</A
>.</P
><P
><A
HREF="#INTLT"
>In yet another context</A
>,
the <SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>&#60;</SPAN
>"</SPAN
> and
<SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>&#62;</SPAN
>"</SPAN
> characters act
as <A
HREF="#ICOMPARISON1"
>integer comparison
operators</A
>. See also <A
HREF="#EX45"
>Example 16-9</A
>.</P
></DD
><DT
><A
NAME="HEREDOCRRREF"
></A
><SPAN
CLASS="TOKEN"
>&#60;&#60;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>redirection used in a <A
HREF="#HEREDOCREF"
>here document</A
>. </B
></P
></DIV
></DD
><DT
><A
NAME="HERESTRINGREF"
></A
><SPAN
CLASS="TOKEN"
>&#60;&#60;&#60;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>redirection used in a <A
HREF="#HERESTRINGSREF"
>here string</A
>. </B
></P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>&#60;</SPAN
>, <SPAN
CLASS="TOKEN"
>&#62;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#LTREF"
>ASCII
comparison</A
>. </B
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>veg1=carrots
veg2=tomatoes
if [[ "$veg1" &#60; "$veg2" ]]
then
echo "Although $veg1 precede $veg2 in the dictionary,"
echo -n "this does not necessarily imply anything "
echo "about my culinary preferences."
else
echo "What kind of dictionary are you using, anyhow?"
fi</PRE
></FONT
></TD
></TR
></TABLE
></P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>\&#60;</SPAN
>, <SPAN
CLASS="TOKEN"
>\&#62;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#ANGLEBRAC"
>word
boundary</A
> in a <A
HREF="#REGEXREF"
>regular
expression</A
>. </B
></P
></DIV
><P
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep '\&#60;the\&#62;' textfile</B
></TT
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>|</SPAN
></DT
><DD
><P
><A
NAME="PIPEREF"
></A
></P
><DIV
CLASS="FORMALPARA"
><P
><B
>pipe. </B
>Passes the output (<TT
CLASS="FILENAME"
>stdout</TT
>)
of a previous command to the input
(<TT
CLASS="FILENAME"
>stdin</TT
>) of the next one, or
to the shell. This is a method of chaining commands
together.</P
></DIV
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo ls -l | sh
# Passes the output of "echo ls -l" to the shell,
#+ with the same result as a simple "ls -l".
cat *.lst | sort | uniq
# Merges and sorts all ".lst" files, then deletes duplicate lines.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN1555"
></A
><P
></P
><P
> A pipe, as a classic method of interprocess
communication, sends the <TT
CLASS="FILENAME"
>stdout</TT
>
of one <A
HREF="#PROCESSREF"
>process</A
> to the
<TT
CLASS="FILENAME"
>stdin</TT
> of another. In a typical case,
a command, such as <A
HREF="#CATREF"
>cat</A
> or
<A
HREF="#ECHOREF"
>echo</A
>, pipes a stream of
data to a
<A
NAME="FILTERDEF"
></A
>
<I
CLASS="FIRSTTERM"
>filter</I
>, a command that
transforms its input for processing.
<A
NAME="AEN1564"
HREF="#FTN.AEN1564"
><SPAN
CLASS="footnote"
>[22]</SPAN
></A
>
</P
><P
>
<TT
CLASS="USERINPUT"
><B
>cat $filename1 $filename2 | grep $search_word</B
></TT
>
</P
><P
>For an interesting note on the complexity of using UNIX
pipes, see <A
HREF="http://www.faqs.org/faqs/unix-faq/faq/part3/"
TARGET="_top"
>the UNIX FAQ,
Part 3</A
>.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
><A
NAME="UCREF"
></A
>The output of a command or commands
may be piped to a script.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# uppercase.sh : Changes input to uppercase.
tr 'a-z' 'A-Z'
# Letter ranges must be quoted
#+ to prevent filename generation from single-letter filenames.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
>
Now, let us pipe the output of <B
CLASS="COMMAND"
>ls -l</B
> to this
script.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l | ./uppercase.sh</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-RW-RW-R-- 1 BOZO BOZO 109 APR 7 19:49 1.TXT
-RW-RW-R-- 1 BOZO BOZO 109 APR 14 16:48 2.TXT
-RW-R--R-- 1 BOZO BOZO 725 APR 20 20:56 DATA-FILE</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <TT
CLASS="FILENAME"
>stdout</TT
> of each process in
a pipe must be read as the <TT
CLASS="FILENAME"
>stdin</TT
>
of the next. If this is not the case, the data stream
will <I
CLASS="FIRSTTERM"
>block</I
>, and the pipe will not
behave as expected.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat file1 file2 | ls -l | sort
# The output from "cat file1 file2" disappears.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>A pipe runs as a <A
HREF="#CHILDREF"
>child
process</A
>, and therefore cannot alter script
variables.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>variable="initial_value"
echo "new_value" | read variable
echo "variable = $variable" # variable = initial_value</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>If one of the commands in the pipe
aborts, this prematurely terminates execution of the
pipe. Called a <I
CLASS="FIRSTTERM"
>broken pipe</I
>, this
condition sends a <TT
CLASS="REPLACEABLE"
><I
>SIGPIPE</I
></TT
> <A
HREF="#SIGNALD"
>signal</A
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>&#62;|</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>force redirection (even if
the <A
HREF="#NOCLOBBERREF"
>noclobber option</A
>
is set). </B
>This will forcibly overwrite an existing file.</P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>||</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#ORREF"
>OR logical operator</A
>. </B
>In a <A
HREF="#TESTCONSTRUCTS1"
>test
construct</A
>, the <SPAN
CLASS="TOKEN"
>||</SPAN
> operator causes
a return of <SPAN
CLASS="RETURNVALUE"
>0</SPAN
> (success) if
<EM
>either</EM
> of the linked test conditions
is true.</P
></DIV
></DD
><DT
><A
NAME="BGJOB"
></A
><SPAN
CLASS="TOKEN"
>&#38;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Run job in background. </B
>A command followed by an <SPAN
CLASS="TOKEN"
>&#38;</SPAN
>
will run in the background.</P
></DIV
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>sleep 10 &#38;</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>[1] 850</TT
>
<TT
CLASS="COMPUTEROUTPUT"
>[1]+ Done sleep 10</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Within a script, commands and even <A
HREF="#FORLOOPREF1"
>loops</A
> may run in the
background.</P
><P
><A
NAME="BGLOOP0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="BGLOOP"
></A
><P
><B
>Example 3-3. Running a loop in the background</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# background-loop.sh
for i in 1 2 3 4 5 6 7 8 9 10 # First loop.
do
echo -n "$i "
done &#38; # Run this loop in background.
# Will sometimes execute after second loop.
echo # This 'echo' sometimes will not display.
for i in 11 12 13 14 15 16 17 18 19 20 # Second loop.
do
echo -n "$i "
done
echo # This 'echo' sometimes will not display.
# ======================================================
# The expected output from the script:
# 1 2 3 4 5 6 7 8 9 10
# 11 12 13 14 15 16 17 18 19 20
# Sometimes, though, you get:
# 11 12 13 14 15 16 17 18 19 20
# 1 2 3 4 5 6 7 8 9 10 bozo $
# (The second 'echo' doesn't execute. Why?)
# Occasionally also:
# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
# (The first 'echo' doesn't execute. Why?)
# Very rarely something like:
# 11 12 13 1 2 3 4 5 6 7 8 9 10 14 15 16 17 18 19 20
# The foreground loop preempts the background one.
exit 0
# Nasimuddin Ansari suggests adding sleep 1
#+ after the echo -n "$i" in lines 6 and 14,
#+ for some real fun.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A command run in the background within a
script may cause the script to hang, waiting
for a keystroke. Fortunately, there is a <A
HREF="#WAITHANG"
>remedy</A
> for this.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="LOGICALAND"
></A
><SPAN
CLASS="TOKEN"
>&#38;&#38;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#LOGOPS1"
>AND logical
operator</A
>. </B
>In a <A
HREF="#TESTCONSTRUCTS1"
>test
construct</A
>, the <SPAN
CLASS="TOKEN"
>&#38;&#38;</SPAN
> operator causes
a return of <SPAN
CLASS="RETURNVALUE"
>0</SPAN
> (success) only if
<EM
>both</EM
> the linked test conditions
are true.</P
></DIV
></DD
><DT
><A
NAME="DASHREF"
></A
><SPAN
CLASS="TOKEN"
>-</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>option, prefix. </B
>Option flag for a command or filter. Prefix for
an operator. Prefix for a <A
HREF="#DEFPARAM1"
>default parameter</A
>
in <A
HREF="#PARAMSUBREF"
>parameter
substitution</A
>.</P
></DIV
><P
><TT
CLASS="USERINPUT"
><B
>COMMAND -[Option1][Option2][...]</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>ls -al</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>sort -dfu $filename</B
></TT
></P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ $file1 -ot $file2 ]
then # ^
echo "File $file1 is older than $file2."
fi
if [ "$a" -eq "$b" ]
then # ^
echo "$a is equal to $b."
fi
if [ "$c" -eq 24 -a "$d" -eq 47 ]
then # ^ ^
echo "$c equals 24 and $d equals 47."
fi
param2=${param1:-$DEFAULTVAL}
# ^</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="DOUBLEDASHREF"
></A
></P
><P
><B
CLASS="COMMAND"
>--</B
></P
><P
>The <I
CLASS="FIRSTTERM"
>double-dash</I
>
<TT
CLASS="OPTION"
>--</TT
> prefixes <I
CLASS="FIRSTTERM"
>long</I
>
(verbatim) options to commands.</P
><P
><TT
CLASS="USERINPUT"
><B
>sort --ignore-leading-blanks</B
></TT
></P
><P
>Used with a <A
HREF="#BUILTINREF"
>Bash
builtin</A
>, it means the <I
CLASS="FIRSTTERM"
>end of
options</I
> to that particular command.</P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This provides a handy means of removing
files whose <EM
>names begin with a dash</EM
>.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-rw-r--r-- 1 bozo bozo 0 Nov 25 12:29 -badname</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>rm -- -badname</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 0</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><P
>The <I
CLASS="FIRSTTERM"
>double-dash</I
> is also used in
conjunction with <A
HREF="#SETREF"
>set</A
>.</P
><P
><TT
CLASS="USERINPUT"
><B
>set -- $variable</B
></TT
> (as in <A
HREF="#SETPOS"
>Example 15-18</A
>)</P
></DD
><DT
><A
NAME="DASHREF2"
></A
><SPAN
CLASS="TOKEN"
>-</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>redirection from/to <TT
CLASS="FILENAME"
>stdin</TT
> or <TT
CLASS="FILENAME"
>stdout</TT
> [dash]. </B
><A
NAME="COXEX"
></A
></P
></DIV
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat -</B
></TT
>
<TT
CLASS="USERINPUT"
><B
>abc</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>abc</TT
>
<TT
CLASS="COMPUTEROUTPUT"
>...</TT
>
<TT
CLASS="USERINPUT"
><B
>Ctl-D</B
></TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>As expected, <TT
CLASS="USERINPUT"
><B
>cat -</B
></TT
> echoes
<TT
CLASS="FILENAME"
>stdin</TT
>, in this case keyboarded user input,
to <TT
CLASS="FILENAME"
>stdout</TT
>. But, does I/O redirection using
<B
CLASS="COMMAND"
>-</B
> have real-world applications?</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>(cd /source/directory &#38;&#38; tar cf - . ) | (cd /dest/directory &#38;&#38; tar xpvf -)
# Move entire file tree from one directory to another
# [courtesy Alan Cox &#60;a.cox@swansea.ac.uk&#62;, with a minor change]
# 1) cd /source/directory
# Source directory, where the files to be moved are.
# 2) &#38;&#38;
# "And-list": if the 'cd' operation successful,
# then execute the next command.
# 3) tar cf - .
# The 'c' option 'tar' archiving command creates a new archive,
# the 'f' (file) option, followed by '-' designates the target file
# as stdout, and do it in current directory tree ('.').
# 4) |
# Piped to ...
# 5) ( ... )
# a subshell
# 6) cd /dest/directory
# Change to the destination directory.
# 7) &#38;&#38;
# "And-list", as above
# 8) tar xpvf -
# Unarchive ('x'), preserve ownership and file permissions ('p'),
# and send verbose messages to stdout ('v'),
# reading data from stdin ('f' followed by '-').
#
# Note that 'x' is a command, and 'p', 'v', 'f' are options.
#
# Whew!
# More elegant than, but equivalent to:
# cd source/directory
# tar cf - . | (cd ../dest/directory; tar xpvf -)
#
# Also having same effect:
# cp -a /source/directory/* /dest/directory
# Or:
# cp -a /source/directory/* /source/directory/.[^.]* /dest/directory
# If there are hidden files in /source/directory.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>bunzip2 -c linux-2.6.16.tar.bz2 | tar xvf -
# --uncompress tar file-- | --then pass it to "tar"--
# If "tar" has not been patched to handle "bunzip2",
#+ this needs to be done in two discrete steps, using a pipe.
# The purpose of the exercise is to unarchive "bzipped" kernel source.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Note that in this context the <SPAN
CLASS="QUOTE"
>"-"</SPAN
> is not
itself a Bash operator, but rather an option recognized by
certain UNIX utilities that write to
<TT
CLASS="FILENAME"
>stdout</TT
>, such as <B
CLASS="COMMAND"
>tar</B
>,
<B
CLASS="COMMAND"
>cat</B
>, etc.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "whatever" | cat -</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>whatever</TT
> </PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Where a filename is expected,
<TT
CLASS="REPLACEABLE"
><I
>-</I
></TT
> redirects output to
<TT
CLASS="FILENAME"
>stdout</TT
> (sometimes seen with
<TT
CLASS="USERINPUT"
><B
>tar cf</B
></TT
>), or accepts input from
<TT
CLASS="FILENAME"
>stdin</TT
>, rather than from a file.
<A
NAME="FILTERDASH"
></A
>
This is a method of using a file-oriented utility as a
filter in a pipe.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>file</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Usage: file [-bciknvzL] [-f namefile] [-m magicfiles] file...</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
By itself on the command-line, <A
HREF="#FILEREF"
>file</A
> fails with an error message.
</P
><P
> Add a <SPAN
CLASS="QUOTE"
>"-"</SPAN
> for a more useful result. This causes the
shell to await user input.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>file -</B
></TT
>
<TT
CLASS="USERINPUT"
><B
>abc</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>standard input: ASCII text</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>file -</B
></TT
>
<TT
CLASS="USERINPUT"
><B
>#!/bin/bash</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>standard input: Bourne-Again shell script text executable</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
Now the command accepts input from <TT
CLASS="FILENAME"
>stdin</TT
>
and analyzes it.
</P
><P
>The <SPAN
CLASS="QUOTE"
>"-"</SPAN
> can be used to pipe
<TT
CLASS="FILENAME"
>stdout</TT
> to other commands. This permits
such stunts as <A
HREF="#PREPENDREF"
>prepending lines
to a file</A
>.</P
><P
>Using <A
HREF="#DIFFREF"
>diff</A
> to
compare a file with a <EM
>section</EM
>
of another:</P
><P
><TT
CLASS="USERINPUT"
><B
>grep Linux file1 | diff file2 -</B
></TT
></P
><P
>Finally, a real-world example using
<TT
CLASS="REPLACEABLE"
><I
>-</I
></TT
> with <A
HREF="#TARREF"
>tar</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX58"
></A
><P
><B
>Example 3-4. Backup of all files changed in last day</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Backs up all files in current directory modified within last 24 hours
#+ in a "tarball" (tarred and gzipped file).
BACKUPFILE=backup-$(date +%m-%d-%Y)
# Embeds date in backup filename.
# Thanks, Joshua Tschida, for the idea.
archive=${1:-$BACKUPFILE}
# If no backup-archive filename specified on command-line,
#+ it will default to "backup-MM-DD-YYYY.tar.gz."
tar cvf - `find . -mtime -1 -type f -print` &#62; $archive.tar
gzip $archive.tar
echo "Directory $PWD backed up in archive file \"$archive.tar.gz\"."
# Stephane Chazelas points out that the above code will fail
#+ if there are too many files found
#+ or if any filenames contain blank characters.
# He suggests the following alternatives:
# -------------------------------------------------------------------
# find . -mtime -1 -type f -print0 | xargs -0 tar rvf "$archive.tar"
# using the GNU version of "find".
# find . -mtime -1 -type f -exec tar rvf "$archive.tar" '{}' \;
# portable to other UNIX flavors, but much slower.
# -------------------------------------------------------------------
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Filenames beginning with
<SPAN
CLASS="QUOTE"
>"-"</SPAN
> may cause problems when coupled with the
<SPAN
CLASS="QUOTE"
>"-"</SPAN
> redirection operator. A script should
check for this and add an appropriate prefix to such
filenames, for example <TT
CLASS="FILENAME"
>./-FILENAME</TT
>,
<TT
CLASS="FILENAME"
>$PWD/-FILENAME</TT
>, or
<TT
CLASS="FILENAME"
>$PATHNAME/-FILENAME</TT
>.</P
><P
>If the value of a variable begins with a
<TT
CLASS="REPLACEABLE"
><I
>-</I
></TT
>, this may likewise create
problems.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>var="-n"
echo $var
# Has the effect of "echo -n", and outputs nothing.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>-</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>previous working directory. </B
>A <B
CLASS="COMMAND"
>cd -</B
> command changes to the
previous working directory. This uses the
<A
HREF="#OLDPWD"
>$OLDPWD</A
> <A
HREF="#ENVREF"
>environmental variable</A
>.</P
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Do not confuse the <SPAN
CLASS="QUOTE"
>"-"</SPAN
> used in this
sense with the <SPAN
CLASS="QUOTE"
>"-"</SPAN
> redirection
operator just discussed. The interpretation of the
<SPAN
CLASS="QUOTE"
>"-"</SPAN
> depends on the context in which it
appears.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>-</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Minus. </B
>Minus sign in an <A
HREF="#AROPS1"
>arithmetic
operation</A
>.</P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>=</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Equals. </B
><A
HREF="#EQREF"
>Assignment operator</A
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>a=28
echo $a # 28</PRE
></FONT
></TD
></TR
></TABLE
></P
></DIV
><P
>In a <A
HREF="#EQUALSIGNREF"
>different context</A
>,
the <SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>=</SPAN
>"</SPAN
> is a <A
HREF="#SCOMPARISON1"
>string comparison</A
>
operator.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>+</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Plus. </B
>Addition <A
HREF="#AROPS1"
>arithmetic
operator</A
>.</P
></DIV
><P
>In a <A
HREF="#PLUSREF"
>different context</A
>,
the <SPAN
CLASS="TOKEN"
>+</SPAN
> is a <A
HREF="#REGEXP"
>Regular
Expression</A
> operator.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>+</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Option. </B
>Option flag for a command or filter.</P
></DIV
><P
>Certain commands and <A
HREF="#BUILTINREF"
>builtins</A
> use the
<TT
CLASS="OPTION"
>+</TT
> to enable certain options and the
<TT
CLASS="OPTION"
>-</TT
> to disable them. In <A
HREF="#PARAMSUBREF"
>parameter substitution</A
>,
the <TT
CLASS="OPTION"
>+</TT
> prefixes an <A
HREF="#PARAMALTV"
> alternate value</A
> that a variable expands to.</P
></DD
><DT
><A
NAME="MODULO00"
></A
><SPAN
CLASS="TOKEN"
>%</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#MODULOREF"
>modulo</A
>. </B
>Modulo (remainder of a division) <A
HREF="#AROPS1"
>arithmetic
operation</A
>.</P
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>let "z = 5 % 3"
echo $z # 2</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>In a <A
HREF="#PCTPATREF"
>different context</A
>,
the <SPAN
CLASS="TOKEN"
>%</SPAN
> is a <A
HREF="#PSUB2"
>pattern
matching</A
> operator.</P
></DD
><DT
><A
NAME="TILDEREF"
></A
><SPAN
CLASS="TOKEN"
>~</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>home directory [tilde]. </B
>This corresponds to the <A
HREF="#HOMEDIRREF"
>$HOME</A
> internal variable.
<TT
CLASS="FILENAME"
>~bozo</TT
> is bozo's home directory,
and <B
CLASS="COMMAND"
>ls ~bozo</B
> lists the contents of it.
<SPAN
CLASS="TOKEN"
>~/</SPAN
> is the current user's home directory,
and <B
CLASS="COMMAND"
>ls ~/</B
> lists the contents of it.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo ~bozo</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/home/bozo</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo ~</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/home/bozo</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo ~/</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/home/bozo/</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo ~:</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/home/bozo:</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo ~nonexistent-user</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>~nonexistent-user</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DIV
></DD
><DT
><A
NAME="WORKINGDIRREF"
></A
><SPAN
CLASS="TOKEN"
>~+</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>current working directory. </B
>This corresponds to the <A
HREF="#PWDREF"
>$PWD</A
> internal variable.</P
></DIV
></DD
><DT
><A
NAME="PREVWORKINGDIR"
></A
><SPAN
CLASS="TOKEN"
>~-</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>previous working directory. </B
>This corresponds to the <A
HREF="#OLDPWD"
>$OLDPWD</A
> internal variable.</P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>=~</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#REGEXMATCHREF"
>regular
expression match</A
>. </B
>This operator was introduced with <A
HREF="#BASH3REF"
>version 3</A
> of Bash.</P
></DIV
></DD
><DT
><A
NAME="BEGLINEREF"
></A
><SPAN
CLASS="TOKEN"
>^</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>beginning-of-line. </B
>In a <A
HREF="#REGEXREF"
>regular expression</A
>, a
<SPAN
CLASS="QUOTE"
>"^"</SPAN
> addresses the <A
HREF="#CARETREF"
>beginning of a line</A
> of text.</P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>^</SPAN
>, <SPAN
CLASS="TOKEN"
>^^</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="#CASEMODPARAMSUB"
>Uppercase
conversion</A
> in <I
CLASS="FIRSTTERM"
>parameter substitution</I
>
(added in <A
HREF="#BASH4REF"
>version 4</A
> of Bash). </B
></P
></DIV
></DD
><DT
><A
NAME="CONTROLCHARREF"
></A
>Control Characters</DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
> change the behavior of the
terminal or text display. </B
>A control character is a <B
CLASS="KEYCAP"
>CONTROL</B
>
+ <B
CLASS="KEYCAP"
>key</B
> combination (pressed
simultaneously).
A control character may also
be written in <I
CLASS="FIRSTTERM"
>octal</I
> or
<I
CLASS="FIRSTTERM"
>hexadecimal</I
> notation,
following an <I
CLASS="FIRSTTERM"
>escape</I
>.</P
></DIV
><P
>Control characters are not normally useful inside a
script.</P
><P
></P
><UL
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-A</B
></TT
></P
><P
>Moves cursor to beginning of line of text
(on the command-line).</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-B</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>Backspace</B
></TT
>
(nondestructive).</P
></LI
><LI
><P
><A
NAME="CTLCREF"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-C</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>Break</B
></TT
>.
Terminate a foreground job.</P
></LI
><LI
><P
><A
NAME="CTLDREF"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-D</B
></TT
></P
><P
><I
CLASS="FIRSTTERM"
>Log out</I
> from a shell (similar to
<A
HREF="#EXITCOMMANDREF"
>exit</A
>).</P
><P
><TT
CLASS="USERINPUT"
><B
>EOF</B
></TT
> (end-of-file). This also
terminates input from <TT
CLASS="FILENAME"
>stdin</TT
>.</P
><P
>When typing text on the console or in an
<I
CLASS="FIRSTTERM"
>xterm</I
> window,
<TT
CLASS="USERINPUT"
><B
>Ctl-D</B
></TT
> erases the character under the
cursor. When there are no characters present,
<TT
CLASS="USERINPUT"
><B
>Ctl-D</B
></TT
> logs out of the session, as
expected. In an <I
CLASS="FIRSTTERM"
>xterm</I
> window,
this has the effect of closing the window.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-E</B
></TT
></P
><P
>Moves cursor to end of line of text
(on the command-line).</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-F</B
></TT
></P
><P
>Moves cursor forward one character position
(on the command-line).</P
></LI
><LI
><P
><A
NAME="CTLGREF"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-G</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>BEL</B
></TT
>. On some
old-time teletype terminals, this would actually ring
a bell. In an <I
CLASS="FIRSTTERM"
>xterm</I
> it might
beep.</P
></LI
><LI
><P
><A
NAME="CTLHREF"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-H</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>Rubout</B
></TT
> (destructive backspace).
Erases characters the cursor backs over while
backspacing.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Embedding Ctl-H in a string.
a="^H^H" # Two Ctl-H's -- backspaces
# ctl-V ctl-H, using vi/vim
echo "abcdef" # abcdef
echo
echo -n "abcdef$a " # abcd f
# Space at end ^ ^ Backspaces twice.
echo
echo -n "abcdef$a" # abcdef
# No space at end ^ Doesn't backspace (why?).
# Results may not be quite as expected.
echo; echo
# Constantin Hagemeier suggests trying:
# a=$'\010\010'
# a=$'\b\b'
# a=$'\x08\x08'
# But, this does not change the results.
########################################
# Now, try this.
rubout="^H^H^H^H^H" # 5 x Ctl-H.
echo -n "12345678"
sleep 2
echo -n "$rubout"
sleep 2</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-I</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>Horizontal tab</B
></TT
>.</P
></LI
><LI
><P
><A
NAME="CTLJREF"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-J</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>Newline</B
></TT
> (line feed).
In a script, may also be expressed in octal notation --
'\012' or in hexadecimal -- '\x0a'.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-K</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>Vertical tab</B
></TT
>.</P
><P
>When typing text on the console or in an
<I
CLASS="FIRSTTERM"
>xterm</I
> window,
<TT
CLASS="USERINPUT"
><B
>Ctl-K</B
></TT
> erases from the character
under the cursor to end of line. Within a script,
<TT
CLASS="USERINPUT"
><B
>Ctl-K</B
></TT
> may behave differently,
as in Lee Lee Maschmeyer's example, below.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-L</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>Formfeed</B
></TT
> (clear the terminal
screen). In a terminal, this has the same effect as the
<A
HREF="#CLEARREF"
>clear</A
> command. When sent
to a printer, a <TT
CLASS="USERINPUT"
><B
>Ctl-L</B
></TT
> causes
an advance to end of the paper sheet.</P
></LI
><LI
><P
><A
NAME="CTLMREF"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-M</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>Carriage return</B
></TT
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Thank you, Lee Maschmeyer, for this example.
read -n 1 -s -p \
$'Control-M leaves cursor at beginning of this line. Press Enter. \x0d'
# Of course, '0d' is the hex equivalent of Control-M.
echo &#62;&#38;2 # The '-s' makes anything typed silent,
#+ so it is necessary to go to new line explicitly.
read -n 1 -s -p $'Control-J leaves cursor on next line. \x0a'
# '0a' is the hex equivalent of Control-J, linefeed.
echo &#62;&#38;2
###
read -n 1 -s -p $'And Control-K\x0bgoes straight down.'
echo &#62;&#38;2 # Control-K is vertical tab.
# A better example of the effect of a vertical tab is:
var=$'\x0aThis is the bottom line\x0bThis is the top line\x0a'
echo "$var"
# This works the same way as the above example. However:
echo "$var" | col
# This causes the right end of the line to be higher than the left end.
# It also explains why we started and ended with a line feed --
#+ to avoid a garbled screen.
# As Lee Maschmeyer explains:
# --------------------------
# In the [first vertical tab example] . . . the vertical tab
#+ makes the printing go straight down without a carriage return.
# This is true only on devices, such as the Linux console,
#+ that can't go "backward."
# The real purpose of VT is to go straight UP, not down.
# It can be used to print superscripts on a printer.
# The col utility can be used to emulate the proper behavior of VT.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-N</B
></TT
></P
><P
>Erases a line of text recalled from
<I
CLASS="FIRSTTERM"
>history buffer</I
>
<A
NAME="AEN2107"
HREF="#FTN.AEN2107"
><SPAN
CLASS="footnote"
>[23]</SPAN
></A
> (on the
command-line).</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-O</B
></TT
></P
><P
>Issues a <I
CLASS="FIRSTTERM"
>newline</I
>
(on the command-line).</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-P</B
></TT
></P
><P
>Recalls last command from <I
CLASS="FIRSTTERM"
>history
buffer</I
> (on the command-line).</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-Q</B
></TT
></P
><P
>Resume (<TT
CLASS="USERINPUT"
><B
>XON</B
></TT
>).</P
><P
>This resumes <TT
CLASS="FILENAME"
>stdin</TT
> in a terminal.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-R</B
></TT
></P
><P
>Backwards search for text in <I
CLASS="FIRSTTERM"
>history
buffer</I
>
(on the command-line).</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-S</B
></TT
></P
><P
>Suspend (<TT
CLASS="USERINPUT"
><B
>XOFF</B
></TT
>).</P
><P
>This freezes <TT
CLASS="FILENAME"
>stdin</TT
> in a terminal.
(Use Ctl-Q to restore input.)</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-T</B
></TT
></P
><P
>Reverses the position of the character the cursor
is on with the previous character (on the
command-line).</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-U</B
></TT
></P
><P
>Erase a line of input, from the cursor backward to
beginning of line. In some settings,
<TT
CLASS="USERINPUT"
><B
>Ctl-U</B
></TT
> erases the entire
line of input, <EM
>regardless of cursor
position</EM
>.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-V</B
></TT
></P
><P
>When inputting text, <TT
CLASS="USERINPUT"
><B
>Ctl-V</B
></TT
>
permits inserting control characters. For example, the
following two are equivalent:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo -e '\x0a'
echo &#60;Ctl-V&#62;&#60;Ctl-J&#62;</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-V</B
></TT
> is primarily useful from
within a text editor.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-W</B
></TT
></P
><P
>When typing text on the console or in an xterm window,
<TT
CLASS="USERINPUT"
><B
>Ctl-W</B
></TT
> erases from the character
under the cursor backwards to the first instance of
<A
HREF="#WHITESPACEREF"
>whitespace</A
>. In
some settings, <TT
CLASS="USERINPUT"
><B
>Ctl-W</B
></TT
> erases
backwards to first non-alphanumeric character.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-X</B
></TT
></P
><P
>In certain word processing programs,
<I
CLASS="FIRSTTERM"
>Cuts</I
> highlighted text
and copies to <I
CLASS="FIRSTTERM"
>clipboard</I
>.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-Y</B
></TT
></P
><P
><I
CLASS="FIRSTTERM"
>Pastes</I
> back text previously
erased (with <TT
CLASS="USERINPUT"
><B
>Ctl-U</B
></TT
> or
<TT
CLASS="USERINPUT"
><B
>Ctl-W</B
></TT
>).</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>Ctl-Z</B
></TT
></P
><P
><I
CLASS="FIRSTTERM"
>Pauses</I
> a foreground job.</P
><P
><I
CLASS="FIRSTTERM"
>Substitute</I
> operation in certain
word processing applications.</P
><P
><TT
CLASS="USERINPUT"
><B
>EOF</B
></TT
> (end-of-file) character
in the MSDOS filesystem.</P
></LI
></UL
></DD
><DT
><A
NAME="WHITESPACEREF"
></A
>Whitespace</DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>functions as a separator between commands and/or
variables. </B
>Whitespace consists of either
<I
CLASS="FIRSTTERM"
>spaces</I
>,
<I
CLASS="FIRSTTERM"
>tabs</I
>, <I
CLASS="FIRSTTERM"
>blank
lines</I
>, or any combination thereof.
<A
NAME="AEN2198"
HREF="#FTN.AEN2198"
><SPAN
CLASS="footnote"
>[24]</SPAN
></A
>
In some contexts, such as <A
HREF="#WSBAD"
>variable
assignment</A
>, whitespace is not permitted, and
results in a syntax error.</P
></DIV
><P
>Blank lines have no effect on the action of a script,
and are therefore useful for visually separating functional
sections.</P
><P
><A
HREF="#IFSREF"
>$IFS</A
>, the special variable
separating <I
CLASS="FIRSTTERM"
>fields</I
> of input to certain
commands. It defaults to whitespace.</P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN2207"
></A
><P
></P
><P
> <A
NAME="FIELDREF"
></A
><TT
CLASS="USERINPUT"
><B
>Definition:</B
></TT
>
A <I
CLASS="FIRSTTERM"
>field</I
> is a discrete chunk of data
expressed as a string of consecutive characters.
Separating each field from adjacent fields is either
<I
CLASS="FIRSTTERM"
>whitespace</I
> or some other designated
character (often determined by the <SPAN
CLASS="TOKEN"
>$IFS</SPAN
>).
In some contexts, a field may be called a
<I
CLASS="FIRSTTERM"
>record</I
>.
</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
><A
NAME="QUOTINGWS"
></A
></P
><P
>To preserve <I
CLASS="FIRSTTERM"
>whitespace</I
>
within a string or in a variable, use <A
HREF="#QUOTINGREF"
>quoting</A
>.</P
><P
>UNIX <A
HREF="#FILTERDEF"
>filters</A
>
can target and operate on <I
CLASS="FIRSTTERM"
>whitespace</I
>
using the <A
HREF="#POSIXREF"
>POSIX</A
> character class
<A
HREF="#WSPOSIX"
>[:space:]</A
>.</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="VARIABLES"
></A
>Chapter 4. Introduction to Variables and Parameters</H1
><P
><I
CLASS="FIRSTTERM"
>Variables</I
> are how programming and
scripting languages represent data. A variable is nothing
more than a <I
CLASS="FIRSTTERM"
>label</I
>, a name assigned to a
location or set of locations in computer memory holding an item
of data.</P
><P
>Variables appear in arithmetic operations and manipulation of
quantities, and in string parsing.</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="VARSUBN"
></A
>4.1. Variable Substitution</H1
><P
>The <I
CLASS="FIRSTTERM"
>name</I
> of a variable is a placeholder
for its <I
CLASS="FIRSTTERM"
>value</I
>, the data it holds.
Referencing (retrieving) its value is called
<I
CLASS="FIRSTTERM"
>variable substitution</I
>.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><SPAN
CLASS="TOKEN"
>$</SPAN
></DT
><DD
><P
><A
NAME="VARNAMEVAL"
></A
></P
><P
>Let us carefully distinguish between the
<I
CLASS="FIRSTTERM"
>name</I
> of a variable
and its <I
CLASS="FIRSTTERM"
>value</I
>. If
<TT
CLASS="USERINPUT"
><B
>variable1</B
></TT
> is the name of a
variable, then <TT
CLASS="USERINPUT"
><B
>$variable1</B
></TT
>
is a reference to its <I
CLASS="FIRSTTERM"
>value</I
>,
the data item it contains.
<A
NAME="AEN2258"
HREF="#FTN.AEN2258"
><SPAN
CLASS="footnote"
>[25]</SPAN
></A
>
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>variable1=23</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo variable1</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>variable1</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $variable1</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>23</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The only times a variable appears <SPAN
CLASS="QUOTE"
>"naked"</SPAN
>
-- without the <SPAN
CLASS="TOKEN"
>$</SPAN
> prefix -- is when
declared or assigned, when <I
CLASS="FIRSTTERM"
>unset</I
>,
when <A
HREF="#EXPORTREF"
>exported</A
>,
in an arithmetic expression within <A
HREF="#DBLPARENS"
>double parentheses
(( ... ))</A
>, or in the special case of a variable
representing a <A
HREF="#SIGNALD"
>signal</A
>
(see <A
HREF="#EX76"
>Example 32-5</A
>). Assignment may be with an
<SPAN
CLASS="TOKEN"
>=</SPAN
> (as in <TT
CLASS="PARAMETER"
><I
>var1=27</I
></TT
>),
in a <A
HREF="#READREF"
>read</A
> statement,
and at the head of a loop (<TT
CLASS="PARAMETER"
><I
>for var2 in 1
2 3</I
></TT
>).</P
><P
><A
NAME="DBLQUO"
></A
>Enclosing a referenced value in
<I
CLASS="FIRSTTERM"
>double quotes</I
> (<SPAN
CLASS="TOKEN"
>" ... "</SPAN
>)
does not interfere with variable substitution. This is
called <I
CLASS="FIRSTTERM"
>partial quoting</I
>, sometimes
referred to as <SPAN
CLASS="QUOTE"
>"weak quoting."</SPAN
> <A
NAME="SNGLQUO"
></A
>Using single quotes (<SPAN
CLASS="TOKEN"
>' ... '</SPAN
>)
causes the variable name to be used literally, and no
substitution will take place. This is <I
CLASS="FIRSTTERM"
>full
quoting</I
>, sometimes referred to as 'strong
quoting.' See <A
HREF="#QUOTING"
>Chapter 5</A
> for a
detailed discussion.</P
><P
>Note that <TT
CLASS="USERINPUT"
><B
>$variable</B
></TT
> is actually a
simplified form of
<TT
CLASS="USERINPUT"
><B
>${variable}</B
></TT
>. In contexts
where the <TT
CLASS="USERINPUT"
><B
>$variable</B
></TT
> syntax
causes an error, the longer form may work (see <A
HREF="#PARAMETER-SUBSTITUTION"
>Section 10.2</A
>, below).</P
><P
><A
NAME="VARUNSETTING"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX9"
></A
><P
><B
>Example 4-1. Variable assignment and substitution</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ex9.sh
# Variables: assignment and substitution
a=375
hello=$a
# ^ ^
#-------------------------------------------------------------------------
# No space permitted on either side of = sign when initializing variables.
# What happens if there is a space?
# "VARIABLE =value"
# ^
#% Script tries to run "VARIABLE" command with one argument, "=value".
# "VARIABLE= value"
# ^
#% Script tries to run "value" command with
#+ the environmental variable "VARIABLE" set to "".
#-------------------------------------------------------------------------
echo hello # hello
# Not a variable reference, just the string "hello" ...
echo $hello # 375
# ^ This *is* a variable reference.
echo ${hello} # 375
# Likewise a variable reference, as above.
# Quoting . . .
echo "$hello" # 375
echo "${hello}" # 375
echo
hello="A B C D"
echo $hello # A B C D
echo "$hello" # A B C D
# As we see, echo $hello and echo "$hello" give different results.
# =======================================
# Quoting a variable preserves whitespace.
# =======================================
echo
echo '$hello' # $hello
# ^ ^
# Variable referencing disabled (escaped) by single quotes,
#+ which causes the "$" to be interpreted literally.
# Notice the effect of different types of quoting.
hello= # Setting it to a null value.
echo "\$hello (null value) = $hello" # $hello (null value) =
# Note that setting a variable to a null value is not the same as
#+ unsetting it, although the end result is the same (see below).
# --------------------------------------------------------------
# It is permissible to set multiple variables on the same line,
#+ if separated by white space.
# Caution, this may reduce legibility, and may not be portable.
var1=21 var2=22 var3=$V3
echo
echo "var1=$var1 var2=$var2 var3=$var3"
# May cause problems with legacy versions of "sh" . . .
# --------------------------------------------------------------
echo; echo
numbers="one two three"
# ^ ^
other_numbers="1 2 3"
# ^ ^
# If there is whitespace embedded within a variable,
#+ then quotes are necessary.
# other_numbers=1 2 3 # Gives an error message.
echo "numbers = $numbers"
echo "other_numbers = $other_numbers" # other_numbers = 1 2 3
# Escaping the whitespace also works.
mixed_bag=2\ ---\ Whatever
# ^ ^ Space after escape (\).
echo "$mixed_bag" # 2 --- Whatever
echo; echo
echo "uninitialized_variable = $uninitialized_variable"
# Uninitialized variable has null value (no value at all!).
uninitialized_variable= # Declaring, but not initializing it --
#+ same as setting it to a null value, as above.
echo "uninitialized_variable = $uninitialized_variable"
# It still has a null value.
uninitialized_variable=23 # Set it.
unset uninitialized_variable # Unset it.
echo "uninitialized_variable = $uninitialized_variable"
# uninitialized_variable =
# It still has a null value.
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="UNINITVAR1"
></A
></P
><P
>An uninitialized variable has a
<SPAN
CLASS="QUOTE"
>"null"</SPAN
> value -- no assigned value at all
(<EM
>not</EM
> zero!).
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ -z "$unassigned" ]
then
echo "\$unassigned is NULL."
fi # $unassigned is NULL.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Using a variable before
assigning a value to it may cause problems.
It is nevertheless possible to perform arithmetic operations
on an uninitialized variable.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo "$uninitialized" # (blank line)
let "uninitialized += 5" # Add 5 to it.
echo "$uninitialized" # 5
# Conclusion:
# An uninitialized variable has no value,
#+ however it evaluates as 0 in an arithmetic operation.</PRE
></FONT
></TD
></TR
></TABLE
>
See also <A
HREF="#SELFSOURCE"
>Example 15-23</A
>.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="VARASSIGNMENT"
></A
>4.2. Variable Assignment</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="EQREF"
></A
><SPAN
CLASS="TOKEN"
>=</SPAN
></DT
><DD
><P
>the assignment operator (<EM
>no space before
and after</EM
>)</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Do not confuse this with <A
HREF="#EQUALSIGNREF"
>=</A
> and
<A
HREF="#EQUALREF"
>-eq</A
>, which
<A
HREF="#IFTHEN"
>test</A
>,
rather than assign!</P
><P
>Note that <SPAN
CLASS="TOKEN"
>=</SPAN
> can be either
an <I
CLASS="FIRSTTERM"
>assignment</I
> or a
<I
CLASS="FIRSTTERM"
>test</I
> operator, depending on
context.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="EX15_0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX15"
></A
><P
><B
>Example 4-2. Plain Variable Assignment</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Naked variables
echo
# When is a variable "naked", i.e., lacking the '$' in front?
# When it is being assigned, rather than referenced.
# Assignment
a=879
echo "The value of \"a\" is $a."
# Assignment using 'let'
let a=16+5
echo "The value of \"a\" is now $a."
echo
# In a 'for' loop (really, a type of disguised assignment):
echo -n "Values of \"a\" in the loop are: "
for a in 7 8 9 11
do
echo -n "$a "
done
echo
echo
# In a 'read' statement (also a type of assignment):
echo -n "Enter \"a\" "
read a
echo "The value of \"a\" is now $a."
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="EX16_0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX16"
></A
><P
><B
>Example 4-3. Variable Assignment, plain and fancy</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
a=23 # Simple case
echo $a
b=$a
echo $b
# Now, getting a little bit fancier (command substitution).
a=`echo Hello!` # Assigns result of 'echo' command to 'a' ...
echo $a
# Note that including an exclamation mark (!) within a
#+ command substitution construct will not work from the command-line,
#+ since this triggers the Bash "history mechanism."
# Inside a script, however, the history functions are disabled by default.
a=`ls -l` # Assigns result of 'ls -l' command to 'a'
echo $a # Unquoted, however, it removes tabs and newlines.
echo
echo "$a" # The quoted variable preserves whitespace.
# (See the chapter on "Quoting.")
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="COMMANDSUBREF0"
></A
></P
><P
>Variable assignment using the <I
CLASS="FIRSTTERM"
>$(...)</I
>
mechanism (a newer method than <A
HREF="#BACKQUOTESREF"
>backquotes</A
>). This is
likewise a form of <A
HREF="#COMMANDSUBREF"
>command
substitution</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># From /etc/rc.d/rc.local
R=$(cat /etc/redhat-release)
arch=$(uname -m)</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="UNTYPED"
></A
>4.3. Bash Variables Are Untyped</H1
><P
><A
NAME="BVUNTYPED"
></A
></P
><P
>Unlike many other programming languages, Bash does not segregate
its variables by <SPAN
CLASS="QUOTE"
>"type."</SPAN
> Essentially, <EM
>Bash
variables are character strings</EM
>, but, depending on
context, Bash permits arithmetic operations and comparisons on
variables. The determining factor is whether the value of a
variable contains only digits.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="INTORSTRING"
></A
><P
><B
>Example 4-4. Integer or string?</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# int-or-string.sh
a=2334 # Integer.
let "a += 1"
echo "a = $a " # a = 2335
echo # Integer, still.
b=${a/23/BB} # Substitute "BB" for "23".
# This transforms $b into a string.
echo "b = $b" # b = BB35
declare -i b # Declaring it an integer doesn't help.
echo "b = $b" # b = BB35
let "b += 1" # BB35 + 1
echo "b = $b" # b = 1
echo # Bash sets the "integer value" of a string to 0.
c=BB34
echo "c = $c" # c = BB34
d=${c/BB/23} # Substitute "23" for "BB".
# This makes $d an integer.
echo "d = $d" # d = 2334
let "d += 1" # 2334 + 1
echo "d = $d" # d = 2335
echo
# What about null variables?
e='' # ... Or e="" ... Or e=
echo "e = $e" # e =
let "e += 1" # Arithmetic operations allowed on a null variable?
echo "e = $e" # e = 1
echo # Null variable transformed into an integer.
# What about undeclared variables?
echo "f = $f" # f =
let "f += 1" # Arithmetic operations allowed?
echo "f = $f" # f = 1
echo # Undeclared variable transformed into an integer.
#
# However ...
let "f /= $undecl_var" # Divide by zero?
# let: f /= : syntax error: operand expected (error token is " ")
# Syntax error! Variable $undecl_var is not set to zero here!
#
# But still ...
let "f /= 0"
# let: f /= 0: division by 0 (error token is "0")
# Expected behavior.
# Bash (usually) sets the "integer value" of null to zero
#+ when performing an arithmetic operation.
# But, don't try this at home, folks!
# It's undocumented and probably non-portable behavior.
# Conclusion: Variables in Bash are untyped,
#+ with all attendant consequences.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Untyped variables are both a blessing and a curse. They permit
more flexibility in scripting and make it easier to grind out
lines of code (and give you enough rope to hang yourself!).
However, they likewise permit subtle errors to creep in
and encourage sloppy programming habits.</P
><P
>To lighten the burden of keeping track of variable
types in a script, Bash <EM
>does</EM
> permit
<A
HREF="#DECLAREREF"
>declaring</A
> variables.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="OTHERTYPESV"
></A
>4.4. Special Variable Types</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><TT
CLASS="REPLACEABLE"
><I
>Local variables</I
></TT
></DT
><DD
><P
>Variables <A
HREF="#SCOPEREF"
>visible</A
> only within a <A
HREF="#CODEBLOCKREF"
>code block</A
> or function (see
also <A
HREF="#LOCALREF"
>local variables</A
> in
<A
HREF="#FUNCTIONREF"
>functions</A
>)</P
></DD
><DT
><A
NAME="ENVREF"
></A
><TT
CLASS="REPLACEABLE"
><I
>Environmental variables</I
></TT
></DT
><DD
><P
>Variables that affect the behavior of the shell and
user interface</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>In a more general context, each <A
HREF="#PROCESSREF"
>process</A
> has an
<SPAN
CLASS="QUOTE"
>"environment"</SPAN
>, that is, a group of
variables that the process may reference. In this sense,
the shell behaves like any other process.</P
><P
>Every time a shell starts, it creates shell variables that
correspond to its own environmental variables. Updating
or adding new environmental variables causes the
shell to update its environment, and all the shell's
<I
CLASS="FIRSTTERM"
>child processes</I
> (the commands it
executes) inherit this environment.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The space allotted to the environment is limited.
Creating too many environmental variables or ones that use up
excessive space may cause problems.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>eval "`seq 10000 | sed -e 's/.*/export var&#38;=ZZZZZZZZZZZZZZ/'`"</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>du</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bash: /usr/bin/du: Argument list too long</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Note: this <SPAN
CLASS="QUOTE"
>"error"</SPAN
> has been fixed, as of
kernel version 2.6.23.</P
><P
>(Thank you, St<53>phane Chazelas for the clarification,
and for providing the above example.)</P
></TD
></TR
></TABLE
></DIV
><P
>If a script sets environmental variables, they need to be
<SPAN
CLASS="QUOTE"
>"exported,"</SPAN
> that is, reported to the
<I
CLASS="FIRSTTERM"
>environment</I
> local to
the script. This is the function of the <A
HREF="#EXPORTREF"
>export</A
> command.</P
><A
NAME="CHILDREF"
></A
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A script can <B
CLASS="COMMAND"
>export</B
> variables only
to child <A
HREF="#PROCESSREF"
>processes</A
>,
that is, only to commands or processes which that
particular script initiates. A script invoked from
the command-line <TT
CLASS="REPLACEABLE"
><I
>cannot</I
></TT
>
export variables back to the command-line environment.
<EM
><A
HREF="#FORKREF"
>Child processes</A
>
cannot export variables back to the parent processes that
spawned them.</EM
></P
><P
><A
NAME="CHILDREF2"
></A
><TT
CLASS="USERINPUT"
><B
>Definition:</B
></TT
>
A <I
CLASS="FIRSTTERM"
>child process</I
> is a
subprocess launched by another process, its <A
HREF="#PARENTREF"
>parent</A
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="POSPARAMREF1"
></A
><TT
CLASS="REPLACEABLE"
><I
>Positional parameters</I
></TT
></DT
><DD
><P
>Arguments passed to the script from the command
line
<A
NAME="AEN2450"
HREF="#FTN.AEN2450"
><SPAN
CLASS="footnote"
>[26]</SPAN
></A
>
: <TT
CLASS="VARNAME"
>$0</TT
>, <TT
CLASS="VARNAME"
>$1</TT
>,
<TT
CLASS="VARNAME"
>$2</TT
>, <TT
CLASS="VARNAME"
>$3</TT
> . . .</P
><P
><A
NAME="SCRNAMEPARAM"
></A
><TT
CLASS="VARNAME"
>$0</TT
> is
the name of the script itself,
<TT
CLASS="VARNAME"
>$1</TT
> is the first argument,
<TT
CLASS="VARNAME"
>$2</TT
> the second, <TT
CLASS="VARNAME"
>$3</TT
>
the third, and so forth.
<A
NAME="AEN2464"
HREF="#FTN.AEN2464"
><SPAN
CLASS="footnote"
>[27]</SPAN
></A
>
<A
NAME="BRACKETNOTATION"
></A
>
After <TT
CLASS="VARNAME"
>$9</TT
>, the arguments must be enclosed
in brackets, for example, <TT
CLASS="VARNAME"
>${10}</TT
>,
<TT
CLASS="VARNAME"
>${11}</TT
>, <TT
CLASS="VARNAME"
>${12}</TT
>.</P
><P
>The special variables <A
HREF="#APPREF"
>$* and $@</A
>
denote <EM
>all</EM
> the positional parameters.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX17"
></A
><P
><B
>Example 4-5. Positional Parameters</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Call this script with at least 10 parameters, for example
# ./scriptname 1 2 3 4 5 6 7 8 9 10
MINPARAMS=10
echo
echo "The name of this script is \"$0\"."
# Adds ./ for current directory
echo "The name of this script is \"`basename $0`\"."
# Strips out path name info (see 'basename')
echo
if [ -n "$1" ] # Tested variable is quoted.
then
echo "Parameter #1 is $1" # Need quotes to escape #
fi
if [ -n "$2" ]
then
echo "Parameter #2 is $2"
fi
if [ -n "$3" ]
then
echo "Parameter #3 is $3"
fi
# ...
if [ -n "${10}" ] # Parameters &#62; $9 must be enclosed in {brackets}.
then
echo "Parameter #10 is ${10}"
fi
echo "-----------------------------------"
echo "All the command-line parameters are: "$*""
if [ $# -lt "$MINPARAMS" ]
then
echo
echo "This script needs at least $MINPARAMS command-line arguments!"
fi
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><I
CLASS="FIRSTTERM"
>Bracket notation</I
> for positional
parameters leads to a fairly simple way of referencing
the <EM
>last</EM
> argument passed to a
script on the command-line. This also requires <A
HREF="#VARREFNEW"
>indirect referencing</A
>.</P
><P
><A
NAME="LASTARGREF"
></A
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>args=$# # Number of args passed.
lastarg=${!args}
# Note: This is an *indirect reference* to $args ...
# Or: lastarg=${!#} (Thanks, Chris Monson.)
# This is an *indirect reference* to the $# variable.
# Note that lastarg=${!$#} doesn't work.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Some scripts can perform different operations,
depending on which name they are invoked with. For this
to work, the script needs to check <TT
CLASS="VARNAME"
>$0</TT
>,
the name it was invoked by.
<A
NAME="AEN2501"
HREF="#FTN.AEN2501"
><SPAN
CLASS="footnote"
>[28]</SPAN
></A
>
There must also exist symbolic links to all the alternate
names of the script. See <A
HREF="#HELLOL"
>Example 16-2</A
>.</P
><P
><A
NAME="NULLVAR"
></A
></P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>If a script expects a command-line parameter
but is invoked without one, this may cause a <I
CLASS="FIRSTTERM"
>null
variable assignment</I
>, generally an undesirable
result. One way to prevent this is to append an extra
character to both sides of the assignment statement using
the expected positional parameter. </P
></TD
></TR
></TABLE
></DIV
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>variable1_=$1_ # Rather than variable1=$1
# This will prevent an error, even if positional parameter is absent.
critical_argument01=$variable1_
# The extra character can be stripped off later, like so.
variable1=${variable1_/_/}
# Side effects only if $variable1_ begins with an underscore.
# This uses one of the parameter substitution templates discussed later.
# (Leaving out the replacement pattern results in a deletion.)
# A more straightforward way of dealing with this is
#+ to simply test whether expected positional parameters have been passed.
if [ -z $1 ]
then
exit $E_MISSING_POS_PARAM
fi
# However, as Fabian Kreutz points out,
#+ the above method may have unexpected side-effects.
# A better method is parameter substitution:
# ${1:-$DefaultVal}
# See the "Parameter Substition" section
#+ in the "Variables Revisited" chapter.</PRE
></FONT
></TD
></TR
></TABLE
><P
>---</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX18"
></A
><P
><B
>Example 4-6. <I
CLASS="FIRSTTERM"
>wh</I
>, <I
CLASS="FIRSTTERM"
> whois</I
> domain name lookup</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ex18.sh
# Does a 'whois domain-name' lookup on any of 3 alternate servers:
# ripe.net, cw.net, radb.net
# Place this script -- renamed 'wh' -- in /usr/local/bin
# Requires symbolic links:
# ln -s /usr/local/bin/wh /usr/local/bin/wh-ripe
# ln -s /usr/local/bin/wh /usr/local/bin/wh-apnic
# ln -s /usr/local/bin/wh /usr/local/bin/wh-tucows
E_NOARGS=75
if [ -z "$1" ]
then
echo "Usage: `basename $0` [domain-name]"
exit $E_NOARGS
fi
# Check script name and call proper server.
case `basename $0` in # Or: case ${0##*/} in
"wh" ) whois $1@whois.tucows.com;;
"wh-ripe" ) whois $1@whois.ripe.net;;
"wh-apnic" ) whois $1@whois.apnic.net;;
"wh-cw" ) whois $1@whois.cw.net;;
* ) echo "Usage: `basename $0` [domain-name]";;
esac
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>---</P
><P
><A
NAME="SHIFTREF"
></A
></P
><P
>
The <B
CLASS="COMMAND"
>shift</B
> command reassigns the positional
parameters, in effect shifting them to the left one notch.</P
><P
><TT
CLASS="VARNAME"
>$1</TT
> &#60;--- <TT
CLASS="VARNAME"
>$2</TT
>, <TT
CLASS="VARNAME"
>$2</TT
> &#60;--- <TT
CLASS="VARNAME"
>$3</TT
>, <TT
CLASS="VARNAME"
>$3</TT
> &#60;--- <TT
CLASS="VARNAME"
>$4</TT
>, etc.</P
><P
>The old <TT
CLASS="VARNAME"
>$1</TT
> disappears, but
<EM
><TT
CLASS="VARNAME"
>$0</TT
> (the script name)
does not change</EM
>. If you use a large number of
positional parameters to a script, <B
CLASS="COMMAND"
>shift</B
>
lets you access those past <TT
CLASS="LITERAL"
>10</TT
>, although
<A
HREF="#BRACKETNOTATION"
>{bracket} notation</A
>
also permits this.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX19"
></A
><P
><B
>Example 4-7. Using <I
CLASS="FIRSTTERM"
>shift</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# shft.sh: Using 'shift' to step through all the positional parameters.
# Name this script something like shft.sh,
#+ and invoke it with some parameters.
#+ For example:
# sh shft.sh a b c def 83 barndoor
until [ -z "$1" ] # Until all parameters used up . . .
do
echo -n "$1 "
shift
done
echo # Extra linefeed.
# But, what happens to the "used-up" parameters?
echo "$2"
# Nothing echoes!
# When $2 shifts into $1 (and there is no $3 to shift into $2)
#+ then $2 remains empty.
# So, it is not a parameter *copy*, but a *move*.
exit
# See also the echo-params.sh script for a "shiftless"
#+ alternative method of stepping through the positional params.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The <B
CLASS="COMMAND"
>shift</B
> command can take a numerical
parameter indicating how many positions to shift.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# shift-past.sh
shift 3 # Shift 3 positions.
# n=3; shift $n
# Has the same effect.
echo "$1"
exit 0
# ======================== #
$ sh shift-past.sh 1 2 3 4 5
4
# However, as Eleni Fragkiadaki, points out,
#+ attempting a 'shift' past the number of
#+ positional parameters ($#) returns an exit status of 1,
#+ and the positional parameters themselves do not change.
# This means possibly getting stuck in an endless loop. . . .
# For example:
# until [ -z "$1" ]
# do
# echo -n "$1 "
# shift 20 # If less than 20 pos params,
# done #+ then loop never ends!
#
# When in doubt, add a sanity check. . . .
# shift 20 || break
# ^^^^^^^^</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>shift</B
> command works in a similar
fashion on parameters passed to a <A
HREF="#FUNCTIONREF"
>function</A
>. See <A
HREF="#MULTIPLICATION"
>Example 36-18</A
>.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="QUOTING"
></A
>Chapter 5. Quoting</H1
><P
><A
NAME="QUOTINGREF"
></A
></P
><P
>Quoting means just that, bracketing a string in quotes. This
has the effect of protecting <A
HREF="#SCHARLIST1"
>special
characters</A
> in the string from reinterpretation
or expansion by the shell or shell script. (A character
is <SPAN
CLASS="QUOTE"
>"special"</SPAN
> if it has an interpretation
other than its literal meaning. For example, the <A
HREF="#ASTERISKREF"
>asterisk *</A
> represents
a <I
CLASS="FIRSTTERM"
>wild card</I
> character in
<A
HREF="#GLOBBINGREF"
>globbing</A
> and <A
HREF="#REGEXREF"
>Regular Expressions</A
>).</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l [Vv]*</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-rw-rw-r-- 1 bozo bozo 324 Apr 2 15:05 VIEWDATA.BAT
-rw-rw-r-- 1 bozo bozo 507 May 4 14:25 vartrace.sh
-rw-rw-r-- 1 bozo bozo 539 Apr 14 17:11 viewdata.sh</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l '[Vv]*'</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>ls: [Vv]*: No such file or directory</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="QUOTINGDEF"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN2596"
></A
><P
></P
><P
>In everyday speech or writing, when we
<SPAN
CLASS="QUOTE"
>"quote"</SPAN
> a phrase, we set it apart and give it special
meaning. In a Bash script, when we <I
CLASS="FIRSTTERM"
>quote</I
> a
string, we set it apart and protect its <I
CLASS="FIRSTTERM"
>literal</I
>
meaning.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
>Certain programs and utilities reinterpret or expand
special characters in a quoted string. An important use of
quoting is protecting a command-line parameter from the shell,
but still letting the calling program expand it.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep '[Ff]irst' *.txt</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>file1.txt:This is the first line of file1.txt.
file2.txt:This is the First line of file2.txt.</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Note that the unquoted <TT
CLASS="USERINPUT"
><B
>grep [Ff]irst *.txt</B
></TT
>
works under the Bash shell.
<A
NAME="AEN2609"
HREF="#FTN.AEN2609"
><SPAN
CLASS="footnote"
>[29]</SPAN
></A
>
</P
><P
>Quoting can also suppress <A
HREF="#ECHOREF"
>echo's</A
>
<SPAN
CLASS="QUOTE"
>"appetite"</SPAN
> for newlines.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $(ls -l)</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 8 -rw-rw-r-- 1 bo bo 13 Aug 21 12:57 t.sh -rw-rw-r-- 1 bo bo 78 Aug 21 12:57 u.sh</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "$(ls -l)"</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 8
-rw-rw-r-- 1 bo bo 13 Aug 21 12:57 t.sh
-rw-rw-r-- 1 bo bo 78 Aug 21 12:57 u.sh</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="QUOTINGVAR"
></A
>5.1. Quoting Variables</H1
><P
>When referencing a variable, it is generally advisable to
enclose its name in double quotes.
This prevents reinterpretation of all special characters within
the quoted string -- except <SPAN
CLASS="TOKEN"
>$</SPAN
>, <SPAN
CLASS="TOKEN"
>`</SPAN
>
(backquote), and <SPAN
CLASS="TOKEN"
>\</SPAN
> (escape).
<A
NAME="AEN2630"
HREF="#FTN.AEN2630"
><SPAN
CLASS="footnote"
>[30]</SPAN
></A
>
Keeping <SPAN
CLASS="TOKEN"
>$</SPAN
> as a special character within
double quotes permits referencing a quoted variable
(<TT
CLASS="REPLACEABLE"
><I
>"$variable"</I
></TT
>), that is, replacing the
variable with its value (see <A
HREF="#EX9"
>Example 4-1</A
>, above).</P
><P
><A
NAME="WSQUO"
></A
></P
><P
>Use double quotes to prevent word splitting.
<A
NAME="AEN2688"
HREF="#FTN.AEN2688"
><SPAN
CLASS="footnote"
>[31]</SPAN
></A
>
An argument enclosed in double quotes presents
itself as a single word, even if it contains <A
HREF="#WHITESPACEREF"
>whitespace</A
> separators.</P
><P
><A
NAME="VARSPLITTING"
></A
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>List="one two three"
for a in $List # Splits the variable in parts at whitespace.
do
echo "$a"
done
# one
# two
# three
echo "---"
for a in "$List" # Preserves whitespace in a single variable.
do # ^ ^
echo "$a"
done
# one two three</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>A more elaborate example:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>variable1="a variable containing five words"
COMMAND This is $variable1 # Executes COMMAND with 7 arguments:
# "This" "is" "a" "variable" "containing" "five" "words"
COMMAND "This is $variable1" # Executes COMMAND with 1 argument:
# "This is a variable containing five words"
variable2="" # Empty.
COMMAND $variable2 $variable2 $variable2
# Executes COMMAND with no arguments.
COMMAND "$variable2" "$variable2" "$variable2"
# Executes COMMAND with 3 empty arguments.
COMMAND "$variable2 $variable2 $variable2"
# Executes COMMAND with 1 argument (2 spaces).
# Thanks, St<53>phane Chazelas.</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Enclosing the arguments to an <B
CLASS="COMMAND"
>echo</B
>
statement in double quotes is necessary only when word splitting
or preservation of <A
HREF="#WHITESPACEREF"
>whitespace</A
>
is an issue.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="WEIRDVARS"
></A
><P
><B
>Example 5-1. Echoing Weird Variables</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# weirdvars.sh: Echoing weird variables.
echo
var="'(]\\{}\$\""
echo $var # '(]\{}$"
echo "$var" # '(]\{}$" Doesn't make a difference.
echo
IFS='\'
echo $var # '(] {}$" \ converted to space. Why?
echo "$var" # '(]\{}$"
# Examples above supplied by Stephane Chazelas.
echo
var2="\\\\\""
echo $var2 # "
echo "$var2" # \\"
echo
# But ... var2="\\\\"" is illegal. Why?
var3='\\\\'
echo "$var3" # \\\\
# Strong quoting works, though.
# ************************************************************ #
# As the first example above shows, nesting quotes is permitted.
echo "$(echo '"')" # "
# ^ ^
# At times this comes in useful.
var1="Two bits"
echo "\$var1 = "$var1"" # $var1 = Two bits
# ^ ^
# Or, as Chris Hiestand points out ...
if [[ "$(du "$My_File1")" -gt "$(du "$My_File2")" ]]
# ^ ^ ^ ^ ^ ^ ^ ^
then
...
fi
# ************************************************************ #</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Single quotes (<SPAN
CLASS="TOKEN"
>' '</SPAN
>) operate similarly to double
quotes, but do not permit referencing variables, since
the special meaning of <SPAN
CLASS="TOKEN"
>$</SPAN
> is turned off.
Within single quotes, <EM
>every</EM
> special
character except <SPAN
CLASS="TOKEN"
>'</SPAN
> gets interpreted literally.
Consider single quotes (<SPAN
CLASS="QUOTE"
>"full quoting"</SPAN
>) to be a
stricter method of quoting than double quotes (<SPAN
CLASS="QUOTE"
>"partial
quoting"</SPAN
>).</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Since even the escape character (<SPAN
CLASS="TOKEN"
>\</SPAN
>)
gets a literal interpretation within single quotes, trying to
enclose a single quote within single quotes will not yield the
expected result.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo "Why can't I write 's between single quotes"
echo
# The roundabout method.
echo 'Why can'\''t I write '"'"'s between single quotes'
# |-------| |----------| |-----------------------|
# Three single-quoted strings, with escaped and quoted single quotes between.
# This example courtesy of St<53>phane Chazelas.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="ESCAPINGSECTION"
></A
>5.2. Escaping</H1
><P
><A
NAME="ESCP"
></A
><I
CLASS="FIRSTTERM"
>Escaping</I
> is a method
of quoting single characters. The <SPAN
CLASS="TOKEN"
>escape</SPAN
>
(<SPAN
CLASS="TOKEN"
>\</SPAN
>) preceding a character tells the shell to
interpret that character literally.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>With certain commands and utilities, such as <A
HREF="#ECHOREF"
>echo</A
> and <A
HREF="#SEDREF"
>sed</A
>, escaping a character may have the
opposite effect - it can toggle on a special meaning for that
character.</P
></TD
></TR
></TABLE
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="SPM"
></A
>Special meanings of certain
escaped characters</B
></P
><DL
><DT
>used with <B
CLASS="COMMAND"
>echo</B
> and
<B
CLASS="COMMAND"
>sed</B
></DT
><DD
><P
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>\n</SPAN
></DT
><DD
><P
>means newline</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>\r</SPAN
></DT
><DD
><P
>means return</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>\t</SPAN
></DT
><DD
><P
>means tab</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>\v</SPAN
></DT
><DD
><P
> means vertical tab</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>\b</SPAN
></DT
><DD
><P
>means backspace</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>\a</SPAN
></DT
><DD
><P
>means <I
CLASS="FIRSTTERM"
>alert</I
> (beep or flash)</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>\0xx</SPAN
></DT
><DD
><P
><A
NAME="OCTALREF"
></A
>translates to the
octal <A
HREF="#ASCIIDEF"
>ASCII</A
>
equivalent of <TT
CLASS="REPLACEABLE"
><I
>0nn</I
></TT
>, where
<TT
CLASS="REPLACEABLE"
><I
>nn</I
></TT
> is a string of digits</P
><DIV
CLASS="IMPORTANT"
><P
></P
><TABLE
CLASS="IMPORTANT"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/important.gif"
HSPACE="5"
ALT="Important"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="STRQ"
></A
></P
><P
>The <TT
CLASS="USERINPUT"
><B
>$' ... '</B
></TT
>
<A
HREF="#QUOTINGREF"
>quoted</A
> string-expansion
construct is a mechanism that uses escaped octal or hex values
to assign ASCII characters to variables, e.g.,
<B
CLASS="COMMAND"
>quote=$'\042'</B
>.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ESCAPED"
></A
><P
><B
>Example 5-2. Escaped Characters</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# escaped.sh: escaped characters
#############################################################
### First, let's show some basic escaped-character usage. ###
#############################################################
# Escaping a newline.
# ------------------
echo ""
echo "This will print
as two lines."
# This will print
# as two lines.
echo "This will print \
as one line."
# This will print as one line.
echo; echo
echo "============="
echo "\v\v\v\v" # Prints \v\v\v\v literally.
# Use the -e option with 'echo' to print escaped characters.
echo "============="
echo "VERTICAL TABS"
echo -e "\v\v\v\v" # Prints 4 vertical tabs.
echo "=============="
echo "QUOTATION MARK"
echo -e "\042" # Prints " (quote, octal ASCII character 42).
echo "=============="
# The $'\X' construct makes the -e option unnecessary.
echo; echo "NEWLINE and (maybe) BEEP"
echo $'\n' # Newline.
echo $'\a' # Alert (beep).
# May only flash, not beep, depending on terminal.
# We have seen $'\nnn" string expansion, and now . . .
# =================================================================== #
# Version 2 of Bash introduced the $'\nnn' string expansion construct.
# =================================================================== #
echo "Introducing the \$\' ... \' string-expansion construct . . . "
echo ". . . featuring more quotation marks."
echo $'\t \042 \t' # Quote (") framed by tabs.
# Note that '\nnn' is an octal value.
# It also works with hexadecimal values, in an $'\xhhh' construct.
echo $'\t \x22 \t' # Quote (") framed by tabs.
# Thank you, Greg Keraunen, for pointing this out.
# Earlier Bash versions allowed '\x022'.
echo
# Assigning ASCII characters to a variable.
# ----------------------------------------
quote=$'\042' # " assigned to a variable.
echo "$quote Quoted string $quote and this lies outside the quotes."
echo
# Concatenating ASCII chars in a variable.
triple_underline=$'\137\137\137' # 137 is octal ASCII code for '_'.
echo "$triple_underline UNDERLINE $triple_underline"
echo
ABC=$'\101\102\103\010' # 101, 102, 103 are octal A, B, C.
echo $ABC
echo
escape=$'\033' # 033 is octal for escape.
echo "\"escape\" echoes as $escape"
# no visible output.
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>A more elaborate example:</P
><DIV
CLASS="EXAMPLE"
><A
NAME="BASHEK"
></A
><P
><B
>Example 5-3. Detecting key-presses</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Author: Sigurd Solaas, 20 Apr 2011
# Used in ABS Guide with permission.
# Requires version 4.2+ of Bash.
key="no value yet"
while true; do
clear
echo "Bash Extra Keys Demo. Keys to try:"
echo
echo "* Insert, Delete, Home, End, Page_Up and Page_Down"
echo "* The four arrow keys"
echo "* Tab, enter, escape, and space key"
echo "* The letter and number keys, etc."
echo
echo " d = show date/time"
echo " q = quit"
echo "================================"
echo
# Convert the separate home-key to home-key_num_7:
if [ "$key" = $'\x1b\x4f\x48' ]; then
key=$'\x1b\x5b\x31\x7e'
# Quoted string-expansion construct.
fi
# Convert the separate end-key to end-key_num_1.
if [ "$key" = $'\x1b\x4f\x46' ]; then
key=$'\x1b\x5b\x34\x7e'
fi
case "$key" in
$'\x1b\x5b\x32\x7e') # Insert
echo Insert Key
;;
$'\x1b\x5b\x33\x7e') # Delete
echo Delete Key
;;
$'\x1b\x5b\x31\x7e') # Home_key_num_7
echo Home Key
;;
$'\x1b\x5b\x34\x7e') # End_key_num_1
echo End Key
;;
$'\x1b\x5b\x35\x7e') # Page_Up
echo Page_Up
;;
$'\x1b\x5b\x36\x7e') # Page_Down
echo Page_Down
;;
$'\x1b\x5b\x41') # Up_arrow
echo Up arrow
;;
$'\x1b\x5b\x42') # Down_arrow
echo Down arrow
;;
$'\x1b\x5b\x43') # Right_arrow
echo Right arrow
;;
$'\x1b\x5b\x44') # Left_arrow
echo Left arrow
;;
$'\x09') # Tab
echo Tab Key
;;
$'\x0a') # Enter
echo Enter Key
;;
$'\x1b') # Escape
echo Escape Key
;;
$'\x20') # Space
echo Space Key
;;
d)
date
;;
q)
echo Time to quit...
echo
exit 0
;;
*)
echo You pressed: \'"$key"\'
;;
esac
echo
echo "================================"
unset K1 K2 K3
read -s -N1 -p "Press a key: "
K1="$REPLY"
read -s -N2 -t 0.001
K2="$REPLY"
read -s -N1 -t 0.001
K3="$REPLY"
key="$K1$K2$K3"
done
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#EX77"
>Example 37-1</A
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>\"</SPAN
></DT
><DD
><P
> gives the quote its literal meaning</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo "Hello" # Hello
echo "\"Hello\" ... he said." # "Hello" ... he said.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>\$</SPAN
></DT
><DD
><P
>gives the dollar sign its literal meaning
(variable name following <SPAN
CLASS="TOKEN"
>\$</SPAN
> will not be
referenced)</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo "\$variable01" # $variable01
echo "The book cost \$7.98." # The book cost $7.98.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>\\</SPAN
></DT
><DD
><P
>gives the backslash its literal meaning</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo "\\" # Results in \
# Whereas . . .
echo "\" # Invokes secondary prompt from the command-line.
# In a script, gives an error message.
# However . . .
echo '\' # Results in \</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
></DL
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The behavior of <SPAN
CLASS="TOKEN"
>\</SPAN
> depends on whether
it is escaped, <A
HREF="#SNGLQUO"
>strong-quoted</A
>,
<A
HREF="#DBLQUO"
>weak-quoted</A
>, or appearing within
<A
HREF="#COMMANDSUBREF"
>command substitution</A
> or a
<A
HREF="#HEREDOCREF"
>here document</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> # Simple escaping and quoting
echo \z # z
echo \\z # \z
echo '\z' # \z
echo '\\z' # \\z
echo "\z" # \z
echo "\\z" # \z
# Command substitution
echo `echo \z` # z
echo `echo \\z` # z
echo `echo \\\z` # \z
echo `echo \\\\z` # \z
echo `echo \\\\\\z` # \z
echo `echo \\\\\\\z` # \\z
echo `echo "\z"` # \z
echo `echo "\\z"` # \z
# Here document
cat &#60;&#60;EOF
\z
EOF # \z
cat &#60;&#60;EOF
\\z
EOF # \z
# These examples supplied by St<53>phane Chazelas.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Elements of a string assigned to a variable may be escaped, but
the escape character alone may not be assigned to a variable.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>variable=\
echo "$variable"
# Will not work - gives an error message:
# test.sh: : command not found
# A "naked" escape cannot safely be assigned to a variable.
#
# What actually happens here is that the "\" escapes the newline and
#+ the effect is variable=echo "$variable"
#+ invalid variable assignment
variable=\
23skidoo
echo "$variable" # 23skidoo
# This works, since the second line
#+ is a valid variable assignment.
variable=\
# \^ escape followed by space
echo "$variable" # space
variable=\\
echo "$variable" # \
variable=\\\
echo "$variable"
# Will not work - gives an error message:
# test.sh: \: command not found
#
# First escape escapes second one, but the third one is left "naked",
#+ with same result as first instance, above.
variable=\\\\
echo "$variable" # \\
# Second and fourth escapes escaped.
# This is o.k.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><P
>Escaping a space can prevent word splitting in a command's argument list.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>file_list="/bin/cat /bin/gzip /bin/more /usr/bin/less /usr/bin/emacs-20.7"
# List of files as argument(s) to a command.
# Add two files to the list, and list all.
ls -l /usr/X11R6/bin/xsetroot /sbin/dump $file_list
echo "-------------------------------------------------------------------------"
# What happens if we escape a couple of spaces?
ls -l /usr/X11R6/bin/xsetroot\ /sbin/dump\ $file_list
# Error: the first three files concatenated into a single argument to 'ls -l'
# because the two escaped spaces prevent argument (word) splitting.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="ESCNEWLINE"
></A
></P
><P
>The <SPAN
CLASS="TOKEN"
>escape</SPAN
> also provides a means of writing a
multi-line command. Normally, each separate line constitutes
a different command, but an <SPAN
CLASS="TOKEN"
>escape</SPAN
> at the end
of a line <EM
>escapes the newline character</EM
>,
and the command sequence continues on to the next line.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>(cd /source/directory &#38;&#38; tar cf - . ) | \
(cd /dest/directory &#38;&#38; tar xpvf -)
# Repeating Alan Cox's directory tree copy command,
# but split into two lines for increased legibility.
# As an alternative:
tar cf - -C /source/directory . |
tar xpvf - -C /dest/directory
# See note below.
# (Thanks, St<53>phane Chazelas.)</PRE
></FONT
></TD
></TR
></TABLE
>
<DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>If a script line ends with a <SPAN
CLASS="TOKEN"
>|</SPAN
>, a pipe
character, then a <SPAN
CLASS="TOKEN"
>\</SPAN
>, an escape, is not strictly
necessary. It is, however, good programming practice to always
escape the end of a line of code that continues to the
following line.</P
></TD
></TR
></TABLE
></DIV
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo "foo
bar"
#foo
#bar
echo
echo 'foo
bar' # No difference yet.
#foo
#bar
echo
echo foo\
bar # Newline escaped.
#foobar
echo
echo "foo\
bar" # Same here, as \ still interpreted as escape within weak quotes.
#foobar
echo
echo 'foo\
bar' # Escape character \ taken literally because of strong quoting.
#foo\
#bar
# Examples suggested by St<53>phane Chazelas.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="EXIT-STATUS"
></A
>Chapter 6. Exit and Exit Status</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>... there are dark corners in the Bourne shell, and people use all
of them.</I
></P
><P
><I
>--Chet Ramey</I
></P
></I
></TD
></TR
></TABLE
><P
><A
NAME="EXITCOMMANDREF"
></A
>The
<B
CLASS="COMMAND"
>
exit
</B
>
command terminates a script, just as in a <B
CLASS="COMMAND"
>C</B
>
program. It can also return a value, which is available to the
script's parent process.</P
><P
><A
NAME="EXITSTATUSREF"
></A
>Every command returns an
<I
CLASS="FIRSTTERM"
>
exit status
</I
>
(sometimes referred to as a
<I
CLASS="FIRSTTERM"
>
return status
</I
> or <I
CLASS="FIRSTTERM"
>exit code</I
>).
<A
NAME="EXITSUCCESS"
></A
>
A successful command returns a <SPAN
CLASS="RETURNVALUE"
>0</SPAN
>, while
an unsuccessful one returns a <SPAN
CLASS="RETURNVALUE"
>non-zero</SPAN
>
value that usually can be interpreted as an <I
CLASS="FIRSTTERM"
>error
code</I
>. Well-behaved UNIX commands, programs, and
utilities return a <SPAN
CLASS="RETURNVALUE"
>0</SPAN
> exit code upon
successful completion, though there are some exceptions.</P
><P
><A
NAME="FUNCTXSTR"
></A
></P
><P
>Likewise, <A
HREF="#FUNCTIONREF"
>functions</A
>
within a script and the script itself return an exit
status. The last command executed in the function or
script determines the exit status. Within a script, an
<TT
CLASS="USERINPUT"
><B
>exit <TT
CLASS="REPLACEABLE"
><I
>nnn</I
></TT
></B
></TT
>
command may be used to deliver an
<SPAN
CLASS="RETURNVALUE"
><TT
CLASS="REPLACEABLE"
><I
>nnn</I
></TT
></SPAN
>
exit status to the shell
(<SPAN
CLASS="RETURNVALUE"
><TT
CLASS="REPLACEABLE"
><I
>nnn</I
></TT
></SPAN
>
must be an integer in the <SPAN
CLASS="RETURNVALUE"
>0</SPAN
> -
<SPAN
CLASS="RETURNVALUE"
>255</SPAN
> range).</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>When a script ends with an <B
CLASS="COMMAND"
>exit</B
> that has
no parameter, the exit status of the script is the exit status
of the last command executed in the script (previous to the
<B
CLASS="COMMAND"
>exit</B
>).</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
COMMAND_1
. . .
COMMAND_LAST
# Will exit with status of last command.
exit</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The equivalent of a bare <B
CLASS="COMMAND"
>exit</B
> is
<B
CLASS="COMMAND"
>exit $?</B
> or even just omitting the
<B
CLASS="COMMAND"
>exit</B
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
COMMAND_1
. . .
COMMAND_LAST
# Will exit with status of last command.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
COMMAND1
. . .
COMMAND_LAST
# Will exit with status of last command.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="EXSREF"
></A
></P
><P
> <TT
CLASS="VARNAME"
> $?</TT
> reads the exit status of the last
command executed. After a function returns,
<TT
CLASS="VARNAME"
>$?</TT
> gives the exit status of the last
command executed in the function. This is Bash's way of giving
functions a <SPAN
CLASS="QUOTE"
>"return value."</SPAN
>
<A
NAME="AEN2981"
HREF="#FTN.AEN2981"
><SPAN
CLASS="footnote"
>[32]</SPAN
></A
>
</P
><P
><A
NAME="PIPEEX"
></A
>Following the execution of a <A
HREF="#PIPEREF"
>pipe</A
>, a <TT
CLASS="VARNAME"
>$?</TT
>
gives the exit status of the last command executed.</P
><P
>After a script terminates, a <TT
CLASS="VARNAME"
>$?</TT
> from the
command-line gives the exit status of the script, that is, the
last command executed in the script, which is, by convention,
<TT
CLASS="USERINPUT"
><B
>0</B
></TT
> on success or an integer in the
range <SPAN
CLASS="RETURNVALUE"
>1 - 255</SPAN
> on error.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX5"
></A
><P
><B
>Example 6-1. exit / exit status</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo hello
echo $? # Exit status 0 returned because command executed successfully.
lskdf # Unrecognized command.
echo $? # Non-zero exit status returned -- command failed to execute.
echo
exit 113 # Will return 113 to shell.
# To verify this, type "echo $?" after script terminates.
# By convention, an 'exit 0' indicates success,
#+ while a non-zero exit value means an error or anomalous condition.
# See the "Exit Codes With Special Meanings" appendix.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
HREF="#XSTATVARREF"
>$?</A
> is especially useful
for testing the result of a command in a script (see <A
HREF="#FILECOMP"
>Example 16-35</A
> and <A
HREF="#LOOKUP"
>Example 16-20</A
>).</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <A
HREF="#NOTREF"
>!</A
>, the <I
CLASS="FIRSTTERM"
>logical
not</I
> qualifier, reverses the outcome of a test or
command, and this affects its <A
HREF="#EXITSTATUSREF"
>exit
status</A
>.
<DIV
CLASS="EXAMPLE"
><A
NAME="NEGCOND"
></A
><P
><B
>Example 6-2. Negating a condition using <SPAN
CLASS="TOKEN"
>!</SPAN
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>true # The "true" builtin.
echo "exit status of \"true\" = $?" # 0
! true
echo "exit status of \"! true\" = $?" # 1
# Note that the "!" needs a space between it and the command.
# !true leads to a "command not found" error
#
# The '!' operator prefixing a command invokes the Bash history mechanism.
true
!true
# No error this time, but no negation either.
# It just repeats the previous command (true).
# =========================================================== #
# Preceding a _pipe_ with ! inverts the exit status returned.
ls | bogus_command # bash: bogus_command: command not found
echo $? # 127
! ls | bogus_command # bash: bogus_command: command not found
echo $? # 0
# Note that the ! does not change the execution of the pipe.
# Only the exit status changes.
# =========================================================== #
# Thanks, St<53>phane Chazelas and Kristopher Newsome.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Certain exit status codes have <A
HREF="#EXITCODESREF"
>reserved meanings</A
> and should not
be user-specified in a script. </P
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="TESTS"
></A
>Chapter 7. Tests</H1
><P
><A
NAME="IFTHEN"
></A
></P
><P
>Every reasonably complete programming language can test
for a condition, then act according to the result of the
test. Bash has the <A
HREF="#TTESTREF"
>test</A
>
command, various <A
HREF="#DBLBRACKETS"
>bracket</A
>
and <A
HREF="#DBLPARENSTST"
>parenthesis</A
> operators,
and the <B
CLASS="COMMAND"
>if/then</B
> construct.</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="TESTCONSTRUCTS"
></A
>7.1. Test Constructs</H1
><P
><A
NAME="TESTCONSTRUCTS1"
></A
></P
><P
></P
><UL
><LI
><P
>An <B
CLASS="COMMAND"
>if/then</B
> construct tests whether the
<A
HREF="#EXITSTATUSREF"
>exit status</A
> of a list
of commands is <SPAN
CLASS="RETURNVALUE"
>0</SPAN
> (since 0 means
<SPAN
CLASS="QUOTE"
>"success"</SPAN
> by UNIX convention), and if so, executes
one or more commands.</P
></LI
><LI
><P
>There exists a dedicated command called <B
CLASS="COMMAND"
> [</B
> (<A
HREF="#LEFTBRACKET"
>left bracket</A
>
special character). It is a synonym for <B
CLASS="COMMAND"
>test</B
>,
and a <A
HREF="#BUILTINREF"
>builtin</A
> for efficiency
reasons. This command considers its arguments as comparison
expressions or file tests and returns an exit status corresponding
to the result of the comparison (0 for true, 1 for false).</P
></LI
><LI
><P
>With version 2.02, Bash introduced the <A
HREF="#DBLBRACKETS"
>[[ ... ]]</A
> <I
CLASS="FIRSTTERM"
>extended
test command</I
>, which performs comparisons
in a manner more familiar to programmers from other
languages. Note that <B
CLASS="COMMAND"
>[[</B
> is a <A
HREF="#KEYWORDREF"
>keyword</A
>, not a command.</P
><P
>Bash sees <TT
CLASS="USERINPUT"
><B
>[[ $a -lt $b ]]</B
></TT
> as a
single element, which returns an exit status.</P
></LI
><LI
><P
><A
NAME="DBLPARENSTST"
></A
></P
><P
>The <A
HREF="#DBLPARENS"
>(( ... ))</A
> and <A
HREF="#LETREF"
>let ...</A
> constructs return an
<A
HREF="#EXITSTATUSREF"
>exit status</A
>,
<EM
>according to whether the arithmetic expressions they
evaluate expand to a non-zero value</EM
>. These
<A
HREF="#ARITHEXPREF"
>arithmetic-expansion</A
>
constructs may therefore be used to perform <A
HREF="#ICOMPARISON1"
>arithmetic comparisons</A
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>(( 0 &#38;&#38; 1 )) # Logical AND
echo $? # 1 ***
# And so ...
let "num = (( 0 &#38;&#38; 1 ))"
echo $num # 0
# But ...
let "num = (( 0 &#38;&#38; 1 ))"
echo $? # 1 ***
(( 200 || 11 )) # Logical OR
echo $? # 0 ***
# ...
let "num = (( 200 || 11 ))"
echo $num # 1
let "num = (( 200 || 11 ))"
echo $? # 0 ***
(( 200 | 11 )) # Bitwise OR
echo $? # 0 ***
# ...
let "num = (( 200 | 11 ))"
echo $num # 203
let "num = (( 200 | 11 ))"
echo $? # 0 ***
# The "let" construct returns the same exit status
#+ as the double-parentheses arithmetic expansion.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="ARXS"
></A
>Again, note that the
<I
CLASS="FIRSTTERM"
>exit status</I
> of an arithmetic expression
is <EM
>not</EM
> an error value.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>var=-2 &#38;&#38; (( var+=2 ))
echo $? # 1
var=-2 &#38;&#38; (( var+=2 )) &#38;&#38; echo $var
# Will not echo $var!</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><A
NAME="IFGREPREF"
></A
></P
><P
>An <B
CLASS="COMMAND"
>if</B
> can test any command, not just
conditions enclosed within brackets.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if cmp a b &#38;&#62; /dev/null # Suppress output.
then echo "Files a and b are identical."
else echo "Files a and b differ."
fi
# The very useful "if-grep" construct:
# -----------------------------------
if grep -q Bash file
then echo "File contains at least one occurrence of Bash."
fi
word=Linux
letter_sequence=inu
if echo "$word" | grep -q "$letter_sequence"
# The "-q" option to grep suppresses output.
then
echo "$letter_sequence found in $word"
else
echo "$letter_sequence not found in $word"
fi
if COMMAND_WHOSE_EXIT_STATUS_IS_0_UNLESS_ERROR_OCCURRED
then echo "Command succeeded."
else echo "Command failed."
fi</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><EM
>These last two examples
courtesy of St<53>phane Chazelas.</EM
></P
></LI
></UL
><DIV
CLASS="EXAMPLE"
><A
NAME="EX10"
></A
><P
><B
>Example 7-1. What is truth?</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Tip:
# If you're unsure how a certain condition might evaluate,
#+ test it in an if-test.
echo
echo "Testing \"0\""
if [ 0 ] # zero
then
echo "0 is true."
else # Or else ...
echo "0 is false."
fi # 0 is true.
echo
echo "Testing \"1\""
if [ 1 ] # one
then
echo "1 is true."
else
echo "1 is false."
fi # 1 is true.
echo
echo "Testing \"-1\""
if [ -1 ] # minus one
then
echo "-1 is true."
else
echo "-1 is false."
fi # -1 is true.
echo
echo "Testing \"NULL\""
if [ ] # NULL (empty condition)
then
echo "NULL is true."
else
echo "NULL is false."
fi # NULL is false.
echo
echo "Testing \"xyz\""
if [ xyz ] # string
then
echo "Random string is true."
else
echo "Random string is false."
fi # Random string is true.
echo
echo "Testing \"\$xyz\""
if [ $xyz ] # Tests if $xyz is null, but...
# it's only an uninitialized variable.
then
echo "Uninitialized variable is true."
else
echo "Uninitialized variable is false."
fi # Uninitialized variable is false.
echo
echo "Testing \"-n \$xyz\""
if [ -n "$xyz" ] # More pedantically correct.
then
echo "Uninitialized variable is true."
else
echo "Uninitialized variable is false."
fi # Uninitialized variable is false.
echo
xyz= # Initialized, but set to null value.
echo "Testing \"-n \$xyz\""
if [ -n "$xyz" ]
then
echo "Null variable is true."
else
echo "Null variable is false."
fi # Null variable is false.
echo
# When is "false" true?
echo "Testing \"false\""
if [ "false" ] # It seems that "false" is just a string ...
then
echo "\"false\" is true." #+ and it tests true.
else
echo "\"false\" is false."
fi # "false" is true.
echo
echo "Testing \"\$false\"" # Again, uninitialized variable.
if [ "$false" ]
then
echo "\"\$false\" is true."
else
echo "\"\$false\" is false."
fi # "$false" is false.
# Now, we get the expected result.
# What would happen if we tested the uninitialized variable "$true"?
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="FORMALPARA"
><P
><B
>Exercise. </B
>Explain the behavior of <A
HREF="#EX10"
>Example 7-1</A
>, above.</P
></DIV
><P
><A
NAME="ELSEREF"
></A
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ condition-true ]
then
command 1
command 2
...
else # Or else ...
# Adds default code block executing if original condition tests false.
command 3
command 4
...
fi</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>When <I
CLASS="FIRSTTERM"
>if</I
> and <I
CLASS="FIRSTTERM"
>then</I
>
are on same line in a condition test, a semicolon must
terminate the <I
CLASS="FIRSTTERM"
>if</I
> statement. Both
<I
CLASS="FIRSTTERM"
>if</I
> and <I
CLASS="FIRSTTERM"
>then</I
>
are <A
HREF="#KEYWORDREF"
>keywords</A
>. Keywords (or
commands) begin statements, and before a new statement on the
same line begins, the old one must terminate.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ -x "$filename" ]; then</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="ELIFREF1"
></A
>Else if and elif</B
></P
><DL
><DT
><SPAN
CLASS="TOKEN"
>elif</SPAN
></DT
><DD
><P
><TT
CLASS="USERINPUT"
><B
>elif</B
></TT
> is a contraction
for <I
CLASS="FIRSTTERM"
>else if</I
>. The effect is to nest an
inner <SPAN
CLASS="TOKEN"
>if/then</SPAN
> construct within an outer
one.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ condition1 ]
then
command1
command2
command3
elif [ condition2 ]
# Same as else if
then
command4
command5
else
default-command
fi</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
></DL
></DIV
><P
>
<A
NAME="IFREF2"
></A
>
The <TT
CLASS="USERINPUT"
><B
>if test condition-true</B
></TT
> construct is the
exact equivalent of <TT
CLASS="USERINPUT"
><B
>if [ condition-true ]</B
></TT
>.
As it happens, the left bracket, <B
CLASS="COMMAND"
>[</B
> , is a
<I
CLASS="FIRSTTERM"
>token</I
>
<A
NAME="AEN3140"
HREF="#FTN.AEN3140"
><SPAN
CLASS="footnote"
>[33]</SPAN
></A
>
which invokes the <B
CLASS="COMMAND"
>test</B
> command. The closing
right bracket, <B
CLASS="COMMAND"
>]</B
> , in an if/test should not
therefore be strictly necessary, however newer versions of Bash
require it.</P
><P
><A
NAME="TTESTREF"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>test</B
> command is a Bash <A
HREF="#BUILTINREF"
>builtin</A
> which tests file
types and compares strings. Therefore, in a Bash script,
<B
CLASS="COMMAND"
>test</B
> does <EM
>not</EM
> call
the external <TT
CLASS="FILENAME"
>/usr/bin/test</TT
> binary,
which is part of the <I
CLASS="FIRSTTERM"
>sh-utils</I
>
package. Likewise, <B
CLASS="COMMAND"
>[</B
> does not call
<TT
CLASS="FILENAME"
>/usr/bin/[</TT
>, which is linked to
<TT
CLASS="FILENAME"
>/usr/bin/test</TT
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>type test</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>test is a shell builtin</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>type '['</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>[ is a shell builtin</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>type '[['</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>[[ is a shell keyword</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>type ']]'</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>]] is a shell keyword</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>type ']'</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bash: type: ]: not found</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="USRBINTEST"
></A
></P
><P
>If, for some reason, you wish to use
<TT
CLASS="FILENAME"
>/usr/bin/test</TT
> in a Bash script,
then specify it by full pathname.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX11"
></A
><P
><B
>Example 7-2. Equivalence of <I
CLASS="FIRSTTERM"
>test</I
>,
<TT
CLASS="FILENAME"
>/usr/bin/test</TT
>, <SPAN
CLASS="TOKEN"
>[ ]</SPAN
>,
and <TT
CLASS="FILENAME"
>/usr/bin/[</TT
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo
if test -z "$1"
then
echo "No command-line arguments."
else
echo "First command-line argument is $1."
fi
echo
if /usr/bin/test -z "$1" # Equivalent to "test" builtin.
# ^^^^^^^^^^^^^ # Specifying full pathname.
then
echo "No command-line arguments."
else
echo "First command-line argument is $1."
fi
echo
if [ -z "$1" ] # Functionally identical to above code blocks.
# if [ -z "$1" should work, but...
#+ Bash responds to a missing close-bracket with an error message.
then
echo "No command-line arguments."
else
echo "First command-line argument is $1."
fi
echo
if /usr/bin/[ -z "$1" ] # Again, functionally identical to above.
# if /usr/bin/[ -z "$1" # Works, but gives an error message.
# # Note:
# This has been fixed in Bash, version 3.x.
then
echo "No command-line arguments."
else
echo "First command-line argument is $1."
fi
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN3206"
></A
><P
></P
><P
><A
NAME="DBLBRACKETS"
></A
>The <SPAN
CLASS="TOKEN"
>[[ ]]</SPAN
> construct
is the more versatile Bash version of <SPAN
CLASS="TOKEN"
>[ ]</SPAN
>. This
is the <I
CLASS="FIRSTTERM"
>extended test command</I
>, adopted from
<I
CLASS="FIRSTTERM"
>ksh88</I
>.</P
><P
>* * *</P
><P
>No filename expansion or word splitting takes place
between <SPAN
CLASS="TOKEN"
>[[</SPAN
> and <SPAN
CLASS="TOKEN"
>]]</SPAN
>, but there is
parameter expansion and command substitution.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>file=/etc/passwd
if [[ -e $file ]]
then
echo "Password file exists."
fi</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Using the <B
CLASS="COMMAND"
>[[ ... ]]</B
> test construct,
rather than <B
CLASS="COMMAND"
>[ ... ]</B
> can prevent many
logic errors in scripts. For example, the <SPAN
CLASS="TOKEN"
>&#38;&#38;</SPAN
>,
<SPAN
CLASS="TOKEN"
>||</SPAN
>, <SPAN
CLASS="TOKEN"
>&#60;</SPAN
>, and <SPAN
CLASS="TOKEN"
>&#62;</SPAN
>
operators work within a <SPAN
CLASS="TOKEN"
>[[ ]]</SPAN
> test, despite
giving an error within a <SPAN
CLASS="TOKEN"
>[ ]</SPAN
> construct.</P
><P
><A
NAME="DBLBRAEV"
></A
></P
><P
><I
CLASS="FIRSTTERM"
>Arithmetic evaluation</I
> of octal /
hexadecimal constants takes place automatically within a
<SPAN
CLASS="TOKEN"
>[[ ... ]]</SPAN
> construct.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># [[ Octal and hexadecimal evaluation ]]
# Thank you, Moritz Gronbach, for pointing this out.
decimal=15
octal=017 # = 15 (decimal)
hex=0x0f # = 15 (decimal)
if [ "$decimal" -eq "$octal" ]
then
echo "$decimal equals $octal"
else
echo "$decimal is not equal to $octal" # 15 is not equal to 017
fi # Doesn't evaluate within [ single brackets ]!
if [[ "$decimal" -eq "$octal" ]]
then
echo "$decimal equals $octal" # 15 equals 017
else
echo "$decimal is not equal to $octal"
fi # Evaluates within [[ double brackets ]]!
if [[ "$decimal" -eq "$hex" ]]
then
echo "$decimal equals $hex" # 15 equals 0x0f
else
echo "$decimal is not equal to $hex"
fi # [[ $hexadecimal ]] also evaluates!</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
></P
></DIV
></TD
></TR
></TABLE
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Following an <B
CLASS="COMMAND"
>if</B
>, neither the
<B
CLASS="COMMAND"
>test</B
> command nor the test brackets ( [ ] or [[ ]] )
are strictly necessary.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>dir=/home/bozo
if cd "$dir" 2&#62;/dev/null; then # "2&#62;/dev/null" hides error message.
echo "Now in $dir."
else
echo "Can't change to $dir."
fi</PRE
></FONT
></TD
></TR
></TABLE
>
The "if COMMAND" construct returns the exit status of COMMAND.</P
><P
>Similarly, a condition within test brackets may stand alone
without an <B
CLASS="COMMAND"
>if</B
>, when used in combination
with a <A
HREF="#LISTCONSREF"
>list construct</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>var1=20
var2=22
[ "$var1" -ne "$var2" ] &#38;&#38; echo "$var1 is not equal to $var2"
home=/home/bozo
[ -d "$home" ] || echo "$home directory does not exist."</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="DBLPRX"
></A
>The <A
HREF="#DBLPARENS"
>(( ))
construct</A
> expands and evaluates an arithmetic
expression. If the expression evaluates as zero, it returns
an <A
HREF="#EXITSTATUSREF"
>exit status</A
> of
<SPAN
CLASS="RETURNVALUE"
>1</SPAN
>, or <SPAN
CLASS="QUOTE"
>"false"</SPAN
>. A non-zero
expression returns an exit status of <SPAN
CLASS="RETURNVALUE"
>0</SPAN
>,
or <SPAN
CLASS="QUOTE"
>"true"</SPAN
>. This is in marked contrast to using
the <B
CLASS="COMMAND"
>test</B
> and <SPAN
CLASS="TOKEN"
>[ ]</SPAN
> constructs
previously discussed.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ARITHTESTS"
></A
><P
><B
>Example 7-3. Arithmetic Tests using <SPAN
CLASS="TOKEN"
>(( ))</SPAN
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# arith-tests.sh
# Arithmetic tests.
# The (( ... )) construct evaluates and tests numerical expressions.
# Exit status opposite from [ ... ] construct!
(( 0 ))
echo "Exit status of \"(( 0 ))\" is $?." # 1
(( 1 ))
echo "Exit status of \"(( 1 ))\" is $?." # 0
(( 5 &#62; 4 )) # true
echo "Exit status of \"(( 5 &#62; 4 ))\" is $?." # 0
(( 5 &#62; 9 )) # false
echo "Exit status of \"(( 5 &#62; 9 ))\" is $?." # 1
(( 5 == 5 )) # true
echo "Exit status of \"(( 5 == 5 ))\" is $?." # 0
# (( 5 = 5 )) gives an error message.
(( 5 - 5 )) # 0
echo "Exit status of \"(( 5 - 5 ))\" is $?." # 1
(( 5 / 4 )) # Division o.k.
echo "Exit status of \"(( 5 / 4 ))\" is $?." # 0
(( 1 / 2 )) # Division result &#60; 1.
echo "Exit status of \"(( 1 / 2 ))\" is $?." # Rounded off to 0.
# 1
(( 1 / 0 )) 2&#62;/dev/null # Illegal division by 0.
# ^^^^^^^^^^^
echo "Exit status of \"(( 1 / 0 ))\" is $?." # 1
# What effect does the "2&#62;/dev/null" have?
# What would happen if it were removed?
# Try removing it, then rerunning the script.
# ======================================= #
# (( ... )) also useful in an if-then test.
var1=5
var2=4
if (( var1 &#62; var2 ))
then #^ ^ Note: Not $var1, $var2. Why?
echo "$var1 is greater than $var2"
fi # 5 is greater than 4
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="FTO"
></A
>7.2. File test operators</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="RTIF"
></A
>Returns true if...</B
></P
><DL
><DT
><SPAN
CLASS="TOKEN"
>-e</SPAN
></DT
><DD
><P
>file exists</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-a</SPAN
></DT
><DD
><P
>file exists</P
><P
>This is identical in effect to <SPAN
CLASS="TOKEN"
>-e</SPAN
>.
It has been <SPAN
CLASS="QUOTE"
>"deprecated,"</SPAN
>
<A
NAME="AEN3289"
HREF="#FTN.AEN3289"
><SPAN
CLASS="footnote"
>[34]</SPAN
></A
>
and its use is
discouraged.</P
></DD
><DT
><A
NAME="REGULARFILE"
></A
><SPAN
CLASS="TOKEN"
>-f</SPAN
></DT
><DD
><P
>file is a <TT
CLASS="REPLACEABLE"
><I
>regular</I
></TT
>
file (not a directory or <A
HREF="#DEVFILEREF"
>device
file</A
>)</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-s</SPAN
></DT
><DD
><P
>file is not zero size</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-d</SPAN
></DT
><DD
><P
>file is a directory</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-b</SPAN
></DT
><DD
><P
>file is a <A
HREF="#BLOCKDEVREF"
>block
device</A
></P
><P
><A
NAME="BLOCKDEVTEST"
></A
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-c</SPAN
></DT
><DD
><P
><A
NAME="CHARDEVTEST"
></A
>file is a <A
HREF="#CHARDEVREF"
>character device</A
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>device0="/dev/sda2" # / (root directory)
if [ -b "$device0" ]
then
echo "$device0 is a block device."
fi
# /dev/sda2 is a block device.
device1="/dev/ttyS1" # PCMCIA modem card.
if [ -c "$device1" ]
then
echo "$device1 is a character device."
fi
# /dev/ttyS1 is a character device.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-p</SPAN
></DT
><DD
><P
>file is a <A
HREF="#PIPEREF"
>pipe</A
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>function show_input_type()
{
[ -p /dev/fd/0 ] &#38;&#38; echo PIPE || echo STDIN
}
show_input_type "Input" # STDIN
echo "Input" | show_input_type # PIPE
# This example courtesy of Carl Anderson.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-h</SPAN
></DT
><DD
><P
>file is a <A
HREF="#SYMLINKREF"
>symbolic
link</A
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-L</SPAN
></DT
><DD
><P
>file is a symbolic link</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-S</SPAN
></DT
><DD
><P
>file is a <A
HREF="#SOCKETREF"
>socket</A
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-t</SPAN
></DT
><DD
><P
><A
NAME="TERMTEST"
></A
>file (<A
HREF="#FDREF"
>descriptor</A
>) is
associated with a terminal device</P
><P
>This test option <A
HREF="#II2TEST"
> may be used
to check</A
> whether the <TT
CLASS="FILENAME"
>stdin</TT
>
<TT
CLASS="USERINPUT"
><B
>[ -t 0 ]</B
></TT
> or
<TT
CLASS="FILENAME"
>stdout</TT
> <TT
CLASS="USERINPUT"
><B
>[ -t 1 ]</B
></TT
>
in a given script is a terminal.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-r</SPAN
></DT
><DD
><P
>file has read permission (<EM
>for the
user running the test</EM
>)</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-w</SPAN
></DT
><DD
><P
>file has write permission (for the user running
the test)</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-x</SPAN
></DT
><DD
><P
>file has execute permission (for the user running
the test)</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-g</SPAN
></DT
><DD
><P
>set-group-id (sgid) flag set on file or directory</P
><P
>If a directory has the <TT
CLASS="REPLACEABLE"
><I
>sgid</I
></TT
>
flag set, then a file created within that directory belongs
to the group that owns the directory, not necessarily to
the group of the user who created the file. This may be
useful for a directory shared by a workgroup.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-u</SPAN
></DT
><DD
><P
><A
NAME="SUIDREF"
></A
></P
><P
>set-user-id (suid) flag set on file</P
><P
>A binary owned by <I
CLASS="FIRSTTERM"
>root</I
>
with <TT
CLASS="REPLACEABLE"
><I
>set-user-id</I
></TT
> flag set
runs with <I
CLASS="FIRSTTERM"
>root</I
> privileges, even
when an ordinary user invokes it.
<A
NAME="AEN3400"
HREF="#FTN.AEN3400"
><SPAN
CLASS="footnote"
>[35]</SPAN
></A
>
This is useful for executables (such as
<B
CLASS="COMMAND"
>pppd</B
> and <B
CLASS="COMMAND"
>cdrecord</B
>)
that need to access system hardware. Lacking the
<I
CLASS="FIRSTTERM"
>suid</I
> flag, these binaries could not
be invoked by a <I
CLASS="FIRSTTERM"
>non-root</I
> user.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
> <TT
CLASS="COMPUTEROUTPUT"
>-rwsr-xr-t 1 root 178236 Oct 2 2000 /usr/sbin/pppd</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>A file with the <TT
CLASS="REPLACEABLE"
><I
>suid</I
></TT
>
flag set shows an <I
CLASS="FIRSTTERM"
>s</I
> in its
permissions.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-k</SPAN
></DT
><DD
><P
><TT
CLASS="REPLACEABLE"
><I
>sticky bit</I
></TT
> set</P
><P
>Commonly known as the <I
CLASS="FIRSTTERM"
>sticky bit,</I
>
the <I
CLASS="FIRSTTERM"
>save-text-mode</I
> flag is a special
type of file permission. If a file has this flag set,
that file will be kept in cache memory, for quicker access.
<A
NAME="AEN3423"
HREF="#FTN.AEN3423"
><SPAN
CLASS="footnote"
>[36]</SPAN
></A
>
If set on a directory, it restricts write permission.
Setting the sticky bit adds a <I
CLASS="FIRSTTERM"
>t</I
>
to the permissions on the file or directory listing.
This restricts altering or deleting specific files
in that directory to the owner of those files.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
> <TT
CLASS="COMPUTEROUTPUT"
>drwxrwxrwt 7 root 1024 May 19 21:26 tmp/</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>If a user does not own a directory that has the sticky
bit set, but has write permission in that directory, she
can only delete those files that she owns in it. This
keeps users from inadvertently overwriting or deleting
each other's files in a publicly accessible directory,
such as <TT
CLASS="FILENAME"
>/tmp</TT
>.
(The <I
CLASS="FIRSTTERM"
>owner</I
> of the directory or
<I
CLASS="FIRSTTERM"
>root</I
> can, of course, delete or
rename files there.)</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-O</SPAN
></DT
><DD
><P
>you are owner of file</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-G</SPAN
></DT
><DD
><P
>group-id of file same as yours</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-N</SPAN
></DT
><DD
><P
>file modified since it was last read</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>f1 -nt f2</SPAN
></DT
><DD
><P
>file <TT
CLASS="REPLACEABLE"
><I
>f1</I
></TT
> is newer than
<TT
CLASS="REPLACEABLE"
><I
>f2</I
></TT
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>f1 -ot f2</SPAN
></DT
><DD
><P
>file <TT
CLASS="REPLACEABLE"
><I
>f1</I
></TT
> is older than
<TT
CLASS="REPLACEABLE"
><I
>f2</I
></TT
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>f1 -ef f2</SPAN
></DT
><DD
><P
>files <TT
CLASS="REPLACEABLE"
><I
>f1</I
></TT
> and
<TT
CLASS="REPLACEABLE"
><I
>f2</I
></TT
> are hard links to the same
file</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>!</SPAN
></DT
><DD
><P
><SPAN
CLASS="QUOTE"
>"not"</SPAN
> -- reverses the sense of the
tests above (returns true if condition absent).</P
></DD
></DL
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="BROKENLINK"
></A
><P
><B
>Example 7-4. Testing for broken links</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# broken-link.sh
# Written by Lee bigelow &#60;ligelowbee@yahoo.com&#62;
# Used in ABS Guide with permission.
# A pure shell script to find dead symlinks and output them quoted
#+ so they can be fed to xargs and dealt with :)
#+ eg. sh broken-link.sh /somedir /someotherdir|xargs rm
#
# This, however, is a better method:
#
# find "somedir" -type l -print0|\
# xargs -r0 file|\
# grep "broken symbolic"|
# sed -e 's/^\|: *broken symbolic.*$/"/g'
#
#+ but that wouldn't be pure Bash, now would it.
# Caution: beware the /proc file system and any circular links!
################################################################
# If no args are passed to the script set directories-to-search
#+ to current directory. Otherwise set the directories-to-search
#+ to the args passed.
######################
[ $# -eq 0 ] &#38;&#38; directorys=`pwd` || directorys=$@
# Setup the function linkchk to check the directory it is passed
#+ for files that are links and don't exist, then print them quoted.
# If one of the elements in the directory is a subdirectory then
#+ send that subdirectory to the linkcheck function.
##########
linkchk () {
for element in $1/*; do
[ -h "$element" -a ! -e "$element" ] &#38;&#38; echo \"$element\"
[ -d "$element" ] &#38;&#38; linkchk $element
# Of course, '-h' tests for symbolic link, '-d' for directory.
done
}
# Send each arg that was passed to the script to the linkchk() function
#+ if it is a valid directoy. If not, then print the error message
#+ and usage info.
##################
for directory in $directorys; do
if [ -d $directory ]
then linkchk $directory
else
echo "$directory is not a directory"
echo "Usage: $0 dir1 dir2 ..."
fi
done
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
HREF="#COOKIES"
>Example 31-1</A
>, <A
HREF="#BINGREP"
>Example 11-8</A
>,
<A
HREF="#FILEINFO"
>Example 11-3</A
>, <A
HREF="#RAMDISK"
>Example 31-3</A
>, and <A
HREF="#MAILFORMAT"
>Example A-1</A
> also illustrate uses of the file test
operators.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="COMPARISON-OPS"
></A
>7.3. Other Comparison Operators</H1
><P
>A <I
CLASS="FIRSTTERM"
>binary</I
> comparison operator
compares two variables or quantities. <EM
>Note
that integer and string comparison use a different set of
operators.</EM
></P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="ICOMPARISON1"
></A
>integer comparison</B
></P
><DL
><DT
><A
NAME="EQUALREF"
></A
><SPAN
CLASS="TOKEN"
>-eq</SPAN
></DT
><DD
><P
>is equal to</P
><P
><TT
CLASS="USERINPUT"
><B
>if [ "$a" -eq "$b" ]</B
></TT
></P
></DD
><DT
><A
NAME="NEQUALREF"
></A
><SPAN
CLASS="TOKEN"
>-ne</SPAN
></DT
><DD
><P
>is not equal to</P
><P
><TT
CLASS="USERINPUT"
><B
>if [ "$a" -ne "$b" ]</B
></TT
></P
></DD
><DT
><A
NAME="GT0REF"
></A
><SPAN
CLASS="TOKEN"
>-gt</SPAN
></DT
><DD
><P
>is greater than</P
><P
><TT
CLASS="USERINPUT"
><B
>if [ "$a" -gt "$b" ]</B
></TT
></P
></DD
><DT
><A
NAME="GE0REF"
></A
><SPAN
CLASS="TOKEN"
>-ge</SPAN
></DT
><DD
><P
>is greater than or equal to</P
><P
><TT
CLASS="USERINPUT"
><B
>if [ "$a" -ge "$b" ]</B
></TT
></P
></DD
><DT
><A
NAME="LT0REF"
></A
><SPAN
CLASS="TOKEN"
>-lt</SPAN
></DT
><DD
><P
>is less than</P
><P
><TT
CLASS="USERINPUT"
><B
>if [ "$a" -lt "$b" ]</B
></TT
></P
></DD
><DT
><A
NAME="LE0REF"
></A
><SPAN
CLASS="TOKEN"
>-le</SPAN
></DT
><DD
><P
>is less than or equal to</P
><P
><TT
CLASS="USERINPUT"
><B
>if [ "$a" -le "$b" ]</B
></TT
></P
></DD
><DT
><A
NAME="INTLT"
></A
><SPAN
CLASS="TOKEN"
>&#60;</SPAN
></DT
><DD
><P
>is less than (within <A
HREF="#DBLPARENS"
>double
parentheses</A
>)</P
><P
><TT
CLASS="USERINPUT"
><B
>(("$a" &#60; "$b"))</B
></TT
></P
></DD
><DT
><A
NAME="LTEQ"
></A
><SPAN
CLASS="TOKEN"
>&#60;=</SPAN
></DT
><DD
><P
>is less than or equal to (within double parentheses)</P
><P
><TT
CLASS="USERINPUT"
><B
>(("$a" &#60;= "$b"))</B
></TT
></P
></DD
><DT
><A
NAME="INTGT"
></A
><SPAN
CLASS="TOKEN"
>&#62;</SPAN
></DT
><DD
><P
>is greater than (within double parentheses)</P
><P
><TT
CLASS="USERINPUT"
><B
>(("$a" &#62; "$b"))</B
></TT
></P
></DD
><DT
><A
NAME="GTEQ"
></A
><SPAN
CLASS="TOKEN"
>&#62;=</SPAN
></DT
><DD
><P
>is greater than or equal to (within double parentheses)</P
><P
><TT
CLASS="USERINPUT"
><B
>(("$a" &#62;= "$b"))</B
></TT
></P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="SCOMPARISON1"
></A
>string comparison</B
></P
><DL
><DT
><SPAN
CLASS="TOKEN"
>=</SPAN
></DT
><DD
><P
><A
NAME="EQUALSIGNREF"
></A
></P
><P
>is equal to</P
><P
><TT
CLASS="USERINPUT"
><B
>if [ "$a" = "$b" ]</B
></TT
></P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Note the <A
HREF="#WHITESPACEREF"
>whitespace</A
>
framing the <B
CLASS="COMMAND"
>=</B
>.</P
><P
><TT
CLASS="USERINPUT"
><B
>if [ "$a"="$b" ]</B
></TT
> is
<EM
>not</EM
> equivalent to the
above.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SCOMPARISON2"
></A
><SPAN
CLASS="TOKEN"
>==</SPAN
></DT
><DD
><P
>is equal to</P
><P
><TT
CLASS="USERINPUT"
><B
>if [ "$a" == "$b" ]</B
></TT
></P
><P
>This is a synonym for <SPAN
CLASS="TOKEN"
>=</SPAN
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
> The <SPAN
CLASS="TOKEN"
>==</SPAN
> comparison operator behaves differently
within a <A
HREF="#DBLBRACKETS"
>double-brackets</A
>
test than within single brackets.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>[[ $a == z* ]] # True if $a starts with an "z" (pattern matching).
[[ $a == "z*" ]] # True if $a is equal to z* (literal matching).
[ $a == z* ] # File globbing and word splitting take place.
[ "$a" == "z*" ] # True if $a is equal to z* (literal matching).
# Thanks, St<53>phane Chazelas</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="NOTEQUAL"
></A
><SPAN
CLASS="TOKEN"
>!=</SPAN
></DT
><DD
><P
>is not equal to</P
><P
><TT
CLASS="USERINPUT"
><B
>if [ "$a" != "$b" ]</B
></TT
></P
><P
>This operator uses pattern matching within a <A
HREF="#DBLBRACKETS"
>[[ ... ]]</A
> construct.</P
></DD
><DT
><A
NAME="LTREF"
></A
><SPAN
CLASS="TOKEN"
>&#60;</SPAN
></DT
><DD
><P
>is less than, in <A
HREF="#ASCIIDEF"
>ASCII</A
> alphabetical
order</P
><P
><TT
CLASS="USERINPUT"
><B
>if [[ "$a" &#60; "$b" ]]</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>if [ "$a" \&#60; "$b" ]</B
></TT
></P
><P
>Note that the <SPAN
CLASS="QUOTE"
>"&#60;"</SPAN
> needs to be
<A
HREF="#ESCP"
>escaped</A
> within a
<TT
CLASS="USERINPUT"
><B
>[ ]</B
></TT
> construct.</P
></DD
><DT
><A
NAME="GTREF"
></A
><SPAN
CLASS="TOKEN"
>&#62;</SPAN
></DT
><DD
><P
>is greater than, in ASCII alphabetical order</P
><P
><TT
CLASS="USERINPUT"
><B
>if [[ "$a" &#62; "$b" ]]</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>if [ "$a" \&#62; "$b" ]</B
></TT
></P
><P
>Note that the <SPAN
CLASS="QUOTE"
>"&#62;"</SPAN
> needs to be
escaped within a <TT
CLASS="USERINPUT"
><B
>[ ]</B
></TT
> construct.</P
><P
>See <A
HREF="#BUBBLE"
>Example 27-11</A
> for an application of this
comparison operator.</P
></DD
><DT
><A
NAME="STRINGNULL"
></A
><SPAN
CLASS="TOKEN"
>-z</SPAN
></DT
><DD
><P
>string is <I
CLASS="FIRSTTERM"
>null</I
>,
that is, has zero length</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> String='' # Zero-length ("null") string variable.
if [ -z "$String" ]
then
echo "\$String is null."
else
echo "\$String is NOT null."
fi # $String is null.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="STRINGNOTNULL"
></A
><SPAN
CLASS="TOKEN"
>-n</SPAN
></DT
><DD
><P
>string is not <I
CLASS="FIRSTTERM"
>null.</I
></P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <TT
CLASS="USERINPUT"
><B
>-n</B
></TT
> test
requires that the string be quoted within the
test brackets. Using an unquoted string with
<I
CLASS="FIRSTTERM"
>! -z</I
>, or even just the
unquoted string alone within test brackets (see <A
HREF="#STRTEST"
>Example 7-6</A
>) normally works, however, this is
an unsafe practice. <EM
>Always</EM
> quote
a tested string.
<A
NAME="AEN3669"
HREF="#FTN.AEN3669"
><SPAN
CLASS="footnote"
>[37]</SPAN
></A
>
</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX13"
></A
><P
><B
>Example 7-5. Arithmetic and string comparisons</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
a=4
b=5
# Here "a" and "b" can be treated either as integers or strings.
# There is some blurring between the arithmetic and string comparisons,
#+ since Bash variables are not strongly typed.
# Bash permits integer operations and comparisons on variables
#+ whose value consists of all-integer characters.
# Caution advised, however.
echo
if [ "$a" -ne "$b" ]
then
echo "$a is not equal to $b"
echo "(arithmetic comparison)"
fi
echo
if [ "$a" != "$b" ]
then
echo "$a is not equal to $b."
echo "(string comparison)"
# "4" != "5"
# ASCII 52 != ASCII 53
fi
# In this particular instance, both "-ne" and "!=" work.
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="STRTEST"
></A
><P
><B
>Example 7-6. Testing whether a string is <I
CLASS="FIRSTTERM"
>null</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# str-test.sh: Testing null strings and unquoted strings,
#+ but not strings and sealing wax, not to mention cabbages and kings . . .
# Using if [ ... ]
# If a string has not been initialized, it has no defined value.
# This state is called "null" (not the same as zero!).
if [ -n $string1 ] # string1 has not been declared or initialized.
then
echo "String \"string1\" is not null."
else
echo "String \"string1\" is null."
fi # Wrong result.
# Shows $string1 as not null, although it was not initialized.
echo
# Let's try it again.
if [ -n "$string1" ] # This time, $string1 is quoted.
then
echo "String \"string1\" is not null."
else
echo "String \"string1\" is null."
fi # Quote strings within test brackets!
echo
if [ $string1 ] # This time, $string1 stands naked.
then
echo "String \"string1\" is not null."
else
echo "String \"string1\" is null."
fi # This works fine.
# The [ ... ] test operator alone detects whether the string is null.
# However it is good practice to quote it (if [ "$string1" ]).
#
# As Stephane Chazelas points out,
# if [ $string1 ] has one argument, "]"
# if [ "$string1" ] has two arguments, the empty "$string1" and "]"
echo
string1=initialized
if [ $string1 ] # Again, $string1 stands unquoted.
then
echo "String \"string1\" is not null."
else
echo "String \"string1\" is null."
fi # Again, gives correct result.
# Still, it is better to quote it ("$string1"), because . . .
string1="a = b"
if [ $string1 ] # Again, $string1 stands unquoted.
then
echo "String \"string1\" is not null."
else
echo "String \"string1\" is null."
fi # Not quoting "$string1" now gives wrong result!
exit 0 # Thank you, also, Florian Wisser, for the "heads-up".</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX14"
></A
><P
><B
>Example 7-7. <I
CLASS="FIRSTTERM"
>zmore</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# zmore
# View gzipped files with 'more' filter.
E_NOARGS=85
E_NOTFOUND=86
E_NOTGZIP=87
if [ $# -eq 0 ] # same effect as: if [ -z "$1" ]
# $1 can exist, but be empty: zmore "" arg2 arg3
then
echo "Usage: `basename $0` filename" &#62;&#38;2
# Error message to stderr.
exit $E_NOARGS
# Returns 85 as exit status of script (error code).
fi
filename=$1
if [ ! -f "$filename" ] # Quoting $filename allows for possible spaces.
then
echo "File $filename not found!" &#62;&#38;2 # Error message to stderr.
exit $E_NOTFOUND
fi
if [ ${filename##*.} != "gz" ]
# Using bracket in variable substitution.
then
echo "File $1 is not a gzipped file!"
exit $E_NOTGZIP
fi
zcat $1 | more
# Uses the 'more' filter.
# May substitute 'less' if desired.
exit $? # Script returns exit status of pipe.
# Actually "exit $?" is unnecessary, as the script will, in any case,
#+ return the exit status of the last command executed.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="CCOMPARISON1"
></A
>compound comparison</B
></P
><DL
><DT
><A
NAME="COMPOUNDAND"
></A
><SPAN
CLASS="TOKEN"
>-a</SPAN
></DT
><DD
><P
>logical and</P
><P
><TT
CLASS="REPLACEABLE"
><I
>exp1 -a exp2</I
></TT
> returns true if
<EM
>both</EM
> exp1 and exp2 are true.</P
></DD
><DT
><A
NAME="COMPOUNDOR"
></A
><SPAN
CLASS="TOKEN"
>-o</SPAN
></DT
><DD
><P
>logical or </P
><P
><TT
CLASS="REPLACEABLE"
><I
>exp1 -o exp2</I
></TT
> returns
true if either exp1 <EM
>or</EM
> exp2 is
true.</P
></DD
></DL
></DIV
><P
> These are similar to the Bash comparison operators
<B
CLASS="COMMAND"
>&#38;&#38;</B
> and <B
CLASS="COMMAND"
>||</B
>, used
within <A
HREF="#DBLBRACKETS"
>double brackets</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>[[ condition1 &#38;&#38; condition2 ]]</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> The <B
CLASS="COMMAND"
>-o</B
> and <B
CLASS="COMMAND"
>-a</B
> operators
work with the <A
HREF="#TTESTREF"
>test</A
> command or
occur within single test brackets.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ "$expr1" -a "$expr2" ]
then
echo "Both expr1 and expr2 are true."
else
echo "Either expr1 or expr2 is false."
fi</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>But, as <EM
>rihad</EM
> points out:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>[ 1 -eq 1 ] &#38;&#38; [ -n "`echo true 1&#62;&#38;2`" ] # true
[ 1 -eq 2 ] &#38;&#38; [ -n "`echo true 1&#62;&#38;2`" ] # (no output)
# ^^^^^^^ False condition. So far, everything as expected.
# However ...
[ 1 -eq 2 -a -n "`echo true 1&#62;&#38;2`" ] # true
# ^^^^^^^ False condition. So, why "true" output?
# Is it because both condition clauses within brackets evaluate?
[[ 1 -eq 2 &#38;&#38; -n "`echo true 1&#62;&#38;2`" ]] # (no output)
# No, that's not it.
# Apparently &#38;&#38; and || "short-circuit" while -a and -o do not.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><P
>Refer to <A
HREF="#ANDOR"
>Example 8-3</A
>, <A
HREF="#TWODIM"
>Example 27-17</A
>,
and <A
HREF="#WHX"
>Example A-29</A
> to see compound comparison operators
in action.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="NESTEDIFTHEN"
></A
>7.4. Nested <TT
CLASS="REPLACEABLE"
><I
>if/then</I
></TT
> Condition Tests</H1
><P
>Condition tests using the <TT
CLASS="REPLACEABLE"
><I
>if/then</I
></TT
>
construct may be nested. The net result is equivalent to using the
<A
HREF="#LOGOPS1"
><I
CLASS="FIRSTTERM"
>&#38;&#38;</I
></A
> compound
comparison operator.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>a=3
if [ "$a" -gt 0 ]
then
if [ "$a" -lt 5 ]
then
echo "The value of \"a\" lies somewhere between 0 and 5."
fi
fi
# Same result as:
if [ "$a" -gt 0 ] &#38;&#38; [ "$a" -lt 5 ]
then
echo "The value of \"a\" lies somewhere between 0 and 5."
fi</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
HREF="#CARDS"
>Example 37-4</A
> and <A
HREF="#BACKLIGHT"
>Example 17-11</A
>
demonstrate nested <TT
CLASS="REPLACEABLE"
><I
>if/then</I
></TT
> condition
tests.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="TESTTEST"
></A
>7.5. Testing Your Knowledge of Tests</H1
><P
>The systemwide <TT
CLASS="FILENAME"
>xinitrc</TT
> file can be used
to launch the X server. This file contains quite a number
of <I
CLASS="FIRSTTERM"
>if/then</I
> tests. The following
is excerpted from an <SPAN
CLASS="QUOTE"
>"ancient"</SPAN
> version of
<TT
CLASS="FILENAME"
>xinitrc</TT
> (<I
CLASS="FIRSTTERM"
>Red Hat 7.1</I
>,
or thereabouts).</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ -f $HOME/.Xclients ]; then
exec $HOME/.Xclients
elif [ -f /etc/X11/xinit/Xclients ]; then
exec /etc/X11/xinit/Xclients
else
# failsafe settings. Although we should never get here
# (we provide fallbacks in Xclients as well) it can't hurt.
xclock -geometry 100x100-5+5 &#38;
xterm -geometry 80x50-50+150 &#38;
if [ -f /usr/bin/netscape -a -f /usr/share/doc/HTML/index.html ]; then
netscape /usr/share/doc/HTML/index.html &#38;
fi
fi</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Explain the <I
CLASS="FIRSTTERM"
>test</I
> constructs in the
above snippet, then examine an updated version of the
file, <TT
CLASS="FILENAME"
>/etc/X11/xinit/xinitrc</TT
>, and
analyze the <I
CLASS="FIRSTTERM"
>if/then</I
> test constructs
there. You may need to refer ahead to the discussions of <A
HREF="#GREPREF"
>grep</A
>, <A
HREF="#SEDREF"
>sed</A
>,
and <A
HREF="#REGEXREF"
>regular expressions</A
>.</P
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="OPERATIONS"
></A
>Chapter 8. Operations and Related Topics</H1
><DIV
CLASS="SECT1"
><H1
CLASS="SECT1"
><A
NAME="OPS"
></A
>8.1. Operators</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="ASNOP1"
></A
>assignment</B
></P
><DL
><DT
><TT
CLASS="REPLACEABLE"
><I
>variable assignment</I
></TT
></DT
><DD
><P
>Initializing or changing the value of a variable</P
></DD
><DT
>=</DT
><DD
><P
>All-purpose assignment operator, which works for both
arithmetic and string assignments.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>var=27
category=minerals # No spaces allowed after the "=".</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Do not confuse the <SPAN
CLASS="QUOTE"
>"="</SPAN
> assignment
operator with the <A
HREF="#EQUALSIGNREF"
>= test
operator</A
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># = as a test operator
if [ "$string1" = "$string2" ]
then
command
fi
# if [ "X$string1" = "X$string2" ] is safer,
#+ to prevent an error message should one of the variables be empty.
# (The prepended "X" characters cancel out.)</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="AROPS1"
></A
>arithmetic operators</B
></P
><DL
><DT
><SPAN
CLASS="TOKEN"
>+</SPAN
></DT
><DD
><P
>plus</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-</SPAN
></DT
><DD
><P
>minus</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>*</SPAN
></DT
><DD
><P
>multiplication</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>/</SPAN
></DT
><DD
><P
>division</P
></DD
><DT
><A
NAME="EXPONENTIATIONREF"
></A
><SPAN
CLASS="TOKEN"
>**</SPAN
></DT
><DD
><P
>exponentiation</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Bash, version 2.02, introduced the "**" exponentiation operator.
let "z=5**3" # 5 * 5 * 5
echo "z = $z" # z = 125</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="MODULOREF"
></A
><SPAN
CLASS="TOKEN"
>%</SPAN
></DT
><DD
><P
>modulo, or mod (returns the
<I
CLASS="FIRSTTERM"
>remainder</I
> of an integer division
operation)</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>expr 5 % 3</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>2</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
<EM
>5/3 = 1, with remainder 2</EM
>
</P
><P
>This operator finds use in, among other things,
generating numbers within a specific range (see <A
HREF="#EX21"
>Example 9-11</A
> and <A
HREF="#RANDOMTEST"
>Example 9-15</A
>) and
formatting program output (see <A
HREF="#QFUNCTION"
>Example 27-16</A
> and
<A
HREF="#COLLATZ"
>Example A-6</A
>). It can even be used to generate
prime numbers, (see <A
HREF="#PRIMES"
>Example A-15</A
>). Modulo turns
up surprisingly often in numerical recipes.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="GCD"
></A
><P
><B
>Example 8-1. Greatest common divisor</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# gcd.sh: greatest common divisor
# Uses Euclid's algorithm
# The "greatest common divisor" (gcd) of two integers
#+ is the largest integer that will divide both, leaving no remainder.
# Euclid's algorithm uses successive division.
# In each pass,
#+ dividend &#60;--- divisor
#+ divisor &#60;--- remainder
#+ until remainder = 0.
# The gcd = dividend, on the final pass.
#
# For an excellent discussion of Euclid's algorithm, see
#+ Jim Loy's site, http://www.jimloy.com/number/euclids.htm.
# ------------------------------------------------------
# Argument check
ARGS=2
E_BADARGS=85
if [ $# -ne "$ARGS" ]
then
echo "Usage: `basename $0` first-number second-number"
exit $E_BADARGS
fi
# ------------------------------------------------------
gcd ()
{
dividend=$1 # Arbitrary assignment.
divisor=$2 #! It doesn't matter which of the two is larger.
# Why not?
remainder=1 # If an uninitialized variable is used inside
#+ test brackets, an error message results.
until [ "$remainder" -eq 0 ]
do # ^^^^^^^^^^ Must be previously initialized!
let "remainder = $dividend % $divisor"
dividend=$divisor # Now repeat with 2 smallest numbers.
divisor=$remainder
done # Euclid's algorithm
} # Last $dividend is the gcd.
gcd $1 $2
echo; echo "GCD of $1 and $2 = $dividend"; echo
# Exercises :
# ---------
# 1) Check command-line arguments to make sure they are integers,
#+ and exit the script with an appropriate error message if not.
# 2) Rewrite the gcd () function to use local variables.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="ARITHOPSCOMB"
></A
><SPAN
CLASS="TOKEN"
>+=</SPAN
></DT
><DD
><P
><I
CLASS="FIRSTTERM"
>plus-equal</I
> (increment variable
by a constant)
<A
NAME="AEN3907"
HREF="#FTN.AEN3907"
><SPAN
CLASS="footnote"
>[38]</SPAN
></A
>
</P
><P
><TT
CLASS="USERINPUT"
><B
>let "var += 5"</B
></TT
> results in
<TT
CLASS="PARAMETER"
><I
>var</I
></TT
> being incremented by
<TT
CLASS="LITERAL"
>5</TT
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-=</SPAN
></DT
><DD
><P
><I
CLASS="FIRSTTERM"
>minus-equal</I
> (decrement
variable by a constant)</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>*=</SPAN
></DT
><DD
><P
><I
CLASS="FIRSTTERM"
>times-equal</I
> (multiply
variable by a constant)</P
><P
><TT
CLASS="USERINPUT"
><B
>let "var *= 4"</B
></TT
> results in <TT
CLASS="PARAMETER"
><I
>var</I
></TT
>
being multiplied by <TT
CLASS="LITERAL"
>4</TT
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>/=</SPAN
></DT
><DD
><P
><I
CLASS="FIRSTTERM"
>slash-equal</I
> (divide
variable by a constant)</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>%=</SPAN
></DT
><DD
><P
><I
CLASS="FIRSTTERM"
>mod-equal</I
>
(<I
CLASS="FIRSTTERM"
>remainder</I
>
of dividing variable by a constant)</P
><P
><EM
>Arithmetic operators often occur in an
<A
HREF="#EXPRREF"
>expr</A
> or <A
HREF="#LETREF"
>let</A
> expression.</EM
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="ARITHOPS"
></A
><P
><B
>Example 8-2. Using Arithmetic Operations</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Counting to 11 in 10 different ways.
n=1; echo -n "$n "
let "n = $n + 1" # let "n = n + 1" also works.
echo -n "$n "
: $((n = $n + 1))
# ":" necessary because otherwise Bash attempts
#+ to interpret "$((n = $n + 1))" as a command.
echo -n "$n "
(( n = n + 1 ))
# A simpler alternative to the method above.
# Thanks, David Lombard, for pointing this out.
echo -n "$n "
n=$(($n + 1))
echo -n "$n "
: $[ n = $n + 1 ]
# ":" necessary because otherwise Bash attempts
#+ to interpret "$[ n = $n + 1 ]" as a command.
# Works even if "n" was initialized as a string.
echo -n "$n "
n=$[ $n + 1 ]
# Works even if "n" was initialized as a string.
#* Avoid this type of construct, since it is obsolete and nonportable.
# Thanks, Stephane Chazelas.
echo -n "$n "
# Now for C-style increment operators.
# Thanks, Frank Wang, for pointing this out.
let "n++" # let "++n" also works.
echo -n "$n "
(( n++ )) # (( ++n )) also works.
echo -n "$n "
: $(( n++ )) # : $(( ++n )) also works.
echo -n "$n "
: $[ n++ ] # : $[ ++n ] also works
echo -n "$n "
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
><A
NAME="INTVARREF"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Integer variables in older versions of Bash were signed
<I
CLASS="FIRSTTERM"
>long</I
> (32-bit) integers, in the range of
-2147483648 to 2147483647. An operation that took a variable
outside these limits gave an erroneous result.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo $BASH_VERSION # 1.14
a=2147483646
echo "a = $a" # a = 2147483646
let "a+=1" # Increment "a".
echo "a = $a" # a = 2147483647
let "a+=1" # increment "a" again, past the limit.
echo "a = $a" # a = -2147483648
# ERROR: out of range,
# + and the leftmost bit, the sign bit,
# + has been set, making the result negative.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>As of version &#62;= 2.05b, Bash supports 64-bit integers.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="NOFLOATINGPOINT"
></A
></P
><P
>Bash does not understand floating point arithmetic. It
treats numbers containing a decimal point as strings.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>a=1.5
let "b = $a + 1.3" # Error.
# t2.sh: let: b = 1.5 + 1.3: syntax error in expression
# (error token is ".5 + 1.3")
echo "b = $b" # b=1</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Use <A
HREF="#BCREF"
>bc</A
> in scripts that that need floating
point calculations or math library functions.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="FORMALPARA"
><P
><B
>bitwise operators. </B
>The bitwise operators seldom make an appearance in shell scripts.
Their chief use seems to be manipulating and testing values read
from ports or <A
HREF="#SOCKETREF"
>sockets</A
>. <SPAN
CLASS="QUOTE"
>"Bit
flipping"</SPAN
> is more relevant to compiled languages, such
as C and C++, which provide direct access to system
hardware. However, see <EM
>vladz's</EM
>
ingenious use of bitwise operators in his
<I
CLASS="FIRSTTERM"
>base64.sh</I
> (<A
HREF="#BASE64"
>Example A-54</A
>)
script. </P
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="BITWSOPS1"
></A
>bitwise operators</B
></P
><DL
><DT
><SPAN
CLASS="TOKEN"
>&#60;&#60;</SPAN
></DT
><DD
><P
>bitwise left shift (multiplies by <TT
CLASS="LITERAL"
>2</TT
>
for each shift position)</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>&#60;&#60;=</SPAN
></DT
><DD
><P
><I
CLASS="FIRSTTERM"
>left-shift-equal</I
></P
><P
><TT
CLASS="USERINPUT"
><B
>let "var &#60;&#60;= 2"</B
></TT
> results in <TT
CLASS="PARAMETER"
><I
>var</I
></TT
>
left-shifted <TT
CLASS="LITERAL"
>2</TT
> bits (multiplied by <TT
CLASS="LITERAL"
>4</TT
>)</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>&#62;&#62;</SPAN
></DT
><DD
><P
>bitwise right shift (divides by <TT
CLASS="LITERAL"
>2</TT
>
for each shift position)</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>&#62;&#62;=</SPAN
></DT
><DD
><P
><I
CLASS="FIRSTTERM"
>right-shift-equal</I
>
(inverse of <SPAN
CLASS="TOKEN"
>&#60;&#60;=</SPAN
>)</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>&#38;</SPAN
></DT
><DD
><P
>bitwise AND</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>&#38;=</SPAN
></DT
><DD
><P
>bitwise <I
CLASS="FIRSTTERM"
>AND-equal</I
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>|</SPAN
></DT
><DD
><P
>bitwise OR</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>|=</SPAN
></DT
><DD
><P
>bitwise <I
CLASS="FIRSTTERM"
>OR-equal</I
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>~</SPAN
></DT
><DD
><P
>bitwise NOT</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>^</SPAN
></DT
><DD
><P
>bitwise XOR</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>^=</SPAN
></DT
><DD
><P
>bitwise <I
CLASS="FIRSTTERM"
>XOR-equal</I
></P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="LOGOPS1"
></A
>logical (boolean) operators</B
></P
><DL
><DT
><SPAN
CLASS="TOKEN"
>!</SPAN
></DT
><DD
><P
>NOT</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ ! -f $FILENAME ]
then
...</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>&#38;&#38;</SPAN
></DT
><DD
><P
>AND</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ $condition1 ] &#38;&#38; [ $condition2 ]
# Same as: if [ $condition1 -a $condition2 ]
# Returns true if both condition1 and condition2 hold true...
if [[ $condition1 &#38;&#38; $condition2 ]] # Also works.
# Note that &#38;&#38; operator not permitted <EM
>inside brackets</EM
>
#+ of [ ... ] construct.</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><SPAN
CLASS="TOKEN"
>&#38;&#38;</SPAN
> may also be used, depending on context,
in an <A
HREF="#LISTCONSREF"
>and list</A
>
to concatenate commands.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="ORREF"
></A
><SPAN
CLASS="TOKEN"
>||</SPAN
></DT
><DD
><P
>OR</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ $condition1 ] || [ $condition2 ]
# Same as: if [ $condition1 -o $condition2 ]
# Returns true if either condition1 or condition2 holds true...
if [[ $condition1 || $condition2 ]] # Also works.
# Note that || operator not permitted <EM
>inside brackets</EM
>
#+ of a [ ... ] construct.</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Bash tests the <A
HREF="#EXITSTATUSREF"
>exit
status</A
> of each statement linked with a logical
operator.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ANDOR"
></A
><P
><B
>Example 8-3. Compound Condition Tests Using &#38;&#38; and ||</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
a=24
b=47
if [ "$a" -eq 24 ] &#38;&#38; [ "$b" -eq 47 ]
then
echo "Test #1 succeeds."
else
echo "Test #1 fails."
fi
# ERROR: if [ "$a" -eq 24 &#38;&#38; "$b" -eq 47 ]
#+ attempts to execute ' [ "$a" -eq 24 '
#+ and fails to finding matching ']'.
#
# Note: if [[ $a -eq 24 &#38;&#38; $b -eq 24 ]] works.
# The double-bracket if-test is more flexible
#+ than the single-bracket version.
# (The "&#38;&#38;" has a different meaning in line 17 than in line 6.)
# Thanks, Stephane Chazelas, for pointing this out.
if [ "$a" -eq 98 ] || [ "$b" -eq 47 ]
then
echo "Test #2 succeeds."
else
echo "Test #2 fails."
fi
# The -a and -o options provide
#+ an alternative compound condition test.
# Thanks to Patrick Callahan for pointing this out.
if [ "$a" -eq 24 -a "$b" -eq 47 ]
then
echo "Test #3 succeeds."
else
echo "Test #3 fails."
fi
if [ "$a" -eq 98 -o "$b" -eq 47 ]
then
echo "Test #4 succeeds."
else
echo "Test #4 fails."
fi
a=rhino
b=crocodile
if [ "$a" = rhino ] &#38;&#38; [ "$b" = crocodile ]
then
echo "Test #5 succeeds."
else
echo "Test #5 fails."
fi
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The <SPAN
CLASS="TOKEN"
>&#38;&#38;</SPAN
> and <SPAN
CLASS="TOKEN"
>||</SPAN
> operators also
find use in an arithmetic context.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $(( 1 &#38;&#38; 2 )) $((3 &#38;&#38; 0)) $((4 || 0)) $((0 || 0))</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>1 0 1 0</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="MISCOP1"
></A
>miscellaneous operators</B
></P
><DL
><DT
><A
NAME="COMMAOP"
></A
><SPAN
CLASS="TOKEN"
>,</SPAN
></DT
><DD
><P
>Comma operator</P
><P
>The <B
CLASS="COMMAND"
>comma operator</B
> chains together
two or more arithmetic operations. All the operations are
evaluated (with possible <I
CLASS="FIRSTTERM"
>side
effects</I
>.
<A
NAME="AEN4242"
HREF="#FTN.AEN4242"
><SPAN
CLASS="footnote"
>[39]</SPAN
></A
>
</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>let "t1 = ((5 + 3, 7 - 1, 15 - 4))"
echo "t1 = $t1" ^^^^^^ # t1 = 11
# Here t1 is set to the result of the last operation. Why?
let "t2 = ((a = 9, 15 / 3))" # Set "a" and calculate "t2".
echo "t2 = $t2 a = $a" # t2 = 5 a = 9</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The comma operator finds use mainly in <A
HREF="#FORLOOPREF1"
>for loops</A
>. See <A
HREF="#FORLOOPC"
>Example 11-13</A
>.</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="NUMERICAL-CONSTANTS"
></A
>8.2. Numerical Constants</H1
><P
><A
NAME="NUMCONSTANTS"
></A
>A shell script interprets a number
as decimal (base 10), unless that number has a
special prefix or notation. A number preceded by a
<TT
CLASS="REPLACEABLE"
><I
>0</I
></TT
> is <TT
CLASS="REPLACEABLE"
><I
>octal</I
></TT
>
(base 8). A number preceded by <TT
CLASS="REPLACEABLE"
><I
>0x</I
></TT
>
is <TT
CLASS="REPLACEABLE"
><I
>hexadecimal</I
></TT
> (base 16). A number
with an embedded <TT
CLASS="REPLACEABLE"
><I
>#</I
></TT
> evaluates as
<TT
CLASS="REPLACEABLE"
><I
>BASE#NUMBER</I
></TT
> (with range and notational
restrictions).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="NUMBERS"
></A
><P
><B
>Example 8-4. Representation of numerical constants</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# numbers.sh: Representation of numbers in different bases.
# Decimal: the default
let "dec = 32"
echo "decimal number = $dec" # 32
# Nothing out of the ordinary here.
# Octal: numbers preceded by '0' (zero)
let "oct = 032"
echo "octal number = $oct" # 26
# Expresses result in decimal.
# --------- ------ -- -------
# Hexadecimal: numbers preceded by '0x' or '0X'
let "hex = 0x32"
echo "hexadecimal number = $hex" # 50
echo $((0x9abc)) # 39612
# ^^ ^^ double-parentheses arithmetic expansion/evaluation
# Expresses result in decimal.
# Other bases: BASE#NUMBER
# BASE between 2 and 64.
# NUMBER must use symbols within the BASE range, see below.
let "bin = 2#111100111001101"
echo "binary number = $bin" # 31181
let "b32 = 32#77"
echo "base-32 number = $b32" # 231
let "b64 = 64#@_"
echo "base-64 number = $b64" # 4031
# This notation only works for a limited range (2 - 64) of ASCII characters.
# 10 digits + 26 lowercase characters + 26 uppercase characters + @ + _
echo
echo $((36#zz)) $((2#10101010)) $((16#AF16)) $((53#1aA))
# 1295 170 44822 3375
# Important note:
# --------------
# Using a digit out of range of the specified base notation
#+ gives an error message.
let "bad_oct = 081"
# (Partial) error message output:
# bad_oct = 081: value too great for base (error token is "081")
# Octal numbers use only digits in the range 0 - 7.
exit $? # Exit value = 1 (error)
# Thanks, Rich Bartell and Stephane Chazelas, for clarification.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="DBLPARENS"
></A
>8.3. The Double-Parentheses Construct</H1
><P
><A
NAME="DBLPARENSREF"
></A
></P
><P
>Similar to the <A
HREF="#LETREF"
>let</A
> command,
the <B
CLASS="COMMAND"
>(( ... ))</B
> construct permits
arithmetic expansion and evaluation. In its simplest
form, <TT
CLASS="USERINPUT"
><B
>a=$(( 5 + 3 ))</B
></TT
> would set
<TT
CLASS="USERINPUT"
><B
>a</B
></TT
> to <TT
CLASS="USERINPUT"
><B
>5 + 3</B
></TT
>, or
<TT
CLASS="USERINPUT"
><B
>8</B
></TT
>. However, this double-parentheses
construct is also a mechanism for allowing C-style
manipulation of variables in Bash, for example,
<TT
CLASS="VARNAME"
>(( var++ ))</TT
>.</P
><P
><A
NAME="PLUSPLUSREF"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="CVARS"
></A
><P
><B
>Example 8-5. C-style manipulation of variables</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# c-vars.sh
# Manipulating a variable, C-style, using the (( ... )) construct.
echo
(( a = 23 )) # Setting a value, C-style,
#+ with spaces on both sides of the "=".
echo "a (initial value) = $a" # 23
(( a++ )) # Post-increment 'a', C-style.
echo "a (after a++) = $a" # 24
(( a-- )) # Post-decrement 'a', C-style.
echo "a (after a--) = $a" # 23
(( ++a )) # Pre-increment 'a', C-style.
echo "a (after ++a) = $a" # 24
(( --a )) # Pre-decrement 'a', C-style.
echo "a (after --a) = $a" # 23
echo
########################################################
# Note that, as in C, pre- and post-decrement operators
#+ have different side-effects.
n=1; let --n &#38;&#38; echo "True" || echo "False" # False
n=1; let n-- &#38;&#38; echo "True" || echo "False" # True
# Thanks, Jeroen Domburg.
########################################################
echo
(( t = a&#60;45?7:11 )) # C-style trinary operator.
# ^ ^ ^
echo "If a &#60; 45, then t = 7, else t = 11." # a = 23
echo "t = $t " # t = 7
echo
# -----------------
# Easter Egg alert!
# -----------------
# Chet Ramey seems to have snuck a bunch of undocumented C-style
#+ constructs into Bash (actually adapted from ksh, pretty much).
# In the Bash docs, Ramey calls (( ... )) shell arithmetic,
#+ but it goes far beyond that.
# Sorry, Chet, the secret is out.
# See also "for" and "while" loops using the (( ... )) construct.
# These work only with version 2.04 or later of Bash.
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#FORLOOPC"
>Example 11-13</A
> and <A
HREF="#NUMBERS"
>Example 8-4</A
>.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="OPPRECEDENCE"
></A
>8.4. Operator Precedence</H1
><P
><A
NAME="OPPRECEDENCE1"
></A
></P
><P
> In a script, operations execute in order of
<I
CLASS="FIRSTTERM"
>precedence</I
>: the higher precedence operations
execute <EM
>before</EM
> the lower precedence ones.
<A
NAME="AEN4290"
HREF="#FTN.AEN4290"
><SPAN
CLASS="footnote"
>[40]</SPAN
></A
>
</P
><DIV
CLASS="TABLE"
><A
NAME="AEN4294"
></A
><P
><B
>Table 8-1. Operator Precedence</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Operator</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Meaning</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Comments</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
></TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><B
CLASS="COMMAND"
>HIGHEST PRECEDENCE</B
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>var++ var--</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>post-increment, post-decrement</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#CSTYLE"
>C-style</A
> operators</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>++var --var</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>pre-increment, pre-decrement</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>! ~</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#NOTREF"
>negation</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>logical / bitwise, inverts sense of following
operator</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>**</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#EXPONENTIATIONREF"
>exponentiation</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#AROPS1"
>arithmetic
operation</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>* / %</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>multiplication, division, modulo</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>arithmetic operation</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>+ -</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>addition, subtraction</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>arithmetic operation</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#60;&#60; &#62;&#62;</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>left, right shift</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#BITWSOPS1"
>bitwise</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-z -n</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><I
CLASS="FIRSTTERM"
>unary</I
> comparison</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>string is/is-not <A
HREF="#STRINGNULL"
>null</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-e -f -t -x, etc.</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><I
CLASS="FIRSTTERM"
>unary</I
> comparison</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#FTO"
>file-test</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#60; -lt &#62; -gt &#60;= -le &#62;= -ge</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><I
CLASS="FIRSTTERM"
>compound</I
> comparison</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>string and integer</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-nt -ot -ef</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><I
CLASS="FIRSTTERM"
>compound</I
> comparison</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>file-test</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>== -eq <A
HREF="#NOTEQUAL"
>!=</A
>
-ne</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>equality / inequality</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>test operators, string and integer</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#38;</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>AND</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>bitwise</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>^</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>XOR</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><I
CLASS="FIRSTTERM"
>exclusive</I
> OR, bitwise</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>|</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>OR</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>bitwise</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#38;&#38; -a</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>AND</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#LOGOPS1"
>logical</A
>,
<I
CLASS="FIRSTTERM"
>compound</I
>
comparison</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>|| -o</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>OR</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>logical, <I
CLASS="FIRSTTERM"
>compound</I
>
comparison</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>?:</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#CSTRINARY"
>trinary
operator</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>C-style</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>=</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#EQREF"
>assignment</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>(do not confuse with equality
<I
CLASS="FIRSTTERM"
>test</I
>)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>*= /= %= += -= &#60;&#60;= &#62;&#62;= &#38;=</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#ARITHOPSCOMB"
>combination
assignment</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>times-equal, divide-equal, mod-equal, etc.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>,</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#COMMAOP"
>comma</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>links a sequence of operations</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
></TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><B
CLASS="COMMAND"
>LOWEST PRECEDENCE</B
></TD
></TR
></TBODY
></TABLE
></DIV
><P
>In practice, all you really need to remember is the
following:</P
><P
></P
><UL
><LI
><P
>The <SPAN
CLASS="QUOTE"
>"My Dear Aunt Sally"</SPAN
> mantra (<EM
>multiply,
divide, add, subtract</EM
>) for the familiar <A
HREF="#AROPS1"
>arithmetic operations</A
>.</P
></LI
><LI
><P
>The <I
CLASS="FIRSTTERM"
>compound</I
> logical operators,
<B
CLASS="COMMAND"
>&#38;&#38;</B
>, <B
CLASS="COMMAND"
>||</B
>, <B
CLASS="COMMAND"
>-a</B
>,
and <B
CLASS="COMMAND"
>-o</B
> have low precedence.</P
></LI
><LI
><P
>The order of evaluation of equal-precedence operators is
usually <I
CLASS="FIRSTTERM"
>left-to-right</I
>.</P
></LI
></UL
><P
>Now, let's utilize our knowledge of operator precedence to
analyze a couple of lines from the
<TT
CLASS="FILENAME"
>/etc/init.d/functions file</TT
>, as found in
the <I
CLASS="FIRSTTERM"
>Fedora Core</I
> Linux distro.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>while [ -n "$remaining" -a "$retry" -gt 0 ]; do
# This looks rather daunting at first glance.
# Separate the conditions:
while [ -n "$remaining" -a "$retry" -gt 0 ]; do
# --condition 1-- ^^ --condition 2-
# If variable "$remaining" is not zero length
#+ AND (-a)
#+ variable "$retry" is greater-than zero
#+ then
#+ the [ expresion-within-condition-brackets ] returns success (0)
#+ and the while-loop executes an iteration.
# ==============================================================
# Evaluate "condition 1" and "condition 2" ***before***
#+ ANDing them. Why? Because the AND (-a) has a lower precedence
#+ than the -n and -gt operators,
#+ and therefore gets evaluated *last*.
#################################################################
if [ -f /etc/sysconfig/i18n -a -z "${NOLOCALE:-}" ] ; then
# Again, separate the conditions:
if [ -f /etc/sysconfig/i18n -a -z "${NOLOCALE:-}" ] ; then
# --condition 1--------- ^^ --condition 2-----
# If file "/etc/sysconfig/i18n" exists
#+ AND (-a)
#+ variable $NOLOCALE is zero length
#+ then
#+ the [ test-expresion-within-condition-brackets ] returns success (0)
#+ and the commands following execute.
#
# As before, the AND (-a) gets evaluated *last*
#+ because it has the lowest precedence of the operators within
#+ the test brackets.
# ==============================================================
# Note:
# ${NOLOCALE:-} is a parameter expansion that seems redundant.
# But, if $NOLOCALE has not been declared, it gets set to *null*,
#+ in effect declaring it.
# This makes a difference in some contexts.</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>To avoid confusion or error in a complex sequence of test
operators, break up the sequence into bracketed sections.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ "$v1" -gt "$v2" -o "$v1" -lt "$v2" -a -e "$filename" ]
# Unclear what's going on here...
if [[ "$v1" -gt "$v2" ]] || [[ "$v1" -lt "$v2" ]] &#38;&#38; [[ -e "$filename" ]]
# Much better -- the condition tests are grouped in logical sections.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
></DIV
></DIV
></DIV
><DIV
CLASS="PART"
><A
NAME="PART3"
></A
><DIV
CLASS="TITLEPAGE"
><H1
CLASS="TITLE"
>Part 3. Beyond the Basics</H1
><DIV
CLASS="TOC"
><DL
><DT
><B
>Table of Contents</B
></DT
><DT
>9. <A
HREF="#VARIABLES2"
>Another Look at Variables</A
></DT
><DD
><DL
><DT
>9.1. <A
HREF="#INTERNALVARIABLES"
>Internal Variables</A
></DT
><DT
>9.2. <A
HREF="#DECLAREREF"
>Typing variables: <B
CLASS="COMMAND"
>declare</B
> or
<B
CLASS="COMMAND"
>typeset</B
></A
></DT
><DT
>9.3. <A
HREF="#RANDOMVAR"
>$RANDOM: generate random integer</A
></DT
></DL
></DD
><DT
>10. <A
HREF="#MANIPULATINGVARS"
>Manipulating Variables</A
></DT
><DD
><DL
><DT
>10.1. <A
HREF="#STRING-MANIPULATION"
>Manipulating Strings</A
></DT
><DT
>10.2. <A
HREF="#PARAMETER-SUBSTITUTION"
>Parameter Substitution</A
></DT
></DL
></DD
><DT
>11. <A
HREF="#LOOPS"
>Loops and Branches</A
></DT
><DD
><DL
><DT
>11.1. <A
HREF="#LOOPS1"
>Loops</A
></DT
><DT
>11.2. <A
HREF="#NESTEDLOOPS"
>Nested Loops</A
></DT
><DT
>11.3. <A
HREF="#LOOPCONTROL"
>Loop Control</A
></DT
><DT
>11.4. <A
HREF="#TESTBRANCH"
>Testing and Branching</A
></DT
></DL
></DD
><DT
>12. <A
HREF="#COMMANDSUB"
>Command Substitution</A
></DT
><DT
>13. <A
HREF="#ARITHEXP"
>Arithmetic Expansion</A
></DT
><DT
>14. <A
HREF="#RECESS-TIME"
>Recess Time</A
></DT
></DL
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="VARIABLES2"
></A
>Chapter 9. Another Look at Variables</H1
><P
>Used properly, variables can add power and flexibility
to scripts. This requires learning their subtleties and
nuances.</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="INTERNALVARIABLES"
></A
>9.1. Internal Variables</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><TT
CLASS="REPLACEABLE"
><I
><A
HREF="#BUILTINREF"
>Builtin</A
> variables:</I
></TT
></DT
><DD
><P
>variables affecting bash script behavior</P
></DD
><DT
><A
NAME="BASHVARREF"
></A
><TT
CLASS="VARNAME"
>$BASH</TT
></DT
><DD
><P
>The path to the <I
CLASS="FIRSTTERM"
>Bash</I
>
binary itself
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $BASH</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/bin/bash</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="BASHENVREF"
></A
><TT
CLASS="VARNAME"
>$BASH_ENV</TT
></DT
><DD
><P
>An <A
HREF="#ENVREF"
>environmental
variable</A
> pointing to a Bash startup file to be read
when a script is invoked</P
></DD
><DT
><A
NAME="BASHSUBSHELLREF"
></A
><TT
CLASS="VARNAME"
>$BASH_SUBSHELL</TT
></DT
><DD
><P
>A variable indicating the <A
HREF="#SUBSHELLSREF"
>subshell</A
> level. This is a
new addition to Bash, <A
HREF="#BASH3REF"
>version 3</A
>.</P
><P
>See <A
HREF="#SUBSHELL"
>Example 21-1</A
> for usage.</P
></DD
><DT
><A
NAME="BASHPIDREF"
></A
><TT
CLASS="VARNAME"
>$BASHPID</TT
></DT
><DD
><P
><I
CLASS="FIRSTTERM"
>Process ID</I
>
of the current instance of Bash. This is not the same as the
<A
HREF="#PROCCID"
>$$</A
> variable, but it often
gives the same result.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash4$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $$</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>11015</TT
>
<TT
CLASS="PROMPT"
>bash4$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $BASHPID</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>11015</TT
>
<TT
CLASS="PROMPT"
>bash4$ </TT
><TT
CLASS="USERINPUT"
><B
>ps ax | grep bash4</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>11015 pts/2 R 0:00 bash4</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="BASHPIDREF2"
></A
>But ...</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
echo "\$\$ outside of subshell = $$" # 9602
echo "\$BASH_SUBSHELL outside of subshell = $BASH_SUBSHELL" # 0
echo "\$BASHPID outside of subshell = $BASHPID" # 9602
echo
( echo "\$\$ inside of subshell = $$" # 9602
echo "\$BASH_SUBSHELL inside of subshell = $BASH_SUBSHELL" # 1
echo "\$BASHPID inside of subshell = $BASHPID" ) # 9603
# Note that $$ returns PID of parent process.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><TT
CLASS="VARNAME"
>$BASH_VERSINFO[n]</TT
></DT
><DD
><P
>A 6-element <A
HREF="#ARRAYREF"
>array</A
>
containing version information about the installed release
of Bash. This is similar to <TT
CLASS="VARNAME"
>$BASH_VERSION</TT
>,
below, but a bit more detailed.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Bash version info:
for n in 0 1 2 3 4 5
do
echo "BASH_VERSINFO[$n] = ${BASH_VERSINFO[$n]}"
done
# BASH_VERSINFO[0] = 3 # Major version no.
# BASH_VERSINFO[1] = 00 # Minor version no.
# BASH_VERSINFO[2] = 14 # Patch level.
# BASH_VERSINFO[3] = 1 # Build version.
# BASH_VERSINFO[4] = release # Release status.
# BASH_VERSINFO[5] = i386-redhat-linux-gnu # Architecture
# (same as $MACHTYPE).</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><TT
CLASS="VARNAME"
>$BASH_VERSION</TT
></DT
><DD
><P
>The version of Bash installed on the system</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $BASH_VERSION</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>3.2.25(1)-release</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>tcsh% </TT
><TT
CLASS="USERINPUT"
><B
>echo $BASH_VERSION</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>BASH_VERSION: Undefined variable.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Checking $BASH_VERSION is a good method of determining which
shell is running. <A
HREF="#SHELLVARREF"
>$SHELL</A
>
does not necessarily give the correct answer.</P
></DD
><DT
><A
NAME="CDPATHREF"
></A
><TT
CLASS="VARNAME"
>$CDPATH</TT
></DT
><DD
><P
>A colon-separated list of search paths
available to the <A
HREF="#CDREF"
>cd</A
>
command, similar in function to the <A
HREF="#PATHREF"
>$PATH</A
> variable for binaries.
The <TT
CLASS="VARNAME"
>$CDPATH</TT
> variable may be set in the
local <A
HREF="#BASHRC"
><TT
CLASS="FILENAME"
>~/.bashrc</TT
></A
>
file.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cd bash-doc</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bash: cd: bash-doc: No such file or directory</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>CDPATH=/usr/share/doc</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cd bash-doc</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/usr/share/doc/bash-doc</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $PWD</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/usr/share/doc/bash-doc</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="DIRSTACKREF"
></A
><TT
CLASS="VARNAME"
>$DIRSTACK</TT
></DT
><DD
><P
>The top value in the directory stack
<A
NAME="AEN4671"
HREF="#FTN.AEN4671"
><SPAN
CLASS="footnote"
>[41]</SPAN
></A
>
(affected by <A
HREF="#PUSHDREF"
>pushd</A
> and <A
HREF="#POPDREF"
>popd</A
>)</P
><P
>This builtin
variable corresponds to the <A
HREF="#DIRSD"
>dirs</A
>
command, however <B
CLASS="COMMAND"
>dirs</B
> shows the entire
contents of the directory stack.</P
></DD
><DT
><TT
CLASS="VARNAME"
>$EDITOR</TT
></DT
><DD
><P
>The default editor invoked by a script, usually
<B
CLASS="COMMAND"
>vi</B
> or <B
CLASS="COMMAND"
>emacs</B
>.</P
></DD
><DT
><A
NAME="EUIDREF"
></A
><TT
CLASS="VARNAME"
>$EUID</TT
></DT
><DD
><P
><SPAN
CLASS="QUOTE"
>"effective"</SPAN
> user ID number</P
><P
>Identification number of whatever identity the
current user has assumed, perhaps by means of <A
HREF="#SUREF"
>su</A
>.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <TT
CLASS="VARNAME"
>$EUID</TT
> is not necessarily
the same as the <A
HREF="#UIDREF"
>$UID</A
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><TT
CLASS="VARNAME"
>$FUNCNAME</TT
></DT
><DD
><P
>Name of the current function</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>xyz23 ()
{
echo "$FUNCNAME now executing." # xyz23 now executing.
}
xyz23
echo "FUNCNAME = $FUNCNAME" # FUNCNAME =
# Null value outside a function.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>See also <A
HREF="#USEGETOPT"
>Example A-50</A
>.</P
></DD
><DT
><TT
CLASS="VARNAME"
>$GLOBIGNORE</TT
></DT
><DD
><P
>A list of filename patterns to be excluded from
matching in <A
HREF="#GLOBBINGREF"
>globbing</A
>.</P
></DD
><DT
><A
NAME="GROUPSREF"
></A
><TT
CLASS="VARNAME"
>$GROUPS</TT
></DT
><DD
><P
>Groups current user belongs to</P
><P
>This is a listing (array) of the group id numbers for
current user, as recorded in
<A
HREF="#DATAFILESREF1"
><TT
CLASS="FILENAME"
>/etc/passwd</TT
></A
>
and <TT
CLASS="FILENAME"
>/etc/group</TT
>.
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>echo $GROUPS</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>0</TT
>
<TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>echo ${GROUPS[1]}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>1</TT
>
<TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>echo ${GROUPS[5]}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>6</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="HOMEDIRREF"
></A
><TT
CLASS="VARNAME"
>$HOME</TT
></DT
><DD
><P
>Home directory of the user, usually <TT
CLASS="FILENAME"
>/home/username</TT
> (see <A
HREF="#EX6"
>Example 10-7</A
>)</P
></DD
><DT
><A
NAME="HOSTNAMEREF"
></A
><TT
CLASS="VARNAME"
>$HOSTNAME</TT
></DT
><DD
><P
>The <A
HREF="#HNAMEREF"
>hostname</A
> command
assigns the system host name at bootup in an init script.
However, the <TT
CLASS="FUNCTION"
>gethostname()</TT
> function
sets the Bash internal variable <TT
CLASS="VARNAME"
>$HOSTNAME</TT
>.
See also <A
HREF="#EX6"
>Example 10-7</A
>.</P
></DD
><DT
><TT
CLASS="VARNAME"
>$HOSTTYPE</TT
></DT
><DD
><P
>host type</P
><P
>Like <A
HREF="#MACHTYPEREF"
>$MACHTYPE</A
>,
identifies the system hardware.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $HOSTTYPE</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>i686</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="IFSREF"
></A
><TT
CLASS="VARNAME"
>$IFS</TT
></DT
><DD
><P
>internal field separator</P
><P
>This variable determines how Bash recognizes <A
HREF="#FIELDREF"
>fields</A
>, or word boundaries,
when it interprets character strings.</P
><P
><A
NAME="IFSWS"
></A
></P
><P
>$IFS defaults to <A
HREF="#WHITESPACEREF"
>whitespace</A
> (space,
tab, and newline), but may be changed, for example,
to parse a comma-separated data file. Note that
<A
HREF="#APPREF"
>$*</A
> uses the first
character held in <TT
CLASS="VARNAME"
>$IFS</TT
>. See <A
HREF="#WEIRDVARS"
>Example 5-1</A
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "$IFS"</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>(With $IFS set to default, a blank line displays.)</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "$IFS" | cat -vte</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> ^I$
$</TT
>
<TT
CLASS="COMPUTEROUTPUT"
>(Show whitespace: here a single space, ^I [horizontal tab],
and newline, and display "$" at end-of-line.)</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>bash -c 'set w x y z; IFS=":-;"; echo "$*"'</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>w:x:y:z</TT
>
<TT
CLASS="COMPUTEROUTPUT"
>(Read commands from string and assign any arguments to pos params.)</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Set <TT
CLASS="VARNAME"
>$IFS</TT
> to eliminate whitespace
in <A
HREF="#PATHNAMEREF"
>pathnames</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>IFS="$(printf '\n\t')" # Per David Wheeler.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><TT
CLASS="VARNAME"
>$IFS</TT
> does not handle whitespace
the same as it does other characters.
<DIV
CLASS="EXAMPLE"
><A
NAME="IFSH"
></A
><P
><B
>Example 9-1. $IFS and whitespace</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ifs.sh
var1="a+b+c"
var2="d-e-f"
var3="g,h,i"
IFS=+
# The plus sign will be interpreted as a separator.
echo $var1 # a b c
echo $var2 # d-e-f
echo $var3 # g,h,i
echo
IFS="-"
# The plus sign reverts to default interpretation.
# The minus sign will be interpreted as a separator.
echo $var1 # a+b+c
echo $var2 # d e f
echo $var3 # g,h,i
echo
IFS=","
# The comma will be interpreted as a separator.
# The minus sign reverts to default interpretation.
echo $var1 # a+b+c
echo $var2 # d-e-f
echo $var3 # g h i
echo
IFS=" "
# The space character will be interpreted as a separator.
# The comma reverts to default interpretation.
echo $var1 # a+b+c
echo $var2 # d-e-f
echo $var3 # g,h,i
# ======================================================== #
# However ...
# $IFS treats whitespace differently than other characters.
output_args_one_per_line()
{
for arg
do
echo "[$arg]"
done # ^ ^ Embed within brackets, for your viewing pleasure.
}
echo; echo "IFS=\" \""
echo "-------"
IFS=" "
var=" a b c "
# ^ ^^ ^^^
output_args_one_per_line $var # output_args_one_per_line `echo " a b c "`
# [a]
# [b]
# [c]
echo; echo "IFS=:"
echo "-----"
IFS=:
var=":a::b:c:::" # Same pattern as above,
# ^ ^^ ^^^ #+ but substituting ":" for " " ...
output_args_one_per_line $var
# []
# [a]
# []
# [b]
# [c]
# []
# []
# Note "empty" brackets.
# The same thing happens with the "FS" field separator in awk.
echo
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
>
</P
></TD
></TR
></TABLE
></DIV
><P
>(Many thanks, St<53>phane Chazelas, for clarification
and above examples.)</P
><P
>See also <A
HREF="#ISSPAMMER"
>Example 16-41</A
>, <A
HREF="#BINGREP"
>Example 11-8</A
>, and <A
HREF="#MAILBOXGREP"
>Example 19-14</A
>
for instructive examples of using
<TT
CLASS="VARNAME"
>$IFS</TT
>.</P
></DD
><DT
><TT
CLASS="VARNAME"
>$IGNOREEOF</TT
></DT
><DD
><P
>Ignore EOF: how many end-of-files (control-D)
the shell will ignore before logging out.</P
></DD
><DT
><TT
CLASS="VARNAME"
>$LC_COLLATE</TT
></DT
><DD
><P
>Often set in the <A
HREF="#SAMPLE-BASHRC"
><TT
CLASS="FILENAME"
>.bashrc</TT
></A
>
or <TT
CLASS="FILENAME"
>/etc/profile</TT
> files, this
variable controls collation order in filename
expansion and pattern matching. If mishandled,
<TT
CLASS="VARNAME"
>LC_COLLATE</TT
> can cause unexpected results in
<A
HREF="#GLOBBINGREF"
>filename globbing</A
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>As of version 2.05 of Bash,
filename globbing no longer distinguishes between lowercase
and uppercase letters in a character range between
brackets. For example, <B
CLASS="COMMAND"
>ls [A-M]*</B
>
would match both <TT
CLASS="FILENAME"
>File1.txt</TT
>
and <TT
CLASS="FILENAME"
>file1.txt</TT
>. To revert to
the customary behavior of bracket matching, set
<TT
CLASS="VARNAME"
>LC_COLLATE</TT
> to <TT
CLASS="OPTION"
>C</TT
>
by an <TT
CLASS="USERINPUT"
><B
>export LC_COLLATE=C</B
></TT
>
in <TT
CLASS="FILENAME"
>/etc/profile</TT
> and/or
<TT
CLASS="FILENAME"
>~/.bashrc</TT
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><TT
CLASS="VARNAME"
>$LC_CTYPE</TT
></DT
><DD
><P
>This internal variable controls character interpretation
in <A
HREF="#GLOBBINGREF"
>globbing</A
> and pattern
matching.</P
></DD
><DT
><A
NAME="LINENOREF"
></A
><TT
CLASS="VARNAME"
>$LINENO</TT
></DT
><DD
><P
>This variable is the line number of the shell
script in which this variable appears. It has significance only
within the script in which it appears, and is chiefly useful for
debugging purposes.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># *** BEGIN DEBUG BLOCK ***
last_cmd_arg=$_ # Save it.
echo "At line number $LINENO, variable \"v1\" = $v1"
echo "Last command argument processed = $last_cmd_arg"
# *** END DEBUG BLOCK ***</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="MACHTYPEREF"
></A
><TT
CLASS="VARNAME"
>$MACHTYPE</TT
></DT
><DD
><P
>machine type</P
><P
>Identifies the system hardware.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $MACHTYPE</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>i686</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="OLDPWD"
></A
><TT
CLASS="VARNAME"
>$OLDPWD</TT
></DT
><DD
><P
>Old working directory
(<SPAN
CLASS="QUOTE"
>"OLD-Print-Working-Directory"</SPAN
>,
previous directory you were in).</P
></DD
><DT
><TT
CLASS="VARNAME"
>$OSTYPE</TT
></DT
><DD
><P
>operating system type</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $OSTYPE</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>linux</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="PATHREF"
></A
><TT
CLASS="VARNAME"
>$PATH</TT
></DT
><DD
><P
>Path to binaries, usually
<TT
CLASS="FILENAME"
>/usr/bin/</TT
>,
<TT
CLASS="FILENAME"
>/usr/X11R6/bin/</TT
>,
<TT
CLASS="FILENAME"
>/usr/local/bin</TT
>, etc.</P
><P
>When given a command, the shell automatically does
a hash table search on the directories listed in the
<I
CLASS="FIRSTTERM"
>path</I
> for the executable. The path
is stored in the <A
HREF="#ENVREF"
>environmental
variable</A
>, <TT
CLASS="VARNAME"
>$PATH</TT
>, a list
of directories, separated by colons. Normally,
the system stores the <TT
CLASS="VARNAME"
>$PATH</TT
>
definition in <TT
CLASS="FILENAME"
>/etc/profile</TT
>
and/or <A
HREF="#SAMPLE-BASHRC"
><TT
CLASS="FILENAME"
>~/.bashrc</TT
></A
>
(see <A
HREF="#FILES"
>Appendix H</A
>).</P
><P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><B
CLASS="COMMAND"
>echo $PATH</B
>
<TT
CLASS="COMPUTEROUTPUT"
>/bin:/usr/bin:/usr/local/bin:/usr/X11R6/bin:/sbin:/usr/sbin</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><TT
CLASS="USERINPUT"
><B
>PATH=${PATH}:/opt/bin</B
></TT
> appends
the <TT
CLASS="FILENAME"
>/opt/bin</TT
>
directory to the current path. In a script, it may be
expedient to temporarily add a directory to the path
in this way. When the script exits, this restores the
original <TT
CLASS="VARNAME"
>$PATH</TT
> (a child process, such
as a script, may not change the environment of the parent
process, the shell).</P
><P
><A
NAME="CURRENTWDREF"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The current <SPAN
CLASS="QUOTE"
>"working directory"</SPAN
>,
<TT
CLASS="FILENAME"
>./</TT
>, is usually
omitted from the <TT
CLASS="VARNAME"
>$PATH</TT
> as a security
measure.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="PIPESTATUSREF"
></A
><TT
CLASS="VARNAME"
>$PIPESTATUS</TT
></DT
><DD
><P
><A
HREF="#ARRAYREF"
>Array</A
> variable holding
<A
HREF="#EXITSTATUSREF"
>exit status</A
>(es) of
last executed <I
CLASS="FIRSTTERM"
>foreground</I
> <A
HREF="#PIPEREF"
>pipe</A
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $PIPESTATUS</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>0</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -al | bogus_command</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bash: bogus_command: command not found</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo ${PIPESTATUS[1]}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>127</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -al | bogus_command</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bash: bogus_command: command not found</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $?</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>127</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The members of the <TT
CLASS="VARNAME"
>$PIPESTATUS</TT
>
array hold the exit status of each respective command
executed in a pipe. <TT
CLASS="VARNAME"
>$PIPESTATUS[0]</TT
>
holds the exit status of the first command in the pipe,
<TT
CLASS="VARNAME"
>$PIPESTATUS[1]</TT
> the exit status of
the second command, and so on.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
> The <TT
CLASS="VARNAME"
>$PIPESTATUS</TT
> variable
may contain an erroneous <SPAN
CLASS="ERRORCODE"
>0</SPAN
> value
in a login shell (in releases prior to 3.0 of Bash).
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>tcsh% </TT
><TT
CLASS="USERINPUT"
><B
>bash</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>who | grep nobody | sort</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo ${PIPESTATUS[*]}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>0</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> The above lines contained in a script would produce the expected
<TT
CLASS="COMPUTEROUTPUT"
>0 1 0</TT
> output.
</P
><P
> Thank you, Wayne Pollock for pointing this out and supplying the
above example.
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <TT
CLASS="VARNAME"
>$PIPESTATUS</TT
> variable gives
unexpected results in some contexts.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $BASH_VERSION</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>3.00.14(1)-release</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>$ ls | bogus_command | wc</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bash: bogus_command: command not found
0 0 0</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo ${PIPESTATUS[@]}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>141 127 0</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Chet Ramey attributes the above output to the behavior of
<A
HREF="#LSREF"
>ls</A
>. If <I
CLASS="FIRSTTERM"
>ls</I
>
writes to a <I
CLASS="FIRSTTERM"
>pipe</I
> whose output is not
read, then <TT
CLASS="REPLACEABLE"
><I
>SIGPIPE</I
></TT
> kills it,
and its <A
HREF="#EXITSTATUSREF"
>exit status</A
>
is <SPAN
CLASS="RETURNVALUE"
>141</SPAN
>. Otherwise
its exit status is <SPAN
CLASS="RETURNVALUE"
>0</SPAN
>,
as expected. This likewise is the case for <A
HREF="#TRREF"
>tr</A
>.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><TT
CLASS="VARNAME"
>$PIPESTATUS</TT
> is a
<SPAN
CLASS="QUOTE"
>"volatile"</SPAN
> variable. It needs to be
captured immediately after the pipe in question, before
any other command intervenes.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>$ ls | bogus_command | wc</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bash: bogus_command: command not found
0 0 0</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo ${PIPESTATUS[@]}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>0 127 0</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo ${PIPESTATUS[@]}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>0</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <A
HREF="#PIPEFAILREF"
>pipefail option</A
>
may be useful in cases where
<TT
CLASS="VARNAME"
>$PIPESTATUS</TT
> does not give the desired
information.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="PPIDREF"
></A
><TT
CLASS="VARNAME"
>$PPID</TT
></DT
><DD
><P
></P
><P
>The <TT
CLASS="VARNAME"
>$PPID</TT
> of a process is
the process ID (<TT
CLASS="VARNAME"
>pid</TT
>) of its parent process.
<A
NAME="AEN5154"
HREF="#FTN.AEN5154"
><SPAN
CLASS="footnote"
>[42]</SPAN
></A
>
</P
><P
>Compare this with the <A
HREF="#PIDOFREF"
>pidof</A
> command.</P
></DD
><DT
><TT
CLASS="VARNAME"
>$PROMPT_COMMAND</TT
></DT
><DD
><P
>A variable holding a command to be executed
just before the primary prompt, <TT
CLASS="VARNAME"
>$PS1</TT
>
is to be displayed.</P
></DD
><DT
><A
NAME="PS1REF"
></A
><TT
CLASS="VARNAME"
>$PS1</TT
></DT
><DD
><P
>This is the main prompt, seen at the command-line.</P
></DD
><DT
><A
NAME="SECPROMPTREF"
></A
><TT
CLASS="VARNAME"
>$PS2</TT
></DT
><DD
><P
>The secondary prompt, seen when additional input is
expected. It displays as <SPAN
CLASS="QUOTE"
>"&#62;"</SPAN
>.</P
></DD
><DT
><TT
CLASS="VARNAME"
>$PS3</TT
></DT
><DD
><P
>The tertiary prompt, displayed in a
<A
HREF="#SELECTREF"
>select</A
> loop (see <A
HREF="#EX31"
>Example 11-30</A
>).</P
></DD
><DT
><TT
CLASS="VARNAME"
>$PS4</TT
></DT
><DD
><P
>The quartenary prompt, shown at the beginning of
each line of output when invoking a script with the
<SPAN
CLASS="TOKEN"
>-x</SPAN
> <EM
>[verbose trace]</EM
>
<A
HREF="#OPTIONSREF"
>option</A
>. It displays as
<SPAN
CLASS="QUOTE"
>"+"</SPAN
>.</P
><P
>As a debugging aid, it may be useful to embed diagnostic
information in <TT
CLASS="VARNAME"
>$PS4</TT
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>P4='$(read time junk &#60; /proc/$$/schedstat; echo "@@@ $time @@@ " )'
# Per suggestion by Erik Brandsberg.
set -x
# Various commands follow ...</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="PWDREF"
></A
><TT
CLASS="VARNAME"
>$PWD</TT
></DT
><DD
><P
>Working directory (directory you are in at the time)</P
><P
>This is the analog to the <A
HREF="#PWD2REF"
>pwd</A
>
builtin command.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
E_WRONG_DIRECTORY=85
clear # Clear the screen.
TargetDirectory=/home/bozo/projects/GreatAmericanNovel
cd $TargetDirectory
echo "Deleting stale files in $TargetDirectory."
if [ "$PWD" != "$TargetDirectory" ]
then # Keep from wiping out wrong directory by accident.
echo "Wrong directory!"
echo "In $PWD, rather than $TargetDirectory!"
echo "Bailing out!"
exit $E_WRONG_DIRECTORY
fi
rm -rf *
rm .[A-Za-z0-9]* # Delete dotfiles.
# rm -f .[^.]* ..?* to remove filenames beginning with multiple dots.
# (shopt -s dotglob; rm -f *) will also work.
# Thanks, S.C. for pointing this out.
# A filename (`basename`) may contain all characters in the 0 - 255 range,
#+ except "/".
# Deleting files beginning with weird characters, such as -
#+ is left as an exercise. (Hint: rm ./-weirdname or rm -- -weirdname)
result=$? # Result of delete operations. If successful = 0.
echo
ls -al # Any files left?
echo "Done."
echo "Old files deleted in $TargetDirectory."
echo
# Various other operations here, as necessary.
exit $result</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="REPLYREF"
></A
><TT
CLASS="VARNAME"
>$REPLY</TT
></DT
><DD
><P
>The default value when a variable is not
supplied to <A
HREF="#READREF"
>read</A
>. Also
applicable to <A
HREF="#SELECTREF"
>select</A
> menus,
but only supplies the item number of the variable chosen,
not the value of the variable itself.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# reply.sh
# REPLY is the default value for a 'read' command.
echo
echo -n "What is your favorite vegetable? "
read
echo "Your favorite vegetable is $REPLY."
# REPLY holds the value of last "read" if and only if
#+ no variable supplied.
echo
echo -n "What is your favorite fruit? "
read fruit
echo "Your favorite fruit is $fruit."
echo "but..."
echo "Value of \$REPLY is still $REPLY."
# $REPLY is still set to its previous value because
#+ the variable $fruit absorbed the new "read" value.
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><TT
CLASS="VARNAME"
>$SECONDS</TT
></DT
><DD
><P
>The number of seconds the script has been running.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
TIME_LIMIT=10
INTERVAL=1
echo
echo "Hit Control-C to exit before $TIME_LIMIT seconds."
echo
while [ "$SECONDS" -le "$TIME_LIMIT" ]
do # $SECONDS is an internal shell variable.
if [ "$SECONDS" -eq 1 ]
then
units=second
else
units=seconds
fi
echo "This script has been running $SECONDS $units."
# On a slow or overburdened machine, the script may skip a count
#+ every once in a while.
sleep $INTERVAL
done
echo -e "\a" # Beep!
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><TT
CLASS="VARNAME"
>$SHELLOPTS</TT
></DT
><DD
><P
>The list of enabled shell <A
HREF="#OPTIONSREF"
>options</A
>, a readonly variable.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $SHELLOPTS</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>braceexpand:hashall:histexpand:monitor:history:interactive-comments:emacs</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="SHLVLREF"
></A
><TT
CLASS="VARNAME"
>$SHLVL</TT
></DT
><DD
><P
>Shell level, how deeply Bash is nested.
<A
NAME="AEN5320"
HREF="#FTN.AEN5320"
><SPAN
CLASS="footnote"
>[43]</SPAN
></A
>
If, at the command-line, $SHLVL is 1, then in a script it
will increment to 2.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This variable is <A
HREF="#SUBSHNLEVREF"
> <EM
>not</EM
> affected by
subshells</A
>. Use <A
HREF="#BASHSUBSHELLREF"
>$BASH_SUBSHELL</A
> when you
need an indication of subshell nesting.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="TMOUTREF"
></A
><TT
CLASS="VARNAME"
>$TMOUT</TT
></DT
><DD
><P
>If the <TT
CLASS="REPLACEABLE"
><I
>$TMOUT</I
></TT
>
environmental variable is set to a non-zero value
<TT
CLASS="VARNAME"
>time</TT
>, then the shell prompt will time out
after <TT
CLASS="VARNAME"
>$time</TT
> seconds. This will cause a
logout.</P
><P
>As of version 2.05b of Bash, it is now possible to use
<TT
CLASS="REPLACEABLE"
><I
>$TMOUT</I
></TT
> in a script in combination
with <A
HREF="#READREF"
>read</A
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Works in scripts for Bash, versions 2.05b and later.
TMOUT=3 # Prompt times out at three seconds.
echo "What is your favorite song?"
echo "Quickly now, you only have $TMOUT seconds to answer!"
read song
if [ -z "$song" ]
then
song="(no answer)"
# Default response.
fi
echo "Your favorite song is $song."</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="TIMINGLOOP"
></A
></P
><P
>There are other, more complex, ways of implementing
timed input in a script. One alternative is to set up
a timing loop to signal the script when it times out.
This also requires a signal handling routine to <A
HREF="#TRAPREF1"
>trap</A
> (see <A
HREF="#EX76"
>Example 32-5</A
>)
the interrupt generated by the timing loop (whew!).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="TMDIN"
></A
><P
><B
>Example 9-2. Timed Input</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# timed-input.sh
# TMOUT=3 Also works, as of newer versions of Bash.
TIMER_INTERRUPT=14
TIMELIMIT=3 # Three seconds in this instance.
# May be set to different value.
PrintAnswer()
{
if [ "$answer" = TIMEOUT ]
then
echo $answer
else # Don't want to mix up the two instances.
echo "Your favorite veggie is $answer"
kill $! # Kills no-longer-needed TimerOn function
#+ running in background.
# $! is PID of last job running in background.
fi
}
TimerOn()
{
sleep $TIMELIMIT &#38;&#38; kill -s 14 $$ &#38;
# Waits 3 seconds, then sends sigalarm to script.
}
Int14Vector()
{
answer="TIMEOUT"
PrintAnswer
exit $TIMER_INTERRUPT
}
trap Int14Vector $TIMER_INTERRUPT
# Timer interrupt (14) subverted for our purposes.
echo "What is your favorite vegetable "
TimerOn
read answer
PrintAnswer
# Admittedly, this is a kludgy implementation of timed input.
# However, the "-t" option to "read" simplifies this task.
# See the "t-out.sh" script.
# However, what about timing not just single user input,
#+ but an entire script?
# If you need something really elegant ...
#+ consider writing the application in C or C++,
#+ using appropriate library functions, such as 'alarm' and 'setitimer.'
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="STTYTO"
></A
></P
><P
>An alternative is using <A
HREF="#STTYREF"
>stty</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="TIMEOUT"
></A
><P
><B
>Example 9-3. Once more, timed input</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# timeout.sh
# Written by Stephane Chazelas,
#+ and modified by the document author.
INTERVAL=5 # timeout interval
timedout_read() {
timeout=$1
varname=$2
old_tty_settings=`stty -g`
stty -icanon min 0 time ${timeout}0
eval read $varname # or just read $varname
stty "$old_tty_settings"
# See man page for "stty."
}
echo; echo -n "What's your name? Quick! "
timedout_read $INTERVAL your_name
# This may not work on every terminal type.
# The maximum timeout depends on the terminal.
#+ (it is often 25.5 seconds).
echo
if [ ! -z "$your_name" ] # If name input before timeout ...
then
echo "Your name is $your_name."
else
echo "Timed out."
fi
echo
# The behavior of this script differs somewhat from "timed-input.sh."
# At each keystroke, the counter resets.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Perhaps the simplest method is using the
<TT
CLASS="OPTION"
>-t</TT
> option to <A
HREF="#READREF"
>read</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="TOUT"
></A
><P
><B
>Example 9-4. Timed <I
CLASS="FIRSTTERM"
>read</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# t-out.sh [time-out]
# Inspired by a suggestion from "syngin seven" (thanks).
TIMELIMIT=4 # 4 seconds
read -t $TIMELIMIT variable &#60;&#38;1
# ^^^
# In this instance, "&#60;&#38;1" is needed for Bash 1.x and 2.x,
# but unnecessary for Bash 3+.
echo
if [ -z "$variable" ] # Is null?
then
echo "Timed out, variable still unset."
else
echo "variable = $variable"
fi
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="UIDREF"
></A
><TT
CLASS="VARNAME"
>$UID</TT
></DT
><DD
><P
>User ID number</P
><P
>Current user's user identification number, as
recorded in <A
HREF="#DATAFILESREF1"
><TT
CLASS="FILENAME"
>/etc/passwd</TT
></A
></P
><P
>This is the current user's real id, even if she has
temporarily assumed another identity through <A
HREF="#SUREF"
>su</A
>. <TT
CLASS="VARNAME"
>$UID</TT
> is a
readonly variable, not subject to change from the command
line or within a script, and is the counterpart to the
<A
HREF="#IDREF"
>id</A
> builtin.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="AMIROOT"
></A
><P
><B
>Example 9-5. Am I root?</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# am-i-root.sh: Am I root or not?
ROOT_UID=0 # Root has $UID 0.
if [ "$UID" -eq "$ROOT_UID" ] # Will the real "root" please stand up?
then
echo "You are root."
else
echo "You are just an ordinary user (but mom loves you just the same)."
fi
exit 0
# ============================================================= #
# Code below will not execute, because the script already exited.
# An alternate method of getting to the root of matters:
ROOTUSER_NAME=root
username=`id -nu` # Or... username=`whoami`
if [ "$username" = "$ROOTUSER_NAME" ]
then
echo "Rooty, toot, toot. You are root."
else
echo "You are just a regular fella."
fi</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#EX2"
>Example 2-3</A
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The variables <TT
CLASS="VARNAME"
>$ENV</TT
>,
<TT
CLASS="VARNAME"
>$LOGNAME</TT
>, <TT
CLASS="VARNAME"
>$MAIL</TT
>,
<TT
CLASS="VARNAME"
>$TERM</TT
>, <TT
CLASS="VARNAME"
>$USER</TT
>, and
<TT
CLASS="VARNAME"
>$USERNAME</TT
> are <EM
>not</EM
>
Bash <A
HREF="#BUILTINREF"
>builtins</A
>. These are,
however, often set as <A
HREF="#ENVREF"
>environmental variables</A
> in
one of the <A
HREF="#FILESREF1"
>Bash</A
> or
<I
CLASS="FIRSTTERM"
>login</I
> startup files. <A
NAME="SHELLVARREF"
></A
><TT
CLASS="VARNAME"
>$SHELL</TT
>,
the name of the user's login shell, may be set from
<TT
CLASS="FILENAME"
>/etc/passwd</TT
> or in an <SPAN
CLASS="QUOTE"
>"init"</SPAN
>
script, and it is likewise not a Bash builtin.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>tcsh% </TT
><TT
CLASS="USERINPUT"
><B
>echo $LOGNAME</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bozo</TT
>
<TT
CLASS="PROMPT"
>tcsh% </TT
><TT
CLASS="USERINPUT"
><B
>echo $SHELL</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/bin/tcsh</TT
>
<TT
CLASS="PROMPT"
>tcsh% </TT
><TT
CLASS="USERINPUT"
><B
>echo $TERM</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>rxvt</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $LOGNAME</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bozo</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $SHELL</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/bin/tcsh</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $TERM</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>rxvt</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
>Positional Parameters</B
></P
><DL
><DT
><A
NAME="POSPARAMREF"
></A
><TT
CLASS="VARNAME"
>$0</TT
>, <TT
CLASS="VARNAME"
>$1</TT
>,
<TT
CLASS="VARNAME"
>$2</TT
>, etc.</DT
><DD
><P
>Positional parameters, passed from command
line to script, passed to a function, or <A
HREF="#SETREF"
>set</A
> to a variable (see <A
HREF="#EX17"
>Example 4-5</A
> and <A
HREF="#EX34"
>Example 15-16</A
>)</P
></DD
><DT
><A
NAME="CLACOUNTREF"
></A
><TT
CLASS="VARNAME"
>$#</TT
></DT
><DD
><P
>Number of command-line arguments
<A
NAME="AEN5479"
HREF="#FTN.AEN5479"
><SPAN
CLASS="footnote"
>[44]</SPAN
></A
>
or positional parameters (see <A
HREF="#EX4"
>Example 36-2</A
>)</P
></DD
><DT
><A
NAME="APPREF"
></A
><TT
CLASS="VARNAME"
>$*</TT
></DT
><DD
><P
>All of the positional parameters, seen as a single word</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><SPAN
CLASS="QUOTE"
>"<TT
CLASS="VARNAME"
>$*</TT
>"</SPAN
> must be
quoted.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="APPREF2"
></A
><TT
CLASS="VARNAME"
>$@</TT
></DT
><DD
><P
>Same as <SPAN
CLASS="TOKEN"
>$*</SPAN
>, but each parameter is a
quoted string, that is, the parameters are passed on
intact, without interpretation or expansion. This means,
among other things, that each parameter in the argument
list is seen as a separate word.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Of course, <SPAN
CLASS="QUOTE"
>"<TT
CLASS="VARNAME"
>$@</TT
>"</SPAN
>
should be quoted.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ARGLIST"
></A
><P
><B
>Example 9-6. <I
CLASS="FIRSTTERM"
>arglist</I
>: Listing arguments
with $* and $@</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# arglist.sh
# Invoke this script with several arguments, such as "one two three" ...
E_BADARGS=85
if [ ! -n "$1" ]
then
echo "Usage: `basename $0` argument1 argument2 etc."
exit $E_BADARGS
fi
echo
index=1 # Initialize count.
echo "Listing args with \"\$*\":"
for arg in "$*" # Doesn't work properly if "$*" isn't quoted.
do
echo "Arg #$index = $arg"
let "index+=1"
done # $* sees all arguments as single word.
echo "Entire arg list seen as single word."
echo
index=1 # Reset count.
# What happens if you forget to do this?
echo "Listing args with \"\$@\":"
for arg in "$@"
do
echo "Arg #$index = $arg"
let "index+=1"
done # $@ sees arguments as separate words.
echo "Arg list seen as separate words."
echo
index=1 # Reset count.
echo "Listing args with \$* (unquoted):"
for arg in $*
do
echo "Arg #$index = $arg"
let "index+=1"
done # Unquoted $* sees arguments as separate words.
echo "Arg list seen as separate words."
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Following a <B
CLASS="COMMAND"
>shift</B
>, the
<TT
CLASS="VARNAME"
>$@</TT
> holds the remaining command-line
parameters, lacking the previous <TT
CLASS="VARNAME"
>$1</TT
>,
which was lost.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Invoke with ./scriptname 1 2 3 4 5
echo "$@" # 1 2 3 4 5
shift
echo "$@" # 2 3 4 5
shift
echo "$@" # 3 4 5
# Each "shift" loses parameter $1.
# "$@" then contains the remaining parameters.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <TT
CLASS="VARNAME"
>$@</TT
> special parameter finds
use as a tool for filtering input into shell scripts. The
<B
CLASS="COMMAND"
>cat "$@"</B
> construction accepts input
to a script either from <TT
CLASS="FILENAME"
>stdin</TT
> or
from files given as parameters to the script. See <A
HREF="#ROT13"
>Example 16-24</A
> and <A
HREF="#CRYPTOQUOTE"
>Example 16-25</A
>.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <TT
CLASS="VARNAME"
>$*</TT
> and <TT
CLASS="VARNAME"
>$@</TT
>
parameters sometimes display inconsistent and
puzzling behavior, depending on the setting of <A
HREF="#IFSREF"
>$IFS</A
>.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="INCOMPAT"
></A
><P
><B
>Example 9-7. Inconsistent <TT
CLASS="VARNAME"
>$*</TT
> and <TT
CLASS="VARNAME"
>$@</TT
> behavior</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Erratic behavior of the "$*" and "$@" internal Bash variables,
#+ depending on whether or not they are quoted.
# Demonstrates inconsistent handling of word splitting and linefeeds.
set -- "First one" "second" "third:one" "" "Fifth: :one"
# Setting the script arguments, $1, $2, $3, etc.
echo
echo 'IFS unchanged, using "$*"'
c=0
for i in "$*" # quoted
do echo "$((c+=1)): [$i]" # This line remains the same in every instance.
# Echo args.
done
echo ---
echo 'IFS unchanged, using $*'
c=0
for i in $* # unquoted
do echo "$((c+=1)): [$i]"
done
echo ---
echo 'IFS unchanged, using "$@"'
c=0
for i in "$@"
do echo "$((c+=1)): [$i]"
done
echo ---
echo 'IFS unchanged, using $@'
c=0
for i in $@
do echo "$((c+=1)): [$i]"
done
echo ---
IFS=:
echo 'IFS=":", using "$*"'
c=0
for i in "$*"
do echo "$((c+=1)): [$i]"
done
echo ---
echo 'IFS=":", using $*'
c=0
for i in $*
do echo "$((c+=1)): [$i]"
done
echo ---
var=$*
echo 'IFS=":", using "$var" (var=$*)'
c=0
for i in "$var"
do echo "$((c+=1)): [$i]"
done
echo ---
echo 'IFS=":", using $var (var=$*)'
c=0
for i in $var
do echo "$((c+=1)): [$i]"
done
echo ---
var="$*"
echo 'IFS=":", using $var (var="$*")'
c=0
for i in $var
do echo "$((c+=1)): [$i]"
done
echo ---
echo 'IFS=":", using "$var" (var="$*")'
c=0
for i in "$var"
do echo "$((c+=1)): [$i]"
done
echo ---
echo 'IFS=":", using "$@"'
c=0
for i in "$@"
do echo "$((c+=1)): [$i]"
done
echo ---
echo 'IFS=":", using $@'
c=0
for i in $@
do echo "$((c+=1)): [$i]"
done
echo ---
var=$@
echo 'IFS=":", using $var (var=$@)'
c=0
for i in $var
do echo "$((c+=1)): [$i]"
done
echo ---
echo 'IFS=":", using "$var" (var=$@)'
c=0
for i in "$var"
do echo "$((c+=1)): [$i]"
done
echo ---
var="$@"
echo 'IFS=":", using "$var" (var="$@")'
c=0
for i in "$var"
do echo "$((c+=1)): [$i]"
done
echo ---
echo 'IFS=":", using $var (var="$@")'
c=0
for i in $var
do echo "$((c+=1)): [$i]"
done
echo
# Try this script with ksh or zsh -y.
exit 0
# This example script written by Stephane Chazelas,
#+ and slightly modified by the document author.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>$@</B
> and <B
CLASS="COMMAND"
>$*</B
>
parameters differ only when between double quotes.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="IFSEMPTY"
></A
><P
><B
>Example 9-8. <TT
CLASS="VARNAME"
>$*</TT
> and <TT
CLASS="VARNAME"
>$@</TT
> when
<TT
CLASS="VARNAME"
>$IFS</TT
> is empty</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# If $IFS set, but empty,
#+ then "$*" and "$@" do not echo positional params as expected.
mecho () # Echo positional parameters.
{
echo "$1,$2,$3";
}
IFS="" # Set, but empty.
set a b c # Positional parameters.
mecho "$*" # abc,,
# ^^
mecho $* # a,b,c
mecho $@ # a,b,c
mecho "$@" # a,b,c
# The behavior of $* and $@ when $IFS is empty depends
#+ on which Bash or sh version being run.
# It is therefore inadvisable to depend on this "feature" in a script.
# Thanks, Stephane Chazelas.
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
>Other Special Parameters</B
></P
><DL
><DT
><A
NAME="FLPREF"
></A
><TT
CLASS="VARNAME"
>$-</TT
></DT
><DD
><P
>Flags passed to script (using <A
HREF="#SETREF"
>set</A
>). See <A
HREF="#EX34"
>Example 15-16</A
>.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This was originally a <I
CLASS="FIRSTTERM"
>ksh</I
>
construct adopted into Bash, and unfortunately it does not
seem to work reliably in Bash scripts. One possible use
for it is to have a script <A
HREF="#IITEST"
>self-test
whether it is interactive</A
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="PIDVARREF"
></A
><TT
CLASS="VARNAME"
>$!</TT
></DT
><DD
><P
><A
HREF="#PROCESSIDDEF"
>PID</A
> (process ID) of last
job run in background</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>LOG=$0.log
COMMAND1="sleep 100"
echo "Logging PIDs background commands for script: $0" &#62;&#62; "$LOG"
# So they can be monitored, and killed as necessary.
echo &#62;&#62; "$LOG"
# Logging commands.
echo -n "PID of \"$COMMAND1\": " &#62;&#62; "$LOG"
${COMMAND1} &#38;
echo $! &#62;&#62; "$LOG"
# PID of "sleep 100": 1506
# Thank you, Jacques Lederer, for suggesting this.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Using <TT
CLASS="VARNAME"
>$!</TT
> for job control:</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>possibly_hanging_job &#38; { sleep ${TIMEOUT}; eval 'kill -9 $!' &#38;&#62; /dev/null; }
# Forces completion of an ill-behaved program.
# Useful, for example, in init scripts.
# Thank you, Sylvain Fourmanoit, for this creative use of the "!" variable.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Or, alternately:</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># This example by Matthew Sage.
# Used with permission.
TIMEOUT=30 # Timeout value in seconds
count=0
possibly_hanging_job &#38; {
while ((count &#60; TIMEOUT )); do
eval '[ ! -d "/proc/$!" ] &#38;&#38; ((count = TIMEOUT))'
# /proc is where information about running processes is found.
# "-d" tests whether it exists (whether directory exists).
# So, we're waiting for the job in question to show up.
((count++))
sleep 1
done
eval '[ -d "/proc/$!" ] &#38;&#38; kill -15 $!'
# If the hanging job is running, kill it.
}
# -------------------------------------------------------------- #
# However, this may not not work as specified if another process
#+ begins to run after the "hanging_job" . . .
# In such a case, the wrong job may be killed.
# Ariel Meragelman suggests the following fix.
TIMEOUT=30
count=0
# Timeout value in seconds
possibly_hanging_job &#38; {
while ((count &#60; TIMEOUT )); do
eval '[ ! -d "/proc/$lastjob" ] &#38;&#38; ((count = TIMEOUT))'
lastjob=$!
((count++))
sleep 1
done
eval '[ -d "/proc/$lastjob" ] &#38;&#38; kill -15 $lastjob'
}
exit</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="UNDERSCOREREF"
></A
><TT
CLASS="VARNAME"
>$_</TT
></DT
><DD
><P
>Special variable set to final argument of previous command
executed.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="USCREF"
></A
><P
><B
>Example 9-9. Underscore variable</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo $_ # /bin/bash
# Just called /bin/bash to run the script.
# Note that this will vary according to
#+ how the script is invoked.
du &#62;/dev/null # So no output from command.
echo $_ # du
ls -al &#62;/dev/null # So no output from command.
echo $_ # -al (last argument)
:
echo $_ # :</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="XSTATVARREF"
></A
><TT
CLASS="VARNAME"
>$?</TT
></DT
><DD
><P
><A
HREF="#EXITSTATUSREF"
>Exit status</A
>
of a command, <A
HREF="#FUNCTIONREF"
>function</A
>,
or the script itself (see <A
HREF="#MAX"
>Example 24-7</A
>)</P
></DD
><DT
><A
NAME="PROCCID"
></A
><TT
CLASS="VARNAME"
>$$</TT
></DT
><DD
><P
>Process ID (<I
CLASS="FIRSTTERM"
>PID</I
>) of
the script itself.
<A
NAME="AEN5654"
HREF="#FTN.AEN5654"
><SPAN
CLASS="footnote"
>[45]</SPAN
></A
>
The <TT
CLASS="VARNAME"
>$$</TT
> variable often
finds use in scripts to construct <SPAN
CLASS="QUOTE"
>"unique"</SPAN
>
temp file names (see <A
HREF="#ONLINE"
>Example 32-6</A
>, <A
HREF="#DERPM"
>Example 16-31</A
>, and <A
HREF="#SELFDESTRUCT"
>Example 15-27</A
>).
This is usually simpler than invoking <A
HREF="#MKTEMPREF"
>mktemp</A
>.</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="DECLAREREF"
></A
>9.2. Typing variables: <B
CLASS="COMMAND"
>declare</B
> or
<B
CLASS="COMMAND"
>typeset</B
></H1
><P
><A
NAME="DECLARE1REF"
></A
></P
><P
>The <I
CLASS="FIRSTTERM"
>declare</I
> or
<I
CLASS="FIRSTTERM"
>typeset</I
> <A
HREF="#BUILTINREF"
>builtins</A
>, which are exact synonyms,
permit modifying the properties of variables. This is
a very weak form of the <I
CLASS="FIRSTTERM"
>typing</I
>
<A
NAME="AEN5685"
HREF="#FTN.AEN5685"
><SPAN
CLASS="footnote"
>[46]</SPAN
></A
>
available in certain programming languages. The
<I
CLASS="FIRSTTERM"
>declare</I
> command is specific to version
2 or later of Bash. The <I
CLASS="FIRSTTERM"
>typeset</I
> command
also works in ksh scripts.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="DECLAREOPSREF1"
></A
>declare/typeset options</B
></P
><DL
><DT
><SPAN
CLASS="TOKEN"
>-r</SPAN
>
<TT
CLASS="REPLACEABLE"
><I
>readonly</I
></TT
></DT
><DD
><P
>(<TT
CLASS="USERINPUT"
><B
>declare -r var1</B
></TT
> works the same as
<TT
CLASS="USERINPUT"
><B
>readonly var1</B
></TT
>)</P
><P
>This is the rough equivalent of the <B
CLASS="COMMAND"
>C</B
>
<I
CLASS="FIRSTTERM"
>const</I
> type qualifier. An attempt
to change the value of a <I
CLASS="FIRSTTERM"
>readonly</I
>
variable fails with an error message.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>declare -r var1=1
echo "var1 = $var1" # var1 = 1
(( var1++ )) # x.sh: line 4: var1: readonly variable</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-i</SPAN
> <TT
CLASS="REPLACEABLE"
><I
>integer</I
></TT
></DT
><DD
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>declare -i number
# The script will treat subsequent occurrences of "number" as an integer.
number=3
echo "Number = $number" # Number = 3
number=three
echo "Number = $number" # Number = 0
# Tries to evaluate the string "three" as an integer.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Certain arithmetic operations are permitted
for declared integer variables without the need
for <A
HREF="#EXPRREF"
>expr</A
> or <A
HREF="#LETREF"
>let</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>n=6/3
echo "n = $n" # n = 6/3
declare -i n
n=6/3
echo "n = $n" # n = 2</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="ARRAYDECLARE"
></A
><SPAN
CLASS="TOKEN"
>-a</SPAN
>
<TT
CLASS="REPLACEABLE"
><I
>array</I
></TT
></DT
><DD
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>declare -a indices</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The variable <TT
CLASS="PARAMETER"
><I
>indices</I
></TT
> will be treated as
an <A
HREF="#ARRAYREF"
>array</A
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-f</SPAN
> <TT
CLASS="REPLACEABLE"
><I
>function(s)</I
></TT
></DT
><DD
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>declare -f</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>A <TT
CLASS="USERINPUT"
><B
>declare -f</B
></TT
> line with no
arguments in a script causes a listing of all the
<A
HREF="#FUNCTIONREF"
>functions</A
> previously
defined in that script.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>declare -f function_name</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>A <TT
CLASS="USERINPUT"
><B
>declare -f function_name</B
></TT
>
in a script lists just the function named.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>-x</SPAN
> <A
HREF="#EXPORTREF"
>export</A
></DT
><DD
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>declare -x var3</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>This declares a variable as available for exporting outside the
environment of the script itself.</P
></DD
><DT
>-x var=$value</DT
><DD
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>declare -x var3=373</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The <B
CLASS="COMMAND"
>declare</B
> command permits
assigning a value to a variable in the same statement
as setting its properties.</P
></DD
></DL
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX20"
></A
><P
><B
>Example 9-10. Using <I
CLASS="FIRSTTERM"
>declare</I
> to type variables</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
func1 ()
{
echo This is a function.
}
declare -f # Lists the function above.
echo
declare -i var1 # var1 is an integer.
var1=2367
echo "var1 declared as $var1"
var1=var1+1 # Integer declaration eliminates the need for 'let'.
echo "var1 incremented by 1 is $var1."
# Attempt to change variable declared as integer.
echo "Attempting to change var1 to floating point value, 2367.1."
var1=2367.1 # Results in error message, with no change to variable.
echo "var1 is still $var1"
echo
declare -r var2=13.36 # 'declare' permits setting a variable property
#+ and simultaneously assigning it a value.
echo "var2 declared as $var2" # Attempt to change readonly variable.
var2=13.37 # Generates error message, and exit from script.
echo "var2 is still $var2" # This line will not execute.
exit 0 # Script will not exit here.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Using the <I
CLASS="FIRSTTERM"
>declare</I
> builtin
restricts the <A
HREF="#SCOPEREF"
>scope</A
>
of a variable.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>foo ()
{
FOO="bar"
}
bar ()
{
foo
echo $FOO
}
bar # Prints bar.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>However . . .
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>foo (){
declare FOO="bar"
}
bar ()
{
foo
echo $FOO
}
bar # Prints nothing.
# Thank you, Michael Iatrou, for pointing this out.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="DECLARE2X"
></A
>9.2.1. Another use for <I
CLASS="FIRSTTERM"
>declare</I
></H2
><P
>The <I
CLASS="FIRSTTERM"
>declare</I
> command can be
helpful in identifying variables, <A
HREF="#ENVREF"
>environmental</A
> or otherwise.
This can be especially useful with <A
HREF="#ARRAYREF"
>arrays</A
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>declare | grep HOME</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>HOME=/home/bozo</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>zzy=68</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>declare | grep zzy</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>zzy=68</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>Colors=([0]="purple" [1]="reddish-orange" [2]="light green")</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo ${Colors[@]}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>purple reddish-orange light green</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>declare | grep Colors</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Colors=([0]="purple" [1]="reddish-orange" [2]="light green")</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="RANDOMVAR"
></A
>9.3. $RANDOM: generate random integer</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>Anyone who attempts to generate random numbers by
deterministic means is, of course, living in a state of
sin.</I
></P
><P
><I
>--John von Neumann</I
></P
></I
></TD
></TR
></TABLE
><P
><A
NAME="RANDOMVAR01"
></A
></P
><P
><TT
CLASS="VARNAME"
>$RANDOM</TT
> is an internal Bash <A
HREF="#FUNCTIONREF"
>function</A
> (not a constant) that
returns a <I
CLASS="FIRSTTERM"
>pseudorandom</I
>
<A
NAME="AEN5817"
HREF="#FTN.AEN5817"
><SPAN
CLASS="footnote"
>[47]</SPAN
></A
>
integer in the range 0 - 32767. It should
<TT
CLASS="REPLACEABLE"
><I
>not</I
></TT
> be used to generate an encryption
key.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX21"
></A
><P
><B
>Example 9-11. Generating random numbers</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# $RANDOM returns a different random integer at each invocation.
# Nominal range: 0 - 32767 (signed 16-bit integer).
MAXCOUNT=10
count=1
echo
echo "$MAXCOUNT random numbers:"
echo "-----------------"
while [ "$count" -le $MAXCOUNT ] # Generate 10 ($MAXCOUNT) random integers.
do
number=$RANDOM
echo $number
let "count += 1" # Increment count.
done
echo "-----------------"
# If you need a random int within a certain range, use the 'modulo' operator.
# This returns the remainder of a division operation.
RANGE=500
echo
number=$RANDOM
let "number %= $RANGE"
# ^^
echo "Random number less than $RANGE --- $number"
echo
# If you need a random integer greater than a lower bound,
#+ then set up a test to discard all numbers below that.
FLOOR=200
number=0 #initialize
while [ "$number" -le $FLOOR ]
do
number=$RANDOM
done
echo "Random number greater than $FLOOR --- $number"
echo
# Let's examine a simple alternative to the above loop, namely
# let "number = $RANDOM + $FLOOR"
# That would eliminate the while-loop and run faster.
# But, there might be a problem with that. What is it?
# Combine above two techniques to retrieve random number between two limits.
number=0 #initialize
while [ "$number" -le $FLOOR ]
do
number=$RANDOM
let "number %= $RANGE" # Scales $number down within $RANGE.
done
echo "Random number between $FLOOR and $RANGE --- $number"
echo
# Generate binary choice, that is, "true" or "false" value.
BINARY=2
T=1
number=$RANDOM
let "number %= $BINARY"
# Note that let "number &#62;&#62;= 14" gives a better random distribution
#+ (right shifts out everything except last binary digit).
if [ "$number" -eq $T ]
then
echo "TRUE"
else
echo "FALSE"
fi
echo
# Generate a toss of the dice.
SPOTS=6 # Modulo 6 gives range 0 - 5.
# Incrementing by 1 gives desired range of 1 - 6.
# Thanks, Paulo Marcel Coelho Aragao, for the simplification.
die1=0
die2=0
# Would it be better to just set SPOTS=7 and not add 1? Why or why not?
# Tosses each die separately, and so gives correct odds.
let "die1 = $RANDOM % $SPOTS +1" # Roll first one.
let "die2 = $RANDOM % $SPOTS +1" # Roll second one.
# Which arithmetic operation, above, has greater precedence --
#+ modulo (%) or addition (+)?
let "throw = $die1 + $die2"
echo "Throw of the dice = $throw"
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="PICKCARD"
></A
><P
><B
>Example 9-12. Picking a random card from a deck</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# pick-card.sh
# This is an example of choosing random elements of an array.
# Pick a card, any card.
Suites="Clubs
Diamonds
Hearts
Spades"
Denominations="2
3
4
5
6
7
8
9
10
Jack
Queen
King
Ace"
# Note variables spread over multiple lines.
suite=($Suites) # Read into array variable.
denomination=($Denominations)
num_suites=${#suite[*]} # Count how many elements.
num_denominations=${#denomination[*]}
echo -n "${denomination[$((RANDOM%num_denominations))]} of "
echo ${suite[$((RANDOM%num_suites))]}
# $bozo sh pick-cards.sh
# Jack of Clubs
# Thank you, "jipe," for pointing out this use of $RANDOM.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="BROWNIANREF"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="BROWNIAN"
></A
><P
><B
>Example 9-13. Brownian Motion Simulation</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# brownian.sh
# Author: Mendel Cooper
# Reldate: 10/26/07
# License: GPL3
# ----------------------------------------------------------------
# This script models Brownian motion:
#+ the random wanderings of tiny particles in a fluid,
#+ as they are buffeted by random currents and collisions.
#+ This is colloquially known as the "Drunkard's Walk."
# It can also be considered as a stripped-down simulation of a
#+ Galton Board, a slanted board with a pattern of pegs,
#+ down which rolls a succession of marbles, one at a time.
#+ At the bottom is a row of slots or catch basins in which
#+ the marbles come to rest at the end of their journey.
# Think of it as a kind of bare-bones Pachinko game.
# As you see by running the script,
#+ most of the marbles cluster around the center slot.
#+ This is consistent with the expected binomial distribution.
# As a Galton Board simulation, the script
#+ disregards such parameters as
#+ board tilt-angle, rolling friction of the marbles,
#+ angles of impact, and elasticity of the pegs.
# To what extent does this affect the accuracy of the simulation?
# ----------------------------------------------------------------
PASSES=500 # Number of particle interactions / marbles.
ROWS=10 # Number of "collisions" (or horiz. peg rows).
RANGE=3 # 0 - 2 output range from $RANDOM.
POS=0 # Left/right position.
RANDOM=$$ # Seeds the random number generator from PID
#+ of script.
declare -a Slots # Array holding cumulative results of passes.
NUMSLOTS=21 # Number of slots at bottom of board.
Initialize_Slots () { # Zero out all elements of the array.
for i in $( seq $NUMSLOTS )
do
Slots[$i]=0
done
echo # Blank line at beginning of run.
}
Show_Slots () {
echo; echo
echo -n " "
for i in $( seq $NUMSLOTS ) # Pretty-print array elements.
do
printf "%3d" ${Slots[$i]} # Allot three spaces per result.
done
echo # Row of slots:
echo " |__|__|__|__|__|__|__|__|__|__|__|__|__|__|__|__|__|__|__|__|__|"
echo " ||"
echo # Note that if the count within any particular slot exceeds 99,
#+ it messes up the display.
# Running only(!) 500 passes usually avoids this.
}
Move () { # Move one unit right / left, or stay put.
Move=$RANDOM # How random is $RANDOM? Well, let's see ...
let "Move %= RANGE" # Normalize into range of 0 - 2.
case "$Move" in
0 ) ;; # Do nothing, i.e., stay in place.
1 ) ((POS--));; # Left.
2 ) ((POS++));; # Right.
* ) echo -n "Error ";; # Anomaly! (Should never occur.)
esac
}
Play () { # Single pass (inner loop).
i=0
while [ "$i" -lt "$ROWS" ] # One event per row.
do
Move
((i++));
done
SHIFT=11 # Why 11, and not 10?
let "POS += $SHIFT" # Shift "zero position" to center.
(( Slots[$POS]++ )) # DEBUG: echo $POS
# echo -n "$POS "
}
Run () { # Outer loop.
p=0
while [ "$p" -lt "$PASSES" ]
do
Play
(( p++ ))
POS=0 # Reset to zero. Why?
done
}
# --------------
# main ()
Initialize_Slots
Run
Show_Slots
# --------------
exit $?
# Exercises:
# ---------
# 1) Show the results in a vertical bar graph, or as an alternative,
#+ a scattergram.
# 2) Alter the script to use /dev/urandom instead of $RANDOM.
# Will this make the results more random?
# 3) Provide some sort of "animation" or graphic output
# for each marble played.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
> <EM
>Jipe</EM
> points out a set of techniques for
generating random numbers within a range.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Generate random number between 6 and 30.
rnumber=$((RANDOM%25+6))
# Generate random number in the same 6 - 30 range,
#+ but the number must be evenly divisible by 3.
rnumber=$(((RANDOM%30/3+1)*3))
# Note that this will not work all the time.
# It fails if $RANDOM%30 returns 0.
# Frank Wang suggests the following alternative:
rnumber=$(( RANDOM%27/3*3+6 ))</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> <EM
>Bill Gradwohl</EM
> came up with an improved
formula that works for positive numbers.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>rnumber=$(((RANDOM%(max-min+divisibleBy))/divisibleBy*divisibleBy+min))</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Here Bill presents a versatile function that returns
a random number between two specified values.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="RANDOMBETWEEN"
></A
><P
><B
>Example 9-14. Random between values</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# random-between.sh
# Random number between two specified values.
# Script by Bill Gradwohl, with minor modifications by the document author.
# Corrections in lines 187 and 189 by Anthony Le Clezio.
# Used with permission.
randomBetween() {
# Generates a positive or negative random number
#+ between $min and $max
#+ and divisible by $divisibleBy.
# Gives a "reasonably random" distribution of return values.
#
# Bill Gradwohl - Oct 1, 2003
syntax() {
# Function embedded within function.
echo
echo "Syntax: randomBetween [min] [max] [multiple]"
echo
echo -n "Expects up to 3 passed parameters, "
echo "but all are completely optional."
echo "min is the minimum value"
echo "max is the maximum value"
echo -n "multiple specifies that the answer must be "
echo "a multiple of this value."
echo " i.e. answer must be evenly divisible by this number."
echo
echo "If any value is missing, defaults area supplied as: 0 32767 1"
echo -n "Successful completion returns 0, "
echo "unsuccessful completion returns"
echo "function syntax and 1."
echo -n "The answer is returned in the global variable "
echo "randomBetweenAnswer"
echo -n "Negative values for any passed parameter are "
echo "handled correctly."
}
local min=${1:-0}
local max=${2:-32767}
local divisibleBy=${3:-1}
# Default values assigned, in case parameters not passed to function.
local x
local spread
# Let's make sure the divisibleBy value is positive.
[ ${divisibleBy} -lt 0 ] &#38;&#38; divisibleBy=$((0-divisibleBy))
# Sanity check.
if [ $# -gt 3 -o ${divisibleBy} -eq 0 -o ${min} -eq ${max} ]; then
syntax
return 1
fi
# See if the min and max are reversed.
if [ ${min} -gt ${max} ]; then
# Swap them.
x=${min}
min=${max}
max=${x}
fi
# If min is itself not evenly divisible by $divisibleBy,
#+ then fix the min to be within range.
if [ $((min/divisibleBy*divisibleBy)) -ne ${min} ]; then
if [ ${min} -lt 0 ]; then
min=$((min/divisibleBy*divisibleBy))
else
min=$((((min/divisibleBy)+1)*divisibleBy))
fi
fi
# If max is itself not evenly divisible by $divisibleBy,
#+ then fix the max to be within range.
if [ $((max/divisibleBy*divisibleBy)) -ne ${max} ]; then
if [ ${max} -lt 0 ]; then
max=$((((max/divisibleBy)-1)*divisibleBy))
else
max=$((max/divisibleBy*divisibleBy))
fi
fi
# ---------------------------------------------------------------------
# Now, to do the real work.
# Note that to get a proper distribution for the end points,
#+ the range of random values has to be allowed to go between
#+ 0 and abs(max-min)+divisibleBy, not just abs(max-min)+1.
# The slight increase will produce the proper distribution for the
#+ end points.
# Changing the formula to use abs(max-min)+1 will still produce
#+ correct answers, but the randomness of those answers is faulty in
#+ that the number of times the end points ($min and $max) are returned
#+ is considerably lower than when the correct formula is used.
# ---------------------------------------------------------------------
spread=$((max-min))
# Omair Eshkenazi points out that this test is unnecessary,
#+ since max and min have already been switched around.
[ ${spread} -lt 0 ] &#38;&#38; spread=$((0-spread))
let spread+=divisibleBy
randomBetweenAnswer=$(((RANDOM%spread)/divisibleBy*divisibleBy+min))
return 0
# However, Paulo Marcel Coelho Aragao points out that
#+ when $max and $min are not divisible by $divisibleBy,
#+ the formula fails.
#
# He suggests instead the following formula:
# rnumber = $(((RANDOM%(max-min+1)+min)/divisibleBy*divisibleBy))
}
# Let's test the function.
min=-14
max=20
divisibleBy=3
# Generate an array of expected answers and check to make sure we get
#+ at least one of each answer if we loop long enough.
declare -a answer
minimum=${min}
maximum=${max}
if [ $((minimum/divisibleBy*divisibleBy)) -ne ${minimum} ]; then
if [ ${minimum} -lt 0 ]; then
minimum=$((minimum/divisibleBy*divisibleBy))
else
minimum=$((((minimum/divisibleBy)+1)*divisibleBy))
fi
fi
# If max is itself not evenly divisible by $divisibleBy,
#+ then fix the max to be within range.
if [ $((maximum/divisibleBy*divisibleBy)) -ne ${maximum} ]; then
if [ ${maximum} -lt 0 ]; then
maximum=$((((maximum/divisibleBy)-1)*divisibleBy))
else
maximum=$((maximum/divisibleBy*divisibleBy))
fi
fi
# We need to generate only positive array subscripts,
#+ so we need a displacement that that will guarantee
#+ positive results.
disp=$((0-minimum))
for ((i=${minimum}; i&#60;=${maximum}; i+=divisibleBy)); do
answer[i+disp]=0
done
# Now loop a large number of times to see what we get.
loopIt=1000 # The script author suggests 100000,
#+ but that takes a good long while.
for ((i=0; i&#60;${loopIt}; ++i)); do
# Note that we are specifying min and max in reversed order here to
#+ make the function correct for this case.
randomBetween ${max} ${min} ${divisibleBy}
# Report an error if an answer is unexpected.
[ ${randomBetweenAnswer} -lt ${min} -o ${randomBetweenAnswer} -gt ${max} ] \
&#38;&#38; echo MIN or MAX error - ${randomBetweenAnswer}!
[ $((randomBetweenAnswer%${divisibleBy})) -ne 0 ] \
&#38;&#38; echo DIVISIBLE BY error - ${randomBetweenAnswer}!
# Store the answer away statistically.
answer[randomBetweenAnswer+disp]=$((answer[randomBetweenAnswer+disp]+1))
done
# Let's check the results
for ((i=${minimum}; i&#60;=${maximum}; i+=divisibleBy)); do
[ ${answer[i+disp]} -eq 0 ] \
&#38;&#38; echo "We never got an answer of $i." \
|| echo "${i} occurred ${answer[i+disp]} times."
done
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Just how random is <TT
CLASS="VARNAME"
>$RANDOM</TT
>? The best
way to test this is to write a script that tracks
the distribution of <SPAN
CLASS="QUOTE"
>"random"</SPAN
> numbers
generated by <TT
CLASS="VARNAME"
>$RANDOM</TT
>. Let's roll a
<TT
CLASS="VARNAME"
>$RANDOM</TT
> die a few times . . .</P
><DIV
CLASS="EXAMPLE"
><A
NAME="RANDOMTEST"
></A
><P
><B
>Example 9-15. Rolling a single die with RANDOM</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# How random is RANDOM?
RANDOM=$$ # Reseed the random number generator using script process ID.
PIPS=6 # A die has 6 pips.
MAXTHROWS=600 # Increase this if you have nothing better to do with your time.
throw=0 # Number of times the dice have been cast.
ones=0 # Must initialize counts to zero,
twos=0 #+ since an uninitialized variable is null, NOT zero.
threes=0
fours=0
fives=0
sixes=0
print_result ()
{
echo
echo "ones = $ones"
echo "twos = $twos"
echo "threes = $threes"
echo "fours = $fours"
echo "fives = $fives"
echo "sixes = $sixes"
echo
}
update_count()
{
case "$1" in
0) ((ones++));; # Since a die has no "zero", this corresponds to 1.
1) ((twos++));; # And this to 2.
2) ((threes++));; # And so forth.
3) ((fours++));;
4) ((fives++));;
5) ((sixes++));;
esac
}
echo
while [ "$throw" -lt "$MAXTHROWS" ]
do
let "die1 = RANDOM % $PIPS"
update_count $die1
let "throw += 1"
done
print_result
exit $?
# The scores should distribute evenly, assuming RANDOM is random.
# With $MAXTHROWS at 600, all should cluster around 100,
#+ plus-or-minus 20 or so.
#
# Keep in mind that RANDOM is a ***pseudorandom*** generator,
#+ and not a spectacularly good one at that.
# Randomness is a deep and complex subject.
# Sufficiently long "random" sequences may exhibit
#+ chaotic and other "non-random" behavior.
# Exercise (easy):
# ---------------
# Rewrite this script to flip a coin 1000 times.
# Choices are "HEADS" and "TAILS."</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>As we have seen in the last example, it is best to
<I
CLASS="FIRSTTERM"
>reseed</I
> the <TT
CLASS="PARAMETER"
><I
>RANDOM</I
></TT
>
generator each time it is invoked. Using the same seed
for <TT
CLASS="PARAMETER"
><I
>RANDOM</I
></TT
> repeats the same series
of numbers.
<A
NAME="AEN5857"
HREF="#FTN.AEN5857"
><SPAN
CLASS="footnote"
>[48]</SPAN
></A
>
(This mirrors the behavior of the
<TT
CLASS="REPLACEABLE"
><I
>random()</I
></TT
> function in
<I
CLASS="FIRSTTERM"
>C</I
>.)</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SEEDINGRANDOM"
></A
><P
><B
>Example 9-16. Reseeding RANDOM</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# seeding-random.sh: Seeding the RANDOM variable.
# v 1.1, reldate 09 Feb 2013
MAXCOUNT=25 # How many numbers to generate.
SEED=
random_numbers ()
{
local count=0
local number
while [ "$count" -lt "$MAXCOUNT" ]
do
number=$RANDOM
echo -n "$number "
let "count++"
done
}
echo; echo
SEED=1
RANDOM=$SEED # Setting RANDOM seeds the random number generator.
echo "Random seed = $SEED"
random_numbers
RANDOM=$SEED # Same seed for RANDOM . . .
echo; echo "Again, with same random seed ..."
echo "Random seed = $SEED"
random_numbers # . . . reproduces the exact same number series.
#
# When is it useful to duplicate a "random" series?
echo; echo
SEED=2
RANDOM=$SEED # Trying again, but with a different seed . . .
echo "Random seed = $SEED"
random_numbers # . . . gives a different number series.
echo; echo
# RANDOM=$$ seeds RANDOM from process id of script.
# It is also possible to seed RANDOM from 'time' or 'date' commands.
# Getting fancy...
SEED=$(head -1 /dev/urandom | od -N 1 | awk '{ print $2 }'| sed s/^0*//)
# Pseudo-random output fetched
#+ from /dev/urandom (system pseudo-random device-file),
#+ then converted to line of printable (octal) numbers by "od",
#+ then "awk" retrieves just one number for SEED,
#+ finally "sed" removes any leading zeros.
RANDOM=$SEED
echo "Random seed = $SEED"
random_numbers
echo; echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="URANDOMREF"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <TT
CLASS="FILENAME"
>/dev/urandom</TT
> pseudo-device file
provides a method of generating much more <SPAN
CLASS="QUOTE"
>"random"</SPAN
>
pseudorandom numbers than the <TT
CLASS="VARNAME"
>$RANDOM</TT
>
variable. <TT
CLASS="USERINPUT"
><B
>dd if=/dev/urandom of=targetfile
bs=1 count=XX</B
></TT
> creates a file of well-scattered
pseudorandom numbers. However, assigning these numbers
to a variable in a script requires a workaround, such
as filtering through <A
HREF="#ODREF"
>od</A
>
(as in above example, <A
HREF="#RND"
>Example 16-14</A
>, and
<A
HREF="#INSERTIONSORT"
>Example A-36</A
>), or even piping to
<A
HREF="#MD5SUMREF"
>md5sum</A
> (see <A
HREF="#HORSERACE"
>Example 36-16</A
>).</P
><P
><A
NAME="AWKRANDOMREF"
></A
></P
><P
>There are also other ways to generate pseudorandom
numbers in a script. <B
CLASS="COMMAND"
>Awk</B
> provides a
convenient means of doing this.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="RANDOM2"
></A
><P
><B
>Example 9-17. Pseudorandom numbers, using <A
HREF="#AWKREF"
>awk</A
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# random2.sh: Returns a pseudorandom number in the range 0 - 1,
#+ to 6 decimal places. For example: 0.822725
# Uses the awk rand() function.
AWKSCRIPT=' { srand(); print rand() } '
# Command(s)/parameters passed to awk
# Note that srand() reseeds awk's random number generator.
echo -n "Random number between 0 and 1 = "
echo | awk "$AWKSCRIPT"
# What happens if you leave out the 'echo'?
exit 0
# Exercises:
# ---------
# 1) Using a loop construct, print out 10 different random numbers.
# (Hint: you must reseed the srand() function with a different seed
#+ in each pass through the loop. What happens if you omit this?)
# 2) Using an integer multiplier as a scaling factor, generate random numbers
#+ in the range of 10 to 100.
# 3) Same as exercise #2, above, but generate random integers this time.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The <A
HREF="#DATEREF"
>date</A
> command also lends
itself to <A
HREF="#DATERANDREF"
>generating pseudorandom
integer sequences</A
>.</P
></TD
></TR
></TABLE
></DIV
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="MANIPULATINGVARS"
></A
>Chapter 10. Manipulating Variables</H1
><DIV
CLASS="SECT1"
><H1
CLASS="SECT1"
><A
NAME="STRING-MANIPULATION"
></A
>10.1. Manipulating Strings</H1
><P
><A
NAME="STRINGMANIP"
></A
></P
><P
>Bash supports a surprising number of string manipulation
operations. Unfortunately, these tools lack
a unified focus. Some are a subset of <A
HREF="#PARAMSUBREF"
>parameter substitution</A
>, and
others fall under the functionality of the UNIX <A
HREF="#EXPRREF"
>expr</A
> command. This results in
inconsistent command syntax and overlap of functionality,
not to mention confusion.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
>String Length</B
></P
><DL
><DT
>${#string}</DT
><DD
><P
></P
></DD
><DT
>expr length $string</DT
><DD
><P
><A
NAME="STRLEN"
></A
>These are the equivalent of
<I
CLASS="FIRSTTERM"
>strlen()</I
> in
<I
CLASS="FIRSTTERM"
>C</I
>.</P
></DD
><DT
>expr "$string" : '.*'</DT
><DD
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>stringZ=abcABC123ABCabc
echo ${#stringZ} # 15
echo `expr length $stringZ` # 15
echo `expr "$stringZ" : '.*'` # 15</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
></DL
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="PARAGRAPHSPACE"
></A
><P
><B
>Example 10-1. Inserting a blank line between paragraphs in a text file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# paragraph-space.sh
# Ver. 2.1, Reldate 29Jul12 [fixup]
# Inserts a blank line between paragraphs of a single-spaced text file.
# Usage: $0 &#60;FILENAME
MINLEN=60 # Change this value? It's a judgment call.
# Assume lines shorter than $MINLEN characters ending in a period
#+ terminate a paragraph. See exercises below.
while read line # For as many lines as the input file has ...
do
echo "$line" # Output the line itself.
len=${#line}
if [[ "$len" -lt "$MINLEN" &#38;&#38; "$line" =~ [*{\.}]$ ]]
# if [[ "$len" -lt "$MINLEN" &#38;&#38; "$line" =~ \[*\.\] ]]
# An update to Bash broke the previous version of this script. Ouch!
# Thank you, Halim Srama, for pointing this out and suggesting a fix.
then echo # Add a blank line immediately
fi #+ after a short line terminated by a period.
done
exit
# Exercises:
# ---------
# 1) The script usually inserts a blank line at the end
#+ of the target file. Fix this.
# 2) Line 17 only considers periods as sentence terminators.
# Modify this to include other common end-of-sentence characters,
#+ such as ?, !, and ".</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
>Length of Matching Substring at Beginning of String</B
></P
><DL
><DT
><A
NAME="EXPRMATCH"
></A
>expr match "$string"
'$substring'</DT
><DD
><P
><TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> is a <A
HREF="#REGEXREF"
>regular expression</A
>.</P
></DD
><DT
>expr "$string" : '$substring'</DT
><DD
><P
><TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> is a regular
expression.</P
><P
>&#13; <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>stringZ=abcABC123ABCabc
# |------|
# 12345678
echo `expr match "$stringZ" 'abc[A-Z]*.2'` # 8
echo `expr "$stringZ" : 'abc[A-Z]*.2'` # 8</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
>Index</B
></P
><DL
><DT
><A
NAME="SUBSTRINGINDEX2"
></A
>expr index $string
$substring</DT
><DD
><P
>Numerical position in $string of first character in
$substring that matches.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>stringZ=abcABC123ABCabc
# 123456 ...
echo `expr index "$stringZ" C12` # 6
# C position.
echo `expr index "$stringZ" 1c` # 3
# 'c' (in #3 position) matches before '1'.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>This is the near equivalent of
<I
CLASS="FIRSTTERM"
>strchr()</I
> in
<I
CLASS="FIRSTTERM"
>C</I
>.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
>Substring Extraction</B
></P
><DL
><DT
><A
NAME="SUBSTREXTR01"
></A
>${string:position}</DT
><DD
><P
>Extracts substring from <TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
> at
<TT
CLASS="REPLACEABLE"
><I
>$position</I
></TT
>.</P
><P
>If the <TT
CLASS="VARNAME"
>$string</TT
> parameter is
<SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>*</SPAN
>"</SPAN
>
or <SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>@</SPAN
>"</SPAN
>, then this extracts the
<A
HREF="#POSPARAMREF"
>positional parameters</A
>,
<A
NAME="AEN5987"
HREF="#FTN.AEN5987"
><SPAN
CLASS="footnote"
>[49]</SPAN
></A
>
starting at <TT
CLASS="VARNAME"
>$position</TT
>.</P
></DD
><DT
><A
NAME="SUBSTREXTR02"
></A
>${string:position:length}</DT
><DD
><P
>Extracts <TT
CLASS="REPLACEABLE"
><I
>$length</I
></TT
> characters
of substring from <TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
> at
<TT
CLASS="REPLACEABLE"
><I
>$position</I
></TT
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>stringZ=abcABC123ABCabc
# 0123456789.....
# 0-based indexing.
echo ${stringZ:0} # abcABC123ABCabc
echo ${stringZ:1} # bcABC123ABCabc
echo ${stringZ:7} # 23ABCabc
echo ${stringZ:7:3} # 23A
# Three characters of substring.
# Is it possible to index from the right end of the string?
echo ${stringZ:-4} # abcABC123ABCabc
# Defaults to full string, as in ${parameter:-default}.
# However . . .
echo ${stringZ:(-4)} # Cabc
echo ${stringZ: -4} # Cabc
# Now, it works.
# Parentheses or added space "escape" the position parameter.
# Thank you, Dan Jacobson, for pointing this out.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <I
CLASS="FIRSTTERM"
>position</I
> and
<I
CLASS="FIRSTTERM"
>length</I
> arguments can be
<SPAN
CLASS="QUOTE"
>"parameterized,"</SPAN
> that is, represented as a
variable, rather than as a numerical constant.</P
><P
><A
NAME="RANDSTRING0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="RANDSTRING"
></A
><P
><B
>Example 10-2. Generating an 8-character <SPAN
CLASS="QUOTE"
>"random"</SPAN
>
string</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# rand-string.sh
# Generating an 8-character "random" string.
if [ -n "$1" ] # If command-line argument present,
then #+ then set start-string to it.
str0="$1"
else # Else use PID of script as start-string.
str0="$$"
fi
POS=2 # Starting from position 2 in the string.
LEN=8 # Extract eight characters.
str1=$( echo "$str0" | md5sum | md5sum )
# Doubly scramble ^^^^^^ ^^^^^^
#+ by piping and repiping to md5sum.
randstring="${str1:$POS:$LEN}"
# Can parameterize ^^^^ ^^^^
echo "$randstring"
exit $?
# bozo$ ./rand-string.sh my-password
# 1bdd88c4
# No, this is is not recommended
#+ as a method of generating hack-proof passwords.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="SUBSTREXTRP"
></A
></P
><P
>If the <TT
CLASS="VARNAME"
>$string</TT
> parameter is
<SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>*</SPAN
>"</SPAN
> or
<SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>@</SPAN
>"</SPAN
>, then this extracts a maximum
of <TT
CLASS="VARNAME"
>$length</TT
> positional parameters, starting
at <TT
CLASS="VARNAME"
>$position</TT
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo ${*:2} # Echoes second and following positional parameters.
echo ${@:2} # Same as above.
echo ${*:2:3} # Echoes three positional parameters, starting at second.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
>expr substr $string $position $length</DT
><DD
><P
>Extracts <TT
CLASS="REPLACEABLE"
><I
>$length</I
></TT
> characters
from <TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
> starting at
<TT
CLASS="REPLACEABLE"
><I
>$position</I
></TT
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>stringZ=abcABC123ABCabc
# 123456789......
# 1-based indexing.
echo `expr substr $stringZ 1 2` # ab
echo `expr substr $stringZ 4 3` # ABC</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="EXPRPAREN"
></A
></P
></DD
><DT
>expr match "$string" '\($substring\)'</DT
><DD
><P
>Extracts <TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
>
at beginning of <TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
>,
where <TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> is a <A
HREF="#REGEXREF"
>regular expression</A
>.</P
></DD
><DT
>expr "$string" : '\($substring\)'</DT
><DD
><P
>Extracts <TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
>
at beginning of <TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
>,
where <TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> is a regular
expression.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>stringZ=abcABC123ABCabc
# =======
echo `expr match "$stringZ" '\(.[b-c]*[A-Z]..[0-9]\)'` # abcABC1
echo `expr "$stringZ" : '\(.[b-c]*[A-Z]..[0-9]\)'` # abcABC1
echo `expr "$stringZ" : '\(.......\)'` # abcABC1
# All of the above forms give an identical result.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
>expr match "$string" '.*\($substring\)'</DT
><DD
><P
>Extracts <TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
>
at <EM
>end</EM
> of
<TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
>, where
<TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> is a regular
expression.</P
></DD
><DT
>expr "$string" : '.*\($substring\)'</DT
><DD
><P
>Extracts <TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
>
at <EM
>end</EM
> of <TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
>,
where <TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> is a regular
expression.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>stringZ=abcABC123ABCabc
# ======
echo `expr match "$stringZ" '.*\([A-C][A-C][A-C][a-c]*\)'` # ABCabc
echo `expr "$stringZ" : '.*\(......\)'` # ABCabc</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
>Substring Removal</B
></P
><DL
><DT
>${string#substring}</DT
><DD
><P
>Deletes shortest match of
<TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> from
<EM
>front</EM
> of
<TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
>.</P
></DD
><DT
>${string##substring}</DT
><DD
><P
>Deletes longest match of
<TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> from
<EM
>front</EM
> of
<TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
>.</P
><P
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>stringZ=abcABC123ABCabc
# |----| shortest
# |----------| longest
echo ${stringZ#a*C} # 123ABCabc
# Strip out shortest match between 'a' and 'C'.
echo ${stringZ##a*C} # abc
# Strip out longest match between 'a' and 'C'.
# You can parameterize the substrings.
X='a*C'
echo ${stringZ#$X} # 123ABCabc
echo ${stringZ##$X} # abc
# As above.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
>${string%substring}</DT
><DD
><P
>Deletes shortest match of
<TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> from
<EM
>back</EM
> of
<TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
>.</P
><P
>For example:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Rename all filenames in $PWD with "TXT" suffix to a "txt" suffix.
# For example, "file1.TXT" becomes "file1.txt" . . .
SUFF=TXT
suff=txt
for i in $(ls *.$SUFF)
do
mv -f $i ${i%.$SUFF}.$suff
# Leave unchanged everything *except* the shortest pattern match
#+ starting from the right-hand-side of the variable $i . . .
done ### This could be condensed into a "one-liner" if desired.
# Thank you, Rory Winston.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
>${string%%substring}</DT
><DD
><P
>Deletes longest match of
<TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> from
<EM
>back</EM
> of
<TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
>.</P
><P
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>stringZ=abcABC123ABCabc
# || shortest
# |------------| longest
echo ${stringZ%b*c} # abcABC123ABCa
# Strip out shortest match between 'b' and 'c', from back of $stringZ.
echo ${stringZ%%b*c} # a
# Strip out longest match between 'b' and 'c', from back of $stringZ.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>This operator is useful for generating filenames.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="CVT"
></A
><P
><B
>Example 10-3. Converting graphic file formats, with filename change</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# cvt.sh:
# Converts all the MacPaint image files in a directory to "pbm" format.
# Uses the "macptopbm" binary from the "netpbm" package,
#+ which is maintained by Brian Henderson (bryanh@giraffe-data.com).
# Netpbm is a standard part of most Linux distros.
OPERATION=macptopbm
SUFFIX=pbm # New filename suffix.
if [ -n "$1" ]
then
directory=$1 # If directory name given as a script argument...
else
directory=$PWD # Otherwise use current working directory.
fi
# Assumes all files in the target directory are MacPaint image files,
#+ with a ".mac" filename suffix.
for file in $directory/* # Filename globbing.
do
filename=${file%.*c} # Strip ".mac" suffix off filename
#+ ('.*c' matches everything
#+ between '.' and 'c', inclusive).
$OPERATION $file &#62; "$filename.$SUFFIX"
# Redirect conversion to new filename.
rm -f $file # Delete original files after converting.
echo "$filename.$SUFFIX" # Log what is happening to stdout.
done
exit 0
# Exercise:
# --------
# As it stands, this script converts *all* the files in the current
#+ working directory.
# Modify it to work *only* on files with a ".mac" suffix.
# *** And here's another way to do it. *** #
#!/bin/bash
# Batch convert into different graphic formats.
# Assumes imagemagick installed (standard in most Linux distros).
INFMT=png # Can be tif, jpg, gif, etc.
OUTFMT=pdf # Can be tif, jpg, gif, pdf, etc.
for pic in *"$INFMT"
do
p2=$(ls "$pic" | sed -e s/\.$INFMT//)
# echo $p2
convert "$pic" $p2.$OUTFMT
done
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="RA2OGG"
></A
><P
><B
>Example 10-4. Converting streaming audio files to
<I
CLASS="FIRSTTERM"
>ogg</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ra2ogg.sh: Convert streaming audio files (*.ra) to ogg.
# Uses the "mplayer" media player program:
# http://www.mplayerhq.hu/homepage
# Uses the "ogg" library and "oggenc":
# http://www.xiph.org/
#
# This script may need appropriate codecs installed, such as sipr.so ...
# Possibly also the compat-libstdc++ package.
OFILEPREF=${1%%ra} # Strip off the "ra" suffix.
OFILESUFF=wav # Suffix for wav file.
OUTFILE="$OFILEPREF""$OFILESUFF"
E_NOARGS=85
if [ -z "$1" ] # Must specify a filename to convert.
then
echo "Usage: `basename $0` [filename]"
exit $E_NOARGS
fi
##########################################################################
mplayer "$1" -ao pcm:file=$OUTFILE
oggenc "$OUTFILE" # Correct file extension automatically added by oggenc.
##########################################################################
rm "$OUTFILE" # Delete intermediate *.wav file.
# If you want to keep it, comment out above line.
exit $?
# Note:
# ----
# On a Website, simply clicking on a *.ram streaming audio file
#+ usually only downloads the URL of the actual *.ra audio file.
# You can then use "wget" or something similar
#+ to download the *.ra file itself.
# Exercises:
# ---------
# As is, this script converts only *.ra filenames.
# Add flexibility by permitting use of *.ram and other filenames.
#
# If you're really ambitious, expand the script
#+ to do automatic downloads and conversions of streaming audio files.
# Given a URL, batch download streaming audio files (using "wget")
#+ and convert them on the fly.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="GETOPTSIMPLE1"
></A
></P
><P
>A simple emulation of <A
HREF="#GETOPTY"
>getopt</A
>
using substring-extraction constructs.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="GETOPTSIMPLE"
></A
><P
><B
>Example 10-5. Emulating <I
CLASS="FIRSTTERM"
>getopt</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# getopt-simple.sh
# Author: Chris Morgan
# Used in the ABS Guide with permission.
getopt_simple()
{
echo "getopt_simple()"
echo "Parameters are '$*'"
until [ -z "$1" ]
do
echo "Processing parameter of: '$1'"
if [ ${1:0:1} = '/' ]
then
tmp=${1:1} # Strip off leading '/' . . .
parameter=${tmp%%=*} # Extract name.
value=${tmp##*=} # Extract value.
echo "Parameter: '$parameter', value: '$value'"
eval $parameter=$value
fi
shift
done
}
# Pass all options to getopt_simple().
getopt_simple $*
echo "test is '$test'"
echo "test2 is '$test2'"
exit 0 # See also, UseGetOpt.sh, a modified version of this script.
---
sh getopt_example.sh /test=value1 /test2=value2
Parameters are '/test=value1 /test2=value2'
Processing parameter of: '/test=value1'
Parameter: 'test', value: 'value1'
Processing parameter of: '/test2=value2'
Parameter: 'test2', value: 'value2'
test is 'value1'
test2 is 'value2'&#13;</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
>Substring Replacement</B
></P
><DL
><DT
><A
NAME="SUBSTRREPL00"
></A
>${string/substring/replacement}</DT
><DD
><P
> Replace first <I
CLASS="FIRSTTERM"
>match</I
> of
<TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> with
<TT
CLASS="REPLACEABLE"
><I
>$replacement</I
></TT
>.
<A
NAME="AEN6164"
HREF="#FTN.AEN6164"
><SPAN
CLASS="footnote"
>[50]</SPAN
></A
>
</P
></DD
><DT
><A
NAME="SUBSTRREPL01"
></A
>${string//substring/replacement}</DT
><DD
><P
>Replace all matches of
<TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> with
<TT
CLASS="REPLACEABLE"
><I
>$replacement</I
></TT
>.</P
><P
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>stringZ=abcABC123ABCabc
echo ${stringZ/abc/xyz} # xyzABC123ABCabc
# Replaces first match of 'abc' with 'xyz'.
echo ${stringZ//abc/xyz} # xyzABC123ABCxyz
# Replaces all matches of 'abc' with # 'xyz'.
echo ---------------
echo "$stringZ" # abcABC123ABCabc
echo ---------------
# The string itself is not altered!
# Can the match and replacement strings be parameterized?
match=abc
repl=000
echo ${stringZ/$match/$repl} # 000ABC123ABCabc
# ^ ^ ^^^
echo ${stringZ//$match/$repl} # 000ABC123ABC000
# Yes! ^ ^ ^^^ ^^^
echo
# What happens if no $replacement string is supplied?
echo ${stringZ/abc} # ABC123ABCabc
echo ${stringZ//abc} # ABC123ABC
# A simple deletion takes place.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="SUBSTRREPL02"
></A
>${string/#substring/replacement}</DT
><DD
><P
>If <TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> matches
<EM
>front</EM
> end of
<TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
>, substitute
<TT
CLASS="REPLACEABLE"
><I
>$replacement</I
></TT
> for
<TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
>.</P
></DD
><DT
><A
NAME="SUBSTRREPL03"
></A
>${string/%substring/replacement}</DT
><DD
><P
>If <TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> matches
<EM
>back</EM
> end of
<TT
CLASS="REPLACEABLE"
><I
>$string</I
></TT
>, substitute
<TT
CLASS="REPLACEABLE"
><I
>$replacement</I
></TT
> for
<TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
>.</P
><P
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>stringZ=abcABC123ABCabc
echo ${stringZ/#abc/XYZ} # XYZABC123ABCabc
# Replaces front-end match of 'abc' with 'XYZ'.
echo ${stringZ/%abc/XYZ} # abcABC123ABCXYZ
# Replaces back-end match of 'abc' with 'XYZ'.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
></DL
></DIV
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="AWKSTRINGMANIP"
></A
>10.1.1. Manipulating strings using awk</H2
><P
><A
NAME="AWKSTRINGMANIP2"
></A
></P
><P
>A Bash script may invoke the string manipulation facilities of
<A
HREF="#AWKREF"
>awk</A
> as an alternative to using its
built-in operations.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SUBSTRINGEX"
></A
><P
><B
>Example 10-6. Alternate ways of extracting and locating substrings</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# substring-extraction.sh
String=23skidoo1
# 012345678 Bash
# 123456789 awk
# Note different string indexing system:
# Bash numbers first character of string as 0.
# Awk numbers first character of string as 1.
echo ${String:2:4} # position 3 (0-1-2), 4 characters long
# skid
# The awk equivalent of ${string:pos:length} is substr(string,pos,length).
echo | awk '
{ print substr("'"${String}"'",3,4) # skid
}
'
# Piping an empty "echo" to awk gives it dummy input,
#+ and thus makes it unnecessary to supply a filename.
echo "----"
# And likewise:
echo | awk '
{ print index("'"${String}"'", "skid") # 3
} # (skid starts at position 3)
' # The awk equivalent of "expr index" ...
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="STRFDISC"
></A
>10.1.2. Further Reference</H2
><P
>For more on string manipulation in scripts, refer to <A
HREF="#PARAMETER-SUBSTITUTION"
>Section 10.2</A
> and the
<A
HREF="#EXPEXTRSUB"
>relevant section</A
> of the <A
HREF="#EXPRREF"
>expr</A
> command listing.</P
><P
>Script examples:
<P
></P
><OL
TYPE="1"
><LI
><P
><A
HREF="#EX45"
>Example 16-9</A
></P
></LI
><LI
><P
><A
HREF="#LENGTH"
>Example 10-9</A
></P
></LI
><LI
><P
><A
HREF="#PATTMATCHING"
>Example 10-10</A
></P
></LI
><LI
><P
><A
HREF="#RFE"
>Example 10-11</A
></P
></LI
><LI
><P
><A
HREF="#VARMATCH"
>Example 10-13</A
></P
></LI
><LI
><P
><A
HREF="#INSERTIONSORT"
>Example A-36</A
></P
></LI
><LI
><P
><A
HREF="#QKY"
>Example A-41</A
></P
></LI
></OL
>
</P
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="PARAMETER-SUBSTITUTION"
></A
>10.2. Parameter Substitution</H1
><P
><A
NAME="PARAMSUBREF"
></A
></P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="PSSUB1"
></A
>Manipulating and/or expanding variables</B
></P
><DL
><DT
><TT
CLASS="USERINPUT"
><B
>${parameter}</B
></TT
></DT
><DD
><P
>Same as <TT
CLASS="REPLACEABLE"
><I
>$parameter</I
></TT
>, i.e.,
value of the variable
<TT
CLASS="REPLACEABLE"
><I
>parameter</I
></TT
>.
In certain contexts, only the less ambiguous
<TT
CLASS="REPLACEABLE"
><I
>${parameter}</I
></TT
> form
works.</P
><P
>May be used for concatenating variables with strings.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>your_id=${USER}-on-${HOSTNAME}
echo "$your_id"
#
echo "Old \$PATH = $PATH"
PATH=${PATH}:/opt/bin # Add /opt/bin to $PATH for duration of script.
echo "New \$PATH = $PATH"</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="DEFPARAM1"
></A
><TT
CLASS="USERINPUT"
><B
>${parameter-default}</B
></TT
>, <TT
CLASS="USERINPUT"
><B
>${parameter:-default}</B
></TT
></DT
><DD
><P
>If parameter not set, use default.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>var1=1
var2=2
# var3 is unset.
echo ${var1-$var2} # 1
echo ${var3-$var2} # 2
# ^ Note the $ prefix.
echo ${username-`whoami`}
# Echoes the result of `whoami`, if variable $username is still unset.</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="UNDDR"
></A
><TT
CLASS="REPLACEABLE"
><I
>${parameter-default}</I
></TT
>
and <TT
CLASS="REPLACEABLE"
><I
>${parameter:-default}</I
></TT
>
are almost equivalent. The extra <SPAN
CLASS="TOKEN"
>:</SPAN
> makes
a difference only when <TT
CLASS="PARAMETER"
><I
>parameter</I
></TT
>
has been declared, but is null. </P
></TD
></TR
></TABLE
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# param-sub.sh
# Whether a variable has been declared
#+ affects triggering of the default option
#+ even if the variable is null.
username0=
echo "username0 has been declared, but is set to null."
echo "username0 = ${username0-`whoami`}"
# Will not echo.
echo
echo username1 has not been declared.
echo "username1 = ${username1-`whoami`}"
# Will echo.
username2=
echo "username2 has been declared, but is set to null."
echo "username2 = ${username2:-`whoami`}"
# ^
# Will echo because of :- rather than just - in condition test.
# Compare to first instance, above.
#
# Once again:
variable=
# variable has been declared, but is set to null.
echo "${variable-0}" # (no output)
echo "${variable:-1}" # 1
# ^
unset variable
echo "${variable-2}" # 2
echo "${variable:-3}" # 3
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The <I
CLASS="FIRSTTERM"
>default parameter</I
> construct
finds use in providing <SPAN
CLASS="QUOTE"
>"missing"</SPAN
> command-line
arguments in scripts.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>DEFAULT_FILENAME=generic.data
filename=${1:-$DEFAULT_FILENAME}
# If not otherwise specified, the following command block operates
#+ on the file "generic.data".
# Begin-Command-Block
# ...
# ...
# ...
# End-Command-Block
# From "hanoi2.bash" example:
DISKS=${1:-E_NOPARAM} # Must specify how many disks.
# Set $DISKS to $1 command-line-parameter,
#+ or to $E_NOPARAM if that is unset.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>See also <A
HREF="#EX58"
>Example 3-4</A
>, <A
HREF="#EX73"
>Example 31-2</A
>, and <A
HREF="#COLLATZ"
>Example A-6</A
>.</P
><P
>Compare this method with <A
HREF="#ANDDEFAULT"
>using an <I
CLASS="FIRSTTERM"
>and
list</I
> to supply a default command-line
argument</A
>.</P
></DD
><DT
><TT
CLASS="USERINPUT"
><B
>${parameter=default}</B
></TT
>, <TT
CLASS="USERINPUT"
><B
>${parameter:=default}</B
></TT
></DT
><DD
><P
><A
NAME="DEFPARAM"
></A
></P
><P
>If parameter not set, set it to
<I
CLASS="FIRSTTERM"
>default</I
>.</P
><P
>Both forms nearly equivalent. The <SPAN
CLASS="TOKEN"
>:</SPAN
>
makes a difference only when <TT
CLASS="VARNAME"
>$parameter</TT
>
has been declared and is null,
<A
NAME="AEN6310"
HREF="#FTN.AEN6310"
><SPAN
CLASS="footnote"
>[51]</SPAN
></A
>
as above.
</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo ${var=abc} # abc
echo ${var=xyz} # abc
# $var had already been set to abc, so it did not change.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="PARAMALTV"
></A
><TT
CLASS="USERINPUT"
><B
>${parameter+alt_value}</B
></TT
>, <TT
CLASS="USERINPUT"
><B
>${parameter:+alt_value}</B
></TT
></DT
><DD
><P
>If parameter set, use
<TT
CLASS="USERINPUT"
><B
>alt_value</B
></TT
>, else use null
string.</P
><P
>Both forms nearly equivalent. The <SPAN
CLASS="TOKEN"
>:</SPAN
>
makes a difference only when
<TT
CLASS="PARAMETER"
><I
>parameter</I
></TT
>
has been declared and is null, see below.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo "###### \${parameter+alt_value} ########"
echo
a=${param1+xyz}
echo "a = $a" # a =
param2=
a=${param2+xyz}
echo "a = $a" # a = xyz
param3=123
a=${param3+xyz}
echo "a = $a" # a = xyz
echo
echo "###### \${parameter:+alt_value} ########"
echo
a=${param4:+xyz}
echo "a = $a" # a =
param5=
a=${param5:+xyz}
echo "a = $a" # a =
# Different result from a=${param5+xyz}
param6=123
a=${param6:+xyz}
echo "a = $a" # a = xyz</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="QERRMSG"
></A
><TT
CLASS="USERINPUT"
><B
>${parameter?err_msg}</B
></TT
>, <TT
CLASS="USERINPUT"
><B
>${parameter:?err_msg}</B
></TT
></DT
><DD
><P
>If parameter set, use it, else print
<I
CLASS="FIRSTTERM"
>err_msg</I
> and <EM
>abort
the script</EM
> with an <A
HREF="#EXITSTATUSREF"
>exit status</A
> of
<SPAN
CLASS="ERRORCODE"
>1</SPAN
>.</P
><P
>Both forms nearly equivalent. The <SPAN
CLASS="TOKEN"
>:</SPAN
>
makes a difference only when <TT
CLASS="PARAMETER"
><I
>parameter</I
></TT
>
has been declared and is null, as above.</P
></DD
></DL
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX6"
></A
><P
><B
>Example 10-7. Using parameter substitution and error messages</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Check some of the system's environmental variables.
# This is good preventative maintenance.
# If, for example, $USER, the name of the person at the console, is not set,
#+ the machine will not recognize you.
: ${HOSTNAME?} ${USER?} ${HOME?} ${MAIL?}
echo
echo "Name of the machine is $HOSTNAME."
echo "You are $USER."
echo "Your home directory is $HOME."
echo "Your mail INBOX is located in $MAIL."
echo
echo "If you are reading this message,"
echo "critical environmental variables have been set."
echo
echo
# ------------------------------------------------------
# The ${variablename?} construction can also check
#+ for variables set within the script.
ThisVariable=Value-of-ThisVariable
# Note, by the way, that string variables may be set
#+ to characters disallowed in their names.
: ${ThisVariable?}
echo "Value of ThisVariable is $ThisVariable".
echo; echo
: ${ZZXy23AB?"ZZXy23AB has not been set."}
# Since ZZXy23AB has not been set,
#+ then the script terminates with an error message.
# You can specify the error message.
# : ${variablename?"ERROR MESSAGE"}
# Same result with: dummy_variable=${ZZXy23AB?}
# dummy_variable=${ZZXy23AB?"ZXy23AB has not been set."}
#
# echo ${ZZXy23AB?} &#62;/dev/null
# Compare these methods of checking whether a variable has been set
#+ with "set -u" . . .
echo "You will not see this message, because script already terminated."
HERE=0
exit $HERE # Will NOT exit here.
# In fact, this script will return an exit status (echo $?) of 1.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="USAGEMESSAGE"
></A
><P
><B
>Example 10-8. Parameter substitution and <SPAN
CLASS="QUOTE"
>"usage"</SPAN
> messages</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# usage-message.sh
: ${1?"Usage: $0 ARGUMENT"}
# Script exits here if command-line parameter absent,
#+ with following error message.
# usage-message.sh: 1: Usage: usage-message.sh ARGUMENT
echo "These two lines echo only if command-line parameter given."
echo "command-line parameter = \"$1\""
exit 0 # Will exit here only if command-line parameter present.
# Check the exit status, both with and without command-line parameter.
# If command-line parameter present, then "$?" is 0.
# If not, then "$?" is 1.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="FORMALPARA"
><P
><B
>Parameter substitution and/or expansion. </B
><A
NAME="PSUB2"
></A
>The following expressions are
the complement to the <B
CLASS="COMMAND"
>match</B
>
<TT
CLASS="REPLACEABLE"
><I
>in</I
></TT
> <B
CLASS="COMMAND"
>expr</B
>
string operations (see <A
HREF="#EX45"
>Example 16-9</A
>).
These particular ones are used mostly in parsing file
path names.</P
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="PSOREX1"
></A
>Variable length / Substring removal</B
></P
><DL
><DT
><TT
CLASS="USERINPUT"
><B
>${#var}</B
></TT
></DT
><DD
><P
><TT
CLASS="USERINPUT"
><B
>String length</B
></TT
> (number
of characters in <TT
CLASS="VARNAME"
>$var</TT
>). For
an <A
HREF="#ARRAYREF"
>array</A
>,
<B
CLASS="COMMAND"
>${#array}</B
> is the length of the
first element in the array.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
> Exceptions:
<P
></P
><UL
><LI
><P
><A
NAME="NUMPOSPARAM"
></A
></P
><P
> <B
CLASS="COMMAND"
>${#*}</B
> and
<B
CLASS="COMMAND"
>${#@}</B
> give the <EM
>number
of positional parameters</EM
>.
</P
></LI
><LI
><P
> For an array, <B
CLASS="COMMAND"
>${#array[*]}</B
> and
<B
CLASS="COMMAND"
>${#array[@]}</B
> give the number
of elements in the array.
</P
></LI
></UL
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="LENGTH"
></A
><P
><B
>Example 10-9. Length of a variable</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# length.sh
E_NO_ARGS=65
if [ $# -eq 0 ] # Must have command-line args to demo script.
then
echo "Please invoke this script with one or more command-line arguments."
exit $E_NO_ARGS
fi
var01=abcdEFGH28ij
echo "var01 = ${var01}"
echo "Length of var01 = ${#var01}"
# Now, let's try embedding a space.
var02="abcd EFGH28ij"
echo "var02 = ${var02}"
echo "Length of var02 = ${#var02}"
echo "Number of command-line arguments passed to script = ${#@}"
echo "Number of command-line arguments passed to script = ${#*}"
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="PSOREX2"
></A
><TT
CLASS="USERINPUT"
><B
>${var#Pattern}</B
></TT
>, <TT
CLASS="USERINPUT"
><B
>${var##Pattern}</B
></TT
></DT
><DD
><P
><A
NAME="PSOREXSH"
></A
></P
><P
><B
CLASS="COMMAND"
>${var#Pattern} </B
>
Remove from <TT
CLASS="VARNAME"
>$var</TT
>
the <EM
>shortest</EM
> part of
<TT
CLASS="VARNAME"
>$Pattern</TT
> that matches
the <TT
CLASS="REPLACEABLE"
><I
>front end</I
></TT
> of
<TT
CLASS="VARNAME"
>$var</TT
>.
</P
><P
><A
NAME="PSOREXLO"
></A
></P
><P
><B
CLASS="COMMAND"
>${var##Pattern} </B
>
Remove from <TT
CLASS="VARNAME"
>$var</TT
>
the <EM
>longest</EM
> part of
<TT
CLASS="VARNAME"
>$Pattern</TT
> that matches
the <TT
CLASS="REPLACEABLE"
><I
>front end</I
></TT
> of
<TT
CLASS="VARNAME"
>$var</TT
>.
</P
><P
>A usage illustration from <A
HREF="#DAYSBETWEEN"
>Example A-7</A
>:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Function from "days-between.sh" example.
# Strips leading zero(s) from argument passed.
strip_leading_zero () # Strip possible leading zero(s)
{ #+ from argument passed.
return=${1#0} # The "1" refers to "$1" -- passed arg.
} # The "0" is what to remove from "$1" -- strips zeros.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Manfred Schwarb's more elaborate variation of the
above:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>strip_leading_zero2 () # Strip possible leading zero(s), since otherwise
{ # Bash will interpret such numbers as octal values.
shopt -s extglob # Turn on extended globbing.
local val=${1##+(0)} # Use local variable, longest matching series of 0's.
shopt -u extglob # Turn off extended globbing.
_strip_leading_zero2=${val:-0}
# If input was 0, return 0 instead of "".
}</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Another usage illustration:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo `basename $PWD` # Basename of current working directory.
echo "${PWD##*/}" # Basename of current working directory.
echo
echo `basename $0` # Name of script.
echo $0 # Name of script.
echo "${0##*/}" # Name of script.
echo
filename=test.data
echo "${filename##*.}" # data
# Extension of filename.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="PCTPATREF"
></A
><TT
CLASS="USERINPUT"
><B
>${var%Pattern}</B
></TT
>, <TT
CLASS="USERINPUT"
><B
>${var%%Pattern}</B
></TT
></DT
><DD
><P
><A
NAME="PCTREP1"
></A
></P
><P
><B
CLASS="COMMAND"
>${var%Pattern}</B
>
Remove from <TT
CLASS="VARNAME"
>$var</TT
>
the <EM
>shortest</EM
> part of
<TT
CLASS="VARNAME"
>$Pattern</TT
> that matches
the <TT
CLASS="REPLACEABLE"
><I
>back end</I
></TT
> of
<TT
CLASS="VARNAME"
>$var</TT
>. </P
><P
><A
NAME="PCTREP2"
></A
></P
><P
><B
CLASS="COMMAND"
>${var%%Pattern}</B
>
Remove from <TT
CLASS="VARNAME"
>$var</TT
>
the <EM
>longest</EM
> part of
<TT
CLASS="VARNAME"
>$Pattern</TT
> that matches
the <TT
CLASS="REPLACEABLE"
><I
>back end</I
></TT
> of
<TT
CLASS="VARNAME"
>$var</TT
>. </P
></DD
></DL
></DIV
><P
><A
HREF="#BASH2REF"
>Version 2</A
> of Bash added
additional options.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="PATTMATCHING"
></A
><P
><B
>Example 10-10. Pattern matching in parameter substitution</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# patt-matching.sh
# Pattern matching using the # ## % %% parameter substitution operators.
var1=abcd12345abc6789
pattern1=a*c # * (wild card) matches everything between a - c.
echo
echo "var1 = $var1" # abcd12345abc6789
echo "var1 = ${var1}" # abcd12345abc6789
# (alternate form)
echo "Number of characters in ${var1} = ${#var1}"
echo
echo "pattern1 = $pattern1" # a*c (everything between 'a' and 'c')
echo "--------------"
echo '${var1#$pattern1} =' "${var1#$pattern1}" # d12345abc6789
# Shortest possible match, strips out first 3 characters abcd12345abc6789
# ^^^^^ |-|
echo '${var1##$pattern1} =' "${var1##$pattern1}" # 6789
# Longest possible match, strips out first 12 characters abcd12345abc6789
# ^^^^^ |----------|
echo; echo; echo
pattern2=b*9 # everything between 'b' and '9'
echo "var1 = $var1" # Still abcd12345abc6789
echo
echo "pattern2 = $pattern2"
echo "--------------"
echo '${var1%pattern2} =' "${var1%$pattern2}" # abcd12345a
# Shortest possible match, strips out last 6 characters abcd12345abc6789
# ^^^^ |----|
echo '${var1%%pattern2} =' "${var1%%$pattern2}" # a
# Longest possible match, strips out last 12 characters abcd12345abc6789
# ^^^^ |-------------|
# Remember, # and ## work from the left end (beginning) of string,
# % and %% work from the right end.
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="RFE"
></A
><P
><B
>Example 10-11. Renaming file extensions<SPAN
CLASS="TOKEN"
>:</SPAN
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# rfe.sh: Renaming file extensions.
#
# rfe old_extension new_extension
#
# Example:
# To rename all *.gif files in working directory to *.jpg,
# rfe gif jpg
E_BADARGS=65
case $# in
0|1) # The vertical bar means "or" in this context.
echo "Usage: `basename $0` old_file_suffix new_file_suffix"
exit $E_BADARGS # If 0 or 1 arg, then bail out.
;;
esac
for filename in *.$1
# Traverse list of files ending with 1st argument.
do
mv $filename ${filename%$1}$2
# Strip off part of filename matching 1st argument,
#+ then append 2nd argument.
done
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="EXPREPL1"
></A
>Variable expansion / Substring
replacement</B
></P
><DL
><DT
></DT
><DD
><P
>These constructs have been adopted from
<I
CLASS="FIRSTTERM"
>ksh</I
>.</P
></DD
><DT
><TT
CLASS="USERINPUT"
><B
>${var:pos}</B
></TT
></DT
><DD
><P
>Variable <TT
CLASS="REPLACEABLE"
><I
>var</I
></TT
> expanded,
starting from offset <TT
CLASS="REPLACEABLE"
><I
>pos</I
></TT
>.
</P
></DD
><DT
><TT
CLASS="USERINPUT"
><B
>${var:pos:len}</B
></TT
></DT
><DD
><P
>Expansion to a max of <TT
CLASS="REPLACEABLE"
><I
>len</I
></TT
>
characters of variable <TT
CLASS="REPLACEABLE"
><I
>var</I
></TT
>, from offset
<TT
CLASS="REPLACEABLE"
><I
>pos</I
></TT
>. See <A
HREF="#PW"
>Example A-13</A
>
for an example of the creative use of this operator.
</P
></DD
><DT
><TT
CLASS="USERINPUT"
><B
>${var/Pattern/Replacement}</B
></TT
></DT
><DD
><P
>First match of <TT
CLASS="REPLACEABLE"
><I
>Pattern</I
></TT
>,
within <TT
CLASS="REPLACEABLE"
><I
>var</I
></TT
> replaced with
<TT
CLASS="REPLACEABLE"
><I
>Replacement</I
></TT
>.</P
><P
>If <TT
CLASS="REPLACEABLE"
><I
>Replacement</I
></TT
> is
omitted, then the first match of
<TT
CLASS="REPLACEABLE"
><I
>Pattern</I
></TT
> is replaced by
<EM
>nothing</EM
>, that is, deleted.</P
></DD
><DT
><TT
CLASS="USERINPUT"
><B
>${var//Pattern/Replacement}</B
></TT
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Global replacement. </B
><A
NAME="PSGLOB"
></A
>
All matches of <TT
CLASS="REPLACEABLE"
><I
>Pattern</I
></TT
>,
within <TT
CLASS="REPLACEABLE"
><I
>var</I
></TT
> replaced with
<TT
CLASS="REPLACEABLE"
><I
>Replacement</I
></TT
>.</P
></DIV
><P
>As above, if <TT
CLASS="REPLACEABLE"
><I
>Replacement</I
></TT
>
is omitted, then all occurrences of
<TT
CLASS="REPLACEABLE"
><I
>Pattern</I
></TT
> are replaced by
<EM
>nothing</EM
>, that is, deleted.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX7"
></A
><P
><B
>Example 10-12. Using pattern matching to parse arbitrary strings</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
var1=abcd-1234-defg
echo "var1 = $var1"
t=${var1#*-*}
echo "var1 (with everything, up to and including first - stripped out) = $t"
# t=${var1#*-} works just the same,
#+ since # matches the shortest string,
#+ and * matches everything preceding, including an empty string.
# (Thanks, Stephane Chazelas, for pointing this out.)
t=${var1##*-*}
echo "If var1 contains a \"-\", returns empty string... var1 = $t"
t=${var1%*-*}
echo "var1 (with everything from the last - on stripped out) = $t"
echo
# -------------------------------------------
path_name=/home/bozo/ideas/thoughts.for.today
# -------------------------------------------
echo "path_name = $path_name"
t=${path_name##/*/}
echo "path_name, stripped of prefixes = $t"
# Same effect as t=`basename $path_name` in this particular case.
# t=${path_name%/}; t=${t##*/} is a more general solution,
#+ but still fails sometimes.
# If $path_name ends with a newline, then `basename $path_name` will not work,
#+ but the above expression will.
# (Thanks, S.C.)
t=${path_name%/*.*}
# Same effect as t=`dirname $path_name`
echo "path_name, stripped of suffixes = $t"
# These will fail in some cases, such as "../", "/foo////", # "foo/", "/".
# Removing suffixes, especially when the basename has no suffix,
#+ but the dirname does, also complicates matters.
# (Thanks, S.C.)
echo
t=${path_name:11}
echo "$path_name, with first 11 chars stripped off = $t"
t=${path_name:11:5}
echo "$path_name, with first 11 chars stripped off, length 5 = $t"
echo
t=${path_name/bozo/clown}
echo "$path_name with \"bozo\" replaced by \"clown\" = $t"
t=${path_name/today/}
echo "$path_name with \"today\" deleted = $t"
t=${path_name//o/O}
echo "$path_name with all o's capitalized = $t"
t=${path_name//o/}
echo "$path_name with all o's deleted = $t"
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><TT
CLASS="USERINPUT"
><B
>${var/#Pattern/Replacement}</B
></TT
></DT
><DD
><P
>If <I
CLASS="FIRSTTERM"
>prefix</I
> of
<TT
CLASS="REPLACEABLE"
><I
>var</I
></TT
> matches
<TT
CLASS="REPLACEABLE"
><I
>Pattern</I
></TT
>, then substitute
<TT
CLASS="REPLACEABLE"
><I
>Replacement</I
></TT
> for
<TT
CLASS="REPLACEABLE"
><I
>Pattern</I
></TT
>.</P
></DD
><DT
><TT
CLASS="USERINPUT"
><B
>${var/%Pattern/Replacement}</B
></TT
></DT
><DD
><P
>If <I
CLASS="FIRSTTERM"
>suffix</I
> of
<TT
CLASS="REPLACEABLE"
><I
>var</I
></TT
> matches
<TT
CLASS="REPLACEABLE"
><I
>Pattern</I
></TT
>, then substitute
<TT
CLASS="REPLACEABLE"
><I
>Replacement</I
></TT
> for
<TT
CLASS="REPLACEABLE"
><I
>Pattern</I
></TT
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="VARMATCH"
></A
><P
><B
>Example 10-13. Matching patterns at prefix or suffix of string</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# var-match.sh:
# Demo of pattern replacement at prefix / suffix of string.
v0=abc1234zip1234abc # Original variable.
echo "v0 = $v0" # abc1234zip1234abc
echo
# Match at prefix (beginning) of string.
v1=${v0/#abc/ABCDEF} # abc1234zip1234abc
# |-|
echo "v1 = $v1" # ABCDEF1234zip1234abc
# |----|
# Match at suffix (end) of string.
v2=${v0/%abc/ABCDEF} # abc1234zip123abc
# |-|
echo "v2 = $v2" # abc1234zip1234ABCDEF
# |----|
echo
# ----------------------------------------------------
# Must match at beginning / end of string,
#+ otherwise no replacement results.
# ----------------------------------------------------
v3=${v0/#123/000} # Matches, but not at beginning.
echo "v3 = $v3" # abc1234zip1234abc
# NO REPLACEMENT.
v4=${v0/%123/000} # Matches, but not at end.
echo "v4 = $v4" # abc1234zip1234abc
# NO REPLACEMENT.
exit 0 </PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="VARPREFIXM"
></A
><TT
CLASS="USERINPUT"
><B
>${!varprefix*}</B
></TT
>, <TT
CLASS="USERINPUT"
><B
>${!varprefix@}</B
></TT
></DT
><DD
><P
>Matches <EM
>names</EM
> of all
previously declared variables beginning
with <TT
CLASS="PARAMETER"
><I
>varprefix</I
></TT
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># This is a variation on indirect reference, but with a * or @.
# Bash, version 2.04, adds this feature.
xyz23=whatever
xyz24=
a=${!xyz*} # Expands to *names* of declared variables
# ^ ^ ^ + beginning with "xyz".
echo "a = $a" # a = xyz23 xyz24
a=${!xyz@} # Same as above.
echo "a = $a" # a = xyz23 xyz24
echo "---"
abc23=something_else
b=${!abc*}
echo "b = $b" # b = abc23
c=${!b} # Now, the more familiar type of indirect reference.
echo $c # something_else</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
></DL
></DIV
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="LOOPS"
></A
>Chapter 11. Loops and Branches</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>What needs this iteration, woman?</I
></P
><P
><I
>--Shakespeare, <TT
CLASS="REPLACEABLE"
><I
>Othello</I
></TT
></I
></P
></I
></TD
></TR
></TABLE
><P
><A
NAME="LOOPREF00"
></A
></P
><P
>Operations on code blocks are the key to structured and organized
shell scripts. Looping and branching constructs provide the tools for
accomplishing this.</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="LOOPS1"
></A
>11.1. Loops</H1
><P
>A <I
CLASS="FIRSTTERM"
>loop</I
> is a block of code that
<I
CLASS="FIRSTTERM"
>iterates</I
>
<A
NAME="AEN6560"
HREF="#FTN.AEN6560"
><SPAN
CLASS="footnote"
>[52]</SPAN
></A
>
a list of commands
as long as the <I
CLASS="FIRSTTERM"
>loop control condition</I
>
is true.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="FORLOOPREF1"
></A
>for loops</B
></P
><DL
><DT
><B
CLASS="COMMAND"
>for <TT
CLASS="PARAMETER"
><I
>arg</I
></TT
> in
<TT
CLASS="REPLACEABLE"
><I
>[list]</I
></TT
></B
></DT
><DD
><P
>This is the basic looping construct. It differs significantly
from its <I
CLASS="FIRSTTERM"
>C</I
> counterpart.</P
><P
><A
NAME="DOINREF"
></A
></P
><P
><P
><B
CLASS="COMMAND"
>for</B
> <TT
CLASS="REPLACEABLE"
><I
>arg</I
></TT
> in [<TT
CLASS="REPLACEABLE"
><I
>list</I
></TT
>]<BR> do <BR> <TT
CLASS="REPLACEABLE"
><I
><3E>command(s)</I
></TT
>... <BR> done </P
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>During each pass through the loop,
<TT
CLASS="REPLACEABLE"
><I
>arg</I
></TT
> takes on the
value of each successive variable in the
<TT
CLASS="REPLACEABLE"
><I
>list</I
></TT
>.</P
></TD
></TR
></TABLE
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>for arg in "$var1" "$var2" "$var3" ... "$varN"
# In pass 1 of the loop, arg = $var1
# In pass 2 of the loop, arg = $var2
# In pass 3 of the loop, arg = $var3
# ...
# In pass N of the loop, arg = $varN
# Arguments in [list] quoted to prevent possible word splitting.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The argument <TT
CLASS="REPLACEABLE"
><I
>list</I
></TT
> may
contain <A
HREF="#ASTERISKREF"
>wild cards</A
>.</P
><P
><A
NAME="NEEDSEMICOLON"
></A
></P
><P
>If <I
CLASS="FIRSTTERM"
>do</I
> is on same line as
<I
CLASS="FIRSTTERM"
>for</I
>, there needs to be a semicolon
after list.</P
><P
><P
><B
CLASS="COMMAND"
>for</B
> <TT
CLASS="REPLACEABLE"
><I
>arg</I
></TT
> in [<TT
CLASS="REPLACEABLE"
><I
>list</I
></TT
>] ; do <BR></P
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX22"
></A
><P
><B
>Example 11-1. Simple <I
CLASS="FIRSTTERM"
>for</I
> loops</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Listing the planets.
for planet in Mercury Venus Earth Mars Jupiter Saturn Uranus Neptune Pluto
do
echo $planet # Each planet on a separate line.
done
echo; echo
for planet in "Mercury Venus Earth Mars Jupiter Saturn Uranus Neptune Pluto"
# All planets on same line.
# Entire 'list' enclosed in quotes creates a single variable.
# Why? Whitespace incorporated into the variable.
do
echo $planet
done
echo; echo "Whoops! Pluto is no longer a planet!"
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="MULTPARAML"
></A
></P
><P
>Each <TT
CLASS="USERINPUT"
><B
>[list]</B
></TT
> element
may contain multiple parameters. This is useful when
processing parameters in groups. In such cases,
use the <A
HREF="#SETREF"
>set</A
> command
(see <A
HREF="#EX34"
>Example 15-16</A
>) to force parsing of each
<TT
CLASS="USERINPUT"
><B
>[list]</B
></TT
> element and assignment of
each component to the positional parameters.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX22A"
></A
><P
><B
>Example 11-2. <I
CLASS="FIRSTTERM"
>for</I
> loop with two parameters in each
[list] element</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Planets revisited.
# Associate the name of each planet with its distance from the sun.
for planet in "Mercury 36" "Venus 67" "Earth 93" "Mars 142" "Jupiter 483"
do
set -- $planet # Parses variable "planet"
#+ and sets positional parameters.
# The "--" prevents nasty surprises if $planet is null or
#+ begins with a dash.
# May need to save original positional parameters,
#+ since they get overwritten.
# One way of doing this is to use an array,
# original_params=("$@")
echo "$1 $2,000,000 miles from the sun"
#-------two tabs---concatenate zeroes onto parameter $2
done
# (Thanks, S.C., for additional clarification.)
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="PARAMLI"
></A
></P
><P
>A variable may supply the <TT
CLASS="USERINPUT"
><B
>[list]</B
></TT
> in a
<I
CLASS="FIRSTTERM"
>for loop</I
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="FILEINFO"
></A
><P
><B
>Example 11-3. <EM
>Fileinfo:</EM
> operating on a file list
contained in a variable</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# fileinfo.sh
FILES="/usr/sbin/accept
/usr/sbin/pwck
/usr/sbin/chroot
/usr/bin/fakefile
/sbin/badblocks
/sbin/ypbind" # List of files you are curious about.
# Threw in a dummy file, /usr/bin/fakefile.
echo
for file in $FILES
do
if [ ! -e "$file" ] # Check if file exists.
then
echo "$file does not exist."; echo
continue # On to next.
fi
ls -l $file | awk '{ print $8 " file size: " $5 }' # Print 2 fields.
whatis `basename $file` # File info.
# Note that the whatis database needs to have been set up for this to work.
# To do this, as root run /usr/bin/makewhatis.
echo
done
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="PARAMLI2"
></A
></P
><P
>The <TT
CLASS="USERINPUT"
><B
>[list]</B
></TT
> in a
<I
CLASS="FIRSTTERM"
>for loop</I
> may be parameterized.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="FILEINFO01"
></A
><P
><B
>Example 11-4. Operating on a parameterized file list</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
filename="*txt"
for file in $filename
do
echo "Contents of $file"
echo "---"
cat "$file"
echo
done</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="LIGLOB"
></A
></P
><P
>If the <TT
CLASS="USERINPUT"
><B
>[list]</B
></TT
> in a
<I
CLASS="FIRSTTERM"
>for loop</I
> contains wild cards
(<SPAN
CLASS="TOKEN"
>*</SPAN
> and <SPAN
CLASS="TOKEN"
>?</SPAN
>) used in filename
expansion, then <A
HREF="#GLOBBINGREF"
>globbing</A
>
takes place.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="LISTGLOB"
></A
><P
><B
>Example 11-5. Operating on files with a <I
CLASS="FIRSTTERM"
>for</I
> loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# list-glob.sh: Generating [list] in a for-loop, using "globbing" ...
# Globbing = filename expansion.
echo
for file in *
# ^ Bash performs filename expansion
#+ on expressions that globbing recognizes.
do
ls -l "$file" # Lists all files in $PWD (current directory).
# Recall that the wild card character "*" matches every filename,
#+ however, in "globbing," it doesn't match dot-files.
# If the pattern matches no file, it is expanded to itself.
# To prevent this, set the nullglob option
#+ (shopt -s nullglob).
# Thanks, S.C.
done
echo; echo
for file in [jx]*
do
rm -f $file # Removes only files beginning with "j" or "x" in $PWD.
echo "Removed file \"$file\"".
done
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="OMITLIST"
></A
></P
><P
>Omitting the <TT
CLASS="USERINPUT"
><B
>in [list]</B
></TT
> part of a
<I
CLASS="FIRSTTERM"
>for loop</I
> causes the loop to operate
on <SPAN
CLASS="TOKEN"
>$@</SPAN
> -- the <A
HREF="#POSPARAMREF"
> positional parameters</A
>. A particularly clever
illustration of this is <A
HREF="#PRIMES"
>Example A-15</A
>. See also <A
HREF="#REVPOSPARAMS"
>Example 15-17</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX23"
></A
><P
><B
>Example 11-6. Missing <TT
CLASS="USERINPUT"
><B
>in [list]</B
></TT
> in a
<I
CLASS="FIRSTTERM"
>for</I
> loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Invoke this script both with and without arguments,
#+ and see what happens.
for a
do
echo -n "$a "
done
# The 'in list' missing, therefore the loop operates on '$@'
#+ (command-line argument list, including whitespace).
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="LOOPCS"
></A
></P
><P
>It is possible to use <A
HREF="#COMMANDSUBREF"
>command substitution</A
>
to generate the <TT
CLASS="USERINPUT"
><B
>[list]</B
></TT
> in a
<I
CLASS="FIRSTTERM"
>for loop</I
>. See also <A
HREF="#EX53"
>Example 16-54</A
>,
<A
HREF="#SYMLINKS"
>Example 11-11</A
> and <A
HREF="#BASE"
>Example 16-48</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="FORLOOPCMD"
></A
><P
><B
>Example 11-7. Generating the <TT
CLASS="USERINPUT"
><B
>[list]</B
></TT
> in
a <I
CLASS="FIRSTTERM"
>for</I
> loop with command substitution</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# for-loopcmd.sh: for-loop with [list]
#+ generated by command substitution.
NUMBERS="9 7 3 8 37.53"
for number in `echo $NUMBERS` # for number in 9 7 3 8 37.53
do
echo -n "$number "
done
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Here is a somewhat more complex example of using command
substitution to create the <TT
CLASS="USERINPUT"
><B
>[list]</B
></TT
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="BINGREP"
></A
><P
><B
>Example 11-8. A <I
CLASS="FIRSTTERM"
>grep</I
> replacement
for binary files</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# bin-grep.sh: Locates matching strings in a binary file.
# A "grep" replacement for binary files.
# Similar effect to "grep -a"
E_BADARGS=65
E_NOFILE=66
if [ $# -ne 2 ]
then
echo "Usage: `basename $0` search_string filename"
exit $E_BADARGS
fi
if [ ! -f "$2" ]
then
echo "File \"$2\" does not exist."
exit $E_NOFILE
fi
IFS=$'\012' # Per suggestion of Anton Filippov.
# was: IFS="\n"
for word in $( strings "$2" | grep "$1" )
# The "strings" command lists strings in binary files.
# Output then piped to "grep", which tests for desired string.
do
echo $word
done
# As S.C. points out, lines 23 - 30 could be replaced with the simpler
# strings "$2" | grep "$1" | tr -s "$IFS" '[\n*]'
# Try something like "./bin-grep.sh mem /bin/ls"
#+ to exercise this script.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>More of the same.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="USERLIST"
></A
><P
><B
>Example 11-9. Listing all users on the system</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# userlist.sh
PASSWORD_FILE=/etc/passwd
n=1 # User number
for name in $(awk 'BEGIN{FS=":"}{print $1}' &#60; "$PASSWORD_FILE" )
# Field separator = : ^^^^^^
# Print first field ^^^^^^^^
# Get input from password file /etc/passwd ^^^^^^^^^^^^^^^^^
do
echo "USER #$n = $name"
let "n += 1"
done
# USER #1 = root
# USER #2 = bin
# USER #3 = daemon
# ...
# USER #33 = bozo
exit $?
# Discussion:
# ----------
# How is it that an ordinary user, or a script run by same,
#+ can read /etc/passwd? (Hint: Check the /etc/passwd file permissions.)
# Is this a security hole? Why or why not?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Yet another example of the <TT
CLASS="USERINPUT"
><B
>[list]</B
></TT
>
resulting from command substitution.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="FINDSTRING"
></A
><P
><B
>Example 11-10. Checking all the binaries in a directory for
authorship</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# findstring.sh:
# Find a particular string in the binaries in a specified directory.
directory=/usr/bin/
fstring="Free Software Foundation" # See which files come from the FSF.
for file in $( find $directory -type f -name '*' | sort )
do
strings -f $file | grep "$fstring" | sed -e "s%$directory%%"
# In the "sed" expression,
#+ it is necessary to substitute for the normal "/" delimiter
#+ because "/" happens to be one of the characters filtered out.
# Failure to do so gives an error message. (Try it.)
done
exit $?
# Exercise (easy):
# ---------------
# Convert this script to take command-line parameters
#+ for $directory and $fstring.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>A final example of <TT
CLASS="USERINPUT"
><B
>[list]</B
></TT
>
/ command substitution, but this time
the <SPAN
CLASS="QUOTE"
>"command"</SPAN
> is a <A
HREF="#FUNCTIONREF"
>function</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>generate_list ()
{
echo "one two three"
}
for word in $(generate_list) # Let "word" grab output of function.
do
echo "$word"
done
# one
# two
# three</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="LOOPREDIR"
></A
></P
><P
>The output of a <I
CLASS="FIRSTTERM"
>for loop</I
> may
be piped to a command or commands.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SYMLINKS"
></A
><P
><B
>Example 11-11. Listing the <I
CLASS="FIRSTTERM"
>symbolic
links</I
> in a directory</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# symlinks.sh: Lists symbolic links in a directory.
directory=${1-`pwd`}
# Defaults to current working directory,
#+ if not otherwise specified.
# Equivalent to code block below.
# ----------------------------------------------------------
# ARGS=1 # Expect one command-line argument.
#
# if [ $# -ne "$ARGS" ] # If not 1 arg...
# then
# directory=`pwd` # current working directory
# else
# directory=$1
# fi
# ----------------------------------------------------------
echo "symbolic links in directory \"$directory\""
for file in "$( find $directory -type l )" # -type l = symbolic links
do
echo "$file"
done | sort # Otherwise file list is unsorted.
# Strictly speaking, a loop isn't really necessary here,
#+ since the output of the "find" command is expanded into a single word.
# However, it's easy to understand and illustrative this way.
# As Dominik 'Aeneas' Schnitzer points out,
#+ failing to quote $( find $directory -type l )
#+ will choke on filenames with embedded whitespace.
# containing whitespace.
exit 0
# --------------------------------------------------------
# Jean Helou proposes the following alternative:
echo "symbolic links in directory \"$directory\""
# Backup of the current IFS. One can never be too cautious.
OLDIFS=$IFS
IFS=:
for file in $(find $directory -type l -printf "%p$IFS")
do # ^^^^^^^^^^^^^^^^
echo "$file"
done|sort
# And, James "Mike" Conley suggests modifying Helou's code thusly:
OLDIFS=$IFS
IFS='' # Null IFS means no word breaks
for file in $( find $directory -type l )
do
echo $file
done | sort
# This works in the "pathological" case of a directory name having
#+ an embedded colon.
# "This also fixes the pathological case of the directory name having
#+ a colon (or space in earlier example) as well."&#13;</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The <TT
CLASS="FILENAME"
>stdout</TT
> of a loop may be <A
HREF="#IOREDIRREF"
>redirected</A
> to a file, as this slight
modification to the previous example shows.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SYMLINKS2"
></A
><P
><B
>Example 11-12. Symbolic links in a directory, saved to a file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# symlinks.sh: Lists symbolic links in a directory.
OUTFILE=symlinks.list # save-file
directory=${1-`pwd`}
# Defaults to current working directory,
#+ if not otherwise specified.
echo "symbolic links in directory \"$directory\"" &#62; "$OUTFILE"
echo "---------------------------" &#62;&#62; "$OUTFILE"
for file in "$( find $directory -type l )" # -type l = symbolic links
do
echo "$file"
done | sort &#62;&#62; "$OUTFILE" # stdout of loop
# ^^^^^^^^^^^^^ redirected to save file.
# echo "Output file = $OUTFILE"
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="LOOPCSTYLE"
></A
></P
><P
>There is an alternative syntax to a <I
CLASS="FIRSTTERM"
>for
loop</I
> that will look very familiar to C
programmers. This requires <A
HREF="#DBLPARENSREF"
>double parentheses</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="FORLOOPC"
></A
><P
><B
>Example 11-13. A C-style <I
CLASS="FIRSTTERM"
>for</I
> loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Multiple ways to count up to 10.
echo
# Standard syntax.
for a in 1 2 3 4 5 6 7 8 9 10
do
echo -n "$a "
done
echo; echo
# +==========================================+
# Using "seq" ...
for a in `seq 10`
do
echo -n "$a "
done
echo; echo
# +==========================================+
# Using brace expansion ...
# Bash, version 3+.
for a in {1..10}
do
echo -n "$a "
done
echo; echo
# +==========================================+
# Now, let's do the same, using C-like syntax.
LIMIT=10
for ((a=1; a &#60;= LIMIT ; a++)) # Double parentheses, and naked "LIMIT"
do
echo -n "$a "
done # A construct borrowed from ksh93.
echo; echo
# +=========================================================================+
# Let's use the C "comma operator" to increment two variables simultaneously.
for ((a=1, b=1; a &#60;= LIMIT ; a++, b++))
do # The comma concatenates operations.
echo -n "$a-$b "
done
echo; echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#QFUNCTION"
>Example 27-16</A
>, <A
HREF="#TWODIM"
>Example 27-17</A
>, and <A
HREF="#COLLATZ"
>Example A-6</A
>.</P
><P
>---</P
><P
>Now, a <I
CLASS="FIRSTTERM"
>for loop</I
> used in a
<SPAN
CLASS="QUOTE"
>"real-life"</SPAN
> context.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX24"
></A
><P
><B
>Example 11-14. Using <I
CLASS="FIRSTTERM"
>efax</I
> in batch mode</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Faxing (must have 'efax' package installed).
EXPECTED_ARGS=2
E_BADARGS=85
MODEM_PORT="/dev/ttyS2" # May be different on your machine.
# ^^^^^ PCMCIA modem card default port.
if [ $# -ne $EXPECTED_ARGS ]
# Check for proper number of command-line args.
then
echo "Usage: `basename $0` phone# text-file"
exit $E_BADARGS
fi
if [ ! -f "$2" ]
then
echo "File $2 is not a text file."
# File is not a regular file, or does not exist.
exit $E_BADARGS
fi
fax make $2 # Create fax-formatted files from text files.
for file in $(ls $2.0*) # Concatenate the converted files.
# Uses wild card (filename "globbing")
#+ in variable list.
do
fil="$fil $file"
done
efax -d "$MODEM_PORT" -t "T$1" $fil # Finally, do the work.
# Trying adding -o1 if above line fails.
# As S.C. points out, the for-loop can be eliminated with
# efax -d /dev/ttyS2 -o1 -t "T$1" $2.0*
#+ but it's not quite as instructive [grin].
exit $? # Also, efax sends diagnostic messages to stdout.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="NODODONE"
></A
>The
<A
HREF="#KEYWORDREF"
>keywords</A
>
<B
CLASS="COMMAND"
>do</B
> and <B
CLASS="COMMAND"
>done</B
> delineate
the <I
CLASS="FIRSTTERM"
>for-loop</I
> command block. However,
these may, in certain contexts, be omitted by framing the
command block within <A
HREF="#CODEBLOCKREF"
>curly
brackets</A
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>for((n=1; n&#60;=10; n++))
# No do!
{
echo -n "* $n *"
}
# No done!
# Outputs:
# * 1 ** 2 ** 3 ** 4 ** 5 ** 6 ** 7 ** 8 ** 9 ** 10 *
# And, echo $? returns 0, so Bash does not register an error.
echo
# But, note that in a classic for-loop: for n in [list] ...
#+ a terminal semicolon is required.
for n in 1 2 3
{ echo -n "$n "; }
# ^
# Thank you, YongYe, for pointing this out.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="WHILELOOPREF"
></A
><B
CLASS="COMMAND"
>while</B
></DT
><DD
><P
>This construct tests for a condition at the top of a
loop, and keeps looping as long as that condition
is true (returns a <SPAN
CLASS="RETURNVALUE"
>0</SPAN
> <A
HREF="#EXITSTATUSREF"
>exit status</A
>). In contrast
to a <A
HREF="#FORLOOPREF1"
>for loop</A
>, a
<I
CLASS="FIRSTTERM"
>while loop</I
> finds use in situations
where the number of loop repetitions is not known
beforehand.</P
><P
><P
><B
CLASS="COMMAND"
>while</B
> [<TT
CLASS="REPLACEABLE"
><I
> condition </I
></TT
>]<BR> do <BR> <TT
CLASS="REPLACEABLE"
><I
><3E>command(s)</I
></TT
>... <BR> done </P
></P
><P
>The bracket construct in a <I
CLASS="FIRSTTERM"
>while
loop</I
> is nothing more than our old friend,
the <A
HREF="#TESTCONSTRUCTS1"
>test brackets</A
>
used in an <I
CLASS="FIRSTTERM"
>if/then</I
> test. In fact,
a <I
CLASS="FIRSTTERM"
>while loop</I
> can legally use the
more versatile <A
HREF="#DBLBRACKETS"
>double-brackets
construct</A
> (while [[ condition ]]).</P
><P
><A
NAME="WHILENEEDSEMI"
></A
></P
><P
><A
HREF="#NEEDSEMICOLON"
>As is the case with
<I
CLASS="FIRSTTERM"
>for loops</I
></A
>, placing the
<I
CLASS="FIRSTTERM"
>do</I
> on the same line as the condition
test requires a semicolon.</P
><P
><P
><B
CLASS="COMMAND"
>while</B
> [<TT
CLASS="REPLACEABLE"
><I
> condition </I
></TT
>] ; do </P
></P
><P
>Note that the <I
CLASS="FIRSTTERM"
>test brackets</I
>
<A
HREF="#WHILENOBRACKETS"
>are <EM
>not</EM
>
mandatory</A
> in a <I
CLASS="FIRSTTERM"
>while</I
> loop.
See, for example, the <A
HREF="#GETOPTSX"
>getopts
construct</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX25"
></A
><P
><B
>Example 11-15. Simple <I
CLASS="FIRSTTERM"
>while</I
> loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
var0=0
LIMIT=10
while [ "$var0" -lt "$LIMIT" ]
# ^ ^
# Spaces, because these are "test-brackets" . . .
do
echo -n "$var0 " # -n suppresses newline.
# ^ Space, to separate printed out numbers.
var0=`expr $var0 + 1` # var0=$(($var0+1)) also works.
# var0=$((var0 + 1)) also works.
# let "var0 += 1" also works.
done # Various other methods also work.
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX26"
></A
><P
><B
>Example 11-16. Another <I
CLASS="FIRSTTERM"
>while</I
> loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo
# Equivalent to:
while [ "$var1" != "end" ] # while test "$var1" != "end"
do
echo "Input variable #1 (end to exit) "
read var1 # Not 'read $var1' (why?).
echo "variable #1 = $var1" # Need quotes because of "#" . . .
# If input is 'end', echoes it here.
# Does not test for termination condition until top of loop.
echo
done
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="WHMULTCOND"
></A
></P
><P
>A <I
CLASS="FIRSTTERM"
>while loop</I
> may have multiple
conditions. Only the final condition determines when the loop
terminates. This necessitates a slightly different loop syntax,
however.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX26A"
></A
><P
><B
>Example 11-17. <I
CLASS="FIRSTTERM"
>while</I
> loop with multiple conditions</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
var1=unset
previous=$var1
while echo "previous-variable = $previous"
echo
previous=$var1
[ "$var1" != end ] # Keeps track of what $var1 was previously.
# Four conditions on *while*, but only the final one controls loop.
# The *last* exit status is the one that counts.
do
echo "Input variable #1 (end to exit) "
read var1
echo "variable #1 = $var1"
done
# Try to figure out how this all works.
# It's a wee bit tricky.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="WLOOPCSTYLE"
></A
></P
><P
>As with a <I
CLASS="FIRSTTERM"
>for loop</I
>, a
<I
CLASS="FIRSTTERM"
>while loop</I
> may employ C-style syntax
by using the double-parentheses construct (see also <A
HREF="#CVARS"
>Example 8-5</A
>).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="WHLOOPC"
></A
><P
><B
>Example 11-18. C-style syntax in a <I
CLASS="FIRSTTERM"
>while</I
> loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# wh-loopc.sh: Count to 10 in a "while" loop.
LIMIT=10 # 10 iterations.
a=1
while [ "$a" -le $LIMIT ]
do
echo -n "$a "
let "a+=1"
done # No surprises, so far.
echo; echo
# +=================================================================+
# Now, we'll repeat with C-like syntax.
((a = 1)) # a=1
# Double parentheses permit space when setting a variable, as in C.
while (( a &#60;= LIMIT )) # Double parentheses,
do #+ and no "$" preceding variables.
echo -n "$a "
((a += 1)) # let "a+=1"
# Yes, indeed.
# Double parentheses permit incrementing a variable with C-like syntax.
done
echo
# C and Java programmers can feel right at home in Bash.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="WHILEFUNC"
></A
></P
><P
> Inside its test brackets, a <I
CLASS="FIRSTTERM"
>while loop</I
>
can call a <A
HREF="#FUNCTIONREF"
>function</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>t=0
condition ()
{
((t++))
if [ $t -lt 5 ]
then
return 0 # true
else
return 1 # false
fi
}
while condition
# ^^^^^^^^^
# Function call -- four loop iterations.
do
echo "Still going: t = $t"
done
# Still going: t = 1
# Still going: t = 2
# Still going: t = 3
# Still going: t = 4</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN6856"
></A
><P
></P
><P
><A
NAME="WHILENOBRACKETS"
></A
></P
><P
>Similar to the <A
HREF="#IFGREPREF"
>if-test</A
>
construct, a <I
CLASS="FIRSTTERM"
>while</I
> loop can omit the test
brackets.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>while condition
do
command(s) ...
done</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
><A
NAME="WHILEREADREF2"
></A
></P
><P
>By coupling the power of the <A
HREF="#READREF"
>read</A
> command with a
<I
CLASS="FIRSTTERM"
>while loop</I
>, we get the handy <A
HREF="#WHILEREADREF"
>while read</A
> construct, useful
for reading and parsing files.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat $filename | # Supply input from a file.
while read line # As long as there is another line to read ...
do
...
done
# =========== Snippet from "sd.sh" example script ========== #
while read value # Read one data point at a time.
do
rt=$(echo "scale=$SC; $rt + $value" | bc)
(( ct++ ))
done
am=$(echo "scale=$SC; $rt / $ct" | bc)
echo $am; return $ct # This function "returns" TWO values!
# Caution: This little trick will not work if $ct &#62; 255!
# To handle a larger number of data points,
#+ simply comment out the "return $ct" above.
} &#60;"$datafile" # Feed in data file.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="WHREDIR"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A <I
CLASS="FIRSTTERM"
>while loop</I
> may have its
<TT
CLASS="FILENAME"
>stdin</TT
> <A
HREF="#REDIRREF"
>redirected to a file</A
> by a
<SPAN
CLASS="TOKEN"
>&#60;</SPAN
> at its end.</P
><P
>A <I
CLASS="FIRSTTERM"
>while loop</I
> may have its
<TT
CLASS="FILENAME"
>stdin</TT
> <A
HREF="#READPIPEREF"
> supplied by a pipe</A
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="UNTILLOOPREF"
></A
><B
CLASS="COMMAND"
>until</B
></DT
><DD
><P
>This construct tests for a condition at the top of a loop, and keeps
looping as long as that condition is
<EM
>false</EM
> (opposite of <I
CLASS="FIRSTTERM"
>while
loop</I
>).</P
><P
><P
><B
CLASS="COMMAND"
>until</B
> [<TT
CLASS="REPLACEABLE"
><I
> condition-is-true </I
></TT
>]<BR> do <BR> <TT
CLASS="REPLACEABLE"
><I
><3E>command(s)</I
></TT
>... <BR> done </P
></P
><P
>Note that an <I
CLASS="FIRSTTERM"
>until loop</I
> tests for the
terminating condition at the <EM
>top</EM
>
of the loop, differing from a similar construct in some
programming languages.</P
><P
>As is the case with <I
CLASS="FIRSTTERM"
>for loops</I
>,
placing the <I
CLASS="FIRSTTERM"
>do</I
> on the same line as
the condition test requires a semicolon.</P
><P
><P
><B
CLASS="COMMAND"
>until</B
> [<TT
CLASS="REPLACEABLE"
><I
> condition-is-true </I
></TT
>] ; do </P
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX27"
></A
><P
><B
>Example 11-19. <I
CLASS="FIRSTTERM"
>until</I
> loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
END_CONDITION=end
until [ "$var1" = "$END_CONDITION" ]
# Tests condition here, at top of loop.
do
echo "Input variable #1 "
echo "($END_CONDITION to exit)"
read var1
echo "variable #1 = $var1"
echo
done
# --- #
# As with "for" and "while" loops,
#+ an "until" loop permits C-like test constructs.
LIMIT=10
var=0
until (( var &#62; LIMIT ))
do # ^^ ^ ^ ^^ No brackets, no $ prefixing variables.
echo -n "$var "
(( var++ ))
done # 0 1 2 3 4 5 6 7 8 9 10
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
><A
NAME="CHOOSELOOP"
></A
></P
><P
>How to choose between a <I
CLASS="FIRSTTERM"
>for</I
> loop or a
<I
CLASS="FIRSTTERM"
>while</I
> loop or
<I
CLASS="FIRSTTERM"
>until</I
> loop? In <B
CLASS="COMMAND"
>C</B
>,
you would typically use a <I
CLASS="FIRSTTERM"
>for</I
> loop
when the number of loop iterations is known beforehand. With
<I
CLASS="FIRSTTERM"
>Bash</I
>, however, the situation is
fuzzier. The Bash <I
CLASS="FIRSTTERM"
>for</I
> loop is more
loosely structured and more flexible than its equivalent in
other languages. Therefore, feel free to use whatever type
of loop gets the job done in the simplest way.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="NESTEDLOOPS"
></A
>11.2. Nested Loops</H1
><P
>A <I
CLASS="FIRSTTERM"
>nested loop</I
> is a loop within a
loop, an inner loop within the body of an outer one. How
this works is that the first pass of the outer loop triggers
the inner loop, which executes to completion. Then the
second pass of the outer loop triggers the inner loop
again. This repeats until the outer loop finishes. Of course,
a <I
CLASS="FIRSTTERM"
>break</I
> within either the inner or outer
loop would interrupt this process.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="NESTEDLOOP"
></A
><P
><B
>Example 11-20. Nested Loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# nested-loop.sh: Nested "for" loops.
outer=1 # Set outer loop counter.
# Beginning of outer loop.
for a in 1 2 3 4 5
do
echo "Pass $outer in outer loop."
echo "---------------------"
inner=1 # Reset inner loop counter.
# ===============================================
# Beginning of inner loop.
for b in 1 2 3 4 5
do
echo "Pass $inner in inner loop."
let "inner+=1" # Increment inner loop counter.
done
# End of inner loop.
# ===============================================
let "outer+=1" # Increment outer loop counter.
echo # Space between output blocks in pass of outer loop.
done
# End of outer loop.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See <A
HREF="#BUBBLE"
>Example 27-11</A
> for an illustration of nested
<A
HREF="#WHILELOOPREF"
>while loops</A
>, and <A
HREF="#EX68"
>Example 27-13</A
> to see a while loop nested inside an <A
HREF="#UNTILLOOPREF"
>until loop</A
>.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="LOOPCONTROL"
></A
>11.3. Loop Control</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>Tournez cent tours, tournez mille tours,</I
></P
><P
><I
>Tournez souvent et tournez toujours . . .</I
></P
><P
><I
>--Verlaine, <SPAN
CLASS="QUOTE"
>"Chevaux de bois"</SPAN
></I
></P
></I
></TD
></TR
></TABLE
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="BRKCONT1"
></A
>Commands affecting loop behavior</B
></P
><DL
><DT
><B
CLASS="COMMAND"
>break</B
>, <B
CLASS="COMMAND"
>continue</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>break</B
> and <B
CLASS="COMMAND"
>continue</B
>
loop control commands
<A
NAME="AEN6981"
HREF="#FTN.AEN6981"
><SPAN
CLASS="footnote"
>[53]</SPAN
></A
>
correspond exactly to their counterparts in other
programming languages. The <B
CLASS="COMMAND"
>break</B
>
command terminates the loop (<EM
>breaks</EM
>
out of it), while <B
CLASS="COMMAND"
>continue</B
> causes a jump
to the next <A
HREF="#ITERATIONREF"
>iteration</A
>
of the loop, skipping all the remaining commands in that
particular loop cycle.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX28"
></A
><P
><B
>Example 11-21. Effects of <I
CLASS="FIRSTTERM"
>break</I
> and
<B
CLASS="COMMAND"
>continue</B
> in a loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
LIMIT=19 # Upper limit
echo
echo "Printing Numbers 1 through 20 (but not 3 and 11)."
a=0
while [ $a -le "$LIMIT" ]
do
a=$(($a+1))
if [ "$a" -eq 3 ] || [ "$a" -eq 11 ] # Excludes 3 and 11.
then
continue # Skip rest of this particular loop iteration.
fi
echo -n "$a " # This will not execute for 3 and 11.
done
# Exercise:
# Why does the loop print up to 20?
echo; echo
echo Printing Numbers 1 through 20, but something happens after 2.
##################################################################
# Same loop, but substituting 'break' for 'continue'.
a=0
while [ "$a" -le "$LIMIT" ]
do
a=$(($a+1))
if [ "$a" -gt 2 ]
then
break # Skip entire rest of loop.
fi
echo -n "$a "
done
echo; echo; echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="BREAKPARAM"
></A
></P
><P
>The <B
CLASS="COMMAND"
>break</B
> command may optionally take a
parameter. A plain <B
CLASS="COMMAND"
>break</B
> terminates
only the innermost loop in which it is embedded,
but a <B
CLASS="COMMAND"
>break N</B
> breaks out of
<TT
CLASS="PARAMETER"
><I
>N</I
></TT
> levels of loop.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="BREAKLEVELS"
></A
><P
><B
>Example 11-22. Breaking out of multiple loop levels</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# break-levels.sh: Breaking out of loops.
# "break N" breaks out of N level loops.
for outerloop in 1 2 3 4 5
do
echo -n "Group $outerloop: "
# --------------------------------------------------------
for innerloop in 1 2 3 4 5
do
echo -n "$innerloop "
if [ "$innerloop" -eq 3 ]
then
break # Try break 2 to see what happens.
# ("Breaks" out of both inner and outer loops.)
fi
done
# --------------------------------------------------------
echo
done
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The <B
CLASS="COMMAND"
>continue</B
> command, similar to
<B
CLASS="COMMAND"
>break</B
>, optionally takes a parameter. A
plain <B
CLASS="COMMAND"
>continue</B
> cuts short the
current iteration within its loop and begins the next.
A <B
CLASS="COMMAND"
>continue N</B
> terminates all remaining
iterations at its loop level and continues with the
next iteration at the loop, <TT
CLASS="OPTION"
>N</TT
> levels
above.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="CONTINUELEVELS"
></A
><P
><B
>Example 11-23. Continuing at a higher loop level</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# The "continue N" command, continuing at the Nth level loop.
for outer in I II III IV V # outer loop
do
echo; echo -n "Group $outer: "
# --------------------------------------------------------------------
for inner in 1 2 3 4 5 6 7 8 9 10 # inner loop
do
if [[ "$inner" -eq 7 &#38;&#38; "$outer" = "III" ]]
then
continue 2 # Continue at loop on 2nd level, that is "outer loop".
# Replace above line with a simple "continue"
# to see normal loop behavior.
fi
echo -n "$inner " # 7 8 9 10 will not echo on "Group III."
done
# --------------------------------------------------------------------
done
echo; echo
# Exercise:
# Come up with a meaningful use for "continue N" in a script.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="CONTINUENEX"
></A
><P
><B
>Example 11-24. Using <I
CLASS="FIRSTTERM"
>continue N</I
> in an actual task</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Albert Reiner gives an example of how to use "continue N":
# ---------------------------------------------------------
# Suppose I have a large number of jobs that need to be run, with
#+ any data that is to be treated in files of a given name pattern
#+ in a directory. There are several machines that access
#+ this directory, and I want to distribute the work over these
#+ different boxen.
# Then I usually nohup something like the following on every box:
while true
do
for n in .iso.*
do
[ "$n" = ".iso.opts" ] &#38;&#38; continue
beta=${n#.iso.}
[ -r .Iso.$beta ] &#38;&#38; continue
[ -r .lock.$beta ] &#38;&#38; sleep 10 &#38;&#38; continue
lockfile -r0 .lock.$beta || continue
echo -n "$beta: " `date`
run-isotherm $beta
date
ls -alF .Iso.$beta
[ -r .Iso.$beta ] &#38;&#38; rm -f .lock.$beta
continue 2
done
break
done
exit 0
# The details, in particular the sleep N, are particular to my
#+ application, but the general pattern is:
while true
do
for job in {pattern}
do
{job already done or running} &#38;&#38; continue
{mark job as running, do job, mark job as done}
continue 2
done
break # Or something like `sleep 600' to avoid termination.
done
# This way the script will stop only when there are no more jobs to do
#+ (including jobs that were added during runtime). Through the use
#+ of appropriate lockfiles it can be run on several machines
#+ concurrently without duplication of calculations [which run a couple
#+ of hours in my case, so I really want to avoid this]. Also, as search
#+ always starts again from the beginning, one can encode priorities in
#+ the file names. Of course, one could also do this without `continue 2',
#+ but then one would have to actually check whether or not some job
#+ was done (so that we should immediately look for the next job) or not
#+ (in which case we terminate or sleep for a long time before checking
#+ for a new job).</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>continue N</B
> construct is
difficult to understand and tricky to use in any meaningful
context. It is probably best avoided.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="TESTBRANCH"
></A
>11.4. Testing and Branching</H1
><P
>The <B
CLASS="COMMAND"
>case</B
> and <B
CLASS="COMMAND"
>select</B
>
constructs are technically not loops, since they do not iterate the
execution of a code block. Like loops, however, they direct
program flow according to conditions at the top or bottom of
the block.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="CASEESAC1"
></A
>Controlling program flow in a code
block</B
></P
><DL
><DT
><B
CLASS="COMMAND"
>case (in) / esac</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>case</B
> construct is the shell
scripting analog to <TT
CLASS="REPLACEABLE"
><I
>switch</I
></TT
>
in <B
CLASS="COMMAND"
>C/C++</B
>.
It permits branching to one of a number of code blocks,
depending on condition tests. It serves as a kind of
shorthand for multiple <SPAN
CLASS="TOKEN"
>if/then/else</SPAN
>
statements and is an appropriate tool for creating
menus.</P
><P
><P
><B
CLASS="COMMAND"
>case</B
> "$<TT
CLASS="REPLACEABLE"
><I
>variable</I
></TT
>" in <BR><BR> <20>"$<TT
CLASS="REPLACEABLE"
><I
>condition1</I
></TT
>" ) <BR> <20><TT
CLASS="REPLACEABLE"
><I
>command</I
></TT
>... <BR> <20>;; <BR><BR> <20>"$<TT
CLASS="REPLACEABLE"
><I
>condition2</I
></TT
>" ) <BR> <20><TT
CLASS="REPLACEABLE"
><I
>command</I
></TT
>... <BR> <20>;; <BR><BR><BR><B
CLASS="COMMAND"
>esac</B
> </P
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
> <P
></P
><UL
><LI
><P
>Quoting the variables is not mandatory,
since word splitting does not take place.</P
></LI
><LI
><P
><A
NAME="CASEPAREN"
></A
>Each test line
ends with a right paren <B
CLASS="COMMAND"
>)</B
>.
<A
NAME="AEN7087"
HREF="#FTN.AEN7087"
><SPAN
CLASS="footnote"
>[54]</SPAN
></A
>
</P
></LI
><LI
><P
>Each condition block ends
with a <EM
>double</EM
> semicolon
<SPAN
CLASS="TOKEN"
>;;</SPAN
>.</P
></LI
><LI
><P
>If a condition tests
<I
CLASS="FIRSTTERM"
>true</I
>, then the associated
commands execute and the <B
CLASS="COMMAND"
>case</B
>
block terminates.</P
></LI
><LI
><P
>The entire <B
CLASS="COMMAND"
>case</B
>
block ends with an <B
CLASS="COMMAND"
>esac</B
>
(<I
CLASS="WORDASWORD"
>case</I
> spelled backwards).</P
></LI
></UL
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX29"
></A
><P
><B
>Example 11-25. Using <I
CLASS="FIRSTTERM"
>case</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Testing ranges of characters.
echo; echo "Hit a key, then hit return."
read Keypress
case "$Keypress" in
[[:lower:]] ) echo "Lowercase letter";;
[[:upper:]] ) echo "Uppercase letter";;
[0-9] ) echo "Digit";;
* ) echo "Punctuation, whitespace, or other";;
esac # Allows ranges of characters in [square brackets],
#+ or POSIX ranges in [[double square brackets.
# In the first version of this example,
#+ the tests for lowercase and uppercase characters were
#+ [a-z] and [A-Z].
# This no longer works in certain locales and/or Linux distros.
# POSIX is more portable.
# Thanks to Frank Wang for pointing this out.
# Exercise:
# --------
# As the script stands, it accepts a single keystroke, then terminates.
# Change the script so it accepts repeated input,
#+ reports on each keystroke, and terminates only when "X" is hit.
# Hint: enclose everything in a "while" loop.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX30"
></A
><P
><B
>Example 11-26. Creating menus using <I
CLASS="FIRSTTERM"
>case</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Crude address database
clear # Clear the screen.
echo " Contact List"
echo " ------- ----"
echo "Choose one of the following persons:"
echo
echo "[E]vans, Roland"
echo "[J]ones, Mildred"
echo "[S]mith, Julie"
echo "[Z]ane, Morris"
echo
read person
case "$person" in
# Note variable is quoted.
"E" | "e" )
# Accept upper or lowercase input.
echo
echo "Roland Evans"
echo "4321 Flash Dr."
echo "Hardscrabble, CO 80753"
echo "(303) 734-9874"
echo "(303) 734-9892 fax"
echo "revans@zzy.net"
echo "Business partner &#38; old friend"
;;
# Note double semicolon to terminate each option.
"J" | "j" )
echo
echo "Mildred Jones"
echo "249 E. 7th St., Apt. 19"
echo "New York, NY 10009"
echo "(212) 533-2814"
echo "(212) 533-9972 fax"
echo "milliej@loisaida.com"
echo "Ex-girlfriend"
echo "Birthday: Feb. 11"
;;
# Add info for Smith &#38; Zane later.
* )
# Default option.
# Empty input (hitting RETURN) fits here, too.
echo
echo "Not yet in database."
;;
esac
echo
# Exercise:
# --------
# Change the script so it accepts multiple inputs,
#+ instead of terminating after displaying just one address.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="CASECL"
></A
></P
><P
>An exceptionally clever use of <B
CLASS="COMMAND"
>case</B
>
involves testing for command-line parameters.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
case "$1" in
"") echo "Usage: ${0##*/} &#60;filename&#62;"; exit $E_PARAM;;
# No command-line parameters,
# or first parameter empty.
# Note that ${0##*/} is ${var##pattern} param substitution.
# Net result is $0.
-*) FILENAME=./$1;; # If filename passed as argument ($1)
#+ starts with a dash,
#+ replace it with ./$1
#+ so further commands don't interpret it
#+ as an option.
* ) FILENAME=$1;; # Otherwise, $1.
esac</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Here is a more straightforward example of
command-line parameter handling:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
while [ $# -gt 0 ]; do # Until you run out of parameters . . .
case "$1" in
-d|--debug)
# "-d" or "--debug" parameter?
DEBUG=1
;;
-c|--conf)
CONFFILE="$2"
shift
if [ ! -f $CONFFILE ]; then
echo "Error: Supplied file doesn't exist!"
exit $E_CONFFILE # File not found error.
fi
;;
esac
shift # Check next set of parameters.
done
# From Stefano Falsetto's "Log2Rot" script,
#+ part of his "rottlog" package.
# Used with permission.</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="CASECMD"
></A
><P
><B
>Example 11-27. Using <I
CLASS="FIRSTTERM"
>command substitution</I
>
to generate the <I
CLASS="FIRSTTERM"
>case</I
> variable</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# case-cmd.sh: Using command substitution to generate a "case" variable.
case $( arch ) in # $( arch ) returns machine architecture.
# Equivalent to 'uname -m' ...
i386 ) echo "80386-based machine";;
i486 ) echo "80486-based machine";;
i586 ) echo "Pentium-based machine";;
i686 ) echo "Pentium2+-based machine";;
* ) echo "Other type of machine";;
esac
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="CSGLOB"
></A
></P
><P
>A <B
CLASS="COMMAND"
>case</B
> construct can filter strings for
<A
HREF="#GLOBBINGREF"
>globbing</A
> patterns.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="MATCHSTRING"
></A
><P
><B
>Example 11-28. Simple string matching</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# match-string.sh: Simple string matching
# using a 'case' construct.
match_string ()
{ # Exact string match.
MATCH=0
E_NOMATCH=90
PARAMS=2 # Function requires 2 arguments.
E_BAD_PARAMS=91
[ $# -eq $PARAMS ] || return $E_BAD_PARAMS
case "$1" in
"$2") return $MATCH;;
* ) return $E_NOMATCH;;
esac
}
a=one
b=two
c=three
d=two
match_string $a # wrong number of parameters
echo $? # 91
match_string $a $b # no match
echo $? # 90
match_string $b $d # match
echo $? # 0
exit 0 </PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ISALPHA"
></A
><P
><B
>Example 11-29. Checking for alphabetic input</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# isalpha.sh: Using a "case" structure to filter a string.
SUCCESS=0
FAILURE=1 # Was FAILURE=-1,
#+ but Bash no longer allows negative return value.
isalpha () # Tests whether *first character* of input string is alphabetic.
{
if [ -z "$1" ] # No argument passed?
then
return $FAILURE
fi
case "$1" in
[a-zA-Z]*) return $SUCCESS;; # Begins with a letter?
* ) return $FAILURE;;
esac
} # Compare this with "isalpha ()" function in C.
isalpha2 () # Tests whether *entire string* is alphabetic.
{
[ $# -eq 1 ] || return $FAILURE
case $1 in
*[!a-zA-Z]*|"") return $FAILURE;;
*) return $SUCCESS;;
esac
}
isdigit () # Tests whether *entire string* is numerical.
{ # In other words, tests for integer variable.
[ $# -eq 1 ] || return $FAILURE
case $1 in
*[!0-9]*|"") return $FAILURE;;
*) return $SUCCESS;;
esac
}
check_var () # Front-end to isalpha ().
{
if isalpha "$@"
then
echo "\"$*\" begins with an alpha character."
if isalpha2 "$@"
then # No point in testing if first char is non-alpha.
echo "\"$*\" contains only alpha characters."
else
echo "\"$*\" contains at least one non-alpha character."
fi
else
echo "\"$*\" begins with a non-alpha character."
# Also "non-alpha" if no argument passed.
fi
echo
}
digit_check () # Front-end to isdigit ().
{
if isdigit "$@"
then
echo "\"$*\" contains only digits [0 - 9]."
else
echo "\"$*\" has at least one non-digit character."
fi
echo
}
a=23skidoo
b=H3llo
c=-What?
d=What?
e=$(echo $b) # Command substitution.
f=AbcDef
g=27234
h=27a34
i=27.34
check_var $a
check_var $b
check_var $c
check_var $d
check_var $e
check_var $f
check_var # No argument passed, so what happens?
#
digit_check $g
digit_check $h
digit_check $i
exit 0 # Script improved by S.C.
# Exercise:
# --------
# Write an 'isfloat ()' function that tests for floating point numbers.
# Hint: The function duplicates 'isdigit ()',
#+ but adds a test for a mandatory decimal point.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SELECTREF"
></A
><B
CLASS="COMMAND"
>select</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>select</B
> construct, adopted from the Korn
Shell, is yet another tool for building menus.</P
><P
><P
><B
CLASS="COMMAND"
>select</B
> <TT
CLASS="REPLACEABLE"
><I
>variable</I
></TT
> [in <TT
CLASS="REPLACEABLE"
><I
>list</I
></TT
>]<BR> do <BR> <20><TT
CLASS="REPLACEABLE"
><I
>command</I
></TT
>... <BR> <20>break <BR> done </P
></P
><P
>This prompts the user to enter one of the choices presented in the
variable list. Note that <B
CLASS="COMMAND"
>select</B
> uses the
<TT
CLASS="VARNAME"
>$PS3</TT
> prompt (<TT
CLASS="PROMPT"
>#? </TT
>) by default,
but this may be changed.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX31"
></A
><P
><B
>Example 11-30. Creating menus using <I
CLASS="FIRSTTERM"
>select</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
PS3='Choose your favorite vegetable: ' # Sets the prompt string.
# Otherwise it defaults to #? .
echo
select vegetable in "beans" "carrots" "potatoes" "onions" "rutabagas"
do
echo
echo "Your favorite veggie is $vegetable."
echo "Yuck!"
echo
break # What happens if there is no 'break' here?
done
exit
# Exercise:
# --------
# Fix this script to accept user input not specified in
#+ the "select" statement.
# For example, if the user inputs "peas,"
#+ the script would respond "Sorry. That is not on the menu."</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="INLISTOMIT"
></A
></P
><P
>If <TT
CLASS="USERINPUT"
><B
>in <TT
CLASS="REPLACEABLE"
><I
>list</I
></TT
></B
></TT
> is
omitted, then <B
CLASS="COMMAND"
>select</B
> uses the list of command
line arguments (<TT
CLASS="VARNAME"
>$@</TT
>) passed to the script or
the function containing the <B
CLASS="COMMAND"
>select</B
>
construct.</P
><P
>Compare this to the behavior of a
<P
><B
CLASS="COMMAND"
>for</B
> <TT
CLASS="REPLACEABLE"
><I
>variable</I
></TT
> [in <TT
CLASS="REPLACEABLE"
><I
>list</I
></TT
>]</P
>
construct with the
<TT
CLASS="USERINPUT"
><B
>in <TT
CLASS="REPLACEABLE"
><I
>list</I
></TT
></B
></TT
>
omitted.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX32"
></A
><P
><B
>Example 11-31. Creating menus using <I
CLASS="FIRSTTERM"
>select</I
>
in a function</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
PS3='Choose your favorite vegetable: '
echo
choice_of()
{
select vegetable
# [in list] omitted, so 'select' uses arguments passed to function.
do
echo
echo "Your favorite veggie is $vegetable."
echo "Yuck!"
echo
break
done
}
choice_of beans rice carrots radishes rutabaga spinach
# $1 $2 $3 $4 $5 $6
# passed to choice_of() function
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#RESISTOR"
>Example 37-3</A
>.</P
></DD
></DL
></DIV
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="COMMANDSUB"
></A
>Chapter 12. Command Substitution</H1
><P
> <A
NAME="COMMANDSUBREF"
></A
><B
CLASS="COMMAND"
>Command
substitution</B
> reassigns the output of a command
<A
NAME="AEN7205"
HREF="#FTN.AEN7205"
><SPAN
CLASS="footnote"
>[55]</SPAN
></A
>
or even multiple commands; it literally plugs the command
output into another context.
<A
NAME="AEN7211"
HREF="#FTN.AEN7211"
><SPAN
CLASS="footnote"
>[56]</SPAN
></A
>
</P
><P
><A
NAME="BACKQUOTESREF"
></A
>The classic form of command
substitution uses <I
CLASS="FIRSTTERM"
>backquotes</I
>
(`...`). Commands within backquotes (backticks) generate
command-line text.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>script_name=`basename $0`
echo "The name of this script is $script_name."</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="FORMALPARA"
><P
><B
>The output of commands can be used as arguments to
another command, to set a variable, and even for generating
the argument list in a <A
HREF="#FORLOOPREF1"
>for</A
>
loop. </B
></P
></DIV
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>rm `cat filename` # <SPAN
CLASS="QUOTE"
>"filename"</SPAN
> contains a list of files to delete.
#
# S. C. points out that "arg list too long" error might result.
# Better is xargs rm -- &#60; filename
# ( -- covers those cases where <SPAN
CLASS="QUOTE"
>"filename"</SPAN
> begins with a <SPAN
CLASS="QUOTE"
>"-"</SPAN
> )
textfile_listing=`ls *.txt`
# Variable contains names of all *.txt files in current working directory.
echo $textfile_listing
textfile_listing2=$(ls *.txt) # The alternative form of command substitution.
echo $textfile_listing2
# Same result.
# A possible problem with putting a list of files into a single string
# is that a newline may creep in.
#
# A safer way to assign a list of files to a parameter is with an array.
# shopt -s nullglob # If no match, filename expands to nothing.
# textfile_listing=( *.txt )
#
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="CSSUBSH"
></A
>Command substitution
invokes a <A
HREF="#SUBSHELLSREF"
>subshell</A
>.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="CSWS"
></A
>Command substitution may
result in <A
HREF="#WSPLITREF"
>word splitting</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>COMMAND `echo a b` # 2 args: a and b
COMMAND "`echo a b`" # 1 arg: "a b"
COMMAND `echo` # no arg
COMMAND "`echo`" # one empty arg
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="CSTRNL"
></A
></P
><P
>Even when there is no word splitting, command
substitution can remove trailing newlines.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># cd "`pwd`" # This should always work.
# However...
mkdir 'dir with trailing newline
'
cd 'dir with trailing newline
'
cd "`pwd`" # Error message:
# bash: cd: /tmp/file with trailing newline: No such file or directory
cd "$PWD" # Works fine.
old_tty_setting=$(stty -g) # Save old terminal setting.
echo "Hit a key "
stty -icanon -echo # Disable "canonical" mode for terminal.
# Also, disable *local* echo.
key=$(dd bs=1 count=1 2&#62; /dev/null) # Using 'dd' to get a keypress.
stty "$old_tty_setting" # Restore old setting.
echo "You hit ${#key} key." # ${#variable} = number of characters in $variable
#
# Hit any key except RETURN, and the output is "You hit 1 key."
# Hit RETURN, and it's "You hit 0 key."
# The newline gets eaten in the command substitution.
#Code snippet by St<53>phane Chazelas.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Using <B
CLASS="COMMAND"
>echo</B
> to output an
<I
CLASS="FIRSTTERM"
>unquoted</I
> variable set with command
substitution removes trailing newlines characters from
the output of the reassigned command(s). This can cause
unpleasant surprises.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>dir_listing=`ls -l`
echo $dir_listing # unquoted
# Expecting a nicely ordered directory listing.
# However, what you get is:
# total 3 -rw-rw-r-- 1 bozo bozo 30 May 13 17:15 1.txt -rw-rw-r-- 1 bozo
# bozo 51 May 15 20:57 t2.sh -rwxr-xr-x 1 bozo bozo 217 Mar 5 21:13 wi.sh
# The newlines disappeared.
echo "$dir_listing" # quoted
# -rw-rw-r-- 1 bozo 30 May 13 17:15 1.txt
# -rw-rw-r-- 1 bozo 51 May 15 20:57 t2.sh
# -rwxr-xr-x 1 bozo 217 Mar 5 21:13 wi.sh</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><P
>Command substitution even permits setting a variable to the
contents of a file, using either <A
HREF="#IOREDIRREF"
>redirection</A
> or the <A
HREF="#CATREF"
>cat</A
> command.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>variable1=`&#60;file1` # Set "variable1" to contents of "file1".
variable2=`cat file2` # Set "variable2" to contents of "file2".
# This, however, forks a new process,
#+ so the line of code executes slower than the above version.
# Note that the variables may contain embedded whitespace,
#+ or even (horrors), control characters.
# It is not necessary to explicitly assign a variable.
echo "` &#60;$0`" # Echoes the script itself to stdout.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Excerpts from system file, /etc/rc.d/rc.sysinit
#+ (on a Red Hat Linux installation)
if [ -f /fsckoptions ]; then
fsckoptions=`cat /fsckoptions`
...
fi
#
#
if [ -e "/proc/ide/${disk[$device]}/media" ] ; then
hdmedia=`cat /proc/ide/${disk[$device]}/media`
...
fi
#
#
if [ ! -n "`uname -r | grep -- "-"`" ]; then
ktag="`cat /proc/version`"
...
fi
#
#
if [ $usb = "1" ]; then
sleep 5
mouseoutput=`cat /proc/bus/usb/devices 2&#62;/dev/null|grep -E "^I.*Cls=03.*Prot=02"`
kbdoutput=`cat /proc/bus/usb/devices 2&#62;/dev/null|grep -E "^I.*Cls=03.*Prot=01"`
...
fi</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Do not set a variable to the contents of a
<EM
>long</EM
> text file unless you have a very good
reason for doing so. Do not set a variable to the contents of a
<I
CLASS="FIRSTTERM"
>binary</I
> file, even as a joke.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="STUPSCR"
></A
><P
><B
>Example 12-1. Stupid script tricks</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# stupid-script-tricks.sh: Don't try this at home, folks.
# From "Stupid Script Tricks," Volume I.
exit 99 ### Comment out this line if you dare.
dangerous_variable=`cat /boot/vmlinuz` # The compressed Linux kernel itself.
echo "string-length of \$dangerous_variable = ${#dangerous_variable}"
# string-length of $dangerous_variable = 794151
# (Newer kernels are bigger.)
# Does not give same count as 'wc -c /boot/vmlinuz'.
# echo "$dangerous_variable"
# Don't try this! It would hang the script.
# The document author is aware of no useful applications for
#+ setting a variable to the contents of a binary file.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Notice that a <I
CLASS="FIRSTTERM"
>buffer overrun</I
>
does not occur. This is one instance where an interpreted
language, such as Bash, provides more protection from
programmer mistakes than a compiled language.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="CSVL"
></A
></P
><P
>Command substitution permits setting a variable to the
output of a <A
HREF="#FORLOOPREF1"
>loop</A
>. The
key to this is grabbing the output of an <A
HREF="#ECHOREF"
>echo</A
> command within the
loop.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="CSUBLOOP"
></A
><P
><B
>Example 12-2. Generating a variable from a loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# csubloop.sh: Setting a variable to the output of a loop.
variable1=`for i in 1 2 3 4 5
do
echo -n "$i" # The 'echo' command is critical
done` #+ to command substitution here.
echo "variable1 = $variable1" # variable1 = 12345
i=0
variable2=`while [ "$i" -lt 10 ]
do
echo -n "$i" # Again, the necessary 'echo'.
let "i += 1" # Increment.
done`
echo "variable2 = $variable2" # variable2 = 0123456789
# Demonstrates that it's possible to embed a loop
#+ inside a variable declaration.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="CSTOOLSET"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN7273"
></A
><P
></P
><P
>Command substitution makes it possible to extend the
toolset available to Bash. It is simply a matter
of writing a program or script that outputs to
<TT
CLASS="FILENAME"
>stdout</TT
> (like a well-behaved UNIX
tool should) and assigning that output to a variable.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#include &#60;stdio.h&#62;
/* "Hello, world." C program */
int main()
{
printf( "Hello, world.\n" );
return (0);
}</PRE
></FONT
></TD
></TR
></TABLE
>
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>gcc -o hello hello.c</B
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# hello.sh
greeting=`./hello`
echo $greeting</PRE
></FONT
></TD
></TR
></TABLE
>
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>sh hello.sh</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Hello, world.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
></P
></DIV
></TD
></TR
></TABLE
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="CSPARENS"
></A
>The <B
CLASS="COMMAND"
>$(...)</B
>
form has superseded backticks for command
substitution.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>output=$(sed -n /"$1"/p $file) # From "grp.sh" example.
# Setting a variable to the contents of a text file.
File_contents1=$(cat $file1)
File_contents2=$(&#60;$file2) # Bash permits this also.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The <B
CLASS="COMMAND"
>$(...)</B
> form of command substitution
treats a double backslash in a different way than
<B
CLASS="COMMAND"
>`...`</B
>.</P
><P
>
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo `echo \\`</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $(echo \\)</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>\</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="CSNEST"
></A
></P
><P
>The <B
CLASS="COMMAND"
>$(...)</B
> form of command
substitution permits nesting.
<A
NAME="AEN7308"
HREF="#FTN.AEN7308"
><SPAN
CLASS="footnote"
>[57]</SPAN
></A
>
</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>word_count=$( wc -w $(echo * | awk '{print $8}') )</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Or, for something a bit more elaborate . . .</P
><DIV
CLASS="EXAMPLE"
><A
NAME="AGRAM2"
></A
><P
><B
>Example 12-3. Finding anagrams</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# agram2.sh
# Example of nested command substitution.
# Uses "anagram" utility
#+ that is part of the author's "yawl" word list package.
# http://ibiblio.org/pub/Linux/libs/yawl-0.3.2.tar.gz
# http://bash.deta.in/yawl-0.3.2.tar.gz
E_NOARGS=86
E_BADARG=87
MINLEN=7
if [ -z "$1" ]
then
echo "Usage $0 LETTERSET"
exit $E_NOARGS # Script needs a command-line argument.
elif [ ${#1} -lt $MINLEN ]
then
echo "Argument must have at least $MINLEN letters."
exit $E_BADARG
fi
FILTER='.......' # Must have at least 7 letters.
# 1234567
Anagrams=( $(echo $(anagram $1 | grep $FILTER) ) )
# $( $( nested command sub. ) )
# ( array assignment )
echo
echo "${#Anagrams[*]} 7+ letter anagrams found"
echo
echo ${Anagrams[0]} # First anagram.
echo ${Anagrams[1]} # Second anagram.
# Etc.
# echo "${Anagrams[*]}" # To list all the anagrams in a single line . . .
# Look ahead to the Arrays chapter for enlightenment on
#+ what's going on here.
# See also the agram.sh script for an exercise in anagram finding.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></TD
></TR
></TABLE
></DIV
><P
>Examples of command substitution in shell scripts:
<P
></P
><OL
TYPE="1"
><LI
><P
><A
HREF="#BINGREP"
>Example 11-8</A
></P
></LI
><LI
><P
><A
HREF="#CASECMD"
>Example 11-27</A
></P
></LI
><LI
><P
><A
HREF="#SEEDINGRANDOM"
>Example 9-16</A
></P
></LI
><LI
><P
><A
HREF="#EX57"
>Example 16-3</A
></P
></LI
><LI
><P
><A
HREF="#LOWERCASE"
>Example 16-22</A
></P
></LI
><LI
><P
><A
HREF="#GRP"
>Example 16-17</A
></P
></LI
><LI
><P
><A
HREF="#EX53"
>Example 16-54</A
></P
></LI
><LI
><P
><A
HREF="#EX24"
>Example 11-14</A
></P
></LI
><LI
><P
><A
HREF="#SYMLINKS"
>Example 11-11</A
></P
></LI
><LI
><P
><A
HREF="#STRIPC"
>Example 16-32</A
></P
></LI
><LI
><P
><A
HREF="#REDIR4"
>Example 20-8</A
></P
></LI
><LI
><P
><A
HREF="#TREE"
>Example A-16</A
></P
></LI
><LI
><P
><A
HREF="#PIDID"
>Example 29-3</A
></P
></LI
><LI
><P
><A
HREF="#MONTHLYPMT"
>Example 16-47</A
></P
></LI
><LI
><P
><A
HREF="#BASE"
>Example 16-48</A
></P
></LI
><LI
><P
><A
HREF="#ALTBC"
>Example 16-49</A
></P
></LI
></OL
>
</P
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="ARITHEXP"
></A
>Chapter 13. Arithmetic Expansion</H1
><P
><A
NAME="ARITHEXPREF"
></A
>Arithmetic expansion provides a
powerful tool for performing (integer) arithmetic
operations in scripts. Translating a string into a
numerical expression is relatively straightforward using
<I
CLASS="FIRSTTERM"
>backticks</I
>, <I
CLASS="FIRSTTERM"
>double
parentheses</I
>, or <I
CLASS="FIRSTTERM"
>let</I
>.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="ARITHEXPVAR1"
></A
>Variations</B
></P
><DL
><DT
>Arithmetic expansion with <A
HREF="#BACKQUOTESREF"
>backticks</A
> (often used in
conjunction with <A
HREF="#EXPRREF"
>expr</A
>)</DT
><DD
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>z=`expr $z + 3` # The 'expr' command performs the expansion.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
>Arithmetic expansion with <A
HREF="#DBLPARENS"
>double
parentheses</A
>, and using <A
HREF="#LETREF"
>let</A
></DT
><DD
><P
>The use of <I
CLASS="FIRSTTERM"
>backticks</I
>
(<I
CLASS="FIRSTTERM"
>backquotes</I
>) in arithmetic
expansion has been superseded by <I
CLASS="FIRSTTERM"
>double
parentheses</I
> --
<TT
CLASS="USERINPUT"
><B
>((...))</B
></TT
> and
<TT
CLASS="USERINPUT"
><B
>$((...))</B
></TT
> -- and also by the very
convenient <A
HREF="#LETREF"
>let</A
> construction.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>z=$(($z+3))
z=$((z+3)) # Also correct.
# Within double parentheses,
#+ parameter dereferencing
#+ is optional.
# $((EXPRESSION)) is arithmetic expansion. # Not to be confused with
#+ command substitution.
# You may also use operations within double parentheses without assignment.
n=0
echo "n = $n" # n = 0
(( n += 1 )) # Increment.
# (( $n += 1 )) is incorrect!
echo "n = $n" # n = 1
let z=z+3
let "z += 3" # Quotes permit the use of spaces in variable assignment.
# The 'let' operator actually performs arithmetic evaluation,
#+ rather than expansion.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Examples of arithmetic expansion in scripts:
<P
></P
><OL
TYPE="1"
><LI
><P
><A
HREF="#EX45"
>Example 16-9</A
></P
></LI
><LI
><P
><A
HREF="#EX25"
>Example 11-15</A
></P
></LI
><LI
><P
><A
HREF="#EX66"
>Example 27-1</A
></P
></LI
><LI
><P
><A
HREF="#BUBBLE"
>Example 27-11</A
></P
></LI
><LI
><P
><A
HREF="#TREE"
>Example A-16</A
></P
></LI
></OL
>
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="RECESS-TIME"
></A
>Chapter 14. Recess Time</H1
><P
><EM
> This bizarre little intermission gives the reader a chance to
relax and maybe laugh a bit.
</EM
></P
><A
NAME="AEN7432"
></A
><BLOCKQUOTE
CLASS="BLOCKQUOTE"
><P
CLASS="LITERALLAYOUT"
>&nbsp;&nbsp;<br>
<br>
&nbsp;&nbsp;Fellow&nbsp;Linux&nbsp;user,&nbsp;greetings!&nbsp;You&nbsp;are&nbsp;reading&nbsp;something&nbsp;which<br>
&nbsp;&nbsp;will&nbsp;bring&nbsp;you&nbsp;luck&nbsp;and&nbsp;good&nbsp;fortune.&nbsp;Just&nbsp;e-mail&nbsp;a&nbsp;copy&nbsp;of<br>
&nbsp;&nbsp;this&nbsp;document&nbsp;to&nbsp;10&nbsp;of&nbsp;your&nbsp;friends.&nbsp;Before&nbsp;making&nbsp;the&nbsp;copies,<br>
&nbsp;&nbsp;send&nbsp;a&nbsp;100-line&nbsp;Bash&nbsp;script&nbsp;to&nbsp;the&nbsp;first&nbsp;person&nbsp;on&nbsp;the&nbsp;list<br>
&nbsp;&nbsp;at&nbsp;the&nbsp;bottom&nbsp;of&nbsp;this&nbsp;letter.&nbsp;Then&nbsp;delete&nbsp;their&nbsp;name&nbsp;and&nbsp;add<br>
&nbsp;&nbsp;yours&nbsp;to&nbsp;the&nbsp;bottom&nbsp;of&nbsp;the&nbsp;list.<br>
<br>
&nbsp;&nbsp;Don't&nbsp;break&nbsp;the&nbsp;chain!&nbsp;Make&nbsp;the&nbsp;copies&nbsp;within&nbsp;48&nbsp;hours.<br>
&nbsp;&nbsp;Wilfred&nbsp;P.&nbsp;of&nbsp;Brooklyn&nbsp;failed&nbsp;to&nbsp;send&nbsp;out&nbsp;his&nbsp;ten&nbsp;copies&nbsp;and<br>
&nbsp;&nbsp;woke&nbsp;the&nbsp;next&nbsp;morning&nbsp;to&nbsp;find&nbsp;his&nbsp;job&nbsp;description&nbsp;changed<br>
&nbsp;&nbsp;to&nbsp;"COBOL&nbsp;programmer."&nbsp;Howard&nbsp;L.&nbsp;of&nbsp;Newport&nbsp;News&nbsp;sent<br>
&nbsp;&nbsp;out&nbsp;his&nbsp;ten&nbsp;copies&nbsp;and&nbsp;within&nbsp;a&nbsp;month&nbsp;had&nbsp;enough&nbsp;hardware<br>
&nbsp;&nbsp;to&nbsp;build&nbsp;a&nbsp;100-node&nbsp;Beowulf&nbsp;cluster&nbsp;dedicated&nbsp;to&nbsp;playing<br>
&nbsp;&nbsp;<EM
>Tuxracer</EM
>.&nbsp;Amelia&nbsp;V.&nbsp;of&nbsp;Chicago&nbsp;laughed&nbsp;at&nbsp;this&nbsp;letter<br>
&nbsp;&nbsp;and&nbsp;broke&nbsp;the&nbsp;chain.&nbsp;Shortly&nbsp;thereafter,&nbsp;a&nbsp;fire&nbsp;broke&nbsp;out<br>
&nbsp;&nbsp;in&nbsp;her&nbsp;terminal&nbsp;and&nbsp;she&nbsp;now&nbsp;spends&nbsp;her&nbsp;days&nbsp;writing<br>
&nbsp;&nbsp;documentation&nbsp;for&nbsp;MS&nbsp;Windows.<br>
<br>
&nbsp;&nbsp;Don't&nbsp;break&nbsp;the&nbsp;chain!&nbsp;&nbsp;Send&nbsp;out&nbsp;your&nbsp;ten&nbsp;copies&nbsp;today!<br>
<br>
</P
></BLOCKQUOTE
><P
><EM
>Courtesy 'NIX "fortune cookies", with some
alterations and many apologies</EM
></P
></DIV
></DIV
><DIV
CLASS="PART"
><A
NAME="PART4"
></A
><DIV
CLASS="TITLEPAGE"
><H1
CLASS="TITLE"
>Part 4. Commands</H1
><DIV
CLASS="PARTINTRO"
><A
NAME="AEN7439"
></A
><P
><A
NAME="PART4A"
></A
></P
><P
>Mastering the commands on your Linux machine is an indispensable
prelude to writing effective shell scripts.</P
><P
>This section covers the following commands:</P
><P
></P
><UL
><LI
><P
><A
HREF="#DOTREF"
>.</A
>
(See also <A
HREF="#SOURCEREF"
>source</A
>)</P
></LI
><LI
><P
><A
HREF="#ACREF"
>ac</A
></P
></LI
><LI
><P
><A
HREF="#USERADDREF"
>adduser</A
></P
></LI
><LI
><P
><A
HREF="#AGETTYREF"
>agetty</A
></P
></LI
><LI
><P
><A
HREF="#AGREPREF"
>agrep</A
></P
></LI
><LI
><P
><A
HREF="#ARREF"
>ar</A
></P
></LI
><LI
><P
><A
HREF="#ARCHREF"
>arch</A
></P
></LI
><LI
><P
><A
HREF="#ATREF"
>at</A
></P
></LI
><LI
><P
><A
HREF="#AUTOLOADREF"
>autoload</A
></P
></LI
><LI
><P
><A
HREF="#AWKREF"
>awk</A
>
(See also <A
HREF="#AWKMATH"
>Using
<B
CLASS="COMMAND"
>awk</B
> for
math operations</A
>)</P
></LI
><LI
><P
><A
HREF="#BADBLOCKSREF"
>badblocks</A
></P
></LI
><LI
><P
><A
HREF="#BANNERREF"
>banner</A
></P
></LI
><LI
><P
><A
HREF="#BASENAMEREF"
>basename</A
></P
></LI
><LI
><P
><A
HREF="#BATCHREF"
>batch</A
></P
></LI
><LI
><P
><A
HREF="#BCREF"
>bc</A
></P
></LI
><LI
><P
><A
HREF="#BGREF"
>bg</A
></P
></LI
><LI
><P
><A
HREF="#BINDREF"
>bind</A
></P
></LI
><LI
><P
><A
HREF="#BISONREF"
>bison</A
></P
></LI
><LI
><P
><A
HREF="#BLTREF"
>builtin</A
></P
></LI
><LI
><P
><A
HREF="#BZGREPREF"
>bzgrep</A
></P
></LI
><LI
><P
><A
HREF="#BZIPREF"
>bzip2</A
></P
></LI
><LI
><P
><A
HREF="#CALREF"
>cal</A
></P
></LI
><LI
><P
><A
HREF="#CALLERREF"
>caller</A
></P
></LI
><LI
><P
><A
HREF="#CATREF"
>cat</A
></P
></LI
><LI
><P
><A
HREF="#CDREF"
>cd</A
></P
></LI
><LI
><P
><A
HREF="#CHATTRREF"
>chattr</A
></P
></LI
><LI
><P
><A
HREF="#CHFNREF"
>chfn</A
></P
></LI
><LI
><P
><A
HREF="#CHGRPREF"
>chgrp</A
></P
></LI
><LI
><P
><A
HREF="#CHKCONFIGREF"
>chkconfig</A
></P
></LI
><LI
><P
><A
HREF="#CHMODREF"
>chmod</A
></P
></LI
><LI
><P
><A
HREF="#CHOWNREF"
>chown</A
></P
></LI
><LI
><P
><A
HREF="#CHROOTREF"
>chroot</A
></P
></LI
><LI
><P
><A
HREF="#CKSUMREF"
>cksum</A
></P
></LI
><LI
><P
><A
HREF="#CLEARREF"
>clear</A
></P
></LI
><LI
><P
><A
HREF="#CLOCKREF"
>clock</A
></P
></LI
><LI
><P
><A
HREF="#CMPREF"
>cmp</A
></P
></LI
><LI
><P
><A
HREF="#COLREF"
>col</A
></P
></LI
><LI
><P
><A
HREF="#COLRMREF"
>colrm</A
></P
></LI
><LI
><P
><A
HREF="#COLUMNREF"
>column</A
></P
></LI
><LI
><P
><A
HREF="#COMMREF"
>comm</A
></P
></LI
><LI
><P
><A
HREF="#COMMANDREF"
>command</A
></P
></LI
><LI
><P
><A
HREF="#COMPGENREF"
>compgen</A
></P
></LI
><LI
><P
><A
HREF="#COMPLETEREF"
>complete</A
></P
></LI
><LI
><P
><A
HREF="#COMPRESSREF"
>compress</A
></P
></LI
><LI
><P
><A
HREF="#COPROCREF"
>coproc</A
></P
></LI
><LI
><P
><A
HREF="#CPREF"
>cp</A
></P
></LI
><LI
><P
><A
HREF="#CPIOREF"
>cpio</A
></P
></LI
><LI
><P
><A
HREF="#CRONREF"
>cron</A
></P
></LI
><LI
><P
><A
HREF="#CRYPTREF"
>crypt</A
></P
></LI
><LI
><P
><A
HREF="#CSPLITREF"
>csplit</A
></P
></LI
><LI
><P
><A
HREF="#CUREF"
>cu</A
></P
></LI
><LI
><P
><A
HREF="#CUTREF"
>cut</A
></P
></LI
><LI
><P
><A
HREF="#DATEREF"
>date</A
></P
></LI
><LI
><P
><A
HREF="#DCREF"
>dc</A
></P
></LI
><LI
><P
><A
HREF="#DDREF"
>dd</A
></P
></LI
><LI
><P
><A
HREF="#DEBUGFSREF"
>debugfs</A
></P
></LI
><LI
><P
><A
HREF="#DECLAREREF"
>declare</A
></P
></LI
><LI
><P
><A
HREF="#DEPMODREF"
>depmod</A
></P
></LI
><LI
><P
><A
HREF="#DFREF"
>df</A
></P
></LI
><LI
><P
><A
HREF="#DIALOGREF"
>dialog</A
></P
></LI
><LI
><P
><A
HREF="#DIFFREF"
>diff</A
></P
></LI
><LI
><P
><A
HREF="#DIFF3REF"
>diff3</A
></P
></LI
><LI
><P
><A
HREF="#DIFFSTATREF"
>diffstat</A
></P
></LI
><LI
><P
><A
HREF="#DIGREF"
>dig</A
></P
></LI
><LI
><P
><A
HREF="#DIRNAMEREF"
>dirname</A
></P
></LI
><LI
><P
><A
HREF="#DIRSD"
>dirs</A
></P
></LI
><LI
><P
><A
HREF="#DISOWNREF"
>disown</A
></P
></LI
><LI
><P
><A
HREF="#DMESGREF"
>dmesg</A
></P
></LI
><LI
><P
><A
HREF="#DOEXECREF"
>doexec</A
></P
></LI
><LI
><P
><A
HREF="#DOS2UNIXREF"
>dos2unix</A
></P
></LI
><LI
><P
><A
HREF="#DUREF"
>du</A
></P
></LI
><LI
><P
><A
HREF="#DUMPREF"
>dump</A
></P
></LI
><LI
><P
><A
HREF="#DUMPE2FSREF"
>dumpe2fs</A
></P
></LI
><LI
><P
><A
HREF="#E2FSCKREF"
>e2fsck</A
></P
></LI
><LI
><P
><A
HREF="#ECHOREF"
>echo</A
></P
></LI
><LI
><P
><A
HREF="#EGREPREF"
>egrep</A
></P
></LI
><LI
><P
><A
HREF="#ENABLEREF"
>enable</A
></P
></LI
><LI
><P
><A
HREF="#ENSCRIPTREF"
>enscript</A
></P
></LI
><LI
><P
><A
HREF="#ENVVREF"
>env</A
></P
></LI
><LI
><P
><A
HREF="#EQNREF"
>eqn</A
></P
></LI
><LI
><P
><A
HREF="#EVALREF"
>eval</A
></P
></LI
><LI
><P
><A
HREF="#EXECREF"
>exec</A
></P
></LI
><LI
><P
><A
HREF="#EXITREF"
>exit</A
>
(Related topic: <A
HREF="#EXITSTATUSREF"
>exit
status</A
>)</P
></LI
><LI
><P
><A
HREF="#EXPANDREF"
>expand</A
></P
></LI
><LI
><P
><A
HREF="#EXPORTREF"
>export</A
></P
></LI
><LI
><P
><A
HREF="#EXPRREF"
>expr</A
></P
></LI
><LI
><P
><A
HREF="#FACTORREF"
>factor</A
></P
></LI
><LI
><P
><A
HREF="#FALSEREF"
>false</A
></P
></LI
><LI
><P
><A
HREF="#FDFORMATREF"
>fdformat</A
></P
></LI
><LI
><P
><A
HREF="#FDISKREF"
>fdisk</A
></P
></LI
><LI
><P
><A
HREF="#FGREF"
>fg</A
></P
></LI
><LI
><P
><A
HREF="#FGREPREF"
>fgrep</A
></P
></LI
><LI
><P
><A
HREF="#FILEREF"
>file</A
></P
></LI
><LI
><P
><A
HREF="#FINDREF"
>find</A
></P
></LI
><LI
><P
><A
HREF="#FINGERREF"
>finger</A
></P
></LI
><LI
><P
><A
HREF="#FLEXREF"
>flex</A
></P
></LI
><LI
><P
><A
HREF="#FLOCKREF"
>flock</A
></P
></LI
><LI
><P
><A
HREF="#FMTREF"
>fmt</A
></P
></LI
><LI
><P
><A
HREF="#FOLDREF"
>fold</A
></P
></LI
><LI
><P
><A
HREF="#FREEREF"
>free</A
></P
></LI
><LI
><P
><A
HREF="#FSCKREF"
>fsck</A
></P
></LI
><LI
><P
><A
HREF="#FTPREF"
>ftp</A
></P
></LI
><LI
><P
><A
HREF="#FUSERREF"
>fuser</A
></P
></LI
><LI
><P
><A
HREF="#GETFACLREF"
>getfacl</A
></P
></LI
><LI
><P
><A
HREF="#GETOPTY"
>getopt</A
></P
></LI
><LI
><P
><A
HREF="#GETOPTSX"
>getopts</A
></P
></LI
><LI
><P
><A
HREF="#GETTEXTREF"
>gettext</A
></P
></LI
><LI
><P
><A
HREF="#GETTYREF"
>getty</A
></P
></LI
><LI
><P
><A
HREF="#GNOMEMOUNTREF"
>gnome-mount</A
></P
></LI
><LI
><P
><A
HREF="#GREPREF"
>grep</A
></P
></LI
><LI
><P
><A
HREF="#GROFFREF"
>groff</A
></P
></LI
><LI
><P
><A
HREF="#GROUPMODREF"
>groupmod</A
></P
></LI
><LI
><P
><A
HREF="#GROUPSCMDREF"
>groups</A
>
(Related topic: the <A
HREF="#GROUPSREF"
>$GROUPS</A
>
variable)</P
></LI
><LI
><P
><A
HREF="#GSREF"
>gs</A
></P
></LI
><LI
><P
><A
HREF="#GZIPREF"
>gzip</A
></P
></LI
><LI
><P
><A
HREF="#HALTREF"
>halt</A
></P
></LI
><LI
><P
><A
HREF="#HASHCMDREF"
>hash</A
></P
></LI
><LI
><P
><A
HREF="#HDPARMREF"
>hdparm</A
></P
></LI
><LI
><P
><A
HREF="#HEADREF"
>head</A
></P
></LI
><LI
><P
><A
HREF="#HELPREF"
>help</A
></P
></LI
><LI
><P
><A
HREF="#HEXDUMPREF"
>hexdump</A
></P
></LI
><LI
><P
><A
HREF="#HOSTREF"
>host</A
></P
></LI
><LI
><P
><A
HREF="#HOSTIDREF"
>hostid</A
></P
></LI
><LI
><P
><A
HREF="#HNAMEREF"
>hostname</A
>
(Related topic: the <A
HREF="#HOSTNAMEREF"
>$HOSTNAME</A
>
variable)</P
></LI
><LI
><P
><A
HREF="#HWCLOCKREF"
>hwclock</A
></P
></LI
><LI
><P
><A
HREF="#ICONVREF"
>iconv</A
></P
></LI
><LI
><P
><A
HREF="#IDREF"
>id</A
>
(Related topic: the <A
HREF="#UIDREF"
>$UID</A
>
variable)</P
></LI
><LI
><P
><A
HREF="#IFCONFIGREF"
>ifconfig</A
></P
></LI
><LI
><P
><A
HREF="#INFOREF"
>info</A
></P
></LI
><LI
><P
><A
HREF="#INFOCMPREF"
>infocmp</A
></P
></LI
><LI
><P
><A
HREF="#INITREF"
>init</A
></P
></LI
><LI
><P
><A
HREF="#INSMODREF"
>insmod</A
></P
></LI
><LI
><P
><A
HREF="#INSTALLREF"
>install</A
></P
></LI
><LI
><P
><A
HREF="#IPREF"
>ip</A
></P
></LI
><LI
><P
><A
HREF="#IPCALCREF"
>ipcalc</A
></P
></LI
><LI
><P
><A
HREF="#IPTABLESREF"
>iptables</A
></P
></LI
><LI
><P
><A
HREF="#IWCONFIGREF"
>iwconfig</A
></P
></LI
><LI
><P
><A
HREF="#JOBSREF"
>jobs</A
></P
></LI
><LI
><P
><A
HREF="#JOINREF"
>join</A
></P
></LI
><LI
><P
><A
HREF="#JOTREF"
>jot</A
></P
></LI
><LI
><P
><A
HREF="#KILLREF"
>kill</A
></P
></LI
><LI
><P
><A
HREF="#KILLALLREF"
>killall</A
></P
></LI
><LI
><P
><A
HREF="#LASTREF"
>last</A
></P
></LI
><LI
><P
><A
HREF="#LASTCOMMREF"
>lastcomm</A
></P
></LI
><LI
><P
><A
HREF="#LASTLOGREF"
>lastlog</A
></P
></LI
><LI
><P
><A
HREF="#LDDREF"
>ldd</A
></P
></LI
><LI
><P
><A
HREF="#LESSREF"
>less</A
></P
></LI
><LI
><P
><A
HREF="#LETREF"
>let</A
></P
></LI
><LI
><P
><A
HREF="#LEXREF"
>lex</A
></P
></LI
><LI
><P
><A
HREF="#LIDREF"
>lid</A
></P
></LI
><LI
><P
><A
HREF="#LINKREF"
>ln</A
></P
></LI
><LI
><P
><A
HREF="#LOCATEREF"
>locate</A
></P
></LI
><LI
><P
><A
HREF="#LOCKFILEREF"
>lockfile</A
></P
></LI
><LI
><P
><A
HREF="#LOGGERREF"
>logger</A
></P
></LI
><LI
><P
><A
HREF="#LOGNAMEREF"
>logname</A
></P
></LI
><LI
><P
><A
HREF="#LOGOUTREF"
>logout</A
></P
></LI
><LI
><P
><A
HREF="#LOGROTATEREF"
>logrotate</A
></P
></LI
><LI
><P
><A
HREF="#LOOKREF"
>look</A
></P
></LI
><LI
><P
><A
HREF="#LOSETUPREF"
>losetup</A
></P
></LI
><LI
><P
><A
HREF="#LPREF"
>lp</A
></P
></LI
><LI
><P
><A
HREF="#LSREF"
>ls</A
></P
></LI
><LI
><P
><A
HREF="#LSDEVREF"
>lsdev</A
></P
></LI
><LI
><P
><A
HREF="#LSMODREF"
>lsmod</A
></P
></LI
><LI
><P
><A
HREF="#LSOFREF"
>lsof</A
></P
></LI
><LI
><P
><A
HREF="#LSPCIREF"
>lspci</A
></P
></LI
><LI
><P
><A
HREF="#LSUSBREF"
>lsusb</A
></P
></LI
><LI
><P
><A
HREF="#LTRACEREF"
>ltrace</A
></P
></LI
><LI
><P
><A
HREF="#LYNXREF"
>lynx</A
></P
></LI
><LI
><P
><A
HREF="#LZMAREF"
>lzcat</A
></P
></LI
><LI
><P
><A
HREF="#LZMAREF"
>lzma</A
></P
></LI
><LI
><P
><A
HREF="#M4REF"
>m4</A
></P
></LI
><LI
><P
><A
HREF="#COMMMAIL1"
>mail</A
></P
></LI
><LI
><P
><A
HREF="#MAILSTATSREF"
>mailstats</A
></P
></LI
><LI
><P
><A
HREF="#MAILTOREF"
>mailto</A
></P
></LI
><LI
><P
><A
HREF="#MAKEREF"
>make</A
></P
></LI
><LI
><P
><A
HREF="#MAKEDEVREF"
>MAKEDEV</A
></P
></LI
><LI
><P
><A
HREF="#MANREF"
>man</A
></P
></LI
><LI
><P
><A
HREF="#MAPFILEREF"
>mapfile</A
></P
></LI
><LI
><P
><A
HREF="#MCOOKIEREF"
>mcookie</A
></P
></LI
><LI
><P
><A
HREF="#MD5SUMREF"
>md5sum</A
></P
></LI
><LI
><P
><A
HREF="#MERGEREF"
>merge</A
></P
></LI
><LI
><P
><A
HREF="#MESGREF"
>mesg</A
></P
></LI
><LI
><P
><A
HREF="#MIMENCODEREF"
>mimencode</A
></P
></LI
><LI
><P
><A
HREF="#MKBOOTDISKREF"
>mkbootdisk</A
></P
></LI
><LI
><P
><A
HREF="#MKDIRREF"
>mkdir</A
></P
></LI
><LI
><P
><A
HREF="#MKDOSFSREF"
>mkdosfs</A
></P
></LI
><LI
><P
><A
HREF="#MKE2FSREF"
>mke2fs</A
></P
></LI
><LI
><P
><A
HREF="#MKFIFOREF"
>mkfifo</A
></P
></LI
><LI
><P
><A
HREF="#MKISOFSREF"
>mkisofs</A
></P
></LI
><LI
><P
><A
HREF="#MKNODREF"
>mknod</A
></P
></LI
><LI
><P
><A
HREF="#MKSWAPREF"
>mkswap</A
></P
></LI
><LI
><P
><A
HREF="#MKTEMPREF"
>mktemp</A
></P
></LI
><LI
><P
><A
HREF="#MMENCODEREF"
>mmencode</A
></P
></LI
><LI
><P
><A
HREF="#MODINFOREF"
>modinfo</A
></P
></LI
><LI
><P
><A
HREF="#MODPROBEREF"
>modprobe</A
></P
></LI
><LI
><P
><A
HREF="#MOREREF"
>more</A
></P
></LI
><LI
><P
><A
HREF="#MOUNTREF"
>mount</A
></P
></LI
><LI
><P
><A
HREF="#MSGFMTREF"
>msgfmt</A
></P
></LI
><LI
><P
><A
HREF="#MVREF"
>mv</A
></P
></LI
><LI
><P
><A
HREF="#NCREF"
>nc</A
></P
></LI
><LI
><P
><A
HREF="#NETCONFIGREF"
>netconfig</A
></P
></LI
><LI
><P
><A
HREF="#NETSTATREF"
>netstat</A
></P
></LI
><LI
><P
><A
HREF="#NEWGRPREF"
>newgrp</A
></P
></LI
><LI
><P
><A
HREF="#NICEREF"
>nice</A
></P
></LI
><LI
><P
><A
HREF="#NLREF"
>nl</A
></P
></LI
><LI
><P
><A
HREF="#NMREF"
>nm</A
></P
></LI
><LI
><P
><A
HREF="#NMAPREF"
>nmap</A
></P
></LI
><LI
><P
><A
HREF="#NOHUPREF"
>nohup</A
></P
></LI
><LI
><P
><A
HREF="#NSLOOKUPREF"
>nslookup</A
></P
></LI
><LI
><P
><A
HREF="#OBJDUMPREF"
>objdump</A
></P
></LI
><LI
><P
><A
HREF="#ODREF"
>od</A
></P
></LI
><LI
><P
><A
HREF="#OPENSSLREF"
>openssl</A
></P
></LI
><LI
><P
><A
HREF="#PASSWDREF"
>passwd</A
></P
></LI
><LI
><P
><A
HREF="#PASTEREF"
>paste</A
></P
></LI
><LI
><P
><A
HREF="#PATCHREF"
>patch</A
>
(Related topic: <A
HREF="#DIFFREF"
>diff</A
>)</P
></LI
><LI
><P
><A
HREF="#PATHCHKREF"
>pathchk</A
></P
></LI
><LI
><P
><A
HREF="#PAXREF"
>pax</A
></P
></LI
><LI
><P
><A
HREF="#PGREPREF"
>pgrep</A
></P
></LI
><LI
><P
><A
HREF="#PIDOFREF"
>pidof</A
></P
></LI
><LI
><P
><A
HREF="#PINGREF"
>ping</A
></P
></LI
><LI
><P
><A
HREF="#PKILLREF"
>pkill</A
></P
></LI
><LI
><P
><A
HREF="#DIRSD"
>popd</A
></P
></LI
><LI
><P
><A
HREF="#PRREF"
>pr</A
></P
></LI
><LI
><P
><A
HREF="#PRINTENVREF"
>printenv</A
></P
></LI
><LI
><P
><A
HREF="#PRINTFREF"
>printf</A
></P
></LI
><LI
><P
><A
HREF="#PROCINFOREF"
>procinfo</A
></P
></LI
><LI
><P
><A
HREF="#PPSSREF"
>ps</A
></P
></LI
><LI
><P
><A
HREF="#PSTREEREF"
>pstree</A
></P
></LI
><LI
><P
><A
HREF="#PTXREF"
>ptx</A
></P
></LI
><LI
><P
><A
HREF="#DIRSD"
>pushd</A
></P
></LI
><LI
><P
><A
HREF="#PWD2REF"
>pwd</A
>
(Related topic: the <A
HREF="#PWDREF"
>$PWD</A
>
variable)</P
></LI
><LI
><P
><A
HREF="#QUOTAREF"
>quota</A
></P
></LI
><LI
><P
><A
HREF="#RCPREF"
>rcp</A
></P
></LI
><LI
><P
><A
HREF="#RDEVREF"
>rdev</A
></P
></LI
><LI
><P
><A
HREF="#RDISTREF"
>rdist</A
></P
></LI
><LI
><P
><A
HREF="#READREF"
>read</A
></P
></LI
><LI
><P
><A
HREF="#READELFREF"
>readelf</A
></P
></LI
><LI
><P
><A
HREF="#READLINKREF"
>readlink</A
></P
></LI
><LI
><P
><A
HREF="#READONLYREF"
>readonly</A
></P
></LI
><LI
><P
><A
HREF="#REBOOTREF"
>reboot</A
></P
></LI
><LI
><P
><A
HREF="#RECODEREF"
>recode</A
></P
></LI
><LI
><P
><A
HREF="#NICE2REF"
>renice</A
></P
></LI
><LI
><P
><A
HREF="#RESETREF"
>reset</A
></P
></LI
><LI
><P
><A
HREF="#RESIZEREF"
>resize</A
></P
></LI
><LI
><P
><A
HREF="#RESTOREREF"
>restore</A
></P
></LI
><LI
><P
><A
HREF="#REVREF"
>rev</A
></P
></LI
><LI
><P
><A
HREF="#RLOGINREF"
>rlogin</A
></P
></LI
><LI
><P
><A
HREF="#RMREF"
>rm</A
></P
></LI
><LI
><P
><A
HREF="#RMDIRREF"
>rmdir</A
></P
></LI
><LI
><P
><A
HREF="#RMMODREF"
>rmmod</A
></P
></LI
><LI
><P
><A
HREF="#ROUTEREF"
>route</A
></P
></LI
><LI
><P
><A
HREF="#RPMREF"
>rpm</A
></P
></LI
><LI
><P
><A
HREF="#RPM2CPIOREF"
>rpm2cpio</A
></P
></LI
><LI
><P
><A
HREF="#RSHREF"
>rsh</A
></P
></LI
><LI
><P
><A
HREF="#RSYNCREF"
>rsync</A
></P
></LI
><LI
><P
><A
HREF="#RUNLEVELREF"
>runlevel</A
></P
></LI
><LI
><P
><A
HREF="#RUNPARTSREF"
>run-parts</A
></P
></LI
><LI
><P
><A
HREF="#RXREF"
>rx</A
></P
></LI
><LI
><P
><A
HREF="#RZREF"
>rz</A
></P
></LI
><LI
><P
><A
HREF="#SARREF"
>sar</A
></P
></LI
><LI
><P
><A
HREF="#SCPREF"
>scp</A
></P
></LI
><LI
><P
><A
HREF="#SCRIPTREF"
>script</A
></P
></LI
><LI
><P
><A
HREF="#SDIFFREF"
>sdiff</A
></P
></LI
><LI
><P
><A
HREF="#SEDREF"
>sed</A
></P
></LI
><LI
><P
><A
HREF="#SEQREF"
>seq</A
></P
></LI
><LI
><P
><A
HREF="#SERVICEREF"
>service</A
></P
></LI
><LI
><P
><A
HREF="#SETREF"
>set</A
></P
></LI
><LI
><P
><A
HREF="#SETFACLREF"
>setfacl</A
></P
></LI
><LI
><P
><A
HREF="#SETQUOTAREF"
>setquota</A
></P
></LI
><LI
><P
><A
HREF="#SETSERIALREF"
>setserial</A
></P
></LI
><LI
><P
><A
HREF="#SETTERMREF"
>setterm</A
></P
></LI
><LI
><P
><A
HREF="#SHA1SUMREF"
>sha1sum</A
></P
></LI
><LI
><P
><A
HREF="#SHARREF"
>shar</A
></P
></LI
><LI
><P
><A
HREF="#SHOPTREF"
>shopt</A
></P
></LI
><LI
><P
><A
HREF="#SHREDREF"
>shred</A
></P
></LI
><LI
><P
><A
HREF="#SHUTDOWNREF"
>shutdown</A
></P
></LI
><LI
><P
><A
HREF="#SIZEREF"
>size</A
></P
></LI
><LI
><P
><A
HREF="#NICE2REF"
>skill</A
></P
></LI
><LI
><P
><A
HREF="#SLEEPREF"
>sleep</A
></P
></LI
><LI
><P
><A
HREF="#SLOCATEREF"
>slocate</A
></P
></LI
><LI
><P
><A
HREF="#NICE2REF"
>snice</A
></P
></LI
><LI
><P
><A
HREF="#SORTREF"
>sort</A
></P
></LI
><LI
><P
><A
HREF="#SOURCEREF"
>source</A
></P
></LI
><LI
><P
><A
HREF="#SOXREF"
>sox</A
></P
></LI
><LI
><P
><A
HREF="#SPLITREF"
>split</A
></P
></LI
><LI
><P
><A
HREF="#SQREF"
>sq</A
></P
></LI
><LI
><P
><A
HREF="#SSHREF"
>ssh</A
></P
></LI
><LI
><P
><A
HREF="#STATREF"
>stat</A
></P
></LI
><LI
><P
><A
HREF="#STRACEREF"
>strace</A
></P
></LI
><LI
><P
><A
HREF="#STRINGSREF"
>strings</A
></P
></LI
><LI
><P
><A
HREF="#STRIPREF"
>strip</A
></P
></LI
><LI
><P
><A
HREF="#STTYREF"
>stty</A
></P
></LI
><LI
><P
><A
HREF="#SUREF"
>su</A
></P
></LI
><LI
><P
><A
HREF="#SUDOREF"
>sudo</A
></P
></LI
><LI
><P
><A
HREF="#SUMREF"
>sum</A
></P
></LI
><LI
><P
><A
HREF="#SUSPENDREF"
>suspend</A
></P
></LI
><LI
><P
><A
HREF="#SWAPONREF"
>swapoff</A
></P
></LI
><LI
><P
><A
HREF="#SWAPONREF"
>swapon</A
></P
></LI
><LI
><P
><A
HREF="#RXREF"
>sx</A
></P
></LI
><LI
><P
><A
HREF="#SYNCREF"
>sync</A
></P
></LI
><LI
><P
><A
HREF="#RZREF"
>sz</A
></P
></LI
><LI
><P
><A
HREF="#CATREF"
>tac</A
></P
></LI
><LI
><P
><A
HREF="#TAILREF"
>tail</A
></P
></LI
><LI
><P
><A
HREF="#TARREF"
>tar</A
></P
></LI
><LI
><P
><A
HREF="#TBLREF"
>tbl</A
></P
></LI
><LI
><P
><A
HREF="#TCPDUMPREF"
>tcpdump</A
></P
></LI
><LI
><P
><A
HREF="#TEEREF"
>tee</A
></P
></LI
><LI
><P
><A
HREF="#TELINITREF"
>telinit</A
></P
></LI
><LI
><P
><A
HREF="#TELNETREF"
>telnet</A
></P
></LI
><LI
><P
><A
HREF="#TEXREF"
>Tex</A
></P
></LI
><LI
><P
><A
HREF="#TEXEXECREF"
>texexec</A
></P
></LI
><LI
><P
><A
HREF="#TIMREF"
>time</A
></P
></LI
><LI
><P
><A
HREF="#TIMESREF"
>times</A
></P
></LI
><LI
><P
><A
HREF="#TMPWATCHREF"
>tmpwatch</A
></P
></LI
><LI
><P
><A
HREF="#TOPREF"
>top</A
></P
></LI
><LI
><P
><A
HREF="#TOUCHREF"
>touch</A
></P
></LI
><LI
><P
><A
HREF="#TPUTREF"
>tput</A
></P
></LI
><LI
><P
><A
HREF="#TRREF"
>tr</A
></P
></LI
><LI
><P
><A
HREF="#TRACEROUTEREF"
>traceroute</A
></P
></LI
><LI
><P
><A
HREF="#TRUEREF"
>true</A
></P
></LI
><LI
><P
><A
HREF="#TSETREF"
>tset</A
></P
></LI
><LI
><P
><A
HREF="#TSORTREF"
>tsort</A
></P
></LI
><LI
><P
><A
HREF="#TTYREF"
>tty</A
></P
></LI
><LI
><P
><A
HREF="#TUNE2FSREF"
>tune2fs</A
></P
></LI
><LI
><P
><A
HREF="#TYPEREF"
>type</A
></P
></LI
><LI
><P
><A
HREF="#DECLAREREF"
>typeset</A
></P
></LI
><LI
><P
><A
HREF="#ULIMITREF"
>ulimit</A
></P
></LI
><LI
><P
><A
HREF="#UMASKREF"
>umask</A
></P
></LI
><LI
><P
><A
HREF="#UMOUNTREF"
>umount</A
></P
></LI
><LI
><P
><A
HREF="#UNAMEREF"
>uname</A
></P
></LI
><LI
><P
><A
HREF="#UNARCREF"
>unarc</A
></P
></LI
><LI
><P
><A
HREF="#UNARCREF"
>unarj</A
></P
></LI
><LI
><P
><A
HREF="#UNCOMPRESSREF"
>uncompress</A
></P
></LI
><LI
><P
><A
HREF="#EXPANDREF"
>unexpand</A
></P
></LI
><LI
><P
><A
HREF="#UNIQREF"
>uniq</A
></P
></LI
><LI
><P
><A
HREF="#UNITSREF"
>units</A
></P
></LI
><LI
><P
><A
HREF="#LZMAREF"
>unlzma</A
></P
></LI
><LI
><P
><A
HREF="#UNARCREF"
>unrar</A
></P
></LI
><LI
><P
><A
HREF="#UNSETREF"
>unset</A
></P
></LI
><LI
><P
><A
HREF="#SQREF"
>unsq</A
></P
></LI
><LI
><P
><A
HREF="#ZIPREF"
>unzip</A
></P
></LI
><LI
><P
><A
HREF="#UPTIMEREF"
>uptime</A
></P
></LI
><LI
><P
><A
HREF="#LSUSBREF"
>usbmodules</A
></P
></LI
><LI
><P
><A
HREF="#USERADDREF"
>useradd</A
></P
></LI
><LI
><P
><A
HREF="#USERADDREF"
>userdel</A
></P
></LI
><LI
><P
><A
HREF="#USERMODREF"
>usermod</A
></P
></LI
><LI
><P
><A
HREF="#USERSREF"
>users</A
></P
></LI
><LI
><P
><A
HREF="#USLEEPREF"
>usleep</A
></P
></LI
><LI
><P
><A
HREF="#UUCPREF"
>uucp</A
></P
></LI
><LI
><P
><A
HREF="#UUDECODEREF"
>uudecode</A
></P
></LI
><LI
><P
><A
HREF="#UUENCODEREF"
>uuencode</A
></P
></LI
><LI
><P
><A
HREF="#UUXREF"
>uux</A
></P
></LI
><LI
><P
><A
HREF="#VACATIONREF"
>vacation</A
></P
></LI
><LI
><P
><A
HREF="#VDIRREF"
>vdir</A
></P
></LI
><LI
><P
><A
HREF="#VMSTATREF"
>vmstat</A
></P
></LI
><LI
><P
><A
HREF="#VRFYREF"
>vrfy</A
></P
></LI
><LI
><P
><A
HREF="#WREF"
>w</A
></P
></LI
><LI
><P
><A
HREF="#WAITREF"
>wait</A
></P
></LI
><LI
><P
><A
HREF="#WALLREF"
>wall</A
></P
></LI
><LI
><P
><A
HREF="#WATCHREF"
>watch</A
></P
></LI
><LI
><P
><A
HREF="#WCREF"
>wc</A
></P
></LI
><LI
><P
><A
HREF="#WGETREF"
>wget</A
></P
></LI
><LI
><P
><A
HREF="#WHATISREF"
>whatis</A
></P
></LI
><LI
><P
><A
HREF="#WHEREISREF"
>whereis</A
></P
></LI
><LI
><P
><A
HREF="#WHICHREF"
>which</A
></P
></LI
><LI
><P
><A
HREF="#WHOREF"
>who</A
></P
></LI
><LI
><P
><A
HREF="#WHOAMIREF"
>whoami</A
></P
></LI
><LI
><P
><A
HREF="#WHOISREF"
>whois</A
></P
></LI
><LI
><P
><A
HREF="#WRITEREF"
>write</A
></P
></LI
><LI
><P
><A
HREF="#XARGSREF"
>xargs</A
></P
></LI
><LI
><P
><A
HREF="#XRANDRREF"
>xrandr</A
></P
></LI
><LI
><P
><A
HREF="#XZREF"
>xz</A
></P
></LI
><LI
><P
><A
HREF="#YACCREF"
>yacc</A
></P
></LI
><LI
><P
><A
HREF="#YESREF"
>yes</A
></P
></LI
><LI
><P
><A
HREF="#ZCATREF"
>zcat</A
></P
></LI
><LI
><P
><A
HREF="#ZDIFFREF"
>zdiff</A
></P
></LI
><LI
><P
><A
HREF="#ZDUMPREF"
>zdump</A
></P
></LI
><LI
><P
><A
HREF="#ZEGREPREF"
>zegrep</A
></P
></LI
><LI
><P
><A
HREF="#ZEGREPREF"
>zfgrep</A
></P
></LI
><LI
><P
><A
HREF="#ZEGREPREF"
>zgrep</A
></P
></LI
><LI
><P
><A
HREF="#ZIPREF"
>zip</A
></P
></LI
></UL
></DIV
><DIV
CLASS="TOC"
><DL
><DT
><B
>Table of Contents</B
></DT
><DT
>15. <A
HREF="#INTERNAL"
>Internal Commands and Builtins</A
></DT
><DD
><DL
><DT
>15.1. <A
HREF="#AEN9644"
>Job Control Commands</A
></DT
></DL
></DD
><DT
>16. <A
HREF="#EXTERNAL"
>External Filters, Programs and Commands</A
></DT
><DD
><DL
><DT
>16.1. <A
HREF="#BASIC"
>Basic Commands</A
></DT
><DT
>16.2. <A
HREF="#MOREADV"
>Complex Commands</A
></DT
><DT
>16.3. <A
HREF="#TIMEDATE"
>Time / Date Commands</A
></DT
><DT
>16.4. <A
HREF="#TEXTPROC"
>Text Processing Commands</A
></DT
><DT
>16.5. <A
HREF="#FILEARCHIV"
>File and Archiving Commands</A
></DT
><DT
>16.6. <A
HREF="#COMMUNICATIONS"
>Communications Commands</A
></DT
><DT
>16.7. <A
HREF="#TERMINALCCMDS"
>Terminal Control Commands</A
></DT
><DT
>16.8. <A
HREF="#MATHC"
>Math Commands</A
></DT
><DT
>16.9. <A
HREF="#EXTMISC"
>Miscellaneous Commands</A
></DT
></DL
></DD
><DT
>17. <A
HREF="#SYSTEM"
>System and Administrative Commands</A
></DT
><DD
><DL
><DT
>17.1. <A
HREF="#SYSSCRIPTS"
>Analyzing a System Script</A
></DT
></DL
></DD
></DL
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="INTERNAL"
></A
>Chapter 15. Internal Commands and Builtins</H1
><P
><A
NAME="BUILTINREF"
></A
>A <I
CLASS="FIRSTTERM"
>builtin</I
>
is a <B
CLASS="COMMAND"
>command</B
> contained within the Bash tool
set, literally <I
CLASS="FIRSTTERM"
>built in</I
>. This is either
for performance reasons -- builtins execute faster than external
commands, which usually require <I
CLASS="FIRSTTERM"
>forking off</I
>
<A
NAME="AEN8607"
HREF="#FTN.AEN8607"
><SPAN
CLASS="footnote"
>[58]</SPAN
></A
>
a separate process -- or because a particular builtin needs
direct access to the shell internals.</P
><P
><A
NAME="FORKREF"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN8611"
></A
><P
></P
><P
><A
NAME="PARENTREF"
></A
>When a command or
the shell itself initiates (or
<I
CLASS="FIRSTTERM"
>spawns</I
>) a new
subprocess to carry out a task, this is called
<I
CLASS="FIRSTTERM"
>forking</I
>. This new process
is the <I
CLASS="FIRSTTERM"
>child</I
>, and the process
that <I
CLASS="FIRSTTERM"
>forked</I
> it off is the
<I
CLASS="FIRSTTERM"
>parent</I
>. While the <I
CLASS="FIRSTTERM"
>child
process</I
> is doing its work, the
<I
CLASS="FIRSTTERM"
>parent process</I
> is still
executing.</P
><P
>Note that while a <I
CLASS="FIRSTTERM"
>parent
process</I
> gets the <I
CLASS="FIRSTTERM"
>process
ID</I
> of the <I
CLASS="FIRSTTERM"
>child
process</I
>, and can thus pass arguments to it,
<EM
>the reverse is not true</EM
>. <A
HREF="#PARCHILDPROBREF"
>This can create problems
that are subtle and hard to track down.</A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="SPAWNSCR"
></A
><P
><B
>Example 15-1. A script that spawns multiple instances of itself</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# spawn.sh
PIDS=$(pidof sh $0) # Process IDs of the various instances of this script.
P_array=( $PIDS ) # Put them in an array (why?).
echo $PIDS # Show process IDs of parent and child processes.
let "instances = ${#P_array[*]} - 1" # Count elements, less 1.
# Why subtract 1?
echo "$instances instance(s) of this script running."
echo "[Hit Ctl-C to exit.]"; echo
sleep 1 # Wait.
sh $0 # Play it again, Sam.
exit 0 # Not necessary; script will never get to here.
# Why not?
# After exiting with a Ctl-C,
#+ do all the spawned instances of the script die?
# If so, why?
# Note:
# ----
# Be careful not to run this script too long.
# It will eventually eat up too many system resources.
# Is having a script spawn multiple instances of itself
#+ an advisable scripting technique.
# Why or why not?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="BLTINFRK"
></A
></P
><P
>Generally, a Bash <I
CLASS="FIRSTTERM"
>builtin</I
>
does not fork a subprocess when it executes within
a script. An external system command or filter in
a script usually <EM
>will</EM
> fork a
subprocess.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
>A builtin may be a synonym to a system command of the same
name, but Bash reimplements it internally. For example,
the Bash <B
CLASS="COMMAND"
>echo</B
> command is not the same as
<TT
CLASS="FILENAME"
>/bin/echo</TT
>, although their behavior is
almost identical.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo "This line uses the \"echo\" builtin."
/bin/echo "This line uses the /bin/echo system command."</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="KEYWORDREF"
></A
>A <I
CLASS="FIRSTTERM"
>keyword</I
>
is a <I
CLASS="FIRSTTERM"
>reserved</I
> word, token or
operator. Keywords have a special meaning to the shell,
and indeed are the building blocks of the shell's
syntax. As examples, <I
CLASS="FIRSTTERM"
>for</I
>,
<I
CLASS="FIRSTTERM"
>while</I
>, <I
CLASS="FIRSTTERM"
>do</I
>,
and <I
CLASS="FIRSTTERM"
>!</I
> are keywords. Similar to a <A
HREF="#BUILTINREF"
>builtin</A
>, a keyword is hard-coded into
Bash, but unlike a <I
CLASS="FIRSTTERM"
>builtin</I
>, a keyword is
not in itself a command, but <EM
>a subunit of a command
construct</EM
>.
<A
NAME="AEN8650"
HREF="#FTN.AEN8650"
><SPAN
CLASS="footnote"
>[59]</SPAN
></A
>
</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="INTIO1"
></A
>I/O</B
></P
><DL
><DT
><A
NAME="ECHOREF"
></A
><B
CLASS="COMMAND"
>echo</B
></DT
><DD
><P
>prints (to <TT
CLASS="FILENAME"
>stdout</TT
>) an expression
or variable (see <A
HREF="#EX9"
>Example 4-1</A
>).
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo Hello
echo $a</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>An <B
CLASS="COMMAND"
>echo</B
> requires the
<TT
CLASS="OPTION"
>-e</TT
> option to print escaped characters. See
<A
HREF="#ESCAPED"
>Example 5-2</A
>.</P
><P
>Normally, each <B
CLASS="COMMAND"
>echo</B
> command prints
a terminal newline, but the <TT
CLASS="OPTION"
>-n</TT
> option
suppresses this.</P
><P
><A
NAME="ECHOGREPREF"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>An <B
CLASS="COMMAND"
>echo</B
> can be used to feed a
sequence of commands down a pipe.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if echo "$VAR" | grep -q txt # if [[ $VAR = *txt* ]]
then
echo "$VAR contains the substring sequence \"txt\""
fi</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ECHOCS"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>An <B
CLASS="COMMAND"
>echo</B
>, in combination with
<A
HREF="#COMMANDSUBREF"
>command substitution</A
>
can set a variable.</P
><P
><TT
CLASS="USERINPUT"
><B
>a=`echo
"HELLO" | tr A-Z a-z`</B
></TT
></P
><P
>See also <A
HREF="#LOWERCASE"
>Example 16-22</A
>, <A
HREF="#EX57"
>Example 16-3</A
>, <A
HREF="#MONTHLYPMT"
>Example 16-47</A
>, and <A
HREF="#BASE"
>Example 16-48</A
>.</P
></TD
></TR
></TABLE
></DIV
><P
>Be aware that <B
CLASS="COMMAND"
>echo `command`</B
>
deletes any linefeeds that the output
of <TT
CLASS="REPLACEABLE"
><I
>command</I
></TT
>
generates.</P
><P
>The <A
HREF="#IFSREF"
>$IFS</A
> (internal field
separator) variable normally contains
<SPAN
CLASS="TOKEN"
>\n</SPAN
> (linefeed) as one of its set of
<A
HREF="#WHITESPACEREF"
>whitespace</A
>
characters. Bash therefore splits the output of
<TT
CLASS="REPLACEABLE"
><I
>command</I
></TT
> at linefeeds
into arguments to <B
CLASS="COMMAND"
>echo</B
>. Then
<B
CLASS="COMMAND"
>echo</B
> outputs these arguments,
separated by spaces.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l /usr/share/apps/kjezz/sounds</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-rw-r--r-- 1 root root 1407 Nov 7 2000 reflect.au
-rw-r--r-- 1 root root 362 Nov 7 2000 seconds.au</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo `ls -l /usr/share/apps/kjezz/sounds`</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 40 -rw-r--r-- 1 root root 716 Nov 7 2000 reflect.au -rw-r--r-- 1 root root ...</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> So, how can we embed a linefeed within an
<A
HREF="#ECHOREF"
>echoed</A
> character string?
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Embedding a linefeed?
echo "Why doesn't this string \n split on two lines?"
# Doesn't split.
# Let's try something else.
echo
echo $"A line of text containing
a linefeed."
# Prints as two distinct lines (embedded linefeed).
# But, is the "$" variable prefix really necessary?
echo
echo "This string splits
on two lines."
# No, the "$" is not needed.
echo
echo "---------------"
echo
echo -n $"Another line of text containing
a linefeed."
# Prints as two distinct lines (embedded linefeed).
# Even the -n option fails to suppress the linefeed here.
echo
echo
echo "---------------"
echo
echo
# However, the following doesn't work as expected.
# Why not? Hint: Assignment to a variable.
string1=$"Yet another line of text containing
a linefeed (maybe)."
echo $string1
# Yet another line of text containing a linefeed (maybe).
# ^
# Linefeed becomes a space.
# Thanks, Steve Parker, for pointing this out.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="BINECHO"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This command is a shell builtin, and not the same as
<TT
CLASS="FILENAME"
>/bin/echo</TT
>, although its behavior is
similar.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>type -a echo</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>echo is a shell builtin
echo is /bin/echo</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="PRINTFREF"
></A
><B
CLASS="COMMAND"
>printf</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>printf</B
>, formatted print, command is an
enhanced <B
CLASS="COMMAND"
>echo</B
>. It is a limited variant
of the <I
CLASS="FIRSTTERM"
>C</I
> language
<TT
CLASS="FUNCTION"
>printf()</TT
> library function, and its
syntax is somewhat different.</P
><P
><B
CLASS="COMMAND"
>printf</B
> <TT
CLASS="REPLACEABLE"
><I
>format-string</I
></TT
>... <TT
CLASS="REPLACEABLE"
><I
>parameter</I
></TT
>... </P
><P
>This is the Bash <I
CLASS="FIRSTTERM"
>builtin</I
> version
of the <TT
CLASS="FILENAME"
>/bin/printf</TT
> or
<TT
CLASS="FILENAME"
>/usr/bin/printf</TT
> command. See the
<B
CLASS="COMMAND"
>printf</B
> <A
HREF="#MANREF"
>manpage</A
> (of the system command)
for in-depth coverage.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Older versions of Bash may not support
<B
CLASS="COMMAND"
>printf</B
>.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX47"
></A
><P
><B
>Example 15-2. <I
CLASS="FIRSTTERM"
>printf</I
> in action</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# printf demo
declare -r PI=3.14159265358979 # Read-only variable, i.e., a constant.
declare -r DecimalConstant=31373
Message1="Greetings,"
Message2="Earthling."
echo
printf "Pi to 2 decimal places = %1.2f" $PI
echo
printf "Pi to 9 decimal places = %1.9f" $PI # It even rounds off correctly.
printf "\n" # Prints a line feed,
# Equivalent to 'echo' . . .
printf "Constant = \t%d\n" $DecimalConstant # Inserts tab (\t).
printf "%s %s \n" $Message1 $Message2
echo
# ==========================================#
# Simulation of C function, sprintf().
# Loading a variable with a formatted string.
echo
Pi12=$(printf "%1.12f" $PI)
echo "Pi to 12 decimal places = $Pi12" # Roundoff error!
Msg=`printf "%s %s \n" $Message1 $Message2`
echo $Msg; echo $Msg
# As it happens, the 'sprintf' function can now be accessed
#+ as a loadable module to Bash,
#+ but this is not portable.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Formatting error messages is a useful application of
<B
CLASS="COMMAND"
>printf</B
></P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>E_BADDIR=85
var=nonexistent_directory
error()
{
printf "$@" &#62;&#38;2
# Formats positional params passed, and sends them to stderr.
echo
exit $E_BADDIR
}
cd $var || error $"Can't cd to %s." "$var"
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>See also <A
HREF="#PROGRESSBAR"
>Example 36-17</A
>.</P
></DD
><DT
><A
NAME="READREF"
></A
><B
CLASS="COMMAND"
>read</B
></DT
><DD
><P
><SPAN
CLASS="QUOTE"
>"Reads"</SPAN
> the value
of a variable from <TT
CLASS="FILENAME"
>stdin</TT
>, that
is, interactively fetches input from the keyboard. The
<TT
CLASS="OPTION"
>-a</TT
> option lets <B
CLASS="COMMAND"
>read</B
>
get array variables (see <A
HREF="#EX67"
>Example 27-6</A
>).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX36"
></A
><P
><B
>Example 15-3. Variable assignment, using <I
CLASS="FIRSTTERM"
>read</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# "Reading" variables.
echo -n "Enter the value of variable 'var1': "
# The -n option to echo suppresses newline.
read var1
# Note no '$' in front of var1, since it is being set.
echo "var1 = $var1"
echo
# A single 'read' statement can set multiple variables.
echo -n "Enter the values of variables 'var2' and 'var3' "
echo =n "(separated by a space or tab): "
read var2 var3
echo "var2 = $var2 var3 = $var3"
# If you input only one value,
#+ the other variable(s) will remain unset (null).
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>A <B
CLASS="COMMAND"
>read</B
> without an associated variable
assigns its input to the dedicated variable <A
HREF="#REPLYREF"
>$REPLY</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="READNOVAR"
></A
><P
><B
>Example 15-4. What happens when <I
CLASS="FIRSTTERM"
>read</I
> has no
variable</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# read-novar.sh
echo
# -------------------------- #
echo -n "Enter a value: "
read var
echo "\"var\" = "$var""
# Everything as expected here.
# -------------------------- #
echo
# ------------------------------------------------------------------- #
echo -n "Enter another value: "
read # No variable supplied for 'read', therefore...
#+ Input to 'read' assigned to default variable, $REPLY.
var="$REPLY"
echo "\"var\" = "$var""
# This is equivalent to the first code block.
# ------------------------------------------------------------------- #
echo
echo "========================="
echo
# This example is similar to the "reply.sh" script.
# However, this one shows that $REPLY is available
#+ even after a 'read' to a variable in the conventional way.
# ================================================================= #
# In some instances, you might wish to discard the first value read.
# In such cases, simply ignore the $REPLY variable.
{ # Code block.
read # Line 1, to be discarded.
read line2 # Line 2, saved in variable.
} &#60;$0
echo "Line 2 of this script is:"
echo "$line2" # # read-novar.sh
echo # #!/bin/bash line discarded.
# See also the soundcard-on.sh script.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Normally, inputting a <TT
CLASS="USERINPUT"
><B
>\</B
></TT
>
suppresses a newline during input to
a <B
CLASS="COMMAND"
>read</B
>. The <TT
CLASS="OPTION"
>-r</TT
>
option causes an inputted <TT
CLASS="USERINPUT"
><B
>\</B
></TT
> to be
interpreted literally.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="READR"
></A
><P
><B
>Example 15-5. Multi-line input to <I
CLASS="FIRSTTERM"
>read</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo
echo "Enter a string terminated by a \\, then press &#60;ENTER&#62;."
echo "Then, enter a second string (no \\ this time), and again press &#60;ENTER&#62;."
read var1 # The "\" suppresses the newline, when reading $var1.
# first line \
# second line
echo "var1 = $var1"
# var1 = first line second line
# For each line terminated by a "\"
#+ you get a prompt on the next line to continue feeding characters into var1.
echo; echo
echo "Enter another string terminated by a \\ , then press &#60;ENTER&#62;."
read -r var2 # The -r option causes the "\" to be read literally.
# first line \
echo "var2 = $var2"
# var2 = first line \
# Data entry terminates with the first &#60;ENTER&#62;.
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="READOPTIONS"
></A
></P
><P
>The <B
CLASS="COMMAND"
>read</B
> command has some interesting
options that permit echoing a prompt and even reading keystrokes
without hitting <B
CLASS="KEYCAP"
>ENTER</B
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Read a keypress without hitting ENTER.
read -s -n1 -p "Hit a key " keypress
echo; echo "Keypress was "\"$keypress\""."
# -s option means do not echo input.
# -n N option means accept only N characters of input.
# -p option means echo the following prompt before reading input.
# Using these options is tricky, since they need to be in the correct order.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="READARROW"
></A
></P
><P
>The <TT
CLASS="OPTION"
>-n</TT
> option to <B
CLASS="COMMAND"
>read</B
>
also allows detection of the <B
CLASS="KEYCAP"
>arrow keys</B
>
and certain of the other unusual keys.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ARROWDETECT"
></A
><P
><B
>Example 15-6. Detecting the arrow keys</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# arrow-detect.sh: Detects the arrow keys, and a few more.
# Thank you, Sandro Magi, for showing me how.
# --------------------------------------------
# Character codes generated by the keypresses.
arrowup='\[A'
arrowdown='\[B'
arrowrt='\[C'
arrowleft='\[D'
insert='\[2'
delete='\[3'
# --------------------------------------------
SUCCESS=0
OTHER=65
echo -n "Press a key... "
# May need to also press ENTER if a key not listed above pressed.
read -n3 key # Read 3 characters.
echo -n "$key" | grep "$arrowup" #Check if character code detected.
if [ "$?" -eq $SUCCESS ]
then
echo "Up-arrow key pressed."
exit $SUCCESS
fi
echo -n "$key" | grep "$arrowdown"
if [ "$?" -eq $SUCCESS ]
then
echo "Down-arrow key pressed."
exit $SUCCESS
fi
echo -n "$key" | grep "$arrowrt"
if [ "$?" -eq $SUCCESS ]
then
echo "Right-arrow key pressed."
exit $SUCCESS
fi
echo -n "$key" | grep "$arrowleft"
if [ "$?" -eq $SUCCESS ]
then
echo "Left-arrow key pressed."
exit $SUCCESS
fi
echo -n "$key" | grep "$insert"
if [ "$?" -eq $SUCCESS ]
then
echo "\"Insert\" key pressed."
exit $SUCCESS
fi
echo -n "$key" | grep "$delete"
if [ "$?" -eq $SUCCESS ]
then
echo "\"Delete\" key pressed."
exit $SUCCESS
fi
echo " Some other key pressed."
exit $OTHER
# ========================================= #
# Mark Alexander came up with a simplified
#+ version of the above script (Thank you!).
# It eliminates the need for grep.
#!/bin/bash
uparrow=$'\x1b[A'
downarrow=$'\x1b[B'
leftarrow=$'\x1b[D'
rightarrow=$'\x1b[C'
read -s -n3 -p "Hit an arrow key: " x
case "$x" in
$uparrow)
echo "You pressed up-arrow"
;;
$downarrow)
echo "You pressed down-arrow"
;;
$leftarrow)
echo "You pressed left-arrow"
;;
$rightarrow)
echo "You pressed right-arrow"
;;
esac
exit $?
# ========================================= #
# Antonio Macchi has a simpler alternative.
#!/bin/bash
while true
do
read -sn1 a
test "$a" == `echo -en "\e"` || continue
read -sn1 a
test "$a" == "[" || continue
read -sn1 a
case "$a" in
A) echo "up";;
B) echo "down";;
C) echo "right";;
D) echo "left";;
esac
done
# ========================================= #
# Exercise:
# --------
# 1) Add detection of the "Home," "End," "PgUp," and "PgDn" keys.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <TT
CLASS="OPTION"
>-n</TT
> option to <B
CLASS="COMMAND"
>read</B
>
will not detect the <B
CLASS="KEYCAP"
>ENTER</B
> (newline)
key.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="READTIMED"
></A
></P
><P
>The <TT
CLASS="OPTION"
>-t</TT
> option to <B
CLASS="COMMAND"
>read</B
>
permits timed input (see <A
HREF="#TOUT"
>Example 9-4</A
> and <A
HREF="#QKY"
>Example A-41</A
>).</P
><P
><A
NAME="READFD"
></A
>The <TT
CLASS="OPTION"
>-u</TT
> option
takes the <A
HREF="#FDREF"
>file descriptor</A
>
of the target file.</P
><P
><A
NAME="READREDIR0"
></A
></P
><P
>The <B
CLASS="COMMAND"
>read</B
> command may also
<SPAN
CLASS="QUOTE"
>"read"</SPAN
> its variable value from a file
<A
HREF="#IOREDIRREF"
>redirected</A
> to
<TT
CLASS="FILENAME"
>stdin</TT
>. If the file contains
more than one line, only the first line is assigned
to the variable. If <B
CLASS="COMMAND"
>read</B
>
has more than one parameter, then each of
these variables gets assigned a successive <A
HREF="#WHITESPACEREF"
>whitespace-delineated</A
>
string. Caution!</P
><DIV
CLASS="EXAMPLE"
><A
NAME="READREDIR"
></A
><P
><B
>Example 15-7. Using <I
CLASS="FIRSTTERM"
>read</I
> with
<A
HREF="#IOREDIRREF"
>file redirection</A
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
read var1 &#60;data-file
echo "var1 = $var1"
# var1 set to the entire first line of the input file "data-file"
read var2 var3 &#60;data-file
echo "var2 = $var2 var3 = $var3"
# Note non-intuitive behavior of "read" here.
# 1) Rewinds back to the beginning of input file.
# 2) Each variable is now set to a corresponding string,
# separated by whitespace, rather than to an entire line of text.
# 3) The final variable gets the remainder of the line.
# 4) If there are more variables to be set than whitespace-terminated strings
# on the first line of the file, then the excess variables remain empty.
echo "------------------------------------------------"
# How to resolve the above problem with a loop:
while read line
do
echo "$line"
done &#60;data-file
# Thanks, Heiner Steven for pointing this out.
echo "------------------------------------------------"
# Use $IFS (Internal Field Separator variable) to split a line of input to
# "read", if you do not want the default to be whitespace.
echo "List of all users:"
OIFS=$IFS; IFS=: # /etc/passwd uses ":" for field separator.
while read name passwd uid gid fullname ignore
do
echo "$name ($fullname)"
done &#60;/etc/passwd # I/O redirection.
IFS=$OIFS # Restore original $IFS.
# This code snippet also by Heiner Steven.
# Setting the $IFS variable within the loop itself
#+ eliminates the need for storing the original $IFS
#+ in a temporary variable.
# Thanks, Dim Segebart, for pointing this out.
echo "------------------------------------------------"
echo "List of all users:"
while IFS=: read name passwd uid gid fullname ignore
do
echo "$name ($fullname)"
done &#60;/etc/passwd # I/O redirection.
echo
echo "\$IFS still $IFS"
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="PIPEREADREF0"
></A
></P
><P
><A
HREF="#PIPEREF"
>Piping</A
> output
to a <I
CLASS="FIRSTTERM"
>read</I
>, using <A
HREF="#ECHOREF"
>echo</A
> to set variables <A
HREF="#BADREAD0"
>will fail</A
>.</P
><P
><A
NAME="READPIPEREF"
></A
>Yet, piping the output of <A
HREF="#CATREF"
>cat</A
> <EM
>seems</EM
> to
work.</P
><P
><A
NAME="WHILEREADREF"
></A
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat file1 file2 |
while read line
do
echo $line
done</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>However, as Bj<42>n Eriksson shows:</P
><DIV
CLASS="EXAMPLE"
><A
NAME="READPIPE"
></A
><P
><B
>Example 15-8. Problems reading from a pipe</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/sh
# readpipe.sh
# This example contributed by Bjon Eriksson.
### shopt -s lastpipe
last="(null)"
cat $0 |
while read line
do
echo "{$line}"
last=$line
done
echo
echo "++++++++++++++++++++++"
printf "\nAll done, last: $last\n" # The output of this line
#+ changes if you uncomment line 5.
# (Bash, version -ge 4.2 required.)
exit 0 # End of code.
# (Partial) output of script follows.
# The 'echo' supplies extra brackets.
#############################################
./readpipe.sh
{#!/bin/sh}
{last="(null)"}
{cat $0 |}
{while read line}
{do}
{echo "{$line}"}
{last=$line}
{done}
{printf "nAll done, last: $lastn"}
All done, last: (null)
The variable (last) is set within the loop/subshell
but its value does not persist outside the loop.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The <I
CLASS="FIRSTTERM"
>gendiff</I
> script, usually
found in <TT
CLASS="FILENAME"
>/usr/bin</TT
> on
many Linux distros, pipes the output of <A
HREF="#FINDREF"
>find</A
> to a <I
CLASS="FIRSTTERM"
>while
read</I
> construct.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>find $1 \( -name "*$2" -o -name ".*$2" \) -print |
while read f; do
. . .</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>It is possible to <I
CLASS="FIRSTTERM"
>paste</I
> text into
the input field of a <I
CLASS="FIRSTTERM"
>read</I
> (but
<EM
>not</EM
> multiple lines!). See <A
HREF="#PADSW"
>Example A-38</A
>.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="INTFILESYSTEM1"
></A
>Filesystem</B
></P
><DL
><DT
><A
NAME="CDREF"
></A
><B
CLASS="COMMAND"
>cd</B
></DT
><DD
><P
>The familiar <B
CLASS="COMMAND"
>cd</B
> change directory
command finds use in scripts where execution of a command
requires being in a specified directory.</P
><P
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>(cd /source/directory &#38;&#38; tar cf - . ) | (cd /dest/directory &#38;&#38; tar xpvf -)</PRE
></FONT
></TD
></TR
></TABLE
>
[from the <A
HREF="#COXEX"
>previously cited</A
>
example by Alan Cox]</P
><P
>The <TT
CLASS="OPTION"
>-P</TT
> (physical) option to
<B
CLASS="COMMAND"
>cd</B
> causes it to ignore symbolic
links.</P
><P
><B
CLASS="COMMAND"
>cd -</B
> changes to <A
HREF="#OLDPWD"
>$OLDPWD</A
>, the previous working
directory.</P
><P
><A
NAME="DOUBLESLASHREF"
></A
></P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>cd</B
> command does not function
as expected when presented with two forward slashes.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cd //</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>pwd</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>//</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
The output should, of course, be <TT
CLASS="COMPUTEROUTPUT"
>/</TT
>.
This is a problem both from the command-line and in a script.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="PWD2REF"
></A
><B
CLASS="COMMAND"
>pwd</B
></DT
><DD
><P
>Print Working Directory. This gives the user's
(or script's) current directory (see <A
HREF="#EX37"
>Example 15-9</A
>). The effect is identical to
reading the value of the builtin variable <A
HREF="#PWDREF"
>$PWD</A
>.</P
></DD
><DT
><A
NAME="DIRSD"
></A
><B
CLASS="COMMAND"
>pushd</B
>, <B
CLASS="COMMAND"
>popd</B
>, <B
CLASS="COMMAND"
>dirs</B
></DT
><DD
><P
>This command set is a mechanism for bookmarking
working directories, a means of moving back and forth
through directories in an orderly manner. A pushdown <A
HREF="#STACKDEFREF"
>stack</A
> is used to keep track
of directory names. Options allow various manipulations
of the directory stack.</P
><P
><A
NAME="PUSHDREF"
></A
><TT
CLASS="USERINPUT"
><B
>pushd
dir-name</B
></TT
> pushes the path
<TT
CLASS="REPLACEABLE"
><I
>dir-name</I
></TT
> onto the directory
stack (to the <I
CLASS="FIRSTTERM"
>top</I
> of the stack)
and simultaneously changes the current working directory
to <TT
CLASS="REPLACEABLE"
><I
>dir-name</I
></TT
></P
><P
><A
NAME="POPDREF"
></A
><B
CLASS="COMMAND"
>popd</B
> removes
(pops) the top directory path name off the directory stack
and simultaneously changes the current working directory
to the directory now at the <I
CLASS="FIRSTTERM"
>top</I
> of
the stack.</P
><P
><A
NAME="DIRSREF"
></A
><B
CLASS="COMMAND"
>dirs</B
> lists
the contents of the directory stack (compare this
with the <A
HREF="#DIRSTACKREF"
>$DIRSTACK</A
>
variable). A successful <B
CLASS="COMMAND"
>pushd</B
>
or <B
CLASS="COMMAND"
>popd</B
> will automatically invoke
<B
CLASS="COMMAND"
>dirs</B
>.</P
><P
>Scripts that require various changes to the current
working directory without hard-coding the directory name
changes can make good use of these commands. Note that
the implicit <TT
CLASS="VARNAME"
>$DIRSTACK</TT
> array variable,
accessible from within a script, holds the contents of
the directory stack.
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX37"
></A
><P
><B
>Example 15-9. Changing the current working directory</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
dir1=/usr/local
dir2=/var/spool
pushd $dir1
# Will do an automatic 'dirs' (list directory stack to stdout).
echo "Now in directory `pwd`." # Uses back-quoted 'pwd'.
# Now, do some stuff in directory 'dir1'.
pushd $dir2
echo "Now in directory `pwd`."
# Now, do some stuff in directory 'dir2'.
echo "The top entry in the DIRSTACK array is $DIRSTACK."
popd
echo "Now back in directory `pwd`."
# Now, do some more stuff in directory 'dir1'.
popd
echo "Now back in original working directory `pwd`."
exit 0
# What happens if you don't 'popd' -- then exit the script?
# Which directory do you end up in? Why?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="INTVAR1"
></A
>Variables</B
></P
><DL
><DT
><A
NAME="LETREF"
></A
><B
CLASS="COMMAND"
>let</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>let</B
> command carries out
<I
CLASS="FIRSTTERM"
>arithmetic</I
> operations on variables.
<A
NAME="AEN9009"
HREF="#FTN.AEN9009"
><SPAN
CLASS="footnote"
>[60]</SPAN
></A
>
In many cases, it functions as a less complex version
of <A
HREF="#EXPRREF"
>expr</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX46"
></A
><P
><B
>Example 15-10. Letting <I
CLASS="FIRSTTERM"
>let</I
> do arithmetic.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo
let a=11 # Same as 'a=11'
let a=a+5 # Equivalent to let "a = a + 5"
# (Double quotes and spaces make it more readable.)
echo "11 + 5 = $a" # 16
let "a &#60;&#60;= 3" # Equivalent to let "a = a &#60;&#60; 3"
echo "\"\$a\" (=16) left-shifted 3 places = $a"
# 128
let "a /= 4" # Equivalent to let "a = a / 4"
echo "128 / 4 = $a" # 32
let "a -= 5" # Equivalent to let "a = a - 5"
echo "32 - 5 = $a" # 27
let "a *= 10" # Equivalent to let "a = a * 10"
echo "27 * 10 = $a" # 270
let "a %= 8" # Equivalent to let "a = a % 8"
echo "270 modulo 8 = $a (270 / 8 = 33, remainder $a)"
# 6
# Does "let" permit C-style operators?
# Yes, just as the (( ... )) double-parentheses construct does.
let a++ # C-style (post) increment.
echo "6++ = $a" # 6++ = 7
let a-- # C-style decrement.
echo "7-- = $a" # 7-- = 6
# Of course, ++a, etc., also allowed . . .
echo
# Trinary operator.
# Note that $a is 6, see above.
let "t = a&#60;7?7:11" # True
echo $t # 7
let a++
let "t = a&#60;7?7:11" # False
echo $t # 11
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="EXITVALANOMALY02"
></A
></P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <I
CLASS="FIRSTTERM"
>let</I
> command can,
in certain contexts, return a surprising <A
HREF="#EXITSTATUSREF"
>exit status</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Evgeniy Ivanov points out:
var=0
echo $? # 0
# As expected.
let var++
echo $? # 1
# The command was successful, so why isn't $?=0 ???
# Anomaly!
let var++
echo $? # 0
# As expected.
# Likewise . . .
let var=0
echo $? # 1
# The command was successful, so why isn't $?=0 ???
# However, as Jeff Gorak points out,
#+ this is part of the design spec for 'let' . . .
# "If the last ARG evaluates to 0, let returns 1;
# let returns 0 otherwise." ['help let']</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="EVALREF"
></A
><B
CLASS="COMMAND"
>eval</B
></DT
><DD
><P
><TT
CLASS="USERINPUT"
><B
>eval arg1 [arg2] ... [argN]</B
></TT
></P
><P
>Combines the arguments in an expression or list of
expressions and <TT
CLASS="REPLACEABLE"
><I
>evaluates</I
></TT
> them.
Any variables within the expression are expanded. The
net result is to <B
CLASS="COMMAND"
>convert a string into a
command</B
>.</P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>eval</B
> command can be used for
code generation from the command-line or within a script.
</P
></TD
></TR
></TABLE
></DIV
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>command_string="ps ax"</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>process="ps ax"</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>eval "$command_string" | grep "$process"</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>26973 pts/3 R+ 0:00 grep --color ps ax
26974 pts/3 R+ 0:00 ps ax</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="EVALFORCED"
></A
></P
><P
>Each invocation of <I
CLASS="FIRSTTERM"
>eval</I
> forces
a re-<EM
>evaluation</EM
> of its arguments.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>a='$b'
b='$c'
c=d
echo $a # $b
# First level.
eval echo $a # $c
# Second level.
eval eval echo $a # d
# Third level.
# Thank you, E. Choroba.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="EVALEFF"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX43"
></A
><P
><B
>Example 15-11. Showing the effect of <I
CLASS="FIRSTTERM"
>eval</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Exercising "eval" ...
y=`eval ls -l` # Similar to y=`ls -l`
echo $y #+ but linefeeds removed because "echoed" variable is unquoted.
echo
echo "$y" # Linefeeds preserved when variable is quoted.
echo; echo
y=`eval df` # Similar to y=`df`
echo $y #+ but linefeeds removed.
# When LF's not preserved, it may make it easier to parse output,
#+ using utilities such as "awk".
echo
echo "==========================================================="
echo
eval "`seq 3 | sed -e 's/.*/echo var&#38;=ABCDEFGHIJ/'`"
# var1=ABCDEFGHIJ
# var2=ABCDEFGHIJ
# var3=ABCDEFGHIJ
echo
echo "==========================================================="
echo
# Now, showing how to do something useful with "eval" . . .
# (Thank you, E. Choroba!)
version=3.4 # Can we split the version into major and minor
#+ part in one command?
echo "version = $version"
eval major=${version/./;minor=} # Replaces '.' in version by ';minor='
# The substitution yields '3; minor=4'
#+ so eval does minor=4, major=3
echo Major: $major, minor: $minor # Major: 3, minor: 4</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ARRCHOICE0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="ARRCHOICE"
></A
><P
><B
>Example 15-12. Using <I
CLASS="FIRSTTERM"
>eval</I
> to select
among variables</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# arr-choice.sh
# Passing arguments to a function to select
#+ one particular variable out of a group.
arr0=( 10 11 12 13 14 15 )
arr1=( 20 21 22 23 24 25 )
arr2=( 30 31 32 33 34 35 )
# 0 1 2 3 4 5 Element number (zero-indexed)
choose_array ()
{
eval array_member=\${arr${array_number}[element_number]}
# ^ ^^^^^^^^^^^^
# Using eval to construct the name of a variable,
#+ in this particular case, an array name.
echo "Element $element_number of array $array_number is $array_member"
} # Function can be rewritten to take parameters.
array_number=0 # First array.
element_number=3
choose_array # 13
array_number=2 # Third array.
element_number=4
choose_array # 34
array_number=3 # Null array (arr3 not allocated).
element_number=4
choose_array # (null)
# Thank you, Antonio Macchi, for pointing this out.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ECHOPARAMS"
></A
><P
><B
>Example 15-13. <I
CLASS="FIRSTTERM"
>Echoing</I
> the
<I
CLASS="FIRSTTERM"
>command-line parameters</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# echo-params.sh
# Call this script with a few command-line parameters.
# For example:
# sh echo-params.sh first second third fourth fifth
params=$# # Number of command-line parameters.
param=1 # Start at first command-line param.
while [ "$param" -le "$params" ]
do
echo -n "Command-line parameter "
echo -n \$$param # Gives only the *name* of variable.
# ^^^ # $1, $2, $3, etc.
# Why?
# \$ escapes the first "$"
#+ so it echoes literally,
#+ and $param dereferences "$param" . . .
#+ . . . as expected.
echo -n " = "
eval echo \$$param # Gives the *value* of variable.
# ^^^^ ^^^ # The "eval" forces the *evaluation*
#+ of \$$
#+ as an indirect variable reference.
(( param ++ )) # On to the next.
done
exit $?
# =================================================
$ sh echo-params.sh first second third fourth fifth
Command-line parameter $1 = first
Command-line parameter $2 = second
Command-line parameter $3 = third
Command-line parameter $4 = fourth
Command-line parameter $5 = fifth</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX44"
></A
><P
><B
>Example 15-14. Forcing a log-off</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Killing ppp to force a log-off.
# For dialup connection, of course.
# Script should be run as root user.
SERPORT=ttyS3
# Depending on the hardware and even the kernel version,
#+ the modem port on your machine may be different --
#+ /dev/ttyS1 or /dev/ttyS2.
killppp="eval kill -9 `ps ax | awk '/ppp/ { print $1 }'`"
# -------- process ID of ppp -------
$killppp # This variable is now a command.
# The following operations must be done as root user.
chmod 666 /dev/$SERPORT # Restore r+w permissions, or else what?
# Since doing a SIGKILL on ppp changed the permissions on the serial port,
#+ we restore permissions to previous state.
rm /var/lock/LCK..$SERPORT # Remove the serial port lock file. Why?
exit $?
# Exercises:
# ---------
# 1) Have script check whether root user is invoking it.
# 2) Do a check on whether the process to be killed
#+ is actually running before attempting to kill it.
# 3) Write an alternate version of this script based on 'fuser':
#+ if [ fuser -s /dev/modem ]; then . . .</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ROT14"
></A
><P
><B
>Example 15-15. A version of <I
CLASS="FIRSTTERM"
>rot13</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# A version of "rot13" using 'eval'.
# Compare to "rot13.sh" example.
setvar_rot_13() # "rot13" scrambling
{
local varname=$1 varvalue=$2
eval $varname='$(echo "$varvalue" | tr a-z n-za-m)'
}
setvar_rot_13 var "foobar" # Run "foobar" through rot13.
echo $var # sbbone
setvar_rot_13 var "$var" # Run "sbbone" through rot13.
# Back to original variable.
echo $var # foobar
# This example by Stephane Chazelas.
# Modified by document author.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Here is another example of using
<I
CLASS="FIRSTTERM"
>eval</I
> to
<EM
>evaluate</EM
> a complex expression,
this one from an earlier version of YongYe's <A
HREF="https://github.com/yongye/shell/blob/master/Tetris_Game.sh"
TARGET="_top"
>Tetris
game script</A
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>eval ${1}+=\"${x} ${y} \"</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
HREF="#SAMORSE"
>Example A-53</A
> uses
<I
CLASS="FIRSTTERM"
>eval</I
> to convert <A
HREF="#ARRAYREF"
>array</A
> elements into a command
list.</P
><P
>The <I
CLASS="FIRSTTERM"
>eval</I
> command occurs
in the older version of <A
HREF="#IVRREF"
>indirect
referencing</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>eval var=\$$var</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <I
CLASS="FIRSTTERM"
>eval</I
> command can
be used to <A
HREF="#BRACEEXPREF3"
>parameterize
<I
CLASS="FIRSTTERM"
>brace expansion</I
></A
>.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="EVALRISK"
></A
></P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>eval</B
> command can be
risky, and normally should be avoided when there
exists a reasonable alternative. An <TT
CLASS="USERINPUT"
><B
>eval
$COMMANDS</B
></TT
> executes the contents of
<TT
CLASS="REPLACEABLE"
><I
>COMMANDS</I
></TT
>, which may
contain such unpleasant surprises as <B
CLASS="COMMAND"
>rm -rf
*</B
>. Running an <B
CLASS="COMMAND"
>eval</B
> on
unfamiliar code written by persons unknown is living
dangerously.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SETREF"
></A
><B
CLASS="COMMAND"
>set</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>set</B
> command changes
the value of internal script variables/options. One use for
this is to toggle <A
HREF="#OPTIONSREF"
>option
flags</A
> which help determine the behavior of the
script. Another application for it is to reset the <A
HREF="#POSPARAMREF"
>positional parameters</A
> that
a script sees as the result of a command (<TT
CLASS="USERINPUT"
><B
>set
`command`</B
></TT
>). The script can then parse the
<A
HREF="#FIELDREF"
>fields</A
> of the command
output.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX34"
></A
><P
><B
>Example 15-16. Using <I
CLASS="FIRSTTERM"
>set</I
> with positional
parameters</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ex34.sh
# Script "set-test"
# Invoke this script with three command-line parameters,
# for example, "sh ex34.sh one two three".
echo
echo "Positional parameters before set \`uname -a\` :"
echo "Command-line argument #1 = $1"
echo "Command-line argument #2 = $2"
echo "Command-line argument #3 = $3"
set `uname -a` # Sets the positional parameters to the output
# of the command `uname -a`
echo
echo +++++
echo $_ # +++++
# Flags set in script.
echo $- # hB
# Anomalous behavior?
echo
echo "Positional parameters after set \`uname -a\` :"
# $1, $2, $3, etc. reinitialized to result of `uname -a`
echo "Field #1 of 'uname -a' = $1"
echo "Field #2 of 'uname -a' = $2"
echo "Field #3 of 'uname -a' = $3"
echo \#\#\#
echo $_ # ###
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>More fun with positional parameters.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="REVPOSPARAMS"
></A
><P
><B
>Example 15-17. Reversing the positional parameters</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# revposparams.sh: Reverse positional parameters.
# Script by Dan Jacobson, with stylistic revisions by document author.
set a\ b c d\ e;
# ^ ^ Spaces escaped
# ^ ^ Spaces not escaped
OIFS=$IFS; IFS=:;
# ^ Saving old IFS and setting new one.
echo
until [ $# -eq 0 ]
do # Step through positional parameters.
echo "### k0 = "$k"" # Before
k=$1:$k; # Append each pos param to loop variable.
# ^
echo "### k = "$k"" # After
echo
shift;
done
set $k # Set new positional parameters.
echo -
echo $# # Count of positional parameters.
echo -
echo
for i # Omitting the "in list" sets the variable -- i --
#+ to the positional parameters.
do
echo $i # Display new positional parameters.
done
IFS=$OIFS # Restore IFS.
# Question:
# Is it necessary to set an new IFS, internal field separator,
#+ in order for this script to work properly?
# What happens if you don't? Try it.
# And, why use the new IFS -- a colon -- in line 17,
#+ to append to the loop variable?
# What is the purpose of this?
exit 0
$ ./revposparams.sh
### k0 =
### k = a b
### k0 = a b
### k = c a b
### k0 = c a b
### k = d e c a b
-
3
-
d e
c
a b</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Invoking <B
CLASS="COMMAND"
>set</B
> without any options or
arguments simply lists all the <A
HREF="#ENVREF"
>environmental</A
> and other variables
that have been initialized.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>set</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>AUTHORCOPY=/home/bozo/posts
BASH=/bin/bash
BASH_VERSION=$'2.05.8(1)-release'
...
XAUTHORITY=/home/bozo/.Xauthority
_=/etc/bashrc
variable22=abc
variable23=xzy</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Using <B
CLASS="COMMAND"
>set</B
> with the <TT
CLASS="OPTION"
>--</TT
>
option explicitly assigns the contents of a variable to
the positional parameters. If no variable follows the
<TT
CLASS="OPTION"
>--</TT
> it <I
CLASS="FIRSTTERM"
>unsets</I
>
the positional parameters.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SETPOS"
></A
><P
><B
>Example 15-18. Reassigning the positional parameters</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
variable="one two three four five"
set -- $variable
# Sets positional parameters to the contents of "$variable".
first_param=$1
second_param=$2
shift; shift # Shift past first two positional params.
# shift 2 also works.
remaining_params="$*"
echo
echo "first parameter = $first_param" # one
echo "second parameter = $second_param" # two
echo "remaining parameters = $remaining_params" # three four five
echo; echo
# Again.
set -- $variable
first_param=$1
second_param=$2
echo "first parameter = $first_param" # one
echo "second parameter = $second_param" # two
# ======================================================
set --
# Unsets positional parameters if no variable specified.
first_param=$1
second_param=$2
echo "first parameter = $first_param" # (null value)
echo "second parameter = $second_param" # (null value)
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#EX22A"
>Example 11-2</A
> and <A
HREF="#EX33A"
>Example 16-56</A
>.</P
></DD
><DT
><A
NAME="UNSETREF"
></A
><B
CLASS="COMMAND"
>unset</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>unset</B
> command deletes a
shell variable, effectively setting it to
<I
CLASS="FIRSTTERM"
>null</I
>. Note that this command does
not affect positional parameters.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>unset PATH</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $PATH</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>&#13;</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="UNS"
></A
><P
><B
>Example 15-19. <SPAN
CLASS="QUOTE"
>"Unsetting"</SPAN
> a variable</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# unset.sh: Unsetting a variable.
variable=hello # Initialized.
echo "variable = $variable"
unset variable # Unset.
# In this particular context,
#+ same effect as: variable=
echo "(unset) variable = $variable" # $variable is null.
if [ -z "$variable" ] # Try a string-length test.
then
echo "\$variable has zero length."
fi
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>In most contexts, an <I
CLASS="FIRSTTERM"
>undeclared</I
>
variable and one that has been <I
CLASS="FIRSTTERM"
>unset</I
>
are equivalent. However, the <A
HREF="#UNDDR"
> ${parameter:-default}</A
> parameter substitution
construct can distinguish between the two.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="EXPORTREF"
></A
><B
CLASS="COMMAND"
>export</B
></DT
><DD
><P
><A
NAME="EXPORTREF2"
></A
></P
><P
>The <B
CLASS="COMMAND"
>export</B
>
<A
NAME="AEN9199"
HREF="#FTN.AEN9199"
><SPAN
CLASS="footnote"
>[61]</SPAN
></A
>
command makes available variables to all child processes
of the running script or shell. One important use
of the <B
CLASS="COMMAND"
>export</B
> command is in <A
HREF="#FILESREF1"
>startup files</A
>, to initialize
and make accessible <A
HREF="#ENVREF"
>environmental
variables</A
> to subsequent user processes.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Unfortunately, <A
HREF="#PARCHILDPROBREF"
> there is no way to export variables back to the parent
process</A
>, to the process that called or invoked the
script or shell.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="EXPORTAWK"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="COLTOTALER3"
></A
><P
><B
>Example 15-20. Using <I
CLASS="FIRSTTERM"
>export</I
> to pass a variable to an
embedded <I
CLASS="FIRSTTERM"
>awk</I
> script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Yet another version of the "column totaler" script (col-totaler.sh)
#+ that adds up a specified column (of numbers) in the target file.
# This uses the environment to pass a script variable to 'awk' . . .
#+ and places the awk script in a variable.
ARGS=2
E_WRONGARGS=85
if [ $# -ne "$ARGS" ] # Check for proper number of command-line args.
then
echo "Usage: `basename $0` filename column-number"
exit $E_WRONGARGS
fi
filename=$1
column_number=$2
#===== Same as original script, up to this point =====#
export column_number
# Export column number to environment, so it's available for retrieval.
# -----------------------------------------------
awkscript='{ total += $ENVIRON["column_number"] }
END { print total }'
# Yes, a variable can hold an awk script.
# -----------------------------------------------
# Now, run the awk script.
awk "$awkscript" "$filename"
# Thanks, Stephane Chazelas.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>It is possible to initialize and export
variables in the same operation, as in <B
CLASS="COMMAND"
>export
var1=xxx</B
>.</P
><P
>However, as Greg Keraunen points out, in certain
situations this may have a different effect than
setting a variable, then exporting it.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>export var=(a b); echo ${var[0]}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>(a b)</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>var=(a b); export var; echo ${var[0]}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>a</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A variable to be exported may require special
treatment. See <A
HREF="#BASHPROF"
>Example M-2</A
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="DECLARE2REF"
></A
><B
CLASS="COMMAND"
>declare</B
>, <B
CLASS="COMMAND"
>typeset</B
></DT
><DD
><P
>The <A
HREF="#DECLAREREF"
>declare</A
> and
<A
HREF="#DECLAREREF"
>typeset</A
> commands specify
and/or restrict properties of variables.</P
></DD
><DT
><A
NAME="READONLYREF"
></A
><B
CLASS="COMMAND"
>readonly</B
></DT
><DD
><P
>Same as <A
HREF="#DECLAREREF"
>declare -r</A
>,
sets a variable as read-only, or, in effect, as a
constant. Attempts to change the variable fail with
an error message. This is the shell analog of the
<I
CLASS="FIRSTTERM"
>C</I
> language <B
CLASS="COMMAND"
>const</B
>
type qualifier.</P
></DD
><DT
><A
NAME="GETOPTSX"
></A
><B
CLASS="COMMAND"
>getopts</B
></DT
><DD
><P
>This powerful tool parses command-line arguments passed
to the script. This is the Bash analog of the <A
HREF="#GETOPTY"
>getopt</A
> external command and the
<I
CLASS="FIRSTTERM"
>getopt</I
> library function familiar to
<I
CLASS="FIRSTTERM"
>C</I
> programmers. It permits passing
and concatenating multiple options
<A
NAME="AEN9289"
HREF="#FTN.AEN9289"
><SPAN
CLASS="footnote"
>[62]</SPAN
></A
>
and associated arguments to a script (for
example <TT
CLASS="USERINPUT"
><B
>scriptname -abc -e
/usr/local</B
></TT
>).</P
><P
><A
NAME="GETOPTSOPT"
></A
></P
><P
>The <B
CLASS="COMMAND"
>getopts</B
> construct uses two implicit
variables. <TT
CLASS="VARNAME"
>$OPTIND</TT
> is the argument
pointer (<I
CLASS="WORDASWORD"
>OPTion INDex</I
>)
and <TT
CLASS="VARNAME"
>$OPTARG</TT
> (<I
CLASS="WORDASWORD"
>OPTion
ARGument</I
>) the (optional) argument attached
to an option. A colon following the option name in the
declaration tags that option as having an associated
argument.</P
><P
>A <B
CLASS="COMMAND"
>getopts</B
> construct usually comes
packaged in a <A
HREF="#WHILELOOPREF"
>while
loop</A
>, which processes the options and
arguments one at a time, then increments the implicit
<TT
CLASS="VARNAME"
>$OPTIND</TT
> variable to point to the
next.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
> <P
></P
><OL
TYPE="1"
><LI
><P
>The arguments passed from the command-line to
the script must be preceded by a
dash (<TT
CLASS="OPTION"
>-</TT
>). It is the
prefixed <TT
CLASS="OPTION"
>-</TT
> that lets
<B
CLASS="COMMAND"
>getopts</B
> recognize command-line
arguments as <I
CLASS="FIRSTTERM"
>options</I
>.
In fact, <B
CLASS="COMMAND"
>getopts</B
> will not process
arguments without the prefixed <TT
CLASS="OPTION"
>-</TT
>,
and will terminate option processing at the first
argument encountered lacking them.</P
></LI
><LI
><P
>The <B
CLASS="COMMAND"
>getopts</B
> template
differs slightly from the standard <A
HREF="#WHILELOOPREF"
>while loop</A
>, in that
it lacks condition brackets.</P
></LI
><LI
><P
>The <B
CLASS="COMMAND"
>getopts</B
> construct is a highly
functional replacement for the traditional
<A
HREF="#GETOPTY"
>getopt</A
> external
command.</P
></LI
></OL
>
</P
></TD
></TR
></TABLE
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>while getopts ":abcde:fg" Option
# Initial declaration.
# a, b, c, d, e, f, and g are the options (flags) expected.
# The : after option 'e' shows it will have an argument passed with it.
do
case $Option in
a ) # Do something with variable 'a'.
b ) # Do something with variable 'b'.
...
e) # Do something with 'e', and also with $OPTARG,
# which is the associated argument passed with option 'e'.
...
g ) # Do something with variable 'g'.
esac
done
shift $(($OPTIND - 1))
# Move argument pointer to next.
# All this is not nearly as complicated as it looks &#60;grin&#62;.</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX33"
></A
><P
><B
>Example 15-21. Using <I
CLASS="FIRSTTERM"
>getopts</I
> to read the
options/arguments passed to a script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ex33.sh: Exercising getopts and OPTIND
# Script modified 10/09/03 at the suggestion of Bill Gradwohl.
# Here we observe how 'getopts' processes command-line arguments to script.
# The arguments are parsed as "options" (flags) and associated arguments.
# Try invoking this script with:
# 'scriptname -mn'
# 'scriptname -oq qOption' (qOption can be some arbitrary string.)
# 'scriptname -qXXX -r'
#
# 'scriptname -qr'
#+ - Unexpected result, takes "r" as the argument to option "q"
# 'scriptname -q -r'
#+ - Unexpected result, same as above
# 'scriptname -mnop -mnop' - Unexpected result
# (OPTIND is unreliable at stating where an option came from.)
#
# If an option expects an argument ("flag:"), then it will grab
#+ whatever is next on the command-line.
NO_ARGS=0
E_OPTERROR=85
if [ $# -eq "$NO_ARGS" ] # Script invoked with no command-line args?
then
echo "Usage: `basename $0` options (-mnopqrs)"
exit $E_OPTERROR # Exit and explain usage.
# Usage: scriptname -options
# Note: dash (-) necessary
fi
while getopts ":mnopq:rs" Option
do
case $Option in
m ) echo "Scenario #1: option -m- [OPTIND=${OPTIND}]";;
n | o ) echo "Scenario #2: option -$Option- [OPTIND=${OPTIND}]";;
p ) echo "Scenario #3: option -p- [OPTIND=${OPTIND}]";;
q ) echo "Scenario #4: option -q-\
with argument \"$OPTARG\" [OPTIND=${OPTIND}]";;
# Note that option 'q' must have an associated argument,
#+ otherwise it falls through to the default.
r | s ) echo "Scenario #5: option -$Option-";;
* ) echo "Unimplemented option chosen.";; # Default.
esac
done
shift $(($OPTIND - 1))
# Decrements the argument pointer so it points to next argument.
# $1 now references the first non-option item supplied on the command-line
#+ if one exists.
exit $?
# As Bill Gradwohl states,
# "The getopts mechanism allows one to specify: scriptname -mnop -mnop
#+ but there is no reliable way to differentiate what came
#+ from where by using OPTIND."
# There are, however, workarounds.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="INTSCRBEH1"
></A
>Script Behavior</B
></P
><DL
><DT
><A
NAME="SOURCEREF"
></A
><B
CLASS="COMMAND"
>source</B
>, <SPAN
CLASS="TOKEN"
>.</SPAN
> (<A
HREF="#DOTREF"
>dot</A
> command)</DT
><DD
><P
>This command, when invoked from the command-line,
executes a script. Within a script, a
<TT
CLASS="USERINPUT"
><B
>source file-name</B
></TT
>
loads the file <TT
CLASS="FILENAME"
>file-name</TT
>.
<I
CLASS="FIRSTTERM"
>Sourcing</I
> a file (dot-command)
<I
CLASS="FIRSTTERM"
>imports</I
>
code into the script, appending to the script (same effect
as the <TT
CLASS="USERINPUT"
><B
>#include</B
></TT
> directive in a
<I
CLASS="FIRSTTERM"
>C</I
> program). The net result is the
same as if the <SPAN
CLASS="QUOTE"
>"sourced"</SPAN
> lines of code were
physically present in the body of the script. This is useful
in situations when multiple scripts use a common data file
or function library.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX38"
></A
><P
><B
>Example 15-22. <SPAN
CLASS="QUOTE"
>"Including"</SPAN
> a data file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Note that this example must be invoked with bash, i.e., bash ex38.sh
#+ not sh ex38.sh !
. data-file # Load a data file.
# Same effect as "source data-file", but more portable.
# The file "data-file" must be present in current working directory,
#+ since it is referred to by its basename.
# Now, let's reference some data from that file.
echo "variable1 (from data-file) = $variable1"
echo "variable3 (from data-file) = $variable3"
let "sum = $variable2 + $variable4"
echo "Sum of variable2 + variable4 (from data-file) = $sum"
echo "message1 (from data-file) is \"$message1\""
# Escaped quotes
echo "message2 (from data-file) is \"$message2\""
print_message This is the message-print function in the data-file.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
><P
>File <TT
CLASS="FILENAME"
>data-file</TT
> for <A
HREF="#EX38"
>Example 15-22</A
>, above. Must be present in same
directory.</P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># This is a data file loaded by a script.
# Files of this type may contain variables, functions, etc.
# It loads with a 'source' or '.' command from a shell script.
# Let's initialize some variables.
variable1=23
variable2=474
variable3=5
variable4=97
message1="Greetings from *** line $LINENO *** of the data file!"
message2="Enough for now. Goodbye."
print_message ()
{ # Echoes any message passed to it.
if [ -z "$1" ]
then
return 1 # Error, if argument missing.
fi
echo
until [ -z "$1" ]
do # Step through arguments passed to function.
echo -n "$1" # Echo args one at a time, suppressing line feeds.
echo -n " " # Insert spaces between words.
shift # Next one.
done
echo
return 0
}</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>If the <I
CLASS="FIRSTTERM"
>sourced</I
> file is itself
an executable script, then it will run, then return
control to the script that called it. A
<I
CLASS="FIRSTTERM"
>sourced</I
> executable script may use a
<A
HREF="#RETURNREF"
>return</A
> for this
purpose.</P
><P
><A
NAME="SOURCEPARAMS"
></A
></P
><P
> Arguments may be (optionally) passed to the
<I
CLASS="FIRSTTERM"
>sourced</I
> file as <A
HREF="#POSPARAMREF1"
>positional parameters</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>source $filename $arg1 arg2</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>It is even possible for a script to
<I
CLASS="FIRSTTERM"
>source</I
> itself, though this does not
seem to have any practical applications.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SELFSOURCE"
></A
><P
><B
>Example 15-23. A (useless) script that sources itself</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# self-source.sh: a script sourcing itself "recursively."
# From "Stupid Script Tricks," Volume II.
MAXPASSCNT=100 # Maximum number of execution passes.
echo -n "$pass_count "
# At first execution pass, this just echoes two blank spaces,
#+ since $pass_count still uninitialized.
let "pass_count += 1"
# Assumes the uninitialized variable $pass_count
#+ can be incremented the first time around.
# This works with Bash and pdksh, but
#+ it relies on non-portable (and possibly dangerous) behavior.
# Better would be to initialize $pass_count to 0 before incrementing.
while [ "$pass_count" -le $MAXPASSCNT ]
do
. $0 # Script "sources" itself, rather than calling itself.
# ./$0 (which would be true recursion) doesn't work here. Why?
done
# What occurs here is not actually recursion,
#+ since the script effectively "expands" itself, i.e.,
#+ generates a new section of code
#+ with each pass through the 'while' loop',
# with each 'source' in line 20.
#
# Of course, the script interprets each newly 'sourced' "#!" line
#+ as a comment, and not as the start of a new script.
echo
exit 0 # The net effect is counting from 1 to 100.
# Very impressive.
# Exercise:
# --------
# Write a script that uses this trick to actually do something useful.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="EXITREF"
></A
><B
CLASS="COMMAND"
>exit</B
></DT
><DD
><P
>Unconditionally terminates a script.
<A
NAME="AEN9393"
HREF="#FTN.AEN9393"
><SPAN
CLASS="footnote"
>[63]</SPAN
></A
>
The <B
CLASS="COMMAND"
>exit</B
> command may optionally take an
integer argument, which is returned to the shell as
the <A
HREF="#EXITSTATUSREF"
>exit status</A
>
of the script. It is good practice to end all but the
simplest scripts with an <TT
CLASS="USERINPUT"
><B
>exit 0</B
></TT
>,
indicating a successful run.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>If a script terminates with an <B
CLASS="COMMAND"
>exit</B
>
lacking an argument, the exit status of the script is the exit
status of the last command executed in the script, not counting
the <B
CLASS="COMMAND"
>exit</B
>. This is equivalent to an
<B
CLASS="COMMAND"
>exit $?</B
>.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>An <B
CLASS="COMMAND"
>exit</B
> command may also be used to
terminate a <A
HREF="#SUBSHELLSREF"
>subshell</A
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="EXECREF"
></A
><B
CLASS="COMMAND"
>exec</B
></DT
><DD
><P
> This shell builtin replaces the current process with
a specified command. Normally, when the shell encounters
a command, it <A
HREF="#FORKREF"
>forks off</A
> a
child process to actually execute the command. Using the
<B
CLASS="COMMAND"
>exec</B
> builtin, the shell does not fork,
and the command <I
CLASS="FIRSTTERM"
>exec</I
>'ed replaces
the shell. When used in a script, therefore, it forces an
exit from the script when the <B
CLASS="COMMAND"
>exec</B
>'ed
command terminates.
<A
NAME="AEN9425"
HREF="#FTN.AEN9425"
><SPAN
CLASS="footnote"
>[64]</SPAN
></A
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX54"
></A
><P
><B
>Example 15-24. Effects of <I
CLASS="FIRSTTERM"
>exec</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
exec echo "Exiting \"$0\" at line $LINENO." # Exit from script here.
# $LINENO is an internal Bash variable set to the line number it's on.
# ----------------------------------
# The following lines never execute.
echo "This echo fails to echo."
exit 99 # This script will not exit here.
# Check exit value after script terminates
#+ with an 'echo $?'.
# It will *not* be 99.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="SELFEXEC"
></A
><P
><B
>Example 15-25. A script that <I
CLASS="FIRSTTERM"
>exec's</I
> itself</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# self-exec.sh
# Note: Set permissions on this script to 555 or 755,
# then call it with ./self-exec.sh or sh ./self-exec.sh.
echo
echo "This line appears ONCE in the script, yet it keeps echoing."
echo "The PID of this instance of the script is still $$."
# Demonstrates that a subshell is not forked off.
echo "==================== Hit Ctl-C to exit ===================="
sleep 1
exec $0 # Spawns another instance of this same script
#+ that replaces the previous one.
echo "This line will never echo!" # Why not?
exit 99 # Will not exit here!
# Exit code will not be 99!</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>An <B
CLASS="COMMAND"
>exec</B
> also serves to <A
HREF="#USINGEXECREF"
>reassign
file descriptors</A
>. For example, <TT
CLASS="USERINPUT"
><B
>exec
&#60;zzz-file</B
></TT
> replaces <TT
CLASS="FILENAME"
>stdin</TT
>
with the file <TT
CLASS="FILENAME"
>zzz-file</TT
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <TT
CLASS="OPTION"
>-exec</TT
> option to
<A
HREF="#FINDREF"
>find</A
> is
<TT
CLASS="REPLACEABLE"
><I
>not</I
></TT
> the same as the
<B
CLASS="COMMAND"
>exec</B
> shell builtin.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SHOPTREF"
></A
><B
CLASS="COMMAND"
>shopt</B
></DT
><DD
><P
>This command permits changing <I
CLASS="FIRSTTERM"
>shell
options</I
> on the fly (see <A
HREF="#AL"
>Example 25-1</A
>
and <A
HREF="#UNAL"
>Example 25-2</A
>). It often appears in the Bash
<A
HREF="#FILESREF1"
>startup files</A
>, but also has
its uses in scripts. Needs <A
HREF="#BASH2REF"
>version
2</A
> or later of Bash.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>shopt -s cdspell
# Allows minor misspelling of directory names with 'cd'
# Option -s sets, -u unsets.
cd /hpme # Oops! Mistyped '/home'.
pwd # /home
# The shell corrected the misspelling.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><B
CLASS="COMMAND"
>caller</B
><A
NAME="CALLERREF"
></A
></DT
><DD
><P
>Putting a <B
CLASS="COMMAND"
>caller</B
> command
inside a <A
HREF="#FUNCTIONREF"
>function</A
>
echoes to <TT
CLASS="FILENAME"
>stdout</TT
> information about
the <I
CLASS="FIRSTTERM"
>caller</I
> of that function.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
function1 ()
{
# Inside function1 ().
caller 0 # Tell me about it.
}
function1 # Line 9 of script.
# 9 main test.sh
# ^ Line number that the function was called from.
# ^^^^ Invoked from "main" part of script.
# ^^^^^^^ Name of calling script.
caller 0 # Has no effect because it's not inside a function.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>A <B
CLASS="COMMAND"
>caller</B
> command can also return
<I
CLASS="FIRSTTERM"
>caller</I
> information from a script <A
HREF="#SOURCEREF"
>sourced</A
> within another
script. Analogous to a function, this is a <SPAN
CLASS="QUOTE"
>"subroutine
call."</SPAN
></P
><P
>You may find this command useful in debugging.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="INTCOMMAND1"
></A
>Commands</B
></P
><DL
><DT
><A
NAME="TRUEREF"
></A
><B
CLASS="COMMAND"
>true</B
></DT
><DD
><P
>A command that returns a successful
(<SPAN
CLASS="RETURNVALUE"
>zero</SPAN
>) <A
HREF="#EXITSTATUSREF"
>exit status</A
>, but does
nothing else.
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>true</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $?</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>0</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Endless loop
while true # alias for ":"
do
operation-1
operation-2
...
operation-n
# Need a way to break out of loop or script will hang.
done</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="FALSEREF"
></A
><B
CLASS="COMMAND"
>false</B
></DT
><DD
><P
>A command that returns an unsuccessful <A
HREF="#EXITSTATUSREF"
>exit status</A
>,
but does nothing else.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>false</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $?</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>1</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Testing "false"
if false
then
echo "false evaluates \"true\""
else
echo "false evaluates \"false\""
fi
# false evaluates "false"
# Looping while "false" (null loop)
while false
do
# The following code will not execute.
operation-1
operation-2
...
operation-n
# Nothing happens!
done </PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="TYPEREF"
></A
><B
CLASS="COMMAND"
>type [cmd]</B
></DT
><DD
><P
>Similar to the <A
HREF="#WHICHREF"
>which</A
> external command,
<B
CLASS="COMMAND"
>type cmd</B
> identifies
<SPAN
CLASS="QUOTE"
>"cmd."</SPAN
> Unlike <B
CLASS="COMMAND"
>which</B
>,
<B
CLASS="COMMAND"
>type</B
> is a Bash builtin. The useful
<TT
CLASS="OPTION"
>-a</TT
> option to <B
CLASS="COMMAND"
>type</B
>
identifies <TT
CLASS="REPLACEABLE"
><I
>keywords</I
></TT
>
and <TT
CLASS="REPLACEABLE"
><I
>builtins</I
></TT
>, and also locates
system commands with identical names.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>type '['</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>[ is a shell builtin</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>type -a '['</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>[ is a shell builtin
[ is /usr/bin/[</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>type type</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>type is a shell builtin</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <B
CLASS="COMMAND"
>type</B
> command can be useful
for <A
HREF="#DEVNULLREDIRECT"
>testing whether a
certain command exists</A
>.</P
></DD
><DT
><A
NAME="HASHCMDREF"
></A
><B
CLASS="COMMAND"
>hash [cmds]</B
></DT
><DD
><P
>Records the <I
CLASS="FIRSTTERM"
>path</I
>
name of specified commands -- in the shell <I
CLASS="FIRSTTERM"
>hash
table</I
>
<A
NAME="AEN9591"
HREF="#FTN.AEN9591"
><SPAN
CLASS="footnote"
>[65]</SPAN
></A
>
-- so the shell or script will not need to search the
<A
HREF="#PATHREF"
>$PATH</A
> on subsequent calls to those
commands. When <B
CLASS="COMMAND"
>hash</B
> is called with no
arguments, it simply lists the commands that have been hashed.
The <TT
CLASS="OPTION"
>-r</TT
> option resets the hash table.</P
></DD
><DT
><A
NAME="BINDREF"
></A
><B
CLASS="COMMAND"
>bind</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>bind</B
> builtin displays or modifies
<I
CLASS="FIRSTTERM"
>readline</I
>
<A
NAME="AEN9621"
HREF="#FTN.AEN9621"
><SPAN
CLASS="footnote"
>[66]</SPAN
></A
>
key bindings.</P
></DD
><DT
><A
NAME="HELPREF"
></A
><B
CLASS="COMMAND"
>help</B
></DT
><DD
><P
>Gets a short usage summary of a shell builtin. This is
the counterpart to <A
HREF="#WHATISREF"
>whatis</A
>,
but for builtins. The display of <I
CLASS="FIRSTTERM"
>help</I
>
information got a much-needed update in the <A
HREF="#BASH4REF"
>version 4 release</A
> of Bash.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>help exit</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>exit: exit [n]
Exit the shell with a status of N. If N is omitted, the exit status
is that of the last command executed.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
></DL
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="AEN9644"
></A
>15.1. Job Control Commands</H1
><P
>Certain of the following job control commands take a
<I
CLASS="FIRSTTERM"
>job identifier</I
> as an argument. See
the <A
HREF="#JOBIDTABLE"
>table</A
> at end of the
chapter.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><A
NAME="JOBSREF"
></A
><B
CLASS="COMMAND"
>jobs</B
></DT
><DD
><P
>Lists the jobs running in the background, giving
the <I
CLASS="FIRSTTERM"
>job number</I
>.
Not as useful as <A
HREF="#PPSSREF"
>ps</A
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>It is all too easy to confuse
<I
CLASS="FIRSTTERM"
>jobs</I
> and
<I
CLASS="FIRSTTERM"
>processes</I
>. Certain <A
HREF="#BUILTINREF"
>builtins</A
>, such as
<B
CLASS="COMMAND"
>kill</B
>, <B
CLASS="COMMAND"
>disown</B
>, and
<B
CLASS="COMMAND"
>wait</B
> accept either a job number or a
process number as an argument. The <A
HREF="#FGREF"
>fg</A
>,
<A
HREF="#BGREF"
>bg</A
> and <B
CLASS="COMMAND"
>jobs</B
>
commands accept only a job number.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>sleep 100 &#38;</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>[1] 1384</TT
>
<TT
CLASS="PROMPT"
>bash $ </TT
><TT
CLASS="USERINPUT"
><B
>jobs</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>[1]+ Running sleep 100 &#38;</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><SPAN
CLASS="QUOTE"
>"1"</SPAN
> is the job number (jobs are
maintained by the current shell). <SPAN
CLASS="QUOTE"
>"1384"</SPAN
>
is the <A
HREF="#PPIDREF"
>PID</A
> or <I
CLASS="FIRSTTERM"
>process ID
number</I
> (processes are maintained by the system). To kill
this job/process, either a <B
CLASS="COMMAND"
>kill %1</B
>
or a <B
CLASS="COMMAND"
>kill 1384</B
> works.</P
><P
><EM
>Thanks, S.C.</EM
></P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="DISOWNREF"
></A
><B
CLASS="COMMAND"
>disown</B
></DT
><DD
><P
>Remove job(s) from the shell's table of active jobs.</P
></DD
><DT
><A
NAME="FGREF"
></A
><B
CLASS="COMMAND"
>fg</B
>, <A
NAME="BGREF"
></A
><B
CLASS="COMMAND"
>bg</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>fg</B
> command switches a job
running in the background into the foreground. The
<B
CLASS="COMMAND"
>bg</B
> command restarts a suspended job, and
runs it in the background. If no job number is specified,
then the <B
CLASS="COMMAND"
>fg</B
> or <B
CLASS="COMMAND"
>bg</B
>
command acts upon the currently running job.</P
></DD
><DT
><A
NAME="WAITREF"
></A
><B
CLASS="COMMAND"
>wait</B
></DT
><DD
><P
>Suspend script execution until all jobs running in
background have terminated, or until the job number or
process ID specified as an option terminates. Returns the <A
HREF="#EXITSTATUSREF"
>exit status</A
> of waited-for
command.</P
><P
>You may use the <B
CLASS="COMMAND"
>wait</B
> command
to prevent a script from exiting before a background
job finishes executing (this would create a dreaded
<A
HREF="#ZOMBIEREF"
>orphan process</A
>).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX39"
></A
><P
><B
>Example 15-26. Waiting for a process to finish before proceeding</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
ROOT_UID=0 # Only users with $UID 0 have root privileges.
E_NOTROOT=65
E_NOPARAMS=66
if [ "$UID" -ne "$ROOT_UID" ]
then
echo "Must be root to run this script."
# "Run along kid, it's past your bedtime."
exit $E_NOTROOT
fi
if [ -z "$1" ]
then
echo "Usage: `basename $0` find-string"
exit $E_NOPARAMS
fi
echo "Updating 'locate' database..."
echo "This may take a while."
updatedb /usr &#38; # Must be run as root.
wait
# Don't run the rest of the script until 'updatedb' finished.
# You want the the database updated before looking up the file name.
locate $1
# Without the 'wait' command, in the worse case scenario,
#+ the script would exit while 'updatedb' was still running,
#+ leaving it as an orphan process.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Optionally, <B
CLASS="COMMAND"
>wait</B
> can take a <I
CLASS="FIRSTTERM"
>job
identifier</I
> as an argument, for example,
<TT
CLASS="REPLACEABLE"
><I
>wait%1</I
></TT
> or <TT
CLASS="REPLACEABLE"
><I
>wait
$PPID</I
></TT
>.
<A
NAME="AEN9753"
HREF="#FTN.AEN9753"
><SPAN
CLASS="footnote"
>[67]</SPAN
></A
>
See the <A
HREF="#JOBIDTABLE"
>job id table</A
>.</P
><P
><A
NAME="WAITHANG"
></A
></P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Within a script, running a command in the background
with an ampersand (&#38;) may cause the script
to hang until <B
CLASS="KEYCAP"
>ENTER</B
> is hit. This
seems to occur with commands that write to
<TT
CLASS="FILENAME"
>stdout</TT
>. It can be a major annoyance.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# test.sh
ls -l &#38;
echo "Done."</PRE
></FONT
></TD
></TR
></TABLE
>
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>./test.sh</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Done.
[bozo@localhost test-scripts]$ total 1
-rwxr-xr-x 1 bozo bozo 34 Oct 11 15:09 test.sh
_</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><A
NAME="AEN9768"
></A
><BLOCKQUOTE
CLASS="BLOCKQUOTE"
><P
CLASS="LITERALLAYOUT"
>&nbsp;&nbsp;&nbsp;&nbsp;As&nbsp;Walter&nbsp;Brameld&nbsp;IV&nbsp;explains&nbsp;it:<br>
<br>
&nbsp;&nbsp;&nbsp;&nbsp;As&nbsp;far&nbsp;as&nbsp;I&nbsp;can&nbsp;tell,&nbsp;such&nbsp;scripts&nbsp;don't&nbsp;actually&nbsp;hang.&nbsp;It&nbsp;just<br>
&nbsp;&nbsp;&nbsp;&nbsp;seems&nbsp;that&nbsp;they&nbsp;do&nbsp;because&nbsp;the&nbsp;background&nbsp;command&nbsp;writes&nbsp;text&nbsp;to<br>
&nbsp;&nbsp;&nbsp;&nbsp;the&nbsp;console&nbsp;after&nbsp;the&nbsp;prompt.&nbsp;The&nbsp;user&nbsp;gets&nbsp;the&nbsp;impression&nbsp;that<br>
&nbsp;&nbsp;&nbsp;&nbsp;the&nbsp;prompt&nbsp;was&nbsp;never&nbsp;displayed.&nbsp;Here's&nbsp;the&nbsp;sequence&nbsp;of&nbsp;events:<br>
<br>
&nbsp;&nbsp;&nbsp;&nbsp;1.&nbsp;Script&nbsp;launches&nbsp;background&nbsp;command.<br>
&nbsp;&nbsp;&nbsp;&nbsp;2.&nbsp;Script&nbsp;exits.<br>
&nbsp;&nbsp;&nbsp;&nbsp;3.&nbsp;Shell&nbsp;displays&nbsp;the&nbsp;prompt.<br>
&nbsp;&nbsp;&nbsp;&nbsp;4.&nbsp;Background&nbsp;command&nbsp;continues&nbsp;running&nbsp;and&nbsp;writing&nbsp;text&nbsp;to&nbsp;the<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;console.<br>
&nbsp;&nbsp;&nbsp;&nbsp;5.&nbsp;Background&nbsp;command&nbsp;finishes.<br>
&nbsp;&nbsp;&nbsp;&nbsp;6.&nbsp;User&nbsp;doesn't&nbsp;see&nbsp;a&nbsp;prompt&nbsp;at&nbsp;the&nbsp;bottom&nbsp;of&nbsp;the&nbsp;output,&nbsp;thinks&nbsp;script<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;is&nbsp;hanging.<br>
</P
></BLOCKQUOTE
><P
>Placing a <B
CLASS="COMMAND"
>wait</B
> after the background
command seems to remedy this.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# test.sh
ls -l &#38;
echo "Done."
wait</PRE
></FONT
></TD
></TR
></TABLE
>
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>./test.sh</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Done.
[bozo@localhost test-scripts]$ total 1
-rwxr-xr-x 1 bozo bozo 34 Oct 11 15:09 test.sh</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
<A
HREF="#IOREDIRREF"
>Redirecting</A
> the
output of the command to a file or even to
<TT
CLASS="FILENAME"
>/dev/null</TT
> also takes care of this
problem.
</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SUSPENDREF"
></A
><B
CLASS="COMMAND"
>suspend</B
></DT
><DD
><P
>This has a similar effect to
<B
CLASS="KEYCAP"
>Control</B
>-<B
CLASS="KEYCAP"
>Z</B
>,
but it suspends the shell (the shell's parent process should
resume it at an appropriate time).</P
></DD
><DT
><A
NAME="LOGOUTREF"
></A
><B
CLASS="COMMAND"
>logout</B
></DT
><DD
><P
>Exit a login shell, optionally specifying an <A
HREF="#EXITSTATUSREF"
>exit status</A
>.</P
></DD
><DT
><A
NAME="TIMESREF"
></A
><B
CLASS="COMMAND"
>times</B
></DT
><DD
><P
>Gives statistics on the system time elapsed when
executing commands, in the following form:
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="COMPUTEROUTPUT"
>0m0.020s 0m0.020s</TT
></PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>This capability is of relatively limited value, since it is not common to
profile and benchmark shell scripts.</P
></DD
><DT
><A
NAME="KILLREF"
></A
><B
CLASS="COMMAND"
>kill</B
></DT
><DD
><P
>Forcibly terminate a process by sending it an
appropriate <I
CLASS="FIRSTTERM"
>terminate</I
> signal
(see <A
HREF="#KILLPROCESS"
>Example 17-6</A
>).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SELFDESTRUCT"
></A
><P
><B
>Example 15-27. A script that kills itself</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# self-destruct.sh
kill $$ # Script kills its own process here.
# Recall that "$$" is the script's PID.
echo "This line will not echo."
# Instead, the shell sends a "Terminated" message to stdout.
exit 0 # Normal exit? No!
# After this script terminates prematurely,
#+ what exit status does it return?
#
# sh self-destruct.sh
# echo $?
# 143
#
# 143 = 128 + 15
# TERM signal</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ZOMBIEREF"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><TT
CLASS="USERINPUT"
><B
>kill -l</B
></TT
> lists all the
<A
HREF="#SIGNALD"
>signals</A
> (as does the
file <TT
CLASS="FILENAME"
>/usr/include/asm/signal.h</TT
>).
A <TT
CLASS="USERINPUT"
><B
>kill -9</B
></TT
> is a <I
CLASS="FIRSTTERM"
>sure
kill</I
>, which will usually terminate a
process that stubbornly refuses to die with a plain
<B
CLASS="COMMAND"
>kill</B
>. Sometimes, a <TT
CLASS="USERINPUT"
><B
>kill
-15</B
></TT
> works. A <I
CLASS="FIRSTTERM"
>zombie</I
> process,
that is, a child process that has terminated, but that
the <A
HREF="#FORKREF"
>parent process</A
>
has not (yet) killed, cannot be killed by a logged-on
user -- you can't kill something that is already dead --
but <B
CLASS="COMMAND"
>init</B
> will generally clean it up
sooner or later.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="KILLALLREF"
></A
><B
CLASS="COMMAND"
>killall</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>killall</B
> command
kills a running process by <I
CLASS="FIRSTTERM"
>name</I
>,
rather than by <A
HREF="#PROCESSIDREF"
>process ID</A
>.
If there are multiple instances of a particular command running,
then doing a <I
CLASS="FIRSTTERM"
>killall</I
> on that command will
terminate them <EM
>all</EM
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This refers to the <B
CLASS="COMMAND"
>killall</B
>
command in <TT
CLASS="FILENAME"
>/usr/bin</TT
>,
<EM
>not</EM
> the <A
HREF="#KILLALL2REF"
>killall script</A
> in <TT
CLASS="FILENAME"
>/etc/rc.d/init.d</TT
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="COMMANDREF"
></A
><B
CLASS="COMMAND"
>command</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>command</B
> directive
disables aliases and functions for the command immediately
following it.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>command ls</B
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This is one of three shell directives that
effect script command processing. The others are
<A
HREF="#BLTREF"
>builtin</A
> and <A
HREF="#ENABLEREF"
>enable</A
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="BLTREF"
></A
><B
CLASS="COMMAND"
>builtin</B
></DT
><DD
><P
>Invoking <B
CLASS="COMMAND"
>builtin
BUILTIN_COMMAND</B
> runs the command
<TT
CLASS="REPLACEABLE"
><I
>BUILTIN_COMMAND</I
></TT
> as a shell <A
HREF="#BUILTINREF"
>builtin</A
>, temporarily disabling
both functions and external system commands with the
same name.</P
></DD
><DT
><A
NAME="ENABLEREF"
></A
><B
CLASS="COMMAND"
>enable</B
></DT
><DD
><P
>This either enables or disables a shell
builtin command. As an example, <TT
CLASS="REPLACEABLE"
><I
>enable -n
kill</I
></TT
> disables the shell builtin <A
HREF="#KILLREF"
>kill</A
>, so that when Bash
subsequently encounters <I
CLASS="FIRSTTERM"
>kill</I
>, it invokes
the external command <TT
CLASS="FILENAME"
>/bin/kill</TT
>.</P
><P
><A
NAME="ENABLEREF1"
></A
>The <TT
CLASS="OPTION"
>-a</TT
>
option to <I
CLASS="FIRSTTERM"
>enable</I
> lists all the
shell builtins, indicating whether or not they
are enabled. The <TT
CLASS="OPTION"
>-f filename</TT
>
option lets <I
CLASS="FIRSTTERM"
>enable</I
> load a <A
HREF="#BUILTINREF"
>builtin</A
> as a shared library
(DLL) module from a properly compiled object file.
<A
NAME="AEN9928"
HREF="#FTN.AEN9928"
><SPAN
CLASS="footnote"
>[68]</SPAN
></A
>.
</P
></DD
><DT
><A
NAME="AUTOLOADREF"
></A
><B
CLASS="COMMAND"
>autoload</B
></DT
><DD
><P
>This is a port to Bash of the
<I
CLASS="FIRSTTERM"
>ksh</I
> autoloader. With
<B
CLASS="COMMAND"
>autoload</B
> in place, a function with
an <I
CLASS="FIRSTTERM"
>autoload</I
> declaration will load from an
external file at its first invocation.
<A
NAME="AEN9949"
HREF="#FTN.AEN9949"
><SPAN
CLASS="footnote"
>[69]</SPAN
></A
>
This saves system resources.</P
><P
>Note that <I
CLASS="FIRSTTERM"
>autoload</I
> is not a part of the
core Bash installation. It needs to be loaded in with
<TT
CLASS="REPLACEABLE"
><I
>enable -f</I
></TT
> (see above).</P
></DD
></DL
></DIV
><P
><A
NAME="JOBIDTABLE0"
></A
></P
><DIV
CLASS="TABLE"
><A
NAME="JOBIDTABLE"
></A
><P
><B
>Table 15-1. Job identifiers</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Notation</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Meaning</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>%N</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Job number [N]</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>%S</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Invocation (command-line) of job begins with string <EM
>S</EM
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>%?S</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Invocation (command-line) of job contains within it string <EM
>S</EM
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>%%</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><SPAN
CLASS="QUOTE"
>"current"</SPAN
> job (last job stopped in
foreground or started in background)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>%+</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><SPAN
CLASS="QUOTE"
>"current"</SPAN
> job (last job stopped in
foreground or started in background)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>%-</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Last job</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>$!</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Last background process</TD
></TR
></TBODY
></TABLE
></DIV
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="EXTERNAL"
></A
>Chapter 16. External Filters, Programs and Commands</H1
><P
><A
NAME="EXTERNALREF"
></A
></P
><P
>Standard UNIX commands make shell scripts more versatile. The
power of scripts comes from coupling system commands and shell
directives with simple programming constructs.</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="BASIC"
></A
>16.1. Basic Commands</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="BASICCOMMANDS1"
></A
>The first commands a novice learns</B
></P
><DL
><DT
><A
NAME="LSREF"
></A
><B
CLASS="COMMAND"
>ls</B
></DT
><DD
><P
>The basic file <SPAN
CLASS="QUOTE"
>"list"</SPAN
> command. It is all too easy
to underestimate the power of this humble command. For
example, using the <TT
CLASS="OPTION"
>-R</TT
>, recursive option,
<B
CLASS="COMMAND"
>ls</B
> provides a tree-like listing of
a directory structure. Other useful options are
<TT
CLASS="OPTION"
>-S</TT
>, sort listing by file size,
<TT
CLASS="OPTION"
>-t</TT
>, sort by file modification time,
<TT
CLASS="OPTION"
>-v</TT
>, sort by (numerical) version numbers
embedded in the filenames,
<A
NAME="AEN10025"
HREF="#FTN.AEN10025"
><SPAN
CLASS="footnote"
>[70]</SPAN
></A
>
<TT
CLASS="OPTION"
>-b</TT
>, show escape characters, and
<TT
CLASS="OPTION"
>-i</TT
>, show file inodes (see <A
HREF="#IDELETE"
>Example 16-4</A
>).</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:44 chapter10.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:44 chapter11.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:44 chapter12.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:44 chapter1.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:44 chapter2.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:44 chapter3.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:49 Chapter_headings.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:49 Preface.txt</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -lv</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> total 0
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:49 Chapter_headings.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:49 Preface.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:44 chapter1.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:44 chapter2.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:44 chapter3.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:44 chapter10.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:44 chapter11.txt
-rw-rw-r-- 1 bozo bozo 0 Sep 14 18:44 chapter12.txt</TT
></PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
> The <I
CLASS="FIRSTTERM"
>ls</I
> command returns a
non-zero <A
HREF="#EXITSTATUSREF"
>exit status</A
> when
attempting to list a non-existent file.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls abc</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>ls: abc: No such file or directory</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $?</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>2</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX40"
></A
><P
><B
>Example 16-1. Using <I
CLASS="FIRSTTERM"
>ls</I
> to create a table of contents
for burning a <SPAN
CLASS="ABBREV"
>CDR</SPAN
> disk</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ex40.sh (burn-cd.sh)
# Script to automate burning a CDR.
SPEED=10 # May use higher speed if your hardware supports it.
IMAGEFILE=cdimage.iso
CONTENTSFILE=contents
# DEVICE=/dev/cdrom For older versions of cdrecord
DEVICE="1,0,0"
DEFAULTDIR=/opt # This is the directory containing the data to be burned.
# Make sure it exists.
# Exercise: Add a test for this.
# Uses Joerg Schilling's "cdrecord" package:
# http://www.fokus.fhg.de/usr/schilling/cdrecord.html
# If this script invoked as an ordinary user, may need to suid cdrecord
#+ chmod u+s /usr/bin/cdrecord, as root.
# Of course, this creates a security hole, though a relatively minor one.
if [ -z "$1" ]
then
IMAGE_DIRECTORY=$DEFAULTDIR
# Default directory, if not specified on command-line.
else
IMAGE_DIRECTORY=$1
fi
# Create a "table of contents" file.
ls -lRF $IMAGE_DIRECTORY &#62; $IMAGE_DIRECTORY/$CONTENTSFILE
# The "l" option gives a "long" file listing.
# The "R" option makes the listing recursive.
# The "F" option marks the file types (directories get a trailing /).
echo "Creating table of contents."
# Create an image file preparatory to burning it onto the CDR.
mkisofs -r -o $IMAGEFILE $IMAGE_DIRECTORY
echo "Creating ISO9660 file system image ($IMAGEFILE)."
# Burn the CDR.
echo "Burning the disk."
echo "Please be patient, this will take a while."
wodim -v -isosize dev=$DEVICE $IMAGEFILE
# In newer Linux distros, the "wodim" utility assumes the
#+ functionality of "cdrecord."
exitcode=$?
echo "Exit code = $exitcode"
exit $exitcode</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="CATREF"
></A
><B
CLASS="COMMAND"
>cat</B
>, <B
CLASS="COMMAND"
>tac</B
></DT
><DD
><P
><B
CLASS="COMMAND"
>cat</B
>, an acronym for
<I
CLASS="WORDASWORD"
>concatenate</I
>,
lists a file to <TT
CLASS="FILENAME"
>stdout</TT
>. When
combined with redirection (<SPAN
CLASS="TOKEN"
>&#62;</SPAN
> or
<SPAN
CLASS="TOKEN"
>&#62;&#62;</SPAN
>), it is commonly used to concatenate
files.
<A
NAME="CATUSES"
></A
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Uses of 'cat'
cat filename # Lists the file.
cat file.1 file.2 file.3 &#62; file.123 # Combines three files into one.</PRE
></FONT
></TD
></TR
></TABLE
>
The <TT
CLASS="OPTION"
>-n</TT
> option to <B
CLASS="COMMAND"
>cat</B
>
inserts consecutive numbers before all lines of the
target file(s). The <TT
CLASS="OPTION"
>-b</TT
> option numbers
only the non-blank lines. The <TT
CLASS="OPTION"
>-v</TT
> option
echoes nonprintable characters, using <SPAN
CLASS="TOKEN"
>^</SPAN
>
notation. The <TT
CLASS="OPTION"
>-s</TT
> option squeezes multiple
consecutive blank lines into a single blank line.</P
><P
>See also <A
HREF="#LNUM"
>Example 16-28</A
> and <A
HREF="#ROT13"
>Example 16-24</A
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="CATLESSEFF"
></A
>
In a <A
HREF="#PIPEREF"
>pipe</A
>, it may be
more efficient to <A
HREF="#IOREDIRREF"
>redirect</A
>
the <TT
CLASS="FILENAME"
>stdin</TT
> to a file, rather than to
<B
CLASS="COMMAND"
>cat</B
> the file.
</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat filename | tr a-z A-Z
tr a-z A-Z &#60; filename # Same effect, but starts one less process,
#+ and also dispenses with the pipe.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><P
><B
CLASS="COMMAND"
>tac</B
>, is the inverse of
<I
CLASS="WORDASWORD"
>cat</I
>, listing a file backwards from its end.</P
></DD
><DT
><A
NAME="REVREF"
></A
><B
CLASS="COMMAND"
>rev</B
></DT
><DD
><P
>reverses each line of a file, and outputs to
<TT
CLASS="FILENAME"
>stdout</TT
>. This does not have the same effect
as <B
CLASS="COMMAND"
>tac</B
>, as it preserves the order of
the lines, but flips each one around (mirror image).</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat file1.txt</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>This is line 1.
This is line 2.</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>tac file1.txt</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>This is line 2.
This is line 1.</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>rev file1.txt</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>.1 enil si sihT
.2 enil si sihT</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="CPREF"
></A
><B
CLASS="COMMAND"
>cp</B
></DT
><DD
><P
>This is the file copy command. <TT
CLASS="USERINPUT"
><B
>cp file1
file2</B
></TT
> copies <TT
CLASS="FILENAME"
>file1</TT
>
to <TT
CLASS="FILENAME"
>file2</TT
>, overwriting
<TT
CLASS="FILENAME"
>file2</TT
> if it already exists (see <A
HREF="#EX42"
>Example 16-6</A
>).</P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Particularly useful are the <TT
CLASS="OPTION"
>-a</TT
>
archive flag (for copying an entire directory tree),
the <TT
CLASS="OPTION"
>-u</TT
> update flag (which prevents
overwriting identically-named newer files), and the
<TT
CLASS="OPTION"
>-r</TT
> and <TT
CLASS="OPTION"
>-R</TT
> recursive
flags.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cp -u source_dir/* dest_dir
# "Synchronize" dest_dir to source_dir
#+ by copying over all newer and not previously existing files.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="MVREF"
></A
><B
CLASS="COMMAND"
>mv</B
></DT
><DD
><P
>This is the file <I
CLASS="FIRSTTERM"
>move</I
> command.
It is equivalent to a combination of <B
CLASS="COMMAND"
>cp</B
>
and <B
CLASS="COMMAND"
>rm</B
>. It may be used to move multiple
files to a directory, or even to rename a directory. For
some examples of using <B
CLASS="COMMAND"
>mv</B
> in a script,
see <A
HREF="#RFE"
>Example 10-11</A
> and <A
HREF="#RN"
>Example A-2</A
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>When used in a non-interactive script,
<B
CLASS="COMMAND"
>mv</B
> takes the <TT
CLASS="OPTION"
>-f</TT
>
(<I
CLASS="FIRSTTERM"
>force</I
>) option to bypass user
input.</P
><P
>When a directory is moved to a preexisting directory,
it becomes a subdirectory of the destination directory.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>mv source_directory target_directory</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -lF target_directory</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 1
drwxrwxr-x 2 bozo bozo 1024 May 28 19:20 source_directory/</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="RMREF"
></A
><B
CLASS="COMMAND"
>rm</B
></DT
><DD
><P
>Delete (remove) a file or files. The <TT
CLASS="OPTION"
>-f</TT
>
option forces removal of even readonly files, and is useful
for bypassing user input in a script.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="DASHREM"
></A
></P
><P
>The <I
CLASS="FIRSTTERM"
>rm</I
> command will, by
itself, fail to remove filenames beginning with
a dash. Why? Because <I
CLASS="FIRSTTERM"
>rm</I
>
sees a dash-prefixed filename as an
<I
CLASS="FIRSTTERM"
>option</I
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>rm -badname</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>rm: invalid option -- b
Try `rm --help' for more information.</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> One clever workaround is to precede
the filename with a <SPAN
CLASS="QUOTE"
>" -- "</SPAN
> (the
<I
CLASS="FIRSTTERM"
>end-of-options</I
> flag).
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>rm -- -badname</B
></TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> Another method to is to preface the filename to be removed
with a <TT
CLASS="FILENAME"
>dot-slash</TT
> .
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>rm ./-badname</B
></TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="WARNING"
><P
></P
><TABLE
CLASS="WARNING"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/warning.gif"
HSPACE="5"
ALT="Warning"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="RMRECURS"
></A
>When used with the
recursive flag <TT
CLASS="OPTION"
>-r</TT
>, this command removes
files all the way down the directory tree from the current
directory. A careless <B
CLASS="COMMAND"
>rm -rf *</B
> can wipe
out a big chunk of a directory structure.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="RMDIRREF"
></A
><B
CLASS="COMMAND"
>rmdir</B
></DT
><DD
><P
>Remove directory. The directory must be empty of
all files -- including <SPAN
CLASS="QUOTE"
>"invisible"</SPAN
>
<I
CLASS="FIRSTTERM"
>dotfiles</I
>
<A
NAME="AEN10228"
HREF="#FTN.AEN10228"
><SPAN
CLASS="footnote"
>[71]</SPAN
></A
>
-- for this command to succeed.</P
></DD
><DT
><A
NAME="MKDIRREF"
></A
><B
CLASS="COMMAND"
>mkdir</B
></DT
><DD
><P
>Make directory, creates a new directory. For example,
<TT
CLASS="USERINPUT"
><B
>mkdir -p project/programs/December</B
></TT
>
creates the named directory. The
<TT
CLASS="REPLACEABLE"
><I
>-p</I
></TT
> option automatically creates
any necessary parent directories.</P
></DD
><DT
><A
NAME="CHMODREF"
></A
><B
CLASS="COMMAND"
>chmod</B
></DT
><DD
><P
>Changes the attributes of an existing file or directory
(see <A
HREF="#EX44"
>Example 15-14</A
>).</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>chmod +x filename
# Makes "filename" executable for all users.
chmod u+s filename
# Sets "suid" bit on "filename" permissions.
# An ordinary user may execute "filename" with same privileges as the file's owner.
# (This does not apply to shell scripts.)</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>chmod 644 filename
# Makes "filename" readable/writable to owner, readable to others
#+ (octal mode).
chmod 444 filename
# Makes "filename" read-only for all.
# Modifying the file (for example, with a text editor)
#+ not allowed for a user who does not own the file (except for root),
#+ and even the file owner must force a file-save
#+ if she modifies the file.
# Same restrictions apply for deleting the file.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>chmod 1777 directory-name
# Gives everyone read, write, and execute permission in directory,
#+ however also sets the "sticky bit".
# This means that only the owner of the directory,
#+ owner of the file, and, of course, root
#+ can delete any particular file in that directory.
chmod 111 directory-name
# Gives everyone execute-only permission in a directory.
# This means that you can execute and READ the files in that directory
#+ (execute permission necessarily includes read permission
#+ because you can't execute a file without being able to read it).
# But you can't list the files or search for them with the "find" command.
# These restrictions do not apply to root.
chmod 000 directory-name
# No permissions at all for that directory.
# Can't read, write, or execute files in it.
# Can't even list files in it or "cd" to it.
# But, you can rename (mv) the directory
#+ or delete it (rmdir) if it is empty.
# You can even symlink to files in the directory,
#+ but you can't read, write, or execute the symlinks.
# These restrictions do not apply to root.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="CHATTRREF"
></A
><B
CLASS="COMMAND"
>chattr</B
></DT
><DD
><P
><B
CLASS="COMMAND"
>Ch</B
>ange file
<B
CLASS="COMMAND"
>attr</B
>ibutes. This is analogous to
<B
CLASS="COMMAND"
>chmod</B
> above, but with different options
and a different invocation syntax, and it works only on
<I
CLASS="FIRSTTERM"
>ext2/ext3</I
> filesystems.</P
><P
>One particularly interesting <B
CLASS="COMMAND"
>chattr</B
>
option is <TT
CLASS="OPTION"
>i</TT
>. A <B
CLASS="COMMAND"
>chattr +i
<TT
CLASS="FILENAME"
>filename</TT
></B
> marks the file
as immutable. The file cannot be modified, linked to, or
deleted, <EM
>not even by root</EM
>. This
file attribute can be set or removed only by
<I
CLASS="FIRSTTERM"
>root</I
>. In a similar fashion,
the <TT
CLASS="OPTION"
>a</TT
> option marks the file as append
only.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>chattr +i file1.txt</B
></TT
>
<TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>rm file1.txt</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>rm: remove write-protected regular file `file1.txt'? y
rm: cannot remove `file1.txt': Operation not permitted</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>If a file has the <TT
CLASS="OPTION"
>s</TT
> (secure)
attribute set, then when it is deleted its block is
overwritten with binary zeroes.
<A
NAME="AEN10301"
HREF="#FTN.AEN10301"
><SPAN
CLASS="footnote"
>[72]</SPAN
></A
>
</P
><P
>If a file has the <TT
CLASS="OPTION"
>u</TT
> (undelete)
attribute set, then when it is deleted, its contents can still
be retrieved (undeleted).</P
><P
>If a file has the <TT
CLASS="OPTION"
>c</TT
> (compress)
attribute set, then it will automatically be compressed
on writes to disk, and uncompressed on reads.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The file attributes set with
<B
CLASS="COMMAND"
>chattr</B
> do not show in a file listing
(<B
CLASS="COMMAND"
>ls -l</B
>).</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="LINKREF"
></A
><B
CLASS="COMMAND"
>ln</B
></DT
><DD
><P
>Creates links to pre-existings files. A <SPAN
CLASS="QUOTE"
>"link"</SPAN
>
is a reference to a file, an alternate name for it.
The <B
CLASS="COMMAND"
>ln</B
> command permits referencing
the linked file by more than one name and is a superior
alternative to aliasing (see <A
HREF="#EX18"
>Example 4-6</A
>).</P
><P
>The <B
CLASS="COMMAND"
>ln</B
> creates only a reference, a
pointer to the file only a few bytes in size.</P
><P
><A
NAME="SYMLINKREF"
></A
></P
><P
>The <B
CLASS="COMMAND"
>ln</B
> command is most often used
with the <TT
CLASS="OPTION"
>-s</TT
>, symbolic or
<SPAN
CLASS="QUOTE"
>"soft"</SPAN
> link flag. Advantages of using the
<TT
CLASS="OPTION"
>-s</TT
> flag are that it permits linking across
file systems or to directories.</P
><P
>The syntax of the command is a bit tricky. For example:
<TT
CLASS="USERINPUT"
><B
>ln -s oldfile newfile</B
></TT
> links the
previously existing <TT
CLASS="FILENAME"
>oldfile</TT
> to the
newly created link, <TT
CLASS="FILENAME"
>newfile</TT
>.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>If a file named <TT
CLASS="FILENAME"
>newfile</TT
> has
previously existed, an error message will
result.</P
></TD
></TR
></TABLE
></DIV
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN10336"
></A
><P
><B
>Which type of link to use?</B
></P
><P
>As John Macdonald explains it:</P
><P
>Both of these [types of links] provide a certain measure of dual reference
-- if you edit the contents of the file using any name,
your changes will affect both the original name and either
a hard or soft new name. The differences between them
occurs when you work at a higher level. The advantage of
a hard link is that the new name is totally independent
of the old name -- if you remove or rename the old name,
that does not affect the hard link, which continues
to point to the data while it would leave a soft link
hanging pointing to the old name which is no longer
there. The advantage of a soft link is that it can refer
to a different file system (since it is just a reference
to a file name, not to actual data). And, unlike a hard
link, a symbolic link can refer to a directory.</P
></DIV
></TD
></TR
></TABLE
><P
><A
NAME="LINKMINVOK"
></A
></P
><P
>Links give the ability to invoke a script (or any other type
of executable) with multiple names, and having that script
behave according to how it was invoked.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="HELLOL"
></A
><P
><B
>Example 16-2. Hello or Good-bye</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# hello.sh: Saying "hello" or "goodbye"
#+ depending on how script is invoked.
# Make a link in current working directory ($PWD) to this script:
# ln -s hello.sh goodbye
# Now, try invoking this script both ways:
# ./hello.sh
# ./goodbye
HELLO_CALL=65
GOODBYE_CALL=66
if [ $0 = "./goodbye" ]
then
echo "Good-bye!"
# Some other goodbye-type commands, as appropriate.
exit $GOODBYE_CALL
fi
echo "Hello!"
# Some other hello-type commands, as appropriate.
exit $HELLO_CALL</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="MANREF"
></A
><B
CLASS="COMMAND"
>man</B
>, <A
NAME="INFOREF"
></A
><B
CLASS="COMMAND"
>info</B
></DT
><DD
><P
>These commands access the manual and information pages on
system commands and installed utilities. When available, the
<I
CLASS="FIRSTTERM"
>info</I
> pages usually contain more detailed
descriptions than do the <I
CLASS="FIRSTTERM"
>man</I
> pages.</P
><P
>There have been various attempts at
<SPAN
CLASS="QUOTE"
>"automating"</SPAN
> the writing of <I
CLASS="FIRSTTERM"
>man
pages</I
>. For a script that makes a tentative first
step in that direction, see <A
HREF="#MANED"
>Example A-39</A
>.</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="MOREADV"
></A
>16.2. Complex Commands</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="CCLISTING1"
></A
>Commands for more advanced users</B
></P
><DL
><DT
><A
NAME="FINDREF"
></A
><B
CLASS="COMMAND"
>find</B
></DT
><DD
><P
><A
NAME="FINDREF0"
></A
></P
><P
>-exec <TT
CLASS="REPLACEABLE"
><I
>COMMAND</I
></TT
> \;</P
><P
>Carries out <TT
CLASS="REPLACEABLE"
><I
>COMMAND</I
></TT
> on
each file that <B
CLASS="COMMAND"
>find</B
> matches. The
command sequence terminates with <SPAN
CLASS="TOKEN"
>;</SPAN
> (the
<SPAN
CLASS="QUOTE"
>";"</SPAN
> is <A
HREF="#ESCP"
>escaped</A
> to
make certain the shell passes it to <B
CLASS="COMMAND"
>find</B
>
literally, without interpreting it as a special character).</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>find ~/ -name '*.txt'</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/home/bozo/.kde/share/apps/karm/karmdata.txt
/home/bozo/misc/irmeyc.txt
/home/bozo/test-scripts/1.txt</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="CURLYBRACKETSREF"
></A
></P
><P
>If <TT
CLASS="REPLACEABLE"
><I
>COMMAND</I
></TT
> contains
<SPAN
CLASS="TOKEN"
>{}</SPAN
>, then <B
CLASS="COMMAND"
>find</B
>
substitutes the full path name of the selected file for
<SPAN
CLASS="QUOTE"
>"{}"</SPAN
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>find ~/ -name 'core*' -exec rm {} \;
# Removes all core dump files from user's home directory.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>find /home/bozo/projects -mtime -1
# ^ Note minus sign!
# Lists all files in /home/bozo/projects directory tree
#+ that were modified within the last day (current_day - 1).
#
find /home/bozo/projects -mtime 1
# Same as above, but modified *exactly* one day ago.
#
# mtime = last modification time of the target file
# ctime = last status change time (via 'chmod' or otherwise)
# atime = last access time
DIR=/home/bozo/junk_files
find "$DIR" -type f -atime +5 -exec rm {} \;
# ^ ^^
# Curly brackets are placeholder for the path name output by "find."
#
# Deletes all files in "/home/bozo/junk_files"
#+ that have not been accessed in *at least* 5 days (plus sign ... +5).
#
# "-type filetype", where
# f = regular file
# d = directory
# l = symbolic link, etc.
#
# (The 'find' manpage and info page have complete option listings.)</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>find /etc -exec grep '[0-9][0-9]*[.][0-9][0-9]*[.][0-9][0-9]*[.][0-9][0-9]*' {} \;
# Finds all IP addresses (xxx.xxx.xxx.xxx) in /etc directory files.
# There a few extraneous hits. Can they be filtered out?
# Possibly by:
find /etc -type f -exec cat '{}' \; | tr -c '.[:digit:]' '\n' \
| grep '^[^.][^.]*\.[^.][^.]*\.[^.][^.]*\.[^.][^.]*$'
#
# [:digit:] is one of the character classes
#+ introduced with the POSIX 1003.2 standard.
# Thanks, St<53>phane Chazelas. </PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <TT
CLASS="OPTION"
>-exec</TT
> option to
<B
CLASS="COMMAND"
>find</B
> should not be confused with the <A
HREF="#EXECREF"
>exec</A
> shell builtin.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX57"
></A
><P
><B
>Example 16-3. <I
CLASS="FIRSTTERM"
>Badname</I
>, eliminate file names
in current directory containing bad characters and <A
HREF="#WHITESPACEREF"
>whitespace</A
>.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# badname.sh
# Delete filenames in current directory containing bad characters.
for filename in *
do
badname=`echo "$filename" | sed -n /[\+\{\;\"\\\=\?~\(\)\&#60;\&#62;\&#38;\*\|\$]/p`
# badname=`echo "$filename" | sed -n '/[+{;"\=?~()&#60;&#62;&#38;*|$]/p'` also works.
# Deletes files containing these nasties: + { ; " \ = ? ~ ( ) &#60; &#62; &#38; * | $
#
rm $badname 2&#62;/dev/null
# ^^^^^^^^^^^ Error messages deep-sixed.
done
# Now, take care of files containing all manner of whitespace.
find . -name "* *" -exec rm -f {} \;
# The path name of the file that _find_ finds replaces the "{}".
# The '\' ensures that the ';' is interpreted literally, as end of command.
exit 0
#---------------------------------------------------------------------
# Commands below this line will not execute because of _exit_ command.
# An alternative to the above script:
find . -name '*[+{;"\\=?~()&#60;&#62;&#38;*|$ ]*' -maxdepth 0 \
-exec rm -f '{}' \;
# The "-maxdepth 0" option ensures that _find_ will not search
#+ subdirectories below $PWD.
# (Thanks, S.C.)</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="IDELETE"
></A
><P
><B
>Example 16-4. Deleting a file by its <I
CLASS="FIRSTTERM"
>inode</I
>
number</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# idelete.sh: Deleting a file by its inode number.
# This is useful when a filename starts with an illegal character,
#+ such as ? or -.
ARGCOUNT=1 # Filename arg must be passed to script.
E_WRONGARGS=70
E_FILE_NOT_EXIST=71
E_CHANGED_MIND=72
if [ $# -ne "$ARGCOUNT" ]
then
echo "Usage: `basename $0` filename"
exit $E_WRONGARGS
fi
if [ ! -e "$1" ]
then
echo "File \""$1"\" does not exist."
exit $E_FILE_NOT_EXIST
fi
inum=`ls -i | grep "$1" | awk '{print $1}'`
# inum = inode (index node) number of file
# -----------------------------------------------------------------------
# Every file has an inode, a record that holds its physical address info.
# -----------------------------------------------------------------------
echo; echo -n "Are you absolutely sure you want to delete \"$1\" (y/n)? "
# The '-v' option to 'rm' also asks this.
read answer
case "$answer" in
[nN]) echo "Changed your mind, huh?"
exit $E_CHANGED_MIND
;;
*) echo "Deleting file \"$1\".";;
esac
find . -inum $inum -exec rm {} \;
# ^^
# Curly brackets are placeholder
#+ for text output by "find."
echo "File "\"$1"\" deleted!"
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The <B
CLASS="COMMAND"
>find</B
> command also works
without the <TT
CLASS="OPTION"
>-exec</TT
> option.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Find suid root files.
# A strange suid file might indicate a security hole,
#+ or even a system intrusion.
directory="/usr/sbin"
# Might also try /sbin, /bin, /usr/bin, /usr/local/bin, etc.
permissions="+4000" # suid root (dangerous!)
for file in $( find "$directory" -perm "$permissions" )
do
ls -ltF --author "$file"
done</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>See <A
HREF="#EX48"
>Example 16-30</A
>, <A
HREF="#EX58"
>Example 3-4</A
>,
and <A
HREF="#FINDSTRING"
>Example 11-10</A
> for scripts using
<B
CLASS="COMMAND"
>find</B
>. Its <A
HREF="#MANREF"
>manpage</A
> provides more detail
on this complex and powerful command.</P
></DD
><DT
><A
NAME="XARGSREF"
></A
><B
CLASS="COMMAND"
>xargs</B
></DT
><DD
><P
>A filter for feeding arguments to a command, and also
a tool for assembling the commands themselves. It breaks
a data stream into small enough chunks for filters and
commands to process. Consider it as a powerful replacement
for <A
HREF="#BACKQUOTESREF"
>backquotes</A
>.
In situations where <A
HREF="#COMMANDSUBREF"
>command
substitution</A
> fails with a <SPAN
CLASS="ERRORNAME"
>too
many arguments</SPAN
> error,
substituting <B
CLASS="COMMAND"
>xargs</B
> often
works.
<A
NAME="AEN10465"
HREF="#FTN.AEN10465"
><SPAN
CLASS="footnote"
>[73]</SPAN
></A
>
Normally, <B
CLASS="COMMAND"
>xargs</B
> reads from
<TT
CLASS="FILENAME"
>stdin</TT
> or from a pipe, but it can also
be given the output of a file.</P
><P
>The default command for <B
CLASS="COMMAND"
>xargs</B
> is
<A
HREF="#ECHOREF"
>echo</A
>. This means that input
piped to <B
CLASS="COMMAND"
>xargs</B
> may have linefeeds and
other whitespace characters stripped out.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 0
-rw-rw-r-- 1 bozo bozo 0 Jan 29 23:58 file1
-rw-rw-r-- 1 bozo bozo 0 Jan 29 23:58 file2</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l | xargs</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 0 -rw-rw-r-- 1 bozo bozo 0 Jan 29 23:58 file1 -rw-rw-r-- 1 bozo bozo 0 Jan...</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>find ~/mail -type f | xargs grep "Linux"</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>./misc:User-Agent: slrn/0.9.8.1 (Linux)
./sent-mail-jul-2005: hosted by the Linux Documentation Project.
./sent-mail-jul-2005: (Linux Documentation Project Site, rtf version)
./sent-mail-jul-2005: Subject: Criticism of Bozo's Windows/Linux article
./sent-mail-jul-2005: while mentioning that the Linux ext2/ext3 filesystem
. . .</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><TT
CLASS="USERINPUT"
><B
>ls | xargs -p -l gzip</B
></TT
> <A
HREF="#GZIPREF"
>gzips</A
> every file in current
directory, one at a time, prompting before each
operation.</P
><P
><A
NAME="XARGSONEATATIME"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Note that <I
CLASS="FIRSTTERM"
>xargs</I
> processes the
arguments passed to it sequentially, <EM
>one at
a time</EM
>.</P
><P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>find /usr/bin | xargs file</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/usr/bin: directory
/usr/bin/foomatic-ppd-options: perl script text executable
. . .</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="XARGSLIMARGS"
></A
></P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>An interesting <I
CLASS="FIRSTTERM"
>xargs</I
>
option is <TT
CLASS="OPTION"
>-n <TT
CLASS="REPLACEABLE"
><I
>NN</I
></TT
></TT
>,
which limits to <TT
CLASS="REPLACEABLE"
><I
>NN</I
></TT
> the number
of arguments passed.</P
><P
><TT
CLASS="USERINPUT"
><B
>ls | xargs -n 8 echo</B
></TT
> lists the files in the
current directory in <TT
CLASS="LITERAL"
>8</TT
> columns.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="XARGSWS"
></A
></P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Another useful option is
<TT
CLASS="OPTION"
>-0</TT
>, in combination with <TT
CLASS="USERINPUT"
><B
>find
-print0</B
></TT
> or <TT
CLASS="USERINPUT"
><B
>grep -lZ</B
></TT
>. This
allows handling arguments containing whitespace or
quotes.</P
><P
> <TT
CLASS="USERINPUT"
><B
>find / -type f -print0 | xargs -0 grep -liwZ GUI | xargs -0 rm -f</B
></TT
>
</P
><P
> <TT
CLASS="USERINPUT"
><B
>grep -rliwZ GUI / | xargs -0 rm -f</B
></TT
>
</P
><P
>Either of the above will remove any file containing <SPAN
CLASS="QUOTE"
>"GUI"</SPAN
>.
<EM
>(Thanks, S.C.)</EM
></P
><P
>Or:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat /proc/"$pid"/"$OPTION" | xargs -0 echo
# Formats output: ^^^^^^^^^^^^^^^
# From Han Holl's fixup of "get-commandline.sh"
#+ script in "/dev and /proc" chapter.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="XARGSMULTIPROCESS"
></A
></P
><P
>The <TT
CLASS="OPTION"
>-P</TT
> option to
<I
CLASS="FIRSTTERM"
>xargs</I
> permits running
processes in parallel. This speeds up execution
in a machine with a multicore CPU.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
ls *gif | xargs -t -n1 -P2 gif2png
# Converts all the gif images in current directory to png.
# Options:
# =======
# -t Print command to stderr.
# -n1 At most 1 argument per command line.
# -P2 Run up to 2 processes simultaneously.
# Thank you, Roberto Polli, for the inspiration.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX41"
></A
><P
><B
>Example 16-5. Logfile: Using <I
CLASS="FIRSTTERM"
>xargs</I
> to monitor system log</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Generates a log file in current directory
# from the tail end of /var/log/messages.
# Note: /var/log/messages must be world readable
# if this script invoked by an ordinary user.
# #root chmod 644 /var/log/messages
LINES=5
( date; uname -a ) &#62;&#62;logfile
# Time and machine name
echo ---------------------------------------------------------- &#62;&#62;logfile
tail -n $LINES /var/log/messages | xargs | fmt -s &#62;&#62;logfile
echo &#62;&#62;logfile
echo &#62;&#62;logfile
exit 0
# Note:
# ----
# As Frank Wang points out,
#+ unmatched quotes (either single or double quotes) in the source file
#+ may give xargs indigestion.
#
# He suggests the following substitution for line 15:
# tail -n $LINES /var/log/messages | tr -d "\"'" | xargs | fmt -s &#62;&#62;logfile
# Exercise:
# --------
# Modify this script to track changes in /var/log/messages at intervals
#+ of 20 minutes.
# Hint: Use the "watch" command. </PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="XARGSCURLYREF"
></A
></P
><P
><A
HREF="#CURLYBRACKETSREF"
>As in
<B
CLASS="COMMAND"
>find</B
></A
>, a curly bracket
pair serves as a placeholder for replacement text.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX42"
></A
><P
><B
>Example 16-6. Copying files in current directory to another</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# copydir.sh
# Copy (verbose) all files in current directory ($PWD)
#+ to directory specified on command-line.
E_NOARGS=85
if [ -z "$1" ] # Exit if no argument given.
then
echo "Usage: `basename $0` directory-to-copy-to"
exit $E_NOARGS
fi
ls . | xargs -i -t cp ./{} $1
# ^^ ^^ ^^
# -t is "verbose" (output command-line to stderr) option.
# -i is "replace strings" option.
# {} is a placeholder for output text.
# This is similar to the use of a curly-bracket pair in "find."
#
# List the files in current directory (ls .),
#+ pass the output of "ls" as arguments to "xargs" (-i -t options),
#+ then copy (cp) these arguments ({}) to new directory ($1).
#
# The net result is the exact equivalent of
#+ cp * $1
#+ unless any of the filenames has embedded "whitespace" characters.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="KILLBYNAME"
></A
><P
><B
>Example 16-7. Killing processes by name</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# kill-byname.sh: Killing processes by name.
# Compare this script with kill-process.sh.
# For instance,
#+ try "./kill-byname.sh xterm" --
#+ and watch all the xterms on your desktop disappear.
# Warning:
# -------
# This is a fairly dangerous script.
# Running it carelessly (especially as root)
#+ can cause data loss and other undesirable effects.
E_BADARGS=66
if test -z "$1" # No command-line arg supplied?
then
echo "Usage: `basename $0` Process(es)_to_kill"
exit $E_BADARGS
fi
PROCESS_NAME="$1"
ps ax | grep "$PROCESS_NAME" | awk '{print $1}' | xargs -i kill {} 2&#38;&#62;/dev/null
# ^^ ^^
# ---------------------------------------------------------------
# Notes:
# -i is the "replace strings" option to xargs.
# The curly brackets are the placeholder for the replacement.
# 2&#38;&#62;/dev/null suppresses unwanted error messages.
#
# Can grep "$PROCESS_NAME" be replaced by pidof "$PROCESS_NAME"?
# ---------------------------------------------------------------
exit $?
# The "killall" command has the same effect as this script,
#+ but using it is not quite as educational.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="WF2"
></A
><P
><B
>Example 16-8. Word frequency analysis using
<I
CLASS="FIRSTTERM"
>xargs</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# wf2.sh: Crude word frequency analysis on a text file.
# Uses 'xargs' to decompose lines of text into single words.
# Compare this example to the "wf.sh" script later on.
# Check for input file on command-line.
ARGS=1
E_BADARGS=85
E_NOFILE=86
if [ $# -ne "$ARGS" ]
# Correct number of arguments passed to script?
then
echo "Usage: `basename $0` filename"
exit $E_BADARGS
fi
if [ ! -f "$1" ] # Does file exist?
then
echo "File \"$1\" does not exist."
exit $E_NOFILE
fi
#####################################################
cat "$1" | xargs -n1 | \
# List the file, one word per line.
tr A-Z a-z | \
# Shift characters to lowercase.
sed -e 's/\.//g' -e 's/\,//g' -e 's/ /\
/g' | \
# Filter out periods and commas, and
#+ change space between words to linefeed,
sort | uniq -c | sort -nr
# Finally remove duplicates, prefix occurrence count
#+ and sort numerically.
#####################################################
# This does the same job as the "wf.sh" example,
#+ but a bit more ponderously, and it runs more slowly (why?).
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="EXPRREF"
></A
><TT
CLASS="USERINPUT"
><B
>expr</B
></TT
></DT
><DD
><P
>All-purpose expression evaluator:
Concatenates and evaluates the arguments according
to the operation given (arguments must be separated
by spaces). Operations may be arithmetic, comparison,
string, or logical.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><TT
CLASS="USERINPUT"
><B
>expr 3 + 5</B
></TT
></DT
><DD
><P
>returns <TT
CLASS="LITERAL"
>8</TT
></P
></DD
><DT
><TT
CLASS="USERINPUT"
><B
>expr 5 % 3</B
></TT
></DT
><DD
><P
>returns 2</P
></DD
><DT
><TT
CLASS="USERINPUT"
><B
>expr 1 / 0</B
></TT
></DT
><DD
><P
>returns the error message, <SPAN
CLASS="ERRORCODE"
>expr: division by
zero</SPAN
></P
><P
>Illegal arithmetic operations not allowed.</P
></DD
><DT
><TT
CLASS="USERINPUT"
><B
>expr 5 \* 3</B
></TT
></DT
><DD
><P
>returns 15</P
><P
>The multiplication operator
must be escaped when used in an arithmetic expression
with <B
CLASS="COMMAND"
>expr</B
>.</P
></DD
><DT
><TT
CLASS="USERINPUT"
><B
>y=`expr $y + 1`</B
></TT
></DT
><DD
><P
>Increment a variable, with the same effect
as <TT
CLASS="USERINPUT"
><B
>let y=y+1</B
></TT
> and
<TT
CLASS="USERINPUT"
><B
>y=$(($y+1))</B
></TT
>. This is an
example of <A
HREF="#ARITHEXPREF"
>arithmetic
expansion</A
>.</P
></DD
><DT
><A
NAME="EXPEXTRSUB"
></A
><TT
CLASS="USERINPUT"
><B
>z=`expr substr
$string $position $length`</B
></TT
></DT
><DD
><P
>Extract substring of $length characters, starting
at $position.</P
></DD
></DL
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX45"
></A
><P
><B
>Example 16-9. Using <I
CLASS="FIRSTTERM"
>expr</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Demonstrating some of the uses of 'expr'
# =======================================
echo
# Arithmetic Operators
# ---------- ---------
echo "Arithmetic Operators"
echo
a=`expr 5 + 3`
echo "5 + 3 = $a"
a=`expr $a + 1`
echo
echo "a + 1 = $a"
echo "(incrementing a variable)"
a=`expr 5 % 3`
# modulo
echo
echo "5 mod 3 = $a"
echo
echo
# Logical Operators
# ------- ---------
# Returns 1 if true, 0 if false,
#+ opposite of normal Bash convention.
echo "Logical Operators"
echo
x=24
y=25
b=`expr $x = $y` # Test equality.
echo "b = $b" # 0 ( $x -ne $y )
echo
a=3
b=`expr $a \&#62; 10`
echo 'b=`expr $a \&#62; 10`, therefore...'
echo "If a &#62; 10, b = 0 (false)"
echo "b = $b" # 0 ( 3 ! -gt 10 )
echo
b=`expr $a \&#60; 10`
echo "If a &#60; 10, b = 1 (true)"
echo "b = $b" # 1 ( 3 -lt 10 )
echo
# Note escaping of operators.
b=`expr $a \&#60;= 3`
echo "If a &#60;= 3, b = 1 (true)"
echo "b = $b" # 1 ( 3 -le 3 )
# There is also a "\&#62;=" operator (greater than or equal to).
echo
echo
# String Operators
# ------ ---------
echo "String Operators"
echo
a=1234zipper43231
echo "The string being operated upon is \"$a\"."
# length: length of string
b=`expr length $a`
echo "Length of \"$a\" is $b."
# index: position of first character in substring
# that matches a character in string
b=`expr index $a 23`
echo "Numerical position of first \"2\" in \"$a\" is \"$b\"."
# substr: extract substring, starting position &#38; length specified
b=`expr substr $a 2 6`
echo "Substring of \"$a\", starting at position 2,\
and 6 chars long is \"$b\"."
# The default behavior of the 'match' operations is to
#+ search for the specified match at the BEGINNING of the string.
#
# Using Regular Expressions ...
b=`expr match "$a" '[0-9]*'` # Numerical count.
echo Number of digits at the beginning of \"$a\" is $b.
b=`expr match "$a" '\([0-9]*\)'` # Note that escaped parentheses
# == == #+ trigger substring match.
echo "The digits at the beginning of \"$a\" are \"$b\"."
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="IMPORTANT"
><P
></P
><TABLE
CLASS="IMPORTANT"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/important.gif"
HSPACE="5"
ALT="Important"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <A
HREF="#NULLREF"
>:
(<I
CLASS="FIRSTTERM"
>null</I
>)</A
> operator
can substitute for <B
CLASS="COMMAND"
>match</B
>. For example,
<TT
CLASS="USERINPUT"
><B
>b=`expr $a : [0-9]*`</B
></TT
> is the
exact equivalent of <TT
CLASS="USERINPUT"
><B
>b=`expr match $a
[0-9]*`</B
></TT
> in the above listing.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo
echo "String operations using \"expr \$string : \" construct"
echo "==================================================="
echo
a=1234zipper5FLIPPER43231
echo "The string being operated upon is \"`expr "$a" : '\(.*\)'`\"."
# Escaped parentheses grouping operator. == ==
# ***************************
#+ Escaped parentheses
#+ match a substring
# ***************************
# If no escaped parentheses ...
#+ then 'expr' converts the string operand to an integer.
echo "Length of \"$a\" is `expr "$a" : '.*'`." # Length of string
echo "Number of digits at the beginning of \"$a\" is `expr "$a" : '[0-9]*'`."
# ------------------------------------------------------------------------- #
echo
echo "The digits at the beginning of \"$a\" are `expr "$a" : '\([0-9]*\)'`."
# == ==
echo "The first 7 characters of \"$a\" are `expr "$a" : '\(.......\)'`."
# ===== == ==
# Again, escaped parentheses force a substring match.
#
echo "The last 7 characters of \"$a\" are `expr "$a" : '.*\(.......\)'`."
# ==== end of string operator ^^
# (In fact, means skip over one or more of any characters until specified
#+ substring found.)
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
>The above script illustrates how
<B
CLASS="COMMAND"
>expr</B
> uses the <I
CLASS="FIRSTTERM"
>escaped
parentheses -- \( ... \) --</I
> grouping operator
in tandem with <A
HREF="#REGEXREF"
>regular
expression</A
> parsing to match a substring.
Here is a another example, this time from <SPAN
CLASS="QUOTE"
>"real
life."</SPAN
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Strip the whitespace from the beginning and end.
LRFDATE=`expr "$LRFDATE" : '[[:space:]]*\(.*\)[[:space:]]*$'`
# From Peter Knowles' "booklistgen.sh" script
#+ for converting files to Sony Librie/PRS-50X format.
# (http://booklistgensh.peterknowles.com)</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
HREF="#PERLREF"
>Perl</A
>,
<A
HREF="#SEDREF"
>sed</A
>, and <A
HREF="#AWKREF"
>awk</A
> have far superior string
parsing facilities. A short <B
CLASS="COMMAND"
>sed</B
> or
<B
CLASS="COMMAND"
>awk</B
> <SPAN
CLASS="QUOTE"
>"subroutine"</SPAN
> within
a script (see <A
HREF="#WRAPPER"
>Section 36.2</A
>) is an attractive
alternative to <B
CLASS="COMMAND"
>expr</B
>.</P
><P
>See <A
HREF="#STRING-MANIPULATION"
>Section 10.1</A
> for more on
using <B
CLASS="COMMAND"
>expr</B
> in string operations.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="TIMEDATE"
></A
>16.3. Time / Date Commands</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="TDLISTING1"
></A
>Time/date and timing</B
></P
><DL
><DT
><A
NAME="DATEREF"
></A
><B
CLASS="COMMAND"
>date</B
></DT
><DD
><P
>Simply invoked, <B
CLASS="COMMAND"
>date</B
> prints the date and
time to <TT
CLASS="FILENAME"
>stdout</TT
>. Where this command gets
interesting is in its formatting and parsing options.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX51"
></A
><P
><B
>Example 16-10. Using <I
CLASS="FIRSTTERM"
>date</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Exercising the 'date' command
echo "The number of days since the year's beginning is `date +%j`."
# Needs a leading '+' to invoke formatting.
# %j gives day of year.
echo "The number of seconds elapsed since 01/01/1970 is `date +%s`."
# %s yields number of seconds since "UNIX epoch" began,
#+ but how is this useful?
prefix=temp
suffix=$(date +%s) # The "+%s" option to 'date' is GNU-specific.
filename=$prefix.$suffix
echo "Temporary filename = $filename"
# It's great for creating "unique and random" temp filenames,
#+ even better than using $$.
# Read the 'date' man page for more formatting options.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The <TT
CLASS="OPTION"
>-u</TT
> option gives the UTC (Universal
Coordinated Time).</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>date</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Fri Mar 29 21:07:39 MST 2002</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>date -u</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Sat Mar 30 04:07:42 UTC 2002</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>This option facilitates calculating the time between
different dates.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="DATECALC"
></A
><P
><B
>Example 16-11. <I
CLASS="FIRSTTERM"
>Date</I
> calculations</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# date-calc.sh
# Author: Nathan Coulter
# Used in ABS Guide with permission (thanks!).
MPHR=60 # Minutes per hour.
HPD=24 # Hours per day.
diff () {
printf '%s' $(( $(date -u -d"$TARGET" +%s) -
$(date -u -d"$CURRENT" +%s)))
# %d = day of month.
}
CURRENT=$(date -u -d '2007-09-01 17:30:24' '+%F %T.%N %Z')
TARGET=$(date -u -d'2007-12-25 12:30:00' '+%F %T.%N %Z')
# %F = full date, %T = %H:%M:%S, %N = nanoseconds, %Z = time zone.
printf '\nIn 2007, %s ' \
"$(date -d"$CURRENT +
$(( $(diff) /$MPHR /$MPHR /$HPD / 2 )) days" '+%d %B')"
# %B = name of month ^ halfway
printf 'was halfway between %s ' "$(date -d"$CURRENT" '+%d %B')"
printf 'and %s\n' "$(date -d"$TARGET" '+%d %B')"
printf '\nOn %s at %s, there were\n' \
$(date -u -d"$CURRENT" +%F) $(date -u -d"$CURRENT" +%T)
DAYS=$(( $(diff) / $MPHR / $MPHR / $HPD ))
CURRENT=$(date -d"$CURRENT +$DAYS days" '+%F %T.%N %Z')
HOURS=$(( $(diff) / $MPHR / $MPHR ))
CURRENT=$(date -d"$CURRENT +$HOURS hours" '+%F %T.%N %Z')
MINUTES=$(( $(diff) / $MPHR ))
CURRENT=$(date -d"$CURRENT +$MINUTES minutes" '+%F %T.%N %Z')
printf '%s days, %s hours, ' "$DAYS" "$HOURS"
printf '%s minutes, and %s seconds ' "$MINUTES" "$(diff)"
printf 'until Christmas Dinner!\n\n'
# Exercise:
# --------
# Rewrite the diff () function to accept passed parameters,
#+ rather than using global variables.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="DATERANDREF"
></A
></P
><P
>The <I
CLASS="FIRSTTERM"
>date</I
> command has quite a
number of <I
CLASS="FIRSTTERM"
>output</I
> options. For
example <TT
CLASS="OPTION"
>%N</TT
> gives the nanosecond portion
of the current time. One interesting use for this is to
generate random integers.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>date +%N | sed -e 's/000$//' -e 's/^0//'
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
# Strip off leading and trailing zeroes, if present.
# Length of generated integer depends on
#+ how many zeroes stripped off.
# 115281032
# 63408725
# 394504284</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>There are many more options (try <B
CLASS="COMMAND"
>man
date</B
>).</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>date +%j
# Echoes day of the year (days elapsed since January 1).
date +%k%M
# Echoes hour and minute in 24-hour format, as a single digit string.
# The 'TZ' parameter permits overriding the default time zone.
date # Mon Mar 28 21:42:16 MST 2005
TZ=EST date # Mon Mar 28 23:42:16 EST 2005
# Thanks, Frank Kannemann and Pete Sjoberg, for the tip.
SixDaysAgo=$(date --date='6 days ago')
OneMonthAgo=$(date --date='1 month ago') # Four weeks back (not a month!)
OneYearAgo=$(date --date='1 year ago')</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>See also <A
HREF="#EX58"
>Example 3-4</A
> and <A
HREF="#STOPWATCH"
>Example A-43</A
>.</P
></DD
><DT
><A
NAME="ZDUMPREF"
></A
><B
CLASS="COMMAND"
>zdump</B
></DT
><DD
><P
>Time zone dump: echoes the time in a specified time zone.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>zdump EST</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>EST Tue Sep 18 22:09:22 2001 EST</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="TIMREF"
></A
><B
CLASS="COMMAND"
>time</B
></DT
><DD
><P
>Outputs verbose timing statistics for executing a command.</P
><P
><TT
CLASS="USERINPUT"
><B
>time ls -l /</B
></TT
> gives something
like this:</P
><P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="COMPUTEROUTPUT"
>real 0m0.067s
user 0m0.004s
sys 0m0.005s</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>See also the very similar <A
HREF="#TIMESREF"
>times</A
> command in the previous
section.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>As of <A
HREF="#BASH2REF"
>version 2.0</A
>
of Bash, <B
CLASS="COMMAND"
>time</B
> became a shell reserved word,
with slightly altered behavior in a pipeline.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="TOUCHREF"
></A
><B
CLASS="COMMAND"
>touch</B
></DT
><DD
><P
>Utility for updating access/modification times of a
file to current system time or other specified time,
but also useful for creating a new file. The command
<TT
CLASS="USERINPUT"
><B
>touch zzz</B
></TT
> will create a new file
of zero length, named <TT
CLASS="FILENAME"
>zzz</TT
>, assuming
that <TT
CLASS="FILENAME"
>zzz</TT
> did not previously exist.
Time-stamping empty files in this way is useful for
storing date information, for example in keeping track of
modification times on a project.
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>touch</B
> command is
equivalent to <TT
CLASS="USERINPUT"
><B
>: &#62;&#62; newfile</B
></TT
>
or <TT
CLASS="USERINPUT"
><B
>&#62;&#62; newfile</B
></TT
> (for ordinary
files).</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Before doing a <A
HREF="#CPREF"
>cp -u</A
>
(<I
CLASS="FIRSTTERM"
>copy/update</I
>), use
<B
CLASS="COMMAND"
>touch</B
> to update the time stamp of files
you don't wish overwritten.</P
><P
>As an example, if the directory <TT
CLASS="FILENAME"
>/home/bozo/tax_audit</TT
> contains the
files <TT
CLASS="FILENAME"
>spreadsheet-051606.data</TT
>,
<TT
CLASS="FILENAME"
>spreadsheet-051706.data</TT
>, and
<TT
CLASS="FILENAME"
>spreadsheet-051806.data</TT
>, then
doing a <B
CLASS="COMMAND"
>touch spreadsheet*.data</B
>
will protect these files from being overwritten
by files with the same names during a
<B
CLASS="COMMAND"
>cp -u /home/bozo/financial_info/spreadsheet*data
/home/bozo/tax_audit</B
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="ATREF"
></A
><B
CLASS="COMMAND"
>at</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>at</B
> job control command executes
a given set of commands at a specified time. Superficially,
it resembles <A
HREF="#CRONREF"
>cron</A
>, however,
<B
CLASS="COMMAND"
>at</B
> is chiefly useful for one-time execution
of a command set.</P
><P
><TT
CLASS="USERINPUT"
><B
>at 2pm January 15</B
></TT
> prompts for a set of
commands to execute at that time. These commands should be
shell-script compatible, since, for all practical
purposes, the user is typing in an executable shell
script a line at a time. Input terminates with a <A
HREF="#CTLDREF"
>Ctl-D</A
>.</P
><P
>Using either the <TT
CLASS="OPTION"
>-f</TT
> option or input
redirection (<SPAN
CLASS="TOKEN"
>&#60;</SPAN
>), <B
CLASS="COMMAND"
>at</B
>
reads a command list from a file. This file is an
executable shell script, though it should, of course,
be non-interactive. Particularly clever is including the
<A
HREF="#RUNPARTSREF"
>run-parts</A
> command in
the file to execute a different set of scripts.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>at 2:30 am Friday &#60; at-jobs.list</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>job 2 at 2000-10-27 02:30</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="BATCHREF"
></A
><B
CLASS="COMMAND"
>batch</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>batch</B
> job control command is similar to
<B
CLASS="COMMAND"
>at</B
>, but it runs a command list when the system
load drops below <TT
CLASS="LITERAL"
>.8</TT
>. Like
<B
CLASS="COMMAND"
>at</B
>, it can read commands from a file with the
<TT
CLASS="OPTION"
>-f</TT
> option.</P
><P
><A
NAME="BATCHPROCREF"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN10811"
></A
><P
></P
><P
>The concept of <I
CLASS="FIRSTTERM"
>batch processing</I
>
dates back to the era of mainframe computers. It means
running a set of commands without user intervention.</P
><P
></P
></DIV
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="CALREF"
></A
><B
CLASS="COMMAND"
>cal</B
></DT
><DD
><P
>Prints a neatly formatted monthly calendar to
<TT
CLASS="FILENAME"
>stdout</TT
>. Will do current year or a large
range of past and future years.</P
></DD
><DT
><A
NAME="SLEEPREF"
></A
><B
CLASS="COMMAND"
>sleep</B
></DT
><DD
><P
>This is the shell equivalent of a <I
CLASS="FIRSTTERM"
>wait
loop</I
>. It pauses for a specified number of
seconds, doing nothing. It can be useful for timing or
in processes running in the background, checking for
a specific event every so often (polling), as in <A
HREF="#ONLINE"
>Example 32-6</A
>. <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>sleep 3 # Pauses 3 seconds.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>sleep</B
> command defaults to
seconds, but minute, hours, or days may also be specified.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>sleep 3 h # Pauses 3 hours!</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <A
HREF="#WATCHREF"
>watch</A
> command may
be a better choice than <B
CLASS="COMMAND"
>sleep</B
> for running
commands at timed intervals.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="USLEEPREF"
></A
><B
CLASS="COMMAND"
>usleep</B
></DT
><DD
><P
><I
CLASS="FIRSTTERM"
>Microsleep</I
> (the
<I
CLASS="FIRSTTERM"
>u</I
> may be read as the Greek
<I
CLASS="FIRSTTERM"
>mu</I
>, or <I
CLASS="FIRSTTERM"
>micro-</I
>
prefix). This is the same as <B
CLASS="COMMAND"
>sleep</B
>,
above, but <SPAN
CLASS="QUOTE"
>"sleeps"</SPAN
> in microsecond
intervals. It can be used for fine-grained timing,
or for polling an ongoing process at very frequent
intervals.</P
><P
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>usleep 30 # Pauses 30 microseconds.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>This command is part of the Red Hat
<I
CLASS="FIRSTTERM"
>initscripts / rc-scripts</I
> package.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>usleep</B
> command does not
provide particularly accurate timing, and is therefore
unsuitable for critical timing loops.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="HWCLOCKREF"
></A
><B
CLASS="COMMAND"
>hwclock</B
>, <A
NAME="CLOCKREF"
></A
><B
CLASS="COMMAND"
>clock</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>hwclock</B
> command accesses or
adjusts the machine's hardware clock. Some options
require <I
CLASS="FIRSTTERM"
>root</I
> privileges. The
<TT
CLASS="FILENAME"
>/etc/rc.d/rc.sysinit</TT
> startup file
uses <B
CLASS="COMMAND"
>hwclock</B
> to set the system time
from the hardware clock at bootup.</P
><P
>The <B
CLASS="COMMAND"
>clock</B
> command is a synonym for
<B
CLASS="COMMAND"
>hwclock</B
>.</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="TEXTPROC"
></A
>16.4. Text Processing Commands</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="TPCOMMANDLISTING1"
></A
>Commands affecting text and
text files</B
></P
><DL
><DT
><A
NAME="SORTREF"
></A
><B
CLASS="COMMAND"
>sort</B
></DT
><DD
><P
>File sort utility, often used as a filter in a pipe. This
command sorts a <I
CLASS="FIRSTTERM"
>text stream</I
>
or file forwards or backwards, or according to various
keys or character positions. Using the <TT
CLASS="OPTION"
>-m</TT
>
option, it merges presorted input files. The <I
CLASS="FIRSTTERM"
>info
page</I
> lists its many capabilities and options. See
<A
HREF="#FINDSTRING"
>Example 11-10</A
>, <A
HREF="#SYMLINKS"
>Example 11-11</A
>,
and <A
HREF="#MAKEDICT"
>Example A-8</A
>.</P
></DD
><DT
><A
NAME="TSORTREF"
></A
><B
CLASS="COMMAND"
>tsort</B
></DT
><DD
><P
><I
CLASS="FIRSTTERM"
>Topological sort</I
>, reading in
pairs of whitespace-separated strings and sorting
according to input patterns. The original purpose of
<B
CLASS="COMMAND"
>tsort</B
> was to sort a list of dependencies
for an obsolete version of the <I
CLASS="FIRSTTERM"
>ld</I
>
linker in an <SPAN
CLASS="QUOTE"
>"ancient"</SPAN
> version of UNIX.</P
><P
>The results of a <I
CLASS="FIRSTTERM"
>tsort</I
> will usually
differ markedly from those of the standard
<B
CLASS="COMMAND"
>sort</B
> command, above.</P
></DD
><DT
><A
NAME="UNIQREF"
></A
><B
CLASS="COMMAND"
>uniq</B
></DT
><DD
><P
>This filter removes duplicate lines from a sorted
file. It is often seen in a pipe coupled with
<A
HREF="#SORTREF"
>sort</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat list-1 list-2 list-3 | sort | uniq &#62; final.list
# Concatenates the list files,
# sorts them,
# removes duplicate lines,
# and finally writes the result to an output file.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The useful <TT
CLASS="OPTION"
>-c</TT
> option prefixes each line of
the input file with its number of occurrences.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat testfile</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>This line occurs only once.
This line occurs twice.
This line occurs twice.
This line occurs three times.
This line occurs three times.
This line occurs three times.</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>uniq -c testfile</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> 1 This line occurs only once.
2 This line occurs twice.
3 This line occurs three times.</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>sort testfile | uniq -c | sort -nr</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> 3 This line occurs three times.
2 This line occurs twice.
1 This line occurs only once.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <TT
CLASS="USERINPUT"
><B
>sort INPUTFILE | uniq -c | sort -nr</B
></TT
>
command string produces a <I
CLASS="FIRSTTERM"
>frequency
of occurrence</I
> listing on the
<TT
CLASS="FILENAME"
>INPUTFILE</TT
> file (the
<TT
CLASS="OPTION"
>-nr</TT
> options to <B
CLASS="COMMAND"
>sort</B
>
cause a reverse numerical sort). This template finds
use in analysis of log files and dictionary lists, and
wherever the lexical structure of a document needs to
be examined.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="WF"
></A
><P
><B
>Example 16-12. Word Frequency Analysis</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# wf.sh: Crude word frequency analysis on a text file.
# This is a more efficient version of the "wf2.sh" script.
# Check for input file on command-line.
ARGS=1
E_BADARGS=85
E_NOFILE=86
if [ $# -ne "$ARGS" ] # Correct number of arguments passed to script?
then
echo "Usage: `basename $0` filename"
exit $E_BADARGS
fi
if [ ! -f "$1" ] # Check if file exists.
then
echo "File \"$1\" does not exist."
exit $E_NOFILE
fi
########################################################
# main ()
sed -e 's/\.//g' -e 's/\,//g' -e 's/ /\
/g' "$1" | tr 'A-Z' 'a-z' | sort | uniq -c | sort -nr
# =========================
# Frequency of occurrence
# Filter out periods and commas, and
#+ change space between words to linefeed,
#+ then shift characters to lowercase, and
#+ finally prefix occurrence count and sort numerically.
# Arun Giridhar suggests modifying the above to:
# . . . | sort | uniq -c | sort +1 [-f] | sort +0 -nr
# This adds a secondary sort key, so instances of
#+ equal occurrence are sorted alphabetically.
# As he explains it:
# "This is effectively a radix sort, first on the
#+ least significant column
#+ (word or string, optionally case-insensitive)
#+ and last on the most significant column (frequency)."
#
# As Frank Wang explains, the above is equivalent to
#+ . . . | sort | uniq -c | sort +0 -nr
#+ and the following also works:
#+ . . . | sort | uniq -c | sort -k1nr -k
########################################################
exit 0
# Exercises:
# ---------
# 1) Add 'sed' commands to filter out other punctuation,
#+ such as semicolons.
# 2) Modify the script to also filter out multiple spaces and
#+ other whitespace.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat testfile</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>This line occurs only once.
This line occurs twice.
This line occurs twice.
This line occurs three times.
This line occurs three times.
This line occurs three times.</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>./wf.sh testfile</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> 6 this
6 occurs
6 line
3 times
3 three
2 twice
1 only
1 once</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="EXPANDREF"
></A
><B
CLASS="COMMAND"
>expand</B
>, <B
CLASS="COMMAND"
>unexpand</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>expand</B
> filter converts tabs to
spaces. It is often used in a <A
HREF="#PIPEREF"
>pipe</A
>.</P
><P
>The <B
CLASS="COMMAND"
>unexpand</B
> filter
converts spaces to tabs. This reverses the effect of
<B
CLASS="COMMAND"
>expand</B
>.</P
></DD
><DT
><A
NAME="CUTREF"
></A
><B
CLASS="COMMAND"
>cut</B
></DT
><DD
><P
>A tool for extracting <A
HREF="#FIELDREF"
>fields</A
> from files. It is similar
to the <TT
CLASS="USERINPUT"
><B
>print $N</B
></TT
> command set in <A
HREF="#AWKREF"
>awk</A
>, but more limited. It may be
simpler to use <I
CLASS="FIRSTTERM"
>cut</I
> in a script than
<I
CLASS="FIRSTTERM"
>awk</I
>. Particularly important are the
<TT
CLASS="OPTION"
>-d</TT
> (delimiter) and <TT
CLASS="OPTION"
>-f</TT
>
(field specifier) options.</P
><P
>Using <B
CLASS="COMMAND"
>cut</B
> to obtain a listing of the
mounted filesystems:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cut -d ' ' -f1,2 /etc/mtab</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Using <B
CLASS="COMMAND"
>cut</B
> to list the OS and kernel version:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>uname -a | cut -d" " -f1,3,11,12</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Using <B
CLASS="COMMAND"
>cut</B
> to extract message headers from
an e-mail folder:
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep '^Subject:' read-messages | cut -c10-80</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Re: Linux suitable for mission-critical apps?
MAKE MILLIONS WORKING AT HOME!!!
Spam complaint
Re: Spam complaint</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Using <B
CLASS="COMMAND"
>cut</B
> to parse a file:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># List all the users in /etc/passwd.
FILENAME=/etc/passwd
for user in $(cut -d: -f1 $FILENAME)
do
echo $user
done
# Thanks, Oleg Philon for suggesting this.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TT
CLASS="USERINPUT"
><B
>cut -d ' ' -f2,3 filename</B
></TT
> is equivalent to
<TT
CLASS="USERINPUT"
><B
>awk -F'[ ]' '{ print $2, $3 }' filename</B
></TT
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>It is even possible to specify a linefeed as a
delimiter. The trick is to actually embed a linefeed
(<B
CLASS="KEYCAP"
>RETURN</B
>) in the command sequence.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cut -d'
' -f3,7,19 testfile</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>This is line 3 of testfile.
This is line 7 of testfile.
This is line 19 of testfile.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Thank you, Jaka Kranjc, for pointing this out.</P
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#BASE"
>Example 16-48</A
>.</P
></DD
><DT
><A
NAME="PASTEREF"
></A
><B
CLASS="COMMAND"
>paste</B
></DT
><DD
><P
>Tool for merging together different files into a single,
multi-column file. In combination with
<A
HREF="#CUTREF"
>cut</A
>, useful for creating system log
files.
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat items</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>alphabet blocks
building blocks
cables</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat prices</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>$1.00/dozen
$2.50 ea.
$3.75</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>paste items prices</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>alphabet blocks $1.00/dozen
building blocks $2.50 ea.
cables $3.75</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="JOINREF"
></A
><B
CLASS="COMMAND"
>join</B
></DT
><DD
><P
>Consider this a special-purpose cousin of
<B
CLASS="COMMAND"
>paste</B
>. This powerful utility allows
merging two files in a meaningful fashion, which essentially
creates a simple version of a relational database.</P
><P
>The <B
CLASS="COMMAND"
>join</B
> command operates on
exactly two files, but pastes together only those lines
with a common tagged <A
HREF="#FIELDREF"
>field</A
>
(usually a numerical label), and writes the result to
<TT
CLASS="FILENAME"
>stdout</TT
>. The files to be joined should
be sorted according to the tagged field for the matchups
to work properly.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>File: 1.data
100 Shoes
200 Laces
300 Socks</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>File: 2.data
100 $40.00
200 $1.00
300 $2.00</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>join 1.data 2.data</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>File: 1.data 2.data
100 Shoes $40.00
200 Laces $1.00
300 Socks $2.00</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The tagged field appears only once in the
output.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="HEADREF"
></A
><B
CLASS="COMMAND"
>head</B
></DT
><DD
><P
>lists the beginning of a file to <TT
CLASS="FILENAME"
>stdout</TT
>.
The default is <TT
CLASS="LITERAL"
>10</TT
> lines, but a different
number can be specified. The command has a number of
interesting options.
<DIV
CLASS="EXAMPLE"
><A
NAME="SCRIPTDETECTOR"
></A
><P
><B
>Example 16-13. Which files are scripts?</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# script-detector.sh: Detects scripts within a directory.
TESTCHARS=2 # Test first 2 characters.
SHABANG='#!' # Scripts begin with a "sha-bang."
for file in * # Traverse all the files in current directory.
do
if [[ `head -c$TESTCHARS "$file"` = "$SHABANG" ]]
# head -c2 #!
# The '-c' option to "head" outputs a specified
#+ number of characters, rather than lines (the default).
then
echo "File \"$file\" is a script."
else
echo "File \"$file\" is *not* a script."
fi
done
exit 0
# Exercises:
# ---------
# 1) Modify this script to take as an optional argument
#+ the directory to scan for scripts
#+ (rather than just the current working directory).
#
# 2) As it stands, this script gives "false positives" for
#+ Perl, awk, and other scripting language scripts.
# Correct this.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
>
<DIV
CLASS="EXAMPLE"
><A
NAME="RND"
></A
><P
><B
>Example 16-14. Generating 10-digit random numbers</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# rnd.sh: Outputs a 10-digit random number
# Script by Stephane Chazelas.
head -c4 /dev/urandom | od -N4 -tu4 | sed -ne '1s/.* //p'
# =================================================================== #
# Analysis
# --------
# head:
# -c4 option takes first 4 bytes.
# od:
# -N4 option limits output to 4 bytes.
# -tu4 option selects unsigned decimal format for output.
# sed:
# -n option, in combination with "p" flag to the "s" command,
# outputs only matched lines.
# The author of this script explains the action of 'sed', as follows.
# head -c4 /dev/urandom | od -N4 -tu4 | sed -ne '1s/.* //p'
# ----------------------------------&#62; |
# Assume output up to "sed" --------&#62; |
# is 0000000 1198195154\n
# sed begins reading characters: 0000000 1198195154\n.
# Here it finds a newline character,
#+ so it is ready to process the first line (0000000 1198195154).
# It looks at its &#60;range&#62;&#60;action&#62;s. The first and only one is
# range action
# 1 s/.* //p
# The line number is in the range, so it executes the action:
#+ tries to substitute the longest string ending with a space in the line
# ("0000000 ") with nothing (//), and if it succeeds, prints the result
# ("p" is a flag to the "s" command here, this is different
#+ from the "p" command).
# sed is now ready to continue reading its input. (Note that before
#+ continuing, if -n option had not been passed, sed would have printed
#+ the line once again).
# Now, sed reads the remainder of the characters, and finds the
#+ end of the file.
# It is now ready to process its 2nd line (which is also numbered '$' as
#+ it's the last one).
# It sees it is not matched by any &#60;range&#62;, so its job is done.
# In few word this sed commmand means:
# "On the first line only, remove any character up to the right-most space,
#+ then print it."
# A better way to do this would have been:
# sed -e 's/.* //;q'
# Here, two &#60;range&#62;&#60;action&#62;s (could have been written
# sed -e 's/.* //' -e q):
# range action
# nothing (matches line) s/.* //
# nothing (matches line) q (quit)
# Here, sed only reads its first line of input.
# It performs both actions, and prints the line (substituted) before
#+ quitting (because of the "q" action) since the "-n" option is not passed.
# =================================================================== #
# An even simpler altenative to the above one-line script would be:
# head -c4 /dev/urandom| od -An -tu4
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
>
See also <A
HREF="#EX52"
>Example 16-39</A
>.</P
></DD
><DT
><A
NAME="TAILREF"
></A
><B
CLASS="COMMAND"
>tail</B
></DT
><DD
><P
>lists the (tail) end of a file to <TT
CLASS="FILENAME"
>stdout</TT
>.
The default is <TT
CLASS="LITERAL"
>10</TT
> lines, but this can
be changed with the <TT
CLASS="OPTION"
>-n</TT
> option.
Commonly used to keep track of
changes to a system logfile, using the <TT
CLASS="OPTION"
>-f</TT
>
option, which outputs lines appended to the file.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX12"
></A
><P
><B
>Example 16-15. Using <I
CLASS="FIRSTTERM"
>tail</I
> to monitor the system log</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
filename=sys.log
cat /dev/null &#62; $filename; echo "Creating / cleaning out file."
# Creates the file if it does not already exist,
#+ and truncates it to zero length if it does.
# : &#62; filename and &#62; filename also work.
tail /var/log/messages &#62; $filename
# /var/log/messages must have world read permission for this to work.
echo "$filename contains tail end of system log."
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>To list a specific line of a text file,
<A
HREF="#PIPEREF"
>pipe</A
> the output of
<B
CLASS="COMMAND"
>head</B
> to <B
CLASS="COMMAND"
>tail -n 1</B
>.
For example <TT
CLASS="USERINPUT"
><B
>head -n 8 database.txt | tail
-n 1</B
></TT
> lists the 8th line of the file
<TT
CLASS="FILENAME"
>database.txt</TT
>.</P
><P
>To set a variable to a given block of a text file:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>var=$(head -n $m $filename | tail -n $n)
# filename = name of file
# m = from beginning of file, number of lines to end of block
# n = number of lines to set variable to (trim from end of block)</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Newer implementations of <B
CLASS="COMMAND"
>tail</B
>
deprecate the older <B
CLASS="COMMAND"
>tail -$LINES
filename</B
> usage. The standard <B
CLASS="COMMAND"
>tail -n $LINES
filename</B
> is correct.</P
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#EX41"
>Example 16-5</A
>, <A
HREF="#EX52"
>Example 16-39</A
> and
<A
HREF="#ONLINE"
>Example 32-6</A
>.</P
></DD
><DT
><A
NAME="GREPREF"
></A
><B
CLASS="COMMAND"
>grep</B
></DT
><DD
><P
>A multi-purpose file search tool that uses
<A
HREF="#REGEXREF"
>Regular Expressions</A
>.
It was originally a command/filter in the
venerable <B
CLASS="COMMAND"
>ed</B
> line editor:
<TT
CLASS="USERINPUT"
><B
>g/re/p</B
></TT
> -- <I
CLASS="FIRSTTERM"
>global -
regular expression - print</I
>.</P
><P
><P
><B
CLASS="COMMAND"
>grep</B
> <TT
CLASS="REPLACEABLE"
><I
>pattern</I
></TT
> [<TT
CLASS="REPLACEABLE"
><I
>file</I
></TT
>...]</P
>Search the target file(s) for
occurrences of <TT
CLASS="REPLACEABLE"
><I
>pattern</I
></TT
>, where
<TT
CLASS="REPLACEABLE"
><I
>pattern</I
></TT
> may be literal text
or a Regular Expression.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep '[rst]ystem.$' osinfo.txt</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>The GPL governs the distribution of the Linux operating system.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>If no target file(s) specified, <B
CLASS="COMMAND"
>grep</B
>
works as a filter on <TT
CLASS="FILENAME"
>stdout</TT
>, as in
a <A
HREF="#PIPEREF"
>pipe</A
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ps ax | grep clock</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>765 tty1 S 0:00 xclock
901 pts/1 S 0:00 grep clock</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <TT
CLASS="OPTION"
>-i</TT
> option causes a case-insensitive
search.</P
><P
>The <TT
CLASS="OPTION"
>-w</TT
> option matches only whole
words.</P
><P
>The <TT
CLASS="OPTION"
>-l</TT
> option lists only the files in which
matches were found, but not the matching lines.</P
><P
>The <TT
CLASS="OPTION"
>-r</TT
> (recursive) option searches files in
the current working directory and all subdirectories below
it.</P
><P
>The <TT
CLASS="OPTION"
>-n</TT
> option lists the matching lines,
together with line numbers.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep -n Linux osinfo.txt</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>2:This is a file containing information about Linux.
6:The GPL governs the distribution of the Linux operating system.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <TT
CLASS="OPTION"
>-v</TT
> (or <TT
CLASS="OPTION"
>--invert-match</TT
>)
option <I
CLASS="FIRSTTERM"
>filters out</I
> matches.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>grep pattern1 *.txt | grep -v pattern2
# Matches all lines in "*.txt" files containing "pattern1",
# but ***not*** "pattern2". </PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The <TT
CLASS="OPTION"
>-c</TT
> (<TT
CLASS="OPTION"
>--count</TT
>)
option gives a numerical count of matches, rather than
actually listing the matches.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>grep -c txt *.sgml # (number of occurrences of "txt" in "*.sgml" files)
# grep -cz .
# ^ dot
# means count (-c) zero-separated (-z) items matching "."
# that is, non-empty ones (containing at least 1 character).
#
printf 'a b\nc d\n\n\n\n\n\000\n\000e\000\000\nf' | grep -cz . # 3
printf 'a b\nc d\n\n\n\n\n\000\n\000e\000\000\nf' | grep -cz '$' # 5
printf 'a b\nc d\n\n\n\n\n\000\n\000e\000\000\nf' | grep -cz '^' # 5
#
printf 'a b\nc d\n\n\n\n\n\000\n\000e\000\000\nf' | grep -c '$' # 9
# By default, newline chars (\n) separate items to match.
# Note that the -z option is GNU "grep" specific.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <TT
CLASS="OPTION"
>--color</TT
> (or <TT
CLASS="OPTION"
>--colour</TT
>)
option marks the matching string in color (on the console
or in an <I
CLASS="FIRSTTERM"
>xterm</I
> window). Since
<I
CLASS="FIRSTTERM"
>grep</I
> prints out each entire line
containing the matching pattern, this lets you see exactly
<EM
>what</EM
> is being matched. See also
the <TT
CLASS="OPTION"
>-o</TT
> option, which shows only the
matching portion of the line(s).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="FROMSH"
></A
><P
><B
>Example 16-16. Printing out the <I
CLASS="FIRSTTERM"
>From</I
> lines in
stored e-mail messages</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# from.sh
# Emulates the useful 'from' utility in Solaris, BSD, etc.
# Echoes the "From" header line in all messages
#+ in your e-mail directory.
MAILDIR=~/mail/* # No quoting of variable. Why?
# Maybe check if-exists $MAILDIR: if [ -d $MAILDIR ] . . .
GREP_OPTS="-H -A 5 --color" # Show file, plus extra context lines
#+ and display "From" in color.
TARGETSTR="^From" # "From" at beginning of line.
for file in $MAILDIR # No quoting of variable.
do
grep $GREP_OPTS "$TARGETSTR" "$file"
# ^^^^^^^^^^ # Again, do not quote this variable.
echo
done
exit $?
# You might wish to pipe the output of this script to 'more'
#+ or redirect it to a file . . .</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>When invoked with more than one target file given,
<B
CLASS="COMMAND"
>grep</B
> specifies which file contains
matches.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep Linux osinfo.txt misc.txt</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>osinfo.txt:This is a file containing information about Linux.
osinfo.txt:The GPL governs the distribution of the Linux operating system.
misc.txt:The Linux operating system is steadily gaining in popularity.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>To force <B
CLASS="COMMAND"
>grep</B
> to show the filename
when searching only one target file, simply give
<TT
CLASS="FILENAME"
>/dev/null</TT
> as the second file.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep Linux osinfo.txt /dev/null</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>osinfo.txt:This is a file containing information about Linux.
osinfo.txt:The GPL governs the distribution of the Linux operating system.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><P
>If there is a successful match, <B
CLASS="COMMAND"
>grep</B
>
returns an <A
HREF="#EXITSTATUSREF"
>exit status</A
>
of 0, which makes it useful in a condition test in a
script, especially in combination with the <TT
CLASS="OPTION"
>-q</TT
>
option to suppress output.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>SUCCESS=0 # if grep lookup succeeds
word=Linux
filename=data.file
grep -q "$word" "$filename" # The "-q" option
#+ causes nothing to echo to stdout.
if [ $? -eq $SUCCESS ]
# if grep -q "$word" "$filename" can replace lines 5 - 7.
then
echo "$word found in $filename"
else
echo "$word not found in $filename"
fi</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
HREF="#ONLINE"
>Example 32-6</A
> demonstrates how to use
<B
CLASS="COMMAND"
>grep</B
> to search for a word pattern in
a system logfile.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="GRP"
></A
><P
><B
>Example 16-17. Emulating <I
CLASS="FIRSTTERM"
>grep</I
> in a script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# grp.sh: Rudimentary reimplementation of grep.
E_BADARGS=85
if [ -z "$1" ] # Check for argument to script.
then
echo "Usage: `basename $0` pattern"
exit $E_BADARGS
fi
echo
for file in * # Traverse all files in $PWD.
do
output=$(sed -n /"$1"/p $file) # Command substitution.
if [ ! -z "$output" ] # What happens if "$output" is not quoted?
then
echo -n "$file: "
echo "$output"
fi # sed -ne "/$1/s|^|${file}: |p" is equivalent to above.
echo
done
echo
exit 0
# Exercises:
# ---------
# 1) Add newlines to output, if more than one match in any given file.
# 2) Add features.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>How can <B
CLASS="COMMAND"
>grep</B
> search for two (or
more) separate patterns? What if you want
<B
CLASS="COMMAND"
>grep</B
> to display all lines in a file
or files that contain both <SPAN
CLASS="QUOTE"
>"pattern1"</SPAN
>
<EM
>and</EM
> <SPAN
CLASS="QUOTE"
>"pattern2"</SPAN
>?</P
><P
>One method is to <A
HREF="#PIPEREF"
>pipe</A
> the result of <B
CLASS="COMMAND"
>grep
pattern1</B
> to <B
CLASS="COMMAND"
>grep pattern2</B
>.</P
><P
>For example, given the following file:</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Filename: tstfile
This is a sample file.
This is an ordinary text file.
This file does not contain any unusual text.
This file is not unusual.
Here is some text.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Now, let's search this file for lines containing
<EM
>both</EM
> <SPAN
CLASS="QUOTE"
>"file"</SPAN
> and
<SPAN
CLASS="QUOTE"
>"text"</SPAN
> . . . </P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep file tstfile</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
># Filename: tstfile
This is a sample file.
This is an ordinary text file.
This file does not contain any unusual text.
This file is not unusual.</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep file tstfile | grep text</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>This is an ordinary text file.
This file does not contain any unusual text.</TT
></PRE
></FONT
></TD
></TR
></TABLE
><P
>Now, for an interesting recreational use
of <I
CLASS="FIRSTTERM"
>grep</I
> . . .</P
><DIV
CLASS="EXAMPLE"
><A
NAME="CWSOLVER"
></A
><P
><B
>Example 16-18. Crossword puzzle solver</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# cw-solver.sh
# This is actually a wrapper around a one-liner (line 46).
# Crossword puzzle and anagramming word game solver.
# You know *some* of the letters in the word you're looking for,
#+ so you need a list of all valid words
#+ with the known letters in given positions.
# For example: w...i....n
# 1???5????10
# w in position 1, 3 unknowns, i in the 5th, 4 unknowns, n at the end.
# (See comments at end of script.)
E_NOPATT=71
DICT=/usr/share/dict/word.lst
# ^^^^^^^^ Looks for word list here.
# ASCII word list, one word per line.
# If you happen to need an appropriate list,
#+ download the author's "yawl" word list package.
# http://ibiblio.org/pub/Linux/libs/yawl-0.3.2.tar.gz
# or
# http://bash.deta.in/yawl-0.3.2.tar.gz
if [ -z "$1" ] # If no word pattern specified
then #+ as a command-line argument . . .
echo #+ . . . then . . .
echo "Usage:" #+ Usage message.
echo
echo ""$0" \"pattern,\""
echo "where \"pattern\" is in the form"
echo "xxx..x.x..."
echo
echo "The x's represent known letters,"
echo "and the periods are unknown letters (blanks)."
echo "Letters and periods can be in any position."
echo "For example, try: sh cw-solver.sh w...i....n"
echo
exit $E_NOPATT
fi
echo
# ===============================================
# This is where all the work gets done.
grep ^"$1"$ "$DICT" # Yes, only one line!
# | |
# ^ is start-of-word regex anchor.
# $ is end-of-word regex anchor.
# From _Stupid Grep Tricks_, vol. 1,
#+ a book the ABS Guide author may yet get around
#+ to writing . . . one of these days . . .
# ===============================================
echo
exit $? # Script terminates here.
# If there are too many words generated,
#+ redirect the output to a file.
$ sh cw-solver.sh w...i....n
wellington
workingman
workingmen</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="EGREPREF"
></A
><B
CLASS="COMMAND"
>egrep</B
>
-- <I
CLASS="FIRSTTERM"
>extended grep</I
> -- is the same
as <B
CLASS="COMMAND"
>grep -E</B
>. This uses a somewhat
different, extended set of <A
HREF="#REGEXREF"
>Regular
Expressions</A
>, which can make the search a bit more
flexible. It also allows the boolean |
(<I
CLASS="FIRSTTERM"
>or</I
>) operator.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash $ </TT
><TT
CLASS="USERINPUT"
><B
>egrep 'matches|Matches' file.txt</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Line 1 matches.
Line 3 Matches.
Line 4 contains matches, but also Matches</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="FGREPREF"
></A
><B
CLASS="COMMAND"
>fgrep</B
> --
<I
CLASS="FIRSTTERM"
>fast grep</I
> -- is the same as
<B
CLASS="COMMAND"
>grep -F</B
>. It does a literal string search
(no <A
HREF="#REGEXREF"
>Regular Expressions</A
>),
which generally speeds things up a bit.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>On some Linux distros, <B
CLASS="COMMAND"
>egrep</B
> and
<B
CLASS="COMMAND"
>fgrep</B
> are symbolic links to, or aliases for
<B
CLASS="COMMAND"
>grep</B
>, but invoked with the
<TT
CLASS="OPTION"
>-E</TT
> and <TT
CLASS="OPTION"
>-F</TT
> options,
respectively.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="DICTLOOKUP"
></A
><P
><B
>Example 16-19. Looking up definitions in <I
CLASS="CITETITLE"
>Webster's 1913 Dictionary</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# dict-lookup.sh
# This script looks up definitions in the 1913 Webster's Dictionary.
# This Public Domain dictionary is available for download
#+ from various sites, including
#+ Project Gutenberg (http://www.gutenberg.org/etext/247).
#
# Convert it from DOS to UNIX format (with only LF at end of line)
#+ before using it with this script.
# Store the file in plain, uncompressed ASCII text.
# Set DEFAULT_DICTFILE variable below to path/filename.
E_BADARGS=85
MAXCONTEXTLINES=50 # Maximum number of lines to show.
DEFAULT_DICTFILE="/usr/share/dict/webster1913-dict.txt"
# Default dictionary file pathname.
# Change this as necessary.
# Note:
# ----
# This particular edition of the 1913 Webster's
#+ begins each entry with an uppercase letter
#+ (lowercase for the remaining characters).
# Only the *very first line* of an entry begins this way,
#+ and that's why the search algorithm below works.
if [[ -z $(echo "$1" | sed -n '/^[A-Z]/p') ]]
# Must at least specify word to look up, and
#+ it must start with an uppercase letter.
then
echo "Usage: `basename $0` Word-to-define [dictionary-file]"
echo
echo "Note: Word to look up must start with capital letter,"
echo "with the rest of the word in lowercase."
echo "--------------------------------------------"
echo "Examples: Abandon, Dictionary, Marking, etc."
exit $E_BADARGS
fi
if [ -z "$2" ] # May specify different dictionary
#+ as an argument to this script.
then
dictfile=$DEFAULT_DICTFILE
else
dictfile="$2"
fi
# ---------------------------------------------------------
Definition=$(fgrep -A $MAXCONTEXTLINES "$1 \\" "$dictfile")
# Definitions in form "Word \..."
#
# And, yes, "fgrep" is fast enough
#+ to search even a very large text file.
# Now, snip out just the definition block.
echo "$Definition" |
sed -n '1,/^[A-Z]/p' |
# Print from first line of output
#+ to the first line of the next entry.
sed '$d' | sed '$d'
# Delete last two lines of output
#+ (blank line and first line of next entry).
# ---------------------------------------------------------
exit $?
# Exercises:
# ---------
# 1) Modify the script to accept any type of alphabetic input
# + (uppercase, lowercase, mixed case), and convert it
# + to an acceptable format for processing.
#
# 2) Convert the script to a GUI application,
# + using something like 'gdialog' or 'zenity' . . .
# The script will then no longer take its argument(s)
# + from the command-line.
#
# 3) Modify the script to parse one of the other available
# + Public Domain Dictionaries, such as the U.S. Census Bureau Gazetteer.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>See also <A
HREF="#QKY"
>Example A-41</A
> for an example
of speedy <I
CLASS="FIRSTTERM"
>fgrep</I
> lookup on a large
text file.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="AGREPREF"
></A
></P
><P
><B
CLASS="COMMAND"
>agrep</B
> (<I
CLASS="FIRSTTERM"
>approximate
grep</I
>) extends the capabilities of
<B
CLASS="COMMAND"
>grep</B
> to approximate matching. The search
string may differ by a specified number of characters
from the resulting matches. This utility is not part of
the core Linux distribution.</P
><P
><A
NAME="ZEGREPREF"
></A
></P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>To search compressed files, use
<B
CLASS="COMMAND"
>zgrep</B
>, <B
CLASS="COMMAND"
>zegrep</B
>, or
<B
CLASS="COMMAND"
>zfgrep</B
>. These also work on non-compressed
files, though slower than plain <B
CLASS="COMMAND"
>grep</B
>,
<B
CLASS="COMMAND"
>egrep</B
>, <B
CLASS="COMMAND"
>fgrep</B
>.
They are handy for searching through a mixed set of files,
some compressed, some not.</P
><P
><A
NAME="BZGREPREF"
></A
></P
><P
>To search <A
HREF="#BZIPREF"
>bzipped</A
>
files, use <B
CLASS="COMMAND"
>bzgrep</B
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="LOOKREF"
></A
><B
CLASS="COMMAND"
>look</B
></DT
><DD
><P
>The command <B
CLASS="COMMAND"
>look</B
> works like
<B
CLASS="COMMAND"
>grep</B
>, but does a lookup on
a <SPAN
CLASS="QUOTE"
>"dictionary,"</SPAN
> a sorted word list.
By default, <B
CLASS="COMMAND"
>look</B
> searches for a match
in <TT
CLASS="FILENAME"
>/usr/dict/words</TT
>, but a different
dictionary file may be specified.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="LOOKUP"
></A
><P
><B
>Example 16-20. Checking words in a list for validity</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# lookup: Does a dictionary lookup on each word in a data file.
file=words.data # Data file from which to read words to test.
echo
echo "Testing file $file"
echo
while [ "$word" != end ] # Last word in data file.
do # ^^^
read word # From data file, because of redirection at end of loop.
look $word &#62; /dev/null # Don't want to display lines in dictionary file.
# Searches for words in the file /usr/share/dict/words
#+ (usually a link to linux.words).
lookup=$? # Exit status of 'look' command.
if [ "$lookup" -eq 0 ]
then
echo "\"$word\" is valid."
else
echo "\"$word\" is invalid."
fi
done &#60;"$file" # Redirects stdin to $file, so "reads" come from there.
echo
exit 0
# ----------------------------------------------------------------
# Code below line will not execute because of "exit" command above.
# Stephane Chazelas proposes the following, more concise alternative:
while read word &#38;&#38; [[ $word != end ]]
do if look "$word" &#62; /dev/null
then echo "\"$word\" is valid."
else echo "\"$word\" is invalid."
fi
done &#60;"$file"
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><B
CLASS="COMMAND"
>sed</B
>, <B
CLASS="COMMAND"
>awk</B
></DT
><DD
><P
>Scripting languages especially suited for parsing text
files and command output. May be embedded singly or in
combination in pipes and shell scripts.</P
></DD
><DT
><B
CLASS="COMMAND"
><A
HREF="#SEDREF"
>sed</A
></B
></DT
><DD
><P
>Non-interactive <SPAN
CLASS="QUOTE"
>"stream editor"</SPAN
>, permits using
many <B
CLASS="COMMAND"
>ex</B
> commands in <A
HREF="#BATCHPROCREF"
>batch</A
> mode. It finds many
uses in shell scripts.</P
></DD
><DT
><B
CLASS="COMMAND"
><A
HREF="#AWKREF"
>awk</A
></B
></DT
><DD
><P
>Programmable file extractor and formatter, good for
manipulating and/or extracting <A
HREF="#FIELDREF"
>fields</A
> (columns) in structured
text files. Its syntax is similar to C.</P
></DD
><DT
><A
NAME="WCREF"
></A
><B
CLASS="COMMAND"
>wc</B
></DT
><DD
><P
><I
CLASS="FIRSTTERM"
>wc</I
> gives a <SPAN
CLASS="QUOTE"
>"word
count"</SPAN
> on a file or I/O stream:
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash $ </TT
><TT
CLASS="USERINPUT"
><B
>wc /usr/share/doc/sed-4.1.2/README</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>13 70 447 README</TT
>
[13 lines 70 words 447 characters]</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TT
CLASS="USERINPUT"
><B
>wc -w</B
></TT
> gives only the word count.</P
><P
><TT
CLASS="USERINPUT"
><B
>wc -l</B
></TT
> gives only the line count.</P
><P
><TT
CLASS="USERINPUT"
><B
>wc -c</B
></TT
> gives only the byte count.</P
><P
><TT
CLASS="USERINPUT"
><B
>wc -m</B
></TT
> gives only the character count.</P
><P
><TT
CLASS="USERINPUT"
><B
>wc -L</B
></TT
> gives only the length of the longest line.</P
><P
>Using <B
CLASS="COMMAND"
>wc</B
> to count how many
<TT
CLASS="FILENAME"
>.txt</TT
> files are in current working directory:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>$ ls *.txt | wc -l
# Will work as long as none of the "*.txt" files
#+ have a linefeed embedded in their name.
# Alternative ways of doing this are:
# find . -maxdepth 1 -name \*.txt -print0 | grep -cz .
# (shopt -s nullglob; set -- *.txt; echo $#)
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Using <B
CLASS="COMMAND"
>wc</B
> to total up the size of all the
files whose names begin with letters in the range d - h
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>wc [d-h]* | grep total | awk '{print $3}'</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>71832</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Using <B
CLASS="COMMAND"
>wc</B
> to count the instances of the
word <SPAN
CLASS="QUOTE"
>"Linux"</SPAN
> in the main source file for
this book.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep Linux abs-book.sgml | wc -l</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>138</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>See also <A
HREF="#EX52"
>Example 16-39</A
> and <A
HREF="#REDIR4"
>Example 20-8</A
>.</P
><P
>Certain commands include some of the
functionality of <B
CLASS="COMMAND"
>wc</B
> as options.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>... | grep foo | wc -l
# This frequently used construct can be more concisely rendered.
... | grep -c foo
# Just use the "-c" (or "--count") option of grep.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="TRREF"
></A
><B
CLASS="COMMAND"
>tr</B
></DT
><DD
><P
>character translation filter.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
HREF="#UCREF"
>Must use quoting and/or
brackets</A
>, as appropriate. Quotes prevent the
shell from reinterpreting the special characters in
<B
CLASS="COMMAND"
>tr</B
> command sequences. Brackets should be
quoted to prevent expansion by the shell. </P
></TD
></TR
></TABLE
></DIV
><P
>Either <TT
CLASS="USERINPUT"
><B
>tr "A-Z" "*" &#60;filename</B
></TT
>
or <TT
CLASS="USERINPUT"
><B
>tr A-Z \* &#60;filename</B
></TT
> changes
all the uppercase letters in <TT
CLASS="FILENAME"
>filename</TT
>
to asterisks (writes to <TT
CLASS="FILENAME"
>stdout</TT
>).
On some systems this may not work, but <TT
CLASS="USERINPUT"
><B
>tr A-Z
'[**]'</B
></TT
> will.</P
><P
><A
NAME="TROPTIONS"
></A
></P
><P
>The <TT
CLASS="OPTION"
>-d</TT
> option deletes a range of
characters.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo "abcdef" # abcdef
echo "abcdef" | tr -d b-d # aef
tr -d 0-9 &#60;filename
# Deletes all digits from the file "filename".</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The <TT
CLASS="OPTION"
>--squeeze-repeats</TT
> (or
<TT
CLASS="OPTION"
>-s</TT
>) option deletes all but the
first instance of a string of consecutive characters.
This option is useful for removing excess <A
HREF="#WHITESPACEREF"
>whitespace</A
>.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "XXXXX" | tr --squeeze-repeats 'X'</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>X</TT
></PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The <TT
CLASS="OPTION"
>-c</TT
> <SPAN
CLASS="QUOTE"
>"complement"</SPAN
>
option <I
CLASS="FIRSTTERM"
>inverts</I
> the character set to
match. With this option, <B
CLASS="COMMAND"
>tr</B
> acts only
upon those characters <EM
>not</EM
> matching
the specified set.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "acfdeb123" | tr -c b-d +</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>+c+d+b++++</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Note that <B
CLASS="COMMAND"
>tr</B
> recognizes <A
HREF="#POSIXREF"
>POSIX character classes</A
>.
<A
NAME="AEN11502"
HREF="#FTN.AEN11502"
><SPAN
CLASS="footnote"
>[74]</SPAN
></A
>
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "abcd2ef1" | tr '[:alpha:]' -</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>----2--1</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX49"
></A
><P
><B
>Example 16-21. <I
CLASS="FIRSTTERM"
>toupper</I
>: Transforms a file
to all uppercase.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Changes a file to all uppercase.
E_BADARGS=85
if [ -z "$1" ] # Standard check for command-line arg.
then
echo "Usage: `basename $0` filename"
exit $E_BADARGS
fi
tr a-z A-Z &#60;"$1"
# Same effect as above, but using POSIX character set notation:
# tr '[:lower:]' '[:upper:]' &#60;"$1"
# Thanks, S.C.
# Or even . . .
# cat "$1" | tr a-z A-Z
# Or dozens of other ways . . .
exit 0
# Exercise:
# Rewrite this script to give the option of changing a file
#+ to *either* upper or lowercase.
# Hint: Use either the "case" or "select" command.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="LOWERCASE"
></A
><P
><B
>Example 16-22. <I
CLASS="FIRSTTERM"
>lowercase</I
>: Changes all
filenames in working directory to lowercase.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
#
# Changes every filename in working directory to all lowercase.
#
# Inspired by a script of John Dubois,
#+ which was translated into Bash by Chet Ramey,
#+ and considerably simplified by the author of the ABS Guide.
for filename in * # Traverse all files in directory.
do
fname=`basename $filename`
n=`echo $fname | tr A-Z a-z` # Change name to lowercase.
if [ "$fname" != "$n" ] # Rename only files not already lowercase.
then
mv $fname $n
fi
done
exit $?
# Code below this line will not execute because of "exit".
#--------------------------------------------------------#
# To run it, delete script above line.
# The above script will not work on filenames containing blanks or newlines.
# Stephane Chazelas therefore suggests the following alternative:
for filename in * # Not necessary to use basename,
# since "*" won't return any file containing "/".
do n=`echo "$filename/" | tr '[:upper:]' '[:lower:]'`
# POSIX char set notation.
# Slash added so that trailing newlines are not
# removed by command substitution.
# Variable substitution:
n=${n%/} # Removes trailing slash, added above, from filename.
[[ $filename == $n ]] || mv "$filename" "$n"
# Checks if filename already lowercase.
done
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="TRD2U"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="DU"
></A
><P
><B
>Example 16-23. <I
CLASS="FIRSTTERM"
>du</I
>: DOS to UNIX text file conversion.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Du.sh: DOS to UNIX text file converter.
E_WRONGARGS=85
if [ -z "$1" ]
then
echo "Usage: `basename $0` filename-to-convert"
exit $E_WRONGARGS
fi
NEWFILENAME=$1.unx
CR='\015' # Carriage return.
# 015 is octal ASCII code for CR.
# Lines in a DOS text file end in CR-LF.
# Lines in a UNIX text file end in LF only.
tr -d $CR &#60; $1 &#62; $NEWFILENAME
# Delete CR's and write to new file.
echo "Original DOS text file is \"$1\"."
echo "Converted UNIX text file is \"$NEWFILENAME\"."
exit 0
# Exercise:
# --------
# Change the above script to convert from UNIX to DOS.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ROT13"
></A
><P
><B
>Example 16-24. <I
CLASS="FIRSTTERM"
>rot13</I
>: ultra-weak encryption.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# rot13.sh: Classic rot13 algorithm,
# encryption that might fool a 3-year old
# for about 10 minutes.
# Usage: ./rot13.sh filename
# or ./rot13.sh &#60;filename
# or ./rot13.sh and supply keyboard input (stdin)
cat "$@" | tr 'a-zA-Z' 'n-za-mN-ZA-M' # "a" goes to "n", "b" to "o" ...
# The cat "$@" construct
#+ permits input either from stdin or from files.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="CRYPTOQUOTE"
></A
><P
><B
>Example 16-25. Generating <SPAN
CLASS="QUOTE"
>"Crypto-Quote"</SPAN
> Puzzles</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# crypto-quote.sh: Encrypt quotes
# Will encrypt famous quotes in a simple monoalphabetic substitution.
# The result is similar to the "Crypto Quote" puzzles
#+ seen in the Op Ed pages of the Sunday paper.
key=ETAOINSHRDLUBCFGJMQPVWZYXK
# The "key" is nothing more than a scrambled alphabet.
# Changing the "key" changes the encryption.
# The 'cat "$@"' construction gets input either from stdin or from files.
# If using stdin, terminate input with a Control-D.
# Otherwise, specify filename as command-line parameter.
cat "$@" | tr "a-z" "A-Z" | tr "A-Z" "$key"
# | to uppercase | encrypt
# Will work on lowercase, uppercase, or mixed-case quotes.
# Passes non-alphabetic characters through unchanged.
# Try this script with something like:
# "Nothing so needs reforming as other people's habits."
# --Mark Twain
#
# Output is:
# "CFPHRCS QF CIIOQ MINFMBRCS EQ FPHIM GIFGUI'Q HETRPQ."
# --BEML PZERC
# To reverse the encryption:
# cat "$@" | tr "$key" "A-Z"
# This simple-minded cipher can be broken by an average 12-year old
#+ using only pencil and paper.
exit 0
# Exercise:
# --------
# Modify the script so that it will either encrypt or decrypt,
#+ depending on command-line argument(s).</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="JABH"
></A
>Of course, <I
CLASS="FIRSTTERM"
>tr</I
>
lends itself to <I
CLASS="FIRSTTERM"
>code
obfuscation</I
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# jabh.sh
x="wftedskaebjgdBstbdbsmnjgz"
echo $x | tr "a-z" 'oh, turtleneck Phrase Jar!'
# Based on the Wikipedia "Just another Perl hacker" article.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="TRVARIANTS"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN11540"
></A
><P
><B
><I
CLASS="FIRSTTERM"
>tr</I
> variants</B
></P
><P
> The <B
CLASS="COMMAND"
>tr</B
> utility has two historic
variants. The BSD version does not use brackets
(<TT
CLASS="USERINPUT"
><B
>tr a-z A-Z</B
></TT
>), but the SysV one does
(<TT
CLASS="USERINPUT"
><B
>tr '[a-z]' '[A-Z]'</B
></TT
>). The GNU version
of <B
CLASS="COMMAND"
>tr</B
> resembles the BSD one.
</P
></DIV
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="FOLDREF"
></A
><B
CLASS="COMMAND"
>fold</B
></DT
><DD
><P
>A filter that wraps lines of input to a specified width.
This is especially useful with the <TT
CLASS="OPTION"
>-s</TT
>
option, which breaks lines at word spaces (see <A
HREF="#EX50"
>Example 16-26</A
> and <A
HREF="#MAILFORMAT"
>Example A-1</A
>).</P
></DD
><DT
><A
NAME="FMTREF"
></A
><B
CLASS="COMMAND"
>fmt</B
></DT
><DD
><P
>Simple-minded file formatter, used as a filter in a
pipe to <SPAN
CLASS="QUOTE"
>"wrap"</SPAN
> long lines of text
output.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX50"
></A
><P
><B
>Example 16-26. Formatted file listing.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
WIDTH=40 # 40 columns wide.
b=`ls /usr/local/bin` # Get a file listing...
echo $b | fmt -w $WIDTH
# Could also have been done by
# echo $b | fold - -s -w $WIDTH
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#EX41"
>Example 16-5</A
>.</P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A powerful alternative to <B
CLASS="COMMAND"
>fmt</B
> is
Kamil Toman's <B
CLASS="COMMAND"
>par</B
>
utility, available from <A
HREF="http://www.cs.berkeley.edu/~amc/Par/"
TARGET="_top"
>http://www.cs.berkeley.edu/~amc/Par/</A
>.
</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="COLREF"
></A
><B
CLASS="COMMAND"
>col</B
></DT
><DD
><P
>This deceptively named filter removes reverse line feeds
from an input stream. It also attempts to replace
whitespace with equivalent tabs. The chief use of
<B
CLASS="COMMAND"
>col</B
> is in filtering the output
from certain text processing utilities, such as
<B
CLASS="COMMAND"
>groff</B
> and <B
CLASS="COMMAND"
>tbl</B
>.</P
></DD
><DT
><A
NAME="COLUMNREF"
></A
><B
CLASS="COMMAND"
>column</B
></DT
><DD
><P
>Column formatter. This filter transforms list-type
text output into a <SPAN
CLASS="QUOTE"
>"pretty-printed"</SPAN
> table
by inserting tabs at appropriate places.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="COL"
></A
><P
><B
>Example 16-27. Using <I
CLASS="FIRSTTERM"
>column</I
> to format a directory
listing</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# colms.sh
# A minor modification of the example file in the "column" man page.
(printf "PERMISSIONS LINKS OWNER GROUP SIZE MONTH DAY HH:MM PROG-NAME\n" \
; ls -l | sed 1d) | column -t
# ^^^^^^ ^^
# The "sed 1d" in the pipe deletes the first line of output,
#+ which would be "total N",
#+ where "N" is the total number of files found by "ls -l".
# The -t option to "column" pretty-prints a table.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="COLRMREF"
></A
><B
CLASS="COMMAND"
>colrm</B
></DT
><DD
><P
>Column removal filter. This removes columns (characters)
from a file and writes the file, lacking the range of
specified columns, back to <TT
CLASS="FILENAME"
>stdout</TT
>.
<TT
CLASS="USERINPUT"
><B
>colrm 2 4 &#60;filename</B
></TT
> removes the
second through fourth characters from each line of the
text file <TT
CLASS="FILENAME"
>filename</TT
>.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>If the file contains tabs or nonprintable
characters, this may cause unpredictable
behavior. In such cases, consider using
<A
HREF="#EXPANDREF"
>expand</A
> and
<B
CLASS="COMMAND"
>unexpand</B
> in a pipe preceding
<B
CLASS="COMMAND"
>colrm</B
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="NLREF"
></A
><B
CLASS="COMMAND"
>nl</B
></DT
><DD
><P
>Line numbering filter: <TT
CLASS="USERINPUT"
><B
>nl filename</B
></TT
>
lists <TT
CLASS="FILENAME"
>filename</TT
> to
<TT
CLASS="FILENAME"
>stdout</TT
>, but inserts consecutive
numbers at the beginning of each non-blank line. If
<TT
CLASS="FILENAME"
>filename</TT
> omitted, operates on
<TT
CLASS="FILENAME"
>stdin.</TT
></P
><P
>The output of <B
CLASS="COMMAND"
>nl</B
> is very similar to
<TT
CLASS="USERINPUT"
><B
>cat -b</B
></TT
>, since, by default
<B
CLASS="COMMAND"
>nl</B
> does not list blank lines.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="LNUM"
></A
><P
><B
>Example 16-28. <I
CLASS="FIRSTTERM"
>nl</I
>: A self-numbering script.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# line-number.sh
# This script echoes itself twice to stdout with its lines numbered.
echo " line number = $LINENO" # 'nl' sees this as line 4
# (nl does not number blank lines).
# 'cat -n' sees it correctly as line #6.
nl `basename $0`
echo; echo # Now, let's try it with 'cat -n'
cat -n `basename $0`
# The difference is that 'cat -n' numbers the blank lines.
# Note that 'nl -ba' will also do so.
exit 0
# -----------------------------------------------------------------</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="PRREF"
></A
><B
CLASS="COMMAND"
>pr</B
></DT
><DD
><P
>Print formatting filter. This will paginate files
(or <TT
CLASS="FILENAME"
>stdout</TT
>) into sections suitable for
hard copy printing or viewing on screen. Various options
permit row and column manipulation, joining lines, setting
margins, numbering lines, adding page headers, and merging
files, among other things. The <B
CLASS="COMMAND"
>pr</B
>
command combines much of the functionality of
<B
CLASS="COMMAND"
>nl</B
>, <B
CLASS="COMMAND"
>paste</B
>,
<B
CLASS="COMMAND"
>fold</B
>, <B
CLASS="COMMAND"
>column</B
>, and
<B
CLASS="COMMAND"
>expand</B
>.</P
><P
><TT
CLASS="USERINPUT"
><B
>pr -o 5 --width=65 fileZZZ | more</B
></TT
>
gives a nice paginated listing to screen of
<TT
CLASS="FILENAME"
>fileZZZ</TT
> with margins set at 5 and
65.</P
><P
>A particularly useful option is <TT
CLASS="OPTION"
>-d</TT
>,
forcing double-spacing (same effect as <B
CLASS="COMMAND"
>sed
-G</B
>).</P
></DD
><DT
><A
NAME="GETTEXTREF"
></A
><B
CLASS="COMMAND"
>gettext</B
></DT
><DD
><P
>The GNU <B
CLASS="COMMAND"
>gettext</B
> package is a set of
utilities for <A
HREF="#LOCALIZATION"
>localizing</A
>
and translating the text output of programs into foreign
languages. While originally intended for C programs, it
now supports quite a number of programming and scripting
languages.</P
><P
>The <B
CLASS="COMMAND"
>gettext</B
>
<EM
>program</EM
> works on shell scripts. See
the <TT
CLASS="REPLACEABLE"
><I
>info page</I
></TT
>.</P
></DD
><DT
><A
NAME="MSGFMTREF"
></A
><B
CLASS="COMMAND"
>msgfmt</B
></DT
><DD
><P
>A program for generating binary
message catalogs. It is used for <A
HREF="#LOCALIZATION"
>localization</A
>.</P
></DD
><DT
><A
NAME="ICONVREF"
></A
><B
CLASS="COMMAND"
>iconv</B
></DT
><DD
><P
>A utility for converting file(s) to a different encoding
(character set). Its chief use is for <A
HREF="#LOCALIZATION"
>localization</A
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Convert a string from UTF-8 to UTF-16 and print to the BookList
function write_utf8_string {
STRING=$1
BOOKLIST=$2
echo -n "$STRING" | iconv -f UTF8 -t UTF16 | \
cut -b 3- | tr -d \\n &#62;&#62; "$BOOKLIST"
}
# From Peter Knowles' "booklistgen.sh" script
#+ for converting files to Sony Librie/PRS-50X format.
# (http://booklistgensh.peterknowles.com)</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="RECODEREF"
></A
><B
CLASS="COMMAND"
>recode</B
></DT
><DD
><P
>Consider this a fancier version of
<B
CLASS="COMMAND"
>iconv</B
>, above. This very versatile utility
for converting a file to a different encoding scheme.
Note that <I
CLASS="FIRSTTERM"
>recode</I
> is not part of the
standard Linux installation.</P
></DD
><DT
><A
NAME="TEXREF"
></A
><B
CLASS="COMMAND"
>TeX</B
>, <A
NAME="GSREF"
></A
><B
CLASS="COMMAND"
>gs</B
></DT
><DD
><P
><B
CLASS="COMMAND"
>TeX</B
> and <B
CLASS="COMMAND"
>Postscript</B
>
are text markup languages used for preparing copy for
printing or formatted video display.</P
><P
><B
CLASS="COMMAND"
>TeX</B
> is Donald Knuth's elaborate
typsetting system. It is often convenient to write a
shell script encapsulating all the options and arguments
passed to one of these markup languages.</P
><P
><I
CLASS="FIRSTTERM"
>Ghostscript</I
>
(<B
CLASS="COMMAND"
>gs</B
>) is a GPL-ed Postscript
interpreter.</P
></DD
><DT
><A
NAME="TEXEXECREF"
></A
><B
CLASS="COMMAND"
>texexec</B
></DT
><DD
><P
>Utility for processing <I
CLASS="FIRSTTERM"
>TeX</I
> and
<I
CLASS="FIRSTTERM"
>pdf</I
> files. Found in
<TT
CLASS="FILENAME"
>/usr/bin</TT
>
on many Linux distros, it is actually a <A
HREF="#SHWRAPPER"
>shell wrapper</A
> that
calls <A
HREF="#PERLREF"
>Perl</A
> to invoke
<I
CLASS="FIRSTTERM"
>Tex</I
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>texexec --pdfarrange --result=Concatenated.pdf *pdf
# Concatenates all the pdf files in the current working directory
#+ into the merged file, Concatenated.pdf . . .
# (The --pdfarrange option repaginates a pdf file. See also --pdfcombine.)
# The above command-line could be parameterized and put into a shell script.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="ENSCRIPTREF"
></A
><B
CLASS="COMMAND"
>enscript</B
></DT
><DD
><P
>Utility for converting plain text file to PostScript</P
><P
>For example, <B
CLASS="COMMAND"
>enscript filename.txt -p filename.ps</B
>
produces the PostScript output file
<TT
CLASS="FILENAME"
>filename.ps</TT
>.</P
></DD
><DT
><A
NAME="GROFFREF"
></A
><B
CLASS="COMMAND"
>groff</B
>, <A
NAME="TBLREF"
></A
><B
CLASS="COMMAND"
>tbl</B
>, <A
NAME="EQNREF"
></A
><B
CLASS="COMMAND"
>eqn</B
></DT
><DD
><P
>Yet another text markup and display formatting language
is <B
CLASS="COMMAND"
>groff</B
>. This is the enhanced GNU version
of the venerable UNIX <B
CLASS="COMMAND"
>roff/troff</B
> display
and typesetting package. <A
HREF="#MANREF"
>Manpages</A
>
use <B
CLASS="COMMAND"
>groff</B
>.</P
><P
>The <B
CLASS="COMMAND"
>tbl</B
> table processing utility
is considered part of <B
CLASS="COMMAND"
>groff</B
>, as its
function is to convert table markup into
<B
CLASS="COMMAND"
>groff</B
> commands.</P
><P
>The <B
CLASS="COMMAND"
>eqn</B
> equation processing utility
is likewise part of <B
CLASS="COMMAND"
>groff</B
>, and
its function is to convert equation markup into
<B
CLASS="COMMAND"
>groff</B
> commands.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="MANVIEW"
></A
><P
><B
>Example 16-29. <I
CLASS="FIRSTTERM"
>manview</I
>: Viewing formatted manpages</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# manview.sh: Formats the source of a man page for viewing.
# This script is useful when writing man page source.
# It lets you look at the intermediate results on the fly
#+ while working on it.
E_WRONGARGS=85
if [ -z "$1" ]
then
echo "Usage: `basename $0` filename"
exit $E_WRONGARGS
fi
# ---------------------------
groff -Tascii -man $1 | less
# From the man page for groff.
# ---------------------------
# If the man page includes tables and/or equations,
#+ then the above code will barf.
# The following line can handle such cases.
#
# gtbl &#60; "$1" | geqn -Tlatin1 | groff -Tlatin1 -mtty-char -man
#
# Thanks, S.C.
exit $? # See also the "maned.sh" script.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#MANED"
>Example A-39</A
>.</P
></DD
><DT
><A
NAME="LEXREF"
></A
><B
CLASS="COMMAND"
>lex</B
>, <A
NAME="YACCREF"
></A
><B
CLASS="COMMAND"
>yacc</B
></DT
><DD
><P
><A
NAME="FLEXREF"
></A
></P
><P
>The <B
CLASS="COMMAND"
>lex</B
> lexical analyzer produces
programs for pattern matching. This has been replaced
by the nonproprietary <B
CLASS="COMMAND"
>flex</B
> on Linux
systems.</P
><P
><A
NAME="BISONREF"
></A
></P
><P
>The <B
CLASS="COMMAND"
>yacc</B
> utility creates a
parser based on a set of specifications. This has been
replaced by the nonproprietary <B
CLASS="COMMAND"
>bison</B
>
on Linux systems.</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="FILEARCHIV"
></A
>16.5. File and Archiving Commands</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="FAARCHIVING1"
></A
>Archiving</B
></P
><DL
><DT
><A
NAME="TARREF"
></A
><B
CLASS="COMMAND"
>tar</B
></DT
><DD
><P
>The standard UNIX archiving utility.
<A
NAME="AEN11885"
HREF="#FTN.AEN11885"
><SPAN
CLASS="footnote"
>[75]</SPAN
></A
>
Originally a
<I
CLASS="WORDASWORD"
>Tape ARchiving</I
> program, it has
developed into a general purpose package that can handle
all manner of archiving with all types of destination
devices, ranging from tape drives to regular files to even
<TT
CLASS="FILENAME"
>stdout</TT
> (see <A
HREF="#EX58"
>Example 3-4</A
>). GNU
<I
CLASS="FIRSTTERM"
>tar</I
> has been patched to accept
various compression filters, for example: <B
CLASS="COMMAND"
>tar
czvf archive_name.tar.gz *</B
>, which recursively
archives and <A
HREF="#GZIPREF"
>gzips</A
>
all files in a directory tree except <A
HREF="#DOTFILESREF"
>dotfiles</A
> in the current
working directory (<A
HREF="#PWDREF"
>$PWD</A
>).
<A
NAME="AEN11896"
HREF="#FTN.AEN11896"
><SPAN
CLASS="footnote"
>[76]</SPAN
></A
>
</P
><P
>Some useful <B
CLASS="COMMAND"
>tar</B
> options:
<P
></P
><OL
TYPE="1"
><LI
><P
><TT
CLASS="OPTION"
>-c</TT
> create (a new
archive)</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-x</TT
> extract (files from
existing archive)</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>--delete</TT
> delete (files
from existing archive)</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This option will not work on magnetic tape
devices.</P
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><TT
CLASS="OPTION"
>-r</TT
> append (files to
existing archive)</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-A</TT
> append
(<I
CLASS="FIRSTTERM"
>tar</I
> files to
existing archive)</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-t</TT
> list (contents of
existing archive)</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-u</TT
> update archive</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-d</TT
> compare archive with
specified filesystem</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>--after-date</TT
> only process
files with a date stamp <EM
>after</EM
>
specified date</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-z</TT
> <A
HREF="#GZIPREF"
>gzip</A
> the archive</P
><P
>(compress or uncompress, depending on whether
combined with the <TT
CLASS="OPTION"
>-c</TT
> or
<TT
CLASS="OPTION"
>-x</TT
>) option</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-j</TT
>
<A
HREF="#BZIPREF"
>bzip2</A
> the
archive</P
></LI
></OL
>
</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>It may be difficult to recover data from a
corrupted <I
CLASS="FIRSTTERM"
>gzipped</I
> tar
archive. When archiving important files, make multiple
backups.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SHARREF"
></A
><B
CLASS="COMMAND"
>shar</B
></DT
><DD
><P
><I
CLASS="FIRSTTERM"
>Shell archiving</I
> utility.
The text and/or binary files in a shell archive are
concatenated without compression, and the resultant
archive is essentially a shell script, complete with
<SPAN
CLASS="TOKEN"
>#!/bin/sh</SPAN
> header, containing all the
necessary unarchiving commands, as well as the files
themselves. Unprintable binary characters in the target
file(s) are converted to printable ASCII characters in the
output <I
CLASS="FIRSTTERM"
>shar</I
> file. <I
CLASS="FIRSTTERM"
>Shar
archives</I
> still show up in Usenet newsgroups,
but otherwise <B
CLASS="COMMAND"
>shar</B
> has been replaced
by <B
CLASS="COMMAND"
>tar</B
>/<B
CLASS="COMMAND"
>gzip</B
>.
The <B
CLASS="COMMAND"
>unshar</B
> command unpacks
<I
CLASS="FIRSTTERM"
>shar</I
> archives.</P
><P
>The
<B
CLASS="COMMAND"
>mailshar</B
> command is a Bash script that
uses <B
CLASS="COMMAND"
>shar</B
> to concatenate multiple files
into a single one for e-mailing.
This script supports compression and <A
HREF="#UUENCODEREF"
>uuencoding</A
>.</P
></DD
><DT
><A
NAME="ARREF"
></A
><B
CLASS="COMMAND"
>ar</B
></DT
><DD
><P
>Creation and manipulation utility for archives, mainly
used for binary object file libraries.</P
></DD
><DT
><A
NAME="RPMREF"
></A
><B
CLASS="COMMAND"
>rpm</B
></DT
><DD
><P
>The <I
CLASS="FIRSTTERM"
>Red Hat Package Manager</I
>, or
<B
CLASS="COMMAND"
>rpm</B
> utility provides a wrapper for
source or binary archives. It includes commands for
installing and checking the integrity of packages, among
other things.</P
><P
>A simple <B
CLASS="COMMAND"
>rpm -i package_name.rpm</B
>
usually suffices to install a package, though there are many
more options available.</P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><TT
CLASS="USERINPUT"
><B
>rpm -qf</B
></TT
> identifies which package a
file originates from.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>rpm -qf /bin/ls</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>coreutils-5.2.1-31</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><TT
CLASS="USERINPUT"
><B
>rpm -qa</B
></TT
> gives a
complete list of all installed <I
CLASS="FIRSTTERM"
>rpm</I
> packages
on a given system. An <TT
CLASS="USERINPUT"
><B
>rpm -qa package_name</B
></TT
>
lists only the package(s) corresponding to
<TT
CLASS="FILENAME"
>package_name</TT
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>rpm -qa</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>redhat-logos-1.1.3-1
glibc-2.2.4-13
cracklib-2.7-12
dosfstools-2.7-1
gdbm-1.8.0-10
ksymoops-2.4.1-1
mktemp-1.5-11
perl-5.6.0-17
reiserfs-utils-3.x.0j-2
...</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>rpm -qa docbook-utils</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>docbook-utils-0.6.9-2</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>rpm -qa docbook | grep docbook</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>docbook-dtd31-sgml-1.0-10
docbook-style-dsssl-1.64-3
docbook-dtd30-sgml-1.0-10
docbook-dtd40-sgml-1.0-11
docbook-utils-pdf-0.6.9-2
docbook-dtd41-sgml-1.0-10
docbook-utils-0.6.9-2</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="CPIOREF"
></A
><B
CLASS="COMMAND"
>cpio</B
></DT
><DD
><P
>This specialized archiving copy command
(<B
CLASS="COMMAND"
>c</B
>o<B
CLASS="COMMAND"
>p</B
>y
<B
CLASS="COMMAND"
>i</B
>nput and <B
CLASS="COMMAND"
>o</B
>utput)
is rarely seen any more, having been supplanted by
<B
CLASS="COMMAND"
>tar</B
>/<B
CLASS="COMMAND"
>gzip</B
>. It still
has its uses, such as moving a directory tree. With an
appropriate block size (for copying) specified, it
can be appreciably faster than <B
CLASS="COMMAND"
>tar</B
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX48"
></A
><P
><B
>Example 16-30. Using <I
CLASS="FIRSTTERM"
>cpio</I
> to move a directory tree</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Copying a directory tree using cpio.
# Advantages of using 'cpio':
# Speed of copying. It's faster than 'tar' with pipes.
# Well suited for copying special files (named pipes, etc.)
#+ that 'cp' may choke on.
ARGS=2
E_BADARGS=65
if [ $# -ne "$ARGS" ]
then
echo "Usage: `basename $0` source destination"
exit $E_BADARGS
fi
source="$1"
destination="$2"
###################################################################
find "$source" -depth | cpio -admvp "$destination"
# ^^^^^ ^^^^^
# Read the 'find' and 'cpio' info pages to decipher these options.
# The above works only relative to $PWD (current directory) . . .
#+ full pathnames are specified.
###################################################################
# Exercise:
# --------
# Add code to check the exit status ($?) of the 'find | cpio' pipe
#+ and output appropriate error messages if anything went wrong.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="RPM2CPIOREF"
></A
><B
CLASS="COMMAND"
>rpm2cpio</B
></DT
><DD
><P
>This command extracts a
<B
CLASS="COMMAND"
>cpio</B
> archive from an <A
HREF="#RPMREF"
>rpm</A
> one.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="DERPM"
></A
><P
><B
>Example 16-31. Unpacking an <I
CLASS="FIRSTTERM"
>rpm</I
> archive</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# de-rpm.sh: Unpack an 'rpm' archive
: ${1?"Usage: `basename $0` target-file"}
# Must specify 'rpm' archive name as an argument.
TEMPFILE=$$.cpio # Tempfile with "unique" name.
# $$ is process ID of script.
rpm2cpio &#60; $1 &#62; $TEMPFILE # Converts rpm archive into
#+ cpio archive.
cpio --make-directories -F $TEMPFILE -i # Unpacks cpio archive.
rm -f $TEMPFILE # Deletes cpio archive.
exit 0
# Exercise:
# Add check for whether 1) "target-file" exists and
#+ 2) it is an rpm archive.
# Hint: Parse output of 'file' command.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="PAXREF"
></A
><B
CLASS="COMMAND"
>pax</B
></DT
><DD
><P
>The <I
CLASS="FIRSTTERM"
>pax</I
>
<B
CLASS="COMMAND"
>p</B
>ortable <B
CLASS="COMMAND"
>a</B
>rchive
e<B
CLASS="COMMAND"
>x</B
>change toolkit facilitates periodic
file backups and is designed to be cross-compatible
between various flavors of UNIX. It was designed
to replace <A
HREF="#TARREF"
>tar</A
> and <A
HREF="#CPIOREF"
>cpio</A
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>pax -wf daily_backup.pax ~/linux-server/files
# Creates a tar archive of all files in the target directory.
# Note that the options to pax must be in the correct order --
#+ pax -fw has an entirely different effect.
pax -f daily_backup.pax
# Lists the files in the archive.
pax -rf daily_backup.pax ~/bsd-server/files
# Restores the backed-up files from the Linux machine
#+ onto a BSD one.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Note that <I
CLASS="FIRSTTERM"
>pax</I
> handles many of
the standard archiving and compression commands.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="FACOMPRESSION1"
></A
>Compression</B
></P
><DL
><DT
><A
NAME="GZIPREF"
></A
><B
CLASS="COMMAND"
>gzip</B
></DT
><DD
><P
>The standard GNU/UNIX compression utility, replacing
the inferior and proprietary
<B
CLASS="COMMAND"
>compress</B
>. The corresponding decompression
command is <B
CLASS="COMMAND"
>gunzip</B
>, which is the equivalent of
<B
CLASS="COMMAND"
>gzip -d</B
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <TT
CLASS="OPTION"
>-c</TT
> option sends the output of
<B
CLASS="COMMAND"
>gzip</B
> to <TT
CLASS="FILENAME"
>stdout</TT
>. This
is useful when <A
HREF="#PIPEREF"
>piping</A
> to other
commands.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ZCATREF"
></A
></P
><P
>The <B
CLASS="COMMAND"
>zcat</B
> filter decompresses a
<I
CLASS="FIRSTTERM"
>gzipped</I
> file to
<TT
CLASS="FILENAME"
>stdout</TT
>, as possible input to a pipe or
redirection. This is, in effect, a <B
CLASS="COMMAND"
>cat</B
>
command that works on compressed files (including files
processed with the older <A
HREF="#COMPRESSREF"
>compress</A
>
utility). The <B
CLASS="COMMAND"
>zcat</B
> command is equivalent to
<B
CLASS="COMMAND"
>gzip -dc</B
>.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>On some commercial UNIX systems, <B
CLASS="COMMAND"
>zcat</B
>
is a synonym for <B
CLASS="COMMAND"
>uncompress -c</B
>,
and will not work on <I
CLASS="FIRSTTERM"
>gzipped</I
>
files.</P
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#EX14"
>Example 7-7</A
>.</P
></DD
><DT
><A
NAME="BZIPREF"
></A
><B
CLASS="COMMAND"
>bzip2</B
></DT
><DD
><P
>An alternate compression utility, usually more efficient
(but slower) than <B
CLASS="COMMAND"
>gzip</B
>, especially on
large files. The corresponding decompression command is
<B
CLASS="COMMAND"
>bunzip2</B
>.</P
><P
>Similar to the <B
CLASS="COMMAND"
>zcat</B
> command,
<B
CLASS="COMMAND"
>bzcat</B
> decompresses a
<I
CLASS="FIRSTTERM"
>bzipped2-ed</I
> file to
<TT
CLASS="FILENAME"
>stdout</TT
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Newer versions of <A
HREF="#TARREF"
>tar</A
> have been patched with
<B
CLASS="COMMAND"
>bzip2</B
> support.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="COMPRESSREF"
></A
><B
CLASS="COMMAND"
>compress</B
>, <A
NAME="UNCOMPRESSREF"
></A
><B
CLASS="COMMAND"
>uncompress</B
></DT
><DD
><P
>This is an older, proprietary compression
utility found in commercial UNIX distributions. The
more efficient <B
CLASS="COMMAND"
>gzip</B
> has largely
replaced it. Linux distributions generally include a
<B
CLASS="COMMAND"
>compress</B
> workalike for compatibility,
although <B
CLASS="COMMAND"
>gunzip</B
> can unarchive files
treated with <B
CLASS="COMMAND"
>compress</B
>.</P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>znew</B
> command transforms
<I
CLASS="FIRSTTERM"
>compressed</I
> files into
<I
CLASS="FIRSTTERM"
>gzipped</I
> ones.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SQREF"
></A
><B
CLASS="COMMAND"
>sq</B
></DT
><DD
><P
>Yet another compression (<B
CLASS="COMMAND"
>sq</B
>ueeze)
utility, a filter that works only on sorted
<A
HREF="#ASCIIDEF"
>ASCII</A
> word lists. It
uses the standard invocation syntax for a filter,
<B
CLASS="COMMAND"
>sq &#60; input-file &#62; output-file</B
>.
Fast, but not nearly as efficient as <A
HREF="#GZIPREF"
>gzip</A
>. The corresponding
uncompression filter is <B
CLASS="COMMAND"
>unsq</B
>, invoked
like <B
CLASS="COMMAND"
>sq</B
>.</P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The output of <B
CLASS="COMMAND"
>sq</B
> may be
piped to <B
CLASS="COMMAND"
>gzip</B
> for further
compression.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="ZIPREF"
></A
><B
CLASS="COMMAND"
>zip</B
>, <B
CLASS="COMMAND"
>unzip</B
></DT
><DD
><P
>Cross-platform file archiving and compression utility
compatible with DOS <I
CLASS="FIRSTTERM"
>pkzip.exe</I
>.
<SPAN
CLASS="QUOTE"
>"Zipped"</SPAN
> archives seem to be a more
common medium of file exchange on the Internet than
<SPAN
CLASS="QUOTE"
>"tarballs."</SPAN
></P
></DD
><DT
><A
NAME="UNARCREF"
></A
><B
CLASS="COMMAND"
>unarc</B
>, <B
CLASS="COMMAND"
>unarj</B
>, <B
CLASS="COMMAND"
>unrar</B
></DT
><DD
><P
>These Linux utilities permit unpacking archives
compressed with the DOS <I
CLASS="FIRSTTERM"
>arc.exe</I
>,
<I
CLASS="FIRSTTERM"
>arj.exe</I
>, and
<I
CLASS="FIRSTTERM"
>rar.exe</I
> programs.</P
></DD
><DT
><A
NAME="LZMAREF"
></A
><B
CLASS="COMMAND"
>lzma</B
>, <B
CLASS="COMMAND"
>unlzma</B
>, <B
CLASS="COMMAND"
>lzcat</B
></DT
><DD
><P
>Highly efficient Lempel-Ziv-Markov compression.
The syntax of <I
CLASS="FIRSTTERM"
>lzma</I
> is similar to
that of <I
CLASS="FIRSTTERM"
>gzip</I
>. The <A
HREF="http://www.7-zip.org/sdk.html"
TARGET="_top"
>7-zip Website</A
>
has more information.</P
></DD
><DT
><A
NAME="XZREF"
></A
><B
CLASS="COMMAND"
>xz</B
>, <B
CLASS="COMMAND"
>unxz</B
>, <B
CLASS="COMMAND"
>xzcat</B
></DT
><DD
><P
>A new high-efficiency compression tool, backward compatible
with <I
CLASS="FIRSTTERM"
>lzma</I
>, and with an invocation
syntax similar to <I
CLASS="FIRSTTERM"
>gzip</I
>. For
more information, see the <A
HREF="http://en.wikipedia.org/wiki/Xz"
TARGET="_top"
>Wikipedia
entry</A
>.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="FAINFORMATION1"
></A
>File Information</B
></P
><DL
><DT
><A
NAME="FILEREF"
></A
><B
CLASS="COMMAND"
>file</B
></DT
><DD
><P
>A utility for identifying file types. The command
<TT
CLASS="USERINPUT"
><B
>file file-name</B
></TT
> will return a
file specification for <TT
CLASS="FILENAME"
>file-name</TT
>,
such as <TT
CLASS="COMPUTEROUTPUT"
>ascii text</TT
> or
<TT
CLASS="COMPUTEROUTPUT"
>data</TT
>. It references
the <A
HREF="#MAGNUMREF"
>magic numbers</A
>
found in <TT
CLASS="FILENAME"
>/usr/share/magic</TT
>,
<TT
CLASS="FILENAME"
>/etc/magic</TT
>, or
<TT
CLASS="FILENAME"
>/usr/lib/magic</TT
>, depending on the
Linux/UNIX distribution.</P
><P
>The <TT
CLASS="OPTION"
>-f</TT
> option causes
<B
CLASS="COMMAND"
>file</B
> to run in <A
HREF="#BATCHPROCREF"
>batch</A
> mode, to read from
a designated file a list of filenames to analyze. The
<TT
CLASS="OPTION"
>-z</TT
> option, when used on a compressed
target file, forces an attempt to analyze the uncompressed
file type.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>file test.tar.gz</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>test.tar.gz: gzip compressed data, deflated,
last modified: Sun Sep 16 13:34:51 2001, os: Unix</TT
>
<TT
CLASS="PROMPT"
>bash </TT
><TT
CLASS="USERINPUT"
><B
>file -z test.tar.gz</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>test.tar.gz: GNU tar archive (gzip compressed data, deflated,
last modified: Sun Sep 16 13:34:51 2001, os: Unix)</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Find sh and Bash scripts in a given directory:
DIRECTORY=/usr/local/bin
KEYWORD=Bourne
# Bourne and Bourne-Again shell scripts
file $DIRECTORY/* | fgrep $KEYWORD
# Output:
# /usr/local/bin/burn-cd: Bourne-Again shell script text executable
# /usr/local/bin/burnit: Bourne-Again shell script text executable
# /usr/local/bin/cassette.sh: Bourne shell script text executable
# /usr/local/bin/copy-cd: Bourne-Again shell script text executable
# . . .</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="STRIPC"
></A
><P
><B
>Example 16-32. Stripping comments from C program files</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# strip-comment.sh: Strips out the comments (/* COMMENT */) in a C program.
E_NOARGS=0
E_ARGERROR=66
E_WRONG_FILE_TYPE=67
if [ $# -eq "$E_NOARGS" ]
then
echo "Usage: `basename $0` C-program-file" &#62;&#38;2 # Error message to stderr.
exit $E_ARGERROR
fi
# Test for correct file type.
type=`file $1 | awk '{ print $2, $3, $4, $5 }'`
# "file $1" echoes file type . . .
# Then awk removes the first field, the filename . . .
# Then the result is fed into the variable "type."
correct_type="ASCII C program text"
if [ "$type" != "$correct_type" ]
then
echo
echo "This script works on C program files only."
echo
exit $E_WRONG_FILE_TYPE
fi
# Rather cryptic sed script:
#--------
sed '
/^\/\*/d
/.*\*\//d
' $1
#--------
# Easy to understand if you take several hours to learn sed fundamentals.
# Need to add one more line to the sed script to deal with
#+ case where line of code has a comment following it on same line.
# This is left as a non-trivial exercise.
# Also, the above code deletes non-comment lines with a "*/" . . .
#+ not a desirable result.
exit 0
# ----------------------------------------------------------------
# Code below this line will not execute because of 'exit 0' above.
# Stephane Chazelas suggests the following alternative:
usage() {
echo "Usage: `basename $0` C-program-file" &#62;&#38;2
exit 1
}
WEIRD=`echo -n -e '\377'` # or WEIRD=$'\377'
[[ $# -eq 1 ]] || usage
case `file "$1"` in
*"C program text"*) sed -e "s%/\*%${WEIRD}%g;s%\*/%${WEIRD}%g" "$1" \
| tr '\377\n' '\n\377' \
| sed -ne 'p;n' \
| tr -d '\n' | tr '\377' '\n';;
*) usage;;
esac
# This is still fooled by things like:
# printf("/*");
# or
# /* /* buggy embedded comment */
#
# To handle all special cases (comments in strings, comments in string
#+ where there is a \", \\" ...),
#+ the only way is to write a C parser (using lex or yacc perhaps?).
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="WHICHREF"
></A
><B
CLASS="COMMAND"
>which</B
></DT
><DD
><P
><B
CLASS="COMMAND"
>which command</B
> gives the full path
to <SPAN
CLASS="QUOTE"
>"command."</SPAN
> This is useful for finding
out whether a particular command or utility is installed
on the system.</P
><P
><TT
CLASS="USERINPUT"
><B
>$bash which rm</B
></TT
>
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="COMPUTEROUTPUT"
>/usr/bin/rm</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>For an interesting use of this command, see <A
HREF="#HORSERACE"
>Example 36-16</A
>.</P
></DD
><DT
><A
NAME="WHEREISREF"
></A
><B
CLASS="COMMAND"
>whereis</B
></DT
><DD
><P
>Similar to <B
CLASS="COMMAND"
>which</B
>, above,
<B
CLASS="COMMAND"
>whereis command</B
> gives the
full path to <SPAN
CLASS="QUOTE"
>"command,"</SPAN
> but also to its
<A
HREF="#MANREF"
>manpage</A
>.</P
><P
><TT
CLASS="USERINPUT"
><B
>$bash whereis rm</B
></TT
>
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="COMPUTEROUTPUT"
>rm: /bin/rm /usr/share/man/man1/rm.1.bz2</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="WHATISREF"
></A
><B
CLASS="COMMAND"
>whatis</B
></DT
><DD
><P
><B
CLASS="COMMAND"
>whatis command</B
> looks up
<SPAN
CLASS="QUOTE"
>"command"</SPAN
> in the
<TT
CLASS="REPLACEABLE"
><I
>whatis</I
></TT
> database. This is useful
for identifying system commands and important configuration
files. Consider it a simplified <B
CLASS="COMMAND"
>man</B
>
command.</P
><P
><TT
CLASS="USERINPUT"
><B
>$bash whatis whatis</B
></TT
>
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="COMPUTEROUTPUT"
>whatis (1) - search the whatis database for complete words</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="WHAT"
></A
><P
><B
>Example 16-33. Exploring <TT
CLASS="FILENAME"
>/usr/X11R6/bin</TT
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# What are all those mysterious binaries in /usr/X11R6/bin?
DIRECTORY="/usr/X11R6/bin"
# Try also "/bin", "/usr/bin", "/usr/local/bin", etc.
for file in $DIRECTORY/*
do
whatis `basename $file` # Echoes info about the binary.
done
exit 0
# Note: For this to work, you must create a "whatis" database
#+ with /usr/sbin/makewhatis.
# You may wish to redirect output of this script, like so:
# ./what.sh &#62;&#62;whatis.db
# or view it a page at a time on stdout,
# ./what.sh | less</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#FILEINFO"
>Example 11-3</A
>.</P
></DD
><DT
><A
NAME="VDIRREF"
></A
><B
CLASS="COMMAND"
>vdir</B
></DT
><DD
><P
>Show a detailed directory listing. The effect is similar to
<A
HREF="#LSREF"
>ls -lb</A
>.</P
><P
>This is one of the GNU
<I
CLASS="FIRSTTERM"
>fileutils</I
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>vdir</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 10
-rw-r--r-- 1 bozo bozo 4034 Jul 18 22:04 data1.xrolo
-rw-r--r-- 1 bozo bozo 4602 May 25 13:58 data1.xrolo.bak
-rw-r--r-- 1 bozo bozo 877 Dec 17 2000 employment.xrolo</TT
>
<TT
CLASS="PROMPT"
>bash </TT
><TT
CLASS="USERINPUT"
><B
>ls -l</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 10
-rw-r--r-- 1 bozo bozo 4034 Jul 18 22:04 data1.xrolo
-rw-r--r-- 1 bozo bozo 4602 May 25 13:58 data1.xrolo.bak
-rw-r--r-- 1 bozo bozo 877 Dec 17 2000 employment.xrolo</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="LOCATEREF"
></A
><B
CLASS="COMMAND"
>locate</B
>, <A
NAME="SLOCATEREF"
></A
><B
CLASS="COMMAND"
>slocate</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>locate</B
> command searches for
files using a database stored for just that purpose. The
<B
CLASS="COMMAND"
>slocate</B
> command is the secure version of
<B
CLASS="COMMAND"
>locate</B
> (which may be aliased to
<B
CLASS="COMMAND"
>slocate</B
>).</P
><P
><TT
CLASS="USERINPUT"
><B
>$bash locate hickson</B
></TT
>
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="COMPUTEROUTPUT"
>/usr/lib/xephem/catalogs/hickson.edb</TT
></PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="GETFACLREF"
></A
><B
CLASS="COMMAND"
>getfacl</B
>, <A
NAME="SETFACLREF"
></A
><B
CLASS="COMMAND"
>setfacl</B
></DT
><DD
><P
>These commands <I
CLASS="FIRSTTERM"
>retrieve</I
> or
<I
CLASS="FIRSTTERM"
>set</I
> the <B
CLASS="COMMAND"
>f</B
>ile
<B
CLASS="COMMAND"
>a</B
>ccess <B
CLASS="COMMAND"
>c</B
>ontrol
<B
CLASS="COMMAND"
>l</B
>ist -- the <I
CLASS="FIRSTTERM"
>owner</I
>,
<I
CLASS="FIRSTTERM"
>group</I
>, and file permissions.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>getfacl *</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
># file: test1.txt
# owner: bozo
# group: bozgrp
user::rw-
group::rw-
other::r--
# file: test2.txt
# owner: bozo
# group: bozgrp
user::rw-
group::rw-
other::r--</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>setfacl -m u:bozo:rw yearly_budget.csv</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>getfacl yearly_budget.csv</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
># file: yearly_budget.csv
# owner: accountant
# group: budgetgrp
user::rw-
user:bozo:rw-
user:accountant:rw-
group::rw-
mask::rw-
other::r--</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="READLINKREF"
></A
><B
CLASS="COMMAND"
>readlink</B
></DT
><DD
><P
>Disclose the file that a symbolic link points to.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>readlink /usr/bin/awk</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>../../bin/gawk</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="STRINGSREF"
></A
><B
CLASS="COMMAND"
>strings</B
></DT
><DD
><P
>Use the <B
CLASS="COMMAND"
>strings</B
> command to find
printable strings in a binary or data file. It will list
sequences of printable characters found in the target
file. This might be handy for a quick 'n dirty examination
of a core dump or for looking at an unknown graphic image
file (<TT
CLASS="USERINPUT"
><B
>strings image-file | more</B
></TT
> might
show something like <I
CLASS="FIRSTTERM"
>JFIF</I
>,
which would identify the file as a <I
CLASS="FIRSTTERM"
>jpeg</I
>
graphic). In a script, you would probably
parse the output of <B
CLASS="COMMAND"
>strings</B
>
with <A
HREF="#GREPREF"
>grep</A
> or <A
HREF="#SEDREF"
>sed</A
>. See <A
HREF="#BINGREP"
>Example 11-8</A
>
and <A
HREF="#FINDSTRING"
>Example 11-10</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="WSTRINGS"
></A
><P
><B
>Example 16-34. An <SPAN
CLASS="QUOTE"
>"improved"</SPAN
>
<I
CLASS="FIRSTTERM"
>strings</I
> command</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# wstrings.sh: "word-strings" (enhanced "strings" command)
#
# This script filters the output of "strings" by checking it
#+ against a standard word list file.
# This effectively eliminates gibberish and noise,
#+ and outputs only recognized words.
# ===========================================================
# Standard Check for Script Argument(s)
ARGS=1
E_BADARGS=85
E_NOFILE=86
if [ $# -ne $ARGS ]
then
echo "Usage: `basename $0` filename"
exit $E_BADARGS
fi
if [ ! -f "$1" ] # Check if file exists.
then
echo "File \"$1\" does not exist."
exit $E_NOFILE
fi
# ===========================================================
MINSTRLEN=3 # Minimum string length.
WORDFILE=/usr/share/dict/linux.words # Dictionary file.
# May specify a different word list file
#+ of one-word-per-line format.
# For example, the "yawl" word-list package,
# http://bash.deta.in/yawl-0.3.2.tar.gz
wlist=`strings "$1" | tr A-Z a-z | tr '[:space:]' Z | \
tr -cs '[:alpha:]' Z | tr -s '\173-\377' Z | tr Z ' '`
# Translate output of 'strings' command with multiple passes of 'tr'.
# "tr A-Z a-z" converts to lowercase.
# "tr '[:space:]'" converts whitespace characters to Z's.
# "tr -cs '[:alpha:]' Z" converts non-alphabetic characters to Z's,
#+ and squeezes multiple consecutive Z's.
# "tr -s '\173-\377' Z" converts all characters past 'z' to Z's
#+ and squeezes multiple consecutive Z's,
#+ which gets rid of all the weird characters that the previous
#+ translation failed to deal with.
# Finally, "tr Z ' '" converts all those Z's to whitespace,
#+ which will be seen as word separators in the loop below.
# ***********************************************************************
# Note the technique of feeding/piping the output of 'tr' back to itself,
#+ but with different arguments and/or options on each successive pass.
# ***********************************************************************
for word in $wlist # Important:
# $wlist must not be quoted here.
# "$wlist" does not work.
# Why not?
do
strlen=${#word} # String length.
if [ "$strlen" -lt "$MINSTRLEN" ] # Skip over short strings.
then
continue
fi
grep -Fw $word "$WORDFILE" # Match whole words only.
# ^^^ # "Fixed strings" and
#+ "whole words" options.
done
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="COMPARISONN1"
></A
>Comparison</B
></P
><DL
><DT
><A
NAME="DIFFREF"
></A
><B
CLASS="COMMAND"
>diff</B
>, <B
CLASS="COMMAND"
>patch</B
></DT
><DD
><P
><B
CLASS="COMMAND"
>diff</B
>: flexible file comparison
utility. It compares the target files line-by-line
sequentially. In some applications, such as comparing
word dictionaries, it may be helpful to filter the
files through <A
HREF="#SORTREF"
>sort</A
>
and <B
CLASS="COMMAND"
>uniq</B
> before piping them
to <B
CLASS="COMMAND"
>diff</B
>. <TT
CLASS="USERINPUT"
><B
>diff file-1
file-2</B
></TT
> outputs the lines in the files that
differ, with carets showing which file each particular
line belongs to.</P
><P
>The <TT
CLASS="OPTION"
>--side-by-side</TT
> option to
<B
CLASS="COMMAND"
>diff</B
> outputs each compared file, line by
line, in separate columns, with non-matching lines marked. The
<TT
CLASS="OPTION"
>-c</TT
> and <TT
CLASS="OPTION"
>-u</TT
> options likewise
make the output of the command easier to interpret.</P
><P
>There are available various fancy frontends for
<B
CLASS="COMMAND"
>diff</B
>, such as <B
CLASS="COMMAND"
>sdiff</B
>,
<B
CLASS="COMMAND"
>wdiff</B
>, <B
CLASS="COMMAND"
>xdiff</B
>, and
<B
CLASS="COMMAND"
>mgdiff</B
>. </P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="DIFFERR2"
></A
>The <B
CLASS="COMMAND"
>diff</B
>
command returns an exit status of <SPAN
CLASS="ERRORCODE"
>0</SPAN
>
if the compared files are identical, and
<SPAN
CLASS="ERRORCODE"
>1</SPAN
> if they differ (or
<SPAN
CLASS="ERRORCODE"
>2</SPAN
> when <I
CLASS="FIRSTTERM"
>binary</I
>
files are being compared). This permits use of
<B
CLASS="COMMAND"
>diff</B
> in a test construct within a shell
script (see below).</P
></TD
></TR
></TABLE
></DIV
><P
>A common use for <B
CLASS="COMMAND"
>diff</B
> is generating
difference files to be used with <B
CLASS="COMMAND"
>patch</B
>
The <TT
CLASS="OPTION"
>-e</TT
> option outputs files suitable
for <B
CLASS="COMMAND"
>ed</B
> or <B
CLASS="COMMAND"
>ex</B
>
scripts.</P
><P
><A
NAME="PATCHREF"
></A
></P
><P
><B
CLASS="COMMAND"
>patch</B
>: flexible versioning
utility. Given a difference file generated by
<B
CLASS="COMMAND"
>diff</B
>, <B
CLASS="COMMAND"
>patch</B
> can
upgrade a previous version of a package to a newer version.
It is much more convenient to distribute a relatively
small <SPAN
CLASS="QUOTE"
>"diff"</SPAN
> file than the entire body of a
newly revised package. Kernel <SPAN
CLASS="QUOTE"
>"patches"</SPAN
> have
become the preferred method of distributing the frequent
releases of the Linux kernel.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>patch -p1 &#60;patch-file
# Takes all the changes listed in 'patch-file'
# and applies them to the files referenced therein.
# This upgrades to a newer version of the package.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Patching the kernel:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cd /usr/src
gzip -cd patchXX.gz | patch -p0
# Upgrading kernel source using 'patch'.
# From the Linux kernel docs "README",
# by anonymous author (Alan Cox?).</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>diff</B
> command can also
recursively compare directories (for the filenames
present).</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>diff -r ~/notes1 ~/notes2</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Only in /home/bozo/notes1: file02
Only in /home/bozo/notes1: file03
Only in /home/bozo/notes2: file04</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="ZDIFFREF"
></A
></P
><P
>Use <B
CLASS="COMMAND"
>zdiff</B
> to compare
<I
CLASS="FIRSTTERM"
>gzipped</I
> files.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="DIFFSTATREF"
></A
></P
><P
>Use <B
CLASS="COMMAND"
>diffstat</B
> to create
a histogram (point-distribution graph) of output from
<B
CLASS="COMMAND"
>diff</B
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="DIFF3REF"
></A
><B
CLASS="COMMAND"
>diff3</B
>, <B
CLASS="COMMAND"
>merge</B
></DT
><DD
><P
>An extended version of <B
CLASS="COMMAND"
>diff</B
> that compares
three files at a time. This command returns an exit value
of 0 upon successful execution, but unfortunately this gives
no information about the results of the comparison.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>diff3 file-1 file-2 file-3</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>====
1:1c
This is line 1 of "file-1".
2:1c
This is line 1 of "file-2".
3:1c
This is line 1 of "file-3"</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="MERGEREF"
></A
>The <B
CLASS="COMMAND"
>merge</B
>
(3-way file merge) command is an interesting adjunct to
<I
CLASS="FIRSTTERM"
>diff3</I
>. Its syntax is
<TT
CLASS="USERINPUT"
><B
>merge Mergefile file1 file2</B
></TT
>.
The result is to output to <TT
CLASS="FILENAME"
>Mergefile</TT
>
the changes that lead from <TT
CLASS="FILENAME"
>file1</TT
>
to <TT
CLASS="FILENAME"
>file2</TT
>. Consider this command
a stripped-down version of <I
CLASS="FIRSTTERM"
>patch</I
>.</P
></DD
><DT
><A
NAME="SDIFFREF"
></A
><B
CLASS="COMMAND"
>sdiff</B
></DT
><DD
><P
>Compare and/or edit two files in order to merge
them into an output file. Because of its interactive nature,
this command would find little use in a script.</P
></DD
><DT
><A
NAME="CMPREF"
></A
><B
CLASS="COMMAND"
>cmp</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>cmp</B
> command is a simpler version of
<B
CLASS="COMMAND"
>diff</B
>, above. Whereas <B
CLASS="COMMAND"
>diff</B
>
reports the differences between two files,
<B
CLASS="COMMAND"
>cmp</B
> merely shows at what point they
differ.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Like <B
CLASS="COMMAND"
>diff</B
>, <B
CLASS="COMMAND"
>cmp</B
>
returns an exit status of 0 if the compared files are
identical, and 1 if they differ. This permits use in a test
construct within a shell script.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="FILECOMP"
></A
><P
><B
>Example 16-35. Using <I
CLASS="FIRSTTERM"
>cmp</I
> to compare two files
within a script.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# file-comparison.sh
ARGS=2 # Two args to script expected.
E_BADARGS=85
E_UNREADABLE=86
if [ $# -ne "$ARGS" ]
then
echo "Usage: `basename $0` file1 file2"
exit $E_BADARGS
fi
if [[ ! -r "$1" || ! -r "$2" ]]
then
echo "Both files to be compared must exist and be readable."
exit $E_UNREADABLE
fi
cmp $1 $2 &#38;&#62; /dev/null
# Redirection to /dev/null buries the output of the "cmp" command.
# cmp -s $1 $2 has same result ("-s" silent flag to "cmp")
# Thank you Anders Gustavsson for pointing this out.
#
# Also works with 'diff', i.e.,
#+ diff $1 $2 &#38;&#62; /dev/null
if [ $? -eq 0 ] # Test exit status of "cmp" command.
then
echo "File \"$1\" is identical to file \"$2\"."
else
echo "File \"$1\" differs from file \"$2\"."
fi
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Use <B
CLASS="COMMAND"
>zcmp</B
> on
<I
CLASS="FIRSTTERM"
>gzipped</I
> files.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="COMMREF"
></A
><B
CLASS="COMMAND"
>comm</B
></DT
><DD
><P
>Versatile file comparison utility. The files must be
sorted for this to be useful.</P
><P
><B
CLASS="COMMAND"
>comm
<TT
CLASS="REPLACEABLE"
><I
>-options</I
></TT
>
<TT
CLASS="REPLACEABLE"
><I
>first-file</I
></TT
>
<TT
CLASS="REPLACEABLE"
><I
>second-file</I
></TT
></B
></P
><P
><TT
CLASS="USERINPUT"
><B
>comm file-1 file-2</B
></TT
> outputs three columns:
<P
></P
><UL
><LI
><P
>column 1 = lines unique to <TT
CLASS="FILENAME"
>file-1</TT
></P
></LI
><LI
><P
>column 2 = lines unique to <TT
CLASS="FILENAME"
>file-2</TT
></P
></LI
><LI
><P
>column 3 = lines common to both.</P
></LI
></UL
></P
><P
>The options allow suppressing output of one or more columns.
<P
></P
><UL
><LI
><P
><TT
CLASS="OPTION"
>-1</TT
> suppresses column
<TT
CLASS="LITERAL"
>1</TT
></P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-2</TT
> suppresses column
<TT
CLASS="LITERAL"
>2</TT
></P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-3</TT
> suppresses column
<TT
CLASS="LITERAL"
>3</TT
></P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-12</TT
> suppresses both columns
<TT
CLASS="LITERAL"
>1</TT
> and <TT
CLASS="LITERAL"
>2</TT
>, etc.</P
></LI
></UL
>
</P
><P
>This command is useful for comparing
<SPAN
CLASS="QUOTE"
>"dictionaries"</SPAN
> or <I
CLASS="FIRSTTERM"
>word
lists</I
> -- sorted text files with one word per
line.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="FAUTILS1"
></A
>Utilities</B
></P
><DL
><DT
><A
NAME="BASENAMEREF"
></A
><B
CLASS="COMMAND"
>basename</B
></DT
><DD
><P
>Strips the path information from a file name, printing
only the file name. The construction <TT
CLASS="USERINPUT"
><B
>basename
$0</B
></TT
> lets the script know its name, that is, the name it
was invoked by. This can be used for <SPAN
CLASS="QUOTE"
>"usage"</SPAN
> messages if,
for example a script is called with missing arguments:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo "Usage: `basename $0` arg1 arg2 ... argn"</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="DIRNAMEREF"
></A
><B
CLASS="COMMAND"
>dirname</B
></DT
><DD
><P
>Strips the <B
CLASS="COMMAND"
>basename</B
> from
a filename, printing only the path information.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><B
CLASS="COMMAND"
>basename</B
> and <B
CLASS="COMMAND"
>dirname</B
>
can operate on any arbitrary string. The argument
does not need to refer to an existing file, or
even be a filename for that matter (see <A
HREF="#DAYSBETWEEN"
>Example A-7</A
>).</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX35"
></A
><P
><B
>Example 16-36. <I
CLASS="FIRSTTERM"
>basename</I
> and
<I
CLASS="FIRSTTERM"
>dirname</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
address=/home/bozo/daily-journal.txt
echo "Basename of /home/bozo/daily-journal.txt = `basename $address`"
echo "Dirname of /home/bozo/daily-journal.txt = `dirname $address`"
echo
echo "My own home is `basename ~/`." # `basename ~` also works.
echo "The home of my home is `dirname ~/`." # `dirname ~` also works.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SPLITREF"
></A
><B
CLASS="COMMAND"
>split</B
>, <A
NAME="CSPLITREF"
></A
><B
CLASS="COMMAND"
>csplit</B
></DT
><DD
><P
>These are utilities for splitting a file into smaller
chunks. Their usual use is for splitting up large files
in order to back them up on floppies or preparatory to
e-mailing or uploading them.</P
><P
>The <B
CLASS="COMMAND"
>csplit</B
> command splits a file
according to <I
CLASS="FIRSTTERM"
>context</I
>, the split occuring
where patterns are matched.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SPLITCOPY"
></A
><P
><B
>Example 16-37. A script that copies itself in sections</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# splitcopy.sh
# A script that splits itself into chunks,
#+ then reassembles the chunks into an exact copy
#+ of the original script.
CHUNKSIZE=4 # Size of first chunk of split files.
OUTPREFIX=xx # csplit prefixes, by default,
#+ files with "xx" ...
csplit "$0" "$CHUNKSIZE"
# Some comment lines for padding . . .
# Line 15
# Line 16
# Line 17
# Line 18
# Line 19
# Line 20
cat "$OUTPREFIX"* &#62; "$0.copy" # Concatenate the chunks.
rm "$OUTPREFIX"* # Get rid of the chunks.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="FAENCENCR1"
></A
>Encoding and Encryption</B
></P
><DL
><DT
><A
NAME="SUMREF"
></A
><B
CLASS="COMMAND"
>sum</B
>, <A
NAME="CKSUMREF"
></A
><B
CLASS="COMMAND"
>cksum</B
>, <A
NAME="MD5SUMREF"
></A
><B
CLASS="COMMAND"
>md5sum</B
>, <A
NAME="SHA1SUMREF"
></A
><B
CLASS="COMMAND"
>sha1sum</B
></DT
><DD
><P
><A
NAME="CHECKSUMREF"
></A
>These are utilities for
generating <I
CLASS="FIRSTTERM"
>checksums</I
>. A
<I
CLASS="FIRSTTERM"
>checksum</I
> is a number
<A
NAME="AEN12840"
HREF="#FTN.AEN12840"
><SPAN
CLASS="footnote"
>[77]</SPAN
></A
>
mathematically calculated from the contents of a file,
for the purpose of checking its integrity. A script might
refer to a list of checksums for security purposes, such
as ensuring that the contents of key system files have not
been altered or corrupted. For security applications, use
the <B
CLASS="COMMAND"
>md5sum</B
> (<B
CLASS="COMMAND"
>m</B
>essage
<B
CLASS="COMMAND"
>d</B
>igest <B
CLASS="COMMAND"
>5</B
>
check<B
CLASS="COMMAND"
>sum</B
>) command, or better yet, the
newer <B
CLASS="COMMAND"
>sha1sum</B
> (Secure Hash Algorithm).
<A
NAME="AEN12849"
HREF="#FTN.AEN12849"
><SPAN
CLASS="footnote"
>[78]</SPAN
></A
>
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cksum /boot/vmlinuz</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>1670054224 804083 /boot/vmlinuz</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo -n "Top Secret" | cksum</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>3391003827 10</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>md5sum /boot/vmlinuz</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>0f43eccea8f09e0a0b2b5cf1dcf333ba /boot/vmlinuz</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo -n "Top Secret" | md5sum</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>8babc97a6f62a4649716f4df8d61728f -</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>cksum</B
> command shows the size,
in bytes, of its target, whether file or
<TT
CLASS="FILENAME"
>stdout</TT
>.</P
><P
>The <B
CLASS="COMMAND"
>md5sum</B
> and
<B
CLASS="COMMAND"
>sha1sum</B
> commands display a
<A
HREF="#DASHREF2"
>dash</A
> when they receive their input from
<TT
CLASS="FILENAME"
>stdout</TT
>.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="FILEINTEGRITY"
></A
><P
><B
>Example 16-38. Checking file integrity</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# file-integrity.sh: Checking whether files in a given directory
# have been tampered with.
E_DIR_NOMATCH=80
E_BAD_DBFILE=81
dbfile=File_record.md5
# Filename for storing records (database file).
set_up_database ()
{
echo ""$directory"" &#62; "$dbfile"
# Write directory name to first line of file.
md5sum "$directory"/* &#62;&#62; "$dbfile"
# Append md5 checksums and filenames.
}
check_database ()
{
local n=0
local filename
local checksum
# ------------------------------------------- #
# This file check should be unnecessary,
#+ but better safe than sorry.
if [ ! -r "$dbfile" ]
then
echo "Unable to read checksum database file!"
exit $E_BAD_DBFILE
fi
# ------------------------------------------- #
while read record[n]
do
directory_checked="${record[0]}"
if [ "$directory_checked" != "$directory" ]
then
echo "Directories do not match up!"
# Tried to use file for a different directory.
exit $E_DIR_NOMATCH
fi
if [ "$n" -gt 0 ] # Not directory name.
then
filename[n]=$( echo ${record[$n]} | awk '{ print $2 }' )
# md5sum writes records backwards,
#+ checksum first, then filename.
checksum[n]=$( md5sum "${filename[n]}" )
if [ "${record[n]}" = "${checksum[n]}" ]
then
echo "${filename[n]} unchanged."
elif [ "`basename ${filename[n]}`" != "$dbfile" ]
# Skip over checksum database file,
#+ as it will change with each invocation of script.
# ---
# This unfortunately means that when running
#+ this script on $PWD, tampering with the
#+ checksum database file will not be detected.
# Exercise: Fix this.
then
echo "${filename[n]} : CHECKSUM ERROR!"
# File has been changed since last checked.
fi
fi
let "n+=1"
done &#60;"$dbfile" # Read from checksum database file.
}
# =================================================== #
# main ()
if [ -z "$1" ]
then
directory="$PWD" # If not specified,
else #+ use current working directory.
directory="$1"
fi
clear # Clear screen.
echo " Running file integrity check on $directory"
echo
# ------------------------------------------------------------------ #
if [ ! -r "$dbfile" ] # Need to create database file?
then
echo "Setting up database file, \""$directory"/"$dbfile"\"."; echo
set_up_database
fi
# ------------------------------------------------------------------ #
check_database # Do the actual work.
echo
# You may wish to redirect the stdout of this script to a file,
#+ especially if the directory checked has many files in it.
exit 0
# For a much more thorough file integrity check,
#+ consider the "Tripwire" package,
#+ http://sourceforge.net/projects/tripwire/.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Also see <A
HREF="#DIRECTORYINFO"
>Example A-19</A
>, <A
HREF="#HORSERACE"
>Example 36-16</A
>, and <A
HREF="#RANDSTRING"
>Example 10-2</A
> for
creative uses of the <B
CLASS="COMMAND"
>md5sum</B
> command.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
> There have been reports that the 128-bit
<B
CLASS="COMMAND"
>md5sum</B
> can be cracked, so the more secure
160-bit <B
CLASS="COMMAND"
>sha1sum</B
> is a welcome new addition
to the checksum toolkit.
</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>md5sum testfile</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>e181e2c8720c60522c4c4c981108e367 testfile</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>sha1sum testfile</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>5d7425a9c08a66c3177f1e31286fa40986ffc996 testfile</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
></TD
></TR
></TABLE
></DIV
><P
>Security consultants have demonstrated that even
<B
CLASS="COMMAND"
>sha1sum</B
> can be compromised. Fortunately,
newer Linux distros include longer bit-length
<B
CLASS="COMMAND"
>sha224sum</B
>,
<B
CLASS="COMMAND"
>sha256sum</B
>,
<B
CLASS="COMMAND"
>sha384sum</B
>, and
<B
CLASS="COMMAND"
>sha512sum</B
> commands.</P
></DD
><DT
><A
NAME="UUENCODEREF"
></A
><B
CLASS="COMMAND"
>uuencode</B
></DT
><DD
><P
>This utility encodes binary files (images, sound files,
compressed files, etc.) into <A
HREF="#ASCIIDEF"
>ASCII</A
> characters, making
them suitable for transmission in the body of an
e-mail message or in a newsgroup posting. This is
especially useful where MIME (multimedia) encoding
is not available.</P
></DD
><DT
><A
NAME="UUDECODEREF"
></A
><B
CLASS="COMMAND"
>uudecode</B
></DT
><DD
><P
>This reverses the encoding, decoding
<I
CLASS="FIRSTTERM"
>uuencoded</I
> files back into the
original binaries.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX52"
></A
><P
><B
>Example 16-39. Uudecoding encoded files</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Uudecodes all uuencoded files in current working directory.
lines=35 # Allow 35 lines for the header (very generous).
for File in * # Test all the files in $PWD.
do
search1=`head -n $lines $File | grep begin | wc -w`
search2=`tail -n $lines $File | grep end | wc -w`
# Uuencoded files have a "begin" near the beginning,
#+ and an "end" near the end.
if [ "$search1" -gt 0 ]
then
if [ "$search2" -gt 0 ]
then
echo "uudecoding - $File -"
uudecode $File
fi
fi
done
# Note that running this script upon itself fools it
#+ into thinking it is a uuencoded file,
#+ because it contains both "begin" and "end".
# Exercise:
# --------
# Modify this script to check each file for a newsgroup header,
#+ and skip to next if not found.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <A
HREF="#FOLDREF"
>fold -s</A
> command
may be useful (possibly in a pipe) to process long uudecoded
text messages downloaded from Usenet newsgroups.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="MIMENCODEREF"
></A
><B
CLASS="COMMAND"
>mimencode</B
>, <A
NAME="MMENCODEREF"
></A
><B
CLASS="COMMAND"
>mmencode</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>mimencode</B
> and
<B
CLASS="COMMAND"
>mmencode</B
> commands process
multimedia-encoded e-mail attachments. Although
<I
CLASS="FIRSTTERM"
>mail user agents</I
> (such as
<I
CLASS="FIRSTTERM"
>pine</I
> or <I
CLASS="FIRSTTERM"
>kmail</I
>)
normally handle this automatically, these particular
utilities permit manipulating such attachments manually from
the command-line or in <A
HREF="#BATCHPROCREF"
>batch
processing mode</A
> by means of a shell script.</P
></DD
><DT
><A
NAME="CRYPTREF"
></A
><B
CLASS="COMMAND"
>crypt</B
></DT
><DD
><P
>At one time, this was the standard UNIX file encryption
utility.
<A
NAME="AEN12969"
HREF="#FTN.AEN12969"
><SPAN
CLASS="footnote"
>[79]</SPAN
></A
>
Politically-motivated government regulations
prohibiting the export of encryption software resulted
in the disappearance of <B
CLASS="COMMAND"
>crypt</B
>
from much of the UNIX world, and it is still
missing from most Linux distributions. Fortunately,
programmers have come up with a number of decent
alternatives to it, among them the author's very own <A
HREF="ftp://metalab.unc.edu/pub/Linux/utils/file/cruft-0.2.tar.gz"
TARGET="_top"
>cruft</A
>
(see <A
HREF="#ENCRYPTEDPW"
>Example A-4</A
>). </P
></DD
><DT
><A
NAME="OPENSSLREF"
></A
><B
CLASS="COMMAND"
>openssl</B
></DT
><DD
><P
>This is an Open Source implementation of
<I
CLASS="FIRSTTERM"
>Secure Sockets Layer</I
> encryption.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># To encrypt a file:
openssl aes-128-ecb -salt -in file.txt -out file.encrypted \
-pass pass:my_password
# ^^^^^^^^^^^ User-selected password.
# aes-128-ecb is the encryption method chosen.
# To decrypt an openssl-encrypted file:
openssl aes-128-ecb -d -salt -in file.encrypted -out file.txt \
-pass pass:my_password
# ^^^^^^^^^^^ User-selected password.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
HREF="#PIPEREF"
>Piping</A
>
<I
CLASS="FIRSTTERM"
>openssl</I
> to/from <A
HREF="#TARREF"
>tar</A
> makes it possible to encrypt
an entire directory tree.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># To encrypt a directory:
sourcedir="/home/bozo/testfiles"
encrfile="encr-dir.tar.gz"
password=my_secret_password
tar czvf - "$sourcedir" |
openssl des3 -salt -out "$encrfile" -pass pass:"$password"
# ^^^^ Uses des3 encryption.
# Writes encrypted file "encr-dir.tar.gz" in current working directory.
# To decrypt the resulting tarball:
openssl des3 -d -salt -in "$encrfile" -pass pass:"$password" |
tar -xzv
# Decrypts and unpacks into current working directory.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Of course, <I
CLASS="FIRSTTERM"
>openssl</I
> has many other uses,
such as obtaining signed <I
CLASS="FIRSTTERM"
>certificates</I
>
for Web sites. See the <A
HREF="#INFOREF"
>info</A
>
page.</P
></DD
><DT
><A
NAME="SHREDREF"
></A
><B
CLASS="COMMAND"
>shred</B
></DT
><DD
><P
>Securely erase a file by overwriting it multiple times with
random bit patterns before deleting it. This command has
the same effect as <A
HREF="#BLOTOUT"
>Example 16-61</A
>, but does it
in a more thorough and elegant manner.</P
><P
>This is one of the GNU
<I
CLASS="FIRSTTERM"
>fileutils</I
>.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Advanced forensic technology may still be able to
recover the contents of a file, even after application of
<B
CLASS="COMMAND"
>shred</B
>.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="FAMISC1"
></A
>Miscellaneous</B
></P
><DL
><DT
><A
NAME="MKTEMPREF"
></A
><B
CLASS="COMMAND"
>mktemp</B
></DT
><DD
><P
>Create a <I
CLASS="FIRSTTERM"
>temporary file</I
>
<A
NAME="AEN13030"
HREF="#FTN.AEN13030"
><SPAN
CLASS="footnote"
>[80]</SPAN
></A
>
with a <SPAN
CLASS="QUOTE"
>"unique"</SPAN
> filename. When invoked
from the command-line without additional arguments,
it creates a zero-length file in the <TT
CLASS="FILENAME"
>/tmp</TT
> directory.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>mktemp</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/tmp/tmp.zzsvql3154</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>PREFIX=filename
tempfile=`mktemp $PREFIX.XXXXXX`
# ^^^^^^ Need at least 6 placeholders
#+ in the filename template.
# If no filename template supplied,
#+ "tmp.XXXXXXXXXX" is the default.
echo "tempfile name = $tempfile"
# tempfile name = filename.QA2ZpY
# or something similar...
# Creates a file of that name in the current working directory
#+ with 600 file permissions.
# A "umask 177" is therefore unnecessary,
#+ but it's good programming practice nevertheless.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="MAKEREF"
></A
><B
CLASS="COMMAND"
>make</B
></DT
><DD
><P
><A
NAME="MAKEFILEREF"
></A
></P
><P
>Utility for building and compiling binary packages.
This can also be used for any set of operations triggered
by incremental changes in source files.</P
><P
>The <I
CLASS="FIRSTTERM"
>make</I
> command checks a
<TT
CLASS="FILENAME"
>Makefile</TT
>, a list of file dependencies and
operations to be carried out.</P
><P
>The <I
CLASS="FIRSTTERM"
>make</I
> utility is, in effect,
a powerful scripting language similar in many ways to
<I
CLASS="FIRSTTERM"
>Bash</I
>, but with the capability of
recognizing <I
CLASS="FIRSTTERM"
>dependencies</I
>. For in-depth
coverage of this useful tool set, see the <A
HREF="http://www.gnu.org/manual/manual.html"
TARGET="_top"
>GNU software
documentation site</A
>.</P
></DD
><DT
><A
NAME="INSTALLREF"
></A
><B
CLASS="COMMAND"
>install</B
></DT
><DD
><P
>Special purpose file copying command, similar to
<A
HREF="#CPREF"
>cp</A
>, but capable of
setting permissions and attributes of the copied
files. This command seems tailormade for installing
software packages, and as such it shows up frequently in
<TT
CLASS="FILENAME"
>Makefiles</TT
> (in the <TT
CLASS="REPLACEABLE"
><I
>make
install :</I
></TT
> section). It could likewise prove
useful in installation scripts.</P
></DD
><DT
><A
NAME="DOS2UNIXREF"
></A
><B
CLASS="COMMAND"
>dos2unix</B
></DT
><DD
><P
>This utility, written by Benjamin Lin and collaborators,
converts DOS-formatted text files (lines terminated by
CR-LF) to UNIX format (lines terminated by LF only),
and <A
HREF="#DOSNEWLINES"
>vice-versa</A
>.</P
></DD
><DT
><A
NAME="PTXREF"
></A
><B
CLASS="COMMAND"
>ptx</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>ptx [targetfile]</B
> command
outputs a permuted index (cross-reference list) of the
targetfile. This may be further filtered and formatted in a
pipe, if necessary.</P
></DD
><DT
><A
NAME="MOREREF"
></A
><B
CLASS="COMMAND"
>more</B
>, <A
NAME="LESSREF"
></A
><B
CLASS="COMMAND"
>less</B
></DT
><DD
><P
>Pagers that display a text file or stream to
<TT
CLASS="FILENAME"
>stdout</TT
>, one screenful at a time.
These may be used to filter the output of
<TT
CLASS="FILENAME"
>stdout</TT
> . . . or of a script.</P
><P
> An interesting application of <I
CLASS="FIRSTTERM"
>more</I
>
is to <SPAN
CLASS="QUOTE"
>"test drive"</SPAN
> a command sequence,
to forestall potentially unpleasant consequences.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>ls /home/bozo | awk '{print "rm -rf " $1}' | more
# ^^^^
# Testing the effect of the following (disastrous) command-line:
# ls /home/bozo | awk '{print "rm -rf " $1}' | sh
# Hand off to the shell to execute . . . ^^</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <I
CLASS="FIRSTTERM"
>less</I
> pager has the
interesting property of doing a formatted display of
<I
CLASS="FIRSTTERM"
>man page</I
> source. See <A
HREF="#MANED"
>Example A-39</A
>.</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="COMMUNICATIONS"
></A
>16.6. Communications Commands</H1
><P
>Certain of the following commands find use in
network data transfer and analysis, as well as in
<A
HREF="#CSPAMMERS"
>chasing spammers</A
>.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="COMMUNINFO1"
></A
>Information and Statistics</B
></P
><DL
><DT
><A
NAME="HOSTREF"
></A
><B
CLASS="COMMAND"
>host</B
></DT
><DD
><P
>Searches for information about an Internet host by name or
IP address, using DNS.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>host surfacemail.com</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>surfacemail.com. has address 202.92.42.236</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="IPCALCREF"
></A
><B
CLASS="COMMAND"
>ipcalc</B
></DT
><DD
><P
>Displays IP information for a host.
With the <TT
CLASS="OPTION"
>-h</TT
> option,
<B
CLASS="COMMAND"
>ipcalc</B
> does a reverse DNS lookup, finding
the name of the host (server) from the IP address.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ipcalc -h 202.92.42.236</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>HOSTNAME=surfacemail.com</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="NSLOOKUPREF"
></A
><B
CLASS="COMMAND"
>nslookup</B
></DT
><DD
><P
>Do an Internet <SPAN
CLASS="QUOTE"
>"name server lookup"</SPAN
>
on a host by IP address. This is essentially equivalent
to <B
CLASS="COMMAND"
>ipcalc -h</B
> or <B
CLASS="COMMAND"
>dig -x
</B
>. The command may be run either interactively
or noninteractively, i.e., from within a script.</P
><P
>The <B
CLASS="COMMAND"
>nslookup</B
> command has allegedly
been <SPAN
CLASS="QUOTE"
>"deprecated,"</SPAN
> but it is still useful.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>nslookup -sil 66.97.104.180</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>nslookup kuhleersparnis.ch
Server: 135.116.137.2
Address: 135.116.137.2#53
Non-authoritative answer:
Name: kuhleersparnis.ch</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="DIGREF"
></A
><B
CLASS="COMMAND"
>dig</B
></DT
><DD
><P
><B
CLASS="COMMAND"
>D</B
>omain <B
CLASS="COMMAND"
>I</B
>nformation
<B
CLASS="COMMAND"
>G</B
>roper. Similar to
<B
CLASS="COMMAND"
>nslookup</B
>, <I
CLASS="FIRSTTERM"
>dig</I
> does
an Internet <I
CLASS="FIRSTTERM"
>name server lookup</I
> on a host.
May be run from the command-line or from within a script.</P
><P
>Some interesting options to <I
CLASS="FIRSTTERM"
>dig</I
> are
<TT
CLASS="OPTION"
>+time=N</TT
> for setting a query timeout to
<TT
CLASS="PARAMETER"
><I
>N</I
></TT
> seconds, <TT
CLASS="OPTION"
>+nofail</TT
> for
continuing to query servers until a reply is received, and
<TT
CLASS="OPTION"
>-x</TT
> for doing a reverse address lookup.</P
><P
>Compare the output of <B
CLASS="COMMAND"
>dig -x</B
> with
<B
CLASS="COMMAND"
>ipcalc -h</B
> and
<B
CLASS="COMMAND"
>nslookup</B
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>dig -x 81.9.6.2</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>;; Got answer:
;; -&#62;&#62;HEADER&#60;&#60;- opcode: QUERY, status: NXDOMAIN, id: 11649
;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 0
;; QUESTION SECTION:
;2.6.9.81.in-addr.arpa. IN PTR
;; AUTHORITY SECTION:
6.9.81.in-addr.arpa. 3600 IN SOA ns.eltel.net. noc.eltel.net.
2002031705 900 600 86400 3600
;; Query time: 537 msec
;; SERVER: 135.116.137.2#53(135.116.137.2)
;; WHEN: Wed Jun 26 08:35:24 2002
;; MSG SIZE rcvd: 91</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="SPAMLOOKUP_0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="SPAMLOOKUP"
></A
><P
><B
>Example 16-40. Finding out where to report a spammer</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# spam-lookup.sh: Look up abuse contact to report a spammer.
# Thanks, Michael Zick.
# Check for command-line arg.
ARGCOUNT=1
E_WRONGARGS=85
if [ $# -ne "$ARGCOUNT" ]
then
echo "Usage: `basename $0` domain-name"
exit $E_WRONGARGS
fi
dig +short $1.contacts.abuse.net -c in -t txt
# Also try:
# dig +nssearch $1
# Tries to find "authoritative name servers" and display SOA records.
# The following also works:
# whois -h whois.abuse.net $1
# ^^ ^^^^^^^^^^^^^^^ Specify host.
# Can even lookup multiple spammers with this, i.e."
# whois -h whois.abuse.net $spamdomain1 $spamdomain2 . . .
# Exercise:
# --------
# Expand the functionality of this script
#+ so that it automatically e-mails a notification
#+ to the responsible ISP's contact address(es).
# Hint: use the "mail" command.
exit $?
# spam-lookup.sh chinatietong.com
# A known spam domain.
# "crnet_mgr@chinatietong.com"
# "crnet_tec@chinatietong.com"
# "postmaster@chinatietong.com"
# For a more elaborate version of this script,
#+ see the SpamViz home page, http://www.spamviz.net/index.html.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ISSPAMMER_0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="ISSPAMMER"
></A
><P
><B
>Example 16-41. Analyzing a spam domain</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
# is-spammer.sh: Identifying spam domains
# $Id: is-spammer, v 1.4 2004/09/01 19:37:52 mszick Exp $
# Above line is RCS ID info.
#
# This is a simplified version of the "is_spammer.bash
#+ script in the Contributed Scripts appendix.
# is-spammer &#60;domain.name&#62;
# Uses an external program: 'dig'
# Tested with version: 9.2.4rc5
# Uses functions.
# Uses IFS to parse strings by assignment into arrays.
# And even does something useful: checks e-mail blacklists.
# Use the domain.name(s) from the text body:
# http://www.good_stuff.spammer.biz/just_ignore_everything_else
# ^^^^^^^^^^^
# Or the domain.name(s) from any e-mail address:
# Really_Good_Offer@spammer.biz
#
# as the only argument to this script.
#(PS: have your Inet connection running)
#
# So, to invoke this script in the above two instances:
# is-spammer.sh spammer.biz
# Whitespace == :Space:Tab:Line Feed:Carriage Return:
WSP_IFS=$'\x20'$'\x09'$'\x0A'$'\x0D'
# No Whitespace == Line Feed:Carriage Return
No_WSP=$'\x0A'$'\x0D'
# Field separator for dotted decimal ip addresses
ADR_IFS=${No_WSP}'.'
# Get the dns text resource record.
# get_txt &#60;error_code&#62; &#60;list_query&#62;
get_txt() {
# Parse $1 by assignment at the dots.
local -a dns
IFS=$ADR_IFS
dns=( $1 )
IFS=$WSP_IFS
if [ "${dns[0]}" == '127' ]
then
# See if there is a reason.
echo $(dig +short $2 -t txt)
fi
}
# Get the dns address resource record.
# chk_adr &#60;rev_dns&#62; &#60;list_server&#62;
chk_adr() {
local reply
local server
local reason
server=${1}${2}
reply=$( dig +short ${server} )
# If reply might be an error code . . .
if [ ${#reply} -gt 6 ]
then
reason=$(get_txt ${reply} ${server} )
reason=${reason:-${reply}}
fi
echo ${reason:-' not blacklisted.'}
}
# Need to get the IP address from the name.
echo 'Get address of: '$1
ip_adr=$(dig +short $1)
dns_reply=${ip_adr:-' no answer '}
echo ' Found address: '${dns_reply}
# A valid reply is at least 4 digits plus 3 dots.
if [ ${#ip_adr} -gt 6 ]
then
echo
declare query
# Parse by assignment at the dots.
declare -a dns
IFS=$ADR_IFS
dns=( ${ip_adr} )
IFS=$WSP_IFS
# Reorder octets into dns query order.
rev_dns="${dns[3]}"'.'"${dns[2]}"'.'"${dns[1]}"'.'"${dns[0]}"'.'
# See: http://www.spamhaus.org (Conservative, well maintained)
echo -n 'spamhaus.org says: '
echo $(chk_adr ${rev_dns} 'sbl-xbl.spamhaus.org')
# See: http://ordb.org (Open mail relays)
echo -n ' ordb.org says: '
echo $(chk_adr ${rev_dns} 'relays.ordb.org')
# See: http://www.spamcop.net/ (You can report spammers here)
echo -n ' spamcop.net says: '
echo $(chk_adr ${rev_dns} 'bl.spamcop.net')
# # # other blacklist operations # # #
# See: http://cbl.abuseat.org.
echo -n ' abuseat.org says: '
echo $(chk_adr ${rev_dns} 'cbl.abuseat.org')
# See: http://dsbl.org/usage (Various mail relays)
echo
echo 'Distributed Server Listings'
echo -n ' list.dsbl.org says: '
echo $(chk_adr ${rev_dns} 'list.dsbl.org')
echo -n ' multihop.dsbl.org says: '
echo $(chk_adr ${rev_dns} 'multihop.dsbl.org')
echo -n 'unconfirmed.dsbl.org says: '
echo $(chk_adr ${rev_dns} 'unconfirmed.dsbl.org')
else
echo
echo 'Could not use that address.'
fi
exit 0
# Exercises:
# --------
# 1) Check arguments to script,
# and exit with appropriate error message if necessary.
# 2) Check if on-line at invocation of script,
# and exit with appropriate error message if necessary.
# 3) Substitute generic variables for "hard-coded" BHL domains.
# 4) Set a time-out for the script using the "+time=" option
to the 'dig' command.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>For a much more elaborate version of the above script, see
<A
HREF="#ISSPAMMER2"
>Example A-28</A
>.</P
></DD
><DT
><A
NAME="TRACEROUTEREF"
></A
><B
CLASS="COMMAND"
>traceroute</B
></DT
><DD
><P
>Trace the route taken by packets sent to a remote host. This
command works within a LAN, WAN, or over the
Internet. The remote host may be specified by an IP
address. The output of this command may be filtered
by <A
HREF="#GREPREF"
>grep</A
> or <A
HREF="#SEDREF"
>sed</A
> in a pipe.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>traceroute 81.9.6.2</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>traceroute to 81.9.6.2 (81.9.6.2), 30 hops max, 38 byte packets
1 tc43.xjbnnbrb.com (136.30.178.8) 191.303 ms 179.400 ms 179.767 ms
2 or0.xjbnnbrb.com (136.30.178.1) 179.536 ms 179.534 ms 169.685 ms
3 192.168.11.101 (192.168.11.101) 189.471 ms 189.556 ms *
...</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="PINGREF"
></A
><B
CLASS="COMMAND"
>ping</B
></DT
><DD
><P
>Broadcast an <TT
CLASS="REPLACEABLE"
><I
>ICMP
ECHO_REQUEST</I
></TT
> packet to another machine,
either on a local or remote network. This is a
diagnostic tool for testing network connections,
and it should be used with caution.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ping localhost</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>PING localhost.localdomain (127.0.0.1) from 127.0.0.1 : 56(84) bytes of data.
64 bytes from localhost.localdomain (127.0.0.1): icmp_seq=0 ttl=255 time=709 usec
64 bytes from localhost.localdomain (127.0.0.1): icmp_seq=1 ttl=255 time=286 usec
--- localhost.localdomain ping statistics ---
2 packets transmitted, 2 packets received, 0% packet loss
round-trip min/avg/max/mdev = 0.286/0.497/0.709/0.212 ms</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>A successful <I
CLASS="FIRSTTERM"
>ping</I
> returns
an <A
HREF="#EXITSTATUSREF"
>exit status</A
> of
<SPAN
CLASS="ERRORCODE"
>0</SPAN
>. This can be tested for in a
script.</P
><P
><A
NAME="PING0"
></A
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> HNAME=news-15.net # Notorious spammer.
# HNAME=$HOST # Debug: test for localhost.
count=2 # Send only two pings.
if [[ `ping -c $count "$HNAME"` ]]
then
echo ""$HNAME" still up and broadcasting spam your way."
else
echo ""$HNAME" seems to be down. Pity."
fi</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="WHOISREF"
></A
><B
CLASS="COMMAND"
>whois</B
></DT
><DD
><P
>Perform a DNS (Domain Name System) lookup.
The <TT
CLASS="OPTION"
>-h</TT
> option permits specifying which
particular <I
CLASS="FIRSTTERM"
>whois</I
> server to query. See
<A
HREF="#EX18"
>Example 4-6</A
> and <A
HREF="#SPAMLOOKUP"
>Example 16-40</A
>.</P
></DD
><DT
><A
NAME="FINGERREF"
></A
><B
CLASS="COMMAND"
>finger</B
></DT
><DD
><P
>Retrieve information about users on a
network. Optionally, this command can display
a user's <TT
CLASS="FILENAME"
>~/.plan</TT
>,
<TT
CLASS="FILENAME"
>~/.project</TT
>, and
<TT
CLASS="FILENAME"
>~/.forward</TT
> files, if present.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>finger</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Login Name Tty Idle Login Time Office Office Phone
bozo Bozo Bozeman tty1 8 Jun 25 16:59 (:0)
bozo Bozo Bozeman ttyp0 Jun 25 16:59 (:0.0)
bozo Bozo Bozeman ttyp1 Jun 25 17:07 (:0.0)</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>finger bozo</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Login: bozo Name: Bozo Bozeman
Directory: /home/bozo Shell: /bin/bash
Office: 2355 Clown St., 543-1234
On since Fri Aug 31 20:13 (MST) on tty1 1 hour 38 minutes idle
On since Fri Aug 31 20:13 (MST) on pts/0 12 seconds idle
On since Fri Aug 31 20:13 (MST) on pts/1
On since Fri Aug 31 20:31 (MST) on pts/2 1 hour 16 minutes idle
Mail last read Tue Jul 3 10:08 2007 (MST)
No Plan.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Out of security considerations, many networks disable
<B
CLASS="COMMAND"
>finger</B
> and its associated daemon.
<A
NAME="AEN13320"
HREF="#FTN.AEN13320"
><SPAN
CLASS="footnote"
>[81]</SPAN
></A
>
</P
></DD
><DT
><A
NAME="CHFNREF"
></A
><B
CLASS="COMMAND"
>chfn</B
></DT
><DD
><P
>Change information disclosed by the
<B
CLASS="COMMAND"
>finger</B
> command.</P
></DD
><DT
><A
NAME="VRFYREF"
></A
><B
CLASS="COMMAND"
>vrfy</B
></DT
><DD
><P
>Verify an Internet e-mail address.</P
><P
>This command seems to be missing from newer Linux
distros.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="COMMREMOTE1"
></A
>Remote Host Access</B
></P
><DL
><DT
><A
NAME="RXREF"
></A
><B
CLASS="COMMAND"
>sx</B
>, <B
CLASS="COMMAND"
>rx</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>sx</B
> and <B
CLASS="COMMAND"
>rx</B
>
command set serves to transfer files to and from a remote
host using the <I
CLASS="FIRSTTERM"
>xmodem</I
> protocol. These
are generally part of a communications package, such as
<B
CLASS="COMMAND"
>minicom</B
>.</P
></DD
><DT
><A
NAME="RZREF"
></A
><B
CLASS="COMMAND"
>sz</B
>, <B
CLASS="COMMAND"
>rz</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>sz</B
> and <B
CLASS="COMMAND"
>rz</B
>
command set serves to transfer files to and from a remote
host using the <I
CLASS="FIRSTTERM"
>zmodem</I
> protocol.
<I
CLASS="FIRSTTERM"
>Zmodem</I
> has certain advantages over
<I
CLASS="FIRSTTERM"
>xmodem</I
>, such as faster transmission
rate and resumption of interrupted file transfers.
Like <B
CLASS="COMMAND"
>sx</B
> and <B
CLASS="COMMAND"
>rx</B
>,
these are generally part of a communications package.</P
></DD
><DT
><A
NAME="FTPREF"
></A
><B
CLASS="COMMAND"
>ftp</B
></DT
><DD
><P
>Utility and protocol for uploading / downloading
files to or from a remote host. An ftp session can be automated
in a script (see <A
HREF="#EX72"
>Example 19-6</A
> and <A
HREF="#ENCRYPTEDPW"
>Example A-4</A
>).</P
></DD
><DT
><A
NAME="UUCPREF"
></A
><B
CLASS="COMMAND"
>uucp</B
>, <A
NAME="UUXREF"
></A
><B
CLASS="COMMAND"
>uux</B
>, <A
NAME="CUREF"
></A
><B
CLASS="COMMAND"
>cu</B
></DT
><DD
><P
><B
CLASS="COMMAND"
>uucp</B
>: <I
CLASS="FIRSTTERM"
>UNIX to UNIX
copy</I
>. This is a communications package for
transferring files between UNIX servers. A shell script
is an effective way to handle a <B
CLASS="COMMAND"
>uucp</B
>
command sequence.</P
><P
>Since the advent of the Internet and e-mail,
<B
CLASS="COMMAND"
>uucp</B
> seems to have faded into obscurity,
but it still exists and remains perfectly workable in
situations where an Internet connection is not available
or appropriate. The advantage of <B
CLASS="COMMAND"
>uucp</B
>
is that it is fault-tolerant, so even if there is a service
interruption the copy operation will resume where it left
off when the connection is restored.</P
><P
>---</P
><P
><B
CLASS="COMMAND"
>uux</B
>: <I
CLASS="FIRSTTERM"
>UNIX to UNIX
execute</I
>. Execute a command on a remote system.
This command is part of the <B
CLASS="COMMAND"
>uucp</B
>
package.</P
><P
>---</P
><P
><B
CLASS="COMMAND"
>cu</B
>: <B
CLASS="COMMAND"
>C</B
>all
<B
CLASS="COMMAND"
>U</B
>p a remote system and connect as a
simple terminal. It is a sort of dumbed-down version of
<A
HREF="#TELNETREF"
>telnet</A
>. This command is
part of the <B
CLASS="COMMAND"
>uucp</B
> package.</P
></DD
><DT
><A
NAME="TELNETREF"
></A
><B
CLASS="COMMAND"
>telnet</B
></DT
><DD
><P
>Utility and protocol for connecting to a remote host.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <I
CLASS="FIRSTTERM"
>telnet</I
> protocol
contains security holes and should therefore probably be
avoided. Its use within a shell script is
<EM
>not</EM
> recommended.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="WGETREF"
></A
><B
CLASS="COMMAND"
>wget</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>wget</B
> utility
<I
CLASS="FIRSTTERM"
>noninteractively</I
> retrieves or
downloads files from a Web or ftp site. It works well in a
script.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>wget -p http://www.xyz23.com/file01.html
# The -p or --page-requisite option causes wget to fetch all files
#+ required to display the specified page.
wget -r ftp://ftp.xyz24.net/~bozo/project_files/ -O $SAVEFILE
# The -r option recursively follows and retrieves all links
#+ on the specified site.
wget -c ftp://ftp.xyz25.net/bozofiles/filename.tar.bz2
# The -c option lets wget resume an interrupted download.
# This works with ftp servers and many HTTP sites.</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="QUOTEFETCH"
></A
><P
><B
>Example 16-42. Getting a stock quote</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# quote-fetch.sh: Download a stock quote.
E_NOPARAMS=86
if [ -z "$1" ] # Must specify a stock (symbol) to fetch.
then echo "Usage: `basename $0` stock-symbol"
exit $E_NOPARAMS
fi
stock_symbol=$1
file_suffix=.html
# Fetches an HTML file, so name it appropriately.
URL='http://finance.yahoo.com/q?s='
# Yahoo finance board, with stock query suffix.
# -----------------------------------------------------------
wget -O ${stock_symbol}${file_suffix} "${URL}${stock_symbol}"
# -----------------------------------------------------------
# To look up stuff on http://search.yahoo.com:
# -----------------------------------------------------------
# URL="http://search.yahoo.com/search?fr=ush-news&#38;p=${query}"
# wget -O "$savefilename" "${URL}"
# -----------------------------------------------------------
# Saves a list of relevant URLs.
exit $?
# Exercises:
# ---------
#
# 1) Add a test to ensure the user running the script is on-line.
# (Hint: parse the output of 'ps -ax' for "ppp" or "connect."
#
# 2) Modify this script to fetch the local weather report,
#+ taking the user's zip code as an argument.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#WGETTER2"
>Example A-30</A
> and <A
HREF="#BASHPODDER"
>Example A-31</A
>.</P
></DD
><DT
><A
NAME="LYNXREF"
></A
><B
CLASS="COMMAND"
>lynx</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>lynx</B
> Web and file browser
can be used inside a script (with the
<TT
CLASS="OPTION"
>-dump</TT
> option) to retrieve a file from a Web or
ftp site noninteractively.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>lynx -dump http://www.xyz23.com/file01.html &#62;$SAVEFILE</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>With the <TT
CLASS="OPTION"
>-traversal</TT
> option,
<B
CLASS="COMMAND"
>lynx</B
> starts at the HTTP URL specified
as an argument, then <SPAN
CLASS="QUOTE"
>"crawls"</SPAN
> through all
links located on that particular server. Used together
with the <TT
CLASS="OPTION"
>-crawl</TT
> option, outputs page text
to a log file.</P
></DD
><DT
><A
NAME="RLOGINREF"
></A
><B
CLASS="COMMAND"
>rlogin</B
></DT
><DD
><P
><TT
CLASS="REPLACEABLE"
><I
>Remote login</I
></TT
>, initates a
session on a remote host. This command has security issues,
so use <A
HREF="#SSHREF"
>ssh</A
> instead.</P
></DD
><DT
><A
NAME="RSHREF"
></A
><B
CLASS="COMMAND"
>rsh</B
></DT
><DD
><P
><TT
CLASS="REPLACEABLE"
><I
>Remote shell</I
></TT
>, executes
command(s) on a remote host. This has security issues,
so use <B
CLASS="COMMAND"
>ssh</B
> instead.</P
></DD
><DT
><A
NAME="RCPREF"
></A
><B
CLASS="COMMAND"
>rcp</B
></DT
><DD
><P
><TT
CLASS="REPLACEABLE"
><I
>Remote copy</I
></TT
>, copies files
between two different networked machines.</P
></DD
><DT
><A
NAME="RSYNCREF"
></A
><B
CLASS="COMMAND"
>rsync</B
></DT
><DD
><P
><TT
CLASS="REPLACEABLE"
><I
>Remote synchronize</I
></TT
>, updates
(synchronizes) files
between two different networked machines.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>rsync -a ~/sourcedir/*txt /node1/subdirectory/</B
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="FC4UPD"
></A
><P
><B
>Example 16-43. Updating FC4</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# fc4upd.sh
# Script author: Frank Wang.
# Slight stylistic modifications by ABS Guide author.
# Used in ABS Guide with permission.
# Download Fedora Core 4 update from mirror site using rsync.
# Should also work for newer Fedora Cores -- 5, 6, . . .
# Only download latest package if multiple versions exist,
#+ to save space.
URL=rsync://distro.ibiblio.org/fedora-linux-core/updates/
# URL=rsync://ftp.kddilabs.jp/fedora/core/updates/
# URL=rsync://rsync.planetmirror.com/fedora-linux-core/updates/
DEST=${1:-/var/www/html/fedora/updates/}
LOG=/tmp/repo-update-$(/bin/date +%Y-%m-%d).txt
PID_FILE=/var/run/${0##*/}.pid
E_RETURN=85 # Something unexpected happened.
# General rsync options
# -r: recursive download
# -t: reserve time
# -v: verbose
OPTS="-rtv --delete-excluded --delete-after --partial"
# rsync include pattern
# Leading slash causes absolute path name match.
INCLUDE=(
"/4/i386/kde-i18n-Chinese*"
# ^ ^
# Quoting is necessary to prevent globbing.
)
# rsync exclude pattern
# Temporarily comment out unwanted pkgs using "#" . . .
EXCLUDE=(
/1
/2
/3
/testing
/4/SRPMS
/4/ppc
/4/x86_64
/4/i386/debug
"/4/i386/kde-i18n-*"
"/4/i386/openoffice.org-langpack-*"
"/4/i386/*i586.rpm"
"/4/i386/GFS-*"
"/4/i386/cman-*"
"/4/i386/dlm-*"
"/4/i386/gnbd-*"
"/4/i386/kernel-smp*"
# "/4/i386/kernel-xen*"
# "/4/i386/xen-*"
)
init () {
# Let pipe command return possible rsync error, e.g., stalled network.
set -o pipefail # Newly introduced in Bash, version 3.
TMP=${TMPDIR:-/tmp}/${0##*/}.$$ # Store refined download list.
trap "{
rm -f $TMP 2&#62;/dev/null
}" EXIT # Clear temporary file on exit.
}
check_pid () {
# Check if process exists.
if [ -s "$PID_FILE" ]; then
echo "PID file exists. Checking ..."
PID=$(/bin/egrep -o "^[[:digit:]]+" $PID_FILE)
if /bin/ps --pid $PID &#38;&#62;/dev/null; then
echo "Process $PID found. ${0##*/} seems to be running!"
/usr/bin/logger -t ${0##*/} \
"Process $PID found. ${0##*/} seems to be running!"
exit $E_RETURN
fi
echo "Process $PID not found. Start new process . . ."
fi
}
# Set overall file update range starting from root or $URL,
#+ according to above patterns.
set_range () {
include=
exclude=
for p in "${INCLUDE[@]}"; do
include="$include --include \"$p\""
done
for p in "${EXCLUDE[@]}"; do
exclude="$exclude --exclude \"$p\""
done
}
# Retrieve and refine rsync update list.
get_list () {
echo $$ &#62; $PID_FILE || {
echo "Can't write to pid file $PID_FILE"
exit $E_RETURN
}
echo -n "Retrieving and refining update list . . ."
# Retrieve list -- 'eval' is needed to run rsync as a single command.
# $3 and $4 is the date and time of file creation.
# $5 is the full package name.
previous=
pre_file=
pre_date=0
eval /bin/nice /usr/bin/rsync \
-r $include $exclude $URL | \
egrep '^dr.x|^-r' | \
awk '{print $3, $4, $5}' | \
sort -k3 | \
{ while read line; do
# Get seconds since epoch, to filter out obsolete pkgs.
cur_date=$(date -d "$(echo $line | awk '{print $1, $2}')" +%s)
# echo $cur_date
# Get file name.
cur_file=$(echo $line | awk '{print $3}')
# echo $cur_file
# Get rpm pkg name from file name, if possible.
if [[ $cur_file == *rpm ]]; then
pkg_name=$(echo $cur_file | sed -r -e \
's/(^([^_-]+[_-])+)[[:digit:]]+\..*[_-].*$/\1/')
else
pkg_name=
fi
# echo $pkg_name
if [ -z "$pkg_name" ]; then # If not a rpm file,
echo $cur_file &#62;&#62; $TMP #+ then append to download list.
elif [ "$pkg_name" != "$previous" ]; then # A new pkg found.
echo $pre_file &#62;&#62; $TMP # Output latest file.
previous=$pkg_name # Save current.
pre_date=$cur_date
pre_file=$cur_file
elif [ "$cur_date" -gt "$pre_date" ]; then
# If same pkg, but newer,
pre_date=$cur_date #+ then update latest pointer.
pre_file=$cur_file
fi
done
echo $pre_file &#62;&#62; $TMP # TMP contains ALL
#+ of refined list now.
# echo "subshell=$BASH_SUBSHELL"
} # Bracket required here to let final "echo $pre_file &#62;&#62; $TMP"
# Remained in the same subshell ( 1 ) with the entire loop.
RET=$? # Get return code of the pipe command.
[ "$RET" -ne 0 ] &#38;&#38; {
echo "List retrieving failed with code $RET"
exit $E_RETURN
}
echo "done"; echo
}
# Real rsync download part.
get_file () {
echo "Downloading..."
/bin/nice /usr/bin/rsync \
$OPTS \
--filter "merge,+/ $TMP" \
--exclude '*' \
$URL $DEST \
| /usr/bin/tee $LOG
RET=$?
# --filter merge,+/ is crucial for the intention.
# + modifier means include and / means absolute path.
# Then sorted list in $TMP will contain ascending dir name and
#+ prevent the following --exclude '*' from "shortcutting the circuit."
echo "Done"
rm -f $PID_FILE 2&#62;/dev/null
return $RET
}
# -------
# Main
init
check_pid
set_range
get_list
get_file
RET=$?
# -------
if [ "$RET" -eq 0 ]; then
/usr/bin/logger -t ${0##*/} "Fedora update mirrored successfully."
else
/usr/bin/logger -t ${0##*/} \
"Fedora update mirrored with failure code: $RET"
fi
exit $RET</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#NIGHTLYBACKUP"
>Example A-32</A
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Using <A
HREF="#RCPREF"
>rcp</A
>, <A
HREF="#RSYNCREF"
>rsync</A
>, and similar
utilities with security implications in a shell
script may not be advisable. Consider, instead, using
<B
CLASS="COMMAND"
>ssh</B
>, <A
HREF="#SCPREF"
>scp</A
>,
or an <B
CLASS="COMMAND"
>expect</B
> script.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SSHREF"
></A
><B
CLASS="COMMAND"
>ssh</B
></DT
><DD
><P
><TT
CLASS="REPLACEABLE"
><I
>Secure shell</I
></TT
>, logs onto
a remote host and executes commands there. This
secure replacement for <B
CLASS="COMMAND"
>telnet</B
>,
<B
CLASS="COMMAND"
>rlogin</B
>, <B
CLASS="COMMAND"
>rcp</B
>, and
<B
CLASS="COMMAND"
>rsh</B
> uses identity authentication
and encryption. See its <A
HREF="#MANREF"
>manpage</A
>
for details.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="REMOTE"
></A
><P
><B
>Example 16-44. Using <I
CLASS="FIRSTTERM"
>ssh</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# remote.bash: Using ssh.
# This example by Michael Zick.
# Used with permission.
# Presumptions:
# ------------
# fd-2 isn't being captured ( '2&#62;/dev/null' ).
# ssh/sshd presumes stderr ('2') will display to user.
#
# sshd is running on your machine.
# For any 'standard' distribution, it probably is,
#+ and without any funky ssh-keygen having been done.
# Try ssh to your machine from the command-line:
#
# $ ssh $HOSTNAME
# Without extra set-up you'll be asked for your password.
# enter password
# when done, $ exit
#
# Did that work? If so, you're ready for more fun.
# Try ssh to your machine as 'root':
#
# $ ssh -l root $HOSTNAME
# When asked for password, enter root's, not yours.
# Last login: Tue Aug 10 20:25:49 2004 from localhost.localdomain
# Enter 'exit' when done.
# The above gives you an interactive shell.
# It is possible for sshd to be set up in a 'single command' mode,
#+ but that is beyond the scope of this example.
# The only thing to note is that the following will work in
#+ 'single command' mode.
# A basic, write stdout (local) command.
ls -l
# Now the same basic command on a remote machine.
# Pass a different 'USERNAME' 'HOSTNAME' if desired:
USER=${USERNAME:-$(whoami)}
HOST=${HOSTNAME:-$(hostname)}
# Now excute the above command-line on the remote host,
#+ with all transmissions encrypted.
ssh -l ${USER} ${HOST} " ls -l "
# The expected result is a listing of your username's home
#+ directory on the remote machine.
# To see any difference, run this script from somewhere
#+ other than your home directory.
# In other words, the Bash command is passed as a quoted line
#+ to the remote shell, which executes it on the remote machine.
# In this case, sshd does ' bash -c "ls -l" ' on your behalf.
# For information on topics such as not having to enter a
#+ password/passphrase for every command-line, see
#+ man ssh
#+ man ssh-keygen
#+ man sshd_config.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Within a loop, <B
CLASS="COMMAND"
>ssh</B
> may cause
unexpected behavior. According to a <A
HREF="http://groups-beta.google.com/group/comp.unix.shell/msg/dcb446b5fff7d230"
TARGET="_top"
> Usenet post</A
> in the comp.unix shell archives,
<B
CLASS="COMMAND"
>ssh</B
> inherits the loop's
<TT
CLASS="FILENAME"
>stdin</TT
>. To remedy this, pass
<B
CLASS="COMMAND"
>ssh</B
> either the <TT
CLASS="OPTION"
>-n</TT
>
or <TT
CLASS="OPTION"
>-f</TT
> option.</P
><P
>Thanks, Jason Bechtel, for pointing this out.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SCPREF"
></A
><B
CLASS="COMMAND"
>scp</B
></DT
><DD
><P
><TT
CLASS="REPLACEABLE"
><I
>Secure copy</I
></TT
>, similar in
function to <B
CLASS="COMMAND"
>rcp</B
>, copies files between
two different networked machines, but does so using
authentication, and with a security level similar to
<B
CLASS="COMMAND"
>ssh</B
>.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="COMMLOCAL1"
></A
>Local Network</B
></P
><DL
><DT
><A
NAME="WRITEREF"
></A
><B
CLASS="COMMAND"
>write</B
></DT
><DD
><P
>This is a utility for terminal-to-terminal communication.
It allows sending lines from your terminal (console or
<I
CLASS="FIRSTTERM"
>xterm</I
>) to that of another user. The
<A
HREF="#MESGREF"
>mesg</A
> command may, of course,
be used to disable write access to a terminal</P
><P
>Since <B
CLASS="COMMAND"
>write</B
> is interactive, it
would not normally find use in a script.</P
></DD
><DT
><A
NAME="NETCONFIGREF"
></A
><B
CLASS="COMMAND"
>netconfig</B
></DT
><DD
><P
>A command-line utility for configuring a network adapter
(using <I
CLASS="FIRSTTERM"
>DHCP</I
>). This command is native
to Red Hat centric Linux distros.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="COMMMAIL1"
></A
>Mail</B
></P
><DL
><DT
><B
CLASS="COMMAND"
>mail</B
></DT
><DD
><P
>Send or read e-mail messages.</P
><P
>This stripped-down command-line mail client
works fine as a command embedded in a script.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SELFMAILER"
></A
><P
><B
>Example 16-45. A script that mails itself</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/sh
# self-mailer.sh: Self-mailing script
adr=${1:-`whoami`} # Default to current user, if not specified.
# Typing 'self-mailer.sh wiseguy@superdupergenius.com'
#+ sends this script to that addressee.
# Just 'self-mailer.sh' (no argument) sends the script
#+ to the person invoking it, for example, bozo@localhost.localdomain.
#
# For more on the ${parameter:-default} construct,
#+ see the "Parameter Substitution" section
#+ of the "Variables Revisited" chapter.
# ============================================================================
cat $0 | mail -s "Script \"`basename $0`\" has mailed itself to you." "$adr"
# ============================================================================
# --------------------------------------------
# Greetings from the self-mailing script.
# A mischievous person has run this script,
#+ which has caused it to mail itself to you.
# Apparently, some people have nothing better
#+ to do with their time.
# --------------------------------------------
echo "At `date`, script \"`basename $0`\" mailed to "$adr"."
exit 0
# Note that the "mailx" command (in "send" mode) may be substituted
#+ for "mail" ... but with somewhat different options.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="MAILTOREF"
></A
><B
CLASS="COMMAND"
>mailto</B
></DT
><DD
><P
>Similar to the <B
CLASS="COMMAND"
>mail</B
> command,
<B
CLASS="COMMAND"
>mailto</B
> sends e-mail messages
from the command-line or in a script. However,
<B
CLASS="COMMAND"
>mailto</B
> also permits sending MIME
(multimedia) messages.</P
></DD
><DT
><A
NAME="MAILSTATSREF"
></A
><B
CLASS="COMMAND"
>mailstats</B
></DT
><DD
><P
>Show <I
CLASS="FIRSTTERM"
>mail statistics</I
>. This command
may be invoked only by <I
CLASS="FIRSTTERM"
>root</I
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>mailstats</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Statistics from Tue Jan 1 20:32:08 2008
M msgsfr bytes_from msgsto bytes_to msgsrej msgsdis msgsqur Mailer
4 1682 24118K 0 0K 0 0 0 esmtp
9 212 640K 1894 25131K 0 0 0 local
=====================================================================
T 1894 24758K 1894 25131K 0 0 0
C 414 0</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="VACATIONREF"
></A
><B
CLASS="COMMAND"
>vacation</B
></DT
><DD
><P
>This utility automatically replies to e-mails that
the intended recipient is on vacation and temporarily
unavailable. It runs on a network, in conjunction with
<B
CLASS="COMMAND"
>sendmail</B
>, and is not applicable to a
dial-up POPmail account.</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="TERMINALCCMDS"
></A
>16.7. Terminal Control Commands</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="TERMCOMMANDLISTING1"
></A
>Command affecting the console
or terminal</B
></P
><DL
><DT
><A
NAME="TPUTREF"
></A
><B
CLASS="COMMAND"
>tput</B
></DT
><DD
><P
>Initialize terminal and/or fetch information about it from
<SPAN
CLASS="DATABASE"
>terminfo</SPAN
> data. Various options permit
certain terminal operations: <B
CLASS="COMMAND"
>tput clear</B
>
is the equivalent of <A
HREF="#CLEARREF"
>clear</A
>;
<B
CLASS="COMMAND"
>tput reset</B
> is the equivalent
of <A
HREF="#RESETREF"
>reset</A
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>tput longname</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>xterm terminal emulator (X Window System)</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Issuing a <B
CLASS="COMMAND"
>tput cup X Y</B
> moves
the cursor to the (X,Y) coordinates in the current
terminal. A <B
CLASS="COMMAND"
>clear</B
> to erase the terminal
screen would normally precede this.</P
><P
>&#13; Some interesting options to <I
CLASS="FIRSTTERM"
>tput</I
> are:
<P
></P
><UL
><LI
><P
><TT
CLASS="OPTION"
>bold</TT
>, for high-intensity
text</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>smul</TT
>, to underline text
in the terminal</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>smso</TT
>, to render text in
reverse</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>sgr0</TT
>, to reset the terminal
parameters (to normal), without clearing the
screen</P
></LI
></UL
>
</P
><P
>Example scripts using <I
CLASS="FIRSTTERM"
>tput</I
>:
<P
></P
><OL
TYPE="1"
><LI
><P
><A
HREF="#COLORECHO"
>Example 36-15</A
></P
></LI
><LI
><P
><A
HREF="#EX30A"
>Example 36-13</A
></P
></LI
><LI
><P
><A
HREF="#HOMEWORK"
>Example A-44</A
></P
></LI
><LI
><P
><A
HREF="#NIM"
>Example A-42</A
></P
></LI
><LI
><P
><A
HREF="#POEM"
>Example 27-2</A
></P
></LI
></OL
>
</P
><P
>Note that <A
HREF="#STTYREF"
>stty</A
> offers
a more powerful command set for controlling a terminal.</P
></DD
><DT
><A
NAME="INFOCMPREF"
></A
><B
CLASS="COMMAND"
>infocmp</B
></DT
><DD
><P
>This command prints out extensive information about the
current terminal. It references the
<I
CLASS="FIRSTTERM"
>terminfo</I
> database.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>infocmp</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
># Reconstructed via infocmp from file:
/usr/share/terminfo/r/rxvt
rxvt|rxvt terminal emulator (X Window System),
am, bce, eo, km, mir, msgr, xenl, xon,
colors#8, cols#80, it#8, lines#24, pairs#64,
acsc=``aaffggjjkkllmmnnooppqqrrssttuuvvwwxxyyzz{{||}}~~,
bel=^G, blink=\E[5m, bold=\E[1m,
civis=\E[?25l,
clear=\E[H\E[2J, cnorm=\E[?25h, cr=^M,
...</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="RESETREF"
></A
><B
CLASS="COMMAND"
>reset</B
></DT
><DD
><P
>Reset terminal parameters and clear text screen. As with
<B
CLASS="COMMAND"
>clear</B
>, the cursor and prompt reappear in the
upper lefthand corner of the terminal.</P
></DD
><DT
><A
NAME="CLEARREF"
></A
><B
CLASS="COMMAND"
>clear</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>clear</B
> command simply clears
the text screen at the console or in an
<I
CLASS="FIRSTTERM"
>xterm</I
>. The prompt and cursor
reappear at the upper lefthand corner of the screen or
xterm window. This command may be used either at the command
line or in a script. See <A
HREF="#EX30"
>Example 11-26</A
>.</P
></DD
><DT
><A
NAME="RESIZEREF"
></A
><B
CLASS="COMMAND"
>resize</B
></DT
><DD
><P
>Echoes commands necessary to set <TT
CLASS="VARNAME"
>$TERM</TT
>
and <TT
CLASS="VARNAME"
>$TERMCAP</TT
> to duplicate the
<I
CLASS="FIRSTTERM"
>size</I
> (dimensions) of the current
terminal.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>resize</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>set noglob;
setenv COLUMNS '80';
setenv LINES '24';
unset noglob;</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="SCRIPTREF"
></A
><B
CLASS="COMMAND"
>script</B
></DT
><DD
><P
>This utility records (saves to a file) all the user keystrokes at
the command-line in a console or an xterm window. This, in effect,
creates a record of a session.</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="MATHC"
></A
>16.8. Math Commands</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="MATHCOMMANDLISTING1"
></A
><SPAN
CLASS="QUOTE"
>"Doing the
numbers"</SPAN
></B
></P
><DL
><DT
><A
NAME="FACTORREF"
></A
><B
CLASS="COMMAND"
>factor</B
></DT
><DD
><P
>Decompose an integer into prime factors.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>factor 27417</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>27417: 3 13 19 37</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="PRIMES2"
></A
><P
><B
>Example 16-46. Generating prime numbers</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# primes2.sh
# Generating prime numbers the quick-and-easy way,
#+ without resorting to fancy algorithms.
CEILING=10000 # 1 to 10000
PRIME=0
E_NOTPRIME=
is_prime ()
{
local factors
factors=( $(factor $1) ) # Load output of `factor` into array.
if [ -z "${factors[2]}" ]
# Third element of "factors" array:
#+ ${factors[2]} is 2nd factor of argument.
# If it is blank, then there is no 2nd factor,
#+ and the argument is therefore prime.
then
return $PRIME # 0
else
return $E_NOTPRIME # null
fi
}
echo
for n in $(seq $CEILING)
do
if is_prime $n
then
printf %5d $n
fi # ^ Five positions per number suffices.
done # For a higher $CEILING, adjust upward, as necessary.
echo
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="BCREF"
></A
><B
CLASS="COMMAND"
>bc</B
></DT
><DD
><P
>Bash can't handle floating point calculations, and
it lacks operators for certain important mathematical
functions. Fortunately, <B
CLASS="COMMAND"
>bc</B
> gallops to
the rescue.</P
><P
>Not just a versatile, arbitrary precision calculation
utility, <B
CLASS="COMMAND"
>bc</B
> offers many of the facilities of
a programming language. It has a syntax vaguely resembling
<B
CLASS="COMMAND"
>C</B
>.</P
><P
>Since it is a fairly well-behaved UNIX utility, and may
therefore be used in a <A
HREF="#PIPEREF"
>pipe</A
>,
<B
CLASS="COMMAND"
>bc</B
> comes in handy in scripts.</P
><P
><A
NAME="BCTEMPLATE"
></A
></P
><P
>Here is a simple template for using
<B
CLASS="COMMAND"
>bc</B
> to calculate a script
variable. This uses <A
HREF="#COMMANDSUBREF"
>command
substitution</A
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
> <TT
CLASS="USERINPUT"
><B
>variable=$(echo "OPTIONS; OPERATIONS" | bc)</B
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="MONTHLYPMT0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="MONTHLYPMT"
></A
><P
><B
>Example 16-47. Monthly Payment on a Mortgage</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# monthlypmt.sh: Calculates monthly payment on a mortgage.
# This is a modification of code in the
#+ "mcalc" (mortgage calculator) package,
#+ by Jeff Schmidt
#+ and
#+ Mendel Cooper (yours truly, the ABS Guide author).
# http://www.ibiblio.org/pub/Linux/apps/financial/mcalc-1.6.tar.gz
echo
echo "Given the principal, interest rate, and term of a mortgage,"
echo "calculate the monthly payment."
bottom=1.0
echo
echo -n "Enter principal (no commas) "
read principal
echo -n "Enter interest rate (percent) " # If 12%, enter "12", not ".12".
read interest_r
echo -n "Enter term (months) "
read term
interest_r=$(echo "scale=9; $interest_r/100.0" | bc) # Convert to decimal.
# ^^^^^^^^^^^^^^^^^ Divide by 100.
# "scale" determines how many decimal places.
interest_rate=$(echo "scale=9; $interest_r/12 + 1.0" | bc)
top=$(echo "scale=9; $principal*$interest_rate^$term" | bc)
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
# Standard formula for figuring interest.
echo; echo "Please be patient. This may take a while."
let "months = $term - 1"
# ====================================================================
for ((x=$months; x &#62; 0; x--))
do
bot=$(echo "scale=9; $interest_rate^$x" | bc)
bottom=$(echo "scale=9; $bottom+$bot" | bc)
# bottom = $(($bottom + $bot"))
done
# ====================================================================
# --------------------------------------------------------------------
# Rick Boivie pointed out a more efficient implementation
#+ of the above loop, which decreases computation time by 2/3.
# for ((x=1; x &#60;= $months; x++))
# do
# bottom=$(echo "scale=9; $bottom * $interest_rate + 1" | bc)
# done
# And then he came up with an even more efficient alternative,
#+ one that cuts down the run time by about 95%!
# bottom=`{
# echo "scale=9; bottom=$bottom; interest_rate=$interest_rate"
# for ((x=1; x &#60;= $months; x++))
# do
# echo 'bottom = bottom * interest_rate + 1'
# done
# echo 'bottom'
# } | bc` # Embeds a 'for loop' within command substitution.
# --------------------------------------------------------------------------
# On the other hand, Frank Wang suggests:
# bottom=$(echo "scale=9; ($interest_rate^$term-1)/($interest_rate-1)" | bc)
# Because . . .
# The algorithm behind the loop
#+ is actually a sum of geometric proportion series.
# The sum formula is e0(1-q^n)/(1-q),
#+ where e0 is the first element and q=e(n+1)/e(n)
#+ and n is the number of elements.
# --------------------------------------------------------------------------
# let "payment = $top/$bottom"
payment=$(echo "scale=2; $top/$bottom" | bc)
# Use two decimal places for dollars and cents.
echo
echo "monthly payment = \$$payment" # Echo a dollar sign in front of amount.
echo
exit 0
# Exercises:
# 1) Filter input to permit commas in principal amount.
# 2) Filter input to permit interest to be entered as percent or decimal.
# 3) If you are really ambitious,
#+ expand this script to print complete amortization tables.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="BASE0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="BASE"
></A
><P
><B
>Example 16-48. Base Conversion</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
###########################################################################
# Shellscript: base.sh - print number to different bases (Bourne Shell)
# Author : Heiner Steven (heiner.steven@odn.de)
# Date : 07-03-95
# Category : Desktop
# $Id: base.sh,v 1.2 2000/02/06 19:55:35 heiner Exp $
# ==&#62; Above line is RCS ID info.
###########################################################################
# Description
#
# Changes
# 21-03-95 stv fixed error occuring with 0xb as input (0.2)
###########################################################################
# ==&#62; Used in ABS Guide with the script author's permission.
# ==&#62; Comments added by ABS Guide author.
NOARGS=85
PN=`basename "$0"` # Program name
VER=`echo '$Revision: 1.2 $' | cut -d' ' -f2` # ==&#62; VER=1.2
Usage () {
echo "$PN - print number to different bases, $VER (stv '95)
usage: $PN [number ...]
If no number is given, the numbers are read from standard input.
A number may be
binary (base 2) starting with 0b (i.e. 0b1100)
octal (base 8) starting with 0 (i.e. 014)
hexadecimal (base 16) starting with 0x (i.e. 0xc)
decimal otherwise (i.e. 12)" &#62;&#38;2
exit $NOARGS
} # ==&#62; Prints usage message.
Msg () {
for i # ==&#62; in [list] missing. Why?
do echo "$PN: $i" &#62;&#38;2
done
}
Fatal () { Msg "$@"; exit 66; }
PrintBases () {
# Determine base of the number
for i # ==&#62; in [list] missing...
do # ==&#62; so operates on command-line arg(s).
case "$i" in
0b*) ibase=2;; # binary
0x*|[a-f]*|[A-F]*) ibase=16;; # hexadecimal
0*) ibase=8;; # octal
[1-9]*) ibase=10;; # decimal
*)
Msg "illegal number $i - ignored"
continue;;
esac
# Remove prefix, convert hex digits to uppercase (bc needs this).
number=`echo "$i" | sed -e 's:^0[bBxX]::' | tr '[a-f]' '[A-F]'`
# ==&#62; Uses ":" as sed separator, rather than "/".
# Convert number to decimal
dec=`echo "ibase=$ibase; $number" | bc` # ==&#62; 'bc' is calculator utility.
case "$dec" in
[0-9]*) ;; # number ok
*) continue;; # error: ignore
esac
# Print all conversions in one line.
# ==&#62; 'here document' feeds command list to 'bc'.
echo `bc &#60;&#60;!
obase=16; "hex="; $dec
obase=10; "dec="; $dec
obase=8; "oct="; $dec
obase=2; "bin="; $dec
!
` | sed -e 's: : :g'
done
}
while [ $# -gt 0 ]
# ==&#62; Is a "while loop" really necessary here,
# ==&#62;+ since all the cases either break out of the loop
# ==&#62;+ or terminate the script.
# ==&#62; (Above comment by Paulo Marcel Coelho Aragao.)
do
case "$1" in
--) shift; break;;
-h) Usage;; # ==&#62; Help message.
-*) Usage;;
*) break;; # First number
esac # ==&#62; Error checking for illegal input might be appropriate.
shift
done
if [ $# -gt 0 ]
then
PrintBases "$@"
else # Read from stdin.
while read line
do
PrintBases $line
done
fi
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="BCHEREDOC"
></A
></P
><P
>An alternate method of invoking <B
CLASS="COMMAND"
>bc</B
>
involves using a <A
HREF="#HEREDOCREF"
>here
document</A
> embedded within a <A
HREF="#COMMANDSUBREF"
>command substitution</A
>
block. This is especially appropriate when a script
needs to pass a list of options and commands to
<B
CLASS="COMMAND"
>bc</B
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>variable=`bc &#60;&#60; LIMIT_STRING
options
statements
operations
LIMIT_STRING
`
...or...
variable=$(bc &#60;&#60; LIMIT_STRING
options
statements
operations
LIMIT_STRING
)</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ALTBC"
></A
><P
><B
>Example 16-49. Invoking <I
CLASS="FIRSTTERM"
>bc</I
> using a <I
CLASS="FIRSTTERM"
>here
document</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Invoking 'bc' using command substitution
# in combination with a 'here document'.
var1=`bc &#60;&#60; EOF
18.33 * 19.78
EOF
`
echo $var1 # 362.56
# $( ... ) notation also works.
v1=23.53
v2=17.881
v3=83.501
v4=171.63
var2=$(bc &#60;&#60; EOF
scale = 4
a = ( $v1 + $v2 )
b = ( $v3 * $v4 )
a * b + 15.35
EOF
)
echo $var2 # 593487.8452
var3=$(bc -l &#60;&#60; EOF
scale = 9
s ( 1.7 )
EOF
)
# Returns the sine of 1.7 radians.
# The "-l" option calls the 'bc' math library.
echo $var3 # .991664810
# Now, try it in a function...
hypotenuse () # Calculate hypotenuse of a right triangle.
{ # c = sqrt( a^2 + b^2 )
hyp=$(bc -l &#60;&#60; EOF
scale = 9
sqrt ( $1 * $1 + $2 * $2 )
EOF
)
# Can't directly return floating point values from a Bash function.
# But, can echo-and-capture:
echo "$hyp"
}
hyp=$(hypotenuse 3.68 7.31)
echo "hypotenuse = $hyp" # 8.184039344
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="CANNONREF"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="CANNON"
></A
><P
><B
>Example 16-50. Calculating PI</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# cannon.sh: Approximating PI by firing cannonballs.
# Author: Mendel Cooper
# License: Public Domain
# Version 2.2, reldate 13oct08.
# This is a very simple instance of a "Monte Carlo" simulation:
#+ a mathematical model of a real-life event,
#+ using pseudorandom numbers to emulate random chance.
# Consider a perfectly square plot of land, 10000 units on a side.
# This land has a perfectly circular lake in its center,
#+ with a diameter of 10000 units.
# The plot is actually mostly water, except for land in the four corners.
# (Think of it as a square with an inscribed circle.)
#
# We will fire iron cannonballs from an old-style cannon
#+ at the square.
# All the shots impact somewhere on the square,
#+ either in the lake or on the dry corners.
# Since the lake takes up most of the area,
#+ most of the shots will SPLASH! into the water.
# Just a few shots will THUD! into solid ground
#+ in the four corners of the square.
#
# If we take enough random, unaimed shots at the square,
#+ Then the ratio of SPLASHES to total shots will approximate
#+ the value of PI/4.
#
# The simplified explanation is that the cannon is actually
#+ shooting only at the upper right-hand quadrant of the square,
#+ i.e., Quadrant I of the Cartesian coordinate plane.
#
#
# Theoretically, the more shots taken, the better the fit.
# However, a shell script, as opposed to a compiled language
#+ with floating-point math built in, requires some compromises.
# This decreases the accuracy of the simulation.
DIMENSION=10000 # Length of each side of the plot.
# Also sets ceiling for random integers generated.
MAXSHOTS=1000 # Fire this many shots.
# 10000 or more would be better, but would take too long.
PMULTIPLIER=4.0 # Scaling factor.
declare -r M_PI=3.141592654
# Actual 9-place value of PI, for comparison purposes.
get_random ()
{
SEED=$(head -n 1 /dev/urandom | od -N 1 | awk '{ print $2 }')
RANDOM=$SEED # From "seeding-random.sh"
#+ example script.
let "rnum = $RANDOM % $DIMENSION" # Range less than 10000.
echo $rnum
}
distance= # Declare global variable.
hypotenuse () # Calculate hypotenuse of a right triangle.
{ # From "alt-bc.sh" example.
distance=$(bc -l &#60;&#60; EOF
scale = 0
sqrt ( $1 * $1 + $2 * $2 )
EOF
)
# Setting "scale" to zero rounds down result to integer value,
#+ a necessary compromise in this script.
# It decreases the accuracy of this simulation.
}
# ==========================================================
# main() {
# "Main" code block, mimicking a C-language main() function.
# Initialize variables.
shots=0
splashes=0
thuds=0
Pi=0
error=0
while [ "$shots" -lt "$MAXSHOTS" ] # Main loop.
do
xCoord=$(get_random) # Get random X and Y coords.
yCoord=$(get_random)
hypotenuse $xCoord $yCoord # Hypotenuse of
#+ right-triangle = distance.
((shots++))
printf "#%4d " $shots
printf "Xc = %4d " $xCoord
printf "Yc = %4d " $yCoord
printf "Distance = %5d " $distance # Distance from
#+ center of lake
#+ -- the "origin" --
#+ coordinate (0,0).
if [ "$distance" -le "$DIMENSION" ]
then
echo -n "SPLASH! "
((splashes++))
else
echo -n "THUD! "
((thuds++))
fi
Pi=$(echo "scale=9; $PMULTIPLIER*$splashes/$shots" | bc)
# Multiply ratio by 4.0.
echo -n "PI ~ $Pi"
echo
done
echo
echo "After $shots shots, PI looks like approximately $Pi"
# Tends to run a bit high,
#+ possibly due to round-off error and imperfect randomness of $RANDOM.
# But still usually within plus-or-minus 5% . . .
#+ a pretty fair rough approximation.
error=$(echo "scale=9; $Pi - $M_PI" | bc)
pct_error=$(echo "scale=2; 100.0 * $error / $M_PI" | bc)
echo -n "Deviation from mathematical value of PI = $error"
echo " ($pct_error% error)"
echo
# End of "main" code block.
# }
# ==========================================================
exit 0
# One might well wonder whether a shell script is appropriate for
#+ an application as complex and computation-intensive as a simulation.
#
# There are at least two justifications.
# 1) As a proof of concept: to show it can be done.
# 2) To prototype and test the algorithms before rewriting
#+ it in a compiled high-level language.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#STDDEV"
>Example A-37</A
>.</P
></DD
><DT
><A
NAME="DCREF"
></A
><B
CLASS="COMMAND"
>dc</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>dc</B
> (<B
CLASS="COMMAND"
>d</B
>esk
<B
CLASS="COMMAND"
>c</B
>alculator) utility is <A
HREF="#STACKDEFREF"
>stack-oriented</A
>
and uses RPN (<I
CLASS="FIRSTTERM"
>Reverse Polish Notation</I
>).
Like <B
CLASS="COMMAND"
>bc</B
>, it has much of the power of
a programming language.</P
><P
>Similar to the procedure with <B
CLASS="COMMAND"
>bc</B
>,
<A
HREF="#ECHOREF"
>echo</A
> a command-string
to <B
CLASS="COMMAND"
>dc</B
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo "[Printing a string ... ]P" | dc
# The P command prints the string between the preceding brackets.
# And now for some simple arithmetic.
echo "7 8 * p" | dc # 56
# Pushes 7, then 8 onto the stack,
#+ multiplies ("*" operator), then prints the result ("p" operator).</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Most persons avoid <B
CLASS="COMMAND"
>dc</B
>, because
of its non-intuitive input and rather cryptic
operators. Yet, it has its uses.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="HEXCONVERT"
></A
><P
><B
>Example 16-51. Converting a decimal number to hexadecimal</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# hexconvert.sh: Convert a decimal number to hexadecimal.
E_NOARGS=85 # Command-line arg missing.
BASE=16 # Hexadecimal.
if [ -z "$1" ]
then # Need a command-line argument.
echo "Usage: $0 number"
exit $E_NOARGS
fi # Exercise: add argument validity checking.
hexcvt ()
{
if [ -z "$1" ]
then
echo 0
return # "Return" 0 if no arg passed to function.
fi
echo ""$1" "$BASE" o p" | dc
# o sets radix (numerical base) of output.
# p prints the top of stack.
# For other options: 'man dc' ...
return
}
hexcvt "$1"
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Studying the <A
HREF="#INFOREF"
>info</A
> page for
<B
CLASS="COMMAND"
>dc</B
> is a painful path to understanding its
intricacies. There seems to be a small, select group of
<EM
>dc wizards</EM
> who delight in showing off
their mastery of this powerful, but arcane utility.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "16i[q]sa[ln0=aln100%Pln100/snlbx]sbA0D68736142snlbxq" | dc</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Bash</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="GOLDENRATIO"
></A
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>dc &#60;&#60;&#60; 10k5v1+2/p # 1.6180339887
# ^^^ Feed operations to dc using a Here String.
# ^^^ Pushes 10 and sets that as the precision (10k).
# ^^ Pushes 5 and takes its square root
# (5v, v = square root).
# ^^ Pushes 1 and adds it to the running total (1+).
# ^^ Pushes 2 and divides the running total by that (2/).
# ^ Pops and prints the result (p)
# The result is 1.6180339887 ...
# ... which happens to be the Pythagorean Golden Ratio, to 10 places.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="FACTR"
></A
><P
><B
>Example 16-52. Factoring</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# factr.sh: Factor a number
MIN=2 # Will not work for number smaller than this.
E_NOARGS=85
E_TOOSMALL=86
if [ -z $1 ]
then
echo "Usage: $0 number"
exit $E_NOARGS
fi
if [ "$1" -lt "$MIN" ]
then
echo "Number to factor must be $MIN or greater."
exit $E_TOOSMALL
fi
# Exercise: Add type checking (to reject non-integer arg).
echo "Factors of $1:"
# -------------------------------------------------------
echo "$1[p]s2[lip/dli%0=1dvsr]s12sid2%0=13sidvsr[dli%0=\
1lrli2+dsi!&#62;.]ds.xd1&#60;2" | dc
# -------------------------------------------------------
# Above code written by Michel Charpentier &#60;charpov@cs.unh.edu&#62;
# (as a one-liner, here broken into two lines for display purposes).
# Used in ABS Guide with permission (thanks!).
exit
# $ sh factr.sh 270138
# 2
# 3
# 11
# 4093</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="AWKMATH"
></A
><B
CLASS="COMMAND"
>awk</B
></DT
><DD
><P
>Yet another way of doing floating point math in
a script is using <A
HREF="#AWKREF"
>awk's</A
>
built-in math functions in a <A
HREF="#SHWRAPPER"
>shell
wrapper</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="HYPOT"
></A
><P
><B
>Example 16-53. Calculating the hypotenuse of a triangle</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# hypotenuse.sh: Returns the "hypotenuse" of a right triangle.
# (square root of sum of squares of the "legs")
ARGS=2 # Script needs sides of triangle passed.
E_BADARGS=85 # Wrong number of arguments.
if [ $# -ne "$ARGS" ] # Test number of arguments to script.
then
echo "Usage: `basename $0` side_1 side_2"
exit $E_BADARGS
fi
AWKSCRIPT=' { printf( "%3.7f\n", sqrt($1*$1 + $2*$2) ) } '
# command(s) / parameters passed to awk
# Now, pipe the parameters to awk.
echo -n "Hypotenuse of $1 and $2 = "
echo $1 $2 | awk "$AWKSCRIPT"
# ^^^^^^^^^^^^
# An echo-and-pipe is an easy way of passing shell parameters to awk.
exit
# Exercise: Rewrite this script using 'bc' rather than awk.
# Which method is more intuitive?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="EXTMISC"
></A
>16.9. Miscellaneous Commands</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="MISCCOMMANDLISTING1"
></A
>Command that fit in no
special category</B
></P
><DL
><DT
><A
NAME="JOTREF"
></A
><B
CLASS="COMMAND"
>jot</B
>, <A
NAME="SEQREF"
></A
><B
CLASS="COMMAND"
>seq</B
></DT
><DD
><P
>These utilities emit a sequence of integers, with a
user-selectable increment.</P
><P
>The default separator character between each integer is a
newline, but this can be changed with the <TT
CLASS="OPTION"
>-s</TT
>
option.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>seq 5</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>1
2
3
4
5</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>seq -s : 5</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>1:2:3:4:5</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Both <B
CLASS="COMMAND"
>jot</B
> and <B
CLASS="COMMAND"
>seq</B
>
come in handy in a <A
HREF="#FORLOOPREF1"
>for
loop</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX53"
></A
><P
><B
>Example 16-54. Using <I
CLASS="FIRSTTERM"
>seq</I
> to generate loop
arguments</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Using "seq"
echo
for a in `seq 80` # or for a in $( seq 80 )
# Same as for a in 1 2 3 4 5 ... 80 (saves much typing!).
# May also use 'jot' (if present on system).
do
echo -n "$a "
done # 1 2 3 4 5 ... 80
# Example of using the output of a command to generate
# the [list] in a "for" loop.
echo; echo
COUNT=80 # Yes, 'seq' also accepts a replaceable parameter.
for a in `seq $COUNT` # or for a in $( seq $COUNT )
do
echo -n "$a "
done # 1 2 3 4 5 ... 80
echo; echo
BEGIN=75
END=80
for a in `seq $BEGIN $END`
# Giving "seq" two arguments starts the count at the first one,
#+ and continues until it reaches the second.
do
echo -n "$a "
done # 75 76 77 78 79 80
echo; echo
BEGIN=45
INTERVAL=5
END=80
for a in `seq $BEGIN $INTERVAL $END`
# Giving "seq" three arguments starts the count at the first one,
#+ uses the second for a step interval,
#+ and continues until it reaches the third.
do
echo -n "$a "
done # 45 50 55 60 65 70 75 80
echo; echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>A simpler example:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Create a set of 10 files,
#+ named file.1, file.2 . . . file.10.
COUNT=10
PREFIX=file
for filename in `seq $COUNT`
do
touch $PREFIX.$filename
# Or, can do other operations,
#+ such as rm, grep, etc.
done</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="LETTERCOUNT"
></A
><P
><B
>Example 16-55. Letter Count"</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# letter-count.sh: Counting letter occurrences in a text file.
# Written by Stefano Palmeri.
# Used in ABS Guide with permission.
# Slightly modified by document author.
MINARGS=2 # Script requires at least two arguments.
E_BADARGS=65
FILE=$1
let LETTERS=$#-1 # How many letters specified (as command-line args).
# (Subtract 1 from number of command-line args.)
show_help(){
echo
echo Usage: `basename $0` file letters
echo Note: `basename $0` arguments are case sensitive.
echo Example: `basename $0` foobar.txt G n U L i N U x.
echo
}
# Checks number of arguments.
if [ $# -lt $MINARGS ]; then
echo
echo "Not enough arguments."
echo
show_help
exit $E_BADARGS
fi
# Checks if file exists.
if [ ! -f $FILE ]; then
echo "File \"$FILE\" does not exist."
exit $E_BADARGS
fi
# Counts letter occurrences .
for n in `seq $LETTERS`; do
shift
if [[ `echo -n "$1" | wc -c` -eq 1 ]]; then # Checks arg.
echo "$1" -\&#62; `cat $FILE | tr -cd "$1" | wc -c` # Counting.
else
echo "$1 is not a single char."
fi
done
exit $?
# This script has exactly the same functionality as letter-count2.sh,
#+ but executes faster.
# Why?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Somewhat more capable than <I
CLASS="FIRSTTERM"
>seq</I
>,
<B
CLASS="COMMAND"
>jot</B
> is a classic UNIX
utility that is not normally included in a standard Linux
distro. However, the source <I
CLASS="FIRSTTERM"
>rpm</I
>
is available for download from the <A
HREF="http://www.mit.edu/afs/athena/system/rhlinux/athena-9.0/free/SRPMS/athena-jot-9.0-3.src.rpm"
TARGET="_top"
> MIT repository</A
>.</P
><P
><A
NAME="JOTRANDOM"
></A
></P
><P
>Unlike <I
CLASS="FIRSTTERM"
>seq</I
>, <B
CLASS="COMMAND"
>jot</B
> can
generate a sequence of random numbers, using the <TT
CLASS="OPTION"
>-r</TT
>
option.</P
><P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>jot -r 3 999</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>1069
1272
1428</TT
></PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="GETOPTY"
></A
><B
CLASS="COMMAND"
>getopt</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>getopt</B
> command
parses command-line options preceded by a <A
HREF="#DASHREF"
>dash</A
>. This external command
corresponds to the <A
HREF="#GETOPTSX"
>getopts</A
>
Bash builtin. Using <B
CLASS="COMMAND"
>getopt</B
> permits
handling long options by means of the <TT
CLASS="OPTION"
>-l</TT
>
flag, and this also allows parameter reshuffling.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX33A"
></A
><P
><B
>Example 16-56. Using <I
CLASS="FIRSTTERM"
>getopt</I
> to parse command-line
options</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Using getopt
# Try the following when invoking this script:
# sh ex33a.sh -a
# sh ex33a.sh -abc
# sh ex33a.sh -a -b -c
# sh ex33a.sh -d
# sh ex33a.sh -dXYZ
# sh ex33a.sh -d XYZ
# sh ex33a.sh -abcd
# sh ex33a.sh -abcdZ
# sh ex33a.sh -z
# sh ex33a.sh a
# Explain the results of each of the above.
E_OPTERR=65
if [ "$#" -eq 0 ]
then # Script needs at least one command-line argument.
echo "Usage $0 -[options a,b,c]"
exit $E_OPTERR
fi
set -- `getopt "abcd:" "$@"`
# Sets positional parameters to command-line arguments.
# What happens if you use "$*" instead of "$@"?
while [ ! -z "$1" ]
do
case "$1" in
-a) echo "Option \"a\"";;
-b) echo "Option \"b\"";;
-c) echo "Option \"c\"";;
-d) echo "Option \"d\" $2";;
*) break;;
esac
shift
done
# It is usually better to use the 'getopts' builtin in a script.
# See "ex33.sh."
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>As <EM
>Peggy Russell</EM
> points out:</P
><P
>It is often necessary to include an <A
HREF="#EVALREF"
>eval</A
> to correctly process
<A
HREF="#WHITESPACEREF"
>whitespace</A
> and
<I
CLASS="FIRSTTERM"
>quotes</I
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>args=$(getopt -o a:bc:d -- "$@")
eval set -- "$args"</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><P
>See <A
HREF="#GETOPTSIMPLE"
>Example 10-5</A
> for a simplified emulation
of <B
CLASS="COMMAND"
>getopt</B
>.</P
></DD
><DT
><A
NAME="RUNPARTSREF"
></A
><B
CLASS="COMMAND"
>run-parts</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>run-parts</B
> command
<A
NAME="AEN14105"
HREF="#FTN.AEN14105"
><SPAN
CLASS="footnote"
>[82]</SPAN
></A
>
executes all the scripts in a target directory, sequentially
in ASCII-sorted filename order. Of course, the scripts
need to have execute permission.</P
><P
>The <A
HREF="#CRONREF"
>cron</A
> <A
HREF="#DAEMONREF"
>daemon</A
> invokes
<B
CLASS="COMMAND"
>run-parts</B
> to run the scripts in
the <TT
CLASS="FILENAME"
>/etc/cron.*</TT
>
directories.</P
></DD
><DT
><A
NAME="YESREF"
></A
><B
CLASS="COMMAND"
>yes</B
></DT
><DD
><P
>In its default behavior the <B
CLASS="COMMAND"
>yes</B
>
command feeds a continuous string of the character
<TT
CLASS="COMPUTEROUTPUT"
>y</TT
> followed
by a line feed to <TT
CLASS="FILENAME"
>stdout</TT
>. A
<B
CLASS="KEYCAP"
>control</B
>-<B
CLASS="KEYCAP"
>C</B
>
terminates the run. A different output string
may be specified, as in <TT
CLASS="USERINPUT"
><B
>yes different
string</B
></TT
>, which would continually output
<TT
CLASS="COMPUTEROUTPUT"
>different string</TT
> to
<TT
CLASS="FILENAME"
>stdout</TT
>.</P
><P
>One might well ask the purpose of this. From the
command-line or in a script, the output of
<B
CLASS="COMMAND"
>yes</B
> can be redirected or piped into a
program expecting user input. In effect, this becomes a sort
of poor man's version of <I
CLASS="FIRSTTERM"
>expect</I
>.</P
><P
><TT
CLASS="USERINPUT"
><B
>yes | fsck /dev/hda1</B
></TT
> runs
<B
CLASS="COMMAND"
>fsck</B
> non-interactively (careful!).</P
><P
><TT
CLASS="USERINPUT"
><B
>yes | rm -r dirname</B
></TT
> has same effect as
<TT
CLASS="USERINPUT"
><B
>rm -rf dirname</B
></TT
> (careful!).</P
><DIV
CLASS="WARNING"
><P
></P
><TABLE
CLASS="WARNING"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/warning.gif"
HSPACE="5"
ALT="Warning"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Caution advised when piping
<I
CLASS="FIRSTTERM"
>yes</I
> to a potentially dangerous
system command, such as <A
HREF="#FSCKREF"
>fsck</A
>
or <A
HREF="#FDISKREF"
>fdisk</A
>. It might have
unintended consequences.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <I
CLASS="FIRSTTERM"
>yes</I
> command parses variables,
or more accurately, it echoes parsed variables.
For example:</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>yes $BASH_VERSION</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>3.1.17(1)-release
3.1.17(1)-release
3.1.17(1)-release
3.1.17(1)-release
3.1.17(1)-release
. . .</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> This particular <SPAN
CLASS="QUOTE"
>"feature"</SPAN
> may be used
to create a <EM
>very large</EM
> ASCII file on the fly:
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>yes $PATH &#62; huge_file.txt</B
></TT
>
<TT
CLASS="USERINPUT"
><B
>Ctl-C</B
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
Hit <TT
CLASS="USERINPUT"
><B
>Ctl-C</B
></TT
> <EM
>very
quickly</EM
>, or you just might get more than you
bargained for. . . .
</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="YESEMU"
></A
>The <I
CLASS="FIRSTTERM"
>yes</I
>
command may be emulated in a very simple script <A
HREF="#FUNCTIONREF"
>function</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>yes ()
{ # Trivial emulation of "yes" ...
local DEFAULT_TEXT="y"
while [ true ] # Endless loop.
do
if [ -z "$1" ]
then
echo "$DEFAULT_TEXT"
else # If argument ...
echo "$1" # ... expand and echo it.
fi
done # The only things missing are the
} #+ --help and --version options.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="BANNERREF"
></A
><B
CLASS="COMMAND"
>banner</B
></DT
><DD
><P
>Prints arguments as a large vertical banner to
<TT
CLASS="FILENAME"
>stdout</TT
>, using an <A
HREF="#ASCIIDEF"
>ASCII</A
> character (default
'#'). This may be redirected to a printer for
hardcopy.</P
><P
>Note that <I
CLASS="FIRSTTERM"
>banner</I
> has been
dropped from many Linux distros, presumably because it
is no longer considered useful.</P
></DD
><DT
><A
NAME="PRINTENVREF"
></A
><B
CLASS="COMMAND"
>printenv</B
></DT
><DD
><P
>Show all the <A
HREF="#ENVREF"
>environmental
variables</A
> set for a particular user.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>printenv | grep HOME</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>HOME=/home/bozo</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="LPREF"
></A
><B
CLASS="COMMAND"
>lp</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>lp</B
> and <B
CLASS="COMMAND"
>lpr</B
>
commands send file(s) to the print queue, to be printed as
hard copy.
<A
NAME="AEN14214"
HREF="#FTN.AEN14214"
><SPAN
CLASS="footnote"
>[83]</SPAN
></A
>
These commands trace the origin of their names to the
line printers of another era.
<A
NAME="AEN14218"
HREF="#FTN.AEN14218"
><SPAN
CLASS="footnote"
>[84]</SPAN
></A
>
</P
><P
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>lp file1.txt</B
></TT
>
or <TT
CLASS="PROMPT"
>bash </TT
><TT
CLASS="USERINPUT"
><B
>lp
&#60;file1.txt</B
></TT
></P
><P
>It is often useful to pipe the formatted output from
<B
CLASS="COMMAND"
>pr</B
> to <B
CLASS="COMMAND"
>lp</B
>.</P
><P
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>pr -options file1.txt | lp</B
></TT
>
</P
><P
>Formatting packages, such as <A
HREF="#GROFFREF"
>groff</A
> and
<I
CLASS="FIRSTTERM"
>Ghostscript</I
> may send their output
directly to <B
CLASS="COMMAND"
>lp</B
>.</P
><P
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>groff -Tascii file.tr | lp</B
></TT
>
</P
><P
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>gs -options | lp file.ps</B
></TT
>
</P
><P
>Related commands are <B
CLASS="COMMAND"
>lpq</B
>, for viewing
the print queue, and <B
CLASS="COMMAND"
>lprm</B
>, for removing
jobs from the print queue.</P
></DD
><DT
><A
NAME="TEEREF"
></A
><B
CLASS="COMMAND"
>tee</B
></DT
><DD
><P
>[UNIX borrows an idea from the plumbing trade.]</P
><P
>This is a redirection operator, but with a difference. Like the
plumber's <I
CLASS="FIRSTTERM"
>tee,</I
> it permits <SPAN
CLASS="QUOTE"
>"siphoning
off"</SPAN
> <EM
>to a file </EM
>the output of a command
or commands within a pipe, but without affecting the result. This is
useful for printing an ongoing process to a file or paper, perhaps to
keep track of it for debugging purposes.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
> (redirection)
|----&#62; to file
|
==========================|====================
command ---&#62; command ---&#62; |tee ---&#62; command ---&#62; ---&#62; output of pipe
===============================================
</PRE
></FONT
></TD
></TR
></TABLE
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat listfile* | sort | tee check.file | uniq &#62; result.file
# ^^^^^^^^^^^^^^ ^^^^
# The file "check.file" contains the concatenated sorted "listfiles,"
#+ before the duplicate lines are removed by 'uniq.'</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="MKFIFOREF"
></A
><B
CLASS="COMMAND"
>mkfifo</B
></DT
><DD
><P
><A
NAME="NAMEDPIPEREF"
></A
>This obscure command
creates a <I
CLASS="FIRSTTERM"
>named pipe</I
>, a temporary
<I
CLASS="FIRSTTERM"
>first-in-first-out buffer</I
> for
transferring data between processes.
<A
NAME="AEN14280"
HREF="#FTN.AEN14280"
><SPAN
CLASS="footnote"
>[85]</SPAN
></A
>
Typically, one process writes to the FIFO, and the other
reads from it. See <A
HREF="#FIFO"
>Example A-14</A
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# This short script by Omair Eshkenazi.
# Used in ABS Guide with permission (thanks!).
mkfifo pipe1 # Yes, pipes can be given names.
mkfifo pipe2 # Hence the designation "named pipe."
(cut -d' ' -f1 | tr "a-z" "A-Z") &#62;pipe2 &#60;pipe1 &#38;
ls -l | tr -s ' ' | cut -d' ' -f3,9- | tee pipe1 |
cut -d' ' -f2 | paste - pipe2
rm -f pipe1
rm -f pipe2
# No need to kill background processes when script terminates (why not?).
exit $?
Now, invoke the script and explain the output:
sh mkfifo-example.sh
4830.tar.gz BOZO
pipe1 BOZO
pipe2 BOZO
mkfifo-example.sh BOZO
Mixed.msg BOZO</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="PATHCHKREF"
></A
><B
CLASS="COMMAND"
>pathchk</B
></DT
><DD
><P
>This command checks the validity of a filename. If the
filename exceeds the maximum allowable length (255
characters) or one or more of the directories in
its path is not searchable, then an error message
results.</P
><P
>Unfortunately, <B
CLASS="COMMAND"
>pathchk</B
> does
not return a recognizable error code, and it is therefore
pretty much useless in a script. Consider instead the
<A
HREF="#RTIF"
>file test operators</A
>.</P
></DD
><DT
><A
NAME="DDREF"
></A
><B
CLASS="COMMAND"
>dd</B
></DT
><DD
><P
>Though this somewhat obscure and much feared
<B
CLASS="COMMAND"
>d</B
>ata <B
CLASS="COMMAND"
>d</B
>uplicator
command originated as a utility for exchanging
data on magnetic tapes between UNIX minicomputers
and IBM mainframes, it still has its uses.
The <B
CLASS="COMMAND"
>dd</B
> command simply copies a
file (or <TT
CLASS="FILENAME"
>stdin/stdout</TT
>), but with
conversions. <A
NAME="DDCONVERSIONS"
></A
>Possible conversions
include ASCII/EBCDIC,
<A
NAME="AEN14318"
HREF="#FTN.AEN14318"
><SPAN
CLASS="footnote"
>[86]</SPAN
></A
>
upper/lower case, swapping of byte pairs between input
and output, and skipping and/or truncating the head or
tail of the input file.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Converting a file to all uppercase:
dd if=$filename conv=ucase &#62; $filename.uppercase
# lcase # For lower case conversion</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="DDOPTIONS"
></A
></P
><P
>Some basic options to <B
CLASS="COMMAND"
>dd</B
> are:
<P
></P
><UL
><LI
><P
>if=INFILE</P
><P
>INFILE is the <I
CLASS="FIRSTTERM"
>source</I
>
file.</P
></LI
><LI
><P
>of=OUTFILE</P
><P
>OUTFILE is the <I
CLASS="FIRSTTERM"
>target</I
>
file, the file that will have the data written to it.</P
></LI
><LI
><P
>bs=BLOCKSIZE</P
><P
>This is the size of each block of data being read
and written, usually a power of 2.</P
></LI
><LI
><P
>skip=BLOCKS</P
><P
>How many blocks of data to skip in INFILE before
starting to copy. This is useful when the INFILE has
<SPAN
CLASS="QUOTE"
>"garbage"</SPAN
> or garbled data in its
header or when it is desirable to copy only a portion
of the INFILE.</P
></LI
><LI
><P
>seek=BLOCKS</P
><P
>How many blocks of data to skip in OUTFILE before
starting to copy, leaving blank data at beginning
of OUTFILE.</P
></LI
><LI
><P
>count=BLOCKS</P
><P
>Copy only this many blocks of data, rather than the
entire INFILE.</P
></LI
><LI
><P
>conv=CONVERSION</P
><P
>Type of conversion to be applied to INFILE data
before copying operation.</P
></LI
></UL
>
</P
><P
>A <TT
CLASS="USERINPUT"
><B
>dd --help</B
></TT
> lists all the
options this powerful utility takes.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SELFCOPY"
></A
><P
><B
>Example 16-57. A script that copies itself</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# self-copy.sh
# This script copies itself.
file_subscript=copy
dd if=$0 of=$0.$file_subscript 2&#62;/dev/null
# Suppress messages from dd: ^^^^^^^^^^^
exit $?
# A program whose only output is its own source code
#+ is called a "quine" per Willard Quine.
# Does this script qualify as a quine?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EXERCISINGDD"
></A
><P
><B
>Example 16-58. Exercising <I
CLASS="FIRSTTERM"
>dd</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# exercising-dd.sh
# Script by Stephane Chazelas.
# Somewhat modified by ABS Guide author.
infile=$0 # This script.
outfile=log.txt # Output file left behind.
n=8
p=11
dd if=$infile of=$outfile bs=1 skip=$((n-1)) count=$((p-n+1)) 2&#62; /dev/null
# Extracts characters n to p (8 to 11) from this script ("bash").
# ----------------------------------------------------------------
echo -n "hello vertical world" | dd cbs=1 conv=unblock 2&#62; /dev/null
# Echoes "hello vertical world" vertically downward.
# Why? A newline follows each character dd emits.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="DDKEYSTROKES"
></A
></P
><P
>To demonstrate just how versatile <B
CLASS="COMMAND"
>dd</B
> is,
let's use it to capture keystrokes.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="DDKEYPRESS"
></A
><P
><B
>Example 16-59. Capturing Keystrokes</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# dd-keypress.sh: Capture keystrokes without needing to press ENTER.
keypresses=4 # Number of keypresses to capture.
old_tty_setting=$(stty -g) # Save old terminal settings.
echo "Press $keypresses keys."
stty -icanon -echo # Disable canonical mode.
# Disable local echo.
keys=$(dd bs=1 count=$keypresses 2&#62; /dev/null)
# 'dd' uses stdin, if "if" (input file) not specified.
stty "$old_tty_setting" # Restore old terminal settings.
echo "You pressed the \"$keys\" keys."
# Thanks, Stephane Chazelas, for showing the way.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="DDRANDOM"
></A
></P
><P
>The <B
CLASS="COMMAND"
>dd</B
> command can do random access on a
data stream.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo -n . | dd bs=1 seek=4 of=file conv=notrunc
# The "conv=notrunc" option means that the output file
#+ will not be truncated.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="DDCOPY"
></A
></P
><P
>The <B
CLASS="COMMAND"
>dd</B
> command can copy raw data
and disk images to and from devices, such as floppies and
tape drives (<A
HREF="#COPYCD"
>Example A-5</A
>). A common use is
creating boot floppies.</P
><P
> <TT
CLASS="USERINPUT"
><B
>dd if=kernel-image of=/dev/fd0H1440</B
></TT
>
</P
><P
>Similarly, <B
CLASS="COMMAND"
>dd</B
> can copy the entire
contents of a floppy, even one formatted with a
<SPAN
CLASS="QUOTE"
>"foreign"</SPAN
> OS, to the hard drive as an
image file.</P
><P
> <TT
CLASS="USERINPUT"
><B
>dd if=/dev/fd0 of=/home/bozo/projects/floppy.img</B
></TT
>
</P
><P
><A
NAME="BFS"
></A
>Likewise, <B
CLASS="COMMAND"
>dd</B
>
can create bootable flash drives and SD cards.</P
><P
><TT
CLASS="USERINPUT"
><B
>dd if=image.iso of=/dev/sdb</B
></TT
></P
><P
><A
NAME="RPSDCARD01"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="RPSDCARD"
></A
><P
><B
>Example 16-60. Preparing a bootable SD card for the
<EM
>Raspberry Pi</EM
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# rp.sdcard.sh
# Preparing an SD card with a bootable image for the Raspberry Pi.
# $1 = imagefile name
# $2 = sdcard (device file)
# Otherwise defaults to the defaults, see below.
DEFAULTbs=4M # Block size, 4 mb default.
DEFAULTif="2013-07-26-wheezy-raspbian.img" # Commonly used distro.
DEFAULTsdcard="/dev/mmcblk0" # May be different. Check!
ROOTUSER_NAME=root # Must run as root!
E_NOTROOT=81
E_NOIMAGE=82
username=$(id -nu) # Who is running this script?
if [ "$username" != "$ROOTUSER_NAME" ]
then
echo "This script must run as root or with root privileges."
exit $E_NOTROOT
fi
if [ -n "$1" ]
then
imagefile="$1"
else
imagefile="$DEFAULTif"
fi
if [ -n "$2" ]
then
sdcard="$2"
else
sdcard="$DEFAULTsdcard"
fi
if [ ! -e $imagefile ]
then
echo "Image file \"$imagefile\" not found!"
exit $E_NOIMAGE
fi
echo "Last chance to change your mind!"; echo
read -s -n1 -p "Hit a key to write $imagefile to $sdcard [Ctl-c to exit]."
echo; echo
echo "Writing $imagefile to $sdcard ..."
dd bs=$DEFAULTbs if=$imagefile of=$sdcard
exit $?
# Exercises:
# ---------
# 1) Provide additional error checking.
# 2) Have script autodetect device file for SD card (difficult!).
# 3) Have script sutodetect image file (*img) in $PWD.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="DDSWAP"
></A
></P
><P
> Other applications of <B
CLASS="COMMAND"
>dd</B
> include
initializing temporary swap files (<A
HREF="#EX73"
>Example 31-2</A
>)
and ramdisks (<A
HREF="#RAMDISK"
>Example 31-3</A
>). It can even do a
low-level copy of an entire hard drive partition, although
this is not necessarily recommended.</P
><P
>People (with presumably nothing better to do with
their time) are constantly thinking of interesting
applications of <B
CLASS="COMMAND"
>dd</B
>.</P
><P
><A
NAME="DDFDEL"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="BLOTOUT"
></A
><P
><B
>Example 16-61. Securely deleting a file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# blot-out.sh: Erase "all" traces of a file.
# This script overwrites a target file alternately
#+ with random bytes, then zeros before finally deleting it.
# After that, even examining the raw disk sectors by conventional methods
#+ will not reveal the original file data.
PASSES=7 # Number of file-shredding passes.
# Increasing this slows script execution,
#+ especially on large target files.
BLOCKSIZE=1 # I/O with /dev/urandom requires unit block size,
#+ otherwise you get weird results.
E_BADARGS=70 # Various error exit codes.
E_NOT_FOUND=71
E_CHANGED_MIND=72
if [ -z "$1" ] # No filename specified.
then
echo "Usage: `basename $0` filename"
exit $E_BADARGS
fi
file=$1
if [ ! -e "$file" ]
then
echo "File \"$file\" not found."
exit $E_NOT_FOUND
fi
echo; echo -n "Are you absolutely sure you want to blot out \"$file\" (y/n)? "
read answer
case "$answer" in
[nN]) echo "Changed your mind, huh?"
exit $E_CHANGED_MIND
;;
*) echo "Blotting out file \"$file\".";;
esac
flength=$(ls -l "$file" | awk '{print $5}') # Field 5 is file length.
pass_count=1
chmod u+w "$file" # Allow overwriting/deleting the file.
echo
while [ "$pass_count" -le "$PASSES" ]
do
echo "Pass #$pass_count"
sync # Flush buffers.
dd if=/dev/urandom of=$file bs=$BLOCKSIZE count=$flength
# Fill with random bytes.
sync # Flush buffers again.
dd if=/dev/zero of=$file bs=$BLOCKSIZE count=$flength
# Fill with zeros.
sync # Flush buffers yet again.
let "pass_count += 1"
echo
done
rm -f $file # Finally, delete scrambled and shredded file.
sync # Flush buffers a final time.
echo "File \"$file\" blotted out and deleted."; echo
exit 0
# This is a fairly secure, if inefficient and slow method
#+ of thoroughly "shredding" a file.
# The "shred" command, part of the GNU "fileutils" package,
#+ does the same thing, although more efficiently.
# The file cannot not be "undeleted" or retrieved by normal methods.
# However . . .
#+ this simple method would *not* likely withstand
#+ sophisticated forensic analysis.
# This script may not play well with a journaled file system.
# Exercise (difficult): Fix it so it does.
# Tom Vier's "wipe" file-deletion package does a much more thorough job
#+ of file shredding than this simple script.
# http://www.ibiblio.org/pub/Linux/utils/file/wipe-2.0.0.tar.bz2
# For an in-depth analysis on the topic of file deletion and security,
#+ see Peter Gutmann's paper,
#+ "Secure Deletion of Data From Magnetic and Solid-State Memory".
# http://www.cs.auckland.ac.nz/~pgut001/pubs/secure_del.html</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also the <A
HREF="#DDLINK"
>dd
thread</A
> entry in the <A
HREF="#BIBLIOREF"
>bibliography</A
>.</P
></DD
><DT
><A
NAME="ODREF"
></A
><B
CLASS="COMMAND"
>od</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>od</B
>, or <I
CLASS="FIRSTTERM"
>octal
dump</I
> filter converts input (or files) to octal
(base-8) or other bases. This is useful for viewing or
processing binary data files or otherwise unreadable system
<A
HREF="#DEVFILEREF"
>device files</A
>, such as
<TT
CLASS="FILENAME"
>/dev/urandom</TT
>, and as a filter for
binary data.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>head -c4 /dev/urandom | od -N4 -tu4 | sed -ne '1s/.* //p'
# Sample output: 1324725719, 3918166450, 2989231420, etc.
# From rnd.sh example script, by St<53>phane Chazelas</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>See also <A
HREF="#SEEDINGRANDOM"
>Example 9-16</A
> and <A
HREF="#INSERTIONSORT"
>Example A-36</A
>.</P
></DD
><DT
><A
NAME="HEXDUMPREF"
></A
><B
CLASS="COMMAND"
>hexdump</B
></DT
><DD
><P
>Performs a hexadecimal, octal, decimal, or ASCII
dump of a binary file. This command is the rough equivalent
of <B
CLASS="COMMAND"
>od</B
>, above, but not nearly as
useful. May be used to view the contents of a binary file,
in combination with <A
HREF="#DDREF"
>dd</A
> and <A
HREF="#LESSREF"
>less</A
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>dd if=/bin/ls | hexdump -C | less
# The -C option nicely formats the output in tabular form.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="OBJDUMPREF"
></A
><B
CLASS="COMMAND"
>objdump</B
></DT
><DD
><P
>Displays information about an object file or binary
executable in either hexadecimal form or as a disassembled
listing (with the <TT
CLASS="OPTION"
>-d</TT
> option).</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>objdump -d /bin/ls</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/bin/ls: file format elf32-i386
Disassembly of section .init:
080490bc &#60;.init&#62;:
80490bc: 55 push %ebp
80490bd: 89 e5 mov %esp,%ebp
. . .</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="MCOOKIEREF"
></A
><B
CLASS="COMMAND"
>mcookie</B
></DT
><DD
><P
>This command generates a <SPAN
CLASS="QUOTE"
>"magic cookie,"</SPAN
> a
128-bit (32-character) pseudorandom hexadecimal number,
normally used as an authorization <SPAN
CLASS="QUOTE"
>"signature"</SPAN
>
by the X server. This also available for use in a script
as a <SPAN
CLASS="QUOTE"
>"quick 'n dirty"</SPAN
> random number.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>random000=$(mcookie)</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Of course, a script could use <A
HREF="#MD5SUMREF"
>md5sum</A
> for the same purpose.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Generate md5 checksum on the script itself.
random001=`md5sum $0 | awk '{print $1}'`
# Uses 'awk' to strip off the filename.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The <B
CLASS="COMMAND"
>mcookie</B
> command gives yet another way
to generate a <SPAN
CLASS="QUOTE"
>"unique"</SPAN
> filename.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="TEMPFILENAME"
></A
><P
><B
>Example 16-62. Filename generator</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# tempfile-name.sh: temp filename generator
BASE_STR=`mcookie` # 32-character magic cookie.
POS=11 # Arbitrary position in magic cookie string.
LEN=5 # Get $LEN consecutive characters.
prefix=temp # This is, after all, a "temp" file.
# For more "uniqueness," generate the
#+ filename prefix using the same method
#+ as the suffix, below.
suffix=${BASE_STR:POS:LEN}
# Extract a 5-character string,
#+ starting at position 11.
temp_filename=$prefix.$suffix
# Construct the filename.
echo "Temp filename = "$temp_filename""
# sh tempfile-name.sh
# Temp filename = temp.e19ea
# Compare this method of generating "unique" filenames
#+ with the 'date' method in ex51.sh.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="UNITSREF"
></A
><B
CLASS="COMMAND"
>units</B
></DT
><DD
><P
>This utility converts between different <I
CLASS="FIRSTTERM"
>units
of measure</I
>. While normally invoked in interactive
mode, <B
CLASS="COMMAND"
>units</B
> may find use in a
script.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="UNITCONVERSION"
></A
><P
><B
>Example 16-63. Converting meters to miles</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# unit-conversion.sh
# Must have 'units' utility installed.
convert_units () # Takes as arguments the units to convert.
{
cf=$(units "$1" "$2" | sed --silent -e '1p' | awk '{print $2}')
# Strip off everything except the actual conversion factor.
echo "$cf"
}
Unit1=miles
Unit2=meters
cfactor=`convert_units $Unit1 $Unit2`
quantity=3.73
result=$(echo $quantity*$cfactor | bc)
echo "There are $result $Unit2 in $quantity $Unit1."
# What happens if you pass incompatible units,
#+ such as "acres" and "miles" to the function?
exit 0
# Exercise: Edit this script to accept command-line parameters,
# with appropriate error checking, of course.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="M4REF"
></A
><B
CLASS="COMMAND"
>m4</B
></DT
><DD
><P
>A hidden treasure, <B
CLASS="COMMAND"
>m4</B
> is a
powerful macro
<A
NAME="AEN14523"
HREF="#FTN.AEN14523"
><SPAN
CLASS="footnote"
>[87]</SPAN
></A
>
processing filter, virtually a complete language.
Although originally written as a pre-processor
for <I
CLASS="FIRSTTERM"
>RatFor</I
>, <B
CLASS="COMMAND"
>m4</B
>
turned out to be useful as a stand-alone utility. In
fact, <B
CLASS="COMMAND"
>m4</B
> combines some of the
functionality of <A
HREF="#EVALREF"
>eval</A
>,
<A
HREF="#TRREF"
>tr</A
>, and <A
HREF="#AWKREF"
>awk</A
>, in addition to its extensive
macro expansion facilities.</P
><P
>The April, 2002 issue of <A
HREF="http://www.linuxjournal.com"
TARGET="_top"
><I
CLASS="CITETITLE"
>Linux Journal</I
></A
>
has a very nice article on <B
CLASS="COMMAND"
>m4</B
> and
its uses.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="M4"
></A
><P
><B
>Example 16-64. Using <I
CLASS="FIRSTTERM"
>m4</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# m4.sh: Using the m4 macro processor
# Strings
string=abcdA01
echo "len($string)" | m4 # 7
echo "substr($string,4)" | m4 # A01
echo "regexp($string,[0-1][0-1],\&#38;Z)" | m4 # 01Z
# Arithmetic
var=99
echo "incr($var)" | m4 # 100
echo "eval($var / 3)" | m4 # 33
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="XMESSAGEREF"
></A
><B
CLASS="COMMAND"
>xmessage</B
></DT
><DD
><P
>This X-based variant of
<A
HREF="#ECHOREF"
>echo</A
> pops up a message/query
window on the desktop.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>xmessage Left click to continue -button okay</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="ZENITYREF"
></A
><B
CLASS="COMMAND"
>zenity</B
></DT
><DD
><P
>The
<A
HREF="http://freshmeat.net/projects/zenity"
TARGET="_top"
>zenity</A
>
utility is adept at displaying
<I
CLASS="FIRSTTERM"
>GTK+</I
> dialog <A
HREF="#WIDGETREF"
>widgets</A
> and <A
HREF="#ZENITYREF2"
>very suitable for scripting
purposes</A
>.</P
></DD
><DT
><A
NAME="DOEXECREF"
></A
><B
CLASS="COMMAND"
>doexec</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>doexec</B
> command enables passing
an arbitrary list of arguments to a <I
CLASS="FIRSTTERM"
>binary
executable</I
>. In particular, passing
<TT
CLASS="PARAMETER"
><I
>argv[0]</I
></TT
> (which corresponds to <A
HREF="#POSPARAMREF1"
>$0</A
> in a script) lets the
executable be invoked by various names, and it can then
carry out different sets of actions, according to the name
by which it was called. What this amounts to is roundabout
way of passing options to an executable.</P
><P
>For example, the <TT
CLASS="FILENAME"
>/usr/local/bin</TT
> directory might
contain a binary called <SPAN
CLASS="QUOTE"
>"aaa"</SPAN
>. Invoking
<B
CLASS="COMMAND"
>doexec /usr/local/bin/aaa list</B
>
would <EM
>list</EM
> all those files
in the current working directory beginning with an
<SPAN
CLASS="QUOTE"
>"a"</SPAN
>, while invoking (the same executable
with) <B
CLASS="COMMAND"
>doexec /usr/local/bin/aaa delete </B
>
would <EM
>delete</EM
> those files.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The various behaviors of the executable
must be defined within the code of the executable itself,
analogous to something like the following in a shell script:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>case `basename $0` in
"name1" ) do_something;;
"name2" ) do_something_else;;
"name3" ) do_yet_another_thing;;
* ) bail_out;;
esac</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><B
CLASS="COMMAND"
>dialog</B
></DT
><DD
><P
>The <A
HREF="#DIALOGREF"
>dialog</A
> family of tools
provide a method of calling interactive
<SPAN
CLASS="QUOTE"
>"dialog"</SPAN
> boxes from a script. The more
elaborate variations of <B
CLASS="COMMAND"
>dialog</B
> --
<B
CLASS="COMMAND"
>gdialog</B
>, <B
CLASS="COMMAND"
>Xdialog</B
>,
and <B
CLASS="COMMAND"
>kdialog</B
> -- actually invoke X-Windows
<A
HREF="#WIDGETREF"
>widgets</A
>.</P
></DD
><DT
><A
NAME="SOXREF"
></A
><B
CLASS="COMMAND"
>sox</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>sox</B
>, or
<SPAN
CLASS="QUOTE"
>"<B
CLASS="COMMAND"
>so</B
>und
e<B
CLASS="COMMAND"
>x</B
>change"</SPAN
> command plays and
performs transformations on sound files. In fact,
the <TT
CLASS="FILENAME"
>/usr/bin/play</TT
> executable
(now deprecated) is nothing but a shell wrapper for
<I
CLASS="FIRSTTERM"
>sox</I
>.</P
><P
>For example, <B
CLASS="COMMAND"
>sox soundfile.wav
soundfile.au</B
> changes a WAV sound file into a
(Sun audio format) AU sound file.</P
><P
>Shell scripts are ideally suited for batch-processing
<B
CLASS="COMMAND"
>sox</B
> operations on
sound files. For examples, see the <A
HREF="http://osl.iu.edu/~tveldhui/radio/"
TARGET="_top"
> Linux Radio
Timeshift HOWTO</A
> and the <A
HREF="http://savannah.nongnu.org/projects/audiodo"
TARGET="_top"
>MP3do
Project</A
>.</P
></DD
></DL
></DIV
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="SYSTEM"
></A
>Chapter 17. System and Administrative Commands</H1
><P
>The startup and shutdown scripts in
<TT
CLASS="FILENAME"
>/etc/rc.d</TT
> illustrate the uses
(and usefulness) of many of these comands. These are usually
invoked by <I
CLASS="FIRSTTERM"
>root</I
> and used for system
maintenance or emergency filesystem repairs. Use with caution, as
some of these commands may damage your system if misused.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="USERSGROUPS1"
></A
>Users and Groups</B
></P
><DL
><DT
><A
NAME="USERSREF"
></A
><B
CLASS="COMMAND"
>users</B
></DT
><DD
><P
>Show all logged on users. This is the approximate
equivalent of <B
CLASS="COMMAND"
>who -q</B
>.</P
></DD
><DT
><A
NAME="GROUPSCMDREF"
></A
><B
CLASS="COMMAND"
>groups</B
></DT
><DD
><P
>Lists the current user and the groups she belongs to.
This corresponds to the <A
HREF="#GROUPSREF"
>$GROUPS</A
> internal variable,
but gives the group names, rather than the numbers.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>groups</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bozita cdrom cdwriter audio xgrp</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $GROUPS</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>501</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="CHOWNREF"
></A
><B
CLASS="COMMAND"
>chown</B
>, <A
NAME="CHGRPREF"
></A
><B
CLASS="COMMAND"
>chgrp</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>chown</B
> command changes the
ownership of a file or files. This command is a useful
method that <I
CLASS="FIRSTTERM"
>root</I
> can use to
shift file ownership from one user to another. An ordinary
user may not change the ownership of files, not even her
own files.
<A
NAME="AEN14695"
HREF="#FTN.AEN14695"
><SPAN
CLASS="footnote"
>[88]</SPAN
></A
>
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>chown bozo *.txt</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <B
CLASS="COMMAND"
>chgrp</B
> command changes the
<TT
CLASS="REPLACEABLE"
><I
>group</I
></TT
> ownership of a file or
files. You must be owner of the file(s) as well as a member
of the destination group (or <I
CLASS="FIRSTTERM"
>root</I
>)
to use this operation.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>chgrp --recursive dunderheads *.data
# The "dunderheads" group will now own all the "*.data" files
#+ all the way down the $PWD directory tree (that's what "recursive" means).</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="USERADDREF"
></A
><B
CLASS="COMMAND"
>useradd</B
>, <B
CLASS="COMMAND"
>userdel</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>useradd</B
> administrative command
adds a user account to the system and creates a home
directory for that particular user, if so specified. The
corresponding <B
CLASS="COMMAND"
>userdel</B
> command removes
a user account from the system
<A
NAME="AEN14727"
HREF="#FTN.AEN14727"
><SPAN
CLASS="footnote"
>[89]</SPAN
></A
>
and deletes associated files.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>adduser</B
> command is a synonym
for <B
CLASS="COMMAND"
>useradd</B
> and is usually a symbolic link to
it.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="USERMODREF"
></A
><B
CLASS="COMMAND"
>usermod</B
></DT
><DD
><P
>Modify a user account. Changes may be made to the password,
group membership, expiration date, and other attributes of
a given user's account. With this command, a user's password
may be locked, which has the effect of disabling the
account.</P
></DD
><DT
><A
NAME="GROUPMODREF"
></A
><B
CLASS="COMMAND"
>groupmod</B
></DT
><DD
><P
>Modify a given group. The group name and/or ID number may be
changed using this command.</P
></DD
><DT
><A
NAME="IDREF"
></A
><B
CLASS="COMMAND"
>id</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>id</B
> command lists the real and
effective user IDs and the group IDs of the user
associated with the current process. This is the
counterpart to the <A
HREF="#UIDREF"
>$UID</A
>,
<A
HREF="#EUIDREF"
>$EUID</A
>, and <A
HREF="#GROUPSREF"
>$GROUPS</A
> internal Bash
variables.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>id</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>uid=501(bozo) gid=501(bozo) groups=501(bozo),22(cdrom),80(cdwriter),81(audio)</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $UID</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>501</TT
></PRE
></FONT
></TD
></TR
></TABLE
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>id</B
> command shows the
<EM
>effective</EM
> IDs only when they differ
from the <EM
>real</EM
> ones.</P
></TD
></TR
></TABLE
></DIV
><P
>Also see <A
HREF="#AMIROOT"
>Example 9-5</A
>.</P
></DD
><DT
><A
NAME="LIDREF"
></A
><B
CLASS="COMMAND"
>lid</B
></DT
><DD
><P
>The <I
CLASS="FIRSTTERM"
>lid</I
> (list ID) command
shows the group(s) that a given user belongs to, or alternately,
the users belonging to a given group. May be invoked only by
root.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>lid bozo</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> bozo(gid=500)</TT
>
<TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>lid daemon</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> bin(gid=1)
daemon(gid=2)
adm(gid=4)
lp(gid=7)</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="WHOREF"
></A
><B
CLASS="COMMAND"
>who</B
></DT
><DD
><P
>Show all users logged on to the system.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>who</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bozo tty1 Apr 27 17:45
bozo pts/0 Apr 27 17:46
bozo pts/1 Apr 27 17:47
bozo pts/2 Apr 27 17:49</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <TT
CLASS="OPTION"
>-m</TT
> gives detailed information about
only the current user. Passing any two arguments to
<B
CLASS="COMMAND"
>who</B
> is the equivalent of <B
CLASS="COMMAND"
>who
-m</B
>, as in <B
CLASS="COMMAND"
>who am i</B
> or <B
CLASS="COMMAND"
>who
The Man</B
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>who -m</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>localhost.localdomain!bozo pts/2 Apr 27 17:49</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="WHOAMIREF"
></A
><B
CLASS="COMMAND"
>whoami</B
> is similar to <B
CLASS="COMMAND"
>who
-m</B
>, but only lists the user name.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>whoami</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bozo</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="WREF"
></A
><B
CLASS="COMMAND"
>w</B
></DT
><DD
><P
>Show all logged on users and the processes belonging to them. This is
an extended version of <B
CLASS="COMMAND"
>who</B
>. The output of <B
CLASS="COMMAND"
>w</B
>
may be piped to <A
HREF="#GREPREF"
>grep</A
> to find
a specific user and/or process.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>w | grep startx</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bozo tty1 - 4:22pm 6:41 4.47s 0.45s startx</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="LOGNAMEREF"
></A
><B
CLASS="COMMAND"
>logname</B
></DT
><DD
><P
>Show current user's login name (as found in
<TT
CLASS="FILENAME"
>/var/run/utmp</TT
>). This is a
near-equivalent to <A
HREF="#WHOAMIREF"
>whoami</A
>,
above.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>logname</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bozo</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>whoami</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bozo</TT
></PRE
></FONT
></TD
></TR
></TABLE
><P
>However . . .</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>su</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Password: ......</TT
>
<TT
CLASS="PROMPT"
>bash# </TT
><TT
CLASS="USERINPUT"
><B
>whoami</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>root</TT
>
<TT
CLASS="PROMPT"
>bash# </TT
><TT
CLASS="USERINPUT"
><B
>logname</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bozo</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>While <B
CLASS="COMMAND"
>logname</B
> prints the name
of the logged in user, <B
CLASS="COMMAND"
>whoami</B
> gives the
name of the user attached to the current process. As we have
just seen, sometimes these are not the same.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SUREF"
></A
><B
CLASS="COMMAND"
>su</B
></DT
><DD
><P
>Runs a program or script as a
<B
CLASS="COMMAND"
>s</B
>ubstitute <B
CLASS="COMMAND"
>u</B
>ser.
<B
CLASS="COMMAND"
>su rjones</B
> starts a shell as user
<EM
>rjones</EM
>. A naked <B
CLASS="COMMAND"
>su</B
>
defaults to <I
CLASS="FIRSTTERM"
>root</I
>. See <A
HREF="#FIFO"
>Example A-14</A
>.</P
></DD
><DT
><A
NAME="SUDOREF"
></A
><B
CLASS="COMMAND"
>sudo</B
></DT
><DD
><P
>Runs a command as <I
CLASS="FIRSTTERM"
>root</I
> (or
another user). This may be used in a script, thus permitting
a <I
CLASS="FIRSTTERM"
>regular user</I
> to run the script.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Some commands.
sudo cp /root/secretfile /home/bozo/secret
# Some more commands.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The file <TT
CLASS="FILENAME"
>/etc/sudoers</TT
> holds
the names of users permitted to invoke
<B
CLASS="COMMAND"
>sudo</B
>.</P
></DD
><DT
><A
NAME="PASSWDREF"
></A
><B
CLASS="COMMAND"
>passwd</B
></DT
><DD
><P
>Sets, changes, or manages a user's password.</P
><P
>The <B
CLASS="COMMAND"
>passwd</B
> command can be used in
a script, but probably <EM
>should not</EM
> be.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SETNEWPW"
></A
><P
><B
>Example 17-1. Setting a new password</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# setnew-password.sh: For demonstration purposes only.
# Not a good idea to actually run this script.
# This script must be run as root.
ROOT_UID=0 # Root has $UID 0.
E_WRONG_USER=65 # Not root?
E_NOSUCHUSER=70
SUCCESS=0
if [ "$UID" -ne "$ROOT_UID" ]
then
echo; echo "Only root can run this script."; echo
exit $E_WRONG_USER
else
echo
echo "You should know better than to run this script, root."
echo "Even root users get the blues... "
echo
fi
username=bozo
NEWPASSWORD=security_violation
# Check if bozo lives here.
grep -q "$username" /etc/passwd
if [ $? -ne $SUCCESS ]
then
echo "User $username does not exist."
echo "No password changed."
exit $E_NOSUCHUSER
fi
echo "$NEWPASSWORD" | passwd --stdin "$username"
# The '--stdin' option to 'passwd' permits
#+ getting a new password from stdin (or a pipe).
echo; echo "User $username's password changed!"
# Using the 'passwd' command in a script is dangerous.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The <B
CLASS="COMMAND"
>passwd</B
> command's <TT
CLASS="OPTION"
>-l</TT
>,
<TT
CLASS="OPTION"
>-u</TT
>, and <TT
CLASS="OPTION"
>-d</TT
> options permit
locking, unlocking, and deleting a user's password. Only
<I
CLASS="FIRSTTERM"
>root</I
> may use these options.</P
></DD
><DT
><A
NAME="ACREF"
></A
><B
CLASS="COMMAND"
>ac</B
></DT
><DD
><P
>Show users' logged in time, as read from
<TT
CLASS="FILENAME"
>/var/log/wtmp</TT
>. This is one of the GNU
accounting utilities.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ac</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> total 68.08</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="LASTREF"
></A
><B
CLASS="COMMAND"
>last</B
></DT
><DD
><P
>List <EM
>last</EM
> logged in users, as read from
<TT
CLASS="FILENAME"
>/var/log/wtmp</TT
>. This command can also
show remote logins.</P
><P
>For example, to show the last few times the system
rebooted:</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>last reboot</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>reboot system boot 2.6.9-1.667 Fri Feb 4 18:18 (00:02)
reboot system boot 2.6.9-1.667 Fri Feb 4 15:20 (01:27)
reboot system boot 2.6.9-1.667 Fri Feb 4 12:56 (00:49)
reboot system boot 2.6.9-1.667 Thu Feb 3 21:08 (02:17)
. . .
wtmp begins Tue Feb 1 12:50:09 2005</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="NEWGRPREF"
></A
><B
CLASS="COMMAND"
>newgrp</B
></DT
><DD
><P
>Change user's <I
CLASS="FIRSTTERM"
>group ID</I
> without
logging out. This permits access to the new group's
files. Since users may be members of multiple groups
simultaneously, this command finds only limited use.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Kurt Glaesemann points out that the
<I
CLASS="FIRSTTERM"
>newgrp</I
> command could prove helpful
in setting the default group permissions for files a user
writes. However, the <A
HREF="#CHGRPREF"
>chgrp</A
>
command might be more convenient for this purpose.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="TERMINALSSYS1"
></A
>Terminals</B
></P
><DL
><DT
><A
NAME="TTYREF"
></A
><B
CLASS="COMMAND"
>tty</B
></DT
><DD
><P
>Echoes the name (filename) of the current user's terminal.
Note that each separate <I
CLASS="FIRSTTERM"
>xterm</I
>
window counts as a different terminal.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>tty</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/dev/pts/1</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="STTYREF"
></A
><B
CLASS="COMMAND"
>stty</B
></DT
><DD
><P
>Shows and/or changes terminal settings. This complex
command, used in a script, can control terminal behavior
and the way output displays. See the info page, and study
it carefully.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ERASE"
></A
><P
><B
>Example 17-2. Setting an <I
CLASS="FIRSTTERM"
>erase</I
> character</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# erase.sh: Using "stty" to set an erase character when reading input.
echo -n "What is your name? "
read name # Try to backspace
#+ to erase characters of input.
# Problems?
echo "Your name is $name."
stty erase '#' # Set "hashmark" (#) as erase character.
echo -n "What is your name? "
read name # Use # to erase last character typed.
echo "Your name is $name."
exit 0
# Even after the script exits, the new key value remains set.
# Exercise: How would you reset the erase character to the default value?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="SECRETPW"
></A
><P
><B
>Example 17-3. <I
CLASS="FIRSTTERM"
>secret password</I
>:
Turning off terminal echoing </B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# secret-pw.sh: secret password
echo
echo -n "Enter password "
read passwd
echo "password is $passwd"
echo -n "If someone had been looking over your shoulder, "
echo "your password would have been compromised."
echo &#38;&#38; echo # Two line-feeds in an "and list."
stty -echo # Turns off screen echo.
# May also be done with
# read -sp passwd
# A big Thank You to Leigh James for pointing this out.
echo -n "Enter password again "
read passwd
echo
echo "password is $passwd"
echo
stty echo # Restores screen echo.
exit 0
# Do an 'info stty' for more on this useful-but-tricky command.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>A creative use of <B
CLASS="COMMAND"
>stty</B
> is detecting a
user keypress (without hitting
<B
CLASS="KEYCAP"
>ENTER</B
>).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="KEYPRESS"
></A
><P
><B
>Example 17-4. Keypress detection</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# keypress.sh: Detect a user keypress ("hot keys").
echo
old_tty_settings=$(stty -g) # Save old settings (why?).
stty -icanon
Keypress=$(head -c1) # or $(dd bs=1 count=1 2&#62; /dev/null)
# on non-GNU systems
echo
echo "Key pressed was \""$Keypress"\"."
echo
stty "$old_tty_settings" # Restore old settings.
# Thanks, Stephane Chazelas.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Also see <A
HREF="#TIMEOUT"
>Example 9-3</A
> and <A
HREF="#STOPWATCH"
>Example A-43</A
>.</P
><P
><A
NAME="TERMINALSREF"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN15053"
></A
><P
><B
>terminals and modes</B
></P
><P
>Normally, a terminal works in the
<I
CLASS="FIRSTTERM"
>canonical</I
> mode. When a user hits a
key, the resulting character does not immediately go to
the program actually running in this terminal. A buffer
local to the terminal stores keystrokes. When the user
hits the <B
CLASS="KEYCAP"
>ENTER</B
> key, this sends all the
stored keystrokes to the program running. There is even
a basic line editor inside the terminal.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>stty -a</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>speed 9600 baud; rows 36; columns 96; line = 0;
intr = ^C; quit = ^\; erase = ^H; kill = ^U; eof = ^D; eol = &#60;undef&#62;; eol2 = &#60;undef&#62;;
start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O;
...
isig icanon iexten echo echoe echok -echonl -noflsh -xcase -tostop -echoprt</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Using canonical mode, it is possible to redefine the
special keys for the local terminal line editor.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat &#62; filexxx</B
></TT
>
<TT
CLASS="USERINPUT"
><B
>wha&#60;ctl-W&#62;I&#60;ctl-H&#62;foo bar&#60;ctl-U&#62;hello world&#60;ENTER&#62;</B
></TT
>
<TT
CLASS="USERINPUT"
><B
>&#60;ctl-D&#62;</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat filexxx</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>hello world</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>wc -c &#60; filexxx</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>12</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
The process controlling the terminal receives only 12
characters (11 alphabetic ones, plus a newline), although
the user hit 26 keys.
</P
><P
>In non-canonical (<SPAN
CLASS="QUOTE"
>"raw"</SPAN
>) mode, every
key hit (including special editing keys such as
<B
CLASS="KEYCAP"
>ctl-H</B
>) sends a character immediately to
the controlling process.</P
><P
>The Bash prompt disables both <TT
CLASS="OPTION"
>icanon</TT
>
and <TT
CLASS="OPTION"
>echo</TT
>, since it replaces the basic
terminal line editor with its own more elaborate one. For
example, when you hit <B
CLASS="KEYCAP"
>ctl-A</B
> at the Bash
prompt, there's no <B
CLASS="KEYCAP"
>^A</B
> echoed by the
terminal, but Bash gets a <B
CLASS="KEYCAP"
>\1</B
> character,
interprets it, and moves the cursor to the begining of
the line.</P
><P
><EM
>St<53>phane Chazelas</EM
></P
></DIV
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="SETTERMREF"
></A
><B
CLASS="COMMAND"
>setterm</B
></DT
><DD
><P
>Set certain terminal attributes. This command writes
to its terminal's <TT
CLASS="FILENAME"
>stdout</TT
> a string that
changes the behavior of that terminal.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>setterm -cursor off</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bash$</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <B
CLASS="COMMAND"
>setterm</B
> command can be used within a
script to change the appearance of text written to
<TT
CLASS="FILENAME"
>stdout</TT
>, although there are certainly
<A
HREF="#COLORIZINGREF"
>better tools</A
> available
for this purpose.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>setterm -bold on
echo bold hello
setterm -bold off
echo normal hello</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="TSETREF"
></A
><B
CLASS="COMMAND"
>tset</B
></DT
><DD
><P
>Show or initialize terminal settings.
This is a less capable version of
<B
CLASS="COMMAND"
>stty</B
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>tset -r</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Terminal type is xterm-xfree86.
Kill is control-U (^U).
Interrupt is control-C (^C).</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="SETSERIALREF"
></A
><B
CLASS="COMMAND"
>setserial</B
></DT
><DD
><P
>Set or display serial port parameters. This command must be
run by <I
CLASS="FIRSTTERM"
>root</I
> and is usually found in a
system setup script.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># From /etc/pcmcia/serial script:
IRQ=`setserial /dev/$DEVICE | sed -e 's/.*IRQ: //'`
setserial /dev/$DEVICE irq 0 ; setserial /dev/$DEVICE irq $IRQ</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="GETTYREF"
></A
><B
CLASS="COMMAND"
>getty</B
>, <A
NAME="AGETTYREF"
></A
><B
CLASS="COMMAND"
>agetty</B
></DT
><DD
><P
>The initialization process for a terminal uses
<B
CLASS="COMMAND"
>getty</B
> or <B
CLASS="COMMAND"
>agetty</B
>
to set it up for login by a user. These commands are not
used within user shell scripts. Their scripting counterpart
is <B
CLASS="COMMAND"
>stty</B
>.</P
></DD
><DT
><A
NAME="MESGREF"
></A
><B
CLASS="COMMAND"
>mesg</B
></DT
><DD
><P
>Enables or disables write access to the current user's
terminal. Disabling access would prevent another user
on the network to <A
HREF="#WRITEREF"
>write</A
>
to the terminal.</P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>It can be quite annoying to have a message
about ordering pizza suddenly appear in the middle of
the text file you are editing. On a multi-user network,
you might therefore wish to disable write access to your
terminal when you need to avoid interruptions.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="WALLREF"
></A
><B
CLASS="COMMAND"
>wall</B
></DT
><DD
><P
>This is an acronym for <SPAN
CLASS="QUOTE"
>"<A
HREF="#WRITEREF"
>write</A
> all,"</SPAN
> i.e., sending
a message to all users at every terminal logged into the
network. It is primarily a system administrator's tool,
useful, for example, when warning everyone that the
system will shortly go down due to a problem (see <A
HREF="#EX70"
>Example 19-1</A
>).</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>wall System going down for maintenance in 5 minutes!</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Broadcast message from bozo (pts/1) Sun Jul 8 13:53:27 2001...
System going down for maintenance in 5 minutes!</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>If write access to a particular terminal has been
disabled with <B
CLASS="COMMAND"
>mesg</B
>, then
<B
CLASS="COMMAND"
>wall</B
> cannot send a message to
that terminal.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="STATISTICSSYS1"
></A
>Information and Statistics</B
></P
><DL
><DT
><A
NAME="UNAMEREF"
></A
><B
CLASS="COMMAND"
>uname</B
></DT
><DD
><P
>Output system specifications (OS, kernel version,
etc.) to <TT
CLASS="FILENAME"
>stdout</TT
>. Invoked with the
<TT
CLASS="OPTION"
>-a</TT
> option, gives verbose system info
(see <A
HREF="#EX41"
>Example 16-5</A
>). The <TT
CLASS="OPTION"
>-s</TT
>
option shows only the OS type.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>uname</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Linux</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>uname -s</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Linux</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>uname -a</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Linux iron.bozo 2.6.15-1.2054_FC5 #1 Tue Mar 14 15:48:33 EST 2006
i686 i686 i386 GNU/Linux</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="ARCHREF"
></A
><B
CLASS="COMMAND"
>arch</B
></DT
><DD
><P
>Show system architecture.
Equivalent to <B
CLASS="COMMAND"
>uname -m</B
>. See <A
HREF="#CASECMD"
>Example 11-27</A
>.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>arch</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>i686</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>uname -m</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>i686</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="LASTCOMMREF"
></A
><B
CLASS="COMMAND"
>lastcomm</B
></DT
><DD
><P
>Gives information about previous commands, as stored
in the <TT
CLASS="FILENAME"
>/var/account/pacct</TT
> file. Command
name and user name can be specified by options. This is
one of the GNU accounting utilities.</P
></DD
><DT
><A
NAME="LASTLOGREF"
></A
><B
CLASS="COMMAND"
>lastlog</B
></DT
><DD
><P
>List the last login time of all system users. This
references the <TT
CLASS="FILENAME"
>/var/log/lastlog</TT
>
file.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>lastlog</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>root tty1 Fri Dec 7 18:43:21 -0700 2001
bin **Never logged in**
daemon **Never logged in**
...
bozo tty1 Sat Dec 8 21:14:29 -0700 2001</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>lastlog | grep root</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>root tty1 Fri Dec 7 18:43:21 -0700 2001</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This command will fail if the user invoking
it does not have read permission for the
<TT
CLASS="FILENAME"
>/var/log/lastlog</TT
> file.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="LSOFREF"
></A
><B
CLASS="COMMAND"
>lsof</B
></DT
><DD
><P
>List open files. This command outputs a detailed
table of all currently open files and gives information
about their owner, size, the processes associated with
them, and more. Of course, <B
CLASS="COMMAND"
>lsof</B
> may
be piped to <A
HREF="#GREPREF"
>grep</A
> and/or
<A
HREF="#AWKREF"
>awk</A
> to parse and analyze
its results.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>lsof</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME
init 1 root mem REG 3,5 30748 30303 /sbin/init
init 1 root mem REG 3,5 73120 8069 /lib/ld-2.1.3.so
init 1 root mem REG 3,5 931668 8075 /lib/libc-2.1.3.so
cardmgr 213 root mem REG 3,5 36956 30357 /sbin/cardmgr
...</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <B
CLASS="COMMAND"
>lsof</B
> command is a useful,
if complex administrative tool. If you are unable to
dismount a filesystem and get an error message that it is
still in use, then running <I
CLASS="FIRSTTERM"
>lsof</I
> helps
determine which files are still open on that filesystem. The
<TT
CLASS="OPTION"
>-i</TT
> option lists open network socket files,
and this can help trace intrusion or hack attempts.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>lsof -an -i tcp</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME
firefox 2330 bozo 32u IPv4 9956 TCP 66.0.118.137:57596-&#62;67.112.7.104:http ...
firefox 2330 bozo 38u IPv4 10535 TCP 66.0.118.137:57708-&#62;216.79.48.24:http ...</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>See <A
HREF="#IPADDRESSES"
>Example 30-2</A
> for an effective use
of <B
CLASS="COMMAND"
>lsof</B
>.</P
></DD
><DT
><A
NAME="STRACEREF"
></A
><B
CLASS="COMMAND"
>strace</B
></DT
><DD
><P
><B
CLASS="COMMAND"
>S</B
>ystem <B
CLASS="COMMAND"
>trace</B
>:
diagnostic and debugging tool for tracing <I
CLASS="FIRSTTERM"
>system
calls</I
> and signals. This command and
<B
CLASS="COMMAND"
>ltrace</B
>, following, are useful for
diagnosing why a given program or package fails to
run . . . perhaps due to missing libraries or related
causes.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>strace df</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>execve("/bin/df", ["df"], [/* 45 vars */]) = 0
uname({sys="Linux", node="bozo.localdomain", ...}) = 0
brk(0) = 0x804f5e4
...</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>This is the Linux equivalent of
the Solaris <B
CLASS="COMMAND"
>truss</B
> command.</P
></DD
><DT
><A
NAME="LTRACEREF"
></A
><B
CLASS="COMMAND"
>ltrace</B
></DT
><DD
><P
><B
CLASS="COMMAND"
>L</B
>ibrary <B
CLASS="COMMAND"
>trace</B
>:
diagnostic and debugging tool that traces <I
CLASS="FIRSTTERM"
>library calls</I
>
invoked by a given command.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ltrace df</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>__libc_start_main(0x804a910, 1, 0xbfb589a4, 0x804fb70, 0x804fb68 &#60;unfinished ...&#62;:
setlocale(6, "") = "en_US.UTF-8"
bindtextdomain("coreutils", "/usr/share/locale") = "/usr/share/locale"
textdomain("coreutils") = "coreutils"
__cxa_atexit(0x804b650, 0, 0, 0x8052bf0, 0xbfb58908) = 0
getenv("DF_BLOCK_SIZE") = NULL
...</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="NCREF"
></A
><B
CLASS="COMMAND"
>nc</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>nc</B
> (<I
CLASS="FIRSTTERM"
>netcat</I
>)
utility is a complete toolkit for connecting to and
listening to TCP and UDP ports. It is useful as a diagnostic
and testing tool and as a component in simple script-based HTTP
clients and servers.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>nc localhost.localdomain 25</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>220 localhost.localdomain ESMTP Sendmail 8.13.1/8.13.1;
Thu, 31 Mar 2005 15:41:35 -0700</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>A real-life <A
HREF="#NETCATEXAMPLE"
>usage
example</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ISCAN"
></A
><P
><B
>Example 17-5. Checking a remote server for
<I
CLASS="FIRSTTERM"
>identd</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/sh
## Duplicate DaveG's ident-scan thingie using netcat. Oooh, he'll be p*ssed.
## Args: target port [port port port ...]
## Hose stdout _and_ stderr together.
##
## Advantages: runs slower than ident-scan, giving remote inetd less cause
##+ for alarm, and only hits the few known daemon ports you specify.
## Disadvantages: requires numeric-only port args, the output sleazitude,
##+ and won't work for r-services when coming from high source ports.
# Script author: Hobbit &#60;hobbit@avian.org&#62;
# Used in ABS Guide with permission.
# ---------------------------------------------------
E_BADARGS=65 # Need at least two args.
TWO_WINKS=2 # How long to sleep.
THREE_WINKS=3
IDPORT=113 # Authentication "tap ident" port.
RAND1=999
RAND2=31337
TIMEOUT0=9
TIMEOUT1=8
TIMEOUT2=4
# ---------------------------------------------------
case "${2}" in
"" ) echo "Need HOST and at least one PORT." ; exit $E_BADARGS ;;
esac
# Ping 'em once and see if they *are* running identd.
nc -z -w $TIMEOUT0 "$1" $IDPORT || \
{ echo "Oops, $1 isn't running identd." ; exit 0 ; }
# -z scans for listening daemons.
# -w $TIMEOUT = How long to try to connect.
# Generate a randomish base port.
RP=`expr $$ % $RAND1 + $RAND2`
TRG="$1"
shift
while test "$1" ; do
nc -v -w $TIMEOUT1 -p ${RP} "$TRG" ${1} &#60; /dev/null &#62; /dev/null &#38;
PROC=$!
sleep $THREE_WINKS
echo "${1},${RP}" | nc -w $TIMEOUT2 -r "$TRG" $IDPORT 2&#62;&#38;1
sleep $TWO_WINKS
# Does this look like a lamer script or what . . . ?
# ABS Guide author comments: "Ain't really all that bad . . .
#+ kinda clever, actually."
kill -HUP $PROC
RP=`expr ${RP} + 1`
shift
done
exit $?
# Notes:
# -----
# Try commenting out line 30 and running this script
#+ with "localhost.localdomain 25" as arguments.
# For more of Hobbit's 'nc' example scripts,
#+ look in the documentation:
#+ the /usr/share/doc/nc-X.XX/scripts directory.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
> And, of course, there's Dr. Andrew Tridgell's notorious
one-line script in the BitKeeper Affair:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo clone | nc thunk.org 5000 &#62; e2fsprogs.dat</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="FREEREF"
></A
><B
CLASS="COMMAND"
>free</B
></DT
><DD
><P
>Shows memory and cache usage in tabular form. The
output of this command lends itself to parsing, using
<A
HREF="#GREPREF"
>grep</A
>, <A
HREF="#AWKREF"
>awk</A
> or <B
CLASS="COMMAND"
>Perl</B
>. The
<B
CLASS="COMMAND"
>procinfo</B
> command shows all the
information that <B
CLASS="COMMAND"
>free</B
> does, and much
more.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><B
CLASS="COMMAND"
>free</B
>
<TT
CLASS="COMPUTEROUTPUT"
> total used free shared buffers cached
Mem: 30504 28624 1880 15820 1608 16376
-/+ buffers/cache: 10640 19864
Swap: 68540 3128 65412</TT
></PRE
></FONT
></TD
></TR
></TABLE
><P
>To show unused RAM memory:</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><B
CLASS="COMMAND"
>free | grep Mem | awk '{ print $4 }'</B
>
<TT
CLASS="COMPUTEROUTPUT"
>1880</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="PROCINFOREF"
></A
><B
CLASS="COMMAND"
>procinfo</B
></DT
><DD
><P
>Extract and list information and statistics from the
<A
HREF="#DEVPROCREF"
><TT
CLASS="FILENAME"
>/proc</TT
>
pseudo-filesystem</A
>. This gives a very extensive and
detailed listing.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>procinfo | grep Bootup</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Bootup: Wed Mar 21 15:15:50 2001 Load average: 0.04 0.21 0.34 3/47 6829</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="LSDEVREF"
></A
><B
CLASS="COMMAND"
>lsdev</B
></DT
><DD
><P
>List devices, that is, show installed hardware.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>lsdev</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Device DMA IRQ I/O Ports
------------------------------------------------
cascade 4 2
dma 0080-008f
dma1 0000-001f
dma2 00c0-00df
fpu 00f0-00ff
ide0 14 01f0-01f7 03f6-03f6
...</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="DUREF"
></A
><B
CLASS="COMMAND"
>du</B
></DT
><DD
><P
>Show (disk) file usage, recursively. Defaults to current
working directory, unless otherwise specified.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><B
CLASS="COMMAND"
>du -ach</B
>
<TT
CLASS="COMPUTEROUTPUT"
>1.0k ./wi.sh
1.0k ./tst.sh
1.0k ./random.file
6.0k .
6.0k total</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="DFREF"
></A
><B
CLASS="COMMAND"
>df</B
></DT
><DD
><P
>Shows filesystem usage in tabular form.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><B
CLASS="COMMAND"
>df</B
>
<TT
CLASS="COMPUTEROUTPUT"
>Filesystem 1k-blocks Used Available Use% Mounted on
/dev/hda5 273262 92607 166547 36% /
/dev/hda8 222525 123951 87085 59% /home
/dev/hda7 1408796 1075744 261488 80% /usr</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="DMESGREF"
></A
><B
CLASS="COMMAND"
>dmesg</B
></DT
><DD
><P
>Lists all system bootup messages to
<TT
CLASS="FILENAME"
>stdout</TT
>. Handy for debugging and
ascertaining which device drivers were installed
and which system interrupts in use. The output
of <B
CLASS="COMMAND"
>dmesg</B
> may, of course, be
parsed with <A
HREF="#GREPREF"
>grep</A
>,
<A
HREF="#SEDREF"
>sed</A
>, or <A
HREF="#AWKREF"
>awk</A
> from within a script.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>dmesg | grep hda</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Kernel command line: ro root=/dev/hda2
hda: IBM-DLGA-23080, ATA DISK drive
hda: 6015744 sectors (3080 MB) w/96KiB Cache, CHS=746/128/63
hda: hda1 hda2 hda3 &#60; hda5 hda6 hda7 &#62; hda4</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="STATREF"
></A
><B
CLASS="COMMAND"
>stat</B
></DT
><DD
><P
>Gives detailed and verbose <EM
>stat</EM
>istics
on a given file (even a directory or device file) or set
of files.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>stat test.cru</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> File: "test.cru"
Size: 49970 Allocated Blocks: 100 Filetype: Regular File
Mode: (0664/-rw-rw-r--) Uid: ( 501/ bozo) Gid: ( 501/ bozo)
Device: 3,8 Inode: 18185 Links: 1
Access: Sat Jun 2 16:40:24 2001
Modify: Sat Jun 2 16:40:24 2001
Change: Sat Jun 2 16:40:24 2001</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>If the target file does not exist, <B
CLASS="COMMAND"
>stat</B
>
returns an error message.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>stat nonexistent-file</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>nonexistent-file: No such file or directory</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>In a script, you can use <B
CLASS="COMMAND"
>stat</B
> to extract
information about files (and filesystems) and set variables
accordingly.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# fileinfo2.sh
# Per suggestion of Jo<4A>l Bourquard and . . .
# http://www.linuxquestions.org/questions/showthread.php?t=410766
FILENAME=testfile.txt
file_name=$(stat -c%n "$FILENAME") # Same as "$FILENAME" of course.
file_owner=$(stat -c%U "$FILENAME")
file_size=$(stat -c%s "$FILENAME")
# Certainly easier than using "ls -l $FILENAME"
#+ and then parsing with sed.
file_inode=$(stat -c%i "$FILENAME")
file_type=$(stat -c%F "$FILENAME")
file_access_rights=$(stat -c%A "$FILENAME")
echo "File name: $file_name"
echo "File owner: $file_owner"
echo "File size: $file_size"
echo "File inode: $file_inode"
echo "File type: $file_type"
echo "File access rights: $file_access_rights"
exit 0
sh fileinfo2.sh
File name: testfile.txt
File owner: bozo
File size: 418
File inode: 1730378
File type: regular file
File access rights: -rw-rw-r--</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="VMSTATREF"
></A
><B
CLASS="COMMAND"
>vmstat</B
></DT
><DD
><P
>Display virtual memory statistics.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>vmstat</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> procs memory swap io system cpu
r b w swpd free buff cache si so bi bo in cs us sy id
0 0 0 0 11040 2636 38952 0 0 33 7 271 88 8 3 89</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="UPTIMEREF"
></A
><B
CLASS="COMMAND"
>uptime</B
></DT
><DD
><P
>Shows how long the system has been running, along with
associated statistics.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>uptime</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>10:28pm up 1:57, 3 users, load average: 0.17, 0.34, 0.27</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A <I
CLASS="FIRSTTERM"
>load average</I
> of 1 or less
indicates that the system handles processes immediately. A load
average greater than 1 means that processes are being queued. When
the load average gets above 3 (on a single-core processor),
then system performance is significantly degraded.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="HNAMEREF"
></A
><B
CLASS="COMMAND"
>hostname</B
></DT
><DD
><P
>Lists the system's host name. This command sets the host
name in an <TT
CLASS="FILENAME"
>/etc/rc.d</TT
>
setup script (<TT
CLASS="FILENAME"
>/etc/rc.d/rc.sysinit</TT
>
or similar). It is equivalent to <B
CLASS="COMMAND"
>uname
-n</B
>, and a counterpart to the <A
HREF="#HOSTNAMEREF"
>$HOSTNAME</A
> internal
variable.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>hostname</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>localhost.localdomain</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $HOSTNAME</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>localhost.localdomain</TT
></PRE
></FONT
></TD
></TR
></TABLE
><P
>Similar to the <B
CLASS="COMMAND"
>hostname</B
> command are the
<B
CLASS="COMMAND"
>domainname</B
>,
<B
CLASS="COMMAND"
>dnsdomainname</B
>,
<B
CLASS="COMMAND"
>nisdomainname</B
>, and
<B
CLASS="COMMAND"
>ypdomainname</B
> commands. Use these to
display or set the system DNS or NIS/YP domain name. Various
options to <B
CLASS="COMMAND"
>hostname</B
> also perform these
functions.</P
></DD
><DT
><A
NAME="HOSTIDREF"
></A
><B
CLASS="COMMAND"
>hostid</B
></DT
><DD
><P
>Echo a 32-bit hexadecimal numerical identifier for the
host machine.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>hostid</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>7f0100</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This command allegedly fetches a <SPAN
CLASS="QUOTE"
>"unique"</SPAN
>
serial number for a particular system. Certain
product registration procedures use this number
to brand a particular user license. Unfortunately,
<B
CLASS="COMMAND"
>hostid</B
> only returns the machine
network address in hexadecimal, with pairs of bytes
transposed.</P
><P
>The network address of a typical non-networked Linux
machine, is found in <TT
CLASS="FILENAME"
>/etc/hosts</TT
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat /etc/hosts</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>127.0.0.1 localhost.localdomain localhost</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>As it happens, transposing the bytes of
<TT
CLASS="USERINPUT"
><B
>127.0.0.1</B
></TT
>, we get
<TT
CLASS="USERINPUT"
><B
>0.127.1.0</B
></TT
>, which translates in
hex to <TT
CLASS="USERINPUT"
><B
>007f0100</B
></TT
>, the exact equivalent
of what <B
CLASS="COMMAND"
>hostid</B
> returns, above. There
exist only a few million other Linux machines with this
identical <I
CLASS="FIRSTTERM"
>hostid</I
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SARREF"
></A
><B
CLASS="COMMAND"
>sar</B
></DT
><DD
><P
>Invoking <B
CLASS="COMMAND"
>sar</B
> (System Activity Reporter)
gives a very detailed rundown on system statistics. The
Santa Cruz Operation (<SPAN
CLASS="QUOTE"
>"Old"</SPAN
> SCO) released
<B
CLASS="COMMAND"
>sar</B
> as Open Source in June, 1999.</P
><P
>This command is not part of the base Linux distribution,
but may be obtained as part of the<A
HREF="http://perso.wanadoo.fr/sebastien.godard/"
TARGET="_top"
> sysstat utilities</A
> package, written by <A
HREF="mailto:sebastien.godard@wanadoo.fr"
TARGET="_top"
>Sebastien
Godard</A
>.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>sar</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Linux 2.4.9 (brooks.seringas.fr) 09/26/03
10:30:00 CPU %user %nice %system %iowait %idle
10:40:00 all 2.21 10.90 65.48 0.00 21.41
10:50:00 all 3.36 0.00 72.36 0.00 24.28
11:00:00 all 1.12 0.00 80.77 0.00 18.11
Average: all 2.23 3.63 72.87 0.00 21.27
14:32:30 LINUX RESTART
15:00:00 CPU %user %nice %system %iowait %idle
15:10:00 all 8.59 2.40 17.47 0.00 71.54
15:20:00 all 4.07 1.00 11.95 0.00 82.98
15:30:00 all 0.79 2.94 7.56 0.00 88.71
Average: all 6.33 1.70 14.71 0.00 77.26</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="READELFREF"
></A
><B
CLASS="COMMAND"
>readelf</B
></DT
><DD
><P
>Show information and statistics about a designated
<I
CLASS="FIRSTTERM"
>elf</I
> binary. This is part of the
<I
CLASS="FIRSTTERM"
>binutils</I
> package.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>readelf -h /bin/bash</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>ELF Header:
Magic: 7f 45 4c 46 01 01 01 00 00 00 00 00 00 00 00 00
Class: ELF32
Data: 2's complement, little endian
Version: 1 (current)
OS/ABI: UNIX - System V
ABI Version: 0
Type: EXEC (Executable file)
. . .</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="SIZEREF"
></A
><B
CLASS="COMMAND"
>size</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>size [/path/to/binary]</B
> command
gives the segment sizes of a binary executable or archive file.
This is mainly of use to programmers.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>size /bin/bash</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> text data bss dec hex filename
495971 22496 17392 535859 82d33 /bin/bash</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="SYSLOG1"
></A
>System Logs</B
></P
><DL
><DT
><A
NAME="LOGGERREF"
></A
><B
CLASS="COMMAND"
>logger</B
></DT
><DD
><P
>Appends a user-generated message to the system log
(<TT
CLASS="FILENAME"
>/var/log/messages</TT
>). You do not have
to be <I
CLASS="FIRSTTERM"
>root</I
> to invoke
<B
CLASS="COMMAND"
>logger</B
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>logger Experiencing instability in network connection at 23:10, 05/21.
# Now, do a 'tail /var/log/messages'.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>By embedding a <B
CLASS="COMMAND"
>logger</B
> command in a script,
it is possible to write debugging information to
<TT
CLASS="FILENAME"
>/var/log/messages</TT
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>logger -t $0 -i Logging at line "$LINENO".
# The "-t" option specifies the tag for the logger entry.
# The "-i" option records the process ID.
# tail /var/log/message
# ...
# Jul 7 20:48:58 localhost ./test.sh[1712]: Logging at line 3.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="LOGROTATEREF"
></A
><B
CLASS="COMMAND"
>logrotate</B
></DT
><DD
><P
>This utility manages the system log files, rotating,
compressing, deleting, and/or e-mailing them, as appropriate.
This keeps the <TT
CLASS="FILENAME"
>/var/log</TT
>
from getting cluttered with old log files.
Usually <A
HREF="#CRONREF"
>cron</A
> runs
<B
CLASS="COMMAND"
>logrotate</B
> on a daily basis.</P
><P
>Adding an appropriate entry to
<TT
CLASS="FILENAME"
>/etc/logrotate.conf</TT
> makes it possible
to manage personal log files, as well as system-wide
ones.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Stefano Falsetto has created <A
HREF="http://www.gnu.org/software/rottlog/"
TARGET="_top"
>rottlog</A
>,
which he considers to be an improved version of
<B
CLASS="COMMAND"
>logrotate</B
>.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="JOBCONTROLSYS1"
></A
>Job Control</B
></P
><DL
><DT
><A
NAME="PPSSREF"
></A
><B
CLASS="COMMAND"
>ps</B
></DT
><DD
><P
><TT
CLASS="REPLACEABLE"
><I
>P</I
></TT
>rocess
<TT
CLASS="REPLACEABLE"
><I
>S</I
></TT
>tatistics: lists currently
executing processes by owner and PID (process ID). This
is usually invoked with <TT
CLASS="OPTION"
>ax</TT
> or
<TT
CLASS="OPTION"
>aux</TT
> options,
and may be piped to <A
HREF="#GREPREF"
>grep</A
>
or <A
HREF="#SEDREF"
>sed</A
> to search for a
specific process (see <A
HREF="#EX44"
>Example 15-14</A
> and <A
HREF="#PIDID"
>Example 29-3</A
>).</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
> ps ax | grep sendmail</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>295 ? S 0:00 sendmail: accepting connections on port 25</TT
></PRE
></FONT
></TD
></TR
></TABLE
><P
>To display system processes in graphical <SPAN
CLASS="QUOTE"
>"tree"</SPAN
>
format: <B
CLASS="COMMAND"
>ps afjx</B
> or
<B
CLASS="COMMAND"
>ps ax --forest</B
>.</P
></DD
><DT
><A
NAME="PGREPREF"
></A
><B
CLASS="COMMAND"
>pgrep</B
>, <A
NAME="PKILLREF"
></A
><B
CLASS="COMMAND"
>pkill</B
></DT
><DD
><P
>Combining the <B
CLASS="COMMAND"
>ps</B
> command
with <A
HREF="#GREPREF"
>grep</A
> or
<A
HREF="#KILLREF"
>kill</A
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ps a | grep mingetty</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>2212 tty2 Ss+ 0:00 /sbin/mingetty tty2
2213 tty3 Ss+ 0:00 /sbin/mingetty tty3
2214 tty4 Ss+ 0:00 /sbin/mingetty tty4
2215 tty5 Ss+ 0:00 /sbin/mingetty tty5
2216 tty6 Ss+ 0:00 /sbin/mingetty tty6
4849 pts/2 S+ 0:00 grep mingetty</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>pgrep mingetty</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>2212 mingetty
2213 mingetty
2214 mingetty
2215 mingetty
2216 mingetty</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Compare the action of <B
CLASS="COMMAND"
>pkill</B
> with <A
HREF="#KILLALLREF"
>killall</A
>.</P
></DD
><DT
><A
NAME="PSTREEREF"
></A
><B
CLASS="COMMAND"
>pstree</B
></DT
><DD
><P
>Lists currently executing processes in
<SPAN
CLASS="QUOTE"
>"tree"</SPAN
> format. The <TT
CLASS="OPTION"
>-p</TT
> option
shows the PIDs, as well as the process names.</P
></DD
><DT
><A
NAME="TOPREF"
></A
><B
CLASS="COMMAND"
>top</B
></DT
><DD
><P
>Continuously updated display of most cpu-intensive
processes. The <TT
CLASS="OPTION"
>-b</TT
> option displays in text
mode, so that the output may be parsed or accessed from
a script.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>top -b</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> 8:30pm up 3 min, 3 users, load average: 0.49, 0.32, 0.13
45 processes: 44 sleeping, 1 running, 0 zombie, 0 stopped
CPU states: 13.6% user, 7.3% system, 0.0% nice, 78.9% idle
Mem: 78396K av, 65468K used, 12928K free, 0K shrd, 2352K buff
Swap: 157208K av, 0K used, 157208K free 37244K cached
PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND
848 bozo 17 0 996 996 800 R 5.6 1.2 0:00 top
1 root 8 0 512 512 444 S 0.0 0.6 0:04 init
2 root 9 0 0 0 0 SW 0.0 0.0 0:00 keventd
...</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="NICEREF"
></A
><B
CLASS="COMMAND"
>nice</B
></DT
><DD
><P
><A
NAME="NICE2REF"
></A
></P
><P
>Run a background job with an altered
priority. Priorities run from 19 (lowest) to -20
(highest). Only <I
CLASS="FIRSTTERM"
>root</I
> may set the
negative (higher) priorities. Related commands are
<B
CLASS="COMMAND"
>renice</B
> and <B
CLASS="COMMAND"
>snice</B
>,
which change the priority of a running process or
processes, and <B
CLASS="COMMAND"
>skill</B
>, which sends a
<A
HREF="#KILLREF"
>kill</A
> signal to a process
or processes.</P
></DD
><DT
><A
NAME="NOHUPREF"
></A
><B
CLASS="COMMAND"
>nohup</B
></DT
><DD
><P
>Keeps a command running even after user logs off.
The command will run as a foreground process unless followed
by <SPAN
CLASS="TOKEN"
>&#38;</SPAN
>. If you use <B
CLASS="COMMAND"
>nohup</B
>
within a script, consider coupling it with a <A
HREF="#WAITREF"
>wait</A
> to avoid creating an
<I
CLASS="FIRSTTERM"
>orphan</I
> or
<A
HREF="#ZOMBIEREF"
>zombie</A
> process.</P
></DD
><DT
><A
NAME="PIDOFREF"
></A
><B
CLASS="COMMAND"
>pidof</B
></DT
><DD
><P
>Identifies <I
CLASS="FIRSTTERM"
>process ID (PID)</I
> of a
running job. Since job control commands, such as <A
HREF="#KILLREF"
>kill</A
> and <A
HREF="#NICE2REF"
>renice</A
> act on the
<I
CLASS="FIRSTTERM"
>PID</I
> of a process (not its
name), it is sometimes necessary to identify that
<I
CLASS="FIRSTTERM"
>PID</I
>. The <B
CLASS="COMMAND"
>pidof</B
>
command is the approximate counterpart to the <A
HREF="#PPIDREF"
>$PPID</A
> internal variable.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>pidof xclock</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>880</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="KILLPROCESS"
></A
><P
><B
>Example 17-6. <I
CLASS="FIRSTTERM"
>pidof</I
> helps kill a process</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# kill-process.sh
NOPROCESS=2
process=xxxyyyzzz # Use nonexistent process.
# For demo purposes only...
# ... don't want to actually kill any actual process with this script.
#
# If, for example, you wanted to use this script to logoff the Internet,
# process=pppd
t=`pidof $process` # Find pid (process id) of $process.
# The pid is needed by 'kill' (can't 'kill' by program name).
if [ -z "$t" ] # If process not present, 'pidof' returns null.
then
echo "Process $process was not running."
echo "Nothing killed."
exit $NOPROCESS
fi
kill $t # May need 'kill -9' for stubborn process.
# Need a check here to see if process allowed itself to be killed.
# Perhaps another " t=`pidof $process` " or ...
# This entire script could be replaced by
# kill $(pidof -x process_name)
# or
# killall process_name
# but it would not be as instructive.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="FUSERREF"
></A
><B
CLASS="COMMAND"
>fuser</B
></DT
><DD
><P
>Identifies the processes (by PID) that are accessing
a given file, set of files, or directory. May also be
invoked with the <TT
CLASS="OPTION"
>-k</TT
> option, which kills
those processes. This has interesting implications for
system security, especially in scripts preventing
unauthorized users from accessing system services.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>fuser -u /usr/bin/vim</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/usr/bin/vim: 3207e(bozo)</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>fuser -u /dev/null</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/dev/null: 3009(bozo) 3010(bozo) 3197(bozo) 3199(bozo)</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>One important application for <B
CLASS="COMMAND"
>fuser</B
> is
when physically inserting or removing storage media, such
as CD ROM disks or USB flash drives. Sometimes trying
a <A
HREF="#UMOUNTREF"
>umount</A
> fails with a
<SPAN
CLASS="ERRORNAME"
>device is busy</SPAN
> error message. This
means that some user(s) and/or process(es) are accessing
the device. An <B
CLASS="COMMAND"
>fuser -um /dev/device_name</B
>
will clear up the mystery, so you can kill any relevant
processes.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>umount /mnt/usbdrive</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>umount: /mnt/usbdrive: device is busy</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>fuser -um /dev/usbdrive</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/mnt/usbdrive: 1772c(bozo)</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>kill -9 1772</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>umount /mnt/usbdrive</B
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> The <B
CLASS="COMMAND"
>fuser</B
> command, invoked with the
<TT
CLASS="OPTION"
>-n</TT
> option identifies the processes
accessing a <I
CLASS="FIRSTTERM"
>port</I
>. This
is especially useful in combination with <A
HREF="#NMAPREF"
>nmap</A
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>nmap localhost.localdomain</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>PORT STATE SERVICE
25/tcp open smtp</TT
>
<TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>fuser -un tcp 25</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>25/tcp: 2095(root)</TT
>
<TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>ps ax | grep 2095 | grep -v grep</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>2095 ? Ss 0:00 sendmail: accepting connections</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="CRONREF"
></A
><B
CLASS="COMMAND"
>cron</B
></DT
><DD
><P
>Administrative program scheduler, performing such
duties as cleaning up and deleting system log files and
updating the <SPAN
CLASS="DATABASE"
>slocate</SPAN
> database. This
is the <I
CLASS="FIRSTTERM"
>superuser</I
> version of <A
HREF="#ATREF"
>at</A
> (although each user may have
their own <TT
CLASS="FILENAME"
>crontab</TT
> file which can be
changed with the <B
CLASS="COMMAND"
>crontab</B
> command).
It runs as a <A
HREF="#DAEMONREF"
>daemon</A
>
and executes scheduled entries from
<TT
CLASS="FILENAME"
>/etc/crontab</TT
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Some flavors of Linux run
<B
CLASS="COMMAND"
>crond</B
>, Matthew Dillon's version of
<B
CLASS="COMMAND"
>cron</B
>.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="RUNCONTROLSYS1"
></A
>Process Control and Booting</B
></P
><DL
><DT
><A
NAME="INITREF"
></A
><B
CLASS="COMMAND"
>init</B
></DT
><DD
><P
><A
NAME="INITTABREF"
></A
></P
><P
>The <B
CLASS="COMMAND"
>init</B
> command is the <A
HREF="#FORKREF"
>parent</A
> of all processes. Called
in the final step of a bootup, <B
CLASS="COMMAND"
>init</B
>
determines the runlevel of the system from
<TT
CLASS="FILENAME"
>/etc/inittab</TT
>. Invoked by its alias
<B
CLASS="COMMAND"
>telinit</B
>, and by
<I
CLASS="FIRSTTERM"
>root</I
> only.</P
></DD
><DT
><A
NAME="TELINITREF"
></A
><B
CLASS="COMMAND"
>telinit</B
></DT
><DD
><P
>Symlinked to <B
CLASS="COMMAND"
>init</B
>, this is a means of changing the system runlevel,
usually done for system maintenance or emergency filesystem
repairs. Invoked only by <I
CLASS="FIRSTTERM"
>root</I
>. This
command can be dangerous -- be certain you understand it
well before using!</P
></DD
><DT
><A
NAME="RUNLEVELREF"
></A
><B
CLASS="COMMAND"
>runlevel</B
></DT
><DD
><P
>Shows the current and last runlevel, that is, whether the system
is halted (runlevel <TT
CLASS="LITERAL"
>0</TT
>), in single-user mode
(<TT
CLASS="LITERAL"
>1</TT
>), in multi-user mode (<TT
CLASS="LITERAL"
>2</TT
>
or <TT
CLASS="LITERAL"
>3</TT
>), in X Windows (<TT
CLASS="LITERAL"
>5</TT
>), or
rebooting (<TT
CLASS="LITERAL"
>6</TT
>). This command accesses the
<TT
CLASS="FILENAME"
>/var/run/utmp</TT
> file.</P
></DD
><DT
><A
NAME="HALTREF"
></A
><B
CLASS="COMMAND"
>halt</B
>, <A
NAME="SHUTDOWNREF"
></A
><B
CLASS="COMMAND"
>shutdown</B
>, <A
NAME="REBOOTREF"
></A
><B
CLASS="COMMAND"
>reboot</B
></DT
><DD
><P
>Command set to shut the system down, usually just prior to a power down.</P
><DIV
CLASS="WARNING"
><P
></P
><TABLE
CLASS="WARNING"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/warning.gif"
HSPACE="5"
ALT="Warning"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>On some Linux distros, the <B
CLASS="COMMAND"
>halt</B
> command
has 755 permissions, so it can be invoked by a non-root user.
A careless <I
CLASS="FIRSTTERM"
>halt</I
> in a terminal or a script
may shut down the system!</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="SERVICEREF"
></A
><B
CLASS="COMMAND"
>service</B
></DT
><DD
><P
>Starts or stops a system <I
CLASS="FIRSTTERM"
>service</I
>.
The startup scripts in <TT
CLASS="FILENAME"
>/etc/init.d</TT
>
and <TT
CLASS="FILENAME"
>/etc/rc.d</TT
> use this
command to start services at bootup.</P
><P
><A
NAME="IPTABLES01"
></A
></P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>root# </TT
><TT
CLASS="USERINPUT"
><B
>/sbin/service iptables stop</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Flushing firewall rules: [ OK ]
Setting chains to policy ACCEPT: filter [ OK ]
Unloading iptables modules: [ OK ]</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="NETWORKSYS1"
></A
>Network</B
></P
><DL
><DT
><A
NAME="NMAPREF"
></A
><B
CLASS="COMMAND"
>nmap</B
></DT
><DD
><P
><B
CLASS="COMMAND"
>N</B
>etwork <B
CLASS="COMMAND"
>map</B
>per
and port scanner. This command scans a server to
locate open ports and the services associated with those
ports. It can also report information about packet filters and
firewalls. This is an important security tool for locking down
a network against hacking attempts.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
SERVER=$HOST # localhost.localdomain (127.0.0.1).
PORT_NUMBER=25 # SMTP port.
nmap $SERVER | grep -w "$PORT_NUMBER" # Is that particular port open?
# grep -w matches whole words only,
#+ so this wouldn't match port 1025, for example.
exit 0
# 25/tcp open smtp</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="IFCONFIGREF"
></A
><B
CLASS="COMMAND"
>ifconfig</B
></DT
><DD
><P
>Network <I
CLASS="FIRSTTERM"
>interface configuration</I
>
and tuning utility.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ifconfig -a</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:10 errors:0 dropped:0 overruns:0 frame:0
TX packets:10 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:700 (700.0 b) TX bytes:700 (700.0 b)</TT
></PRE
></FONT
></TD
></TR
></TABLE
><P
>The <B
CLASS="COMMAND"
>ifconfig</B
> command is most often used
at bootup to set up the interfaces, or to shut them down
when rebooting.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Code snippets from /etc/rc.d/init.d/network
# ...
# Check that networking is up.
[ ${NETWORKING} = "no" ] &#38;&#38; exit 0
[ -x /sbin/ifconfig ] || exit 0
# ...
for i in $interfaces ; do
if ifconfig $i 2&#62;/dev/null | grep -q "UP" &#62;/dev/null 2&#62;&#38;1 ; then
action "Shutting down interface $i: " ./ifdown $i boot
fi
# The GNU-specific "-q" option to "grep" means "quiet", i.e.,
#+ producing no output.
# Redirecting output to /dev/null is therefore not strictly necessary.
# ...
echo "Currently active devices:"
echo `/sbin/ifconfig | grep ^[a-z] | awk '{print $1}'`
# ^^^^^ should be quoted to prevent globbing.
# The following also work.
# echo $(/sbin/ifconfig | awk '/^[a-z]/ { print $1 })'
# echo $(/sbin/ifconfig | sed -e 's/ .*//')
# Thanks, S.C., for additional comments.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>See also <A
HREF="#ONLINE"
>Example 32-6</A
>.</P
></DD
><DT
><A
NAME="NETSTATREF"
></A
><B
CLASS="COMMAND"
>netstat</B
></DT
><DD
><P
>Show current network statistics and information,
such as routing tables and active connections. This utility
accesses information in <TT
CLASS="FILENAME"
>/proc/net</TT
>
(<A
HREF="#DEVPROC"
>Chapter 29</A
>). See <A
HREF="#CONSTAT"
>Example 29-4</A
>.</P
><P
><B
CLASS="COMMAND"
>netstat -r</B
> is equivalent to <A
HREF="#ROUTEREF"
>route</A
>.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>netstat</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Active Internet connections (w/o servers)
Proto Recv-Q Send-Q Local Address Foreign Address State
Active UNIX domain sockets (w/o servers)
Proto RefCnt Flags Type State I-Node Path
unix 11 [ ] DGRAM 906 /dev/log
unix 3 [ ] STREAM CONNECTED 4514 /tmp/.X11-unix/X0
unix 3 [ ] STREAM CONNECTED 4513
. . .</TT
></PRE
></FONT
></TD
></TR
></TABLE
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A <B
CLASS="COMMAND"
>netstat -lptu</B
> shows <A
HREF="#SOCKETREF"
>sockets</A
> that are listening
to ports, and the associated processes. This can be useful
for determining whether a computer has been hacked or
compromised.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="IWCONFIGREF"
></A
><B
CLASS="COMMAND"
>iwconfig</B
></DT
><DD
><P
>This is the command set for configuring a wireless network.
It is the wireless equivalent of <B
CLASS="COMMAND"
>ifconfig</B
>,
above.</P
></DD
><DT
><A
NAME="IPREF"
></A
><B
CLASS="COMMAND"
>ip</B
></DT
><DD
><P
>General purpose utility for setting up, changing, and
analyzing <I
CLASS="FIRSTTERM"
>IP</I
> (Internet Protocol)
networks and attached devices. This command is part of
the <I
CLASS="FIRSTTERM"
>iproute2</I
> package.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ip link show</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>1: lo: &#60;LOOPBACK,UP&#62; mtu 16436 qdisc noqueue
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: eth0: &#60;BROADCAST,MULTICAST&#62; mtu 1500 qdisc pfifo_fast qlen 1000
link/ether 00:d0:59:ce:af:da brd ff:ff:ff:ff:ff:ff
3: sit0: &#60;NOARP&#62; mtu 1480 qdisc noop
link/sit 0.0.0.0 brd 0.0.0.0</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ip route list</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>169.254.0.0/16 dev lo scope link</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Or, in a script:</P
><P
><A
NAME="IPSCRIPT0"
></A
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Script by Juan Nicolas Ruiz
# Used with his kind permission.
# Setting up (and stopping) a GRE tunnel.
# --- start-tunnel.sh ---
LOCAL_IP="192.168.1.17"
REMOTE_IP="10.0.5.33"
OTHER_IFACE="192.168.0.100"
REMOTE_NET="192.168.3.0/24"
/sbin/ip tunnel add netb mode gre remote $REMOTE_IP \
local $LOCAL_IP ttl 255
/sbin/ip addr add $OTHER_IFACE dev netb
/sbin/ip link set netb up
/sbin/ip route add $REMOTE_NET dev netb
exit 0 #############################################
# --- stop-tunnel.sh ---
REMOTE_NET="192.168.3.0/24"
/sbin/ip route del $REMOTE_NET dev netb
/sbin/ip link set netb down
/sbin/ip tunnel del netb
exit 0</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="ROUTEREF"
></A
><B
CLASS="COMMAND"
>route</B
></DT
><DD
><P
>Show info about or make changes to the kernel routing table.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>route</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Destination Gateway Genmask Flags MSS Window irtt Iface
pm3-67.bozosisp * 255.255.255.255 UH 40 0 0 ppp0
127.0.0.0 * 255.0.0.0 U 40 0 0 lo
default pm3-67.bozosisp 0.0.0.0 UG 40 0 0 ppp0</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="IPTABLESREF"
></A
><B
CLASS="COMMAND"
>iptables</B
></DT
><DD
><P
> The <B
CLASS="COMMAND"
>iptables</B
> command set is
a packet filtering tool used mainly for such security
purposes as setting up network firewalls. This
is a complex tool, and a detailed explanation of
its use is beyond the scope of this document. <A
HREF="http://www.frozentux.net/iptables-tutorial/iptables-tutorial.html"
TARGET="_top"
>Oskar
Andreasson's tutorial</A
> is a reasonable starting
point.</P
><P
>See also <A
HREF="#IPTABLES01"
>shutting down
<I
CLASS="FIRSTTERM"
>iptables</I
></A
> and <A
HREF="#IPADDRESSES"
>Example 30-2</A
>.</P
></DD
><DT
><A
NAME="CHKCONFIGREF"
></A
><B
CLASS="COMMAND"
>chkconfig</B
></DT
><DD
><P
>Check network and system configuration. This command
lists and
manages the network and system services started at bootup in
the <TT
CLASS="FILENAME"
>/etc/rc?.d</TT
>
directory.</P
><P
>Originally a port from IRIX to Red Hat Linux,
<B
CLASS="COMMAND"
>chkconfig</B
> may not be part of the core
installation of some Linux flavors.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>chkconfig --list</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>atd 0:off 1:off 2:off 3:on 4:on 5:on 6:off
rwhod 0:off 1:off 2:off 3:off 4:off 5:off 6:off
...</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="TCPDUMPREF"
></A
><B
CLASS="COMMAND"
>tcpdump</B
></DT
><DD
><P
>Network packet <SPAN
CLASS="QUOTE"
>"sniffer."</SPAN
> This is a tool for
analyzing and troubleshooting traffic on a network by dumping
packet headers that match specified criteria.</P
><P
>Dump ip packet traffic between hosts
<EM
>bozoville</EM
> and
<EM
>caduceus</EM
>:</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>tcpdump ip host bozoville and caduceus</B
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Of course, the output of <B
CLASS="COMMAND"
>tcpdump</B
> can be
parsed with certain of the previously discussed <A
HREF="#TPCOMMANDLISTING1"
>text processing
utilities</A
>.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="FILESYSTEMSYS1"
></A
>Filesystem</B
></P
><DL
><DT
><A
NAME="MOUNTREF"
></A
><B
CLASS="COMMAND"
>mount</B
></DT
><DD
><P
>Mount a filesystem, usually on an external device,
such as a floppy or CDROM. <A
NAME="FSTABREF"
></A
>The file
<TT
CLASS="FILENAME"
>/etc/fstab</TT
> provides a handy listing
of available filesystems, partitions, and devices,
including options, that may be automatically or manually
mounted. The file <TT
CLASS="FILENAME"
>/etc/mtab</TT
> shows
the currently mounted filesystems and partitions
(including the virtual ones, such as <TT
CLASS="FILENAME"
>/proc</TT
>).</P
><P
><B
CLASS="COMMAND"
>mount -a</B
> mounts all filesystems and
partitions listed in <TT
CLASS="FILENAME"
>/etc/fstab</TT
>,
except those with a <TT
CLASS="OPTION"
>noauto</TT
>
option. At bootup, a startup script in
<TT
CLASS="FILENAME"
>/etc/rc.d</TT
>
(<TT
CLASS="FILENAME"
>rc.sysinit</TT
> or something similar)
invokes this to get everything mounted.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>mount -t iso9660 /dev/cdrom /mnt/cdrom
# Mounts CD ROM. ISO 9660 is a standard CD ROM filesystem.
mount /mnt/cdrom
# Shortcut, if /mnt/cdrom listed in /etc/fstab</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="ISOMOUNTREF0"
></A
></P
><P
>The versatile <I
CLASS="FIRSTTERM"
>mount</I
> command can even
mount an ordinary file on a block device, and the file will
act as if it were a filesystem. <I
CLASS="FIRSTTERM"
>Mount</I
>
accomplishes that by associating the file with a <A
HREF="#LOOPBACKREF"
>loopback device</A
>. One application of
this is to mount and examine an ISO9660 filesystem image before
burning it onto a CDR.
<A
NAME="AEN16255"
HREF="#FTN.AEN16255"
><SPAN
CLASS="footnote"
>[90]</SPAN
></A
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ISOMOUNTREF"
></A
><P
><B
>Example 17-7. Checking a CD image</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># As root...
mkdir /mnt/cdtest # Prepare a mount point, if not already there.
mount -r -t iso9660 -o loop cd-image.iso /mnt/cdtest # Mount the image.
# "-o loop" option equivalent to "losetup /dev/loop0"
cd /mnt/cdtest # Now, check the image.
ls -alR # List the files in the directory tree there.
# And so forth.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="UMOUNTREF"
></A
><B
CLASS="COMMAND"
>umount</B
></DT
><DD
><P
>Unmount a currently mounted filesystem. Before physically removing a
previously mounted floppy or CDROM disk, the device must be
<B
CLASS="COMMAND"
>umount</B
>ed, else filesystem corruption may result.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>umount /mnt/cdrom
# You may now press the eject button and safely remove the disk.</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>automount</B
> utility, if
properly installed, can mount and unmount floppies or
CDROM disks as they are accessed or removed. On
<SPAN
CLASS="QUOTE"
>"multispindle"</SPAN
> laptops with swappable
floppy and optical drives, this can cause problems,
however.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="GNOMEMOUNTREF"
></A
><B
CLASS="COMMAND"
>gnome-mount</B
></DT
><DD
><P
>The newer Linux distros have deprecated
<B
CLASS="COMMAND"
>mount</B
> and <B
CLASS="COMMAND"
>umount</B
>.
The successor, for command-line mounting of removable storage
devices, is <B
CLASS="COMMAND"
>gnome-mount</B
>. It can take the
<TT
CLASS="OPTION"
>-d</TT
> option to mount a <A
HREF="#DEVFILEREF"
>device file</A
> by its listing in
<TT
CLASS="FILENAME"
>/dev</TT
>.</P
><P
>For example, to mount a USB flash drive:</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>gnome-mount -d /dev/sda1</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>gnome-mount 0.4</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>df</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>. . .
/dev/sda1 63584 12034 51550 19% /media/disk</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="SYNCREF"
></A
><B
CLASS="COMMAND"
>sync</B
></DT
><DD
><P
>Forces an immediate write of all updated data from
buffers to hard drive (synchronize drive
with buffers). While not strictly necessary, a
<B
CLASS="COMMAND"
>sync</B
> assures the sys admin or
user that the data just changed will survive a sudden
power failure. In the olden days, a <TT
CLASS="USERINPUT"
><B
>sync;
sync</B
></TT
> (twice, just to make absolutely sure) was a
useful precautionary measure before a system reboot.</P
><P
>At times, you may wish to force an immediate buffer
flush, as when securely deleting a file (see <A
HREF="#BLOTOUT"
>Example 16-61</A
>) or when the lights begin to
flicker.</P
></DD
><DT
><A
NAME="LOSETUPREF"
></A
><B
CLASS="COMMAND"
>losetup</B
></DT
><DD
><P
>Sets up and configures <A
HREF="#LOOPBACKREF"
> loopback devices</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="CREATEFS"
></A
><P
><B
>Example 17-8. Creating a filesystem in a file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>SIZE=1000000 # 1 meg
head -c $SIZE &#60; /dev/zero &#62; file # Set up file of designated size.
losetup /dev/loop0 file # Set it up as loopback device.
mke2fs /dev/loop0 # Create filesystem.
mount -o loop /dev/loop0 /mnt # Mount it.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="MKSWAPREF"
></A
><B
CLASS="COMMAND"
>mkswap</B
></DT
><DD
><P
>Creates a swap partition or file. The swap area must
subsequently be enabled with
<B
CLASS="COMMAND"
>swapon</B
>.</P
></DD
><DT
><A
NAME="SWAPONREF"
></A
><B
CLASS="COMMAND"
>swapon</B
>, <A
NAME="SWAPOFFREF"
></A
><B
CLASS="COMMAND"
>swapoff</B
></DT
><DD
><P
>Enable / disable swap partitition or file.
These commands usually take effect at bootup and
shutdown.</P
></DD
><DT
><A
NAME="MKE2FSREF"
></A
><B
CLASS="COMMAND"
>mke2fs</B
></DT
><DD
><P
>Create a Linux <I
CLASS="FIRSTTERM"
>ext2</I
>
filesystem. This command must be invoked as
<I
CLASS="FIRSTTERM"
>root</I
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ADDDRV"
></A
><P
><B
>Example 17-9. Adding a new hard drive</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Adding a second hard drive to system.
# Software configuration. Assumes hardware already mounted.
# From an article by the author of the ABS Guide.
# In issue #38 of _Linux Gazette_, http://www.linuxgazette.com.
ROOT_UID=0 # This script must be run as root.
E_NOTROOT=67 # Non-root exit error.
if [ "$UID" -ne "$ROOT_UID" ]
then
echo "Must be root to run this script."
exit $E_NOTROOT
fi
# Use with extreme caution!
# If something goes wrong, you may wipe out your current filesystem.
NEWDISK=/dev/hdb # Assumes /dev/hdb vacant. Check!
MOUNTPOINT=/mnt/newdisk # Or choose another mount point.
fdisk $NEWDISK
mke2fs -cv $NEWDISK1 # Check for bad blocks (verbose output).
# Note: ^ /dev/hdb1, *not* /dev/hdb!
mkdir $MOUNTPOINT
chmod 777 $MOUNTPOINT # Makes new drive accessible to all users.
# Now, test ...
# mount -t ext2 /dev/hdb1 /mnt/newdisk
# Try creating a directory.
# If it works, umount it, and proceed.
# Final step:
# Add the following line to /etc/fstab.
# /dev/hdb1 /mnt/newdisk ext2 defaults 1 1
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#CREATEFS"
>Example 17-8</A
> and <A
HREF="#RAMDISK"
>Example 31-3</A
>.</P
></DD
><DT
><A
NAME="MKDOSFSREF"
></A
><B
CLASS="COMMAND"
>mkdosfs</B
></DT
><DD
><P
>Create a DOS <I
CLASS="FIRSTTERM"
>FAT</I
>
filesystem.</P
></DD
><DT
><A
NAME="TUNE2FSREF"
></A
><B
CLASS="COMMAND"
>tune2fs</B
></DT
><DD
><P
>Tune <I
CLASS="FIRSTTERM"
>ext2</I
> filesystem. May be
used to change filesystem parameters, such as maximum
mount count. This must be invoked as
<I
CLASS="FIRSTTERM"
>root</I
>.</P
><DIV
CLASS="WARNING"
><P
></P
><TABLE
CLASS="WARNING"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/warning.gif"
HSPACE="5"
ALT="Warning"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This is an extremely dangerous command. Use it at
your own risk, as you may inadvertently destroy your filesystem.
</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="DUMPE2FSREF"
></A
><B
CLASS="COMMAND"
>dumpe2fs</B
></DT
><DD
><P
>Dump (list to <TT
CLASS="FILENAME"
>stdout</TT
>) very verbose
filesystem info. This must be invoked as
<I
CLASS="FIRSTTERM"
>root</I
>.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>root# </TT
><B
CLASS="COMMAND"
>dumpe2fs /dev/hda7 | grep 'ount count'</B
>
<TT
CLASS="COMPUTEROUTPUT"
>dumpe2fs 1.19, 13-Jul-2000 for EXT2 FS 0.5b, 95/08/09
Mount count: 6
Maximum mount count: 20</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="HDPARMREF"
></A
><B
CLASS="COMMAND"
>hdparm</B
></DT
><DD
><P
>List or change hard disk parameters. This command must be
invoked as <I
CLASS="FIRSTTERM"
>root</I
>, and it may be
dangerous if misused.</P
></DD
><DT
><A
NAME="FDISKREF"
></A
><B
CLASS="COMMAND"
>fdisk</B
></DT
><DD
><P
>Create or change a partition table on a storage device,
usually a hard drive. This command must be invoked as
<I
CLASS="FIRSTTERM"
>root</I
>.</P
><DIV
CLASS="WARNING"
><P
></P
><TABLE
CLASS="WARNING"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/warning.gif"
HSPACE="5"
ALT="Warning"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Use this command with extreme caution. If something
goes wrong, you may destroy an existing
filesystem.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="FSCKREF"
></A
><B
CLASS="COMMAND"
>fsck</B
>, <A
NAME="E2FSCKREF"
></A
><B
CLASS="COMMAND"
>e2fsck</B
>, <A
NAME="DEBUGFSREF"
></A
><B
CLASS="COMMAND"
>debugfs</B
></DT
><DD
><P
>Filesystem check, repair, and debug command set.</P
><P
><B
CLASS="COMMAND"
>fsck</B
>: a front end for checking a UNIX
filesystem (may invoke other utilities). The actual
filesystem type generally defaults to
<I
CLASS="FIRSTTERM"
>ext2</I
>.</P
><P
><B
CLASS="COMMAND"
>e2fsck</B
>: ext2 filesystem checker.</P
><P
><B
CLASS="COMMAND"
>debugfs</B
>: ext2 filesystem debugger.
One of the uses of this versatile, but dangerous command
is to (attempt to) recover deleted files. For advanced users
only!</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>All of these should be invoked as
<I
CLASS="FIRSTTERM"
>root</I
>, and they can damage or destroy
a filesystem if misused.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="BADBLOCKSREF"
></A
><B
CLASS="COMMAND"
>badblocks</B
></DT
><DD
><P
>Checks for bad blocks (physical media flaws) on a
storage device. This command finds use when formatting
a newly installed hard drive or testing the integrity
of backup media.
<A
NAME="AEN16504"
HREF="#FTN.AEN16504"
><SPAN
CLASS="footnote"
>[91]</SPAN
></A
>
As an example, <B
CLASS="COMMAND"
>badblocks /dev/fd0</B
>
tests a floppy disk.</P
><P
>The <B
CLASS="COMMAND"
>badblocks</B
> command
may be invoked destructively (overwrite all data) or
in non-destructive read-only mode. If <I
CLASS="FIRSTTERM"
>root
user</I
> owns the device to be tested, as is
generally the case, then <I
CLASS="FIRSTTERM"
>root</I
>
must invoke this command.</P
></DD
><DT
><A
NAME="LSUSBREF"
></A
><B
CLASS="COMMAND"
>lsusb</B
>, <B
CLASS="COMMAND"
>usbmodules</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>lsusb</B
> command lists all USB
(Universal Serial Bus) buses and the devices hooked up to
them.</P
><P
>The <B
CLASS="COMMAND"
>usbmodules</B
> command outputs
information about the driver modules for connected USB
devices.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>lsusb</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Bus 001 Device 001: ID 0000:0000
Device Descriptor:
bLength 18
bDescriptorType 1
bcdUSB 1.00
bDeviceClass 9 Hub
bDeviceSubClass 0
bDeviceProtocol 0
bMaxPacketSize0 8
idVendor 0x0000
idProduct 0x0000
. . .</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="LSPCIREF"
></A
><B
CLASS="COMMAND"
>lspci</B
></DT
><DD
><P
>Lists <I
CLASS="FIRSTTERM"
>pci</I
> busses present.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>lspci</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>00:00.0 Host bridge: Intel Corporation 82845 845
(Brookdale) Chipset Host Bridge (rev 04)
00:01.0 PCI bridge: Intel Corporation 82845 845
(Brookdale) Chipset AGP Bridge (rev 04)
00:1d.0 USB Controller: Intel Corporation 82801CA/CAM USB (Hub #1) (rev 02)
00:1d.1 USB Controller: Intel Corporation 82801CA/CAM USB (Hub #2) (rev 02)
00:1d.2 USB Controller: Intel Corporation 82801CA/CAM USB (Hub #3) (rev 02)
00:1e.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev 42)
. . .</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="MKBOOTDISKREF"
></A
><B
CLASS="COMMAND"
>mkbootdisk</B
></DT
><DD
><P
>Creates a boot floppy which can be used to bring up the
system if, for example, the MBR (master boot record) becomes
corrupted. Of special interest is the <TT
CLASS="OPTION"
>--iso</TT
>
option, which uses <B
CLASS="COMMAND"
>mkisofs</B
> to create a
bootable <I
CLASS="FIRSTTERM"
>ISO9660</I
> filesystem image
suitable for burning a bootable CDR.</P
><P
>The <B
CLASS="COMMAND"
>mkbootdisk</B
> command is actually
a Bash script, written by Erik Troan, in the <TT
CLASS="FILENAME"
>/sbin</TT
> directory.</P
></DD
><DT
><A
NAME="MKISOFSREF"
></A
><B
CLASS="COMMAND"
>mkisofs</B
></DT
><DD
><P
>Creates an <I
CLASS="FIRSTTERM"
>ISO9660</I
> filesystem
suitable for a CDR image.</P
></DD
><DT
><A
NAME="CHROOTREF"
></A
><B
CLASS="COMMAND"
>chroot</B
></DT
><DD
><P
>CHange ROOT directory. Normally commands are fetched
from <A
HREF="#PATHREF"
>$PATH</A
>, relative to
<TT
CLASS="FILENAME"
>/</TT
>, the default
<I
CLASS="FIRSTTERM"
>root
directory</I
>. This changes the
<I
CLASS="FIRSTTERM"
>root</I
> directory to a different one
(and also changes the working directory to there). This is
useful for security purposes, for instance when the system
administrator wishes to restrict certain users, such as
those <A
HREF="#TELNETREF"
>telnetting</A
> in,
to a secured portion of the filesystem (this is sometimes
referred to as confining a guest user to a <SPAN
CLASS="QUOTE"
>"chroot
jail"</SPAN
>). Note that after a <B
CLASS="COMMAND"
>chroot</B
>,
the execution path for system binaries is no longer
valid.</P
><P
>A <TT
CLASS="USERINPUT"
><B
>chroot /opt</B
></TT
> would cause
references to <TT
CLASS="FILENAME"
>/usr/bin</TT
>
to be translated to <TT
CLASS="FILENAME"
>/opt/usr/bin</TT
>. Likewise,
<TT
CLASS="USERINPUT"
><B
>chroot /aaa/bbb /bin/ls</B
></TT
> would
redirect future instances of <B
CLASS="COMMAND"
>ls</B
>
to <TT
CLASS="FILENAME"
>/aaa/bbb</TT
> as the base directory,
rather than <TT
CLASS="FILENAME"
>/</TT
> as is
normally the case. An <B
CLASS="COMMAND"
>alias XX 'chroot /aaa/bbb
ls'</B
> in a user's <A
HREF="#SAMPLE-BASHRC"
><TT
CLASS="FILENAME"
>~/.bashrc</TT
></A
>
effectively restricts which portion of the filesystem
she may run command <SPAN
CLASS="QUOTE"
>"XX"</SPAN
> on.</P
><P
>The <B
CLASS="COMMAND"
>chroot</B
> command is also handy
when running from an emergency boot floppy
(<B
CLASS="COMMAND"
>chroot</B
> to <TT
CLASS="FILENAME"
>/dev/fd0</TT
>),
or as an option to <B
CLASS="COMMAND"
>lilo</B
> when recovering
from a system crash. Other uses include installation from a
different filesystem (an <A
HREF="#RPMREF"
>rpm</A
>
option) or running a readonly filesystem from a CD ROM.
Invoke only as <I
CLASS="FIRSTTERM"
>root</I
>, and use with
care.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>It might be necessary to copy certain system
files to a <I
CLASS="FIRSTTERM"
>chrooted</I
> directory,
since the normal <TT
CLASS="VARNAME"
>$PATH</TT
> can no longer
be relied upon.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="LOCKFILEREF"
></A
><B
CLASS="COMMAND"
>lockfile</B
></DT
><DD
><P
>This utility is part of the <B
CLASS="COMMAND"
>procmail</B
>
package (<A
HREF="http://www.procmail.org"
TARGET="_top"
>www.procmail.org</A
>).
It creates a <I
CLASS="FIRSTTERM"
>lock file</I
>, a
<I
CLASS="FIRSTTERM"
>semaphore</I
> that controls access to
a file, device, or resource.</P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN16645"
></A
><P
></P
><P
><A
NAME="SEMAPHOREREF"
></A
>
<TT
CLASS="USERINPUT"
><B
>Definition:</B
></TT
>
A <I
CLASS="FIRSTTERM"
>semaphore</I
> is a flag or
signal. (The usage originated in railroading, where a
colored flag, lantern, or striped movable arm
<I
CLASS="FIRSTTERM"
>semaphore</I
> indicated whether a
particular track was in use and therefore unavailable
for another train.) A UNIX process can check the
appropriate semaphore to determine whether a particular
resource is available/accessible.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
>The lock file serves as a flag that this particular
file, device, or resource is in use by a process (and
is therefore <SPAN
CLASS="QUOTE"
>"busy"</SPAN
>). The presence of a
lock file permits only restricted access (or no access)
to other processes.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>lockfile /home/bozo/lockfiles/$0.lock
# Creates a write-protected lockfile prefixed with the name of the script.
lockfile /home/bozo/lockfiles/${0##*/}.lock
# A safer version of the above, as pointed out by E. Choroba.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Lock files are used in such applications as protecting
system mail folders from simultaneously being changed
by multiple users, indicating that a modem port
is being accessed, and showing that an instance of
<SPAN
CLASS="APPLICATION"
>Firefox</SPAN
> is using its cache.
Scripts may check for the existence of a lock file created
by a certain process to check if that process is running.
Note that if a script attempts to create a lock file that
already exists, the script will likely hang.</P
><P
>Normally, applications create and check for lock files
in the <TT
CLASS="FILENAME"
>/var/lock</TT
>
directory.
<A
NAME="AEN16659"
HREF="#FTN.AEN16659"
><SPAN
CLASS="footnote"
>[92]</SPAN
></A
>
A script can test for the presence of a lock file by
something like the following.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>appname=xyzip
# Application "xyzip" created lock file "/var/lock/xyzip.lock".
if [ -e "/var/lock/$appname.lock" ]
then #+ Prevent other programs &#38; scripts
# from accessing files/resources used by xyzip.
...</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="FLOCKREF"
></A
><B
CLASS="COMMAND"
>flock</B
></DT
><DD
><P
>Much less useful than the <B
CLASS="COMMAND"
>lockfile</B
>
command is <B
CLASS="COMMAND"
>flock</B
>. It sets an
<SPAN
CLASS="QUOTE"
>"advisory"</SPAN
> lock on a file and then executes
a command while the lock is on. This is to prevent
any other process from setting a lock on that file until
completion of the specified command.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>flock $0 cat $0 &#62; lockfile__$0
# Set a lock on the script the above line appears in,
#+ while listing the script to stdout.</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Unlike <B
CLASS="COMMAND"
>lockfile</B
>,
<B
CLASS="COMMAND"
>flock</B
> does <EM
>not</EM
>
automatically create a lock file.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="MKNODREF"
></A
><B
CLASS="COMMAND"
>mknod</B
></DT
><DD
><P
>Creates block or character <A
HREF="#DEVFILEREF"
>device files</A
> (may be
necessary when installing new hardware on the system). The
<B
CLASS="COMMAND"
>MAKEDEV</B
> utility has virtually
all of the functionality of <B
CLASS="COMMAND"
>mknod</B
>,
and is easier to use.</P
></DD
><DT
><A
NAME="MAKEDEVREF"
></A
><B
CLASS="COMMAND"
>MAKEDEV</B
></DT
><DD
><P
>Utility for creating device files. It must be run as
<I
CLASS="FIRSTTERM"
>root</I
>, and in the <TT
CLASS="FILENAME"
>/dev</TT
> directory. It is a sort
of advanced version of <B
CLASS="COMMAND"
>mknod</B
>.</P
></DD
><DT
><A
NAME="TMPWATCHREF"
></A
><B
CLASS="COMMAND"
>tmpwatch</B
></DT
><DD
><P
>Automatically deletes files which have not been accessed
within a specified period of time. Usually invoked by
<A
HREF="#CRONREF"
>cron</A
> to remove stale log
files.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="PERIPHSYS1"
></A
>Backup</B
></P
><DL
><DT
><A
NAME="DUMPREF"
></A
><B
CLASS="COMMAND"
>dump</B
>, <A
NAME="RESTOREREF"
></A
><B
CLASS="COMMAND"
>restore</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>dump</B
> command is an elaborate
filesystem backup utility, generally used on larger
installations and networks.
<A
NAME="AEN16748"
HREF="#FTN.AEN16748"
><SPAN
CLASS="footnote"
>[93]</SPAN
></A
>
It reads raw disk partitions and writes a backup file
in a binary format. Files to be backed up may be saved
to a variety of storage media, including disks and tape
drives. The <B
CLASS="COMMAND"
>restore</B
> command restores
backups made with <B
CLASS="COMMAND"
>dump</B
>.</P
></DD
><DT
><A
NAME="FDFORMATREF"
></A
><B
CLASS="COMMAND"
>fdformat</B
></DT
><DD
><P
>Perform a low-level format on a floppy disk
(<TT
CLASS="FILENAME"
>/dev/fd0*</TT
>).</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="SYSRESOURCES1"
></A
>System Resources</B
></P
><DL
><DT
><A
NAME="ULIMITREF"
></A
><B
CLASS="COMMAND"
>ulimit</B
></DT
><DD
><P
>Sets an <I
CLASS="FIRSTTERM"
>upper limit</I
> on use
of system resources. Usually invoked with the
<TT
CLASS="OPTION"
>-f</TT
> option, which sets a limit on file size
(<B
CLASS="COMMAND"
>ulimit -f 1000</B
> limits files to 1 meg
maximum).
<A
NAME="AEN16782"
HREF="#FTN.AEN16782"
><SPAN
CLASS="footnote"
>[94]</SPAN
></A
>
The <TT
CLASS="OPTION"
>-t</TT
> option limits the coredump
size (<B
CLASS="COMMAND"
>ulimit -c 0</B
> eliminates coredumps).
Normally, the value of <B
CLASS="COMMAND"
>ulimit</B
>
would be set in <TT
CLASS="FILENAME"
>/etc/profile</TT
>
and/or <TT
CLASS="FILENAME"
>~/.bash_profile</TT
> (see <A
HREF="#FILES"
>Appendix H</A
>).</P
><DIV
CLASS="IMPORTANT"
><P
></P
><TABLE
CLASS="IMPORTANT"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/important.gif"
HSPACE="5"
ALT="Important"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Judicious use of <B
CLASS="COMMAND"
>ulimit</B
> can
protect a system against the dreaded <I
CLASS="FIRSTTERM"
>fork
bomb</I
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# This script is for illustrative purposes only.
# Run it at your own peril -- it WILL freeze your system.
while true # Endless loop.
do
$0 &#38; # This script invokes itself . . .
#+ forks an infinite number of times . . .
#+ until the system freezes up because all resources exhausted.
done # This is the notorious <SPAN
CLASS="QUOTE"
>"sorcerer's appentice"</SPAN
> scenario.
exit 0 # Will not exit here, because this script will never terminate.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>A <B
CLASS="COMMAND"
>ulimit -Hu XX</B
> (where
<EM
>XX</EM
> is the user process limit) in
<TT
CLASS="FILENAME"
>/etc/profile</TT
> would abort
this script when it exceeded the preset limit.
</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="QUOTAREF"
></A
><B
CLASS="COMMAND"
>quota</B
></DT
><DD
><P
>Display user or group disk quotas.</P
></DD
><DT
><A
NAME="SETQUOTAREF"
></A
><B
CLASS="COMMAND"
>setquota</B
></DT
><DD
><P
>Set user or group disk quotas from the command-line.</P
></DD
><DT
><A
NAME="UMASKREF"
></A
><B
CLASS="COMMAND"
>umask</B
></DT
><DD
><P
>User file creation permissions
<I
CLASS="FIRSTTERM"
>mask</I
>. Limit the default file
attributes for a particular user. All files created
by that user take on the attributes specified by
<B
CLASS="COMMAND"
>umask</B
>. The (octal) value passed to
<B
CLASS="COMMAND"
>umask</B
> defines the file permissions
<I
CLASS="FIRSTTERM"
>disabled</I
>. For example, <B
CLASS="COMMAND"
>umask
022</B
> ensures that new files will have at most
755 permissions (777 NAND 022).
<A
NAME="AEN16847"
HREF="#FTN.AEN16847"
><SPAN
CLASS="footnote"
>[95]</SPAN
></A
>
Of course, the user may later change the
attributes of particular files with <A
HREF="#CHMODREF"
>chmod</A
>. The usual practice
is to set the value of <B
CLASS="COMMAND"
>umask</B
>
in <TT
CLASS="FILENAME"
>/etc/profile</TT
> and/or
<TT
CLASS="FILENAME"
>~/.bash_profile</TT
> (see <A
HREF="#FILES"
>Appendix H</A
>).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ROT13A"
></A
><P
><B
>Example 17-10. Using <I
CLASS="FIRSTTERM"
>umask</I
> to hide an output file
from prying eyes</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# rot13a.sh: Same as "rot13.sh" script, but writes output to "secure" file.
# Usage: ./rot13a.sh filename
# or ./rot13a.sh &#60;filename
# or ./rot13a.sh and supply keyboard input (stdin)
umask 177 # File creation mask.
# Files created by this script
#+ will have 600 permissions.
OUTFILE=decrypted.txt # Results output to file "decrypted.txt"
#+ which can only be read/written
# by invoker of script (or root).
cat "$@" | tr 'a-zA-Z' 'n-za-mN-ZA-M' &#62; $OUTFILE
# ^^ Input from stdin or a file. ^^^^^^^^^^ Output redirected to file.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="RDEVREF"
></A
><B
CLASS="COMMAND"
>rdev</B
></DT
><DD
><P
>Get info about or make changes to root device, swap space, or video
mode. The functionality of <B
CLASS="COMMAND"
>rdev</B
> has generally been taken over by
<B
CLASS="COMMAND"
>lilo</B
>, but <B
CLASS="COMMAND"
>rdev</B
> remains
useful for setting up a ram disk. This is a dangerous command, if misused.
</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="MODULESSYS1"
></A
>Modules</B
></P
><DL
><DT
><A
NAME="LSMODREF"
></A
><B
CLASS="COMMAND"
>lsmod</B
></DT
><DD
><P
>List installed kernel modules.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>lsmod</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Module Size Used by
autofs 9456 2 (autoclean)
opl3 11376 0
serial_cs 5456 0 (unused)
sb 34752 0
uart401 6384 0 [sb]
sound 58368 0 [opl3 sb uart401]
soundlow 464 0 [sound]
soundcore 2800 6 [sb sound]
ds 6448 2 [serial_cs]
i82365 22928 2
pcmcia_core 45984 0 [serial_cs ds i82365]</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Doing a <B
CLASS="COMMAND"
>cat /proc/modules</B
> gives the
same information.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="INSMODREF"
></A
><B
CLASS="COMMAND"
>insmod</B
></DT
><DD
><P
>Force installation of a kernel module (use
<B
CLASS="COMMAND"
>modprobe</B
> instead, when possible). Must
be invoked as <I
CLASS="FIRSTTERM"
>root</I
>.</P
></DD
><DT
><A
NAME="RMMODREF"
></A
><B
CLASS="COMMAND"
>rmmod</B
></DT
><DD
><P
>Force unloading of a kernel module. Must be invoked
as <I
CLASS="FIRSTTERM"
>root</I
>.</P
></DD
><DT
><A
NAME="MODPROBEREF"
></A
><B
CLASS="COMMAND"
>modprobe</B
></DT
><DD
><P
>Module loader that is normally invoked automatically
in a startup script. Must be invoked as
<I
CLASS="FIRSTTERM"
>root</I
>.</P
></DD
><DT
><A
NAME="DEPMODREF"
></A
><B
CLASS="COMMAND"
>depmod</B
></DT
><DD
><P
>Creates module dependency file. Usually invoked from a
startup script.</P
></DD
><DT
><A
NAME="MODINFOREF"
></A
><B
CLASS="COMMAND"
>modinfo</B
></DT
><DD
><P
>Output information about a loadable module.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>modinfo hid</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>filename: /lib/modules/2.4.20-6/kernel/drivers/usb/hid.o
description: "USB HID support drivers"
author: "Andreas Gal, Vojtech Pavlik &#60;vojtech@suse.cz&#62;"
license: "GPL"</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="MISCSYS1"
></A
>Miscellaneous</B
></P
><DL
><DT
><A
NAME="ENVVREF"
></A
><B
CLASS="COMMAND"
>env</B
></DT
><DD
><P
> Runs a program or script with certain <A
HREF="#ENVREF"
>environmental variables</A
>
set or changed (without changing the overall system
environment). The <TT
CLASS="OPTION"
>[varname=xxx]</TT
>
permits changing the environmental variable
<TT
CLASS="VARNAME"
>varname</TT
> for the duration of the
script. With no options specified, this command lists all
the environmental variable settings.
<A
NAME="AEN16975"
HREF="#FTN.AEN16975"
><SPAN
CLASS="footnote"
>[96]</SPAN
></A
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="ENVV2REF"
></A
>The first line of a script (the
<SPAN
CLASS="QUOTE"
>"sha-bang"</SPAN
> line) may use <B
CLASS="COMMAND"
>env</B
>
when the path to the shell or interpreter is unknown.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /usr/bin/env perl
print "This Perl script will run,\n";
print "even when I don't know where to find Perl.\n";
# Good for portable cross-platform scripts,
# where the Perl binaries may not be in the expected place.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Or even ... </P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/env bash
# Queries the $PATH enviromental variable for the location of bash.
# Therefore ...
# This script will run where Bash is not in its usual place, in /bin.
...</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="LDDREF"
></A
><B
CLASS="COMMAND"
>ldd</B
></DT
><DD
><P
>Show shared lib dependencies for an executable file.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ldd /bin/ls</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>libc.so.6 =&#62; /lib/libc.so.6 (0x4000c000)
/lib/ld-linux.so.2 =&#62; /lib/ld-linux.so.2 (0x80000000)</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DD
><DT
><A
NAME="WATCHREF"
></A
><B
CLASS="COMMAND"
>watch</B
></DT
><DD
><P
>Run a command repeatedly, at specified time intervals.</P
><P
>The default is two-second intervals, but this may be changed
with the <TT
CLASS="OPTION"
>-n</TT
> option.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>watch -n 5 tail /var/log/messages
# Shows tail end of system log, /var/log/messages, every five seconds.</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Unfortunately, <A
HREF="#PIPEREF"
>piping</A
>
the output of <B
CLASS="COMMAND"
>watch command</B
> to <A
HREF="#GREPREF"
>grep</A
> does not work.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="STRIPREF"
></A
><B
CLASS="COMMAND"
>strip</B
></DT
><DD
><P
>Remove the debugging symbolic references from an executable
binary. This decreases its size, but makes debugging it
impossible.</P
><P
>This command often occurs in a <A
HREF="#MAKEFILEREF"
>Makefile</A
>,
but rarely in a shell script.</P
></DD
><DT
><A
NAME="NMREF"
></A
><B
CLASS="COMMAND"
>nm</B
></DT
><DD
><P
>List symbols in an unstripped compiled binary.</P
></DD
><DT
><A
NAME="XRANDRREF"
></A
><B
CLASS="COMMAND"
>xrandr</B
></DT
><DD
><P
>Command-line tool for manipulating the root window
of the screen.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="BACKLIGHT"
></A
><P
><B
>Example 17-11. <I
CLASS="FIRSTTERM"
>Backlight</I
>: changes
the brightness of the (laptop) screen backlight</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# backlight.sh
# reldate 02dec2011
# A bug in Fedora Core 16/17 messes up the keyboard backlight controls.
# This script is a quick-n-dirty workaround, essentially a shell wrapper
#+ for xrandr. It gives more control than on-screen sliders and widgets.
OUTPUT=$(xrandr | grep LV | awk '{print $1}') # Get display name!
INCR=.05 # For finer-grained control, set INCR to .03 or .02.
old_brightness=$(xrandr --verbose | grep rightness | awk '{ print $2 }')
if [ -z "$1" ]
then
bright=1 # If no command-line arg, set brightness to 1.0 (default).
else
if [ "$1" = "+" ]
then
bright=$(echo "scale=2; $old_brightness + $INCR" | bc) # +.05
else
if [ "$1" = "-" ]
then
bright=$(echo "scale=2; $old_brightness - $INCR" | bc) # -.05
else
if [ "$1" = "#" ] # Echoes current brightness; does not change it.
then
bright=$old_brightness
else
if [[ "$1" = "h" || "$1" = "H" ]]
then
echo
echo "Usage:"
echo "$0 [No args] Sets/resets brightness to default (1.0)."
echo "$0 + Increments brightness by 0.5."
echo "$0 - Decrements brightness by 0.5."
echo "$0 # Echoes current brightness without changing it."
echo "$0 N (number) Sets brightness to N (useful range .7 - 1.2)."
echo "$0 h [H] Echoes this help message."
echo "$0 any-other Gives xrandr usage message."
bright=$old_brightness
else
bright="$1"
fi
fi
fi
fi
fi
xrandr --output "$OUTPUT" --brightness "$bright" # See xrandr manpage.
# As root!
E_CHANGE0=$?
echo "Current brightness = $bright"
exit $E_CHANGE0
# =========== Or, alternately . . . ==================== #
#!/bin/bash
# backlight2.sh
# reldate 20jun2012
# A bug in Fedora Core 16/17 messes up the keyboard backlight controls.
# This is a quick-n-dirty workaround, an alternate to backlight.sh.
target_dir=\
/sys/devices/pci0000:00/0000:00:01.0/0000:01:00.0/backlight/acpi_video0
# Hardware directory.
actual_brightness=$(cat $target_dir/actual_brightness)
max_brightness=$(cat $target_dir/max_brightness)
Brightness=$target_dir/brightness
let "req_brightness = actual_brightness" # Requested brightness.
if [ "$1" = "-" ]
then # Decrement brightness 1 notch.
let "req_brightness = $actual_brightness - 1"
else
if [ "$1" = "+" ]
then # Increment brightness 1 notch.
let "req_brightness = $actual_brightness + 1"
fi
fi
if [ $req_brightness -gt $max_brightness ]
then
req_brightness=$max_brightness
fi # Do not exceed max. hardware design brightness.
echo
echo "Old brightness = $actual_brightness"
echo "Max brightness = $max_brightness"
echo "Requested brightness = $req_brightness"
echo
# =====================================
echo $req_brightness &#62; $Brightness
# Must be root for this to take effect.
E_CHANGE1=$? # Successful?
# =====================================
if [ "$?" -eq 0 ]
then
echo "Changed brightness!"
else
echo "Failed to change brightness!"
fi
act_brightness=$(cat $Brightness)
echo "Actual brightness = $act_brightness"
scale0=2
sf=100 # Scale factor.
pct=$(echo "scale=$scale0; $act_brightness / $max_brightness * $sf" | bc)
echo "Percentage brightness = $pct%"
exit $E_CHANGE1</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="RDISTREF"
></A
><B
CLASS="COMMAND"
>rdist</B
></DT
><DD
><P
>Remote distribution client: synchronizes, clones,
or backs up a file system on a remote server.</P
></DD
></DL
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="SYSSCRIPTS"
></A
>17.1. Analyzing a System Script</H1
><P
><A
NAME="KILLALL2REF"
></A
></P
><P
>Using our knowledge of administrative commands, let us examine a system
script. One of the shortest and simplest to understand scripts is
<SPAN
CLASS="QUOTE"
>"killall,"</SPAN
>
<A
NAME="AEN17079"
HREF="#FTN.AEN17079"
><SPAN
CLASS="footnote"
>[97]</SPAN
></A
>
used to suspend running processes at system shutdown.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX55"
></A
><P
><B
>Example 17-12. <I
CLASS="FIRSTTERM"
>killall</I
>, from <TT
CLASS="FILENAME"
>/etc/rc.d/init.d</TT
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/sh
# --&#62; Comments added by the author of this document marked by "# --&#62;".
# --&#62; This is part of the 'rc' script package
# --&#62; by Miquel van Smoorenburg, &#60;miquels@drinkel.nl.mugnet.org&#62;.
# --&#62; This particular script seems to be Red Hat / FC specific
# --&#62; (may not be present in other distributions).
# Bring down all unneeded services that are still running
#+ (there shouldn't be any, so this is just a sanity check)
for i in /var/lock/subsys/*; do
# --&#62; Standard for/in loop, but since "do" is on same line,
# --&#62; it is necessary to add ";".
# Check if the script is there.
[ ! -f $i ] &#38;&#38; continue
# --&#62; This is a clever use of an "and list", equivalent to:
# --&#62; if [ ! -f "$i" ]; then continue
# Get the subsystem name.
subsys=${i#/var/lock/subsys/}
# --&#62; Match variable name, which, in this case, is the file name.
# --&#62; This is the exact equivalent of subsys=`basename $i`.
# --&#62; It gets it from the lock file name
# --&#62;+ (if there is a lock file,
# --&#62;+ that's proof the process has been running).
# --&#62; See the "lockfile" entry, above.
# Bring the subsystem down.
if [ -f /etc/rc.d/init.d/$subsys.init ]; then
/etc/rc.d/init.d/$subsys.init stop
else
/etc/rc.d/init.d/$subsys stop
# --&#62; Suspend running jobs and daemons.
# --&#62; Note that "stop" is a positional parameter,
# --&#62;+ not a shell builtin.
fi
done</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>That wasn't so bad. Aside from a little fancy footwork with variable
matching, there is no new material there.</P
><DIV
CLASS="FORMALPARA"
><P
><B
>Exercise 1. </B
>In <TT
CLASS="FILENAME"
>/etc/rc.d/init.d</TT
>,
analyze the <B
CLASS="COMMAND"
>halt</B
> script. It is a bit longer
than <B
CLASS="COMMAND"
>killall</B
>, but similar in concept. Make
a copy of this script somewhere in your home directory and
experiment with it (do <EM
>not</EM
> run it as
<I
CLASS="FIRSTTERM"
>root</I
>). Do a simulated run
with the <TT
CLASS="OPTION"
>-vn</TT
> flags (<TT
CLASS="USERINPUT"
><B
>sh
-vn scriptname</B
></TT
>). Add extensive
comments. Change the commands to <A
HREF="#ECHOREF"
>echos</A
>.</P
></DIV
><DIV
CLASS="FORMALPARA"
><P
><B
>Exercise 2. </B
>Look at some of the more complex scripts in
<TT
CLASS="FILENAME"
>/etc/rc.d/init.d</TT
>.
Try to understand at least portions of them. Follow
the above procedure to analyze them. For some
additional insight, you might also examine the
file <TT
CLASS="FILENAME"
>sysvinitfiles</TT
> in <TT
CLASS="FILENAME"
>/usr/share/doc/initscripts-?.??</TT
>,
which is part of the <SPAN
CLASS="QUOTE"
>"initscripts"</SPAN
>
documentation.</P
></DIV
></DIV
></DIV
></DIV
><DIV
CLASS="PART"
><A
NAME="PART5"
></A
><DIV
CLASS="TITLEPAGE"
><H1
CLASS="TITLE"
>Part 5. Advanced Topics</H1
><DIV
CLASS="PARTINTRO"
><A
NAME="AEN17110"
></A
><P
>At this point, we are ready to delve into certain of the
difficult and unusual aspects of scripting. Along the way, we
will attempt to <SPAN
CLASS="QUOTE"
>"push the envelope"</SPAN
> in various
ways and examine <I
CLASS="FIRSTTERM"
>boundary conditions</I
>
(what happens when we move into uncharted territory?).</P
></DIV
><DIV
CLASS="TOC"
><DL
><DT
><B
>Table of Contents</B
></DT
><DT
>18. <A
HREF="#REGEXP"
>Regular Expressions</A
></DT
><DD
><DL
><DT
>18.1. <A
HREF="#AEN17129"
>A Brief Introduction to Regular Expressions</A
></DT
><DT
>18.2. <A
HREF="#GLOBBINGREF"
>Globbing</A
></DT
></DL
></DD
><DT
>19. <A
HREF="#HERE-DOCS"
>Here Documents</A
></DT
><DD
><DL
><DT
>19.1. <A
HREF="#AEN17837"
>Here Strings</A
></DT
></DL
></DD
><DT
>20. <A
HREF="#IO-REDIRECTION"
>I/O Redirection</A
></DT
><DD
><DL
><DT
>20.1. <A
HREF="#AEN17974"
>Using <I
CLASS="FIRSTTERM"
>exec</I
></A
></DT
><DT
>20.2. <A
HREF="#REDIRCB"
>Redirecting Code Blocks</A
></DT
><DT
>20.3. <A
HREF="#REDIRAPPS"
>Applications</A
></DT
></DL
></DD
><DT
>21. <A
HREF="#SUBSHELLS"
>Subshells</A
></DT
><DT
>22. <A
HREF="#RESTRICTED-SH"
>Restricted Shells</A
></DT
><DT
>23. <A
HREF="#PROCESS-SUB"
>Process Substitution</A
></DT
><DT
>24. <A
HREF="#FUNCTIONS"
>Functions</A
></DT
><DD
><DL
><DT
>24.1. <A
HREF="#COMPLEXFUNCT"
>Complex Functions and Function Complexities</A
></DT
><DT
>24.2. <A
HREF="#LOCALVAR"
>Local Variables</A
></DT
><DT
>24.3. <A
HREF="#RECURNOLOCVAR"
>Recursion Without Local Variables</A
></DT
></DL
></DD
><DT
>25. <A
HREF="#ALIASES"
>Aliases</A
></DT
><DT
>26. <A
HREF="#LIST-CONS"
>List Constructs</A
></DT
><DT
>27. <A
HREF="#ARRAYS"
>Arrays</A
></DT
><DT
>28. <A
HREF="#IVR"
>Indirect References</A
></DT
><DT
>29. <A
HREF="#DEVPROC"
><TT
CLASS="FILENAME"
>/dev</TT
> and <TT
CLASS="FILENAME"
>/proc</TT
></A
></DT
><DD
><DL
><DT
>29.1. <A
HREF="#DEVREF1"
><TT
CLASS="FILENAME"
>/dev</TT
></A
></DT
><DT
>29.2. <A
HREF="#PROCREF1"
><TT
CLASS="FILENAME"
>/proc</TT
></A
></DT
></DL
></DD
><DT
>30. <A
HREF="#NETWORKPROGRAMMING"
>Network Programming</A
></DT
><DT
>31. <A
HREF="#ZEROS"
>Of Zeros and Nulls</A
></DT
><DT
>32. <A
HREF="#DEBUGGING"
>Debugging</A
></DT
><DT
>33. <A
HREF="#OPTIONS"
>Options</A
></DT
><DT
>34. <A
HREF="#GOTCHAS"
>Gotchas</A
></DT
><DT
>35. <A
HREF="#SCRSTYLE"
>Scripting With Style</A
></DT
><DD
><DL
><DT
>35.1. <A
HREF="#UNOFFICIALST"
>Unofficial Shell Scripting Stylesheet</A
></DT
></DL
></DD
><DT
>36. <A
HREF="#MISCELLANY"
>Miscellany</A
></DT
><DD
><DL
><DT
>36.1. <A
HREF="#INTANDNONINT"
>Interactive and non-interactive shells and scripts</A
></DT
><DT
>36.2. <A
HREF="#WRAPPER"
>Shell Wrappers</A
></DT
><DT
>36.3. <A
HREF="#TESTSANDCOMPARISONS"
>Tests and Comparisons: Alternatives</A
></DT
><DT
>36.4. <A
HREF="#RECURSIONSCT"
>Recursion: a script calling itself</A
></DT
><DT
>36.5. <A
HREF="#COLORIZING"
><SPAN
CLASS="QUOTE"
>"Colorizing"</SPAN
> Scripts</A
></DT
><DT
>36.6. <A
HREF="#OPTIMIZATIONS"
>Optimizations</A
></DT
><DT
>36.7. <A
HREF="#ASSORTEDTIPS"
>Assorted Tips</A
></DT
><DT
>36.8. <A
HREF="#SECURITYISSUES"
>Security Issues</A
></DT
><DT
>36.9. <A
HREF="#PORTABILITYISSUES"
>Portability Issues</A
></DT
><DT
>36.10. <A
HREF="#WINSCRIPT"
>Shell Scripting Under Windows</A
></DT
></DL
></DD
><DT
>37. <A
HREF="#BASH2"
>Bash, versions 2, 3, and 4</A
></DT
><DD
><DL
><DT
>37.1. <A
HREF="#BASHVER2"
>Bash, version 2</A
></DT
><DT
>37.2. <A
HREF="#BASHVER3"
>Bash, version 3</A
></DT
><DT
>37.3. <A
HREF="#BASHVER4"
>Bash, version 4</A
></DT
></DL
></DD
></DL
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="REGEXP"
></A
>Chapter 18. Regular Expressions</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>. . . the intellectual activity associated with software
development is largely one of gaining insight.</I
></P
><P
><I
>--Stowe Boyd</I
></P
></I
></TD
></TR
></TABLE
><P
><A
NAME="REGEXREF"
></A
></P
><P
>To fully utilize the power of shell scripting, you need to
master Regular Expressions. Certain commands
and utilities commonly used in scripts, such
as <A
HREF="#GREPREF"
>grep</A
>, <A
HREF="#EXPRREF"
>expr</A
>, <A
HREF="#SEDREF"
>sed</A
>
and <A
HREF="#AWKREF"
>awk</A
>, interpret and use REs. As of
<A
HREF="#BASH3REF"
>version 3</A
>, Bash has acquired its
own <A
HREF="#REGEXMATCHREF"
>RE-match operator</A
>:
<B
CLASS="COMMAND"
>=~</B
>.</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="AEN17129"
></A
>18.1. A Brief Introduction to Regular Expressions</H1
><P
>An expression is a string of characters. Those characters
having an interpretation above and beyond their literal
meaning are called <I
CLASS="FIRSTTERM"
>metacharacters</I
>.
A quote symbol, for example, may denote speech by a person,
<I
CLASS="FIRSTTERM"
>ditto</I
>, or a meta-meaning
<A
NAME="AEN17134"
HREF="#FTN.AEN17134"
><SPAN
CLASS="footnote"
>[98]</SPAN
></A
>
for the symbols that follow. Regular Expressions are sets
of characters and/or metacharacters that match (or specify)
patterns.</P
><P
>A Regular Expression contains one or more of the
following:</P
><P
></P
><UL
><LI
><P
><I
CLASS="FIRSTTERM"
>A character set</I
>. These are the
characters retaining their literal meaning. The
simplest type of Regular Expression consists
<EM
>only</EM
> of a character set, with no
metacharacters.</P
></LI
><LI
><P
><A
NAME="ANCHORREF"
></A
></P
><P
><I
CLASS="FIRSTTERM"
>An anchor</I
>. These designate
(<I
CLASS="FIRSTTERM"
>anchor</I
>) the position in the line of
text that the RE is to match. For example, <SPAN
CLASS="TOKEN"
>^</SPAN
>,
and <SPAN
CLASS="TOKEN"
>$</SPAN
> are anchors.</P
></LI
><LI
><P
><I
CLASS="FIRSTTERM"
>Modifiers</I
>. These expand or narrow
(<I
CLASS="FIRSTTERM"
>modify</I
>) the range of text the RE is
to match. Modifiers include the asterisk, brackets, and
the backslash.</P
></LI
></UL
><P
>The main uses for Regular Expressions
(<I
CLASS="FIRSTTERM"
>RE</I
>s) are text searches and string
manipulation. An RE <I
CLASS="FIRSTTERM"
>matches</I
> a single
character or a set of characters -- a string or a part of
a string.</P
><P
></P
><UL
><LI
><P
><A
NAME="ASTERISKREG"
></A
>The asterisk --
<SPAN
CLASS="TOKEN"
>*</SPAN
> -- matches any number of
repeats of the character string or RE preceding it,
including <EM
>zero</EM
> instances.</P
><P
><SPAN
CLASS="QUOTE"
>"1133*"</SPAN
> matches <TT
CLASS="REPLACEABLE"
><I
>11 +
one or more 3's</I
></TT
>:
<TT
CLASS="REPLACEABLE"
><I
>113</I
></TT
>, <TT
CLASS="REPLACEABLE"
><I
>1133</I
></TT
>,
<TT
CLASS="REPLACEABLE"
><I
>1133333</I
></TT
>, and so forth.</P
></LI
><LI
><P
><A
NAME="REGEXDOT"
></A
>The <I
CLASS="FIRSTTERM"
>dot</I
>
-- <SPAN
CLASS="TOKEN"
>.</SPAN
> -- matches
any one character, except a newline.
<A
NAME="AEN17189"
HREF="#FTN.AEN17189"
><SPAN
CLASS="footnote"
>[99]</SPAN
></A
>
</P
><P
><SPAN
CLASS="QUOTE"
>"13."</SPAN
> matches <TT
CLASS="REPLACEABLE"
><I
>13 + at
least one of any character (including a
space)</I
></TT
>: <TT
CLASS="REPLACEABLE"
><I
>1133</I
></TT
>,
<TT
CLASS="REPLACEABLE"
><I
>11333</I
></TT
>, but not
<TT
CLASS="REPLACEABLE"
><I
>13</I
></TT
> (additional character
missing).</P
><P
>See <A
HREF="#CWSOLVER"
>Example 16-18</A
> for a demonstration
of <I
CLASS="FIRSTTERM"
>dot single-character</I
>
matching.</P
></LI
><LI
><P
><A
NAME="CARETREF"
></A
>The caret -- <SPAN
CLASS="TOKEN"
>^</SPAN
>
-- matches the beginning of a line, but sometimes, depending
on context, negates the meaning of a set of characters in
an RE.</P
></LI
><LI
><P
><A
NAME="DOLLARSIGNREF"
></A
></P
><P
>The dollar sign -- <SPAN
CLASS="TOKEN"
>$</SPAN
> -- at the end of an
RE matches the end of a line.</P
><P
><SPAN
CLASS="QUOTE"
>"XXX$"</SPAN
> matches <SPAN
CLASS="TOKEN"
>XXX</SPAN
> at the
end of a line.</P
><P
><SPAN
CLASS="QUOTE"
>"^$"</SPAN
> matches blank lines.</P
></LI
><LI
><P
><A
NAME="BRACKETSREF"
></A
></P
><P
>Brackets -- <SPAN
CLASS="TOKEN"
>[...]</SPAN
> -- enclose a set of characters
to match in a single RE.</P
><P
><SPAN
CLASS="QUOTE"
>"[xyz]"</SPAN
> matches any one of the characters
<TT
CLASS="REPLACEABLE"
><I
>x</I
></TT
>, <TT
CLASS="REPLACEABLE"
><I
>y</I
></TT
>,
or <TT
CLASS="REPLACEABLE"
><I
>z</I
></TT
>.</P
><P
><SPAN
CLASS="QUOTE"
>"[c-n]"</SPAN
> matches any one of the
characters in the range <TT
CLASS="REPLACEABLE"
><I
>c</I
></TT
>
to <TT
CLASS="REPLACEABLE"
><I
>n</I
></TT
>.</P
><P
><SPAN
CLASS="QUOTE"
>"[B-Pk-y]"</SPAN
> matches any one of the
characters in the ranges <TT
CLASS="REPLACEABLE"
><I
>B</I
></TT
>
to <TT
CLASS="REPLACEABLE"
><I
>P</I
></TT
>
and <TT
CLASS="REPLACEABLE"
><I
>k</I
></TT
> to
<TT
CLASS="REPLACEABLE"
><I
>y</I
></TT
>.</P
><P
><SPAN
CLASS="QUOTE"
>"[a-z0-9]"</SPAN
> matches any single lowercase
letter or any digit.</P
><P
><SPAN
CLASS="QUOTE"
>"[^b-d]"</SPAN
> matches any character
<EM
>except</EM
> those in
the range <TT
CLASS="REPLACEABLE"
><I
>b</I
></TT
> to
<TT
CLASS="REPLACEABLE"
><I
>d</I
></TT
>. This is an instance of
<SPAN
CLASS="TOKEN"
>^</SPAN
> negating or inverting the meaning
of the following RE (taking on a role similar to
<SPAN
CLASS="TOKEN"
>!</SPAN
> in a different context).</P
><P
>Combined sequences of bracketed characters match
common word patterns. <SPAN
CLASS="QUOTE"
>"[Yy][Ee][Ss]"</SPAN
> matches
<TT
CLASS="REPLACEABLE"
><I
>yes</I
></TT
>, <TT
CLASS="REPLACEABLE"
><I
>Yes</I
></TT
>,
<TT
CLASS="REPLACEABLE"
><I
>YES</I
></TT
>, <TT
CLASS="REPLACEABLE"
><I
>yEs</I
></TT
>,
and so forth.
<SPAN
CLASS="QUOTE"
>"[0-9][0-9][0-9]-[0-9][0-9]-[0-9][0-9][0-9][0-9]"</SPAN
>
matches any Social Security number.</P
></LI
><LI
><P
><A
NAME="REGEXBS"
></A
></P
><P
>The backslash -- <SPAN
CLASS="TOKEN"
>\</SPAN
> -- <A
HREF="#ESCP"
>escapes</A
> a special character, which
means that character gets interpreted literally (and is
therefore no longer <I
CLASS="FIRSTTERM"
>special</I
>).</P
><P
>A <SPAN
CLASS="QUOTE"
>"\$"</SPAN
> reverts back to its
literal meaning of <SPAN
CLASS="QUOTE"
>"$"</SPAN
>, rather than its
RE meaning of end-of-line. Likewise a <SPAN
CLASS="QUOTE"
>"\\"</SPAN
>
has the literal meaning of <SPAN
CLASS="QUOTE"
>"\"</SPAN
>.</P
></LI
><LI
><P
><A
NAME="ANGLEBRAC"
></A
></P
><P
><A
HREF="#ESCP"
>Escaped</A
> <SPAN
CLASS="QUOTE"
>"angle
brackets"</SPAN
> -- <SPAN
CLASS="TOKEN"
>\&#60;...\&#62;</SPAN
> -- mark word
boundaries.</P
><P
>The angle brackets must be escaped, since otherwise
they have only their literal character meaning.</P
><P
><SPAN
CLASS="QUOTE"
>"\&#60;the\&#62;"</SPAN
> matches the word
<SPAN
CLASS="QUOTE"
>"the,"</SPAN
> but not the words <SPAN
CLASS="QUOTE"
>"them,"</SPAN
>
<SPAN
CLASS="QUOTE"
>"there,"</SPAN
> <SPAN
CLASS="QUOTE"
>"other,"</SPAN
> etc.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat textfile</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>This is line 1, of which there is only one instance.
This is the only instance of line 2.
This is line 3, another line.
This is line 4.</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep 'the' textfile</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>This is line 1, of which there is only one instance.
This is the only instance of line 2.
This is line 3, another line.</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep '\&#60;the\&#62;' textfile</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>This is the only instance of line 2.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
></UL
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN17316"
></A
><P
></P
><P
>The only way to be certain that a particular RE works is to
test it.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>TEST FILE: tstfile # No match.
# No match.
Run grep "1133*" on this file. # Match.
# No match.
# No match.
This line contains the number 113. # Match.
This line contains the number 13. # No match.
This line contains the number 133. # No match.
This line contains the number 1133. # Match.
This line contains the number 113312. # Match.
This line contains the number 1112. # No match.
This line contains the number 113312312. # Match.
This line contains no numbers at all. # No match.</PRE
></FONT
></TD
></TR
></TABLE
></P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep "1133*" tstfile</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Run grep "1133*" on this file. # Match.
This line contains the number 113. # Match.
This line contains the number 1133. # Match.
This line contains the number 113312. # Match.
This line contains the number 113312312. # Match.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
><P
></P
></DIV
></TD
></TR
></TABLE
><P
></P
><UL
><LI
STYLE="list-style-type: square"
><DIV
CLASS="FORMALPARA"
><P
><B
><A
NAME="EXTREGEX"
></A
>Extended REs. </B
>Additional metacharacters added to the basic set. Used
in <A
HREF="#EGREPREF"
>egrep</A
>,
<A
HREF="#AWKREF"
>awk</A
>, and <A
HREF="#PERLREF"
>Perl</A
>.</P
></DIV
></LI
><LI
><P
><A
NAME="QUEXREGEX"
></A
></P
><P
>The question mark -- <SPAN
CLASS="TOKEN"
>?</SPAN
> -- matches zero or
one of the previous RE. It is generally used for matching
single characters.</P
></LI
><LI
><P
><A
NAME="PLUSREF"
></A
></P
><P
>The plus -- <SPAN
CLASS="TOKEN"
>+</SPAN
> -- matches one or more of the
previous RE. It serves a role similar to the <SPAN
CLASS="TOKEN"
>*</SPAN
>, but
does <EM
>not</EM
> match zero occurrences.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># GNU versions of sed and awk can use "+",
# but it needs to be escaped.
echo a111b | sed -ne '/a1\+b/p'
echo a111b | grep 'a1\+b'
echo a111b | gawk '/a1+b/'
# All of above are equivalent.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="ESCPCB"
></A
></P
></LI
><LI
><P
><A
HREF="#ESCP"
>Escaped</A
> <SPAN
CLASS="QUOTE"
>"curly
brackets"</SPAN
> -- <SPAN
CLASS="TOKEN"
>\{ \}</SPAN
> -- indicate the number
of occurrences of a preceding RE to match.</P
><P
>It is necessary to escape the curly brackets since
they have only their literal character meaning
otherwise. This usage is technically not part of the basic
RE set.</P
><P
><SPAN
CLASS="QUOTE"
>"[0-9]\{5\}"</SPAN
> matches exactly five digits
(characters in the range of 0 to 9).</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Curly brackets are not available as an RE in the
<SPAN
CLASS="QUOTE"
>"classic"</SPAN
> (non-POSIX compliant) version
of <A
HREF="#AWKREF"
>awk</A
>.
<A
NAME="GNUGAWK"
></A
>However, the GNU extended version
of <I
CLASS="FIRSTTERM"
>awk</I
>, <B
CLASS="COMMAND"
>gawk</B
>,
has the <TT
CLASS="OPTION"
>--re-interval</TT
> option that permits
them (without being escaped).</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo 2222 | gawk --re-interval '/2{3}/'</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>2222</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><B
CLASS="COMMAND"
>Perl</B
> and some
<B
CLASS="COMMAND"
>egrep</B
> versions do not require escaping
the curly brackets.</P
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><A
NAME="PARENGRPS"
></A
></P
><P
>Parentheses -- <B
CLASS="COMMAND"
>( )</B
> -- enclose a group of
REs. They are useful with the following
<SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>|</SPAN
>"</SPAN
> operator and in <A
HREF="#EXPRPAREN"
>substring extraction</A
> using <A
HREF="#EXPRREF"
>expr</A
>.</P
></LI
><LI
><P
>The -- <B
CLASS="COMMAND"
>|</B
> -- <SPAN
CLASS="QUOTE"
>"or"</SPAN
> RE operator
matches any of a set of alternate characters.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>egrep 're(a|e)d' misc.txt</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>People who read seem to be better informed than those who do not.
The clarinet produces sound by the vibration of its reed.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
></UL
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Some versions of <B
CLASS="COMMAND"
>sed</B
>,
<B
CLASS="COMMAND"
>ed</B
>, and <B
CLASS="COMMAND"
>ex</B
> support
escaped versions of the extended Regular Expressions
described above, as do the GNU utilities.</P
></TD
></TR
></TABLE
></DIV
><P
></P
><UL
><LI
STYLE="list-style-type: square"
><DIV
CLASS="FORMALPARA"
><P
><B
><A
NAME="POSIXREF"
></A
>POSIX Character Classes. </B
><TT
CLASS="USERINPUT"
><B
>[:class:]</B
></TT
></P
></DIV
><P
>This is an alternate method of specifying a range of
characters to match.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>[:alnum:]</B
></TT
> matches alphabetic or
numeric characters. This is equivalent to
<TT
CLASS="USERINPUT"
><B
>A-Za-z0-9</B
></TT
>.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>[:alpha:]</B
></TT
> matches alphabetic
characters. This is equivalent to
<TT
CLASS="USERINPUT"
><B
>A-Za-z</B
></TT
>.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>[:blank:]</B
></TT
> matches a space or a
tab.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>[:cntrl:]</B
></TT
> matches control
characters.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>[:digit:]</B
></TT
> matches (decimal)
digits. This is equivalent to
<TT
CLASS="USERINPUT"
><B
>0-9</B
></TT
>.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>[:graph:]</B
></TT
> (graphic printable
characters). Matches characters in the range of <A
HREF="#ASCIIDEF"
>ASCII</A
> 33 - 126. This is
the same as <TT
CLASS="USERINPUT"
><B
>[:print:]</B
></TT
>, below,
but excluding the space character.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>[:lower:]</B
></TT
> matches lowercase
alphabetic characters. This is equivalent to
<TT
CLASS="USERINPUT"
><B
>a-z</B
></TT
>.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>[:print:]</B
></TT
> (printable
characters). Matches characters in the range of ASCII 32 -
126. This is the same as <TT
CLASS="USERINPUT"
><B
>[:graph:]</B
></TT
>,
above, but adding the space character.</P
></LI
><LI
><P
><A
NAME="WSPOSIX"
></A
><TT
CLASS="USERINPUT"
><B
>[:space:]</B
></TT
>
matches whitespace characters (space and horizontal
tab).</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>[:upper:]</B
></TT
> matches uppercase
alphabetic characters. This is equivalent to
<TT
CLASS="USERINPUT"
><B
>A-Z</B
></TT
>.</P
></LI
><LI
><P
><TT
CLASS="USERINPUT"
><B
>[:xdigit:]</B
></TT
> matches hexadecimal
digits. This is equivalent to
<TT
CLASS="USERINPUT"
><B
>0-9A-Fa-f</B
></TT
>.</P
><DIV
CLASS="IMPORTANT"
><P
></P
><TABLE
CLASS="IMPORTANT"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/important.gif"
HSPACE="5"
ALT="Important"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>POSIX character classes generally require quoting
or <A
HREF="#DBLBRACKETS"
>double brackets</A
>
([[ ]]).</P
></TD
></TR
></TABLE
></DIV
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep [[:digit:]] test.file</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>abc=723</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># ...
if [[ $arow =~ [[:digit:]] ]] # Numerical input?
then # POSIX char class
if [[ $acol =~ [[:alpha:]] ]] # Number followed by a letter? Illegal!
# ...
# From ktour.sh example script.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>These character classes may even be used with <A
HREF="#GLOBBINGREF"
>globbing</A
>, to a limited
extent.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l ?[[:digit:]][[:digit:]]?</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-rw-rw-r-- 1 bozo bozo 0 Aug 21 14:47 a33b</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>POSIX character classes are used in
<A
HREF="#EX49"
>Example 16-21</A
> and <A
HREF="#LOWERCASE"
>Example 16-22</A
>.</P
></LI
></UL
><P
><A
HREF="#SEDREF"
>Sed</A
>, <A
HREF="#AWKREF"
>awk</A
>, and <A
HREF="#PERLREF"
>Perl</A
>, used as filters in scripts, take
REs as arguments when "sifting" or transforming files or I/O
streams. See <A
HREF="#BEHEAD"
>Example A-12</A
> and <A
HREF="#TREE"
>Example A-16</A
>
for illustrations of this.</P
><P
>The standard reference on this complex topic is Friedl's
<I
CLASS="CITETITLE"
>Mastering Regular
Expressions</I
>. <I
CLASS="CITETITLE"
>Sed &#38;
Awk</I
>, by Dougherty and Robbins, also gives a very
lucid treatment of REs. See the <A
HREF="#BIBLIO"
><I
>Bibliography</I
></A
> for
more information on these books.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="GLOBBINGREF"
></A
>18.2. Globbing</H1
><P
><A
NAME="GLOBBINGREF2"
></A
></P
><P
>Bash itself cannot recognize Regular Expressions. Inside
scripts, it is commands and utilities -- such as
<A
HREF="#SEDREF"
>sed</A
> and <A
HREF="#AWKREF"
>awk</A
> -- that interpret RE's.</P
><P
>Bash <EM
>does</EM
> carry out <I
CLASS="FIRSTTERM"
>filename
expansion</I
>
<A
NAME="AEN17572"
HREF="#FTN.AEN17572"
><SPAN
CLASS="footnote"
>[100]</SPAN
></A
>
-- a process known as <I
CLASS="FIRSTTERM"
>globbing</I
> -- but
this does <EM
>not</EM
> use the standard RE set.
Instead, globbing recognizes and expands <I
CLASS="FIRSTTERM"
>wild
cards</I
>. Globbing interprets the standard wild
card characters
<A
NAME="AEN17581"
HREF="#FTN.AEN17581"
><SPAN
CLASS="footnote"
>[101]</SPAN
></A
>
-- <A
HREF="#ASTERISKREF"
>*</A
> and
<A
HREF="#WILDCARDQU"
>?</A
>, character lists in
square brackets, and certain other special characters (such
as <SPAN
CLASS="TOKEN"
>^</SPAN
> for negating the sense of a match).
<A
NAME="WDOTFILEWC"
></A
>There are important limitations on wild
card characters in globbing, however. Strings containing
<TT
CLASS="REPLACEABLE"
><I
>*</I
></TT
> will not match filenames that
start with a dot, as, for example, <A
HREF="#SAMPLE-BASHRC"
><TT
CLASS="FILENAME"
>.bashrc</TT
></A
>.
<A
NAME="AEN17592"
HREF="#FTN.AEN17592"
><SPAN
CLASS="footnote"
>[102]</SPAN
></A
>
Likewise, the <TT
CLASS="REPLACEABLE"
><I
>?</I
></TT
> has a different
meaning in globbing than as part of an RE.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 2
-rw-rw-r-- 1 bozo bozo 0 Aug 6 18:42 a.1
-rw-rw-r-- 1 bozo bozo 0 Aug 6 18:42 b.1
-rw-rw-r-- 1 bozo bozo 0 Aug 6 18:42 c.1
-rw-rw-r-- 1 bozo bozo 466 Aug 6 17:48 t2.sh
-rw-rw-r-- 1 bozo bozo 758 Jul 30 09:02 test1.txt</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l t?.sh</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-rw-rw-r-- 1 bozo bozo 466 Aug 6 17:48 t2.sh</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l [ab]*</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-rw-rw-r-- 1 bozo bozo 0 Aug 6 18:42 a.1
-rw-rw-r-- 1 bozo bozo 0 Aug 6 18:42 b.1</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l [a-c]*</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-rw-rw-r-- 1 bozo bozo 0 Aug 6 18:42 a.1
-rw-rw-r-- 1 bozo bozo 0 Aug 6 18:42 b.1
-rw-rw-r-- 1 bozo bozo 0 Aug 6 18:42 c.1</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l [^ab]*</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-rw-rw-r-- 1 bozo bozo 0 Aug 6 18:42 c.1
-rw-rw-r-- 1 bozo bozo 466 Aug 6 17:48 t2.sh
-rw-rw-r-- 1 bozo bozo 758 Jul 30 09:02 test1.txt</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>ls -l {b*,c*,*est*}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-rw-rw-r-- 1 bozo bozo 0 Aug 6 18:42 b.1
-rw-rw-r-- 1 bozo bozo 0 Aug 6 18:42 c.1
-rw-rw-r-- 1 bozo bozo 758 Jul 30 09:02 test1.txt</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Bash performs filename expansion on unquoted command-line
arguments. The <A
HREF="#ECHOREF"
>echo</A
> command
demonstrates this.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo *</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>a.1 b.1 c.1 t2.sh test1.txt</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo t*</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>t2.sh test1.txt</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo t?.sh</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>t2.sh</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>It is possible to modify the way Bash interprets
special characters in globbing. A <B
CLASS="COMMAND"
>set -f</B
>
command disables globbing, and the
<TT
CLASS="OPTION"
>nocaseglob</TT
> and <TT
CLASS="OPTION"
>nullglob</TT
>
options to <A
HREF="#SHOPTREF"
>shopt</A
> change
globbing behavior.</P
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#LISTGLOB"
>Example 11-5</A
>.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="HANDLINGFNAMES"
></A
>Filenames with
embedded <A
HREF="#WHITESPACEREF"
>whitespace</A
>
can cause <I
CLASS="FIRSTTERM"
>globbing</I
> to choke.
<A
HREF="http://www.dwheeler.com/essays/filenames-in-shell.html"
TARGET="_top"
>David
Wheeler</A
> shows how to avoid many such pitfalls.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>IFS="$(printf '\n\t')" # Remove space.
# Correct glob use:
# Always use for-loop, prefix glob, check if exists file.
for file in ./* ; do # Use ./* ... NEVER bare *
if [ -e "$file" ] ; then # Check whether file exists.
COMMAND ... "$file" ...
fi
done
# This example taken from David Wheeler's site, with permission.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="HERE-DOCS"
></A
>Chapter 19. Here Documents</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>Here and now, boys.</I
></P
><P
><I
>--Aldous Huxley, <I
CLASS="FIRSTTERM"
>Island</I
></I
></P
></I
></TD
></TR
></TABLE
><P
><A
NAME="HEREDOCREF"
></A
></P
><P
>A <I
CLASS="FIRSTTERM"
>here document</I
> is a special-purpose
code block. It uses a form of <A
HREF="#IOREDIRREF"
>I/O
redirection</A
> to feed a command list to
an interactive program or a command, such as <A
HREF="#FTPREF"
>ftp</A
>, <A
HREF="#CATREF"
>cat</A
>,
or the <I
CLASS="FIRSTTERM"
>ex</I
> text editor.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>COMMAND &#60;&#60;InputComesFromHERE
...
...
...
InputComesFromHERE</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="LIMITSTRINGREF"
></A
></P
><P
>A <I
CLASS="FIRSTTERM"
>limit string</I
> delineates (frames)
the command list. The special symbol <SPAN
CLASS="TOKEN"
>&#60;&#60;</SPAN
> precedes
the limit string. This has the effect of redirecting the output
of a command block into the <TT
CLASS="FILENAME"
>stdin</TT
> of the program
or command. It is similar to <TT
CLASS="USERINPUT"
><B
>interactive-program &#60;
command-file</B
></TT
>, where <TT
CLASS="FILENAME"
>command-file</TT
>
contains
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>command #1
command #2
...</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>The <I
CLASS="FIRSTTERM"
>here document</I
> equivalent looks
like this:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>interactive-program &#60;&#60;LimitString
command #1
command #2
...
LimitString</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Choose a <I
CLASS="FIRSTTERM"
>limit string</I
> sufficiently
unusual that it will not occur anywhere in the command list and
confuse matters.</P
><P
>Note that <I
CLASS="FIRSTTERM"
>here documents</I
> may sometimes
be used to good effect with non-interactive utilities and commands,
such as, for example, <A
HREF="#WALLREF"
>wall</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX70"
></A
><P
><B
>Example 19-1. <I
CLASS="FIRSTTERM"
>broadcast</I
>: Sends message to everyone
logged in</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
wall &#60;&#60;zzz23EndOfMessagezzz23
E-mail your noontime orders for pizza to the system administrator.
(Add an extra dollar for anchovy or mushroom topping.)
# Additional message text goes here.
# Note: 'wall' prints comment lines.
zzz23EndOfMessagezzz23
# Could have been done more efficiently by
# wall &#60;message-file
# However, embedding the message template in a script
#+ is a quick-and-dirty one-off solution.
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="VIHERE"
></A
></P
><P
>Even such unlikely candidates as the <I
CLASS="FIRSTTERM"
>vi</I
>
text editor lend themselves to <I
CLASS="FIRSTTERM"
>here
documents</I
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX69"
></A
><P
><B
>Example 19-2. <I
CLASS="FIRSTTERM"
>dummyfile</I
>: Creates a 2-line dummy
file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Noninteractive use of 'vi' to edit a file.
# Emulates 'sed'.
E_BADARGS=85
if [ -z "$1" ]
then
echo "Usage: `basename $0` filename"
exit $E_BADARGS
fi
TARGETFILE=$1
# Insert 2 lines in file, then save.
#--------Begin here document-----------#
vi $TARGETFILE &#60;&#60;x23LimitStringx23
i
This is line 1 of the example file.
This is line 2 of the example file.
^[
ZZ
x23LimitStringx23
#----------End here document-----------#
# Note that ^[ above is a literal escape
#+ typed by Control-V &#60;Esc&#62;.
# Bram Moolenaar points out that this may not work with 'vim'
#+ because of possible problems with terminal interaction.
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
> The above script could just as effectively have been implemented with
<B
CLASS="COMMAND"
>ex</B
>, rather than
<B
CLASS="COMMAND"
>vi</B
>. <A
NAME="EXSCRIPTREF"
></A
><I
CLASS="FIRSTTERM"
>Here
documents</I
> containing a list of <B
CLASS="COMMAND"
>ex</B
>
commands are common enough to form their own category, known as
<I
CLASS="FIRSTTERM"
>ex scripts</I
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Replace all instances of "Smith" with "Jones"
#+ in files with a ".txt" filename suffix.
ORIGINAL=Smith
REPLACEMENT=Jones
for word in $(fgrep -l $ORIGINAL *.txt)
do
# -------------------------------------
ex $word &#60;&#60;EOF
:%s/$ORIGINAL/$REPLACEMENT/g
:wq
EOF
# :%s is the "ex" substitution command.
# :wq is write-and-quit.
# -------------------------------------
done</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="CATSCRIPTREF"
></A
></P
><P
>Analogous to <SPAN
CLASS="QUOTE"
>"ex scripts"</SPAN
> are <I
CLASS="FIRSTTERM"
>cat
scripts</I
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX71"
></A
><P
><B
>Example 19-3. Multi-line message using <I
CLASS="FIRSTTERM"
>cat</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# 'echo' is fine for printing single line messages,
#+ but somewhat problematic for for message blocks.
# A 'cat' here document overcomes this limitation.
cat &#60;&#60;End-of-message
-------------------------------------
This is line 1 of the message.
This is line 2 of the message.
This is line 3 of the message.
This is line 4 of the message.
This is the last line of the message.
-------------------------------------
End-of-message
# Replacing line 7, above, with
#+ cat &#62; $Newfile &#60;&#60;End-of-message
#+ ^^^^^^^^^^
#+ writes the output to the file $Newfile, rather than to stdout.
exit 0
#--------------------------------------------
# Code below disabled, due to "exit 0" above.
# S.C. points out that the following also works.
echo "-------------------------------------
This is line 1 of the message.
This is line 2 of the message.
This is line 3 of the message.
This is line 4 of the message.
This is the last line of the message.
-------------------------------------"
# However, text may not include double quotes unless they are escaped.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="LIMITSTRDASH"
></A
></P
><P
>The <TT
CLASS="OPTION"
>-</TT
> option to mark a here document limit string
(<TT
CLASS="USERINPUT"
><B
>&#60;&#60;-LimitString</B
></TT
>) suppresses leading
tabs (but not spaces) in the output. This may be useful in making
a script more readable.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX71A"
></A
><P
><B
>Example 19-4. Multi-line message, with tabs suppressed</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Same as previous example, but...
# The - option to a here document &#60;&#60;-
#+ suppresses leading tabs in the body of the document,
#+ but *not* spaces.
cat &#60;&#60;-ENDOFMESSAGE
This is line 1 of the message.
This is line 2 of the message.
This is line 3 of the message.
This is line 4 of the message.
This is the last line of the message.
ENDOFMESSAGE
# The output of the script will be flush left.
# Leading tab in each line will not show.
# Above 5 lines of "message" prefaced by a tab, not spaces.
# Spaces not affected by &#60;&#60;- .
# Note that this option has no effect on *embedded* tabs.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="HEREPASSP"
></A
></P
><P
>A <I
CLASS="FIRSTTERM"
>here document</I
> supports parameter and
command substitution. It is therefore possible to pass different
parameters to the body of the here document, changing its output
accordingly.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX71B"
></A
><P
><B
>Example 19-5. Here document with replaceable parameters</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Another 'cat' here document, using parameter substitution.
# Try it with no command-line parameters, ./scriptname
# Try it with one command-line parameter, ./scriptname Mortimer
# Try it with one two-word quoted command-line parameter,
# ./scriptname "Mortimer Jones"
CMDLINEPARAM=1 # Expect at least command-line parameter.
if [ $# -ge $CMDLINEPARAM ]
then
NAME=$1 # If more than one command-line param,
#+ then just take the first.
else
NAME="John Doe" # Default, if no command-line parameter.
fi
RESPONDENT="the author of this fine script"
cat &#60;&#60;Endofmessage
Hello, there, $NAME.
Greetings to you, $NAME, from $RESPONDENT.
# This comment shows up in the output (why?).
Endofmessage
# Note that the blank lines show up in the output.
# So does the comment.
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="HEREPARAMSUB"
></A
></P
><P
>This is a useful script containing a <I
CLASS="FIRSTTERM"
>here
document</I
> with parameter substitution.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX72"
></A
><P
><B
>Example 19-6. Upload a file pair to <I
CLASS="FIRSTTERM"
>Sunsite</I
> incoming
directory</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# upload.sh
# Upload file pair (Filename.lsm, Filename.tar.gz)
#+ to incoming directory at Sunsite/UNC (ibiblio.org).
# Filename.tar.gz is the tarball itself.
# Filename.lsm is the descriptor file.
# Sunsite requires "lsm" file, otherwise will bounce contributions.
E_ARGERROR=85
if [ -z "$1" ]
then
echo "Usage: `basename $0` Filename-to-upload"
exit $E_ARGERROR
fi
Filename=`basename $1` # Strips pathname out of file name.
Server="ibiblio.org"
Directory="/incoming/Linux"
# These need not be hard-coded into script,
#+ but may instead be changed to command-line argument.
Password="your.e-mail.address" # Change above to suit.
ftp -n $Server &#60;&#60;End-Of-Session
# -n option disables auto-logon
user anonymous "$Password" # If this doesn't work, then try:
# quote user anonymous "$Password"
binary
bell # Ring 'bell' after each file transfer.
cd $Directory
put "$Filename.lsm"
put "$Filename.tar.gz"
bye
End-Of-Session
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="HEREESC"
></A
></P
><P
>Quoting or escaping the <SPAN
CLASS="QUOTE"
>"limit string"</SPAN
> at the
head of a here document disables parameter substitution within its
body. The reason for this is that <I
CLASS="FIRSTTERM"
>quoting/escaping the
limit string</I
> effectively <A
HREF="#ESCP"
>escapes</A
> the <SPAN
CLASS="TOKEN"
>$</SPAN
>,
<SPAN
CLASS="TOKEN"
>`</SPAN
>, and <SPAN
CLASS="TOKEN"
>\</SPAN
> <A
HREF="#SCHARLIST"
>special characters</A
>, and causes them to
be interpreted literally. (Thank you, Allen Halsey, for pointing
this out.)</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX71C"
></A
><P
><B
>Example 19-7. Parameter substitution turned off</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# A 'cat' here-document, but with parameter substitution disabled.
NAME="John Doe"
RESPONDENT="the author of this fine script"
cat &#60;&#60;'Endofmessage'
Hello, there, $NAME.
Greetings to you, $NAME, from $RESPONDENT.
Endofmessage
# No parameter substitution when the "limit string" is quoted or escaped.
# Either of the following at the head of the here document would have
#+ the same effect.
# cat &#60;&#60;"Endofmessage"
# cat &#60;&#60;\Endofmessage
# And, likewise:
cat &#60;&#60;"SpecialCharTest"
Directory listing would follow
if limit string were not quoted.
`ls -l`
Arithmetic expansion would take place
if limit string were not quoted.
$((5 + 3))
A a single backslash would echo
if limit string were not quoted.
\\
SpecialCharTest
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="HERELIT"
></A
></P
><P
>Disabling parameter substitution permits outputting literal text.
Generating scripts or even program code is one use for this.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="GENERATESCRIPT"
></A
><P
><B
>Example 19-8. A script that generates another script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# generate-script.sh
# Based on an idea by Albert Reiner.
OUTFILE=generated.sh # Name of the file to generate.
# -----------------------------------------------------------
# 'Here document containing the body of the generated script.
(
cat &#60;&#60;'EOF'
#!/bin/bash
echo "This is a generated shell script."
# Note that since we are inside a subshell,
#+ we can't access variables in the "outside" script.
echo "Generated file will be named: $OUTFILE"
# Above line will not work as normally expected
#+ because parameter expansion has been disabled.
# Instead, the result is literal output.
a=7
b=3
let "c = $a * $b"
echo "c = $c"
exit 0
EOF
) &#62; $OUTFILE
# -----------------------------------------------------------
# Quoting the 'limit string' prevents variable expansion
#+ within the body of the above 'here document.'
# This permits outputting literal strings in the output file.
if [ -f "$OUTFILE" ]
then
chmod 755 $OUTFILE
# Make the generated file executable.
else
echo "Problem in creating file: \"$OUTFILE\""
fi
# This method also works for generating
#+ C programs, Perl programs, Python programs, Makefiles,
#+ and the like.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="HERECS"
></A
></P
><P
> It is possible to set a variable from the output of a here document.
This is actually a devious form of <A
HREF="#COMMANDSUBREF"
>command substitution</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>variable=$(cat &#60;&#60;SETVAR
This variable
runs over multiple lines.
SETVAR
)
echo "$variable"</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="HEREFUNC"
></A
></P
><P
>A here document can supply input to a function in the same
script.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="HF"
></A
><P
><B
>Example 19-9. Here documents and functions</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# here-function.sh
GetPersonalData ()
{
read firstname
read lastname
read address
read city
read state
read zipcode
} # This certainly appears to be an interactive function, but . . .
# Supply input to the above function.
GetPersonalData &#60;&#60;RECORD001
Bozo
Bozeman
2726 Nondescript Dr.
Bozeman
MT
21226
RECORD001
echo
echo "$firstname $lastname"
echo "$address"
echo "$city, $state $zipcode"
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ANONHEREDOC0"
></A
></P
><P
>It is possible to use <SPAN
CLASS="TOKEN"
>:</SPAN
> as a dummy command
accepting output from a here document. This, in effect, creates an
<SPAN
CLASS="QUOTE"
>"anonymous"</SPAN
> here document.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ANONHEREDOC"
></A
><P
><B
>Example 19-10. <SPAN
CLASS="QUOTE"
>"Anonymous"</SPAN
> Here Document</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
: &#60;&#60;TESTVARIABLES
${HOSTNAME?}${USER?}${MAIL?} # Print error message if one of the variables not set.
TESTVARIABLES
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="CBLOCK1"
></A
></P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A variation of the above technique permits <SPAN
CLASS="QUOTE"
>"commenting
out"</SPAN
> blocks of code.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="COMMENTBLOCK"
></A
><P
><B
>Example 19-11. Commenting out a block of code</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# commentblock.sh
: &#60;&#60;COMMENTBLOCK
echo "This line will not echo."
This is a comment line missing the "#" prefix.
This is another comment line missing the "#" prefix.
&#38;*@!!++=
The above line will cause no error message,
because the Bash interpreter will ignore it.
COMMENTBLOCK
echo "Exit value of above \"COMMENTBLOCK\" is $?." # 0
# No error shown.
echo
# The above technique also comes in useful for commenting out
#+ a block of working code for debugging purposes.
# This saves having to put a "#" at the beginning of each line,
#+ then having to go back and delete each "#" later.
# Note that the use of of colon, above, is optional.
echo "Just before commented-out code block."
# The lines of code between the double-dashed lines will not execute.
# ===================================================================
: &#60;&#60;DEBUGXXX
for file in *
do
cat "$file"
done
DEBUGXXX
# ===================================================================
echo "Just after commented-out code block."
exit 0
######################################################################
# Note, however, that if a bracketed variable is contained within
#+ the commented-out code block,
#+ then this could cause problems.
# for example:
#/!/bin/bash
: &#60;&#60;COMMENTBLOCK
echo "This line will not echo."
&#38;*@!!++=
${foo_bar_bazz?}
$(rm -rf /tmp/foobar/)
$(touch my_build_directory/cups/Makefile)
COMMENTBLOCK
$ sh commented-bad.sh
commented-bad.sh: line 3: foo_bar_bazz: parameter null or not set
# The remedy for this is to strong-quote the 'COMMENTBLOCK' in line 49, above.
: &#60;&#60;'COMMENTBLOCK'
# Thank you, Kurt Pfeifle, for pointing this out.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="HSELFDOC"
></A
></P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Yet another twist of this nifty trick makes
<SPAN
CLASS="QUOTE"
>"self-documenting"</SPAN
> scripts possible.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="SELFDOCUMENT"
></A
><P
><B
>Example 19-12. A self-documenting script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# self-document.sh: self-documenting script
# Modification of "colm.sh".
DOC_REQUEST=70
if [ "$1" = "-h" -o "$1" = "--help" ] # Request help.
then
echo; echo "Usage: $0 [directory-name]"; echo
sed --silent -e '/DOCUMENTATIONXX$/,/^DOCUMENTATIONXX$/p' "$0" |
sed -e '/DOCUMENTATIONXX$/d'; exit $DOC_REQUEST; fi
: &#60;&#60;DOCUMENTATIONXX
List the statistics of a specified directory in tabular format.
---------------------------------------------------------------
The command-line parameter gives the directory to be listed.
If no directory specified or directory specified cannot be read,
then list the current working directory.
DOCUMENTATIONXX
if [ -z "$1" -o ! -r "$1" ]
then
directory=.
else
directory="$1"
fi
echo "Listing of "$directory":"; echo
(printf "PERMISSIONS LINKS OWNER GROUP SIZE MONTH DAY HH:MM PROG-NAME\n" \
; ls -l "$directory" | sed 1d) | column -t
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Using a <A
HREF="#CATSCRIPTREF"
>cat script</A
> is an
alternate way of accomplishing this.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>DOC_REQUEST=70
if [ "$1" = "-h" -o "$1" = "--help" ] # Request help.
then # Use a "cat script" . . .
cat &#60;&#60;DOCUMENTATIONXX
List the statistics of a specified directory in tabular format.
---------------------------------------------------------------
The command-line parameter gives the directory to be listed.
If no directory specified or directory specified cannot be read,
then list the current working directory.
DOCUMENTATIONXX
exit $DOC_REQUEST
fi</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>See also <A
HREF="#ISSPAMMER2"
>Example A-28</A
>, <A
HREF="#PETALS"
>Example A-40</A
>,
<A
HREF="#QKY"
>Example A-41</A
>, and <A
HREF="#NIM"
>Example A-42</A
> for more examples
of self-documenting scripts.</P
><P
><A
NAME="HERETEMP"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Here documents create temporary files, but these
files are deleted after opening and are not accessible to
any other process.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>bash -c 'lsof -a -p $$ -d0' &#60;&#60; EOF</B
></TT
>
<TT
CLASS="PROMPT"
>&#62; </TT
><TT
CLASS="USERINPUT"
><B
>EOF</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>lsof 1213 bozo 0r REG 3,5 0 30386 /tmp/t1213-0-sh (deleted)</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Some utilities will not work inside a
<I
CLASS="FIRSTTERM"
>here document</I
>.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="INDENTEDLS"
></A
></P
><DIV
CLASS="WARNING"
><P
></P
><TABLE
CLASS="WARNING"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/warning.gif"
HSPACE="5"
ALT="Warning"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The closing <I
CLASS="FIRSTTERM"
>limit string</I
>,
on the final line of a here document, must start in the
<EM
>first</EM
> character position. There can
be <EM
>no leading whitespace</EM
>. Trailing
whitespace after the limit string likewise causes unexpected
behavior. The whitespace prevents the limit string from being
recognized.
<A
NAME="AEN17822"
HREF="#FTN.AEN17822"
><SPAN
CLASS="footnote"
>[103]</SPAN
></A
>
</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo "----------------------------------------------------------------------"
cat &#60;&#60;LimitString
echo "This is line 1 of the message inside the here document."
echo "This is line 2 of the message inside the here document."
echo "This is the final line of the message inside the here document."
LimitString
#^^^^Indented limit string. Error! This script will not behave as expected.
echo "----------------------------------------------------------------------"
# These comments are outside the 'here document',
#+ and should not echo.
echo "Outside the here document."
exit 0
echo "This line had better not echo." # Follows an 'exit' command.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="EXCLLS"
></A
>Some people very cleverly use a
single <SPAN
CLASS="TOKEN"
>!</SPAN
> as a limit string. But, that's not
necessarily a good idea.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># This works.
cat &#60;&#60;!
Hello!
! Three more exclamations !!!
!
# But . . .
cat &#60;&#60;!
Hello!
Single exclamation point follows!
!
!
# Crashes with an error message.
# However, the following will work.
cat &#60;&#60;EOF
Hello!
Single exclamation point follows!
!
EOF
# It's safer to use a multi-character limit string.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><P
>For those tasks too complex for a <I
CLASS="FIRSTTERM"
>here
document</I
>, consider using the
<TT
CLASS="REPLACEABLE"
><I
>expect</I
></TT
> scripting language, which
was specifically designed for feeding input into interactive
programs.</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="AEN17837"
></A
>19.1. Here Strings</H1
><P
><A
NAME="HERESTRINGSREF"
></A
></P
><A
NAME="AEN17841"
></A
><BLOCKQUOTE
CLASS="BLOCKQUOTE"
><P
CLASS="LITERALLAYOUT"
>A&nbsp;<I
CLASS="FIRSTTERM"
>here string</I
>&nbsp;can&nbsp;be&nbsp;considered&nbsp;as&nbsp;a&nbsp;stripped-down&nbsp;form&nbsp;of&nbsp;a&nbsp;<I
CLASS="FIRSTTERM"
>here document</I
>.<br>
It&nbsp;consists&nbsp;of&nbsp;nothing&nbsp;more&nbsp;than&nbsp;<B
CLASS="COMMAND"
>COMMAND &#60;&#60;&#60; $WORD</B
>,<br>
where&nbsp;<TT
CLASS="VARNAME"
>$WORD</TT
>&nbsp;is&nbsp;expanded&nbsp;and&nbsp;fed&nbsp;to&nbsp;the&nbsp;<TT
CLASS="FILENAME"
>stdin</TT
>&nbsp;of&nbsp;<B
CLASS="COMMAND"
>COMMAND</B
>.<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</P
></BLOCKQUOTE
><P
>As a simple example, consider this alternative to the <A
HREF="#ECHOGREPREF"
>echo-grep</A
> construction.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Instead of:
if echo "$VAR" | grep -q txt # if [[ $VAR = *txt* ]]
# etc.
# Try:
if grep -q "txt" &#60;&#60;&#60; "$VAR"
then # ^^^
echo "$VAR contains the substring sequence \"txt\""
fi
# Thank you, Sebastian Kaminski, for the suggestion.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="HSREAD"
></A
></P
><P
>Or, in combination with <A
HREF="#READREF"
>read</A
>:</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>String="This is a string of words."
read -r -a Words &#60;&#60;&#60; "$String"
# The -a option to "read"
#+ assigns the resulting values to successive members of an array.
echo "First word in String is: ${Words[0]}" # This
echo "Second word in String is: ${Words[1]}" # is
echo "Third word in String is: ${Words[2]}" # a
echo "Fourth word in String is: ${Words[3]}" # string
echo "Fifth word in String is: ${Words[4]}" # of
echo "Sixth word in String is: ${Words[5]}" # words.
echo "Seventh word in String is: ${Words[6]}" # (null)
# Past end of $String.
# Thank you, Francisco Lobo, for the suggestion.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="HSLOOP"
></A
>It is, of course, possible to feed
the output of a <I
CLASS="FIRSTTERM"
>here string</I
>
into the <TT
CLASS="FILENAME"
>stdin</TT
> of a <A
HREF="#LOOPREF00"
>loop</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># As Seamus points out . . .
ArrayVar=( element0 element1 element2 {A..D} )
while read element ; do
echo "$element" 1&#62;&#38;2
done &#60;&#60;&#60; $(echo ${ArrayVar[*]})
# element0 element1 element2 A B C D</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="HSPRE"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="PREPENDEX"
></A
><P
><B
>Example 19-13. Prepending a line to a file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# prepend.sh: Add text at beginning of file.
#
# Example contributed by Kenny Stauffer,
#+ and slightly modified by document author.
E_NOSUCHFILE=85
read -p "File: " file # -p arg to 'read' displays prompt.
if [ ! -e "$file" ]
then # Bail out if no such file.
echo "File $file not found."
exit $E_NOSUCHFILE
fi
read -p "Title: " title
cat - $file &#60;&#60;&#60;$title &#62; $file.new
echo "Modified file is $file.new"
exit # Ends script execution.
from 'man bash':
Here Strings
A variant of here documents, the format is:
&#60;&#60;&#60;word
The word is expanded and supplied to the command on its standard input.
Of course, the following also works:
sed -e '1i\
Title: ' $file</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="MAILBOXGREP"
></A
><P
><B
>Example 19-14. Parsing a mailbox</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Script by Francisco Lobo,
#+ and slightly modified and commented by ABS Guide author.
# Used in ABS Guide with permission. (Thank you!)
# This script will not run under Bash versions -lt 3.0.
E_MISSING_ARG=87
if [ -z "$1" ]
then
echo "Usage: $0 mailbox-file"
exit $E_MISSING_ARG
fi
mbox_grep() # Parse mailbox file.
{
declare -i body=0 match=0
declare -a date sender
declare mail header value
while IFS= read -r mail
# ^^^^ Reset $IFS.
# Otherwise "read" will strip leading &#38; trailing space from its input.
do
if [[ $mail =~ ^From ]] # Match "From" field in message.
then
(( body = 0 )) # "Zero out" variables.
(( match = 0 ))
unset date
elif (( body ))
then
(( match ))
# echo "$mail"
# Uncomment above line if you want entire body
#+ of message to display.
elif [[ $mail ]]; then
IFS=: read -r header value &#60;&#60;&#60; "$mail"
# ^^^ "here string"
case "$header" in
[Ff][Rr][Oo][Mm] ) [[ $value =~ "$2" ]] &#38;&#38; (( match++ )) ;;
# Match "From" line.
[Dd][Aa][Tt][Ee] ) read -r -a date &#60;&#60;&#60; "$value" ;;
# ^^^
# Match "Date" line.
[Rr][Ee][Cc][Ee][Ii][Vv][Ee][Dd] ) read -r -a sender &#60;&#60;&#60; "$value" ;;
# ^^^
# Match IP Address (may be spoofed).
esac
else
(( body++ ))
(( match )) &#38;&#38;
echo "MESSAGE ${date:+of: ${date[*]} }"
# Entire $date array ^
echo "IP address of sender: ${sender[1]}"
# Second field of "Received" line ^
fi
done &#60; "$1" # Redirect stdout of file into loop.
}
mbox_grep "$1" # Send mailbox file to function.
exit $?
# Exercises:
# ---------
# 1) Break the single function, above, into multiple functions,
#+ for the sake of readability.
# 2) Add additional parsing to the script, checking for various keywords.
$ mailbox_grep.sh scam_mail
MESSAGE of Thu, 5 Jan 2006 08:00:56 -0500 (EST)
IP address of sender: 196.3.62.4</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Exercise: Find other uses for <I
CLASS="FIRSTTERM"
>here
strings</I
>, such as, for example, <A
HREF="#GOLDENRATIO"
>feeding input to
<I
CLASS="FIRSTTERM"
>dc</I
></A
>.</P
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="IO-REDIRECTION"
></A
>Chapter 20. I/O Redirection</H1
><P
><A
NAME="IOREDIRREF"
></A
></P
><P
>There are always three default <I
CLASS="FIRSTTERM"
>files</I
>
<A
NAME="AEN17884"
HREF="#FTN.AEN17884"
><SPAN
CLASS="footnote"
>[104]</SPAN
></A
>
open, <TT
CLASS="FILENAME"
>stdin</TT
> (the keyboard),
<TT
CLASS="FILENAME"
>stdout</TT
> (the screen), and
<TT
CLASS="FILENAME"
>stderr</TT
> (error messages output to the
screen). These, and any other open files, can be redirected.
Redirection simply means capturing output from a file, command,
program, script, or even code block within a script (see <A
HREF="#EX8"
>Example 3-1</A
> and <A
HREF="#RPMCHECK"
>Example 3-2</A
>) and sending it as
input to another file, command, program, or script.</P
><P
><A
NAME="FDREF"
></A
>Each open file gets assigned a file descriptor.
<A
NAME="AEN17894"
HREF="#FTN.AEN17894"
><SPAN
CLASS="footnote"
>[105]</SPAN
></A
>
The file descriptors for <TT
CLASS="FILENAME"
>stdin</TT
>,
<TT
CLASS="FILENAME"
>stdout</TT
>, and <TT
CLASS="FILENAME"
>stderr</TT
> are
0, 1, and 2, respectively. For opening additional files, there
remain descriptors 3 to 9. It is sometimes useful to assign one of
these additional file descriptors to <TT
CLASS="FILENAME"
>stdin</TT
>,
<TT
CLASS="FILENAME"
>stdout</TT
>, or <TT
CLASS="FILENAME"
>stderr</TT
>
as a temporary duplicate link.
<A
NAME="AEN17906"
HREF="#FTN.AEN17906"
><SPAN
CLASS="footnote"
>[106]</SPAN
></A
>
This simplifies restoration to normal after complex redirection
and reshuffling (see <A
HREF="#REDIR1"
>Example 20-1</A
>).</P
><P
><A
NAME="IOREDIRECTIONREF"
></A
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> COMMAND_OUTPUT &#62;
# Redirect stdout to a file.
# Creates the file if not present, otherwise overwrites it.
ls -lR &#62; dir-tree.list
# Creates a file containing a listing of the directory tree.
: &#62; filename
# The &#62; truncates file "filename" to zero length.
# If file not present, creates zero-length file (same effect as 'touch').
# The : serves as a dummy placeholder, producing no output.
&#62; filename
# The &#62; truncates file "filename" to zero length.
# If file not present, creates zero-length file (same effect as 'touch').
# (Same result as ": &#62;", above, but this does not work with some shells.)
COMMAND_OUTPUT &#62;&#62;
# Redirect stdout to a file.
# Creates the file if not present, otherwise appends to it.
# Single-line redirection commands (affect only the line they are on):
# --------------------------------------------------------------------
1&#62;filename
# Redirect stdout to file "filename."
1&#62;&#62;filename
# Redirect and append stdout to file "filename."
2&#62;filename
# Redirect stderr to file "filename."
2&#62;&#62;filename
# Redirect and append stderr to file "filename."
&#38;&#62;filename
# Redirect both stdout and stderr to file "filename."
# This operator is now functional, as of Bash 4, final release.
M&#62;N
# "M" is a file descriptor, which defaults to 1, if not explicitly set.
# "N" is a filename.
# File descriptor "M" is redirect to file "N."
M&#62;&#38;N
# "M" is a file descriptor, which defaults to 1, if not set.
# "N" is another file descriptor.
#==============================================================================
# Redirecting stdout, one line at a time.
LOGFILE=script.log
echo "This statement is sent to the log file, \"$LOGFILE\"." 1&#62;$LOGFILE
echo "This statement is appended to \"$LOGFILE\"." 1&#62;&#62;$LOGFILE
echo "This statement is also appended to \"$LOGFILE\"." 1&#62;&#62;$LOGFILE
echo "This statement is echoed to stdout, and will not appear in \"$LOGFILE\"."
# These redirection commands automatically "reset" after each line.
# Redirecting stderr, one line at a time.
ERRORFILE=script.errors
bad_command1 2&#62;$ERRORFILE # Error message sent to $ERRORFILE.
bad_command2 2&#62;&#62;$ERRORFILE # Error message appended to $ERRORFILE.
bad_command3 # Error message echoed to stderr,
#+ and does not appear in $ERRORFILE.
# These redirection commands also automatically "reset" after each line.
#=======================================================================</PRE
></FONT
></TD
></TR
></TABLE
><P
><A
NAME="IOREDIRECTIONREF1"
></A
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> 2&#62;&#38;1
# Redirects stderr to stdout.
# Error messages get sent to same place as standard output.
&#62;&#62;filename 2&#62;&#38;1
bad_command &#62;&#62;filename 2&#62;&#38;1
# Appends both stdout and stderr to the file "filename" ...
2&#62;&#38;1 | [command(s)]
bad_command 2&#62;&#38;1 | awk '{print $5}' # found
# Sends stderr through a pipe.
# |&#38; was added to Bash 4 as an abbreviation for 2&#62;&#38;1 |.
i&#62;&#38;j
# Redirects file descriptor <EM
>i</EM
> to <EM
>j</EM
>.
# All output of file pointed to by <EM
>i</EM
> gets sent to file pointed to by <EM
>j</EM
>.
&#62;&#38;j
# Redirects, by default, file descriptor <EM
>1</EM
> (stdout) to <EM
>j</EM
>.
# All stdout gets sent to file pointed to by <EM
>j</EM
>.</PRE
></FONT
></TD
></TR
></TABLE
><P
><A
NAME="IOREDIRECTIONREF2"
></A
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> 0&#60; FILENAME
&#60; FILENAME
# Accept input from a file.
# Companion command to <SPAN
CLASS="QUOTE"
>"&#62;"</SPAN
>, and often used in combination with it.
#
# grep search-word &#60;filename
[j]&#60;&#62;filename
# Open file "filename" for reading and writing,
#+ and assign file descriptor "j" to it.
# If "filename" does not exist, create it.
# If file descriptor "j" is not specified, default to fd 0, stdin.
#
# An application of this is writing at a specified place in a file.
echo 1234567890 &#62; File # Write string to "File".
exec 3&#60;&#62; File # Open "File" and assign fd 3 to it.
read -n 4 &#60;&#38;3 # Read only 4 characters.
echo -n . &#62;&#38;3 # Write a decimal point there.
exec 3&#62;&#38;- # Close fd 3.
cat File # ==&#62; 1234.67890
# Random access, by golly.
|
# Pipe.
# General purpose process and command chaining tool.
# Similar to <SPAN
CLASS="QUOTE"
>"&#62;"</SPAN
>, but more general in effect.
# Useful for chaining commands, scripts, files, and programs together.
cat *.txt | sort | uniq &#62; result-file
# Sorts the output of all the .txt files and deletes duplicate lines,
# finally saves results to <SPAN
CLASS="QUOTE"
>"result-file"</SPAN
>.</PRE
></FONT
></TD
></TR
></TABLE
><P
>Multiple instances of input and output redirection
and/or pipes can be combined in a single command
line.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>command &#60; input-file &#62; output-file
# Or the equivalent:
&#60; input-file command &#62; output-file # Although this is non-standard.
command1 | command2 | command3 &#62; output-file</PRE
></FONT
></TD
></TR
></TABLE
>
See <A
HREF="#DERPM"
>Example 16-31</A
> and <A
HREF="#FIFO"
>Example A-14</A
>.</P
><P
>Multiple output streams may be redirected to one file.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>ls -yz &#62;&#62; command.log 2&#62;&#38;1
# Capture result of illegal options "yz" in file "command.log."
# Because stderr is redirected to the file,
#+ any error messages will also be there.
# Note, however, that the following does *not* give the same result.
ls -yz 2&#62;&#38;1 &#62;&#62; command.log
# Outputs an error message, but does not write to file.
# More precisely, the command output (in this case, null)
#+ writes to the file, but the error message goes only to stdout.
# If redirecting both stdout and stderr,
#+ the order of the commands makes a difference.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="CFD"
></A
>Closing File Descriptors</B
></P
><DL
><DT
><SPAN
CLASS="TOKEN"
>n&#60;&#38;-</SPAN
></DT
><DD
><P
>Close input file descriptor
<TT
CLASS="REPLACEABLE"
><I
>n</I
></TT
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>0&#60;&#38;-</SPAN
>, <SPAN
CLASS="TOKEN"
>&#60;&#38;-</SPAN
></DT
><DD
><P
>Close <TT
CLASS="FILENAME"
>stdin</TT
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>n&#62;&#38;-</SPAN
></DT
><DD
><P
>Close output file descriptor <TT
CLASS="REPLACEABLE"
><I
>n</I
></TT
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>1&#62;&#38;-</SPAN
>, <SPAN
CLASS="TOKEN"
>&#62;&#38;-</SPAN
></DT
><DD
><P
>Close <TT
CLASS="FILENAME"
>stdout</TT
>.</P
></DD
></DL
></DIV
><P
><A
NAME="FDREF2"
></A
></P
><P
>Child processes inherit open file descriptors. This is
why pipes work. To prevent an fd from being inherited, close it.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Redirecting only stderr to a pipe.
exec 3&#62;&#38;1 # Save current "value" of stdout.
ls -l 2&#62;&#38;1 &#62;&#38;3 3&#62;&#38;- | grep bad 3&#62;&#38;- # Close fd 3 for 'grep' (but not 'ls').
# ^^^^ ^^^^
exec 3&#62;&#38;- # Now close it for the remainder of the script.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>For a more detailed introduction to I/O redirection see
<A
HREF="#IOREDIRINTRO"
>Appendix F</A
>.</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="AEN17974"
></A
>20.1. Using <I
CLASS="FIRSTTERM"
>exec</I
></H1
><P
><A
NAME="USINGEXECREF"
></A
></P
><P
>An <B
CLASS="COMMAND"
>exec &#60;filename</B
> command redirects
<TT
CLASS="FILENAME"
>stdin</TT
> to a file. From that point on, all
<TT
CLASS="FILENAME"
>stdin</TT
> comes from that file, rather than
its normal source (usually keyboard input). This provides a
method of reading a file line by line and possibly parsing
each line of input using <A
HREF="#SEDREF"
>sed</A
>
and/or <A
HREF="#AWKREF"
>awk</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="REDIR1"
></A
><P
><B
>Example 20-1. Redirecting <TT
CLASS="FILENAME"
>stdin</TT
> using
<I
CLASS="FIRSTTERM"
>exec</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Redirecting stdin using 'exec'.
exec 6&#60;&#38;0 # Link file descriptor #6 with stdin.
# Saves stdin.
exec &#60; data-file # stdin replaced by file "data-file"
read a1 # Reads first line of file "data-file".
read a2 # Reads second line of file "data-file."
echo
echo "Following lines read from file."
echo "-------------------------------"
echo $a1
echo $a2
echo; echo; echo
exec 0&#60;&#38;6 6&#60;&#38;-
# Now restore stdin from fd #6, where it had been saved,
#+ and close fd #6 ( 6&#60;&#38;- ) to free it for other processes to use.
#
# &#60;&#38;6 6&#60;&#38;- also works.
echo -n "Enter data "
read b1 # Now "read" functions as expected, reading from normal stdin.
echo "Input read from stdin."
echo "----------------------"
echo "b1 = $b1"
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Similarly, an <B
CLASS="COMMAND"
>exec &#62;filename</B
>
command redirects <TT
CLASS="FILENAME"
>stdout</TT
> to a designated
file. This sends all command output that would normally go
to <TT
CLASS="FILENAME"
>stdout</TT
> to that file.</P
><DIV
CLASS="IMPORTANT"
><P
></P
><TABLE
CLASS="IMPORTANT"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/important.gif"
HSPACE="5"
ALT="Important"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
> <B
CLASS="COMMAND"
>exec N &#62; filename</B
> affects the entire
script or <I
CLASS="FIRSTTERM"
>current shell</I
>. Redirection in
the <A
HREF="#PROCESSIDREF"
>PID</A
> of the script or shell
from that point on has changed. However . . .
</P
><P
> <B
CLASS="COMMAND"
>N &#62; filename</B
> affects only the newly-forked process,
not the entire script or shell.
</P
><P
>Thank you, Ahmed Darwish, for pointing this out.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="REASSIGNSTDOUT"
></A
><P
><B
>Example 20-2. Redirecting <TT
CLASS="FILENAME"
>stdout</TT
> using
<I
CLASS="FIRSTTERM"
>exec</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# reassign-stdout.sh
LOGFILE=logfile.txt
exec 6&#62;&#38;1 # Link file descriptor #6 with stdout.
# Saves stdout.
exec &#62; $LOGFILE # stdout replaced with file "logfile.txt".
# ----------------------------------------------------------- #
# All output from commands in this block sent to file $LOGFILE.
echo -n "Logfile: "
date
echo "-------------------------------------"
echo
echo "Output of \"ls -al\" command"
echo
ls -al
echo; echo
echo "Output of \"df\" command"
echo
df
# ----------------------------------------------------------- #
exec 1&#62;&#38;6 6&#62;&#38;- # Restore stdout and close file descriptor #6.
echo
echo "== stdout now restored to default == "
echo
ls -al
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="UPPERCONV"
></A
><P
><B
>Example 20-3. Redirecting both <TT
CLASS="FILENAME"
>stdin</TT
> and
<TT
CLASS="FILENAME"
>stdout</TT
> in the same script with
<I
CLASS="FIRSTTERM"
>exec</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# upperconv.sh
# Converts a specified input file to uppercase.
E_FILE_ACCESS=70
E_WRONG_ARGS=71
if [ ! -r "$1" ] # Is specified input file readable?
then
echo "Can't read from input file!"
echo "Usage: $0 input-file output-file"
exit $E_FILE_ACCESS
fi # Will exit with same error
#+ even if input file ($1) not specified (why?).
if [ -z "$2" ]
then
echo "Need to specify output file."
echo "Usage: $0 input-file output-file"
exit $E_WRONG_ARGS
fi
exec 4&#60;&#38;0
exec &#60; $1 # Will read from input file.
exec 7&#62;&#38;1
exec &#62; $2 # Will write to output file.
# Assumes output file writable (add check?).
# -----------------------------------------------
cat - | tr a-z A-Z # Uppercase conversion.
# ^^^^^ # Reads from stdin.
# ^^^^^^^^^^ # Writes to stdout.
# However, both stdin and stdout were redirected.
# Note that the 'cat' can be omitted.
# -----------------------------------------------
exec 1&#62;&#38;7 7&#62;&#38;- # Restore stout.
exec 0&#60;&#38;4 4&#60;&#38;- # Restore stdin.
# After restoration, the following line prints to stdout as expected.
echo "File \"$1\" written to \"$2\" as uppercase conversion."
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>I/O redirection is a clever way of avoiding the dreaded <A
HREF="#PARVIS"
>inaccessible variables within a subshell</A
>
problem.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="AVOIDSUBSHELL"
></A
><P
><B
>Example 20-4. Avoiding a subshell</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# avoid-subshell.sh
# Suggested by Matthew Walker.
Lines=0
echo
cat myfile.txt | while read line;
do {
echo $line
(( Lines++ )); # Incremented values of this variable
#+ inaccessible outside loop.
# Subshell problem.
}
done
echo "Number of lines read = $Lines" # 0
# Wrong!
echo "------------------------"
exec 3&#60;&#62; myfile.txt
while read line &#60;&#38;3
do {
echo "$line"
(( Lines++ )); # Incremented values of this variable
#+ accessible outside loop.
# No subshell, no problem.
}
done
exec 3&#62;&#38;-
echo "Number of lines read = $Lines" # 8
echo
exit 0
# Lines below not seen by script.
$ cat myfile.txt
Line 1.
Line 2.
Line 3.
Line 4.
Line 5.
Line 6.
Line 7.
Line 8.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="REDIRCB"
></A
>20.2. Redirecting Code Blocks</H1
><P
><A
NAME="REDIRREF"
></A
>Blocks of code, such as <A
HREF="#WHILELOOPREF"
>while</A
>, <A
HREF="#UNTILLOOPREF"
>until</A
>, and <A
HREF="#FORLOOPREF1"
>for</A
> loops, even <A
HREF="#IFTHEN"
>if/then</A
> test blocks can also incorporate
redirection of <TT
CLASS="FILENAME"
>stdin</TT
>. Even a function may
use this form of redirection (see <A
HREF="#REALNAME"
>Example 24-11</A
>).
The <SPAN
CLASS="TOKEN"
>&#60;</SPAN
> operator at the end of the code block
accomplishes this.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="REDIR2"
></A
><P
><B
>Example 20-5. Redirected <I
CLASS="FIRSTTERM"
>while</I
> loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# redir2.sh
if [ -z "$1" ]
then
Filename=names.data # Default, if no filename specified.
else
Filename=$1
fi
#+ Filename=${1:-names.data}
# can replace the above test (parameter substitution).
count=0
echo
while [ "$name" != Smith ] # Why is variable $name in quotes?
do
read name # Reads from $Filename, rather than stdin.
echo $name
let "count += 1"
done &#60;"$Filename" # Redirects stdin to file $Filename.
# ^^^^^^^^^^^^
echo; echo "$count names read"; echo
exit 0
# Note that in some older shell scripting languages,
#+ the redirected loop would run as a subshell.
# Therefore, $count would return 0, the initialized value outside the loop.
# Bash and ksh avoid starting a subshell *whenever possible*,
#+ so that this script, for example, runs correctly.
# (Thanks to Heiner Steven for pointing this out.)
# However . . .
# Bash *can* sometimes start a subshell in a PIPED "while-read" loop,
#+ as distinct from a REDIRECTED "while" loop.
abc=hi
echo -e "1\n2\n3" | while read l
do abc="$l"
echo $abc
done
echo $abc
# Thanks, Bruno de Oliveira Schneider, for demonstrating this
#+ with the above snippet of code.
# And, thanks, Brian Onn, for correcting an annotation error.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="REDIR2A"
></A
><P
><B
>Example 20-6. Alternate form of redirected <I
CLASS="FIRSTTERM"
>while</I
> loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# This is an alternate form of the preceding script.
# Suggested by Heiner Steven
#+ as a workaround in those situations when a redirect loop
#+ runs as a subshell, and therefore variables inside the loop
# +do not keep their values upon loop termination.
if [ -z "$1" ]
then
Filename=names.data # Default, if no filename specified.
else
Filename=$1
fi
exec 3&#60;&#38;0 # Save stdin to file descriptor 3.
exec 0&#60;"$Filename" # Redirect standard input.
count=0
echo
while [ "$name" != Smith ]
do
read name # Reads from redirected stdin ($Filename).
echo $name
let "count += 1"
done # Loop reads from file $Filename
#+ because of line 20.
# The original version of this script terminated the "while" loop with
#+ done &#60;"$Filename"
# Exercise:
# Why is this unnecessary?
exec 0&#60;&#38;3 # Restore old stdin.
exec 3&#60;&#38;- # Close temporary fd 3.
echo; echo "$count names read"; echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="REDIR3"
></A
><P
><B
>Example 20-7. Redirected <I
CLASS="FIRSTTERM"
>until</I
> loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Same as previous example, but with "until" loop.
if [ -z "$1" ]
then
Filename=names.data # Default, if no filename specified.
else
Filename=$1
fi
# while [ "$name" != Smith ]
until [ "$name" = Smith ] # Change != to =.
do
read name # Reads from $Filename, rather than stdin.
echo $name
done &#60;"$Filename" # Redirects stdin to file $Filename.
# ^^^^^^^^^^^^
# Same results as with "while" loop in previous example.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="REDIR4"
></A
><P
><B
>Example 20-8. Redirected <I
CLASS="FIRSTTERM"
>for</I
> loop</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
if [ -z "$1" ]
then
Filename=names.data # Default, if no filename specified.
else
Filename=$1
fi
line_count=`wc $Filename | awk '{ print $1 }'`
# Number of lines in target file.
#
# Very contrived and kludgy, nevertheless shows that
#+ it's possible to redirect stdin within a "for" loop...
#+ if you're clever enough.
#
# More concise is line_count=$(wc -l &#60; "$Filename")
for name in `seq $line_count` # Recall that "seq" prints sequence of numbers.
# while [ "$name" != Smith ] -- more complicated than a "while" loop --
do
read name # Reads from $Filename, rather than stdin.
echo $name
if [ "$name" = Smith ] # Need all this extra baggage here.
then
break
fi
done &#60;"$Filename" # Redirects stdin to file $Filename.
# ^^^^^^^^^^^^
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>We can modify the previous example to also redirect the output of
the loop.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="REDIR4A"
></A
><P
><B
>Example 20-9. Redirected <I
CLASS="FIRSTTERM"
>for</I
> loop (both
<TT
CLASS="FILENAME"
>stdin</TT
> and <TT
CLASS="FILENAME"
>stdout</TT
>
redirected)</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
if [ -z "$1" ]
then
Filename=names.data # Default, if no filename specified.
else
Filename=$1
fi
Savefile=$Filename.new # Filename to save results in.
FinalName=Jonah # Name to terminate "read" on.
line_count=`wc $Filename | awk '{ print $1 }'` # Number of lines in target file.
for name in `seq $line_count`
do
read name
echo "$name"
if [ "$name" = "$FinalName" ]
then
break
fi
done &#60; "$Filename" &#62; "$Savefile" # Redirects stdin to file $Filename,
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^ and saves it to backup file.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="REDIR5"
></A
><P
><B
>Example 20-10. Redirected <I
CLASS="FIRSTTERM"
>if/then</I
> test</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
if [ -z "$1" ]
then
Filename=names.data # Default, if no filename specified.
else
Filename=$1
fi
TRUE=1
if [ "$TRUE" ] # if true and if : also work.
then
read name
echo $name
fi &#60;"$Filename"
# ^^^^^^^^^^^^
# Reads only first line of file.
# An "if/then" test has no way of iterating unless embedded in a loop.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="NAMESDATA"
></A
><P
><B
>Example 20-11. Data file <I
CLASS="FIRSTTERM"
>names.data</I
> for above
examples</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>Aristotle
Arrhenius
Belisarius
Capablanca
Dickens
Euler
Goethe
Hegel
Jonah
Laplace
Maroczy
Purcell
Schmidt
Schopenhauer
Semmelweiss
Smith
Steinmetz
Tukhashevsky
Turing
Venn
Warshawski
Znosko-Borowski
# This is a data file for
#+ "redir2.sh", "redir3.sh", "redir4.sh", "redir4a.sh", "redir5.sh".</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Redirecting the <TT
CLASS="FILENAME"
>stdout</TT
> of a code
block has the effect of saving its output to a file. See <A
HREF="#RPMCHECK"
>Example 3-2</A
>.</P
><P
><A
HREF="#HEREDOCREF"
>Here documents</A
>
are a special case of redirected code blocks. That being the case,
it should be possible to feed the output of a <I
CLASS="FIRSTTERM"
>here
document</I
> into the <TT
CLASS="FILENAME"
>stdin</TT
> for a
<I
CLASS="FIRSTTERM"
>while loop</I
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># This example by Albert Siersema
# Used with permission (thanks!).
function doesOutput()
# Could be an external command too, of course.
# Here we show you can use a function as well.
{
ls -al *.jpg | awk '{print $5,$9}'
}
nr=0 # We want the while loop to be able to manipulate these and
totalSize=0 #+ to be able to see the changes after the 'while' finished.
while read fileSize fileName ; do
echo "$fileName is $fileSize bytes"
let nr++
totalSize=$((totalSize+fileSize)) # Or: "let totalSize+=fileSize"
done&#60;&#60;EOF
$(doesOutput)
EOF
echo "$nr files totaling $totalSize bytes"</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="REDIRAPPS"
></A
>20.3. Applications</H1
><P
>Clever use of I/O redirection permits parsing and stitching
together snippets of command output (see <A
HREF="#READREDIR"
>Example 15-7</A
>). This permits
generating report and log files.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="LOGEVENTS"
></A
><P
><B
>Example 20-12. Logging events</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# logevents.sh
# Author: Stephane Chazelas.
# Used in ABS Guide with permission.
# Event logging to a file.
# Must be run as root (for write access in /var/log).
ROOT_UID=0 # Only users with $UID 0 have root privileges.
E_NOTROOT=67 # Non-root exit error.
if [ "$UID" -ne "$ROOT_UID" ]
then
echo "Must be root to run this script."
exit $E_NOTROOT
fi
FD_DEBUG1=3
FD_DEBUG2=4
FD_DEBUG3=5
# === Uncomment one of the two lines below to activate script. ===
# LOG_EVENTS=1
# LOG_VARS=1
log() # Writes time and date to log file.
{
echo "$(date) $*" &#62;&#38;7 # This *appends* the date to the file.
# ^^^^^^^ command substitution
# See below.
}
case $LOG_LEVEL in
1) exec 3&#62;&#38;2 4&#62; /dev/null 5&#62; /dev/null;;
2) exec 3&#62;&#38;2 4&#62;&#38;2 5&#62; /dev/null;;
3) exec 3&#62;&#38;2 4&#62;&#38;2 5&#62;&#38;2;;
*) exec 3&#62; /dev/null 4&#62; /dev/null 5&#62; /dev/null;;
esac
FD_LOGVARS=6
if [[ $LOG_VARS ]]
then exec 6&#62;&#62; /var/log/vars.log
else exec 6&#62; /dev/null # Bury output.
fi
FD_LOGEVENTS=7
if [[ $LOG_EVENTS ]]
then
# exec 7 &#62;(exec gawk '{print strftime(), $0}' &#62;&#62; /var/log/event.log)
# Above line fails in versions of Bash more recent than 2.04. Why?
exec 7&#62;&#62; /var/log/event.log # Append to "event.log".
log # Write time and date.
else exec 7&#62; /dev/null # Bury output.
fi
echo "DEBUG3: beginning" &#62;&#38;${FD_DEBUG3}
ls -l &#62;&#38;5 2&#62;&#38;4 # command1 &#62;&#38;5 2&#62;&#38;4
echo "Done" # command2
echo "sending mail" &#62;&#38;${FD_LOGEVENTS}
# Writes "sending mail" to file descriptor #7.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="SUBSHELLS"
></A
>Chapter 21. Subshells</H1
><P
><A
NAME="SUBSHELLSREF"
></A
></P
><P
>Running a shell script launches a new process, a
<I
CLASS="FIRSTTERM"
>subshell</I
>.</P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN18083"
></A
><P
></P
><P
><TT
CLASS="USERINPUT"
><B
>Definition:</B
></TT
>
A <I
CLASS="FIRSTTERM"
>subshell</I
> is a
<A
HREF="#CHILDREF2"
>child process</A
> launched by a
shell (or <I
CLASS="FIRSTTERM"
>shell script</I
>).</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
>A subshell is a separate instance of the command processor
-- the <I
CLASS="FIRSTTERM"
>shell</I
> that gives you the prompt at
the console or in an <I
CLASS="FIRSTTERM"
>xterm</I
> window. Just
as your commands are interpreted at the command-line prompt,
similarly does a script <A
HREF="#BATCHPROCREF"
>batch-process</A
> a list of
commands. Each shell script running is, in effect, a subprocess
(<I
CLASS="FIRSTTERM"
>child process</I
>) of the <A
HREF="#FORKREF"
>parent</A
> shell.</P
><P
>A shell script can itself launch subprocesses. These
<I
CLASS="FIRSTTERM"
>subshells</I
> let the script do
parallel processing, in effect executing multiple subtasks
simultaneously.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# subshell-test.sh
(
# Inside parentheses, and therefore a subshell . . .
while [ 1 ] # Endless loop.
do
echo "Subshell running . . ."
done
)
# Script will run forever,
#+ or at least until terminated by a Ctl-C.
exit $? # End of script (but will never get here).
Now, run the script:
sh subshell-test.sh
And, while the script is running, from a different xterm:
ps -ef | grep subshell-test.sh
UID PID PPID C STIME TTY TIME CMD
500 2698 2502 0 14:26 pts/4 00:00:00 sh subshell-test.sh
500 2699 2698 21 14:26 pts/4 00:00:24 sh subshell-test.sh
^^^^
Analysis:
PID 2698, the script, launched PID 2699, the subshell.
Note: The "UID ..." line would be filtered out by the "grep" command,
but is shown here for illustrative purposes.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>In general, an <A
HREF="#EXTERNALREF"
>external
command</A
> in a script <A
HREF="#FORKREF"
>forks
off</A
> a subprocess,
<A
NAME="AEN18102"
HREF="#FTN.AEN18102"
><SPAN
CLASS="footnote"
>[107]</SPAN
></A
>
whereas a Bash <A
HREF="#BUILTINREF"
>builtin</A
>
does not. For this reason, builtins execute more quickly
and use fewer system resources than their external command
equivalents.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="SUBSHELLPARENS1"
></A
>Command List within
Parentheses</B
></P
><DL
><DT
>( command1; command2; command3; ... )</DT
><DD
><P
>A command list embedded between
<TT
CLASS="REPLACEABLE"
><I
>parentheses</I
></TT
> runs as a
subshell.</P
></DD
></DL
></DIV
><P
><A
NAME="PARVIS"
></A
>Variables in a subshell are
<EM
>not</EM
> visible outside the block of code
in the subshell. They are not accessible to the <A
HREF="#FORKREF"
>parent process</A
>, to the shell
that launched the subshell. These are, in effect,
variables <A
HREF="#LOCALREF"
>local</A
> to the
<I
CLASS="FIRSTTERM"
>child process</I
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SUBSHELL"
></A
><P
><B
>Example 21-1. Variable scope in a subshell</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# subshell.sh
echo
echo "We are outside the subshell."
echo "Subshell level OUTSIDE subshell = $BASH_SUBSHELL"
# Bash, version 3, adds the new $BASH_SUBSHELL variable.
echo; echo
outer_variable=Outer
global_variable=
# Define global variable for "storage" of
#+ value of subshell variable.
(
echo "We are inside the subshell."
echo "Subshell level INSIDE subshell = $BASH_SUBSHELL"
inner_variable=Inner
echo "From inside subshell, \"inner_variable\" = $inner_variable"
echo "From inside subshell, \"outer\" = $outer_variable"
global_variable="$inner_variable" # Will this allow "exporting"
#+ a subshell variable?
)
echo; echo
echo "We are outside the subshell."
echo "Subshell level OUTSIDE subshell = $BASH_SUBSHELL"
echo
if [ -z "$inner_variable" ]
then
echo "inner_variable undefined in main body of shell"
else
echo "inner_variable defined in main body of shell"
fi
echo "From main body of shell, \"inner_variable\" = $inner_variable"
# $inner_variable will show as blank (uninitialized)
#+ because variables defined in a subshell are "local variables".
# Is there a remedy for this?
echo "global_variable = "$global_variable"" # Why doesn't this work?
echo
# =======================================================================
# Additionally ...
echo "-----------------"; echo
var=41 # Global variable.
( let "var+=1"; echo "\$var INSIDE subshell = $var" ) # 42
echo "\$var OUTSIDE subshell = $var" # 41
# Variable operations inside a subshell, even to a GLOBAL variable
#+ do not affect the value of the variable outside the subshell!
exit 0
# Question:
# --------
# Once having exited a subshell,
#+ is there any way to reenter that very same subshell
#+ to modify or access the subshell variables?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#BASHPIDREF"
>$BASHPID</A
> and
<A
HREF="#SUBPIT"
>Example 34-2</A
>.</P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN18127"
></A
><P
></P
><P
><A
NAME="SCOPEREF"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>Definition:</B
></TT
> The
<I
CLASS="FIRSTTERM"
>scope</I
> of a variable is the
context in which it has meaning, in which it has a
<I
CLASS="FIRSTTERM"
>value</I
> that can be referenced. For
example, the scope of a <A
HREF="#LOCALREF1"
>local
variable</A
> lies only within the function,
block of code, or subshell within which it is defined,
while the scope of a <I
CLASS="FIRSTTERM"
>global</I
> variable
is the entire script in which it appears.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
><A
NAME="SUBSHNLEVREF"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>While the <A
HREF="#BASHSUBSHELLREF"
>$BASH_SUBSHELL</A
>
internal variable indicates the nesting level of a
subshell, the <A
HREF="#SHLVLREF"
>$SHLVL</A
>
variable <EM
>shows no change</EM
> within
a subshell.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo " \$BASH_SUBSHELL outside subshell = $BASH_SUBSHELL" # 0
( echo " \$BASH_SUBSHELL inside subshell = $BASH_SUBSHELL" ) # 1
( ( echo " \$BASH_SUBSHELL inside nested subshell = $BASH_SUBSHELL" ) ) # 2
# ^ ^ *** nested *** ^ ^
echo
echo " \$SHLVL outside subshell = $SHLVL" # 3
( echo " \$SHLVL inside subshell = $SHLVL" ) # 3 (No change!)</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><P
>Directory changes made in a subshell do not carry over to the
parent shell.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ALLPROFS"
></A
><P
><B
>Example 21-2. List User Profiles</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# allprofs.sh: Print all user profiles.
# This script written by Heiner Steven, and modified by the document author.
FILE=.bashrc # File containing user profile,
#+ was ".profile" in original script.
for home in `awk -F: '{print $6}' /etc/passwd`
do
[ -d "$home" ] || continue # If no home directory, go to next.
[ -r "$home" ] || continue # If not readable, go to next.
(cd $home; [ -e $FILE ] &#38;&#38; less $FILE)
done
# When script terminates, there is no need to 'cd' back to original directory,
#+ because 'cd $home' takes place in a subshell.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>A subshell may be used to set up a <SPAN
CLASS="QUOTE"
>"dedicated
environment"</SPAN
> for a command group.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>COMMAND1
COMMAND2
COMMAND3
(
IFS=:
PATH=/bin
unset TERMINFO
set -C
shift 5
COMMAND4
COMMAND5
exit 3 # Only exits the subshell!
)
# The parent shell has not been affected, and the environment is preserved.
COMMAND6
COMMAND7</PRE
></FONT
></TD
></TR
></TABLE
>
As seen here, the <A
HREF="#EXITREF"
>exit</A
>
command only terminates the subshell in which it is running,
<EM
>not</EM
> the parent shell or script.</P
><P
>One application of such a <SPAN
CLASS="QUOTE"
>"dedicated environment"</SPAN
>
is testing whether a variable is defined.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if (set -u; : $variable) 2&#62; /dev/null
then
echo "Variable is set."
fi # Variable has been set in current script,
#+ or is an an internal Bash variable,
#+ or is present in environment (has been exported).
# Could also be written [[ ${variable-x} != x || ${variable-y} != y ]]
# or [[ ${variable-x} != x$variable ]]
# or [[ ${variable+x} = x ]]
# or [[ ${variable-x} != x ]]</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Another application is checking for a lock file:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if (set -C; : &#62; lock_file) 2&#62; /dev/null
then
: # lock_file didn't exist: no user running the script
else
echo "Another user is already running that script."
exit 65
fi
# Code snippet by St<53>phane Chazelas,
#+ with modifications by Paulo Marcel Coelho Aragao.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>+</P
><P
>Processes may execute in parallel within different
subshells. This permits breaking a complex task into subcomponents
processed concurrently.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="PARALLEL-PROCESSES"
></A
><P
><B
>Example 21-3. Running parallel processes in subshells</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> (cat list1 list2 list3 | sort | uniq &#62; list123) &#38;
(cat list4 list5 list6 | sort | uniq &#62; list456) &#38;
# Merges and sorts both sets of lists simultaneously.
# Running in background ensures parallel execution.
#
# Same effect as
# cat list1 list2 list3 | sort | uniq &#62; list123 &#38;
# cat list4 list5 list6 | sort | uniq &#62; list456 &#38;
wait # Don't execute the next command until subshells finish.
diff list123 list456</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Redirecting I/O to a subshell uses the <SPAN
CLASS="QUOTE"
>"|"</SPAN
> pipe
operator, as in <TT
CLASS="USERINPUT"
><B
>ls -al | (command)</B
></TT
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A code block between <A
HREF="#CODEBLOCKREF"
>curly brackets</A
> does
<EM
>not</EM
> launch a subshell.</P
><P
>{ command1; command2; command3; . . . commandN; }</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>var1=23
echo "$var1" # 23
{ var1=76; }
echo "$var1" # 76</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="RESTRICTED-SH"
></A
>Chapter 22. Restricted Shells</H1
><P
><A
NAME="RESTRICTEDSHREF"
></A
></P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="DISABLEDCOMMREF"
></A
>Disabled commands in restricted
shells</B
></P
><DL
><DT
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
> . </B
>Running a script or portion of a script in
<I
CLASS="FIRSTTERM"
>restricted mode</I
> disables certain commands
that would otherwise be available. This is a security measure
intended to limit the privileges of the script user and to
minimize possible damage from running the script.</P
></DIV
></DD
></DL
></DIV
><P
>The following commands and actions are disabled:</P
><P
></P
><UL
><LI
><P
>Using <TT
CLASS="REPLACEABLE"
><I
>cd</I
></TT
> to change the working
directory.</P
></LI
><LI
><P
>Changing the values of the
<TT
CLASS="REPLACEABLE"
><I
>$PATH</I
></TT
>,
<TT
CLASS="REPLACEABLE"
><I
>$SHELL</I
></TT
>,
<TT
CLASS="REPLACEABLE"
><I
>$BASH_ENV</I
></TT
>,
or <TT
CLASS="REPLACEABLE"
><I
>$ENV</I
></TT
> <A
HREF="#ENVREF"
>environmental variables</A
>.</P
></LI
><LI
><P
>Reading or changing the <TT
CLASS="REPLACEABLE"
><I
>$SHELLOPTS</I
></TT
>,
shell environmental options.</P
></LI
><LI
><P
>Output redirection.</P
></LI
><LI
><P
>Invoking commands containing one or more
<SPAN
CLASS="TOKEN"
>/</SPAN
>'s.</P
></LI
><LI
><P
>Invoking <A
HREF="#EXECREF"
>exec</A
> to substitute
a different process for the shell.</P
></LI
><LI
><P
>Various other commands that would enable monkeying
with or attempting to subvert the script for an unintended
purpose.</P
></LI
><LI
><P
>Getting out of restricted mode within the script.</P
></LI
></UL
><DIV
CLASS="EXAMPLE"
><A
NAME="RESTRICTED"
></A
><P
><B
>Example 22-1. Running a script in restricted mode</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Starting the script with "#!/bin/bash -r"
#+ runs entire script in restricted mode.
echo
echo "Changing directory."
cd /usr/local
echo "Now in `pwd`"
echo "Coming back home."
cd
echo "Now in `pwd`"
echo
# Everything up to here in normal, unrestricted mode.
set -r
# set --restricted has same effect.
echo "==&#62; Now in restricted mode. &#60;=="
echo
echo
echo "Attempting directory change in restricted mode."
cd ..
echo "Still in `pwd`"
echo
echo
echo "\$SHELL = $SHELL"
echo "Attempting to change shell in restricted mode."
SHELL="/bin/ash"
echo
echo "\$SHELL= $SHELL"
echo
echo
echo "Attempting to redirect output in restricted mode."
ls -l /usr/bin &#62; bin.files
ls -l bin.files # Try to list attempted file creation effort.
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="PROCESS-SUB"
></A
>Chapter 23. Process Substitution</H1
><P
><A
NAME="PROCESSSUBREF"
></A
><A
HREF="#PIPEREF"
>Piping</A
> the <TT
CLASS="FILENAME"
>stdout</TT
>
of a command into the <TT
CLASS="FILENAME"
>stdin</TT
> of another
is a powerful technique. But, what if you need to pipe the
<TT
CLASS="FILENAME"
>stdout</TT
> of <EM
>multiple</EM
>
commands? This is where <TT
CLASS="REPLACEABLE"
><I
>process
substitution</I
></TT
> comes in.</P
><P
><I
CLASS="FIRSTTERM"
>Process substitution</I
> feeds the
output of a <A
HREF="#PROCESSREF"
>process</A
> (or
processes) into the <TT
CLASS="FILENAME"
>stdin</TT
> of another
process.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="COMMANDSPARENS1"
></A
>Template</B
></P
><DL
><DT
>Command list enclosed within parentheses</DT
><DD
><P
><B
CLASS="COMMAND"
>&#62;(command_list)</B
></P
><P
><B
CLASS="COMMAND"
>&#60;(command_list)</B
></P
><P
>Process substitution uses
<TT
CLASS="FILENAME"
>/dev/fd/&#60;n&#62;</TT
> files to send the
results of the process(es) within parentheses to another process.
<A
NAME="AEN18244"
HREF="#FTN.AEN18244"
><SPAN
CLASS="footnote"
>[108]</SPAN
></A
>
</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>There is <EM
>no</EM
> space between the
the <SPAN
CLASS="QUOTE"
>"&#60;"</SPAN
> or <SPAN
CLASS="QUOTE"
>"&#62;"</SPAN
> and the parentheses.
Space there would give an error message.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo &#62;(true)</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/dev/fd/63</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo &#60;(true)</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/dev/fd/63</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo &#62;(true) &#60;(true)</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/dev/fd/63 /dev/fd/62</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>wc &#60;(cat /usr/share/dict/linux.words)</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> 483523 483523 4992010 /dev/fd/63</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep script /usr/share/dict/linux.words | wc</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> 262 262 3601</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>wc &#60;(grep script /usr/share/dict/linux.words)</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> 262 262 3601 /dev/fd/63</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
> Bash creates a pipe with two <A
HREF="#FDREF"
>file
descriptors</A
>, <TT
CLASS="FILENAME"
>--fIn</TT
> and
<TT
CLASS="FILENAME"
>fOut--</TT
>. The <TT
CLASS="FILENAME"
>stdin</TT
>
of <A
HREF="#TRUEREF"
>true</A
> connects
to <TT
CLASS="FILENAME"
>fOut</TT
> (dup2(fOut, 0)),
then Bash passes a <TT
CLASS="FILENAME"
>/dev/fd/fIn</TT
>
argument to <B
CLASS="COMMAND"
>echo</B
>. On systems lacking
<TT
CLASS="FILENAME"
>/dev/fd/&#60;n&#62;</TT
> files, Bash may use
temporary files. (Thanks, S.C.)
</P
></TD
></TR
></TABLE
></DIV
><P
>Process substitution can compare the output of two
different commands, or even the output of different options
to the same command.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>comm &#60;(ls -l) &#60;(ls -al)</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 12
-rw-rw-r-- 1 bozo bozo 78 Mar 10 12:58 File0
-rw-rw-r-- 1 bozo bozo 42 Mar 10 12:58 File2
-rw-rw-r-- 1 bozo bozo 103 Mar 10 12:58 t2.sh
total 20
drwxrwxrwx 2 bozo bozo 4096 Mar 10 18:10 .
drwx------ 72 bozo bozo 4096 Mar 10 17:58 ..
-rw-rw-r-- 1 bozo bozo 78 Mar 10 12:58 File0
-rw-rw-r-- 1 bozo bozo 42 Mar 10 12:58 File2
-rw-rw-r-- 1 bozo bozo 103 Mar 10 12:58 t2.sh</TT
></PRE
></FONT
></TD
></TR
></TABLE
><P
><A
NAME="PCC2DIR"
></A
></P
><P
> Process substitution can compare the contents
of two directories -- to see which filenames are in one,
but not the other.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>diff &#60;(ls $first_directory) &#60;(ls $second_directory)</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Some other usages and uses of process substitution:</P
><P
><A
NAME="PSFDSTDIN"
></A
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>read -a list &#60; &#60;( od -Ad -w24 -t u2 /dev/urandom )
# Read a list of random numbers from /dev/urandom,
#+ process with "od"
#+ and feed into stdin of "read" . . .
# From "insertion-sort.bash" example script.
# Courtesy of JuanJo Ciarlante.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="NETCATEXAMPLE"
></A
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>PORT=6881 # bittorrent
# Scan the port to make sure nothing nefarious is going on.
netcat -l $PORT | tee&#62;(md5sum -&#62;mydata-orig.md5) |
gzip | tee&#62;(md5sum - | sed 's/-$/mydata.lz2/'&#62;mydata-gz.md5)&#62;mydata.gz
# Check the decompression:
gzip -d&#60;mydata.gz | md5sum -c mydata-orig.md5)
# The MD5sum of the original checks stdin and detects compression issues.
# Bill Davidsen contributed this example
#+ (with light edits by the ABS Guide author).</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat &#60;(ls -l)
# Same as ls -l | cat
sort -k 9 &#60;(ls -l /bin) &#60;(ls -l /usr/bin) &#60;(ls -l /usr/X11R6/bin)
# Lists all the files in the 3 main 'bin' directories, and sorts by filename.
# Note that three (count 'em) distinct commands are fed to 'sort'.
diff &#60;(command1) &#60;(command2) # Gives difference in command output.
tar cf &#62;(bzip2 -c &#62; file.tar.bz2) $directory_name
# Calls "tar cf /dev/fd/?? $directory_name", and "bzip2 -c &#62; file.tar.bz2".
#
# Because of the /dev/fd/&#60;n&#62; system feature,
# the pipe between both commands does not need to be named.
#
# This can be emulated.
#
bzip2 -c &#60; pipe &#62; file.tar.bz2&#38;
tar cf pipe $directory_name
rm pipe
# or
exec 3&#62;&#38;1
tar cf /dev/fd/4 $directory_name 4&#62;&#38;1 &#62;&#38;3 3&#62;&#38;- | bzip2 -c &#62; file.tar.bz2 3&#62;&#38;-
exec 3&#62;&#38;-
# Thanks, St<53>phane Chazelas</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="GOODREAD0"
></A
>Here is a method of circumventing the
problem of an <A
HREF="#BADREAD0"
><I
CLASS="FIRSTTERM"
>echo</I
>
piped to a <I
CLASS="FIRSTTERM"
>while-read loop</I
></A
> running
in a subshell.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="WRPS"
></A
><P
><B
>Example 23-1. Code block redirection without forking</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# wr-ps.bash: while-read loop with process substitution.
# This example contributed by Tomas Pospisek.
# (Heavily edited by the ABS Guide author.)
echo
echo "random input" | while read i
do
global=3D": Not available outside the loop."
# ... because it runs in a subshell.
done
echo "\$global (from outside the subprocess) = $global"
# $global (from outside the subprocess) =
echo; echo "--"; echo
while read i
do
echo $i
global=3D": Available outside the loop."
# ... because it does NOT run in a subshell.
done &#60; &#60;( echo "random input" )
# ^ ^
echo "\$global (using process substitution) = $global"
# Random input
# $global (using process substitution) = 3D: Available outside the loop.
echo; echo "##########"; echo
# And likewise . . .
declare -a inloop
index=0
cat $0 | while read line
do
inloop[$index]="$line"
((index++))
# It runs in a subshell, so ...
done
echo "OUTPUT = "
echo ${inloop[*]} # ... nothing echoes.
echo; echo "--"; echo
declare -a outloop
index=0
while read line
do
outloop[$index]="$line"
((index++))
# It does NOT run in a subshell, so ...
done &#60; &#60;( cat $0 )
echo "OUTPUT = "
echo ${outloop[*]} # ... the entire script echoes.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="PSUBPIPING"
></A
>This is a similar example.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="PSUBP"
></A
><P
><B
>Example 23-2. Redirecting the output of <I
CLASS="FIRSTTERM"
>process
substitution</I
> into a loop.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# psub.bash
# As inspired by Diego Molina (thanks!).
declare -a array0
while read
do
array0[${#array0[@]}]="$REPLY"
done &#60; &#60;( sed -e 's/bash/CRASH-BANG!/' $0 | grep bin | awk '{print $1}' )
# Sets the default 'read' variable, $REPLY, by process substitution,
#+ then copies it into an array.
echo "${array0[@]}"
exit $?
# ====================================== #
bash psub.bash
#!/bin/CRASH-BANG! done #!/bin/CRASH-BANG!</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>A reader sent in the following interesting example of process
substitution.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Script fragment taken from SuSE distribution:
# --------------------------------------------------------------#
while read des what mask iface; do
# Some commands ...
done &#60; &#60;(route -n)
# ^ ^ First &#60; is redirection, second is process substitution.
# To test it, let's make it do something.
while read des what mask iface; do
echo $des $what $mask $iface
done &#60; &#60;(route -n)
# Output:
# Kernel IP routing table
# Destination Gateway Genmask Flags Metric Ref Use Iface
# 127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo
# --------------------------------------------------------------#
# As St<53>phane Chazelas points out,
#+ an easier-to-understand equivalent is:
route -n |
while read des what mask iface; do # Variables set from output of pipe.
echo $des $what $mask $iface
done # This yields the same output as above.
# However, as Ulrich Gayer points out . . .
#+ this simplified equivalent uses a subshell for the while loop,
#+ and therefore the variables disappear when the pipe terminates.
# --------------------------------------------------------------#
# However, Filip Moritz comments that there is a subtle difference
#+ between the above two examples, as the following shows.
(
route -n | while read x; do ((y++)); done
echo $y # $y is still unset
while read x; do ((y++)); done &#60; &#60;(route -n)
echo $y # $y has the number of lines of output of route -n
)
More generally spoken
(
: | x=x
# seems to start a subshell like
: | ( x=x )
# while
x=x &#60; &#60;(:)
# does not
)
# This is useful, when parsing csv and the like.
# That is, in effect, what the original SuSE code fragment does.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="FUNCTIONS"
></A
>Chapter 24. Functions</H1
><P
><A
NAME="FUNCTIONREF"
></A
></P
><P
>Like <SPAN
CLASS="QUOTE"
>"real"</SPAN
> programming languages,
Bash has functions, though in a somewhat limited implementation.
A function is a subroutine, a <A
HREF="#CODEBLOCKREF"
>code
block</A
> that implements a set of operations, a <SPAN
CLASS="QUOTE"
>"black
box"</SPAN
> that performs a specified task. Wherever there is
repetitive code, when a task repeats with only slight variations in
procedure, then consider using a function.</P
><P
><P
><B
CLASS="COMMAND"
>function</B
> <TT
CLASS="REPLACEABLE"
><I
>function_name</I
></TT
> { <BR> <TT
CLASS="REPLACEABLE"
><I
>command</I
></TT
>... <BR> } <BR></P
>
or
<P
> <TT
CLASS="REPLACEABLE"
><I
>function_name</I
></TT
> () { <BR> <TT
CLASS="REPLACEABLE"
><I
>command</I
></TT
>... <BR> } <BR></P
>
</P
><P
>This second form will cheer the hearts of C programmers
(and is more <A
HREF="#PORTABILITYISSUES"
>portable</A
>).</P
><P
>As in C, the function's opening bracket may optionally appear
on the second line.</P
><P
><P
> <TT
CLASS="REPLACEABLE"
><I
>function_name</I
></TT
> () <BR> { <BR> <TT
CLASS="REPLACEABLE"
><I
>command</I
></TT
>... <BR> } <BR></P
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A function may be <SPAN
CLASS="QUOTE"
>"compacted"</SPAN
> into a single
line.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>fun () { echo "This is a function"; echo; }
# ^ ^</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>In this case, however, a <I
CLASS="FIRSTTERM"
>semicolon</I
>
must follow the final command in the function.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>fun () { echo "This is a function"; echo } # Error!
# ^
fun2 () { echo "Even a single-command function? Yes!"; }
# ^</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><P
>Functions are called, <I
CLASS="FIRSTTERM"
>triggered</I
>, simply by
invoking their names. <EM
>A function call is equivalent to
a command.</EM
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX59"
></A
><P
><B
>Example 24-1. Simple functions</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ex59.sh: Exercising functions (simple).
JUST_A_SECOND=1
funky ()
{ # This is about as simple as functions get.
echo "This is a funky function."
echo "Now exiting funky function."
} # Function declaration must precede call.
fun ()
{ # A somewhat more complex function.
i=0
REPEATS=30
echo
echo "And now the fun really begins."
echo
sleep $JUST_A_SECOND # Hey, wait a second!
while [ $i -lt $REPEATS ]
do
echo "----------FUNCTIONS----------&#62;"
echo "&#60;------------ARE-------------"
echo "&#60;------------FUN------------&#62;"
echo
let "i+=1"
done
}
# Now, call the functions.
funky
fun
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="FUNCTDEFMUST"
></A
></P
><P
>The function definition must precede the first call to
it. There is no method of <SPAN
CLASS="QUOTE"
>"declaring"</SPAN
> the function,
as, for example, in C.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>f1
# Will give an error message, since function "f1" not yet defined.
declare -f f1 # This doesn't help either.
f1 # Still an error message.
# However...
f1 ()
{
echo "Calling function \"f2\" from within function \"f1\"."
f2
}
f2 ()
{
echo "Function \"f2\"."
}
f1 # Function "f2" is not actually called until this point,
#+ although it is referenced before its definition.
# This is permissible.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="EMPTYFUNC"
></A
>Functions may not be empty!
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# empty-function.sh
empty ()
{
}
exit 0 # Will not exit here!
# $ sh empty-function.sh
# empty-function.sh: line 6: syntax error near unexpected token `}'
# empty-function.sh: line 6: `}'
# $ echo $?
# 2
# Note that a function containing only comments is empty.
func ()
{
# Comment 1.
# Comment 2.
# This is still an empty function.
# Thank you, Mark Bova, for pointing this out.
}
# Results in same error message as above.
# However ...
not_quite_empty ()
{
illegal_command
} # A script containing this function will *not* bomb
#+ as long as the function is not called.
not_empty ()
{
:
} # Contains a : (null command), and this is okay.
# Thank you, Dominick Geyer and Thiemo Kellner.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><P
>It is even possible to nest a function within another function,
although this is not very useful.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>f1 ()
{
f2 () # nested
{
echo "Function \"f2\", inside \"f1\"."
}
}
f2 # Gives an error message.
# Even a preceding "declare -f f2" wouldn't help.
echo
f1 # Does nothing, since calling "f1" does not automatically call "f2".
f2 # Now, it's all right to call "f2",
#+ since its definition has been made visible by calling "f1".
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Function declarations can appear in unlikely places, even where a
command would otherwise go.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>ls -l | foo() { echo "foo"; } # Permissible, but useless.
if [ "$USER" = bozo ]
then
bozo_greet () # Function definition embedded in an if/then construct.
{
echo "Hello, Bozo."
}
fi
bozo_greet # Works only for Bozo, and other users get an error.
# Something like this might be useful in some contexts.
NO_EXIT=1 # Will enable function definition below.
[[ $NO_EXIT -eq 1 ]] &#38;&#38; exit() { true; } # Function definition in an "and-list".
# If $NO_EXIT is 1, declares "exit ()".
# This disables the "exit" builtin by aliasing it to "true".
exit # Invokes "exit ()" function, not "exit" builtin.
# Or, similarly:
filename=file1
[ -f "$filename" ] &#38;&#38;
foo () { rm -f "$filename"; echo "File "$filename" deleted."; } ||
foo () { echo "File "$filename" not found."; touch bar; }
foo
# Thanks, S.C. and Christopher Head</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="FSTRANGEREF"
></A
>Function names can take strange
forms.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> _(){ for i in {1..10}; do echo -n "$FUNCNAME"; done; echo; }
# ^^^ No space between function name and parentheses.
# This doesn't always work. Why not?
# Now, let's invoke the function.
_ # __________
# ^^^^^^^^^^ 10 underscores (10 x function name)!
# A "naked" underscore is an acceptable function name.
# In fact, a colon is likewise an acceptable function name.
:(){ echo ":"; }; :
# Of what use is this?
# It's a devious way to obfuscate the code in a script.</PRE
></FONT
></TD
></TR
></TABLE
>
See also <A
HREF="#GRONSFELD"
>Example A-56</A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>What happens when different versions of the same function
appear in a script?
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># As Yan Chen points out,
# when a function is defined multiple times,
# the final version is what is invoked.
# This is not, however, particularly useful.
func ()
{
echo "First version of func ()."
}
func ()
{
echo "Second version of func ()."
}
func # Second version of func ().
exit $?
# It is even possible to use functions to override
#+ or preempt system commands.
# Of course, this is *not* advisable.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="COMPLEXFUNCT"
></A
>24.1. Complex Functions and Function Complexities</H1
><P
>Functions may process arguments passed to them and return
an <A
HREF="#EXITSTATUSREF"
>exit status</A
> to the script
for further processing.</P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>function_name $arg1 $arg2</PRE
></FONT
></TD
></TR
></TABLE
><P
><A
NAME="PASSEDARGS"
></A
></P
><P
>The function refers to the passed arguments by position (as if they were
<A
HREF="#POSPARAMREF"
>positional parameters</A
>),
that is, <TT
CLASS="VARNAME"
>$1</TT
>, <TT
CLASS="VARNAME"
>$2</TT
>, and
so forth.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX60"
></A
><P
><B
>Example 24-2. Function Taking Parameters</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Functions and parameters
DEFAULT=default # Default param value.
func2 () {
if [ -z "$1" ] # Is parameter #1 zero length?
then
echo "-Parameter #1 is zero length.-" # Or no parameter passed.
else
echo "-Parameter #1 is \"$1\".-"
fi
variable=${1-$DEFAULT} # What does
echo "variable = $variable" #+ parameter substitution show?
# ---------------------------
# It distinguishes between
#+ no param and a null param.
if [ "$2" ]
then
echo "-Parameter #2 is \"$2\".-"
fi
return 0
}
echo
echo "Nothing passed."
func2 # Called with no params
echo
echo "Zero-length parameter passed."
func2 "" # Called with zero-length param
echo
echo "Null parameter passed."
func2 "$uninitialized_param" # Called with uninitialized param
echo
echo "One parameter passed."
func2 first # Called with one param
echo
echo "Two parameters passed."
func2 first second # Called with two params
echo
echo "\"\" \"second\" passed."
func2 "" second # Called with zero-length first parameter
echo # and ASCII string as a second one.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="FSHIFTREF"
></A
></P
><DIV
CLASS="IMPORTANT"
><P
></P
><TABLE
CLASS="IMPORTANT"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/important.gif"
HSPACE="5"
ALT="Important"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <A
HREF="#SHIFTREF"
>shift</A
>
command works on arguments passed to functions (see <A
HREF="#MULTIPLICATION"
>Example 36-18</A
>).</P
></TD
></TR
></TABLE
></DIV
><P
>But, what about command-line arguments passed to the script?
Does a function see them? Well, let's clear up the confusion.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="FUNCCMDLINEARG"
></A
><P
><B
>Example 24-3. Functions and command-line args passed to the script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# func-cmdlinearg.sh
# Call this script with a command-line argument,
#+ something like $0 arg1.
func ()
{
echo "$1" # Echoes first arg passed to the function.
} # Does a command-line arg qualify?
echo "First call to function: no arg passed."
echo "See if command-line arg is seen."
func
# No! Command-line arg not seen.
echo "============================================================"
echo
echo "Second call to function: command-line arg passed explicitly."
func $1
# Now it's seen!
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>In contrast to certain other programming languages,
shell scripts normally pass only value parameters to
functions. Variable names (which are actually
<I
CLASS="FIRSTTERM"
>pointers</I
>), if
passed as parameters to functions, will be treated as string
literals. <EM
>Functions interpret their arguments
literally.</EM
></P
><P
><A
NAME="FUNCPOINTERS"
></A
></P
><P
><A
HREF="#IVRREF"
>Indirect variable
references</A
> (see <A
HREF="#EX78"
>Example 37-2</A
>) provide a clumsy
sort of mechanism for passing variable pointers to
functions.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="INDFUNC"
></A
><P
><B
>Example 24-4. Passing an indirect reference to a function</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ind-func.sh: Passing an indirect reference to a function.
echo_var ()
{
echo "$1"
}
message=Hello
Hello=Goodbye
echo_var "$message" # Hello
# Now, let's pass an indirect reference to the function.
echo_var "${!message}" # Goodbye
echo "-------------"
# What happens if we change the contents of "hello" variable?
Hello="Hello, again!"
echo_var "$message" # Hello
echo_var "${!message}" # Hello, again!
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The next logical question is whether parameters can be
dereferenced <EM
>after</EM
> being passed to a
function.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="DEREFERENCECL"
></A
><P
><B
>Example 24-5. Dereferencing a parameter passed to a function</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# dereference.sh
# Dereferencing parameter passed to a function.
# Script by Bruce W. Clare.
dereference ()
{
y=\$"$1" # Name of variable (not value!).
echo $y # $Junk
x=`eval "expr \"$y\" "`
echo $1=$x
eval "$1=\"Some Different Text \"" # Assign new value.
}
Junk="Some Text"
echo $Junk "before" # Some Text before
dereference Junk
echo $Junk "after" # Some Different Text after
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="REFPARAMS"
></A
><P
><B
>Example 24-6. Again, dereferencing a parameter passed to a function</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ref-params.sh: Dereferencing a parameter passed to a function.
# (Complex Example)
ITERATIONS=3 # How many times to get input.
icount=1
my_read () {
# Called with my_read varname,
#+ outputs the previous value between brackets as the default value,
#+ then asks for a new value.
local local_var
echo -n "Enter a value "
eval 'echo -n "[$'$1'] "' # Previous value.
# eval echo -n "[\$$1] " # Easier to understand,
#+ but loses trailing space in user prompt.
read local_var
[ -n "$local_var" ] &#38;&#38; eval $1=\$local_var
# "And-list": if "local_var" then set "$1" to its value.
}
echo
while [ "$icount" -le "$ITERATIONS" ]
do
my_read var
echo "Entry #$icount = $var"
let "icount += 1"
echo
done
# Thanks to Stephane Chazelas for providing this instructive example.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="EXITRETURN1"
></A
>Exit and Return</B
></P
><DL
><DT
><B
CLASS="COMMAND"
>exit status</B
></DT
><DD
><P
>Functions return a value, called an <I
CLASS="FIRSTTERM"
>exit
status</I
>. This is analogous to the <A
HREF="#EXITSTATUSREF"
>exit status</A
> returned by a
command. The exit status may be explicitly specified
by a <B
CLASS="COMMAND"
>return</B
> statement, otherwise it
is the exit status of the last command in the function
(<SPAN
CLASS="RETURNVALUE"
>0</SPAN
> if successful, and a non-zero
error code if not). This <A
HREF="#EXITSTATUSREF"
>exit
status</A
> may be used in the script by referencing it
as <A
HREF="#XSTATVARREF"
>$?</A
>. This mechanism
effectively permits script functions to have a <SPAN
CLASS="QUOTE"
>"return
value"</SPAN
> similar to C functions.</P
></DD
><DT
><B
CLASS="COMMAND"
>return</B
></DT
><DD
><P
><A
NAME="RETURNREF"
></A
></P
><P
>Terminates a function. A <B
CLASS="COMMAND"
>return</B
> command
<A
NAME="AEN18474"
HREF="#FTN.AEN18474"
><SPAN
CLASS="footnote"
>[109]</SPAN
></A
>
optionally takes an <I
CLASS="FIRSTTERM"
>integer</I
>
argument, which is returned to the calling script as
the <SPAN
CLASS="QUOTE"
>"exit status"</SPAN
> of the function, and
this exit status is assigned to the variable <A
HREF="#XSTATVARREF"
>$?</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="MAX"
></A
><P
><B
>Example 24-7. Maximum of two numbers</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# max.sh: Maximum of two integers.
E_PARAM_ERR=250 # If less than 2 params passed to function.
EQUAL=251 # Return value if both params equal.
# Error values out of range of any
#+ params that might be fed to the function.
max2 () # Returns larger of two numbers.
{ # Note: numbers compared must be less than 250.
if [ -z "$2" ]
then
return $E_PARAM_ERR
fi
if [ "$1" -eq "$2" ]
then
return $EQUAL
else
if [ "$1" -gt "$2" ]
then
return $1
else
return $2
fi
fi
}
max2 33 34
return_val=$?
if [ "$return_val" -eq $E_PARAM_ERR ]
then
echo "Need to pass two parameters to the function."
elif [ "$return_val" -eq $EQUAL ]
then
echo "The two numbers are equal."
else
echo "The larger of the two numbers is $return_val."
fi
exit 0
# Exercise (easy):
# ---------------
# Convert this to an interactive script,
#+ that is, have the script ask for input (two numbers).</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>For a function to return a string or array, use a
dedicated variable.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>count_lines_in_etc_passwd()
{
[[ -r /etc/passwd ]] &#38;&#38; REPLY=$(echo $(wc -l &#60; /etc/passwd))
# If /etc/passwd is readable, set REPLY to line count.
# Returns both a parameter value and status information.
# The 'echo' seems unnecessary, but . . .
#+ it removes excess whitespace from the output.
}
if count_lines_in_etc_passwd
then
echo "There are $REPLY lines in /etc/passwd."
else
echo "Cannot count lines in /etc/passwd."
fi
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX61"
></A
><P
><B
>Example 24-8. Converting numbers to Roman numerals</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Arabic number to Roman numeral conversion
# Range: 0 - 200
# It's crude, but it works.
# Extending the range and otherwise improving the script is left as an exercise.
# Usage: roman number-to-convert
LIMIT=200
E_ARG_ERR=65
E_OUT_OF_RANGE=66
if [ -z "$1" ]
then
echo "Usage: `basename $0` number-to-convert"
exit $E_ARG_ERR
fi
num=$1
if [ "$num" -gt $LIMIT ]
then
echo "Out of range!"
exit $E_OUT_OF_RANGE
fi
to_roman () # Must declare function before first call to it.
{
number=$1
factor=$2
rchar=$3
let "remainder = number - factor"
while [ "$remainder" -ge 0 ]
do
echo -n $rchar
let "number -= factor"
let "remainder = number - factor"
done
return $number
# Exercises:
# ---------
# 1) Explain how this function works.
# Hint: division by successive subtraction.
# 2) Extend to range of the function.
# Hint: use "echo" and command-substitution capture.
}
to_roman $num 100 C
num=$?
to_roman $num 90 LXXXX
num=$?
to_roman $num 50 L
num=$?
to_roman $num 40 XL
num=$?
to_roman $num 10 X
num=$?
to_roman $num 9 IX
num=$?
to_roman $num 5 V
num=$?
to_roman $num 4 IV
num=$?
to_roman $num 1 I
# Successive calls to conversion function!
# Is this really necessary??? Can it be simplified?
echo
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#ISALPHA"
>Example 11-29</A
>.</P
><DIV
CLASS="IMPORTANT"
><P
></P
><TABLE
CLASS="IMPORTANT"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/important.gif"
HSPACE="5"
ALT="Important"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The largest positive integer a function can return is
255. The <B
CLASS="COMMAND"
>return</B
> command is closely tied
to the concept of <A
HREF="#EXITSTATUSREF"
>exit
status</A
>, which accounts for this particular
limitation. Fortunately, there are various <A
HREF="#RVT"
>workarounds</A
> for those situations
requiring a large integer return value from a
function.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="RETURNTEST"
></A
><P
><B
>Example 24-9. Testing large return values in a function</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# return-test.sh
# The largest positive value a function can return is 255.
return_test () # Returns whatever passed to it.
{
return $1
}
return_test 27 # o.k.
echo $? # Returns 27.
return_test 255 # Still o.k.
echo $? # Returns 255.
return_test 257 # Error!
echo $? # Returns 1 (return code for miscellaneous error).
# =========================================================
return_test -151896 # Do large negative numbers work?
echo $? # Will this return -151896?
# No! It returns 168.
# Version of Bash before 2.05b permitted
#+ large negative integer return values.
# It happened to be a useful feature.
# Newer versions of Bash unfortunately plug this loophole.
# This may break older scripts.
# Caution!
# =========================================================
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>A workaround for obtaining large integer <SPAN
CLASS="QUOTE"
>"return
values"</SPAN
> is to simply assign the <SPAN
CLASS="QUOTE"
>"return
value"</SPAN
> to a global variable.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>Return_Val= # Global variable to hold oversize return value of function.
alt_return_test ()
{
fvar=$1
Return_Val=$fvar
return # Returns 0 (success).
}
alt_return_test 1
echo $? # 0
echo "return value = $Return_Val" # 1
alt_return_test 256
echo "return value = $Return_Val" # 256
alt_return_test 257
echo "return value = $Return_Val" # 257
alt_return_test 25701
echo "return value = $Return_Val" #25701</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="CAPTURERETVAL"
></A
></P
><P
>A more elegant method is to have the function
<B
CLASS="COMMAND"
>echo</B
> its <SPAN
CLASS="QUOTE"
>"return
value to <TT
CLASS="FILENAME"
>stdout</TT
>,"</SPAN
> and
then capture it by <A
HREF="#COMMANDSUBREF"
>command
substitution</A
>. See the <A
HREF="#RVT"
>discussion
of this</A
> in <A
HREF="#ASSORTEDTIPS"
>Section 36.7</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="MAX2"
></A
><P
><B
>Example 24-10. Comparing two large integers</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# max2.sh: Maximum of two LARGE integers.
# This is the previous "max.sh" example,
#+ modified to permit comparing large integers.
EQUAL=0 # Return value if both params equal.
E_PARAM_ERR=-99999 # Not enough params passed to function.
# ^^^^^^ Out of range of any params that might be passed.
max2 () # "Returns" larger of two numbers.
{
if [ -z "$2" ]
then
echo $E_PARAM_ERR
return
fi
if [ "$1" -eq "$2" ]
then
echo $EQUAL
return
else
if [ "$1" -gt "$2" ]
then
retval=$1
else
retval=$2
fi
fi
echo $retval # Echoes (to stdout), rather than returning value.
# Why?
}
return_val=$(max2 33001 33997)
# ^^^^ Function name
# ^^^^^ ^^^^^ Params passed
# This is actually a form of command substitution:
#+ treating a function as if it were a command,
#+ and assigning the stdout of the function to the variable "return_val."
# ========================= OUTPUT ========================
if [ "$return_val" -eq "$E_PARAM_ERR" ]
then
echo "Error in parameters passed to comparison function!"
elif [ "$return_val" -eq "$EQUAL" ]
then
echo "The two numbers are equal."
else
echo "The larger of the two numbers is $return_val."
fi
# =========================================================
exit 0
# Exercises:
# ---------
# 1) Find a more elegant way of testing
#+ the parameters passed to the function.
# 2) Simplify the if/then structure at "OUTPUT."
# 3) Rewrite the script to take input from command-line parameters.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Here is another example of capturing a function
<SPAN
CLASS="QUOTE"
>"return value."</SPAN
> Understanding it requires some
knowledge of <A
HREF="#AWKREF"
>awk</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>month_length () # Takes month number as an argument.
{ # Returns number of days in month.
monthD="31 28 31 30 31 30 31 31 30 31 30 31" # Declare as local?
echo "$monthD" | awk '{ print $'"${1}"' }' # Tricky.
# ^^^^^^^^^
# Parameter passed to function ($1 -- month number), then to awk.
# Awk sees this as "print $1 . . . print $12" (depending on month number)
# Template for passing a parameter to embedded awk script:
# $'"${script_parameter}"'
# Here's a slightly simpler awk construct:
# echo $monthD | awk -v month=$1 '{print $(month)}'
# Uses the -v awk option, which assigns a variable value
#+ prior to execution of the awk program block.
# Thank you, Rich.
# Needs error checking for correct parameter range (1-12)
#+ and for February in leap year.
}
# ----------------------------------------------
# Usage example:
month=4 # April, for example (4th month).
days_in=$(month_length $month)
echo $days_in # 30
# ----------------------------------------------</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>See also <A
HREF="#DAYSBETWEEN"
>Example A-7</A
>
and <A
HREF="#STDDEV"
>Example A-37</A
>.</P
><P
><TT
CLASS="USERINPUT"
><B
>Exercise:</B
></TT
> Using what we have
just learned, extend the previous <A
HREF="#EX61"
>Roman numerals example</A
> to accept
arbitrarily large input.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="REDSTDINFUNC1"
></A
>Redirection</B
></P
><DL
><DT
><TT
CLASS="REPLACEABLE"
><I
>Redirecting the stdin
of a function</I
></TT
></DT
><DD
><P
>A function is essentially a <A
HREF="#CODEBLOCKREF"
>code block</A
>, which means its
<TT
CLASS="FILENAME"
>stdin</TT
> can be redirected (as in <A
HREF="#EX8"
>Example 3-1</A
>).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="REALNAME"
></A
><P
><B
>Example 24-11. Real name from username</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# realname.sh
#
# From username, gets "real name" from /etc/passwd.
ARGCOUNT=1 # Expect one arg.
E_WRONGARGS=85
file=/etc/passwd
pattern=$1
if [ $# -ne "$ARGCOUNT" ]
then
echo "Usage: `basename $0` USERNAME"
exit $E_WRONGARGS
fi
file_excerpt () # Scan file for pattern,
{ #+ then print relevant portion of line.
while read line # "while" does not necessarily need [ condition ]
do
echo "$line" | grep $1 | awk -F":" '{ print $5 }'
# Have awk use ":" delimiter.
done
} &#60;$file # Redirect into function's stdin.
file_excerpt $pattern
# Yes, this entire script could be reduced to
# grep PATTERN /etc/passwd | awk -F":" '{ print $5 }'
# or
# awk -F: '/PATTERN/ {print $5}'
# or
# awk -F: '($1 == "username") { print $5 }' # real name from username
# However, it might not be as instructive.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>There is an alternate, and perhaps less confusing
method of redirecting a function's
<TT
CLASS="FILENAME"
>stdin</TT
>. This involves redirecting the
<TT
CLASS="FILENAME"
>stdin</TT
> to an embedded bracketed code
block within the function.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Instead of:
Function ()
{
...
} &#60; file
# Try this:
Function ()
{
{
...
} &#60; file
}
# Similarly,
Function () # This works.
{
{
echo $*
} | tr a b
}
Function () # This doesn't work.
{
echo $*
} | tr a b # A nested code block is mandatory here.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Emmanuel Rouat's <A
HREF="#SAMPLE-BASHRC"
>sample <TT
CLASS="FILENAME"
>bashrc</TT
>
file</A
> contains some instructive examples of
functions.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="LOCALVAR"
></A
>24.2. Local Variables</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="LOCALREF1"
></A
>What makes a variable
<I
CLASS="FIRSTTERM"
>local</I
>?</B
></P
><DL
><DT
>local variables</DT
><DD
><P
>A variable declared as <I
CLASS="FIRSTTERM"
>local</I
>
is one that is visible only within the <A
HREF="#CODEBLOCKREF"
>block of code</A
> in which it
appears. It has local <A
HREF="#SCOPEREF"
>scope</A
>.
In a function, a <I
CLASS="FIRSTTERM"
>local variable</I
> has
meaning only within that function block.
<A
NAME="AEN18568"
HREF="#FTN.AEN18568"
><SPAN
CLASS="footnote"
>[110]</SPAN
></A
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX62"
></A
><P
><B
>Example 24-12. Local variable visibility</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ex62.sh: Global and local variables inside a function.
func ()
{
local loc_var=23 # Declared as local variable.
echo # Uses the 'local' builtin.
echo "\"loc_var\" in function = $loc_var"
global_var=999 # Not declared as local.
# Therefore, defaults to global.
echo "\"global_var\" in function = $global_var"
}
func
# Now, to see if local variable "loc_var" exists outside the function.
echo
echo "\"loc_var\" outside function = $loc_var"
# $loc_var outside function =
# No, $loc_var not visible globally.
echo "\"global_var\" outside function = $global_var"
# $global_var outside function = 999
# $global_var is visible globally.
echo
exit 0
# In contrast to C, a Bash variable declared inside a function
#+ is local ONLY if declared as such.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Before a function is called, <EM
>all</EM
>
variables declared within the function are invisible outside
the body of the function, not just those explicitly declared
as <I
CLASS="FIRSTTERM"
>local</I
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
func ()
{
global_var=37 # Visible only within the function block
#+ before the function has been called.
} # END OF FUNCTION
echo "global_var = $global_var" # global_var =
# Function "func" has not yet been called,
#+ so $global_var is not visible here.
func
echo "global_var = $global_var" # global_var = 37
# Has been set by function call.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="EXITVALANOMALY01"
></A
></P
><P
>As Evgeniy Ivanov points out, when declaring and
setting a local variable in a single command, apparently the
order of operations is to <EM
>first set the variable,
and only afterwards restrict it to local scope</EM
>.
This is reflected in the <A
HREF="#EXITSTATUSREF"
>return value</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo "==OUTSIDE Function (global)=="
t=$(exit 1)
echo $? # 1
# As expected.
echo
function0 ()
{
echo "==INSIDE Function=="
echo "Global"
t0=$(exit 1)
echo $? # 1
# As expected.
echo
echo "Local declared &#38; assigned in same command."
local t1=$(exit 1)
echo $? # 0
# Unexpected!
# Apparently, the variable assignment takes place before
#+ the local declaration.
#+ The return value is for the latter.
echo
echo "Local declared, then assigned (separate commands)."
local t2
t2=$(exit 1)
echo $? # 1
# As expected.
}
function0</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="LOCVARRECUR"
></A
>24.2.1. Local variables and recursion.</H2
><P
><A
NAME="RECURSIONREF0"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN18598"
></A
><P
></P
><P
><A
NAME="RECURSIONREF"
></A
></P
><P
><I
CLASS="FIRSTTERM"
>Recursion</I
> is an interesting
and sometimes useful form of
<I
CLASS="FIRSTTERM"
>self-reference</I
>. <A
HREF="#MAYERREF"
>Herbert Mayer</A
> defines it
as <SPAN
CLASS="QUOTE"
>". . . expressing an algorithm by using a
simpler version of that same algorithm . . ."</SPAN
></P
><P
>Consider a definition defined in terms of itself,
<A
NAME="AEN18607"
HREF="#FTN.AEN18607"
><SPAN
CLASS="footnote"
>[111]</SPAN
></A
>
an expression implicit in its own expression,
<A
NAME="AEN18610"
HREF="#FTN.AEN18610"
><SPAN
CLASS="footnote"
>[112]</SPAN
></A
>
<EM
>a snake swallowing its own
tail</EM
>,
<A
NAME="AEN18614"
HREF="#FTN.AEN18614"
><SPAN
CLASS="footnote"
>[113]</SPAN
></A
>
or . . . a function that calls itself.
<A
NAME="AEN18617"
HREF="#FTN.AEN18617"
><SPAN
CLASS="footnote"
>[114]</SPAN
></A
>
</P
><P
><A
NAME="RECURSIONDEMO0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="RECURSIONDEMO"
></A
><P
><B
>Example 24-13. Demonstration of a simple recursive function</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# recursion-demo.sh
# Demonstration of recursion.
RECURSIONS=9 # How many times to recurse.
r_count=0 # Must be global. Why?
recurse ()
{
var="$1"
while [ "$var" -ge 0 ]
do
echo "Recursion count = "$r_count" +-+ \$var = "$var""
(( var-- )); (( r_count++ ))
recurse "$var" # Function calls itself (recurses)
done #+ until what condition is met?
}
recurse $RECURSIONS
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="RECURSIONDEMO02"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="RECURSIONDEMO2"
></A
><P
><B
>Example 24-14. Another simple demonstration</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# recursion-def.sh
# A script that defines "recursion" in a rather graphic way.
RECURSIONS=10
r_count=0
sp=" "
define_recursion ()
{
((r_count++))
sp="$sp"" "
echo -n "$sp"
echo "\"The act of recurring ... \"" # Per 1913 Webster's dictionary.
while [ $r_count -le $RECURSIONS ]
do
define_recursion
done
}
echo
echo "Recursion: "
define_recursion
echo
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
></P
></DIV
></TD
></TR
></TABLE
><P
>Local variables are a useful tool for writing recursive
code, but this practice generally involves a great deal of
computational overhead and is definitely
<EM
>not</EM
> recommended in a shell script.
<A
NAME="AEN18632"
HREF="#FTN.AEN18632"
><SPAN
CLASS="footnote"
>[115]</SPAN
></A
>
</P
><P
><A
NAME="FACTORIALREF"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX63"
></A
><P
><B
>Example 24-15. Recursion, using a local variable</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# factorial
# ---------
# Does bash permit recursion?
# Well, yes, but...
# It's so slow that you gotta have rocks in your head to try it.
MAX_ARG=5
E_WRONG_ARGS=85
E_RANGE_ERR=86
if [ -z "$1" ]
then
echo "Usage: `basename $0` number"
exit $E_WRONG_ARGS
fi
if [ "$1" -gt $MAX_ARG ]
then
echo "Out of range ($MAX_ARG is maximum)."
# Let's get real now.
# If you want greater range than this,
#+ rewrite it in a Real Programming Language.
exit $E_RANGE_ERR
fi
fact ()
{
local number=$1
# Variable "number" must be declared as local,
#+ otherwise this doesn't work.
if [ "$number" -eq 0 ]
then
factorial=1 # Factorial of 0 = 1.
else
let "decrnum = number - 1"
fact $decrnum # Recursive function call (the function calls itself).
let "factorial = $number * $?"
fi
return $factorial
}
fact $1
echo "Factorial of $1 is $?."
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Also see <A
HREF="#PRIMES"
>Example A-15</A
> for an example of
recursion in a script. Be aware that recursion is
resource-intensive and executes slowly, and is therefore
generally not appropriate in a script.</P
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="RECURNOLOCVAR"
></A
>24.3. Recursion Without Local Variables</H1
><P
>A function may recursively call itself even without use of
local variables.</P
><P
><A
NAME="FIBOREF"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="FIBO"
></A
><P
><B
>Example 24-16. <I
CLASS="FIRSTTERM"
>The Fibonacci Sequence</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# fibo.sh : Fibonacci sequence (recursive)
# Author: M. Cooper
# License: GPL3
# ----------algorithm--------------
# Fibo(0) = 0
# Fibo(1) = 1
# else
# Fibo(j) = Fibo(j-1) + Fibo(j-2)
# ---------------------------------
MAXTERM=15 # Number of terms (+1) to generate.
MINIDX=2 # If idx is less than 2, then Fibo(idx) = idx.
Fibonacci ()
{
idx=$1 # Doesn't need to be local. Why not?
if [ "$idx" -lt "$MINIDX" ]
then
echo "$idx" # First two terms are 0 1 ... see above.
else
(( --idx )) # j-1
term1=$( Fibonacci $idx ) # Fibo(j-1)
(( --idx )) # j-2
term2=$( Fibonacci $idx ) # Fibo(j-2)
echo $(( term1 + term2 ))
fi
# An ugly, ugly kludge.
# The more elegant implementation of recursive fibo in C
#+ is a straightforward translation of the algorithm in lines 7 - 10.
}
for i in $(seq 0 $MAXTERM)
do # Calculate $MAXTERM+1 terms.
FIBO=$(Fibonacci $i)
echo -n "$FIBO "
done
# 0 1 1 2 3 5 8 13 21 34 55 89 144 233 377 610
# Takes a while, doesn't it? Recursion in a script is slow.
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="HANOIREF"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="HANOI"
></A
><P
><B
>Example 24-17. <I
CLASS="FIRSTTERM"
>The Towers of Hanoi</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
#
# The Towers Of Hanoi
# Bash script
# Copyright (C) 2000 Amit Singh. All Rights Reserved.
# http://hanoi.kernelthread.com
#
# Tested under Bash version 2.05b.0(13)-release.
# Also works under Bash version 3.x.
#
# Used in "Advanced Bash Scripting Guide"
#+ with permission of script author.
# Slightly modified and commented by ABS author.
#=================================================================#
# The Tower of Hanoi is a mathematical puzzle attributed to
#+ Edouard Lucas, a nineteenth-century French mathematician.
#
# There are three vertical posts set in a base.
# The first post has a set of annular rings stacked on it.
# These rings are disks with a hole drilled out of the center,
#+ so they can slip over the posts and rest flat.
# The rings have different diameters, and they stack in ascending
#+ order, according to size.
# The smallest ring is on top, and the largest on the bottom.
#
# The task is to transfer the stack of rings
#+ to one of the other posts.
# You can move only one ring at a time to another post.
# You are permitted to move rings back to the original post.
# You may place a smaller ring atop a larger one,
#+ but *not* vice versa.
# Again, it is forbidden to place a larger ring atop a smaller one.
#
# For a small number of rings, only a few moves are required.
#+ For each additional ring,
#+ the required number of moves approximately doubles,
#+ and the "strategy" becomes increasingly complicated.
#
# For more information, see http://hanoi.kernelthread.com
#+ or pp. 186-92 of _The Armchair Universe_ by A.K. Dewdney.
#
#
# ... ... ...
# | | | | | |
# _|_|_ | | | |
# |_____| | | | |
# |_______| | | | |
# |_________| | | | |
# |___________| | | | |
# | | | | | |
# .--------------------------------------------------------------.
# |**************************************************************|
# #1 #2 #3
#
#=================================================================#
E_NOPARAM=66 # No parameter passed to script.
E_BADPARAM=67 # Illegal number of disks passed to script.
Moves= # Global variable holding number of moves.
# Modification to original script.
dohanoi() { # Recursive function.
case $1 in
0)
;;
*)
dohanoi "$(($1-1))" $2 $4 $3
echo move $2 "--&#62;" $3
((Moves++)) # Modification to original script.
dohanoi "$(($1-1))" $4 $3 $2
;;
esac
}
case $# in
1) case $(($1&#62;0)) in # Must have at least one disk.
1) # Nested case statement.
dohanoi $1 1 3 2
echo "Total moves = $Moves" # 2^n - 1, where n = # of disks.
exit 0;
;;
*)
echo "$0: illegal value for number of disks";
exit $E_BADPARAM;
;;
esac
;;
*)
echo "usage: $0 N"
echo " Where \"N\" is the number of disks."
exit $E_NOPARAM;
;;
esac
# Exercises:
# ---------
# 1) Would commands beyond this point ever be executed?
# Why not? (Easy)
# 2) Explain the workings of the workings of the "dohanoi" function.
# (Difficult -- see the Dewdney reference, above.)</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="ALIASES"
></A
>Chapter 25. Aliases</H1
><P
><A
NAME="ALIASREF"
></A
></P
><P
>A Bash <I
CLASS="FIRSTTERM"
>alias</I
> is essentially nothing
more than a keyboard shortcut, an abbreviation, a means of
avoiding typing a long command sequence. If, for example,
we include <B
CLASS="COMMAND"
>alias lm="ls -l | more"</B
> in
the <A
HREF="#SAMPLE-BASHRC"
><TT
CLASS="FILENAME"
>~/.bashrc</TT
>
file</A
>, then each <TT
CLASS="USERINPUT"
><B
>lm</B
></TT
>
<A
NAME="AEN18669"
HREF="#FTN.AEN18669"
><SPAN
CLASS="footnote"
>[116]</SPAN
></A
>
typed at the command-line will automatically be replaced by a
<B
CLASS="COMMAND"
>ls -l | more</B
>. This can save a great deal of
typing at the command-line and avoid having to remember complex
combinations of commands and options. Setting <B
CLASS="COMMAND"
>alias
rm="rm -i"</B
> (interactive mode delete) may save a
good deal of grief, since it can prevent inadvertently deleting
important files.</P
><P
>In a script, aliases have very limited usefulness. It would be
nice if aliases could assume some of the functionality of
the <B
CLASS="COMMAND"
>C</B
> preprocessor, such as macro expansion,
but unfortunately Bash does not expand arguments within the
alias body.
<A
NAME="AEN18676"
HREF="#FTN.AEN18676"
><SPAN
CLASS="footnote"
>[117]</SPAN
></A
>
Moreover, a script fails to expand an alias itself
within <SPAN
CLASS="QUOTE"
>"compound constructs,"</SPAN
> such as <A
HREF="#IFTHEN"
>if/then</A
> statements, loops, and
functions. An added limitation is that an alias will not expand
recursively. Almost invariably, whatever we would like an alias
to do could be accomplished much more effectively with a <A
HREF="#FUNCTIONREF"
>function</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="AL"
></A
><P
><B
>Example 25-1. Aliases within a script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# alias.sh
shopt -s expand_aliases
# Must set this option, else script will not expand aliases.
# First, some fun.
alias Jesse_James='echo "\"Alias Jesse James\" was a 1959 comedy starring Bob Hope."'
Jesse_James
echo; echo; echo;
alias ll="ls -l"
# May use either single (') or double (") quotes to define an alias.
echo "Trying aliased \"ll\":"
ll /usr/X11R6/bin/mk* #* Alias works.
echo
directory=/usr/X11R6/bin/
prefix=mk* # See if wild card causes problems.
echo "Variables \"directory\" + \"prefix\" = $directory$prefix"
echo
alias lll="ls -l $directory$prefix"
echo "Trying aliased \"lll\":"
lll # Long listing of all files in /usr/X11R6/bin stating with mk.
# An alias can handle concatenated variables -- including wild card -- o.k.
TRUE=1
echo
if [ TRUE ]
then
alias rr="ls -l"
echo "Trying aliased \"rr\" within if/then statement:"
rr /usr/X11R6/bin/mk* #* Error message results!
# Aliases not expanded within compound statements.
echo "However, previously expanded alias still recognized:"
ll /usr/X11R6/bin/mk*
fi
echo
count=0
while [ $count -lt 3 ]
do
alias rrr="ls -l"
echo "Trying aliased \"rrr\" within \"while\" loop:"
rrr /usr/X11R6/bin/mk* #* Alias will not expand here either.
# alias.sh: line 57: rrr: command not found
let count+=1
done
echo; echo
alias xyz='cat $0' # Script lists itself.
# Note strong quotes.
xyz
# This seems to work,
#+ although the Bash documentation suggests that it shouldn't.
#
# However, as Steve Jacobson points out,
#+ the "$0" parameter expands immediately upon declaration of the alias.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="UNALIASREF"
></A
></P
><P
>The <B
CLASS="COMMAND"
>unalias</B
> command removes a previously
set <I
CLASS="FIRSTTERM"
>alias</I
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="UNAL"
></A
><P
><B
>Example 25-2. <I
CLASS="FIRSTTERM"
>unalias</I
>: Setting and unsetting
an alias</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# unalias.sh
shopt -s expand_aliases # Enables alias expansion.
alias llm='ls -al | more'
llm
echo
unalias llm # Unset alias.
llm
# Error message results, since 'llm' no longer recognized.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>./unalias.sh</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>total 6
drwxrwxr-x 2 bozo bozo 3072 Feb 6 14:04 .
drwxr-xr-x 40 bozo bozo 2048 Feb 6 14:04 ..
-rwxr-xr-x 1 bozo bozo 199 Feb 6 14:04 unalias.sh
./unalias.sh: llm: command not found</TT
></PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="LIST-CONS"
></A
>Chapter 26. List Constructs</H1
><P
><A
NAME="LISTCONSREF"
></A
></P
><P
>The <I
CLASS="FIRSTTERM"
>and list</I
> and <I
CLASS="FIRSTTERM"
>or
list</I
> constructs provide a means of processing a
number of commands consecutively. These can effectively replace
complex nested <A
HREF="#TESTCONSTRUCTS1"
>if/then</A
>
or even <A
HREF="#CASEESAC1"
>case</A
> statements.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="LCONS1"
></A
>Chaining together commands</B
></P
><DL
><DT
>and list</DT
><DD
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>command-1 &#38;&#38; command-2 &#38;&#38; command-3 &#38;&#38; ... command-n</PRE
></FONT
></TD
></TR
></TABLE
>
Each command executes in turn, provided that
the previous command has given a return value of
<TT
CLASS="REPLACEABLE"
><I
>true</I
></TT
> (zero). At the first
<TT
CLASS="REPLACEABLE"
><I
>false</I
></TT
> (non-zero) return, the
command chain terminates (the first command returning
<TT
CLASS="REPLACEABLE"
><I
>false</I
></TT
> is the last one to
execute).</P
><P
>An interesting use of a two-condition <I
CLASS="FIRSTTERM"
>and
list</I
> from an early version of YongYe's <A
HREF="http://bash.deta.in/Tetris_Game.sh"
TARGET="_top"
>Tetris
game script</A
>:</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>equation()
{ # core algorithm used for doubling and halving the coordinates
[[ ${cdx} ]] &#38;&#38; ((y=cy+(ccy-cdy)${2}2))
eval ${1}+=\"${x} ${y} \"
}</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX64"
></A
><P
><B
>Example 26-1. Using an <I
CLASS="FIRSTTERM"
>and list</I
> to test
for command-line arguments</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# and list
if [ ! -z "$1" ] &#38;&#38; echo "Argument #1 = $1" &#38;&#38; [ ! -z "$2" ] &#38;&#38; \
# ^^ ^^ ^^
echo "Argument #2 = $2"
then
echo "At least 2 arguments passed to script."
# All the chained commands return true.
else
echo "Fewer than 2 arguments passed to script."
# At least one of the chained commands returns false.
fi
# Note that "if [ ! -z $1 ]" works, but its alleged equivalent,
# "if [ -n $1 ]" does not.
# However, quoting fixes this.
# if "[ -n "$1" ]" works.
# ^ ^ Careful!
# It is always best to QUOTE the variables being tested.
# This accomplishes the same thing, using "pure" if/then statements.
if [ ! -z "$1" ]
then
echo "Argument #1 = $1"
fi
if [ ! -z "$2" ]
then
echo "Argument #2 = $2"
echo "At least 2 arguments passed to script."
else
echo "Fewer than 2 arguments passed to script."
fi
# It's longer and more ponderous than using an "and list".
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ANDLIST2"
></A
><P
><B
>Example 26-2. Another command-line arg test using an <I
CLASS="FIRSTTERM"
>and
list</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
ARGS=1 # Number of arguments expected.
E_BADARGS=85 # Exit value if incorrect number of args passed.
test $# -ne $ARGS &#38;&#38; \
# ^^^^^^^^^^^^ condition #1
echo "Usage: `basename $0` $ARGS argument(s)" &#38;&#38; exit $E_BADARGS
# ^^
# If condition #1 tests true (wrong number of args passed to script),
#+ then the rest of the line executes, and script terminates.
# Line below executes only if the above test fails.
echo "Correct number of arguments passed to this script."
exit 0
# To check exit value, do a "echo $?" after script termination.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ANDDEFAULT"
></A
></P
><P
> Of course, an <I
CLASS="FIRSTTERM"
>and list</I
> can also
<I
CLASS="FIRSTTERM"
>set</I
> variables to a default value.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>arg1=$@ &#38;&#38; [ -z "$arg1" ] &#38;&#38; arg1=DEFAULT
# Set $arg1 to command-line arguments, if any.
# But . . . set to DEFAULT if not specified on command-line.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
><DT
><A
NAME="ORLISTREF"
></A
>or list</DT
><DD
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>command-1 || command-2 || command-3 || ... command-n</PRE
></FONT
></TD
></TR
></TABLE
>
Each command executes in turn for as long as the previous
command returns <SPAN
CLASS="RETURNVALUE"
>false</SPAN
>. At
the first <SPAN
CLASS="RETURNVALUE"
>true</SPAN
> return, the
command chain terminates (the first command returning
<SPAN
CLASS="RETURNVALUE"
>true</SPAN
> is the last one to
execute). This is obviously the inverse of the <SPAN
CLASS="QUOTE"
>"and
list"</SPAN
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX65"
></A
><P
><B
>Example 26-3. Using <I
CLASS="FIRSTTERM"
>or lists</I
> in combination
with an <I
CLASS="FIRSTTERM"
>and list</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# delete.sh, a not-so-cunning file deletion utility.
# Usage: delete filename
E_BADARGS=85
if [ -z "$1" ]
then
echo "Usage: `basename $0` filename"
exit $E_BADARGS # No arg? Bail out.
else
file=$1 # Set filename.
fi
[ ! -f "$file" ] &#38;&#38; echo "File \"$file\" not found. \
Cowardly refusing to delete a nonexistent file."
# AND LIST, to give error message if file not present.
# Note echo message continuing on to a second line after an escape.
[ ! -f "$file" ] || (rm -f $file; echo "File \"$file\" deleted.")
# OR LIST, to delete file if present.
# Note logic inversion above.
# AND LIST executes on true, OR LIST on false.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>If the first command in an <I
CLASS="FIRSTTERM"
>or
list</I
> returns <SPAN
CLASS="RETURNVALUE"
>true</SPAN
>,
it <TT
CLASS="REPLACEABLE"
><I
>will</I
></TT
> execute.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># ==&#62; The following snippets from the /etc/rc.d/init.d/single
#+==&#62; script by Miquel van Smoorenburg
#+==&#62; illustrate use of "and" and "or" lists.
# ==&#62; "Arrowed" comments added by document author.
[ -x /usr/bin/clear ] &#38;&#38; /usr/bin/clear
# ==&#62; If /usr/bin/clear exists, then invoke it.
# ==&#62; Checking for the existence of a command before calling it
#+==&#62; avoids error messages and other awkward consequences.
# ==&#62; . . .
# If they want to run something in single user mode, might as well run it...
for i in /etc/rc1.d/S[0-9][0-9]* ; do
# Check if the script is there.
[ -x "$i" ] || continue
# ==&#62; If corresponding file in $PWD *not* found,
#+==&#62; then "continue" by jumping to the top of the loop.
# Reject backup files and files generated by rpm.
case "$1" in
*.rpmsave|*.rpmorig|*.rpmnew|*~|*.orig)
continue;;
esac
[ "$i" = "/etc/rc1.d/S00single" ] &#38;&#38; continue
# ==&#62; Set script name, but don't execute it yet.
$i start
done
# ==&#62; . . .</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="IMPORTANT"
><P
></P
><TABLE
CLASS="IMPORTANT"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/important.gif"
HSPACE="5"
ALT="Important"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <A
HREF="#EXITSTATUSREF"
>exit
status</A
> of an <TT
CLASS="USERINPUT"
><B
>and list</B
></TT
> or an
<TT
CLASS="USERINPUT"
><B
>or list</B
></TT
> is the exit status of the last
command executed.</P
></TD
></TR
></TABLE
></DIV
><P
>Clever combinations of <I
CLASS="FIRSTTERM"
>and</I
> and
<I
CLASS="FIRSTTERM"
>or</I
> lists are possible, but the logic may
easily become convoluted and require close attention to <A
HREF="#OPPRECEDENCE1"
>operator precedence rules</A
>, and
possibly extensive debugging.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>false &#38;&#38; true || echo false # false
# Same result as
( false &#38;&#38; true ) || echo false # false
# But NOT
false &#38;&#38; ( true || echo false ) # (nothing echoed)
# Note left-to-right grouping and evaluation of statements.
# It's usually best to avoid such complexities.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>See <A
HREF="#DAYSBETWEEN"
>Example A-7</A
> and <A
HREF="#BROKENLINK"
>Example 7-4</A
> for illustrations of using <TT
CLASS="USERINPUT"
><B
>and
/ or list</B
></TT
> constructs to test variables.</P
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="ARRAYS"
></A
>Chapter 27. Arrays</H1
><P
><A
NAME="ARRAYREF"
></A
></P
><P
>Newer versions of Bash support one-dimensional arrays.
<A
NAME="BRACKARRAY"
></A
>
Array elements may be initialized with the
<TT
CLASS="USERINPUT"
><B
>variable[xx]</B
></TT
> notation. Alternatively,
a script may introduce the entire array by an explicit
<TT
CLASS="USERINPUT"
><B
>declare -a variable</B
></TT
> statement. To
dereference (retrieve the contents of) an array element, use
<I
CLASS="FIRSTTERM"
>curly bracket</I
> notation, that is,
<TT
CLASS="USERINPUT"
><B
>${element[xx]}</B
></TT
>.</P
><P
><A
NAME="ARRAYNOTATION"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX66"
></A
><P
><B
>Example 27-1. Simple array usage</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
area[11]=23
area[13]=37
area[51]=UFOs
# Array members need not be consecutive or contiguous.
# Some members of the array can be left uninitialized.
# Gaps in the array are okay.
# In fact, arrays with sparse data ("sparse arrays")
#+ are useful in spreadsheet-processing software.
echo -n "area[11] = "
echo ${area[11]} # {curly brackets} needed.
echo -n "area[13] = "
echo ${area[13]}
echo "Contents of area[51] are ${area[51]}."
# Contents of uninitialized array variable print blank (null variable).
echo -n "area[43] = "
echo ${area[43]}
echo "(area[43] unassigned)"
echo
# Sum of two array variables assigned to third
area[5]=`expr ${area[11]} + ${area[13]}`
echo "area[5] = area[11] + area[13]"
echo -n "area[5] = "
echo ${area[5]}
area[6]=`expr ${area[11]} + ${area[51]}`
echo "area[6] = area[11] + area[51]"
echo -n "area[6] = "
echo ${area[6]}
# This fails because adding an integer to a string is not permitted.
echo; echo; echo
# -----------------------------------------------------------------
# Another array, "area2".
# Another way of assigning array variables...
# array_name=( XXX YYY ZZZ ... )
area2=( zero one two three four )
echo -n "area2[0] = "
echo ${area2[0]}
# Aha, zero-based indexing (first element of array is [0], not [1]).
echo -n "area2[1] = "
echo ${area2[1]} # [1] is second element of array.
# -----------------------------------------------------------------
echo; echo; echo
# -----------------------------------------------
# Yet another array, "area3".
# Yet another way of assigning array variables...
# array_name=([xx]=XXX [yy]=YYY ...)
area3=([17]=seventeen [24]=twenty-four)
echo -n "area3[17] = "
echo ${area3[17]}
echo -n "area3[24] = "
echo ${area3[24]}
# -----------------------------------------------
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ARRAYINIT0"
></A
></P
><P
>As we have seen, a convenient way of initializing an entire array
is the <TT
CLASS="VARNAME"
>array=( element1 element2 ... elementN )</TT
>
notation.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>base64_charset=( {A..Z} {a..z} {0..9} + / = )
# Using extended brace expansion
#+ to initialize the elements of the array.
# Excerpted from vladz's "base64.sh" script
#+ in the "Contributed Scripts" appendix.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="ARRAYOPSVARS"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN18812"
></A
><P
></P
><P
>Bash permits array operations on variables, even if
the variables are not explicitly declared as arrays.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>string=abcABC123ABCabc
echo ${string[@]} # abcABC123ABCabc
echo ${string[*]} # abcABC123ABCabc
echo ${string[0]} # abcABC123ABCabc
echo ${string[1]} # No output!
# Why?
echo ${#string[@]} # 1
# One element in the array.
# The string itself.
# Thank you, Michael Zick, for pointing this out.</PRE
></FONT
></TD
></TR
></TABLE
>
Once again this demonstrates that <A
HREF="#BVUNTYPED"
>Bash
variables are untyped</A
>.
</P
><P
></P
></DIV
></TD
></TR
></TABLE
><DIV
CLASS="EXAMPLE"
><A
NAME="POEM"
></A
><P
><B
>Example 27-2. Formatting a poem</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# poem.sh: Pretty-prints one of the ABS Guide author's favorite poems.
# Lines of the poem (single stanza).
Line[1]="I do not know which to prefer,"
Line[2]="The beauty of inflections"
Line[3]="Or the beauty of innuendoes,"
Line[4]="The blackbird whistling"
Line[5]="Or just after."
# Note that quoting permits embedding whitespace.
# Attribution.
Attrib[1]=" Wallace Stevens"
Attrib[2]="\"Thirteen Ways of Looking at a Blackbird\""
# This poem is in the Public Domain (copyright expired).
echo
tput bold # Bold print.
for index in 1 2 3 4 5 # Five lines.
do
printf " %s\n" "${Line[index]}"
done
for index in 1 2 # Two attribution lines.
do
printf " %s\n" "${Attrib[index]}"
done
tput sgr0 # Reset terminal.
# See 'tput' docs.
echo
exit 0
# Exercise:
# --------
# Modify this script to pretty-print a poem from a text data file.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ARRAYSYNTAX"
></A
></P
><P
>Array variables have a syntax all their own, and even
standard Bash commands and operators have special options adapted
for array use.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ARRAYOPS"
></A
><P
><B
>Example 27-3. Various array operations</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# array-ops.sh: More fun with arrays.
array=( zero one two three four five )
# Element 0 1 2 3 4 5
echo ${array[0]} # zero
echo ${array:0} # zero
# Parameter expansion of first element,
#+ starting at position # 0 (1st character).
echo ${array:1} # ero
# Parameter expansion of first element,
#+ starting at position # 1 (2nd character).
echo "--------------"
echo ${#array[0]} # 4
# Length of first element of array.
echo ${#array} # 4
# Length of first element of array.
# (Alternate notation)
echo ${#array[1]} # 3
# Length of second element of array.
# Arrays in Bash have zero-based indexing.
echo ${#array[*]} # 6
# Number of elements in array.
echo ${#array[@]} # 6
# Number of elements in array.
echo "--------------"
array2=( [0]="first element" [1]="second element" [3]="fourth element" )
# ^ ^ ^ ^ ^ ^ ^ ^ ^
# Quoting permits embedding whitespace within individual array elements.
echo ${array2[0]} # first element
echo ${array2[1]} # second element
echo ${array2[2]} #
# Skipped in initialization, and therefore null.
echo ${array2[3]} # fourth element
echo ${#array2[0]} # 13 (length of first element)
echo ${#array2[*]} # 3 (number of elements in array)
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ARRAYSTRINGOPS"
></A
></P
><P
>Many of the standard <A
HREF="#STRINGMANIP"
>string
operations</A
> work on arrays.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ARRAYSTROPS"
></A
><P
><B
>Example 27-4. String operations on arrays</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# array-strops.sh: String operations on arrays.
# Script by Michael Zick.
# Used in ABS Guide with permission.
# Fixups: 05 May 08, 04 Aug 08.
# In general, any string operation using the ${name ... } notation
#+ can be applied to all string elements in an array,
#+ with the ${name[@] ... } or ${name[*] ...} notation.
arrayZ=( one two three four five five )
echo
# Trailing Substring Extraction
echo ${arrayZ[@]:0} # one two three four five five
# ^ All elements.
echo ${arrayZ[@]:1} # two three four five five
# ^ All elements following element[0].
echo ${arrayZ[@]:1:2} # two three
# ^ Only the two elements after element[0].
echo "---------"
# Substring Removal
# Removes shortest match from front of string(s).
echo ${arrayZ[@]#f*r} # one two three five five
# ^ # Applied to all elements of the array.
# Matches "four" and removes it.
# Longest match from front of string(s)
echo ${arrayZ[@]##t*e} # one two four five five
# ^^ # Applied to all elements of the array.
# Matches "three" and removes it.
# Shortest match from back of string(s)
echo ${arrayZ[@]%h*e} # one two t four five five
# ^ # Applied to all elements of the array.
# Matches "hree" and removes it.
# Longest match from back of string(s)
echo ${arrayZ[@]%%t*e} # one two four five five
# ^^ # Applied to all elements of the array.
# Matches "three" and removes it.
echo "----------------------"
# Substring Replacement
# Replace first occurrence of substring with replacement.
echo ${arrayZ[@]/fiv/XYZ} # one two three four XYZe XYZe
# ^ # Applied to all elements of the array.
# Replace all occurrences of substring.
echo ${arrayZ[@]//iv/YY} # one two three four fYYe fYYe
# Applied to all elements of the array.
# Delete all occurrences of substring.
# Not specifing a replacement defaults to 'delete' ...
echo ${arrayZ[@]//fi/} # one two three four ve ve
# ^^ # Applied to all elements of the array.
# Replace front-end occurrences of substring.
echo ${arrayZ[@]/#fi/XY} # one two three four XYve XYve
# ^ # Applied to all elements of the array.
# Replace back-end occurrences of substring.
echo ${arrayZ[@]/%ve/ZZ} # one two three four fiZZ fiZZ
# ^ # Applied to all elements of the array.
echo ${arrayZ[@]/%o/XX} # one twXX three four five five
# ^ # Why?
echo "-----------------------------"
replacement() {
echo -n "!!!"
}
echo ${arrayZ[@]/%e/$(replacement)}
# ^ ^^^^^^^^^^^^^^
# on!!! two thre!!! four fiv!!! fiv!!!
# The stdout of replacement() is the replacement string.
# Q.E.D: The replacement action is, in effect, an 'assignment.'
echo "------------------------------------"
# Accessing the "for-each":
echo ${arrayZ[@]//*/$(replacement optional_arguments)}
# ^^ ^^^^^^^^^^^^^
# !!! !!! !!! !!! !!! !!!
# Now, if Bash would only pass the matched string
#+ to the function being called . . .
echo
exit 0
# Before reaching for a Big Hammer -- Perl, Python, or all the rest --
# recall:
# $( ... ) is command substitution.
# A function runs as a sub-process.
# A function writes its output (if echo-ed) to stdout.
# Assignment, in conjunction with "echo" and command substitution,
#+ can read a function's stdout.
# The name[@] notation specifies (the equivalent of) a "for-each"
#+ operation.
# Bash is more powerful than you think!</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
HREF="#COMMANDSUBREF"
>Command substitution</A
> can
construct the individual elements of an array.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SCRIPTARRAY"
></A
><P
><B
>Example 27-5. Loading the contents of a script into an array</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# script-array.sh: Loads this script into an array.
# Inspired by an e-mail from Chris Martin (thanks!).
script_contents=( $(cat "$0") ) # Stores contents of this script ($0)
#+ in an array.
for element in $(seq 0 $((${#script_contents[@]} - 1)))
do # ${#script_contents[@]}
#+ gives number of elements in the array.
#
# Question:
# Why is seq 0 necessary?
# Try changing it to seq 1.
echo -n "${script_contents[$element]}"
# List each field of this script on a single line.
# echo -n "${script_contents[element]}" also works because of ${ ... }.
echo -n " -- " # Use " -- " as a field separator.
done
echo
exit 0
# Exercise:
# --------
# Modify this script so it lists itself
#+ in its original format,
#+ complete with whitespace, line breaks, etc.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>In an array context, some Bash <A
HREF="#BUILTINREF"
>builtins</A
> have a slightly
altered meaning. <A
NAME="ARRAYUNSET"
></A
>For example, <A
HREF="#UNSETREF"
>unset</A
> deletes array elements, or even
an entire array.</P
><P
><A
NAME="ARRAYSPECIALPROPS"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX67"
></A
><P
><B
>Example 27-6. Some special properties of arrays</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
declare -a colors
# All subsequent commands in this script will treat
#+ the variable "colors" as an array.
echo "Enter your favorite colors (separated from each other by a space)."
read -a colors # Enter at least 3 colors to demonstrate features below.
# Special option to 'read' command,
#+ allowing assignment of elements in an array.
echo
element_count=${#colors[@]}
# Special syntax to extract number of elements in array.
# element_count=${#colors[*]} works also.
#
# The "@" variable allows word splitting within quotes
#+ (extracts variables separated by whitespace).
#
# This corresponds to the behavior of "$@" and "$*"
#+ in positional parameters.
index=0
while [ "$index" -lt "$element_count" ]
do # List all the elements in the array.
echo ${colors[$index]}
# ${colors[index]} also works because it's within ${ ... } brackets.
let "index = $index + 1"
# Or:
# ((index++))
done
# Each array element listed on a separate line.
# If this is not desired, use echo -n "${colors[$index]} "
#
# Doing it with a "for" loop instead:
# for i in "${colors[@]}"
# do
# echo "$i"
# done
# (Thanks, S.C.)
echo
# Again, list all the elements in the array, but using a more elegant method.
echo ${colors[@]} # echo ${colors[*]} also works.
echo
# The "unset" command deletes elements of an array, or entire array.
unset colors[1] # Remove 2nd element of array.
# Same effect as colors[1]=
echo ${colors[@]} # List array again, missing 2nd element.
unset colors # Delete entire array.
# unset colors[*] and
#+ unset colors[@] also work.
echo; echo -n "Colors gone."
echo ${colors[@]} # List array again, now empty.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ARRAYNUMELEMENTS"
></A
></P
><P
>As seen in the previous example, either
<B
CLASS="COMMAND"
>${array_name[@]}</B
> or
<B
CLASS="COMMAND"
>${array_name[*]}</B
> refers to
<EM
>all</EM
> the elements of the array.
Similarly, to get a count of the number of elements in an
array, use either <B
CLASS="COMMAND"
>${#array_name[@]}</B
>
or <B
CLASS="COMMAND"
>${#array_name[*]}</B
>.
<B
CLASS="COMMAND"
>${#array_name}</B
> is the length (number of
characters) of <B
CLASS="COMMAND"
>${array_name[0]}</B
>, the first
element of the array.</P
><P
><A
NAME="EMPTYARRAY0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EMPTYARRAY"
></A
><P
><B
>Example 27-7. Of empty arrays and empty elements</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# empty-array.sh
# Thanks to Stephane Chazelas for the original example,
#+ and to Michael Zick and Omair Eshkenazi, for extending it.
# And to Nathan Coulter for clarifications and corrections.
# An empty array is not the same as an array with empty elements.
array0=( first second third )
array1=( '' ) # "array1" consists of one empty element.
array2=( ) # No elements . . . "array2" is empty.
array3=( ) # What about this array?
echo
ListArray()
{
echo
echo "Elements in array0: ${array0[@]}"
echo "Elements in array1: ${array1[@]}"
echo "Elements in array2: ${array2[@]}"
echo "Elements in array3: ${array3[@]}"
echo
echo "Length of first element in array0 = ${#array0}"
echo "Length of first element in array1 = ${#array1}"
echo "Length of first element in array2 = ${#array2}"
echo "Length of first element in array3 = ${#array3}"
echo
echo "Number of elements in array0 = ${#array0[*]}" # 3
echo "Number of elements in array1 = ${#array1[*]}" # 1 (Surprise!)
echo "Number of elements in array2 = ${#array2[*]}" # 0
echo "Number of elements in array3 = ${#array3[*]}" # 0
}
# ===================================================================
ListArray
# Try extending those arrays.
# Adding an element to an array.
array0=( "${array0[@]}" "new1" )
array1=( "${array1[@]}" "new1" )
array2=( "${array2[@]}" "new1" )
array3=( "${array3[@]}" "new1" )
ListArray
# or
array0[${#array0[*]}]="new2"
array1[${#array1[*]}]="new2"
array2[${#array2[*]}]="new2"
array3[${#array3[*]}]="new2"
ListArray
# When extended as above, arrays are 'stacks' ...
# Above is the 'push' ...
# The stack 'height' is:
height=${#array2[@]}
echo
echo "Stack height for array2 = $height"
# The 'pop' is:
unset array2[${#array2[@]}-1] # Arrays are zero-based,
height=${#array2[@]} #+ which means first element has index 0.
echo
echo "POP"
echo "New stack height for array2 = $height"
ListArray
# List only 2nd and 3rd elements of array0.
from=1 # Zero-based numbering.
to=2
array3=( ${array0[@]:1:2} )
echo
echo "Elements in array3: ${array3[@]}"
# Works like a string (array of characters).
# Try some other "string" forms.
# Replacement:
array4=( ${array0[@]/second/2nd} )
echo
echo "Elements in array4: ${array4[@]}"
# Replace all matching wildcarded string.
array5=( ${array0[@]//new?/old} )
echo
echo "Elements in array5: ${array5[@]}"
# Just when you are getting the feel for this . . .
array6=( ${array0[@]#*new} )
echo # This one might surprise you.
echo "Elements in array6: ${array6[@]}"
array7=( ${array0[@]#new1} )
echo # After array6 this should not be a surprise.
echo "Elements in array7: ${array7[@]}"
# Which looks a lot like . . .
array8=( ${array0[@]/new1/} )
echo
echo "Elements in array8: ${array8[@]}"
# So what can one say about this?
# The string operations are performed on
#+ each of the elements in var[@] in succession.
# Therefore : Bash supports string vector operations.
# If the result is a zero length string,
#+ that element disappears in the resulting assignment.
# However, if the expansion is in quotes, the null elements remain.
# Michael Zick: Question, are those strings hard or soft quotes?
# Nathan Coulter: There is no such thing as "soft quotes."
#! What's really happening is that
#!+ the pattern matching happens after
#!+ all the other expansions of [word]
#!+ in cases like ${parameter#word}.
zap='new*'
array9=( ${array0[@]/$zap/} )
echo
echo "Number of elements in array9: ${#array9[@]}"
array9=( "${array0[@]/$zap/}" )
echo "Elements in array9: ${array9[@]}"
# This time the null elements remain.
echo "Number of elements in array9: ${#array9[@]}"
# Just when you thought you were still in Kansas . . .
array10=( ${array0[@]#$zap} )
echo
echo "Elements in array10: ${array10[@]}"
# But, the asterisk in zap won't be interpreted if quoted.
array10=( ${array0[@]#"$zap"} )
echo
echo "Elements in array10: ${array10[@]}"
# Well, maybe we _are_ still in Kansas . . .
# (Revisions to above code block by Nathan Coulter.)
# Compare array7 with array10.
# Compare array8 with array9.
# Reiterating: No such thing as soft quotes!
# Nathan Coulter explains:
# Pattern matching of 'word' in ${parameter#word} is done after
#+ parameter expansion and *before* quote removal.
# In the normal case, pattern matching is done *after* quote removal.
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The relationship of <B
CLASS="COMMAND"
>${array_name[@]}</B
>
and <B
CLASS="COMMAND"
>${array_name[*]}</B
> is analogous to that
between <A
HREF="#APPREF"
>$@ and $*</A
>. This powerful
array notation has a number of uses.</P
><P
><A
NAME="COPYARRAY0"
></A
></P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Copying an array.
array2=( "${array1[@]}" )
# or
array2="${array1[@]}"
#
# However, this fails with "sparse" arrays,
#+ arrays with holes (missing elements) in them,
#+ as Jochen DeSmet points out.
# ------------------------------------------
array1[0]=0
# array1[1] not assigned
array1[2]=2
array2=( "${array1[@]}" ) # Copy it?
echo ${array2[0]} # 0
echo ${array2[2]} # (null), should be 2
# ------------------------------------------
# Adding an element to an array.
array=( "${array[@]}" "new element" )
# or
array[${#array[*]}]="new element"
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="ARRAYINITCS"
></A
></P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <B
CLASS="COMMAND"
>array=( element1 element2 ... elementN )</B
>
initialization operation, with the help of <A
HREF="#COMMANDSUBREF"
>command substitution</A
>, makes it
possible to load the contents of a text file into an array.</P
><P
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
filename=sample_file
# cat sample_file
#
# 1 a b c
# 2 d e fg
declare -a array1
array1=( `cat "$filename"`) # Loads contents
# List file to stdout #+ of $filename into array1.
#
# array1=( `cat "$filename" | tr '\n' ' '`)
# change linefeeds in file to spaces.
# Not necessary because Bash does word splitting,
#+ changing linefeeds to spaces.
echo ${array1[@]} # List the array.
# 1 a b c 2 d e fg
#
# Each whitespace-separated "word" in the file
#+ has been assigned to an element of the array.
element_count=${#array1[*]}
echo $element_count # 8</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><P
>Clever scripting makes it possible to add array operations.</P
><P
><A
NAME="ARRAYASSIGN0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="ARRAYASSIGN"
></A
><P
><B
>Example 27-8. Initializing arrays</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
# array-assign.bash
# Array operations are Bash-specific,
#+ hence the ".bash" in the script name.
# Copyright (c) Michael S. Zick, 2003, All rights reserved.
# License: Unrestricted reuse in any form, for any purpose.
# Version: $ID$
#
# Clarification and additional comments by William Park.
# Based on an example provided by Stephane Chazelas
#+ which appeared in an earlier version of the
#+ Advanced Bash Scripting Guide.
# Output format of the 'times' command:
# User CPU &#60;space&#62; System CPU
# User CPU of dead children &#60;space&#62; System CPU of dead children
# Bash has two versions of assigning all elements of an array
#+ to a new array variable.
# Both drop 'null reference' elements
#+ in Bash versions 2.04 and later.
# An additional array assignment that maintains the relationship of
#+ [subscript]=value for arrays may be added to newer versions.
# Constructs a large array using an internal command,
#+ but anything creating an array of several thousand elements
#+ will do just fine.
declare -a bigOne=( /dev/* ) # All the files in /dev . . .
echo
echo 'Conditions: Unquoted, default IFS, All-Elements-Of'
echo "Number of elements in array is ${#bigOne[@]}"
# set -vx
echo
echo '- - testing: =( ${array[@]} ) - -'
times
declare -a bigTwo=( ${bigOne[@]} )
# Note parens: ^ ^
times
echo
echo '- - testing: =${array[@]} - -'
times
declare -a bigThree=${bigOne[@]}
# No parentheses this time.
times
# Comparing the numbers shows that the second form, pointed out
#+ by Stephane Chazelas, is faster.
#
# As William Park explains:
#+ The bigTwo array assigned element by element (because of parentheses),
#+ whereas bigThree assigned as a single string.
# So, in essence, you have:
# bigTwo=( [0]="..." [1]="..." [2]="..." ... )
# bigThree=( [0]="... ... ..." )
#
# Verify this by: echo ${bigTwo[0]}
# echo ${bigThree[0]}
# I will continue to use the first form in my example descriptions
#+ because I think it is a better illustration of what is happening.
# The reusable portions of my examples will actual contain
#+ the second form where appropriate because of the speedup.
# MSZ: Sorry about that earlier oversight folks.
# Note:
# ----
# The "declare -a" statements in lines 32 and 44
#+ are not strictly necessary, since it is implicit
#+ in the Array=( ... ) assignment form.
# However, eliminating these declarations slows down
#+ the execution of the following sections of the script.
# Try it, and see.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Adding a superfluous <B
CLASS="COMMAND"
>declare -a</B
>
statement to an array declaration may speed up execution of
subsequent operations on the array.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ARRAYAPPEND0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="COPYARRAY"
></A
><P
><B
>Example 27-9. Copying and concatenating arrays</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
# CopyArray.sh
#
# This script written by Michael Zick.
# Used here with permission.
# How-To "Pass by Name &#38; Return by Name"
#+ or "Building your own assignment statement".
CpArray_Mac() {
# Assignment Command Statement Builder
echo -n 'eval '
echo -n "$2" # Destination name
echo -n '=( ${'
echo -n "$1" # Source name
echo -n '[@]} )'
# That could all be a single command.
# Matter of style only.
}
declare -f CopyArray # Function "Pointer"
CopyArray=CpArray_Mac # Statement Builder
Hype()
{
# Hype the array named $1.
# (Splice it together with array containing "Really Rocks".)
# Return in array named $2.
local -a TMP
local -a hype=( Really Rocks )
$($CopyArray $1 TMP)
TMP=( ${TMP[@]} ${hype[@]} )
$($CopyArray TMP $2)
}
declare -a before=( Advanced Bash Scripting )
declare -a after
echo "Array Before = ${before[@]}"
Hype before after
echo "Array After = ${after[@]}"
# Too much hype?
echo "What ${after[@]:3:2}?"
declare -a modest=( ${after[@]:2:1} ${after[@]:3:2} )
# ---- substring extraction ----
echo "Array Modest = ${modest[@]}"
# What happened to 'before' ?
echo "Array Before = ${before[@]}"
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ARRAYAPPEND"
></A
><P
><B
>Example 27-10. More on concatenating arrays</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
# array-append.bash
# Copyright (c) Michael S. Zick, 2003, All rights reserved.
# License: Unrestricted reuse in any form, for any purpose.
# Version: $ID$
#
# Slightly modified in formatting by M.C.
# Array operations are Bash-specific.
# Legacy UNIX /bin/sh lacks equivalents.
# Pipe the output of this script to 'more'
#+ so it doesn't scroll off the terminal.
# Or, redirect output to a file.
declare -a array1=( zero1 one1 two1 )
# Subscript packed.
declare -a array2=( [0]=zero2 [2]=two2 [3]=three2 )
# Subscript sparse -- [1] is not defined.
echo
echo '- Confirm that the array is really subscript sparse. -'
echo "Number of elements: 4" # Hard-coded for illustration.
for (( i = 0 ; i &#60; 4 ; i++ ))
do
echo "Element [$i]: ${array2[$i]}"
done
# See also the more general code example in basics-reviewed.bash.
declare -a dest
# Combine (append) two arrays into a third array.
echo
echo 'Conditions: Unquoted, default IFS, All-Elements-Of operator'
echo '- Undefined elements not present, subscripts not maintained. -'
# # The undefined elements do not exist; they are not being dropped.
dest=( ${array1[@]} ${array2[@]} )
# dest=${array1[@]}${array2[@]} # Strange results, possibly a bug.
# Now, list the result.
echo
echo '- - Testing Array Append - -'
cnt=${#dest[@]}
echo "Number of elements: $cnt"
for (( i = 0 ; i &#60; cnt ; i++ ))
do
echo "Element [$i]: ${dest[$i]}"
done
# Assign an array to a single array element (twice).
dest[0]=${array1[@]}
dest[1]=${array2[@]}
# List the result.
echo
echo '- - Testing modified array - -'
cnt=${#dest[@]}
echo "Number of elements: $cnt"
for (( i = 0 ; i &#60; cnt ; i++ ))
do
echo "Element [$i]: ${dest[$i]}"
done
# Examine the modified second element.
echo
echo '- - Reassign and list second element - -'
declare -a subArray=${dest[1]}
cnt=${#subArray[@]}
echo "Number of elements: $cnt"
for (( i = 0 ; i &#60; cnt ; i++ ))
do
echo "Element [$i]: ${subArray[$i]}"
done
# The assignment of an entire array to a single element
#+ of another array using the '=${ ... }' array assignment
#+ has converted the array being assigned into a string,
#+ with the elements separated by a space (the first character of IFS).
# If the original elements didn't contain whitespace . . .
# If the original array isn't subscript sparse . . .
# Then we could get the original array structure back again.
# Restore from the modified second element.
echo
echo '- - Listing restored element - -'
declare -a subArray=( ${dest[1]} )
cnt=${#subArray[@]}
echo "Number of elements: $cnt"
for (( i = 0 ; i &#60; cnt ; i++ ))
do
echo "Element [$i]: ${subArray[$i]}"
done
echo '- - Do not depend on this behavior. - -'
echo '- - This behavior is subject to change - -'
echo '- - in versions of Bash newer than version 2.05b - -'
# MSZ: Sorry about any earlier confusion folks.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>--</P
><P
>Arrays permit deploying old familiar algorithms as shell scripts.
Whether this is necessarily a good idea is left for the reader to
decide.</P
><P
><A
NAME="BUBBLESORT"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="BUBBLE"
></A
><P
><B
>Example 27-11. The Bubble Sort</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# bubble.sh: Bubble sort, of sorts.
# Recall the algorithm for a bubble sort. In this particular version...
# With each successive pass through the array to be sorted,
#+ compare two adjacent elements, and swap them if out of order.
# At the end of the first pass, the "heaviest" element has sunk to bottom.
# At the end of the second pass, the next "heaviest" one has sunk next to bottom.
# And so forth.
# This means that each successive pass needs to traverse less of the array.
# You will therefore notice a speeding up in the printing of the later passes.
exchange()
{
# Swaps two members of the array.
local temp=${Countries[$1]} # Temporary storage
#+ for element getting swapped out.
Countries[$1]=${Countries[$2]}
Countries[$2]=$temp
return
}
declare -a Countries # Declare array,
#+ optional here since it's initialized below.
# Is it permissable to split an array variable over multiple lines
#+ using an escape (\)?
# Yes.
Countries=(Netherlands Ukraine Zaire Turkey Russia Yemen Syria \
Brazil Argentina Nicaragua Japan Mexico Venezuela Greece England \
Israel Peru Canada Oman Denmark Wales France Kenya \
Xanadu Qatar Liechtenstein Hungary)
# "Xanadu" is the mythical place where, according to Coleridge,
#+ Kubla Khan did a pleasure dome decree.
clear # Clear the screen to start with.
echo "0: ${Countries[*]}" # List entire array at pass 0.
number_of_elements=${#Countries[@]}
let "comparisons = $number_of_elements - 1"
count=1 # Pass number.
while [ "$comparisons" -gt 0 ] # Beginning of outer loop
do
index=0 # Reset index to start of array after each pass.
while [ "$index" -lt "$comparisons" ] # Beginning of inner loop
do
if [ ${Countries[$index]} \&#62; ${Countries[`expr $index + 1`]} ]
# If out of order...
# Recalling that \&#62; is ASCII comparison operator
#+ within single brackets.
# if [[ ${Countries[$index]} &#62; ${Countries[`expr $index + 1`]} ]]
#+ also works.
then
exchange $index `expr $index + 1` # Swap.
fi
let "index += 1" # Or, index+=1 on Bash, ver. 3.1 or newer.
done # End of inner loop
# ----------------------------------------------------------------------
# Paulo Marcel Coelho Aragao suggests for-loops as a simpler altenative.
#
# for (( last = $number_of_elements - 1 ; last &#62; 0 ; last-- ))
## Fix by C.Y. Hunt ^ (Thanks!)
# do
# for (( i = 0 ; i &#60; last ; i++ ))
# do
# [[ "${Countries[$i]}" &#62; "${Countries[$((i+1))]}" ]] \
# &#38;&#38; exchange $i $((i+1))
# done
# done
# ----------------------------------------------------------------------
let "comparisons -= 1" # Since "heaviest" element bubbles to bottom,
#+ we need do one less comparison each pass.
echo
echo "$count: ${Countries[@]}" # Print resultant array at end of each pass.
echo
let "count += 1" # Increment pass count.
done # End of outer loop
# All done.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>--</P
><P
><A
NAME="ARRAYNEST"
></A
></P
><P
>Is it possible to nest arrays within arrays?</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# "Nested" array.
# Michael Zick provided this example,
#+ with corrections and clarifications by William Park.
AnArray=( $(ls --inode --ignore-backups --almost-all \
--directory --full-time --color=none --time=status \
--sort=time -l ${PWD} ) ) # Commands and options.
# Spaces are significant . . . and don't quote anything in the above.
SubArray=( ${AnArray[@]:11:1} ${AnArray[@]:6:5} )
# This array has six elements:
#+ SubArray=( [0]=${AnArray[11]} [1]=${AnArray[6]} [2]=${AnArray[7]}
# [3]=${AnArray[8]} [4]=${AnArray[9]} [5]=${AnArray[10]} )
#
# Arrays in Bash are (circularly) linked lists
#+ of type string (char *).
# So, this isn't actually a nested array,
#+ but it's functionally similar.
echo "Current directory and date of last status change:"
echo "${SubArray[@]}"
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>--</P
><P
>Embedded arrays in combination with <A
HREF="#VARREFNEW"
>indirect references</A
> create some fascinating
possibilities</P
><P
><A
NAME="ARRAYINDIR"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EMBARR"
></A
><P
><B
>Example 27-12. Embedded arrays and indirect references</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# embedded-arrays.sh
# Embedded arrays and indirect references.
# This script by Dennis Leeuw.
# Used with permission.
# Modified by document author.
ARRAY1=(
VAR1_1=value11
VAR1_2=value12
VAR1_3=value13
)
ARRAY2=(
VARIABLE="test"
STRING="VAR1=value1 VAR2=value2 VAR3=value3"
ARRAY21=${ARRAY1[*]}
) # Embed ARRAY1 within this second array.
function print () {
OLD_IFS="$IFS"
IFS=$'\n' # To print each array element
#+ on a separate line.
TEST1="ARRAY2[*]"
local ${!TEST1} # See what happens if you delete this line.
# Indirect reference.
# This makes the components of $TEST1
#+ accessible to this function.
# Let's see what we've got so far.
echo
echo "\$TEST1 = $TEST1" # Just the name of the variable.
echo; echo
echo "{\$TEST1} = ${!TEST1}" # Contents of the variable.
# That's what an indirect
#+ reference does.
echo
echo "-------------------------------------------"; echo
echo
# Print variable
echo "Variable VARIABLE: $VARIABLE"
# Print a string element
IFS="$OLD_IFS"
TEST2="STRING[*]"
local ${!TEST2} # Indirect reference (as above).
echo "String element VAR2: $VAR2 from STRING"
# Print an array element
TEST2="ARRAY21[*]"
local ${!TEST2} # Indirect reference (as above).
echo "Array element VAR1_1: $VAR1_1 from ARRAY21"
}
print
echo
exit 0
# As the author of the script notes,
#+ "you can easily expand it to create named-hashes in bash."
# (Difficult) exercise for the reader: implement this.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>--</P
><P
><A
NAME="PRIMES0"
></A
></P
><P
>Arrays enable implementing a shell script version of the
<I
CLASS="FIRSTTERM"
>Sieve of Eratosthenes</I
>. Of course, a
resource-intensive application of this nature should really be
written in a compiled language, such as C. It runs excruciatingly
slowly as a script.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX68"
></A
><P
><B
>Example 27-13. The Sieve of Eratosthenes</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# sieve.sh (ex68.sh)
# Sieve of Eratosthenes
# Ancient algorithm for finding prime numbers.
# This runs a couple of orders of magnitude slower
#+ than the equivalent program written in C.
LOWER_LIMIT=1 # Starting with 1.
UPPER_LIMIT=1000 # Up to 1000.
# (You may set this higher . . . if you have time on your hands.)
PRIME=1
NON_PRIME=0
let SPLIT=UPPER_LIMIT/2
# Optimization:
# Need to test numbers only halfway to upper limit. Why?
declare -a Primes
# Primes[] is an array.
initialize ()
{
# Initialize the array.
i=$LOWER_LIMIT
until [ "$i" -gt "$UPPER_LIMIT" ]
do
Primes[i]=$PRIME
let "i += 1"
done
# Assume all array members guilty (prime)
#+ until proven innocent.
}
print_primes ()
{
# Print out the members of the Primes[] array tagged as prime.
i=$LOWER_LIMIT
until [ "$i" -gt "$UPPER_LIMIT" ]
do
if [ "${Primes[i]}" -eq "$PRIME" ]
then
printf "%8d" $i
# 8 spaces per number gives nice, even columns.
fi
let "i += 1"
done
}
sift () # Sift out the non-primes.
{
let i=$LOWER_LIMIT+1
# Let's start with 2.
until [ "$i" -gt "$UPPER_LIMIT" ]
do
if [ "${Primes[i]}" -eq "$PRIME" ]
# Don't bother sieving numbers already sieved (tagged as non-prime).
then
t=$i
while [ "$t" -le "$UPPER_LIMIT" ]
do
let "t += $i "
Primes[t]=$NON_PRIME
# Tag as non-prime all multiples.
done
fi
let "i += 1"
done
}
# ==============================================
# main ()
# Invoke the functions sequentially.
initialize
sift
print_primes
# This is what they call structured programming.
# ==============================================
echo
exit 0
# -------------------------------------------------------- #
# Code below line will not execute, because of 'exit.'
# This improved version of the Sieve, by Stephane Chazelas,
#+ executes somewhat faster.
# Must invoke with command-line argument (limit of primes).
UPPER_LIMIT=$1 # From command-line.
let SPLIT=UPPER_LIMIT/2 # Halfway to max number.
Primes=( '' $(seq $UPPER_LIMIT) )
i=1
until (( ( i += 1 ) &#62; SPLIT )) # Need check only halfway.
do
if [[ -n ${Primes[i]} ]]
then
t=$i
until (( ( t += i ) &#62; UPPER_LIMIT ))
do
Primes[t]=
done
fi
done
echo ${Primes[*]}
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX68A"
></A
><P
><B
>Example 27-14. The Sieve of Eratosthenes, Optimized</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Optimized Sieve of Eratosthenes
# Script by Jared Martin, with very minor changes by ABS Guide author.
# Used in ABS Guide with permission (thanks!).
# Based on script in Advanced Bash Scripting Guide.
# http://tldp.org/LDP/abs/html/arrays.html#PRIMES0 (ex68.sh).
# http://www.cs.hmc.edu/~oneill/papers/Sieve-JFP.pdf (reference)
# Check results against http://primes.utm.edu/lists/small/1000.txt
# Necessary but not sufficient would be, e.g.,
# (($(sieve 7919 | wc -w) == 1000)) &#38;&#38; echo "7919 is the 1000th prime"
UPPER_LIMIT=${1:?"Need an upper limit of primes to search."}
Primes=( '' $(seq ${UPPER_LIMIT}) )
typeset -i i t
Primes[i=1]='' # 1 is not a prime.
until (( ( i += 1 ) &#62; (${UPPER_LIMIT}/i) )) # Need check only ith-way.
do # Why?
if ((${Primes[t=i*(i-1), i]}))
# Obscure, but instructive, use of arithmetic expansion in subscript.
then
until (( ( t += i ) &#62; ${UPPER_LIMIT} ))
do Primes[t]=; done
fi
done
# echo ${Primes[*]}
echo # Change to original script for pretty-printing (80-col. display).
printf "%8d" ${Primes[*]}
echo; echo
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Compare these array-based prime number generators with
alternatives that do not use arrays, <A
HREF="#PRIMES"
>Example A-15</A
>,
and <A
HREF="#PRIMES2"
>Example 16-46</A
>.</P
><P
>--</P
><P
>Arrays lend themselves, to some extent, to emulating data
structures for which Bash has no native support.</P
><P
><A
NAME="STACKEX0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="STACKEX"
></A
><P
><B
>Example 27-15. Emulating a push-down stack</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# stack.sh: push-down stack simulation
# Similar to the CPU stack, a push-down stack stores data items
#+ sequentially, but releases them in reverse order, last-in first-out.
BP=100 # Base Pointer of stack array.
# Begin at element 100.
SP=$BP # Stack Pointer.
# Initialize it to "base" (bottom) of stack.
Data= # Contents of stack location.
# Must use global variable,
#+ because of limitation on function return range.
# 100 Base pointer &#60;-- Base Pointer
# 99 First data item
# 98 Second data item
# ... More data
# Last data item &#60;-- Stack pointer
declare -a stack
push() # Push item on stack.
{
if [ -z "$1" ] # Nothing to push?
then
return
fi
let "SP -= 1" # Bump stack pointer.
stack[$SP]=$1
return
}
pop() # Pop item off stack.
{
Data= # Empty out data item.
if [ "$SP" -eq "$BP" ] # Stack empty?
then
return
fi # This also keeps SP from getting past 100,
#+ i.e., prevents a runaway stack.
Data=${stack[$SP]}
let "SP += 1" # Bump stack pointer.
return
}
status_report() # Find out what's happening.
{
echo "-------------------------------------"
echo "REPORT"
echo "Stack Pointer = $SP"
echo "Just popped \""$Data"\" off the stack."
echo "-------------------------------------"
echo
}
# =======================================================
# Now, for some fun.
echo
# See if you can pop anything off empty stack.
pop
status_report
echo
push garbage
pop
status_report # Garbage in, garbage out.
value1=23; push $value1
value2=skidoo; push $value2
value3=LAST; push $value3
pop # LAST
status_report
pop # skidoo
status_report
pop # 23
status_report # Last-in, first-out!
# Notice how the stack pointer decrements with each push,
#+ and increments with each pop.
echo
exit 0
# =======================================================
# Exercises:
# ---------
# 1) Modify the "push()" function to permit pushing
# + multiple element on the stack with a single function call.
# 2) Modify the "pop()" function to permit popping
# + multiple element from the stack with a single function call.
# 3) Add error checking to the critical functions.
# That is, return an error code, depending on
# + successful or unsuccessful completion of the operation,
# + and take appropriate action.
# 4) Using this script as a starting point,
# + write a stack-based 4-function calculator.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>--</P
><P
>Fancy manipulation of array <SPAN
CLASS="QUOTE"
>"subscripts"</SPAN
> may require
intermediate variables. For projects involving this, again consider
using a more powerful programming language, such as Perl or C.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="QFUNCTION"
></A
><P
><B
>Example 27-16. Complex array application:
<EM
>Exploring a weird mathematical series</EM
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Douglas Hofstadter's notorious "Q-series":
# Q(1) = Q(2) = 1
# Q(n) = Q(n - Q(n-1)) + Q(n - Q(n-2)), for n&#62;2
# This is a "chaotic" integer series with strange
#+ and unpredictable behavior.
# The first 20 terms of the series are:
# 1 1 2 3 3 4 5 5 6 6 6 8 8 8 10 9 10 11 11 12
# See Hofstadter's book, _Goedel, Escher, Bach: An Eternal Golden Braid_,
#+ p. 137, ff.
LIMIT=100 # Number of terms to calculate.
LINEWIDTH=20 # Number of terms printed per line.
Q[1]=1 # First two terms of series are 1.
Q[2]=1
echo
echo "Q-series [$LIMIT terms]:"
echo -n "${Q[1]} " # Output first two terms.
echo -n "${Q[2]} "
for ((n=3; n &#60;= $LIMIT; n++)) # C-like loop expression.
do # Q[n] = Q[n - Q[n-1]] + Q[n - Q[n-2]] for n&#62;2
# Need to break the expression into intermediate terms,
#+ since Bash doesn't handle complex array arithmetic very well.
let "n1 = $n - 1" # n-1
let "n2 = $n - 2" # n-2
t0=`expr $n - ${Q[n1]}` # n - Q[n-1]
t1=`expr $n - ${Q[n2]}` # n - Q[n-2]
T0=${Q[t0]} # Q[n - Q[n-1]]
T1=${Q[t1]} # Q[n - Q[n-2]]
Q[n]=`expr $T0 + $T1` # Q[n - Q[n-1]] + Q[n - Q[n-2]]
echo -n "${Q[n]} "
if [ `expr $n % $LINEWIDTH` -eq 0 ] # Format output.
then # ^ modulo
echo # Break lines into neat chunks.
fi
done
echo
exit 0
# This is an iterative implementation of the Q-series.
# The more intuitive recursive implementation is left as an exercise.
# Warning: calculating this series recursively takes a VERY long time
#+ via a script. C/C++ would be orders of magnitude faster.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>--</P
><P
><A
NAME="ARRAYMULTIDIM"
></A
></P
><P
>Bash supports only one-dimensional arrays, though a little
trickery permits simulating multi-dimensional ones.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="TWODIM"
></A
><P
><B
>Example 27-17. Simulating a two-dimensional array, then tilting it</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# twodim.sh: Simulating a two-dimensional array.
# A one-dimensional array consists of a single row.
# A two-dimensional array stores rows sequentially.
Rows=5
Columns=5
# 5 X 5 Array.
declare -a alpha # char alpha [Rows] [Columns];
# Unnecessary declaration. Why?
load_alpha ()
{
local rc=0
local index
for i in A B C D E F G H I J K L M N O P Q R S T U V W X Y
do # Use different symbols if you like.
local row=`expr $rc / $Columns`
local column=`expr $rc % $Rows`
let "index = $row * $Rows + $column"
alpha[$index]=$i
# alpha[$row][$column]
let "rc += 1"
done
# Simpler would be
#+ declare -a alpha=( A B C D E F G H I J K L M N O P Q R S T U V W X Y )
#+ but this somehow lacks the "flavor" of a two-dimensional array.
}
print_alpha ()
{
local row=0
local index
echo
while [ "$row" -lt "$Rows" ] # Print out in "row major" order:
do #+ columns vary,
#+ while row (outer loop) remains the same.
local column=0
echo -n " " # Lines up "square" array with rotated one.
while [ "$column" -lt "$Columns" ]
do
let "index = $row * $Rows + $column"
echo -n "${alpha[index]} " # alpha[$row][$column]
let "column += 1"
done
let "row += 1"
echo
done
# The simpler equivalent is
# echo ${alpha[*]} | xargs -n $Columns
echo
}
filter () # Filter out negative array indices.
{
echo -n " " # Provides the tilt.
# Explain how.
if [[ "$1" -ge 0 &#38;&#38; "$1" -lt "$Rows" &#38;&#38; "$2" -ge 0 &#38;&#38; "$2" -lt "$Columns" ]]
then
let "index = $1 * $Rows + $2"
# Now, print it rotated.
echo -n " ${alpha[index]}"
# alpha[$row][$column]
fi
}
rotate () # Rotate the array 45 degrees --
{ #+ "balance" it on its lower lefthand corner.
local row
local column
for (( row = Rows; row &#62; -Rows; row-- ))
do # Step through the array backwards. Why?
for (( column = 0; column &#60; Columns; column++ ))
do
if [ "$row" -ge 0 ]
then
let "t1 = $column - $row"
let "t2 = $column"
else
let "t1 = $column"
let "t2 = $column + $row"
fi
filter $t1 $t2 # Filter out negative array indices.
# What happens if you don't do this?
done
echo; echo
done
# Array rotation inspired by examples (pp. 143-146) in
#+ "Advanced C Programming on the IBM PC," by Herbert Mayer
#+ (see bibliography).
# This just goes to show that much of what can be done in C
#+ can also be done in shell scripting.
}
#--------------- Now, let the show begin. ------------#
load_alpha # Load the array.
print_alpha # Print it out.
rotate # Rotate it 45 degrees counterclockwise.
#-----------------------------------------------------#
exit 0
# This is a rather contrived, not to mention inelegant simulation.
# Exercises:
# ---------
# 1) Rewrite the array loading and printing functions
# in a more intuitive and less kludgy fashion.
#
# 2) Figure out how the array rotation functions work.
# Hint: think about the implications of backwards-indexing an array.
#
# 3) Rewrite this script to handle a non-square array,
# such as a 6 X 4 one.
# Try to minimize "distortion" when the array is rotated.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>A two-dimensional array is essentially equivalent to a
one-dimensional one, but with additional addressing modes
for referencing and manipulating the individual elements by
<I
CLASS="FIRSTTERM"
>row</I
> and <I
CLASS="FIRSTTERM"
>column</I
>
position.</P
><P
>For an even more elaborate example of simulating a
two-dimensional array, see <A
HREF="#LIFESLOW"
>Example A-10</A
>.</P
><P
>--</P
><P
>For more interesting scripts using arrays, see:
<P
></P
><UL
><LI
><P
><A
HREF="#AGRAM2"
>Example 12-3</A
></P
></LI
><LI
><P
><A
HREF="#PRIMES2"
>Example 16-46</A
></P
></LI
><LI
><P
><A
HREF="#HASHEX2"
>Example A-22</A
></P
></LI
><LI
><P
><A
HREF="#HOMEWORK"
>Example A-44</A
></P
></LI
><LI
><P
><A
HREF="#QKY"
>Example A-41</A
></P
></LI
><LI
><P
><A
HREF="#NIM"
>Example A-42</A
></P
></LI
></UL
>
</P
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="IVR"
></A
>Chapter 28. Indirect References</H1
><P
><A
NAME="IVRREF"
></A
></P
><P
>We have seen that <A
HREF="#VARSUBN"
>referencing
a variable</A
>, <TT
CLASS="VARNAME"
>$var</TT
>, fetches its
<I
CLASS="FIRSTTERM"
>value</I
>. <A
NAME="EVALINDREF"
></A
>But,
what about the <EM
>value of a value</EM
>? What
about <TT
CLASS="VARNAME"
>$$var</TT
>?</P
><P
>The actual notation is
<TT
CLASS="REPLACEABLE"
><I
>\$$var</I
></TT
>, usually preceded by
an <A
HREF="#EVALREF"
>eval</A
> (and sometimes an
<A
HREF="#ECHOREF"
>echo</A
>). This is called an
<I
CLASS="FIRSTTERM"
>indirect reference</I
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="INDREF"
></A
><P
><B
>Example 28-1. Indirect Variable References</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ind-ref.sh: Indirect variable referencing.
# Accessing the contents of the contents of a variable.
# First, let's fool around a little.
var=23
echo "\$var = $var" # $var = 23
# So far, everything as expected. But ...
echo "\$\$var = $$var" # $$var = 4570var
# Not useful ...
# \$\$ expanded to PID of the script
# -- refer to the entry on the $$ variable --
#+ and "var" is echoed as plain text.
# (Thank you, Jakob Bohm, for pointing this out.)
echo "\\\$\$var = \$$var" # \$$var = $23
# As expected. The first $ is escaped and pasted on to
#+ the value of var ($var = 23 ).
# Meaningful, but still not useful.
# Now, let's start over and do it the right way.
# ============================================== #
a=letter_of_alphabet # Variable "a" holds the name of another variable.
letter_of_alphabet=z
echo
# Direct reference.
echo "a = $a" # a = letter_of_alphabet
# Indirect reference.
eval a=\$$a
# ^^^ Forcing an eval(uation), and ...
# ^ Escaping the first $ ...
# ------------------------------------------------------------------------
# The 'eval' forces an update of $a, sets it to the updated value of \$$a.
# So, we see why 'eval' so often shows up in indirect reference notation.
# ------------------------------------------------------------------------
echo "Now a = $a" # Now a = z
echo
# Now, let's try changing the second-order reference.
t=table_cell_3
table_cell_3=24
echo "\"table_cell_3\" = $table_cell_3" # "table_cell_3" = 24
echo -n "dereferenced \"t\" = "; eval echo \$$t # dereferenced "t" = 24
# In this simple case, the following also works (why?).
# eval t=\$$t; echo "\"t\" = $t"
echo
t=table_cell_3
NEW_VAL=387
table_cell_3=$NEW_VAL
echo "Changing value of \"table_cell_3\" to $NEW_VAL."
echo "\"table_cell_3\" now $table_cell_3"
echo -n "dereferenced \"t\" now "; eval echo \$$t
# "eval" takes the two arguments "echo" and "\$$t" (set equal to $table_cell_3)
echo
# (Thanks, Stephane Chazelas, for clearing up the above behavior.)
# A more straightforward method is the ${!t} notation, discussed in the
#+ "Bash, version 2" section.
# See also ex78.sh.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="IRRREF"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN18998"
></A
><P
></P
><P
>Indirect referencing in Bash
is a multi-step process. First, take the name of a variable:
<TT
CLASS="VARNAME"
>varname</TT
>. Then, reference it:
<TT
CLASS="VARNAME"
>$varname</TT
>. Then, reference the reference:
<TT
CLASS="VARNAME"
>$$varname</TT
>. Then, <I
CLASS="FIRSTTERM"
>escape</I
>
the first <SPAN
CLASS="TOKEN"
>$</SPAN
>: <TT
CLASS="VARNAME"
>\$$varname</TT
>.
Finally, force a reevaluation of the expression and assign it:
<B
CLASS="COMMAND"
>eval newvar=\$$varname</B
>.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
>Of what practical use is indirect referencing of
variables? It gives Bash a little of the functionality
of <A
HREF="#POINTERREF"
>pointers</A
>
in <I
CLASS="FIRSTTERM"
>C</I
>, for instance, in <A
HREF="#RESISTOR"
>table lookup</A
>. And, it also has some
other very interesting applications. . . .</P
><P
>Nils Radtke shows how to build <SPAN
CLASS="QUOTE"
>"dynamic"</SPAN
>
variable names and evaluate their contents. This can be useful
when <A
HREF="#SOURCEREF"
>sourcing</A
> configuration
files.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ---------------------------------------------
# This could be "sourced" from a separate file.
isdnMyProviderRemoteNet=172.16.0.100
isdnYourProviderRemoteNet=10.0.0.10
isdnOnlineService="MyProvider"
# ---------------------------------------------
remoteNet=$(eval "echo \$$(echo isdn${isdnOnlineService}RemoteNet)")
remoteNet=$(eval "echo \$$(echo isdnMyProviderRemoteNet)")
remoteNet=$(eval "echo \$isdnMyProviderRemoteNet")
remoteNet=$(eval "echo $isdnMyProviderRemoteNet")
echo "$remoteNet" # 172.16.0.100
# ================================================================
# And, it gets even better.
# Consider the following snippet given a variable named getSparc,
#+ but no such variable getIa64:
chkMirrorArchs () {
arch="$1";
if [ "$(eval "echo \${$(echo get$(echo -ne $arch |
sed 's/^\(.\).*/\1/g' | tr 'a-z' 'A-Z'; echo $arch |
sed 's/^.\(.*\)/\1/g')):-false}")" = true ]
then
return 0;
else
return 1;
fi;
}
getSparc="true"
unset getIa64
chkMirrorArchs sparc
echo $? # 0
# True
chkMirrorArchs Ia64
echo $? # 1
# False
# Notes:
# -----
# Even the to-be-substituted variable name part is built explicitly.
# The parameters to the chkMirrorArchs calls are all lower case.
# The variable name is composed of two parts: "get" and "Sparc" . . .</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="COLTOTALER2"
></A
><P
><B
>Example 28-2. Passing an indirect reference to <I
CLASS="FIRSTTERM"
>awk</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Another version of the "column totaler" script
#+ that adds up a specified column (of numbers) in the target file.
# This one uses indirect references.
ARGS=2
E_WRONGARGS=85
if [ $# -ne "$ARGS" ] # Check for proper number of command-line args.
then
echo "Usage: `basename $0` filename column-number"
exit $E_WRONGARGS
fi
filename=$1 # Name of file to operate on.
column_number=$2 # Which column to total up.
#===== Same as original script, up to this point =====#
# A multi-line awk script is invoked by
# awk "
# ...
# ...
# ...
# "
# Begin awk script.
# -------------------------------------------------
awk "
{ total += \$${column_number} # Indirect reference
}
END {
print total
}
" "$filename"
# Note that awk doesn't need an eval preceding \$$.
# -------------------------------------------------
# End awk script.
# Indirect variable reference avoids the hassles
#+ of referencing a shell variable within the embedded awk script.
# Thanks, Stephane Chazelas.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This method of indirect referencing is a bit tricky.
If the second order variable changes its value, then the first
order variable must be properly dereferenced (as in the above
example). <A
NAME="IVR2"
></A
>Fortunately, the
<TT
CLASS="REPLACEABLE"
><I
>${!variable}</I
></TT
> notation introduced
with <A
HREF="#BASH2REF"
>version 2</A
> of Bash
(see <A
HREF="#EX78"
>Example 37-2</A
> and <A
HREF="#HASHEX2"
>Example A-22</A
>) makes
indirect referencing more intuitive.</P
></TD
></TR
></TABLE
></DIV
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN19027"
></A
><P
></P
><P
>Bash does not support pointer arithmetic, and this severely
limits the usefulness of indirect referencing. In fact, indirect
referencing in a scripting language is, at best, something of
an afterthought.</P
><P
></P
></DIV
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="DEVPROC"
></A
>Chapter 29. <TT
CLASS="FILENAME"
>/dev</TT
> and <TT
CLASS="FILENAME"
>/proc</TT
></H1
><P
><A
NAME="DEVPROCREF"
></A
></P
><P
>A Linux or UNIX filesystem typically has the
<TT
CLASS="FILENAME"
>/dev</TT
> and
<TT
CLASS="FILENAME"
>/proc</TT
> special-purpose
directories.</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="DEVREF1"
></A
>29.1. <TT
CLASS="FILENAME"
>/dev</TT
></H1
><P
><A
NAME="DEVFILEREF"
></A
>The <TT
CLASS="FILENAME"
>/dev</TT
> directory contains entries for
the <I
CLASS="FIRSTTERM"
>physical devices</I
> that may or may not
be present in the hardware.
<A
NAME="AEN19045"
HREF="#FTN.AEN19045"
><SPAN
CLASS="footnote"
>[118]</SPAN
></A
>
Appropriately enough, these are called <I
CLASS="FIRSTTERM"
>device
files</I
>.
As an example, the hard drive partitions containing
the mounted filesystem(s) have entries in <TT
CLASS="FILENAME"
>/dev</TT
>, as <A
HREF="#DFREF"
>df</A
> shows.</P
><P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>df</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Filesystem 1k-blocks Used Available Use%
Mounted on
/dev/hda6 495876 222748 247527 48% /
/dev/hda1 50755 3887 44248 9% /boot
/dev/hda8 367013 13262 334803 4% /home
/dev/hda5 1714416 1123624 503704 70% /usr</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="LOOPBACKREF"
></A
>Among other things, the <TT
CLASS="FILENAME"
>/dev</TT
> directory
contains <I
CLASS="FIRSTTERM"
>loopback</I
> devices, such as
<TT
CLASS="FILENAME"
>/dev/loop0</TT
>. A loopback device is a gimmick
that allows an ordinary file to be accessed as if it were a
block device.
<A
NAME="AEN19065"
HREF="#FTN.AEN19065"
><SPAN
CLASS="footnote"
>[119]</SPAN
></A
>
This permits mounting an entire filesystem within a
single large file. See <A
HREF="#CREATEFS"
>Example 17-8</A
> and <A
HREF="#ISOMOUNTREF"
>Example 17-7</A
>.</P
><P
>A few of the pseudo-devices in <TT
CLASS="FILENAME"
>/dev</TT
>
have other specialized uses, such as <A
HREF="#ZEROSREF"
><TT
CLASS="FILENAME"
>/dev/null</TT
></A
>, <A
HREF="#ZEROSREF1"
><TT
CLASS="FILENAME"
>/dev/zero</TT
></A
>, <A
HREF="#URANDOMREF"
><TT
CLASS="FILENAME"
>/dev/urandom</TT
></A
>,
<TT
CLASS="FILENAME"
>/dev/sda1</TT
> (hard drive partition),
<TT
CLASS="FILENAME"
>/dev/udp</TT
> (<I
CLASS="FIRSTTERM"
>User
Datagram Packet</I
> port), and <A
HREF="#DEVTCP"
><TT
CLASS="FILENAME"
>/dev/tcp</TT
></A
>.</P
><P
>For instance:</P
><P
>To manually <A
HREF="#MOUNTREF"
>mount</A
>
a USB flash drive, append the following line to
<A
HREF="#FSTABREF"
><TT
CLASS="FILENAME"
>/etc/fstab</TT
></A
>.
<A
NAME="AEN19093"
HREF="#FTN.AEN19093"
><SPAN
CLASS="footnote"
>[120]</SPAN
></A
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>/dev/sda1 /mnt/flashdrive auto noauto,user,noatime 0 0</PRE
></FONT
></TD
></TR
></TABLE
>
(See also <A
HREF="#USBINST"
>Example A-23</A
>.)</P
><P
>Checking whether a disk is in the CD-burner
(soft-linked to <TT
CLASS="FILENAME"
>/dev/hdc</TT
>):
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>head -1 /dev/hdc
# head: cannot open '/dev/hdc' for reading: No medium found
# (No disc in the drive.)
# head: error reading '/dev/hdc': Input/output error
# (There is a disk in the drive, but it can't be read;
#+ possibly it's an unrecorded CDR blank.)
# Stream of characters and assorted gibberish
# (There is a pre-recorded disk in the drive,
#+ and this is raw output -- a stream of ASCII and binary data.)
# Here we see the wisdom of using 'head' to limit the output
#+ to manageable proportions, rather than 'cat' or something similar.
# Now, it's just a matter of checking/parsing the output and taking
#+ appropriate action.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="SOCKETREF"
></A
></P
><P
>When executing a command on a
<TT
CLASS="FILENAME"
>/dev/tcp/$host/$port</TT
> pseudo-device file, Bash
opens a TCP connection to the associated
<I
CLASS="FIRSTTERM"
>socket</I
>.</P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN19112"
></A
><P
></P
><P
>A <I
CLASS="FIRSTTERM"
>socket</I
> is a
communications node associated with a specific I/O
port. (This is analogous to a <I
CLASS="FIRSTTERM"
>hardware
socket</I
>, or <I
CLASS="FIRSTTERM"
>receptacle</I
>,
for a connecting cable.) It permits data transfer between
hardware devices on the same machine, between machines
on the same network, between machines across different
networks, and, of course, between machines at different
locations on the Internet.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
><A
NAME="NPREF"
></A
>The following examples assume an active Internet
connection.</P
><P
>Getting the time from <TT
CLASS="FILENAME"
>nist.gov</TT
>:</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat &#60;/dev/tcp/time.nist.gov/13</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>53082 04-03-18 04:26:54 68 0 0 502.3 UTC(NIST) *</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
><P
>[Mark contributed this example.]</P
><P
>Generalizing the above into a script:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# This script must run with root permissions.
URL="time.nist.gov/13"
Time=$(cat &#60;/dev/tcp/"$URL")
UTC=$(echo "$Time" | awk '{print$3}') # Third field is UTC (GMT) time.
# Exercise: modify this for different time zones.
echo "UTC Time = "$UTC""</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="NW001"
></A
>Downloading a URL:</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>exec 5&#60;&#62;/dev/tcp/www.net.cn/80</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo -e "GET / HTTP/1.0\n" &#62;&#38;5</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat &#60;&#38;5</B
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
><P
>[Thanks, Mark and Mihai Maties.]</P
><DIV
CLASS="EXAMPLE"
><A
NAME="DEVTCP"
></A
><P
><B
>Example 29-1. Using <TT
CLASS="FILENAME"
>/dev/tcp</TT
> for
troubleshooting</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# dev-tcp.sh: /dev/tcp redirection to check Internet connection.
# Script by Troy Engel.
# Used with permission.
TCP_HOST=news-15.net # A known spam-friendly ISP.
TCP_PORT=80 # Port 80 is http.
# Try to connect. (Somewhat similar to a 'ping' . . .)
echo "HEAD / HTTP/1.0" &#62;/dev/tcp/${TCP_HOST}/${TCP_PORT}
MYEXIT=$?
: &#60;&#60;EXPLANATION
If bash was compiled with --enable-net-redirections, it has the capability of
using a special character device for both TCP and UDP redirections. These
redirections are used identically as STDIN/STDOUT/STDERR. The device entries
are 30,36 for /dev/tcp:
mknod /dev/tcp c 30 36
&#62;From the bash reference:
/dev/tcp/host/port
If host is a valid hostname or Internet address, and port is an integer
port number or service name, Bash attempts to open a TCP connection to the
corresponding socket.
EXPLANATION
if [ "X$MYEXIT" = "X0" ]; then
echo "Connection successful. Exit code: $MYEXIT"
else
echo "Connection unsuccessful. Exit code: $MYEXIT"
fi
exit $MYEXIT</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="MUSICSCR"
></A
><P
><B
>Example 29-2. Playing music</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# music.sh
# Music without external files
# Author: Antonio Macchi
# Used in ABS Guide with permission.
# /dev/dsp default = 8000 frames per second, 8 bits per frame (1 byte),
#+ 1 channel (mono)
duration=2000 # If 8000 bytes = 1 second, then 2000 = 1/4 second.
volume=$'\xc0' # Max volume = \xff (or \x00).
mute=$'\x80' # No volume = \x80 (the middle).
function mknote () # $1=Note Hz in bytes (e.g. A = 440Hz ::
{ #+ 8000 fps / 440 = 16 :: A = 16 bytes per second)
for t in `seq 0 $duration`
do
test $(( $t % $1 )) = 0 &#38;&#38; echo -n $volume || echo -n $mute
done
}
e=`mknote 49`
g=`mknote 41`
a=`mknote 36`
b=`mknote 32`
c=`mknote 30`
cis=`mknote 29`
d=`mknote 27`
e2=`mknote 24`
n=`mknote 32767`
# European notation.
echo -n "$g$e2$d$c$d$c$a$g$n$g$e$n$g$e2$d$c$c$b$c$cis$n$cis$d \
$n$g$e2$d$c$d$c$a$g$n$g$e$n$g$a$d$c$b$a$b$c" &#62; /dev/dsp
# dsp = Digital Signal Processor
exit # A "bonny" example of an elegant shell script!</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="PROCREF1"
></A
>29.2. <TT
CLASS="FILENAME"
>/proc</TT
></H1
><P
><A
NAME="PROCREF2"
></A
></P
><P
>The <TT
CLASS="FILENAME"
>/proc</TT
> directory
is actually a pseudo-filesystem. The files in <TT
CLASS="FILENAME"
>/proc</TT
> mirror currently running
system and kernel <A
HREF="#PROCESSREF"
>processes</A
>
and contain information and statistics about them.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat /proc/devices</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Character devices:
1 mem
2 pty
3 ttyp
4 ttyS
5 cua
7 vcs
10 misc
14 sound
29 fb
36 netlink
128 ptm
136 pts
162 raw
254 pcmcia
Block devices:
1 ramdisk
2 fd
3 ide0
9 md</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat /proc/interrupts</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> CPU0
0: 84505 XT-PIC timer
1: 3375 XT-PIC keyboard
2: 0 XT-PIC cascade
5: 1 XT-PIC soundblaster
8: 1 XT-PIC rtc
12: 4231 XT-PIC PS/2 Mouse
14: 109373 XT-PIC ide0
NMI: 0
ERR: 0</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat /proc/partitions</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>major minor #blocks name rio rmerge rsect ruse wio wmerge wsect wuse running use aveq
3 0 3007872 hda 4472 22260 114520 94240 3551 18703 50384 549710 0 111550 644030
3 1 52416 hda1 27 395 844 960 4 2 14 180 0 800 1140
3 2 1 hda2 0 0 0 0 0 0 0 0 0 0 0
3 4 165280 hda4 10 0 20 210 0 0 0 0 0 210 210
...</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat /proc/loadavg</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>0.13 0.42 0.27 2/44 1119</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat /proc/apm</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>1.16 1.2 0x03 0x01 0xff 0x80 -1% -1 ?</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat /proc/acpi/battery/BAT0/info</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>present: yes
design capacity: 43200 mWh
last full capacity: 36640 mWh
battery technology: rechargeable
design voltage: 10800 mV
design capacity warning: 1832 mWh
design capacity low: 200 mWh
capacity granularity 1: 1 mWh
capacity granularity 2: 1 mWh
model number: IBM-02K6897
serial number: 1133
battery type: LION
OEM info: Panasonic</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>fgrep Mem /proc/meminfo</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>MemTotal: 515216 kB
MemFree: 266248 kB</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Shell scripts may extract data from certain of the files in
<TT
CLASS="FILENAME"
>/proc</TT
>.
<A
NAME="AEN19180"
HREF="#FTN.AEN19180"
><SPAN
CLASS="footnote"
>[121]</SPAN
></A
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>FS=iso # ISO filesystem support in kernel?
grep $FS /proc/filesystems # iso9660</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>kernel_version=$( awk '{ print $3 }' /proc/version )</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>CPU=$( awk '/model name/ {print $5}' &#60; /proc/cpuinfo )
if [ "$CPU" = "Pentium(R)" ]
then
run_some_commands
...
else
run_other_commands
...
fi
cpu_speed=$( fgrep "cpu MHz" /proc/cpuinfo | awk '{print $4}' )
# Current operating speed (in MHz) of the cpu on your machine.
# On a laptop this may vary, depending on use of battery
#+ or AC power.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# get-commandline.sh
# Get the command-line parameters of a process.
OPTION=cmdline
# Identify PID.
pid=$( echo $(pidof "$1") | awk '{ print $1 }' )
# Get only first ^^^^^^^^^^^^^^^^^^ of multiple instances.
echo
echo "Process ID of (first instance of) "$1" = $pid"
echo -n "Command-line arguments: "
cat /proc/"$pid"/"$OPTION" | xargs -0 echo
# Formats output: ^^^^^^^^^^^^^^^
# (Thanks, Han Holl, for the fixup!)
echo; echo
# For example:
# sh get-commandline.sh xterm</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>+</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>devfile="/proc/bus/usb/devices"
text="Spd"
USB1="Spd=12"
USB2="Spd=480"
bus_speed=$(fgrep -m 1 "$text" $devfile | awk '{print $9}')
# ^^^^ Stop after first match.
if [ "$bus_speed" = "$USB1" ]
then
echo "USB 1.1 port found."
# Do something appropriate for USB 1.1.
fi</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>It is even possible to control certain peripherals with commands
sent to the <TT
CLASS="FILENAME"
>/proc</TT
> directory.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
> <TT
CLASS="PROMPT"
>root# </TT
><B
CLASS="COMMAND"
>echo on &#62; /proc/acpi/ibm/light</B
>
</PRE
></FONT
></TD
></TR
></TABLE
>
This turns on the <EM
>Thinklight</EM
> in certain models
of IBM/Lenovo Thinkpads. (May not work on all Linux distros.)</P
><P
>Of course, caution is advised when writing to <TT
CLASS="FILENAME"
>/proc</TT
>.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="PROCRUNNING"
></A
></P
><P
>The <TT
CLASS="FILENAME"
>/proc</TT
> directory
contains subdirectories with unusual numerical
names. Every one of these names maps to the <A
HREF="#PPIDREF"
>process ID</A
> of a currently running
process. Within each of these subdirectories, there are
a number of files that hold useful information about the
corresponding process. The <TT
CLASS="FILENAME"
>stat</TT
> and
<TT
CLASS="FILENAME"
>status</TT
> files keep running statistics
on the process, the <TT
CLASS="FILENAME"
>cmdline</TT
> file holds
the command-line arguments the process was invoked with, and
the <TT
CLASS="FILENAME"
>exe</TT
> file is a symbolic link to the
complete path name of the invoking process. There are a few
more such files, but these seem to be the most interesting
from a scripting standpoint.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="PIDID"
></A
><P
><B
>Example 29-3. Finding the process associated with a PID</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# pid-identifier.sh:
# Gives complete path name to process associated with pid.
ARGNO=1 # Number of arguments the script expects.
E_WRONGARGS=65
E_BADPID=66
E_NOSUCHPROCESS=67
E_NOPERMISSION=68
PROCFILE=exe
if [ $# -ne $ARGNO ]
then
echo "Usage: `basename $0` PID-number" &#62;&#38;2 # Error message &#62;stderr.
exit $E_WRONGARGS
fi
pidno=$( ps ax | grep $1 | awk '{ print $1 }' | grep $1 )
# Checks for pid in "ps" listing, field #1.
# Then makes sure it is the actual process, not the process invoked by this script.
# The last "grep $1" filters out this possibility.
#
# pidno=$( ps ax | awk '{ print $1 }' | grep $1 )
# also works, as Teemu Huovila, points out.
if [ -z "$pidno" ] # If, after all the filtering, the result is a zero-length string,
then #+ no running process corresponds to the pid given.
echo "No such process running."
exit $E_NOSUCHPROCESS
fi
# Alternatively:
# if ! ps $1 &#62; /dev/null 2&#62;&#38;1
# then # no running process corresponds to the pid given.
# echo "No such process running."
# exit $E_NOSUCHPROCESS
# fi
# To simplify the entire process, use "pidof".
if [ ! -r "/proc/$1/$PROCFILE" ] # Check for read permission.
then
echo "Process $1 running, but..."
echo "Can't get read permission on /proc/$1/$PROCFILE."
exit $E_NOPERMISSION # Ordinary user can't access some files in /proc.
fi
# The last two tests may be replaced by:
# if ! kill -0 $1 &#62; /dev/null 2&#62;&#38;1 # '0' is not a signal, but
# this will test whether it is possible
# to send a signal to the process.
# then echo "PID doesn't exist or you're not its owner" &#62;&#38;2
# exit $E_BADPID
# fi
exe_file=$( ls -l /proc/$1 | grep "exe" | awk '{ print $11 }' )
# Or exe_file=$( ls -l /proc/$1/exe | awk '{print $11}' )
#
# /proc/pid-number/exe is a symbolic link
#+ to the complete path name of the invoking process.
if [ -e "$exe_file" ] # If /proc/pid-number/exe exists,
then #+ then the corresponding process exists.
echo "Process #$1 invoked by $exe_file."
else
echo "No such process running."
fi
# This elaborate script can *almost* be replaced by
# ps ax | grep $1 | awk '{ print $5 }'
# However, this will not work...
#+ because the fifth field of 'ps' is argv[0] of the process,
#+ not the executable file path.
#
# However, either of the following would work.
# find /proc/$1/exe -printf '%l\n'
# lsof -aFn -p $1 -d txt | sed -ne 's/^n//p'
# Additional commentary by Stephane Chazelas.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="CONSTAT"
></A
><P
><B
>Example 29-4. On-line connect status</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# connect-stat.sh
# Note that this script may need modification
#+ to work with a wireless connection.
PROCNAME=pppd # ppp daemon
PROCFILENAME=status # Where to look.
NOTCONNECTED=85
INTERVAL=2 # Update every 2 seconds.
pidno=$( ps ax | grep -v "ps ax" | grep -v grep | grep $PROCNAME |
awk '{ print $1 }' )
# Finding the process number of 'pppd', the 'ppp daemon'.
# Have to filter out the process lines generated by the search itself.
#
# However, as Oleg Philon points out,
#+ this could have been considerably simplified by using "pidof".
# pidno=$( pidof $PROCNAME )
#
# Moral of the story:
#+ When a command sequence gets too complex, look for a shortcut.
if [ -z "$pidno" ] # If no pid, then process is not running.
then
echo "Not connected."
# exit $NOTCONNECTED
else
echo "Connected."; echo
fi
while [ true ] # Endless loop, script can be improved here.
do
if [ ! -e "/proc/$pidno/$PROCFILENAME" ]
# While process running, then "status" file exists.
then
echo "Disconnected."
# exit $NOTCONNECTED
fi
netstat -s | grep "packets received" # Get some connect statistics.
netstat -s | grep "packets delivered"
sleep $INTERVAL
echo; echo
done
exit 0
# As it stands, this script must be terminated with a Control-C.
# Exercises:
# ---------
# Improve the script so it exits on a "q" keystroke.
# Make the script more user-friendly in other ways.
# Fix the script to work with wireless/DSL connections.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="PROCWARNING"
></A
></P
><DIV
CLASS="WARNING"
><P
></P
><TABLE
CLASS="WARNING"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/warning.gif"
HSPACE="5"
ALT="Warning"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>In general, it is dangerous to
<EM
>write</EM
> to the files in <TT
CLASS="FILENAME"
>/proc</TT
>, as this can corrupt the
filesystem or crash the machine.</P
></TD
></TR
></TABLE
></DIV
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="NETWORKPROGRAMMING"
></A
>Chapter 30. Network Programming</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>The Net's a cross between an elephant and a white
elephant sale: it never forgets, and it's always crap.</I
></P
><P
><I
>--Nemo</I
></P
></I
></TD
></TR
></TABLE
><P
>A Linux system has quite a number of tools for accessing,
manipulating, and troubleshooting network connections. We can
incorporate some of these tools into scripts -- scripts that
expand our knowledge of networking, useful scripts that can
facilitate the administration of a network.</P
><P
><A
NAME="CGISCRIPT"
></A
>Here is a simple CGI
script that demonstrates connecting to a remote server.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="TESTCGI"
></A
><P
><B
>Example 30-1. Print the server environment</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# test-cgi.sh
# by Michael Zick
# Used with permission
# May have to change the location for your site.
# (At the ISP's servers, Bash may not be in the usual place.)
# Other places: /usr/bin or /usr/local/bin
# Might even try it without any path in sha-bang.
# Disable filename globbing.
set -f
# Header tells browser what to expect.
echo Content-type: text/plain
echo
echo CGI/1.0 test script report:
echo
echo environment settings:
set
echo
echo whereis bash?
whereis bash
echo
echo who are we?
echo ${BASH_VERSINFO[*]}
echo
echo argc is $#. argv is "$*".
echo
# CGI/1.0 expected environment variables.
echo SERVER_SOFTWARE = $SERVER_SOFTWARE
echo SERVER_NAME = $SERVER_NAME
echo GATEWAY_INTERFACE = $GATEWAY_INTERFACE
echo SERVER_PROTOCOL = $SERVER_PROTOCOL
echo SERVER_PORT = $SERVER_PORT
echo REQUEST_METHOD = $REQUEST_METHOD
echo HTTP_ACCEPT = "$HTTP_ACCEPT"
echo PATH_INFO = "$PATH_INFO"
echo PATH_TRANSLATED = "$PATH_TRANSLATED"
echo SCRIPT_NAME = "$SCRIPT_NAME"
echo QUERY_STRING = "$QUERY_STRING"
echo REMOTE_HOST = $REMOTE_HOST
echo REMOTE_ADDR = $REMOTE_ADDR
echo REMOTE_USER = $REMOTE_USER
echo AUTH_TYPE = $AUTH_TYPE
echo CONTENT_TYPE = $CONTENT_TYPE
echo CONTENT_LENGTH = $CONTENT_LENGTH
exit 0
# Here document to give short instructions.
:&#60;&#60;-'_test_CGI_'
1) Drop this in your http://domain.name/cgi-bin directory.
2) Then, open http://domain.name/cgi-bin/test-cgi.sh.
_test_CGI_</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>For security purposes, it may be helpful to identify the IP
addresses a computer is accessing.</P
><P
><A
NAME="IPTABLES02"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="IPADDRESSES"
></A
><P
><B
>Example 30-2. IP addresses</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ip-addresses.sh
# List the IP addresses your computer is connected to.
# Inspired by Greg Bledsoe's ddos.sh script,
# Linux Journal, 09 March 2011.
# URL:
# http://www.linuxjournal.com/content/back-dead-simple-bash-complex-ddos
# Greg licensed his script under the GPL2,
#+ and as a derivative, this script is likewise GPL2.
connection_type=TCP # Also try UDP.
field=2 # Which field of the output we're interested in.
no_match=LISTEN # Filter out records containing this. Why?
lsof_args=-ni # -i lists Internet-associated files.
# -n preserves numerical IP addresses.
# What happens without the -n option? Try it.
router="[0-9][0-9][0-9][0-9][0-9]-&#62;"
# Delete the router info.
lsof "$lsof_args" | grep $connection_type | grep -v "$no_match" |
awk '{print $9}' | cut -d : -f $field | sort | uniq |
sed s/"^$router"//
# Bledsoe's script assigns the output of a filtered IP list,
# (similar to lines 19-22, above) to a variable.
# He checks for multiple connections to a single IP address,
# then uses:
#
# iptables -I INPUT -s $ip -p tcp -j REJECT --reject-with tcp-reset
#
# ... within a 60-second delay loop to bounce packets from DDOS attacks.
# Exercise:
# --------
# Use the 'iptables' command to extend this script
#+ to reject connection attempts from well-known spammer IP domains.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>More examples of network programming:
<P
></P
><OL
TYPE="1"
><LI
><P
><A
HREF="#NPREF"
>Getting the time from
<I
CLASS="FIRSTTERM"
>nist.gov</I
></A
></P
></LI
><LI
><P
><A
HREF="#NW001"
>Downloading a
URL</A
></P
></LI
><LI
><P
><A
HREF="#IPSCRIPT0"
>A GRE
tunnel</A
></P
></LI
><LI
><P
><A
HREF="#PING0"
>Checking
if an Internet server is up</A
></P
></LI
><LI
><P
><A
HREF="#ISSPAMMER"
>Example 16-41</A
></P
></LI
><LI
><P
><A
HREF="#ISSPAMMER2"
>Example A-28</A
></P
></LI
><LI
><P
><A
HREF="#WHX"
>Example A-29</A
></P
></LI
><LI
><P
><A
HREF="#DEVTCP"
>Example 29-1</A
></P
></LI
></OL
>
</P
><P
>See also the <A
HREF="#NETWORKSYS1"
>networking commands</A
>
in the <A
HREF="#SYSTEM"
>System and
Administrative Commands</A
> chapter and the <A
HREF="#COMMUNICATIONS"
>communications commands</A
> in
the <A
HREF="#EXTERNAL"
>External Filters, Programs and
Commands</A
> chapter.</P
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="ZEROS"
></A
>Chapter 31. Of Zeros and Nulls</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>Faultily faultless, icily regular, splendidly null</I
></P
><P
><I
>Dead perfection; no more.</I
></P
><P
><I
>--Alfred Lord Tennyson</I
></P
></I
></TD
></TR
></TABLE
><P
><A
NAME="ZEROSREF"
></A
></P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="ZERONULL1"
></A
><TT
CLASS="FILENAME"
>/dev/zero</TT
>
... <TT
CLASS="FILENAME"
>/dev/null</TT
></B
></P
><DL
><DT
><A
NAME="DEVNULLREF"
></A
>Uses of
<TT
CLASS="FILENAME"
>/dev/null</TT
></DT
><DD
><P
>Think of <TT
CLASS="FILENAME"
>/dev/null</TT
> as a <I
CLASS="FIRSTTERM"
>black
hole</I
>. It is essentially the equivalent of
a write-only file. Everything written to it disappears.
Attempts to read or output from it result in nothing. All
the same, <TT
CLASS="FILENAME"
>/dev/null</TT
> can be quite
useful from both the command-line and in scripts.</P
><P
>Suppressing <TT
CLASS="FILENAME"
>stdout</TT
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat $filename &#62;/dev/null
# Contents of the file will not list to stdout.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Suppressing <TT
CLASS="FILENAME"
>stderr</TT
>
(from <A
HREF="#EX57"
>Example 16-3</A
>).
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>rm $badname 2&#62;/dev/null
# So error messages [stderr] deep-sixed.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Suppressing output from <EM
>both</EM
>
<TT
CLASS="FILENAME"
>stdout</TT
> and <TT
CLASS="FILENAME"
>stderr</TT
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat $filename 2&#62;/dev/null &#62;/dev/null
# If "$filename" does not exist, there will be no error message output.
# If "$filename" does exist, the contents of the file will not list to stdout.
# Therefore, no output at all will result from the above line of code.
#
# This can be useful in situations where the return code from a command
#+ needs to be tested, but no output is desired.
#
# cat $filename &#38;&#62;/dev/null
# also works, as Baris Cicek points out.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Deleting contents of a file, but preserving the file itself, with
all attendant permissions (from <A
HREF="#EX1"
>Example 2-1</A
> and <A
HREF="#EX2"
>Example 2-3</A
>):
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat /dev/null &#62; /var/log/messages
# : &#62; /var/log/messages has same effect, but does not spawn a new process.
cat /dev/null &#62; /var/log/wtmp</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Automatically emptying the contents of a logfile
(especially good for dealing with those nasty
<SPAN
CLASS="QUOTE"
>"cookies"</SPAN
> sent by commercial Web sites):</P
><DIV
CLASS="EXAMPLE"
><A
NAME="COOKIES"
></A
><P
><B
>Example 31-1. Hiding the cookie jar</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Obsolete Netscape browser.
# Same principle applies to newer browsers.
if [ -f ~/.netscape/cookies ] # Remove, if exists.
then
rm -f ~/.netscape/cookies
fi
ln -s /dev/null ~/.netscape/cookies
# All cookies now get sent to a black hole, rather than saved to disk.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="ZEROSREF1"
></A
>Uses of <TT
CLASS="FILENAME"
>/dev/zero</TT
></DT
><DD
><P
>Like <TT
CLASS="FILENAME"
>/dev/null</TT
>,
<TT
CLASS="FILENAME"
>/dev/zero</TT
> is a pseudo-device file, but
it actually produces a stream of nulls
(<EM
>binary</EM
> zeros, not the <A
HREF="#ASCIIDEF"
>ASCII</A
> kind). Output written
to <TT
CLASS="FILENAME"
>/dev/zero</TT
> disappears, and it is
fairly difficult to actually read the nulls emitted there,
though it can be done with <A
HREF="#ODREF"
>od</A
>
or a hex editor. <A
NAME="SWAPFILEREF"
></A
>The chief use of
<TT
CLASS="FILENAME"
>/dev/zero</TT
> is creating an initialized
dummy file of predetermined length intended as a temporary
swap file.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX73"
></A
><P
><B
>Example 31-2. Setting up a swapfile using <TT
CLASS="FILENAME"
>/dev/zero</TT
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Creating a swap file.
# A swap file provides a temporary storage cache
#+ which helps speed up certain filesystem operations.
ROOT_UID=0 # Root has $UID 0.
E_WRONG_USER=85 # Not root?
FILE=/swap
BLOCKSIZE=1024
MINBLOCKS=40
SUCCESS=0
# This script must be run as root.
if [ "$UID" -ne "$ROOT_UID" ]
then
echo; echo "You must be root to run this script."; echo
exit $E_WRONG_USER
fi
blocks=${1:-$MINBLOCKS} # Set to default of 40 blocks,
#+ if nothing specified on command-line.
# This is the equivalent of the command block below.
# --------------------------------------------------
# if [ -n "$1" ]
# then
# blocks=$1
# else
# blocks=$MINBLOCKS
# fi
# --------------------------------------------------
if [ "$blocks" -lt $MINBLOCKS ]
then
blocks=$MINBLOCKS # Must be at least 40 blocks long.
fi
######################################################################
echo "Creating swap file of size $blocks blocks (KB)."
dd if=/dev/zero of=$FILE bs=$BLOCKSIZE count=$blocks # Zero out file.
mkswap $FILE $blocks # Designate it a swap file.
swapon $FILE # Activate swap file.
retcode=$? # Everything worked?
# Note that if one or more of these commands fails,
#+ then it could cause nasty problems.
######################################################################
# Exercise:
# Rewrite the above block of code so that if it does not execute
#+ successfully, then:
# 1) an error message is echoed to stderr,
# 2) all temporary files are cleaned up, and
# 3) the script exits in an orderly fashion with an
#+ appropriate error code.
echo "Swap file created and activated."
exit $retcode</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Another application of <TT
CLASS="FILENAME"
>/dev/zero</TT
>
is to <SPAN
CLASS="QUOTE"
>"zero out"</SPAN
> a file of a designated
size for a special purpose, such as mounting a filesystem
on a <A
HREF="#LOOPBACKREF"
>loopback device</A
>
(see <A
HREF="#CREATEFS"
>Example 17-8</A
>) or <SPAN
CLASS="QUOTE"
>"securely"</SPAN
>
deleting a file (see <A
HREF="#BLOTOUT"
>Example 16-61</A
>).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="RAMDISK"
></A
><P
><B
>Example 31-3. Creating a ramdisk</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ramdisk.sh
# A "ramdisk" is a segment of system RAM memory
#+ which acts as if it were a filesystem.
# Its advantage is very fast access (read/write time).
# Disadvantages: volatility, loss of data on reboot or powerdown,
#+ less RAM available to system.
#
# Of what use is a ramdisk?
# Keeping a large dataset, such as a table or dictionary on ramdisk,
#+ speeds up data lookup, since memory access is much faster than disk access.
E_NON_ROOT_USER=70 # Must run as root.
ROOTUSER_NAME=root
MOUNTPT=/mnt/ramdisk # Create with mkdir /mnt/ramdisk.
SIZE=2000 # 2K blocks (change as appropriate)
BLOCKSIZE=1024 # 1K (1024 byte) block size
DEVICE=/dev/ram0 # First ram device
username=`id -nu`
if [ "$username" != "$ROOTUSER_NAME" ]
then
echo "Must be root to run \"`basename $0`\"."
exit $E_NON_ROOT_USER
fi
if [ ! -d "$MOUNTPT" ] # Test whether mount point already there,
then #+ so no error if this script is run
mkdir $MOUNTPT #+ multiple times.
fi
##############################################################################
dd if=/dev/zero of=$DEVICE count=$SIZE bs=$BLOCKSIZE # Zero out RAM device.
# Why is this necessary?
mke2fs $DEVICE # Create an ext2 filesystem on it.
mount $DEVICE $MOUNTPT # Mount it.
chmod 777 $MOUNTPT # Enables ordinary user to access ramdisk.
# However, must be root to unmount it.
##############################################################################
# Need to test whether above commands succeed. Could cause problems otherwise.
# Exercise: modify this script to make it safer.
echo "\"$MOUNTPT\" now available for use."
# The ramdisk is now accessible for storing files, even by an ordinary user.
# Caution, the ramdisk is volatile, and its contents will disappear
#+ on reboot or power loss.
# Copy anything you want saved to a regular directory.
# After reboot, run this script to again set up ramdisk.
# Remounting /mnt/ramdisk without the other steps will not work.
# Suitably modified, this script can by invoked in /etc/rc.d/rc.local,
#+ to set up ramdisk automatically at bootup.
# That may be appropriate on, for example, a database server.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>In addition to all the above,
<TT
CLASS="FILENAME"
>/dev/zero</TT
> is needed by ELF
(<I
CLASS="FIRSTTERM"
>Executable and Linking Format</I
>)
UNIX/Linux binaries.</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="DEBUGGING"
></A
>Chapter 32. Debugging</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible,
you are, by definition, not smart enough to debug it.</I
></P
><P
><I
>--Brian Kernighan</I
></P
></I
></TD
></TR
></TABLE
><P
>The Bash shell contains no built-in debugger, and only bare-bones
debugging-specific commands and constructs. Syntax errors or
outright typos in the script generate cryptic error messages that
are often of no help in debugging a non-functional script.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX74"
></A
><P
><B
>Example 32-1. A buggy script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ex74.sh
# This is a buggy script.
# Where, oh where is the error?
a=37
if [$a -gt 27 ]
then
echo $a
fi
exit $? # 0! Why?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Output from script:
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="COMPUTEROUTPUT"
>./ex74.sh: [37: command not found</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
What's wrong with the above script? Hint: after the
<I
CLASS="FIRSTTERM"
>if</I
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="MISSINGKEYWORD"
></A
><P
><B
>Example 32-2. Missing <A
HREF="#KEYWORDREF"
>keyword</A
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# missing-keyword.sh
# What error message will this script generate? And why?
for a in 1 2 3
do
echo "$a"
# done # Required keyword 'done' commented out in line 8.
exit 0 # Will not exit here!
# === #
# From command line, after script terminates:
echo $? # 2</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Output from script:
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="COMPUTEROUTPUT"
>missing-keyword.sh: line 10: syntax error: unexpected end of file</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
Note that the error message does <EM
>not</EM
> necessarily
reference the line in which the error occurs, but the line where the
Bash interpreter finally becomes aware of the error.
</P
><P
>Error messages may disregard comment lines in a script when
reporting the line number of a syntax error.</P
><P
>What if the script executes, but does not work as expected? This is the
all too familiar logic error.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX75"
></A
><P
><B
>Example 32-3. <I
CLASS="FIRSTTERM"
>test24</I
>: another buggy script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# This script is supposed to delete all filenames in current directory
#+ containing embedded spaces.
# It doesn't work.
# Why not?
badname=`ls | grep ' '`
# Try this:
# echo "$badname"
rm "$badname"
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Try to find out what's wrong with <A
HREF="#EX75"
>Example 32-3</A
>
by uncommenting the <TT
CLASS="USERINPUT"
><B
>echo "$badname"</B
></TT
> line. Echo
statements are useful for seeing whether what you expect is
actually what you get.</P
><P
>In this particular case, <TT
CLASS="USERINPUT"
><B
>rm "$badname"</B
></TT
>
will not give the desired results because
<TT
CLASS="VARNAME"
>$badname</TT
> should not be quoted. Placing it
in quotes ensures that <B
CLASS="COMMAND"
>rm</B
> has only one
argument (it will match only one filename). A partial fix
is to remove to quotes from <TT
CLASS="VARNAME"
>$badname</TT
> and
to reset <TT
CLASS="VARNAME"
>$IFS</TT
> to contain only a newline,
<TT
CLASS="USERINPUT"
><B
>IFS=$'\n'</B
></TT
>. However, there are simpler
ways of going about it.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Correct methods of deleting filenames containing spaces.
rm *\ *
rm *" "*
rm *' '*
# Thank you. S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Summarizing the symptoms of a buggy script,
<P
></P
><OL
TYPE="1"
><LI
><P
>It bombs with a <SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="ERRORNAME"
>syntax error</SPAN
>"</SPAN
> message, or</P
></LI
><LI
><P
>It runs, but does not work as expected
(<SPAN
CLASS="ERRORNAME"
>logic error</SPAN
>).</P
></LI
><LI
><P
>It runs, works as expected, but has nasty side effects
(<SPAN
CLASS="ERRORNAME"
>logic bomb</SPAN
>).</P
></LI
></OL
>
</P
><P
><A
NAME="DEBUGTOOLS"
></A
></P
><P
>Tools for debugging non-working scripts include
<P
></P
><OL
TYPE="1"
><LI
><P
>Inserting <A
HREF="#ECHOREF"
>echo</A
>
statements at critical points in the script to trace the
variables, and otherwise give a snapshot of what is going
on.</P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Even better is an <B
CLASS="COMMAND"
>echo</B
> that echoes
only when <I
CLASS="FIRSTTERM"
>debug</I
> is on.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>### debecho (debug-echo), by Stefano Falsetto ###
### Will echo passed parameters only if DEBUG is set to a value. ###
debecho () {
if [ ! -z "$DEBUG" ]; then
echo "$1" &#62;&#38;2
# ^^^ to stderr
fi
}
DEBUG=on
Whatever=whatnot
debecho $Whatever # whatnot
DEBUG=
Whatever=notwhat
debecho $Whatever # (Will not echo.)</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
>Using the <A
HREF="#TEEREF"
>tee</A
> filter
to check processes or data flows at critical points.</P
></LI
><LI
><P
>Setting option flags <TT
CLASS="OPTION"
>-n -v -x</TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>sh -n scriptname</B
></TT
> checks for
syntax errors without actually running the script. This is
the equivalent of inserting <TT
CLASS="USERINPUT"
><B
>set -n</B
></TT
> or
<TT
CLASS="USERINPUT"
><B
>set -o noexec</B
></TT
> into the script. Note
that certain types of syntax errors can slip past this
check.</P
><P
><TT
CLASS="USERINPUT"
><B
>sh -v scriptname</B
></TT
> echoes each
command before executing it. This is the equivalent of
inserting <TT
CLASS="USERINPUT"
><B
>set -v</B
></TT
> or <TT
CLASS="USERINPUT"
><B
>set
-o verbose</B
></TT
> in the script.</P
><P
>The <TT
CLASS="OPTION"
>-n</TT
> and <TT
CLASS="OPTION"
>-v</TT
>
flags work well together. <TT
CLASS="USERINPUT"
><B
>sh -nv
scriptname</B
></TT
> gives a verbose syntax check.</P
><P
><TT
CLASS="USERINPUT"
><B
>sh -x scriptname</B
></TT
> echoes the result each
command, but in an abbreviated manner. This is the equivalent of
inserting <TT
CLASS="USERINPUT"
><B
>set -x</B
></TT
> or
<TT
CLASS="USERINPUT"
><B
>set -o xtrace</B
></TT
> in the script.</P
><P
><A
NAME="UNDVARERR"
></A
></P
><P
>Inserting <TT
CLASS="USERINPUT"
><B
>set -u</B
></TT
> or
<TT
CLASS="USERINPUT"
><B
>set -o nounset</B
></TT
> in the script runs it, but
gives an <SPAN
CLASS="ERRORNAME"
>unbound variable</SPAN
> error message
and aborts the script.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>set -u # Or set -o nounset
# Setting a variable to null will not trigger the error/abort.
# unset_var=
echo $unset_var # Unset (and undeclared) variable.
echo "Should not echo!"
# sh t2.sh
# t2.sh: line 6: unset_var: unbound variable</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
>Using an <SPAN
CLASS="QUOTE"
>"assert"</SPAN
> function to test a
variable or condition at critical points in a script. (This is
an idea borrowed from C.)</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ASSERT"
></A
><P
><B
>Example 32-4. Testing a condition with an
<I
CLASS="FIRSTTERM"
>assert</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# assert.sh
#######################################################################
assert () # If condition false,
{ #+ exit from script
#+ with appropriate error message.
E_PARAM_ERR=98
E_ASSERT_FAILED=99
if [ -z "$2" ] # Not enough parameters passed
then #+ to assert() function.
return $E_PARAM_ERR # No damage done.
fi
lineno=$2
if [ ! $1 ]
then
echo "Assertion failed: \"$1\""
echo "File \"$0\", line $lineno" # Give name of file and line number.
exit $E_ASSERT_FAILED
# else
# return
# and continue executing the script.
fi
} # Insert a similar assert() function into a script you need to debug.
#######################################################################
a=5
b=4
condition="$a -lt $b" # Error message and exit from script.
# Try setting "condition" to something else
#+ and see what happens.
assert "$condition" $LINENO
# The remainder of the script executes only if the "assert" does not fail.
# Some commands.
# Some more commands . . .
echo "This statement echoes only if the \"assert\" does not fail."
# . . .
# More commands . . .
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
>Using the <A
HREF="#LINENOREF"
>$LINENO</A
>
variable and the <A
HREF="#CALLERREF"
>caller</A
>
builtin.</P
></LI
><LI
><P
><A
NAME="DEBUGTRAP"
></A
>Trapping at exit.</P
><P
>The <A
HREF="#EXITREF"
>exit</A
> command in a script
triggers a signal <SPAN
CLASS="RETURNVALUE"
>0</SPAN
>, terminating
the process, that is, the script itself.
<A
NAME="AEN19460"
HREF="#FTN.AEN19460"
><SPAN
CLASS="footnote"
>[122]</SPAN
></A
>
It is often useful to trap the
<I
CLASS="FIRSTTERM"
>exit</I
>, forcing a <SPAN
CLASS="QUOTE"
>"printout"</SPAN
>
of variables, for example. The <I
CLASS="FIRSTTERM"
>trap</I
>
must be the first command in the script.</P
></LI
></OL
>
</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="TRAPREF1"
></A
>Trapping signals</B
></P
><DL
><DT
><B
CLASS="COMMAND"
>trap</B
></DT
><DD
><P
>Specifies an action on receipt of a
signal; also useful for debugging.</P
><P
><A
NAME="SIGNALD"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN19477"
></A
><P
></P
><P
>A <I
CLASS="FIRSTTERM"
>signal</I
> is a message
sent to a process, either by the kernel or another
process, telling it to take some specified action
(usually to terminate). For example, hitting a
<A
HREF="#CTLCREF"
>Control-C</A
>
sends a user interrupt, an INT signal, to a running
program.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
><EM
>A simple instance:</EM
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>trap '' 2
# Ignore interrupt 2 (Control-C), with no action specified.
trap 'echo "Control-C disabled."' 2
# Message when Control-C pressed.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DD
></DL
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX76"
></A
><P
><B
>Example 32-5. Trapping at exit</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Hunting variables with a trap.
trap 'echo Variable Listing --- a = $a b = $b' EXIT
# EXIT is the name of the signal generated upon exit from a script.
#
# The command specified by the "trap" doesn't execute until
#+ the appropriate signal is sent.
echo "This prints before the \"trap\" --"
echo "even though the script sees the \"trap\" first."
echo
a=39
b=36
exit 0
# Note that commenting out the 'exit' command makes no difference,
#+ since the script exits in any case after running out of commands.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ONLINE"
></A
><P
><B
>Example 32-6. Cleaning up after <B
CLASS="KEYCAP"
>Control-C</B
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# logon.sh: A quick 'n dirty script to check whether you are on-line yet.
umask 177 # Make sure temp files are not world readable.
TRUE=1
LOGFILE=/var/log/messages
# Note that $LOGFILE must be readable
#+ (as root, chmod 644 /var/log/messages).
TEMPFILE=temp.$$
# Create a "unique" temp file name, using process id of the script.
# Using 'mktemp' is an alternative.
# For example:
# TEMPFILE=`mktemp temp.XXXXXX`
KEYWORD=address
# At logon, the line "remote IP address xxx.xxx.xxx.xxx"
# appended to /var/log/messages.
ONLINE=22
USER_INTERRUPT=13
CHECK_LINES=100
# How many lines in log file to check.
trap 'rm -f $TEMPFILE; exit $USER_INTERRUPT' TERM INT
# Cleans up the temp file if script interrupted by control-c.
echo
while [ $TRUE ] #Endless loop.
do
tail -n $CHECK_LINES $LOGFILE&#62; $TEMPFILE
# Saves last 100 lines of system log file as temp file.
# Necessary, since newer kernels generate many log messages at log on.
search=`grep $KEYWORD $TEMPFILE`
# Checks for presence of the "IP address" phrase,
#+ indicating a successful logon.
if [ ! -z "$search" ] # Quotes necessary because of possible spaces.
then
echo "On-line"
rm -f $TEMPFILE # Clean up temp file.
exit $ONLINE
else
echo -n "." # The -n option to echo suppresses newline,
#+ so you get continuous rows of dots.
fi
sleep 1
done
# Note: if you change the KEYWORD variable to "Exit",
#+ this script can be used while on-line
#+ to check for an unexpected logoff.
# Exercise: Change the script, per the above note,
# and prettify it.
exit 0
# Nick Drage suggests an alternate method:
while true
do ifconfig ppp0 | grep UP 1&#62; /dev/null &#38;&#38; echo "connected" &#38;&#38; exit 0
echo -n "." # Prints dots (.....) until connected.
sleep 2
done
# Problem: Hitting Control-C to terminate this process may be insufficient.
#+ (Dots may keep on echoing.)
# Exercise: Fix this.
# Stephane Chazelas has yet another alternative:
CHECK_INTERVAL=1
while ! tail -n 1 "$LOGFILE" | grep -q "$KEYWORD"
do echo -n .
sleep $CHECK_INTERVAL
done
echo "On-line"
# Exercise: Discuss the relative strengths and weaknesses
# of each of these various approaches.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="PROGRESSBAR2"
></A
><P
><B
>Example 32-7. A Simple Implementation of a Progress Bar</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
# progress-bar2.sh
# Author: Graham Ewart (with reformatting by ABS Guide author).
# Used in ABS Guide with permission (thanks!).
# Invoke this script with bash. It doesn't work with sh.
interval=1
long_interval=10
{
trap "exit" SIGUSR1
sleep $interval; sleep $interval
while true
do
echo -n '.' # Use dots.
sleep $interval
done; } &#38; # Start a progress bar as a background process.
pid=$!
trap "echo !; kill -USR1 $pid; wait $pid" EXIT # To handle ^C.
echo -n 'Long-running process '
sleep $long_interval
echo ' Finished!'
kill -USR1 $pid
wait $pid # Stop the progress bar.
trap EXIT
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <TT
CLASS="OPTION"
>DEBUG</TT
> argument to
<B
CLASS="COMMAND"
>trap</B
> causes a specified action to execute
after every command in a script. This permits tracing variables,
for example.
<DIV
CLASS="EXAMPLE"
><A
NAME="VARTRACE"
></A
><P
><B
>Example 32-8. Tracing a variable</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
trap 'echo "VARIABLE-TRACE&#62; \$variable = \"$variable\""' DEBUG
# Echoes the value of $variable after every command.
variable=29; line=$LINENO
echo " Just initialized \$variable to $variable in line number $line."
let "variable *= 3"; line=$LINENO
echo " Just multiplied \$variable by 3 in line number $line."
exit 0
# The "trap 'command1 . . . command2 . . .' DEBUG" construct is
#+ more appropriate in the context of a complex script,
#+ where inserting multiple "echo $variable" statements might be
#+ awkward and time-consuming.
# Thanks, Stephane Chazelas for the pointer.
Output of script:
VARIABLE-TRACE&#62; $variable = ""
VARIABLE-TRACE&#62; $variable = "29"
Just initialized $variable to 29.
VARIABLE-TRACE&#62; $variable = "29"
VARIABLE-TRACE&#62; $variable = "87"
Just multiplied $variable by 3.
VARIABLE-TRACE&#62; $variable = "87"</PRE
></FONT
></TD
></TR
></TABLE
></DIV
>
</P
></TD
></TR
></TABLE
></DIV
><P
>Of course, the <B
CLASS="COMMAND"
>trap</B
> command has other uses
aside from debugging, such as disabling certain keystrokes within a
script (see <A
HREF="#STOPWATCH"
>Example A-43</A
>).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="MULTIPLEPROC"
></A
><P
><B
>Example 32-9. Running multiple processes (on an SMP box)</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# parent.sh
# Running multiple processes on an SMP box.
# Author: Tedman Eng
# This is the first of two scripts,
#+ both of which must be present in the current working directory.
LIMIT=$1 # Total number of process to start
NUMPROC=4 # Number of concurrent threads (forks?)
PROCID=1 # Starting Process ID
echo "My PID is $$"
function start_thread() {
if [ $PROCID -le $LIMIT ] ; then
./child.sh $PROCID&#38;
let "PROCID++"
else
echo "Limit reached."
wait
exit
fi
}
while [ "$NUMPROC" -gt 0 ]; do
start_thread;
let "NUMPROC--"
done
while true
do
trap "start_thread" SIGRTMIN
done
exit 0
# ======== Second script follows ========
#!/bin/bash
# child.sh
# Running multiple processes on an SMP box.
# This script is called by parent.sh.
# Author: Tedman Eng
temp=$RANDOM
index=$1
shift
let "temp %= 5"
let "temp += 4"
echo "Starting $index Time:$temp" "$@"
sleep ${temp}
echo "Ending $index"
kill -s SIGRTMIN $PPID
exit 0
# ======================= SCRIPT AUTHOR'S NOTES ======================= #
# It's not completely bug free.
# I ran it with limit = 500 and after the first few hundred iterations,
#+ one of the concurrent threads disappeared!
# Not sure if this is collisions from trap signals or something else.
# Once the trap is received, there's a brief moment while executing the
#+ trap handler but before the next trap is set. During this time, it may
#+ be possible to miss a trap signal, thus miss spawning a child process.
# No doubt someone may spot the bug and will be writing
#+ . . . in the future.
# ===================================================================== #
# ----------------------------------------------------------------------#
#################################################################
# The following is the original script written by Vernia Damiano.
# Unfortunately, it doesn't work properly.
#################################################################
#!/bin/bash
# Must call script with at least one integer parameter
#+ (number of concurrent processes).
# All other parameters are passed through to the processes started.
INDICE=8 # Total number of process to start
TEMPO=5 # Maximum sleep time per process
E_BADARGS=65 # No arg(s) passed to script.
if [ $# -eq 0 ] # Check for at least one argument passed to script.
then
echo "Usage: `basename $0` number_of_processes [passed params]"
exit $E_BADARGS
fi
NUMPROC=$1 # Number of concurrent process
shift
PARAMETRI=( "$@" ) # Parameters of each process
function avvia() {
local temp
local index
temp=$RANDOM
index=$1
shift
let "temp %= $TEMPO"
let "temp += 1"
echo "Starting $index Time:$temp" "$@"
sleep ${temp}
echo "Ending $index"
kill -s SIGRTMIN $$
}
function parti() {
if [ $INDICE -gt 0 ] ; then
avvia $INDICE "${PARAMETRI[@]}" &#38;
let "INDICE--"
else
trap : SIGRTMIN
fi
}
trap parti SIGRTMIN
while [ "$NUMPROC" -gt 0 ]; do
parti;
let "NUMPROC--"
done
wait
trap - SIGRTMIN
exit $?
: &#60;&#60;SCRIPT_AUTHOR_COMMENTS
I had the need to run a program, with specified options, on a number of
different files, using a SMP machine. So I thought [I'd] keep running
a specified number of processes and start a new one each time . . . one
of these terminates.
The "wait" instruction does not help, since it waits for a given process
or *all* process started in background. So I wrote [this] bash script
that can do the job, using the "trap" instruction.
--Vernia Damiano
SCRIPT_AUTHOR_COMMENTS</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><TT
CLASS="USERINPUT"
><B
>trap '' SIGNAL</B
></TT
> (two adjacent
apostrophes) disables SIGNAL for the remainder of the
script. <TT
CLASS="USERINPUT"
><B
>trap SIGNAL</B
></TT
> restores
the functioning of SIGNAL once more. This is useful to
protect a critical portion of a script from an undesirable
interrupt.</P
></TD
></TR
></TABLE
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> trap '' 2 # Signal 2 is Control-C, now disabled.
command
command
command
trap 2 # Reenables Control-C
</PRE
></FONT
></TD
></TR
></TABLE
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN19513"
></A
><P
></P
><P
><A
HREF="#BASH3REF"
>Version 3</A
> of Bash adds the
following <A
HREF="#INTERNALVARIABLES1"
>internal
variables</A
> for use by the debugger.
<P
></P
><OL
TYPE="1"
><LI
><P
><TT
CLASS="VARNAME"
>$BASH_ARGC</TT
></P
><P
>Number of command-line arguments passed to script,
similar to <A
HREF="#CLACOUNTREF"
><TT
CLASS="VARNAME"
>$#</TT
></A
>.</P
></LI
><LI
><P
><TT
CLASS="VARNAME"
>$BASH_ARGV</TT
></P
><P
>Final command-line parameter passed to script, equivalent
<A
HREF="#LASTARGREF"
><TT
CLASS="VARNAME"
>${!#}</TT
></A
>.</P
></LI
><LI
><P
><TT
CLASS="VARNAME"
>$BASH_COMMAND</TT
></P
><P
>Command currently executing.</P
></LI
><LI
><P
><TT
CLASS="VARNAME"
>$BASH_EXECUTION_STRING</TT
></P
><P
>The <I
CLASS="FIRSTTERM"
>option string</I
> following the
<TT
CLASS="OPTION"
>-c</TT
> <A
HREF="#CLOPTS"
>option</A
>
to Bash.</P
></LI
><LI
><P
><TT
CLASS="VARNAME"
>$BASH_LINENO</TT
></P
><P
>In a <A
HREF="#FUNCTIONREF"
>function</A
>,
indicates the line number of the function call.</P
></LI
><LI
><P
><TT
CLASS="VARNAME"
>$BASH_REMATCH</TT
></P
><P
>Array variable associated with <B
CLASS="COMMAND"
>=~</B
>
<A
HREF="#REGEXMATCHREF"
>conditional regex
matching</A
>.</P
></LI
><LI
><P
><A
NAME="BASHSOURCEREF"
></A
></P
><P
><TT
CLASS="VARNAME"
>$BASH_SOURCE</TT
></P
><P
>This is the name of the script, usually the same as
<A
HREF="#ARG0"
>$0</A
>.</P
></LI
><LI
><P
> <A
HREF="#BASHSUBSHELLREF"
><TT
CLASS="VARNAME"
>$BASH_SUBSHELL</TT
></A
></P
></LI
></OL
></P
><P
></P
></DIV
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="OPTIONS"
></A
>Chapter 33. Options</H1
><P
><A
NAME="OPTIONSREF"
></A
></P
><P
>Options are settings that change shell and/or script
behavior.</P
><P
>The <A
HREF="#SETREF"
>set</A
> command
enables options within a script. At the point in the script
where you want the options to take effect, use <B
CLASS="COMMAND"
>set
-o option-name</B
> or, in short form, <B
CLASS="COMMAND"
>set
-option-abbrev</B
>. These two forms are equivalent.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> #!/bin/bash
set -o verbose
# Echoes all commands before executing.
</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> #!/bin/bash
set -v
# Exact same effect as above.
</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>To <I
CLASS="FIRSTTERM"
>disable</I
> an option within a script,
use <B
CLASS="COMMAND"
>set +o option-name</B
> or <B
CLASS="COMMAND"
>set
+option-abbrev</B
>.</P
></TD
></TR
></TABLE
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> #!/bin/bash
set -o verbose
# Command echoing on.
command
...
command
set +o verbose
# Command echoing off.
command
# Not echoed.
set -v
# Command echoing on.
command
...
command
set +v
# Command echoing off.
command
exit 0
</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>An alternate method of enabling options in a script is
to specify them immediately following the
<TT
CLASS="REPLACEABLE"
><I
>#!</I
></TT
> script header.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> #!/bin/bash -x
#
# Body of script follows.
</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="INVOCATIONOPTIONSREF"
></A
></P
><P
>It is also possible to enable script options from the command
line. Some options that will not work with
<B
CLASS="COMMAND"
>set</B
> are available this way. Among these
are <TT
CLASS="REPLACEABLE"
><I
>-i</I
></TT
>, force script to run
interactive.</P
><P
><TT
CLASS="USERINPUT"
><B
>bash -v script-name</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>bash -o verbose script-name</B
></TT
></P
><P
>The following is a listing of some useful options. They may be
specified in either abbreviated form (preceded by a single dash)
or by complete name (preceded by a <EM
>double</EM
>
dash or by <TT
CLASS="OPTION"
>-o</TT
>).</P
><P
><A
NAME="OPTIONSTABLE"
></A
></P
><DIV
CLASS="TABLE"
><A
NAME="AEN19601"
></A
><P
><B
>Table 33-1. Bash options</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Abbreviation</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Name</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Effect</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-B</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>brace expansion</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><EM
>Enable</EM
>
<A
HREF="#BRACEEXPREF"
>brace
expansion</A
> (default setting =
<EM
>on</EM
>)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>+B</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>brace expansion</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><EM
>Disable</EM
>
brace expansion</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
NAME="NOCLOBBERREF"
></A
><TT
CLASS="OPTION"
>-C</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>noclobber</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Prevent overwriting of files by redirection (may be
overridden by <SPAN
CLASS="TOKEN"
>&#62;|</SPAN
>)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-D</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>(none)</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>List double-quoted strings prefixed by <SPAN
CLASS="TOKEN"
>$</SPAN
>,
but do not execute commands in script</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-a</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>allexport</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Export all defined variables</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-b</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>notify</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Notify when jobs running in background terminate (not of
much use in a script)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-c ...</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>(none)</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Read commands from <B
CLASS="COMMAND"
>...</B
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>checkjobs</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Informs user of any open <A
HREF="#JOBSREF"
>jobs</A
> upon shell exit.
Introduced in <A
HREF="#BASH4REF"
>version 4</A
>
of Bash, and still <SPAN
CLASS="QUOTE"
>"experimental."</SPAN
>
<EM
>Usage:</EM
> shopt -s checkjobs
(<EM
>Caution:</EM
> may hang!)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-e</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>errexit</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Abort script at first error, when a command
exits with non-zero status (except in <A
HREF="#UNTILLOOPREF"
>until</A
> or <A
HREF="#WHILELOOPREF"
>while loops</A
>, <A
HREF="#TESTCONSTRUCTS1"
>if-tests</A
>, <A
HREF="#LCONS1"
>list constructs</A
>)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-f</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>noglob</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Filename expansion (globbing) disabled</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>globstar</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#GLOBSTARREF"
><I
CLASS="FIRSTTERM"
>globbing</I
>
star-match</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Enables the <SPAN
CLASS="TOKEN"
>**</SPAN
> <A
HREF="#GLOBBINGREF"
>globbing</A
> operator
(<A
HREF="#BASH4REF"
>version 4+</A
> of Bash).
<EM
>Usage:</EM
> shopt -s globstar</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-i</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>interactive</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Script runs in <I
CLASS="FIRSTTERM"
>interactive</I
> mode</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-n</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>noexec</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Read commands in script, but do not execute them (syntax check)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-o Option-Name</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>(none)</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Invoke the <I
CLASS="FIRSTTERM"
>Option-Name</I
>
option</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-o posix</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>POSIX</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Change the behavior of Bash, or invoked script, to
conform to <A
HREF="#POSIX2REF"
>POSIX</A
>
standard.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-o pipefail</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>pipe failure</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Causes a pipeline to return the <A
HREF="#EXITSTATUSREF"
>exit status</A
> of
the last command in the pipe that returned a non-zero
return value.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-p</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>privileged</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Script runs as <SPAN
CLASS="QUOTE"
>"suid"</SPAN
> (caution!)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-r</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>restricted</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Script runs in <I
CLASS="FIRSTTERM"
>restricted</I
>
mode (see <A
HREF="#RESTRICTED-SH"
>Chapter 22</A
>).</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-s</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>stdin</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Read commands from <TT
CLASS="FILENAME"
>stdin</TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-t</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>(none)</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Exit after first command</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-u</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>nounset</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Attempt to use undefined variable
outputs error message, and forces an exit</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-v</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>verbose</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Print each command to <TT
CLASS="FILENAME"
>stdout</TT
> before executing it</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-x</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>xtrace</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Similar to <TT
CLASS="OPTION"
>-v</TT
>, but expands commands</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>(none)</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>End of options flag. All other arguments
are <A
HREF="#POSPARAMREF"
>positional
parameters</A
>.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>--</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>(none)</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Unset positional parameters.
If arguments given (<TT
CLASS="PARAMETER"
><I
>-- arg1 arg2</I
></TT
>),
positional parameters set to arguments.</TD
></TR
></TBODY
></TABLE
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="GOTCHAS"
></A
>Chapter 34. Gotchas</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>Turandot: <I
CLASS="FOREIGNPHRASE"
>Gli enigmi sono tre, la morte
una!</I
></I
></P
><P
><I
>Caleph: <I
CLASS="FOREIGNPHRASE"
>No, no! Gli enigmi sono tre, una la
vita!</I
></I
></P
><P
><I
>--Puccini</I
></P
></I
></TD
></TR
></TABLE
><P
><A
NAME="BASH3GOTCHA"
></A
></P
><P
>Here are some (non-recommended!) scripting practices that
will bring excitement into an otherwise dull life.</P
><P
></P
><UL
><LI
><P
><A
NAME="INAPPVN"
></A
></P
><P
>Assigning reserved words or characters to variable names.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>case=value0 # Causes problems.
23skidoo=value1 # Also problems.
# Variable names starting with a digit are reserved by the shell.
# Try _23skidoo=value1. Starting variables with an underscore is okay.
# However . . . using just an underscore will not work.
_=25
echo $_ # $_ is a special variable set to last arg of last command.
# But . . . _ is a valid function name!
xyz((!*=value2 # Causes severe problems.
# As of version 3 of Bash, periods are not allowed within variable names.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
>Using a hyphen or other reserved characters in a variable name (or
function name).</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>var-1=23
# Use 'var_1' instead.
function-whatever () # Error
# Use 'function_whatever ()' instead.
# As of version 3 of Bash, periods are not allowed within function names.
function.whatever () # Error
# Use 'functionWhatever ()' instead.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
>Using the same name for a variable and a function. This can make a
script difficult to understand.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>do_something ()
{
echo "This function does something with \"$1\"."
}
do_something=do_something
do_something do_something
# All this is legal, but highly confusing.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><A
NAME="WSBAD"
></A
>Using <A
HREF="#WHITESPACEREF"
>whitespace</A
> inappropriately.
In contrast to other programming languages, Bash can be quite
finicky about whitespace.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>var1 = 23 # 'var1=23' is correct.
# On line above, Bash attempts to execute command "var1"
# with the arguments "=" and "23".
let c = $a - $b # Instead: let c=$a-$b or let "c = $a - $b"
if [ $a -le 5] # if [ $a -le 5 ] is correct.
# ^^ if [ "$a" -le 5 ] is even better.
# [[ $a -le 5 ]] also works.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><A
NAME="OMITSEMICOLON"
></A
></P
><P
>Not terminating with a <A
HREF="#SEMICOLONREF"
>semicolon</A
> the final command
in a <A
HREF="#CODEBLOCKREF"
>code block within curly
brackets</A
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>{ ls -l; df; echo "Done." }
# bash: syntax error: unexpected end of file
{ ls -l; df; echo "Done."; }
# ^ ### Final command needs semicolon.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><A
NAME="UNINITVAR"
></A
></P
><P
> Assuming uninitialized variables (variables before a value is
assigned to them) are <SPAN
CLASS="QUOTE"
>"zeroed out"</SPAN
>. An
uninitialized variable has a value of <I
CLASS="FIRSTTERM"
>null</I
>,
<EM
>not</EM
> zero.</P
><P
><A
NAME="BASH4.2-UNINITIALIZED"
></A
></P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo "uninitialized_var = $uninitialized_var"
# uninitialized_var =
# However . . .
# if $BASH_VERSION &#8805; 4.2; then
if [[ ! -v uninitialized_var ]]
then
uninitialized_var=0 # Initialize it to zero!
fi
&#13;</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><A
NAME="EQDIF"
></A
></P
><P
>Mixing up <I
CLASS="FIRSTTERM"
>=</I
> and
<I
CLASS="FIRSTTERM"
>-eq</I
> in a test. Remember,
<I
CLASS="FIRSTTERM"
>=</I
> is for comparing literal variables
and <I
CLASS="FIRSTTERM"
>-eq</I
> for integers.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ "$a" = 273 ] # Is $a an integer or string?
if [ "$a" -eq 273 ] # If $a is an integer.
# Sometimes you can interchange -eq and = without adverse consequences.
# However . . .
a=273.0 # Not an integer.
if [ "$a" = 273 ]
then
echo "Comparison works."
else
echo "Comparison does not work."
fi # Comparison does not work.
# Same with a=" 273" and a="0273".
# Likewise, problems trying to use "-eq" with non-integer values.
if [ "$a" -eq 273.0 ]
then
echo "a = $a"
fi # Aborts with an error message.
# test.sh: [: 273.0: integer expression expected</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><A
NAME="NUMSTRCOMPNE"
></A
></P
><P
>Misusing <A
HREF="#SCOMPARISON1"
>string comparison</A
>
operators.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="BADOP"
></A
><P
><B
>Example 34-1. Numerical and string comparison are not equivalent</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# bad-op.sh: Trying to use a string comparison on integers.
echo
number=1
# The following while-loop has two errors:
#+ one blatant, and the other subtle.
while [ "$number" &#60; 5 ] # Wrong! Should be: while [ "$number" -lt 5 ]
do
echo -n "$number "
let "number += 1"
done
# Attempt to run this bombs with the error message:
#+ bad-op.sh: line 10: 5: No such file or directory
# Within single brackets, "&#60;" must be escaped,
#+ and even then, it's still wrong for comparing integers.
echo "---------------------"
while [ "$number" \&#60; 5 ] # 1 2 3 4
do #
echo -n "$number " # It *seems* to work, but . . .
let "number += 1" #+ it actually does an ASCII comparison,
done #+ rather than a numerical one.
echo; echo "---------------------"
# This can cause problems. For example:
lesser=5
greater=105
if [ "$greater" \&#60; "$lesser" ]
then
echo "$greater is less than $lesser"
fi # 105 is less than 5
# In fact, "105" actually is less than "5"
#+ in a string comparison (ASCII sort order).
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><A
NAME="LETBAD"
></A
></P
><P
>Attempting to use <A
HREF="#LETREF"
>let</A
>
to set string variables.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>let "a = hello, you"
echo "$a" # 0</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
><A
NAME="FAILQUOTE"
></A
></P
><P
>Sometimes variables within <SPAN
CLASS="QUOTE"
>"test"</SPAN
> brackets
([ ]) need to be quoted (double quotes). Failure to do so may
cause unexpected behavior. See <A
HREF="#STRTEST"
>Example 7-6</A
>, <A
HREF="#REDIR2"
>Example 20-5</A
>, and <A
HREF="#ARGLIST"
>Example 9-6</A
>.</P
></LI
><LI
><P
><A
NAME="FAILNOTQUOTE"
></A
></P
><P
>Quoting a variable containing whitespace <A
HREF="#WSQUO"
>prevents splitting</A
>. Sometimes
this produces <A
HREF="#VARSPLITTING"
>unintended
consequences</A
>.</P
></LI
><LI
><P
><A
NAME="EXECPERM"
></A
></P
><P
>Commands issued from a script may fail to execute because
the script owner lacks execute permission for them. If a user
cannot invoke a command from the command-line, then putting it
into a script will likewise fail. Try changing the attributes of
the command in question, perhaps even setting the suid bit
(as <I
CLASS="FIRSTTERM"
>root</I
>, of course).</P
></LI
><LI
><P
><A
NAME="DASHNREDR"
></A
></P
><P
>Attempting to use <B
CLASS="COMMAND"
>-</B
> as a redirection
operator (which it is not) will usually result in an unpleasant
surprise.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>command1 2&#62; - | command2
# Trying to redirect error output of command1 into a pipe . . .
# . . . will not work.
command1 2&#62;&#38; - | command2 # Also futile.
Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
><A
NAME="LATEVERF"
></A
></P
><P
>Using Bash <A
HREF="#BASH2REF"
>version 2+</A
>
functionality may cause a bailout with error messages. Older
Linux machines may have version 1.XX of Bash as the default
installation.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
minimum_version=2
# Since Chet Ramey is constantly adding features to Bash,
# you may set $minimum_version to 2.XX, 3.XX, or whatever is appropriate.
E_BAD_VERSION=80
if [ "$BASH_VERSION" \&#60; "$minimum_version" ]
then
echo "This script works only with Bash, version $minimum or greater."
echo "Upgrade strongly recommended."
exit $E_BAD_VERSION
fi
...</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
>Using Bash-specific functionality in a <A
HREF="#BASHDEF"
>Bourne shell</A
> script
(<TT
CLASS="USERINPUT"
><B
>#!/bin/sh</B
></TT
>) on a non-Linux machine
<A
HREF="#BINSH"
>may cause unexpected behavior</A
>.
A Linux system usually aliases <B
CLASS="COMMAND"
>sh</B
> to
<B
CLASS="COMMAND"
>bash</B
>, but this does not necessarily hold true
for a generic UNIX machine.</P
></LI
><LI
><P
><A
NAME="UNDOCF"
></A
></P
><P
>Using undocumented features in Bash turns out to be a
dangerous practice. In previous releases of this
book there were several scripts that depended on the
<SPAN
CLASS="QUOTE"
>"feature"</SPAN
> that, although the maximum value
of an <A
HREF="#EXITSTATUSREF"
>exit</A
> or <A
HREF="#RETURNREF"
>return</A
> value was 255, that limit
did not apply to <EM
>negative</EM
> integers.
Unfortunately, in version 2.05b and later, that loophole
disappeared. See <A
HREF="#RETURNTEST"
>Example 24-9</A
>.</P
></LI
><LI
><P
><A
NAME="GOTCHAEXITVALANAMALIES"
></A
></P
><P
>In certain contexts, a misleading <A
HREF="#EXITSTATUSREF"
>exit status</A
>
may be returned. This may occur when <A
HREF="#EXITVALANOMALY01"
>setting a local variable within a
function</A
> or when <A
HREF="#EXITVALANOMALY02"
>assigning
an arithmetic value to a variable</A
>.</P
></LI
><LI
><P
><A
NAME="ARXS1"
></A
>The <A
HREF="#ARXS"
>exit
status of an arithmetic expression</A
> is
<EM
>not</EM
> equivalent to an <I
CLASS="FIRSTTERM"
>error
code</I
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>var=1 &#38;&#38; ((--var)) &#38;&#38; echo $var
# ^^^^^^^^^ Here the and-list terminates with exit status 1.
# $var doesn't echo!
echo $? # 1</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
><A
NAME="DOSNEWLINES"
></A
></P
><P
> A script with DOS-type newlines (<TT
CLASS="REPLACEABLE"
><I
>\r\n</I
></TT
>)
will fail to execute, since <TT
CLASS="USERINPUT"
><B
>#!/bin/bash\r\n</B
></TT
>
is <EM
>not</EM
> recognized, <EM
>not</EM
>
the same as the expected <TT
CLASS="USERINPUT"
><B
>#!/bin/bash\n</B
></TT
>. The
fix is to convert the script to UNIX-style newlines.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo "Here"
unix2dos $0 # Script changes itself to DOS format.
chmod 755 $0 # Change back to execute permission.
# The 'unix2dos' command removes execute permission.
./$0 # Script tries to run itself again.
# But it won't work as a DOS file.
echo "There"
exit 0</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><A
NAME="BINSH"
></A
></P
><P
>A shell script headed by <TT
CLASS="USERINPUT"
><B
>#!/bin/sh</B
></TT
>
will not run in full Bash-compatibility mode. Some Bash-specific
functions might be disabled. Scripts that need complete
access to all the Bash-specific extensions should start with
<TT
CLASS="USERINPUT"
><B
>#!/bin/bash</B
></TT
>.</P
></LI
><LI
><P
><A
HREF="#INDENTEDLS"
>Putting whitespace in front of
the terminating limit string</A
> of a <A
HREF="#HEREDOCREF"
>here document</A
> will cause unexpected
behavior in a script.</P
></LI
><LI
><P
><A
NAME="RVTCAUTION2"
></A
>Putting more than one
<I
CLASS="FIRSTTERM"
>echo</I
> statement in a function <A
HREF="#RVT"
>whose output is captured</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>add2 ()
{
echo "Whatever ... " # Delete this line!
let "retval = $1 + $2"
echo $retval
}
num1=12
num2=43
echo "Sum of $num1 and $num2 = $(add2 $num1 $num2)"
# Sum of 12 and 43 = Whatever ...
# 55
# The "echoes" concatenate.</PRE
></FONT
></TD
></TR
></TABLE
>
This <A
HREF="#RVTCAUTION"
>will not work</A
>.</P
></LI
><LI
><P
><A
NAME="PARCHILDPROBREF"
></A
></P
><P
>A script may not <B
CLASS="COMMAND"
>export</B
> variables back
to its <A
HREF="#FORKREF"
>parent process</A
>, the shell,
or to the environment. Just as we learned in biology, a child
process can inherit from a parent, but not vice versa.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>WHATEVER=/home/bozo
export WHATEVER
exit 0</PRE
></FONT
></TD
></TR
></TABLE
>
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><B
CLASS="COMMAND"
>echo $WHATEVER</B
>
<TT
CLASS="COMPUTEROUTPUT"
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
></PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> Sure enough, back at the command prompt, $WHATEVER remains unset.
</P
></LI
><LI
><P
><A
NAME="VARSUBSH"
></A
></P
><P
>Setting and manipulating variables in a <A
HREF="#SUBSHELLSREF"
>subshell</A
>, then attempting
to use those same variables outside the scope of the subshell will
result an unpleasant surprise.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SUBPIT"
></A
><P
><B
>Example 34-2. Subshell Pitfalls</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Pitfalls of variables in a subshell.
outer_variable=outer
echo
echo "outer_variable = $outer_variable"
echo
(
# Begin subshell
echo "outer_variable inside subshell = $outer_variable"
inner_variable=inner # Set
echo "inner_variable inside subshell = $inner_variable"
outer_variable=inner # Will value change globally?
echo "outer_variable inside subshell = $outer_variable"
# Will 'exporting' make a difference?
# export inner_variable
# export outer_variable
# Try it and see.
# End subshell
)
echo
echo "inner_variable outside subshell = $inner_variable" # Unset.
echo "outer_variable outside subshell = $outer_variable" # Unchanged.
echo
exit 0
# What happens if you uncomment lines 19 and 20?
# Does it make a difference?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><A
NAME="BADREAD0"
></A
></P
><P
><A
HREF="#PIPEREF"
>Piping</A
>
<B
CLASS="COMMAND"
>echo</B
> output to a <A
HREF="#READREF"
>read</A
> may produce unexpected
results. In this scenario, the <B
CLASS="COMMAND"
>read</B
>
acts as if it were running in a subshell. Instead, use
the <A
HREF="#SETREF"
>set</A
> command (as in <A
HREF="#SETPOS"
>Example 15-18</A
>).</P
><DIV
CLASS="EXAMPLE"
><A
NAME="BADREAD"
></A
><P
><B
>Example 34-3. Piping the output of <I
CLASS="FIRSTTERM"
>echo</I
> to a
<I
CLASS="FIRSTTERM"
>read</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# badread.sh:
# Attempting to use 'echo and 'read'
#+ to assign variables non-interactively.
# shopt -s lastpipe
a=aaa
b=bbb
c=ccc
echo "one two three" | read a b c
# Try to reassign a, b, and c.
echo
echo "a = $a" # a = aaa
echo "b = $b" # b = bbb
echo "c = $c" # c = ccc
# Reassignment failed.
### However . . .
## Uncommenting line 6:
# shopt -s lastpipe
##+ fixes the problem!
### This is a new feature in Bash, version 4.2.
# ------------------------------
# Try the following alternative.
var=`echo "one two three"`
set -- $var
a=$1; b=$2; c=$3
echo "-------"
echo "a = $a" # a = one
echo "b = $b" # b = two
echo "c = $c" # c = three
# Reassignment succeeded.
# ------------------------------
# Note also that an echo to a 'read' works within a subshell.
# However, the value of the variable changes *only* within the subshell.
a=aaa # Starting all over again.
b=bbb
c=ccc
echo; echo
echo "one two three" | ( read a b c;
echo "Inside subshell: "; echo "a = $a"; echo "b = $b"; echo "c = $c" )
# a = one
# b = two
# c = three
echo "-----------------"
echo "Outside subshell: "
echo "a = $a" # a = aaa
echo "b = $b" # b = bbb
echo "c = $c" # c = ccc
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="PIPELOOP"
></A
></P
><P
>In fact, as Anthony Richardson points out, piping to
<EM
>any</EM
> loop can cause a similar problem.</P
><P
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Loop piping troubles.
# This example by Anthony Richardson,
#+ with addendum by Wilbert Berendsen.
foundone=false
find $HOME -type f -atime +30 -size 100k |
while true
do
read f
echo "$f is over 100KB and has not been accessed in over 30 days"
echo "Consider moving the file to archives."
foundone=true
# ------------------------------------
echo "Subshell level = $BASH_SUBSHELL"
# Subshell level = 1
# Yes, we're inside a subshell.
# ------------------------------------
done
# foundone will always be false here since it is
#+ set to true inside a subshell
if [ $foundone = false ]
then
echo "No files need archiving."
fi
# =====================Now, here is the correct way:=================
foundone=false
for f in $(find $HOME -type f -atime +30 -size 100k) # No pipe here.
do
echo "$f is over 100KB and has not been accessed in over 30 days"
echo "Consider moving the file to archives."
foundone=true
done
if [ $foundone = false ]
then
echo "No files need archiving."
fi
# ==================And here is another alternative==================
# Places the part of the script that reads the variables
#+ within a code block, so they share the same subshell.
# Thank you, W.B.
find $HOME -type f -atime +30 -size 100k | {
foundone=false
while read f
do
echo "$f is over 100KB and has not been accessed in over 30 days"
echo "Consider moving the file to archives."
foundone=true
done
if ! $foundone
then
echo "No files need archiving."
fi
}</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="PTAILGREP"
></A
></P
><P
> A lookalike problem occurs when trying to write the
<TT
CLASS="FILENAME"
>stdout</TT
> of a <B
CLASS="COMMAND"
>tail -f</B
>
piped to <A
HREF="#GREPREF"
>grep</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>tail -f /var/log/messages | grep "$ERROR_MSG" &#62;&#62; error.log
# The "error.log" file will not have anything written to it.
# As Samuli Kaipiainen points out, this results from grep
#+ buffering its output.
# The fix is to add the "--line-buffered" parameter to grep.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><A
NAME="SUIDSCR"
></A
></P
><P
>Using <SPAN
CLASS="QUOTE"
>"suid"</SPAN
> commands within scripts is risky,
as it may compromise system security.
<A
NAME="AEN19993"
HREF="#FTN.AEN19993"
><SPAN
CLASS="footnote"
>[123]</SPAN
></A
>
</P
></LI
><LI
><P
><A
NAME="CGIREF"
></A
></P
><P
>Using shell scripts for CGI programming may be problematic. Shell
script variables are not <SPAN
CLASS="QUOTE"
>"typesafe,"</SPAN
> and this can cause
undesirable behavior as far as CGI is concerned. Moreover, it is
difficult to <SPAN
CLASS="QUOTE"
>"cracker-proof"</SPAN
> shell scripts.</P
></LI
><LI
><P
>Bash does not handle the <A
HREF="#DOUBLESLASHREF"
>double slash
(<SPAN
CLASS="TOKEN"
>//</SPAN
>) string</A
> correctly.</P
></LI
><LI
><P
><A
NAME="GNUREF"
></A
></P
><P
>Bash scripts written for Linux or BSD systems may need
fixups to run on a commercial UNIX machine. Such
scripts often employ the GNU set of commands and filters,
which have greater functionality than their generic UNIX
counterparts. This is particularly true of such text processing
utilites as <A
HREF="#TRREF"
>tr</A
>.</P
></LI
><LI
><P
><A
NAME="UPDATEBREAKS"
></A
></P
><P
>Sadly, updates to Bash itself have broken older scripts
that <A
HREF="#PARAGRAPHSPACE"
>used to work perfectly
fine</A
>. Let us recall <A
HREF="#UNDOCF"
>how
risky it is to use undocumented Bash features</A
>.</P
></LI
></UL
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>Danger is near thee --</I
></P
><P
><I
>Beware, beware, beware, beware.</I
></P
><P
><I
>Many brave hearts are asleep in the deep.</I
></P
><P
><I
>So beware --</I
></P
><P
><I
>Beware.</I
></P
><P
><I
>--A.J. Lamb and H.W. Petrie</I
></P
></I
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="SCRSTYLE"
></A
>Chapter 35. Scripting With Style</H1
><P
>Get into the habit of writing shell scripts in a structured and
systematic manner. Even on-the-fly and <SPAN
CLASS="QUOTE"
>"written on the
back of an envelope"</SPAN
> scripts will benefit if you take a
few minutes to plan and organize your thoughts before sitting
down and coding.</P
><P
>Herewith are a few stylistic guidelines. This is not
(necessarily) intended as an <EM
>Official Shell Scripting
Stylesheet</EM
>.</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="UNOFFICIALST"
></A
>35.1. Unofficial Shell Scripting Stylesheet</H1
><P
></P
><UL
><LI
><P
>Comment your code. This makes it easier for others to
understand (and appreciate), and easier for you to maintain.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>PASS="$PASS${MATRIX:$(($RANDOM%${#MATRIX})):1}"
# It made perfect sense when you wrote it last year,
#+ but now it's a complete mystery.
# (From Antek Sawicki's "pw.sh" script.)</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Add descriptive headers to your scripts and functions.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
#************************************************#
# xyz.sh #
# written by Bozo Bozeman #
# July 05, 2001 #
# #
# Clean up project files. #
#************************************************#
E_BADDIR=85 # No such directory.
projectdir=/home/bozo/projects # Directory to clean up.
# --------------------------------------------------------- #
# cleanup_pfiles () #
# Removes all files in designated directory. #
# Parameter: $target_directory #
# Returns: 0 on success, $E_BADDIR if something went wrong. #
# --------------------------------------------------------- #
cleanup_pfiles ()
{
if [ ! -d "$1" ] # Test if target directory exists.
then
echo "$1 is not a directory."
return $E_BADDIR
fi
rm -f "$1"/*
return 0 # Success.
}
cleanup_pfiles $projectdir
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
>Avoid using <SPAN
CLASS="QUOTE"
>"magic numbers,"</SPAN
>
<A
NAME="AEN20041"
HREF="#FTN.AEN20041"
><SPAN
CLASS="footnote"
>[124]</SPAN
></A
>
that is, <SPAN
CLASS="QUOTE"
>"hard-wired"</SPAN
> literal constants. Use
meaningful variable names instead. This makes the script
easier to understand and permits making changes and updates
without breaking the application.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ -f /var/log/messages ]
then
...
fi
# A year later, you decide to change the script to check /var/log/syslog.
# It is now necessary to manually change the script, instance by instance,
#+ and hope nothing breaks.
# A better way:
LOGFILE=/var/log/messages # Only line that needs to be changed.
if [ -f "$LOGFILE" ]
then
...
fi</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
>Choose descriptive names for variables and functions.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>fl=`ls -al $dirname` # Cryptic.
file_listing=`ls -al $dirname` # Better.
MAXVAL=10 # All caps used for a script constant.
while [ "$index" -le "$MAXVAL" ]
...
E_NOTFOUND=95 # Uppercase for an errorcode,
#+ and name prefixed with E_.
if [ ! -e "$filename" ]
then
echo "File $filename not found."
exit $E_NOTFOUND
fi
MAIL_DIRECTORY=/var/spool/mail/bozo # Uppercase for an environmental
export MAIL_DIRECTORY #+ variable.
GetAnswer () # Mixed case works well for a
{ #+ function name, especially
prompt=$1 #+ when it improves legibility.
echo -n $prompt
read answer
return $answer
}
GetAnswer "What is your favorite number? "
favorite_number=$?
echo $favorite_number
_uservariable=23 # Permissible, but not recommended.
# It's better for user-defined variables not to start with an underscore.
# Leave that for system variables.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
>Use <A
HREF="#EXITCOMMANDREF"
>exit codes</A
>
in a systematic and meaningful way.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>E_WRONG_ARGS=95
...
...
exit $E_WRONG_ARGS</PRE
></FONT
></TD
></TR
></TABLE
>
See also <A
HREF="#EXITCODES"
>Appendix E</A
>.</P
><P
><EM
>Ender</EM
> suggests using the <A
HREF="#SYSEXITSREF"
>exit codes
in <TT
CLASS="FILENAME"
>/usr/include/sysexits.h</TT
></A
> in shell
scripts, though these are primarily intended for C and C++
programming.</P
></LI
><LI
><P
>Use standardized parameter flags for script invocation.
<EM
>Ender</EM
> proposes the following set
of flags.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>-a All: Return all information (including hidden file info).
-b Brief: Short version, usually for other scripts.
-c Copy, concatenate, etc.
-d Daily: Use information from the whole day, and not merely
information for a specific instance/user.
-e Extended/Elaborate: (often does not include hidden file info).
-h Help: Verbose usage w/descs, aux info, discussion, help.
See also -V.
-l Log output of script.
-m Manual: Launch man-page for base command.
-n Numbers: Numerical data only.
-r Recursive: All files in a directory (and/or all sub-dirs).
-s Setup &#38; File Maintenance: Config files for this script.
-u Usage: List of invocation flags for the script.
-v Verbose: Human readable output, more or less formatted.
-V Version / License / Copy(right|left) / Contribs (email too).</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>See also <A
HREF="#STANDARD-OPTIONS"
>Section G.1</A
>.</P
></LI
><LI
><P
>Break complex scripts into simpler modules. Use functions
where appropriate. See <A
HREF="#CARDS"
>Example 37-4</A
>.</P
></LI
><LI
><P
>Don't use a complex construct where a simpler one will do.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>COMMAND
if [ $? -eq 0 ]
...
# Redundant and non-intuitive.
if COMMAND
...
# More concise (if perhaps not quite as legible).</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
></UL
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>... reading the UNIX source code to the Bourne shell (/bin/sh). I
was shocked at how much simple algorithms could be made cryptic, and
therefore useless, by a poor choice of code style. I asked myself,
<SPAN
CLASS="QUOTE"
>"Could someone be proud of this code?"</SPAN
></I
></P
><P
><I
>--Landon Noll</I
></P
></I
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="MISCELLANY"
></A
>Chapter 36. Miscellany</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>Nobody really knows what the Bourne shell's grammar is. Even
examination of the source code is little help.</I
></P
><P
><I
>--Tom Duff</I
></P
></I
></TD
></TR
></TABLE
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="INTANDNONINT"
></A
>36.1. Interactive and non-interactive shells and scripts</H1
><P
>An <I
CLASS="FIRSTTERM"
>interactive</I
> shell reads
commands from user input on a <TT
CLASS="FILENAME"
>tty</TT
>. Among
other things, such a shell reads startup files on activation,
displays a prompt, and enables job control by default. The
user can <I
CLASS="FIRSTTERM"
>interact</I
> with the shell.</P
><P
>A shell running a script is always a non-interactive
shell. All the same, the script can still access its
<TT
CLASS="FILENAME"
>tty</TT
>. It is even possible to emulate an
interactive shell in a script.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
MY_PROMPT='$ '
while :
do
echo -n "$MY_PROMPT"
read line
eval "$line"
done
exit 0
# This example script, and much of the above explanation supplied by
# St<53>phane Chazelas (thanks again).</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Let us consider an <I
CLASS="FIRSTTERM"
>interactive</I
>
script to be one that requires input from the user, usually
with <A
HREF="#READREF"
>read</A
> statements (see <A
HREF="#EX36"
>Example 15-3</A
>). <SPAN
CLASS="QUOTE"
>"Real life"</SPAN
> is actually a
bit messier than that. For now, assume an interactive script
is bound to a tty, a script that a user has invoked from the
console or an <I
CLASS="FIRSTTERM"
>xterm</I
>.</P
><P
>Init and startup scripts are necessarily non-interactive,
since they must run without human intervention. Many
administrative and system maintenance scripts are likewise
non-interactive. Unvarying repetitive tasks cry out for
automation by non-interactive scripts.</P
><P
>Non-interactive scripts can run in the background, but
interactive ones hang, waiting for input that never comes.
Handle that difficulty by having an <B
CLASS="COMMAND"
>expect</B
>
script or embedded <A
HREF="#HEREDOCREF"
>here
document</A
> feed input to an interactive script running
as a background job. In the simplest case, redirect a
file to supply input to a <B
CLASS="COMMAND"
>read</B
> statement
(<B
CLASS="COMMAND"
>read variable &#60;file</B
>). These particular
workarounds make possible general purpose scripts that run
in either interactive or non-interactive modes.</P
><P
>If a script needs to test whether it is running in an
interactive shell, it is simply a matter of finding
whether the <I
CLASS="FIRSTTERM"
>prompt</I
> variable, <A
HREF="#PS1REF"
>$PS1</A
> is set. (If the user is being
prompted for input, then the script needs to display a
prompt.)</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if [ -z $PS1 ] # no prompt?
### if [ -v PS1 ] # On Bash 4.2+ ...
then
# non-interactive
...
else
# interactive
...
fi</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="IITEST"
></A
>Alternatively, the script can test
for the presence of option <SPAN
CLASS="QUOTE"
>"i"</SPAN
> in the <A
HREF="#FLPREF"
>$-</A
> flag.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>case $- in
*i*) # interactive shell
;;
*) # non-interactive shell
;;
# (Courtesy of "UNIX F.A.Q.," 1993)</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="II2TEST"
></A
>However, John Lange describes
an alternative method, using the <A
HREF="#TERMTEST"
><SPAN
CLASS="TOKEN"
>-t</SPAN
>
<I
CLASS="FIRSTTERM"
>test</I
> operator</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Test for a terminal!
fd=0 # stdin
# As we recall, the -t test option checks whether the stdin, [ -t 0 ],
#+ or stdout, [ -t 1 ], in a given script is running in a terminal.
if [ -t "$fd" ]
then
echo interactive
else
echo non-interactive
fi
# But, as John points out:
# if [ -t 0 ] works ... when you're logged in locally
# but fails when you invoke the command remotely via ssh.
# So for a true test you also have to test for a socket.
if [[ -t "$fd" || -p /dev/stdin ]]
then
echo interactive
else
echo non-interactive
fi</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Scripts may be forced to run in interactive
mode with the <SPAN
CLASS="TOKEN"
>-i</SPAN
> option or with a
<TT
CLASS="USERINPUT"
><B
>#!/bin/bash -i</B
></TT
> header. Be aware that
this can cause erratic script behavior or show error messages
even when no error is present.</P
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="WRAPPER"
></A
>36.2. Shell Wrappers</H1
><P
><A
NAME="SHWRAPPER"
></A
></P
><P
>A <I
CLASS="FIRSTTERM"
>wrapper</I
> is a shell script that embeds
a system command or utility, that accepts and passes a set of
parameters to that command.
<A
NAME="AEN20130"
HREF="#FTN.AEN20130"
><SPAN
CLASS="footnote"
>[125]</SPAN
></A
>
Wrapping a script around a complex command-line
simplifies invoking it. This is expecially useful
with <A
HREF="#SEDREF"
>sed</A
> and <A
HREF="#AWKREF"
>awk</A
>.</P
><P
>A
<B
CLASS="COMMAND"
> sed</B
> or
<B
CLASS="COMMAND"
>
awk</B
> script would normally be invoked
from the command-line by a <TT
CLASS="USERINPUT"
><B
>sed -e
<TT
CLASS="REPLACEABLE"
><I
>'commands'</I
></TT
></B
></TT
>
or <TT
CLASS="USERINPUT"
><B
>awk
<TT
CLASS="REPLACEABLE"
><I
>'commands'</I
></TT
></B
></TT
>. Embedding
such a script in a Bash script permits calling it more simply,
and makes it <I
CLASS="FIRSTTERM"
>reusable</I
>. This also
enables combining the functionality of <I
CLASS="FIRSTTERM"
>sed</I
>
and <I
CLASS="FIRSTTERM"
>awk</I
>, for example <A
HREF="#PIPEREF"
>piping</A
> the output of a set of
<I
CLASS="FIRSTTERM"
>sed</I
> commands to
<I
CLASS="FIRSTTERM"
>awk</I
>. As a saved executable file,
you can then repeatedly invoke it in its original form or
modified, without the inconvenience of retyping it on the
command-line.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX3"
></A
><P
><B
>Example 36-1. <I
CLASS="FIRSTTERM"
>shell wrapper</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# This simple script removes blank lines from a file.
# No argument checking.
#
# You might wish to add something like:
#
# E_NOARGS=85
# if [ -z "$1" ]
# then
# echo "Usage: `basename $0` target-file"
# exit $E_NOARGS
# fi
sed -e /^$/d "$1"
# Same as
# sed -e '/^$/d' filename
# invoked from the command-line.
# The '-e' means an "editing" command follows (optional here).
# '^' indicates the beginning of line, '$' the end.
# This matches lines with nothing between the beginning and the end --
#+ blank lines.
# The 'd' is the delete command.
# Quoting the command-line arg permits
#+ whitespace and special characters in the filename.
# Note that this script doesn't actually change the target file.
# If you need to do that, redirect its output.
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="EX4"
></A
><P
><B
>Example 36-2. A slightly more complex <I
CLASS="FIRSTTERM"
>shell
wrapper</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# subst.sh: a script that substitutes one pattern for
#+ another in a file,
#+ i.e., "sh subst.sh Smith Jones letter.txt".
# Jones replaces Smith.
ARGS=3 # Script requires 3 arguments.
E_BADARGS=85 # Wrong number of arguments passed to script.
if [ $# -ne "$ARGS" ]
then
echo "Usage: `basename $0` old-pattern new-pattern filename"
exit $E_BADARGS
fi
old_pattern=$1
new_pattern=$2
if [ -f "$3" ]
then
file_name=$3
else
echo "File \"$3\" does not exist."
exit $E_BADARGS
fi
# -----------------------------------------------
# Here is where the heavy work gets done.
sed -e "s/$old_pattern/$new_pattern/g" $file_name
# -----------------------------------------------
# 's' is, of course, the substitute command in sed,
#+ and /pattern/ invokes address matching.
# The 'g,' or global flag causes substitution for EVERY
#+ occurence of $old_pattern on each line, not just the first.
# Read the 'sed' docs for an in-depth explanation.
exit $? # Redirect the output of this script to write to a file.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="LOGGINGWRAPPER"
></A
><P
><B
>Example 36-3. A generic <I
CLASS="FIRSTTERM"
>shell wrapper</I
> that
writes to a logfile</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# logging-wrapper.sh
# Generic shell wrapper that performs an operation
#+ and logs it.
DEFAULT_LOGFILE=logfile.txt
# Set the following two variables.
OPERATION=
# Can be a complex chain of commands,
#+ for example an awk script or a pipe . . .
LOGFILE=
if [ -z "$LOGFILE" ]
then # If not set, default to ...
LOGFILE="$DEFAULT_LOGFILE"
fi
# Command-line arguments, if any, for the operation.
OPTIONS="$@"
# Log it.
echo "`date` + `whoami` + $OPERATION "$@"" &#62;&#62; $LOGFILE
# Now, do it.
exec $OPERATION "$@"
# It's necessary to do the logging before the operation.
# Why?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="PRASC"
></A
><P
><B
>Example 36-4. A <I
CLASS="FIRSTTERM"
>shell wrapper</I
> around an awk
script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# pr-ascii.sh: Prints a table of ASCII characters.
START=33 # Range of printable ASCII characters (decimal).
END=127 # Will not work for unprintable characters (&#62; 127).
echo " Decimal Hex Character" # Header.
echo " ------- --- ---------"
for ((i=START; i&#60;=END; i++))
do
echo $i | awk '{printf(" %3d %2x %c\n", $1, $1, $1)}'
# The Bash printf builtin will not work in this context:
# printf "%c" "$i"
done
exit 0
# Decimal Hex Character
# ------- --- ---------
# 33 21 !
# 34 22 "
# 35 23 #
# 36 24 $
#
# . . .
#
# 122 7a z
# 123 7b {
# 124 7c |
# 125 7d }
# Redirect the output of this script to a file
#+ or pipe it to "more": sh pr-asc.sh | more</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="COLTOTALER"
></A
><P
><B
>Example 36-5. A <I
CLASS="FIRSTTERM"
>shell wrapper</I
> around another
awk script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Adds up a specified column (of numbers) in the target file.
# Floating-point (decimal) numbers okay, because awk can handle them.
ARGS=2
E_WRONGARGS=85
if [ $# -ne "$ARGS" ] # Check for proper number of command-line args.
then
echo "Usage: `basename $0` filename column-number"
exit $E_WRONGARGS
fi
filename=$1
column_number=$2
# Passing shell variables to the awk part of the script is a bit tricky.
# One method is to strong-quote the Bash-script variable
#+ within the awk script.
# $'$BASH_SCRIPT_VAR'
# ^ ^
# This is done in the embedded awk script below.
# See the awk documentation for more details.
# A multi-line awk script is here invoked by
# awk '
# ...
# ...
# ...
# '
# Begin awk script.
# -----------------------------
awk '
{ total += $'"${column_number}"'
}
END {
print total
}
' "$filename"
# -----------------------------
# End awk script.
# It may not be safe to pass shell variables to an embedded awk script,
#+ so Stephane Chazelas proposes the following alternative:
# ---------------------------------------
# awk -v column_number="$column_number" '
# { total += $column_number
# }
# END {
# print total
# }' "$filename"
# ---------------------------------------
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="PERLREF"
></A
>For those scripts needing a single
do-it-all tool, a Swiss army knife, there is
<I
CLASS="FIRSTTERM"
>Perl</I
>. Perl combines the
capabilities of <A
HREF="#SEDREF"
>sed</A
> and <A
HREF="#AWKREF"
>awk</A
>, and throws in a large subset of
<B
CLASS="COMMAND"
>C</B
>, to boot. It is modular and contains support
for everything ranging from object-oriented programming up to and
including the kitchen sink. Short Perl scripts lend themselves to
embedding within shell scripts, and there may be some substance
to the claim that Perl can totally replace shell scripting
(though the author of the <EM
>ABS Guide</EM
> remains
skeptical).</P
><P
><A
NAME="PERLEMB"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX56"
></A
><P
><B
>Example 36-6. Perl embedded in a <I
CLASS="FIRSTTERM"
>Bash</I
> script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Shell commands may precede the Perl script.
echo "This precedes the embedded Perl script within \"$0\"."
echo "==============================================================="
perl -e 'print "This line prints from an embedded Perl script.\n";'
# Like sed, Perl also uses the "-e" option.
echo "==============================================================="
echo "However, the script may also contain shell and system commands."
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>It is even possible to combine a Bash script and Perl script
within the same file. Depending on how the script is invoked, either
the Bash part or the Perl part will execute.</P
><P
><A
NAME="BASHANDPERL0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="BASHANDPERL"
></A
><P
><B
>Example 36-7. Bash and Perl scripts combined</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# bashandperl.sh
echo "Greetings from the Bash part of the script, $0."
# More Bash commands may follow here.
exit
# End of Bash part of the script.
# =======================================================
#!/usr/bin/perl
# This part of the script must be invoked with
# perl -x bashandperl.sh
print "Greetings from the Perl part of the script, $0.\n";
# Perl doesn't seem to like "echo" ...
# More Perl commands may follow here.
# End of Perl part of the script.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>bash bashandperl.sh</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Greetings from the Bash part of the script.</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>perl -x bashandperl.sh</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>Greetings from the Perl part of the script.</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>It is, of course, possible to embed even more exotic scripting
languages within shell wrappers. <I
CLASS="FIRSTTERM"
>Python</I
>,
for example ...</P
><P
><A
NAME="PYTHONEMB"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX56PY"
></A
><P
><B
>Example 36-8. Python embedded in a <I
CLASS="FIRSTTERM"
>Bash</I
> script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ex56py.sh
# Shell commands may precede the Python script.
echo "This precedes the embedded Python script within \"$0.\""
echo "==============================================================="
python -c 'print "This line prints from an embedded Python script.\n";'
# Unlike sed and perl, Python uses the "-c" option.
python -c 'k = raw_input( "Hit a key to exit to outer script. " )'
echo "==============================================================="
echo "However, the script may also contain shell and system commands."
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Wrapping a script around <I
CLASS="FIRSTTERM"
>mplayer</I
>
and the Google's translation server, you can create something
that talks back to you.</P
><P
><A
NAME="SPEECH00"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="SPEECH0"
></A
><P
><B
>Example 36-9. A script that speaks</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Courtesy of:
# http://elinux.org/RPi_Text_to_Speech_(Speech_Synthesis)
# You must be on-line for this script to work,
#+ so you can access the Google translation server.
# Of course, mplayer must be present on your computer.
speak()
{
local IFS=+
# Invoke mplayer, then connect to Google translation server.
/usr/bin/mplayer -ao alsa -really-quiet -noconsolecontrols \
"http://translate.google.com/translate_tts?tl=en&#38;q="$*""
# Google translates, but can also speak.
}
LINES=4
spk=$(tail -$LINES $0) # Tail end of same script!
speak "$spk"
exit
# Browns. Nice talking to you.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>One interesting example of a complex shell wrapper is Martin
Matusiak's <A
HREF="http://sourceforge.net/projects/undvd/"
TARGET="_top"
><I
CLASS="FIRSTTERM"
>undvd</I
>
script</A
>, which provides an easy-to-use
command-line interface to the complex <A
HREF="http://www.mplayerhq.hu/DOCS/HTML/en/mencoder.html"
TARGET="_top"
>mencoder</A
>
utility. Another example is Itzchak Rehberg's <A
HREF="http://projects.izzysoft.de/trac/ext3undel"
TARGET="_top"
>Ext3Undel</A
>,
a set of scripts to recover deleted file on an
<I
CLASS="FIRSTTERM"
>ext3</I
> filesystem.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="TESTSANDCOMPARISONS"
></A
>36.3. Tests and Comparisons: Alternatives</H1
><P
>For tests, the <A
HREF="#DBLBRACKETS"
>[[ ]]</A
>
construct may be more appropriate than <TT
CLASS="USERINPUT"
><B
>[
]</B
></TT
>. Likewise, <A
HREF="#ICOMPARISON1"
>arithmetic
comparisons</A
> might benefit from the <A
HREF="#DBLPARENS"
>(( ))</A
> construct.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>a=8
# All of the comparisons below are equivalent.
test "$a" -lt 16 &#38;&#38; echo "yes, $a &#60; 16" # "and list"
/bin/test "$a" -lt 16 &#38;&#38; echo "yes, $a &#60; 16"
[ "$a" -lt 16 ] &#38;&#38; echo "yes, $a &#60; 16"
[[ $a -lt 16 ]] &#38;&#38; echo "yes, $a &#60; 16" # Quoting variables within
(( a &#60; 16 )) &#38;&#38; echo "yes, $a &#60; 16" # [[ ]] and (( )) not necessary.
city="New York"
# Again, all of the comparisons below are equivalent.
test "$city" \&#60; Paris &#38;&#38; echo "Yes, Paris is greater than $city"
# Greater ASCII order.
/bin/test "$city" \&#60; Paris &#38;&#38; echo "Yes, Paris is greater than $city"
[ "$city" \&#60; Paris ] &#38;&#38; echo "Yes, Paris is greater than $city"
[[ $city &#60; Paris ]] &#38;&#38; echo "Yes, Paris is greater than $city"
# Need not quote $city.
# Thank you, S.C.</PRE
></FONT
></TD
></TR
></TABLE
></P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="RECURSIONSCT"
></A
>36.4. Recursion: a script calling itself</H1
><P
><A
NAME="SCRIPTRECURSION"
></A
></P
><P
>Can a script <A
HREF="#RECURSIONREF"
>recursively</A
>
call itself? Indeed.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="RECURSE"
></A
><P
><B
>Example 36-10. A (useless) script that recursively calls itself</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# recurse.sh
# Can a script recursively call itself?
# Yes, but is this of any practical use?
# (See the following.)
RANGE=10
MAXVAL=9
i=$RANDOM
let "i %= $RANGE" # Generate a random number between 0 and $RANGE - 1.
if [ "$i" -lt "$MAXVAL" ]
then
echo "i = $i"
./$0 # Script recursively spawns a new instance of itself.
fi # Each child script does the same, until
#+ a generated $i equals $MAXVAL.
# Using a "while" loop instead of an "if/then" test causes problems.
# Explain why.
exit 0
# Note:
# ----
# This script must have execute permission for it to work properly.
# This is the case even if it is invoked by an "sh" command.
# Explain why.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="PBOOK"
></A
><P
><B
>Example 36-11. A (useful) script that recursively calls itself</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# pb.sh: phone book
# Written by Rick Boivie, and used with permission.
# Modifications by ABS Guide author.
MINARGS=1 # Script needs at least one argument.
DATAFILE=./phonebook
# A data file in current working directory
#+ named "phonebook" must exist.
PROGNAME=$0
E_NOARGS=70 # No arguments error.
if [ $# -lt $MINARGS ]; then
echo "Usage: "$PROGNAME" data-to-look-up"
exit $E_NOARGS
fi
if [ $# -eq $MINARGS ]; then
grep $1 "$DATAFILE"
# 'grep' prints an error message if $DATAFILE not present.
else
( shift; "$PROGNAME" $* ) | grep $1
# Script recursively calls itself.
fi
exit 0 # Script exits here.
# Therefore, it's o.k. to put
#+ non-hashmarked comments and data after this point.
# ------------------------------------------------------------------------
Sample "phonebook" datafile:
John Doe 1555 Main St., Baltimore, MD 21228 (410) 222-3333
Mary Moe 9899 Jones Blvd., Warren, NH 03787 (603) 898-3232
Richard Roe 856 E. 7th St., New York, NY 10009 (212) 333-4567
Sam Roe 956 E. 8th St., New York, NY 10009 (212) 444-5678
Zoe Zenobia 4481 N. Baker St., San Francisco, SF 94338 (415) 501-1631
# ------------------------------------------------------------------------
$bash pb.sh Roe
Richard Roe 856 E. 7th St., New York, NY 10009 (212) 333-4567
Sam Roe 956 E. 8th St., New York, NY 10009 (212) 444-5678
$bash pb.sh Roe Sam
Sam Roe 956 E. 8th St., New York, NY 10009 (212) 444-5678
# When more than one argument is passed to this script,
#+ it prints *only* the line(s) containing all the arguments.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="USRMNT"
></A
><P
><B
>Example 36-12. Another (useful) script that recursively calls itself</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# usrmnt.sh, written by Anthony Richardson
# Used in ABS Guide with permission.
# usage: usrmnt.sh
# description: mount device, invoking user must be listed in the
# MNTUSERS group in the /etc/sudoers file.
# ----------------------------------------------------------
# This is a usermount script that reruns itself using sudo.
# A user with the proper permissions only has to type
# usermount /dev/fd0 /mnt/floppy
# instead of
# sudo usermount /dev/fd0 /mnt/floppy
# I use this same technique for all of my
#+ sudo scripts, because I find it convenient.
# ----------------------------------------------------------
# If SUDO_COMMAND variable is not set we are not being run through
#+ sudo, so rerun ourselves. Pass the user's real and group id . . .
if [ -z "$SUDO_COMMAND" ]
then
mntusr=$(id -u) grpusr=$(id -g) sudo $0 $*
exit 0
fi
# We will only get here if we are being run by sudo.
/bin/mount $* -o uid=$mntusr,gid=$grpusr
exit 0
# Additional notes (from the author of this script):
# -------------------------------------------------
# 1) Linux allows the "users" option in the /etc/fstab
# file so that any user can mount removable media.
# But, on a server, I like to allow only a few
# individuals access to removable media.
# I find using sudo gives me more control.
# 2) I also find sudo to be more convenient than
# accomplishing this task through groups.
# 3) This method gives anyone with proper permissions
# root access to the mount command, so be careful
# about who you allow access.
# You can get finer control over which access can be mounted
# by using this same technique in separate mntfloppy, mntcdrom,
# and mntsamba scripts.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Too many levels of recursion can exhaust the
script's stack space, causing a segfault.</P
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="COLORIZING"
></A
>36.5. <SPAN
CLASS="QUOTE"
>"Colorizing"</SPAN
> Scripts</H1
><P
><A
NAME="COLORIZINGREF"
></A
></P
><P
>The ANSI
<A
NAME="AEN20259"
HREF="#FTN.AEN20259"
><SPAN
CLASS="footnote"
>[126]</SPAN
></A
>
escape sequences set screen attributes, such as bold
text, and color of foreground and background. <A
HREF="#DOSBATCH1"
>DOS batch files</A
> commonly used
ANSI escape codes for <EM
>color</EM
> output,
and so can Bash scripts.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX30A"
></A
><P
><B
>Example 36-13. A <SPAN
CLASS="QUOTE"
>"colorized"</SPAN
> address database</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ex30a.sh: "Colorized" version of ex30.sh.
# Crude address database
clear # Clear the screen.
echo -n " "
echo -e '\E[37;44m'"\033[1mContact List\033[0m"
# White on blue background
echo; echo
echo -e "\033[1mChoose one of the following persons:\033[0m"
# Bold
tput sgr0 # Reset attributes.
echo "(Enter only the first letter of name.)"
echo
echo -en '\E[47;34m'"\033[1mE\033[0m" # Blue
tput sgr0 # Reset colors to "normal."
echo "vans, Roland" # "[E]vans, Roland"
echo -en '\E[47;35m'"\033[1mJ\033[0m" # Magenta
tput sgr0
echo "ambalaya, Mildred"
echo -en '\E[47;32m'"\033[1mS\033[0m" # Green
tput sgr0
echo "mith, Julie"
echo -en '\E[47;31m'"\033[1mZ\033[0m" # Red
tput sgr0
echo "ane, Morris"
echo
read person
case "$person" in
# Note variable is quoted.
"E" | "e" )
# Accept upper or lowercase input.
echo
echo "Roland Evans"
echo "4321 Flash Dr."
echo "Hardscrabble, CO 80753"
echo "(303) 734-9874"
echo "(303) 734-9892 fax"
echo "revans@zzy.net"
echo "Business partner &#38; old friend"
;;
"J" | "j" )
echo
echo "Mildred Jambalaya"
echo "249 E. 7th St., Apt. 19"
echo "New York, NY 10009"
echo "(212) 533-2814"
echo "(212) 533-9972 fax"
echo "milliej@loisaida.com"
echo "Girlfriend"
echo "Birthday: Feb. 11"
;;
# Add info for Smith &#38; Zane later.
* )
# Default option.
# Empty input (hitting RETURN) fits here, too.
echo
echo "Not yet in database."
;;
esac
tput sgr0 # Reset colors to "normal."
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="DRAW-BOX"
></A
><P
><B
>Example 36-14. Drawing a box</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Draw-box.sh: Drawing a box using ASCII characters.
# Script by Stefano Palmeri, with minor editing by document author.
# Minor edits suggested by Jim Angstadt.
# Used in the ABS Guide with permission.
######################################################################
### draw_box function doc ###
# The "draw_box" function lets the user
#+ draw a box in a terminal.
#
# Usage: draw_box ROW COLUMN HEIGHT WIDTH [COLOR]
# ROW and COLUMN represent the position
#+ of the upper left angle of the box you're going to draw.
# ROW and COLUMN must be greater than 0
#+ and less than current terminal dimension.
# HEIGHT is the number of rows of the box, and must be &#62; 0.
# HEIGHT + ROW must be &#60;= than current terminal height.
# WIDTH is the number of columns of the box and must be &#62; 0.
# WIDTH + COLUMN must be &#60;= than current terminal width.
#
# E.g.: If your terminal dimension is 20x80,
# draw_box 2 3 10 45 is good
# draw_box 2 3 19 45 has bad HEIGHT value (19+2 &#62; 20)
# draw_box 2 3 18 78 has bad WIDTH value (78+3 &#62; 80)
#
# COLOR is the color of the box frame.
# This is the 5th argument and is optional.
# 0=black 1=red 2=green 3=tan 4=blue 5=purple 6=cyan 7=white.
# If you pass the function bad arguments,
#+ it will just exit with code 65,
#+ and no messages will be printed on stderr.
#
# Clear the terminal before you start to draw a box.
# The clear command is not contained within the function.
# This allows the user to draw multiple boxes, even overlapping ones.
### end of draw_box function doc ###
######################################################################
draw_box(){
#=============#
HORZ="-"
VERT="|"
CORNER_CHAR="+"
MINARGS=4
E_BADARGS=65
#=============#
if [ $# -lt "$MINARGS" ]; then # If args are less than 4, exit.
exit $E_BADARGS
fi
# Looking for non digit chars in arguments.
# Probably it could be done better (exercise for the reader?).
if echo $@ | tr -d [:blank:] | tr -d [:digit:] | grep . &#38;&#62; /dev/null; then
exit $E_BADARGS
fi
BOX_HEIGHT=`expr $3 - 1` # -1 correction needed because angle char "+"
BOX_WIDTH=`expr $4 - 1` #+ is a part of both box height and width.
T_ROWS=`tput lines` # Define current terminal dimension
T_COLS=`tput cols` #+ in rows and columns.
if [ $1 -lt 1 ] || [ $1 -gt $T_ROWS ]; then # Start checking if arguments
exit $E_BADARGS #+ are correct.
fi
if [ $2 -lt 1 ] || [ $2 -gt $T_COLS ]; then
exit $E_BADARGS
fi
if [ `expr $1 + $BOX_HEIGHT + 1` -gt $T_ROWS ]; then
exit $E_BADARGS
fi
if [ `expr $2 + $BOX_WIDTH + 1` -gt $T_COLS ]; then
exit $E_BADARGS
fi
if [ $3 -lt 1 ] || [ $4 -lt 1 ]; then
exit $E_BADARGS
fi # End checking arguments.
plot_char(){ # Function within a function.
echo -e "\E[${1};${2}H"$3
}
echo -ne "\E[3${5}m" # Set box frame color, if defined.
# start drawing the box
count=1 # Draw vertical lines using
for (( r=$1; count&#60;=$BOX_HEIGHT; r++)); do #+ plot_char function.
plot_char $r $2 $VERT
let count=count+1
done
count=1
c=`expr $2 + $BOX_WIDTH`
for (( r=$1; count&#60;=$BOX_HEIGHT; r++)); do
plot_char $r $c $VERT
let count=count+1
done
count=1 # Draw horizontal lines using
for (( c=$2; count&#60;=$BOX_WIDTH; c++)); do #+ plot_char function.
plot_char $1 $c $HORZ
let count=count+1
done
count=1
r=`expr $1 + $BOX_HEIGHT`
for (( c=$2; count&#60;=$BOX_WIDTH; c++)); do
plot_char $r $c $HORZ
let count=count+1
done
plot_char $1 $2 $CORNER_CHAR # Draw box angles.
plot_char $1 `expr $2 + $BOX_WIDTH` $CORNER_CHAR
plot_char `expr $1 + $BOX_HEIGHT` $2 $CORNER_CHAR
plot_char `expr $1 + $BOX_HEIGHT` `expr $2 + $BOX_WIDTH` $CORNER_CHAR
echo -ne "\E[0m" # Restore old colors.
P_ROWS=`expr $T_ROWS - 1` # Put the prompt at bottom of the terminal.
echo -e "\E[${P_ROWS};1H"
}
# Now, let's try drawing a box.
clear # Clear the terminal.
R=2 # Row
C=3 # Column
H=10 # Height
W=45 # Width
col=1 # Color (red)
draw_box $R $C $H $W $col # Draw the box.
exit 0
# Exercise:
# --------
# Add the option of printing text within the drawn box.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The simplest, and perhaps most useful ANSI escape sequence is
bold text, <B
CLASS="COMMAND"
>\033[1m ... \033[0m</B
>. The
<SPAN
CLASS="TOKEN"
>\033</SPAN
> represents an <A
HREF="#ESCP"
>escape</A
>, the <SPAN
CLASS="QUOTE"
>"[1"</SPAN
> turns on the
bold attribute, while the <SPAN
CLASS="QUOTE"
>"[0"</SPAN
> switches it off. The
<SPAN
CLASS="QUOTE"
>"m"</SPAN
> terminates each term of the escape sequence.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo -e "\033[1mThis is bold text.\033[0m"</B
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>A similar escape sequence switches on the underline
attribute (on an <I
CLASS="FIRSTTERM"
>rxvt</I
> and an
<I
CLASS="FIRSTTERM"
>aterm</I
>).
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo -e "\033[4mThis is underlined text.\033[0m"</B
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>With an <B
CLASS="COMMAND"
>echo</B
>, the
<TT
CLASS="OPTION"
>-e</TT
> option enables the escape
sequences.</P
></TD
></TR
></TABLE
></DIV
><P
>Other escape sequences change the text and/or background
color.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo -e '\E[34;47mThis prints in blue.'; tput sgr0</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo -e '\E[33;44m'"yellow text on blue background"; tput sgr0</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo -e '\E[1;33;44m'"BOLD yellow text on blue background"; tput sgr0</B
></TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>It's usually advisable to set the
<I
CLASS="FIRSTTERM"
>bold</I
> attribute for light-colored foreground
text.</P
></TD
></TR
></TABLE
></DIV
><P
>The <B
CLASS="COMMAND"
>tput sgr0</B
> restores the
terminal settings to normal. Omitting this lets all
subsequent output from that particular terminal remain
blue.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Since <B
CLASS="COMMAND"
>tput sgr0</B
> fails to restore
terminal settings under certain circumstances,
<B
CLASS="COMMAND"
>echo -ne \E[0m</B
> may be a better choice.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="COLORIZTEMPL"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN20311"
></A
><P
></P
><P
>Use the following template for writing colored text on a colored
background.</P
><P
> <TT
CLASS="USERINPUT"
><B
>echo -e '\E[COLOR1;COLOR2mSome text goes here.'</B
></TT
>
</P
><P
>The <SPAN
CLASS="QUOTE"
>"\E["</SPAN
> begins the escape sequence.
The semicolon-separated numbers <SPAN
CLASS="QUOTE"
>"COLOR1"</SPAN
> and
<SPAN
CLASS="QUOTE"
>"COLOR2"</SPAN
> specify a foreground and a background
color, according to the table below. (The order of the
numbers does not matter, since the foreground and background
numbers fall in non-overlapping ranges.) The <SPAN
CLASS="QUOTE"
>"m"</SPAN
>
terminates the escape sequence, and the text begins immediately
after that.</P
><P
>Note also that <A
HREF="#SNGLQUO"
>single quotes</A
>
enclose the remainder of the command sequence following the
<B
CLASS="COMMAND"
>echo -e</B
>.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
>The numbers in the following table work for an
<I
CLASS="FIRSTTERM"
>rxvt</I
> terminal. Results may vary for other
terminal emulators.</P
><P
><A
NAME="COLORIZTABLE"
></A
></P
><DIV
CLASS="TABLE"
><A
NAME="AEN20327"
></A
><P
><B
>Table 36-1. Numbers representing colors in Escape Sequences</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Color</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Foreground</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Background</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>black</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>30</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>40</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>red</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>31</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>41</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>green</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>32</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>42</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>yellow</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>33</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>43</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>blue</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>34</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>44</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>magenta</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>35</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>45</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>cyan</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>36</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>46</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>white</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>37</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>47</TD
></TR
></TBODY
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="COLORECHO"
></A
><P
><B
>Example 36-15. Echoing colored text</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# color-echo.sh: Echoing text messages in color.
# Modify this script for your own purposes.
# It's easier than hand-coding color.
black='\E[30;47m'
red='\E[31;47m'
green='\E[32;47m'
yellow='\E[33;47m'
blue='\E[34;47m'
magenta='\E[35;47m'
cyan='\E[36;47m'
white='\E[37;47m'
alias Reset="tput sgr0" # Reset text attributes to normal
#+ without clearing screen.
cecho () # Color-echo.
# Argument $1 = message
# Argument $2 = color
{
local default_msg="No message passed."
# Doesn't really need to be a local variable.
message=${1:-$default_msg} # Defaults to default message.
color=${2:-$black} # Defaults to black, if not specified.
echo -e "$color"
echo "$message"
Reset # Reset to normal.
return
}
# Now, let's try it out.
# ----------------------------------------------------
cecho "Feeling blue..." $blue
cecho "Magenta looks more like purple." $magenta
cecho "Green with envy." $green
cecho "Seeing red?" $red
cecho "Cyan, more familiarly known as aqua." $cyan
cecho "No color passed (defaults to black)."
# Missing $color argument.
cecho "\"Empty\" color passed (defaults to black)." ""
# Empty $color argument.
cecho
# Missing $message and $color arguments.
cecho "" ""
# Empty $message and $color arguments.
# ----------------------------------------------------
echo
exit 0
# Exercises:
# ---------
# 1) Add the "bold" attribute to the 'cecho ()' function.
# 2) Add options for colored backgrounds.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="HORSERACEREF"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="HORSERACE"
></A
><P
><B
>Example 36-16. A <SPAN
CLASS="QUOTE"
>"horserace"</SPAN
> game</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# horserace.sh: Very simple horserace simulation.
# Author: Stefano Palmeri
# Used with permission.
################################################################
# Goals of the script:
# playing with escape sequences and terminal colors.
#
# Exercise:
# Edit the script to make it run less randomly,
#+ set up a fake betting shop . . .
# Um . . . um . . . it's starting to remind me of a movie . . .
#
# The script gives each horse a random handicap.
# The odds are calculated upon horse handicap
#+ and are expressed in European(?) style.
# E.g., odds=3.75 means that if you bet $1 and win,
#+ you receive $3.75.
#
# The script has been tested with a GNU/Linux OS,
#+ using xterm and rxvt, and konsole.
# On a machine with an AMD 900 MHz processor,
#+ the average race time is 75 seconds.
# On faster computers the race time would be lower.
# So, if you want more suspense, reset the USLEEP_ARG variable.
#
# Script by Stefano Palmeri.
################################################################
E_RUNERR=65
# Check if md5sum and bc are installed.
if ! which bc &#38;&#62; /dev/null; then
echo bc is not installed.
echo "Can\'t run . . . "
exit $E_RUNERR
fi
if ! which md5sum &#38;&#62; /dev/null; then
echo md5sum is not installed.
echo "Can\'t run . . . "
exit $E_RUNERR
fi
# Set the following variable to slow down script execution.
# It will be passed as the argument for usleep (man usleep)
#+ and is expressed in microseconds (500000 = half a second).
USLEEP_ARG=0
# Clean up the temp directory, restore terminal cursor and
#+ terminal colors -- if script interrupted by Ctl-C.
trap 'echo -en "\E[?25h"; echo -en "\E[0m"; stty echo;\
tput cup 20 0; rm -fr $HORSE_RACE_TMP_DIR' TERM EXIT
# See the chapter on debugging for an explanation of 'trap.'
# Set a unique (paranoid) name for the temp directory the script needs.
HORSE_RACE_TMP_DIR=$HOME/.horserace-`date +%s`-`head -c10 /dev/urandom \
| md5sum | head -c30`
# Create the temp directory and move right in.
mkdir $HORSE_RACE_TMP_DIR
cd $HORSE_RACE_TMP_DIR
# This function moves the cursor to line $1 column $2 and then prints $3.
# E.g.: "move_and_echo 5 10 linux" is equivalent to
#+ "tput cup 4 9; echo linux", but with one command instead of two.
# Note: "tput cup" defines 0 0 the upper left angle of the terminal,
#+ echo defines 1 1 the upper left angle of the terminal.
move_and_echo() {
echo -ne "\E[${1};${2}H""$3"
}
# Function to generate a pseudo-random number between 1 and 9.
random_1_9 ()
{
head -c10 /dev/urandom | md5sum | tr -d [a-z] | tr -d 0 | cut -c1
}
# Two functions that simulate "movement," when drawing the horses.
draw_horse_one() {
echo -n " "//$MOVE_HORSE//
}
draw_horse_two(){
echo -n " "\\\\$MOVE_HORSE\\\\
}
# Define current terminal dimension.
N_COLS=`tput cols`
N_LINES=`tput lines`
# Need at least a 20-LINES X 80-COLUMNS terminal. Check it.
if [ $N_COLS -lt 80 ] || [ $N_LINES -lt 20 ]; then
echo "`basename $0` needs a 80-cols X 20-lines terminal."
echo "Your terminal is ${N_COLS}-cols X ${N_LINES}-lines."
exit $E_RUNERR
fi
# Start drawing the race field.
# Need a string of 80 chars. See below.
BLANK80=`seq -s "" 100 | head -c80`
clear
# Set foreground and background colors to white.
echo -ne '\E[37;47m'
# Move the cursor on the upper left angle of the terminal.
tput cup 0 0
# Draw six white lines.
for n in `seq 5`; do
echo $BLANK80 # Use the 80 chars string to colorize the terminal.
done
# Sets foreground color to black.
echo -ne '\E[30m'
move_and_echo 3 1 "START 1"
move_and_echo 3 75 FINISH
move_and_echo 1 5 "|"
move_and_echo 1 80 "|"
move_and_echo 2 5 "|"
move_and_echo 2 80 "|"
move_and_echo 4 5 "| 2"
move_and_echo 4 80 "|"
move_and_echo 5 5 "V 3"
move_and_echo 5 80 "V"
# Set foreground color to red.
echo -ne '\E[31m'
# Some ASCII art.
move_and_echo 1 8 "..@@@..@@@@@...@@@@@.@...@..@@@@..."
move_and_echo 2 8 ".@...@...@.......@...@...@.@......."
move_and_echo 3 8 ".@@@@@...@.......@...@@@@@.@@@@...."
move_and_echo 4 8 ".@...@...@.......@...@...@.@......."
move_and_echo 5 8 ".@...@...@.......@...@...@..@@@@..."
move_and_echo 1 43 "@@@@...@@@...@@@@..@@@@..@@@@."
move_and_echo 2 43 "@...@.@...@.@.....@.....@....."
move_and_echo 3 43 "@@@@..@@@@@.@.....@@@@...@@@.."
move_and_echo 4 43 "@..@..@...@.@.....@.........@."
move_and_echo 5 43 "@...@.@...@..@@@@..@@@@.@@@@.."
# Set foreground and background colors to green.
echo -ne '\E[32;42m'
# Draw eleven green lines.
tput cup 5 0
for n in `seq 11`; do
echo $BLANK80
done
# Set foreground color to black.
echo -ne '\E[30m'
tput cup 5 0
# Draw the fences.
echo "++++++++++++++++++++++++++++++++++++++\
++++++++++++++++++++++++++++++++++++++++++"
tput cup 15 0
echo "++++++++++++++++++++++++++++++++++++++\
++++++++++++++++++++++++++++++++++++++++++"
# Set foreground and background colors to white.
echo -ne '\E[37;47m'
# Draw three white lines.
for n in `seq 3`; do
echo $BLANK80
done
# Set foreground color to black.
echo -ne '\E[30m'
# Create 9 files to stores handicaps.
for n in `seq 10 7 68`; do
touch $n
done
# Set the first type of "horse" the script will draw.
HORSE_TYPE=2
# Create position-file and odds-file for every "horse".
#+ In these files, store the current position of the horse,
#+ the type and the odds.
for HN in `seq 9`; do
touch horse_${HN}_position
touch odds_${HN}
echo \-1 &#62; horse_${HN}_position
echo $HORSE_TYPE &#62;&#62; horse_${HN}_position
# Define a random handicap for horse.
HANDICAP=`random_1_9`
# Check if the random_1_9 function returned a good value.
while ! echo $HANDICAP | grep [1-9] &#38;&#62; /dev/null; do
HANDICAP=`random_1_9`
done
# Define last handicap position for horse.
LHP=`expr $HANDICAP \* 7 + 3`
for FILE in `seq 10 7 $LHP`; do
echo $HN &#62;&#62; $FILE
done
# Calculate odds.
case $HANDICAP in
1) ODDS=`echo $HANDICAP \* 0.25 + 1.25 | bc`
echo $ODDS &#62; odds_${HN}
;;
2 | 3) ODDS=`echo $HANDICAP \* 0.40 + 1.25 | bc`
echo $ODDS &#62; odds_${HN}
;;
4 | 5 | 6) ODDS=`echo $HANDICAP \* 0.55 + 1.25 | bc`
echo $ODDS &#62; odds_${HN}
;;
7 | 8) ODDS=`echo $HANDICAP \* 0.75 + 1.25 | bc`
echo $ODDS &#62; odds_${HN}
;;
9) ODDS=`echo $HANDICAP \* 0.90 + 1.25 | bc`
echo $ODDS &#62; odds_${HN}
esac
done
# Print odds.
print_odds() {
tput cup 6 0
echo -ne '\E[30;42m'
for HN in `seq 9`; do
echo "#$HN odds-&#62;" `cat odds_${HN}`
done
}
# Draw the horses at starting line.
draw_horses() {
tput cup 6 0
echo -ne '\E[30;42m'
for HN in `seq 9`; do
echo /\\$HN/\\" "
done
}
print_odds
echo -ne '\E[47m'
# Wait for a enter key press to start the race.
# The escape sequence '\E[?25l' disables the cursor.
tput cup 17 0
echo -e '\E[?25l'Press [enter] key to start the race...
read -s
# Disable normal echoing in the terminal.
# This avoids key presses that might "contaminate" the screen
#+ during the race.
stty -echo
# --------------------------------------------------------
# Start the race.
draw_horses
echo -ne '\E[37;47m'
move_and_echo 18 1 $BLANK80
echo -ne '\E[30m'
move_and_echo 18 1 Starting...
sleep 1
# Set the column of the finish line.
WINNING_POS=74
# Define the time the race started.
START_TIME=`date +%s`
# COL variable needed by following "while" construct.
COL=0
while [ $COL -lt $WINNING_POS ]; do
MOVE_HORSE=0
# Check if the random_1_9 function has returned a good value.
while ! echo $MOVE_HORSE | grep [1-9] &#38;&#62; /dev/null; do
MOVE_HORSE=`random_1_9`
done
# Define old type and position of the "randomized horse".
HORSE_TYPE=`cat horse_${MOVE_HORSE}_position | tail -n 1`
COL=$(expr `cat horse_${MOVE_HORSE}_position | head -n 1`)
ADD_POS=1
# Check if the current position is an handicap position.
if seq 10 7 68 | grep -w $COL &#38;&#62; /dev/null; then
if grep -w $MOVE_HORSE $COL &#38;&#62; /dev/null; then
ADD_POS=0
grep -v -w $MOVE_HORSE $COL &#62; ${COL}_new
rm -f $COL
mv -f ${COL}_new $COL
else ADD_POS=1
fi
else ADD_POS=1
fi
COL=`expr $COL + $ADD_POS`
echo $COL &#62; horse_${MOVE_HORSE}_position # Store new position.
# Choose the type of horse to draw.
case $HORSE_TYPE in
1) HORSE_TYPE=2; DRAW_HORSE=draw_horse_two
;;
2) HORSE_TYPE=1; DRAW_HORSE=draw_horse_one
esac
echo $HORSE_TYPE &#62;&#62; horse_${MOVE_HORSE}_position
# Store current type.
# Set foreground color to black and background to green.
echo -ne '\E[30;42m'
# Move the cursor to new horse position.
tput cup `expr $MOVE_HORSE + 5` \
`cat horse_${MOVE_HORSE}_position | head -n 1`
# Draw the horse.
$DRAW_HORSE
usleep $USLEEP_ARG
# When all horses have gone beyond field line 15, reprint odds.
touch fieldline15
if [ $COL = 15 ]; then
echo $MOVE_HORSE &#62;&#62; fieldline15
fi
if [ `wc -l fieldline15 | cut -f1 -d " "` = 9 ]; then
print_odds
: &#62; fieldline15
fi
# Define the leading horse.
HIGHEST_POS=`cat *position | sort -n | tail -1`
# Set background color to white.
echo -ne '\E[47m'
tput cup 17 0
echo -n Current leader: `grep -w $HIGHEST_POS *position | cut -c7`\
" "
done
# Define the time the race finished.
FINISH_TIME=`date +%s`
# Set background color to green and enable blinking text.
echo -ne '\E[30;42m'
echo -en '\E[5m'
# Make the winning horse blink.
tput cup `expr $MOVE_HORSE + 5` \
`cat horse_${MOVE_HORSE}_position | head -n 1`
$DRAW_HORSE
# Disable blinking text.
echo -en '\E[25m'
# Set foreground and background color to white.
echo -ne '\E[37;47m'
move_and_echo 18 1 $BLANK80
# Set foreground color to black.
echo -ne '\E[30m'
# Make winner blink.
tput cup 17 0
echo -e "\E[5mWINNER: $MOVE_HORSE\E[25m"" Odds: `cat odds_${MOVE_HORSE}`"\
" Race time: `expr $FINISH_TIME - $START_TIME` secs"
# Restore cursor and old colors.
echo -en "\E[?25h"
echo -en "\E[0m"
# Restore echoing.
stty echo
# Remove race temp directory.
rm -rf $HORSE_RACE_TMP_DIR
tput cup 19 0
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#HASHEXAMPLE"
>Example A-21</A
>, <A
HREF="#HOMEWORK"
>Example A-44</A
>, <A
HREF="#SHOWALLC"
>Example A-52</A
>, and <A
HREF="#PETALS"
>Example A-40</A
>.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>There is, however, a major problem with all
this. <EM
>ANSI escape sequences are emphatically
<A
HREF="#PORTABILITYISSUES"
>non-portable</A
>.</EM
>
What works fine on some terminal emulators (or the
console) may work differently, or not at all, on others.
A <SPAN
CLASS="QUOTE"
>"colorized"</SPAN
> script that looks stunning on the
script author's machine may produce unreadable output on
someone else's. This somewhat compromises the usefulness of
colorizing scripts, and possibly relegates this technique
to the status of a gimmick. Colorized scripts are probably
inappropriate in a commercial setting, i.e., your supervisor
might disapprove.</P
></TD
></TR
></TABLE
></DIV
><P
>Alister's <A
HREF="http://code.google.com/p/ansi-color/"
TARGET="_top"
> ansi-color</A
> utility (based on <A
HREF="http://bash.deta.in/color-1.1.tar.gz"
TARGET="_top"
>Moshe
Jacobson's color utility</A
> considerably simplifies using
ANSI escape sequences. It substitutes a clean and logical
syntax for the clumsy constructs just discussed.</P
><P
>Henry/teikedvl has likewise created a utility (<A
HREF="http://scriptechocolor.sourceforge.net/"
TARGET="_top"
>http://scriptechocolor.sourceforge.net/</A
>) to simplify creation of colorized scripts.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="OPTIMIZATIONS"
></A
>36.6. Optimizations</H1
><P
>Most shell scripts are quick 'n dirty solutions to non-complex
problems. As such, optimizing them for speed is not much of an
issue. Consider the case, though, where a script carries out
an important task, does it well, but runs too slowly. Rewriting
it in a compiled language may not be a palatable option. The
simplest fix would be to rewrite the parts of the script
that slow it down. Is it possible to apply principles of code
optimization even to a lowly shell script?</P
><P
>Check the loops in the script. Time consumed by repetitive
operations adds up quickly. If at all possible, remove
time-consuming operations from within loops.</P
><P
>Use <A
HREF="#BUILTINREF"
>builtin</A
> commands in
preference to system commands. Builtins execute faster and
usually do not launch a subshell when invoked.</P
><P
><A
NAME="CATABUSE"
></A
></P
><P
>Avoid unnecessary commands, particularly in a <A
HREF="#PIPEREF"
>pipe</A
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat "$file" | grep "$word"
grep "$word" "$file"
# The above command-lines have an identical effect,
#+ but the second runs faster since it launches one fewer subprocess.</PRE
></FONT
></TD
></TR
></TABLE
>
The <A
HREF="#CATREF"
>cat</A
> command seems especially
prone to overuse in scripts.</P
><P
><A
NAME="LCALL"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN20414"
></A
><P
></P
><P
>Disabling certain Bash options can speed up scripts.</P
><P
>As Erik Brandsberg points out:</P
><P
>If you don't need <A
HREF="#UNICODEREF"
>Unicode</A
> support, you can
get potentially a 2x or more improvement in speed by
simply setting the <TT
CLASS="USERINPUT"
><B
>LC_ALL</B
></TT
> variable.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> export LC_ALL=C
[specifies the locale as ANSI C,
thereby disabling Unicode support]
[In an example script ...]
Without [Unicode support]:
erik@erik-desktop:~/capture$ time ./cap-ngrep.sh
live2.pcap &#62; out.txt
real 0m20.483s
user 1m34.470s
sys 0m12.869s
With [Unicode support]:
erik@erik-desktop:~/capture$ time ./cap-ngrep.sh
live2.pcap &#62; out.txt
real 0m50.232s
user 3m51.118s
sys 0m11.221s
A large part of the overhead that is optimized is, I believe,
regex match using [[ string =~ REGEX ]],
but it may help with other portions of the code as well.
I hadn't [seen it] mentioned that this optimization helped
with Bash, but I had seen it helped with "grep,"
so why not try?</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
><A
NAME="OPTIMES"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Certain operators, notably <A
HREF="#EXPRREF"
>expr</A
>, are very inefficient
and might be replaced by <A
HREF="#DBLPARENS"
>double
parentheses</A
> arithmetic expansion.
See <A
HREF="#TESTEXECTIME"
>Example A-59</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>Math tests
math via $(( ))
real 0m0.294s
user 0m0.288s
sys 0m0.008s
math via expr:
real 1m17.879s # Much slower!
user 0m3.600s
sys 0m8.765s
math via let:
real 0m0.364s
user 0m0.372s
sys 0m0.000s</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
HREF="#IFTHEN"
>Condition testing</A
>
constructs in scripts deserve close scrutiny. Substitute
<A
HREF="#CASEESAC1"
>case</A
> for <A
HREF="#IFTHEN"
>if-then</A
> constructs and combine tests
when possible, to minimize script execution time. Again,
refer to <A
HREF="#TESTEXECTIME"
>Example A-59</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>Test using "case" construct:
real 0m0.329s
user 0m0.320s
sys 0m0.000s
Test with if [], no quotes:
real 0m0.438s
user 0m0.432s
sys 0m0.008s
Test with if [], quotes:
real 0m0.476s
user 0m0.452s
sys 0m0.024s
Test with if [], using -eq:
real 0m0.457s
user 0m0.456s
sys 0m0.000s</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="ASSOCARRTST"
></A
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Erik Brandsberg recommends using <A
HREF="#ASSOCARR"
>associative arrays</A
> in preference to
conventional numeric-indexed arrays in most cases. When
overwriting values in a numeric array, there is a significant
performance penalty vs. associative arrays. Running a test
script confirms this. See <A
HREF="#ASSOCARRTEST"
>Example A-60</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>Assignment tests
Assigning a simple variable
real 0m0.418s
user 0m0.416s
sys 0m0.004s
Assigning a numeric index array entry
real 0m0.582s
user 0m0.564s
sys 0m0.016s
Overwriting a numeric index array entry
real 0m21.931s
user 0m21.913s
sys 0m0.016s
Linear reading of numeric index array
real 0m0.422s
user 0m0.416s
sys 0m0.004s
Assigning an associative array entry
real 0m1.800s
user 0m1.796s
sys 0m0.004s
Overwriting an associative array entry
real 0m1.798s
user 0m1.784s
sys 0m0.012s
Linear reading an associative array entry
real 0m0.420s
user 0m0.420s
sys 0m0.000s
Assigning a random number to a simple variable
real 0m0.402s
user 0m0.388s
sys 0m0.016s
Assigning a sparse numeric index array entry randomly into 64k cells
real 0m12.678s
user 0m12.649s
sys 0m0.028s
Reading sparse numeric index array entry
real 0m0.087s
user 0m0.084s
sys 0m0.000s
Assigning a sparse associative array entry randomly into 64k cells
real 0m0.698s
user 0m0.696s
sys 0m0.004s
Reading sparse associative index array entry
real 0m0.083s
user 0m0.084s
sys 0m0.000s</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><P
>Use the <A
HREF="#TIMREF"
>time</A
> and <A
HREF="#TIMESREF"
>times</A
> tools to profile
computation-intensive commands. Consider rewriting time-critical
code sections in C, or even in assembler.</P
><P
>Try to minimize file I/O. Bash is not particularly
efficient at handling files, so consider using
more appropriate tools for this within the script,
such as <A
HREF="#AWKREF"
>awk</A
> or <A
HREF="#PERLREF"
>Perl</A
>.</P
><P
>Write your scripts in a modular and coherent form,
<A
NAME="AEN20452"
HREF="#FTN.AEN20452"
><SPAN
CLASS="footnote"
>[127]</SPAN
></A
>
so they can be reorganized and tightened up as necessary. Some
of the optimization techniques applicable to high-level
languages may work for scripts, but others, such as
<I
CLASS="FIRSTTERM"
>loop unrolling</I
>, are mostly
irrelevant. Above all, use common sense.</P
><P
>For an excellent demonstration of how optimization can
dramatically reduce the execution time of a script, see <A
HREF="#MONTHLYPMT"
>Example 16-47</A
>.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="ASSORTEDTIPS"
></A
>36.7. Assorted Tips</H1
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="AEN20460"
></A
>36.7.1. Ideas for more powerful scripts</H2
><P
></P
><UL
><LI
><P
><A
NAME="PSEUDOCODEREF"
></A
></P
><P
>You have a problem that you want to solve by writing a Bash
script. Unfortunately, you don't know quite where to start.
One method is to plunge right in and code those parts
of the script that come easily, and write the hard parts as
<I
CLASS="FIRSTTERM"
>pseudo-code</I
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
ARGCOUNT=1 # Need name as argument.
E_WRONGARGS=65
if [ number-of-arguments is-not-equal-to "$ARGCOUNT" ]
# ^^^^^^^^^^^^^^^^^^^ ^^^^^^^^^^^^^^^
# Can't figure out how to code this . . .
#+ . . . so write it in pseudo-code.
then
echo "Usage: name-of-script name"
# ^^^^^^^^^^^^^^ More pseudo-code.
exit $E_WRONGARGS
fi
. . .
exit 0
# Later on, substitute working code for the pseudo-code.
# Line 6 becomes:
if [ $# -ne "$ARGCOUNT" ]
# Line 12 becomes:
echo "Usage: `basename $0` name"</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>For an example of using pseudo-code, see the <A
HREF="#NEWTONSQRT"
>Square Root</A
> exercise.</P
></LI
><LI
><P
><A
NAME="TRACKINGSCR"
></A
></P
><P
>To keep a record of which user scripts have run
during a particular session or over a number of sessions,
add the following lines to each script you want to keep track
of. This will keep a continuing file record of the script
names and invocation times. </P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Append (&#62;&#62;) following to end of each script tracked.
whoami&#62;&#62; $SAVE_FILE # User invoking the script.
echo $0&#62;&#62; $SAVE_FILE # Script name.
date&#62;&#62; $SAVE_FILE # Date and time.
echo&#62;&#62; $SAVE_FILE # Blank line as separator.
# Of course, SAVE_FILE defined and exported as environmental variable in ~/.bashrc
#+ (something like ~/.scripts-run)</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><A
NAME="PREPENDREF"
></A
></P
><P
>The <SPAN
CLASS="TOKEN"
>&#62;&#62;</SPAN
> operator
<I
CLASS="FIRSTTERM"
>appends</I
> lines to a file.
What if you wish to <I
CLASS="FIRSTTERM"
>prepend</I
> a
line to an existing file, that is, to paste it in at the
beginning?</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>file=data.txt
title="***This is the title line of data text file***"
echo $title | cat - $file &#62;$file.new
# "cat -" concatenates stdout to $file.
# End result is
#+ to write a new file with $title appended at *beginning*.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>This is a simplified variant of the <A
HREF="#PREPENDEX"
>Example 19-13</A
> script given earlier. And, of course,
<A
HREF="#SEDREF"
>sed</A
> can also do this.</P
></LI
><LI
><P
><A
NAME="SCRIPTASEMB"
></A
></P
><P
>A shell script may act as an embedded command inside
another shell script, a <I
CLASS="FIRSTTERM"
>Tcl</I
> or
<I
CLASS="FIRSTTERM"
>wish</I
> script, or even a <A
HREF="#MAKEFILEREF"
>Makefile</A
>. It can be invoked
as an external shell command in a C program using the
<TT
CLASS="REPLACEABLE"
><I
>system()</I
></TT
> call, i.e.,
<TT
CLASS="REPLACEABLE"
><I
>system("script_name");</I
></TT
>.</P
></LI
><LI
><P
><A
NAME="SETVAREMB"
></A
></P
><P
>Setting a variable to the contents of an embedded
<I
CLASS="FIRSTTERM"
>sed</I
> or <I
CLASS="FIRSTTERM"
>awk</I
>
script increases the readability of the surrounding <A
HREF="#SHWRAPPER"
>shell wrapper</A
>. See <A
HREF="#MAILFORMAT"
>Example A-1</A
> and <A
HREF="#COLTOTALER3"
>Example 15-20</A
>.</P
></LI
><LI
><P
><A
NAME="LIBROUTINES"
></A
></P
><P
>Put together files containing your favorite and most useful
definitions and functions. As necessary,
<SPAN
CLASS="QUOTE"
>"include"</SPAN
> one or more of these
<SPAN
CLASS="QUOTE"
>"library files"</SPAN
> in scripts with either the
<A
HREF="#DOTREF"
>dot</A
> (<B
CLASS="COMMAND"
>.</B
>)
or <A
HREF="#SOURCEREF"
>source</A
> command.</P
><P
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># SCRIPT LIBRARY
# ------ -------
# Note:
# No "#!" here.
# No "live code" either.
# Useful variable definitions
ROOT_UID=0 # Root has $UID 0.
E_NOTROOT=101 # Not root user error.
MAXRETVAL=255 # Maximum (positive) return value of a function.
SUCCESS=0
FAILURE=-1
# Functions
Usage () # "Usage:" message.
{
if [ -z "$1" ] # No arg passed.
then
msg=filename
else
msg=$@
fi
echo "Usage: `basename $0` "$msg""
}
Check_if_root () # Check if root running script.
{ # From "ex39.sh" example.
if [ "$UID" -ne "$ROOT_UID" ]
then
echo "Must be root to run this script."
exit $E_NOTROOT
fi
}
CreateTempfileName () # Creates a "unique" temp filename.
{ # From "ex51.sh" example.
prefix=temp
suffix=`eval date +%s`
Tempfilename=$prefix.$suffix
}
isalpha2 () # Tests whether *entire string* is alphabetic.
{ # From "isalpha.sh" example.
[ $# -eq 1 ] || return $FAILURE
case $1 in
*[!a-zA-Z]*|"") return $FAILURE;;
*) return $SUCCESS;;
esac # Thanks, S.C.
}
abs () # Absolute value.
{ # Caution: Max return value = 255.
E_ARGERR=-999999
if [ -z "$1" ] # Need arg passed.
then
return $E_ARGERR # Obvious error value returned.
fi
if [ "$1" -ge 0 ] # If non-negative,
then #
absval=$1 # stays as-is.
else # Otherwise,
let "absval = (( 0 - $1 ))" # change sign.
fi
return $absval
}
tolower () # Converts string(s) passed as argument(s)
{ #+ to lowercase.
if [ -z "$1" ] # If no argument(s) passed,
then #+ send error message
echo "(null)" #+ (C-style void-pointer error message)
return #+ and return from function.
fi
echo "$@" | tr A-Z a-z
# Translate all passed arguments ($@).
return
# Use command substitution to set a variable to function output.
# For example:
# oldvar="A seT of miXed-caSe LEtTerS"
# newvar=`tolower "$oldvar"`
# echo "$newvar" # a set of mixed-case letters
#
# Exercise: Rewrite this function to change lowercase passed argument(s)
# to uppercase ... toupper() [easy].
}</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><A
NAME="COMMENTH"
></A
></P
><P
>Use special-purpose comment headers to increase clarity
and legibility in scripts.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>## Caution.
rm -rf *.zzy ## The "-rf" options to "rm" are very dangerous,
##+ especially with wild cards.
#+ Line continuation.
# This is line 1
#+ of a multi-line comment,
#+ and this is the final line.
#* Note.
#o List item.
#&#62; Another point of view.
while [ "$var1" != "end" ] #&#62; while test "$var1" != "end"</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
><A
NAME="PROGBAR"
></A
></P
><P
>Dotan Barak contributes template code for a
<I
CLASS="FIRSTTERM"
>progress bar</I
> in a script.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="PROGRESSBAR"
></A
><P
><B
>Example 36-17. A Progress Bar</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# progress-bar.sh
# Author: Dotan Barak (very minor revisions by ABS Guide author).
# Used in ABS Guide with permission (thanks!).
BAR_WIDTH=50
BAR_CHAR_START="["
BAR_CHAR_END="]"
BAR_CHAR_EMPTY="."
BAR_CHAR_FULL="="
BRACKET_CHARS=2
LIMIT=100
print_progress_bar()
{
# Calculate how many characters will be full.
let "full_limit = ((($1 - $BRACKET_CHARS) * $2) / $LIMIT)"
# Calculate how many characters will be empty.
let "empty_limit = ($1 - $BRACKET_CHARS) - ${full_limit}"
# Prepare the bar.
bar_line="${BAR_CHAR_START}"
for ((j=0; j&#60;full_limit; j++)); do
bar_line="${bar_line}${BAR_CHAR_FULL}"
done
for ((j=0; j&#60;empty_limit; j++)); do
bar_line="${bar_line}${BAR_CHAR_EMPTY}"
done
bar_line="${bar_line}${BAR_CHAR_END}"
printf "%3d%% %s" $2 ${bar_line}
}
# Here is a sample of code that uses it.
MAX_PERCENT=100
for ((i=0; i&#60;=MAX_PERCENT; i++)); do
#
usleep 10000
# ... Or run some other commands ...
#
print_progress_bar ${BAR_WIDTH} ${i}
echo -en "\r"
done
echo ""
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><A
NAME="COMOUTBL"
></A
></P
><P
>A particularly clever use of <A
HREF="#TESTCONSTRUCTS1"
>if-test</A
> constructs
is for comment blocks.</P
><P
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
COMMENT_BLOCK=
# Try setting the above variable to some value
#+ for an unpleasant surprise.
if [ $COMMENT_BLOCK ]; then
Comment block --
=================================
This is a comment line.
This is another comment line.
This is yet another comment line.
=================================
echo "This will not echo."
Comment blocks are error-free! Whee!
fi
echo "No more comments, please."
exit 0</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Compare this with <A
HREF="#CBLOCK1"
>using
here documents to comment out code blocks</A
>.</P
></LI
><LI
><P
><A
NAME="INTPARAM"
></A
></P
><P
>Using the <A
HREF="#XSTATVARREF"
>$? exit status
variable</A
>, a script may test if a parameter contains
only digits, so it can be treated as an integer.</P
><P
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
SUCCESS=0
E_BADINPUT=85
test "$1" -ne 0 -o "$1" -eq 0 2&#62;/dev/null
# An integer is either equal to 0 or not equal to 0.
# 2&#62;/dev/null suppresses error message.
if [ $? -ne "$SUCCESS" ]
then
echo "Usage: `basename $0` integer-input"
exit $E_BADINPUT
fi
let "sum = $1 + 25" # Would give error if $1 not integer.
echo "Sum = $sum"
# Any variable, not just a command-line parameter, can be tested this way.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><A
NAME="RVT"
></A
>The 0 - 255 range for function return
values is a severe limitation. Global variables and
other workarounds are often problematic. An alternative
method for a function to communicate a value back to
the main body of the script is to have the function
write to <TT
CLASS="FILENAME"
>stdout</TT
> (usually with
<A
HREF="#ECHOREF"
>echo</A
>) the <SPAN
CLASS="QUOTE"
>"return
value,"</SPAN
> and assign this to a variable. This is
actually a variant of <A
HREF="#COMMANDSUBREF"
>command
substitution.</A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="MULTIPLICATION"
></A
><P
><B
>Example 36-18. Return value trickery</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# multiplication.sh
multiply () # Multiplies params passed.
{ # Will accept a variable number of args.
local product=1
until [ -z "$1" ] # Until uses up arguments passed...
do
let "product *= $1"
shift
done
echo $product # Will not echo to stdout,
} #+ since this will be assigned to a variable.
mult1=15383; mult2=25211
val1=`multiply $mult1 $mult2`
# Assigns stdout (echo) of function to the variable val1.
echo "$mult1 X $mult2 = $val1" # 387820813
mult1=25; mult2=5; mult3=20
val2=`multiply $mult1 $mult2 $mult3`
echo "$mult1 X $mult2 X $mult3 = $val2" # 2500
mult1=188; mult2=37; mult3=25; mult4=47
val3=`multiply $mult1 $mult2 $mult3 $mult4`
echo "$mult1 X $mult2 X $mult3 X $mult4 = $val3" # 8173300
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>The same technique also works for alphanumeric
strings. This means that a function can <SPAN
CLASS="QUOTE"
>"return"</SPAN
>
a non-numeric value.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>capitalize_ichar () # Capitalizes initial character
{ #+ of argument string(s) passed.
string0="$@" # Accepts multiple arguments.
firstchar=${string0:0:1} # First character.
string1=${string0:1} # Rest of string(s).
FirstChar=`echo "$firstchar" | tr a-z A-Z`
# Capitalize first character.
echo "$FirstChar$string1" # Output to stdout.
}
newstring=`capitalize_ichar "every sentence should start with a capital letter."`
echo "$newstring" # Every sentence should start with a capital letter.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>It is even possible for a function to <SPAN
CLASS="QUOTE"
>"return"</SPAN
>
multiple values with this method.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="SUMPRODUCT"
></A
><P
><B
>Example 36-19. Even more return value trickery</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# sum-product.sh
# A function may "return" more than one value.
sum_and_product () # Calculates both sum and product of passed args.
{
echo $(( $1 + $2 )) $(( $1 * $2 ))
# Echoes to stdout each calculated value, separated by space.
}
echo
echo "Enter first number "
read first
echo
echo "Enter second number "
read second
echo
retval=`sum_and_product $first $second` # Assigns output of function.
sum=`echo "$retval" | awk '{print $1}'` # Assigns first field.
product=`echo "$retval" | awk '{print $2}'` # Assigns second field.
echo "$first + $second = $sum"
echo "$first * $second = $product"
echo
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
NAME="RVTCAUTION"
></A
>There can be only
<B
CLASS="COMMAND"
>one</B
> <I
CLASS="FIRSTTERM"
>echo</I
> statement
in the function for this to work. If you alter the previous
example:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>sum_and_product ()
{
echo "This is the sum_and_product function." # This messes things up!
echo $(( $1 + $2 )) $(( $1 * $2 ))
}
...
retval=`sum_and_product $first $second` # Assigns output of function.
# Now, this will not work correctly.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><A
NAME="PASSARRAY"
></A
></P
><P
>Next in our bag of tricks are techniques for passing
an <A
HREF="#ARRAYREF"
>array</A
> to a
<A
HREF="#FUNCTIONREF"
>function</A
>, then
<SPAN
CLASS="QUOTE"
>"returning"</SPAN
> an array back to the main body of
the script.</P
><P
>Passing an array involves loading the space-separated
elements of the array into a variable with <A
HREF="#COMMANDSUBREF"
>command substitution</A
>. <A
NAME="RETARRAY"
></A
>Getting an array back as the <SPAN
CLASS="QUOTE"
>"return
value"</SPAN
> from a function uses the previously mentioned
strategem of <A
HREF="#ECHOREF"
>echoing</A
> the
array in the function, then invoking command substitution
and the <B
CLASS="COMMAND"
>( ... )</B
> operator to assign it to
an array.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ARRFUNC"
></A
><P
><B
>Example 36-20. Passing and returning arrays</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# array-function.sh: Passing an array to a function and ...
# "returning" an array from a function
Pass_Array ()
{
local passed_array # Local variable!
passed_array=( `echo "$1"` )
echo "${passed_array[@]}"
# List all the elements of the new array
#+ declared and set within the function.
}
original_array=( element1 element2 element3 element4 element5 )
echo
echo "original_array = ${original_array[@]}"
# List all elements of original array.
# This is the trick that permits passing an array to a function.
# **********************************
argument=`echo ${original_array[@]}`
# **********************************
# Pack a variable
#+ with all the space-separated elements of the original array.
#
# Attempting to just pass the array itself will not work.
# This is the trick that allows grabbing an array as a "return value".
# *****************************************
returned_array=( `Pass_Array "$argument"` )
# *****************************************
# Assign 'echoed' output of function to array variable.
echo "returned_array = ${returned_array[@]}"
echo "============================================================="
# Now, try it again,
#+ attempting to access (list) the array from outside the function.
Pass_Array "$argument"
# The function itself lists the array, but ...
#+ accessing the array from outside the function is forbidden.
echo "Passed array (within function) = ${passed_array[@]}"
# NULL VALUE since the array is a variable local to the function.
echo
############################################
# And here is an even more explicit example:
ret_array ()
{
for element in {11..20}
do
echo "$element " # Echo individual elements
done #+ of what will be assembled into an array.
}
arr=( $(ret_array) ) # Assemble into array.
echo "Capturing array \"arr\" from function ret_array () ..."
echo "Third element of array \"arr\" is ${arr[2]}." # 13 (zero-indexed)
echo -n "Entire array is: "
echo ${arr[@]} # 11 12 13 14 15 16 17 18 19 20
echo
exit 0
# Nathan Coulter points out that passing arrays with elements containing
#+ whitespace breaks this example.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>For a more elaborate example of passing arrays to
functions, see <A
HREF="#LIFESLOW"
>Example A-10</A
>.</P
></LI
><LI
><P
><A
NAME="CSTYLE"
></A
></P
><P
>Using the <A
HREF="#DBLPARENS"
>double-parentheses
construct</A
>, it is possible to use C-style syntax
for setting and incrementing/decrementing variables
and in <A
HREF="#FORLOOPREF1"
>for</A
> and <A
HREF="#WHILELOOPREF"
>while</A
> loops. See <A
HREF="#FORLOOPC"
>Example 11-13</A
> and <A
HREF="#WHLOOPC"
>Example 11-18</A
>.</P
></LI
><LI
><P
><A
NAME="SETPUM"
></A
></P
><P
>Setting the <A
HREF="#PATHREF"
>path</A
> and <A
HREF="#UMASKREF"
>umask</A
> at the beginning of a script makes
it more <A
HREF="#PORTABILITYISSUES"
>portable</A
>
-- more likely to run on a <SPAN
CLASS="QUOTE"
>"foreign"</SPAN
> machine
whose user may have bollixed up the <TT
CLASS="VARNAME"
>$PATH</TT
>
and <B
CLASS="COMMAND"
>umask</B
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
PATH=/bin:/usr/bin:/usr/local/bin ; export PATH
umask 022 # Files that the script creates will have 755 permission.
# Thanks to Ian D. Allen, for this tip.</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
><A
NAME="FILTEROUTP"
></A
></P
><P
>A useful scripting technique is to
<EM
>repeatedly</EM
> feed the output of a filter
(by piping) back to the <EM
>same filter</EM
>, but
with a different set of arguments and/or options. Especially
suitable for this are <A
HREF="#TRREF"
>tr</A
> and
<A
HREF="#GREPREF"
>grep</A
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># From "wstrings.sh" example.
wlist=`strings "$1" | tr A-Z a-z | tr '[:space:]' Z | \
tr -cs '[:alpha:]' Z | tr -s '\173-\377' Z | tr Z ' '`</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="AGRAM"
></A
><P
><B
>Example 36-21. Fun with anagrams</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# agram.sh: Playing games with anagrams.
# Find anagrams of...
LETTERSET=etaoinshrdlu
FILTER='.......' # How many letters minimum?
# 1234567
anagram "$LETTERSET" | # Find all anagrams of the letterset...
grep "$FILTER" | # With at least 7 letters,
grep '^is' | # starting with 'is'
grep -v 's$' | # no plurals
grep -v 'ed$' # no past tense verbs
# Possible to add many combinations of conditions and filters.
# Uses "anagram" utility
#+ that is part of the author's "yawl" word list package.
# http://ibiblio.org/pub/Linux/libs/yawl-0.3.2.tar.gz
# http://bash.deta.in/yawl-0.3.2.tar.gz
exit 0 # End of code.
bash$ sh agram.sh
islander
isolate
isolead
isotheral
# Exercises:
# ---------
# Modify this script to take the LETTERSET as a command-line parameter.
# Parameterize the filters in lines 11 - 13 (as with $FILTER),
#+ so that they can be specified by passing arguments to a function.
# For a slightly different approach to anagramming,
#+ see the agram2.sh script.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See also <A
HREF="#CONSTAT"
>Example 29-4</A
>, <A
HREF="#CRYPTOQUOTE"
>Example 16-25</A
>, and <A
HREF="#SOUNDEX"
>Example A-9</A
>.</P
></LI
><LI
><P
><A
NAME="COMMBLAHD"
></A
></P
><P
>Use <SPAN
CLASS="QUOTE"
>"<A
HREF="#ANONHEREDOC0"
>anonymous here
documents</A
>"</SPAN
> to comment out blocks of code,
to save having to individually comment out each line with
a <SPAN
CLASS="TOKEN"
>#</SPAN
>. See <A
HREF="#COMMENTBLOCK"
>Example 19-11</A
>.</P
></LI
><LI
><P
><A
NAME="WHATISREF3"
></A
></P
><P
>Running a script on a machine that relies on a command
that might not be installed is dangerous. Use <A
HREF="#WHATISREF"
>whatis</A
> to avoid potential problems
with this.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>CMD=command1 # First choice.
PlanB=command2 # Fallback option.
command_test=$(whatis "$CMD" | grep 'nothing appropriate')
# If 'command1' not found on system , 'whatis' will return
#+ "command1: nothing appropriate."
#
# A safer alternative is:
# command_test=$(whereis "$CMD" | grep \/)
# But then the sense of the following test would have to be reversed,
#+ since the $command_test variable holds content only if
#+ the $CMD exists on the system.
# (Thanks, bojster.)
if [[ -z "$command_test" ]] # Check whether command present.
then
$CMD option1 option2 # Run command1 with options.
else # Otherwise,
$PlanB #+ run command2.
fi</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><A
NAME="IFGREPFIX"
></A
></P
><P
>An <A
HREF="#IFGREPREF"
>if-grep test</A
> may not
return expected results in an error case, when text is output to
<TT
CLASS="FILENAME"
>stderr</TT
>, rather that
<TT
CLASS="FILENAME"
>stdout</TT
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if ls -l nonexistent_filename | grep -q 'No such file or directory'
then echo "File \"nonexistent_filename\" does not exist."
fi</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
HREF="#IOREDIRREF"
>Redirecting</A
>
<TT
CLASS="FILENAME"
>stderr</TT
> to <TT
CLASS="FILENAME"
>stdout</TT
> fixes
this.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>if ls -l nonexistent_filename 2&#62;&#38;1 | grep -q 'No such file or directory'
# ^^^^
then echo "File \"nonexistent_filename\" does not exist."
fi
# Thanks, Chris Martin, for pointing this out.</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
><A
NAME="SUBSHTMP"
></A
>
If you absolutely must access a subshell variable outside the
subshell, here's a way to do it.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>TMPFILE=tmpfile # Create a temp file to store the variable.
( # Inside the subshell ...
inner_variable=Inner
echo $inner_variable
echo $inner_variable &#62;&#62;$TMPFILE # Append to temp file.
)
# Outside the subshell ...
echo; echo "-----"; echo
echo $inner_variable # Null, as expected.
echo "-----"; echo
# Now ...
read inner_variable &#60;$TMPFILE # Read back shell variable.
rm -f "$TMPFILE" # Get rid of temp file.
echo "$inner_variable" # It's an ugly kludge, but it works.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><A
NAME="RUNPARTSREF2"
></A
></P
><P
>The <A
HREF="#RUNPARTSREF"
>run-parts</A
>
command is handy for running a set of command
scripts in a particular sequence, especially in
combination with <A
HREF="#CRONREF"
>cron</A
> or
<A
HREF="#ATREF"
>at</A
>.</P
></LI
><LI
><P
><A
NAME="RCSREF"
></A
></P
><P
>For doing multiple revisions on a complex script, use the
<I
CLASS="FIRSTTERM"
>rcs</I
> Revision Control System package.</P
><P
> Among other benefits of this is automatically updated ID
header tags. The <B
CLASS="COMMAND"
>co</B
> command in
<I
CLASS="FIRSTTERM"
>rcs</I
> does a parameter replacement of
certain reserved key words, for example, replacing
<TT
CLASS="PARAMETER"
><I
># $Id$</I
></TT
> in a script with something like:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># $Id: hello-world.sh,v 1.1 2004/10/16 02:43:05 bozo Exp $</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
></UL
></DIV
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="AEN20679"
></A
>36.7.2. Widgets</H2
><P
><A
NAME="WIDGETREF"
></A
></P
><P
>It would be nice to be able to invoke X-Windows widgets
from a shell script. There happen to exist several packages
that purport to do so, namely <I
CLASS="FIRSTTERM"
>Xscript</I
>,
<I
CLASS="FIRSTTERM"
>Xmenu</I
>, and <I
CLASS="FIRSTTERM"
>widtools</I
>.
The first two of these no longer seem
to be maintained. Fortunately, it is still
possible to obtain <I
CLASS="FIRSTTERM"
>widtools</I
> <A
HREF="http://www.batse.msfc.nasa.gov/~mallozzi/home/software/xforms/src/widtools-2.0.tgz"
TARGET="_top"
>here</A
>.
</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The <I
CLASS="FIRSTTERM"
>widtools</I
> (widget tools)
package requires the <I
CLASS="FIRSTTERM"
>XForms</I
> library to
be installed. Additionally, the <A
HREF="#MAKEFILEREF"
>Makefile</A
> needs some judicious
editing before the package will build on a typical Linux
system. Finally, three of the six widgets offered do not work
(and, in fact, segfault).</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="DIALOGREF"
></A
></P
><P
>The <I
CLASS="FIRSTTERM"
>dialog</I
> family of tools offers a method
of calling <SPAN
CLASS="QUOTE"
>"dialog"</SPAN
> widgets from a shell script. The
original <I
CLASS="FIRSTTERM"
>dialog</I
> utility works in a text
console, but its successors, <I
CLASS="FIRSTTERM"
>gdialog</I
>,
<I
CLASS="FIRSTTERM"
>Xdialog</I
>, and <I
CLASS="FIRSTTERM"
>kdialog</I
>
use X-Windows-based widget sets.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="DIALOG"
></A
><P
><B
>Example 36-22. Widgets invoked from a shell script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# dialog.sh: Using 'gdialog' widgets.
# Must have 'gdialog' installed on your system to run this script.
# Or, you can replace all instance of 'gdialog' below with 'kdialog' ...
# Version 1.1 (corrected 04/05/05)
# This script was inspired by the following article.
# "Scripting for X Productivity," by Marco Fioretti,
# LINUX JOURNAL, Issue 113, September 2003, pp. 86-9.
# Thank you, all you good people at LJ.
# Input error in dialog box.
E_INPUT=85
# Dimensions of display, input widgets.
HEIGHT=50
WIDTH=60
# Output file name (constructed out of script name).
OUTFILE=$0.output
# Display this script in a text widget.
gdialog --title "Displaying: $0" --textbox $0 $HEIGHT $WIDTH
# Now, we'll try saving input in a file.
echo -n "VARIABLE=" &#62; $OUTFILE
gdialog --title "User Input" --inputbox "Enter variable, please:" \
$HEIGHT $WIDTH 2&#62;&#62; $OUTFILE
if [ "$?" -eq 0 ]
# It's good practice to check exit status.
then
echo "Executed \"dialog box\" without errors."
else
echo "Error(s) in \"dialog box\" execution."
# Or, clicked on "Cancel", instead of "OK" button.
rm $OUTFILE
exit $E_INPUT
fi
# Now, we'll retrieve and display the saved variable.
. $OUTFILE # 'Source' the saved file.
echo "The variable input in the \"input box\" was: "$VARIABLE""
rm $OUTFILE # Clean up by removing the temp file.
# Some applications may need to retain this file.
exit $?
# Exercise: Rewrite this script using the 'zenity' widget set.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="XMESSAGEREF2"
></A
>
The <A
HREF="#XMESSAGEREF"
>xmessage</A
> command is
a simple method of popping up a message/query window. For
example:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>xmessage Fatal error in script! -button exit</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="ZENITYREF2"
></A
>
The latest entry in the widget sweepstakes is
<A
HREF="#ZENITYREF"
>zenity</A
>.
This utility pops up
<I
CLASS="FIRSTTERM"
>GTK+</I
> dialog widgets-and-windows,
and it works very nicely within a script.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>get_info ()
{
zenity --entry # Pops up query window . . .
#+ and prints user entry to stdout.
# Also try the --calendar and --scale options.
}
answer=$( get_info ) # Capture stdout in $answer variable.
echo "User entered: "$answer""</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>For other methods of scripting with widgets, try
<I
CLASS="FIRSTTERM"
>Tk</I
> or <I
CLASS="FIRSTTERM"
>wish</I
>
(<I
CLASS="FIRSTTERM"
>Tcl</I
> derivatives),
<I
CLASS="FIRSTTERM"
>PerlTk</I
> (<I
CLASS="FIRSTTERM"
>Perl</I
>
with <I
CLASS="FIRSTTERM"
>Tk</I
> extensions),
<I
CLASS="FIRSTTERM"
>tksh</I
> (<I
CLASS="FIRSTTERM"
>ksh</I
>
with <I
CLASS="FIRSTTERM"
>Tk</I
> extensions),
<I
CLASS="FIRSTTERM"
>XForms4Perl</I
>
(<I
CLASS="FIRSTTERM"
>Perl</I
> with
<I
CLASS="FIRSTTERM"
>XForms</I
> extensions),
<I
CLASS="FIRSTTERM"
>Gtk-Perl</I
> (<I
CLASS="FIRSTTERM"
>Perl</I
>
with <I
CLASS="FIRSTTERM"
>Gtk</I
> extensions), or
<I
CLASS="FIRSTTERM"
>PyQt</I
> (<I
CLASS="FIRSTTERM"
>Python</I
>
with <I
CLASS="FIRSTTERM"
>Qt</I
> extensions).</P
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="SECURITYISSUES"
></A
>36.8. Security Issues</H1
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="INFECTEDSCRIPTS"
></A
>36.8.1. Infected Shell Scripts</H2
><P
><A
NAME="INFECTEDSCRIPTS1"
></A
></P
><P
>A brief warning about script security is indicated.
A shell script may contain a <I
CLASS="FIRSTTERM"
>worm</I
>,
<I
CLASS="FIRSTTERM"
>trojan</I
>, or even a
<I
CLASS="FIRSTTERM"
>virus</I
>. For that reason, never run
as <I
CLASS="FIRSTTERM"
>root</I
> a script (or permit it to
be inserted into the system startup scripts in <TT
CLASS="FILENAME"
>/etc/rc.d</TT
>) unless you have obtained
said script from a trusted source or you have carefully analyzed
it to make certain it does nothing harmful.</P
><P
>Various researchers at Bell Labs and other sites, including M.
Douglas McIlroy, Tom Duff, and Fred Cohen have investigated the
implications of shell script viruses. They conclude that it is
all too easy for even a novice, a <SPAN
CLASS="QUOTE"
>"script kiddie,"</SPAN
>
to write one.
<A
NAME="AEN20748"
HREF="#FTN.AEN20748"
><SPAN
CLASS="footnote"
>[128]</SPAN
></A
>
</P
><P
>Here is yet another reason to learn scripting. Being able to
look at and understand scripts may protect your system from
being compromised by a rogue script.</P
></DIV
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="HIDINGSOURCE"
></A
>36.8.2. Hiding Shell Script Source</H2
><P
>For security purposes, it may be necessary to render a script
unreadable. If only there were a utility to create a stripped
binary executable from a script. Francisco Rosales' <A
HREF="http://www.datsi.fi.upm.es/~frosal/sources/"
TARGET="_top"
>shc --
generic shell script compiler</A
> does exactly that.</P
><P
>Unfortunately, according to <A
HREF="http://www.linuxjournal.com/article/8256"
TARGET="_top"
>an article</A
> in
the October, 2005 <EM
>Linux Journal</EM
>,
the binary can, in at least some cases, be decrypted to recover
the original script source. Still, this could be a useful
method of keeping scripts secure from all but the most skilled
hackers.</P
></DIV
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="SECURITYTIPS"
></A
>36.8.3. Writing Secure Shell Scripts</H2
><P
><A
NAME="SECURITYTIPS1"
></A
></P
><P
><EM
>Dan Stromberg</EM
> suggests the following
guidelines for writing (relatively) secure shell scripts.</P
><P
> <P
></P
><UL
><LI
><P
>Don't put secret data in <A
HREF="#ENVREF"
>environment variables</A
>.</P
></LI
><LI
><P
>Don't pass secret data in an external
command's arguments (pass them in via a <A
HREF="#PIPEREF"
>pipe</A
> or <A
HREF="#IOREDIRREF"
>redirection</A
> instead).</P
></LI
><LI
><P
>Set your <A
HREF="#PATHREF"
>$PATH</A
>
carefully. Don't just trust whatever path you
inherit from the caller if your script is running as
<I
CLASS="FIRSTTERM"
>root</I
>. In fact, whenever you use
an environment variable inherited from the caller, think
about what could happen if the caller put something
misleading in the variable, e.g., if the caller set
<A
HREF="#HOMEDIRREF"
>$HOME</A
> to <TT
CLASS="FILENAME"
>/etc</TT
>.</P
></LI
></UL
>
</P
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="PORTABILITYISSUES"
></A
>36.9. Portability Issues</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>It is easier to port a shell than a shell script.</I
></P
><P
><I
>--Larry Wall</I
></P
></I
></TD
></TR
></TABLE
><P
>This book deals specifically with Bash scripting on
a GNU/Linux system. All the same, users of <B
CLASS="COMMAND"
>sh</B
>
and <B
CLASS="COMMAND"
>ksh</B
> will find much of value here.</P
><P
><A
NAME="POSIX3REF"
></A
>As it happens, many of the various
shells and scripting languages seem to be converging toward the
<A
HREF="#POSIX2REF"
>POSIX</A
> 1003.2 standard. Invoking
Bash with the <TT
CLASS="OPTION"
>--posix</TT
> option or inserting
a <B
CLASS="COMMAND"
>set -o posix</B
> at the head of a script
causes Bash to conform very closely to this standard. Another
alternative is to use a <I
CLASS="FIRSTTERM"
>#!/bin/sh</I
> <A
HREF="#SHABANGREF"
>sha-bang header</A
> in the script,
rather than <I
CLASS="FIRSTTERM"
>#!/bin/bash</I
>.
<A
NAME="AEN20799"
HREF="#FTN.AEN20799"
><SPAN
CLASS="footnote"
>[129]</SPAN
></A
>
Note that <TT
CLASS="FILENAME"
>/bin/sh</TT
> is a <A
HREF="#LINKREF"
>link</A
> to <TT
CLASS="FILENAME"
>/bin/bash</TT
>
in Linux and certain other flavors of UNIX, and a script invoked
this way disables extended Bash functionality.</P
><P
>Most Bash scripts will run as-is under
<B
CLASS="COMMAND"
>ksh</B
>, and vice-versa, since Chet Ramey has
been busily porting <B
CLASS="COMMAND"
>ksh</B
> features to the
latest versions of Bash.</P
><P
>On a commercial UNIX machine, scripts using GNU-specific
features of standard commands may not work. This has become less
of a problem in the last few years, as the GNU utilities have
pretty much displaced their proprietary
counterparts even on <SPAN
CLASS="QUOTE"
>"big-iron"</SPAN
> UNIX.
<A
HREF="http://linux.oreillynet.com/pub/a/linux/2002/02/28/caldera.html"
TARGET="_top"
>Caldera's
release of the source</A
> to many of the original UNIX
utilities has accelerated the trend.</P
><P
><A
NAME="BASHCOMPAT"
></A
></P
><P
>Bash has certain features that the traditional <A
HREF="#BASHDEF"
>Bourne shell</A
> lacks. Among these are:
<P
></P
><UL
><LI
><P
>Certain extended <A
HREF="#INVOCATIONOPTIONSREF"
>invocation options</A
></P
></LI
><LI
><P
><A
HREF="#COMMANDSUBREF"
>Command substitution</A
> using
<B
CLASS="COMMAND"
>$( )</B
> notation</P
></LI
><LI
><P
><A
HREF="#BRACEEXPREF3"
>Brace expansion</A
></P
></LI
><LI
><P
>Certain <A
HREF="#ARRAYREF"
>array</A
> operations,
and <A
HREF="#ASSOCARR"
>associative arrays</A
></P
></LI
><LI
><P
>The <A
HREF="#DBLBRACKETS"
>double brackets</A
>
extended test construct</P
></LI
><LI
><P
>The <A
HREF="#DBLPARENSREF"
>double-parentheses</A
>
arithmetic-evaluation construct</P
></LI
><LI
><P
>Certain <A
HREF="#STRINGMANIP"
>string manipulation</A
>
operations</P
></LI
><LI
><P
><A
HREF="#PROCESSSUBREF"
>Process substitution</A
></P
></LI
><LI
><P
>A Regular Expression <A
HREF="#REGEXMATCHREF"
>matching
operator</A
></P
></LI
><LI
><P
>Bash-specific <A
HREF="#BUILTINREF"
>builtins</A
></P
></LI
><LI
><P
><A
HREF="#COPROCREF"
>Coprocesses</A
></P
></LI
></UL
>
</P
><P
>See the <A
HREF="ftp://ftp.cwru.edu/pub/bash/FAQ"
TARGET="_top"
>Bash
F.A.Q.</A
> for a complete listing.</P
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="AEN20853"
></A
>36.9.1. A Test Suite</H2
><P
><A
NAME="TESTSUITE0"
></A
>Let us illustrate some of the
incompatibilities between Bash and the classic
Bourne shell. Download and install the <A
HREF="http://freshmeat.net/projects/bournesh"
TARGET="_top"
><SPAN
CLASS="QUOTE"
>"Heirloom
Bourne Shell"</SPAN
></A
> and run the following
script, first using Bash, then the classic
<I
CLASS="FIRSTTERM"
>sh</I
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="TESTSUITE"
></A
><P
><B
>Example 36-23. Test Suite</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# test-suite.sh
# A partial Bash compatibility test suite.
# Run this on your version of Bash, or some other shell.
default_option=FAIL # Tests below will fail unless . . .
echo
echo -n "Testing "
sleep 1; echo -n ". "
sleep 1; echo -n ". "
sleep 1; echo ". "
echo
# Double brackets
String="Double brackets supported?"
echo -n "Double brackets test: "
if [[ "$String" = "Double brackets supported?" ]]
then
echo "PASS"
else
echo "FAIL"
fi
# Double brackets and regex matching
String="Regex matching supported?"
echo -n "Regex matching: "
if [[ "$String" =~ R.....matching* ]]
then
echo "PASS"
else
echo "FAIL"
fi
# Arrays
test_arr=$default_option # FAIL
Array=( If supports arrays will print PASS )
test_arr=${Array[5]}
echo "Array test: $test_arr"
# Command Substitution
csub_test ()
{
echo "PASS"
}
test_csub=$default_option # FAIL
test_csub=$(csub_test)
echo "Command substitution test: $test_csub"
echo
# Completing this script is an exercise for the reader.
# Add to the above similar tests for double parentheses,
#+ brace expansion, process substitution, etc.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="WINSCRIPT"
></A
>36.10. Shell Scripting Under Windows</H1
><P
>Even users running <EM
>that other</EM
> OS can
run UNIX-like shell scripts, and therefore benefit
from many of the lessons of this book. The <A
HREF="http://sourceware.cygnus.com/cygwin/"
TARGET="_top"
> Cygwin</A
> package from Cygnus and the <A
HREF="http://www.mkssoftware.com/"
TARGET="_top"
>MKS utilities</A
> from
Mortice Kern Associates add shell scripting capabilities to
Windows.</P
><P
>Another alternative is <A
HREF="http://www2.research.att.com/~gsf/download/uwin/uwin.html"
TARGET="_top"
> UWIN</A
>, written by David Korn of AT&#38;T, of <A
HREF="#KORNSHELLREF"
>Korn Shell</A
> fame.</P
><P
>In 2006, Microsoft released the <SPAN
CLASS="TRADEMARK"
>Windows Powershell</SPAN
>&reg;,
which contains limited Bash-like command-line scripting
capabilities.</P
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="BASH2"
></A
>Chapter 37. Bash, versions 2, 3, and 4</H1
><DIV
CLASS="SECT1"
><H1
CLASS="SECT1"
><A
NAME="BASHVER2"
></A
>37.1. Bash, version 2</H1
><P
><A
NAME="BASH2REF"
></A
></P
><P
> The current version of <I
CLASS="FIRSTTERM"
>Bash</I
>, the one
you have running on your machine, is most likely version 2.xx.yy,
3.xx.yy, or 4.xx.yy.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $BASH_VERSION</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>3.2.25(1)-release</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The version 2 update of the classic Bash scripting language
added array variables, string and parameter expansion, and
a better method of indirect variable references, among other
features.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX77"
></A
><P
><B
>Example 37-1. String expansion</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# String expansion.
# Introduced with version 2 of Bash.
# Strings of the form $'xxx'
#+ have the standard escaped characters interpreted.
echo $'Ringing bell 3 times \a \a \a'
# May only ring once with certain terminals.
# Or ...
# May not ring at all, depending on terminal settings.
echo $'Three form feeds \f \f \f'
echo $'10 newlines \n\n\n\n\n\n\n\n\n\n'
echo $'\102\141\163\150'
# B a s h
# Octal equivalent of characters.
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="VARREFNEW"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="EX78"
></A
><P
><B
>Example 37-2. Indirect variable references - the new way</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Indirect variable referencing.
# This has a few of the attributes of references in C++.
a=letter_of_alphabet
letter_of_alphabet=z
echo "a = $a" # Direct reference.
echo "Now a = ${!a}" # Indirect reference.
# The ${!variable} notation is more intuitive than the old
#+ eval var1=\$$var2
echo
t=table_cell_3
table_cell_3=24
echo "t = ${!t}" # t = 24
table_cell_3=387
echo "Value of t changed to ${!t}" # 387
# No 'eval' necessary.
# This is useful for referencing members of an array or table,
#+ or for simulating a multi-dimensional array.
# An indexing option (analogous to pointer arithmetic)
#+ would have been nice. Sigh.
exit 0
# See also, ind-ref.sh example.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="RESISTOR"
></A
><P
><B
>Example 37-3. Simple database application, using indirect variable
referencing</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# resistor-inventory.sh
# Simple database / table-lookup application.
# ============================================================== #
# Data
B1723_value=470 # Ohms
B1723_powerdissip=.25 # Watts
B1723_colorcode="yellow-violet-brown" # Color bands
B1723_loc=173 # Where they are
B1723_inventory=78 # How many
B1724_value=1000
B1724_powerdissip=.25
B1724_colorcode="brown-black-red"
B1724_loc=24N
B1724_inventory=243
B1725_value=10000
B1725_powerdissip=.125
B1725_colorcode="brown-black-orange"
B1725_loc=24N
B1725_inventory=89
# ============================================================== #
echo
PS3='Enter catalog number: '
echo
select catalog_number in "B1723" "B1724" "B1725"
do
Inv=${catalog_number}_inventory
Val=${catalog_number}_value
Pdissip=${catalog_number}_powerdissip
Loc=${catalog_number}_loc
Ccode=${catalog_number}_colorcode
echo
echo "Catalog number $catalog_number:"
# Now, retrieve value, using indirect referencing.
echo "There are ${!Inv} of [${!Val} ohm / ${!Pdissip} watt]\
resistors in stock." # ^ ^
# As of Bash 4.2, you can replace "ohm" with \u2126 (using echo -e).
echo "These are located in bin # ${!Loc}."
echo "Their color code is \"${!Ccode}\"."
break
done
echo; echo
# Exercises:
# ---------
# 1) Rewrite this script to read its data from an external file.
# 2) Rewrite this script to use arrays,
#+ rather than indirect variable referencing.
# Which method is more straightforward and intuitive?
# Which method is easier to code?
# Notes:
# -----
# Shell scripts are inappropriate for anything except the most simple
#+ database applications, and even then it involves workarounds and kludges.
# Much better is to use a language with native support for data structures,
#+ such as C++ or Java (or even Perl).
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="CARDS"
></A
><P
><B
>Example 37-4. Using arrays and other miscellaneous trickery
to deal four random hands from a deck of cards</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# cards.sh
# Deals four random hands from a deck of cards.
UNPICKED=0
PICKED=1
DUPE_CARD=99
LOWER_LIMIT=0
UPPER_LIMIT=51
CARDS_IN_SUIT=13
CARDS=52
declare -a Deck
declare -a Suits
declare -a Cards
# It would have been easier to implement and more intuitive
#+ with a single, 3-dimensional array.
# Perhaps a future version of Bash will support multidimensional arrays.
initialize_Deck ()
{
i=$LOWER_LIMIT
until [ "$i" -gt $UPPER_LIMIT ]
do
Deck[i]=$UNPICKED # Set each card of "Deck" as unpicked.
let "i += 1"
done
echo
}
initialize_Suits ()
{
Suits[0]=C #Clubs
Suits[1]=D #Diamonds
Suits[2]=H #Hearts
Suits[3]=S #Spades
}
initialize_Cards ()
{
Cards=(2 3 4 5 6 7 8 9 10 J Q K A)
# Alternate method of initializing an array.
}
pick_a_card ()
{
card_number=$RANDOM
let "card_number %= $CARDS" # Restrict range to 0 - 51, i.e., 52 cards.
if [ "${Deck[card_number]}" -eq $UNPICKED ]
then
Deck[card_number]=$PICKED
return $card_number
else
return $DUPE_CARD
fi
}
parse_card ()
{
number=$1
let "suit_number = number / CARDS_IN_SUIT"
suit=${Suits[suit_number]}
echo -n "$suit-"
let "card_no = number % CARDS_IN_SUIT"
Card=${Cards[card_no]}
printf %-4s $Card
# Print cards in neat columns.
}
seed_random () # Seed random number generator.
{ # What happens if you don't do this?
seed=`eval date +%s`
let "seed %= 32766"
RANDOM=$seed
} # Consider other methods of seeding the random number generator.
deal_cards ()
{
echo
cards_picked=0
while [ "$cards_picked" -le $UPPER_LIMIT ]
do
pick_a_card
t=$?
if [ "$t" -ne $DUPE_CARD ]
then
parse_card $t
u=$cards_picked+1
# Change back to 1-based indexing, temporarily. Why?
let "u %= $CARDS_IN_SUIT"
if [ "$u" -eq 0 ] # Nested if/then condition test.
then
echo
echo
fi # Each hand set apart with a blank line.
let "cards_picked += 1"
fi
done
echo
return 0
}
# Structured programming:
# Entire program logic modularized in functions.
#===============
seed_random
initialize_Deck
initialize_Suits
initialize_Cards
deal_cards
#===============
exit
# Exercise 1:
# Add comments to thoroughly document this script.
# Exercise 2:
# Add a routine (function) to print out each hand sorted in suits.
# You may add other bells and whistles if you like.
# Exercise 3:
# Simplify and streamline the logic of the script.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="BASHVER3"
></A
>37.2. Bash, version 3</H1
><P
><A
NAME="BASH3REF"
></A
></P
><P
>On July 27, 2004, Chet Ramey released version 3 of Bash.
This update fixed quite a number of bugs and added new
features.</P
><P
>Some of the more important added features:
<P
></P
><UL
><LI
><P
><A
NAME="BRACEEXPREF3"
></A
></P
><P
>A new, more generalized <B
CLASS="COMMAND"
>{a..z}</B
> <A
HREF="#BRACEEXPREF"
>brace expansion</A
> operator.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
for i in {1..10}
# Simpler and more straightforward than
#+ for i in $(seq 10)
do
echo -n "$i "
done
echo
# 1 2 3 4 5 6 7 8 9 10
# Or just . . .
echo {a..z} # a b c d e f g h i j k l m n o p q r s t u v w x y z
echo {e..m} # e f g h i j k l m
echo {z..a} # z y x w v u t s r q p o n m l k j i h g f e d c b a
# Works backwards, too.
echo {25..30} # 25 26 27 28 29 30
echo {3..-2} # 3 2 1 0 -1 -2
echo {X..d} # X Y Z [ ] ^ _ ` a b c d
# Shows (some of) the ASCII characters between Z and a,
#+ but don't rely on this type of behavior because . . .
echo {]..a} # {]..a}
# Why?
# You can tack on prefixes and suffixes.
echo "Number #"{1..4}, "..."
# Number #1, Number #2, Number #3, Number #4, ...
# You can concatenate brace-expansion sets.
echo {1..3}{x..z}" +" "..."
# 1x + 1y + 1z + 2x + 2y + 2z + 3x + 3y + 3z + ...
# Generates an algebraic expression.
# This could be used to find permutations.
# You can nest brace-expansion sets.
echo {{a..c},{1..3}}
# a b c 1 2 3
# The "comma operator" splices together strings.
# ########## ######### ############ ########### ######### ###############
# Unfortunately, brace expansion does not lend itself to parameterization.
var1=1
var2=5
echo {$var1..$var2} # {1..5}
# Yet, as Emiliano G. points out, using "eval" overcomes this limitation.
start=0
end=10
for index in $(eval echo {$start..$end})
do
echo -n "$index " # 0 1 2 3 4 5 6 7 8 9 10
done
echo</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
>The <B
CLASS="COMMAND"
>${!array[@]}</B
> operator, which
expands to all the indices of a given <A
HREF="#ARRAYREF"
>array</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
Array=(element-zero element-one element-two element-three)
echo ${Array[0]} # element-zero
# First element of array.
echo ${!Array[@]} # 0 1 2 3
# All the indices of Array.
for i in ${!Array[@]}
do
echo ${Array[i]} # element-zero
# element-one
# element-two
# element-three
#
# All the elements in Array.
done</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
><A
NAME="REGEXMATCHREF"
></A
></P
><P
>The <B
CLASS="COMMAND"
>=~</B
> <A
HREF="#REGEXREF"
>Regular
Expression</A
> matching operator within a <A
HREF="#DBLBRACKETS"
>double brackets</A
> test expression.
(Perl has a similar operator.)</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
variable="This is a fine mess."
echo "$variable"
# Regex matching with =~ operator within [[ double brackets ]].
if [[ "$variable" =~ T.........fin*es* ]]
# NOTE: As of version 3.2 of Bash, expression to match no longer quoted.
then
echo "match found"
# match found
fi</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Or, more usefully:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
input=$1
if [[ "$input" =~ "[0-9][0-9][0-9]-[0-9][0-9]-[0-9][0-9][0-9][0-9]" ]]
# ^ NOTE: Quoting not necessary, as of version 3.2 of Bash.
# NNN-NN-NNNN (where each N is a digit).
then
echo "Social Security number."
# Process SSN.
else
echo "Not a Social Security number!"
# Or, ask for corrected input.
fi</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>For additional examples of using the
<B
CLASS="COMMAND"
>=~</B
> operator, see <A
HREF="#WHX"
>Example A-29</A
>,
<A
HREF="#MAILBOXGREP"
>Example 19-14</A
>, <A
HREF="#FINDSPLIT"
>Example A-35</A
>, and <A
HREF="#TOHTML"
>Example A-24</A
>.</P
></LI
><LI
><P
><A
NAME="PIPEFAILREF"
></A
></P
><P
>The new <TT
CLASS="OPTION"
>set -o pipefail</TT
> option is
useful for debugging <A
HREF="#PIPEREF"
>pipes</A
>. If
this option is set, then the <A
HREF="#EXITSTATUSREF"
>exit status</A
> of a pipe
is the exit status of the last command in the pipe to
<EM
>fail</EM
> (return a non-zero value), rather
than the actual final command in the pipe.</P
><P
>See <A
HREF="#FC4UPD"
>Example 16-43</A
>.</P
></LI
></UL
>
</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The update to version 3 of Bash breaks a few scripts
that worked under earlier versions. <EM
>Test critical legacy
scripts to make sure they still work!</EM
></P
><P
>As it happens, a couple of the scripts in the
<EM
>Advanced Bash Scripting Guide</EM
> had to be
fixed up (see <A
HREF="#TOUT"
>Example 9-4</A
>, for instance).</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="AEN20956"
></A
>37.2.1. Bash, version 3.1</H2
><P
>The version 3.1 update of Bash introduces a number of bugfixes
and a few minor changes.</P
><P
></P
><UL
><LI
><P
>The <SPAN
CLASS="TOKEN"
>+=</SPAN
> operator is now permitted in
in places where previously only the <SPAN
CLASS="TOKEN"
>=</SPAN
>
assignment operator was recognized.</P
><P
><A
NAME="PLUSEQSTR"
></A
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>a=1
echo $a # 1
a+=5 # Won't work under versions of Bash earlier than 3.1.
echo $a # 15
a+=Hello
echo $a # 15Hello</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Here, <SPAN
CLASS="TOKEN"
>+=</SPAN
> functions as a <I
CLASS="FIRSTTERM"
>string
concatenation</I
> operator. Note that its behavior
in this particular context is different than within a
<A
HREF="#LETREF"
>let</A
> construct.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>a=1
echo $a # 1
let a+=5 # Integer arithmetic, rather than string concatenation.
echo $a # 6
let a+=Hello # Doesn't "add" anything to a.
echo $a # 6</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="PATHAPPEND"
></A
>Jeffrey Haemer points out
that this concatenation operator can be quite
useful. In this instance, we append a directory to the
<TT
CLASS="VARNAME"
>$PATH</TT
>.</P
><P
>&#13; <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $PATH</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/usr/bin:/bin:/usr/local/bin:/usr/X11R6/bin/:/usr/games</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>PATH+=:/opt/bin</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $PATH</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>/usr/bin:/bin:/usr/local/bin:/usr/X11R6/bin/:/usr/games:/opt/bin</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
></UL
></DIV
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="AEN20987"
></A
>37.2.2. Bash, version 3.2</H2
><P
>This is pretty much a bugfix update.</P
><P
></P
><UL
><LI
><P
>In <A
HREF="#PSGLOB"
><I
CLASS="FIRSTTERM"
>global</I
>
parameter substitutions</A
>, the pattern no longer anchors
at the start of the string.</P
></LI
><LI
><P
>The <TT
CLASS="OPTION"
>--wordexp</TT
> option disables
<A
HREF="#PROCESSSUBREF"
>process substitution</A
>.</P
></LI
><LI
><P
>The <B
CLASS="COMMAND"
>=~</B
> <A
HREF="#REGEXMATCHREF"
>Regular Expression
match operator</A
> no longer requires
<A
HREF="#QUOTINGREF"
>quoting</A
> of the
<I
CLASS="FIRSTTERM"
>pattern</I
> within <A
HREF="#DBLBRACKETS"
>[[ ... ]]</A
>.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>In fact, quoting in this context is
<EM
>not</EM
> advisable as it may
cause <I
CLASS="FIRSTTERM"
>regex</I
> evaluation to fail.
Chet Ramey states in the <A
HREF="#BASHFAQ"
>Bash
FAQ</A
> that quoting explicitly disables regex evaluation.
See also the <A
HREF="https://bugs.launchpad.net/ubuntu-website/+bug/109931"
TARGET="_top"
> Ubuntu Bug List</A
> and <A
HREF="http://en.wikinerds.org/index.php/Bash_syntax_and_semantics"
TARGET="_top"
> Wikinerds on Bash syntax</A
>.</P
><P
>Setting <EM
>shopt -s compat31</EM
>
in a script causes reversion to the original
behavior.</P
></TD
></TR
></TABLE
></DIV
></LI
></UL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="BASHVER4"
></A
>37.3. Bash, version 4</H1
><P
><A
NAME="BASH4REF"
></A
></P
><P
>Chet Ramey announced Version 4 of Bash on the 20th
of February, 2009. This release has a number of significant
new features, as well as some important bugfixes.</P
><P
>Among the new goodies:</P
><P
></P
><UL
><LI
><P
><A
NAME="ASSOCARR"
></A
>Associative arrays.
<A
NAME="AEN21025"
HREF="#FTN.AEN21025"
><SPAN
CLASS="footnote"
>[130]</SPAN
></A
>
</P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN21029"
></A
><P
></P
><P
>An <I
CLASS="FIRSTTERM"
>associative</I
> array can
be thought of as a set of two linked arrays -- one holding
the <I
CLASS="FIRSTTERM"
>data</I
>, and the other the
<I
CLASS="FIRSTTERM"
>keys</I
> that index the individual elements
of the <I
CLASS="FIRSTTERM"
>data</I
> array.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><DIV
CLASS="EXAMPLE"
><A
NAME="FETCHADDRESS"
></A
><P
><B
>Example 37-5. A simple address database</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
# fetch_address.sh
declare -A address
# -A option declares associative array.
address[Charles]="414 W. 10th Ave., Baltimore, MD 21236"
address[John]="202 E. 3rd St., New York, NY 10009"
address[Wilma]="1854 Vermont Ave, Los Angeles, CA 90023"
echo "Charles's address is ${address[Charles]}."
# Charles's address is 414 W. 10th Ave., Baltimore, MD 21236.
echo "Wilma's address is ${address[Wilma]}."
# Wilma's address is 1854 Vermont Ave, Los Angeles, CA 90023.
echo "John's address is ${address[John]}."
# John's address is 202 E. 3rd St., New York, NY 10009.
echo
echo "${!address[*]}" # The array indices ...
# Charles John Wilma</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="FETCHADDRESS2"
></A
><P
><B
>Example 37-6. A somewhat more elaborate address database</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
# fetch_address-2.sh
# A more elaborate version of fetch_address.sh.
SUCCESS=0
E_DB=99 # Error code for missing entry.
declare -A address
# -A option declares associative array.
store_address ()
{
address[$1]="$2"
return $?
}
fetch_address ()
{
if [[ -z "${address[$1]}" ]]
then
echo "$1's address is not in database."
return $E_DB
fi
echo "$1's address is ${address[$1]}."
return $?
}
store_address "Lucas Fayne" "414 W. 13th Ave., Baltimore, MD 21236"
store_address "Arvid Boyce" "202 E. 3rd St., New York, NY 10009"
store_address "Velma Winston" "1854 Vermont Ave, Los Angeles, CA 90023"
# Exercise:
# Rewrite the above store_address calls to read data from a file,
#+ then assign field 1 to name, field 2 to address in the array.
# Each line in the file would have a format corresponding to the above.
# Use a while-read loop to read from file, sed or awk to parse the fields.
fetch_address "Lucas Fayne"
# Lucas Fayne's address is 414 W. 13th Ave., Baltimore, MD 21236.
fetch_address "Velma Winston"
# Velma Winston's address is 1854 Vermont Ave, Los Angeles, CA 90023.
fetch_address "Arvid Boyce"
# Arvid Boyce's address is 202 E. 3rd St., New York, NY 10009.
fetch_address "Bozo Bozeman"
# Bozo Bozeman's address is not in database.
exit $? # In this case, exit code = 99, since that is function return.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>See <A
HREF="#SAMORSE"
>Example A-53</A
> for an interesting
usage of an <I
CLASS="FIRSTTERM"
>associative array</I
>.</P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Elements of the <I
CLASS="FIRSTTERM"
>index</I
> array
may include embedded <A
HREF="#WHITESPACEREF"
>space
characters</A
>, or even leading and/or trailing space
characters. However, index array elements containing
<EM
>only</EM
> <I
CLASS="FIRSTTERM"
>whitespace</I
>
are <EM
>not</EM
> permitted.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>address[ ]="Blank" # Error!</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><A
NAME="NCTERM"
></A
>Enhancements to the
<A
HREF="#CASEESAC1"
>case</A
> construct:
the <TT
CLASS="REPLACEABLE"
><I
>;;&#38;</I
></TT
> and
<TT
CLASS="REPLACEABLE"
><I
>;&#38;</I
></TT
> terminators.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="CASE4"
></A
><P
><B
>Example 37-7. Testing characters</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
test_char ()
{
case "$1" in
[[:print:]] ) echo "$1 is a printable character.";;&#38; # |
# The ;;&#38; terminator continues to the next pattern test. |
[[:alnum:]] ) echo "$1 is an alpha/numeric character.";;&#38; # v
[[:alpha:]] ) echo "$1 is an alphabetic character.";;&#38; # v
[[:lower:]] ) echo "$1 is a lowercase alphabetic character.";;&#38;
[[:digit:]] ) echo "$1 is an numeric character.";&#38; # |
# The ;&#38; terminator executes the next statement ... # |
%%%@@@@@ ) echo "********************************";; # v
# ^^^^^^^^ ... even with a dummy pattern.
esac
}
echo
test_char 3
# 3 is a printable character.
# 3 is an alpha/numeric character.
# 3 is an numeric character.
# ********************************
echo
test_char m
# m is a printable character.
# m is an alpha/numeric character.
# m is an alphabetic character.
# m is a lowercase alphabetic character.
echo
test_char /
# / is a printable character.
echo
# The ;;&#38; terminator can save complex if/then conditions.
# The ;&#38; is somewhat less useful.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><A
NAME="COPROCREF"
></A
>The new <B
CLASS="COMMAND"
>coproc</B
>
builtin enables two parallel <A
HREF="#PROCESSREF"
>processes</A
> to communicate and
interact. As Chet Ramey states in the
<A
HREF="#BASHFAQ"
>Bash FAQ</A
>
<A
NAME="AEN21068"
HREF="#FTN.AEN21068"
><SPAN
CLASS="footnote"
>[131]</SPAN
></A
>
, ver. 4.01:
</P
><A
NAME="AEN21070"
></A
><BLOCKQUOTE
CLASS="BLOCKQUOTE"
><P
CLASS="LITERALLAYOUT"
>&nbsp;&nbsp;&nbsp;&nbsp;There&nbsp;is&nbsp;a&nbsp;new&nbsp;'coproc'&nbsp;reserved&nbsp;word&nbsp;that&nbsp;specifies&nbsp;a&nbsp;coprocess:<br>
&nbsp;&nbsp;&nbsp;&nbsp;an&nbsp;asynchronous&nbsp;command&nbsp;run&nbsp;with&nbsp;two&nbsp;pipes&nbsp;connected&nbsp;to&nbsp;the&nbsp;creating<br>
&nbsp;&nbsp;&nbsp;&nbsp;shell.&nbsp;Coprocs&nbsp;can&nbsp;be&nbsp;named.&nbsp;The&nbsp;input&nbsp;and&nbsp;output&nbsp;file&nbsp;descriptors<br>
&nbsp;&nbsp;&nbsp;&nbsp;and&nbsp;the&nbsp;PID&nbsp;of&nbsp;the&nbsp;coprocess&nbsp;are&nbsp;available&nbsp;to&nbsp;the&nbsp;calling&nbsp;shell&nbsp;in<br>
&nbsp;&nbsp;&nbsp;&nbsp;variables&nbsp;with&nbsp;coproc-specific&nbsp;names.<br>
<br>
&nbsp;&nbsp;&nbsp;&nbsp;George&nbsp;Dimitriu&nbsp;explains,<br>
&nbsp;&nbsp;&nbsp;&nbsp;"...&nbsp;coproc&nbsp;...&nbsp;is&nbsp;a&nbsp;feature&nbsp;used&nbsp;in&nbsp;Bash&nbsp;process&nbsp;substitution,<br>
&nbsp;&nbsp;&nbsp;&nbsp;which&nbsp;now&nbsp;is&nbsp;made&nbsp;publicly&nbsp;available."<br>
&nbsp;&nbsp;&nbsp;&nbsp;This&nbsp;means&nbsp;it&nbsp;can&nbsp;be&nbsp;explicitly&nbsp;invoked&nbsp;in&nbsp;a&nbsp;script,&nbsp;rather&nbsp;than<br>
&nbsp;&nbsp;&nbsp;&nbsp;just&nbsp;being&nbsp;a&nbsp;behind-the-scenes&nbsp;mechanism&nbsp;used&nbsp;by&nbsp;Bash.<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</P
></BLOCKQUOTE
><P
>Coprocesses use <I
CLASS="FIRSTTERM"
>file descriptors</I
>.
<A
HREF="#FDREF2"
>File descriptors enable processes and
pipes to communicate</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
# A coprocess communicates with a while-read loop.
coproc { cat mx_data.txt; sleep 2; }
# ^^^^^^^
# Try running this without "sleep 2" and see what happens.
while read -u ${COPROC[0]} line # ${COPROC[0]} is the
do #+ file descriptor of the coprocess.
echo "$line" | sed -e 's/line/NOT-ORIGINAL-TEXT/'
done
kill $COPROC_PID # No longer need the coprocess,
#+ so kill its PID.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>But, be careful!</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
echo; echo
a=aaa
b=bbb
c=ccc
coproc echo "one two three"
while read -u ${COPROC[0]} a b c; # Note that this loop
do #+ runs in a subshell.
echo "Inside while-read loop: ";
echo "a = $a"; echo "b = $b"; echo "c = $c"
echo "coproc file descriptor: ${COPROC[0]}"
done
# a = one
# b = two
# c = three
# So far, so good, but ...
echo "-----------------"
echo "Outside while-read loop: "
echo "a = $a" # a =
echo "b = $b" # b =
echo "c = $c" # c =
echo "coproc file descriptor: ${COPROC[0]}"
echo
# The coproc is still running, but ...
#+ it still doesn't enable the parent process
#+ to "inherit" variables from the child process, the while-read loop.
# Compare this to the "badread.sh" script.</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The coprocess is <I
CLASS="FIRSTTERM"
>asynchronous</I
>,
and this might cause a problem. It may terminate before another
process has finished communicating with it.</P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
coproc cpname { for i in {0..10}; do echo "index = $i"; done; }
# ^^^^^^ This is a *named* coprocess.
read -u ${cpname[0]}
echo $REPLY # index = 0
echo ${COPROC[0]} #+ No output ... the coprocess timed out
# after the first loop iteration.
# However, George Dimitriu has a partial fix.
coproc cpname { for i in {0..10}; do echo "index = $i"; done; sleep 1;
echo hi &#62; myo; cat - &#62;&#62; myo; }
# ^^^^^ This is a *named* coprocess.
echo "I am main"$'\04' &#62;&#38;${cpname[1]}
myfd=${cpname[0]}
echo myfd=$myfd
### while read -u $myfd
### do
### echo $REPLY;
### done
echo $cpname_PID
# Run this with and without the commented-out while-loop, and it is
#+ apparent that each process, the executing shell and the coprocess,
#+ waits for the other to finish writing in its own write-enabled pipe.</PRE
></FONT
></TD
></TR
></TABLE
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><A
NAME="MAPFILEREF"
></A
>The new <B
CLASS="COMMAND"
>mapfile</B
>
builtin makes it possible to load an array with the contents
of a text file without using a loop or <A
HREF="#ARRAYINITCS"
>command substitution</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
mapfile Arr1 &#60; $0
# Same result as Arr1=( $(cat $0) )
echo "${Arr1[@]}" # Copies this entire script out to stdout.
echo "--"; echo
# But, not the same as read -a !!!
read -a Arr2 &#60; $0
echo "${Arr2[@]}" # Reads only first line of script into the array.
exit</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
>The <A
HREF="#READREF"
>read</A
> builtin got
a minor facelift. The <TT
CLASS="OPTION"
>-t</TT
>
<A
HREF="#READTIMED"
>timeout</A
> option now accepts
(decimal) fractional values
<A
NAME="AEN21096"
HREF="#FTN.AEN21096"
><SPAN
CLASS="footnote"
>[132]</SPAN
></A
>
and the <TT
CLASS="OPTION"
>-i</TT
> option
permits preloading the edit buffer.
<A
NAME="AEN21101"
HREF="#FTN.AEN21101"
><SPAN
CLASS="footnote"
>[133]</SPAN
></A
>
Unfortunately, these enhancements are still a work in progress
and not (yet) usable in scripts.</P
></LI
><LI
><P
><A
NAME="CASEMODPARAMSUB"
></A
>
<A
HREF="#PARAMSUBREF"
>Parameter substitution</A
>
gets <I
CLASS="FIRSTTERM"
>case-modification</I
> operators.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
var=veryMixedUpVariable
echo ${var} # veryMixedUpVariable
echo ${var^} # VeryMixedUpVariable
# * First char --&#62; uppercase.
echo ${var^^} # VERYMIXEDUPVARIABLE
# ** All chars --&#62; uppercase.
echo ${var,} # veryMixedUpVariable
# * First char --&#62; lowercase.
echo ${var,,} # verymixedupvariable
# ** All chars --&#62; lowercase.</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
><A
NAME="DECLARECASEMOD"
></A
></P
><P
>The <A
HREF="#DECLAREREF"
>declare</A
> builtin now
accepts the <TT
CLASS="OPTION"
>-l</TT
> <I
CLASS="FIRSTTERM"
>lowercase</I
>
and <TT
CLASS="OPTION"
>-c</TT
> <I
CLASS="FIRSTTERM"
>capitalize</I
>
options.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
declare -l var1 # Will change to lowercase
var1=MixedCaseVARIABLE
echo "$var1" # mixedcasevariable
# Same effect as echo $var1 | tr A-Z a-z
declare -c var2 # Changes only initial char to uppercase.
var2=originally_lowercase
echo "$var2" # Originally_lowercase
# NOT the same effect as echo $var2 | tr a-z A-Z</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
><A
NAME="BRACEEXPREF4"
></A
>
<A
HREF="#BRACEEXPREF"
>Brace expansion</A
> has more options.</P
><P
><I
CLASS="FIRSTTERM"
>Increment/decrement</I
>, specified in the
final term within braces.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
echo {40..60..2}
# 40 42 44 46 48 50 52 54 56 58 60
# All the even numbers, between 40 and 60.
echo {60..40..2}
# 60 58 56 54 52 50 48 46 44 42 40
# All the even numbers, between 40 and 60, counting backwards.
# In effect, a decrement.
echo {60..40..-2}
# The same output. The minus sign is not necessary.
# But, what about letters and symbols?
echo {X..d}
# X Y Z [ ] ^ _ ` a b c d
# Does not echo the \ which escapes a space.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><I
CLASS="FIRSTTERM"
>Zero-padding</I
>, specified in the
first term within braces, prefixes each term in the output
with the <EM
>same number</EM
> of zeroes.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash4$ </TT
><TT
CLASS="USERINPUT"
><B
>echo {010..15}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>010 011 012 013 014 015</TT
>
<TT
CLASS="PROMPT"
>bash4$ </TT
><TT
CLASS="USERINPUT"
><B
>echo {000..10}</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>000 001 002 003 004 005 006 007 008 009 010</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
></LI
><LI
><P
><A
NAME="SUBSTREXTREF4"
></A
></P
><P
><A
HREF="#SUBSTREXTREF4"
><I
CLASS="FIRSTTERM"
>Substring
extraction</I
> on <I
CLASS="FIRSTTERM"
>positional
parameters</I
></A
> now starts with <A
HREF="#SCRNAMEPARAM"
>$0</A
> as the
<I
CLASS="FIRSTTERM"
>zero-index</I
>. (This corrects an
inconsistency in the treatment of positional parameters.)</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# show-params.bash
# Requires version 4+ of Bash.
# Invoke this scripts with at least one positional parameter.
E_BADPARAMS=99
if [ -z "$1" ]
then
echo "Usage $0 param1 ..."
exit $E_BADPARAMS
fi
echo ${@:0}
# bash3 show-params.bash4 one two three
# one two three
# bash4 show-params.bash4 one two three
# show-params.bash4 one two three
# $0 $1 $2 $3</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
><A
NAME="GLOBSTARREF"
></A
>The new <SPAN
CLASS="TOKEN"
>**</SPAN
>
<A
HREF="#GLOBBINGREF"
>globbing</A
> operator
matches filenames and directories
<A
HREF="#RECURSIONREF0"
>recursively</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
# filelist.bash4
shopt -s globstar # Must enable globstar, otherwise ** doesn't work.
# The globstar shell option is new to version 4 of Bash.
echo "Using *"; echo
for filename in *
do
echo "$filename"
done # Lists only files in current directory ($PWD).
echo; echo "--------------"; echo
echo "Using **"
for filename in **
do
echo "$filename"
done # Lists complete file tree, recursively.
exit
Using *
allmyfiles
filelist.bash4
--------------
Using **
allmyfiles
allmyfiles/file.index.txt
allmyfiles/my_music
allmyfiles/my_music/me-singing-60s-folksongs.ogg
allmyfiles/my_music/me-singing-opera.ogg
allmyfiles/my_music/piano-lesson.1.ogg
allmyfiles/my_pictures
allmyfiles/my_pictures/at-beach-with-Jade.png
allmyfiles/my_pictures/picnic-with-Melissa.png
filelist.bash4</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
><LI
><P
>The new <A
HREF="#BASHPIDREF"
>$BASHPID</A
>
internal variable.</P
></LI
><LI
><P
><A
NAME="CNFH"
></A
></P
><P
>There is a new <A
HREF="#BUILTINREF"
>builtin</A
>
error-handling function named
<B
CLASS="COMMAND"
>command_not_found_handle</B
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash4
command_not_found_handle ()
{ # Accepts implicit parameters.
echo "The following command is not valid: \""$1\"""
echo "With the following argument(s): \""$2\"" \""$3\""" # $4, $5 ...
} # $1, $2, etc. are not explicitly passed to the function.
bad_command arg1 arg2
# The following command is not valid: "bad_command"
# With the following argument(s): "arg1" "arg2"</PRE
></FONT
></TD
></TR
></TABLE
></P
></LI
></UL
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN21170"
></A
><P
></P
><P
><EM
>Editorial comment</EM
></P
><P
>Associative arrays? Coprocesses? Whatever happened
to the lean and mean Bash we have come to know and love?
Could it be suffering from (horrors!) <SPAN
CLASS="QUOTE"
>"feature
creep"</SPAN
>? Or perhaps even Korn shell envy?</P
><P
><EM
>Note to Chet Ramey:</EM
> Please add only
<EM
>essential</EM
> features in future Bash
releases -- perhaps <I
CLASS="FIRSTTERM"
>for-each</I
>
loops and support for multi-dimensional arrays.
<A
NAME="AEN21179"
HREF="#FTN.AEN21179"
><SPAN
CLASS="footnote"
>[134]</SPAN
></A
>
Most Bash users won't need, won't use, and likely won't greatly
appreciate complex <SPAN
CLASS="QUOTE"
>"features"</SPAN
> like built-in
debuggers, Perl interfaces, and bolt-on rocket boosters.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="AEN21183"
></A
>37.3.1. Bash, version 4.1</H2
><P
><A
NAME="BASH41"
></A
>Version 4.1 of Bash, released in May,
2010, was primarily a bugfix update.</P
><P
></P
><UL
><LI
><P
>The <A
HREF="#PRINTFREF"
>printf</A
> command
now accepts a <TT
CLASS="OPTION"
>-v</TT
> option for setting <A
HREF="#ARRAYREF"
>array</A
> indices.</P
></LI
><LI
><P
>Within <A
HREF="#DBLBRACKETS"
>double brackets</A
>,
the <B
CLASS="COMMAND"
>&#62;</B
> and <B
CLASS="COMMAND"
>&#60;</B
>
string comparison operators now conform to the <A
HREF="#LOCALEREF"
>locale</A
>. Since the locale setting may
affect the sorting order of string expressions, this
has side-effects on comparison tests within
<EM
>[[ ... ]]</EM
> expressions.</P
></LI
><LI
><P
>The <A
HREF="#READREF"
>read</A
> builtin
now takes a <TT
CLASS="OPTION"
>-N</TT
> option (<I
CLASS="FIRSTTERM"
>read -N
chars</I
>), which causes the <I
CLASS="FIRSTTERM"
>read</I
>
to terminate after <I
CLASS="FIRSTTERM"
>chars</I
>
characters.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="READN"
></A
><P
><B
>Example 37-8. Reading N characters</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Requires Bash version -ge 4.1 ...
num_chars=61
read -N $num_chars var &#60; $0 # Read first 61 characters of script!
echo "$var"
exit
####### Output of Script #######
#!/bin/bash
# Requires Bash version -ge 4.1 ...
num_chars=61</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><A
HREF="#HEREDOCREF"
>Here documents</A
>
embedded in <A
HREF="#COMMANDSUBREF0"
> <TT
CLASS="USERINPUT"
><B
>$( ... )</B
></TT
> command substitution</A
>
constructs may terminate with a simple
<B
CLASS="COMMAND"
>)</B
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="HERECOMMSUB"
></A
><P
><B
>Example 37-9. Using a <I
CLASS="FIRSTTERM"
>here document</I
>
to set a variable</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# here-commsub.sh
# Requires Bash version -ge 4.1 ...
multi_line_var=$( cat &#60;&#60;ENDxxx
------------------------------
This is line 1 of the variable
This is line 2 of the variable
This is line 3 of the variable
------------------------------
ENDxxx)
# Rather than what Bash 4.0 requires:
#+ that the terminating limit string and
#+ the terminating close-parenthesis be on separate lines.
# ENDxxx
# )
echo "$multi_line_var"
# Bash still emits a warning, though.
# warning: here-document at line 10 delimited
#+ by end-of-file (wanted `ENDxxx')</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></LI
></UL
></DIV
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="AEN21220"
></A
>37.3.2. Bash, version 4.2</H2
><P
><A
NAME="BASH42"
></A
>Version 4.2 of Bash, released
in February, 2011, contains a number of new features and
enhancements, in addition to bugfixes.</P
><P
></P
><UL
><LI
><P
>Bash now supports the the
<TT
CLASS="REPLACEABLE"
><I
>\u</I
></TT
>
and <TT
CLASS="REPLACEABLE"
><I
>\U</I
></TT
>
<I
CLASS="FIRSTTERM"
>Unicode</I
> escape.</P
><P
><A
NAME="UNICODEREF"
></A
></P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN21232"
></A
><P
></P
><P
>Unicode is a cross-platform standard for encoding
into numerical values letters and graphic symbols.
This permits representing and displaying characters
in foreign alphabets and unusual fonts.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
><A
NAME="UNICODEREF2"
></A
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>echo -e '\u2630' # Horizontal triple bar character.
# Equivalent to the more roundabout:
echo -e "\xE2\x98\xB0"
# Recognized by earlier Bash versions.
echo -e '\u220F' # PI (Greek letter and mathematical symbol)
echo -e '\u0416' # Capital "ZHE" (Cyrillic letter)
echo -e '\u2708' # Airplane (Dingbat font) symbol
echo -e '\u2622' # Radioactivity trefoil
echo -e "The amplifier circuit requires a 100 \u2126 pull-up resistor."
unicode_var='\u2640'
echo -e $unicode_var # Female symbol
printf "$unicode_var \n" # Female symbol, with newline
# And for something a bit more elaborate . . .
# We can store Unicode symbols in an associative array,
#+ then retrieve them by name.
# Run this in a gnome-terminal or a terminal with a large, bold font
#+ for better legibility.
declare -A symbol # Associative array.
symbol[script_E]='\u2130'
symbol[script_F]='\u2131'
symbol[script_J]='\u2110'
symbol[script_M]='\u2133'
symbol[Rx]='\u211E'
symbol[TEL]='\u2121'
symbol[FAX]='\u213B'
symbol[care_of]='\u2105'
symbol[account]='\u2100'
symbol[trademark]='\u2122'
echo -ne "${symbol[script_E]} "
echo -ne "${symbol[script_F]} "
echo -ne "${symbol[script_J]} "
echo -ne "${symbol[script_M]} "
echo -ne "${symbol[Rx]} "
echo -ne "${symbol[TEL]} "
echo -ne "${symbol[FAX]} "
echo -ne "${symbol[care_of]} "
echo -ne "${symbol[account]} "
echo -ne "${symbol[trademark]} "
echo</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The above example uses the
<A
HREF="#STRQ"
><B
CLASS="COMMAND"
>$' ... '</B
></A
>
<I
CLASS="FIRSTTERM"
>string-expansion</I
> construct.</P
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><A
NAME="LASTPIPEREF"
></A
></P
><P
>When the <TT
CLASS="REPLACEABLE"
><I
>lastpipe</I
></TT
> shell option
is set, the last command in a <A
HREF="#PIPEREF"
>pipe</A
> <EM
>doesn't run in a
subshell</EM
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="LASTPIPEOPT"
></A
><P
><B
>Example 37-10. Piping input to a <A
HREF="#READREF"
>read</A
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# lastpipe-option.sh
line='' # Null value.
echo "\$line = "$line"" # $line =
echo
shopt -s lastpipe # Error on Bash version -lt 4.2.
echo "Exit status of attempting to set \"lastpipe\" option is $?"
# 1 if Bash version -lt 4.2, 0 otherwise.
echo
head -1 $0 | read line # Pipe the first line of the script to read.
# ^^^^^^^^^ Not in a subshell!!!
echo "\$line = "$line""
# Older Bash releases $line =
# Bash version 4.2 $line = #!/bin/bash</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>This option offers possible <SPAN
CLASS="QUOTE"
>"fixups"</SPAN
>
for these example scripts:
<A
HREF="#BADREAD"
>Example 34-3</A
> and
<A
HREF="#READPIPE"
>Example 15-8</A
>.</P
></LI
><LI
><P
>Negative <A
HREF="#ARRAYREF"
>array</A
> indices
permit counting backwards from the end of an array.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="NEGARRAY"
></A
><P
><B
>Example 37-11. Negative array indices</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# neg-array.sh
# Requires Bash, version -ge 4.2.
array=( zero one two three four five ) # Six-element array.
# 0 1 2 3 4 5
# -6 -5 -4 -3 -2 -1
# Negative array indices now permitted.
echo ${array[-1]} # five
echo ${array[-2]} # four
# ...
echo ${array[-6]} # zero
# Negative array indices count backward from the last element+1.
# But, you cannot index past the beginning of the array.
echo ${array[-7]} # array: bad array subscript
# So, what is this new feature good for?
echo "The last element in the array is "${array[-1]}""
# Which is quite a bit more straightforward than:
echo "The last element in the array is "${array[${#array[*]}-1]}""
echo
# And ...
index=0
let "neg_element_count = 0 - ${#array[*]}"
# Number of elements, converted to a negative number.
while [ $index -gt $neg_element_count ]; do
((index--)); echo -n "${array[index]} "
done # Lists the elements in the array, backwards.
# We have just simulated the "tac" command on this array.
echo
# See also neg-offset.sh.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></LI
><LI
><P
><A
HREF="#SUBSTREXTR01"
>Substring extraction</A
>
uses a negative <I
CLASS="FIRSTTERM"
>length</I
> parameter to
specify an offset from the <I
CLASS="FIRSTTERM"
>end</I
> of the
target string.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="NEGOFFSET"
></A
><P
><B
>Example 37-12. Negative parameter in string-extraction
construct</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Bash, version -ge 4.2
# Negative length-index in substring extraction.
# Important: It changes the interpretation of this construct!
stringZ=abcABC123ABCabc
echo ${stringZ} # abcABC123ABCabc
# Position within string: 0123456789.....
echo ${stringZ:2:3} # cAB
# Count 2 chars forward from string beginning, and extract 3 chars.
# ${string:position:length}
# So far, nothing new, but now ...
# abcABC123ABCabc
# Position within string: 0123....6543210
echo ${stringZ:3:-6} # ABC123
# ^
# Index 3 chars forward from beginning and 6 chars backward from end,
#+ and extract everything in between.
# ${string:offset-from-front:offset-from-end}
# When the "length" parameter is negative,
#+ it serves as an offset-from-end parameter.
# See also neg-array.sh.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></LI
></UL
></DIV
></DIV
></DIV
></DIV
><DIV
CLASS="CHAPTER"
><HR><H1
><A
NAME="ENDNOTES"
></A
>Chapter 38. Endnotes</H1
><DIV
CLASS="SECT1"
><H1
CLASS="SECT1"
><A
NAME="AUTHORSNOTE"
></A
>38.1. Author's Note</H1
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
><I
CLASS="FOREIGNPHRASE"
>doce ut discas</I
></I
></P
><P
><I
>(Teach, that you yourself may learn.)</I
></P
></I
></TD
></TR
></TABLE
><P
>How did I come to write a scripting book? It's a strange
tale. It seems that a few years back I needed to learn
shell scripting -- and what better way to do that than to read a
good book on the subject? I was looking to buy a tutorial and
reference covering all aspects of the subject. I was looking for a
book that would take difficult concepts, turn them inside out, and
explain them in excruciating detail, with well-commented examples.
<A
NAME="AEN21281"
HREF="#FTN.AEN21281"
><SPAN
CLASS="footnote"
>[135]</SPAN
></A
>
In fact, I was looking for <EM
>this very book</EM
>,
or something very much like it. Unfortunately, <A
HREF="#KOCHANREF"
>it didn't exist</A
>, and if I wanted it,
I'd have to write it. And so, here we are, folks.</P
><P
>That reminds me of the apocryphal story about a mad
professor. Crazy as a loon, the fellow was. At the sight of a
book, any book -- at the library, at a bookstore, anywhere --
he would become totally obsessed with the idea that he could have
written it, should have written it -- and done a better job of it
to boot. He would thereupon rush home and proceed to do just that,
write a book with the very same title. When he died some years
later, he allegedly had several thousand books to his credit,
probably putting even Asimov to shame. The books might not have
been any good, who knows, but does that really matter? Here's
a fellow who lived his dream, even if he was obsessed by it,
driven by it . . . and somehow I can't help admiring the old
coot.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="ABOUTAUTHOR"
></A
>38.2. About the Author</H1
><FONT
COLOR="RED"
>Who is this guy anyhow?</FONT
><P
><A
NAME="NOCREDS"
></A
>The author claims no credentials or
special qualifications,
<A
NAME="AEN21292"
HREF="#FTN.AEN21292"
><SPAN
CLASS="footnote"
>[136]</SPAN
></A
>
other than a compulsion to write.
<A
NAME="AEN21295"
HREF="#FTN.AEN21295"
><SPAN
CLASS="footnote"
>[137]</SPAN
></A
>
</P
><P
>This book is somewhat of a departure from his other major work,
<A
HREF="http://bash.deta.in/hmw60.zip"
TARGET="_top"
> HOW-2 Meet Women: The Shy Man's Guide to
Relationships</A
>. He has also written the <A
HREF="http://tldp.org/HOWTO/Software-Building-HOWTO.html"
TARGET="_top"
>Software-Building
HOWTO</A
>. Of late, he has been trying his
(heavy) hand at fiction: <A
HREF="http://bash.deta.in/dave-dawson-over-berlin.epub"
TARGET="_top"
>Dave Dawson
Over Berlin (First Installment)</A
>
<A
HREF="http://bash.deta.in/dave-dawson-over-berlin.II.epub"
TARGET="_top"
>Dave Dawson
Over Berlin (Second Installment)</A
> and
<A
HREF="http://bash.deta.in/dave-dawson-over-berlin.III.epub"
TARGET="_top"
>Dave Dawson
Over Berlin (Third Installment)</A
>
. He also has a few
<EM
>Instructables</EM
> (<A
HREF="http://www.instructables.com/id/Arduino-Morse-Code-Shield/"
TARGET="_top"
>here</A
>,
<A
HREF="http://www.instructables.com/id/Haywired-Hackduino/"
TARGET="_top"
>here</A
>,
<A
HREF="http://www.instructables.com/id/Arduino-DIY-SD-Card-Logging-Shield/"
TARGET="_top"
>here</A
>,
<A
HREF="http://www.instructables.com/id/Binguino-An-Arduino-based-Bingo-Number-Generato/"
TARGET="_top"
>here</A
>,
<A
HREF="http://www.instructables.com/id/The-Raspberry-Pi-Lapdock-Connection/"
TARGET="_top"
>here</A
>,
<A
HREF="http://www.instructables.com/id/The-Raspberry-Pi-Arduino-Connection/"
TARGET="_top"
>here</A
>, and
<A
HREF="http://www.instructables.com/id/Switchable-Dual-Voltage-33v5v-Hacduino/"
TARGET="_top"
>here</A
>
to his (dis)credit.</P
><P
>A Linux user since 1995 (Slackware 2.2, kernel 1.2.1),
the author has emitted a few
software truffles, including the <A
HREF="http://ibiblio.org/pub/Linux/utils/file/cruft-0.2.tar.gz"
TARGET="_top"
>cruft</A
>
one-time pad encryption utility, the <A
HREF="http://ibiblio.org/pub/Linux/apps/financial/mcalc-1.6.tar.gz"
TARGET="_top"
>mcalc</A
>
mortgage calculator, the <A
HREF="http://ibiblio.org/pub/Linux/games/amusements/judge-1.0.tar.gz"
TARGET="_top"
>judge</A
>
Scrabble<6C> adjudicator, the <A
HREF="http://ibiblio.org/pub/Linux/libs/yawl-0.3.2.tar.gz"
TARGET="_top"
>yawl</A
>
word gaming list package, and the <A
HREF="http://bash.deta.in/qky.README.html"
TARGET="_top"
>Quacky</A
>
anagramming gaming package. He got off to a rather shaky start in the
computer game -- programming FORTRAN IV on a CDC 3800 (on paper coding
pads, with occasional forays on a keypunch machine and a Friden
Flexowriter) -- and is not the least bit nostalgic for those
days.</P
><P
>Living in an out-of-the-way community with wife and orange
tabby, he cherishes human frailty, especially his own.
<A
NAME="AEN21318"
HREF="#FTN.AEN21318"
><SPAN
CLASS="footnote"
>[138]</SPAN
></A
>
</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="WHEREHELP"
></A
>38.3. Where to Go For Help</H1
><P
><A
HREF="mailto:thegrendel.abs@gmail.com"
TARGET="_top"
>The author</A
>
is no longer supporting or updating this document. He will not
answer questions about this book or about general scripting topics.
</P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN21326"
></A
><P
></P
><P
>If you need assistance with a schoolwork assignment,
read the pertinent sections of this and other reference works.
Do your best to solve the problem using your own wits and
resources. <EM
>Please do not waste the author's time.</EM
>
You will get neither help nor sympathy.
<A
NAME="AEN21329"
HREF="#FTN.AEN21329"
><SPAN
CLASS="footnote"
>[139]</SPAN
></A
>
</P
><P
>Likewise, kindly refrain from annoying the author
with solicitations, offers of employment, or <SPAN
CLASS="QUOTE"
>"business
opportunities."</SPAN
> He is doing just fine, and requires
neither help nor sympathy, thank you.</P
><P
>Please note that the author will <EM
>not</EM
> answer
scripting questions for Sun/Solaris/Oracle or Apple systems. The
endarkened execs and the arachnoid corporate attorneys of
those particular outfits have been using litigation in a
predatory manner and/or as a weapon against the Open Source
Community. Any Solaris or Apple users needing scripting help
will therefore kindly direct their concerns to corporate
customer service.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>... sophisticated in mechanism but possibly agile
operating under noises being extremely suppressed ...</I
></P
><P
><I
>--<EM
>CI-300 printer manual</EM
></I
></P
></I
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="TOOLSUSED"
></A
>38.4. Tools Used to Produce This Book</H1
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="SOFTWARE-HARDWARE"
></A
>38.4.1. Hardware</H2
><P
>A used IBM Thinkpad, model 760XL laptop (P166, 104 meg RAM)
running Red Hat 7.1/7.3. Sure, it's slow and has a funky
keyboard, but it beats the heck out of a No. 2 pencil and a
Big Chief tablet.</P
><P
> <EM
>Update:</EM
> upgraded to a
770Z Thinkpad (P2-366, 192 meg RAM) running FC3. Anyone
feel like donating a later-model laptop to a starving writer
&#60;g&#62;?
</P
><P
> <EM
>Update:</EM
> upgraded to a T61 Thinkpad
running Mandriva 2011. No longer starving &#60;g&#62;,
but not too proud to accept donations.
</P
></DIV
><DIV
CLASS="SECT2"
><HR><H2
CLASS="SECT2"
><A
NAME="SOFTWARE-PRINTWARE"
></A
>38.4.2. Software and Printware</H2
><P
></P
><OL
TYPE="i"
><LI
><P
>Bram Moolenaar's powerful SGML-aware <A
HREF="http://www.vim.org"
TARGET="_top"
>vim</A
> text editor.</P
></LI
><LI
><P
><A
HREF="http://www.netfolder.com/DSSSL/"
TARGET="_top"
>OpenJade</A
>,
a DSSSL rendering engine for converting SGML documents into other
formats.</P
></LI
><LI
><P
><A
HREF="http://nwalsh.com/docbook/dsssl/"
TARGET="_top"
> Norman
Walsh's DSSSL stylesheets</A
>.</P
></LI
><LI
><P
><I
CLASS="CITETITLE"
>DocBook, The Definitive
Guide</I
>, by Norman Walsh and Leonard Muellner
(O'Reilly, ISBN 1-56592-580-7). This is still the standard
reference for anyone attempting to write a document in Docbook
SGML format.</P
></LI
></OL
></DIV
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="CREDITS"
></A
>38.5. Credits</H1
><P
><EM
>Community participation made this project
possible.</EM
> The author gratefully acknowledges that
writing this book would have been unthinkable without
help and feedback from all you people out there.</P
><P
><A
HREF="mailto:feloy@free.fr"
TARGET="_top"
>Philippe Martin</A
>
translated the first version (0.1) of this document into
DocBook/SGML. While not on the job at a small French company as a
software developer, he enjoys working on GNU/Linux documentation
and software, reading literature, playing music, and, for his
peace of mind, making merry with friends. You may run across him
somewhere in France or in the Basque Country, or you can email him
at <A
HREF="mailto:feloy@free.fr"
TARGET="_top"
>feloy@free.fr</A
>.</P
><P
>Philippe Martin also pointed out that positional parameters
past $9 are possible using {bracket} notation. (See <A
HREF="#EX17"
>Example 4-5</A
>).</P
><P
><A
HREF="mailto:stephane_chazelas@yahoo.fr"
TARGET="_top"
>St<53>phane
Chazelas</A
> sent a long list of corrections, additions,
and example scripts. More than a contributor, he had, in effect,
for a while taken on the role of <EM
>co-editor</EM
>
for this document. <I
CLASS="FOREIGNPHRASE"
>Merci
beaucoup!</I
></P
><P
>Paulo Marcel Coelho Aragao offered many corrections, both major
and minor, and contributed quite a number of helpful
suggestions.</P
><P
>I would like to especially thank <EM
>Patrick
Callahan</EM
>, <EM
>Mike Novak</EM
>, and
<EM
>Pal Domokos</EM
> for catching bugs, pointing out
ambiguities, and for suggesting clarifications and changes in the
preliminary version (0.1) of this document. Their lively
discussion of shell scripting and general documentation issues
inspired me to try to make this document more readable.</P
><P
>I'm grateful to Jim Van Zandt for pointing out errors and
omissions in version 0.2 of this document. He also contributed
an instructive <A
HREF="#ZFIFO"
>example script</A
>.</P
><P
>Many thanks to <A
HREF="mailto:mikaku@fiwix.org"
TARGET="_top"
>Jordi Sanfeliu</A
>
for giving permission to use his fine tree script (<A
HREF="#TREE"
>Example A-16</A
>), and to Rick Boivie for revising it.</P
><P
>Likewise, thanks to <A
HREF="mailto:charpov@cs.unh.edu"
TARGET="_top"
>Michel Charpentier</A
> for
permission to use his <A
HREF="#DCREF"
>dc</A
> factoring script
(<A
HREF="#FACTR"
>Example 16-52</A
>).</P
><P
>Kudos to <A
HREF="mailto:friedman@prep.ai.mit.edu"
TARGET="_top"
>Noah Friedman</A
>
for permission to use his string function script (<A
HREF="#STRING"
>Example A-18</A
>).</P
><P
><A
HREF="mailto:emmanuel.rouat@wanadoo.fr"
TARGET="_top"
>Emmanuel
Rouat</A
> suggested corrections and additions on <A
HREF="#COMMANDSUBREF"
>command substitution</A
>,
<A
HREF="#ALIASREF"
>aliases</A
>, and <A
HREF="#PATHMANAGEMENT"
>path management</A
>. He also
contributed a very nice sample <TT
CLASS="FILENAME"
>.bashrc</TT
> file
(<A
HREF="#SAMPLE-BASHRC"
>Appendix M</A
>).</P
><P
><A
HREF="mailto:heiner.steven@odn.de"
TARGET="_top"
>Heiner Steven</A
>
kindly gave permission to use his base conversion script, <A
HREF="#BASE"
>Example 16-48</A
>. He also made a number of corrections and many
helpful suggestions. Special thanks.</P
><P
>Rick Boivie contributed the delightfully recursive
<EM
>pb.sh</EM
> script (<A
HREF="#PBOOK"
>Example 36-11</A
>),
revised the <EM
>tree.sh</EM
> script (<A
HREF="#TREE"
>Example A-16</A
>), and suggested performance improvements
for the <EM
>monthlypmt.sh</EM
> script (<A
HREF="#MONTHLYPMT"
>Example 16-47</A
>).</P
><P
>Florian Wisser enlightened me on some of the fine points of
testing strings (see <A
HREF="#STRTEST"
>Example 7-6</A
>), and on other
matters.</P
><P
>Oleg Philon sent suggestions concerning <A
HREF="#CUTREF"
>cut</A
> and <A
HREF="#PIDOFREF"
>pidof</A
>.</P
><P
>Michael Zick extended the <A
HREF="#EMPTYARRAY"
>empty
array</A
> example to demonstrate some surprising array
properties. He also contributed the <EM
>isspammer</EM
>
scripts (<A
HREF="#ISSPAMMER"
>Example 16-41</A
> and <A
HREF="#ISSPAMMER2"
>Example A-28</A
>).</P
><P
>Marc-Jano Knopp sent corrections and clarifications on DOS
batch files.</P
><P
>Hyun Jin Cha found several typos in the document in the
process of doing a Korean translation. Thanks for pointing
these out.</P
><P
>Andreas Abraham sent in a long list of typographical
errors and other corrections. Special thanks!</P
><P
>Others contributing scripts, making helpful suggestions, and
pointing out errors were Gabor Kiss, Leopold Toetsch,
Peter Tillier, Marcus Berglof, Tony Richardson, Nick Drage
(script ideas!), Rich Bartell, Jess Thrysoee, Adam Lazur, Bram
Moolenaar, Baris Cicek, Greg Keraunen, Keith Matthews, Sandro
Magi, Albert Reiner, Dim Segebart, Rory Winston, Lee Bigelow,
Wayne Pollock, <SPAN
CLASS="QUOTE"
>"jipe,"</SPAN
> <SPAN
CLASS="QUOTE"
>"bojster,"</SPAN
>
<SPAN
CLASS="QUOTE"
>"nyal,"</SPAN
> <SPAN
CLASS="QUOTE"
>"Hobbit,"</SPAN
> <SPAN
CLASS="QUOTE"
>"Ender,"</SPAN
>
<SPAN
CLASS="QUOTE"
>"Little Monster"</SPAN
> (Alexis), <SPAN
CLASS="QUOTE"
>"Mark,"</SPAN
>
<SPAN
CLASS="QUOTE"
>"Patsie,"</SPAN
> <SPAN
CLASS="QUOTE"
>"vladz,"</SPAN
> Peggy Russell,
Emilio Conti, Ian. D. Allen, Hans-Joerg Diers, Arun Giridhar,
Dennis Leeuw, Dan Jacobson, Aurelio Marinho Jargas, Edward
Scholtz, Jean Helou, Chris Martin, Lee Maschmeyer, Bruno Haible,
Wilbert Berendsen, Sebastien Godard, Bj<42>n Eriksson, John
MacDonald, John Lange, Joshua Tschida, Troy Engel, Manfred
Schwarb, Amit Singh, Bill Gradwohl, E. Choroba, David Lombard,
Jason Parker, Steve Parker, Bruce W. Clare, William Park, Vernia
Damiano, Mihai Maties, Mark Alexander, Jeremy Impson, Ken Fuchs,
Jared Martin, Frank Wang, Sylvain Fourmanoit, Matthew Sage,
Matthew Walker, Kenny Stauffer, Filip Moritz, Andrzej Stefanski,
Daniel Albers, Jeffrey Haemer, Stefano Palmeri, Nils Radtke,
Sigurd Solaas, Serghey Rodin, Jeroen Domburg, Alfredo Pironti,
Phil Braham, Bruno de Oliveira Schneider, Stefano Falsetto,
Chris Morgan, Walter Dnes, Linc Fessenden, Michael Iatrou, Pharis
Monalo, Jesse Gough, Fabian Kreutz, Mark Norman, Harald Koenig,
Dan Stromberg, Peter Knowles, Francisco Lobo, Mariusz Gniazdowski,
Sebastian Arming, Chetankumar Phulpagare, Benno Schulenberg,
Tedman Eng, Jochen DeSmet, Juan Nicolas Ruiz, Oliver Beckstein,
Achmed Darwish, Dotan Barak, Richard Neill, Albert Siersema,
Omair Eshkenazi, Geoff Lee, Graham Ewart, JuanJo Ciarlante,
Cliff Bamford, Nathan Coulter, Ramses Rodriguez Martinez,
Evgeniy Ivanov, Craig Barnes, George Dimitriu, Kevin LeBlanc,
Antonio Macchi, Tomas Pospisek, David Wheeler, Erik Brandsberg,
YongYe, Andreas K<>hne, P<>draig Brady, Joseph
Steinhauser, and David Lawyer (himself an author of four
HOWTOs).</P
><P
>My gratitude to <A
HREF="mailto:chet@po.cwru.edu"
TARGET="_top"
>Chet
Ramey</A
> and Brian Fox for writing <I
CLASS="FIRSTTERM"
>Bash</I
>,
and building into it elegant and powerful scripting
capabilities rivaling those of <I
CLASS="FIRSTTERM"
>ksh</I
>.</P
><P
>Very special thanks to the hard-working volunteers at
the <A
HREF="http://www.tldp.org"
TARGET="_top"
>Linux Documentation
Project</A
>. The LDP hosts a repository of Linux knowledge
and lore, and has, to a great extent, enabled the publication
of this book.</P
><P
>Thanks and appreciation to IBM, Red Hat, Google, the <A
HREF="http://www.fsf.org"
TARGET="_top"
>Free Software Foundation</A
>, and
all the good people fighting the good fight to keep Open Source
software free and open.</P
><P
>Belated thanks to my fourth grade teacher, Miss Spencer,
for emotional support and for convincing me that maybe, just
maybe I wasn't a total loss.</P
><P
>Thanks most of all to my wife, Anita, for her encouragement,
inspiration, and emotional support.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="DISCLAIMER"
></A
>38.6. Disclaimer</H1
><P
>(This is a variant of the standard <A
HREF="http://www.tldp.org"
TARGET="_top"
>LDP</A
> disclaimer.)</P
><P
>No liability for the contents of this document can be
accepted. Use the concepts, examples and information at your
own risk. There may be errors, omissions, and inaccuracies
that could cause you to lose data, harm your system, or induce
involuntary electrocution, so <EM
>proceed with appropriate
caution</EM
>. The author takes no responsibility for any
damages, incidental or otherwise.</P
><P
>As it happens, it is highly unlikely that either you or
your system will suffer ill effects, aside from uncontrollable
hiccups. In fact, the <I
CLASS="FOREIGNPHRASE"
>raison
d'etre</I
> of this book is to enable its readers
to analyze shell scripts and determine whether they have <A
HREF="#GOTCHAS"
>unanticipated consequences</A
>.</P
></DIV
></DIV
><A
NAME="BIBLIO"
></A
><HR><H1
><A
NAME="BIBLIO"
></A
>Bibliography</H1
><A
NAME="BIBLIOREF"
></A
><TABLE
BORDER="0"
WIDTH="100%"
CELLSPACING="0"
CELLPADDING="0"
CLASS="EPIGRAPH"
><TR
><TD
WIDTH="45%"
>&nbsp;</TD
><TD
WIDTH="45%"
ALIGN="LEFT"
VALIGN="TOP"
><I
><P
><I
>Those who do not understand UNIX are condemned to reinvent it,
poorly.</I
></P
><P
><I
>--Henry Spencer</I
></P
><P
><I
><A
NAME="DENNINGREF"
></A
></I
></P
></I
></TD
></TR
></TABLE
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21461"
></A
><P
><SPAN
CLASS="EDITOR"
>Edited by Peter Denning</SPAN
>, <I
>Computers Under Attack: Intruders, Worms, and Viruses</I
>, ACM Press, 1990, 0-201-53067-8.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>This compendium contains a couple of articles on
shell script viruses.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21475"
></A
><P
><SPAN
CLASS="AUTHOR"
>Ken Burtch</SPAN
>, <I
><A
HREF="http://www.samspublishing.com/title/0672326426"
TARGET="_top"
>Linux Shell Scripting with Bash</A
></I
>, 1st edition, Sams Publishing (Pearson), 2004, 0672326426.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
> Covers much of the same material as the <EM
>ABS
Guide</EM
>, though in a different style.</P
><P
>*</P
><P
><A
NAME="DGSEDREF"
></A
></P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21494"
></A
><P
><SPAN
CLASS="AUTHOR"
>Daniel Goldman</SPAN
>, <I
><A
HREF="http://www.sed-book.com/"
TARGET="_top"
>Definitive Guide
to Sed</A
></I
>, 1st edition, 2013.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
> This ebook is an excellent introduction to
<I
CLASS="FIRSTTERM"
>sed</I
>. Rather than being a conversion from
a printed volume, it was specifically designed and formatted
for viewing on an ebook reader. Well-written, informative,
and useful as a reference as well as a tutorial. Highly
recommended.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21508"
></A
><P
><SPAN
CLASS="AUTHOR"
>Dale Dougherty </SPAN
><SPAN
CLASS="AUTHOR"
>and Arnold Robbins</SPAN
>, <I
>Sed and Awk</I
>, 2nd edition, O'Reilly and Associates, 1997, 1-156592-225-5.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
> Unfolding the full power of shell scripting requires at least a passing
familiarity with <A
HREF="#SEDREF"
><I
CLASS="FIRSTTERM"
>sed</I
>
and <I
CLASS="FIRSTTERM"
>awk</I
></A
>. This is the classic
tutorial. It includes an excellent introduction to
<I
CLASS="FIRSTTERM"
>Regular Expressions</I
>. Recommended.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21530"
></A
><P
><SPAN
CLASS="AUTHOR"
>Jeffrey Friedl</SPAN
>, <I
>Mastering Regular Expressions</I
>, O'Reilly and Associates, 2002, 0-596-00289-0.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Still the best all-around reference on <A
HREF="#REGEXREF"
>Regular Expressions</A
>.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21545"
></A
><P
><SPAN
CLASS="AUTHOR"
>Aeleen Frisch</SPAN
>, <I
>Essential System Administration</I
>, 3rd edition, O'Reilly and Associates, 2002, 0-596-00343-9.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>This excellent manual provides a decent introduction
to shell scripting from a sys admin point of view. It includes
comprehensive explanations of the startup and initialization
scripts in a UNIX system.</P
><P
>*</P
><P
><A
NAME="KOCHANREF"
></A
></P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21562"
></A
><P
><SPAN
CLASS="AUTHOR"
>Stephen Kochan </SPAN
><SPAN
CLASS="AUTHOR"
>and Patrick Wood</SPAN
>, <I
>Unix Shell Programming</I
>, Hayden, 1990, 067248448X.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Still considered a standard reference, though somewhat dated, and
a bit <SPAN
CLASS="QUOTE"
>"wooden"</SPAN
> stylistically speaking.
<A
NAME="AEN21579"
HREF="#FTN.AEN21579"
><SPAN
CLASS="footnote"
>[140]</SPAN
></A
>
In fact, this book was the <I
CLASS="FIRSTTERM"
>ABS Guide</I
> author's
first exposure to UNIX shell scripting, lo these many years ago.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21583"
></A
><P
><SPAN
CLASS="AUTHOR"
>Neil Matthew </SPAN
><SPAN
CLASS="AUTHOR"
>and Richard Stones</SPAN
>, <I
>Beginning Linux Programming</I
>, Wrox Press, 1996, 1874416680.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Surprisingly good in-depth coverage of various
programming languages available for Linux, including a fairly
strong chapter on shell scripting.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="MAYERREF"
></A
><P
><SPAN
CLASS="AUTHOR"
>Herbert Mayer</SPAN
>, <I
>Advanced C Programming on the IBM PC</I
>, Windcrest Books, 1989, 0830693637.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Excellent coverage of algorithms and general
programming practices. Highly recommended, but unfortunately
out of print.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21614"
></A
><P
><SPAN
CLASS="AUTHOR"
>David Medinets</SPAN
>, <I
>Unix Shell Programming Tools</I
>, McGraw-Hill, 1999, 0070397333.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Pretty good treatment of shell scripting, with
examples, and a short intro to Tcl and Perl.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21628"
></A
><P
><SPAN
CLASS="AUTHOR"
>Cameron Newham </SPAN
><SPAN
CLASS="AUTHOR"
>and Bill Rosenblatt</SPAN
>, <I
>Learning the Bash Shell</I
>, 2nd edition, O'Reilly and Associates, 1998, 1-56592-347-2.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>This is a valiant effort at a decent shell primer,
but sadly deficient in its coverage of writing scripts and
lacking sufficient examples.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21646"
></A
><P
><SPAN
CLASS="AUTHOR"
>Anatole Olczak</SPAN
>, <I
>Bourne Shell Quick Reference Guide</I
>, ASP, Inc., 1991, 093573922X.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>A very handy pocket reference, despite lacking
coverage of Bash-specific features.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21660"
></A
><P
><SPAN
CLASS="AUTHOR"
>Jerry Peek, </SPAN
><SPAN
CLASS="AUTHOR"
>Tim O'Reilly, </SPAN
><SPAN
CLASS="AUTHOR"
>and Mike Loukides</SPAN
>, <I
>Unix Power Tools</I
>, 3rd edition, O'Reilly and Associates, Random House, 2002, 0-596-00330-7.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Contains a couple of sections of very informative
in-depth articles on shell programming, but falls short of being
a self-teaching manual. It reproduces much of the <A
HREF="#REGEXREF"
>Regular Expressions</A
> tutorial from the
Dougherty and Robbins book, above. The comprehensive coverage
of UNIX commands makes this book worthy of a place on your
bookshelf.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21684"
></A
><P
><SPAN
CLASS="AUTHOR"
>Clifford Pickover</SPAN
>, <I
>Computers, Pattern, Chaos, and Beauty</I
>, St. Martin's Press, 1990, 0-312-04123-3.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>A treasure trove of ideas and recipes for
computer-based exploration of mathematical oddities.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21698"
></A
><P
><SPAN
CLASS="AUTHOR"
>George Polya</SPAN
>, <I
>How To Solve It</I
>, Princeton University Press, 1973, 0-691-02356-5.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>The classic tutorial on problem-solving methods
(algorithms), with special emphasis on how to teach them.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21712"
></A
><P
><SPAN
CLASS="AUTHOR"
>Chet Ramey </SPAN
><SPAN
CLASS="AUTHOR"
>and Brian Fox</SPAN
>, <I
><A
HREF="http://www.network-theory.co.uk/bash/manual/"
TARGET="_top"
>The GNU Bash Reference Manual</A
></I
>, Network Theory Ltd, 2003, 0-9541617-7-7.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>This manual is the definitive reference for
GNU Bash. The authors of this manual, Chet Ramey and Brian Fox,
are the original developers of GNU Bash. For each copy sold,
the publisher donates $1 to the Free Software Foundation.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21730"
></A
><P
><SPAN
CLASS="AUTHOR"
>Arnold Robbins</SPAN
>, <I
>Bash Reference Card</I
>, SSC, 1998, 1-58731-010-5.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Excellent Bash pocket reference (don't leave home without it,
especially if you're a sysadmin). A bargain at $4.95, but
unfortunately no longer available for free download.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21744"
></A
><P
><SPAN
CLASS="AUTHOR"
>Arnold Robbins</SPAN
>, <I
>Effective Awk Programming</I
>, Free Software Foundation / O'Reilly and Associates, 2000, 1-882114-26-4.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>The absolute best <A
HREF="#AWKREF"
>awk</A
>
tutorial and reference. The free electronic version of this
book is part of the <I
CLASS="FIRSTTERM"
>awk</I
> documentation,
and printed copies of the latest version are available from
O'Reilly and Associates.</P
><P
>This book served as an inspiration for the author
of the <I
CLASS="FIRSTTERM"
>ABS Guide</I
>.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21762"
></A
><P
><SPAN
CLASS="AUTHOR"
>Bill Rosenblatt</SPAN
>, <I
>Learning the Korn Shell</I
>, O'Reilly and Associates, 1993, 1-56592-054-6.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><A
NAME="KORNSHELLREF"
></A
>This well-written book contains
some excellent pointers on shell scripting in general.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21777"
></A
><P
><SPAN
CLASS="AUTHOR"
>Paul Sheer</SPAN
>, <I
>LINUX: Rute User's Tutorial and Exposition</I
>, 1st edition, , 2002, 0-13-033351-4.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Very detailed and readable introduction to Linux system
administration.</P
><P
>The book is available in print, or
<A
HREF="http://burks.brighton.ac.uk/burks/linux/rute/rute.htm"
TARGET="_top"
>on-line</A
>.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21794"
></A
><P
><SPAN
CLASS="AUTHOR"
>Ellen Siever </SPAN
><SPAN
CLASS="AUTHOR"
>and the staff of O'Reilly and Associates</SPAN
>, <I
>Linux in a Nutshell</I
>, 2nd edition, O'Reilly and Associates, 1999, 1-56592-585-8.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>The all-around best Linux command reference.
It even has a Bash section.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21811"
></A
><P
><SPAN
CLASS="AUTHOR"
>Dave Taylor</SPAN
>, <I
>Wicked Cool Shell Scripts: 101 Scripts for Linux, Mac OS X, and Unix Systems</I
>, 1st edition, No Starch Press, 2004, 1-59327-012-7.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Pretty much what the title promises . . .</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21826"
></A
><P
><I
>The UNIX CD Bookshelf</I
>, 3rd edition, O'Reilly and Associates, 2003, 0-596-00392-7.</P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>An array of seven UNIX books on CD ROM, including
<I
CLASS="CITETITLE"
>UNIX Power Tools</I
>,
<I
CLASS="CITETITLE"
>Sed and Awk</I
>, and <I
CLASS="CITETITLE"
>Learning the Korn Shell</I
>. A complete
set of all the UNIX references and tutorials you would ever need
at about $130. Buy this one, even if it means going into debt
and not paying the rent.</P
><P
>Update: Seems to have somehow fallen out of print.
Ah, well. You can still buy the dead-tree editions of these books.</P
><P
>*</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21841"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>The O'Reilly books on Perl. (Actually,
<EM
>any</EM
> O'Reilly books.)</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21845"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>* * *</P
><P
><B
CLASS="COMMAND"
>Other Resources</B
></P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21850"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Fioretti, Marco, <SPAN
CLASS="QUOTE"
>"Scripting for X
Productivity,"</SPAN
> <A
HREF="linuxjournal.com"
TARGET="_top"
><I
CLASS="CITETITLE"
>Linux Journal</I
></A
>, Issue 113,
September, 2003, pp. 86-9.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21856"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Ben Okopnik's well-written <I
CLASS="CITETITLE"
>introductory Bash scripting</I
>
articles in issues 53, 54, 55, 57, and 59 of the
<A
HREF="http://www.linuxgazette.net"
TARGET="_top"
><I
CLASS="CITETITLE"
>Linux Gazette</I
></A
>, and his
explanation of <SPAN
CLASS="QUOTE"
>"The Deep, Dark Secrets of Bash"</SPAN
>
in issue 56.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21863"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Chet Ramey's <I
CLASS="CITETITLE"
>Bash -
The GNU Shell</I
>, a two-part series published in issues 3
and 4 of the <A
HREF="http://www.linuxjournal.com"
TARGET="_top"
><I
CLASS="CITETITLE"
>Linux Journal</I
></A
>, July-August
1994.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21869"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Mike G's <A
HREF="http://www.tldp.org/HOWTO/Bash-Prog-Intro-HOWTO.html"
TARGET="_top"
>Bash-Programming-Intro
HOWTO</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21873"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Richard's <A
HREF="http://www.injunea.demon.co.uk/index.htm"
TARGET="_top"
>Unix
Scripting Universe</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21877"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><A
NAME="BASHFAQ"
></A
>Chet Ramey's
<A
HREF="http://tiswww.case.edu/php/chet/bash/FAQ"
TARGET="_top"
>Bash
FAQ</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21882"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><A
HREF="http://wooledge.org:8000/BashFAQ"
TARGET="_top"
> Greg's WIKI: Bash FAQ</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21886"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Example shell scripts at <A
HREF="http://alge.anart.no/linux/scripts/"
TARGET="_top"
>Lucc's Shell Scripts
</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21890"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Example shell scripts at <A
HREF="http://www.shelldorado.com"
TARGET="_top"
>SHELLdorado </A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21894"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Example shell scripts at <A
HREF="http://www.splode.com/~friedman/software/scripts/src/"
TARGET="_top"
>Noah
Friedman's script site</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21898"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><A
HREF="http://bashcookbook.com/bashinfo/"
TARGET="_top"
>Examples</A
>
from the <I
CLASS="CITETITLE"
>The Bash Scripting
Cookbook</I
>, by Albing, Vossen, and Newham.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21903"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Example shell scripts at <A
HREF="http://www.zazzybob.com"
TARGET="_top"
>zazzybob</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21907"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Steve Parker's <A
HREF="http://steve-parker.org/sh/sh.shtml"
TARGET="_top"
>Shell Programming
Stuff</A
>. In fact, all of his shell scripting
books are highly recommended. See also Steve's <A
HREF="http://nixshell.wordpress.com/2011/07/13/arcade-games-written-in-a-shell-script/"
TARGET="_top"
>Arcade
Games written in a shell script</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21912"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>An excellent collection of Bash scripting tips, tricks,
and resources at the <A
HREF="http://www.bash-hackers.org/wiki.doku.php"
TARGET="_top"
>Bash Hackers
Wiki</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21916"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Giles Orr's <A
HREF="http://www.tldp.org/HOWTO/Bash-Prompt-HOWTO/"
TARGET="_top"
>Bash-Prompt
HOWTO</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21920"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>The <A
HREF="http://www.pixelbeat.org/cmdline.html"
TARGET="_top"
><EM
>Pixelbeat</EM
>
command-line reference</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21925"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Very nice <B
CLASS="COMMAND"
>sed</B
>,
<B
CLASS="COMMAND"
>awk</B
>, and regular expression tutorials at
<A
HREF="http://www.grymoire.com/Unix/index.html"
TARGET="_top"
>The UNIX
Grymoire</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21931"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>The GNU <A
HREF="http://www.gnu.org/software/sed/manual/"
TARGET="_top"
>sed</A
>
and
<A
HREF="http://www.gnu.org/software/gawk/manual/"
TARGET="_top"
> gawk</A
> manuals. As you recall, <A
HREF="#GNUGAWK"
>gawk</A
> is the enhanced GNU version of
<B
CLASS="COMMAND"
>awk</B
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21938"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Many interesting sed scripts at the <A
HREF="http://sed.sourceforge.net/grabbag/"
TARGET="_top"
> seder's grab bag</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21942"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Tips and tricks at <A
HREF="http://linuxreviews.org"
TARGET="_top"
> Linux Reviews</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21946"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Trent Fisher's <A
HREF="http://www.cs.pdx.edu/~trent/gnu/groff/groff.html"
TARGET="_top"
>groff
tutorial</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21950"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>David Wheeler's <A
HREF="http://www.dwheeler.com/essays/filenames-in-shell.html"
TARGET="_top"
>Filenames
in Shell</A
> essay.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21954"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><SPAN
CLASS="QUOTE"
>"Shelltris"</SPAN
> and <SPAN
CLASS="QUOTE"
>"shellitaire"</SPAN
>
at <A
HREF="http://www.shellscriptgames.com"
TARGET="_top"
>Shell Script
Games</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21960"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>YongYe's wonderfully complex <A
HREF="http://bash.deta.in/Tetris_Game.sh"
TARGET="_top"
>Tetris game
script</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21964"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Mark Komarinski's <A
HREF="http://www.tldp.org/HOWTO/Printing-Usage-HOWTO.html"
TARGET="_top"
>Printing-Usage
HOWTO</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21968"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><A
HREF="http://www.linux-usb.org/USB-guide/book1.html"
TARGET="_top"
>The
Linux USB subsystem</A
> (helpful in writing scripts affecting
USB peripherals).</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21972"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>There is some nice material on <A
HREF="#IOREDIRREF"
>I/O redirection</A
> in <A
HREF="http://sunsite.ualberta.ca/Documentation/Gnu/textutils-2.0/html_chapter/textutils_10.html"
TARGET="_top"
> chapter 10 of the textutils documentation</A
> at the <A
HREF="http://sunsite.ualberta.ca/Documentation"
TARGET="_top"
> University of
Alberta site</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21978"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><A
HREF="mailto:humbubba@smarty.smart.net"
TARGET="_top"
>Rick
Hohensee</A
> has written the
<I
CLASS="FIRSTTERM"
>osimpa</I
> i386 assembler
entirely as Bash scripts.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21983"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><I
CLASS="FIRSTTERM"
>dgatwood</I
>
has a very nice <A
HREF="http://www.shellscriptgames.com/"
TARGET="_top"
> shell script games</A
> site, featuring a Tetris<69>
clone and solitaire.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21988"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Aurelio Marinho Jargas has written a <A
HREF="http://txt2regex.sf.net"
TARGET="_top"
>Regular expression
wizard</A
>. He has also written an informative <A
HREF="http://guia-er.sf.net"
TARGET="_top"
>book</A
> on Regular Expressions,
in Portuguese.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21993"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><A
HREF="mailto:brtompkins@comcast.net"
TARGET="_top"
>Ben
Tomkins</A
> has created the <A
HREF="http://bashnavigator.sourceforge.net"
TARGET="_top"
> Bash Navigator</A
> directory management tool.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN21998"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><A
HREF="mailto:opengeometry@yahoo.ca"
TARGET="_top"
>William Park</A
>
has been working on a project
to incorporate certain <I
CLASS="FIRSTTERM"
>Awk</I
> and
<I
CLASS="FIRSTTERM"
>Python</I
> features into Bash. Among these is
a <I
CLASS="FIRSTTERM"
>gdbm</I
> interface. He has released
<I
CLASS="FIRSTTERM"
>bashdiff</I
>
on <A
HREF="http://freshmeat.net"
TARGET="_top"
>Freshmeat.net</A
>. He
has an <A
HREF="http://linuxgazette.net/108/park.html"
TARGET="_top"
>article</A
>
in the November, 2004 issue of the <A
HREF="http://www.linuxgazette.net"
TARGET="_top"
><I
CLASS="CITETITLE"
>Linux Gazette</I
></A
>
on adding string functions to Bash, with a <A
HREF="http://linuxgazette.net/109/park.html"
TARGET="_top"
>followup
article</A
> in the December issue, and <A
HREF="http://linuxgazette.net/110/park.htm"
TARGET="_top"
>yet another</A
>
in the January, 2005 issue.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22012"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Peter Knowles has written an
<A
HREF="http://booklistgensh.peterknowles.com/"
TARGET="_top"
>elaborate
Bash script</A
> that generates a book list on the <A
HREF="http://www.dottocomu.com/b/archives/002571.html"
TARGET="_top"
>Sony
Librie</A
> e-book reader. This useful tool facilitates
loading non-DRM user content on the <EM
>Librie</EM
>
(and the newer <EM
>PRS-xxx-series</EM
> devices).</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22019"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Tim Waugh's <A
HREF="http://cyberelk.net/tim/xmlto/"
TARGET="_top"
>xmlto</A
> is an
elaborate Bash script for converting Docbook XML documents to
other formats.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22023"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Philip Patterson's <A
HREF="http://www.gossiplabs.org"
TARGET="_top"
>logforbash</A
>
logging/debugging script.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22027"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><A
HREF="http://auctiongallery.sourceforge.net"
TARGET="_top"
>AuctionGallery</A
>,
an application for eBay <SPAN
CLASS="QUOTE"
>"power sellers"</SPAN
> coded
in Bash.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22032"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Of historical interest are Colin Needham's
<I
CLASS="FIRSTTERM"
>original International Movie Database (IMDB)
reader polling scripts</I
>, which nicely illustrate
the use of <A
HREF="#AWKREF"
>awk</A
> for string
parsing. Unfortunately, the URL link is broken.</P
><P
>---</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22038"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Fritz Mehner has written a <A
HREF="http://vim.sourceforge.net/scripts/script.php?script_id=365"
TARGET="_top"
>bash-support
plugin</A
> for the <I
CLASS="FIRSTTERM"
>vim</I
> text editor.
He has also also come up with his own <A
HREF="http://lug.fh-swf.de/vim/vim-bash/StyleGuideShell.en.pdf"
TARGET="_top"
>stylesheet
for Bash</A
>. Compare it with the <A
HREF="#UNOFFICIALST"
>ABS Guide
Unofficial Stylesheet</A
>.</P
><P
>---</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22046"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><EM
>Penguin Pete</EM
> has quite a number of
shell scripting tips and hints on <A
HREF="http://www.penguinpetes.com"
TARGET="_top"
>his superb
site</A
>. Highly recommended.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22051"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>The excellent <I
CLASS="CITETITLE"
> Bash Reference
Manual</I
>, by Chet Ramey and Brian Fox, distributed as
part of the <I
CLASS="FIRSTTERM"
>bash-2-doc</I
> package (available
as an <A
HREF="#RPMREF"
>rpm</A
>). See especially the
instructive example scripts in this package.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22057"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>John Lion's classic, <A
HREF="http://www.lemis.com/grog/Documentation/Lions/index.html"
TARGET="_top"
> <EM
>A Commentary on the Sixth Edition UNIX Operating
System</EM
></A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22062"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>The <A
HREF="news:comp.unix.shell"
TARGET="_top"
>comp.os.unix.shell</A
>
newsgroup.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22066"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
><A
NAME="DDLINK"
></A
></P
><P
>The <A
HREF="http://www.linuxquestions.org/questions/showthread.php?t=362506"
TARGET="_top"
><I
CLASS="FIRSTTERM"
>dd</I
>
thread</A
> on <A
HREF="http://www.linuxquestions.org"
TARGET="_top"
>Linux Questions</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22074"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>The
<A
HREF="http://www.newsville.com/cgi-bin/getfaq?file=comp.unix.shell/comp.unix.shell_FAQ_-_Answers_to_Frequently_Asked_Questions"
TARGET="_top"
>comp.os.unix.shell
FAQ</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22078"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>Assorted comp.os.unix <A
HREF="http://www.faqs.org/faqs/by-newsgroup/comp/comp.unix.shell.html"
TARGET="_top"
> FAQs</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22082"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>The <A
HREF="http://en.wikipedia.org/wiki/Dc_(Unix)"
TARGET="_top"
><I
CLASS="FIRSTTERM"
>Wikipedia</I
>
article</A
> covering <A
HREF="#DCREF"
>dc</A
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="BIBLIOENTRY"
><A
NAME="AEN22088"
></A
><P
></P
><DIV
CLASS="BIBLIOENTRYBLOCK"
STYLE="margin-left: 0.5in"
><DIV
CLASS="ABSTRACT"
><P
>The <A
HREF="#MANREF"
>manpages</A
> for
<B
CLASS="COMMAND"
>bash</B
> and <B
CLASS="COMMAND"
>bash2</B
>,
<B
CLASS="COMMAND"
>date</B
>, <B
CLASS="COMMAND"
>expect</B
>,
<B
CLASS="COMMAND"
>expr</B
>, <B
CLASS="COMMAND"
>find</B
>,
<B
CLASS="COMMAND"
>grep</B
>, <B
CLASS="COMMAND"
>gzip</B
>,
<B
CLASS="COMMAND"
>ln</B
>, <B
CLASS="COMMAND"
>patch</B
>,
<B
CLASS="COMMAND"
>tar</B
>, <B
CLASS="COMMAND"
>tr</B
>,
<B
CLASS="COMMAND"
>bc</B
>, <B
CLASS="COMMAND"
>xargs</B
>.
The <I
CLASS="FIRSTTERM"
>texinfo</I
> documentation on
<B
CLASS="COMMAND"
>bash</B
>, <B
CLASS="COMMAND"
>dd</B
>,
<B
CLASS="COMMAND"
>m4</B
>, <B
CLASS="COMMAND"
>gawk</B
>, and
<B
CLASS="COMMAND"
>sed</B
>.</P
></DIV
></DIV
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="CONTRIBUTED-SCRIPTS"
></A
>Appendix A. Contributed Scripts</H1
><P
>These scripts, while not fitting into the text of this document, do
illustrate some interesting shell programming techniques. Some are useful,
too. Have fun analyzing and running them.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="MAILFORMAT"
></A
><P
><B
>Example A-1. <I
CLASS="FIRSTTERM"
>mailformat</I
>: Formatting an e-mail
message</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# mail-format.sh (ver. 1.1): Format e-mail messages.
# Gets rid of carets, tabs, and also folds excessively long lines.
# =================================================================
# Standard Check for Script Argument(s)
ARGS=1
E_BADARGS=85
E_NOFILE=86
if [ $# -ne $ARGS ] # Correct number of arguments passed to script?
then
echo "Usage: `basename $0` filename"
exit $E_BADARGS
fi
if [ -f "$1" ] # Check if file exists.
then
file_name=$1
else
echo "File \"$1\" does not exist."
exit $E_NOFILE
fi
# -----------------------------------------------------------------
MAXWIDTH=70 # Width to fold excessively long lines to.
# =================================
# A variable can hold a sed script.
# It's a useful technique.
sedscript='s/^&#62;//
s/^ *&#62;//
s/^ *//
s/ *//'
# =================================
# Delete carets and tabs at beginning of lines,
#+ then fold lines to $MAXWIDTH characters.
sed "$sedscript" $1 | fold -s --width=$MAXWIDTH
# -s option to "fold"
#+ breaks lines at whitespace, if possible.
# This script was inspired by an article in a well-known trade journal
#+ extolling a 164K MS Windows utility with similar functionality.
#
# An nice set of text processing utilities and an efficient
#+ scripting language provide an alternative to the bloated executables
#+ of a clunky operating system.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="RN"
></A
><P
><B
>Example A-2. <I
CLASS="FIRSTTERM"
>rn</I
>: A simple-minded file renaming
utility</B
></P
><P
>This script is a modification of <A
HREF="#LOWERCASE"
>Example 16-22</A
>.</P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
# rn.sh
# Very simpleminded filename "rename" utility (based on "lowercase.sh").
#
# The "ren" utility, by Vladimir Lanin (lanin@csd2.nyu.edu),
#+ does a much better job of this.
ARGS=2
E_BADARGS=85
ONE=1 # For getting singular/plural right (see below).
if [ $# -ne "$ARGS" ]
then
echo "Usage: `basename $0` old-pattern new-pattern"
# As in "rn gif jpg", which renames all gif files in working directory to jpg.
exit $E_BADARGS
fi
number=0 # Keeps track of how many files actually renamed.
for filename in *$1* #Traverse all matching files in directory.
do
if [ -f "$filename" ] # If finds match...
then
fname=`basename $filename` # Strip off path.
n=`echo $fname | sed -e "s/$1/$2/"` # Substitute new for old in filename.
mv $fname $n # Rename.
let "number += 1"
fi
done
if [ "$number" -eq "$ONE" ] # For correct grammar.
then
echo "$number file renamed."
else
echo "$number files renamed."
fi
exit $?
# Exercises:
# ---------
# What types of files will this not work on?
# How can this be fixed?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="BLANKRENAME"
></A
><P
><B
>Example A-3. <I
CLASS="FIRSTTERM"
>blank-rename</I
>: Renames filenames containing
blanks</B
></P
><P
>This is an even simpler-minded version of previous script.</P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
# blank-rename.sh
#
# Substitutes underscores for blanks in all the filenames in a directory.
ONE=1 # For getting singular/plural right (see below).
number=0 # Keeps track of how many files actually renamed.
FOUND=0 # Successful return value.
for filename in * #Traverse all files in directory.
do
echo "$filename" | grep -q " " # Check whether filename
if [ $? -eq $FOUND ] #+ contains space(s).
then
fname=$filename # Yes, this filename needs work.
n=`echo $fname | sed -e "s/ /_/g"` # Substitute underscore for blank.
mv "$fname" "$n" # Do the actual renaming.
let "number += 1"
fi
done
if [ "$number" -eq "$ONE" ] # For correct grammar.
then
echo "$number file renamed."
else
echo "$number files renamed."
fi
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ENCRYPTEDPW"
></A
><P
><B
>Example A-4. <I
CLASS="FIRSTTERM"
>encryptedpw</I
>: Uploading to an ftp site,
using a locally encrypted password</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Example "ex72.sh" modified to use encrypted password.
# Note that this is still rather insecure,
#+ since the decrypted password is sent in the clear.
# Use something like "ssh" if this is a concern.
E_BADARGS=85
if [ -z "$1" ]
then
echo "Usage: `basename $0` filename"
exit $E_BADARGS
fi
Username=bozo # Change to suit.
pword=/home/bozo/secret/password_encrypted.file
# File containing encrypted password.
Filename=`basename $1` # Strips pathname out of file name.
Server="XXX"
Directory="YYY" # Change above to actual server name &#38; directory.
Password=`cruft &#60;$pword` # Decrypt password.
# Uses the author's own "cruft" file encryption package,
#+ based on the classic "onetime pad" algorithm,
#+ and obtainable from:
#+ Primary-site: ftp://ibiblio.org/pub/Linux/utils/file
#+ cruft-0.2.tar.gz [16k]
ftp -n $Server &#60;&#60;End-Of-Session
user $Username $Password
binary
bell
cd $Directory
put $Filename
bye
End-Of-Session
# -n option to "ftp" disables auto-logon.
# Note that "bell" rings 'bell' after each file transfer.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="COPYCD"
></A
><P
><B
>Example A-5. <I
CLASS="FIRSTTERM"
>copy-cd</I
>: Copying a data CD</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# copy-cd.sh: copying a data CD
CDROM=/dev/cdrom # CD ROM device
OF=/home/bozo/projects/cdimage.iso # output file
# /xxxx/xxxxxxxx/ Change to suit your system.
BLOCKSIZE=2048
# SPEED=10 # If unspecified, uses max spd.
# DEVICE=/dev/cdrom older version.
DEVICE="1,0,0"
echo; echo "Insert source CD, but do *not* mount it."
echo "Press ENTER when ready. "
read ready # Wait for input, $ready not used.
echo; echo "Copying the source CD to $OF."
echo "This may take a while. Please be patient."
dd if=$CDROM of=$OF bs=$BLOCKSIZE # Raw device copy.
echo; echo "Remove data CD."
echo "Insert blank CDR."
echo "Press ENTER when ready. "
read ready # Wait for input, $ready not used.
echo "Copying $OF to CDR."
# cdrecord -v -isosize speed=$SPEED dev=$DEVICE $OF # Old version.
wodim -v -isosize dev=$DEVICE $OF
# Uses Joerg Schilling's "cdrecord" package (see its docs).
# http://www.fokus.gmd.de/nthp/employees/schilling/cdrecord.html
# Newer Linux distros may use "wodim" rather than "cdrecord" ...
echo; echo "Done copying $OF to CDR on device $CDROM."
echo "Do you want to erase the image file (y/n)? " # Probably a huge file.
read answer
case "$answer" in
[yY]) rm -f $OF
echo "$OF erased."
;;
*) echo "$OF not erased.";;
esac
echo
# Exercise:
# Change the above "case" statement to also accept "yes" and "Yes" as input.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="COLLATZ"
></A
><P
><B
>Example A-6. Collatz series</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# collatz.sh
# The notorious "hailstone" or Collatz series.
# -------------------------------------------
# 1) Get the integer "seed" from the command-line.
# 2) NUMBER &#60;-- seed
# 3) Print NUMBER.
# 4) If NUMBER is even, divide by 2, or
# 5)+ if odd, multiply by 3 and add 1.
# 6) NUMBER &#60;-- result
# 7) Loop back to step 3 (for specified number of iterations).
#
# The theory is that every such sequence,
#+ no matter how large the initial value,
#+ eventually settles down to repeating "4,2,1..." cycles,
#+ even after fluctuating through a wide range of values.
#
# This is an instance of an "iterate,"
#+ an operation that feeds its output back into its input.
# Sometimes the result is a "chaotic" series.
MAX_ITERATIONS=200
# For large seed numbers (&#62;32000), try increasing MAX_ITERATIONS.
h=${1:-$$} # Seed.
# Use $PID as seed,
#+ if not specified as command-line arg.
echo
echo "C($h) -*- $MAX_ITERATIONS Iterations"
echo
for ((i=1; i&#60;=MAX_ITERATIONS; i++))
do
# echo -n "$h "
# ^^^
# tab
# printf does it better ...
COLWIDTH=%7d
printf $COLWIDTH $h
let "remainder = h % 2"
if [ "$remainder" -eq 0 ] # Even?
then
let "h /= 2" # Divide by 2.
else
let "h = h*3 + 1" # Multiply by 3 and add 1.
fi
COLUMNS=10 # Output 10 values per line.
let "line_break = i % $COLUMNS"
if [ "$line_break" -eq 0 ]
then
echo
fi
done
echo
# For more information on this strange mathematical function,
#+ see _Computers, Pattern, Chaos, and Beauty_, by Pickover, p. 185 ff.,
#+ as listed in the bibliography.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="DAYSBETWEEN0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="DAYSBETWEEN"
></A
><P
><B
>Example A-7. <I
CLASS="FIRSTTERM"
>days-between</I
>: Days between two
dates</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# days-between.sh: Number of days between two dates.
# Usage: ./days-between.sh [M]M/[D]D/YYYY [M]M/[D]D/YYYY
#
# Note: Script modified to account for changes in Bash, v. 2.05b +,
#+ that closed the loophole permitting large negative
#+ integer return values.
ARGS=2 # Two command-line parameters expected.
E_PARAM_ERR=85 # Param error.
REFYR=1600 # Reference year.
CENTURY=100
DIY=365
ADJ_DIY=367 # Adjusted for leap year + fraction.
MIY=12
DIM=31
LEAPCYCLE=4
MAXRETVAL=255 # Largest permissible
#+ positive return value from a function.
diff= # Declare global variable for date difference.
value= # Declare global variable for absolute value.
day= # Declare globals for day, month, year.
month=
year=
Param_Error () # Command-line parameters wrong.
{
echo "Usage: `basename $0` [M]M/[D]D/YYYY [M]M/[D]D/YYYY"
echo " (date must be after 1/3/1600)"
exit $E_PARAM_ERR
}
Parse_Date () # Parse date from command-line params.
{
month=${1%%/**}
dm=${1%/**} # Day and month.
day=${dm#*/}
let "year = `basename $1`" # Not a filename, but works just the same.
}
check_date () # Checks for invalid date(s) passed.
{
[ "$day" -gt "$DIM" ] || [ "$month" -gt "$MIY" ] ||
[ "$year" -lt "$REFYR" ] &#38;&#38; Param_Error
# Exit script on bad value(s).
# Uses or-list / and-list.
#
# Exercise: Implement more rigorous date checking.
}
strip_leading_zero () # Better to strip possible leading zero(s)
{ #+ from day and/or month
return ${1#0} #+ since otherwise Bash will interpret them
} #+ as octal values (POSIX.2, sect 2.9.2.1).
day_index () # Gauss' Formula:
{ # Days from March 1, 1600 to date passed as param.
# ^^^^^^^^^^^^^
day=$1
month=$2
year=$3
let "month = $month - 2"
if [ "$month" -le 0 ]
then
let "month += 12"
let "year -= 1"
fi
let "year -= $REFYR"
let "indexyr = $year / $CENTURY"
let "Days = $DIY*$year + $year/$LEAPCYCLE - $indexyr \
+ $indexyr/$LEAPCYCLE + $ADJ_DIY*$month/$MIY + $day - $DIM"
# For an in-depth explanation of this algorithm, see
#+ http://weblogs.asp.net/pgreborio/archive/2005/01/06/347968.aspx
echo $Days
}
calculate_difference () # Difference between two day indices.
{
let "diff = $1 - $2" # Global variable.
}
abs () # Absolute value
{ # Uses global "value" variable.
if [ "$1" -lt 0 ] # If negative
then #+ then
let "value = 0 - $1" #+ change sign,
else #+ else
let "value = $1" #+ leave it alone.
fi
}
if [ $# -ne "$ARGS" ] # Require two command-line params.
then
Param_Error
fi
Parse_Date $1
check_date $day $month $year # See if valid date.
strip_leading_zero $day # Remove any leading zeroes
day=$? #+ on day and/or month.
strip_leading_zero $month
month=$?
let "date1 = `day_index $day $month $year`"
Parse_Date $2
check_date $day $month $year
strip_leading_zero $day
day=$?
strip_leading_zero $month
month=$?
date2=$(day_index $day $month $year) # Command substitution.
calculate_difference $date1 $date2
abs $diff # Make sure it's positive.
diff=$value
echo $diff
exit 0
# Exercise:
# --------
# If given only one command-line parameter, have the script
#+ use today's date as the second.
# Compare this script with
#+ the implementation of Gauss' Formula in a C program at
#+ http://buschencrew.hypermart.net/software/datedif</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="MAKEDICT"
></A
><P
><B
>Example A-8. Making a <I
CLASS="FIRSTTERM"
>dictionary</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# makedict.sh [make dictionary]
# Modification of /usr/sbin/mkdict (/usr/sbin/cracklib-forman) script.
# Original script copyright 1993, by Alec Muffett.
#
# This modified script included in this document in a manner
#+ consistent with the "LICENSE" document of the "Crack" package
#+ that the original script is a part of.
# This script processes text files to produce a sorted list
#+ of words found in the files.
# This may be useful for compiling dictionaries
#+ and for other lexicographic purposes.
E_BADARGS=85
if [ ! -r "$1" ] # Need at least one
then #+ valid file argument.
echo "Usage: $0 files-to-process"
exit $E_BADARGS
fi
# SORT="sort" # No longer necessary to define
#+ options to sort. Changed from
#+ original script.
cat $* | # Dump specified files to stdout.
tr A-Z a-z | # Convert to lowercase.
tr ' ' '\012' | # New: change spaces to newlines.
# tr -cd '\012[a-z][0-9]' | # Get rid of everything
#+ non-alphanumeric (in orig. script).
tr -c '\012a-z' '\012' | # Rather than deleting non-alpha
#+ chars, change them to newlines.
sort | # $SORT options unnecessary now.
uniq | # Remove duplicates.
grep -v '^#' | # Delete lines starting with #.
grep -v '^$' # Delete blank lines.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="SOUNDEX0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="SOUNDEX"
></A
><P
><B
>Example A-9. Soundex conversion</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# soundex.sh: Calculate "soundex" code for names
# =======================================================
# Soundex script
# by
# Mendel Cooper
# thegrendel.abs@gmail.com
# reldate: 23 January, 2002
#
# Placed in the Public Domain.
#
# A slightly different version of this script appeared in
#+ Ed Schaefer's July, 2002 "Shell Corner" column
#+ in "Unix Review" on-line,
#+ http://www.unixreview.com/documents/uni1026336632258/
# =======================================================
ARGCOUNT=1 # Need name as argument.
E_WRONGARGS=90
if [ $# -ne "$ARGCOUNT" ]
then
echo "Usage: `basename $0` name"
exit $E_WRONGARGS
fi
assign_value () # Assigns numerical value
{ #+ to letters of name.
val1=bfpv # 'b,f,p,v' = 1
val2=cgjkqsxz # 'c,g,j,k,q,s,x,z' = 2
val3=dt # etc.
val4=l
val5=mn
val6=r
# Exceptionally clever use of 'tr' follows.
# Try to figure out what is going on here.
value=$( echo "$1" \
| tr -d wh \
| tr $val1 1 | tr $val2 2 | tr $val3 3 \
| tr $val4 4 | tr $val5 5 | tr $val6 6 \
| tr -s 123456 \
| tr -d aeiouy )
# Assign letter values.
# Remove duplicate numbers, except when separated by vowels.
# Ignore vowels, except as separators, so delete them last.
# Ignore 'w' and 'h', even as separators, so delete them first.
#
# The above command substitution lays more pipe than a plumber &#60;g&#62;.
}
input_name="$1"
echo
echo "Name = $input_name"
# Change all characters of name input to lowercase.
# ------------------------------------------------
name=$( echo $input_name | tr A-Z a-z )
# ------------------------------------------------
# Just in case argument to script is mixed case.
# Prefix of soundex code: first letter of name.
# --------------------------------------------
char_pos=0 # Initialize character position.
prefix0=${name:$char_pos:1}
prefix=`echo $prefix0 | tr a-z A-Z`
# Uppercase 1st letter of soundex.
let "char_pos += 1" # Bump character position to 2nd letter of name.
name1=${name:$char_pos}
# ++++++++++++++++++++++++++ Exception Patch ++++++++++++++++++++++++++++++
# Now, we run both the input name and the name shifted one char
#+ to the right through the value-assigning function.
# If we get the same value out, that means that the first two characters
#+ of the name have the same value assigned, and that one should cancel.
# However, we also need to test whether the first letter of the name
#+ is a vowel or 'w' or 'h', because otherwise this would bollix things up.
char1=`echo $prefix | tr A-Z a-z` # First letter of name, lowercased.
assign_value $name
s1=$value
assign_value $name1
s2=$value
assign_value $char1
s3=$value
s3=9$s3 # If first letter of name is a vowel
#+ or 'w' or 'h',
#+ then its "value" will be null (unset).
#+ Therefore, set it to 9, an otherwise
#+ unused value, which can be tested for.
if [[ "$s1" -ne "$s2" || "$s3" -eq 9 ]]
then
suffix=$s2
else
suffix=${s2:$char_pos}
fi
# ++++++++++++++++++++++ end Exception Patch ++++++++++++++++++++++++++++++
padding=000 # Use at most 3 zeroes to pad.
soun=$prefix$suffix$padding # Pad with zeroes.
MAXLEN=4 # Truncate to maximum of 4 chars.
soundex=${soun:0:$MAXLEN}
echo "Soundex = $soundex"
echo
# The soundex code is a method of indexing and classifying names
#+ by grouping together the ones that sound alike.
# The soundex code for a given name is the first letter of the name,
#+ followed by a calculated three-number code.
# Similar sounding names should have almost the same soundex codes.
# Examples:
# Smith and Smythe both have a "S-530" soundex.
# Harrison = H-625
# Hargison = H-622
# Harriman = H-655
# This works out fairly well in practice, but there are numerous anomalies.
#
#
# The U.S. Census and certain other governmental agencies use soundex,
# as do genealogical researchers.
#
# For more information,
#+ see the "National Archives and Records Administration home page",
#+ http://www.nara.gov/genealogy/soundex/soundex.html
# Exercise:
# --------
# Simplify the "Exception Patch" section of this script.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="LIFEREF"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="LIFESLOW"
></A
><P
><B
>Example A-10. <I
CLASS="FIRSTTERM"
>Game of Life</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# life.sh: "Life in the Slow Lane"
# Author: Mendel Cooper
# License: GPL3
# Version 0.2: Patched by Daniel Albers
#+ to allow non-square grids as input.
# Version 0.2.1: Added 2-second delay between generations.
# ##################################################################### #
# This is the Bash script version of John Conway's "Game of Life". #
# "Life" is a simple implementation of cellular automata. #
# --------------------------------------------------------------------- #
# On a rectangular grid, let each "cell" be either "living" or "dead." #
# Designate a living cell with a dot, and a dead one with a blank space.#
# Begin with an arbitrarily drawn dot-and-blank grid, #
#+ and let this be the starting generation: generation 0. #
# Determine each successive generation by the following rules: #
# 1) Each cell has 8 neighbors, the adjoining cells #
#+ left, right, top, bottom, and the 4 diagonals. #
# #
# 123 #
# 4*5 The * is the cell under consideration. #
# 678 #
# #
# 2) A living cell with either 2 or 3 living neighbors remains alive. #
SURVIVE=2 #
# 3) A dead cell with 3 living neighbors comes alive, a "birth." #
BIRTH=3 #
# 4) All other cases result in a dead cell for the next generation. #
# ##################################################################### #
startfile=gen0 # Read the starting generation from the file "gen0" ...
# Default, if no other file specified when invoking script.
#
if [ -n "$1" ] # Specify another "generation 0" file.
then
startfile="$1"
fi
############################################
# Abort script if "startfile" not specified
#+ and
#+ default file "gen0" not present.
E_NOSTARTFILE=86
if [ ! -e "$startfile" ]
then
echo "Startfile \""$startfile"\" missing!"
exit $E_NOSTARTFILE
fi
############################################
ALIVE1=.
DEAD1=_
# Represent living and dead cells in the start-up file.
# -----------------------------------------------------#
# This script uses a 10 x 10 grid (may be increased,
#+ but a large grid will slow down execution).
ROWS=10
COLS=10
# Change above two variables to match desired grid size.
# -----------------------------------------------------#
GENERATIONS=10 # How many generations to cycle through.
# Adjust this upwards
#+ if you have time on your hands.
NONE_ALIVE=85 # Exit status on premature bailout,
#+ if no cells left alive.
DELAY=2 # Pause between generations.
TRUE=0
FALSE=1
ALIVE=0
DEAD=1
avar= # Global; holds current generation.
generation=0 # Initialize generation count.
# =================================================================
let "cells = $ROWS * $COLS" # How many cells.
# Arrays containing "cells."
declare -a initial
declare -a current
display ()
{
alive=0 # How many cells alive at any given time.
# Initially zero.
declare -a arr
arr=( `echo "$1"` ) # Convert passed arg to array.
element_count=${#arr[*]}
local i
local rowcheck
for ((i=0; i&#60;$element_count; i++))
do
# Insert newline at end of each row.
let "rowcheck = $i % COLS"
if [ "$rowcheck" -eq 0 ]
then
echo # Newline.
echo -n " " # Indent.
fi
cell=${arr[i]}
if [ "$cell" = . ]
then
let "alive += 1"
fi
echo -n "$cell" | sed -e 's/_/ /g'
# Print out array, changing underscores to spaces.
done
return
}
IsValid () # Test if cell coordinate valid.
{
if [ -z "$1" -o -z "$2" ] # Mandatory arguments missing?
then
return $FALSE
fi
local row
local lower_limit=0 # Disallow negative coordinate.
local upper_limit
local left
local right
let "upper_limit = $ROWS * $COLS - 1" # Total number of cells.
if [ "$1" -lt "$lower_limit" -o "$1" -gt "$upper_limit" ]
then
return $FALSE # Out of array bounds.
fi
row=$2
let "left = $row * $COLS" # Left limit.
let "right = $left + $COLS - 1" # Right limit.
if [ "$1" -lt "$left" -o "$1" -gt "$right" ]
then
return $FALSE # Beyond row boundary.
fi
return $TRUE # Valid coordinate.
}
IsAlive () # Test whether cell is alive.
# Takes array, cell number, and
{ #+ state of cell as arguments.
GetCount "$1" $2 # Get alive cell count in neighborhood.
local nhbd=$?
if [ "$nhbd" -eq "$BIRTH" ] # Alive in any case.
then
return $ALIVE
fi
if [ "$3" = "." -a "$nhbd" -eq "$SURVIVE" ]
then # Alive only if previously alive.
return $ALIVE
fi
return $DEAD # Defaults to dead.
}
GetCount () # Count live cells in passed cell's neighborhood.
# Two arguments needed:
# $1) variable holding array
# $2) cell number
{
local cell_number=$2
local array
local top
local center
local bottom
local r
local row
local i
local t_top
local t_cen
local t_bot
local count=0
local ROW_NHBD=3
array=( `echo "$1"` )
let "top = $cell_number - $COLS - 1" # Set up cell neighborhood.
let "center = $cell_number - 1"
let "bottom = $cell_number + $COLS - 1"
let "r = $cell_number / $COLS"
for ((i=0; i&#60;$ROW_NHBD; i++)) # Traverse from left to right.
do
let "t_top = $top + $i"
let "t_cen = $center + $i"
let "t_bot = $bottom + $i"
let "row = $r" # Count center row.
IsValid $t_cen $row # Valid cell position?
if [ $? -eq "$TRUE" ]
then
if [ ${array[$t_cen]} = "$ALIVE1" ] # Is it alive?
then # If yes, then ...
let "count += 1" # Increment count.
fi
fi
let "row = $r - 1" # Count top row.
IsValid $t_top $row
if [ $? -eq "$TRUE" ]
then
if [ ${array[$t_top]} = "$ALIVE1" ] # Redundancy here.
then # Can it be optimized?
let "count += 1"
fi
fi
let "row = $r + 1" # Count bottom row.
IsValid $t_bot $row
if [ $? -eq "$TRUE" ]
then
if [ ${array[$t_bot]} = "$ALIVE1" ]
then
let "count += 1"
fi
fi
done
if [ ${array[$cell_number]} = "$ALIVE1" ]
then
let "count -= 1" # Make sure value of tested cell itself
fi #+ is not counted.
return $count
}
next_gen () # Update generation array.
{
local array
local i=0
array=( `echo "$1"` ) # Convert passed arg to array.
while [ "$i" -lt "$cells" ]
do
IsAlive "$1" $i ${array[$i]} # Is the cell alive?
if [ $? -eq "$ALIVE" ]
then # If alive, then
array[$i]=. #+ represent the cell as a period.
else
array[$i]="_" # Otherwise underscore
fi #+ (will later be converted to space).
let "i += 1"
done
# let "generation += 1" # Increment generation count.
### Why was the above line commented out?
# Set variable to pass as parameter to "display" function.
avar=`echo ${array[@]}` # Convert array back to string variable.
display "$avar" # Display it.
echo; echo
echo "Generation $generation - $alive alive"
if [ "$alive" -eq 0 ]
then
echo
echo "Premature exit: no more cells alive!"
exit $NONE_ALIVE # No point in continuing
fi #+ if no live cells.
}
# =========================================================
# main ()
# {
# Load initial array with contents of startup file.
initial=( `cat "$startfile" | sed -e '/#/d' | tr -d '\n' |\
# Delete lines containing '#' comment character.
sed -e 's/\./\. /g' -e 's/_/_ /g'` )
# Remove linefeeds and insert space between elements.
clear # Clear screen.
echo # Title
setterm -reverse on
echo "======================="
setterm -reverse off
echo " $GENERATIONS generations"
echo " of"
echo "\"Life in the Slow Lane\""
setterm -reverse on
echo "======================="
setterm -reverse off
sleep $DELAY # Display "splash screen" for 2 seconds.
# -------- Display first generation. --------
Gen0=`echo ${initial[@]}`
display "$Gen0" # Display only.
echo; echo
echo "Generation $generation - $alive alive"
sleep $DELAY
# -------------------------------------------
let "generation += 1" # Bump generation count.
echo
# ------- Display second generation. -------
Cur=`echo ${initial[@]}`
next_gen "$Cur" # Update &#38; display.
sleep $DELAY
# ------------------------------------------
let "generation += 1" # Increment generation count.
# ------ Main loop for displaying subsequent generations ------
while [ "$generation" -le "$GENERATIONS" ]
do
Cur="$avar"
next_gen "$Cur"
let "generation += 1"
sleep $DELAY
done
# ==============================================================
echo
# }
exit 0 # CEOF:EOF
# The grid in this script has a "boundary problem."
# The the top, bottom, and sides border on a void of dead cells.
# Exercise: Change the script to have the grid wrap around,
# + so that the left and right sides will "touch,"
# + as will the top and bottom.
#
# Exercise: Create a new "gen0" file to seed this script.
# Use a 12 x 16 grid, instead of the original 10 x 10 one.
# Make the necessary changes to the script,
#+ so it will run with the altered file.
#
# Exercise: Modify this script so that it can determine the grid size
#+ from the "gen0" file, and set any variables necessary
#+ for the script to run.
# This would make unnecessary any changes to variables
#+ in the script for an altered grid size.
#
# Exercise: Optimize this script.
# It has redundant code.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="GEN0DATA"
></A
><P
><B
>Example A-11. Data file for <I
CLASS="FIRSTTERM"
>Game of Life</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># gen0
#
# This is an example "generation 0" start-up file for "life.sh".
# --------------------------------------------------------------
# The "gen0" file is a 10 x 10 grid using a period (.) for live cells,
#+ and an underscore (_) for dead ones. We cannot simply use spaces
#+ for dead cells in this file because of a peculiarity in Bash arrays.
# [Exercise for the reader: explain this.]
#
# Lines beginning with a '#' are comments, and the script ignores them.
__.__..___
__.._.____
____.___..
_._______.
____._____
..__...___
____._____
___...____
__.._..___
_..___..__</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>+++</P
><P
>The following script is by Mark Moraes of the University
of Toronto. See the file <TT
CLASS="FILENAME"
>Moraes-COPYRIGHT</TT
>
for permissions and restrictions. This file is included in the
combined <A
HREF="#WHERE_TARBALL"
>HTML/source tarball</A
>
of the <EM
>ABS Guide</EM
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="BEHEAD"
></A
><P
><B
>Example A-12. <I
CLASS="FIRSTTERM"
>behead</I
>: Removing mail and news
message headers</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/sh
# Strips off the header from a mail/News message i.e. till the first
#+ empty line.
# Author: Mark Moraes, University of Toronto
# ==&#62; These comments added by author of this document.
if [ $# -eq 0 ]; then
# ==&#62; If no command-line args present, then works on file redirected to stdin.
sed -e '1,/^$/d' -e '/^[ ]*$/d'
# --&#62; Delete empty lines and all lines until
# --&#62; first one beginning with white space.
else
# ==&#62; If command-line args present, then work on files named.
for i do
sed -e '1,/^$/d' -e '/^[ ]*$/d' $i
# --&#62; Ditto, as above.
done
fi
exit
# ==&#62; Exercise: Add error checking and other options.
# ==&#62;
# ==&#62; Note that the small sed script repeats, except for the arg passed.
# ==&#62; Does it make sense to embed it in a function? Why or why not?
/*
* Copyright University of Toronto 1988, 1989.
* Written by Mark Moraes
*
* Permission is granted to anyone to use this software for any purpose on
* any computer system, and to alter it and redistribute it freely, subject
* to the following restrictions:
*
* 1. The author and the University of Toronto are not responsible
* for the consequences of use of this software, no matter how awful,
* even if they arise from flaws in it.
*
* 2. The origin of this software must not be misrepresented, either by
* explicit claim or by omission. Since few users ever read sources,
* credits must appear in the documentation.
*
* 3. Altered versions must be plainly marked as such, and must not be
* misrepresented as being the original software. Since few users
* ever read sources, credits must appear in the documentation.
*
* 4. This notice may not be removed or altered.
*/</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>+</P
><P
>Antek Sawicki contributed the following script, which makes very
clever use of the parameter substitution operators discussed in
<A
HREF="#PARAMETER-SUBSTITUTION"
>Section 10.2</A
>.</P
><P
><A
NAME="PW0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="PW"
></A
><P
><B
>Example A-13. <I
CLASS="FIRSTTERM"
>password</I
>: Generating random
8-character passwords</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
#
#
# Random password generator for Bash 2.x +
#+ by Antek Sawicki &#60;tenox@tenox.tc&#62;,
#+ who generously gave usage permission to the ABS Guide author.
#
# ==&#62; Comments added by document author ==&#62;
MATRIX="0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"
# ==&#62; Password will consist of alphanumeric characters.
LENGTH="8"
# ==&#62; May change 'LENGTH' for longer password.
while [ "${n:=1}" -le "$LENGTH" ]
# ==&#62; Recall that := is "default substitution" operator.
# ==&#62; So, if 'n' has not been initialized, set it to 1.
do
PASS="$PASS${MATRIX:$(($RANDOM%${#MATRIX})):1}"
# ==&#62; Very clever, but tricky.
# ==&#62; Starting from the innermost nesting...
# ==&#62; ${#MATRIX} returns length of array MATRIX.
# ==&#62; $RANDOM%${#MATRIX} returns random number between 1
# ==&#62; and [length of MATRIX] - 1.
# ==&#62; ${MATRIX:$(($RANDOM%${#MATRIX})):1}
# ==&#62; returns expansion of MATRIX at random position, by length 1.
# ==&#62; See {var:pos:len} parameter substitution in Chapter 9.
# ==&#62; and the associated examples.
# ==&#62; PASS=... simply pastes this result onto previous PASS (concatenation).
# ==&#62; To visualize this more clearly, uncomment the following line
# echo "$PASS"
# ==&#62; to see PASS being built up,
# ==&#62; one character at a time, each iteration of the loop.
let n+=1
# ==&#62; Increment 'n' for next pass.
done
echo "$PASS" # ==&#62; Or, redirect to a file, as desired.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>+</P
><P
><A
NAME="ZFIFO"
></A
>James R. Van Zandt contributed this script
which uses named pipes and, in his words, <SPAN
CLASS="QUOTE"
>"really exercises
quoting and escaping."</SPAN
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="FIFO"
></A
><P
><B
>Example A-14. <I
CLASS="FIRSTTERM"
>fifo</I
>: Making daily backups, using
named pipes</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ==&#62; Script by James R. Van Zandt, and used here with his permission.
# ==&#62; Comments added by author of this document.
HERE=`uname -n` # ==&#62; hostname
THERE=bilbo
echo "starting remote backup to $THERE at `date +%r`"
# ==&#62; `date +%r` returns time in 12-hour format, i.e. "08:08:34 PM".
# make sure /pipe really is a pipe and not a plain file
rm -rf /pipe
mkfifo /pipe # ==&#62; Create a "named pipe", named "/pipe" ...
# ==&#62; 'su xyz' runs commands as user "xyz".
# ==&#62; 'ssh' invokes secure shell (remote login client).
su xyz -c "ssh $THERE \"cat &#62; /home/xyz/backup/${HERE}-daily.tar.gz\" &#60; /pipe"&#38;
cd /
tar -czf - bin boot dev etc home info lib man root sbin share usr var &#62; /pipe
# ==&#62; Uses named pipe, /pipe, to communicate between processes:
# ==&#62; 'tar/gzip' writes to /pipe and 'ssh' reads from /pipe.
# ==&#62; The end result is this backs up the main directories, from / on down.
# ==&#62; What are the advantages of a "named pipe" in this situation,
# ==&#62;+ as opposed to an "anonymous pipe", with |?
# ==&#62; Will an anonymous pipe even work here?
# ==&#62; Is it necessary to delete the pipe before exiting the script?
# ==&#62; How could that be done?
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>+</P
><P
><A
NAME="PRIMES1"
></A
></P
><P
>St<53>phane Chazelas used the following script to
demonstrate generating prime numbers without arrays.</P
><P
><A
NAME="PRIMES00"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="PRIMES"
></A
><P
><B
>Example A-15. Generating prime numbers using the modulo operator</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# primes.sh: Generate prime numbers, without using arrays.
# Script contributed by Stephane Chazelas.
# This does *not* use the classic "Sieve of Eratosthenes" algorithm,
#+ but instead the more intuitive method of testing each candidate number
#+ for factors (divisors), using the "%" modulo operator.
LIMIT=1000 # Primes, 2 ... 1000.
Primes()
{
(( n = $1 + 1 )) # Bump to next integer.
shift # Next parameter in list.
# echo "_n=$n i=$i_"
if (( n == LIMIT ))
then echo $*
return
fi
for i; do # "i" set to "@", previous values of $n.
# echo "-n=$n i=$i-"
(( i * i &#62; n )) &#38;&#38; break # Optimization.
(( n % i )) &#38;&#38; continue # Sift out non-primes using modulo operator.
Primes $n $@ # Recursion inside loop.
return
done
Primes $n $@ $n # Recursion outside loop.
# Successively accumulate
#+ positional parameters.
# "$@" is the accumulating list of primes.
}
Primes 1
exit $?
# Pipe output of the script to 'fmt' for prettier printing.
# Uncomment lines 16 and 24 to help figure out what is going on.
# Compare the speed of this algorithm for generating primes
#+ with the Sieve of Eratosthenes (ex68.sh).
# Exercise: Rewrite this script without recursion.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>+</P
><P
>Rick Boivie's revision of Jordi Sanfeliu's
<EM
>tree</EM
> script.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="TREE"
></A
><P
><B
>Example A-16. <I
CLASS="FIRSTTERM"
>tree</I
>: Displaying a directory tree</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# tree.sh
# Written by Rick Boivie.
# Used with permission.
# This is a revised and simplified version of a script
#+ by Jordi Sanfeliu (the original author), and patched by Ian Kjos.
# This script replaces the earlier version used in
#+ previous releases of the Advanced Bash Scripting Guide.
# Copyright (c) 2002, by Jordi Sanfeliu, Rick Boivie, and Ian Kjos.
# ==&#62; Comments added by the author of this document.
search () {
for dir in `echo *`
# ==&#62; `echo *` lists all the files in current working directory,
#+ ==&#62; without line breaks.
# ==&#62; Similar effect to for dir in *
# ==&#62; but "dir in `echo *`" will not handle filenames with blanks.
do
if [ -d "$dir" ] ; then # ==&#62; If it is a directory (-d)...
zz=0 # ==&#62; Temp variable, keeping track of
# directory level.
while [ $zz != $1 ] # Keep track of inner nested loop.
do
echo -n "| " # ==&#62; Display vertical connector symbol,
# ==&#62; with 2 spaces &#38; no line feed
# in order to indent.
zz=`expr $zz + 1` # ==&#62; Increment zz.
done
if [ -L "$dir" ] ; then # ==&#62; If directory is a symbolic link...
echo "+---$dir" `ls -l $dir | sed 's/^.*'$dir' //'`
# ==&#62; Display horiz. connector and list directory name, but...
# ==&#62; delete date/time part of long listing.
else
echo "+---$dir" # ==&#62; Display horizontal connector symbol...
# ==&#62; and print directory name.
numdirs=`expr $numdirs + 1` # ==&#62; Increment directory count.
if cd "$dir" ; then # ==&#62; If can move to subdirectory...
search `expr $1 + 1` # with recursion ;-)
# ==&#62; Function calls itself.
cd ..
fi
fi
fi
done
}
if [ $# != 0 ] ; then
cd $1 # Move to indicated directory.
#else # stay in current directory
fi
echo "Initial directory = `pwd`"
numdirs=0
search 0
echo "Total directories = $numdirs"
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Patsie's version of a directory <I
CLASS="FIRSTTERM"
>tree</I
>
script.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="TREE2"
></A
><P
><B
>Example A-17. <I
CLASS="FIRSTTERM"
>tree2</I
>: Alternate directory tree script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# tree2.sh
# Lightly modified/reformatted by ABS Guide author.
# Included in ABS Guide with permission of script author (thanks!).
## Recursive file/dirsize checking script, by Patsie
##
## This script builds a list of files/directories and their size (du -akx)
## and processes this list to a human readable tree shape
## The 'du -akx' is only as good as the permissions the owner has.
## So preferably run as root* to get the best results, or use only on
## directories for which you have read permissions. Anything you can't
## read is not in the list.
#* ABS Guide author advises caution when running scripts as root!
########## THIS IS CONFIGURABLE ##########
TOP=5 # Top 5 biggest (sub)directories.
MAXRECURS=5 # Max 5 subdirectories/recursions deep.
E_BL=80 # Blank line already returned.
E_DIR=81 # Directory not specified.
########## DON'T CHANGE ANYTHING BELOW THIS LINE ##########
PID=$$ # Our own process ID.
SELF=`basename $0` # Our own program name.
TMP="/tmp/${SELF}.${PID}.tmp" # Temporary 'du' result.
# Convert number to dotted thousand.
function dot { echo " $*" |
sed -e :a -e 's/\(.*[0-9]\)\([0-9]\{3\}\)/\1,\2/;ta' |
tail -c 12; }
# Usage: tree &#60;recursion&#62; &#60;indent prefix&#62; &#60;min size&#62; &#60;directory&#62;
function tree {
recurs="$1" # How deep nested are we?
prefix="$2" # What do we display before file/dirname?
minsize="$3" # What is the minumum file/dirsize?
dirname="$4" # Which directory are we checking?
# Get ($TOP) biggest subdirs/subfiles from TMP file.
LIST=`egrep "[[:space:]]${dirname}/[^/]*$" "$TMP" |
awk '{if($1&#62;'$minsize') print;}' | sort -nr | head -$TOP`
[ -z "$LIST" ] &#38;&#38; return # Empty list, then go back.
cnt=0
num=`echo "$LIST" | wc -l` # How many entries in the list.
## Main loop
echo "$LIST" | while read size name; do
((cnt+=1)) # Count entry number.
bname=`basename "$name"` # We only need a basename of the entry.
[ -d "$name" ] &#38;&#38; bname="$bname/"
# If it's a directory, append a slash.
echo "`dot $size`$prefix +-$bname"
# Display the result.
# Call ourself recursively if it's a directory
#+ and we're not nested too deep ($MAXRECURS).
# The recursion goes up: $((recurs+1))
# The prefix gets a space if it's the last entry,
#+ or a pipe if there are more entries.
# The minimum file/dirsize becomes
#+ a tenth of his parent: $((size/10)).
# Last argument is the full directory name to check.
if [ -d "$name" -a $recurs -lt $MAXRECURS ]; then
[ $cnt -lt $num ] \
|| (tree $((recurs+1)) "$prefix " $((size/10)) "$name") \
&#38;&#38; (tree $((recurs+1)) "$prefix |" $((size/10)) "$name")
fi
done
[ $? -eq 0 ] &#38;&#38; echo " $prefix"
# Every time we jump back add a 'blank' line.
return $E_BL
# We return 80 to tell we added a blank line already.
}
### ###
### main program ###
### ###
rootdir="$@"
[ -d "$rootdir" ] ||
{ echo "$SELF: Usage: $SELF &#60;directory&#62;" &#62;&#38;2; exit $E_DIR; }
# We should be called with a directory name.
echo "Building inventory list, please wait ..."
# Show "please wait" message.
du -akx "$rootdir" 1&#62;"$TMP" 2&#62;/dev/null
# Build a temporary list of all files/dirs and their size.
size=`tail -1 "$TMP" | awk '{print $1}'`
# What is our rootdirectory's size?
echo "`dot $size` $rootdir"
# Display rootdirectory's entry.
tree 0 "" 0 "$rootdir"
# Display the tree below our rootdirectory.
rm "$TMP" 2&#62;/dev/null
# Clean up TMP file.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Noah Friedman permitted use of his <EM
>string
function</EM
> script. It essentially reproduces some
of the <I
CLASS="FIRSTTERM"
>C</I
>-library string manipulation
functions.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="STRING"
></A
><P
><B
>Example A-18. <I
CLASS="FIRSTTERM"
>string functions</I
>: C-style string
functions</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# string.bash --- bash emulation of string(3) library routines
# Author: Noah Friedman &#60;friedman@prep.ai.mit.edu&#62;
# ==&#62; Used with his kind permission in this document.
# Created: 1992-07-01
# Last modified: 1993-09-29
# Public domain
# Conversion to bash v2 syntax done by Chet Ramey
# Commentary:
# Code:
#:docstring strcat:
# Usage: strcat s1 s2
#
# Strcat appends the value of variable s2 to variable s1.
#
# Example:
# a="foo"
# b="bar"
# strcat a b
# echo $a
# =&#62; foobar
#
#:end docstring:
###;;;autoload ==&#62; Autoloading of function commented out.
function strcat ()
{
local s1_val s2_val
s1_val=${!1} # indirect variable expansion
s2_val=${!2}
eval "$1"=\'"${s1_val}${s2_val}"\'
# ==&#62; eval $1='${s1_val}${s2_val}' avoids problems,
# ==&#62; if one of the variables contains a single quote.
}
#:docstring strncat:
# Usage: strncat s1 s2 $n
#
# Line strcat, but strncat appends a maximum of n characters from the value
# of variable s2. It copies fewer if the value of variabl s2 is shorter
# than n characters. Echoes result on stdout.
#
# Example:
# a=foo
# b=barbaz
# strncat a b 3
# echo $a
# =&#62; foobar
#
#:end docstring:
###;;;autoload
function strncat ()
{
local s1="$1"
local s2="$2"
local -i n="$3"
local s1_val s2_val
s1_val=${!s1} # ==&#62; indirect variable expansion
s2_val=${!s2}
if [ ${#s2_val} -gt ${n} ]; then
s2_val=${s2_val:0:$n} # ==&#62; substring extraction
fi
eval "$s1"=\'"${s1_val}${s2_val}"\'
# ==&#62; eval $1='${s1_val}${s2_val}' avoids problems,
# ==&#62; if one of the variables contains a single quote.
}
#:docstring strcmp:
# Usage: strcmp $s1 $s2
#
# Strcmp compares its arguments and returns an integer less than, equal to,
# or greater than zero, depending on whether string s1 is lexicographically
# less than, equal to, or greater than string s2.
#:end docstring:
###;;;autoload
function strcmp ()
{
[ "$1" = "$2" ] &#38;&#38; return 0
[ "${1}" '&#60;' "${2}" ] &#62; /dev/null &#38;&#38; return -1
return 1
}
#:docstring strncmp:
# Usage: strncmp $s1 $s2 $n
#
# Like strcmp, but makes the comparison by examining a maximum of n
# characters (n less than or equal to zero yields equality).
#:end docstring:
###;;;autoload
function strncmp ()
{
if [ -z "${3}" -o "${3}" -le "0" ]; then
return 0
fi
if [ ${3} -ge ${#1} -a ${3} -ge ${#2} ]; then
strcmp "$1" "$2"
return $?
else
s1=${1:0:$3}
s2=${2:0:$3}
strcmp $s1 $s2
return $?
fi
}
#:docstring strlen:
# Usage: strlen s
#
# Strlen returns the number of characters in string literal s.
#:end docstring:
###;;;autoload
function strlen ()
{
eval echo "\${#${1}}"
# ==&#62; Returns the length of the value of the variable
# ==&#62; whose name is passed as an argument.
}
#:docstring strspn:
# Usage: strspn $s1 $s2
#
# Strspn returns the length of the maximum initial segment of string s1,
# which consists entirely of characters from string s2.
#:end docstring:
###;;;autoload
function strspn ()
{
# Unsetting IFS allows whitespace to be handled as normal chars.
local IFS=
local result="${1%%[!${2}]*}"
echo ${#result}
}
#:docstring strcspn:
# Usage: strcspn $s1 $s2
#
# Strcspn returns the length of the maximum initial segment of string s1,
# which consists entirely of characters not from string s2.
#:end docstring:
###;;;autoload
function strcspn ()
{
# Unsetting IFS allows whitspace to be handled as normal chars.
local IFS=
local result="${1%%[${2}]*}"
echo ${#result}
}
#:docstring strstr:
# Usage: strstr s1 s2
#
# Strstr echoes a substring starting at the first occurrence of string s2 in
# string s1, or nothing if s2 does not occur in the string. If s2 points to
# a string of zero length, strstr echoes s1.
#:end docstring:
###;;;autoload
function strstr ()
{
# if s2 points to a string of zero length, strstr echoes s1
[ ${#2} -eq 0 ] &#38;&#38; { echo "$1" ; return 0; }
# strstr echoes nothing if s2 does not occur in s1
case "$1" in
*$2*) ;;
*) return 1;;
esac
# use the pattern matching code to strip off the match and everything
# following it
first=${1/$2*/}
# then strip off the first unmatched portion of the string
echo "${1##$first}"
}
#:docstring strtok:
# Usage: strtok s1 s2
#
# Strtok considers the string s1 to consist of a sequence of zero or more
# text tokens separated by spans of one or more characters from the
# separator string s2. The first call (with a non-empty string s1
# specified) echoes a string consisting of the first token on stdout. The
# function keeps track of its position in the string s1 between separate
# calls, so that subsequent calls made with the first argument an empty
# string will work through the string immediately following that token. In
# this way subsequent calls will work through the string s1 until no tokens
# remain. The separator string s2 may be different from call to call.
# When no token remains in s1, an empty value is echoed on stdout.
#:end docstring:
###;;;autoload
function strtok ()
{
:
}
#:docstring strtrunc:
# Usage: strtrunc $n $s1 {$s2} {$...}
#
# Used by many functions like strncmp to truncate arguments for comparison.
# Echoes the first n characters of each string s1 s2 ... on stdout.
#:end docstring:
###;;;autoload
function strtrunc ()
{
n=$1 ; shift
for z; do
echo "${z:0:$n}"
done
}
# provide string
# string.bash ends here
# ========================================================================== #
# ==&#62; Everything below here added by the document author.
# ==&#62; Suggested use of this script is to delete everything below here,
# ==&#62; and "source" this file into your own scripts.
# strcat
string0=one
string1=two
echo
echo "Testing \"strcat\" function:"
echo "Original \"string0\" = $string0"
echo "\"string1\" = $string1"
strcat string0 string1
echo "New \"string0\" = $string0"
echo
# strlen
echo
echo "Testing \"strlen\" function:"
str=123456789
echo "\"str\" = $str"
echo -n "Length of \"str\" = "
strlen str
echo
# Exercise:
# --------
# Add code to test all the other string functions above.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Michael Zick's complex array example uses the <A
HREF="#MD5SUMREF"
>md5sum</A
> check sum command to encode directory
information.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="DIRECTORYINFO"
></A
><P
><B
>Example A-19. Directory information</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
# directory-info.sh
# Parses and lists directory information.
# NOTE: Change lines 273 and 353 per "README" file.
# Michael Zick is the author of this script.
# Used here with his permission.
# Controls
# If overridden by command arguments, they must be in the order:
# Arg1: "Descriptor Directory"
# Arg2: "Exclude Paths"
# Arg3: "Exclude Directories"
#
# Environment Settings override Defaults.
# Command arguments override Environment Settings.
# Default location for content addressed file descriptors.
MD5UCFS=${1:-${MD5UCFS:-'/tmpfs/ucfs'}}
# Directory paths never to list or enter
declare -a \
EXCLUDE_PATHS=${2:-${EXCLUDE_PATHS:-'(/proc /dev /devfs /tmpfs)'}}
# Directories never to list or enter
declare -a \
EXCLUDE_DIRS=${3:-${EXCLUDE_DIRS:-'(ucfs lost+found tmp wtmp)'}}
# Files never to list or enter
declare -a \
EXCLUDE_FILES=${3:-${EXCLUDE_FILES:-'(core "Name with Spaces")'}}
# Here document used as a comment block.
: &#60;&#60;LSfieldsDoc
# # # # # List Filesystem Directory Information # # # # #
#
# ListDirectory "FileGlob" "Field-Array-Name"
# or
# ListDirectory -of "FileGlob" "Field-Array-Filename"
# '-of' meaning 'output to filename'
# # # # #
String format description based on: ls (GNU fileutils) version 4.0.36
Produces a line (or more) formatted:
inode permissions hard-links owner group ...
32736 -rw------- 1 mszick mszick
size day month date hh:mm:ss year path
2756608 Sun Apr 20 08:53:06 2003 /home/mszick/core
Unless it is formatted:
inode permissions hard-links owner group ...
266705 crw-rw---- 1 root uucp
major minor day month date hh:mm:ss year path
4, 68 Sun Apr 20 09:27:33 2003 /dev/ttyS4
NOTE: that pesky comma after the major number
NOTE: the 'path' may be multiple fields:
/home/mszick/core
/proc/982/fd/0 -&#62; /dev/null
/proc/982/fd/1 -&#62; /home/mszick/.xsession-errors
/proc/982/fd/13 -&#62; /tmp/tmpfZVVOCs (deleted)
/proc/982/fd/7 -&#62; /tmp/kde-mszick/ksycoca
/proc/982/fd/8 -&#62; socket:[11586]
/proc/982/fd/9 -&#62; pipe:[11588]
If that isn't enough to keep your parser guessing,
either or both of the path components may be relative:
../Built-Shared -&#62; Built-Static
../linux-2.4.20.tar.bz2 -&#62; ../../../SRCS/linux-2.4.20.tar.bz2
The first character of the 11 (10?) character permissions field:
's' Socket
'd' Directory
'b' Block device
'c' Character device
'l' Symbolic link
NOTE: Hard links not marked - test for identical inode numbers
on identical filesystems.
All information about hard linked files are shared, except
for the names and the name's location in the directory system.
NOTE: A "Hard link" is known as a "File Alias" on some systems.
'-' An undistingushed file
Followed by three groups of letters for: User, Group, Others
Character 1: '-' Not readable; 'r' Readable
Character 2: '-' Not writable; 'w' Writable
Character 3, User and Group: Combined execute and special
'-' Not Executable, Not Special
'x' Executable, Not Special
's' Executable, Special
'S' Not Executable, Special
Character 3, Others: Combined execute and sticky (tacky?)
'-' Not Executable, Not Tacky
'x' Executable, Not Tacky
't' Executable, Tacky
'T' Not Executable, Tacky
Followed by an access indicator
Haven't tested this one, it may be the eleventh character
or it may generate another field
' ' No alternate access
'+' Alternate access
LSfieldsDoc
ListDirectory()
{
local -a T
local -i of=0 # Default return in variable
# OLD_IFS=$IFS # Using BASH default ' \t\n'
case "$#" in
3) case "$1" in
-of) of=1 ; shift ;;
* ) return 1 ;;
esac ;;
2) : ;; # Poor man's "continue"
*) return 1 ;;
esac
# NOTE: the (ls) command is NOT quoted (")
T=( $(ls --inode --ignore-backups --almost-all --directory \
--full-time --color=none --time=status --sort=none \
--format=long $1) )
case $of in
# Assign T back to the array whose name was passed as $2
0) eval $2=\( \"\$\{T\[@\]\}\" \) ;;
# Write T into filename passed as $2
1) echo "${T[@]}" &#62; "$2" ;;
esac
return 0
}
# # # # # Is that string a legal number? # # # # #
#
# IsNumber "Var"
# # # # # There has to be a better way, sigh...
IsNumber()
{
local -i int
if [ $# -eq 0 ]
then
return 1
else
(let int=$1) 2&#62;/dev/null
return $? # Exit status of the let thread
fi
}
# # # # # Index Filesystem Directory Information # # # # #
#
# IndexList "Field-Array-Name" "Index-Array-Name"
# or
# IndexList -if Field-Array-Filename Index-Array-Name
# IndexList -of Field-Array-Name Index-Array-Filename
# IndexList -if -of Field-Array-Filename Index-Array-Filename
# # # # #
: &#60;&#60;IndexListDoc
Walk an array of directory fields produced by ListDirectory
Having suppressed the line breaks in an otherwise line oriented
report, build an index to the array element which starts each line.
Each line gets two index entries, the first element of each line
(inode) and the element that holds the pathname of the file.
The first index entry pair (Line-Number==0) are informational:
Index-Array-Name[0] : Number of "Lines" indexed
Index-Array-Name[1] : "Current Line" pointer into Index-Array-Name
The following index pairs (if any) hold element indexes into
the Field-Array-Name per:
Index-Array-Name[Line-Number * 2] : The "inode" field element.
NOTE: This distance may be either +11 or +12 elements.
Index-Array-Name[(Line-Number * 2) + 1] : The "pathname" element.
NOTE: This distance may be a variable number of elements.
Next line index pair for Line-Number+1.
IndexListDoc
IndexList()
{
local -a LIST # Local of listname passed
local -a -i INDEX=( 0 0 ) # Local of index to return
local -i Lidx Lcnt
local -i if=0 of=0 # Default to variable names
case "$#" in # Simplistic option testing
0) return 1 ;;
1) return 1 ;;
2) : ;; # Poor man's continue
3) case "$1" in
-if) if=1 ;;
-of) of=1 ;;
* ) return 1 ;;
esac ; shift ;;
4) if=1 ; of=1 ; shift ; shift ;;
*) return 1
esac
# Make local copy of list
case "$if" in
0) eval LIST=\( \"\$\{$1\[@\]\}\" \) ;;
1) LIST=( $(cat $1) ) ;;
esac
# Grok (grope?) the array
Lcnt=${#LIST[@]}
Lidx=0
until (( Lidx &#62;= Lcnt ))
do
if IsNumber ${LIST[$Lidx]}
then
local -i inode name
local ft
inode=Lidx
local m=${LIST[$Lidx+2]} # Hard Links field
ft=${LIST[$Lidx+1]:0:1} # Fast-Stat
case $ft in
b) ((Lidx+=12)) ;; # Block device
c) ((Lidx+=12)) ;; # Character device
*) ((Lidx+=11)) ;; # Anything else
esac
name=Lidx
case $ft in
-) ((Lidx+=1)) ;; # The easy one
b) ((Lidx+=1)) ;; # Block device
c) ((Lidx+=1)) ;; # Character device
d) ((Lidx+=1)) ;; # The other easy one
l) ((Lidx+=3)) ;; # At LEAST two more fields
# A little more elegance here would handle pipes,
#+ sockets, deleted files - later.
*) until IsNumber ${LIST[$Lidx]} || ((Lidx &#62;= Lcnt))
do
((Lidx+=1))
done
;; # Not required
esac
INDEX[${#INDEX[*]}]=$inode
INDEX[${#INDEX[*]}]=$name
INDEX[0]=${INDEX[0]}+1 # One more "line" found
# echo "Line: ${INDEX[0]} Type: $ft Links: $m Inode: \
# ${LIST[$inode]} Name: ${LIST[$name]}"
else
((Lidx+=1))
fi
done
case "$of" in
0) eval $2=\( \"\$\{INDEX\[@\]\}\" \) ;;
1) echo "${INDEX[@]}" &#62; "$2" ;;
esac
return 0 # What could go wrong?
}
# # # # # Content Identify File # # # # #
#
# DigestFile Input-Array-Name Digest-Array-Name
# or
# DigestFile -if Input-FileName Digest-Array-Name
# # # # #
# Here document used as a comment block.
: &#60;&#60;DigestFilesDoc
The key (no pun intended) to a Unified Content File System (UCFS)
is to distinguish the files in the system based on their content.
Distinguishing files by their name is just so 20th Century.
The content is distinguished by computing a checksum of that content.
This version uses the md5sum program to generate a 128 bit checksum
representative of the file's contents.
There is a chance that two files having different content might
generate the same checksum using md5sum (or any checksum). Should
that become a problem, then the use of md5sum can be replace by a
cyrptographic signature. But until then...
The md5sum program is documented as outputting three fields (and it
does), but when read it appears as two fields (array elements). This
is caused by the lack of whitespace between the second and third field.
So this function gropes the md5sum output and returns:
[0] 32 character checksum in hexidecimal (UCFS filename)
[1] Single character: ' ' text file, '*' binary file
[2] Filesystem (20th Century Style) name
Note: That name may be the character '-' indicating STDIN read.
DigestFilesDoc
DigestFile()
{
local if=0 # Default, variable name
local -a T1 T2
case "$#" in
3) case "$1" in
-if) if=1 ; shift ;;
* ) return 1 ;;
esac ;;
2) : ;; # Poor man's "continue"
*) return 1 ;;
esac
case $if in
0) eval T1=\( \"\$\{$1\[@\]\}\" \)
T2=( $(echo ${T1[@]} | md5sum -) )
;;
1) T2=( $(md5sum $1) )
;;
esac
case ${#T2[@]} in
0) return 1 ;;
1) return 1 ;;
2) case ${T2[1]:0:1} in # SanScrit-2.0.5
\*) T2[${#T2[@]}]=${T2[1]:1}
T2[1]=\*
;;
*) T2[${#T2[@]}]=${T2[1]}
T2[1]=" "
;;
esac
;;
3) : ;; # Assume it worked
*) return 1 ;;
esac
local -i len=${#T2[0]}
if [ $len -ne 32 ] ; then return 1 ; fi
eval $2=\( \"\$\{T2\[@\]\}\" \)
}
# # # # # Locate File # # # # #
#
# LocateFile [-l] FileName Location-Array-Name
# or
# LocateFile [-l] -of FileName Location-Array-FileName
# # # # #
# A file location is Filesystem-id and inode-number
# Here document used as a comment block.
: &#60;&#60;StatFieldsDoc
Based on stat, version 2.2
stat -t and stat -lt fields
[0] name
[1] Total size
File - number of bytes
Symbolic link - string length of pathname
[2] Number of (512 byte) blocks allocated
[3] File type and Access rights (hex)
[4] User ID of owner
[5] Group ID of owner
[6] Device number
[7] Inode number
[8] Number of hard links
[9] Device type (if inode device) Major
[10] Device type (if inode device) Minor
[11] Time of last access
May be disabled in 'mount' with noatime
atime of files changed by exec, read, pipe, utime, mknod (mmap?)
atime of directories changed by addition/deletion of files
[12] Time of last modification
mtime of files changed by write, truncate, utime, mknod
mtime of directories changed by addtition/deletion of files
[13] Time of last change
ctime reflects time of changed inode information (owner, group
permissions, link count
-*-*- Per:
Return code: 0
Size of array: 14
Contents of array
Element 0: /home/mszick
Element 1: 4096
Element 2: 8
Element 3: 41e8
Element 4: 500
Element 5: 500
Element 6: 303
Element 7: 32385
Element 8: 22
Element 9: 0
Element 10: 0
Element 11: 1051221030
Element 12: 1051214068
Element 13: 1051214068
For a link in the form of linkname -&#62; realname
stat -t linkname returns the linkname (link) information
stat -lt linkname returns the realname information
stat -tf and stat -ltf fields
[0] name
[1] ID-0? # Maybe someday, but Linux stat structure
[2] ID-0? # does not have either LABEL nor UUID
# fields, currently information must come
# from file-system specific utilities
These will be munged into:
[1] UUID if possible
[2] Volume Label if possible
Note: 'mount -l' does return the label and could return the UUID
[3] Maximum length of filenames
[4] Filesystem type
[5] Total blocks in the filesystem
[6] Free blocks
[7] Free blocks for non-root user(s)
[8] Block size of the filesystem
[9] Total inodes
[10] Free inodes
-*-*- Per:
Return code: 0
Size of array: 11
Contents of array
Element 0: /home/mszick
Element 1: 0
Element 2: 0
Element 3: 255
Element 4: ef53
Element 5: 2581445
Element 6: 2277180
Element 7: 2146050
Element 8: 4096
Element 9: 1311552
Element 10: 1276425
StatFieldsDoc
# LocateFile [-l] FileName Location-Array-Name
# LocateFile [-l] -of FileName Location-Array-FileName
LocateFile()
{
local -a LOC LOC1 LOC2
local lk="" of=0
case "$#" in
0) return 1 ;;
1) return 1 ;;
2) : ;;
*) while (( "$#" &#62; 2 ))
do
case "$1" in
-l) lk=-1 ;;
-of) of=1 ;;
*) return 1 ;;
esac
shift
done ;;
esac
# More Sanscrit-2.0.5
# LOC1=( $(stat -t $lk $1) )
# LOC2=( $(stat -tf $lk $1) )
# Uncomment above two lines if system has "stat" command installed.
LOC=( ${LOC1[@]:0:1} ${LOC1[@]:3:11}
${LOC2[@]:1:2} ${LOC2[@]:4:1} )
case "$of" in
0) eval $2=\( \"\$\{LOC\[@\]\}\" \) ;;
1) echo "${LOC[@]}" &#62; "$2" ;;
esac
return 0
# Which yields (if you are lucky, and have "stat" installed)
# -*-*- Location Discriptor -*-*-
# Return code: 0
# Size of array: 15
# Contents of array
# Element 0: /home/mszick 20th Century name
# Element 1: 41e8 Type and Permissions
# Element 2: 500 User
# Element 3: 500 Group
# Element 4: 303 Device
# Element 5: 32385 inode
# Element 6: 22 Link count
# Element 7: 0 Device Major
# Element 8: 0 Device Minor
# Element 9: 1051224608 Last Access
# Element 10: 1051214068 Last Modify
# Element 11: 1051214068 Last Status
# Element 12: 0 UUID (to be)
# Element 13: 0 Volume Label (to be)
# Element 14: ef53 Filesystem type
}
# And then there was some test code
ListArray() # ListArray Name
{
local -a Ta
eval Ta=\( \"\$\{$1\[@\]\}\" \)
echo
echo "-*-*- List of Array -*-*-"
echo "Size of array $1: ${#Ta[*]}"
echo "Contents of array $1:"
for (( i=0 ; i&#60;${#Ta[*]} ; i++ ))
do
echo -e "\tElement $i: ${Ta[$i]}"
done
return 0
}
declare -a CUR_DIR
# For small arrays
ListDirectory "${PWD}" CUR_DIR
ListArray CUR_DIR
declare -a DIR_DIG
DigestFile CUR_DIR DIR_DIG
echo "The new \"name\" (checksum) for ${CUR_DIR[9]} is ${DIR_DIG[0]}"
declare -a DIR_ENT
# BIG_DIR # For really big arrays - use a temporary file in ramdisk
# BIG-DIR # ListDirectory -of "${CUR_DIR[11]}/*" "/tmpfs/junk2"
ListDirectory "${CUR_DIR[11]}/*" DIR_ENT
declare -a DIR_IDX
# BIG-DIR # IndexList -if "/tmpfs/junk2" DIR_IDX
IndexList DIR_ENT DIR_IDX
declare -a IDX_DIG
# BIG-DIR # DIR_ENT=( $(cat /tmpfs/junk2) )
# BIG-DIR # DigestFile -if /tmpfs/junk2 IDX_DIG
DigestFile DIR_ENT IDX_DIG
# Small (should) be able to parallize IndexList &#38; DigestFile
# Large (should) be able to parallize IndexList &#38; DigestFile &#38; the assignment
echo "The \"name\" (checksum) for the contents of ${PWD} is ${IDX_DIG[0]}"
declare -a FILE_LOC
LocateFile ${PWD} FILE_LOC
ListArray FILE_LOC
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>St<53>phane Chazelas demonstrates object-oriented programming in a
Bash script.</P
><P
>Mariusz Gniazdowski contributed a <A
HREF="#HASHREF"
>hash</A
>
library for use in scripts.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="HASHLIB"
></A
><P
><B
>Example A-20. Library of hash functions</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># Hash:
# Hash function library
# Author: Mariusz Gniazdowski &#60;mariusz.gn-at-gmail.com&#62;
# Date: 2005-04-07
# Functions making emulating hashes in Bash a little less painful.
# Limitations:
# * Only global variables are supported.
# * Each hash instance generates one global variable per value.
# * Variable names collisions are possible
#+ if you define variable like __hash__hashname_key
# * Keys must use chars that can be part of a Bash variable name
#+ (no dashes, periods, etc.).
# * The hash is created as a variable:
# ... hashname_keyname
# So if somone will create hashes like:
# myhash_ + mykey = myhash__mykey
# myhash + _mykey = myhash__mykey
# Then there will be a collision.
# (This should not pose a major problem.)
Hash_config_varname_prefix=__hash__
# Emulates: hash[key]=value
#
# Params:
# 1 - hash
# 2 - key
# 3 - value
function hash_set {
eval "${Hash_config_varname_prefix}${1}_${2}=\"${3}\""
}
# Emulates: value=hash[key]
#
# Params:
# 1 - hash
# 2 - key
# 3 - value (name of global variable to set)
function hash_get_into {
eval "$3=\"\$${Hash_config_varname_prefix}${1}_${2}\""
}
# Emulates: echo hash[key]
#
# Params:
# 1 - hash
# 2 - key
# 3 - echo params (like -n, for example)
function hash_echo {
eval "echo $3 \"\$${Hash_config_varname_prefix}${1}_${2}\""
}
# Emulates: hash1[key1]=hash2[key2]
#
# Params:
# 1 - hash1
# 2 - key1
# 3 - hash2
# 4 - key2
function hash_copy {
eval "${Hash_config_varname_prefix}${1}_${2}\
=\"\$${Hash_config_varname_prefix}${3}_${4}\""
}
# Emulates: hash[keyN-1]=hash[key2]=...hash[key1]
#
# Copies first key to rest of keys.
#
# Params:
# 1 - hash1
# 2 - key1
# 3 - key2
# . . .
# N - keyN
function hash_dup {
local hashName="$1" keyName="$2"
shift 2
until [ ${#} -le 0 ]; do
eval "${Hash_config_varname_prefix}${hashName}_${1}\
=\"\$${Hash_config_varname_prefix}${hashName}_${keyName}\""
shift;
done;
}
# Emulates: unset hash[key]
#
# Params:
# 1 - hash
# 2 - key
function hash_unset {
eval "unset ${Hash_config_varname_prefix}${1}_${2}"
}
# Emulates something similar to: ref=&#38;hash[key]
#
# The reference is name of the variable in which value is held.
#
# Params:
# 1 - hash
# 2 - key
# 3 - ref - Name of global variable to set.
function hash_get_ref_into {
eval "$3=\"${Hash_config_varname_prefix}${1}_${2}\""
}
# Emulates something similar to: echo &#38;hash[key]
#
# That reference is name of variable in which value is held.
#
# Params:
# 1 - hash
# 2 - key
# 3 - echo params (like -n for example)
function hash_echo_ref {
eval "echo $3 \"${Hash_config_varname_prefix}${1}_${2}\""
}
# Emulates something similar to: $$hash[key](param1, param2, ...)
#
# Params:
# 1 - hash
# 2 - key
# 3,4, ... - Function parameters
function hash_call {
local hash key
hash=$1
key=$2
shift 2
eval "eval \"\$${Hash_config_varname_prefix}${hash}_${key} \\\"\\\$@\\\"\""
}
# Emulates something similar to: isset(hash[key]) or hash[key]==NULL
#
# Params:
# 1 - hash
# 2 - key
# Returns:
# 0 - there is such key
# 1 - there is no such key
function hash_is_set {
eval "if [[ \"\${${Hash_config_varname_prefix}${1}_${2}-a}\" = \"a\" &#38;&#38;
\"\${${Hash_config_varname_prefix}${1}_${2}-b}\" = \"b\" ]]
then return 1; else return 0; fi"
}
# Emulates something similar to:
# foreach($hash as $key =&#62; $value) { fun($key,$value); }
#
# It is possible to write different variations of this function.
# Here we use a function call to make it as "generic" as possible.
#
# Params:
# 1 - hash
# 2 - function name
function hash_foreach {
local keyname oldIFS="$IFS"
IFS=' '
for i in $(eval "echo \${!${Hash_config_varname_prefix}${1}_*}"); do
keyname=$(eval "echo \${i##${Hash_config_varname_prefix}${1}_}")
eval "$2 $keyname \"\$$i\""
done
IFS="$oldIFS"
}
# NOTE: In lines 103 and 116, ampersand changed.
# But, it doesn't matter, because these are comment lines anyhow.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Here is an example script using the foregoing hash library.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="HASHEXAMPLE"
></A
><P
><B
>Example A-21. Colorizing text using hash functions</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# hash-example.sh: Colorizing text.
# Author: Mariusz Gniazdowski &#60;mariusz.gn-at-gmail.com&#62;
. Hash.lib # Load the library of functions.
hash_set colors red "\033[0;31m"
hash_set colors blue "\033[0;34m"
hash_set colors light_blue "\033[1;34m"
hash_set colors light_red "\033[1;31m"
hash_set colors cyan "\033[0;36m"
hash_set colors light_green "\033[1;32m"
hash_set colors light_gray "\033[0;37m"
hash_set colors green "\033[0;32m"
hash_set colors yellow "\033[1;33m"
hash_set colors light_purple "\033[1;35m"
hash_set colors purple "\033[0;35m"
hash_set colors reset_color "\033[0;00m"
# $1 - keyname
# $2 - value
try_colors() {
echo -en "$2"
echo "This line is $1."
}
hash_foreach colors try_colors
hash_echo colors reset_color -en
echo -e '\nLet us overwrite some colors with yellow.\n'
# It's hard to read yellow text on some terminals.
hash_dup colors yellow red light_green blue green light_gray cyan
hash_foreach colors try_colors
hash_echo colors reset_color -en
echo -e '\nLet us delete them and try colors once more . . .\n'
for i in red light_green blue green light_gray cyan; do
hash_unset colors $i
done
hash_foreach colors try_colors
hash_echo colors reset_color -en
hash_set other txt "Other examples . . ."
hash_echo other txt
hash_get_into other txt text
echo $text
hash_set other my_fun try_colors
hash_call other my_fun purple "`hash_echo colors purple`"
hash_echo colors reset_color -en
echo; echo "Back to normal?"; echo
exit $?
# On some terminals, the "light" colors print in bold,
# and end up looking darker than the normal ones.
# Why is this?&#13;</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="HASHEX2_0"
></A
>An example illustrating the mechanics
of hashing, but from a different point of view.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="HASHEX2"
></A
><P
><B
>Example A-22. More on hash functions</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# $Id: ha.sh,v 1.2 2005/04/21 23:24:26 oliver Exp $
# Copyright 2005 Oliver Beckstein
# Released under the GNU Public License
# Author of script granted permission for inclusion in ABS Guide.
# (Thank you!)
#----------------------------------------------------------------
# pseudo hash based on indirect parameter expansion
# API: access through functions:
#
# create the hash:
#
# newhash Lovers
#
# add entries (note single quotes for spaces)
#
# addhash Lovers Tristan Isolde
# addhash Lovers 'Romeo Montague' 'Juliet Capulet'
#
# access value by key
#
# gethash Lovers Tristan ----&#62; Isolde
#
# show all keys
#
# keyshash Lovers ----&#62; 'Tristan' 'Romeo Montague'
#
#
# Convention: instead of perls' foo{bar} = boing' syntax,
# use
# '_foo_bar=boing' (two underscores, no spaces)
#
# 1) store key in _NAME_keys[]
# 2) store value in _NAME_values[] using the same integer index
# The integer index for the last entry is _NAME_ptr
#
# NOTE: No error or sanity checks, just bare bones.
function _inihash () {
# private function
# call at the beginning of each procedure
# defines: _keys _values _ptr
#
# Usage: _inihash NAME
local name=$1
_keys=_${name}_keys
_values=_${name}_values
_ptr=_${name}_ptr
}
function newhash () {
# Usage: newhash NAME
# NAME should not contain spaces or dots.
# Actually: it must be a legal name for a Bash variable.
# We rely on Bash automatically recognising arrays.
local name=$1
local _keys _values _ptr
_inihash ${name}
eval ${_ptr}=0
}
function addhash () {
# Usage: addhash NAME KEY 'VALUE with spaces'
# arguments with spaces need to be quoted with single quotes ''
local name=$1 k="$2" v="$3"
local _keys _values _ptr
_inihash ${name}
#echo "DEBUG(addhash): ${_ptr}=${!_ptr}"
eval let ${_ptr}=${_ptr}+1
eval "$_keys[${!_ptr}]=\"${k}\""
eval "$_values[${!_ptr}]=\"${v}\""
}
function gethash () {
# Usage: gethash NAME KEY
# Returns boing
# ERR=0 if entry found, 1 otherwise
# That's not a proper hash --
#+ we simply linearly search through the keys.
local name=$1 key="$2"
local _keys _values _ptr
local k v i found h
_inihash ${name}
# _ptr holds the highest index in the hash
found=0
for i in $(seq 1 ${!_ptr}); do
h="\${${_keys}[${i}]}" # Safer to do it in two steps,
eval k=${h} #+ especially when quoting for spaces.
if [ "${k}" = "${key}" ]; then found=1; break; fi
done;
[ ${found} = 0 ] &#38;&#38; return 1;
# else: i is the index that matches the key
h="\${${_values}[${i}]}"
eval echo "${h}"
return 0;
}
function keyshash () {
# Usage: keyshash NAME
# Returns list of all keys defined for hash name.
local name=$1 key="$2"
local _keys _values _ptr
local k i h
_inihash ${name}
# _ptr holds the highest index in the hash
for i in $(seq 1 ${!_ptr}); do
h="\${${_keys}[${i}]}" # Safer to do it in two steps,
eval k=${h} #+ especially when quoting for spaces.
echo -n "'${k}' "
done;
}
# -----------------------------------------------------------------------
# Now, let's test it.
# (Per comments at the beginning of the script.)
newhash Lovers
addhash Lovers Tristan Isolde
addhash Lovers 'Romeo Montague' 'Juliet Capulet'
# Output results.
echo
gethash Lovers Tristan # Isolde
echo
keyshash Lovers # 'Tristan' 'Romeo Montague'
echo; echo
exit 0
# Exercise:
# --------
# Add error checks to the functions.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Now for a script that installs and mounts
those cute USB keychain solid-state <SPAN
CLASS="QUOTE"
>"hard drives."</SPAN
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="USBINST"
></A
><P
><B
>Example A-23. Mounting USB keychain storage devices</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ==&#62; usb.sh
# ==&#62; Script for mounting and installing pen/keychain USB storage devices.
# ==&#62; Runs as root at system startup (see below).
# ==&#62;
# ==&#62; Newer Linux distros (2004 or later) autodetect
# ==&#62; and install USB pen drives, and therefore don't need this script.
# ==&#62; But, it's still instructive.
# This code is free software covered by GNU GPL license version 2 or above.
# Please refer to http://www.gnu.org/ for the full license text.
#
# Some code lifted from usb-mount by Michael Hamilton's usb-mount (LGPL)
#+ see http://users.actrix.co.nz/michael/usbmount.html
#
# INSTALL
# -------
# Put this in /etc/hotplug/usb/diskonkey.
# Then look in /etc/hotplug/usb.distmap, and copy all usb-storage entries
#+ into /etc/hotplug/usb.usermap, substituting "usb-storage" for "diskonkey".
# Otherwise this code is only run during the kernel module invocation/removal
#+ (at least in my tests), which defeats the purpose.
#
# TODO
# ----
# Handle more than one diskonkey device at one time (e.g. /dev/diskonkey1
#+ and /mnt/diskonkey1), etc. The biggest problem here is the handling in
#+ devlabel, which I haven't yet tried.
#
# AUTHOR and SUPPORT
# ------------------
# Konstantin Riabitsev, &#60;icon linux duke edu&#62;.
# Send any problem reports to my email address at the moment.
#
# ==&#62; Comments added by ABS Guide author.
SYMLINKDEV=/dev/diskonkey
MOUNTPOINT=/mnt/diskonkey
DEVLABEL=/sbin/devlabel
DEVLABELCONFIG=/etc/sysconfig/devlabel
IAM=$0
##
# Functions lifted near-verbatim from usb-mount code.
#
function allAttachedScsiUsb {
find /proc/scsi/ -path '/proc/scsi/usb-storage*' -type f |
xargs grep -l 'Attached: Yes'
}
function scsiDevFromScsiUsb {
echo $1 | awk -F"[-/]" '{ n=$(NF-1);
print "/dev/sd" substr("abcdefghijklmnopqrstuvwxyz", n+1, 1) }'
}
if [ "${ACTION}" = "add" ] &#38;&#38; [ -f "${DEVICE}" ]; then
##
# lifted from usbcam code.
#
if [ -f /var/run/console.lock ]; then
CONSOLEOWNER=`cat /var/run/console.lock`
elif [ -f /var/lock/console.lock ]; then
CONSOLEOWNER=`cat /var/lock/console.lock`
else
CONSOLEOWNER=
fi
for procEntry in $(allAttachedScsiUsb); do
scsiDev=$(scsiDevFromScsiUsb $procEntry)
# Some bug with usb-storage?
# Partitions are not in /proc/partitions until they are accessed
#+ somehow.
/sbin/fdisk -l $scsiDev &#62;/dev/null
##
# Most devices have partitioning info, so the data would be on
#+ /dev/sd?1. However, some stupider ones don't have any partitioning
#+ and use the entire device for data storage. This tries to
#+ guess semi-intelligently if we have a /dev/sd?1 and if not, then
#+ it uses the entire device and hopes for the better.
#
if grep -q `basename $scsiDev`1 /proc/partitions; then
part="$scsiDev""1"
else
part=$scsiDev
fi
##
# Change ownership of the partition to the console user so they can
#+ mount it.
#
if [ ! -z "$CONSOLEOWNER" ]; then
chown $CONSOLEOWNER:disk $part
fi
##
# This checks if we already have this UUID defined with devlabel.
# If not, it then adds the device to the list.
#
prodid=`$DEVLABEL printid -d $part`
if ! grep -q $prodid $DEVLABELCONFIG; then
# cross our fingers and hope it works
$DEVLABEL add -d $part -s $SYMLINKDEV 2&#62;/dev/null
fi
##
# Check if the mount point exists and create if it doesn't.
#
if [ ! -e $MOUNTPOINT ]; then
mkdir -p $MOUNTPOINT
fi
##
# Take care of /etc/fstab so mounting is easy.
#
if ! grep -q "^$SYMLINKDEV" /etc/fstab; then
# Add an fstab entry
echo -e \
"$SYMLINKDEV\t\t$MOUNTPOINT\t\tauto\tnoauto,owner,kudzu 0 0" \
&#62;&#62; /etc/fstab
fi
done
if [ ! -z "$REMOVER" ]; then
##
# Make sure this script is triggered on device removal.
#
mkdir -p `dirname $REMOVER`
ln -s $IAM $REMOVER
fi
elif [ "${ACTION}" = "remove" ]; then
##
# If the device is mounted, unmount it cleanly.
#
if grep -q "$MOUNTPOINT" /etc/mtab; then
# unmount cleanly
umount -l $MOUNTPOINT
fi
##
# Remove it from /etc/fstab if it's there.
#
if grep -q "^$SYMLINKDEV" /etc/fstab; then
grep -v "^$SYMLINKDEV" /etc/fstab &#62; /etc/.fstab.new
mv -f /etc/.fstab.new /etc/fstab
fi
fi
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Converting a text file to HTML format.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="TOHTML"
></A
><P
><B
>Example A-24. Converting to HTML</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# tohtml.sh [v. 0.2.01, reldate: 04/13/12, a teeny bit less buggy]
# Convert a text file to HTML format.
# Author: Mendel Cooper
# License: GPL3
# Usage: sh tohtml.sh &#60; textfile &#62; htmlfile
# Script can easily be modified to accept source and target filenames.
# Assumptions:
# 1) Paragraphs in (target) text file are separated by a blank line.
# 2) Jpeg images (*.jpg) are located in "images" subdirectory.
# In the target file, the image names are enclosed in square brackets,
# for example, [image01.jpg].
# 3) Emphasized (italic) phrases begin with a space+underscore
#+ or the first character on the line is an underscore,
#+ and end with an underscore+space or underscore+end-of-line.
# Settings
FNTSIZE=2 # Small-medium font size
IMGDIR="images" # Image directory
# Headers
HDR01='&#60;!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"&#62;'
HDR02='&#60;!-- Converted to HTML by ***tohtml.sh*** script --&#62;'
HDR03='&#60;!-- script author: M. Leo Cooper &#60;thegrendel.abs@gmail.com&#62; --&#62;'
HDR10='&#60;html&#62;'
HDR11='&#60;head&#62;'
HDR11a='&#60;/head&#62;'
HDR12a='&#60;title&#62;'
HDR12b='&#60;/title&#62;'
HDR121='&#60;META NAME="GENERATOR" CONTENT="tohtml.sh script"&#62;'
HDR13='&#60;body bgcolor="#dddddd"&#62;' # Change background color to suit.
HDR14a='&#60;font size='
HDR14b='&#62;'
# Footers
FTR10='&#60;/body&#62;'
FTR11='&#60;/html&#62;'
# Tags
BOLD="&#60;b&#62;"
CENTER="&#60;center&#62;"
END_CENTER="&#60;/center&#62;"
LF="&#60;br&#62;"
write_headers ()
{
echo "$HDR01"
echo
echo "$HDR02"
echo "$HDR03"
echo
echo
echo "$HDR10"
echo "$HDR11"
echo "$HDR121"
echo "$HDR11a"
echo "$HDR13"
echo
echo -n "$HDR14a"
echo -n "$FNTSIZE"
echo "$HDR14b"
echo
echo "$BOLD" # Everything in bold (more easily readable).
}
process_text ()
{
while read line # Read one line at a time.
do
{
if [ ! "$line" ] # Blank line?
then # Then new paragraph must follow.
echo
echo "$LF" # Insert two &#60;br&#62; tags.
echo "$LF"
echo
continue # Skip the underscore test.
else # Otherwise . . .
if [[ "$line" =~ \[*jpg\] ]] # Is a graphic?
then # Strip away brackets.
temp=$( echo "$line" | sed -e 's/\[//' -e 's/\]//' )
line=""$CENTER" &#60;img src="\"$IMGDIR"/$temp\"&#62; "$END_CENTER" "
# Add image tag.
# And, center it.
fi
fi
echo "$line" | grep -q _
if [ "$?" -eq 0 ] # If line contains underscore ...
then
# ===================================================
# Convert underscored phrase to italics.
temp=$( echo "$line" |
sed -e 's/ _/ &#60;i&#62;/' -e 's/_/&#60;\/i&#62; /' |
sed -e 's/^_/&#60;i&#62;/' -e 's/_/&#60;\/i&#62;/' )
# Process only underscores prefixed by space,
#+ or at beginning or end of line.
# Do not convert underscores embedded within a word!
line="$temp"
# Slows script execution. Can be optimized?
# ===================================================
fi
# echo
echo "$line"
# echo
# Don't want extra blank lines in generated text!
} # End while
done
} # End process_text ()
write_footers () # Termination tags.
{
echo "$FTR10"
echo "$FTR11"
}
# main () {
# =========
write_headers
process_text
write_footers
# =========
# }
exit $?
# Exercises:
# ---------
# 1) Fixup: Check for closing underscore before a comma or period.
# 2) Add a test for the presence of a closing underscore
#+ in phrases to be italicized.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Here is something to warm the hearts of webmasters and mistresses:
a script that saves weblogs.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ARCHIVWEBLOGS"
></A
><P
><B
>Example A-25. Preserving weblogs</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# archiveweblogs.sh v1.0
# Troy Engel &#60;tengel@fluid.com&#62;
# Slightly modified by document author.
# Used with permission.
#
# This script will preserve the normally rotated and
#+ thrown away weblogs from a default RedHat/Apache installation.
# It will save the files with a date/time stamp in the filename,
#+ bzipped, to a given directory.
#
# Run this from crontab nightly at an off hour,
#+ as bzip2 can suck up some serious CPU on huge logs:
# 0 2 * * * /opt/sbin/archiveweblogs.sh
PROBLEM=66
# Set this to your backup dir.
BKP_DIR=/opt/backups/weblogs
# Default Apache/RedHat stuff
LOG_DAYS="4 3 2 1"
LOG_DIR=/var/log/httpd
LOG_FILES="access_log error_log"
# Default RedHat program locations
LS=/bin/ls
MV=/bin/mv
ID=/usr/bin/id
CUT=/bin/cut
COL=/usr/bin/column
BZ2=/usr/bin/bzip2
# Are we root?
USER=`$ID -u`
if [ "X$USER" != "X0" ]; then
echo "PANIC: Only root can run this script!"
exit $PROBLEM
fi
# Backup dir exists/writable?
if [ ! -x $BKP_DIR ]; then
echo "PANIC: $BKP_DIR doesn't exist or isn't writable!"
exit $PROBLEM
fi
# Move, rename and bzip2 the logs
for logday in $LOG_DAYS; do
for logfile in $LOG_FILES; do
MYFILE="$LOG_DIR/$logfile.$logday"
if [ -w $MYFILE ]; then
DTS=`$LS -lgo --time-style=+%Y%m%d $MYFILE | $COL -t | $CUT -d ' ' -f7`
$MV $MYFILE $BKP_DIR/$logfile.$DTS
$BZ2 $BKP_DIR/$logfile.$DTS
else
# Only spew an error if the file exits (ergo non-writable).
if [ -f $MYFILE ]; then
echo "ERROR: $MYFILE not writable. Skipping."
fi
fi
done
done
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="PROTECTLITERAL0"
></A
>How to keep the shell from
expanding and reinterpreting text strings.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="PROTECTLITERAL"
></A
><P
><B
>Example A-26. Protecting literal strings</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
# protect_literal.sh
# set -vx
:&#60;&#60;-'_Protect_Literal_String_Doc'
Copyright (c) Michael S. Zick, 2003; All Rights Reserved
License: Unrestricted reuse in any form, for any purpose.
Warranty: None
Revision: $ID$
Documentation redirected to the Bash no-operation.
Bash will '/dev/null' this block when the script is first read.
(Uncomment the above set command to see this action.)
Remove the first (Sha-Bang) line when sourcing this as a library
procedure. Also comment out the example use code in the two
places where shown.
Usage:
_protect_literal_str 'Whatever string meets your ${fancy}'
Just echos the argument to standard out, hard quotes
restored.
$(_protect_literal_str 'Whatever string meets your ${fancy}')
as the right-hand-side of an assignment statement.
Does:
As the right-hand-side of an assignment, preserves the
hard quotes protecting the contents of the literal during
assignment.
Notes:
The strange names (_*) are used to avoid trampling on
the user's chosen names when this is sourced as a
library.
_Protect_Literal_String_Doc
# The 'for illustration' function form
_protect_literal_str() {
# Pick an un-used, non-printing character as local IFS.
# Not required, but shows that we are ignoring it.
local IFS=$'\x1B' # \ESC character
# Enclose the All-Elements-Of in hard quotes during assignment.
local tmp=$'\x27'$@$'\x27'
# local tmp=$'\''$@$'\'' # Even uglier.
local len=${#tmp} # Info only.
echo $tmp is $len long. # Output AND information.
}
# This is the short-named version.
_pls() {
local IFS=$'x1B' # \ESC character (not required)
echo $'\x27'$@$'\x27' # Hard quoted parameter glob
}
# :&#60;&#60;-'_Protect_Literal_String_Test'
# # # Remove the above "# " to disable this code. # # #
# See how that looks when printed.
echo
echo "- - Test One - -"
_protect_literal_str 'Hello $user'
_protect_literal_str 'Hello "${username}"'
echo
# Which yields:
# - - Test One - -
# 'Hello $user' is 13 long.
# 'Hello "${username}"' is 21 long.
# Looks as expected, but why all of the trouble?
# The difference is hidden inside the Bash internal order
#+ of operations.
# Which shows when you use it on the RHS of an assignment.
# Declare an array for test values.
declare -a arrayZ
# Assign elements with various types of quotes and escapes.
arrayZ=( zero "$(_pls 'Hello ${Me}')" 'Hello ${You}' "\'Pass: ${pw}\'" )
# Now list that array and see what is there.
echo "- - Test Two - -"
for (( i=0 ; i&#60;${#arrayZ[*]} ; i++ ))
do
echo Element $i: ${arrayZ[$i]} is: ${#arrayZ[$i]} long.
done
echo
# Which yields:
# - - Test Two - -
# Element 0: zero is: 4 long. # Our marker element
# Element 1: 'Hello ${Me}' is: 13 long. # Our "$(_pls '...' )"
# Element 2: Hello ${You} is: 12 long. # Quotes are missing
# Element 3: \'Pass: \' is: 10 long. # ${pw} expanded to nothing
# Now make an assignment with that result.
declare -a array2=( ${arrayZ[@]} )
# And print what happened.
echo "- - Test Three - -"
for (( i=0 ; i&#60;${#array2[*]} ; i++ ))
do
echo Element $i: ${array2[$i]} is: ${#array2[$i]} long.
done
echo
# Which yields:
# - - Test Three - -
# Element 0: zero is: 4 long. # Our marker element.
# Element 1: Hello ${Me} is: 11 long. # Intended result.
# Element 2: Hello is: 5 long. # ${You} expanded to nothing.
# Element 3: 'Pass: is: 6 long. # Split on the whitespace.
# Element 4: ' is: 1 long. # The end quote is here now.
# Our Element 1 has had its leading and trailing hard quotes stripped.
# Although not shown, leading and trailing whitespace is also stripped.
# Now that the string contents are set, Bash will always, internally,
#+ hard quote the contents as required during its operations.
# Why?
# Considering our "$(_pls 'Hello ${Me}')" construction:
# " ... " -&#62; Expansion required, strip the quotes.
# $( ... ) -&#62; Replace with the result of..., strip this.
# _pls ' ... ' -&#62; called with literal arguments, strip the quotes.
# The result returned includes hard quotes; BUT the above processing
#+ has already been done, so they become part of the value assigned.
#
# Similarly, during further usage of the string variable, the ${Me}
#+ is part of the contents (result) and survives any operations
# (Until explicitly told to evaluate the string).
# Hint: See what happens when the hard quotes ($'\x27') are replaced
#+ with soft quotes ($'\x22') in the above procedures.
# Interesting also is to remove the addition of any quoting.
# _Protect_Literal_String_Test
# # # Remove the above "# " to disable this code. # # #
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="UNPROTECTLITERAL0"
></A
>But, what if you
<EM
>want</EM
> the shell to expand
and reinterpret strings?</P
><DIV
CLASS="EXAMPLE"
><A
NAME="UNPROTECTLITERAL"
></A
><P
><B
>Example A-27. Unprotecting literal strings</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
# unprotect_literal.sh
# set -vx
:&#60;&#60;-'_UnProtect_Literal_String_Doc'
Copyright (c) Michael S. Zick, 2003; All Rights Reserved
License: Unrestricted reuse in any form, for any purpose.
Warranty: None
Revision: $ID$
Documentation redirected to the Bash no-operation. Bash will
'/dev/null' this block when the script is first read.
(Uncomment the above set command to see this action.)
Remove the first (Sha-Bang) line when sourcing this as a library
procedure. Also comment out the example use code in the two
places where shown.
Usage:
Complement of the "$(_pls 'Literal String')" function.
(See the protect_literal.sh example.)
StringVar=$(_upls ProtectedSringVariable)
Does:
When used on the right-hand-side of an assignment statement;
makes the substitions embedded in the protected string.
Notes:
The strange names (_*) are used to avoid trampling on
the user's chosen names when this is sourced as a
library.
_UnProtect_Literal_String_Doc
_upls() {
local IFS=$'x1B' # \ESC character (not required)
eval echo $@ # Substitution on the glob.
}
# :&#60;&#60;-'_UnProtect_Literal_String_Test'
# # # Remove the above "# " to disable this code. # # #
_pls() {
local IFS=$'x1B' # \ESC character (not required)
echo $'\x27'$@$'\x27' # Hard quoted parameter glob
}
# Declare an array for test values.
declare -a arrayZ
# Assign elements with various types of quotes and escapes.
arrayZ=( zero "$(_pls 'Hello ${Me}')" 'Hello ${You}' "\'Pass: ${pw}\'" )
# Now make an assignment with that result.
declare -a array2=( ${arrayZ[@]} )
# Which yielded:
# - - Test Three - -
# Element 0: zero is: 4 long # Our marker element.
# Element 1: Hello ${Me} is: 11 long # Intended result.
# Element 2: Hello is: 5 long # ${You} expanded to nothing.
# Element 3: 'Pass: is: 6 long # Split on the whitespace.
# Element 4: ' is: 1 long # The end quote is here now.
# set -vx
# Initialize 'Me' to something for the embedded ${Me} substitution.
# This needs to be done ONLY just prior to evaluating the
#+ protected string.
# (This is why it was protected to begin with.)
Me="to the array guy."
# Set a string variable destination to the result.
newVar=$(_upls ${array2[1]})
# Show what the contents are.
echo $newVar
# Do we really need a function to do this?
newerVar=$(eval echo ${array2[1]})
echo $newerVar
# I guess not, but the _upls function gives us a place to hang
#+ the documentation on.
# This helps when we forget what a # construction like:
#+ $(eval echo ... ) means.
# What if Me isn't set when the protected string is evaluated?
unset Me
newestVar=$(_upls ${array2[1]})
echo $newestVar
# Just gone, no hints, no runs, no errors.
# Why in the world?
# Setting the contents of a string variable containing character
#+ sequences that have a meaning in Bash is a general problem in
#+ script programming.
#
# This problem is now solved in eight lines of code
#+ (and four pages of description).
# Where is all this going?
# Dynamic content Web pages as an array of Bash strings.
# Content set per request by a Bash 'eval' command
#+ on the stored page template.
# Not intended to replace PHP, just an interesting thing to do.
###
# Don't have a webserver application?
# No problem, check the example directory of the Bash source;
#+ there is a Bash script for that also.
# _UnProtect_Literal_String_Test
# # # Remove the above "# " to disable this code. # # #
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>This interesting script helps hunt down spammers.</P
><P
><A
NAME="ISSPAMMER2_0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="ISSPAMMER2"
></A
><P
><B
>Example A-28. Spammer Identification</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# $Id: is_spammer.bash,v 1.12.2.11 2004/10/01 21:42:33 mszick Exp $
# Above line is RCS info.
# The latest version of this script is available from http://www.morethan.org.
#
# Spammer-identification
# by Michael S. Zick
# Used in the ABS Guide with permission.
#######################################################
# Documentation
# See also "Quickstart" at end of script.
#######################################################
:&#60;&#60;-'__is_spammer_Doc_'
Copyright (c) Michael S. Zick, 2004
License: Unrestricted reuse in any form, for any purpose.
Warranty: None -{Its a script; the user is on their own.}-
Impatient?
Application code: goto "# # # Hunt the Spammer' program code # # #"
Example output: ":&#60;&#60;-'_is_spammer_outputs_'"
How to use: Enter script name without arguments.
Or goto "Quickstart" at end of script.
Provides
Given a domain name or IP(v4) address as input:
Does an exhaustive set of queries to find the associated
network resources (short of recursing into TLDs).
Checks the IP(v4) addresses found against Blacklist
nameservers.
If found to be a blacklisted IP(v4) address,
reports the blacklist text records.
(Usually hyper-links to the specific report.)
Requires
A working Internet connection.
(Exercise: Add check and/or abort if not on-line when running script.)
Bash with arrays (2.05b+).
The external program 'dig' --
a utility program provided with the 'bind' set of programs.
Specifically, the version which is part of Bind series 9.x
See: http://www.isc.org
All usages of 'dig' are limited to wrapper functions,
which may be rewritten as required.
See: dig_wrappers.bash for details.
("Additional documentation" -- below)
Usage
Script requires a single argument, which may be:
1) A domain name;
2) An IP(v4) address;
3) A filename, with one name or address per line.
Script accepts an optional second argument, which may be:
1) A Blacklist server name;
2) A filename, with one Blacklist server name per line.
If the second argument is not provided, the script uses
a built-in set of (free) Blacklist servers.
See also, the Quickstart at the end of this script (after 'exit').
Return Codes
0 - All OK
1 - Script failure
2 - Something is Blacklisted
Optional environment variables
SPAMMER_TRACE
If set to a writable file,
script will log an execution flow trace.
SPAMMER_DATA
If set to a writable file, script will dump its
discovered data in the form of GraphViz file.
See: http://www.research.att.com/sw/tools/graphviz
SPAMMER_LIMIT
Limits the depth of resource tracing.
Default is 2 levels.
A setting of 0 (zero) means 'unlimited' . . .
Caution: script might recurse the whole Internet!
A limit of 1 or 2 is most useful when processing
a file of domain names and addresses.
A higher limit can be useful when hunting spam gangs.
Additional documentation
Download the archived set of scripts
explaining and illustrating the function contained within this script.
http://bash.deta.in/mszick_clf.tar.bz2
Study notes
This script uses a large number of functions.
Nearly all general functions have their own example script.
Each of the example scripts have tutorial level comments.
Scripting project
Add support for IP(v6) addresses.
IP(v6) addresses are recognized but not processed.
Advanced project
Add the reverse lookup detail to the discovered information.
Report the delegation chain and abuse contacts.
Modify the GraphViz file output to include the
newly discovered information.
__is_spammer_Doc_
#######################################################
#### Special IFS settings used for string parsing. ####
# Whitespace == :Space:Tab:Line Feed:Carriage Return:
WSP_IFS=$'\x20'$'\x09'$'\x0A'$'\x0D'
# No Whitespace == Line Feed:Carriage Return
NO_WSP=$'\x0A'$'\x0D'
# Field separator for dotted decimal IP addresses
ADR_IFS=${NO_WSP}'.'
# Array to dotted string conversions
DOT_IFS='.'${WSP_IFS}
# # # Pending operations stack machine # # #
# This set of functions described in func_stack.bash.
# (See "Additional documentation" above.)
# # #
# Global stack of pending operations.
declare -f -a _pending_
# Global sentinel for stack runners
declare -i _p_ctrl_
# Global holder for currently executing function
declare -f _pend_current_
# # # Debug version only - remove for regular use # # #
#
# The function stored in _pend_hook_ is called
# immediately before each pending function is
# evaluated. Stack clean, _pend_current_ set.
#
# This thingy demonstrated in pend_hook.bash.
declare -f _pend_hook_
# # #
# The do nothing function
pend_dummy() { : ; }
# Clear and initialize the function stack.
pend_init() {
unset _pending_[@]
pend_func pend_stop_mark
_pend_hook_='pend_dummy' # Debug only.
}
# Discard the top function on the stack.
pend_pop() {
if [ ${#_pending_[@]} -gt 0 ]
then
local -i _top_
_top_=${#_pending_[@]}-1
unset _pending_[$_top_]
fi
}
# pend_func function_name [$(printf '%q\n' arguments)]
pend_func() {
local IFS=${NO_WSP}
set -f
_pending_[${#_pending_[@]}]=$@
set +f
}
# The function which stops the release:
pend_stop_mark() {
_p_ctrl_=0
}
pend_mark() {
pend_func pend_stop_mark
}
# Execute functions until 'pend_stop_mark' . . .
pend_release() {
local -i _top_ # Declare _top_ as integer.
_p_ctrl_=${#_pending_[@]}
while [ ${_p_ctrl_} -gt 0 ]
do
_top_=${#_pending_[@]}-1
_pend_current_=${_pending_[$_top_]}
unset _pending_[$_top_]
$_pend_hook_ # Debug only.
eval $_pend_current_
done
}
# Drop functions until 'pend_stop_mark' . . .
pend_drop() {
local -i _top_
local _pd_ctrl_=${#_pending_[@]}
while [ ${_pd_ctrl_} -gt 0 ]
do
_top_=$_pd_ctrl_-1
if [ "${_pending_[$_top_]}" == 'pend_stop_mark' ]
then
unset _pending_[$_top_]
break
else
unset _pending_[$_top_]
_pd_ctrl_=$_top_
fi
done
if [ ${#_pending_[@]} -eq 0 ]
then
pend_func pend_stop_mark
fi
}
#### Array editors ####
# This function described in edit_exact.bash.
# (See "Additional documentation," above.)
# edit_exact &#60;excludes_array_name&#62; &#60;target_array_name&#62;
edit_exact() {
[ $# -eq 2 ] ||
[ $# -eq 3 ] || return 1
local -a _ee_Excludes
local -a _ee_Target
local _ee_x
local _ee_t
local IFS=${NO_WSP}
set -f
eval _ee_Excludes=\( \$\{$1\[@\]\} \)
eval _ee_Target=\( \$\{$2\[@\]\} \)
local _ee_len=${#_ee_Target[@]} # Original length.
local _ee_cnt=${#_ee_Excludes[@]} # Exclude list length.
[ ${_ee_len} -ne 0 ] || return 0 # Can't edit zero length.
[ ${_ee_cnt} -ne 0 ] || return 0 # Can't edit zero length.
for (( x = 0; x &#60; ${_ee_cnt} ; x++ ))
do
_ee_x=${_ee_Excludes[$x]}
for (( n = 0 ; n &#60; ${_ee_len} ; n++ ))
do
_ee_t=${_ee_Target[$n]}
if [ x"${_ee_t}" == x"${_ee_x}" ]
then
unset _ee_Target[$n] # Discard match.
[ $# -eq 2 ] &#38;&#38; break # If 2 arguments, then done.
fi
done
done
eval $2=\( \$\{_ee_Target\[@\]\} \)
set +f
return 0
}
# This function described in edit_by_glob.bash.
# edit_by_glob &#60;excludes_array_name&#62; &#60;target_array_name&#62;
edit_by_glob() {
[ $# -eq 2 ] ||
[ $# -eq 3 ] || return 1
local -a _ebg_Excludes
local -a _ebg_Target
local _ebg_x
local _ebg_t
local IFS=${NO_WSP}
set -f
eval _ebg_Excludes=\( \$\{$1\[@\]\} \)
eval _ebg_Target=\( \$\{$2\[@\]\} \)
local _ebg_len=${#_ebg_Target[@]}
local _ebg_cnt=${#_ebg_Excludes[@]}
[ ${_ebg_len} -ne 0 ] || return 0
[ ${_ebg_cnt} -ne 0 ] || return 0
for (( x = 0; x &#60; ${_ebg_cnt} ; x++ ))
do
_ebg_x=${_ebg_Excludes[$x]}
for (( n = 0 ; n &#60; ${_ebg_len} ; n++ ))
do
[ $# -eq 3 ] &#38;&#38; _ebg_x=${_ebg_x}'*' # Do prefix edit
if [ ${_ebg_Target[$n]:=} ] #+ if defined &#38; set.
then
_ebg_t=${_ebg_Target[$n]/#${_ebg_x}/}
[ ${#_ebg_t} -eq 0 ] &#38;&#38; unset _ebg_Target[$n]
fi
done
done
eval $2=\( \$\{_ebg_Target\[@\]\} \)
set +f
return 0
}
# This function described in unique_lines.bash.
# unique_lines &#60;in_name&#62; &#60;out_name&#62;
unique_lines() {
[ $# -eq 2 ] || return 1
local -a _ul_in
local -a _ul_out
local -i _ul_cnt
local -i _ul_pos
local _ul_tmp
local IFS=${NO_WSP}
set -f
eval _ul_in=\( \$\{$1\[@\]\} \)
_ul_cnt=${#_ul_in[@]}
for (( _ul_pos = 0 ; _ul_pos &#60; ${_ul_cnt} ; _ul_pos++ ))
do
if [ ${_ul_in[${_ul_pos}]:=} ] # If defined &#38; not empty
then
_ul_tmp=${_ul_in[${_ul_pos}]}
_ul_out[${#_ul_out[@]}]=${_ul_tmp}
for (( zap = _ul_pos ; zap &#60; ${_ul_cnt} ; zap++ ))
do
[ ${_ul_in[${zap}]:=} ] &#38;&#38;
[ 'x'${_ul_in[${zap}]} == 'x'${_ul_tmp} ] &#38;&#38;
unset _ul_in[${zap}]
done
fi
done
eval $2=\( \$\{_ul_out\[@\]\} \)
set +f
return 0
}
# This function described in char_convert.bash.
# to_lower &#60;string&#62;
to_lower() {
[ $# -eq 1 ] || return 1
local _tl_out
_tl_out=${1//A/a}
_tl_out=${_tl_out//B/b}
_tl_out=${_tl_out//C/c}
_tl_out=${_tl_out//D/d}
_tl_out=${_tl_out//E/e}
_tl_out=${_tl_out//F/f}
_tl_out=${_tl_out//G/g}
_tl_out=${_tl_out//H/h}
_tl_out=${_tl_out//I/i}
_tl_out=${_tl_out//J/j}
_tl_out=${_tl_out//K/k}
_tl_out=${_tl_out//L/l}
_tl_out=${_tl_out//M/m}
_tl_out=${_tl_out//N/n}
_tl_out=${_tl_out//O/o}
_tl_out=${_tl_out//P/p}
_tl_out=${_tl_out//Q/q}
_tl_out=${_tl_out//R/r}
_tl_out=${_tl_out//S/s}
_tl_out=${_tl_out//T/t}
_tl_out=${_tl_out//U/u}
_tl_out=${_tl_out//V/v}
_tl_out=${_tl_out//W/w}
_tl_out=${_tl_out//X/x}
_tl_out=${_tl_out//Y/y}
_tl_out=${_tl_out//Z/z}
echo ${_tl_out}
return 0
}
#### Application helper functions ####
# Not everybody uses dots as separators (APNIC, for example).
# This function described in to_dot.bash
# to_dot &#60;string&#62;
to_dot() {
[ $# -eq 1 ] || return 1
echo ${1//[#|@|%]/.}
return 0
}
# This function described in is_number.bash.
# is_number &#60;input&#62;
is_number() {
[ "$#" -eq 1 ] || return 1 # is blank?
[ x"$1" == 'x0' ] &#38;&#38; return 0 # is zero?
local -i tst
let tst=$1 2&#62;/dev/null # else is numeric!
return $?
}
# This function described in is_address.bash.
# is_address &#60;input&#62;
is_address() {
[ $# -eq 1 ] || return 1 # Blank ==&#62; false
local -a _ia_input
local IFS=${ADR_IFS}
_ia_input=( $1 )
if [ ${#_ia_input[@]} -eq 4 ] &#38;&#38;
is_number ${_ia_input[0]} &#38;&#38;
is_number ${_ia_input[1]} &#38;&#38;
is_number ${_ia_input[2]} &#38;&#38;
is_number ${_ia_input[3]} &#38;&#38;
[ ${_ia_input[0]} -lt 256 ] &#38;&#38;
[ ${_ia_input[1]} -lt 256 ] &#38;&#38;
[ ${_ia_input[2]} -lt 256 ] &#38;&#38;
[ ${_ia_input[3]} -lt 256 ]
then
return 0
else
return 1
fi
}
# This function described in split_ip.bash.
# split_ip &#60;IP_address&#62;
#+ &#60;array_name_norm&#62; [&#60;array_name_rev&#62;]
split_ip() {
[ $# -eq 3 ] || # Either three
[ $# -eq 2 ] || return 1 #+ or two arguments
local -a _si_input
local IFS=${ADR_IFS}
_si_input=( $1 )
IFS=${WSP_IFS}
eval $2=\(\ \$\{_si_input\[@\]\}\ \)
if [ $# -eq 3 ]
then
# Build query order array.
local -a _dns_ip
_dns_ip[0]=${_si_input[3]}
_dns_ip[1]=${_si_input[2]}
_dns_ip[2]=${_si_input[1]}
_dns_ip[3]=${_si_input[0]}
eval $3=\(\ \$\{_dns_ip\[@\]\}\ \)
fi
return 0
}
# This function described in dot_array.bash.
# dot_array &#60;array_name&#62;
dot_array() {
[ $# -eq 1 ] || return 1 # Single argument required.
local -a _da_input
eval _da_input=\(\ \$\{$1\[@\]\}\ \)
local IFS=${DOT_IFS}
local _da_output=${_da_input[@]}
IFS=${WSP_IFS}
echo ${_da_output}
return 0
}
# This function described in file_to_array.bash
# file_to_array &#60;file_name&#62; &#60;line_array_name&#62;
file_to_array() {
[ $# -eq 2 ] || return 1 # Two arguments required.
local IFS=${NO_WSP}
local -a _fta_tmp_
_fta_tmp_=( $(cat $1) )
eval $2=\( \$\{_fta_tmp_\[@\]\} \)
return 0
}
# Columnized print of an array of multi-field strings.
# col_print &#60;array_name&#62; &#60;min_space&#62; &#60;
#+ tab_stop [tab_stops]&#62;
col_print() {
[ $# -gt 2 ] || return 0
local -a _cp_inp
local -a _cp_spc
local -a _cp_line
local _cp_min
local _cp_mcnt
local _cp_pos
local _cp_cnt
local _cp_tab
local -i _cp
local -i _cpf
local _cp_fld
# WARNING: FOLLOWING LINE NOT BLANK -- IT IS QUOTED SPACES.
local _cp_max=' '
set -f
local IFS=${NO_WSP}
eval _cp_inp=\(\ \$\{$1\[@\]\}\ \)
[ ${#_cp_inp[@]} -gt 0 ] || return 0 # Empty is easy.
_cp_mcnt=$2
_cp_min=${_cp_max:1:${_cp_mcnt}}
shift
shift
_cp_cnt=$#
for (( _cp = 0 ; _cp &#60; _cp_cnt ; _cp++ ))
do
_cp_spc[${#_cp_spc[@]}]="${_cp_max:2:$1}" #"
shift
done
_cp_cnt=${#_cp_inp[@]}
for (( _cp = 0 ; _cp &#60; _cp_cnt ; _cp++ ))
do
_cp_pos=1
IFS=${NO_WSP}$'\x20'
_cp_line=( ${_cp_inp[${_cp}]} )
IFS=${NO_WSP}
for (( _cpf = 0 ; _cpf &#60; ${#_cp_line[@]} ; _cpf++ ))
do
_cp_tab=${_cp_spc[${_cpf}]:${_cp_pos}}
if [ ${#_cp_tab} -lt ${_cp_mcnt} ]
then
_cp_tab="${_cp_min}"
fi
echo -n "${_cp_tab}"
(( _cp_pos = ${_cp_pos} + ${#_cp_tab} ))
_cp_fld="${_cp_line[${_cpf}]}"
echo -n ${_cp_fld}
(( _cp_pos = ${_cp_pos} + ${#_cp_fld} ))
done
echo
done
set +f
return 0
}
# # # # 'Hunt the Spammer' data flow # # # #
# Application return code
declare -i _hs_RC
# Original input, from which IP addresses are removed
# After which, domain names to check
declare -a uc_name
# Original input IP addresses are moved here
# After which, IP addresses to check
declare -a uc_address
# Names against which address expansion run
# Ready for name detail lookup
declare -a chk_name
# Addresses against which name expansion run
# Ready for address detail lookup
declare -a chk_address
# Recursion is depth-first-by-name.
# The expand_input_address maintains this list
#+ to prohibit looking up addresses twice during
#+ domain name recursion.
declare -a been_there_addr
been_there_addr=( '127.0.0.1' ) # Whitelist localhost
# Names which we have checked (or given up on)
declare -a known_name
# Addresses which we have checked (or given up on)
declare -a known_address
# List of zero or more Blacklist servers to check.
# Each 'known_address' will be checked against each server,
#+ with negative replies and failures suppressed.
declare -a list_server
# Indirection limit - set to zero == no limit
indirect=${SPAMMER_LIMIT:=2}
# # # # 'Hunt the Spammer' information output data # # # #
# Any domain name may have multiple IP addresses.
# Any IP address may have multiple domain names.
# Therefore, track unique address-name pairs.
declare -a known_pair
declare -a reverse_pair
# In addition to the data flow variables; known_address
#+ known_name and list_server, the following are output to the
#+ external graphics interface file.
# Authority chain, parent -&#62; SOA fields.
declare -a auth_chain
# Reference chain, parent name -&#62; child name
declare -a ref_chain
# DNS chain - domain name -&#62; address
declare -a name_address
# Name and service pairs - domain name -&#62; service
declare -a name_srvc
# Name and resource pairs - domain name -&#62; Resource Record
declare -a name_resource
# Parent and Child pairs - parent name -&#62; child name
# This MAY NOT be the same as the ref_chain followed!
declare -a parent_child
# Address and Blacklist hit pairs - address-&#62;server
declare -a address_hits
# Dump interface file data
declare -f _dot_dump
_dot_dump=pend_dummy # Initially a no-op
# Data dump is enabled by setting the environment variable SPAMMER_DATA
#+ to the name of a writable file.
declare _dot_file
# Helper function for the dump-to-dot-file function
# dump_to_dot &#60;array_name&#62; &#60;prefix&#62;
dump_to_dot() {
local -a _dda_tmp
local -i _dda_cnt
local _dda_form=' '${2}'%04u %s\n'
local IFS=${NO_WSP}
eval _dda_tmp=\(\ \$\{$1\[@\]\}\ \)
_dda_cnt=${#_dda_tmp[@]}
if [ ${_dda_cnt} -gt 0 ]
then
for (( _dda = 0 ; _dda &#60; _dda_cnt ; _dda++ ))
do
printf "${_dda_form}" \
"${_dda}" "${_dda_tmp[${_dda}]}" &#62;&#62;${_dot_file}
done
fi
}
# Which will also set _dot_dump to this function . . .
dump_dot() {
local -i _dd_cnt
echo '# Data vintage: '$(date -R) &#62;${_dot_file}
echo '# ABS Guide: is_spammer.bash; v2, 2004-msz' &#62;&#62;${_dot_file}
echo &#62;&#62;${_dot_file}
echo 'digraph G {' &#62;&#62;${_dot_file}
if [ ${#known_name[@]} -gt 0 ]
then
echo &#62;&#62;${_dot_file}
echo '# Known domain name nodes' &#62;&#62;${_dot_file}
_dd_cnt=${#known_name[@]}
for (( _dd = 0 ; _dd &#60; _dd_cnt ; _dd++ ))
do
printf ' N%04u [label="%s"] ;\n' \
"${_dd}" "${known_name[${_dd}]}" &#62;&#62;${_dot_file}
done
fi
if [ ${#known_address[@]} -gt 0 ]
then
echo &#62;&#62;${_dot_file}
echo '# Known address nodes' &#62;&#62;${_dot_file}
_dd_cnt=${#known_address[@]}
for (( _dd = 0 ; _dd &#60; _dd_cnt ; _dd++ ))
do
printf ' A%04u [label="%s"] ;\n' \
"${_dd}" "${known_address[${_dd}]}" &#62;&#62;${_dot_file}
done
fi
echo &#62;&#62;${_dot_file}
echo '/*' &#62;&#62;${_dot_file}
echo ' * Known relationships :: User conversion to' &#62;&#62;${_dot_file}
echo ' * graphic form by hand or program required.' &#62;&#62;${_dot_file}
echo ' *' &#62;&#62;${_dot_file}
if [ ${#auth_chain[@]} -gt 0 ]
then
echo &#62;&#62;${_dot_file}
echo '# Authority ref. edges followed &#38; field source.' &#62;&#62;${_dot_file}
dump_to_dot auth_chain AC
fi
if [ ${#ref_chain[@]} -gt 0 ]
then
echo &#62;&#62;${_dot_file}
echo '# Name ref. edges followed and field source.' &#62;&#62;${_dot_file}
dump_to_dot ref_chain RC
fi
if [ ${#name_address[@]} -gt 0 ]
then
echo &#62;&#62;${_dot_file}
echo '# Known name-&#62;address edges' &#62;&#62;${_dot_file}
dump_to_dot name_address NA
fi
if [ ${#name_srvc[@]} -gt 0 ]
then
echo &#62;&#62;${_dot_file}
echo '# Known name-&#62;service edges' &#62;&#62;${_dot_file}
dump_to_dot name_srvc NS
fi
if [ ${#name_resource[@]} -gt 0 ]
then
echo &#62;&#62;${_dot_file}
echo '# Known name-&#62;resource edges' &#62;&#62;${_dot_file}
dump_to_dot name_resource NR
fi
if [ ${#parent_child[@]} -gt 0 ]
then
echo &#62;&#62;${_dot_file}
echo '# Known parent-&#62;child edges' &#62;&#62;${_dot_file}
dump_to_dot parent_child PC
fi
if [ ${#list_server[@]} -gt 0 ]
then
echo &#62;&#62;${_dot_file}
echo '# Known Blacklist nodes' &#62;&#62;${_dot_file}
_dd_cnt=${#list_server[@]}
for (( _dd = 0 ; _dd &#60; _dd_cnt ; _dd++ ))
do
printf ' LS%04u [label="%s"] ;\n' \
"${_dd}" "${list_server[${_dd}]}" &#62;&#62;${_dot_file}
done
fi
unique_lines address_hits address_hits
if [ ${#address_hits[@]} -gt 0 ]
then
echo &#62;&#62;${_dot_file}
echo '# Known address-&#62;Blacklist_hit edges' &#62;&#62;${_dot_file}
echo '# CAUTION: dig warnings can trigger false hits.' &#62;&#62;${_dot_file}
dump_to_dot address_hits AH
fi
echo &#62;&#62;${_dot_file}
echo ' *' &#62;&#62;${_dot_file}
echo ' * That is a lot of relationships. Happy graphing.' &#62;&#62;${_dot_file}
echo ' */' &#62;&#62;${_dot_file}
echo '}' &#62;&#62;${_dot_file}
return 0
}
# # # # 'Hunt the Spammer' execution flow # # # #
# Execution trace is enabled by setting the
#+ environment variable SPAMMER_TRACE to the name of a writable file.
declare -a _trace_log
declare _log_file
# Function to fill the trace log
trace_logger() {
_trace_log[${#_trace_log[@]}]=${_pend_current_}
}
# Dump trace log to file function variable.
declare -f _log_dump
_log_dump=pend_dummy # Initially a no-op.
# Dump the trace log to a file.
dump_log() {
local -i _dl_cnt
_dl_cnt=${#_trace_log[@]}
for (( _dl = 0 ; _dl &#60; _dl_cnt ; _dl++ ))
do
echo ${_trace_log[${_dl}]} &#62;&#62; ${_log_file}
done
_dl_cnt=${#_pending_[@]}
if [ ${_dl_cnt} -gt 0 ]
then
_dl_cnt=${_dl_cnt}-1
echo '# # # Operations stack not empty # # #' &#62;&#62; ${_log_file}
for (( _dl = ${_dl_cnt} ; _dl &#62;= 0 ; _dl-- ))
do
echo ${_pending_[${_dl}]} &#62;&#62; ${_log_file}
done
fi
}
# # # Utility program 'dig' wrappers # # #
#
# These wrappers are derived from the
#+ examples shown in dig_wrappers.bash.
#
# The major difference is these return
#+ their results as a list in an array.
#
# See dig_wrappers.bash for details and
#+ use that script to develop any changes.
#
# # #
# Short form answer: 'dig' parses answer.
# Forward lookup :: Name -&#62; Address
# short_fwd &#60;domain_name&#62; &#60;array_name&#62;
short_fwd() {
local -a _sf_reply
local -i _sf_rc
local -i _sf_cnt
IFS=${NO_WSP}
echo -n '.'
# echo 'sfwd: '${1}
_sf_reply=( $(dig +short ${1} -c in -t a 2&#62;/dev/null) )
_sf_rc=$?
if [ ${_sf_rc} -ne 0 ]
then
_trace_log[${#_trace_log[@]}]='## Lookup error '${_sf_rc}' on '${1}' ##'
# [ ${_sf_rc} -ne 9 ] &#38;&#38; pend_drop
return ${_sf_rc}
else
# Some versions of 'dig' return warnings on stdout.
_sf_cnt=${#_sf_reply[@]}
for (( _sf = 0 ; _sf &#60; ${_sf_cnt} ; _sf++ ))
do
[ 'x'${_sf_reply[${_sf}]:0:2} == 'x;;' ] &#38;&#38;
unset _sf_reply[${_sf}]
done
eval $2=\( \$\{_sf_reply\[@\]\} \)
fi
return 0
}
# Reverse lookup :: Address -&#62; Name
# short_rev &#60;ip_address&#62; &#60;array_name&#62;
short_rev() {
local -a _sr_reply
local -i _sr_rc
local -i _sr_cnt
IFS=${NO_WSP}
echo -n '.'
# echo 'srev: '${1}
_sr_reply=( $(dig +short -x ${1} 2&#62;/dev/null) )
_sr_rc=$?
if [ ${_sr_rc} -ne 0 ]
then
_trace_log[${#_trace_log[@]}]='## Lookup error '${_sr_rc}' on '${1}' ##'
# [ ${_sr_rc} -ne 9 ] &#38;&#38; pend_drop
return ${_sr_rc}
else
# Some versions of 'dig' return warnings on stdout.
_sr_cnt=${#_sr_reply[@]}
for (( _sr = 0 ; _sr &#60; ${_sr_cnt} ; _sr++ ))
do
[ 'x'${_sr_reply[${_sr}]:0:2} == 'x;;' ] &#38;&#38;
unset _sr_reply[${_sr}]
done
eval $2=\( \$\{_sr_reply\[@\]\} \)
fi
return 0
}
# Special format lookup used to query blacklist servers.
# short_text &#60;ip_address&#62; &#60;array_name&#62;
short_text() {
local -a _st_reply
local -i _st_rc
local -i _st_cnt
IFS=${NO_WSP}
# echo 'stxt: '${1}
_st_reply=( $(dig +short ${1} -c in -t txt 2&#62;/dev/null) )
_st_rc=$?
if [ ${_st_rc} -ne 0 ]
then
_trace_log[${#_trace_log[@]}]='##Text lookup error '${_st_rc}' on '${1}'##'
# [ ${_st_rc} -ne 9 ] &#38;&#38; pend_drop
return ${_st_rc}
else
# Some versions of 'dig' return warnings on stdout.
_st_cnt=${#_st_reply[@]}
for (( _st = 0 ; _st &#60; ${#_st_cnt} ; _st++ ))
do
[ 'x'${_st_reply[${_st}]:0:2} == 'x;;' ] &#38;&#38;
unset _st_reply[${_st}]
done
eval $2=\( \$\{_st_reply\[@\]\} \)
fi
return 0
}
# The long forms, a.k.a., the parse it yourself versions
# RFC 2782 Service lookups
# dig +noall +nofail +answer _ldap._tcp.openldap.org -t srv
# _&#60;service&#62;._&#60;protocol&#62;.&#60;domain_name&#62;
# _ldap._tcp.openldap.org. 3600 IN SRV 0 0 389 ldap.openldap.org.
# domain TTL Class SRV Priority Weight Port Target
# Forward lookup :: Name -&#62; poor man's zone transfer
# long_fwd &#60;domain_name&#62; &#60;array_name&#62;
long_fwd() {
local -a _lf_reply
local -i _lf_rc
local -i _lf_cnt
IFS=${NO_WSP}
echo -n ':'
# echo 'lfwd: '${1}
_lf_reply=( $(
dig +noall +nofail +answer +authority +additional \
${1} -t soa ${1} -t mx ${1} -t any 2&#62;/dev/null) )
_lf_rc=$?
if [ ${_lf_rc} -ne 0 ]
then
_trace_log[${#_trace_log[@]}]='# Zone lookup err '${_lf_rc}' on '${1}' #'
# [ ${_lf_rc} -ne 9 ] &#38;&#38; pend_drop
return ${_lf_rc}
else
# Some versions of 'dig' return warnings on stdout.
_lf_cnt=${#_lf_reply[@]}
for (( _lf = 0 ; _lf &#60; ${_lf_cnt} ; _lf++ ))
do
[ 'x'${_lf_reply[${_lf}]:0:2} == 'x;;' ] &#38;&#38;
unset _lf_reply[${_lf}]
done
eval $2=\( \$\{_lf_reply\[@\]\} \)
fi
return 0
}
# The reverse lookup domain name corresponding to the IPv6 address:
# 4321:0:1:2:3:4:567:89ab
# would be (nibble, I.E: Hexdigit) reversed:
# b.a.9.8.7.6.5.0.4.0.0.0.3.0.0.0.2.0.0.0.1.0.0.0.0.0.0.0.1.2.3.4.IP6.ARPA.
# Reverse lookup :: Address -&#62; poor man's delegation chain
# long_rev &#60;rev_ip_address&#62; &#60;array_name&#62;
long_rev() {
local -a _lr_reply
local -i _lr_rc
local -i _lr_cnt
local _lr_dns
_lr_dns=${1}'.in-addr.arpa.'
IFS=${NO_WSP}
echo -n ':'
# echo 'lrev: '${1}
_lr_reply=( $(
dig +noall +nofail +answer +authority +additional \
${_lr_dns} -t soa ${_lr_dns} -t any 2&#62;/dev/null) )
_lr_rc=$?
if [ ${_lr_rc} -ne 0 ]
then
_trace_log[${#_trace_log[@]}]='# Deleg lkp error '${_lr_rc}' on '${1}' #'
# [ ${_lr_rc} -ne 9 ] &#38;&#38; pend_drop
return ${_lr_rc}
else
# Some versions of 'dig' return warnings on stdout.
_lr_cnt=${#_lr_reply[@]}
for (( _lr = 0 ; _lr &#60; ${_lr_cnt} ; _lr++ ))
do
[ 'x'${_lr_reply[${_lr}]:0:2} == 'x;;' ] &#38;&#38;
unset _lr_reply[${_lr}]
done
eval $2=\( \$\{_lr_reply\[@\]\} \)
fi
return 0
}
# # # Application specific functions # # #
# Mung a possible name; suppresses root and TLDs.
# name_fixup &#60;string&#62;
name_fixup(){
local -a _nf_tmp
local -i _nf_end
local _nf_str
local IFS
_nf_str=$(to_lower ${1})
_nf_str=$(to_dot ${_nf_str})
_nf_end=${#_nf_str}-1
[ ${_nf_str:${_nf_end}} != '.' ] &#38;&#38;
_nf_str=${_nf_str}'.'
IFS=${ADR_IFS}
_nf_tmp=( ${_nf_str} )
IFS=${WSP_IFS}
_nf_end=${#_nf_tmp[@]}
case ${_nf_end} in
0) # No dots, only dots.
echo
return 1
;;
1) # Only a TLD.
echo
return 1
;;
2) # Maybe okay.
echo ${_nf_str}
return 0
# Needs a lookup table?
if [ ${#_nf_tmp[1]} -eq 2 ]
then # Country coded TLD.
echo
return 1
else
echo ${_nf_str}
return 0
fi
;;
esac
echo ${_nf_str}
return 0
}
# Grope and mung original input(s).
split_input() {
[ ${#uc_name[@]} -gt 0 ] || return 0
local -i _si_cnt
local -i _si_len
local _si_str
unique_lines uc_name uc_name
_si_cnt=${#uc_name[@]}
for (( _si = 0 ; _si &#60; _si_cnt ; _si++ ))
do
_si_str=${uc_name[$_si]}
if is_address ${_si_str}
then
uc_address[${#uc_address[@]}]=${_si_str}
unset uc_name[$_si]
else
if ! uc_name[$_si]=$(name_fixup ${_si_str})
then
unset ucname[$_si]
fi
fi
done
uc_name=( ${uc_name[@]} )
_si_cnt=${#uc_name[@]}
_trace_log[${#_trace_log[@]}]='#Input '${_si_cnt}' unchkd name input(s).#'
_si_cnt=${#uc_address[@]}
_trace_log[${#_trace_log[@]}]='#Input '${_si_cnt}' unchkd addr input(s).#'
return 0
}
# # # Discovery functions -- recursively interlocked by external data # # #
# # # The leading 'if list is empty; return 0' in each is required. # # #
# Recursion limiter
# limit_chk() &#60;next_level&#62;
limit_chk() {
local -i _lc_lmt
# Check indirection limit.
if [ ${indirect} -eq 0 ] || [ $# -eq 0 ]
then
# The 'do-forever' choice
echo 1 # Any value will do.
return 0 # OK to continue.
else
# Limiting is in effect.
if [ ${indirect} -lt ${1} ]
then
echo ${1} # Whatever.
return 1 # Stop here.
else
_lc_lmt=${1}+1 # Bump the given limit.
echo ${_lc_lmt} # Echo it.
return 0 # OK to continue.
fi
fi
}
# For each name in uc_name:
# Move name to chk_name.
# Add addresses to uc_address.
# Pend expand_input_address.
# Repeat until nothing new found.
# expand_input_name &#60;indirection_limit&#62;
expand_input_name() {
[ ${#uc_name[@]} -gt 0 ] || return 0
local -a _ein_addr
local -a _ein_new
local -i _ucn_cnt
local -i _ein_cnt
local _ein_tst
_ucn_cnt=${#uc_name[@]}
if ! _ein_cnt=$(limit_chk ${1})
then
return 0
fi
for (( _ein = 0 ; _ein &#60; _ucn_cnt ; _ein++ ))
do
if short_fwd ${uc_name[${_ein}]} _ein_new
then
for (( _ein_cnt = 0 ; _ein_cnt &#60; ${#_ein_new[@]}; _ein_cnt++ ))
do
_ein_tst=${_ein_new[${_ein_cnt}]}
if is_address ${_ein_tst}
then
_ein_addr[${#_ein_addr[@]}]=${_ein_tst}
fi
done
fi
done
unique_lines _ein_addr _ein_addr # Scrub duplicates.
edit_exact chk_address _ein_addr # Scrub pending detail.
edit_exact known_address _ein_addr # Scrub already detailed.
if [ ${#_ein_addr[@]} -gt 0 ] # Anything new?
then
uc_address=( ${uc_address[@]} ${_ein_addr[@]} )
pend_func expand_input_address ${1}
_trace_log[${#_trace_log[@]}]='#Add '${#_ein_addr[@]}' unchkd addr inp.#'
fi
edit_exact chk_name uc_name # Scrub pending detail.
edit_exact known_name uc_name # Scrub already detailed.
if [ ${#uc_name[@]} -gt 0 ]
then
chk_name=( ${chk_name[@]} ${uc_name[@]} )
pend_func detail_each_name ${1}
fi
unset uc_name[@]
return 0
}
# For each address in uc_address:
# Move address to chk_address.
# Add names to uc_name.
# Pend expand_input_name.
# Repeat until nothing new found.
# expand_input_address &#60;indirection_limit&#62;
expand_input_address() {
[ ${#uc_address[@]} -gt 0 ] || return 0
local -a _eia_addr
local -a _eia_name
local -a _eia_new
local -i _uca_cnt
local -i _eia_cnt
local _eia_tst
unique_lines uc_address _eia_addr
unset uc_address[@]
edit_exact been_there_addr _eia_addr
_uca_cnt=${#_eia_addr[@]}
[ ${_uca_cnt} -gt 0 ] &#38;&#38;
been_there_addr=( ${been_there_addr[@]} ${_eia_addr[@]} )
for (( _eia = 0 ; _eia &#60; _uca_cnt ; _eia++ ))
do
if short_rev ${_eia_addr[${_eia}]} _eia_new
then
for (( _eia_cnt = 0 ; _eia_cnt &#60; ${#_eia_new[@]} ; _eia_cnt++ ))
do
_eia_tst=${_eia_new[${_eia_cnt}]}
if _eia_tst=$(name_fixup ${_eia_tst})
then
_eia_name[${#_eia_name[@]}]=${_eia_tst}
fi
done
fi
done
unique_lines _eia_name _eia_name # Scrub duplicates.
edit_exact chk_name _eia_name # Scrub pending detail.
edit_exact known_name _eia_name # Scrub already detailed.
if [ ${#_eia_name[@]} -gt 0 ] # Anything new?
then
uc_name=( ${uc_name[@]} ${_eia_name[@]} )
pend_func expand_input_name ${1}
_trace_log[${#_trace_log[@]}]='#Add '${#_eia_name[@]}' unchkd name inp.#'
fi
edit_exact chk_address _eia_addr # Scrub pending detail.
edit_exact known_address _eia_addr # Scrub already detailed.
if [ ${#_eia_addr[@]} -gt 0 ] # Anything new?
then
chk_address=( ${chk_address[@]} ${_eia_addr[@]} )
pend_func detail_each_address ${1}
fi
return 0
}
# The parse-it-yourself zone reply.
# The input is the chk_name list.
# detail_each_name &#60;indirection_limit&#62;
detail_each_name() {
[ ${#chk_name[@]} -gt 0 ] || return 0
local -a _den_chk # Names to check
local -a _den_name # Names found here
local -a _den_address # Addresses found here
local -a _den_pair # Pairs found here
local -a _den_rev # Reverse pairs found here
local -a _den_tmp # Line being parsed
local -a _den_auth # SOA contact being parsed
local -a _den_new # The zone reply
local -a _den_pc # Parent-Child gets big fast
local -a _den_ref # So does reference chain
local -a _den_nr # Name-Resource can be big
local -a _den_na # Name-Address
local -a _den_ns # Name-Service
local -a _den_achn # Chain of Authority
local -i _den_cnt # Count of names to detail
local -i _den_lmt # Indirection limit
local _den_who # Named being processed
local _den_rec # Record type being processed
local _den_cont # Contact domain
local _den_str # Fixed up name string
local _den_str2 # Fixed up reverse
local IFS=${WSP_IFS}
# Local, unique copy of names to check
unique_lines chk_name _den_chk
unset chk_name[@] # Done with globals.
# Less any names already known
edit_exact known_name _den_chk
_den_cnt=${#_den_chk[@]}
# If anything left, add to known_name.
[ ${_den_cnt} -gt 0 ] &#38;&#38;
known_name=( ${known_name[@]} ${_den_chk[@]} )
# for the list of (previously) unknown names . . .
for (( _den = 0 ; _den &#60; _den_cnt ; _den++ ))
do
_den_who=${_den_chk[${_den}]}
if long_fwd ${_den_who} _den_new
then
unique_lines _den_new _den_new
if [ ${#_den_new[@]} -eq 0 ]
then
_den_pair[${#_den_pair[@]}]='0.0.0.0 '${_den_who}
fi
# Parse each line in the reply.
for (( _line = 0 ; _line &#60; ${#_den_new[@]} ; _line++ ))
do
IFS=${NO_WSP}$'\x09'$'\x20'
_den_tmp=( ${_den_new[${_line}]} )
IFS=${WSP_IFS}
# If usable record and not a warning message . . .
if [ ${#_den_tmp[@]} -gt 4 ] &#38;&#38; [ 'x'${_den_tmp[0]} != 'x;;' ]
then
_den_rec=${_den_tmp[3]}
_den_nr[${#_den_nr[@]}]=${_den_who}' '${_den_rec}
# Begin at RFC1033 (+++)
case ${_den_rec} in
#&#60;name&#62; [&#60;ttl&#62;] [&#60;class&#62;] SOA &#60;origin&#62; &#60;person&#62;
SOA) # Start Of Authority
if _den_str=$(name_fixup ${_den_tmp[0]})
then
_den_name[${#_den_name[@]}]=${_den_str}
_den_achn[${#_den_achn[@]}]=${_den_who}' '${_den_str}' SOA'
# SOA origin -- domain name of master zone record
if _den_str2=$(name_fixup ${_den_tmp[4]})
then
_den_name[${#_den_name[@]}]=${_den_str2}
_den_achn[${#_den_achn[@]}]=${_den_who}' '${_den_str2}' SOA.O'
fi
# Responsible party e-mail address (possibly bogus).
# Possibility of first.last@domain.name ignored.
set -f
if _den_str2=$(name_fixup ${_den_tmp[5]})
then
IFS=${ADR_IFS}
_den_auth=( ${_den_str2} )
IFS=${WSP_IFS}
if [ ${#_den_auth[@]} -gt 2 ]
then
_den_cont=${_den_auth[1]}
for (( _auth = 2 ; _auth &#60; ${#_den_auth[@]} ; _auth++ ))
do
_den_cont=${_den_cont}'.'${_den_auth[${_auth}]}
done
_den_name[${#_den_name[@]}]=${_den_cont}'.'
_den_achn[${#_den_achn[@]}]=${_den_who}' '${_den_cont}'. SOA.C'
fi
fi
set +f
fi
;;
A) # IP(v4) Address Record
if _den_str=$(name_fixup ${_den_tmp[0]})
then
_den_name[${#_den_name[@]}]=${_den_str}
_den_pair[${#_den_pair[@]}]=${_den_tmp[4]}' '${_den_str}
_den_na[${#_den_na[@]}]=${_den_str}' '${_den_tmp[4]}
_den_ref[${#_den_ref[@]}]=${_den_who}' '${_den_str}' A'
else
_den_pair[${#_den_pair[@]}]=${_den_tmp[4]}' unknown.domain'
_den_na[${#_den_na[@]}]='unknown.domain '${_den_tmp[4]}
_den_ref[${#_den_ref[@]}]=${_den_who}' unknown.domain A'
fi
_den_address[${#_den_address[@]}]=${_den_tmp[4]}
_den_pc[${#_den_pc[@]}]=${_den_who}' '${_den_tmp[4]}
;;
NS) # Name Server Record
# Domain name being serviced (may be other than current)
if _den_str=$(name_fixup ${_den_tmp[0]})
then
_den_name[${#_den_name[@]}]=${_den_str}
_den_ref[${#_den_ref[@]}]=${_den_who}' '${_den_str}' NS'
# Domain name of service provider
if _den_str2=$(name_fixup ${_den_tmp[4]})
then
_den_name[${#_den_name[@]}]=${_den_str2}
_den_ref[${#_den_ref[@]}]=${_den_who}' '${_den_str2}' NSH'
_den_ns[${#_den_ns[@]}]=${_den_str2}' NS'
_den_pc[${#_den_pc[@]}]=${_den_str}' '${_den_str2}
fi
fi
;;
MX) # Mail Server Record
# Domain name being serviced (wildcards not handled here)
if _den_str=$(name_fixup ${_den_tmp[0]})
then
_den_name[${#_den_name[@]}]=${_den_str}
_den_ref[${#_den_ref[@]}]=${_den_who}' '${_den_str}' MX'
fi
# Domain name of service provider
if _den_str=$(name_fixup ${_den_tmp[5]})
then
_den_name[${#_den_name[@]}]=${_den_str}
_den_ref[${#_den_ref[@]}]=${_den_who}' '${_den_str}' MXH'
_den_ns[${#_den_ns[@]}]=${_den_str}' MX'
_den_pc[${#_den_pc[@]}]=${_den_who}' '${_den_str}
fi
;;
PTR) # Reverse address record
# Special name
if _den_str=$(name_fixup ${_den_tmp[0]})
then
_den_ref[${#_den_ref[@]}]=${_den_who}' '${_den_str}' PTR'
# Host name (not a CNAME)
if _den_str2=$(name_fixup ${_den_tmp[4]})
then
_den_rev[${#_den_rev[@]}]=${_den_str}' '${_den_str2}
_den_ref[${#_den_ref[@]}]=${_den_who}' '${_den_str2}' PTRH'
_den_pc[${#_den_pc[@]}]=${_den_who}' '${_den_str}
fi
fi
;;
AAAA) # IP(v6) Address Record
if _den_str=$(name_fixup ${_den_tmp[0]})
then
_den_name[${#_den_name[@]}]=${_den_str}
_den_pair[${#_den_pair[@]}]=${_den_tmp[4]}' '${_den_str}
_den_na[${#_den_na[@]}]=${_den_str}' '${_den_tmp[4]}
_den_ref[${#_den_ref[@]}]=${_den_who}' '${_den_str}' AAAA'
else
_den_pair[${#_den_pair[@]}]=${_den_tmp[4]}' unknown.domain'
_den_na[${#_den_na[@]}]='unknown.domain '${_den_tmp[4]}
_den_ref[${#_den_ref[@]}]=${_den_who}' unknown.domain'
fi
# No processing for IPv6 addresses
_den_pc[${#_den_pc[@]}]=${_den_who}' '${_den_tmp[4]}
;;
CNAME) # Alias name record
# Nickname
if _den_str=$(name_fixup ${_den_tmp[0]})
then
_den_name[${#_den_name[@]}]=${_den_str}
_den_ref[${#_den_ref[@]}]=${_den_who}' '${_den_str}' CNAME'
_den_pc[${#_den_pc[@]}]=${_den_who}' '${_den_str}
fi
# Hostname
if _den_str=$(name_fixup ${_den_tmp[4]})
then
_den_name[${#_den_name[@]}]=${_den_str}
_den_ref[${#_den_ref[@]}]=${_den_who}' '${_den_str}' CHOST'
_den_pc[${#_den_pc[@]}]=${_den_who}' '${_den_str}
fi
;;
# TXT)
# ;;
esac
fi
done
else # Lookup error == 'A' record 'unknown address'
_den_pair[${#_den_pair[@]}]='0.0.0.0 '${_den_who}
fi
done
# Control dot array growth.
unique_lines _den_achn _den_achn # Works best, all the same.
edit_exact auth_chain _den_achn # Works best, unique items.
if [ ${#_den_achn[@]} -gt 0 ]
then
IFS=${NO_WSP}
auth_chain=( ${auth_chain[@]} ${_den_achn[@]} )
IFS=${WSP_IFS}
fi
unique_lines _den_ref _den_ref # Works best, all the same.
edit_exact ref_chain _den_ref # Works best, unique items.
if [ ${#_den_ref[@]} -gt 0 ]
then
IFS=${NO_WSP}
ref_chain=( ${ref_chain[@]} ${_den_ref[@]} )
IFS=${WSP_IFS}
fi
unique_lines _den_na _den_na
edit_exact name_address _den_na
if [ ${#_den_na[@]} -gt 0 ]
then
IFS=${NO_WSP}
name_address=( ${name_address[@]} ${_den_na[@]} )
IFS=${WSP_IFS}
fi
unique_lines _den_ns _den_ns
edit_exact name_srvc _den_ns
if [ ${#_den_ns[@]} -gt 0 ]
then
IFS=${NO_WSP}
name_srvc=( ${name_srvc[@]} ${_den_ns[@]} )
IFS=${WSP_IFS}
fi
unique_lines _den_nr _den_nr
edit_exact name_resource _den_nr
if [ ${#_den_nr[@]} -gt 0 ]
then
IFS=${NO_WSP}
name_resource=( ${name_resource[@]} ${_den_nr[@]} )
IFS=${WSP_IFS}
fi
unique_lines _den_pc _den_pc
edit_exact parent_child _den_pc
if [ ${#_den_pc[@]} -gt 0 ]
then
IFS=${NO_WSP}
parent_child=( ${parent_child[@]} ${_den_pc[@]} )
IFS=${WSP_IFS}
fi
# Update list known_pair (Address and Name).
unique_lines _den_pair _den_pair
edit_exact known_pair _den_pair
if [ ${#_den_pair[@]} -gt 0 ] # Anything new?
then
IFS=${NO_WSP}
known_pair=( ${known_pair[@]} ${_den_pair[@]} )
IFS=${WSP_IFS}
fi
# Update list of reverse pairs.
unique_lines _den_rev _den_rev
edit_exact reverse_pair _den_rev
if [ ${#_den_rev[@]} -gt 0 ] # Anything new?
then
IFS=${NO_WSP}
reverse_pair=( ${reverse_pair[@]} ${_den_rev[@]} )
IFS=${WSP_IFS}
fi
# Check indirection limit -- give up if reached.
if ! _den_lmt=$(limit_chk ${1})
then
return 0
fi
# Execution engine is LIFO. Order of pend operations is important.
# Did we define any new addresses?
unique_lines _den_address _den_address # Scrub duplicates.
edit_exact known_address _den_address # Scrub already processed.
edit_exact un_address _den_address # Scrub already waiting.
if [ ${#_den_address[@]} -gt 0 ] # Anything new?
then
uc_address=( ${uc_address[@]} ${_den_address[@]} )
pend_func expand_input_address ${_den_lmt}
_trace_log[${#_trace_log[@]}]='# Add '${#_den_address[@]}' unchkd addr. #'
fi
# Did we find any new names?
unique_lines _den_name _den_name # Scrub duplicates.
edit_exact known_name _den_name # Scrub already processed.
edit_exact uc_name _den_name # Scrub already waiting.
if [ ${#_den_name[@]} -gt 0 ] # Anything new?
then
uc_name=( ${uc_name[@]} ${_den_name[@]} )
pend_func expand_input_name ${_den_lmt}
_trace_log[${#_trace_log[@]}]='#Added '${#_den_name[@]}' unchkd name#'
fi
return 0
}
# The parse-it-yourself delegation reply
# Input is the chk_address list.
# detail_each_address &#60;indirection_limit&#62;
detail_each_address() {
[ ${#chk_address[@]} -gt 0 ] || return 0
unique_lines chk_address chk_address
edit_exact known_address chk_address
if [ ${#chk_address[@]} -gt 0 ]
then
known_address=( ${known_address[@]} ${chk_address[@]} )
unset chk_address[@]
fi
return 0
}
# # # Application specific output functions # # #
# Pretty print the known pairs.
report_pairs() {
echo
echo 'Known network pairs.'
col_print known_pair 2 5 30
if [ ${#auth_chain[@]} -gt 0 ]
then
echo
echo 'Known chain of authority.'
col_print auth_chain 2 5 30 55
fi
if [ ${#reverse_pair[@]} -gt 0 ]
then
echo
echo 'Known reverse pairs.'
col_print reverse_pair 2 5 55
fi
return 0
}
# Check an address against the list of blacklist servers.
# A good place to capture for GraphViz: address-&#62;status(server(reports))
# check_lists &#60;ip_address&#62;
check_lists() {
[ $# -eq 1 ] || return 1
local -a _cl_fwd_addr
local -a _cl_rev_addr
local -a _cl_reply
local -i _cl_rc
local -i _ls_cnt
local _cl_dns_addr
local _cl_lkup
split_ip ${1} _cl_fwd_addr _cl_rev_addr
_cl_dns_addr=$(dot_array _cl_rev_addr)'.'
_ls_cnt=${#list_server[@]}
echo ' Checking address '${1}
for (( _cl = 0 ; _cl &#60; _ls_cnt ; _cl++ ))
do
_cl_lkup=${_cl_dns_addr}${list_server[${_cl}]}
if short_text ${_cl_lkup} _cl_reply
then
if [ ${#_cl_reply[@]} -gt 0 ]
then
echo ' Records from '${list_server[${_cl}]}
address_hits[${#address_hits[@]}]=${1}' '${list_server[${_cl}]}
_hs_RC=2
for (( _clr = 0 ; _clr &#60; ${#_cl_reply[@]} ; _clr++ ))
do
echo ' '${_cl_reply[${_clr}]}
done
fi
fi
done
return 0
}
# # # The usual application glue # # #
# Who did it?
credits() {
echo
echo 'Advanced Bash Scripting Guide: is_spammer.bash, v2, 2004-msz'
}
# How to use it?
# (See also, "Quickstart" at end of script.)
usage() {
cat &#60;&#60;-'_usage_statement_'
The script is_spammer.bash requires either one or two arguments.
arg 1) May be one of:
a) A domain name
b) An IPv4 address
c) The name of a file with any mix of names
and addresses, one per line.
arg 2) May be one of:
a) A Blacklist server domain name
b) The name of a file with Blacklist server
domain names, one per line.
c) If not present, a default list of (free)
Blacklist servers is used.
d) If a filename of an empty, readable, file
is given,
Blacklist server lookup is disabled.
All script output is written to stdout.
Return codes: 0 -&#62; All OK, 1 -&#62; Script failure,
2 -&#62; Something is Blacklisted.
Requires the external program 'dig' from the 'bind-9'
set of DNS programs. See: http://www.isc.org
The domain name lookup depth limit defaults to 2 levels.
Set the environment variable SPAMMER_LIMIT to change.
SPAMMER_LIMIT=0 means 'unlimited'
Limit may also be set on the command-line.
If arg#1 is an integer, the limit is set to that value
and then the above argument rules are applied.
Setting the environment variable 'SPAMMER_DATA' to a filename
will cause the script to write a GraphViz graphic file.
For the development version;
Setting the environment variable 'SPAMMER_TRACE' to a filename
will cause the execution engine to log a function call trace.
_usage_statement_
}
# The default list of Blacklist servers:
# Many choices, see: http://www.spews.org/lists.html
declare -a default_servers
# See: http://www.spamhaus.org (Conservative, well maintained)
default_servers[0]='sbl-xbl.spamhaus.org'
# See: http://ordb.org (Open mail relays)
default_servers[1]='relays.ordb.org'
# See: http://www.spamcop.net/ (You can report spammers here)
default_servers[2]='bl.spamcop.net'
# See: http://www.spews.org (An 'early detect' system)
default_servers[3]='l2.spews.dnsbl.sorbs.net'
# See: http://www.dnsbl.us.sorbs.net/using.shtml
default_servers[4]='dnsbl.sorbs.net'
# See: http://dsbl.org/usage (Various mail relay lists)
default_servers[5]='list.dsbl.org'
default_servers[6]='multihop.dsbl.org'
default_servers[7]='unconfirmed.dsbl.org'
# User input argument #1
setup_input() {
if [ -e ${1} ] &#38;&#38; [ -r ${1} ] # Name of readable file
then
file_to_array ${1} uc_name
echo 'Using filename &#62;'${1}'&#60; as input.'
else
if is_address ${1} # IP address?
then
uc_address=( ${1} )
echo 'Starting with address &#62;'${1}'&#60;'
else # Must be a name.
uc_name=( ${1} )
echo 'Starting with domain name &#62;'${1}'&#60;'
fi
fi
return 0
}
# User input argument #2
setup_servers() {
if [ -e ${1} ] &#38;&#38; [ -r ${1} ] # Name of a readable file
then
file_to_array ${1} list_server
echo 'Using filename &#62;'${1}'&#60; as blacklist server list.'
else
list_server=( ${1} )
echo 'Using blacklist server &#62;'${1}'&#60;'
fi
return 0
}
# User environment variable SPAMMER_TRACE
live_log_die() {
if [ ${SPAMMER_TRACE:=} ] # Wants trace log?
then
if [ ! -e ${SPAMMER_TRACE} ]
then
if ! touch ${SPAMMER_TRACE} 2&#62;/dev/null
then
pend_func echo $(printf '%q\n' \
'Unable to create log file &#62;'${SPAMMER_TRACE}'&#60;')
pend_release
exit 1
fi
_log_file=${SPAMMER_TRACE}
_pend_hook_=trace_logger
_log_dump=dump_log
else
if [ ! -w ${SPAMMER_TRACE} ]
then
pend_func echo $(printf '%q\n' \
'Unable to write log file &#62;'${SPAMMER_TRACE}'&#60;')
pend_release
exit 1
fi
_log_file=${SPAMMER_TRACE}
echo '' &#62; ${_log_file}
_pend_hook_=trace_logger
_log_dump=dump_log
fi
fi
return 0
}
# User environment variable SPAMMER_DATA
data_capture() {
if [ ${SPAMMER_DATA:=} ] # Wants a data dump?
then
if [ ! -e ${SPAMMER_DATA} ]
then
if ! touch ${SPAMMER_DATA} 2&#62;/dev/null
then
pend_func echo $(printf '%q]n' \
'Unable to create data output file &#62;'${SPAMMER_DATA}'&#60;')
pend_release
exit 1
fi
_dot_file=${SPAMMER_DATA}
_dot_dump=dump_dot
else
if [ ! -w ${SPAMMER_DATA} ]
then
pend_func echo $(printf '%q\n' \
'Unable to write data output file &#62;'${SPAMMER_DATA}'&#60;')
pend_release
exit 1
fi
_dot_file=${SPAMMER_DATA}
_dot_dump=dump_dot
fi
fi
return 0
}
# Grope user specified arguments.
do_user_args() {
if [ $# -gt 0 ] &#38;&#38; is_number $1
then
indirect=$1
shift
fi
case $# in # Did user treat us well?
1)
if ! setup_input $1 # Needs error checking.
then
pend_release
$_log_dump
exit 1
fi
list_server=( ${default_servers[@]} )
_list_cnt=${#list_server[@]}
echo 'Using default blacklist server list.'
echo 'Search depth limit: '${indirect}
;;
2)
if ! setup_input $1 # Needs error checking.
then
pend_release
$_log_dump
exit 1
fi
if ! setup_servers $2 # Needs error checking.
then
pend_release
$_log_dump
exit 1
fi
echo 'Search depth limit: '${indirect}
;;
*)
pend_func usage
pend_release
$_log_dump
exit 1
;;
esac
return 0
}
# A general purpose debug tool.
# list_array &#60;array_name&#62;
list_array() {
[ $# -eq 1 ] || return 1 # One argument required.
local -a _la_lines
set -f
local IFS=${NO_WSP}
eval _la_lines=\(\ \$\{$1\[@\]\}\ \)
echo
echo "Element count "${#_la_lines[@]}" array "${1}
local _ln_cnt=${#_la_lines[@]}
for (( _i = 0; _i &#60; ${_ln_cnt}; _i++ ))
do
echo 'Element '$_i' &#62;'${_la_lines[$_i]}'&#60;'
done
set +f
return 0
}
# # # 'Hunt the Spammer' program code # # #
pend_init # Ready stack engine.
pend_func credits # Last thing to print.
# # # Deal with user # # #
live_log_die # Setup debug trace log.
data_capture # Setup data capture file.
echo
do_user_args $@
# # # Haven't exited yet - There is some hope # # #
# Discovery group - Execution engine is LIFO - pend
# in reverse order of execution.
_hs_RC=0 # Hunt the Spammer return code
pend_mark
pend_func report_pairs # Report name-address pairs.
# The two detail_* are mutually recursive functions.
# They also pend expand_* functions as required.
# These two (the last of ???) exit the recursion.
pend_func detail_each_address # Get all resources of addresses.
pend_func detail_each_name # Get all resources of names.
# The two expand_* are mutually recursive functions,
#+ which pend additional detail_* functions as required.
pend_func expand_input_address 1 # Expand input names by address.
pend_func expand_input_name 1 # #xpand input addresses by name.
# Start with a unique set of names and addresses.
pend_func unique_lines uc_address uc_address
pend_func unique_lines uc_name uc_name
# Separate mixed input of names and addresses.
pend_func split_input
pend_release
# # # Pairs reported -- Unique list of IP addresses found
echo
_ip_cnt=${#known_address[@]}
if [ ${#list_server[@]} -eq 0 ]
then
echo 'Blacklist server list empty, none checked.'
else
if [ ${_ip_cnt} -eq 0 ]
then
echo 'Known address list empty, none checked.'
else
_ip_cnt=${_ip_cnt}-1 # Start at top.
echo 'Checking Blacklist servers.'
for (( _ip = _ip_cnt ; _ip &#62;= 0 ; _ip-- ))
do
pend_func check_lists $( printf '%q\n' ${known_address[$_ip]} )
done
fi
fi
pend_release
$_dot_dump # Graphics file dump
$_log_dump # Execution trace
echo
##############################
# Example output from script #
##############################
:&#60;&#60;-'_is_spammer_outputs_'
./is_spammer.bash 0 web4.alojamentos7.com
Starting with domain name &#62;web4.alojamentos7.com&#60;
Using default blacklist server list.
Search depth limit: 0
.:....::::...:::...:::.......::..::...:::.......::
Known network pairs.
66.98.208.97 web4.alojamentos7.com.
66.98.208.97 ns1.alojamentos7.com.
69.56.202.147 ns2.alojamentos.ws.
66.98.208.97 alojamentos7.com.
66.98.208.97 web.alojamentos7.com.
69.56.202.146 ns1.alojamentos.ws.
69.56.202.146 alojamentos.ws.
66.235.180.113 ns1.alojamentos.org.
66.235.181.192 ns2.alojamentos.org.
66.235.180.113 alojamentos.org.
66.235.180.113 web6.alojamentos.org.
216.234.234.30 ns1.theplanet.com.
12.96.160.115 ns2.theplanet.com.
216.185.111.52 mail1.theplanet.com.
69.56.141.4 spooling.theplanet.com.
216.185.111.40 theplanet.com.
216.185.111.40 www.theplanet.com.
216.185.111.52 mail.theplanet.com.
Checking Blacklist servers.
Checking address 66.98.208.97
Records from dnsbl.sorbs.net
"Spam Received See: http://www.dnsbl.sorbs.net/lookup.shtml?66.98.208.97"
Checking address 69.56.202.147
Checking address 69.56.202.146
Checking address 66.235.180.113
Checking address 66.235.181.192
Checking address 216.185.111.40
Checking address 216.234.234.30
Checking address 12.96.160.115
Checking address 216.185.111.52
Checking address 69.56.141.4
Advanced Bash Scripting Guide: is_spammer.bash, v2, 2004-msz
_is_spammer_outputs_
exit ${_hs_RC}
####################################################
# The script ignores everything from here on down #
#+ because of the 'exit' command, just above. #
####################################################
Quickstart
==========
Prerequisites
Bash version 2.05b or 3.00 (bash --version)
A version of Bash which supports arrays. Array
support is included by default Bash configurations.
'dig,' version 9.x.x (dig $HOSTNAME, see first line of output)
A version of dig which supports the +short options.
See: dig_wrappers.bash for details.
Optional Prerequisites
'named,' a local DNS caching program. Any flavor will do.
Do twice: dig $HOSTNAME
Check near bottom of output for: SERVER: 127.0.0.1#53
That means you have one running.
Optional Graphics Support
'date,' a standard *nix thing. (date -R)
dot Program to convert graphic description file to a
diagram. (dot -V)
A part of the Graph-Viz set of programs.
See: [http://www.research.att.com/sw/tools/graphviz||GraphViz]
'dotty,' a visual editor for graphic description files.
Also a part of the Graph-Viz set of programs.
Quick Start
In the same directory as the is_spammer.bash script;
Do: ./is_spammer.bash
Usage Details
1. Blacklist server choices.
(a) To use default, built-in list: Do nothing.
(b) To use your own list:
i. Create a file with a single Blacklist server
domain name per line.
ii. Provide that filename as the last argument to
the script.
(c) To use a single Blacklist server: Last argument
to the script.
(d) To disable Blacklist lookups:
i. Create an empty file (touch spammer.nul)
Your choice of filename.
ii. Provide the filename of that empty file as the
last argument to the script.
2. Search depth limit.
(a) To use the default value of 2: Do nothing.
(b) To set a different limit:
A limit of 0 means: no limit.
i. export SPAMMER_LIMIT=1
or whatever limit you want.
ii. OR provide the desired limit as the first
argument to the script.
3. Optional execution trace log.
(a) To use the default setting of no log output: Do nothing.
(b) To write an execution trace log:
export SPAMMER_TRACE=spammer.log
or whatever filename you want.
4. Optional graphic description file.
(a) To use the default setting of no graphic file: Do nothing.
(b) To write a Graph-Viz graphic description file:
export SPAMMER_DATA=spammer.dot
or whatever filename you want.
5. Where to start the search.
(a) Starting with a single domain name:
i. Without a command-line search limit: First
argument to script.
ii. With a command-line search limit: Second
argument to script.
(b) Starting with a single IP address:
i. Without a command-line search limit: First
argument to script.
ii. With a command-line search limit: Second
argument to script.
(c) Starting with (mixed) multiple name(s) and/or address(es):
Create a file with one name or address per line.
Your choice of filename.
i. Without a command-line search limit: Filename as
first argument to script.
ii. With a command-line search limit: Filename as
second argument to script.
6. What to do with the display output.
(a) To view display output on screen: Do nothing.
(b) To save display output to a file: Redirect stdout to a filename.
(c) To discard display output: Redirect stdout to /dev/null.
7. Temporary end of decision making.
press RETURN
wait (optionally, watch the dots and colons).
8. Optionally check the return code.
(a) Return code 0: All OK
(b) Return code 1: Script setup failure
(c) Return code 2: Something was blacklisted.
9. Where is my graph (diagram)?
The script does not directly produce a graph (diagram).
It only produces a graphic description file. You can
process the graphic descriptor file that was output
with the 'dot' program.
Until you edit that descriptor file, to describe the
relationships you want shown, all that you will get is
a bunch of labeled name and address nodes.
All of the script's discovered relationships are within
a comment block in the graphic descriptor file, each
with a descriptive heading.
The editing required to draw a line between a pair of
nodes from the information in the descriptor file may
be done with a text editor.
Given these lines somewhere in the descriptor file:
# Known domain name nodes
N0000 [label="guardproof.info."] ;
N0002 [label="third.guardproof.info."] ;
# Known address nodes
A0000 [label="61.141.32.197"] ;
/*
# Known name-&#62;address edges
NA0000 third.guardproof.info. 61.141.32.197
# Known parent-&#62;child edges
PC0000 guardproof.info. third.guardproof.info.
*/
Turn that into the following lines by substituting node
identifiers into the relationships:
# Known domain name nodes
N0000 [label="guardproof.info."] ;
N0002 [label="third.guardproof.info."] ;
# Known address nodes
A0000 [label="61.141.32.197"] ;
# PC0000 guardproof.info. third.guardproof.info.
N0000-&#62;N0002 ;
# NA0000 third.guardproof.info. 61.141.32.197
N0002-&#62;A0000 ;
/*
# Known name-&#62;address edges
NA0000 third.guardproof.info. 61.141.32.197
# Known parent-&#62;child edges
PC0000 guardproof.info. third.guardproof.info.
*/
Process that with the 'dot' program, and you have your
first network diagram.
In addition to the conventional graphic edges, the
descriptor file includes similar format pair-data that
describes services, zone records (sub-graphs?),
blacklisted addresses, and other things which might be
interesting to include in your graph. This additional
information could be displayed as different node
shapes, colors, line sizes, etc.
The descriptor file can also be read and edited by a
Bash script (of course). You should be able to find
most of the functions required within the
"is_spammer.bash" script.
# End Quickstart.
Additional Note
========== ====
Michael Zick points out that there is a "makeviz.bash" interactive
Web site at rediris.es. Can't give the full URL, since this is not
a publically accessible site.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Another anti-spam script.</P
><P
><A
NAME="WHX0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="WHX"
></A
><P
><B
>Example A-29. Spammer Hunt</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# whx.sh: "whois" spammer lookup
# Author: Walter Dnes
# Slight revisions (first section) by ABS Guide author.
# Used in ABS Guide with permission.
# Needs version 3.x or greater of Bash to run (because of =~ operator).
# Commented by script author and ABS Guide author.
E_BADARGS=85 # Missing command-line arg.
E_NOHOST=86 # Host not found.
E_TIMEOUT=87 # Host lookup timed out.
E_UNDEF=88 # Some other (undefined) error.
HOSTWAIT=10 # Specify up to 10 seconds for host query reply.
# The actual wait may be a bit longer.
OUTFILE=whois.txt # Output file.
PORT=4321
if [ -z "$1" ] # Check for (required) command-line arg.
then
echo "Usage: $0 domain name or IP address"
exit $E_BADARGS
fi
if [[ "$1" =~ [a-zA-Z][a-zA-Z]$ ]] # Ends in two alpha chars?
then # It's a domain name &#38;&#38;
#+ must do host lookup.
IPADDR=$(host -W $HOSTWAIT $1 | awk '{print $4}')
# Doing host lookup
#+ to get IP address.
# Extract final field.
else
IPADDR="$1" # Command-line arg was IP address.
fi
echo; echo "IP Address is: "$IPADDR""; echo
if [ -e "$OUTFILE" ]
then
rm -f "$OUTFILE"
echo "Stale output file \"$OUTFILE\" removed."; echo
fi
# Sanity checks.
# (This section needs more work.)
# ===============================
if [ -z "$IPADDR" ]
# No response.
then
echo "Host not found!"
exit $E_NOHOST # Bail out.
fi
if [[ "$IPADDR" =~ ^[;;] ]]
# ;; Connection timed out; no servers could be reached.
then
echo "Host lookup timed out!"
exit $E_TIMEOUT # Bail out.
fi
if [[ "$IPADDR" =~ [(NXDOMAIN)]$ ]]
# Host xxxxxxxxx.xxx not found: 3(NXDOMAIN)
then
echo "Host not found!"
exit $E_NOHOST # Bail out.
fi
if [[ "$IPADDR" =~ [(SERVFAIL)]$ ]]
# Host xxxxxxxxx.xxx not found: 2(SERVFAIL)
then
echo "Host not found!"
exit $E_NOHOST # Bail out.
fi
# ======================== Main body of script ========================
AFRINICquery() {
# Define the function that queries AFRINIC. Echo a notification to the
#+ screen, and then run the actual query, redirecting output to $OUTFILE.
echo "Searching for $IPADDR in whois.afrinic.net"
whois -h whois.afrinic.net "$IPADDR" &#62; $OUTFILE
# Check for presence of reference to an rwhois.
# Warn about non-functional rwhois.infosat.net server
#+ and attempt rwhois query.
if grep -e "^remarks: .*rwhois\.[^ ]\+" "$OUTFILE"
then
echo " " &#62;&#62; $OUTFILE
echo "***" &#62;&#62; $OUTFILE
echo "***" &#62;&#62; $OUTFILE
echo "Warning: rwhois.infosat.net was not working \
as of 2005/02/02" &#62;&#62; $OUTFILE
echo " when this script was written." &#62;&#62; $OUTFILE
echo "***" &#62;&#62; $OUTFILE
echo "***" &#62;&#62; $OUTFILE
echo " " &#62;&#62; $OUTFILE
RWHOIS=`grep "^remarks: .*rwhois\.[^ ]\+" "$OUTFILE" | tail -n 1 |\
sed "s/\(^.*\)\(rwhois\..*\)\(:4.*\)/\2/"`
whois -h ${RWHOIS}:${PORT} "$IPADDR" &#62;&#62; $OUTFILE
fi
}
APNICquery() {
echo "Searching for $IPADDR in whois.apnic.net"
whois -h whois.apnic.net "$IPADDR" &#62; $OUTFILE
# Just about every country has its own internet registrar.
# I don't normally bother consulting them, because the regional registry
#+ usually supplies sufficient information.
# There are a few exceptions, where the regional registry simply
#+ refers to the national registry for direct data.
# These are Japan and South Korea in APNIC, and Brasil in LACNIC.
# The following if statement checks $OUTFILE (whois.txt) for the presence
#+ of "KR" (South Korea) or "JP" (Japan) in the country field.
# If either is found, the query is re-run against the appropriate
#+ national registry.
if grep -E "^country:[ ]+KR$" "$OUTFILE"
then
echo "Searching for $IPADDR in whois.krnic.net"
whois -h whois.krnic.net "$IPADDR" &#62;&#62; $OUTFILE
elif grep -E "^country:[ ]+JP$" "$OUTFILE"
then
echo "Searching for $IPADDR in whois.nic.ad.jp"
whois -h whois.nic.ad.jp "$IPADDR"/e &#62;&#62; $OUTFILE
fi
}
ARINquery() {
echo "Searching for $IPADDR in whois.arin.net"
whois -h whois.arin.net "$IPADDR" &#62; $OUTFILE
# Several large internet providers listed by ARIN have their own
#+ internal whois service, referred to as "rwhois".
# A large block of IP addresses is listed with the provider
#+ under the ARIN registry.
# To get the IP addresses of 2nd-level ISPs or other large customers,
#+ one has to refer to the rwhois server on port 4321.
# I originally started with a bunch of "if" statements checking for
#+ the larger providers.
# This approach is unwieldy, and there's always another rwhois server
#+ that I didn't know about.
# A more elegant approach is to check $OUTFILE for a reference
#+ to a whois server, parse that server name out of the comment section,
#+ and re-run the query against the appropriate rwhois server.
# The parsing looks a bit ugly, with a long continued line inside
#+ backticks.
# But it only has to be done once, and will work as new servers are added.
#@ ABS Guide author comment: it isn't all that ugly, and is, in fact,
#@+ an instructive use of Regular Expressions.
if grep -E "^Comment: .*rwhois.[^ ]+" "$OUTFILE"
then
RWHOIS=`grep -e "^Comment:.*rwhois\.[^ ]\+" "$OUTFILE" | tail -n 1 |\
sed "s/^\(.*\)\(rwhois\.[^ ]\+\)\(.*$\)/\2/"`
echo "Searching for $IPADDR in ${RWHOIS}"
whois -h ${RWHOIS}:${PORT} "$IPADDR" &#62;&#62; $OUTFILE
fi
}
LACNICquery() {
echo "Searching for $IPADDR in whois.lacnic.net"
whois -h whois.lacnic.net "$IPADDR" &#62; $OUTFILE
# The following if statement checks $OUTFILE (whois.txt) for
#+ the presence of "BR" (Brasil) in the country field.
# If it is found, the query is re-run against whois.registro.br.
if grep -E "^country:[ ]+BR$" "$OUTFILE"
then
echo "Searching for $IPADDR in whois.registro.br"
whois -h whois.registro.br "$IPADDR" &#62;&#62; $OUTFILE
fi
}
RIPEquery() {
echo "Searching for $IPADDR in whois.ripe.net"
whois -h whois.ripe.net "$IPADDR" &#62; $OUTFILE
}
# Initialize a few variables.
# * slash8 is the most significant octet
# * slash16 consists of the two most significant octets
# * octet2 is the second most significant octet
slash8=`echo $IPADDR | cut -d. -f 1`
if [ -z "$slash8" ] # Yet another sanity check.
then
echo "Undefined error!"
exit $E_UNDEF
fi
slash16=`echo $IPADDR | cut -d. -f 1-2`
# ^ Period specified as 'cut" delimiter.
if [ -z "$slash16" ]
then
echo "Undefined error!"
exit $E_UNDEF
fi
octet2=`echo $slash16 | cut -d. -f 2`
if [ -z "$octet2" ]
then
echo "Undefined error!"
exit $E_UNDEF
fi
# Check for various odds and ends of reserved space.
# There is no point in querying for those addresses.
if [ $slash8 == 0 ]; then
echo $IPADDR is '"This Network"' space\; Not querying
elif [ $slash8 == 10 ]; then
echo $IPADDR is RFC1918 space\; Not querying
elif [ $slash8 == 14 ]; then
echo $IPADDR is '"Public Data Network"' space\; Not querying
elif [ $slash8 == 127 ]; then
echo $IPADDR is loopback space\; Not querying
elif [ $slash16 == 169.254 ]; then
echo $IPADDR is link-local space\; Not querying
elif [ $slash8 == 172 ] &#38;&#38; [ $octet2 -ge 16 ] &#38;&#38; [ $octet2 -le 31 ];then
echo $IPADDR is RFC1918 space\; Not querying
elif [ $slash16 == 192.168 ]; then
echo $IPADDR is RFC1918 space\; Not querying
elif [ $slash8 -ge 224 ]; then
echo $IPADDR is either Multicast or reserved space\; Not querying
elif [ $slash8 -ge 200 ] &#38;&#38; [ $slash8 -le 201 ]; then LACNICquery "$IPADDR"
elif [ $slash8 -ge 202 ] &#38;&#38; [ $slash8 -le 203 ]; then APNICquery "$IPADDR"
elif [ $slash8 -ge 210 ] &#38;&#38; [ $slash8 -le 211 ]; then APNICquery "$IPADDR"
elif [ $slash8 -ge 218 ] &#38;&#38; [ $slash8 -le 223 ]; then APNICquery "$IPADDR"
# If we got this far without making a decision, query ARIN.
# If a reference is found in $OUTFILE to APNIC, AFRINIC, LACNIC, or RIPE,
#+ query the appropriate whois server.
else
ARINquery "$IPADDR"
if grep "whois.afrinic.net" "$OUTFILE"; then
AFRINICquery "$IPADDR"
elif grep -E "^OrgID:[ ]+RIPE$" "$OUTFILE"; then
RIPEquery "$IPADDR"
elif grep -E "^OrgID:[ ]+APNIC$" "$OUTFILE"; then
APNICquery "$IPADDR"
elif grep -E "^OrgID:[ ]+LACNIC$" "$OUTFILE"; then
LACNICquery "$IPADDR"
fi
fi
#@ ---------------------------------------------------------------
# Try also:
# wget http://logi.cc/nw/whois.php3?ACTION=doQuery&#38;DOMAIN=$IPADDR
#@ ---------------------------------------------------------------
# We've now finished the querying.
# Echo a copy of the final result to the screen.
cat $OUTFILE
# Or "less $OUTFILE" . . .
exit 0
#@ ABS Guide author comments:
#@ Nothing fancy here, but still a very useful tool for hunting spammers.
#@ Sure, the script can be cleaned up some, and it's still a bit buggy,
#@+ (exercise for reader), but all the same, it's a nice piece of coding
#@+ by Walter Dnes.
#@ Thank you!</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><SPAN
CLASS="QUOTE"
>"Little Monster's"</SPAN
> front end to <A
HREF="#WGETREF"
>wget</A
>.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="WGETTER2"
></A
><P
><B
>Example A-30. Making <I
CLASS="FIRSTTERM"
>wget</I
> easier to use</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# wgetter2.bash
# Author: Little Monster [monster@monstruum.co.uk]
# ==&#62; Used in ABS Guide with permission of script author.
# ==&#62; This script still needs debugging and fixups (exercise for reader).
# ==&#62; It could also use some additional editing in the comments.
# This is wgetter2 --
#+ a Bash script to make wget a bit more friendly, and save typing.
# Carefully crafted by Little Monster.
# More or less complete on 02/02/2005.
# If you think this script can be improved,
#+ email me at: monster@monstruum.co.uk
# ==&#62; and cc: to the author of the ABS Guide, please.
# This script is licenced under the GPL.
# You are free to copy, alter and re-use it,
#+ but please don't try to claim you wrote it.
# Log your changes here instead.
# =======================================================================
# changelog:
# 07/02/2005. Fixups by Little Monster.
# 02/02/2005. Minor additions by Little Monster.
# (See after # +++++++++++ )
# 29/01/2005. Minor stylistic edits and cleanups by author of ABS Guide.
# Added exit error codes.
# 22/11/2004. Finished initial version of second version of wgetter:
# wgetter2 is born.
# 01/12/2004. Changed 'runn' function so it can be run 2 ways --
# either ask for a file name or have one input on the CL.
# 01/12/2004. Made sensible handling of no URL's given.
# 01/12/2004. Made loop of main options, so you don't
# have to keep calling wgetter 2 all the time.
# Runs as a session instead.
# 01/12/2004. Added looping to 'runn' function.
# Simplified and improved.
# 01/12/2004. Added state to recursion setting.
# Enables re-use of previous value.
# 05/12/2004. Modified the file detection routine in the 'runn' function
# so it's not fooled by empty values, and is cleaner.
# 01/02/2004. Added cookie finding routine from later version (which
# isn't ready yet), so as not to have hard-coded paths.
# =======================================================================
# Error codes for abnormal exit.
E_USAGE=67 # Usage message, then quit.
E_NO_OPTS=68 # No command-line args entered.
E_NO_URLS=69 # No URLs passed to script.
E_NO_SAVEFILE=70 # No save filename passed to script.
E_USER_EXIT=71 # User decides to quit.
# Basic default wget command we want to use.
# This is the place to change it, if required.
# NB: if using a proxy, set http_proxy = yourproxy in .wgetrc.
# Otherwise delete --proxy=on, below.
# ====================================================================
CommandA="wget -nc -c -t 5 --progress=bar --random-wait --proxy=on -r"
# ====================================================================
# --------------------------------------------------------------------
# Set some other variables and explain them.
pattern=" -A .jpg,.JPG,.jpeg,.JPEG,.gif,.GIF,.htm,.html,.shtml,.php"
# wget's option to only get certain types of file.
# comment out if not using
today=`date +%F` # Used for a filename.
home=$HOME # Set HOME to an internal variable.
# In case some other path is used, change it here.
depthDefault=3 # Set a sensible default recursion.
Depth=$depthDefault # Otherwise user feedback doesn't tie in properly.
RefA="" # Set blank referring page.
Flag="" # Default to not saving anything,
#+ or whatever else might be wanted in future.
lister="" # Used for passing a list of urls directly to wget.
Woptions="" # Used for passing wget some options for itself.
inFile="" # Used for the run function.
newFile="" # Used for the run function.
savePath="$home/w-save"
Config="$home/.wgetter2rc"
# This is where some variables can be stored,
#+ if permanently changed from within the script.
Cookie_List="$home/.cookielist"
# So we know where the cookies are kept . . .
cFlag="" # Part of the cookie file selection routine.
# Define the options available. Easy to change letters here if needed.
# These are the optional options; you don't just wait to be asked.
save=s # Save command instead of executing it.
cook=c # Change cookie file for this session.
help=h # Usage guide.
list=l # Pass wget the -i option and URL list.
runn=r # Run saved commands as an argument to the option.
inpu=i # Run saved commands interactively.
wopt=w # Allow to enter options to pass directly to wget.
# --------------------------------------------------------------------
if [ -z "$1" ]; then # Make sure we get something for wget to eat.
echo "You must at least enter a URL or option!"
echo "-$help for usage."
exit $E_NO_OPTS
fi
# +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
# added added added added added added added added added added added added
if [ ! -e "$Config" ]; then # See if configuration file exists.
echo "Creating configuration file, $Config"
echo "# This is the configuration file for wgetter2" &#62; "$Config"
echo "# Your customised settings will be saved in this file" &#62;&#62; "$Config"
else
source $Config # Import variables we set outside the script.
fi
if [ ! -e "$Cookie_List" ]; then
# Set up a list of cookie files, if there isn't one.
echo "Hunting for cookies . . ."
find -name cookies.txt &#62;&#62; $Cookie_List # Create the list of cookie files.
fi # Isolate this in its own 'if' statement,
#+ in case we got interrupted while searching.
if [ -z "$cFlag" ]; then # If we haven't already done this . . .
echo # Make a nice space after the command prompt.
echo "Looks like you haven't set up your source of cookies yet."
n=0 # Make sure the counter
#+ doesn't contain random values.
while read; do
Cookies[$n]=$REPLY # Put the cookie files we found into an array.
echo "$n) ${Cookies[$n]}" # Create a menu.
n=$(( n + 1 )) # Increment the counter.
done &#60; $Cookie_List # Feed the read statement.
echo "Enter the number of the cookie file you want to use."
echo "If you won't be using cookies, just press RETURN."
echo
echo "I won't be asking this again. Edit $Config"
echo "If you decide to change at a later date"
echo "or use the -${cook} option for per session changes."
read
if [ ! -z $REPLY ]; then # User didn't just press return.
Cookie=" --load-cookies ${Cookies[$REPLY]}"
# Set the variable here as well as in the config file.
echo "Cookie=\" --load-cookies ${Cookies[$REPLY]}\"" &#62;&#62; $Config
fi
echo "cFlag=1" &#62;&#62; $Config # So we know not to ask again.
fi
# end added section end added section end added section end added section
# +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
# Another variable.
# This one may or may not be subject to variation.
# A bit like the small print.
CookiesON=$Cookie
# echo "cookie file is $CookiesON" # For debugging.
# echo "home is ${home}" # For debugging.
# Got caught with this one!
wopts()
{
echo "Enter options to pass to wget."
echo "It is assumed you know what you're doing."
echo
echo "You can pass their arguments here too."
# That is to say, everything passed here is passed to wget.
read Wopts
# Read in the options to be passed to wget.
Woptions=" $Wopts"
# ^ Why the leading space?
# Assign to another variable.
# Just for fun, or something . . .
echo "passing options ${Wopts} to wget"
# Mainly for debugging.
# Is cute.
return
}
save_func()
{
echo "Settings will be saved."
if [ ! -d $savePath ]; then # See if directory exists.
mkdir $savePath # Create the directory to save things in
#+ if it isn't already there.
fi
Flag=S
# Tell the final bit of code what to do.
# Set a flag since stuff is done in main.
return
}
usage() # Tell them how it works.
{
echo "Welcome to wgetter. This is a front end to wget."
echo "It will always run wget with these options:"
echo "$CommandA"
echo "and the pattern to match: $pattern \
(which you can change at the top of this script)."
echo "It will also ask you for recursion depth, \
and if you want to use a referring page."
echo "Wgetter accepts the following options:"
echo ""
echo "-$help : Display this help."
echo "-$save : Save the command to a file $savePath/wget-($today) \
instead of running it."
echo "-$runn : Run saved wget commands instead of starting a new one -"
echo "Enter filename as argument to this option."
echo "-$inpu : Run saved wget commands interactively --"
echo "The script will ask you for the filename."
echo "-$cook : Change the cookies file for this session."
echo "-$list : Tell wget to use URL's from a list instead of \
from the command-line."
echo "-$wopt : Pass any other options direct to wget."
echo ""
echo "See the wget man page for additional options \
you can pass to wget."
echo ""
exit $E_USAGE # End here. Don't process anything else.
}
list_func() # Gives the user the option to use the -i option to wget,
#+ and a list of URLs.
{
while [ 1 ]; do
echo "Enter the name of the file containing URL's (press q to change
your mind)."
read urlfile
if [ ! -e "$urlfile" ] &#38;&#38; [ "$urlfile" != q ]; then
# Look for a file, or the quit option.
echo "That file does not exist!"
elif [ "$urlfile" = q ]; then # Check quit option.
echo "Not using a url list."
return
else
echo "using $urlfile."
echo "If you gave url's on the command-line, I'll use those first."
# Report wget standard behaviour to the user.
lister=" -i $urlfile" # This is what we want to pass to wget.
return
fi
done
}
cookie_func() # Give the user the option to use a different cookie file.
{
while [ 1 ]; do
echo "Change the cookies file. Press return if you don't want to change
it."
read Cookies
# NB: this is not the same as Cookie, earlier.
# There is an 's' on the end.
# Bit like chocolate chips.
if [ -z "$Cookies" ]; then # Escape clause for wusses.
return
elif [ ! -e "$Cookies" ]; then
echo "File does not exist. Try again." # Keep em going . . .
else
CookiesON=" --load-cookies $Cookies" # File is good -- use it!
return
fi
done
}
run_func()
{
if [ -z "$OPTARG" ]; then
# Test to see if we used the in-line option or the query one.
if [ ! -d "$savePath" ]; then # If directory doesn't exist . . .
echo "$savePath does not appear to exist."
echo "Please supply path and filename of saved wget commands:"
read newFile
until [ -f "$newFile" ]; do # Keep going till we get something.
echo "Sorry, that file does not exist. Please try again."
# Try really hard to get something.
read newFile
done
# -----------------------------------------------------------------------
# if [ -z ( grep wget ${newfile} ) ]; then
# Assume they haven't got the right file and bail out.
# echo "Sorry, that file does not contain wget commands. Aborting."
# exit
# fi
#
# This is bogus code.
# It doesn't actually work.
# If anyone wants to fix it, feel free!
# -----------------------------------------------------------------------
filePath="${newFile}"
else
echo "Save path is $savePath"
echo "Please enter name of the file which you want to use."
echo "You have a choice of:"
ls $savePath # Give them a choice.
read inFile
until [ -f "$savePath/$inFile" ]; do # Keep going till
#+ we get something.
if [ ! -f "${savePath}/${inFile}" ]; then # If file doesn't exist.
echo "Sorry, that file does not exist. Please choose from:"
ls $savePath # If a mistake is made.
read inFile
fi
done
filePath="${savePath}/${inFile}" # Make one variable . . .
fi
else filePath="${savePath}/${OPTARG}" # Which can be many things . . .
fi
if [ ! -f "$filePath" ]; then # If a bogus file got through.
echo "You did not specify a suitable file."
echo "Run this script with the -${save} option first."
echo "Aborting."
exit $E_NO_SAVEFILE
fi
echo "Using: $filePath"
while read; do
eval $REPLY
echo "Completed: $REPLY"
done &#60; $filePath # Feed the actual file we are using into a 'while' loop.
exit
}
# Fish out any options we are using for the script.
# This is based on the demo in "Learning The Bash Shell" (O'Reilly).
while getopts ":$save$cook$help$list$runn:$inpu$wopt" opt
do
case $opt in
$save) save_func;; # Save some wgetter sessions for later.
$cook) cookie_func;; # Change cookie file.
$help) usage;; # Get help.
$list) list_func;; # Allow wget to use a list of URLs.
$runn) run_func;; # Useful if you are calling wgetter from,
#+ for example, a cron script.
$inpu) run_func;; # When you don't know what your files are named.
$wopt) wopts;; # Pass options directly to wget.
\?) echo "Not a valid option."
echo "Use -${wopt} to pass options directly to wget,"
echo "or -${help} for help";; # Catch anything else.
esac
done
shift $((OPTIND - 1)) # Do funky magic stuff with $#.
if [ -z "$1" ] &#38;&#38; [ -z "$lister" ]; then
# We should be left with at least one URL
#+ on the command-line, unless a list is
#+ being used -- catch empty CL's.
echo "No URL's given! You must enter them on the same line as wgetter2."
echo "E.g., wgetter2 http://somesite http://anothersite."
echo "Use $help option for more information."
exit $E_NO_URLS # Bail out, with appropriate error code.
fi
URLS=" $@"
# Use this so that URL list can be changed if we stay in the option loop.
while [ 1 ]; do
# This is where we ask for the most used options.
# (Mostly unchanged from version 1 of wgetter)
if [ -z $curDepth ]; then
Current=""
else Current=" Current value is $curDepth"
fi
echo "How deep should I go? \
(integer: Default is $depthDefault.$Current)"
read Depth # Recursion -- how far should we go?
inputB="" # Reset this to blank on each pass of the loop.
echo "Enter the name of the referring page (default is none)."
read inputB # Need this for some sites.
echo "Do you want to have the output logged to the terminal"
echo "(y/n, default is yes)?"
read noHide # Otherwise wget will just log it to a file.
case $noHide in # Now you see me, now you don't.
y|Y ) hide="";;
n|N ) hide=" -b";;
* ) hide="";;
esac
if [ -z ${Depth} ]; then
# User accepted either default or current depth,
#+ in which case Depth is now empty.
if [ -z ${curDepth} ]; then
# See if a depth was set on a previous iteration.
Depth="$depthDefault"
# Set the default recursion depth if nothing
#+ else to use.
else Depth="$curDepth" # Otherwise, set the one we used before.
fi
fi
Recurse=" -l $Depth" # Set how deep we want to go.
curDepth=$Depth # Remember setting for next time.
if [ ! -z $inputB ]; then
RefA=" --referer=$inputB" # Option to use referring page.
fi
WGETTER="${CommandA}${pattern}${hide}${RefA}${Recurse}\
${CookiesON}${lister}${Woptions}${URLS}"
# Just string the whole lot together . . .
# NB: no embedded spaces.
# They are in the individual elements so that if any are empty,
#+ we don't get an extra space.
if [ -z "${CookiesON}" ] &#38;&#38; [ "$cFlag" = "1" ] ; then
echo "Warning -- can't find cookie file"
# This should be changed,
#+ in case the user has opted to not use cookies.
fi
if [ "$Flag" = "S" ]; then
echo "$WGETTER" &#62;&#62; $savePath/wget-${today}
# Create a unique filename for today, or append to it if it exists.
echo "$inputB" &#62;&#62; $savePath/site-list-${today}
# Make a list, so it's easy to refer back to,
#+ since the whole command is a bit confusing to look at.
echo "Command saved to the file $savePath/wget-${today}"
# Tell the user.
echo "Referring page URL saved to the file$ \
savePath/site-list-${today}"
# Tell the user.
Saver=" with save option"
# Stick this somewhere, so it appears in the loop if set.
else
echo "*****************"
echo "*****Getting*****"
echo "*****************"
echo ""
echo "$WGETTER"
echo ""
echo "*****************"
eval "$WGETTER"
fi
echo ""
echo "Starting over$Saver."
echo "If you want to stop, press q."
echo "Otherwise, enter some URL's:"
# Let them go again. Tell about save option being set.
read
case $REPLY in
# Need to change this to a 'trap' clause.
q|Q ) exit $E_USER_EXIT;; # Exercise for the reader?
* ) URLS=" $REPLY";;
esac
echo ""
done
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="BASHPODDER"
></A
><P
><B
>Example A-31. A <I
CLASS="FIRSTTERM"
>podcasting</I
> script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# bashpodder.sh:
# By Linc 10/1/2004
# Find the latest script at
#+ http://linc.homeunix.org:8080/scripts/bashpodder
# Last revision 12/14/2004 - Many Contributors!
# If you use this and have made improvements or have comments
#+ drop me an email at linc dot fessenden at gmail dot com
# I'd appreciate it!
# ==&#62; ABS Guide extra comments.
# ==&#62; Author of this script has kindly granted permission
# ==&#62;+ for inclusion in ABS Guide.
# ==&#62; ################################################################
#
# ==&#62; What is "podcasting"?
# ==&#62; It's broadcasting "radio shows" over the Internet.
# ==&#62; These shows can be played on iPods and other music file players.
# ==&#62; This script makes it possible.
# ==&#62; See documentation at the script author's site, above.
# ==&#62; ################################################################
# Make script crontab friendly:
cd $(dirname $0)
# ==&#62; Change to directory where this script lives.
# datadir is the directory you want podcasts saved to:
datadir=$(date +%Y-%m-%d)
# ==&#62; Will create a date-labeled directory, named: YYYY-MM-DD
# Check for and create datadir if necessary:
if test ! -d $datadir
then
mkdir $datadir
fi
# Delete any temp file:
rm -f temp.log
# Read the bp.conf file and wget any url not already
#+ in the podcast.log file:
while read podcast
do # ==&#62; Main action follows.
file=$(wget -q $podcast -O - | tr '\r' '\n' | tr \' \" | \
sed -n 's/.*url="\([^"]*\)".*/\1/p')
for url in $file
do
echo $url &#62;&#62; temp.log
if ! grep "$url" podcast.log &#62; /dev/null
then
wget -q -P $datadir "$url"
fi
done
done &#60; bp.conf
# Move dynamically created log file to permanent log file:
cat podcast.log &#62;&#62; temp.log
sort temp.log | uniq &#62; podcast.log
rm temp.log
# Create an m3u playlist:
ls $datadir | grep -v m3u &#62; $datadir/podcast.m3u
exit 0
#################################################
For a different scripting approach to Podcasting,
see Phil Salkie's article,
"Internet Radio to Podcast with Shell Tools"
in the September, 2005 issue of LINUX JOURNAL,
http://www.linuxjournal.com/article/8171
#################################################</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="NIGHTLYBACKUP"
></A
><P
><B
>Example A-32. Nightly backup to a firewire HD</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# nightly-backup.sh
# http://www.richardneill.org/source.php#nightly-backup-rsync
# Copyright (c) 2005 Richard Neill &#60;backup@richardneill.org&#62;.
# This is Free Software licensed under the GNU GPL.
# ==&#62; Included in ABS Guide with script author's kind permission.
# ==&#62; (Thanks!)
# This does a backup from the host computer to a locally connected
#+ firewire HDD using rsync and ssh.
# (Script should work with USB-connected device (see lines 40-43).
# It then rotates the backups.
# Run it via cron every night at 5am.
# This only backs up the home directory.
# If ownerships (other than the user's) should be preserved,
#+ then run the rsync process as root (and re-instate the -o).
# We save every day for 7 days, then every week for 4 weeks,
#+ then every month for 3 months.
# See: http://www.mikerubel.org/computers/rsync_snapshots/
#+ for more explanation of the theory.
# Save as: $HOME/bin/nightly-backup_firewire-hdd.sh
# Known bugs:
# ----------
# i) Ideally, we want to exclude ~/.tmp and the browser caches.
# ii) If the user is sitting at the computer at 5am,
#+ and files are modified while the rsync is occurring,
#+ then the BACKUP_JUSTINCASE branch gets triggered.
# To some extent, this is a
#+ feature, but it also causes a "disk-space leak".
##### BEGIN CONFIGURATION SECTION ############################################
LOCAL_USER=rjn # User whose home directory should be backed up.
MOUNT_POINT=/backup # Mountpoint of backup drive.
# NO trailing slash!
# This must be unique (eg using a udev symlink)
# MOUNT_POINT=/media/disk # For USB-connected device.
SOURCE_DIR=/home/$LOCAL_USER # NO trailing slash - it DOES matter to rsync.
BACKUP_DEST_DIR=$MOUNT_POINT/backup/`hostname -s`.${LOCAL_USER}.nightly_backup
DRY_RUN=false #If true, invoke rsync with -n, to do a dry run.
# Comment out or set to false for normal use.
VERBOSE=false # If true, make rsync verbose.
# Comment out or set to false otherwise.
COMPRESS=false # If true, compress.
# Good for internet, bad on LAN.
# Comment out or set to false otherwise.
### Exit Codes ###
E_VARS_NOT_SET=64
E_COMMANDLINE=65
E_MOUNT_FAIL=70
E_NOSOURCEDIR=71
E_UNMOUNTED=72
E_BACKUP=73
##### END CONFIGURATION SECTION ##############################################
# Check that all the important variables have been set:
if [ -z "$LOCAL_USER" ] ||
[ -z "$SOURCE_DIR" ] ||
[ -z "$MOUNT_POINT" ] ||
[ -z "$BACKUP_DEST_DIR" ]
then
echo 'One of the variables is not set! Edit the file: $0. BACKUP FAILED.'
exit $E_VARS_NOT_SET
fi
if [ "$#" != 0 ] # If command-line param(s) . . .
then # Here document(ation).
cat &#60;&#60;-ENDOFTEXT
Automatic Nightly backup run from cron.
Read the source for more details: $0
The backup directory is $BACKUP_DEST_DIR .
It will be created if necessary; initialisation is no longer required.
WARNING: Contents of $BACKUP_DEST_DIR are rotated.
Directories named 'backup.\$i' will eventually be DELETED.
We keep backups from every day for 7 days (1-8),
then every week for 4 weeks (9-12),
then every month for 3 months (13-15).
You may wish to add this to your crontab using 'crontab -e'
# Back up files: $SOURCE_DIR to $BACKUP_DEST_DIR
#+ every night at 3:15 am
15 03 * * * /home/$LOCAL_USER/bin/nightly-backup_firewire-hdd.sh
Don't forget to verify the backups are working,
especially if you don't read cron's mail!"
ENDOFTEXT
exit $E_COMMANDLINE
fi
# Parse the options.
# ==================
if [ "$DRY_RUN" == "true" ]; then
DRY_RUN="-n"
echo "WARNING:"
echo "THIS IS A 'DRY RUN'!"
echo "No data will actually be transferred!"
else
DRY_RUN=""
fi
if [ "$VERBOSE" == "true" ]; then
VERBOSE="-v"
else
VERBOSE=""
fi
if [ "$COMPRESS" == "true" ]; then
COMPRESS="-z"
else
COMPRESS=""
fi
# Every week (actually of 8 days) and every month,
#+ extra backups are preserved.
DAY_OF_MONTH=`date +%d` # Day of month (01..31).
if [ $DAY_OF_MONTH = 01 ]; then # First of month.
MONTHSTART=true
elif [ $DAY_OF_MONTH = 08 \
-o $DAY_OF_MONTH = 16 \
-o $DAY_OF_MONTH = 24 ]; then
# Day 8,16,24 (use 8, not 7 to better handle 31-day months)
WEEKSTART=true
fi
# Check that the HDD is mounted.
# At least, check that *something* is mounted here!
# We can use something unique to the device, rather than just guessing
#+ the scsi-id by having an appropriate udev rule in
#+ /etc/udev/rules.d/10-rules.local
#+ and by putting a relevant entry in /etc/fstab.
# Eg: this udev rule:
# BUS="scsi", KERNEL="sd*", SYSFS{vendor}="WDC WD16",
# SYSFS{model}="00JB-00GVA0 ", NAME="%k", SYMLINK="lacie_1394d%n"
if mount | grep $MOUNT_POINT &#62;/dev/null; then
echo "Mount point $MOUNT_POINT is indeed mounted. OK"
else
echo -n "Attempting to mount $MOUNT_POINT..."
# If it isn't mounted, try to mount it.
sudo mount $MOUNT_POINT 2&#62;/dev/null
if mount | grep $MOUNT_POINT &#62;/dev/null; then
UNMOUNT_LATER=TRUE
echo "OK"
# Note: Ensure that this is also unmounted
#+ if we exit prematurely with failure.
else
echo "FAILED"
echo -e "Nothing is mounted at $MOUNT_POINT. BACKUP FAILED!"
exit $E_MOUNT_FAIL
fi
fi
# Check that source dir exists and is readable.
if [ ! -r $SOURCE_DIR ] ; then
echo "$SOURCE_DIR does not exist, or cannot be read. BACKUP FAILED."
exit $E_NOSOURCEDIR
fi
# Check that the backup directory structure is as it should be.
# If not, create it.
# Create the subdirectories.
# Note that backup.0 will be created as needed by rsync.
for ((i=1;i&#60;=15;i++)); do
if [ ! -d $BACKUP_DEST_DIR/backup.$i ]; then
if /bin/mkdir -p $BACKUP_DEST_DIR/backup.$i ; then
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ No [ ] test brackets. Why?
echo "Warning: directory $BACKUP_DEST_DIR/backup.$i is missing,"
echo "or was not initialised. (Re-)creating it."
else
echo "ERROR: directory $BACKUP_DEST_DIR/backup.$i"
echo "is missing and could not be created."
if [ "$UNMOUNT_LATER" == "TRUE" ]; then
# Before we exit, unmount the mount point if necessary.
cd
sudo umount $MOUNT_POINT &#38;&#38;
echo "Unmounted $MOUNT_POINT again. Giving up."
fi
exit $E_UNMOUNTED
fi
fi
done
# Set the permission to 700 for security
#+ on an otherwise permissive multi-user system.
if ! /bin/chmod 700 $BACKUP_DEST_DIR ; then
echo "ERROR: Could not set permissions on $BACKUP_DEST_DIR to 700."
if [ "$UNMOUNT_LATER" == "TRUE" ]; then
# Before we exit, unmount the mount point if necessary.
cd ; sudo umount $MOUNT_POINT \
&#38;&#38; echo "Unmounted $MOUNT_POINT again. Giving up."
fi
exit $E_UNMOUNTED
fi
# Create the symlink: current -&#62; backup.1 if required.
# A failure here is not critical.
cd $BACKUP_DEST_DIR
if [ ! -h current ] ; then
if ! /bin/ln -s backup.1 current ; then
echo "WARNING: could not create symlink current -&#62; backup.1"
fi
fi
# Now, do the rsync.
echo "Now doing backup with rsync..."
echo "Source dir: $SOURCE_DIR"
echo -e "Backup destination dir: $BACKUP_DEST_DIR\n"
/usr/bin/rsync $DRY_RUN $VERBOSE -a -S --delete --modify-window=60 \
--link-dest=../backup.1 $SOURCE_DIR $BACKUP_DEST_DIR/backup.0/
# Only warn, rather than exit if the rsync failed,
#+ since it may only be a minor problem.
# E.g., if one file is not readable, rsync will fail.
# This shouldn't prevent the rotation.
# Not using, e.g., `date +%a` since these directories
#+ are just full of links and don't consume *that much* space.
if [ $? != 0 ]; then
BACKUP_JUSTINCASE=backup.`date +%F_%T`.justincase
echo "WARNING: the rsync process did not entirely succeed."
echo "Something might be wrong."
echo "Saving an extra copy at: $BACKUP_JUSTINCASE"
echo "WARNING: if this occurs regularly, a LOT of space will be consumed,"
echo "even though these are just hard-links!"
fi
# Save a readme in the backup parent directory.
# Save another one in the recent subdirectory.
echo "Backup of $SOURCE_DIR on `hostname` was last run on \
`date`" &#62; $BACKUP_DEST_DIR/README.txt
echo "This backup of $SOURCE_DIR on `hostname` was created on \
`date`" &#62; $BACKUP_DEST_DIR/backup.0/README.txt
# If we are not in a dry run, rotate the backups.
[ -z "$DRY_RUN" ] &#38;&#38;
# Check how full the backup disk is.
# Warn if 90%. if 98% or more, we'll probably fail, so give up.
# (Note: df can output to more than one line.)
# We test this here, rather than before
#+ so that rsync may possibly have a chance.
DISK_FULL_PERCENT=`/bin/df $BACKUP_DEST_DIR |
tr "\n" ' ' | awk '{print $12}' | grep -oE [0-9]+ `
echo "Disk space check on backup partition \
$MOUNT_POINT $DISK_FULL_PERCENT% full."
if [ $DISK_FULL_PERCENT -gt 90 ]; then
echo "Warning: Disk is greater than 90% full."
fi
if [ $DISK_FULL_PERCENT -gt 98 ]; then
echo "Error: Disk is full! Giving up."
if [ "$UNMOUNT_LATER" == "TRUE" ]; then
# Before we exit, unmount the mount point if necessary.
cd; sudo umount $MOUNT_POINT &#38;&#38;
echo "Unmounted $MOUNT_POINT again. Giving up."
fi
exit $E_UNMOUNTED
fi
# Create an extra backup.
# If this copy fails, give up.
if [ -n "$BACKUP_JUSTINCASE" ]; then
if ! /bin/cp -al $BACKUP_DEST_DIR/backup.0 \
$BACKUP_DEST_DIR/$BACKUP_JUSTINCASE
then
echo "ERROR: Failed to create extra copy \
$BACKUP_DEST_DIR/$BACKUP_JUSTINCASE"
if [ "$UNMOUNT_LATER" == "TRUE" ]; then
# Before we exit, unmount the mount point if necessary.
cd ;sudo umount $MOUNT_POINT &#38;&#38;
echo "Unmounted $MOUNT_POINT again. Giving up."
fi
exit $E_UNMOUNTED
fi
fi
# At start of month, rotate the oldest 8.
if [ "$MONTHSTART" == "true" ]; then
echo -e "\nStart of month. \
Removing oldest backup: $BACKUP_DEST_DIR/backup.15" &#38;&#38;
/bin/rm -rf $BACKUP_DEST_DIR/backup.15 &#38;&#38;
echo "Rotating monthly,weekly backups: \
$BACKUP_DEST_DIR/backup.[8-14] -&#62; $BACKUP_DEST_DIR/backup.[9-15]" &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.14 $BACKUP_DEST_DIR/backup.15 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.13 $BACKUP_DEST_DIR/backup.14 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.12 $BACKUP_DEST_DIR/backup.13 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.11 $BACKUP_DEST_DIR/backup.12 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.10 $BACKUP_DEST_DIR/backup.11 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.9 $BACKUP_DEST_DIR/backup.10 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.8 $BACKUP_DEST_DIR/backup.9
# At start of week, rotate the second-oldest 4.
elif [ "$WEEKSTART" == "true" ]; then
echo -e "\nStart of week. \
Removing oldest weekly backup: $BACKUP_DEST_DIR/backup.12" &#38;&#38;
/bin/rm -rf $BACKUP_DEST_DIR/backup.12 &#38;&#38;
echo "Rotating weekly backups: \
$BACKUP_DEST_DIR/backup.[8-11] -&#62; $BACKUP_DEST_DIR/backup.[9-12]" &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.11 $BACKUP_DEST_DIR/backup.12 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.10 $BACKUP_DEST_DIR/backup.11 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.9 $BACKUP_DEST_DIR/backup.10 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.8 $BACKUP_DEST_DIR/backup.9
else
echo -e "\nRemoving oldest daily backup: $BACKUP_DEST_DIR/backup.8" &#38;&#38;
/bin/rm -rf $BACKUP_DEST_DIR/backup.8
fi &#38;&#38;
# Every day, rotate the newest 8.
echo "Rotating daily backups: \
$BACKUP_DEST_DIR/backup.[1-7] -&#62; $BACKUP_DEST_DIR/backup.[2-8]" &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.7 $BACKUP_DEST_DIR/backup.8 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.6 $BACKUP_DEST_DIR/backup.7 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.5 $BACKUP_DEST_DIR/backup.6 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.4 $BACKUP_DEST_DIR/backup.5 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.3 $BACKUP_DEST_DIR/backup.4 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.2 $BACKUP_DEST_DIR/backup.3 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.1 $BACKUP_DEST_DIR/backup.2 &#38;&#38;
/bin/mv $BACKUP_DEST_DIR/backup.0 $BACKUP_DEST_DIR/backup.1 &#38;&#38;
SUCCESS=true
if [ "$UNMOUNT_LATER" == "TRUE" ]; then
# Unmount the mount point if it wasn't mounted to begin with.
cd ; sudo umount $MOUNT_POINT &#38;&#38; echo "Unmounted $MOUNT_POINT again."
fi
if [ "$SUCCESS" == "true" ]; then
echo 'SUCCESS!'
exit 0
fi
# Should have already exited if backup worked.
echo 'BACKUP FAILED! Is this just a dry run? Is the disk full?) '
exit $E_BACKUP</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="CDLL"
></A
><P
><B
>Example A-33. An expanded <I
CLASS="FIRSTTERM"
>cd</I
> command</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>###########################################################################
#
# cdll
# by Phil Braham
#
# ############################################
# Latest version of this script available from
# http://freshmeat.net/projects/cd/
# ############################################
#
# .cd_new
#
# An enhancement of the Unix cd command
#
# There are unlimited stack entries and special entries. The stack
# entries keep the last cd_maxhistory
# directories that have been used. The special entries can be
# assigned to commonly used directories.
#
# The special entries may be pre-assigned by setting the environment
# variables CDSn or by using the -u or -U command.
#
# The following is a suggestion for the .profile file:
#
# . cdll # Set up the cd command
# alias cd='cd_new' # Replace the cd command
# cd -U # Upload pre-assigned entries for
# #+ the stack and special entries
# cd -D # Set non-default mode
# alias @="cd_new @" # Allow @ to be used to get history
#
# For help type:
#
# cd -h or
# cd -H
#
#
###########################################################################
#
# Version 1.2.1
#
# Written by Phil Braham - Realtime Software Pty Ltd
# (realtime@mpx.com.au)
# Please send any suggestions or enhancements to the author (also at
# phil@braham.net)
#
############################################################################
cd_hm ()
{
${PRINTF} "%s" "cd [dir] [0-9] [@[s|h] [-g [&#60;dir&#62;]] [-d] \
[-D] [-r&#60;n&#62;] [dir|0-9] [-R&#60;n&#62;] [&#60;dir&#62;|0-9]
[-s&#60;n&#62;] [-S&#60;n&#62;] [-u] [-U] [-f] [-F] [-h] [-H] [-v]
&#60;dir&#62; Go to directory
0-n Go to previous directory (0 is previous, 1 is last but 1 etc)
n is up to max history (default is 50)
@ List history and special entries
@h List history entries
@s List special entries
-g [&#60;dir&#62;] Go to literal name (bypass special names)
This is to allow access to dirs called '0','1','-h' etc
-d Change default action - verbose. (See note)
-D Change default action - silent. (See note)
-s&#60;n&#62; Go to the special entry &#60;n&#62;*
-S&#60;n&#62; Go to the special entry &#60;n&#62;
and replace it with the current dir*
-r&#60;n&#62; [&#60;dir&#62;] Go to directory &#60;dir&#62;
and then put it on special entry &#60;n&#62;*
-R&#60;n&#62; [&#60;dir&#62;] Go to directory &#60;dir&#62;
and put current dir on special entry &#60;n&#62;*
-a&#60;n&#62; Alternative suggested directory. See note below.
-f [&#60;file&#62;] File entries to &#60;file&#62;.
-u [&#60;file&#62;] Update entries from &#60;file&#62;.
If no filename supplied then default file
(${CDPath}${2:-"$CDFile"}) is used
-F and -U are silent versions
-v Print version number
-h Help
-H Detailed help
*The special entries (0 - 9) are held until log off, replaced by another
entry or updated with the -u command
Alternative suggested directories:
If a directory is not found then CD will suggest any
possibilities. These are directories starting with the same letters
and if any are found they are listed prefixed with -a&#60;n&#62;
where &#60;n&#62; is a number.
It's possible to go to the directory by entering cd -a&#60;n&#62;
on the command line.
The directory for -r&#60;n&#62; or -R&#60;n&#62; may be a number.
For example:
$ cd -r3 4 Go to history entry 4 and put it on special entry 3
$ cd -R3 4 Put current dir on the special entry 3
and go to history entry 4
$ cd -s3 Go to special entry 3
Note that commands R,r,S and s may be used without a number
and refer to 0:
$ cd -s Go to special entry 0
$ cd -S Go to special entry 0 and make special
entry 0 current dir
$ cd -r 1 Go to history entry 1 and put it on special entry 0
$ cd -r Go to history entry 0 and put it on special entry 0
"
if ${TEST} "$CD_MODE" = "PREV"
then
${PRINTF} "$cd_mnset"
else
${PRINTF} "$cd_mset"
fi
}
cd_Hm ()
{
cd_hm
${PRINTF} "%s" "
The previous directories (0-$cd_maxhistory) are stored in the
environment variables CD[0] - CD[$cd_maxhistory]
Similarly the special directories S0 - $cd_maxspecial are in
the environment variable CDS[0] - CDS[$cd_maxspecial]
and may be accessed from the command line
The default pathname for the -f and -u commands is $CDPath
The default filename for the -f and -u commands is $CDFile
Set the following environment variables:
CDL_PROMPTLEN - Set to the length of prompt you require.
Prompt string is set to the right characters of the
current directory.
If not set then prompt is left unchanged
CDL_PROMPT_PRE - Set to the string to prefix the prompt.
Default is:
non-root: \"\\[\\e[01;34m\\]\" (sets colour to blue).
root: \"\\[\\e[01;31m\\]\" (sets colour to red).
CDL_PROMPT_POST - Set to the string to suffix the prompt.
Default is:
non-root: \"\\[\\e[00m\\]$\"
(resets colour and displays $).
root: \"\\[\\e[00m\\]#\"
(resets colour and displays #).
CDPath - Set the default path for the -f &#38; -u options.
Default is home directory
CDFile - Set the default filename for the -f &#38; -u options.
Default is cdfile
"
cd_version
}
cd_version ()
{
printf "Version: ${VERSION_MAJOR}.${VERSION_MINOR} Date: ${VERSION_DATE}\n"
}
#
# Truncate right.
#
# params:
# p1 - string
# p2 - length to truncate to
#
# returns string in tcd
#
cd_right_trunc ()
{
local tlen=${2}
local plen=${#1}
local str="${1}"
local diff
local filler="&#60;--"
if ${TEST} ${plen} -le ${tlen}
then
tcd="${str}"
else
let diff=${plen}-${tlen}
elen=3
if ${TEST} ${diff} -le 2
then
let elen=${diff}
fi
tlen=-${tlen}
let tlen=${tlen}+${elen}
tcd=${filler:0:elen}${str:tlen}
fi
}
#
# Three versions of do history:
# cd_dohistory - packs history and specials side by side
# cd_dohistoryH - Shows only hstory
# cd_dohistoryS - Shows only specials
#
cd_dohistory ()
{
cd_getrc
${PRINTF} "History:\n"
local -i count=${cd_histcount}
while ${TEST} ${count} -ge 0
do
cd_right_trunc "${CD[count]}" ${cd_lchar}
${PRINTF} "%2d %-${cd_lchar}.${cd_lchar}s " ${count} "${tcd}"
cd_right_trunc "${CDS[count]}" ${cd_rchar}
${PRINTF} "S%d %-${cd_rchar}.${cd_rchar}s\n" ${count} "${tcd}"
count=${count}-1
done
}
cd_dohistoryH ()
{
cd_getrc
${PRINTF} "History:\n"
local -i count=${cd_maxhistory}
while ${TEST} ${count} -ge 0
do
${PRINTF} "${count} %-${cd_flchar}.${cd_flchar}s\n" ${CD[$count]}
count=${count}-1
done
}
cd_dohistoryS ()
{
cd_getrc
${PRINTF} "Specials:\n"
local -i count=${cd_maxspecial}
while ${TEST} ${count} -ge 0
do
${PRINTF} "S${count} %-${cd_flchar}.${cd_flchar}s\n" ${CDS[$count]}
count=${count}-1
done
}
cd_getrc ()
{
cd_flchar=$(stty -a | awk -F \;
'/rows/ { print $2 $3 }' | awk -F \ '{ print $4 }')
if ${TEST} ${cd_flchar} -ne 0
then
cd_lchar=${cd_flchar}/2-5
cd_rchar=${cd_flchar}/2-5
cd_flchar=${cd_flchar}-5
else
cd_flchar=${FLCHAR:=75}
# cd_flchar is used for for the @s &#38; @h history
cd_lchar=${LCHAR:=35}
cd_rchar=${RCHAR:=35}
fi
}
cd_doselection ()
{
local -i nm=0
cd_doflag="TRUE"
if ${TEST} "${CD_MODE}" = "PREV"
then
if ${TEST} -z "$cd_npwd"
then
cd_npwd=0
fi
fi
tm=$(echo "${cd_npwd}" | cut -b 1)
if ${TEST} "${tm}" = "-"
then
pm=$(echo "${cd_npwd}" | cut -b 2)
nm=$(echo "${cd_npwd}" | cut -d $pm -f2)
case "${pm}" in
a) cd_npwd=${cd_sugg[$nm]} ;;
s) cd_npwd="${CDS[$nm]}" ;;
S) cd_npwd="${CDS[$nm]}" ; CDS[$nm]=`pwd` ;;
r) cd_npwd="$2" ; cd_specDir=$nm ; cd_doselection "$1" "$2";;
R) cd_npwd="$2" ; CDS[$nm]=`pwd` ; cd_doselection "$1" "$2";;
esac
fi
if ${TEST} "${cd_npwd}" != "." -a "${cd_npwd}" \
!= ".." -a "${cd_npwd}" -le ${cd_maxhistory} &#62;&#62;/dev/null 2&#62;&#38;1
then
cd_npwd=${CD[$cd_npwd]}
else
case "$cd_npwd" in
@) cd_dohistory ; cd_doflag="FALSE" ;;
@h) cd_dohistoryH ; cd_doflag="FALSE" ;;
@s) cd_dohistoryS ; cd_doflag="FALSE" ;;
-h) cd_hm ; cd_doflag="FALSE" ;;
-H) cd_Hm ; cd_doflag="FALSE" ;;
-f) cd_fsave "SHOW" $2 ; cd_doflag="FALSE" ;;
-u) cd_upload "SHOW" $2 ; cd_doflag="FALSE" ;;
-F) cd_fsave "NOSHOW" $2 ; cd_doflag="FALSE" ;;
-U) cd_upload "NOSHOW" $2 ; cd_doflag="FALSE" ;;
-g) cd_npwd="$2" ;;
-d) cd_chdefm 1; cd_doflag="FALSE" ;;
-D) cd_chdefm 0; cd_doflag="FALSE" ;;
-r) cd_npwd="$2" ; cd_specDir=0 ; cd_doselection "$1" "$2";;
-R) cd_npwd="$2" ; CDS[0]=`pwd` ; cd_doselection "$1" "$2";;
-s) cd_npwd="${CDS[0]}" ;;
-S) cd_npwd="${CDS[0]}" ; CDS[0]=`pwd` ;;
-v) cd_version ; cd_doflag="FALSE";;
esac
fi
}
cd_chdefm ()
{
if ${TEST} "${CD_MODE}" = "PREV"
then
CD_MODE=""
if ${TEST} $1 -eq 1
then
${PRINTF} "${cd_mset}"
fi
else
CD_MODE="PREV"
if ${TEST} $1 -eq 1
then
${PRINTF} "${cd_mnset}"
fi
fi
}
cd_fsave ()
{
local sfile=${CDPath}${2:-"$CDFile"}
if ${TEST} "$1" = "SHOW"
then
${PRINTF} "Saved to %s\n" $sfile
fi
${RM} -f ${sfile}
local -i count=0
while ${TEST} ${count} -le ${cd_maxhistory}
do
echo "CD[$count]=\"${CD[$count]}\"" &#62;&#62; ${sfile}
count=${count}+1
done
count=0
while ${TEST} ${count} -le ${cd_maxspecial}
do
echo "CDS[$count]=\"${CDS[$count]}\"" &#62;&#62; ${sfile}
count=${count}+1
done
}
cd_upload ()
{
local sfile=${CDPath}${2:-"$CDFile"}
if ${TEST} "${1}" = "SHOW"
then
${PRINTF} "Loading from %s\n" ${sfile}
fi
. ${sfile}
}
cd_new ()
{
local -i count
local -i choose=0
cd_npwd="${1}"
cd_specDir=-1
cd_doselection "${1}" "${2}"
if ${TEST} ${cd_doflag} = "TRUE"
then
if ${TEST} "${CD[0]}" != "`pwd`"
then
count=$cd_maxhistory
while ${TEST} $count -gt 0
do
CD[$count]=${CD[$count-1]}
count=${count}-1
done
CD[0]=`pwd`
fi
command cd "${cd_npwd}" 2&#62;/dev/null
if ${TEST} $? -eq 1
then
${PRINTF} "Unknown dir: %s\n" "${cd_npwd}"
local -i ftflag=0
for i in "${cd_npwd}"*
do
if ${TEST} -d "${i}"
then
if ${TEST} ${ftflag} -eq 0
then
${PRINTF} "Suggest:\n"
ftflag=1
fi
${PRINTF} "\t-a${choose} %s\n" "$i"
cd_sugg[$choose]="${i}"
choose=${choose}+1
fi
done
fi
fi
if ${TEST} ${cd_specDir} -ne -1
then
CDS[${cd_specDir}]=`pwd`
fi
if ${TEST} ! -z "${CDL_PROMPTLEN}"
then
cd_right_trunc "${PWD}" ${CDL_PROMPTLEN}
cd_rp=${CDL_PROMPT_PRE}${tcd}${CDL_PROMPT_POST}
export PS1="$(echo -ne ${cd_rp})"
fi
}
#########################################################################
# #
# Initialisation here #
# #
#########################################################################
#
VERSION_MAJOR="1"
VERSION_MINOR="2.1"
VERSION_DATE="24-MAY-2003"
#
alias cd=cd_new
#
# Set up commands
RM=/bin/rm
TEST=test
PRINTF=printf # Use builtin printf
#########################################################################
# #
# Change this to modify the default pre- and post prompt strings. #
# These only come into effect if CDL_PROMPTLEN is set. #
# #
#########################################################################
if ${TEST} ${EUID} -eq 0
then
# CDL_PROMPT_PRE=${CDL_PROMPT_PRE:="$HOSTNAME@"}
CDL_PROMPT_PRE=${CDL_PROMPT_PRE:="\\[\\e[01;31m\\]"} # Root is in red
CDL_PROMPT_POST=${CDL_PROMPT_POST:="\\[\\e[00m\\]#"}
else
CDL_PROMPT_PRE=${CDL_PROMPT_PRE:="\\[\\e[01;34m\\]"} # Users in blue
CDL_PROMPT_POST=${CDL_PROMPT_POST:="\\[\\e[00m\\]$"}
fi
#########################################################################
#
# cd_maxhistory defines the max number of history entries allowed.
typeset -i cd_maxhistory=50
#########################################################################
#
# cd_maxspecial defines the number of special entries.
typeset -i cd_maxspecial=9
#
#
#########################################################################
#
# cd_histcount defines the number of entries displayed in
#+ the history command.
typeset -i cd_histcount=9
#
#########################################################################
export CDPath=${HOME}/
# Change these to use a different #
#+ default path and filename #
export CDFile=${CDFILE:=cdfile} # for the -u and -f commands #
#
#########################################################################
#
typeset -i cd_lchar cd_rchar cd_flchar
# This is the number of chars to allow for the #
cd_flchar=${FLCHAR:=75} #+ cd_flchar is used for for the @s &#38; @h history#
typeset -ax CD CDS
#
cd_mset="\n\tDefault mode is now set - entering cd with no parameters \
has the default action\n\tUse cd -d or -D for cd to go to \
previous directory with no parameters\n"
cd_mnset="\n\tNon-default mode is now set - entering cd with no \
parameters is the same as entering cd 0\n\tUse cd -d or \
-D to change default cd action\n"
# ==================================================================== #
: &#60;&#60;DOCUMENTATION
Written by Phil Braham. Realtime Software Pty Ltd.
Released under GNU license. Free to use. Please pass any modifications
or comments to the author Phil Braham:
realtime@mpx.com.au
=======================================================================
cdll is a replacement for cd and incorporates similar functionality to
the bash pushd and popd commands but is independent of them.
This version of cdll has been tested on Linux using Bash. It will work
on most Linux versions but will probably not work on other shells without
modification.
Introduction
============
cdll allows easy moving about between directories. When changing to a new
directory the current one is automatically put onto a stack. By default
50 entries are kept, but this is configurable. Special directories can be
kept for easy access - by default up to 10, but this is configurable. The
most recent stack entries and the special entries can be easily viewed.
The directory stack and special entries can be saved to, and loaded from,
a file. This allows them to be set up on login, saved before logging out
or changed when moving project to project.
In addition, cdll provides a flexible command prompt facility that allows,
for example, a directory name in colour that is truncated from the left
if it gets too long.
Setting up cdll
===============
Copy cdll to either your local home directory or a central directory
such as /usr/bin (this will require root access).
Copy the file cdfile to your home directory. It will require read and
write access. This a default file that contains a directory stack and
special entries.
To replace the cd command you must add commands to your login script.
The login script is one or more of:
/etc/profile
~/.bash_profile
~/.bash_login
~/.profile
~/.bashrc
/etc/bash.bashrc.local
To setup your login, ~/.bashrc is recommended, for global (and root) setup
add the commands to /etc/bash.bashrc.local
To set up on login, add the command:
. &#60;dir&#62;/cdll
For example if cdll is in your local home directory:
. ~/cdll
If in /usr/bin then:
. /usr/bin/cdll
If you want to use this instead of the buitin cd command then add:
alias cd='cd_new'
We would also recommend the following commands:
alias @='cd_new @'
cd -U
cd -D
If you want to use cdll's prompt facilty then add the following:
CDL_PROMPTLEN=nn
Where nn is a number described below. Initially 99 would be suitable
number.
Thus the script looks something like this:
######################################################################
# CD Setup
######################################################################
CDL_PROMPTLEN=21 # Allow a prompt length of up to 21 characters
. /usr/bin/cdll # Initialise cdll
alias cd='cd_new' # Replace the built in cd command
alias @='cd_new @' # Allow @ at the prompt to display history
cd -U # Upload directories
cd -D # Set default action to non-posix
######################################################################
The full meaning of these commands will become clear later.
There are a couple of caveats. If another program changes the directory
without calling cdll, then the directory won't be put on the stack and
also if the prompt facility is used then this will not be updated. Two
programs that can do this are pushd and popd. To update the prompt and
stack simply enter:
cd .
Note that if the previous entry on the stack is the current directory
then the stack is not updated.
Usage
=====
cd [dir] [0-9] [@[s|h] [-g &#60;dir&#62;] [-d] [-D] [-r&#60;n&#62;]
[dir|0-9] [-R&#60;n&#62;] [&#60;dir&#62;|0-9] [-s&#60;n&#62;] [-S&#60;n&#62;]
[-u] [-U] [-f] [-F] [-h] [-H] [-v]
&#60;dir&#62; Go to directory
0-n Goto previous directory (0 is previous,
1 is last but 1, etc.)
n is up to max history (default is 50)
@ List history and special entries (Usually available as $ @)
@h List history entries
@s List special entries
-g [&#60;dir&#62;] Go to literal name (bypass special names)
This is to allow access to dirs called '0','1','-h' etc
-d Change default action - verbose. (See note)
-D Change default action - silent. (See note)
-s&#60;n&#62; Go to the special entry &#60;n&#62;
-S&#60;n&#62; Go to the special entry &#60;n&#62;
and replace it with the current dir
-r&#60;n&#62; [&#60;dir&#62;] Go to directory &#60;dir&#62;
and then put it on special entry &#60;n&#62;
-R&#60;n&#62; [&#60;dir&#62;] Go to directory &#60;dir&#62;
and put current dir on special entry &#60;n&#62;
-a&#60;n&#62; Alternative suggested directory. See note below.
-f [&#60;file&#62;] File entries to &#60;file&#62;.
-u [&#60;file&#62;] Update entries from &#60;file&#62;.
If no filename supplied then default file (~/cdfile) is used
-F and -U are silent versions
-v Print version number
-h Help
-H Detailed help
Examples
========
These examples assume non-default mode is set (that is, cd with no
parameters will go to the most recent stack directory), that aliases
have been set up for cd and @ as described above and that cd's prompt
facility is active and the prompt length is 21 characters.
/home/phil$ @
# List the entries with the @
History:
# Output of the @ command
.....
# Skipped these entries for brevity
1 /home/phil/ummdev S1 /home/phil/perl
# Most recent two history entries
0 /home/phil/perl/eg S0 /home/phil/umm/ummdev
# and two special entries are shown
/home/phil$ cd /home/phil/utils/Cdll
# Now change directories
/home/phil/utils/Cdll$ @
# Prompt reflects the directory.
History:
# New history
.....
1 /home/phil/perl/eg S1 /home/phil/perl
# History entry 0 has moved to 1
0 /home/phil S0 /home/phil/umm/ummdev
# and the most recent has entered
To go to a history entry:
/home/phil/utils/Cdll$ cd 1
# Go to history entry 1.
/home/phil/perl/eg$
# Current directory is now what was 1
To go to a special entry:
/home/phil/perl/eg$ cd -s1
# Go to special entry 1
/home/phil/umm/ummdev$
# Current directory is S1
To go to a directory called, for example, 1:
/home/phil$ cd -g 1
# -g ignores the special meaning of 1
/home/phil/1$
To put current directory on the special list as S1:
cd -r1 . # OR
cd -R1 . # These have the same effect if the directory is
#+ . (the current directory)
To go to a directory and add it as a special
The directory for -r&#60;n&#62; or -R&#60;n&#62; may be a number.
For example:
$ cd -r3 4 Go to history entry 4 and put it on special entry 3
$ cd -R3 4 Put current dir on the special entry 3 and go to
history entry 4
$ cd -s3 Go to special entry 3
Note that commands R,r,S and s may be used without a number and
refer to 0:
$ cd -s Go to special entry 0
$ cd -S Go to special entry 0 and make special entry 0
current dir
$ cd -r 1 Go to history entry 1 and put it on special entry 0
$ cd -r Go to history entry 0 and put it on special entry 0
Alternative suggested directories:
If a directory is not found, then CD will suggest any
possibilities. These are directories starting with the same letters
and if any are found they are listed prefixed with -a&#60;n&#62;
where &#60;n&#62; is a number. It's possible to go to the directory
by entering cd -a&#60;n&#62; on the command line.
Use cd -d or -D to change default cd action. cd -H will show
current action.
The history entries (0-n) are stored in the environment variables
CD[0] - CD[n]
Similarly the special directories S0 - 9 are in the environment
variable CDS[0] - CDS[9]
and may be accessed from the command line, for example:
ls -l ${CDS[3]}
cat ${CD[8]}/file.txt
The default pathname for the -f and -u commands is ~
The default filename for the -f and -u commands is cdfile
Configuration
=============
The following environment variables can be set:
CDL_PROMPTLEN - Set to the length of prompt you require.
Prompt string is set to the right characters of the current
directory. If not set, then prompt is left unchanged. Note
that this is the number of characters that the directory is
shortened to, not the total characters in the prompt.
CDL_PROMPT_PRE - Set to the string to prefix the prompt.
Default is:
non-root: "\\[\\e[01;34m\\]" (sets colour to blue).
root: "\\[\\e[01;31m\\]" (sets colour to red).
CDL_PROMPT_POST - Set to the string to suffix the prompt.
Default is:
non-root: "\\[\\e[00m\\]$"
(resets colour and displays $).
root: "\\[\\e[00m\\]#"
(resets colour and displays #).
Note:
CDL_PROMPT_PRE &#38; _POST only t
CDPath - Set the default path for the -f &#38; -u options.
Default is home directory
CDFile - Set the default filename for the -f &#38; -u options.
Default is cdfile
There are three variables defined in the file cdll which control the
number of entries stored or displayed. They are in the section labeled
'Initialisation here' towards the end of the file.
cd_maxhistory - The number of history entries stored.
Default is 50.
cd_maxspecial - The number of special entries allowed.
Default is 9.
cd_histcount - The number of history and special entries
displayed. Default is 9.
Note that cd_maxspecial should be &#62;= cd_histcount to avoid displaying
special entries that can't be set.
Version: 1.2.1 Date: 24-MAY-2003
DOCUMENTATION</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="SOUNDCARDON"
></A
><P
><B
>Example A-34. A soundcard setup script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# soundcard-on.sh
# Script author: Mkarcher
# http://www.thinkwiki.org/wiki ...
# /Script_for_configuring_the_CS4239_sound_chip_in_PnP_mode
# ABS Guide author made minor changes and added comments.
# Couldn't contact script author to ask for permission to use, but ...
#+ the script was released under the FDL,
#+ so its use here should be both legal and ethical.
# Sound-via-pnp-script for Thinkpad 600E
#+ and possibly other computers with onboard CS4239/CS4610
#+ that do not work with the PCI driver
#+ and are not recognized by the PnP code of snd-cs4236.
# Also for some 770-series Thinkpads, such as the 770x.
# Run as root user, of course.
#
# These are old and very obsolete laptop computers,
#+ but this particular script is very instructive,
#+ as it shows how to set up and hack device files.
# Search for sound card pnp device:
for dev in /sys/bus/pnp/devices/*
do
grep CSC0100 $dev/id &#62; /dev/null &#38;&#38; WSSDEV=$dev
grep CSC0110 $dev/id &#62; /dev/null &#38;&#38; CTLDEV=$dev
done
# On 770x:
# WSSDEV = /sys/bus/pnp/devices/00:07
# CTLDEV = /sys/bus/pnp/devices/00:06
# These are symbolic links to /sys/devices/pnp0/ ...
# Activate devices:
# Thinkpad boots with devices disabled unless "fast boot" is turned off
#+ (in BIOS).
echo activate &#62; $WSSDEV/resources
echo activate &#62; $CTLDEV/resources
# Parse resource settings.
{ read # Discard "state = active" (see below).
read bla port1
read bla port2
read bla port3
read bla irq
read bla dma1
read bla dma2
# The "bla's" are labels in the first field: "io," "state," etc.
# These are discarded.
# Hack: with PnPBIOS: ports are: port1: WSS, port2:
#+ OPL, port3: sb (unneeded)
# with ACPI-PnP:ports are: port1: OPL, port2: sb, port3: WSS
# (ACPI bios seems to be wrong here, the PnP-card-code in snd-cs4236.c
#+ uses the PnPBIOS port order)
# Detect port order using the fixed OPL port as reference.
if [ ${port2%%-*} = 0x388 ]
# ^^^^ Strip out everything following hyphen in port address.
# So, if port1 is 0x530-0x537
#+ we're left with 0x530 -- the start address of the port.
then
# PnPBIOS: usual order
port=${port1%%-*}
oplport=${port2%%-*}
else
# ACPI: mixed-up order
port=${port3%%-*}
oplport=${port1%%-*}
fi
} &#60; $WSSDEV/resources
# To see what's going on here:
# ---------------------------
# cat /sys/devices/pnp0/00:07/resources
#
# state = active
# io 0x530-0x537
# io 0x388-0x38b
# io 0x220-0x233
# irq 5
# dma 1
# dma 0
# ^^^ "bla" labels in first field (discarded).
{ read # Discard first line, as above.
read bla port1
cport=${port1%%-*}
# ^^^^
# Just want _start_ address of port.
} &#60; $CTLDEV/resources
# Load the module:
modprobe --ignore-install snd-cs4236 port=$port cport=$cport\
fm_port=$oplport irq=$irq dma1=$dma1 dma2=$dma2 isapnp=0 index=0
# See the modprobe manpage.
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="FINDSPLIT0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="FINDSPLIT"
></A
><P
><B
>Example A-35. Locating split paragraphs in a text file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# find-splitpara.sh
# Finds split paragraphs in a text file,
#+ and tags the line numbers.
ARGCOUNT=1 # Expect one arg.
OFF=0 # Flag states.
ON=1
E_WRONGARGS=85
file="$1" # Target filename.
lineno=1 # Line number. Start at 1.
Flag=$OFF # Blank line flag.
if [ $# -ne "$ARGCOUNT" ]
then
echo "Usage: `basename $0` FILENAME"
exit $E_WRONGARGS
fi
file_read () # Scan file for pattern, then print line.
{
while read line
do
if [[ "$line" =~ ^[a-z] &#38;&#38; $Flag -eq $ON ]]
then # Line begins with lowercase character, following blank line.
echo -n "$lineno:: "
echo "$line"
fi
if [[ "$line" =~ ^$ ]]
then # If blank line,
Flag=$ON #+ set flag.
else
Flag=$OFF
fi
((lineno++))
done
} &#60; $file # Redirect file into function's stdin.
file_read
exit $?
# ----------------------------------------------------------------
This is line one of an example paragraph, bla, bla, bla.
This is line two, and line three should follow on next line, but
there is a blank line separating the two parts of the paragraph.
# ----------------------------------------------------------------
Running this script on a file containing the above paragraph
yields:
4:: there is a blank line separating the two parts of the paragraph.
There will be additional output for all the other split paragraphs
in the target file.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="INSERTIONSORT0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="INSERTIONSORT"
></A
><P
><B
>Example A-36. Insertion sort</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# insertion-sort.bash: Insertion sort implementation in Bash
# Heavy use of Bash array features:
#+ (string) slicing, merging, etc
# URL: http://www.lugmen.org.ar/~jjo/jjotip/insertion-sort.bash.d
#+ /insertion-sort.bash.sh
#
# Author: JuanJo Ciarlante &#60;jjo@irrigacion.gov.ar&#62;
# Lightly reformatted by ABS Guide author.
# License: GPLv2
# Used in ABS Guide with author's permission (thanks!).
#
# Test with: ./insertion-sort.bash -t
# Or: bash insertion-sort.bash -t
# The following *doesn't* work:
# sh insertion-sort.bash -t
# Why not? Hint: which Bash-specific features are disabled
#+ when running a script by 'sh script.sh'?
#
: ${DEBUG:=0} # Debug, override with: DEBUG=1 ./scriptname . . .
# Parameter substitution -- set DEBUG to 0 if not previously set.
# Global array: "list"
typeset -a list
# Load whitespace-separated numbers from stdin.
if [ "$1" = "-t" ]; then
DEBUG=1
read -a list &#60; &#60;( od -Ad -w24 -t u2 /dev/urandom ) # Random list.
# ^ ^ process substition
else
read -a list
fi
numelem=${#list[*]}
# Shows the list, marking the element whose index is $1
#+ by surrounding it with the two chars passed as $2.
# Whole line prefixed with $3.
showlist()
{
echo "$3"${list[@]:0:$1} ${2:0:1}${list[$1]}${2:1:1} ${list[@]:$1+1};
}
# Loop _pivot_ -- from second element to end of list.
for(( i=1; i&#60;numelem; i++ )) do
((DEBUG))&#38;&#38;showlist i "[]" " "
# From current _pivot_, back to first element.
for(( j=i; j; j-- )) do
# Search for the 1st elem. less than current "pivot" . . .
[[ "${list[j-1]}" -le "${list[i]}" ]] &#38;&#38; break
done
(( i==j )) &#38;&#38; continue ## No insertion was needed for this element.
# . . . Move list[i] (pivot) to the left of list[j]:
list=(${list[@]:0:j} ${list[i]} ${list[j]}\
# {0,j-1} {i} {j}
${list[@]:j+1:i-(j+1)} ${list[@]:i+1})
# {j+1,i-1} {i+1,last}
((DEBUG))&#38;&#38;showlist j "&#60;&#62;" "*"
done
echo
echo "------"
echo $'Result:\n'${list[@]}
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="STDDEV"
></A
><P
><B
>Example A-37. Standard Deviation</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# sd.sh: Standard Deviation
# The Standard Deviation indicates how consistent a set of data is.
# It shows to what extent the individual data points deviate from the
#+ arithmetic mean, i.e., how much they "bounce around" (or cluster).
# It is essentially the average deviation-distance of the
#+ data points from the mean.
# =========================================================== #
# To calculate the Standard Deviation:
#
# 1 Find the arithmetic mean (average) of all the data points.
# 2 Subtract each data point from the arithmetic mean,
# and square that difference.
# 3 Add all of the individual difference-squares in # 2.
# 4 Divide the sum in # 3 by the number of data points.
# This is known as the "variance."
# 5 The square root of # 4 gives the Standard Deviation.
# =========================================================== #
count=0 # Number of data points; global.
SC=9 # Scale to be used by bc. Nine decimal places.
E_DATAFILE=90 # Data file error.
# ----------------- Set data file ---------------------
if [ ! -z "$1" ] # Specify filename as cmd-line arg?
then
datafile="$1" # ASCII text file,
else #+ one (numerical) data point per line!
datafile=sample.dat
fi # See example data file, below.
if [ ! -e "$datafile" ]
then
echo "\""$datafile"\" does not exist!"
exit $E_DATAFILE
fi
# -----------------------------------------------------
arith_mean ()
{
local rt=0 # Running total.
local am=0 # Arithmetic mean.
local ct=0 # Number of data points.
while read value # Read one data point at a time.
do
rt=$(echo "scale=$SC; $rt + $value" | bc)
(( ct++ ))
done
am=$(echo "scale=$SC; $rt / $ct" | bc)
echo $am; return $ct # This function "returns" TWO values!
# Caution: This little trick will not work if $ct &#62; 255!
# To handle a larger number of data points,
#+ simply comment out the "return $ct" above.
} &#60;"$datafile" # Feed in data file.
sd ()
{
mean1=$1 # Arithmetic mean (passed to function).
n=$2 # How many data points.
sum2=0 # Sum of squared differences ("variance").
avg2=0 # Average of $sum2.
sdev=0 # Standard Deviation.
while read value # Read one line at a time.
do
diff=$(echo "scale=$SC; $mean1 - $value" | bc)
# Difference between arith. mean and data point.
dif2=$(echo "scale=$SC; $diff * $diff" | bc) # Squared.
sum2=$(echo "scale=$SC; $sum2 + $dif2" | bc) # Sum of squares.
done
avg2=$(echo "scale=$SC; $sum2 / $n" | bc) # Avg. of sum of squares.
sdev=$(echo "scale=$SC; sqrt($avg2)" | bc) # Square root =
echo $sdev # Standard Deviation.
} &#60;"$datafile" # Rewinds data file.
# ======================================================= #
mean=$(arith_mean); count=$? # Two returns from function!
std_dev=$(sd $mean $count)
echo
echo "Number of data points in \""$datafile"\" = $count"
echo "Arithmetic mean (average) = $mean"
echo "Standard Deviation = $std_dev"
echo
# ======================================================= #
exit
# This script could stand some drastic streamlining,
#+ but not at the cost of reduced legibility, please.
# ++++++++++++++++++++++++++++++++++++++++ #
# A sample data file (sample1.dat):
# 18.35
# 19.0
# 18.88
# 18.91
# 18.64
# $ sh sd.sh sample1.dat
# Number of data points in "sample1.dat" = 5
# Arithmetic mean (average) = 18.756000000
# Standard Deviation = .235338054
# ++++++++++++++++++++++++++++++++++++++++ #</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="PADSW"
></A
><P
><B
>Example A-38. A <I
CLASS="FIRSTTERM"
>pad</I
> file generator for shareware
authors</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# pad.sh
#######################################################
# PAD (xml) file creator
#+ Written by Mendel Cooper &#60;thegrendel.abs@gmail.com&#62;.
#+ Released to the Public Domain.
#
# Generates a "PAD" descriptor file for shareware
#+ packages, according to the specifications
#+ of the ASP.
# http://www.asp-shareware.org/pad
#######################################################
# Accepts (optional) save filename as a command-line argument.
if [ -n "$1" ]
then
savefile=$1
else
savefile=save_file.xml # Default save_file name.
fi
# ===== PAD file headers =====
HDR1="&#60;?xml version=\"1.0\" encoding=\"Windows-1252\" ?&#62;"
HDR2="&#60;XML_DIZ_INFO&#62;"
HDR3="&#60;MASTER_PAD_VERSION_INFO&#62;"
HDR4="\t&#60;MASTER_PAD_VERSION&#62;1.15&#60;/MASTER_PAD_VERSION&#62;"
HDR5="\t&#60;MASTER_PAD_INFO&#62;Portable Application Description, or PAD
for short, is a data set that is used by shareware authors to
disseminate information to anyone interested in their software products.
To find out more go to http://www.asp-shareware.org/pad&#60;/MASTER_PAD_INFO&#62;"
HDR6="&#60;/MASTER_PAD_VERSION_INFO&#62;"
# ============================
fill_in ()
{
if [ -z "$2" ]
then
echo -n "$1? " # Get user input.
else
echo -n "$1 $2? " # Additional query?
fi
read var # May paste to fill in field.
# This shows how flexible "read" can be.
if [ -z "$var" ]
then
echo -e "\t\t&#60;$1 /&#62;" &#62;&#62;$savefile # Indent with 2 tabs.
return
else
echo -e "\t\t&#60;$1&#62;$var&#60;/$1&#62;" &#62;&#62;$savefile
return ${#var} # Return length of input string.
fi
}
check_field_length () # Check length of program description fields.
{
# $1 = maximum field length
# $2 = actual field length
if [ "$2" -gt "$1" ]
then
echo "Warning: Maximum field length of $1 characters exceeded!"
fi
}
clear # Clear screen.
echo "PAD File Creator"
echo "--- ---- -------"
echo
# Write File Headers to file.
echo $HDR1 &#62;$savefile
echo $HDR2 &#62;&#62;$savefile
echo $HDR3 &#62;&#62;$savefile
echo -e $HDR4 &#62;&#62;$savefile
echo -e $HDR5 &#62;&#62;$savefile
echo $HDR6 &#62;&#62;$savefile
# Company_Info
echo "COMPANY INFO"
CO_HDR="Company_Info"
echo "&#60;$CO_HDR&#62;" &#62;&#62;$savefile
fill_in Company_Name
fill_in Address_1
fill_in Address_2
fill_in City_Town
fill_in State_Province
fill_in Zip_Postal_Code
fill_in Country
# If applicable:
# fill_in ASP_Member "[Y/N]"
# fill_in ASP_Member_Number
# fill_in ESC_Member "[Y/N]"
fill_in Company_WebSite_URL
clear # Clear screen between sections.
# Contact_Info
echo "CONTACT INFO"
CONTACT_HDR="Contact_Info"
echo "&#60;$CONTACT_HDR&#62;" &#62;&#62;$savefile
fill_in Author_First_Name
fill_in Author_Last_Name
fill_in Author_Email
fill_in Contact_First_Name
fill_in Contact_Last_Name
fill_in Contact_Email
echo -e "\t&#60;/$CONTACT_HDR&#62;" &#62;&#62;$savefile
# END Contact_Info
clear
# Support_Info
echo "SUPPORT INFO"
SUPPORT_HDR="Support_Info"
echo "&#60;$SUPPORT_HDR&#62;" &#62;&#62;$savefile
fill_in Sales_Email
fill_in Support_Email
fill_in General_Email
fill_in Sales_Phone
fill_in Support_Phone
fill_in General_Phone
fill_in Fax_Phone
echo -e "\t&#60;/$SUPPORT_HDR&#62;" &#62;&#62;$savefile
# END Support_Info
echo "&#60;/$CO_HDR&#62;" &#62;&#62;$savefile
# END Company_Info
clear
# Program_Info
echo "PROGRAM INFO"
PROGRAM_HDR="Program_Info"
echo "&#60;$PROGRAM_HDR&#62;" &#62;&#62;$savefile
fill_in Program_Name
fill_in Program_Version
fill_in Program_Release_Month
fill_in Program_Release_Day
fill_in Program_Release_Year
fill_in Program_Cost_Dollars
fill_in Program_Cost_Other
fill_in Program_Type "[Shareware/Freeware/GPL]"
fill_in Program_Release_Status "[Beta, Major Upgrade, etc.]"
fill_in Program_Install_Support
fill_in Program_OS_Support "[Win9x/Win2k/Linux/etc.]"
fill_in Program_Language "[English/Spanish/etc.]"
echo; echo
# File_Info
echo "FILE INFO"
FILEINFO_HDR="File_Info"
echo "&#60;$FILEINFO_HDR&#62;" &#62;&#62;$savefile
fill_in Filename_Versioned
fill_in Filename_Previous
fill_in Filename_Generic
fill_in Filename_Long
fill_in File_Size_Bytes
fill_in File_Size_K
fill_in File_Size_MB
echo -e "\t&#60;/$FILEINFO_HDR&#62;" &#62;&#62;$savefile
# END File_Info
clear
# Expire_Info
echo "EXPIRE INFO"
EXPIRE_HDR="Expire_Info"
echo "&#60;$EXPIRE_HDR&#62;" &#62;&#62;$savefile
fill_in Has_Expire_Info "Y/N"
fill_in Expire_Count
fill_in Expire_Based_On
fill_in Expire_Other_Info
fill_in Expire_Month
fill_in Expire_Day
fill_in Expire_Year
echo -e "\t&#60;/$EXPIRE_HDR&#62;" &#62;&#62;$savefile
# END Expire_Info
clear
# More Program_Info
echo "ADDITIONAL PROGRAM INFO"
fill_in Program_Change_Info
fill_in Program_Specific_Category
fill_in Program_Categories
fill_in Includes_JAVA_VM "[Y/N]"
fill_in Includes_VB_Runtime "[Y/N]"
fill_in Includes_DirectX "[Y/N]"
# END More Program_Info
echo "&#60;/$PROGRAM_HDR&#62;" &#62;&#62;$savefile
# END Program_Info
clear
# Program Description
echo "PROGRAM DESCRIPTIONS"
PROGDESC_HDR="Program_Descriptions"
echo "&#60;$PROGDESC_HDR&#62;" &#62;&#62;$savefile
LANG="English"
echo "&#60;$LANG&#62;" &#62;&#62;$savefile
fill_in Keywords "[comma + space separated]"
echo
echo "45, 80, 250, 450, 2000 word program descriptions"
echo "(may cut and paste into field)"
# It would be highly appropriate to compose the following
#+ "Char_Desc" fields with a text editor,
#+ then cut-and-paste the text into the answer fields.
echo
echo " |---------------45 characters---------------|"
fill_in Char_Desc_45
check_field_length 45 "$?"
echo
fill_in Char_Desc_80
check_field_length 80 "$?"
fill_in Char_Desc_250
check_field_length 250 "$?"
fill_in Char_Desc_450
fill_in Char_Desc_2000
echo "&#60;/$LANG&#62;" &#62;&#62;$savefile
echo "&#60;/$PROGDESC_HDR&#62;" &#62;&#62;$savefile
# END Program Description
clear
echo "Done."; echo; echo
echo "Save file is: \""$savefile"\""
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="MANED"
></A
><P
><B
>Example A-39. A <I
CLASS="FIRSTTERM"
>man page</I
> editor</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# maned.sh
# A rudimentary man page editor
# Version: 0.1 (Alpha, probably buggy)
# Author: Mendel Cooper &#60;thegrendel.abs@gmail.com&#62;
# Reldate: 16 June 2008
# License: GPL3
savefile= # Global, used in multiple functions.
E_NOINPUT=90 # User input missing (error). May or may not be critical.
# =========== Markup Tags ============ #
TopHeader=".TH"
NameHeader=".SH NAME"
SyntaxHeader=".SH SYNTAX"
SynopsisHeader=".SH SYNOPSIS"
InstallationHeader=".SH INSTALLATION"
DescHeader=".SH DESCRIPTION"
OptHeader=".SH OPTIONS"
FilesHeader=".SH FILES"
EnvHeader=".SH ENVIRONMENT"
AuthHeader=".SH AUTHOR"
BugsHeader=".SH BUGS"
SeeAlsoHeader=".SH SEE ALSO"
BOLD=".B"
# Add more tags, as needed.
# See groff docs for markup meanings.
# ==================================== #
start ()
{
clear # Clear screen.
echo "ManEd"
echo "-----"
echo
echo "Simple man page creator"
echo "Author: Mendel Cooper"
echo "License: GPL3"
echo; echo; echo
}
progname ()
{
echo -n "Program name? "
read name
echo -n "Manpage section? [Hit RETURN for default (\"1\") ] "
read section
if [ -z "$section" ]
then
section=1 # Most man pages are in section 1.
fi
if [ -n "$name" ]
then
savefile=""$name"."$section"" # Filename suffix = section.
echo -n "$1 " &#62;&#62;$savefile
name1=$(echo "$name" | tr a-z A-Z) # Change to uppercase,
#+ per man page convention.
echo -n "$name1" &#62;&#62;$savefile
else
echo "Error! No input." # Mandatory input.
exit $E_NOINPUT # Critical!
# Exercise: The script-abort if no filename input is a bit clumsy.
# Rewrite this section so a default filename is used
#+ if no input.
fi
echo -n " \"$section\""&#62;&#62;$savefile # Append, always append.
echo -n "Version? "
read ver
echo -n " \"Version $ver \""&#62;&#62;$savefile
echo &#62;&#62;$savefile
echo -n "Short description [0 - 5 words]? "
read sdesc
echo "$NameHeader"&#62;&#62;$savefile
echo ""$BOLD" "$name""&#62;&#62;$savefile
echo "\- "$sdesc""&#62;&#62;$savefile
}
fill_in ()
{ # This function more or less copied from "pad.sh" script.
echo -n "$2? " # Get user input.
read var # May paste (a single line only!) to fill in field.
if [ -n "$var" ]
then
echo "$1 " &#62;&#62;$savefile
echo -n "$var" &#62;&#62;$savefile
else # Don't append empty field to file.
return $E_NOINPUT # Not critical here.
fi
echo &#62;&#62;$savefile
}
end ()
{
clear
echo -n "Would you like to view the saved man page (y/n)? "
read ans
if [ "$ans" = "n" -o "$ans" = "N" ]; then exit; fi
exec less "$savefile" # Exit script and hand off control to "less" ...
#+ ... which formats for viewing man page source.
}
# ---------------------------------------- #
start
progname "$TopHeader"
fill_in "$SynopsisHeader" "Synopsis"
fill_in "$DescHeader" "Long description"
# May paste in *single line* of text.
fill_in "$OptHeader" "Options"
fill_in "$FilesHeader" "Files"
fill_in "$AuthHeader" "Author"
fill_in "$BugsHeader" "Bugs"
fill_in "$SeeAlsoHeader" "See also"
# fill_in "$OtherHeader" ... as necessary.
end # ... exit not needed.
# ---------------------------------------- #
# Note that the generated man page will usually
#+ require manual fine-tuning with a text editor.
# However, it's a distinct improvement upon
#+ writing man source from scratch
#+ or even editing a blank man page template.
# The main deficiency of the script is that it permits
#+ pasting only a single text line into the input fields.
# This may be a long, cobbled-together line, which groff
# will automatically wrap and hyphenate.
# However, if you want multiple (newline-separated) paragraphs,
#+ these must be inserted by manual text editing on the
#+ script-generated man page.
# Exercise (difficult): Fix this!
# This script is not nearly as elaborate as the
#+ full-featured "manedit" package
#+ http://freshmeat.net/projects/manedit/
#+ but it's much easier to use.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="PETALS"
></A
><P
><B
>Example A-40. Petals Around the Rose</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash -i
# petals.sh
#########################################################################
# Petals Around the Rose #
# #
# Version 0.1 Created by Serghey Rodin #
# Version 0.2 Modded by ABS Guide Author #
# #
# License: GPL3 #
# Used in ABS Guide with permission. #
# ##################################################################### #
hits=0 # Correct guesses.
WIN=6 # Mastered the game.
ALMOST=5 # One short of mastery.
EXIT=exit # Give up early?
RANDOM=$$ # Seeds the random number generator from PID of script.
# Bones (ASCII graphics for dice)
bone1[1]="| |"
bone1[2]="| o |"
bone1[3]="| o |"
bone1[4]="| o o |"
bone1[5]="| o o |"
bone1[6]="| o o |"
bone2[1]="| o |"
bone2[2]="| |"
bone2[3]="| o |"
bone2[4]="| |"
bone2[5]="| o |"
bone2[6]="| o o |"
bone3[1]="| |"
bone3[2]="| o |"
bone3[3]="| o |"
bone3[4]="| o o |"
bone3[5]="| o o |"
bone3[6]="| o o |"
bone="+---------+"
# Functions
instructions () {
clear
echo -n "Do you need instructions? (y/n) "; read ans
if [ "$ans" = "y" -o "$ans" = "Y" ]; then
clear
echo -e '\E[34;47m' # Blue type.
# "cat document"
cat &#60;&#60;INSTRUCTIONSZZZ
The name of the game is Petals Around the Rose,
and that name is significant.
Five dice will roll and you must guess the "answer" for each roll.
It will be zero or an even number.
After your guess, you will be told the answer for the roll, but . . .
that's ALL the information you will get.
Six consecutive correct guesses admits you to the
Fellowship of the Rose.
INSTRUCTIONSZZZ
echo -e "\033[0m" # Turn off blue.
else clear
fi
}
fortune ()
{
RANGE=7
FLOOR=0
number=0
while [ "$number" -le $FLOOR ]
do
number=$RANDOM
let "number %= $RANGE" # 1 - 6.
done
return $number
}
throw () { # Calculate each individual die.
fortune; B1=$?
fortune; B2=$?
fortune; B3=$?
fortune; B4=$?
fortune; B5=$?
calc () { # Function embedded within a function!
case "$1" in
3 ) rose=2;;
5 ) rose=4;;
* ) rose=0;;
esac # Simplified algorithm.
# Doesn't really get to the heart of the matter.
return $rose
}
answer=0
calc "$B1"; answer=$(expr $answer + $(echo $?))
calc "$B2"; answer=$(expr $answer + $(echo $?))
calc "$B3"; answer=$(expr $answer + $(echo $?))
calc "$B4"; answer=$(expr $answer + $(echo $?))
calc "$B5"; answer=$(expr $answer + $(echo $?))
}
game ()
{ # Generate graphic display of dice throw.
throw
echo -e "\033[1m" # Bold.
echo -e "\n"
echo -e "$bone\t$bone\t$bone\t$bone\t$bone"
echo -e \
"${bone1[$B1]}\t${bone1[$B2]}\t${bone1[$B3]}\t${bone1[$B4]}\t${bone1[$B5]}"
echo -e \
"${bone2[$B1]}\t${bone2[$B2]}\t${bone2[$B3]}\t${bone2[$B4]}\t${bone2[$B5]}"
echo -e \
"${bone3[$B1]}\t${bone3[$B2]}\t${bone3[$B3]}\t${bone3[$B4]}\t${bone3[$B5]}"
echo -e "$bone\t$bone\t$bone\t$bone\t$bone"
echo -e "\n\n\t\t"
echo -e "\033[0m" # Turn off bold.
echo -n "There are how many petals around the rose? "
}
# ============================================================== #
instructions
while [ "$petal" != "$EXIT" ] # Main loop.
do
game
read petal
echo "$petal" | grep [0-9] &#62;/dev/null # Filter response for digit.
# Otherwise just roll dice again.
if [ "$?" -eq 0 ] # If-loop #1.
then
if [ "$petal" == "$answer" ]; then # If-loop #2.
echo -e "\nCorrect. There are $petal petals around the rose.\n"
(( hits++ ))
if [ "$hits" -eq "$WIN" ]; then # If-loop #3.
echo -e '\E[31;47m' # Red type.
echo -e "\033[1m" # Bold.
echo "You have unraveled the mystery of the Rose Petals!"
echo "Welcome to the Fellowship of the Rose!!!"
echo "(You are herewith sworn to secrecy.)"; echo
echo -e "\033[0m" # Turn off red &#38; bold.
break # Exit!
else echo "You have $hits correct so far."; echo
if [ "$hits" -eq "$ALMOST" ]; then
echo "Just one more gets you to the heart of the mystery!"; echo
fi
fi # Close if-loop #3.
else
echo -e "\nWrong. There are $answer petals around the rose.\n"
hits=0 # Reset number of correct guesses.
fi # Close if-loop #2.
echo -n "Hit ENTER for the next roll, or type \"exit\" to end. "
read
if [ "$REPLY" = "$EXIT" ]; then exit
fi
fi # Close if-loop #1.
clear
done # End of main (while) loop.
###
exit $?
# Resources:
# ---------
# 1) http://en.wikipedia.org/wiki/Petals_Around_the_Rose
# (Wikipedia entry.)
# 2) http://www.borrett.id.au/computing/petals-bg.htm
# (How Bill Gates coped with the Petals Around the Rose challenge.)</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="QKY"
></A
><P
><B
>Example A-41. Quacky: a Perquackey-type word game</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# qky.sh
##############################################################
# QUACKEY: a somewhat simplified version of Perquackey [TM]. #
# #
# Author: Mendel Cooper &#60;thegrendel.abs@gmail.com&#62; #
# version 0.1.02 03 May, 2008 #
# License: GPL3 #
##############################################################
WLIST=/usr/share/dict/word.lst
# ^^^^^^^^ Word list file found here.
# ASCII word list, one word per line, UNIX format.
# A suggested list is the script author's "yawl" word list package.
# http://bash.deta.in/yawl-0.3.2.tar.gz
# or
# http://ibiblio.org/pub/Linux/libs/yawl-0.3.2.tar.gz
NONCONS=0 # Word not constructable from letter set.
CONS=1 # Constructable.
SUCCESS=0
NG=1
FAILURE=''
NULL=0 # Zero out value of letter (if found).
MINWLEN=3 # Minimum word length.
MAXCAT=5 # Maximum number of words in a given category.
PENALTY=200 # General-purpose penalty for unacceptable words.
total=
E_DUP=70 # Duplicate word error.
TIMEOUT=10 # Time for word input.
NVLET=10 # 10 letters for non-vulnerable.
VULET=13 # 13 letters for vulnerable (not yet implemented!).
declare -a Words
declare -a Status
declare -a Score=( 0 0 0 0 0 0 0 0 0 0 0 )
letters=( a n s r t m l k p r b c i d s i d z e w u e t f
e y e r e f e g t g h h i t r s c i t i d i j a t a o l a
m n a n o v n w o s e l n o s p a q e e r a b r s a o d s
t g t i t l u e u v n e o x y m r k )
# Letter distribution table shamelessly borrowed from "Wordy" game,
#+ ca. 1992, written by a certain fine fellow named Mendel Cooper.
declare -a LS
numelements=${#letters[@]}
randseed="$1"
instructions ()
{
clear
echo "Welcome to QUACKEY, the anagramming word construction game."; echo
echo -n "Do you need instructions? (y/n) "; read ans
if [ "$ans" = "y" -o "$ans" = "Y" ]; then
clear
echo -e '\E[31;47m' # Red foreground. '\E[34;47m' for blue.
cat &#60;&#60;INSTRUCTION1
QUACKEY is a variant of Perquackey [TM].
The rules are the same, but the scoring is simplified
and plurals of previously played words are allowed.
"Vulnerable" play is not yet implemented,
but it is otherwise feature-complete.
As the game begins, the player gets 10 letters.
The object is to construct valid dictionary words
of at least 3-letter length from the letterset.
Each word-length category
-- 3-letter, 4-letter, 5-letter, ... --
fills up with the fifth word entered,
and no further words in that category are accepted.
The penalty for too-short (two-letter), duplicate, unconstructable,
and invalid (not in dictionary) words is -200. The same penalty applies
to attempts to enter a word in a filled-up category.
INSTRUCTION1
echo -n "Hit ENTER for next page of instructions. "; read az1
cat &#60;&#60;INSTRUCTION2
The scoring mostly corresponds to classic Perquackey:
The first 3-letter word scores 60, plus 10 for each additional one.
The first 4-letter word scores 120, plus 20 for each additional one.
The first 5-letter word scores 200, plus 50 for each additional one.
The first 6-letter word scores 300, plus 100 for each additional one.
The first 7-letter word scores 500, plus 150 for each additional one.
The first 8-letter word scores 750, plus 250 for each additional one.
The first 9-letter word scores 1000, plus 500 for each additional one.
The first 10-letter word scores 2000, plus 2000 for each additional one.
Category completion bonuses are:
3-letter words 100
4-letter words 200
5-letter words 400
6-letter words 800
7-letter words 2000
8-letter words 10000
This is a simplification of the absurdly baroque Perquackey bonus
scoring system.
INSTRUCTION2
echo -n "Hit ENTER for final page of instructions. "; read az1
cat &#60;&#60;INSTRUCTION3
Hitting just ENTER for a word entry ends the game.
Individual word entry is timed to a maximum of 10 seconds.
*** Timing out on an entry ends the game. ***
Aside from that, the game is untimed.
--------------------------------------------------
Game statistics are automatically saved to a file.
--------------------------------------------------
For competitive ("duplicate") play, a previous letterset
may be duplicated by repeating the script's random seed,
command-line parameter \$1.
For example, "qky 7633" specifies the letterset
c a d i f r h u s k ...
INSTRUCTION3
echo; echo -n "Hit ENTER to begin game. "; read az1
echo -e "\033[0m" # Turn off red.
else clear
fi
clear
}
seed_random ()
{ # Seed random number generator.
if [ -n "$randseed" ] # Can specify random seed.
then #+ for play in competitive mode.
# RANDOM="$randseed"
echo "RANDOM seed set to "$randseed""
else
randseed="$$" # Or get random seed from process ID.
echo "RANDOM seed not specified, set to Process ID of script ($$)."
fi
RANDOM="$randseed"
echo
}
get_letset ()
{
element=0
echo -n "Letterset:"
for lset in $(seq $NVLET)
do # Pick random letters to fill out letterset.
LS[element]="${letters[$((RANDOM%numelements))]}"
((element++))
done
echo
echo "${LS[@]}"
}
add_word ()
{
wrd="$1"
local idx=0
Status[0]=""
Status[3]=""
Status[4]=""
while [ "${Words[idx]}" != '' ]
do
if [ "${Words[idx]}" = "$wrd" ]
then
Status[3]="Duplicate-word-PENALTY"
let "Score[0]= 0 - $PENALTY"
let "Score[1]-=$PENALTY"
return $E_DUP
fi
((idx++))
done
Words[idx]="$wrd"
get_score
}
get_score()
{
local wlen=0
local score=0
local bonus=0
local first_word=0
local add_word=0
local numwords=0
wlen=${#wrd}
numwords=${Score[wlen]}
Score[2]=0
Status[4]="" # Initialize "bonus" to 0.
case "$wlen" in
3) first_word=60
add_word=10;;
4) first_word=120
add_word=20;;
5) first_word=200
add_word=50;;
6) first_word=300
add_word=100;;
7) first_word=500
add_word=150;;
8) first_word=750
add_word=250;;
9) first_word=1000
add_word=500;;
10) first_word=2000
add_word=2000;; # This category modified from original rules!
esac
((Score[wlen]++))
if [ ${Score[wlen]} -eq $MAXCAT ]
then # Category completion bonus scoring simplified!
case $wlen in
3 ) bonus=100;;
4 ) bonus=200;;
5 ) bonus=400;;
6 ) bonus=800;;
7 ) bonus=2000;;
8 ) bonus=10000;;
esac # Needn't worry about 9's and 10's.
Status[4]="Category-$wlen-completion***BONUS***"
Score[2]=$bonus
else
Status[4]="" # Erase it.
fi
let "score = $first_word + $add_word * $numwords"
if [ "$numwords" -eq 0 ]
then
Score[0]=$score
else
Score[0]=$add_word
fi # All this to distinguish last-word score
#+ from total running score.
let "Score[1] += ${Score[0]}"
let "Score[1] += ${Score[2]}"
}
get_word ()
{
local wrd=''
read -t $TIMEOUT wrd # Timed read.
echo $wrd
}
is_constructable ()
{ # This is the most complex and difficult-to-write function.
local -a local_LS=( "${LS[@]}" ) # Local copy of letter set.
local is_found=0
local idx=0
local pos
local strlen
local local_word=( "$1" )
strlen=${#local_word}
while [ "$idx" -lt "$strlen" ]
do
is_found=$(expr index "${local_LS[*]}" "${local_word:idx:1}")
if [ "$is_found" -eq "$NONCONS" ] # Not constructable!
then
echo "$FAILURE"; return
else
((pos = ($is_found - 1) / 2)) # Compensate for spaces betw. letters!
local_LS[pos]=$NULL # Zero out used letters.
((idx++)) # Bump index.
fi
done
echo "$SUCCESS"
return
}
is_valid ()
{ # Surprisingly easy to check if word in dictionary ...
fgrep -qw "$1" "$WLIST" # ... courtesy of 'grep' ...
echo $?
}
check_word ()
{
if [ -z "$1" ]
then
return
fi
Status[1]=""
Status[2]=""
Status[3]=""
Status[4]=""
iscons=$(is_constructable "$1")
if [ "$iscons" ]
then
Status[1]="constructable"
v=$(is_valid "$1")
if [ "$v" -eq "$SUCCESS" ]
then
Status[2]="valid"
strlen=${#1}
if [ ${Score[strlen]} -eq "$MAXCAT" ] # Category full!
then
Status[3]="Category-$strlen-overflow-PENALTY"
return $NG
fi
case "$strlen" in
1 | 2 )
Status[3]="Two-letter-word-PENALTY"
return $NG;;
* )
Status[3]=""
return $SUCCESS;;
esac
else
Status[3]="Not-valid-PENALTY"
return $NG
fi
else
Status[3]="Not-constructable-PENALTY"
return $NG
fi
### FIXME: Streamline the above code block.
}
display_words ()
{
local idx=0
local wlen0
clear
echo "Letterset: ${LS[@]}"
echo "Threes: Fours: Fives: Sixes: Sevens: Eights:"
echo "------------------------------------------------------------"
while [ "${Words[idx]}" != '' ]
do
wlen0=${#Words[idx]}
case "$wlen0" in
3) ;;
4) echo -n " " ;;
5) echo -n " " ;;
6) echo -n " " ;;
7) echo -n " " ;;
8) echo -n " " ;;
esac
echo "${Words[idx]}"
((idx++))
done
### FIXME: The word display is pretty crude.
}
play ()
{
word="Start game" # Dummy word, to start ...
while [ "$word" ] # If player just hits return (null word),
do #+ then game ends.
echo "$word: "${Status[@]}""
echo -n "Last score: [${Score[0]}] TOTAL score: [${Score[1]}]: Next word: "
total=${Score[1]}
word=$(get_word)
check_word "$word"
if [ "$?" -eq "$SUCCESS" ]
then
add_word "$word"
else
let "Score[0]= 0 - $PENALTY"
let "Score[1]-=$PENALTY"
fi
display_words
done # Exit game.
### FIXME: The play () function calls too many other functions.
### This verges on "spaghetti code" !!!
}
end_of_game ()
{ # Save and display stats.
#######################Autosave##########################
savefile=qky.save.$$
# ^^ PID of script
echo `date` &#62;&#62; $savefile
echo "Letterset # $randseed (random seed) "&#62;&#62; $savefile
echo -n "Letterset: " &#62;&#62; $savefile
echo "${LS[@]}" &#62;&#62; $savefile
echo "---------" &#62;&#62; $savefile
echo "Words constructed:" &#62;&#62; $savefile
echo "${Words[@]}" &#62;&#62; $savefile
echo &#62;&#62; $savefile
echo "Score: $total" &#62;&#62; $savefile
echo "Statistics for this round saved in \""$savefile"\""
#########################################################
echo "Score for this round: $total"
echo "Words: ${Words[@]}"
}
# ---------#
instructions
seed_random
get_letset
play
end_of_game
# ---------#
exit $?
# TODO:
#
# 1) Clean up code!
# 2) Prettify the display_words () function (maybe with widgets?).
# 3) Improve the time-out ... maybe change to untimed entry,
#+ but with a time limit for the overall round.
# 4) An on-screen countdown timer would be nice.
# 5) Implement "vulnerable" mode of play for compatibility with classic
#+ version of the game.
# 6) Improve save-to-file capability (and maybe make it optional).
# 7) Fix bugs!!!
# For more info, reference:
# http://bash.deta.in/qky.README.html</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="NIM"
></A
><P
><B
>Example A-42. Nim</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# nim.sh: Game of Nim
# Author: Mendel Cooper
# Reldate: 15 July 2008
# License: GPL3
ROWS=5 # Five rows of pegs (or matchsticks).
WON=91 # Exit codes to keep track of wins/losses.
LOST=92 # Possibly useful if running in batch mode.
QUIT=99
peg_msg= # Peg/Pegs?
Rows=( 0 5 4 3 2 1 ) # Array holding play info.
# ${Rows[0]} holds total number of pegs, updated after each turn.
# Other array elements hold number of pegs in corresponding row.
instructions ()
{
clear
tput bold
echo "Welcome to the game of Nim."; echo
echo -n "Do you need instructions? (y/n) "; read ans
if [ "$ans" = "y" -o "$ans" = "Y" ]; then
clear
echo -e '\E[33;41m' # Yellow fg., over red bg.; bold.
cat &#60;&#60;INSTRUCTIONS
Nim is a game with roots in the distant past.
This particular variant starts with five rows of pegs.
1: | | | | |
2: | | | |
3: | | |
4: | |
5: |
The number at the left identifies the row.
The human player moves first, and alternates turns with the bot.
A turn consists of removing at least one peg from a single row.
It is permissable to remove ALL the pegs from a row.
For example, in row 2, above, the player can remove 1, 2, 3, or 4 pegs.
The player who removes the last peg loses.
The strategy consists of trying to be the one who removes
the next-to-last peg(s), leaving the loser with the final peg.
To exit the game early, hit ENTER during your turn.
INSTRUCTIONS
echo; echo -n "Hit ENTER to begin game. "; read azx
echo -e "\033[0m" # Restore display.
else tput sgr0; clear
fi
clear
}
tally_up ()
{
let "Rows[0] = ${Rows[1]} + ${Rows[2]} + ${Rows[3]} + ${Rows[4]} + \
${Rows[5]}" # Add up how many pegs remaining.
}
display ()
{
index=1 # Start with top row.
echo
while [ "$index" -le "$ROWS" ]
do
p=${Rows[index]}
echo -n "$index: " # Show row number.
# ------------------------------------------------
# Two concurrent inner loops.
indent=$index
while [ "$indent" -gt 0 ]
do
echo -n " " # Staggered rows.
((indent--)) # Spacing between pegs.
done
while [ "$p" -gt 0 ]
do
echo -n "| "
((p--))
done
# -----------------------------------------------
echo
((index++))
done
tally_up
rp=${Rows[0]}
if [ "$rp" -eq 1 ]
then
peg_msg=peg
final_msg="Game over."
else # Game not yet over . . .
peg_msg=pegs
final_msg="" # . . . So "final message" is blank.
fi
echo " $rp $peg_msg remaining."
echo " "$final_msg""
echo
}
player_move ()
{
echo "Your move:"
echo -n "Which row? "
while read idx
do # Validity check, etc.
if [ -z "$idx" ] # Hitting return quits.
then
echo "Premature exit."; echo
tput sgr0 # Restore display.
exit $QUIT
fi
if [ "$idx" -gt "$ROWS" -o "$idx" -lt 1 ] # Bounds check.
then
echo "Invalid row number!"
echo -n "Which row? "
else
break
fi
# TODO:
# Add check for non-numeric input.
# Also, script crashes on input outside of range of long double.
# Fix this.
done
echo -n "Remove how many? "
while read num
do # Validity check.
if [ -z "$num" ]
then
echo "Premature exit."; echo
tput sgr0 # Restore display.
exit $QUIT
fi
if [ "$num" -gt ${Rows[idx]} -o "$num" -lt 1 ]
then
echo "Cannot remove $num!"
echo -n "Remove how many? "
else
break
fi
done
# TODO:
# Add check for non-numeric input.
# Also, script crashes on input outside of range of long double.
# Fix this.
let "Rows[idx] -= $num"
display
tally_up
if [ ${Rows[0]} -eq 1 ]
then
echo " Human wins!"
echo " Congratulations!"
tput sgr0 # Restore display.
echo
exit $WON
fi
if [ ${Rows[0]} -eq 0 ]
then # Snatching defeat from the jaws of victory . . .
echo " Fool!"
echo " You just removed the last peg!"
echo " Bot wins!"
tput sgr0 # Restore display.
echo
exit $LOST
fi
}
bot_move ()
{
row_b=0
while [[ $row_b -eq 0 || ${Rows[row_b]} -eq 0 ]]
do
row_b=$RANDOM # Choose random row.
let "row_b %= $ROWS"
done
num_b=0
r0=${Rows[row_b]}
if [ "$r0" -eq 1 ]
then
num_b=1
else
let "num_b = $r0 - 1"
# Leave only a single peg in the row.
fi # Not a very strong strategy,
#+ but probably a bit better than totally random.
let "Rows[row_b] -= $num_b"
echo -n "Bot: "
echo "Removing from row $row_b ... "
if [ "$num_b" -eq 1 ]
then
peg_msg=peg
else
peg_msg=pegs
fi
echo " $num_b $peg_msg."
display
tally_up
if [ ${Rows[0]} -eq 1 ]
then
echo " Bot wins!"
tput sgr0 # Restore display.
exit $WON
fi
}
# ================================================== #
instructions # If human player needs them . . .
tput bold # Bold characters for easier viewing.
display # Show game board.
while [ true ] # Main loop.
do # Alternate human and bot turns.
player_move
bot_move
done
# ================================================== #
# Exercise:
# --------
# Improve the bot's strategy.
# There is, in fact, a Nim strategy that can force a win.
# See the Wikipedia article on Nim: http://en.wikipedia.org/wiki/Nim
# Recode the bot to use this strategy (rather difficult).
# Curiosities:
# -----------
# Nim played a prominent role in Alain Resnais' 1961 New Wave film,
#+ Last Year at Marienbad.
#
# In 1978, Leo Christopherson wrote an animated version of Nim,
#+ Android Nim, for the TRS-80 Model I.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="STOPWATCH"
></A
><P
><B
>Example A-43. A command-line stopwatch</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/sh
# sw.sh
# A command-line Stopwatch
# Author: P<>draig Brady
# http://www.pixelbeat.org/scripts/sw
# (Minor reformatting by ABS Guide author.)
# Used in ABS Guide with script author's permission.
# Notes:
# This script starts a few processes per lap, in addition to
# the shell loop processing, so the assumption is made that
# this takes an insignificant amount of time compared to
# the response time of humans (~.1s) (or the keyboard
# interrupt rate (~.05s)).
# '?' for splits must be entered twice if characters
# (erroneously) entered before it (on the same line).
# '?' since not generating a signal may be slightly delayed
# on heavily loaded systems.
# Lap timings on ubuntu may be slightly delayed due to:
# https://bugs.launchpad.net/bugs/62511
# Changes:
# V1.0, 23 Aug 2005, Initial release
# V1.1, 26 Jul 2007, Allow both splits and laps from single invocation.
# Only start timer after a key is pressed.
# Indicate lap number
# Cache programs at startup so there is less error
# due to startup delays.
# V1.2, 01 Aug 2007, Work around `date` commands that don't have
# nanoseconds.
# Use stty to change interrupt keys to space for
# laps etc.
# Ignore other input as it causes problems.
# V1.3, 01 Aug 2007, Testing release.
# V1.4, 02 Aug 2007, Various tweaks to get working under ubuntu
# and Mac OS X.
# V1.5, 27 Jun 2008, set LANG=C as got vague bug report about it.
export LANG=C
ulimit -c 0 # No coredumps from SIGQUIT.
trap '' TSTP # Ignore Ctrl-Z just in case.
save_tty=`stty -g` &#38;&#38; trap "stty $save_tty" EXIT # Restore tty on exit.
stty quit ' ' # Space for laps rather than Ctrl-\.
stty eof '?' # ? for splits rather than Ctrl-D.
stty -echo # Don't echo input.
cache_progs() {
stty &#62; /dev/null
date &#62; /dev/null
grep . &#60; /dev/null
(echo "import time" | python) 2&#62; /dev/null
bc &#60; /dev/null
sed '' &#60; /dev/null
printf '1' &#62; /dev/null
/usr/bin/time false 2&#62; /dev/null
cat &#60; /dev/null
}
cache_progs # To minimise startup delay.
date +%s.%N | grep -qF 'N' &#38;&#38; use_python=1 # If `date` lacks nanoseconds.
now() {
if [ "$use_python" ]; then
echo "import time; print time.time()" 2&#62;/dev/null | python
else
printf "%.2f" `date +%s.%N`
fi
}
fmt_seconds() {
seconds=$1
mins=`echo $seconds/60 | bc`
if [ "$mins" != "0" ]; then
seconds=`echo "$seconds - ($mins*60)" | bc`
echo "$mins:$seconds"
else
echo "$seconds"
fi
}
total() {
end=`now`
total=`echo "$end - $start" | bc`
fmt_seconds $total
}
stop() {
[ "$lapped" ] &#38;&#38; lap "$laptime" "display"
total
exit
}
lap() {
laptime=`echo "$1" | sed -n 's/.*real[^0-9.]*\(.*\)/\1/p'`
[ ! "$laptime" -o "$laptime" = "0.00" ] &#38;&#38; return
# Signals too frequent.
laptotal=`echo $laptime+0$laptotal | bc`
if [ "$2" = "display" ]; then
lapcount=`echo 0$lapcount+1 | bc`
laptime=`fmt_seconds $laptotal`
echo $laptime "($lapcount)"
lapped="true"
laptotal="0"
fi
}
echo -n "Space for lap | ? for split | Ctrl-C to stop | Space to start..."&#62;&#38;2
while true; do
trap true INT QUIT # Set signal handlers.
laptime=`/usr/bin/time -p 2&#62;&#38;1 cat &#62;/dev/null`
ret=$?
trap '' INT QUIT # Ignore signals within this script.
if [ $ret -eq 1 -o $ret -eq 2 -o $ret -eq 130 ]; then # SIGINT = stop
[ ! "$start" ] &#38;&#38; { echo &#62;&#38;2; exit; }
stop
elif [ $ret -eq 3 -o $ret -eq 131 ]; then # SIGQUIT = lap
if [ ! "$start" ]; then
start=`now` || exit 1
echo &#62;&#38;2
continue
fi
lap "$laptime" "display"
else # eof = split
[ ! "$start" ] &#38;&#38; continue
total
lap "$laptime" # Update laptotal.
fi
done
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="HOMEWORK"
></A
><P
><B
>Example A-44. An all-purpose shell scripting homework assignment solution</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# homework.sh: All-purpose homework assignment solution.
# Author: M. Leo Cooper
# If you substitute your own name as author, then it is plagiarism,
#+ possibly a lesser sin than cheating on your homework!
# License: Public Domain
# This script may be turned in to your instructor
#+ in fulfillment of ALL shell scripting homework assignments.
# It's sparsely commented, but you, the student, can easily remedy that.
# The script author repudiates all responsibility!
DLA=1
P1=2
P2=4
P3=7
PP1=0
PP2=8
MAXL=9
E_LZY=99
declare -a L
L[0]="3 4 0 17 29 8 13 18 19 17 20 2 19 14 17 28"
L[1]="8 29 12 14 18 19 29 4 12 15 7 0 19 8 2 0 11 11 24 29 17 4 6 17 4 19"
L[2]="29 19 7 0 19 29 8 29 7 0 21 4 29 13 4 6 11 4 2 19 4 3"
L[3]="19 14 29 2 14 12 15 11 4 19 4 29 19 7 8 18 29"
L[4]="18 2 7 14 14 11 22 14 17 10 29 0 18 18 8 6 13 12 4 13 19 26"
L[5]="15 11 4 0 18 4 29 0 2 2 4 15 19 29 12 24 29 7 20 12 1 11 4 29"
L[6]="4 23 2 20 18 4 29 14 5 29 4 6 17 4 6 8 14 20 18 29"
L[7]="11 0 25 8 13 4 18 18 27"
L[8]="0 13 3 29 6 17 0 3 4 29 12 4 29 0 2 2 14 17 3 8 13 6 11 24 26"
L[9]="19 7 0 13 10 29 24 14 20 26"
declare -a \
alph=( A B C D E F G H I J K L M N O P Q R S T U V W X Y Z . , : ' ' )
pt_lt ()
{
echo -n "${alph[$1]}"
echo -n -e "\a"
sleep $DLA
}
b_r ()
{
echo -e '\E[31;48m\033[1m'
}
cr ()
{
echo -e "\a"
sleep $DLA
}
restore ()
{
echo -e '\033[0m' # Bold off.
tput sgr0 # Normal.
}
p_l ()
{
for ltr in $1
do
pt_lt "$ltr"
done
}
# ----------------------
b_r
for i in $(seq 0 $MAXL)
do
p_l "${L[i]}"
if [[ "$i" -eq "$P1" || "$i" -eq "$P2" || "$i" -eq "$P3" ]]
then
cr
elif [[ "$i" -eq "$PP1" || "$i" -eq "$PP2" ]]
then
cr; cr
fi
done
restore
# ----------------------
echo
exit $E_LZY
# A typical example of an obfuscated script that is difficult
#+ to understand, and frustrating to maintain.
# In your career as a sysadmin, you'll run into these critters
#+ all too often.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="KTOUR0"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="KTOUR"
></A
><P
><B
>Example A-45. The Knight's Tour</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ktour.sh
# author: mendel cooper
# reldate: 12 Jan 2009
# license: public domain
# (Not much sense GPLing something that's pretty much in the common
#+ domain anyhow.)
###################################################################
# The Knight's Tour, a classic problem. #
# ===================================== #
# The knight must move onto every square of the chess board, #
# but cannot revisit any square he has already visited. #
# #
# And just why is Sir Knight unwelcome for a return visit? #
# Could it be that he has a habit of partying into the wee hours #
#+ of the morning? #
# Possibly he leaves pizza crusts in the bed, empty beer bottles #
#+ all over the floor, and clogs the plumbing. . . . #
# #
# ------------------------------------------------------------- #
# #
# Usage: ktour.sh [start-square] [stupid] #
# #
# Note that start-square can be a square number #
#+ in the range 0 - 63 ... or #
# a square designator in conventional chess notation, #
# such as a1, f5, h3, etc. #
# #
# If start-square-number not supplied, #
#+ then starts on a random square somewhere on the board. #
# #
# "stupid" as second parameter sets the stupid strategy. #
# #
# Examples: #
# ktour.sh 23 starts on square #23 (h3) #
# ktour.sh g6 stupid starts on square #46, #
# using "stupid" (non-Warnsdorff) strategy. #
###################################################################
DEBUG= # Set this to echo debugging info to stdout.
SUCCESS=0
FAIL=99
BADMOVE=-999
FAILURE=1
LINELEN=21 # How many moves to display per line.
# ---------------------------------------- #
# Board array params
ROWS=8 # 8 x 8 board.
COLS=8
let "SQUARES = $ROWS * $COLS"
let "MAX = $SQUARES - 1"
MIN=0
# 64 squares on board, indexed from 0 to 63.
VISITED=1
UNVISITED=-1
UNVSYM="##"
# ---------------------------------------- #
# Global variables.
startpos= # Starting position (square #, 0 - 63).
currpos= # Current position.
movenum= # Move number.
CRITPOS=37 # Have to patch for f5 starting position!
declare -i board
# Use a one-dimensional array to simulate a two-dimensional one.
# This can make life difficult and result in ugly kludges; see below.
declare -i moves # Offsets from current knight position.
initialize_board ()
{
local idx
for idx in {0..63}
do
board[$idx]=$UNVISITED
done
}
print_board ()
{
local idx
echo " _____________________________________"
for row in {7..0} # Reverse order of rows ...
do #+ so it prints in chessboard order.
let "rownum = $row + 1" # Start numbering rows at 1.
echo -n "$rownum |" # Mark board edge with border and
for column in {0..7} #+ "algebraic notation."
do
let "idx = $ROWS*$row + $column"
if [ ${board[idx]} -eq $UNVISITED ]
then
echo -n "$UNVSYM " ##
else # Mark square with move number.
printf "%02d " "${board[idx]}"; echo -n " "
fi
done
echo -e -n "\b\b\b|" # \b is a backspace.
echo # -e enables echoing escaped chars.
done
echo " -------------------------------------"
echo " a b c d e f g h"
}
failure()
{ # Whine, then bail out.
echo
print_board
echo
echo " Waah!!! Ran out of squares to move to!"
echo -n " Knight's Tour attempt ended"
echo " on $(to_algebraic $currpos) [square #$currpos]"
echo " after just $movenum moves!"
echo
exit $FAIL
}
xlat_coords () # Translate x/y coordinates to board position
{ #+ (board-array element #).
# For user input of starting board position as x/y coords.
# This function not used in initial release of ktour.sh.
# May be used in an updated version, for compatibility with
#+ standard implementation of the Knight's Tour in C, Python, etc.
if [ -z "$1" -o -z "$2" ]
then
return $FAIL
fi
local xc=$1
local yc=$2
let "board_index = $xc * $ROWS + yc"
if [ $board_index -lt $MIN -o $board_index -gt $MAX ]
then
return $FAIL # Strayed off the board!
else
return $board_index
fi
}
to_algebraic () # Translate board position (board-array element #)
{ #+ to standard algebraic notation used by chess players.
if [ -z "$1" ]
then
return $FAIL
fi
local element_no=$1 # Numerical board position.
local col_arr=( a b c d e f g h )
local row_arr=( 1 2 3 4 5 6 7 8 )
let "row_no = $element_no / $ROWS"
let "col_no = $element_no % $ROWS"
t1=${col_arr[col_no]}; t2=${row_arr[row_no]}
local apos=$t1$t2 # Concatenate.
echo $apos
}
from_algebraic () # Translate standard algebraic chess notation
{ #+ to numerical board position (board-array element #).
# Or recognize numerical input &#38; return it unchanged.
if [ -z "$1" ]
then
return $FAIL
fi # If no command-line arg, then will default to random start pos.
local ix
local ix_count=0
local b_index # Board index [0-63]
local alpos="$1"
arow=${alpos:0:1} # position = 0, length = 1
acol=${alpos:1:1}
if [[ $arow =~ [[:digit:]] ]] # Numerical input?
then # POSIX char class
if [[ $acol =~ [[:alpha:]] ]] # Number followed by a letter? Illegal!
then return $FAIL
else if [ $alpos -gt $MAX ] # Off board?
then return $FAIL
else return $alpos # Return digit(s) unchanged . . .
fi #+ if within range.
fi
fi
if [[ $acol -eq $MIN || $acol -gt $ROWS ]]
then # Outside of range 1 - 8?
return $FAIL
fi
for ix in a b c d e f g h
do # Convert column letter to column number.
if [ "$arow" = "$ix" ]
then
break
fi
((ix_count++)) # Find index count.
done
((acol--)) # Decrementing converts to zero-based array.
let "b_index = $ix_count + $acol * $ROWS"
if [ $b_index -gt $MAX ] # Off board?
then
return $FAIL
fi
return $b_index
}
generate_moves () # Calculate all valid knight moves,
{ #+ relative to current position ($1),
#+ and store in ${moves} array.
local kt_hop=1 # One square :: short leg of knight move.
local kt_skip=2 # Two squares :: long leg of knight move.
local valmov=0 # Valid moves.
local row_pos; let "row_pos = $1 % $COLS"
let "move1 = -$kt_skip + $ROWS" # 2 sideways to-the-left, 1 up
if [[ `expr $row_pos - $kt_skip` -lt $MIN ]] # An ugly, ugly kludge!
then # Can't move off board.
move1=$BADMOVE # Not even temporarily.
else
((valmov++))
fi
let "move2 = -$kt_hop + $kt_skip * $ROWS" # 1 sideways to-the-left, 2 up
if [[ `expr $row_pos - $kt_hop` -lt $MIN ]] # Kludge continued ...
then
move2=$BADMOVE
else
((valmov++))
fi
let "move3 = $kt_hop + $kt_skip * $ROWS" # 1 sideways to-the-right, 2 up
if [[ `expr $row_pos + $kt_hop` -ge $COLS ]]
then
move3=$BADMOVE
else
((valmov++))
fi
let "move4 = $kt_skip + $ROWS" # 2 sideways to-the-right, 1 up
if [[ `expr $row_pos + $kt_skip` -ge $COLS ]]
then
move4=$BADMOVE
else
((valmov++))
fi
let "move5 = $kt_skip - $ROWS" # 2 sideways to-the-right, 1 dn
if [[ `expr $row_pos + $kt_skip` -ge $COLS ]]
then
move5=$BADMOVE
else
((valmov++))
fi
let "move6 = $kt_hop - $kt_skip * $ROWS" # 1 sideways to-the-right, 2 dn
if [[ `expr $row_pos + $kt_hop` -ge $COLS ]]
then
move6=$BADMOVE
else
((valmov++))
fi
let "move7 = -$kt_hop - $kt_skip * $ROWS" # 1 sideways to-the-left, 2 dn
if [[ `expr $row_pos - $kt_hop` -lt $MIN ]]
then
move7=$BADMOVE
else
((valmov++))
fi
let "move8 = -$kt_skip - $ROWS" # 2 sideways to-the-left, 1 dn
if [[ `expr $row_pos - $kt_skip` -lt $MIN ]]
then
move8=$BADMOVE
else
((valmov++))
fi # There must be a better way to do this.
local m=( $valmov $move1 $move2 $move3 $move4 $move5 $move6 $move7 $move8 )
# ${moves[0]} = number of valid moves.
# ${moves[1]} ... ${moves[8]} = possible moves.
echo "${m[*]}" # Elements of array to stdout for capture in a var.
}
is_on_board () # Is position actually on the board?
{
if [[ "$1" -lt "$MIN" || "$1" -gt "$MAX" ]]
then
return $FAILURE
else
return $SUCCESS
fi
}
do_move () # Move the knight!
{
local valid_moves=0
local aapos
currposl="$1"
lmin=$ROWS
iex=0
squarel=
mpm=
mov=
declare -a p_moves
########################## DECIDE-MOVE #############################
if [ $startpos -ne $CRITPOS ]
then # CRITPOS = square #37
decide_move
else # Needs a special patch for startpos=37 !!!
decide_move_patched # Why this particular move and no other ???
fi
####################################################################
(( ++movenum )) # Increment move count.
let "square = $currposl + ${moves[iex]}"
################## DEBUG ###############
if [ "$DEBUG" ]
then debug # Echo debugging information.
fi
##############################################
if [[ "$square" -gt $MAX || "$square" -lt $MIN ||
${board[square]} -ne $UNVISITED ]]
then
(( --movenum )) # Decrement move count,
echo "RAN OUT OF SQUARES!!!" #+ since previous one was invalid.
return $FAIL
fi
board[square]=$movenum
currpos=$square # Update current position.
((valid_moves++)); # moves[0]=$valid_moves
aapos=$(to_algebraic $square)
echo -n "$aapos "
test $(( $Moves % $LINELEN )) -eq 0 &#38;&#38; echo
# Print LINELEN=21 moves per line. A valid tour shows 3 complete lines.
return $valid_moves # Found a square to move to!
}
do_move_stupid() # Dingbat algorithm,
{ #+ courtesy of script author, *not* Warnsdorff.
local valid_moves=0
local movloc
local squareloc
local aapos
local cposloc="$1"
for movloc in {1..8}
do # Move to first-found unvisited square.
let "squareloc = $cposloc + ${moves[movloc]}"
is_on_board $squareloc
if [ $? -eq $SUCCESS ] &#38;&#38; [ ${board[squareloc]} -eq $UNVISITED ]
then # Add conditions to above if-test to improve algorithm.
(( ++movenum ))
board[squareloc]=$movenum
currpos=$squareloc # Update current position.
((valid_moves++)); # moves[0]=$valid_moves
aapos=$(to_algebraic $squareloc)
echo -n "$aapos "
test $(( $Moves % $LINELEN )) -eq 0 &#38;&#38; echo # Print 21 moves/line.
return $valid_moves # Found a square to move to!
fi
done
return $FAIL
# If no square found in all 8 loop iterations,
#+ then Knight's Tour attempt ends in failure.
# Dingbat algorithm will typically fail after about 30 - 40 moves,
#+ but executes _much_ faster than Warnsdorff's in do_move() function.
}
decide_move () # Which move will we make?
{ # But, fails on startpos=37 !!!
for mov in {1..8}
do
let "squarel = $currposl + ${moves[mov]}"
is_on_board $squarel
if [[ $? -eq $SUCCESS &#38;&#38; ${board[squarel]} -eq $UNVISITED ]]
then # Find accessible square with least possible future moves.
# This is Warnsdorff's algorithm.
# What happens is that the knight wanders toward the outer edge
#+ of the board, then pretty much spirals inward.
# Given two or more possible moves with same value of
#+ least-possible-future-moves, this implementation chooses
#+ the _first_ of those moves.
# This means that there is not necessarily a unique solution
#+ for any given starting position.
possible_moves $squarel
mpm=$?
p_moves[mov]=$mpm
if [ $mpm -lt $lmin ] # If less than previous minimum ...
then # ^^
lmin=$mpm # Update minimum.
iex=$mov # Save index.
fi
fi
done
}
decide_move_patched () # Decide which move to make,
{ # ^^^^^^^ #+ but only if startpos=37 !!!
for mov in {1..8}
do
let "squarel = $currposl + ${moves[mov]}"
is_on_board $squarel
if [[ $? -eq $SUCCESS &#38;&#38; ${board[squarel]} -eq $UNVISITED ]]
then
possible_moves $squarel
mpm=$?
p_moves[mov]=$mpm
if [ $mpm -le $lmin ] # If less-than-or equal to prev. minimum!
then # ^^
lmin=$mpm
iex=$mov
fi
fi
done # There has to be a better way to do this.
}
possible_moves () # Calculate number of possible moves,
{ #+ given the current position.
if [ -z "$1" ]
then
return $FAIL
fi
local curr_pos=$1
local valid_movl=0
local icx=0
local movl
local sq
declare -a movesloc
movesloc=( $(generate_moves $curr_pos) )
for movl in {1..8}
do
let "sq = $curr_pos + ${movesloc[movl]}"
is_on_board $sq
if [ $? -eq $SUCCESS ] &#38;&#38; [ ${board[sq]} -eq $UNVISITED ]
then
((valid_movl++));
fi
done
return $valid_movl # Found a square to move to!
}
strategy ()
{
echo
if [ -n "$STUPID" ]
then
for Moves in {1..63}
do
cposl=$1
moves=( $(generate_moves $currpos) )
do_move_stupid "$currpos"
if [ $? -eq $FAIL ]
then
failure
fi
done
fi
# Don't need an "else" clause here,
#+ because Stupid Strategy will always fail and exit!
for Moves in {1..63}
do
cposl=$1
moves=( $(generate_moves $currpos) )
do_move "$currpos"
if [ $? -eq $FAIL ]
then
failure
fi
done
# Could have condensed above two do-loops into a single one,
echo #+ but this would have slowed execution.
print_board
echo
echo "Knight's Tour ends on $(to_algebraic $currpos) [square #$currpos]."
return $SUCCESS
}
debug ()
{ # Enable this by setting DEBUG=1 near beginning of script.
local n
echo "================================="
echo " At move number $movenum:"
echo " *** possible moves = $mpm ***"
# echo "### square = $square ###"
echo "lmin = $lmin"
echo "${moves[@]}"
for n in {1..8}
do
echo -n "($n):${p_moves[n]} "
done
echo
echo "iex = $iex :: moves[iex] = ${moves[iex]}"
echo "square = $square"
echo "================================="
echo
} # Gives pretty complete status after ea. move.
# =============================================================== #
# int main () {
from_algebraic "$1"
startpos=$?
if [ "$startpos" -eq "$FAIL" ] # Okay even if no $1.
then # ^^^^^^^^^^^ Okay even if input -lt 0.
echo "No starting square specified (or illegal input)."
let "startpos = $RANDOM % $SQUARES" # 0 - 63 permissable range.
fi
if [ "$2" = "stupid" ]
then
STUPID=1
echo -n " ### Stupid Strategy ###"
else
STUPID=''
echo -n " *** Warnsdorff's Algorithm ***"
fi
initialize_board
movenum=0
board[startpos]=$movenum # Mark each board square with move number.
currpos=$startpos
algpos=$(to_algebraic $startpos)
echo; echo "Starting from $algpos [square #$startpos] ..."; echo
echo -n "Moves:"
strategy "$currpos"
echo
exit 0 # return 0;
# } # End of main() pseudo-function.
# =============================================================== #
# Exercises:
# ---------
#
# 1) Extend this example to a 10 x 10 board or larger.
# 2) Improve the "stupid strategy" by modifying the
# do_move_stupid function.
# Hint: Prevent straying into corner squares in early moves
# (the exact opposite of Warnsdorff's algorithm!).
# 3) This script could stand considerable improvement and
# streamlining, especially in the poorly-written
# generate_moves() function
# and in the DECIDE-MOVE patch in the do_move() function.
# Must figure out why standard algorithm fails for startpos=37 ...
#+ but _not_ on any other, including symmetrical startpos=26.
# Possibly, when calculating possible moves, counts the move back
#+ to the originating square. If so, it might be a relatively easy fix.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="MSQUARE"
></A
><P
><B
>Example A-46. Magic Squares</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# msquare.sh
# Magic Square generator (odd-order squares only!)
# Author: mendel cooper
# reldate: 19 Jan. 2009
# License: Public Domain
# A C-program by the very talented Kwon Young Shin inspired this script.
# http://user.chollian.net/~brainstm/MagicSquare.htm
# Definition: A "magic square" is a two-dimensional array
# of integers in which all the rows, columns,
# and *long* diagonals add up to the same number.
# Being "square," the array has the same number
# of rows and columns. That number is the "order."
# An example of a magic square of order 3 is:
# 8 1 6
# 3 5 7
# 4 9 2
# All the rows, columns, and the two long diagonals add up to 15.
# Globals
EVEN=2
MAXSIZE=31 # 31 rows x 31 cols.
E_usage=90 # Invocation error.
dimension=
declare -i square
usage_message ()
{
echo "Usage: $0 order"
echo " ... where \"order\" (square size) is an ODD integer"
echo " in the range 3 - 31."
# Actually works for squares up to order 159,
#+ but large squares will not display pretty-printed in a term window.
# Try increasing MAXSIZE, above.
exit $E_usage
}
calculate () # Here's where the actual work gets done.
{
local row col index dimadj j k cell_val=1
dimension=$1
let "dimadj = $dimension * 3"; let "dimadj /= 2" # x 1.5, then truncate.
for ((j=0; j &#60; dimension; j++))
do
for ((k=0; k &#60; dimension; k++))
do # Calculate indices, then convert to 1-dim. array index.
# Bash doesn't support multidimensional arrays. Pity.
let "col = $k - $j + $dimadj"; let "col %= $dimension"
let "row = $j * 2 - $k + $dimension"; let "row %= $dimension"
let "index = $row*($dimension) + $col"
square[$index]=cell_val; ((cell_val++))
done
done
} # Plain math, visualization not required.
print_square () # Output square, one row at a time.
{
local row col idx d1
let "d1 = $dimension - 1" # Adjust for zero-indexed array.
for row in $(seq 0 $d1)
do
for col in $(seq 0 $d1)
do
let "idx = $row * $dimension + $col"
printf "%3d " "${square[idx]}"; echo -n " "
done # Displays up to 13th order neatly in 80-column term window.
echo # Newline after each row.
done
}
#################################################
if [[ -z "$1" ]] || [[ "$1" -gt $MAXSIZE ]]
then
usage_message
fi
let "test_even = $1 % $EVEN"
if [ $test_even -eq 0 ]
then # Can't handle even-order squares.
usage_message
fi
calculate $1
print_square # echo "${square[@]}" # DEBUG
exit $?
#################################################
# Exercises:
# ---------
# 1) Add a function to calculate the sum of each row, column,
# and *long* diagonal. The sums must match.
# This is the "magic constant" of that particular order square.
# 2) Have the print_square function auto-calculate how much space
# to allot between square elements for optimized display.
# This might require parameterizing the "printf" line.
# 3) Add appropriate functions for generating magic squares
# with an *even* number of rows/columns.
# This is non-trivial(!).
# See the URL for Kwon Young Shin, above, for help.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="FIFTEEN"
></A
><P
><B
>Example A-47. Fifteen Puzzle</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# fifteen.sh
# Classic "Fifteen Puzzle"
# Author: Antonio Macchi
# Lightly edited and commented by ABS Guide author.
# Used in ABS Guide with permission. (Thanks!)
# The invention of the Fifteen Puzzle is attributed to either
#+ Sam Loyd or Noyes Palmer Chapman.
# The puzzle was wildly popular in the late 19th-century.
# Object: Rearrange the numbers so they read in order,
#+ from 1 - 15: ________________
# | 1 2 3 4 |
# | 5 6 7 8 |
# | 9 10 11 12 |
# | 13 14 15 |
# ----------------
#######################
# Constants #
SQUARES=16 #
FAIL=70 #
E_PREMATURE_EXIT=80 #
#######################
########
# Data #
########
Puzzle=( 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 " " )
#############
# Functions #
#############
function swap
{
local tmp
tmp=${Puzzle[$1]}
Puzzle[$1]=${Puzzle[$2]}
Puzzle[$2]=$tmp
}
function Jumble
{ # Scramble the pieces at beginning of round.
local i pos1 pos2
for i in {1..100}
do
pos1=$(( $RANDOM % $SQUARES))
pos2=$(( $RANDOM % $SQUARES ))
swap $pos1 $pos2
done
}
function PrintPuzzle
{
local i1 i2 puzpos
puzpos=0
clear
echo "Enter quit to exit."; echo # Better that than Ctl-C.
echo ",----.----.----.----." # Top border.
for i1 in {1..4}
do
for i2 in {1..4}
do
printf "| %2s " "${Puzzle[$puzpos]}"
(( puzpos++ ))
done
echo "|" # Right-side border.
test $i1 = 4 || echo "+----+----+----+----+"
done
echo "'----'----'----'----'" # Bottom border.
}
function GetNum
{ # Test for valid input.
local puznum garbage
while true
do
echo "Moves: $moves" # Also counts invalid moves.
read -p "Number to move: " puznum garbage
if [ "$puznum" = "quit" ]; then echo; exit $E_PREMATURE_EXIT; fi
test -z "$puznum" -o -n "${puznum//[0-9]/}" &#38;&#38; continue
test $puznum -gt 0 -a $puznum -lt $SQUARES &#38;&#38; break
done
return $puznum
}
function GetPosFromNum
{ # $1 = puzzle-number
local puzpos
for puzpos in {0..15}
do
test "${Puzzle[$puzpos]}" = "$1" &#38;&#38; break
done
return $puzpos
}
function Move
{ # $1=Puzzle-pos
test $1 -gt 3 &#38;&#38; test "${Puzzle[$(( $1 - 4 ))]}" = " "\
&#38;&#38; swap $1 $(( $1 - 4 )) &#38;&#38; return 0
test $(( $1%4 )) -ne 3 &#38;&#38; test "${Puzzle[$(( $1 + 1 ))]}" = " "\
&#38;&#38; swap $1 $(( $1 + 1 )) &#38;&#38; return 0
test $1 -lt 12 &#38;&#38; test "${Puzzle[$(( $1 + 4 ))]}" = " "\
&#38;&#38; swap $1 $(( $1 + 4 )) &#38;&#38; return 0
test $(( $1%4 )) -ne 0 &#38;&#38; test "${Puzzle[$(( $1 - 1 ))]}" = " " &#38;&#38;\
swap $1 $(( $1 - 1 )) &#38;&#38; return 0
return 1
}
function Solved
{
local pos
for pos in {0..14}
do
test "${Puzzle[$pos]}" = $(( $pos + 1 )) || return $FAIL
# Check whether number in each square = square number.
done
return 0 # Successful solution.
}
################### MAIN () #######################{
moves=0
Jumble
while true # Loop continuously until puzzle solved.
do
echo; echo
PrintPuzzle
echo
while true
do
GetNum
puznum=$?
GetPosFromNum $puznum
puzpos=$?
((moves++))
Move $puzpos &#38;&#38; break
done
Solved &#38;&#38; break
done
echo;echo
PrintPuzzle
echo; echo "BRAVO!"; echo
exit 0
###################################################}
# Exercise:
# --------
# Rewrite the script to display the letters A - O,
#+ rather than the numbers 1 - 15.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="HANOI2REF"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="HANOI2"
></A
><P
><B
>Example A-48. <I
CLASS="FIRSTTERM"
>The Towers of Hanoi, graphic
version</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
# The Towers Of Hanoi
# Original script (hanoi.bash) copyright (C) 2000 Amit Singh.
# All Rights Reserved.
# http://hanoi.kernelthread.com
# hanoi2.bash
# Version 2.00: modded for ASCII-graphic display.
# Version 2.01: fixed no command-line param bug.
# Uses code contributed by Antonio Macchi,
#+ with heavy editing by ABS Guide author.
# This variant falls under the original copyright, see above.
# Used in ABS Guide with Amit Singh's permission (thanks!).
### Variables &#38;&#38; sanity check ###
E_NOPARAM=86
E_BADPARAM=87 # Illegal no. of disks passed to script.
E_NOEXIT=88
DISKS=${1:-$E_NOPARAM} # Must specify how many disks.
Moves=0
MWIDTH=7
MARGIN=2
# Arbitrary "magic" constants; work okay for relatively small # of disks.
# BASEWIDTH=51 # Original code.
let "basewidth = $MWIDTH * $DISKS + $MARGIN" # "Base" beneath rods.
# Above "algorithm" could likely stand improvement.
### Display variables ###
let "disks1 = $DISKS - 1"
let "spaces1 = $DISKS"
let "spaces2 = 2 * $DISKS"
let "lastmove_t = $DISKS - 1" # Final move?
declare -a Rod1 Rod2 Rod3
### ######################### ###
function repeat { # $1=char $2=number of repetitions
local n # Repeat-print a character.
for (( n=0; n&#60;$2; n++ )); do
echo -n "$1"
done
}
function FromRod {
local rod summit weight sequence
while true; do
rod=$1
test ${rod/[^123]/} || continue
sequence=$(echo $(seq 0 $disks1 | tac))
for summit in $sequence; do
eval weight=\${Rod${rod}[$summit]}
test $weight -ne 0 &#38;&#38;
{ echo "$rod $summit $weight"; return; }
done
done
}
function ToRod { # $1=previous (FromRod) weight
local rod firstfree weight sequence
while true; do
rod=$2
test ${rod/[^123]} || continue
sequence=$(echo $(seq 0 $disks1 | tac))
for firstfree in $sequence; do
eval weight=\${Rod${rod}[$firstfree]}
test $weight -gt 0 &#38;&#38; { (( firstfree++ )); break; }
done
test $weight -gt $1 -o $firstfree = 0 &#38;&#38;
{ echo "$rod $firstfree"; return; }
done
}
function PrintRods {
local disk rod empty fill sp sequence
repeat " " $spaces1
echo -n "|"
repeat " " $spaces2
echo -n "|"
repeat " " $spaces2
echo "|"
sequence=$(echo $(seq 0 $disks1 | tac))
for disk in $sequence; do
for rod in {1..3}; do
eval empty=$(( $DISKS - (Rod${rod}[$disk] / 2) ))
eval fill=\${Rod${rod}[$disk]}
repeat " " $empty
test $fill -gt 0 &#38;&#38; repeat "*" $fill || echo -n "|"
repeat " " $empty
done
echo
done
repeat "=" $basewidth # Print "base" beneath rods.
echo
}
display ()
{
echo
PrintRods
# Get rod-number, summit and weight
first=( `FromRod $1` )
eval Rod${first[0]}[${first[1]}]=0
# Get rod-number and first-free position
second=( `ToRod ${first[2]} $2` )
eval Rod${second[0]}[${second[1]}]=${first[2]}
echo; echo; echo
if [ "${Rod3[lastmove_t]}" = 1 ]
then # Last move? If yes, then display final position.
echo "+ Final Position: $Moves moves"; echo
PrintRods
fi
}
# From here down, almost the same as original (hanoi.bash) script.
dohanoi() { # Recursive function.
case $1 in
0)
;;
*)
dohanoi "$(($1-1))" $2 $4 $3
if [ "$Moves" -ne 0 ]
then
echo "+ Position after move $Moves"
fi
((Moves++))
echo -n " Next move will be: "
echo $2 "--&#62;" $3
display $2 $3
dohanoi "$(($1-1))" $4 $3 $2
;;
esac
}
setup_arrays ()
{
local dim n elem
let "dim1 = $1 - 1"
elem=$dim1
for n in $(seq 0 $dim1)
do
let "Rod1[$elem] = 2 * $n + 1"
Rod2[$n]=0
Rod3[$n]=0
((elem--))
done
}
### Main ###
setup_arrays $DISKS
echo; echo "+ Start Position"
case $# in
1) case $(($1&#62;0)) in # Must have at least one disk.
1)
disks=$1
dohanoi $1 1 3 2
# Total moves = 2^n - 1, where n = number of disks.
echo
exit 0;
;;
*)
echo "$0: Illegal value for number of disks";
exit $E_BADPARAM;
;;
esac
;;
*)
clear
echo "usage: $0 N"
echo " Where \"N\" is the number of disks."
exit $E_NOPARAM;
;;
esac
exit $E_NOEXIT # Shouldn't exit here.
# Note:
# Redirect script output to a file, otherwise it scrolls off display.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="HANOI2AREF"
></A
></P
><DIV
CLASS="EXAMPLE"
><A
NAME="HANOI2A"
></A
><P
><B
>Example A-49. <I
CLASS="FIRSTTERM"
>The Towers of Hanoi, alternate graphic
version</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/bash
# The Towers Of Hanoi
# Original script (hanoi.bash) copyright (C) 2000 Amit Singh.
# All Rights Reserved.
# http://hanoi.kernelthread.com
# hanoi2.bash
# Version 2: modded for ASCII-graphic display.
# Uses code contributed by Antonio Macchi,
#+ with heavy editing by ABS Guide author.
# This variant also falls under the original copyright, see above.
# Used in ABS Guide with Amit Singh's permission (thanks!).
# Variables #
E_NOPARAM=86
E_BADPARAM=87 # Illegal no. of disks passed to script.
E_NOEXIT=88
DELAY=2 # Interval, in seconds, between moves. Change, if desired.
DISKS=$1
Moves=0
MWIDTH=7
MARGIN=2
# Arbitrary "magic" constants, work okay for relatively small # of disks.
# BASEWIDTH=51 # Original code.
let "basewidth = $MWIDTH * $DISKS + $MARGIN" # "Base" beneath rods.
# Above "algorithm" could likely stand improvement.
# Display variables.
let "disks1 = $DISKS - 1"
let "spaces1 = $DISKS"
let "spaces2 = 2 * $DISKS"
let "lastmove_t = $DISKS - 1" # Final move?
declare -a Rod1 Rod2 Rod3
#################
function repeat { # $1=char $2=number of repetitions
local n # Repeat-print a character.
for (( n=0; n&#60;$2; n++ )); do
echo -n "$1"
done
}
function FromRod {
local rod summit weight sequence
while true; do
rod=$1
test ${rod/[^123]/} || continue
sequence=$(echo $(seq 0 $disks1 | tac))
for summit in $sequence; do
eval weight=\${Rod${rod}[$summit]}
test $weight -ne 0 &#38;&#38;
{ echo "$rod $summit $weight"; return; }
done
done
}
function ToRod { # $1=previous (FromRod) weight
local rod firstfree weight sequence
while true; do
rod=$2
test ${rod/[^123]} || continue
sequence=$(echo $(seq 0 $disks1 | tac))
for firstfree in $sequence; do
eval weight=\${Rod${rod}[$firstfree]}
test $weight -gt 0 &#38;&#38; { (( firstfree++ )); break; }
done
test $weight -gt $1 -o $firstfree = 0 &#38;&#38;
{ echo "$rod $firstfree"; return; }
done
}
function PrintRods {
local disk rod empty fill sp sequence
tput cup 5 0
repeat " " $spaces1
echo -n "|"
repeat " " $spaces2
echo -n "|"
repeat " " $spaces2
echo "|"
sequence=$(echo $(seq 0 $disks1 | tac))
for disk in $sequence; do
for rod in {1..3}; do
eval empty=$(( $DISKS - (Rod${rod}[$disk] / 2) ))
eval fill=\${Rod${rod}[$disk]}
repeat " " $empty
test $fill -gt 0 &#38;&#38; repeat "*" $fill || echo -n "|"
repeat " " $empty
done
echo
done
repeat "=" $basewidth # Print "base" beneath rods.
echo
}
display ()
{
echo
PrintRods
# Get rod-number, summit and weight
first=( `FromRod $1` )
eval Rod${first[0]}[${first[1]}]=0
# Get rod-number and first-free position
second=( `ToRod ${first[2]} $2` )
eval Rod${second[0]}[${second[1]}]=${first[2]}
if [ "${Rod3[lastmove_t]}" = 1 ]
then # Last move? If yes, then display final position.
tput cup 0 0
echo; echo "+ Final Position: $Moves moves"
PrintRods
fi
sleep $DELAY
}
# From here down, almost the same as original (hanoi.bash) script.
dohanoi() { # Recursive function.
case $1 in
0)
;;
*)
dohanoi "$(($1-1))" $2 $4 $3
if [ "$Moves" -ne 0 ]
then
tput cup 0 0
echo; echo "+ Position after move $Moves"
fi
((Moves++))
echo -n " Next move will be: "
echo $2 "--&#62;" $3
display $2 $3
dohanoi "$(($1-1))" $4 $3 $2
;;
esac
}
setup_arrays ()
{
local dim n elem
let "dim1 = $1 - 1"
elem=$dim1
for n in $(seq 0 $dim1)
do
let "Rod1[$elem] = 2 * $n + 1"
Rod2[$n]=0
Rod3[$n]=0
((elem--))
done
}
### Main ###
trap "tput cnorm" 0
tput civis
clear
setup_arrays $DISKS
tput cup 0 0
echo; echo "+ Start Position"
case $# in
1) case $(($1&#62;0)) in # Must have at least one disk.
1)
disks=$1
dohanoi $1 1 3 2
# Total moves = 2^n - 1, where n = # of disks.
echo
exit 0;
;;
*)
echo "$0: Illegal value for number of disks";
exit $E_BADPARAM;
;;
esac
;;
*)
echo "usage: $0 N"
echo " Where \"N\" is the number of disks."
exit $E_NOPARAM;
;;
esac
exit $E_NOEXIT # Shouldn't exit here.
# Exercise:
# --------
# There is a minor bug in the script that causes the display of
#+ the next-to-last move to be skipped.
#+ Fix this.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="USEGETOPT"
></A
><P
><B
>Example A-50. An alternate version of the
<A
HREF="#GETOPTSIMPLE"
>getopt-simple.sh</A
> script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# UseGetOpt.sh
# Author: Peggy Russell &#60;prusselltechgroup@gmail.com&#62;
UseGetOpt () {
declare inputOptions
declare -r E_OPTERR=85
declare -r ScriptName=${0##*/}
declare -r ShortOpts="adf:hlt"
declare -r LongOpts="aoption,debug,file:,help,log,test"
DoSomething () {
echo "The function name is '${FUNCNAME}'"
# Recall that $FUNCNAME is an internal variable
#+ holding the name of the function it is in.
}
inputOptions=$(getopt -o "${ShortOpts}" --long \
"${LongOpts}" --name "${ScriptName}" -- "${@}")
if [[ ($? -ne 0) || ($# -eq 0) ]]; then
echo "Usage: ${ScriptName} [-dhlt] {OPTION...}"
exit $E_OPTERR
fi
eval set -- "${inputOptions}"
# Only for educational purposes. Can be removed.
#-----------------------------------------------
echo "++ Test: Number of arguments: [$#]"
echo '++ Test: Looping through "$@"'
for a in "$@"; do
echo " ++ [$a]"
done
#-----------------------------------------------
while true; do
case "${1}" in
--aoption | -a) # Argument found.
echo "Option [$1]"
;;
--debug | -d) # Enable informational messages.
echo "Option [$1] Debugging enabled"
;;
--file | -f) # Check for optional argument.
case "$2" in #+ Double colon is optional argument.
"") # Not there.
echo "Option [$1] Use default"
shift
;;
*) # Got it
echo "Option [$1] Using input [$2]"
shift
;;
esac
DoSomething
;;
--log | -l) # Enable Logging.
echo "Option [$1] Logging enabled"
;;
--test | -t) # Enable testing.
echo "Option [$1] Testing enabled"
;;
--help | -h)
echo "Option [$1] Display help"
break
;;
--) # Done! $# is argument number for "--", $@ is "--"
echo "Option [$1] Dash Dash"
break
;;
*)
echo "Major internal error!"
exit 8
;;
esac
echo "Number of arguments: [$#]"
shift
done
shift
# Only for educational purposes. Can be removed.
#----------------------------------------------------------------------
echo "++ Test: Number of arguments after \"--\" is [$#] They are: [$@]"
echo '++ Test: Looping through "$@"'
for a in "$@"; do
echo " ++ [$a]"
done
#----------------------------------------------------------------------
}
################################### M A I N ########################
# If you remove "function UseGetOpt () {" and corresponding "}",
#+ you can uncomment the "exit 0" line below, and invoke this script
#+ with the various options from the command-line.
#-------------------------------------------------------------------
# exit 0
echo "Test 1"
UseGetOpt -f myfile one "two three" four
echo;echo "Test 2"
UseGetOpt -h
echo;echo "Test 3 - Short Options"
UseGetOpt -adltf myfile anotherfile
echo;echo "Test 4 - Long Options"
UseGetOpt --aoption --debug --log --test --file myfile anotherfile
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="USEGETOPT2"
></A
><P
><B
>Example A-51. The version of the
<I
CLASS="FIRSTTERM"
>UseGetOpt.sh</I
> example used in the <A
HREF="#TABEXPANSION"
>Tab Expansion appendix</A
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# UseGetOpt-2.sh
# Modified version of the script for illustrating tab-expansion
#+ of command-line options.
# See the "Introduction to Tab Expansion" appendix.
# Possible options: -a -d -f -l -t -h
#+ --aoption, --debug --file --log --test -- help --
# Author of original script: Peggy Russell &#60;prusselltechgroup@gmail.com&#62;
# UseGetOpt () {
declare inputOptions
declare -r E_OPTERR=85
declare -r ScriptName=${0##*/}
declare -r ShortOpts="adf:hlt"
declare -r LongOpts="aoption,debug,file:,help,log,test"
DoSomething () {
echo "The function name is '${FUNCNAME}'"
}
inputOptions=$(getopt -o "${ShortOpts}" --long \
"${LongOpts}" --name "${ScriptName}" -- "${@}")
if [[ ($? -ne 0) || ($# -eq 0) ]]; then
echo "Usage: ${ScriptName} [-dhlt] {OPTION...}"
exit $E_OPTERR
fi
eval set -- "${inputOptions}"
while true; do
case "${1}" in
--aoption | -a) # Argument found.
echo "Option [$1]"
;;
--debug | -d) # Enable informational messages.
echo "Option [$1] Debugging enabled"
;;
--file | -f) # Check for optional argument.
case "$2" in #+ Double colon is optional argument.
"") # Not there.
echo "Option [$1] Use default"
shift
;;
*) # Got it
echo "Option [$1] Using input [$2]"
shift
;;
esac
DoSomething
;;
--log | -l) # Enable Logging.
echo "Option [$1] Logging enabled"
;;
--test | -t) # Enable testing.
echo "Option [$1] Testing enabled"
;;
--help | -h)
echo "Option [$1] Display help"
break
;;
--) # Done! $# is argument number for "--", $@ is "--"
echo "Option [$1] Dash Dash"
break
;;
*)
echo "Major internal error!"
exit 8
;;
esac
echo "Number of arguments: [$#]"
shift
done
shift
# }
exit</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="SHOWALLC"
></A
><P
><B
>Example A-52. Cycling through all the possible color backgrounds</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# show-all-colors.sh
# Displays all 256 possible background colors, using ANSI escape sequences.
# Author: Chetankumar Phulpagare
# Used in ABS Guide with permission.
T1=8
T2=6
T3=36
offset=0
for num1 in {0..7}
do {
for num2 in {0,1}
do {
shownum=`echo "$offset + $T1 * ${num2} + $num1" | bc`
echo -en "\E[0;48;5;${shownum}m color ${shownum} \E[0m"
}
done
echo
}
done
offset=16
for num1 in {0..5}
do {
for num2 in {0..5}
do {
for num3 in {0..5}
do {
shownum=`echo "$offset + $T2 * ${num3} \
+ $num2 + $T3 * ${num1}" | bc`
echo -en "\E[0;48;5;${shownum}m color ${shownum} \E[0m"
}
done
echo
}
done
}
done
offset=232
for num1 in {0..23}
do {
shownum=`expr $offset + $num1`
echo -en "\E[0;48;5;${shownum}m ${shownum}\E[0m"
}
done
echo</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="SAMORSE"
></A
><P
><B
>Example A-53. Morse Code Practice</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# sam.sh, v. .01a
# Still Another Morse (code training script)
# With profuse apologies to Sam (F.B.) Morse.
# Author: Mendel Cooper
# License: GPL3
# Reldate: 05/25/11
# Morse code training script.
# Converts arguments to audible dots and dashes.
# Note: lowercase input only at this time.
# Get the wav files from the source tarball:
# http://bash.deta.in/abs-guide-latest.tar.bz2
DOT='soundfiles/dot.wav'
DASH='soundfiles/dash.wav'
# Maybe move soundfiles to /usr/local/sounds?
LETTERSPACE=300000 # Microseconds.
WORDSPACE=980000
# Nice and slow, for beginners. Maybe 5 wpm?
EXIT_MSG="May the Morse be with you!"
E_NOARGS=75 # No command-line args?
declare -A morse # Associative array!
# ======================================= #
morse[a]="dot; dash"
morse[b]="dash; dot; dot; dot"
morse[c]="dash; dot; dash; dot"
morse[d]="dash; dot; dot"
morse[e]="dot"
morse[f]="dot; dot; dash; dot"
morse[g]="dash; dash; dot"
morse[h]="dot; dot; dot; dot"
morse[i]="dot; dot;"
morse[j]="dot; dash; dash; dash"
morse[k]="dash; dot; dash"
morse[l]="dot; dash; dot; dot"
morse[m]="dash; dash"
morse[n]="dash; dot"
morse[o]="dash; dash; dash"
morse[p]="dot; dash; dash; dot"
morse[q]="dash; dash; dot; dash"
morse[r]="dot; dash; dot"
morse[s]="dot; dot; dot"
morse[t]="dash"
morse[u]="dot; dot; dash"
morse[v]="dot; dot; dot; dash"
morse[w]="dot; dash; dash"
morse[x]="dash; dot; dot; dash"
morse[y]="dash; dot; dash; dash"
morse[z]="dash; dash; dot; dot"
morse[0]="dash; dash; dash; dash; dash"
morse[1]="dot; dash; dash; dash; dash"
morse[2]="dot; dot; dash; dash; dash"
morse[3]="dot; dot; dot; dash; dash"
morse[4]="dot; dot; dot; dot; dash"
morse[5]="dot; dot; dot; dot; dot"
morse[6]="dash; dot; dot; dot; dot"
morse[7]="dash; dash; dot; dot; dot"
morse[8]="dash; dash; dash; dot; dot"
morse[9]="dash; dash; dash; dash; dot"
# The following must be escaped or quoted.
morse[?]="dot; dot; dash; dash; dot; dot"
morse[.]="dot; dash; dot; dash; dot; dash"
morse[,]="dash; dash; dot; dot; dash; dash"
morse[/]="dash; dot; dot; dash; dot"
morse[\@]="dot; dash; dash; dot; dash; dot"
# ======================================= #
play_letter ()
{
eval ${morse[$1]} # Play dots, dashes from appropriate sound files.
# Why is 'eval' necessary here?
usleep $LETTERSPACE # Pause in between letters.
}
extract_letters ()
{ # Slice string apart, letter by letter.
local pos=0 # Starting at left end of string.
local len=1 # One letter at a time.
strlen=${#1}
while [ $pos -lt $strlen ]
do
letter=${1:pos:len}
# ^^^^^^^^^^^^ See Chapter 10.1.
play_letter $letter
echo -n "*" # Mark letter just played.
((pos++))
done
}
######### Play the sounds ############
dot() { aplay "$DOT" 2&#38;&#62;/dev/null; }
dash() { aplay "$DASH" 2&#38;&#62;/dev/null; }
######################################
no_args ()
{
declare -a usage
usage=( $0 word1 word2 ... )
echo "Usage:"; echo
echo ${usage[*]}
for index in 0 1 2 3
do
extract_letters ${usage[index]}
usleep $WORDSPACE
echo -n " " # Print space between words.
done
# echo "Usage: $0 word1 word2 ... "
echo; echo
}
# int main()
# {
clear # Clear the terminal screen.
echo " SAM"
echo "Still Another Morse code trainer"
echo " Author: Mendel Cooper"
echo; echo;
if [ -z "$1" ]
then
no_args
echo; echo; echo "$EXIT_MSG"; echo
exit $E_NOARGS
fi
echo; echo "$*" # Print text that will be played.
until [ -z "$1" ]
do
extract_letters $1
shift # On to next word.
usleep $WORDSPACE
echo -n " " # Print space between words.
done
echo; echo; echo "$EXIT_MSG"; echo
exit 0
# }
# Exercises:
# ---------
# 1) Have the script accept either lowercase or uppercase words
#+ as arguments. Hint: Use 'tr' . . .
# 2) Have the script optionally accept input from a text file.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="BASE64"
></A
><P
><B
>Example A-54. Base64 encoding/decoding</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# base64.sh: Bash implementation of Base64 encoding and decoding.
#
# Copyright (c) 2011 vladz &#60;vladz@devzero.fr&#62;
# Used in ABSG with permission (thanks!).
#
# Encode or decode original Base64 (and also Base64url)
#+ from STDIN to STDOUT.
#
# Usage:
#
# Encode
# $ ./base64.sh &#60; binary-file &#62; binary-file.base64
# Decode
# $ ./base64.sh -d &#60; binary-file.base64 &#62; binary-file
#
# Reference:
#
# [1] RFC4648 - "The Base16, Base32, and Base64 Data Encodings"
# http://tools.ietf.org/html/rfc4648#section-5
# The base64_charset[] array contains entire base64 charset,
# and additionally the character "=" ...
base64_charset=( {A..Z} {a..z} {0..9} + / = )
# Nice illustration of brace expansion.
# Uncomment the ### line below to use base64url encoding instead of
#+ original base64.
### base64_charset=( {A..Z} {a..z} {0..9} - _ = )
# Output text width when encoding
#+ (64 characters, just like openssl output).
text_width=64
function display_base64_char {
# Convert a 6-bit number (between 0 and 63) into its corresponding values
#+ in Base64, then display the result with the specified text width.
printf "${base64_charset[$1]}"; (( width++ ))
(( width % text_width == 0 )) &#38;&#38; printf "\n"
}
function encode_base64 {
# Encode three 8-bit hexadecimal codes into four 6-bit numbers.
# We need two local int array variables:
# c8[]: to store the codes of the 8-bit characters to encode
# c6[]: to store the corresponding encoded values on 6-bit
declare -a -i c8 c6
# Convert hexadecimal to decimal.
c8=( $(printf "ibase=16; ${1:0:2}\n${1:2:2}\n${1:4:2}\n" | bc) )
# Let's play with bitwise operators
#+ (3x8-bit into 4x6-bits conversion).
(( c6[0] = c8[0] &#62;&#62; 2 ))
(( c6[1] = ((c8[0] &#38; 3) &#60;&#60; 4) | (c8[1] &#62;&#62; 4) ))
# The following operations depend on the c8 element number.
case ${#c8[*]} in
3) (( c6[2] = ((c8[1] &#38; 15) &#60;&#60; 2) | (c8[2] &#62;&#62; 6) ))
(( c6[3] = c8[2] &#38; 63 )) ;;
2) (( c6[2] = (c8[1] &#38; 15) &#60;&#60; 2 ))
(( c6[3] = 64 )) ;;
1) (( c6[2] = c6[3] = 64 )) ;;
esac
for char in ${c6[@]}; do
display_base64_char ${char}
done
}
function decode_base64 {
# Decode four base64 characters into three hexadecimal ASCII characters.
# c8[]: to store the codes of the 8-bit characters
# c6[]: to store the corresponding Base64 values on 6-bit
declare -a -i c8 c6
# Find decimal value corresponding to the current base64 character.
for current_char in ${1:0:1} ${1:1:1} ${1:2:1} ${1:3:1}; do
[ "${current_char}" = "=" ] &#38;&#38; break
position=0
while [ "${current_char}" != "${base64_charset[${position}]}" ]; do
(( position++ ))
done
c6=( ${c6[*]} ${position} )
done
# Let's play with bitwise operators
#+ (4x8-bit into 3x6-bits conversion).
(( c8[0] = (c6[0] &#60;&#60; 2) | (c6[1] &#62;&#62; 4) ))
# The next operations depends on the c6 elements number.
case ${#c6[*]} in
3) (( c8[1] = ( (c6[1] &#38; 15) &#60;&#60; 4) | (c6[2] &#62;&#62; 2) ))
(( c8[2] = (c6[2] &#38; 3) &#60;&#60; 6 )); unset c8[2] ;;
4) (( c8[1] = ( (c6[1] &#38; 15) &#60;&#60; 4) | (c6[2] &#62;&#62; 2) ))
(( c8[2] = ( (c6[2] &#38; 3) &#60;&#60; 6) | c6[3] )) ;;
esac
for char in ${c8[*]}; do
printf "\x$(printf "%x" ${char})"
done
}
# main ()
if [ "$1" = "-d" ]; then # decode
# Reformat STDIN in pseudo 4x6-bit groups.
content=$(cat - | tr -d "\n" | sed -r "s/(.{4})/\1 /g")
for chars in ${content}; do decode_base64 ${chars}; done
else
# Make a hexdump of stdin and reformat in 3-byte groups.
content=$(cat - | xxd -ps -u | sed -r "s/(\w{6})/\1 /g" |
tr -d "\n")
for chars in ${content}; do encode_base64 ${chars}; done
echo
fi</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="SEDAPPEND"
></A
><P
><B
>Example A-55. Inserting text in a file using
<I
CLASS="FIRSTTERM"
>sed</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Prepends a string at a specified line
#+ in files with names ending in "sample"
#+ in the current working directory.
# 000000000000000000000000000000000000
# This script overwrites files!
# Be careful running it in a directory
#+ where you have important files!!!
# 000000000000000000000000000000000000
# Create a couple of files to operate on ...
# 01sample
# 02sample
# ... etc.
# These files must not be empty, else the prepend will not work.
lineno=1 # Append at line 1 (prepend).
filespec="*sample" # Filename pattern to operate on.
string=$(whoami) # Will set your username as string to insert.
# It could just as easily be any other string.
for file in $filespec # Specify which files to alter.
do # ^^^^^^^^^
sed -i ""$lineno"i "$string"" $file
# ^^ -i option edits files in-place.
# ^ Insert (i) command.
echo ""$file" altered!"
done
echo "Warning: files possibly clobbered!"
exit 0
# Exercise:
# Add error checking to this script.
# It needs it badly.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="GRONSFELD"
></A
><P
><B
>Example A-56. The Gronsfeld Cipher</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# gronsfeld.bash
# License: GPL3
# Reldate 06/23/11
# This is an implementation of the Gronsfeld Cipher.
# It's essentially a stripped-down variant of the
#+ polyalphabetic Vigen<65>re Tableau, but with only 10 alphabets.
# The classic Gronsfeld has a numeric sequence as the key word,
#+ but here we substitute a letter string, for ease of use.
# Allegedly, this cipher was invented by the eponymous Count Gronsfeld
#+ in the 17th Century. It was at one time considered to be unbreakable.
# Note that this is ###not### a secure cipher by modern standards.
# Global Variables #
Enc_suffix="29379" # Encrypted text output with this 5-digit suffix.
# This functions as a decryption flag,
#+ and when used to generate passwords adds security.
Default_key="gronsfeldk"
# The script uses this if key not entered below
# (at "Keychain").
# Change the above two values frequently
#+ for added security.
GROUPLEN=5 # Output in groups of 5 letters, per tradition.
alpha1=( abcdefghijklmnopqrstuvwxyz )
alpha2=( {A..Z} ) # Output in all caps, per tradition.
# Use alpha2=( {a..z} ) for password generator.
wraplen=26 # Wrap around if past end of alphabet.
dflag= # Decrypt flag (set if $Enc_suffix present).
E_NOARGS=76 # Missing command-line args?
DEBUG=77 # Debugging flag.
declare -a offsets # This array holds the numeric shift values for
#+ encryption/decryption.
########Keychain#########
key= ### Put key here!!!
# 10 characters!
#########################
# Function
: ()
{ # Encrypt or decrypt, depending on whether $dflag is set.
# Why ": ()" as a function name? Just to prove that it can be done.
local idx keydx mlen off1 shft
local plaintext="$1"
local mlen=${#plaintext}
for (( idx=0; idx&#60;$mlen; idx++ ))
do
let "keydx = $idx % $keylen"
shft=${offsets[keydx]}
if [ -n "$dflag" ]
then # Decrypt!
let "off1 = $(expr index "${alpha1[*]}" ${plaintext:idx:1}) - $shft"
# Shift backward to decrypt.
else # Encrypt!
let "off1 = $(expr index "${alpha1[*]}" ${plaintext:idx:1}) + $shft"
# Shift forward to encrypt.
test $(( $idx % $GROUPLEN)) = 0 &#38;&#38; echo -n " " # Groups of 5 letters.
# Comment out above line for output as a string without whitespace,
#+ for example, if using the script as a password generator.
fi
((off1--)) # Normalize. Why is this necessary?
if [ $off1 -lt 0 ]
then # Catch negative indices.
let "off1 += $wraplen"
fi
((off1 %= $wraplen)) # Wrap around if past end of alphabet.
echo -n "${alpha2[off1]}"
done
if [ -z "$dflag" ]
then
echo " $Enc_suffix"
# echo "$Enc_suffix" # For password generator.
else
echo
fi
} # End encrypt/decrypt function.
# int main () {
# Check for command-line args.
if [ -z "$1" ]
then
echo "Usage: $0 TEXT TO ENCODE/DECODE"
exit $E_NOARGS
fi
if [ ${!#} == "$Enc_suffix" ]
# ^^^^^ Final command-line arg.
then
dflag=ON
echo -n "+" # Flag decrypted text with a "+" for easy ID.
fi
if [ -z "$key" ]
then
key="$Default_key" # "gronsfeldk" per above.
fi
keylen=${#key}
for (( idx=0; idx&#60;$keylen; idx++ ))
do # Calculate shift values for encryption/decryption.
offsets[idx]=$(expr index "${alpha1[*]}" ${key:idx:1}) # Normalize.
((offsets[idx]--)) # Necessary because "expr index" starts at 1,
#+ whereas array count starts at 0.
# Generate array of numerical offsets corresponding to the key.
# There are simpler ways to accomplish this.
done
args=$(echo "$*" | sed -e 's/ //g' | tr A-Z a-z | sed -e 's/[0-9]//g')
# Remove whitespace and digits from command-line args.
# Can modify to also remove punctuation characters, if desired.
# Debug:
# echo "$args"; exit $DEBUG
: "$args" # Call the function named ":".
# : is a null operator, except . . . when it's a function name!
exit $? # } End-of-script
# ************************************************************** #
# This script can function as a password generator,
#+ with several minor mods, see above.
# That would allow an easy-to-remember password, even the word
#+ "password" itself, which encrypts to vrgfotvo29379
#+ a fairly secure password not susceptible to a dictionary attack.
# Or, you could use your own name (surely that's easy to remember!).
# For example, Bozo Bozeman encrypts to hfnbttdppkt29379.
# ************************************************************** #</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="BINGO"
></A
><P
><B
>Example A-57. Bingo Number Generator</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# bingo.sh
# Bingo number generator
# Reldate 20Aug12, License: Public Domain
#######################################################################
# This script generates bingo numbers.
# Hitting a key generates a new number.
# Hitting 'q' terminates the script.
# In a given run of the script, there will be no duplicate numbers.
# When the script terminates, it prints a log of the numbers generated.
#######################################################################
MIN=1 # Lowest allowable bingo number.
MAX=75 # Highest allowable bingo number.
COLS=15 # Numbers in each column (B I N G O).
SINGLE_DIGIT_MAX=9
declare -a Numbers
Prefix=(B I N G O)
initialize_Numbers ()
{ # Zero them out to start.
# They'll be incremented if chosen.
local index=0
until [ "$index" -gt $MAX ]
do
Numbers[index]=0
((index++))
done
Numbers[0]=1 # Flag zero, so it won't be selected.
}
generate_number ()
{
local number
while [ 1 ]
do
let "number = $(expr $RANDOM % $MAX)"
if [ ${Numbers[number]} -eq 0 ] # Number not yet called.
then
let "Numbers[number]+=1" # Flag it in the array.
break # And terminate loop.
fi # Else if already called, loop and generate another number.
done
# Exercise: Rewrite this more elegantly as an until-loop.
return $number
}
print_numbers_called ()
{ # Print out the called number log in neat columns.
# echo ${Numbers[@]}
local pre2=0 # Prefix a zero, so columns will align
#+ on single-digit numbers.
echo "Number Stats"
for (( index=1; index&#60;=MAX; index++))
do
count=${Numbers[index]}
let "t = $index - 1" # Normalize, since array begins with index 0.
let "column = $(expr $t / $COLS)"
pre=${Prefix[column]}
# echo -n "${Prefix[column]} "
if [ $(expr $t % $COLS) -eq 0 ]
then
echo # Newline at end of row.
fi
if [ "$index" -gt $SINGLE_DIGIT_MAX ] # Check for single-digit number.
then
echo -n "$pre$index#$count "
else # Prefix a zero.
echo -n "$pre$pre2$index#$count "
fi
done
}
# main () {
RANDOM=$$ # Seed random number generator.
initialize_Numbers # Zero out the number tracking array.
clear
echo "Bingo Number Caller"; echo
while [[ "$key" != "q" ]] # Main loop.
do
read -s -n1 -p "Hit a key for the next number [q to exit] " key
# Usually 'q' exits, but not always.
# Can always hit Ctl-C if q fails.
echo
generate_number; new_number=$?
let "column = $(expr $new_number / $COLS)"
echo -n "${Prefix[column]} " # B-I-N-G-O
echo $new_number
done
echo; echo
# Game over ...
print_numbers_called
echo; echo "[#0 = not called . . . #1 = called]"
echo
exit 0
# }
# Certainly, this script could stand some improvement.
#See also the author's Instructable:
#www.instructables.com/id/Binguino-An-Arduino-based-Bingo-Number-Generato/</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="BASICSREV0"
></A
>To end this section, a review of the
basics . . . and more.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="BASICSREVIEWED"
></A
><P
><B
>Example A-58. Basics Reviewed</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# basics-reviewed.bash
# File extension == *.bash == specific to Bash
# Copyright (c) Michael S. Zick, 2003; All rights reserved.
# License: Use in any form, for any purpose.
# Revision: $ID$
#
# Edited for layout by M.C.
# (author of the "Advanced Bash Scripting Guide")
# Fixes and updates (04/08) by Cliff Bamford.
# This script tested under Bash versions 2.04, 2.05a and 2.05b.
# It may not work with earlier versions.
# This demonstration script generates one --intentional--
#+ "command not found" error message. See line 436.
# The current Bash maintainer, Chet Ramey, has fixed the items noted
#+ for later versions of Bash.
###-------------------------------------------###
### Pipe the output of this script to 'more' ###
###+ else it will scroll off the page. ###
### ###
### You may also redirect its output ###
###+ to a file for examination. ###
###-------------------------------------------###
# Most of the following points are described at length in
#+ the text of the foregoing "Advanced Bash Scripting Guide."
# This demonstration script is mostly just a reorganized presentation.
# -- msz
# Variables are not typed unless otherwise specified.
# Variables are named. Names must contain a non-digit.
# File descriptor names (as in, for example: 2&#62;&#38;1)
#+ contain ONLY digits.
# Parameters and Bash array elements are numbered.
# (Parameters are very similar to Bash arrays.)
# A variable name may be undefined (null reference).
unset VarNull
# A variable name may be defined but empty (null contents).
VarEmpty='' # Two, adjacent, single quotes.
# A variable name may be defined and non-empty.
VarSomething='Literal'
# A variable may contain:
# * A whole number as a signed 32-bit (or larger) integer
# * A string
# A variable may also be an array.
# A string may contain embedded blanks and may be treated
#+ as if it where a function name with optional arguments.
# The names of variables and the names of functions
#+ are in different namespaces.
# A variable may be defined as a Bash array either explicitly or
#+ implicitly by the syntax of the assignment statement.
# Explicit:
declare -a ArrayVar
# The echo command is a builtin.
echo $VarSomething
# The printf command is a builtin.
# Translate %s as: String-Format
printf %s $VarSomething # No linebreak specified, none output.
echo # Default, only linebreak output.
# The Bash parser word breaks on whitespace.
# Whitespace, or the lack of it is significant.
# (This holds true in general; there are, of course, exceptions.)
# Translate the DOLLAR_SIGN character as: Content-Of.
# Extended-Syntax way of writing Content-Of:
echo ${VarSomething}
# The ${ ... } Extended-Syntax allows more than just the variable
#+ name to be specified.
# In general, $VarSomething can always be written as: ${VarSomething}.
# Call this script with arguments to see the following in action.
# Outside of double-quotes, the special characters @ and *
#+ specify identical behavior.
# May be pronounced as: All-Elements-Of.
# Without specification of a name, they refer to the
#+ pre-defined parameter Bash-Array.
# Glob-Pattern references
echo $* # All parameters to script or function
echo ${*} # Same
# Bash disables filename expansion for Glob-Patterns.
# Only character matching is active.
# All-Elements-Of references
echo $@ # Same as above
echo ${@} # Same as above
# Within double-quotes, the behavior of Glob-Pattern references
#+ depends on the setting of IFS (Input Field Separator).
# Within double-quotes, All-Elements-Of references behave the same.
# Specifying only the name of a variable holding a string refers
#+ to all elements (characters) of a string.
# To specify an element (character) of a string,
#+ the Extended-Syntax reference notation (see below) MAY be used.
# Specifying only the name of a Bash array references
#+ the subscript zero element,
#+ NOT the FIRST DEFINED nor the FIRST WITH CONTENTS element.
# Additional qualification is needed to reference other elements,
#+ which means that the reference MUST be written in Extended-Syntax.
# The general form is: ${name[subscript]}.
# The string forms may also be used: ${name:subscript}
#+ for Bash-Arrays when referencing the subscript zero element.
# Bash-Arrays are implemented internally as linked lists,
#+ not as a fixed area of storage as in some programming languages.
# Characteristics of Bash arrays (Bash-Arrays):
# --------------------------------------------
# If not otherwise specified, Bash-Array subscripts begin with
#+ subscript number zero. Literally: [0]
# This is called zero-based indexing.
###
# If not otherwise specified, Bash-Arrays are subscript packed
#+ (sequential subscripts without subscript gaps).
###
# Negative subscripts are not allowed.
###
# Elements of a Bash-Array need not all be of the same type.
###
# Elements of a Bash-Array may be undefined (null reference).
# That is, a Bash-Array may be "subscript sparse."
###
# Elements of a Bash-Array may be defined and empty (null contents).
###
# Elements of a Bash-Array may contain:
# * A whole number as a signed 32-bit (or larger) integer
# * A string
# * A string formated so that it appears to be a function name
# + with optional arguments
###
# Defined elements of a Bash-Array may be undefined (unset).
# That is, a subscript packed Bash-Array may be changed
# + into a subscript sparse Bash-Array.
###
# Elements may be added to a Bash-Array by defining an element
#+ not previously defined.
###
# For these reasons, I have been calling them "Bash-Arrays".
# I'll return to the generic term "array" from now on.
# -- msz
echo "========================================================="
# Lines 202 - 334 supplied by Cliff Bamford. (Thanks!)
# Demo --- Interaction with Arrays, quoting, IFS, echo, * and @ ---
#+ all affect how things work
ArrayVar[0]='zero' # 0 normal
ArrayVar[1]=one # 1 unquoted literal
ArrayVar[2]='two' # 2 normal
ArrayVar[3]='three' # 3 normal
ArrayVar[4]='I am four' # 4 normal with spaces
ArrayVar[5]='five' # 5 normal
unset ArrayVar[6] # 6 undefined
ArrayValue[7]='seven' # 7 normal
ArrayValue[8]='' # 8 defined but empty
ArrayValue[9]='nine' # 9 normal
echo '--- Here is the array we are using for this test'
echo
echo "ArrayVar[0]='zero' # 0 normal"
echo "ArrayVar[1]=one # 1 unquoted literal"
echo "ArrayVar[2]='two' # 2 normal"
echo "ArrayVar[3]='three' # 3 normal"
echo "ArrayVar[4]='I am four' # 4 normal with spaces"
echo "ArrayVar[5]='five' # 5 normal"
echo "unset ArrayVar[6] # 6 undefined"
echo "ArrayValue[7]='seven' # 7 normal"
echo "ArrayValue[8]='' # 8 defined but empty"
echo "ArrayValue[9]='nine' # 9 normal"
echo
echo
echo '---Case0: No double-quotes, Default IFS of space,tab,newline ---'
IFS=$'\x20'$'\x09'$'\x0A' # In exactly this order.
echo 'Here is: printf %q {${ArrayVar[*]}'
printf %q ${ArrayVar[*]}
echo
echo 'Here is: printf %q {${ArrayVar[@]}'
printf %q ${ArrayVar[@]}
echo
echo 'Here is: echo ${ArrayVar[*]}'
echo ${ArrayVar[@]}
echo 'Here is: echo {${ArrayVar[@]}'
echo ${ArrayVar[@]}
echo
echo '---Case1: Within double-quotes - Default IFS of space-tab-
newline ---'
IFS=$'\x20'$'\x09'$'\x0A' # These three bytes,
echo 'Here is: printf %q "{${ArrayVar[*]}"'
printf %q "${ArrayVar[*]}"
echo
echo 'Here is: printf %q "{${ArrayVar[@]}"'
printf %q "${ArrayVar[@]}"
echo
echo 'Here is: echo "${ArrayVar[*]}"'
echo "${ArrayVar[@]}"
echo 'Here is: echo "{${ArrayVar[@]}"'
echo "${ArrayVar[@]}"
echo
echo '---Case2: Within double-quotes - IFS is q'
IFS='q'
echo 'Here is: printf %q "{${ArrayVar[*]}"'
printf %q "${ArrayVar[*]}"
echo
echo 'Here is: printf %q "{${ArrayVar[@]}"'
printf %q "${ArrayVar[@]}"
echo
echo 'Here is: echo "${ArrayVar[*]}"'
echo "${ArrayVar[@]}"
echo 'Here is: echo "{${ArrayVar[@]}"'
echo "${ArrayVar[@]}"
echo
echo '---Case3: Within double-quotes - IFS is ^'
IFS='^'
echo 'Here is: printf %q "{${ArrayVar[*]}"'
printf %q "${ArrayVar[*]}"
echo
echo 'Here is: printf %q "{${ArrayVar[@]}"'
printf %q "${ArrayVar[@]}"
echo
echo 'Here is: echo "${ArrayVar[*]}"'
echo "${ArrayVar[@]}"
echo 'Here is: echo "{${ArrayVar[@]}"'
echo "${ArrayVar[@]}"
echo
echo '---Case4: Within double-quotes - IFS is ^ followed by
space,tab,newline'
IFS=$'^'$'\x20'$'\x09'$'\x0A' # ^ + space tab newline
echo 'Here is: printf %q "{${ArrayVar[*]}"'
printf %q "${ArrayVar[*]}"
echo
echo 'Here is: printf %q "{${ArrayVar[@]}"'
printf %q "${ArrayVar[@]}"
echo
echo 'Here is: echo "${ArrayVar[*]}"'
echo "${ArrayVar[@]}"
echo 'Here is: echo "{${ArrayVar[@]}"'
echo "${ArrayVar[@]}"
echo
echo '---Case6: Within double-quotes - IFS set and empty '
IFS=''
echo 'Here is: printf %q "{${ArrayVar[*]}"'
printf %q "${ArrayVar[*]}"
echo
echo 'Here is: printf %q "{${ArrayVar[@]}"'
printf %q "${ArrayVar[@]}"
echo
echo 'Here is: echo "${ArrayVar[*]}"'
echo "${ArrayVar[@]}"
echo 'Here is: echo "{${ArrayVar[@]}"'
echo "${ArrayVar[@]}"
echo
echo '---Case7: Within double-quotes - IFS is unset'
unset IFS
echo 'Here is: printf %q "{${ArrayVar[*]}"'
printf %q "${ArrayVar[*]}"
echo
echo 'Here is: printf %q "{${ArrayVar[@]}"'
printf %q "${ArrayVar[@]}"
echo
echo 'Here is: echo "${ArrayVar[*]}"'
echo "${ArrayVar[@]}"
echo 'Here is: echo "{${ArrayVar[@]}"'
echo "${ArrayVar[@]}"
echo
echo '---End of Cases---'
echo "========================================================="; echo
# Put IFS back to the default.
# Default is exactly these three bytes.
IFS=$'\x20'$'\x09'$'\x0A' # In exactly this order.
# Interpretation of the above outputs:
# A Glob-Pattern is I/O; the setting of IFS matters.
###
# An All-Elements-Of does not consider IFS settings.
###
# Note the different output using the echo command and the
#+ quoted format operator of the printf command.
# Recall:
# Parameters are similar to arrays and have the similar behaviors.
###
# The above examples demonstrate the possible variations.
# To retain the shape of a sparse array, additional script
#+ programming is required.
###
# The source code of Bash has a routine to output the
#+ [subscript]=value array assignment format.
# As of version 2.05b, that routine is not used,
#+ but that might change in future releases.
# The length of a string, measured in non-null elements (characters):
echo
echo '- - Non-quoted references - -'
echo 'Non-Null character count: '${#VarSomething}' characters.'
# test='Lit'$'\x00''eral' # $'\x00' is a null character.
# echo ${#test} # See that?
# The length of an array, measured in defined elements,
#+ including null content elements.
echo
echo 'Defined content count: '${#ArrayVar[@]}' elements.'
# That is NOT the maximum subscript (4).
# That is NOT the range of the subscripts (1 . . 4 inclusive).
# It IS the length of the linked list.
###
# Both the maximum subscript and the range of the subscripts may
#+ be found with additional script programming.
# The length of a string, measured in non-null elements (characters):
echo
echo '- - Quoted, Glob-Pattern references - -'
echo 'Non-Null character count: '"${#VarSomething}"' characters.'
# The length of an array, measured in defined elements,
#+ including null-content elements.
echo
echo 'Defined element count: '"${#ArrayVar[*]}"' elements.'
# Interpretation: Substitution does not effect the ${# ... } operation.
# Suggestion:
# Always use the All-Elements-Of character
#+ if that is what is intended (independence from IFS).
# Define a simple function.
# I include an underscore in the name
#+ to make it distinctive in the examples below.
###
# Bash separates variable names and function names
#+ in different namespaces.
# The Mark-One eyeball isn't that advanced.
###
_simple() {
echo -n 'SimpleFunc'$@ # Newlines are swallowed in
} #+ result returned in any case.
# The ( ... ) notation invokes a command or function.
# The $( ... ) notation is pronounced: Result-Of.
# Invoke the function _simple
echo
echo '- - Output of function _simple - -'
_simple # Try passing arguments.
echo
# or
(_simple) # Try passing arguments.
echo
echo '- Is there a variable of that name? -'
echo $_simple not defined # No variable by that name.
# Invoke the result of function _simple (Error msg intended)
###
$(_simple) # Gives an error message:
# line 436: SimpleFunc: command not found
# ---------------------------------------
echo
###
# The first word of the result of function _simple
#+ is neither a valid Bash command nor the name of a defined function.
###
# This demonstrates that the output of _simple is subject to evaluation.
###
# Interpretation:
# A function can be used to generate in-line Bash commands.
# A simple function where the first word of result IS a bash command:
###
_print() {
echo -n 'printf %q '$@
}
echo '- - Outputs of function _print - -'
_print parm1 parm2 # An Output NOT A Command.
echo
$(_print parm1 parm2) # Executes: printf %q parm1 parm2
# See above IFS examples for the
#+ various possibilities.
echo
$(_print $VarSomething) # The predictable result.
echo
# Function variables
# ------------------
echo
echo '- - Function variables - -'
# A variable may represent a signed integer, a string or an array.
# A string may be used like a function name with optional arguments.
# set -vx # Enable if desired
declare -f funcVar #+ in namespace of functions
funcVar=_print # Contains name of function.
$funcVar parm1 # Same as _print at this point.
echo
funcVar=$(_print ) # Contains result of function.
$funcVar # No input, No output.
$funcVar $VarSomething # The predictable result.
echo
funcVar=$(_print $VarSomething) # $VarSomething replaced HERE.
$funcVar # The expansion is part of the
echo #+ variable contents.
funcVar="$(_print $VarSomething)" # $VarSomething replaced HERE.
$funcVar # The expansion is part of the
echo #+ variable contents.
# The difference between the unquoted and the double-quoted versions
#+ above can be seen in the "protect_literal.sh" example.
# The first case above is processed as two, unquoted, Bash-Words.
# The second case above is processed as one, quoted, Bash-Word.
# Delayed replacement
# -------------------
echo
echo '- - Delayed replacement - -'
funcVar="$(_print '$VarSomething')" # No replacement, single Bash-Word.
eval $funcVar # $VarSomething replaced HERE.
echo
VarSomething='NewThing'
eval $funcVar # $VarSomething replaced HERE.
echo
# Restore the original setting trashed above.
VarSomething=Literal
# There are a pair of functions demonstrated in the
#+ "protect_literal.sh" and "unprotect_literal.sh" examples.
# These are general purpose functions for delayed replacement literals
#+ containing variables.
# REVIEW:
# ------
# A string can be considered a Classic-Array of elements (characters).
# A string operation applies to all elements (characters) of the string
#+ (in concept, anyway).
###
# The notation: ${array_name[@]} represents all elements of the
#+ Bash-Array: array_name.
###
# The Extended-Syntax string operations can be applied to all
#+ elements of an array.
###
# This may be thought of as a For-Each operation on a vector of strings.
###
# Parameters are similar to an array.
# The initialization of a parameter array for a script
#+ and a parameter array for a function only differ
#+ in the initialization of ${0}, which never changes its setting.
###
# Subscript zero of the script's parameter array contains
#+ the name of the script.
###
# Subscript zero of a function's parameter array DOES NOT contain
#+ the name of the function.
# The name of the current function is accessed by the $FUNCNAME variable.
###
# A quick, review list follows (quick, not short).
echo
echo '- - Test (but not change) - -'
echo '- null reference -'
echo -n ${VarNull-'NotSet'}' ' # NotSet
echo ${VarNull} # NewLine only
echo -n ${VarNull:-'NotSet'}' ' # NotSet
echo ${VarNull} # Newline only
echo '- null contents -'
echo -n ${VarEmpty-'Empty'}' ' # Only the space
echo ${VarEmpty} # Newline only
echo -n ${VarEmpty:-'Empty'}' ' # Empty
echo ${VarEmpty} # Newline only
echo '- contents -'
echo ${VarSomething-'Content'} # Literal
echo ${VarSomething:-'Content'} # Literal
echo '- Sparse Array -'
echo ${ArrayVar[@]-'not set'}
# ASCII-Art time
# State Y==yes, N==no
# - :-
# Unset Y Y ${# ... } == 0
# Empty N Y ${# ... } == 0
# Contents N N ${# ... } &#62; 0
# Either the first and/or the second part of the tests
#+ may be a command or a function invocation string.
echo
echo '- - Test 1 for undefined - -'
declare -i t
_decT() {
t=$t-1
}
# Null reference, set: t == -1
t=${#VarNull} # Results in zero.
${VarNull- _decT } # Function executes, t now -1.
echo $t
# Null contents, set: t == 0
t=${#VarEmpty} # Results in zero.
${VarEmpty- _decT } # _decT function NOT executed.
echo $t
# Contents, set: t == number of non-null characters
VarSomething='_simple' # Set to valid function name.
t=${#VarSomething} # non-zero length
${VarSomething- _decT } # Function _simple executed.
echo $t # Note the Append-To action.
# Exercise: clean up that example.
unset t
unset _decT
VarSomething=Literal
echo
echo '- - Test and Change - -'
echo '- Assignment if null reference -'
echo -n ${VarNull='NotSet'}' ' # NotSet NotSet
echo ${VarNull}
unset VarNull
echo '- Assignment if null reference -'
echo -n ${VarNull:='NotSet'}' ' # NotSet NotSet
echo ${VarNull}
unset VarNull
echo '- No assignment if null contents -'
echo -n ${VarEmpty='Empty'}' ' # Space only
echo ${VarEmpty}
VarEmpty=''
echo '- Assignment if null contents -'
echo -n ${VarEmpty:='Empty'}' ' # Empty Empty
echo ${VarEmpty}
VarEmpty=''
echo '- No change if already has contents -'
echo ${VarSomething='Content'} # Literal
echo ${VarSomething:='Content'} # Literal
# "Subscript sparse" Bash-Arrays
###
# Bash-Arrays are subscript packed, beginning with
#+ subscript zero unless otherwise specified.
###
# The initialization of ArrayVar was one way
#+ to "otherwise specify". Here is the other way:
###
echo
declare -a ArraySparse
ArraySparse=( [1]=one [2]='' [4]='four' )
# [0]=null reference, [2]=null content, [3]=null reference
echo '- - Array-Sparse List - -'
# Within double-quotes, default IFS, Glob-Pattern
IFS=$'\x20'$'\x09'$'\x0A'
printf %q "${ArraySparse[*]}"
echo
# Note that the output does not distinguish between "null content"
#+ and "null reference".
# Both print as escaped whitespace.
###
# Note also that the output does NOT contain escaped whitespace
#+ for the "null reference(s)" prior to the first defined element.
###
# This behavior of 2.04, 2.05a and 2.05b has been reported
#+ and may change in a future version of Bash.
# To output a sparse array and maintain the [subscript]=value
#+ relationship without change requires a bit of programming.
# One possible code fragment:
###
# local l=${#ArraySparse[@]} # Count of defined elements
# local f=0 # Count of found subscripts
# local i=0 # Subscript to test
( # Anonymous in-line function
for (( l=${#ArraySparse[@]}, f = 0, i = 0 ; f &#60; l ; i++ ))
do
# 'if defined then...'
${ArraySparse[$i]+ eval echo '\ ['$i']='${ArraySparse[$i]} ; (( f++ )) }
done
)
# The reader coming upon the above code fragment cold
#+ might want to review "command lists" and "multiple commands on a line"
#+ in the text of the foregoing "Advanced Bash Scripting Guide."
###
# Note:
# The "read -a array_name" version of the "read" command
#+ begins filling array_name at subscript zero.
# ArraySparse does not define a value at subscript zero.
###
# The user needing to read/write a sparse array to either
#+ external storage or a communications socket must invent
#+ a read/write code pair suitable for their purpose.
###
# Exercise: clean it up.
unset ArraySparse
echo
echo '- - Conditional alternate (But not change)- -'
echo '- No alternate if null reference -'
echo -n ${VarNull+'NotSet'}' '
echo ${VarNull}
unset VarNull
echo '- No alternate if null reference -'
echo -n ${VarNull:+'NotSet'}' '
echo ${VarNull}
unset VarNull
echo '- Alternate if null contents -'
echo -n ${VarEmpty+'Empty'}' ' # Empty
echo ${VarEmpty}
VarEmpty=''
echo '- No alternate if null contents -'
echo -n ${VarEmpty:+'Empty'}' ' # Space only
echo ${VarEmpty}
VarEmpty=''
echo '- Alternate if already has contents -'
# Alternate literal
echo -n ${VarSomething+'Content'}' ' # Content Literal
echo ${VarSomething}
# Invoke function
echo -n ${VarSomething:+ $(_simple) }' ' # SimpleFunc Literal
echo ${VarSomething}
echo
echo '- - Sparse Array - -'
echo ${ArrayVar[@]+'Empty'} # An array of 'Empty'(ies)
echo
echo '- - Test 2 for undefined - -'
declare -i t
_incT() {
t=$t+1
}
# Note:
# This is the same test used in the sparse array
#+ listing code fragment.
# Null reference, set: t == -1
t=${#VarNull}-1 # Results in minus-one.
${VarNull+ _incT } # Does not execute.
echo $t' Null reference'
# Null contents, set: t == 0
t=${#VarEmpty}-1 # Results in minus-one.
${VarEmpty+ _incT } # Executes.
echo $t' Null content'
# Contents, set: t == (number of non-null characters)
t=${#VarSomething}-1 # non-null length minus-one
${VarSomething+ _incT } # Executes.
echo $t' Contents'
# Exercise: clean up that example.
unset t
unset _incT
# ${name?err_msg} ${name:?err_msg}
# These follow the same rules but always exit afterwards
#+ if an action is specified following the question mark.
# The action following the question mark may be a literal
#+ or a function result.
###
# ${name?} ${name:?} are test-only, the return can be tested.
# Element operations
# ------------------
echo
echo '- - Trailing sub-element selection - -'
# Strings, Arrays and Positional parameters
# Call this script with multiple arguments
#+ to see the parameter selections.
echo '- All -'
echo ${VarSomething:0} # all non-null characters
echo ${ArrayVar[@]:0} # all elements with content
echo ${@:0} # all parameters with content;
# ignoring parameter[0]
echo
echo '- All after -'
echo ${VarSomething:1} # all non-null after character[0]
echo ${ArrayVar[@]:1} # all after element[0] with content
echo ${@:2} # all after param[1] with content
echo
echo '- Range after -'
echo ${VarSomething:4:3} # ral
# Three characters after
# character[3]
echo '- Sparse array gotch -'
echo ${ArrayVar[@]:1:2} # four - The only element with content.
# Two elements after (if that many exist).
# the FIRST WITH CONTENTS
#+ (the FIRST WITH CONTENTS is being
#+ considered as if it
#+ were subscript zero).
# Executed as if Bash considers ONLY array elements with CONTENT
# printf %q "${ArrayVar[@]:0:3}" # Try this one
# In versions 2.04, 2.05a and 2.05b,
#+ Bash does not handle sparse arrays as expected using this notation.
#
# The current Bash maintainer, Chet Ramey, has corrected this.
echo '- Non-sparse array -'
echo ${@:2:2} # Two parameters following parameter[1]
# New victims for string vector examples:
stringZ=abcABC123ABCabc
arrayZ=( abcabc ABCABC 123123 ABCABC abcabc )
sparseZ=( [1]='abcabc' [3]='ABCABC' [4]='' [5]='123123' )
echo
echo ' - - Victim string - -'$stringZ'- - '
echo ' - - Victim array - -'${arrayZ[@]}'- - '
echo ' - - Sparse array - -'${sparseZ[@]}'- - '
echo ' - [0]==null ref, [2]==null ref, [4]==null content - '
echo ' - [1]=abcabc [3]=ABCABC [5]=123123 - '
echo ' - non-null-reference count: '${#sparseZ[@]}' elements'
echo
echo '- - Prefix sub-element removal - -'
echo '- - Glob-Pattern match must include the first character. - -'
echo '- - Glob-Pattern may be a literal or a function result. - -'
echo
# Function returning a simple, Literal, Glob-Pattern
_abc() {
echo -n 'abc'
}
echo '- Shortest prefix -'
echo ${stringZ#123} # Unchanged (not a prefix).
echo ${stringZ#$(_abc)} # ABC123ABCabc
echo ${arrayZ[@]#abc} # Applied to each element.
# echo ${sparseZ[@]#abc} # Version-2.05b core dumps.
# Has since been fixed by Chet Ramey.
# The -it would be nice- First-Subscript-Of
# echo ${#sparseZ[@]#*} # This is NOT valid Bash.
echo
echo '- Longest prefix -'
echo ${stringZ##1*3} # Unchanged (not a prefix)
echo ${stringZ##a*C} # abc
echo ${arrayZ[@]##a*c} # ABCABC 123123 ABCABC
# echo ${sparseZ[@]##a*c} # Version-2.05b core dumps.
# Has since been fixed by Chet Ramey.
echo
echo '- - Suffix sub-element removal - -'
echo '- - Glob-Pattern match must include the last character. - -'
echo '- - Glob-Pattern may be a literal or a function result. - -'
echo
echo '- Shortest suffix -'
echo ${stringZ%1*3} # Unchanged (not a suffix).
echo ${stringZ%$(_abc)} # abcABC123ABC
echo ${arrayZ[@]%abc} # Applied to each element.
# echo ${sparseZ[@]%abc} # Version-2.05b core dumps.
# Has since been fixed by Chet Ramey.
# The -it would be nice- Last-Subscript-Of
# echo ${#sparseZ[@]%*} # This is NOT valid Bash.
echo
echo '- Longest suffix -'
echo ${stringZ%%1*3} # Unchanged (not a suffix)
echo ${stringZ%%b*c} # a
echo ${arrayZ[@]%%b*c} # a ABCABC 123123 ABCABC a
# echo ${sparseZ[@]%%b*c} # Version-2.05b core dumps.
# Has since been fixed by Chet Ramey.
echo
echo '- - Sub-element replacement - -'
echo '- - Sub-element at any location in string. - -'
echo '- - First specification is a Glob-Pattern - -'
echo '- - Glob-Pattern may be a literal or Glob-Pattern function result. - -'
echo '- - Second specification may be a literal or function result. - -'
echo '- - Second specification may be unspecified. Pronounce that'
echo ' as: Replace-With-Nothing (Delete) - -'
echo
# Function returning a simple, Literal, Glob-Pattern
_123() {
echo -n '123'
}
echo '- Replace first occurrence -'
echo ${stringZ/$(_123)/999} # Changed (123 is a component).
echo ${stringZ/ABC/xyz} # xyzABC123ABCabc
echo ${arrayZ[@]/ABC/xyz} # Applied to each element.
echo ${sparseZ[@]/ABC/xyz} # Works as expected.
echo
echo '- Delete first occurrence -'
echo ${stringZ/$(_123)/}
echo ${stringZ/ABC/}
echo ${arrayZ[@]/ABC/}
echo ${sparseZ[@]/ABC/}
# The replacement need not be a literal,
#+ since the result of a function invocation is allowed.
# This is general to all forms of replacement.
echo
echo '- Replace first occurrence with Result-Of -'
echo ${stringZ/$(_123)/$(_simple)} # Works as expected.
echo ${arrayZ[@]/ca/$(_simple)} # Applied to each element.
echo ${sparseZ[@]/ca/$(_simple)} # Works as expected.
echo
echo '- Replace all occurrences -'
echo ${stringZ//[b2]/X} # X-out b's and 2's
echo ${stringZ//abc/xyz} # xyzABC123ABCxyz
echo ${arrayZ[@]//abc/xyz} # Applied to each element.
echo ${sparseZ[@]//abc/xyz} # Works as expected.
echo
echo '- Delete all occurrences -'
echo ${stringZ//[b2]/}
echo ${stringZ//abc/}
echo ${arrayZ[@]//abc/}
echo ${sparseZ[@]//abc/}
echo
echo '- - Prefix sub-element replacement - -'
echo '- - Match must include the first character. - -'
echo
echo '- Replace prefix occurrences -'
echo ${stringZ/#[b2]/X} # Unchanged (neither is a prefix).
echo ${stringZ/#$(_abc)/XYZ} # XYZABC123ABCabc
echo ${arrayZ[@]/#abc/XYZ} # Applied to each element.
echo ${sparseZ[@]/#abc/XYZ} # Works as expected.
echo
echo '- Delete prefix occurrences -'
echo ${stringZ/#[b2]/}
echo ${stringZ/#$(_abc)/}
echo ${arrayZ[@]/#abc/}
echo ${sparseZ[@]/#abc/}
echo
echo '- - Suffix sub-element replacement - -'
echo '- - Match must include the last character. - -'
echo
echo '- Replace suffix occurrences -'
echo ${stringZ/%[b2]/X} # Unchanged (neither is a suffix).
echo ${stringZ/%$(_abc)/XYZ} # abcABC123ABCXYZ
echo ${arrayZ[@]/%abc/XYZ} # Applied to each element.
echo ${sparseZ[@]/%abc/XYZ} # Works as expected.
echo
echo '- Delete suffix occurrences -'
echo ${stringZ/%[b2]/}
echo ${stringZ/%$(_abc)/}
echo ${arrayZ[@]/%abc/}
echo ${sparseZ[@]/%abc/}
echo
echo '- - Special cases of null Glob-Pattern - -'
echo
echo '- Prefix all -'
# null substring pattern means 'prefix'
echo ${stringZ/#/NEW} # NEWabcABC123ABCabc
echo ${arrayZ[@]/#/NEW} # Applied to each element.
echo ${sparseZ[@]/#/NEW} # Applied to null-content also.
# That seems reasonable.
echo
echo '- Suffix all -'
# null substring pattern means 'suffix'
echo ${stringZ/%/NEW} # abcABC123ABCabcNEW
echo ${arrayZ[@]/%/NEW} # Applied to each element.
echo ${sparseZ[@]/%/NEW} # Applied to null-content also.
# That seems reasonable.
echo
echo '- - Special case For-Each Glob-Pattern - -'
echo '- - - - This is a nice-to-have dream - - - -'
echo
_GenFunc() {
echo -n ${0} # Illustration only.
# Actually, that would be an arbitrary computation.
}
# All occurrences, matching the AnyThing pattern.
# Currently //*/ does not match null-content nor null-reference.
# /#/ and /%/ does match null-content but not null-reference.
echo ${sparseZ[@]//*/$(_GenFunc)}
# A possible syntax would be to make
#+ the parameter notation used within this construct mean:
# ${1} - The full element
# ${2} - The prefix, if any, to the matched sub-element
# ${3} - The matched sub-element
# ${4} - The suffix, if any, to the matched sub-element
#
# echo ${sparseZ[@]//*/$(_GenFunc ${3})} # Same as ${1} here.
# Perhaps it will be implemented in a future version of Bash.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="TESTEXECTIME"
></A
><P
><B
>Example A-59. Testing execution times of various commands</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# test-execution-time.sh
# Example by Erik Brandsberg, for testing execution time
#+ of certain operations.
# Referenced in the "Optimizations" section of "Miscellany" chapter.
count=50000
echo "Math tests"
echo "Math via \$(( ))"
time for (( i=0; i&#60; $count; i++))
do
result=$(( $i%2 ))
done
echo "Math via *expr*:"
time for (( i=0; i&#60; $count; i++))
do
result=`expr "$i%2"`
done
echo "Math via *let*:"
time for (( i=0; i&#60; $count; i++))
do
let result=$i%2
done
echo
echo "Conditional testing tests"
echo "Test via case:"
time for (( i=0; i&#60; $count; i++))
do
case $(( $i%2 )) in
0) : ;;
1) : ;;
esac
done
echo "Test with if [], no quotes:"
time for (( i=0; i&#60; $count; i++))
do
if [ $(( $i%2 )) = 0 ]; then
:
else
:
fi
done
echo "Test with if [], quotes:"
time for (( i=0; i&#60; $count; i++))
do
if [ "$(( $i%2 ))" = "0" ]; then
:
else
:
fi
done
echo "Test with if [], using -eq:"
time for (( i=0; i&#60; $count; i++))
do
if [ $(( $i%2 )) -eq 0 ]; then
:
else
:
fi
done
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ASSOCARRTEST"
></A
><P
><B
>Example A-60. Associative arrays vs. conventional arrays (execution
times)</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# assoc-arr-test.sh
# Benchmark test script to compare execution times of
# numeric-indexed array vs. associative array.
# Thank you, Erik Brandsberg.
count=100000 # May take a while for some of the tests below.
declare simple # Can change to 20000, if desired.
declare -a array1
declare -A array2
declare -a array3
declare -A array4
echo "===Assignment tests==="
echo
echo "Assigning a simple variable:"
# References $i twice to equalize lookup times.
time for (( i=0; i&#60; $count; i++)); do
simple=$i$i
done
echo "---"
echo "Assigning a numeric index array entry:"
time for (( i=0; i&#60; $count; i++)); do
array1[$i]=$i
done
echo "---"
echo "Overwriting a numeric index array entry:"
time for (( i=0; i&#60; $count; i++)); do
array1[$i]=$i
done
echo "---"
echo "Linear reading of numeric index array:"
time for (( i=0; i&#60; $count; i++)); do
simple=array1[$i]
done
echo "---"
echo "Assigning an associative array entry:"
time for (( i=0; i&#60; $count; i++)); do
array2[$i]=$i
done
echo "---"
echo "Overwriting an associative array entry:"
time for (( i=0; i&#60; $count; i++)); do
array2[$i]=$i
done
echo "---"
echo "Linear reading an associative array entry:"
time for (( i=0; i&#60; $count; i++)); do
simple=array2[$i]
done
echo "---"
echo "Assigning a random number to a simple variable:"
time for (( i=0; i&#60; $count; i++)); do
simple=$RANDOM
done
echo "---"
echo "Assign a sparse numeric index array entry randomly into 64k cells:"
time for (( i=0; i&#60; $count; i++)); do
array3[$RANDOM]=$i
done
echo "---"
echo "Reading sparse numeric index array entry:"
time for value in "${array3[@]}"i; do
simple=$value
done
echo "---"
echo "Assigning a sparse associative array entry randomly into 64k cells:"
time for (( i=0; i&#60; $count; i++)); do
array4[$RANDOM]=$i
done
echo "---"
echo "Reading sparse associative index array entry:"
time for value in "${array4[@]}"; do
simple=$value
done
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="REFCARDS"
></A
>Appendix B. Reference Cards</H1
><P
>The following reference cards provide a useful
<EM
>summary</EM
> of certain scripting concepts.
The foregoing text treats these matters in more depth, as well as
giving usage examples.</P
><P
><A
NAME="SPECSHVARTAB"
></A
></P
><DIV
CLASS="TABLE"
><A
NAME="AEN22402"
></A
><P
><B
>Table B-1. Special Shell Variables</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Variable</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Meaning</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>$0</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Filename of script</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>$1</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Positional parameter #1</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>$2 - $9</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Positional parameters #2 - #9</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${10}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Positional parameter #10</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>$#</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Number of positional parameters</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>"$*"</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>All the positional parameters (as a single word) *</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>"$@"</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>All the positional parameters (as separate strings)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${#*}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Number of positional parameters</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${#@}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Number of positional parameters</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>$?</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Return value</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>$$</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Process ID (PID) of script</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>$-</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Flags passed to script (using
<I
CLASS="FIRSTTERM"
>set</I
>)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>$_</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Last argument of previous command</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>$!</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Process ID (PID) of last job run in background</TD
></TR
></TBODY
></TABLE
></DIV
><P
><B
CLASS="COMMAND"
>*</B
> <EM
>Must be quoted</EM
>,
otherwise it defaults to
<TT
CLASS="VARNAME"
>$@</TT
>.</P
><P
><A
NAME="BINCOMPTAB"
></A
></P
><DIV
CLASS="TABLE"
><A
NAME="AEN22473"
></A
><P
><B
>Table B-2. TEST Operators: Binary Comparison</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Operator</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Meaning</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>-----</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Operator</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Meaning</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#ICOMPARISON1"
>Arithmetic
Comparison</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#SCOMPARISON1"
>String
Comparison</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-eq</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Equal to</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>=</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Equal to</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>==</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Equal to</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-ne</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Not equal to</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>!=</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Not equal to</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-lt</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Less than</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>\&#60;</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Less than (<A
HREF="#ASCIIDEF"
>ASCII</A
>) *</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-le</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Less than or equal to</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-gt</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Greater than</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>\&#62;</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Greater than (ASCII) *</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-ge</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Greater than or equal to</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-z</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>String is empty</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-n</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>String is not empty</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Arithmetic Comparison</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#DBLPRX"
>within double
parentheses</A
> (( ... ))</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#62;</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Greater than</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#62;=</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Greater than or equal to</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#60;</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Less than</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#60;=</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Less than or equal to</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
></TBODY
></TABLE
></DIV
><P
><B
CLASS="COMMAND"
>*</B
> <EM
>If within a
double-bracket</EM
> <SPAN
CLASS="TOKEN"
>[[ ... ]]</SPAN
> <EM
>test construct,
then no escape</EM
> <SPAN
CLASS="TOKEN"
>\</SPAN
> <EM
>is
needed.</EM
></P
><P
><A
NAME="FILESTAB"
></A
></P
><DIV
CLASS="TABLE"
><A
NAME="AEN22593"
></A
><P
><B
>Table B-3. TEST Operators: Files</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Operator</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Tests Whether</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>-----</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Operator</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Tests Whether</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-e</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File exists</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-s</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File is not zero size</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-f</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File is a <I
CLASS="FIRSTTERM"
>regular</I
> file</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-d</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File is a <I
CLASS="FIRSTTERM"
>directory</I
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-r</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File has <I
CLASS="FIRSTTERM"
>read</I
>
permission</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-h</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File is a <A
HREF="#SYMLINKREF"
>symbolic link</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-w</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File has <I
CLASS="FIRSTTERM"
>write</I
>
permission</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-L</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File is a <I
CLASS="FIRSTTERM"
>symbolic link</I
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-x</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File has <I
CLASS="FIRSTTERM"
>execute</I
>
permission</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-b</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File is a <A
HREF="#BLOCKDEVREF"
>block
device</A
></TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-c</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File is a <A
HREF="#CHARDEVREF"
>character
device</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-g</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><I
CLASS="FIRSTTERM"
>sgid</I
> flag set</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-p</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File is a <A
HREF="#PIPEREF"
>pipe</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-u</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><I
CLASS="FIRSTTERM"
>suid</I
> flag set</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-S</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File is a <A
HREF="#SOCKETREF"
>socket</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-k</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><SPAN
CLASS="QUOTE"
>"sticky bit"</SPAN
> set</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-t</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File is associated with a
<I
CLASS="FIRSTTERM"
>terminal</I
></TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-N</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File modified since it was last read</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>F1 -nt F2</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File F1 is <EM
>newer</EM
> than F2 *</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-O</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>You own the file</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>F1 -ot F2</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>File F1 is <EM
>older</EM
> than F2 *</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>-G</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><I
CLASS="FIRSTTERM"
>Group id</I
> of file same as
yours</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>F1 -ef F2</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Files F1 and F2 are <I
CLASS="FIRSTTERM"
>hard links</I
>
to the same file *</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>!</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>NOT (inverts sense of above tests)</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
><TD
>&nbsp;</TD
></TR
></TBODY
></TABLE
></DIV
><P
><B
CLASS="COMMAND"
>*</B
> <I
CLASS="FIRSTTERM"
>Binary</I
> operator
(requires two operands).</P
><P
><A
NAME="PARSUBTAB"
></A
></P
><DIV
CLASS="TABLE"
><A
NAME="AEN22728"
></A
><P
><B
>Table B-4. Parameter Substitution and Expansion</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Expression</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Meaning</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${var}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Value of <TT
CLASS="PARAMETER"
><I
>var</I
></TT
> (same as
<TT
CLASS="PARAMETER"
><I
>$var</I
></TT
>)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${var-$DEFAULT}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>If <TT
CLASS="PARAMETER"
><I
>var</I
></TT
> not set, <A
HREF="#EVALREF"
>evaluate</A
> expression
as <TT
CLASS="PARAMETER"
><I
>$DEFAULT</I
></TT
> *</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${var:-$DEFAULT}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>If <TT
CLASS="PARAMETER"
><I
>var</I
></TT
> not set or is empty,
<I
CLASS="FIRSTTERM"
>evaluate</I
> expression as
<TT
CLASS="PARAMETER"
><I
>$DEFAULT</I
></TT
>
*</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${var=$DEFAULT}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>If <TT
CLASS="PARAMETER"
><I
>var</I
></TT
> not set, evaluate expression
as <TT
CLASS="PARAMETER"
><I
>$DEFAULT</I
></TT
> *</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${var:=$DEFAULT}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>If <TT
CLASS="PARAMETER"
><I
>var</I
></TT
> not set or is empty, evaluate expression
as <TT
CLASS="PARAMETER"
><I
>$DEFAULT</I
></TT
> *</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${var+$OTHER}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>If <TT
CLASS="PARAMETER"
><I
>var</I
></TT
> set, evaluate expression as
<TT
CLASS="PARAMETER"
><I
>$OTHER</I
></TT
>, otherwise as null string</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${var:+$OTHER}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>If <TT
CLASS="PARAMETER"
><I
>var</I
></TT
> set, evaluate expression as
<TT
CLASS="PARAMETER"
><I
>$OTHER</I
></TT
>, otherwise as null string</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${var?$ERR_MSG}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>If <TT
CLASS="PARAMETER"
><I
>var</I
></TT
> not set, print
<TT
CLASS="PARAMETER"
><I
>$ERR_MSG</I
></TT
> and abort script
with an exit status of <SPAN
CLASS="ERRORCODE"
>1</SPAN
>.*</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${var:?$ERR_MSG}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>If <TT
CLASS="PARAMETER"
><I
>var</I
></TT
> not set, print
<TT
CLASS="PARAMETER"
><I
>$ERR_MSG</I
></TT
> and abort script
with an exit status of <SPAN
CLASS="ERRORCODE"
>1</SPAN
>.*</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${!varprefix*}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Matches all previously declared variables beginning with
<TT
CLASS="PARAMETER"
><I
>varprefix</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${!varprefix@}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Matches all previously declared variables beginning with
<TT
CLASS="PARAMETER"
><I
>varprefix</I
></TT
></TD
></TR
></TBODY
></TABLE
></DIV
><P
><B
CLASS="COMMAND"
>*</B
> If <TT
CLASS="PARAMETER"
><I
>var</I
></TT
>
<EM
>is</EM
> set, evaluate the expression as
<TT
CLASS="PARAMETER"
><I
>$var</I
></TT
> with no side-effects.</P
><P
><B
CLASS="COMMAND"
># Note</B
> that some of the above behavior
of operators has changed from earlier versions of Bash.</P
><P
><A
NAME="STRINGOPSTAB"
></A
></P
><DIV
CLASS="TABLE"
><A
NAME="AEN22828"
></A
><P
><B
>Table B-5. String Operations</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Expression</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Meaning</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${#string}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Length of <TT
CLASS="PARAMETER"
><I
>$string</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${string:position}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Extract substring from <TT
CLASS="PARAMETER"
><I
>$string</I
></TT
>
at <TT
CLASS="PARAMETER"
><I
>$position</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${string:position:length}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Extract <TT
CLASS="PARAMETER"
><I
>$length</I
></TT
>
characters substring from <TT
CLASS="PARAMETER"
><I
>$string</I
></TT
>
at <TT
CLASS="PARAMETER"
><I
>$position</I
></TT
> [zero-indexed,
first character is at position 0]</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${string#substring}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Strip shortest match of
<TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
> from front of
<TT
CLASS="PARAMETER"
><I
>$string</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${string##substring}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Strip longest match of
<TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
> from front of
<TT
CLASS="PARAMETER"
><I
>$string</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${string%substring}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Strip shortest match of
<TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
> from back of
<TT
CLASS="PARAMETER"
><I
>$string</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${string%%substring}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Strip longest match of
<TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
> from back of
<TT
CLASS="PARAMETER"
><I
>$string</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${string/substring/replacement}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Replace first match of
<TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
> with
<TT
CLASS="PARAMETER"
><I
>$replacement</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${string//substring/replacement}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Replace <EM
>all</EM
> matches of
<TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
> with
<TT
CLASS="PARAMETER"
><I
>$replacement</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${string/#substring/replacement}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>If <TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
>
matches <EM
>front</EM
> end of
<TT
CLASS="PARAMETER"
><I
>$string</I
></TT
>, substitute
<TT
CLASS="PARAMETER"
><I
>$replacement</I
></TT
> for
<TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${string/%substring/replacement}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>If <TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
>
matches <EM
>back</EM
> end of
<TT
CLASS="PARAMETER"
><I
>$string</I
></TT
>, substitute
<TT
CLASS="PARAMETER"
><I
>$replacement</I
></TT
> for
<TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>expr match "$string" '$substring'</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Length of matching <TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
>*
at beginning of <TT
CLASS="PARAMETER"
><I
>$string</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>expr "$string" : '$substring'</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Length of matching <TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
>*
at beginning of <TT
CLASS="PARAMETER"
><I
>$string</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>expr index "$string" $substring</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Numerical position in <TT
CLASS="PARAMETER"
><I
>$string</I
></TT
>
of first character in <TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
>*
that matches [0 if no match, first character counts as
position 1]</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>expr substr $string $position
$length</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Extract <TT
CLASS="PARAMETER"
><I
>$length</I
></TT
> characters
from <TT
CLASS="PARAMETER"
><I
>$string</I
></TT
> starting at
<TT
CLASS="PARAMETER"
><I
>$position</I
></TT
> [0 if no match, first
character counts as position 1]</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>expr match "$string"
'\($substring\)'</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Extract <TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
>*, searching
from beginning of <TT
CLASS="PARAMETER"
><I
>$string</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>expr "$string" :
'\($substring\)'</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Extract <TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
>* , searching
from beginning of <TT
CLASS="PARAMETER"
><I
>$string</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>expr match "$string"
'.*\($substring\)'</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Extract <TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
>*, searching
from end of <TT
CLASS="PARAMETER"
><I
>$string</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>expr "$string" :
'.*\($substring\)'</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Extract <TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
>*, searching
from end of <TT
CLASS="PARAMETER"
><I
>$string</I
></TT
></TD
></TR
></TBODY
></TABLE
></DIV
><P
><B
CLASS="COMMAND"
>*</B
> Where <TT
CLASS="PARAMETER"
><I
>$substring</I
></TT
> is a
<A
HREF="#REGEXREF"
>Regular Expression</A
>.</P
><P
><A
NAME="MISCTAB"
></A
></P
><DIV
CLASS="TABLE"
><A
NAME="AEN22979"
></A
><P
><B
>Table B-6. Miscellaneous Constructs</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Expression</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Interpretation</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#BRACKETSREF"
>Brackets</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>if [ CONDITION ]</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#LEFTBRACKET"
>Test construct</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>if [[ CONDITION ]]</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#DBLBRACKETS"
>Extended test construct</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>Array[1]=element1</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#ARRAYREF"
>Array initialization</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
> [a-z]</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#BRACKETSREF"
>Range of
characters</A
> within a <A
HREF="#REGEXREF"
>Regular
Expression</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Curly Brackets</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${variable}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#PARAMSUBREF"
>Parameter substitution</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>${!variable}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#IVRREF"
>Indirect variable reference</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>{ command1; command2; . . . commandN; }</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#CODEBLOCKREF"
>Block of code</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>{string1,string2,string3,...}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#BRACEEXPREF"
>Brace expansion</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>{a..z}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#BRACEEXPREF3"
>Extended brace expansion</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>{}</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Text replacement, after <A
HREF="#CURLYBRACKETSREF"
>find</A
> and <A
HREF="#XARGSCURLYREF"
>xargs</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#PARENSREF"
>Parentheses</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>( command1; command2 )</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Command group executed within a <A
HREF="#SUBSHELLSREF"
>subshell</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>Array=(element1 element2 element3)</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#ARRAYINIT0"
>Array initialization</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>result=$(COMMAND)</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#CSPARENS"
>Command substitution</A
>,
new style</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#62;(COMMAND)</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#PROCESSSUBREF"
>Process substitution</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#60;(COMMAND)</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Process substitution</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#DBLPARENS"
>Double Parentheses</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>(( var = 78 ))</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#DBLPARENSREF"
>Integer arithmetic</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>var=$(( 20 + 5 ))</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Integer arithmetic, with variable assignment</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>(( var++ ))</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><I
CLASS="FIRSTTERM"
>C-style</I
> <A
HREF="#PLUSPLUSREF"
> variable increment</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>(( var-- ))</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><I
CLASS="FIRSTTERM"
>C-style</I
> <A
HREF="#PLUSPLUSREF"
> variable decrement</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>(( var0 = var1&#60;98?9:21 ))</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><I
CLASS="FIRSTTERM"
>C-style</I
> <A
HREF="#CSTRINARY"
> ternary</A
> operation</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#QUOTINGREF"
>Quoting</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>"$variable"</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#DBLQUO"
>"Weak" quoting</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>'string'</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#SNGLQUO"
>'Strong' quoting</A
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#BACKQUOTESREF"
>Back Quotes</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&nbsp;</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>result=`COMMAND`</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#COMMANDSUBREF"
>Command
substitution</A
>, classic style</TD
></TR
></TBODY
></TABLE
></DIV
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="SEDAWK"
></A
>Appendix C. A Sed and Awk Micro-Primer</H1
><P
><A
NAME="SEDREF"
></A
></P
><P
>This is a very brief introduction to the <B
CLASS="COMMAND"
>sed</B
>
and <B
CLASS="COMMAND"
>awk</B
> text processing utilities. We will
deal with only a few basic commands here, but that will suffice
for understanding simple sed and awk constructs within shell
scripts.</P
><P
><B
CLASS="COMMAND"
>sed</B
>: a non-interactive
text file editor</P
><P
><B
CLASS="COMMAND"
>awk</B
>: a field-oriented pattern processing
language with a <B
CLASS="COMMAND"
>C</B
>-style syntax</P
><P
>For all their differences, the two utilities share a similar
invocation syntax, use <A
HREF="#REGEXREF"
>regular
expressions </A
>, read input by default
from <TT
CLASS="FILENAME"
>stdin</TT
>, and output to
<TT
CLASS="FILENAME"
>stdout</TT
>. These are well-behaved UNIX tools,
and they work together well. The output from one can be piped
to the other, and their combined capabilities give shell scripts
some of the power of <A
HREF="#PERLREF"
>Perl</A
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>One important difference between the utilities is
that while shell scripts can easily pass arguments to sed, it
is more cumbersome for awk (see <A
HREF="#COLTOTALER"
>Example 36-5</A
>
and <A
HREF="#COLTOTALER2"
>Example 28-2</A
>).
</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="AEN23170"
></A
>C.1. Sed</H1
><P
><I
CLASS="FIRSTTERM"
>Sed</I
> is a non-interactive
<A
NAME="AEN23174"
HREF="#FTN.AEN23174"
><SPAN
CLASS="footnote"
>[141]</SPAN
></A
>
<B
CLASS="COMMAND"
>s</B
>tream <B
CLASS="COMMAND"
>ed</B
>itor. It
receives text input, whether from <TT
CLASS="FILENAME"
>stdin</TT
>
or from a file, performs certain operations on specified lines
of the input, one line at a time, then outputs the result to
<TT
CLASS="FILENAME"
>stdout</TT
> or to a file. Within a shell script,
<I
CLASS="FIRSTTERM"
>sed</I
> is usually one of several tool
components in a pipe.</P
><P
><I
CLASS="FIRSTTERM"
>Sed</I
> determines which lines of
its input that it will operate on from the <I
CLASS="FIRSTTERM"
>address
range</I
> passed to it.
<A
NAME="AEN23185"
HREF="#FTN.AEN23185"
><SPAN
CLASS="footnote"
>[142]</SPAN
></A
>
Specify this address range either by line number or by a
pattern to match. For example, <TT
CLASS="REPLACEABLE"
><I
>3d</I
></TT
>
signals <I
CLASS="FIRSTTERM"
>sed</I
> to delete line 3 of the
input, and <TT
CLASS="REPLACEABLE"
><I
>/Windows/d</I
></TT
> tells sed
that you want every line of the input containing a match to
<SPAN
CLASS="QUOTE"
>"Windows"</SPAN
> deleted.</P
><P
>Of all the operations in the <I
CLASS="FIRSTTERM"
>sed</I
>
toolkit, we will focus primarily on the three most commonly
used ones. These are <B
CLASS="COMMAND"
>p</B
>rinting (to
<TT
CLASS="FILENAME"
>stdout</TT
>), <B
CLASS="COMMAND"
>d</B
>eletion,
and <B
CLASS="COMMAND"
>s</B
>ubstitution.</P
><P
><A
NAME="SEDBASICTABLE"
></A
></P
><DIV
CLASS="TABLE"
><A
NAME="AEN23200"
></A
><P
><B
>Table C-1. Basic sed operators</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Operator</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Name</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Effect</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>[address-range]/p</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>print</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Print [specified address range]</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>[address-range]/d</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>delete</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Delete [specified address range]</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>s/pattern1/pattern2/</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>substitute</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Substitute pattern2 for first instance of pattern1 in a line</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>[address-range]/s/pattern1/pattern2/</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>substitute</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Substitute pattern2 for first instance of pattern1 in a
line, over <TT
CLASS="REPLACEABLE"
><I
>address-range</I
></TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>[address-range]/y/pattern1/pattern2/</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>transform</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>replace any character in pattern1 with the
corresponding character in pattern2, over
<TT
CLASS="REPLACEABLE"
><I
>address-range</I
></TT
> (equivalent of
<B
CLASS="COMMAND"
>tr</B
>)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>[address] i pattern Filename</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>insert</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Insert pattern at address indicated in file Filename.
Usually used with <TT
CLASS="OPTION"
>-i</TT
>
<TT
CLASS="REPLACEABLE"
><I
>in-place</I
></TT
> option.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>g</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>global</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Operate on <EM
>every</EM
> pattern match
within each matched line of input</TD
></TR
></TBODY
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Unless the <TT
CLASS="OPTION"
>g</TT
>
(<I
CLASS="FIRSTTERM"
>global</I
>) operator is appended to a
<I
CLASS="FIRSTTERM"
>substitute</I
> command, the substitution
operates only on the <EM
>first</EM
> instance of a
pattern match within each line.</P
></TD
></TR
></TABLE
></DIV
><P
>From the command-line and in a shell script, a sed operation may
require quoting and certain options.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>sed -e '/^$/d' $filename
# The -e option causes the next string to be interpreted as an editing instruction.
# (If passing only a single instruction to sed, the "-e" is optional.)
# The "strong" quotes ('') protect the RE characters in the instruction
#+ from reinterpretation as special characters by the body of the script.
# (This reserves RE expansion of the instruction for sed.)
#
# Operates on the text contained in file $filename.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>In certain cases, a <I
CLASS="FIRSTTERM"
>sed</I
> editing command will
not work with single quotes.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>filename=file1.txt
pattern=BEGIN
sed "/^$pattern/d" "$filename" # Works as specified.
# sed '/^$pattern/d' "$filename" has unexpected results.
# In this instance, with strong quoting (' ... '),
#+ "$pattern" will not expand to "BEGIN".</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><I
CLASS="FIRSTTERM"
>Sed</I
> uses the <TT
CLASS="OPTION"
>-e</TT
>
option to specify that the following string is an instruction
or set of instructions. If there is only a single instruction
contained in the string, then this may be omitted.</P
></TD
></TR
></TABLE
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>sed -n '/xzy/p' $filename
# The -n option tells sed to print only those lines matching the pattern.
# Otherwise all input lines would print.
# The -e option not necessary here since there is only a single editing instruction.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
><A
NAME="SEDOPTABLE"
></A
></P
><DIV
CLASS="TABLE"
><A
NAME="AEN23271"
></A
><P
><B
>Table C-2. Examples of sed operators</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Notation</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Effect</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>8d</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Delete 8th line of input.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>/^$/d</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Delete all blank lines.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>1,/^$/d</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Delete from beginning of input up to, and including
first blank line.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>/Jones/p</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Print only lines containing <SPAN
CLASS="QUOTE"
>"Jones"</SPAN
> (with
<SPAN
CLASS="TOKEN"
>-n</SPAN
> option).</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>s/Windows/Linux/</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Substitute <SPAN
CLASS="QUOTE"
>"Linux"</SPAN
> for first instance
of <SPAN
CLASS="QUOTE"
>"Windows"</SPAN
> found in each input line.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>s/BSOD/stability/g</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Substitute <SPAN
CLASS="QUOTE"
>"stability"</SPAN
> for every instance
of <SPAN
CLASS="QUOTE"
>"BSOD"</SPAN
> found in each input line.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>s/ *$//</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Delete all spaces at the end of every line.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>s/00*/0/g</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Compress all consecutive sequences of zeroes into
a single zero.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>echo "Working on it." | sed -e '1i How far are you along?'</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Prints "How far are you along?" as first line,
"Working on it" as second.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>5i 'Linux is great.' file.txt</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Inserts 'Linux is great.' at line 5 of the file
file.txt.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>/GUI/d</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Delete all lines containing <SPAN
CLASS="QUOTE"
>"GUI"</SPAN
>.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>s/GUI//g</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Delete all instances of <SPAN
CLASS="QUOTE"
>"GUI"</SPAN
>, leaving the
remainder of each line intact.</TD
></TR
></TBODY
></TABLE
></DIV
><P
>Substituting a zero-length string for another is equivalent
to deleting that string within a line of input. This leaves the
remainder of the line intact. Applying <TT
CLASS="USERINPUT"
><B
>s/GUI//</B
></TT
>
to the line
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="USERINPUT"
><B
>The most important parts of any application are its GUI and sound effects</B
></TT
></PRE
></FONT
></TD
></TR
></TABLE
>
results in
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="COMPUTEROUTPUT"
>The most important parts of any application are its and sound effects</TT
></PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>A backslash forces the <B
CLASS="COMMAND"
>sed</B
> replacement
command to continue on to the next line. This has the effect of
using the <I
CLASS="FIRSTTERM"
>newline</I
> at the end of the first
line as the <I
CLASS="FIRSTTERM"
>replacement string</I
>.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>s/^ */\
/g</PRE
></FONT
></TD
></TR
></TABLE
>
This substitution replaces line-beginning spaces with a
newline. The net result is to replace paragraph indents with a
blank line between paragraphs.</P
><P
>An address range followed by one or more operations may require
open and closed curly brackets, with appropriate newlines.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>/[0-9A-Za-z]/,/^$/{
/^$/d
}</PRE
></FONT
></TD
></TR
></TABLE
>
This deletes only the first of each set of consecutive blank
lines. That might be useful for single-spacing a text file,
but retaining the blank line(s) between paragraphs.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>The usual delimiter that <I
CLASS="FIRSTTERM"
>sed</I
> uses is
<SPAN
CLASS="TOKEN"
>/</SPAN
>. However, <EM
>sed</EM
> allows other
delimiters, such as <SPAN
CLASS="TOKEN"
>%</SPAN
>. This is useful when
<SPAN
CLASS="TOKEN"
>/</SPAN
> is part of a replacement string, as in a file pathname.
See <A
HREF="#FINDSTRING"
>Example 11-10</A
> and <A
HREF="#STRIPC"
>Example 16-32</A
>.</P
></TD
></TR
></TABLE
></DIV
><P
><A
NAME="DOUBLESPACE"
></A
></P
><DIV
CLASS="TIP"
><P
></P
><TABLE
CLASS="TIP"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/tip.gif"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>A quick way to double-space a text file is <TT
CLASS="USERINPUT"
><B
>sed G
filename</B
></TT
>.</P
></TD
></TR
></TABLE
></DIV
><P
>For illustrative examples of sed within shell scripts, see:
<P
></P
><OL
TYPE="1"
><LI
><P
><A
HREF="#EX3"
>Example 36-1</A
></P
></LI
><LI
><P
><A
HREF="#EX4"
>Example 36-2</A
></P
></LI
><LI
><P
><A
HREF="#EX57"
>Example 16-3</A
></P
></LI
><LI
><P
><A
HREF="#RN"
>Example A-2</A
></P
></LI
><LI
><P
><A
HREF="#GRP"
>Example 16-17</A
></P
></LI
><LI
><P
><A
HREF="#COL"
>Example 16-27</A
></P
></LI
><LI
><P
><A
HREF="#BEHEAD"
>Example A-12</A
></P
></LI
><LI
><P
><A
HREF="#TREE"
>Example A-16</A
></P
></LI
><LI
><P
><A
HREF="#TREE2"
>Example A-17</A
></P
></LI
><LI
><P
><A
HREF="#STRIPC"
>Example 16-32</A
></P
></LI
><LI
><P
><A
HREF="#FINDSTRING"
>Example 11-10</A
></P
></LI
><LI
><P
><A
HREF="#BASE"
>Example 16-48</A
></P
></LI
><LI
><P
><A
HREF="#MAILFORMAT"
>Example A-1</A
></P
></LI
><LI
><P
><A
HREF="#RND"
>Example 16-14</A
></P
></LI
><LI
><P
><A
HREF="#WF"
>Example 16-12</A
></P
></LI
><LI
><P
><A
HREF="#LIFESLOW"
>Example A-10</A
></P
></LI
><LI
><P
><A
HREF="#SELFDOCUMENT"
>Example 19-12</A
></P
></LI
><LI
><P
><A
HREF="#DICTLOOKUP"
>Example 16-19</A
></P
></LI
><LI
><P
><A
HREF="#WHX"
>Example A-29</A
></P
></LI
><LI
><P
><A
HREF="#BASHPODDER"
>Example A-31</A
></P
></LI
><LI
><P
><A
HREF="#TOHTML"
>Example A-24</A
></P
></LI
><LI
><P
><A
HREF="#STOPWATCH"
>Example A-43</A
></P
></LI
><LI
><P
><A
HREF="#SEDAPPEND"
>Example A-55</A
></P
></LI
></OL
>
</P
><P
>For a more extensive treatment of <I
CLASS="FIRSTTERM"
>sed</I
>,
refer to the <A
HREF="#DGSEDREF"
>pertinent references</A
>
in the <A
HREF="#BIBLIO"
><I
>Bibliography</I
></A
>.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="AWK"
></A
>C.2. Awk</H1
><P
><A
NAME="AWKREF"
></A
></P
><P
><I
CLASS="FIRSTTERM"
>Awk</I
>
<A
NAME="AEN23443"
HREF="#FTN.AEN23443"
><SPAN
CLASS="footnote"
>[143]</SPAN
></A
>
is a full-featured text processing language with a syntax
reminiscent of <I
CLASS="FIRSTTERM"
>C</I
>. While it possesses an
extensive set of operators and capabilities, we will cover only
a few of these here - the ones most useful in shell scripts.</P
><P
>Awk breaks each line of input passed to it into
<A
NAME="FIELDREF2"
></A
>
<A
HREF="#FIELDREF"
>fields</A
>. By default, a field
is a string of consecutive characters delimited by <A
HREF="#WHITESPACEREF"
>whitespace</A
>, though there are options
for changing this. Awk parses and operates on each separate
field. This makes it ideal for handling structured text files
-- especially tables -- data organized into consistent chunks,
such as rows and columns.</P
><P
><A
HREF="#SNGLQUO"
>Strong quoting</A
> and <A
HREF="#CODEBLOCKREF"
>curly brackets</A
> enclose blocks of
awk code within a shell script.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># $1 is field #1, $2 is field #2, etc.
echo one two | awk '{print $1}'
# one
echo one two | awk '{print $2}'
# two
# But what is field #0 ($0)?
echo one two | awk '{print $0}'
# one two
# All the fields!
awk '{print $3}' $filename
# Prints field #3 of file $filename to stdout.
awk '{print $1 $5 $6}' $filename
# Prints fields #1, #5, and #6 of file $filename.
awk '{print $0}' $filename
# Prints the entire file!
# Same effect as: cat $filename . . . or . . . sed '' $filename</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>We have just seen the awk <I
CLASS="FIRSTTERM"
>print</I
> command
in action. The only other feature of awk we need to deal with
here is variables. Awk handles variables similarly to shell
scripts, though a bit more flexibly.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>{ total += ${column_number} }</PRE
></FONT
></TD
></TR
></TABLE
>
This adds the value of <TT
CLASS="PARAMETER"
><I
>column_number</I
></TT
> to
the running total of <TT
CLASS="PARAMETER"
><I
>total</I
></TT
>&#62;. Finally, to print
<SPAN
CLASS="QUOTE"
>"total"</SPAN
>, there is an <B
CLASS="COMMAND"
>END</B
> command
block, executed after the script has processed all its input.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>END { print total }</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Corresponding to the <B
CLASS="COMMAND"
>END</B
>, there is a
<B
CLASS="COMMAND"
>BEGIN</B
>, for a code block to be performed before awk
starts processing its input.</P
><P
>The following example illustrates how <B
CLASS="COMMAND"
>awk</B
> can
add text-parsing tools to a shell script.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="LETTERCOUNT2"
></A
><P
><B
>Example C-1. Counting Letter Occurrences</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /bin/sh
# letter-count2.sh: Counting letter occurrences in a text file.
#
# Script by nyal [nyal@voila.fr].
# Used in ABS Guide with permission.
# Recommented and reformatted by ABS Guide author.
# Version 1.1: Modified to work with gawk 3.1.3.
# (Will still work with earlier versions.)
INIT_TAB_AWK=""
# Parameter to initialize awk script.
count_case=0
FILE_PARSE=$1
E_PARAMERR=85
usage()
{
echo "Usage: letter-count.sh file letters" 2&#62;&#38;1
# For example: ./letter-count2.sh filename.txt a b c
exit $E_PARAMERR # Too few arguments passed to script.
}
if [ ! -f "$1" ] ; then
echo "$1: No such file." 2&#62;&#38;1
usage # Print usage message and exit.
fi
if [ -z "$2" ] ; then
echo "$2: No letters specified." 2&#62;&#38;1
usage
fi
shift # Letters specified.
for letter in `echo $@` # For each one . . .
do
INIT_TAB_AWK="$INIT_TAB_AWK tab_search[${count_case}] = \
\"$letter\"; final_tab[${count_case}] = 0; "
# Pass as parameter to awk script below.
count_case=`expr $count_case + 1`
done
# DEBUG:
# echo $INIT_TAB_AWK;
cat $FILE_PARSE |
# Pipe the target file to the following awk script.
# ---------------------------------------------------------------------
# Earlier version of script:
# awk -v tab_search=0 -v final_tab=0 -v tab=0 -v \
# nb_letter=0 -v chara=0 -v chara2=0 \
awk \
"BEGIN { $INIT_TAB_AWK } \
{ split(\$0, tab, \"\"); \
for (chara in tab) \
{ for (chara2 in tab_search) \
{ if (tab_search[chara2] == tab[chara]) { final_tab[chara2]++ } } } } \
END { for (chara in final_tab) \
{ print tab_search[chara] \" =&#62; \" final_tab[chara] } }"
# ---------------------------------------------------------------------
# Nothing all that complicated, just . . .
#+ for-loops, if-tests, and a couple of specialized functions.
exit $?
# Compare this script to letter-count.sh.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>For simpler examples of awk within shell scripts, see:
<P
></P
><OL
TYPE="1"
><LI
><P
><A
HREF="#EX44"
>Example 15-14</A
></P
></LI
><LI
><P
><A
HREF="#REDIR4"
>Example 20-8</A
></P
></LI
><LI
><P
><A
HREF="#STRIPC"
>Example 16-32</A
></P
></LI
><LI
><P
><A
HREF="#COLTOTALER"
>Example 36-5</A
></P
></LI
><LI
><P
><A
HREF="#COLTOTALER2"
>Example 28-2</A
></P
></LI
><LI
><P
><A
HREF="#COLTOTALER3"
>Example 15-20</A
></P
></LI
><LI
><P
><A
HREF="#PIDID"
>Example 29-3</A
></P
></LI
><LI
><P
><A
HREF="#CONSTAT"
>Example 29-4</A
></P
></LI
><LI
><P
><A
HREF="#FILEINFO"
>Example 11-3</A
></P
></LI
><LI
><P
><A
HREF="#BLOTOUT"
>Example 16-61</A
></P
></LI
><LI
><P
><A
HREF="#SEEDINGRANDOM"
>Example 9-16</A
></P
></LI
><LI
><P
><A
HREF="#IDELETE"
>Example 16-4</A
></P
></LI
><LI
><P
><A
HREF="#SUBSTRINGEX"
>Example 10-6</A
></P
></LI
><LI
><P
><A
HREF="#SUMPRODUCT"
>Example 36-19</A
></P
></LI
><LI
><P
><A
HREF="#USERLIST"
>Example 11-9</A
></P
></LI
><LI
><P
><A
HREF="#PRASC"
>Example 36-4</A
></P
></LI
><LI
><P
><A
HREF="#HYPOT"
>Example 16-53</A
></P
></LI
><LI
><P
><A
HREF="#ASCII3SH"
>Example T-3</A
></P
></LI
></OL
>
</P
><P
>That's all the awk we'll cover here, folks, but there's lots
more to learn. See the appropriate references in the <A
HREF="#BIBLIO"
><I
>Bibliography</I
></A
>.</P
></DIV
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="PATHMANAGEMENT"
></A
>Appendix D. Parsing and Managing Pathnames</H1
><P
>Emmanual Rouat contributed the following example of parsing
and transforming <I
CLASS="FIRSTTERM"
>filenames</I
> and, in
particular, <A
HREF="#PATHNAMEREF"
>pathnames</A
>. It draws
heavily on the functionality of <I
CLASS="FIRSTTERM"
>sed</I
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/usr/bin/env bash
#-----------------------------------------------------------
# Management of PATH, LD_LIBRARY_PATH, MANPATH variables...
# By Emmanuel Rouat &#60;no-email&#62;
# (Inspired by the bash documentation 'pathfuncs' and on
# discussions found on stackoverflow:
# http://stackoverflow.com/questions/370047/
# http://stackoverflow.com/questions/273909/#346860 )
# Last modified: Sat Sep 22 12:01:55 CEST 2012
#
# The following functions handle spaces correctly.
# These functions belong in .bash_profile rather than in
# .bashrc, I guess.
#
# The modular aspect of these functions should make it easy
# to expand them to handle path substitutions instead
# of path removal etc....
#
# See http://www.catonmat.net/blog/awk-one-liners-explained-part-two/
# (item 43) for an explanation of the 'duplicate-entries' removal
# (it's a nice trick!)
#-----------------------------------------------------------
# Show $@ (usually PATH) as list.
function p_show() { local p="$@" &#38;&#38; for p; do [[ ${!p} ]] &#38;&#38;
echo -e ${!p//:/\\n}; done }
# Filter out empty lines, multiple/trailing slashes, and duplicate entries.
function p_filter()
{ awk '/^[ \t]*$/ {next} {sub(/\/+$/, "");gsub(/\/+/, "/")}!x[$0]++' ;}
# Rebuild list of items into ':' separated word (PATH-like).
function p_build() { paste -sd: ;}
# Clean $1 (typically PATH) and rebuild it
function p_clean()
{ local p=${1} &#38;&#38; eval ${p}='$(p_show ${p} | p_filter | p_build)' ;}
# Remove $1 from $2 (found on stackoverflow, with modifications).
function p_rm()
{ local d=$(echo $1 | p_filter) p=${2} &#38;&#38;
eval ${p}='$(p_show ${p} | p_filter | grep -xv "${d}" | p_build)' ;}
# Same as previous, but filters on a pattern (dangerous...
#+ don't use 'bin' or '/' as pattern!).
function p_rmpat()
{ local d=$(echo $1 | p_filter) p=${2} &#38;&#38; eval ${p}='$(p_show ${p} |
p_filter | grep -v "${d}" | p_build)' ;}
# Delete $1 from $2 and append it cleanly.
function p_append()
{ local d=$(echo $1 | p_filter) p=${2} &#38;&#38; p_rm "${d}" ${p} &#38;&#38;
eval ${p}='$(p_show ${p} d | p_build)' ;}
# Delete $1 from $2 and prepend it cleanly.
function p_prepend()
{ local d=$(echo $1 | p_filter) p=${2} &#38;&#38; p_rm "${d}" ${p} &#38;&#38;
eval ${p}='$(p_show d ${p} | p_build)' ;}
# Some tests:
echo
MYPATH="/bin:/usr/bin/:/bin://bin/"
p_append "/project//my project/bin" MYPATH
echo "Append '/project//my project/bin' to '/bin:/usr/bin/:/bin://bin/'"
echo "(result should be: /bin:/usr/bin:/project/my project/bin)"
echo $MYPATH
echo
MYOTHERPATH="/bin:/usr/bin/:/bin:/project//my project/bin"
p_prepend "/project//my project/bin" MYOTHERPATH
echo "Prepend '/project//my project/bin' \
to '/bin:/usr/bin/:/bin:/project//my project/bin/'"
echo "(result should be: /project/my project/bin:/bin:/usr/bin)"
echo $MYOTHERPATH
echo
p_prepend "/project//my project/bin" FOOPATH # FOOPATH doesn't exist.
echo "Prepend '/project//my project/bin' to an unset variable"
echo "(result should be: /project/my project/bin)"
echo $FOOPATH
echo
BARPATH="/a:/b/://b c://a:/my local pub"
p_clean BARPATH
echo "Clean BARPATH='/a:/b/://b c://a:/my local pub'"
echo "(result should be: /a:/b:/b c:/my local pub)"
echo $BARPATH</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>***</P
><P
>David Wheeler kindly permitted me to use his instructive
examples.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>Doing it correctly: A quick summary
by David Wheeler
http://www.dwheeler.com/essays/filenames-in-shell.html
So, how can you process filenames correctly in shell? Here's a quick
summary about how to do it correctly, for the impatient who "just want the
answer". In short: Double-quote to use "$variable" instead of $variable,
set IFS to just newline and tab, prefix all globs/filenames so they cannot
begin with "-" when expanded, and use one of a few templates that work
correctly. Here are some of those templates that work correctly:
IFS="$(printf '\n\t')"
# Remove SPACE, so filenames with spaces work well.
# Correct glob use:
#+ always use "for" loop, prefix glob, check for existence:
for file in ./* ; do # Use "./*" ... NEVER bare "*" ...
if [ -e "$file" ] ; then # Make sure it isn't an empty match.
COMMAND ... "$file" ...
fi
done
# Correct glob use, but requires nonstandard bash extension.
shopt -s nullglob # Bash extension,
#+ so that empty glob matches will work.
for file in ./* ; do # Use "./*", NEVER bare "*"
COMMAND ... "$file" ...
done
# These handle all filenames correctly;
#+ can be unwieldy if COMMAND is large:
find ... -exec COMMAND... {} \;
find ... -exec COMMAND... {} \+ # If multiple files are okay for COMMAND.
# This skips filenames with control characters
#+ (including tab and newline).
IFS="$(printf '\n\t')"
controlchars="$(printf '*[\001-\037\177]*')"
for file in $(find . ! -name "$controlchars"') ; do
COMMAND "$file" ...
done
# Okay if filenames can't contain tabs or newlines --
#+ beware the assumption.
IFS="$(printf '\n\t')"
for file in $(find .) ; do
COMMAND "$file" ...
done
# Requires nonstandard but common extensions in find and xargs:
find . -print0 | xargs -0 COMMAND
# Requires nonstandard extensions to find and to shell (bash works).
# variables might not stay set once the loop ends:
find . -print0 | while IFS="" read -r -d "" file ; do ...
COMMAND "$file" # Use quoted "$file", not $file, everywhere.
done
# Requires nonstandard extensions to find and to shell (bash works).
# Underlying system must include named pipes (FIFOs)
#+ or the /dev/fd mechanism.
# In this version, variables *do* stay set after the loop ends,
# and you can read from stdin.
#+ (Change the 4 to another number if fd 4 is needed.)
while IFS="" read -r -d "" file &#60;&#38;4 ; do
COMMAND "$file" # Use quoted "$file" -- not $file, everywhere.
done 4&#60; &#60;(find . -print0)
# Named pipe version.
# Requires nonstandard extensions to find and to shell's read (bash ok).
# Underlying system must include named pipes (FIFOs).
# Again, in this version, variables *do* stay set after the loop ends,
# and you can read from stdin.
# (Change the 4 to something else if fd 4 needed).
mkfifo mypipe
find . -print0 &#62; mypipe &#38;
while IFS="" read -r -d "" file &#60;&#38;4 ; do
COMMAND "$file" # Use quoted "$file", not $file, everywhere.
done 4&#60; mypipe</PRE
></FONT
></TD
></TR
></TABLE
></P
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="EXITCODES"
></A
>Appendix E. Exit Codes With Special Meanings</H1
><P
><A
NAME="EXITCODESREF"
></A
></P
><DIV
CLASS="TABLE"
><A
NAME="AEN23549"
></A
><P
><B
>Table E-1. <I
CLASS="FIRSTTERM"
>Reserved</I
> Exit Codes</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Exit Code Number</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Meaning</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Example</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Comments</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>1</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Catchall for general errors</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>let "var1 = 1/0"</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Miscellaneous errors, such as <SPAN
CLASS="QUOTE"
>"divide by
zero"</SPAN
> and other impermissible operations</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>2</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Misuse of shell builtins (according to Bash documentation)</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>empty_function() {}</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><A
HREF="#MISSINGKEYWORD"
>Missing keyword</A
>
or command, or permission problem (and <A
HREF="#DIFFERR2"
><I
CLASS="FIRSTTERM"
>diff</I
> return code
on a failed binary file comparison</A
>).</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>126</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Command invoked cannot execute</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>/dev/null</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Permission problem or command is not an executable</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>127</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><SPAN
CLASS="QUOTE"
>"command not found"</SPAN
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>illegal_command</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Possible problem with <TT
CLASS="VARNAME"
>$PATH</TT
> or a typo</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>128</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Invalid argument to <A
HREF="#EXITCOMMANDREF"
>exit</A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>exit 3.14159</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><B
CLASS="COMMAND"
>exit</B
> takes only integer args in the
range <SPAN
CLASS="RETURNVALUE"
>0 - 255</SPAN
> (see
first footnote)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>128+n</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Fatal error signal <SPAN
CLASS="QUOTE"
>"n"</SPAN
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><I
CLASS="FIRSTTERM"
>kill -9</I
> <TT
CLASS="VARNAME"
>$PPID</TT
> of script</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="USERINPUT"
><B
>$?</B
></TT
> returns
<SPAN
CLASS="ERRORCODE"
>137</SPAN
> (128 + 9)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>130</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Script terminated by Control-C</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><EM
>Ctl-C</EM
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Control-C is fatal error signal
<SPAN
CLASS="ERRORCODE"
>2</SPAN
>, (130 = 128 + 2, see above)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>255*</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Exit status out of range</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>exit <SPAN
CLASS="RETURNVALUE"
>-1</SPAN
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><B
CLASS="COMMAND"
>exit</B
> takes only integer args in the
range <SPAN
CLASS="ERRORCODE"
>0 - 255</SPAN
></TD
></TR
></TBODY
></TABLE
></DIV
><P
>According to the above table, exit codes <SPAN
CLASS="ERRORCODE"
>1 - 2,
126 - 165, and 255</SPAN
>
<A
NAME="AEN23629"
HREF="#FTN.AEN23629"
><SPAN
CLASS="footnote"
>[144]</SPAN
></A
>
have special meanings, and should therefore be avoided for
user-specified exit parameters. Ending a script with <I
CLASS="FIRSTTERM"
>exit
127</I
> would certainly cause confusion when troubleshooting
(is the error code a <SPAN
CLASS="QUOTE"
>"command not found"</SPAN
> or a
user-defined one?). However, many scripts use an <I
CLASS="FIRSTTERM"
>exit
1</I
> as a general bailout-upon-error. Since exit code
<SPAN
CLASS="ERRORCODE"
>1</SPAN
> signifies so many possible errors,
it is not particularly useful in debugging.</P
><P
><A
NAME="SYSEXITSREF"
></A
></P
><P
>There has been an attempt to systematize exit status numbers
(see <TT
CLASS="FILENAME"
>/usr/include/sysexits.h</TT
>),
but this is intended for C and C++ programmers. A similar
standard for scripting might be appropriate. The author of
this document proposes restricting user-defined exit codes to
the range <SPAN
CLASS="RETURNVALUE"
>64 - 113</SPAN
> (in addition to
<SPAN
CLASS="RETURNVALUE"
>0</SPAN
>, for success), to conform with
the C/C++ standard. This would allot 50 valid codes, and make
troubleshooting scripts more straightforward.
<A
NAME="AEN23647"
HREF="#FTN.AEN23647"
><SPAN
CLASS="footnote"
>[145]</SPAN
></A
>
All user-defined exit codes in the accompanying examples to
this document conform to this standard, except where overriding
circumstances exist, as in <A
HREF="#TMDIN"
>Example 9-2</A
>.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Issuing a <A
HREF="#XSTATVARREF"
>$?</A
> from
the command-line after a shell script exits gives
results consistent with the table above only from the
Bash or <I
CLASS="FIRSTTERM"
>sh</I
> prompt. Running the
<I
CLASS="FIRSTTERM"
>C-shell</I
> or <I
CLASS="FIRSTTERM"
>tcsh</I
>
may give different values in some cases.</P
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="IOREDIRINTRO"
></A
>Appendix F. A Detailed Introduction to I/O and I/O Redirection</H1
><P
><EM
>written by St<53>phane Chazelas, and revised
by the document author</EM
></P
><P
><A
NAME="STDINOUTDEF"
></A
></P
><P
>A command expects the first three <A
HREF="#FDREF"
>file
descriptors</A
> to be available. The first, <I
CLASS="FIRSTTERM"
>fd
0</I
> (standard input, <TT
CLASS="FILENAME"
>stdin</TT
>),
is for reading. The other two (<I
CLASS="FIRSTTERM"
>fd 1</I
>,
<TT
CLASS="FILENAME"
>stdout</TT
> and <I
CLASS="FIRSTTERM"
>fd 2</I
>,
<TT
CLASS="FILENAME"
>stderr</TT
>) are for writing.</P
><P
>There is a <TT
CLASS="FILENAME"
>stdin</TT
>, <TT
CLASS="FILENAME"
>stdout</TT
>,
and a <TT
CLASS="FILENAME"
>stderr</TT
> associated with each command.
<TT
CLASS="USERINPUT"
><B
>ls 2&#62;&#38;1</B
></TT
> means temporarily connecting the
<TT
CLASS="FILENAME"
>stderr</TT
> of the <B
CLASS="COMMAND"
>ls</B
> command to the
same <SPAN
CLASS="QUOTE"
>"resource"</SPAN
> as the shell's
<TT
CLASS="FILENAME"
>stdout</TT
>.</P
><P
>By convention, a command reads its input from fd 0
(<TT
CLASS="FILENAME"
>stdin</TT
>), prints normal output to fd
1 (<TT
CLASS="FILENAME"
>stdout</TT
>), and error ouput to fd 2
(<TT
CLASS="FILENAME"
>stderr</TT
>). If one of those three fd's is
not open, you may encounter problems:</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>cat /etc/passwd &#62;&#38;-</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>cat: standard output: Bad file descriptor</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
><P
>For example, when <B
CLASS="COMMAND"
>xterm</B
> runs, it first
initializes itself. Before running the user's shell,
<B
CLASS="COMMAND"
>xterm</B
> opens the terminal device
(/dev/pts/&#60;n&#62; or something similar) three times.</P
><P
>At this point, Bash inherits these three file descriptors,
and each command (child process) run by Bash inherits
them in turn, except when you redirect the command. <A
HREF="#IOREDIRREF"
>Redirection</A
> means reassigning
one of the file descriptors to another file (or a pipe, or
anything permissible). File descriptors may be reassigned
locally (for a command, a command group, a <A
HREF="#SUBSHELLSREF"
>subshell</A
>, a <A
HREF="#REDIRREF"
>while or if or case or for loop</A
>...),
or globally, for the remainder of the shell (using <A
HREF="#EXECREF"
>exec</A
>).</P
><P
><TT
CLASS="USERINPUT"
><B
>ls &#62; /dev/null</B
></TT
> means
running <B
CLASS="COMMAND"
>ls</B
> with its fd 1 connected to
<TT
CLASS="FILENAME"
>/dev/null</TT
>.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>lsof -a -p $$ -d0,1,2</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME
bash 363 bozo 0u CHR 136,1 3 /dev/pts/1
bash 363 bozo 1u CHR 136,1 3 /dev/pts/1
bash 363 bozo 2u CHR 136,1 3 /dev/pts/1</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>exec 2&#62; /dev/null</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>lsof -a -p $$ -d0,1,2</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME
bash 371 bozo 0u CHR 136,1 3 /dev/pts/1
bash 371 bozo 1u CHR 136,1 3 /dev/pts/1
bash 371 bozo 2w CHR 1,3 120 /dev/null</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>bash -c 'lsof -a -p $$ -d0,1,2' | cat</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME
lsof 379 root 0u CHR 136,1 3 /dev/pts/1
lsof 379 root 1w FIFO 0,0 7118 pipe
lsof 379 root 2u CHR 136,1 3 /dev/pts/1</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "$(bash -c 'lsof -a -p $$ -d0,1,2' 2&#62;&#38;1)"</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME
lsof 426 root 0u CHR 136,1 3 /dev/pts/1
lsof 426 root 1w FIFO 0,0 7520 pipe
lsof 426 root 2w FIFO 0,0 7520 pipe</TT
></PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>This works for different types of redirection.</P
><P
><TT
CLASS="USERINPUT"
><B
>Exercise:</B
></TT
> Analyze the following script.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#! /usr/bin/env bash
mkfifo /tmp/fifo1 /tmp/fifo2
while read a; do echo "FIFO1: $a"; done &#60; /tmp/fifo1 &#38; exec 7&#62; /tmp/fifo1
exec 8&#62; &#62;(while read a; do echo "FD8: $a, to fd7"; done &#62;&#38;7)
exec 3&#62;&#38;1
(
(
(
while read a; do echo "FIFO2: $a"; done &#60; /tmp/fifo2 | tee /dev/stderr \
| tee /dev/fd/4 | tee /dev/fd/5 | tee /dev/fd/6 &#62;&#38;7 &#38; exec 3&#62; /tmp/fifo2
echo 1st, to stdout
sleep 1
echo 2nd, to stderr &#62;&#38;2
sleep 1
echo 3rd, to fd 3 &#62;&#38;3
sleep 1
echo 4th, to fd 4 &#62;&#38;4
sleep 1
echo 5th, to fd 5 &#62;&#38;5
sleep 1
echo 6th, through a pipe | sed 's/.*/PIPE: &#38;, to fd 5/' &#62;&#38;5
sleep 1
echo 7th, to fd 6 &#62;&#38;6
sleep 1
echo 8th, to fd 7 &#62;&#38;7
sleep 1
echo 9th, to fd 8 &#62;&#38;8
) 4&#62;&#38;1 &#62;&#38;3 3&#62;&#38;- | while read a; do echo "FD4: $a"; done 1&#62;&#38;3 5&#62;&#38;- 6&#62;&#38;-
) 5&#62;&#38;1 &#62;&#38;3 | while read a; do echo "FD5: $a"; done 1&#62;&#38;3 6&#62;&#38;-
) 6&#62;&#38;1 &#62;&#38;3 | while read a; do echo "FD6: $a"; done 3&#62;&#38;-
rm -f /tmp/fifo1 /tmp/fifo2
# For each command and subshell, figure out which fd points to what.
# Good luck!
exit 0</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="COMMAND-LINE-OPTIONS"
></A
>Appendix G. Command-Line Options</H1
><P
>Many executables, whether binaries or script files, accept
options to modify their run-time behavior. For example: from
the command-line, typing <B
CLASS="COMMAND"
>command -o</B
>
would invoke <EM
>command</EM
>, with option
<TT
CLASS="OPTION"
>o</TT
>.</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="STANDARD-OPTIONS"
></A
>G.1. Standard Command-Line Options</H1
><P
>Over time, there has evolved a loose standard for the
meanings of command-line option flags. The GNU utilities conform
more closely to this <SPAN
CLASS="QUOTE"
>"standard"</SPAN
> than older UNIX
utilities.</P
><P
>Traditionally, UNIX command-line options consist of a dash,
followed by one or more lowercase letters. The GNU utilities
added a double-dash, followed by a complete word or compound
word.</P
><P
>The two most widely-accepted options are:</P
><P
></P
><UL
><LI
><P
><TT
CLASS="OPTION"
>-h</TT
></P
><P
><TT
CLASS="OPTION"
>--help</TT
></P
><P
><I
CLASS="FIRSTTERM"
>Help</I
>: Give usage message and exit.</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-v</TT
></P
><P
><TT
CLASS="OPTION"
>--version</TT
></P
><P
><I
CLASS="FIRSTTERM"
>Version</I
>: Show program version and exit.</P
></LI
></UL
><P
>Other common options are:</P
><P
></P
><UL
><LI
><P
><TT
CLASS="OPTION"
>-a</TT
></P
><P
><TT
CLASS="OPTION"
>--all</TT
></P
><P
><I
CLASS="FIRSTTERM"
>All</I
>: show <EM
>all</EM
>
information or operate on <EM
>all</EM
> arguments.</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-l</TT
></P
><P
><TT
CLASS="OPTION"
>--list</TT
></P
><P
><I
CLASS="FIRSTTERM"
>List</I
>: list files or arguments without
taking other action.</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-o</TT
></P
><P
><I
CLASS="FIRSTTERM"
>Output</I
> filename</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-q</TT
></P
><P
><TT
CLASS="OPTION"
>--quiet</TT
></P
><P
><I
CLASS="FIRSTTERM"
>Quiet</I
>: suppress
<TT
CLASS="FILENAME"
>stdout</TT
>.</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-r</TT
></P
><P
><TT
CLASS="OPTION"
>-R</TT
></P
><P
><TT
CLASS="OPTION"
>--recursive</TT
></P
><P
><I
CLASS="FIRSTTERM"
>Recursive</I
>: Operate recursively (down
directory tree).</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-v</TT
></P
><P
><TT
CLASS="OPTION"
>--verbose</TT
></P
><P
><I
CLASS="FIRSTTERM"
>Verbose</I
>: output additional information to
<TT
CLASS="FILENAME"
>stdout</TT
> or <TT
CLASS="FILENAME"
>stderr</TT
>.</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-z</TT
></P
><P
><TT
CLASS="OPTION"
>--compress</TT
></P
><P
><I
CLASS="FIRSTTERM"
>Compress</I
>: apply compression (usually
<A
HREF="#GZIPREF"
>gzip</A
>).</P
></LI
></UL
><P
>However:</P
><P
></P
><UL
><LI
><P
>In <B
CLASS="COMMAND"
>tar</B
> and <B
CLASS="COMMAND"
>gawk</B
>:</P
><P
><TT
CLASS="OPTION"
>-f</TT
></P
><P
><TT
CLASS="OPTION"
>--file</TT
></P
><P
><I
CLASS="FIRSTTERM"
>File</I
>: filename follows.</P
></LI
><LI
><P
>In <B
CLASS="COMMAND"
>cp</B
>, <B
CLASS="COMMAND"
>mv</B
>,
<B
CLASS="COMMAND"
>rm</B
>:</P
><P
><TT
CLASS="OPTION"
>-f</TT
></P
><P
><TT
CLASS="OPTION"
>--force</TT
></P
><P
><I
CLASS="FIRSTTERM"
>Force</I
>: force overwrite of target file(s).</P
></LI
></UL
><DIV
CLASS="CAUTION"
><P
></P
><TABLE
CLASS="CAUTION"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/caution.gif"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Many UNIX and Linux utilities deviate from this
<SPAN
CLASS="QUOTE"
>"standard,"</SPAN
> so it is dangerous to
<EM
>assume</EM
> that a given option will behave in a
standard way. Always check the man page for the command in question
when in doubt.</P
></TD
></TR
></TABLE
></DIV
><P
>A complete table of recommended options for the GNU utilities
is available at <A
HREF="http://www.gnu.org/prep/standards/"
TARGET="_top"
>the GNU standards page</A
>.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="BASH-OPTIONS"
></A
>G.2. Bash Command-Line Options</H1
><P
><A
NAME="CLOPTS"
></A
></P
><P
><I
CLASS="FIRSTTERM"
>Bash</I
> itself has a number of command-line
options. Here are some of the more useful ones.</P
><P
></P
><UL
><LI
><P
><TT
CLASS="OPTION"
>-c</TT
></P
><P
><EM
>Read commands from the following string and assign any
arguments to the <A
HREF="#POSPARAMREF"
>positional
parameters</A
>.</EM
></P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>bash -c 'set a b c d; IFS="+-;"; echo "$*"'</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>a+b+c+d</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></LI
><LI
><P
><TT
CLASS="OPTION"
>-r</TT
></P
><P
><TT
CLASS="OPTION"
>--restricted</TT
></P
><P
><EM
>Runs the shell, or a script, in <A
HREF="#RESTRICTEDSHREF"
>restricted mode</A
>.</EM
></P
></LI
><LI
><P
><TT
CLASS="OPTION"
>--posix</TT
></P
><P
><EM
>Forces Bash to conform to <A
HREF="#POSIX2REF"
>POSIX</A
> mode.</EM
></P
></LI
><LI
><P
><TT
CLASS="OPTION"
>--version</TT
></P
><P
><EM
>Display Bash version information and
exit.</EM
></P
></LI
><LI
><P
><TT
CLASS="OPTION"
>--</TT
></P
><P
><EM
>End of options. Anything further on the command
line is an argument, not an option.</EM
></P
></LI
></UL
></DIV
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="FILES"
></A
>Appendix H. Important Files</H1
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="FILESREF1"
></A
>startup files</B
></P
><DL
><DT
></DT
><DD
><P
>These files contain the aliases and <A
HREF="#ENVREF"
>environmental variables</A
>
made available to Bash running as a user shell and to all
Bash scripts invoked after system initialization.</P
></DD
><DT
><TT
CLASS="FILENAME"
>/etc/profile</TT
></DT
><DD
><P
>Systemwide defaults, mostly setting the environment
(all Bourne-type shells, not just Bash
<A
NAME="AEN23892"
HREF="#FTN.AEN23892"
><SPAN
CLASS="footnote"
>[146]</SPAN
></A
>)</P
></DD
><DT
><TT
CLASS="FILENAME"
>/etc/bashrc</TT
></DT
><DD
><P
>systemwide functions and <A
HREF="#ALIASREF"
>aliases</A
> for Bash</P
></DD
><DT
><TT
CLASS="FILENAME"
><TT
CLASS="VARNAME"
>$HOME</TT
>/.bash_profile</TT
></DT
><DD
><P
>user-specific Bash environmental default settings,
found in each user's home directory (the local counterpart
to <TT
CLASS="FILENAME"
>/etc/profile</TT
>)</P
></DD
><DT
><TT
CLASS="FILENAME"
><TT
CLASS="VARNAME"
>$HOME</TT
>/.bashrc</TT
></DT
><DD
><P
>user-specific Bash init file, found in each user's home
directory (the local counterpart to
<TT
CLASS="FILENAME"
>/etc/bashrc</TT
>). Only interactive
shells and user scripts read this file. See
<A
HREF="#SAMPLE-BASHRC"
>Appendix M</A
> for a sample
<TT
CLASS="FILENAME"
>.bashrc</TT
> file.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="LOGOUTFILEREF1"
></A
>logout file</B
></P
><DL
><DT
><TT
CLASS="FILENAME"
><TT
CLASS="VARNAME"
>$HOME</TT
>/.bash_logout</TT
></DT
><DD
><P
>user-specific instruction file, found in
each user's home directory. Upon exit from a login (Bash)
shell, the commands in this file execute.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="DATAFILESREF1"
></A
>data files</B
></P
><DL
><DT
><TT
CLASS="FILENAME"
>/etc/passwd</TT
></DT
><DD
><P
>A listing of all the user accounts on the system,
their identities, their home directories, the groups they
belong to, and their default shell. Note that the user
passwords are <EM
>not</EM
>
stored in this file,
<A
NAME="AEN23937"
HREF="#FTN.AEN23937"
><SPAN
CLASS="footnote"
>[147]</SPAN
></A
>
but in <TT
CLASS="FILENAME"
>/etc/shadow</TT
> in encrypted form.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="SYSCONFREF1"
></A
>system configuration files</B
></P
><DL
><DT
><TT
CLASS="FILENAME"
>/etc/sysconfig/hwconf</TT
></DT
><DD
><P
>Listing and description of attached hardware devices.
This information is in text form and can be extracted and
parsed.</P
><P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep -A 5 AUDIO /etc/sysconfig/hwconf</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>class: AUDIO
bus: PCI
detached: 0
driver: snd-intel8x0
desc: "Intel Corporation 82801CA/CAM AC'97 Audio Controller"
vendorId: 8086</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>This file is present on Red Hat and Fedora Core
installations, but may be missing from other
distros.</P
></TD
></TR
></TABLE
></DIV
></DD
></DL
></DIV
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="SYSTEMDIRS"
></A
>Appendix I. Important System Directories</H1
><P
>Sysadmins and anyone else writing administrative scripts
should be intimately familiar with the following system
directories.</P
><P
></P
><UL
><LI
><P
><TT
CLASS="FILENAME"
>/bin</TT
></P
><P
>Binaries (executables). Basic system programs
and utilities (such as <B
CLASS="COMMAND"
>bash</B
>).</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/usr/bin</TT
>
<A
NAME="AEN23969"
HREF="#FTN.AEN23969"
><SPAN
CLASS="footnote"
>[148]</SPAN
></A
>
</P
><P
>More system binaries.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/usr/local/bin</TT
></P
><P
>Miscellaneous binaries local to the particular machine.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/sbin</TT
></P
><P
>System binaries. Basic system administrative programs
and utilities (such as <B
CLASS="COMMAND"
>fsck</B
>).</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/usr/sbin</TT
></P
><P
>More system administrative programs and utilities.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/etc</TT
></P
><P
><I
CLASS="FIRSTTERM"
>Et cetera</I
>. Systemwide configuration
scripts.</P
><P
>Of particular interest are the
<A
HREF="#FSTABREF"
><TT
CLASS="FILENAME"
>/etc/fstab</TT
></A
>
(filesystem table),
<TT
CLASS="FILENAME"
>/etc/mtab</TT
>
(mounted filesystem table), and the <A
HREF="#INITTABREF"
><TT
CLASS="FILENAME"
>/etc/inittab</TT
></A
>
files.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/etc/rc.d</TT
></P
><P
>Boot scripts, on Red Hat and derivative distributions
of Linux.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/usr/share/doc</TT
></P
><P
>Documentation for installed packages.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/usr/man</TT
></P
><P
>The systemwide <A
HREF="#MANREF"
>manpages</A
>.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/dev</TT
></P
><P
>Device directory. Entries (but <EM
>not</EM
>
mount points) for physical and virtual devices.
See <A
HREF="#DEVPROC"
>Chapter 29</A
>.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/proc</TT
></P
><P
>Process directory. Contains information and statistics
about running processes and kernel parameters.
See <A
HREF="#DEVPROC"
>Chapter 29</A
>.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/sys</TT
></P
><P
>Systemwide device directory. Contains information and
statistics about device and device names. This is newly
added to Linux with the 2.6.X kernels.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/mnt</TT
></P
><P
><I
CLASS="FIRSTTERM"
>Mount</I
>. Directory for mounting
hard drive partitions, such as <TT
CLASS="FILENAME"
>/mnt/dos</TT
>, and physical
devices. In newer Linux distros, the <TT
CLASS="FILENAME"
>/media</TT
> directory has taken
over as the preferred mount point for I/O devices.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/media</TT
></P
><P
>In newer Linux distros, the preferred mount point for
I/O devices, such as CD/DVD drives or USB flash drives.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/var</TT
></P
><P
><I
CLASS="FIRSTTERM"
>Variable</I
> (changeable) system
files. This is a catchall <SPAN
CLASS="QUOTE"
>"scratchpad"</SPAN
>
directory for data generated while a Linux/UNIX machine
is running.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/var/log</TT
></P
><P
>Systemwide log files.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/var/spool/mail</TT
></P
><P
>User mail spool.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/lib</TT
></P
><P
>Systemwide library files.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/usr/lib</TT
></P
><P
>More systemwide library files.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/tmp</TT
></P
><P
>System temporary files.</P
></LI
><LI
><P
><TT
CLASS="FILENAME"
>/boot</TT
></P
><P
>System <I
CLASS="FIRSTTERM"
>boot</I
> directory. The kernel,
module links, system map, and boot manager reside here.</P
><DIV
CLASS="WARNING"
><P
></P
><TABLE
CLASS="WARNING"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/warning.gif"
HSPACE="5"
ALT="Warning"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Altering files in this directory may result in an
unbootable system.</P
></TD
></TR
></TABLE
></DIV
></LI
></UL
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="TABEXPANSION"
></A
>Appendix J. An Introduction to Programmable Completion</H1
><P
>The <I
CLASS="FIRSTTERM"
>programmable completion</I
> feature in
Bash permits typing a partial command, then pressing the
<B
CLASS="KEYCAP"
>[Tab]</B
> key to auto-complete the command sequence.
<A
NAME="AEN24082"
HREF="#FTN.AEN24082"
><SPAN
CLASS="footnote"
>[149]</SPAN
></A
>
If multiple completions are possible, then <B
CLASS="KEYCAP"
>[Tab]</B
>
lists them all. Let's see how it works.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>xtra[Tab]</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>xtraceroute xtrapin xtrapproto
xtraceroute.real xtrapinfo xtrapreset
xtrapchar xtrapout xtrapstats</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>xtrac[Tab]</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>xtraceroute xtraceroute.real</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>xtraceroute.r[Tab]</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>xtraceroute.real</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Tab completion also works for variables and path names.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $BASH[Tab]</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>$BASH $BASH_COMPLETION $BASH_SUBSHELL
$BASH_ARGC $BASH_COMPLETION_DIR $BASH_VERSINFO
$BASH_ARGV $BASH_LINENO $BASH_VERSION
$BASH_COMMAND $BASH_SOURCE</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo /usr/local/[Tab]</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bin/ etc/ include/ libexec/ sbin/ src/
doc/ games/ lib/ man/ share/</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
><A
NAME="COMPLETEREF"
></A
></P
><P
>The Bash <B
CLASS="COMMAND"
>complete</B
> and
<B
CLASS="COMMAND"
>compgen</B
> <A
HREF="#BUILTINREF"
>builtins</A
> make it
possible for <I
CLASS="FIRSTTERM"
>tab completion</I
> to
recognize partial <I
CLASS="FIRSTTERM"
>parameters</I
> and
<I
CLASS="FIRSTTERM"
>options</I
> to commands. In a very simple case,
we can use <B
CLASS="COMMAND"
>complete</B
> from the command-line to
specify a short list of acceptable parameters.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>touch sample_command</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>touch file1.txt file2.txt file2.doc file30.txt file4.zzz</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>chmod +x sample_command</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>complete -f -X '!*.txt' sample_command</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>./sample[Tab][Tab]</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>sample_command</TT
>
<TT
CLASS="COMPUTEROUTPUT"
>file1.txt file2.txt file30.txt</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
><P
>The <TT
CLASS="OPTION"
>-f</TT
> option to
<I
CLASS="FIRSTTERM"
>complete</I
> specifies filenames,
and <TT
CLASS="OPTION"
>-X</TT
> the filter pattern.</P
><P
><A
NAME="COMPGENREF"
></A
></P
><P
>For anything more complex, we could write a script that
specifies a list of acceptable command-line parameters.
The <B
CLASS="COMMAND"
>compgen</B
> builtin expands a list of
<I
CLASS="FIRSTTERM"
>arguments</I
> to <I
CLASS="FIRSTTERM"
>generate</I
>
completion matches. </P
><P
>Let us take a <A
HREF="#USEGETOPT2"
>modified version</A
>
of the <EM
>UseGetOpt.sh</EM
> script as an example
command. This script accepts a number of command-line parameters,
preceded by either a single or double dash. And here is the
corresponding <I
CLASS="FIRSTTERM"
>completion script</I
>, by
convention given a filename corresponding to its associated
command.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="USEGETOPTEX"
></A
><P
><B
>Example J-1. Completion script for
<I
CLASS="FIRSTTERM"
>UseGetOpt.sh</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># file: UseGetOpt-2
# UseGetOpt-2.sh parameter-completion
_UseGetOpt-2 () # By convention, the function name
{ #+ starts with an underscore.
local cur
# Pointer to current completion word.
# By convention, it's named "cur" but this isn't strictly necessary.
COMPREPLY=() # Array variable storing the possible completions.
cur=${COMP_WORDS[COMP_CWORD]}
case "$cur" in
-*)
COMPREPLY=( $( compgen -W '-a -d -f -l -t -h --aoption --debug \
--file --log --test --help --' -- $cur ) );;
# Generate the completion matches and load them into $COMPREPLY array.
# xx) May add more cases here.
# yy)
# zz)
esac
return 0
}
complete -F _UseGetOpt-2 -o filenames ./UseGetOpt-2.sh
# ^^ ^^^^^^^^^^^^ Invokes the function _UseGetOpt-2.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Now, let's try it.</P
><TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>source UseGetOpt-2</B
></TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>./UseGetOpt-2.sh -[Tab]</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-- --aoption --debug --file --help --log --test
-a -d -f -h -l -t</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>./UseGetOpt-2.sh --[Tab]</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>-- --aoption --debug --file --help --log --test</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
><P
> We begin by <A
HREF="#SOURCEREF"
>sourcing</A
> the <SPAN
CLASS="QUOTE"
>"completion
script."</SPAN
> This sets the command-line parameters.
<A
NAME="AEN24160"
HREF="#FTN.AEN24160"
><SPAN
CLASS="footnote"
>[150]</SPAN
></A
>
</P
><P
>In the first instance, hitting <B
CLASS="KEYCAP"
>[Tab]</B
> after
a single dash, the output is all the possible parameters preceded by
<EM
>one or more</EM
> dashes. Hitting <B
CLASS="KEYCAP"
>[Tab]</B
>
after <EM
>two</EM
> dashes gives the possible parameters
preceded by <EM
>two or more</EM
> dashes.</P
><P
>Now, just what is the point of having to jump through flaming
hoops to enable command-line tab completion? <EM
>It saves
keystrokes.</EM
>
<A
NAME="AEN24173"
HREF="#FTN.AEN24173"
><SPAN
CLASS="footnote"
>[151]</SPAN
></A
>
</P
><P
>--</P
><P
><EM
>Resources:</EM
></P
><P
>Bash <A
HREF="http://freshmeat.net/projects/bashcompletion"
TARGET="_top"
> programmable completion</A
> project</P
><P
>Mitch Frazier's <A
HREF="http://www.linuxjournal.com"
TARGET="_top"
><I
CLASS="CITETITLE"
>Linux Journal</I
></A
> article, <A
HREF="http://www.linuxjournal.com/content/more-using-bash-complete-command"
TARGET="_top"
><EM
>More
on Using the Bash Complete Command</EM
></A
></P
><P
>Steve's excellent two-part article, <SPAN
CLASS="QUOTE"
>"An Introduction to Bash
Completion"</SPAN
>:
<A
HREF="http://www.debian-administration.org/article/An_introduction_to_bash_completion_part_1"
TARGET="_top"
>Part
1</A
> and
<A
HREF="http://www.debian-administration.org/article/An_introduction_to_bash_completion_part_2"
TARGET="_top"
>Part 2</A
></P
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="LOCALIZATION"
></A
>Appendix K. Localization</H1
><P
>Localization is an undocumented Bash feature.</P
><P
><A
NAME="LOCALEREF"
></A
>A localized shell script echoes
its text output in the language defined as the system's locale.
A Linux user in Berlin, Germany, would get script output in German,
whereas his cousin in Berlin, Maryland, would get output from
the same script in English.</P
><P
>To create a localized script, use the following template to
write all messages to the user (error messages, prompts,
etc.).</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# localized.sh
# Script by St<53>phane Chazelas,
#+ modified by Bruno Haible, bugfixed by Alfredo Pironti.
. gettext.sh
E_CDERROR=65
error()
{
printf "$@" &#62;&#38;2
exit $E_CDERROR
}
cd $var || error "`eval_gettext \"Can\'t cd to \\\$var.\"`"
# The triple backslashes (escapes) in front of $var needed
#+ "because eval_gettext expects a string
#+ where the variable values have not yet been substituted."
# -- per Bruno Haible
read -p "`gettext \"Enter the value: \"`" var
# ...
# ------------------------------------------------------------------
# Alfredo Pironti comments:
# This script has been modified to not use the $"..." syntax in
#+ favor of the "`gettext \"...\"`" syntax.
# This is ok, but with the new localized.sh program, the commands
#+ "bash -D filename" and "bash --dump-po-string filename"
#+ will produce no output
#+ (because those command are only searching for the $"..." strings)!
# The ONLY way to extract strings from the new file is to use the
# 'xgettext' program. However, the xgettext program is buggy.
# Note that 'xgettext' has another bug.
#
# The shell fragment:
# gettext -s "I like Bash"
# will be correctly extracted, but . . .
# xgettext -s "I like Bash"
# . . . fails!
# 'xgettext' will extract "-s" because
#+ the command only extracts the
#+ very first argument after the 'gettext' word.
# Escape characters:
#
# To localize a sentence like
# echo -e "Hello\tworld!"
#+ you must use
# echo -e "`gettext \"Hello\\tworld\"`"
# The "double escape character" before the `t' is needed because
#+ 'gettext' will search for a string like: 'Hello\tworld'
# This is because gettext will read one literal `\')
#+ and will output a string like "Bonjour\tmonde",
#+ so the 'echo' command will display the message correctly.
#
# You may not use
# echo "`gettext -e \"Hello\tworld\"`"
#+ due to the xgettext bug explained above.
# Let's localize the following shell fragment:
# echo "-h display help and exit"
#
# First, one could do this:
# echo "`gettext \"-h display help and exit\"`"
# This way 'xgettext' will work ok,
#+ but the 'gettext' program will read "-h" as an option!
#
# One solution could be
# echo "`gettext -- \"-h display help and exit\"`"
# This way 'gettext' will work,
#+ but 'xgettext' will extract "--", as referred to above.
#
# The workaround you may use to get this string localized is
# echo -e "`gettext \"\\0-h display help and exit\"`"
# We have added a \0 (NULL) at the beginning of the sentence.
# This way 'gettext' works correctly, as does 'xgettext.'
# Moreover, the NULL character won't change the behavior
#+ of the 'echo' command.
# ------------------------------------------------------------------</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>bash -D localized.sh</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>"Can't cd to %s."
"Enter the value: "</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
This lists all the localized text. (The <TT
CLASS="OPTION"
>-D</TT
>
option lists double-quoted strings prefixed by a <SPAN
CLASS="TOKEN"
>$</SPAN
>,
without executing the script.)</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>bash --dump-po-strings localized.sh</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>#: a:6
msgid "Can't cd to %s."
msgstr ""
#: a:7
msgid "Enter the value: "
msgstr ""</TT
></PRE
></FONT
></TD
></TR
></TABLE
>
The <TT
CLASS="OPTION"
>--dump-po-strings</TT
> option to Bash
resembles the <TT
CLASS="OPTION"
>-D</TT
> option, but uses <A
HREF="#GETTEXTREF"
>gettext</A
> <SPAN
CLASS="QUOTE"
>"po"</SPAN
> format.
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Bruno Haible points out:</P
><P
>Starting with gettext-0.12.2, <B
CLASS="COMMAND"
>xgettext -o - localized.sh</B
>
is recommended instead of <B
CLASS="COMMAND"
>bash --dump-po-strings
localized.sh</B
>, because <B
CLASS="COMMAND"
>xgettext</B
> . . .</P
><P
>1. understands the gettext and eval_gettext commands
(whereas bash --dump-po-strings understands only its deprecated
$"..." syntax)</P
><P
>2. can extract comments placed by the programmer, intended
to be read by the translator.</P
><P
>This shell code is then not specific to Bash any
more; it works the same way with Bash 1.x and other /bin/sh
implementations.</P
></TD
></TR
></TABLE
></DIV
><P
>Now, build a <TT
CLASS="FILENAME"
>language.po</TT
>
file for each language that the script will be translated
into, specifying the <TT
CLASS="REPLACEABLE"
><I
>msgstr</I
></TT
>. Alfredo
Pironti gives the following example:</P
><P
>fr.po:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#: a:6
msgid "Can't cd to $var."
msgstr "Impossible de se positionner dans le repertoire $var."
#: a:7
msgid "Enter the value: "
msgstr "Entrez la valeur : "
# The string are dumped with the variable names, not with the %s syntax,
#+ similar to C programs.
#+ This is a very cool feature if the programmer uses
#+ variable names that make sense!</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Then, run <A
HREF="#MSGFMTREF"
>msgfmt</A
>.</P
><P
><TT
CLASS="USERINPUT"
><B
>msgfmt -o localized.sh.mo fr.po</B
></TT
></P
><P
>Place the resulting <TT
CLASS="FILENAME"
>localized.sh.mo</TT
> file in the
<TT
CLASS="FILENAME"
>/usr/local/share/locale/fr/LC_MESSAGES</TT
>
directory, and at the beginning of the script, insert the lines:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>TEXTDOMAINDIR=/usr/local/share/locale
TEXTDOMAIN=localized.sh</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>If a user on a French system runs the script, she will get
French messages.</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>With older versions of Bash or other shells, localization requires
<A
HREF="#GETTEXTREF"
>gettext</A
>, using the
<TT
CLASS="OPTION"
>-s</TT
> option. In this case, the script becomes:</P
><P
><A
NAME="GETTEXTEXAMPLE"
></A
>
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# localized.sh
E_CDERROR=65
error() {
local format=$1
shift
printf "$(gettext -s "$format")" "$@" &#62;&#38;2
exit $E_CDERROR
}
cd $var || error "Can't cd to %s." "$var"
read -p "$(gettext -s "Enter the value: ")" var
# ...</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></DIV
><P
>The <TT
CLASS="VARNAME"
>TEXTDOMAIN</TT
> and
<TT
CLASS="VARNAME"
>TEXTDOMAINDIR</TT
> variables need to be set and
exported to the environment. This should be done within the
script itself.</P
><P
>---</P
><P
>This appendix written by St<53>phane Chazelas,
with modifications suggested by Alfredo Pironti,
and by Bruno Haible, maintainer of GNU <A
HREF="#GETTEXTREF"
>gettext</A
>.</P
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="HISTCOMMANDS"
></A
>Appendix L. History Commands</H1
><P
>The Bash shell provides command-line tools for editing and
manipulating a user's <I
CLASS="FIRSTTERM"
>command history</I
>. This
is primarily a convenience, a means of saving keystrokes.</P
><P
>Bash history commands:
<P
></P
><OL
TYPE="1"
><LI
><P
><B
CLASS="COMMAND"
>history</B
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>fc</B
></P
></LI
></OL
>
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>history</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
> 1 mount /mnt/cdrom
2 cd /mnt/cdrom
3 ls
...</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Internal variables associated with Bash history commands:
<P
></P
><OL
TYPE="1"
><LI
><P
>$HISTCMD</P
></LI
><LI
><P
>$HISTCONTROL</P
></LI
><LI
><P
>$HISTIGNORE</P
></LI
><LI
><P
>$HISTFILE</P
></LI
><LI
><P
>$HISTFILESIZE</P
></LI
><LI
><P
>$HISTSIZE</P
></LI
><LI
><P
>$HISTTIMEFORMAT (Bash, ver. 3.0 or later)</P
></LI
><LI
><P
>!!</P
></LI
><LI
><P
>!$</P
></LI
><LI
><P
>!#</P
></LI
><LI
><P
>!N</P
></LI
><LI
><P
>!-N</P
></LI
><LI
><P
>!STRING</P
></LI
><LI
><P
>!?STRING?</P
></LI
><LI
><P
>^STRING^string^</P
></LI
></OL
>
</P
><P
>Unfortunately, the Bash history tools find no use in
scripting.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# history.sh
# A (vain) attempt to use the 'history' command in a script.
history # No output.
var=$(history); echo "$var" # $var is empty.
# History commands are, by default, disabled within a script.
# However, as dhw points out,
#+ set -o history
#+ enables the history mechanism.
set -o history
var=$(history); echo "$var" # 1 var=$(history)</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>./history.sh</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>(no output)</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>The <A
HREF="http://samrowe.com/wordpress/advancing-in-the-bash-shell/"
TARGET="_top"
>Advancing
in the Bash Shell</A
> site gives a good introduction to
the use of history commands in Bash.</P
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="SAMPLE-BASHRC"
></A
>Appendix M. Sample <TT
CLASS="FILENAME"
>.bashrc</TT
> and
<TT
CLASS="FILENAME"
>.bash_profile</TT
> Files</H1
><P
>The <TT
CLASS="FILENAME"
>~/.bashrc</TT
> file determines the
behavior of interactive shells. A good look at this file can
lead to a better understanding of Bash.</P
><P
><A
HREF="mailto:emmanuel.rouat@wanadoo.fr"
TARGET="_top"
>Emmanuel
Rouat</A
> contributed the following very elaborate
<TT
CLASS="FILENAME"
>.bashrc</TT
> file, written for a Linux system.
He welcomes reader feedback on it.</P
><P
>Study the file carefully, and feel free to reuse code
snippets and functions from it in your own
<TT
CLASS="FILENAME"
>.bashrc</TT
> file or even in your scripts.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="BASHRC"
></A
><P
><B
>Example M-1. Sample <TT
CLASS="FILENAME"
>.bashrc</TT
> file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># =============================================================== #
#
# PERSONAL $HOME/.bashrc FILE for bash-3.0 (or later)
# By Emmanuel Rouat [no-email]
#
# Last modified: Tue Nov 20 22:04:47 CET 2012
# This file is normally read by interactive shells only.
#+ Here is the place to define your aliases, functions and
#+ other interactive features like your prompt.
#
# The majority of the code here assumes you are on a GNU
#+ system (most likely a Linux box) and is often based on code
#+ found on Usenet or Internet.
#
# See for instance:
# http://tldp.org/LDP/abs/html/index.html
# http://www.caliban.org/bash
# http://www.shelldorado.com/scripts/categories.html
# http://www.dotfiles.org
#
# The choice of colors was done for a shell with a dark background
#+ (white on black), and this is usually also suited for pure text-mode
#+ consoles (no X server available). If you use a white background,
#+ you'll have to do some other choices for readability.
#
# This bashrc file is a bit overcrowded.
# Remember, it is just just an example.
# Tailor it to your needs.
#
# =============================================================== #
# --&#62; Comments added by HOWTO author.
# If not running interactively, don't do anything
[ -z "$PS1" ] &#38;&#38; return
#-------------------------------------------------------------
# Source global definitions (if any)
#-------------------------------------------------------------
if [ -f /etc/bashrc ]; then
. /etc/bashrc # --&#62; Read /etc/bashrc, if present.
fi
#--------------------------------------------------------------
# Automatic setting of $DISPLAY (if not set already).
# This works for me - your mileage may vary. . . .
# The problem is that different types of terminals give
#+ different answers to 'who am i' (rxvt in particular can be
#+ troublesome) - however this code seems to work in a majority
#+ of cases.
#--------------------------------------------------------------
function get_xserver ()
{
case $TERM in
xterm )
XSERVER=$(who am i | awk '{print $NF}' | tr -d ')''(' )
# Ane-Pieter Wieringa suggests the following alternative:
# I_AM=$(who am i)
# SERVER=${I_AM#*(}
# SERVER=${SERVER%*)}
XSERVER=${XSERVER%%:*}
;;
aterm | rxvt)
# Find some code that works here. ...
;;
esac
}
if [ -z ${DISPLAY:=""} ]; then
get_xserver
if [[ -z ${XSERVER} || ${XSERVER} == $(hostname) ||
${XSERVER} == "unix" ]]; then
DISPLAY=":0.0" # Display on local host.
else
DISPLAY=${XSERVER}:0.0 # Display on remote host.
fi
fi
export DISPLAY
#-------------------------------------------------------------
# Some settings
#-------------------------------------------------------------
#set -o nounset # These two options are useful for debugging.
#set -o xtrace
alias debug="set -o nounset; set -o xtrace"
ulimit -S -c 0 # Don't want coredumps.
set -o notify
set -o noclobber
set -o ignoreeof
# Enable options:
shopt -s cdspell
shopt -s cdable_vars
shopt -s checkhash
shopt -s checkwinsize
shopt -s sourcepath
shopt -s no_empty_cmd_completion
shopt -s cmdhist
shopt -s histappend histreedit histverify
shopt -s extglob # Necessary for programmable completion.
# Disable options:
shopt -u mailwarn
unset MAILCHECK # Don't want my shell to warn me of incoming mail.
#-------------------------------------------------------------
# Greeting, motd etc. ...
#-------------------------------------------------------------
# Color definitions (taken from Color Bash Prompt HowTo).
# Some colors might look different of some terminals.
# For example, I see 'Bold Red' as 'orange' on my screen,
# hence the 'Green' 'BRed' 'Red' sequence I often use in my prompt.
# Normal Colors
Black='\e[0;30m' # Black
Red='\e[0;31m' # Red
Green='\e[0;32m' # Green
Yellow='\e[0;33m' # Yellow
Blue='\e[0;34m' # Blue
Purple='\e[0;35m' # Purple
Cyan='\e[0;36m' # Cyan
White='\e[0;37m' # White
# Bold
BBlack='\e[1;30m' # Black
BRed='\e[1;31m' # Red
BGreen='\e[1;32m' # Green
BYellow='\e[1;33m' # Yellow
BBlue='\e[1;34m' # Blue
BPurple='\e[1;35m' # Purple
BCyan='\e[1;36m' # Cyan
BWhite='\e[1;37m' # White
# Background
On_Black='\e[40m' # Black
On_Red='\e[41m' # Red
On_Green='\e[42m' # Green
On_Yellow='\e[43m' # Yellow
On_Blue='\e[44m' # Blue
On_Purple='\e[45m' # Purple
On_Cyan='\e[46m' # Cyan
On_White='\e[47m' # White
NC="\e[m" # Color Reset
ALERT=${BWhite}${On_Red} # Bold White on red background
echo -e "${BCyan}This is BASH ${BRed}${BASH_VERSION%.*}${BCyan}\
- DISPLAY on ${BRed}$DISPLAY${NC}\n"
date
if [ -x /usr/games/fortune ]; then
/usr/games/fortune -s # Makes our day a bit more fun.... :-)
fi
function _exit() # Function to run upon exit of shell.
{
echo -e "${BRed}Hasta la vista, baby${NC}"
}
trap _exit EXIT
#-------------------------------------------------------------
# Shell Prompt - for many examples, see:
# http://www.debian-administration.org/articles/205
# http://www.askapache.com/linux/bash-power-prompt.html
# http://tldp.org/HOWTO/Bash-Prompt-HOWTO
# https://github.com/nojhan/liquidprompt
#-------------------------------------------------------------
# Current Format: [TIME USER@HOST PWD] &#62;
# TIME:
# Green == machine load is low
# Orange == machine load is medium
# Red == machine load is high
# ALERT == machine load is very high
# USER:
# Cyan == normal user
# Orange == SU to user
# Red == root
# HOST:
# Cyan == local session
# Green == secured remote connection (via ssh)
# Red == unsecured remote connection
# PWD:
# Green == more than 10% free disk space
# Orange == less than 10% free disk space
# ALERT == less than 5% free disk space
# Red == current user does not have write privileges
# Cyan == current filesystem is size zero (like /proc)
# &#62;:
# White == no background or suspended jobs in this shell
# Cyan == at least one background job in this shell
# Orange == at least one suspended job in this shell
#
# Command is added to the history file each time you hit enter,
# so it's available to all shells (using 'history -a').
# Test connection type:
if [ -n "${SSH_CONNECTION}" ]; then
CNX=${Green} # Connected on remote machine, via ssh (good).
elif [[ "${DISPLAY%%:0*}" != "" ]]; then
CNX=${ALERT} # Connected on remote machine, not via ssh (bad).
else
CNX=${BCyan} # Connected on local machine.
fi
# Test user type:
if [[ ${USER} == "root" ]]; then
SU=${Red} # User is root.
elif [[ ${USER} != $(logname) ]]; then
SU=${BRed} # User is not login user.
else
SU=${BCyan} # User is normal (well ... most of us are).
fi
NCPU=$(grep -c 'processor' /proc/cpuinfo) # Number of CPUs
SLOAD=$(( 100*${NCPU} )) # Small load
MLOAD=$(( 200*${NCPU} )) # Medium load
XLOAD=$(( 400*${NCPU} )) # Xlarge load
# Returns system load as percentage, i.e., '40' rather than '0.40)'.
function load()
{
local SYSLOAD=$(cut -d " " -f1 /proc/loadavg | tr -d '.')
# System load of the current host.
echo $((10#$SYSLOAD)) # Convert to decimal.
}
# Returns a color indicating system load.
function load_color()
{
local SYSLOAD=$(load)
if [ ${SYSLOAD} -gt ${XLOAD} ]; then
echo -en ${ALERT}
elif [ ${SYSLOAD} -gt ${MLOAD} ]; then
echo -en ${Red}
elif [ ${SYSLOAD} -gt ${SLOAD} ]; then
echo -en ${BRed}
else
echo -en ${Green}
fi
}
# Returns a color according to free disk space in $PWD.
function disk_color()
{
if [ ! -w "${PWD}" ] ; then
echo -en ${Red}
# No 'write' privilege in the current directory.
elif [ -s "${PWD}" ] ; then
local used=$(command df -P "$PWD" |
awk 'END {print $5} {sub(/%/,"")}')
if [ ${used} -gt 95 ]; then
echo -en ${ALERT} # Disk almost full (&#62;95%).
elif [ ${used} -gt 90 ]; then
echo -en ${BRed} # Free disk space almost gone.
else
echo -en ${Green} # Free disk space is ok.
fi
else
echo -en ${Cyan}
# Current directory is size '0' (like /proc, /sys etc).
fi
}
# Returns a color according to running/suspended jobs.
function job_color()
{
if [ $(jobs -s | wc -l) -gt "0" ]; then
echo -en ${BRed}
elif [ $(jobs -r | wc -l) -gt "0" ] ; then
echo -en ${BCyan}
fi
}
# Adds some text in the terminal frame (if applicable).
# Now we construct the prompt.
PROMPT_COMMAND="history -a"
case ${TERM} in
*term | rxvt | linux)
PS1="\[\$(load_color)\][\A\[${NC}\] "
# Time of day (with load info):
PS1="\[\$(load_color)\][\A\[${NC}\] "
# User@Host (with connection type info):
PS1=${PS1}"\[${SU}\]\u\[${NC}\]@\[${CNX}\]\h\[${NC}\] "
# PWD (with 'disk space' info):
PS1=${PS1}"\[\$(disk_color)\]\W]\[${NC}\] "
# Prompt (with 'job' info):
PS1=${PS1}"\[\$(job_color)\]&#62;\[${NC}\] "
# Set title of current xterm:
PS1=${PS1}"\[\e]0;[\u@\h] \w\a\]"
;;
*)
PS1="(\A \u@\h \W) &#62; " # --&#62; PS1="(\A \u@\h \w) &#62; "
# --&#62; Shows full pathname of current dir.
;;
esac
export TIMEFORMAT=$'\nreal %3R\tuser %3U\tsys %3S\tpcpu %P\n'
export HISTIGNORE="&#38;:bg:fg:ll:h"
export HISTTIMEFORMAT="$(echo -e ${BCyan})[%d/%m %H:%M:%S]$(echo -e ${NC}) "
export HISTCONTROL=ignoredups
export HOSTFILE=$HOME/.hosts # Put a list of remote hosts in ~/.hosts
#============================================================
#
# ALIASES AND FUNCTIONS
#
# Arguably, some functions defined here are quite big.
# If you want to make this file smaller, these functions can
#+ be converted into scripts and removed from here.
#
#============================================================
#-------------------
# Personnal Aliases
#-------------------
alias rm='rm -i'
alias cp='cp -i'
alias mv='mv -i'
# -&#62; Prevents accidentally clobbering files.
alias mkdir='mkdir -p'
alias h='history'
alias j='jobs -l'
alias which='type -a'
alias ..='cd ..'
# Pretty-print of some PATH variables:
alias path='echo -e ${PATH//:/\\n}'
alias libpath='echo -e ${LD_LIBRARY_PATH//:/\\n}'
alias du='du -kh' # Makes a more readable output.
alias df='df -kTh'
#-------------------------------------------------------------
# The 'ls' family (this assumes you use a recent GNU ls).
#-------------------------------------------------------------
# Add colors for filetype and human-readable sizes by default on 'ls':
alias ls='ls -h --color'
alias lx='ls -lXB' # Sort by extension.
alias lk='ls -lSr' # Sort by size, biggest last.
alias lt='ls -ltr' # Sort by date, most recent last.
alias lc='ls -ltcr' # Sort by/show change time,most recent last.
alias lu='ls -ltur' # Sort by/show access time,most recent last.
# The ubiquitous 'll': directories first, with alphanumeric sorting:
alias ll="ls -lv --group-directories-first"
alias lm='ll |more' # Pipe through 'more'
alias lr='ll -R' # Recursive ls.
alias la='ll -A' # Show hidden files.
alias tree='tree -Csuh' # Nice alternative to 'recursive ls' ...
#-------------------------------------------------------------
# Tailoring 'less'
#-------------------------------------------------------------
alias more='less'
export PAGER=less
export LESSCHARSET='latin1'
export LESSOPEN='|/usr/bin/lesspipe.sh %s 2&#62;&#38;-'
# Use this if lesspipe.sh exists.
export LESS='-i -N -w -z-4 -g -e -M -X -F -R -P%t?f%f \
:stdin .?pb%pb\%:?lbLine %lb:?bbByte %bb:-...'
# LESS man page colors (makes Man pages more readable).
export LESS_TERMCAP_mb=$'\E[01;31m'
export LESS_TERMCAP_md=$'\E[01;31m'
export LESS_TERMCAP_me=$'\E[0m'
export LESS_TERMCAP_se=$'\E[0m'
export LESS_TERMCAP_so=$'\E[01;44;33m'
export LESS_TERMCAP_ue=$'\E[0m'
export LESS_TERMCAP_us=$'\E[01;32m'
#-------------------------------------------------------------
# Spelling typos - highly personnal and keyboard-dependent :-)
#-------------------------------------------------------------
alias xs='cd'
alias vf='cd'
alias moer='more'
alias moew='more'
alias kk='ll'
#-------------------------------------------------------------
# A few fun ones
#-------------------------------------------------------------
# Adds some text in the terminal frame (if applicable).
function xtitle()
{
case "$TERM" in
*term* | rxvt)
echo -en "\e]0;$*\a" ;;
*) ;;
esac
}
# Aliases that use xtitle
alias top='xtitle Processes on $HOST &#38;&#38; top'
alias make='xtitle Making $(basename $PWD) ; make'
# .. and functions
function man()
{
for i ; do
xtitle The $(basename $1|tr -d .[:digit:]) manual
command man -a "$i"
done
}
#-------------------------------------------------------------
# Make the following commands run in background automatically:
#-------------------------------------------------------------
function te() # wrapper around xemacs/gnuserv
{
if [ "$(gnuclient -batch -eval t 2&#62;&#38;-)" == "t" ]; then
gnuclient -q "$@";
else
( xemacs "$@" &#38;);
fi
}
function soffice() { command soffice "$@" &#38; }
function firefox() { command firefox "$@" &#38; }
function xpdf() { command xpdf "$@" &#38; }
#-------------------------------------------------------------
# File &#38; strings related functions:
#-------------------------------------------------------------
# Find a file with a pattern in name:
function ff() { find . -type f -iname '*'"$*"'*' -ls ; }
# Find a file with pattern $1 in name and Execute $2 on it:
function fe() { find . -type f -iname '*'"${1:-}"'*' \
-exec ${2:-file} {} \; ; }
# Find a pattern in a set of files and highlight them:
#+ (needs a recent version of egrep).
function fstr()
{
OPTIND=1
local mycase=""
local usage="fstr: find string in files.
Usage: fstr [-i] \"pattern\" [\"filename pattern\"] "
while getopts :it opt
do
case "$opt" in
i) mycase="-i " ;;
*) echo "$usage"; return ;;
esac
done
shift $(( $OPTIND - 1 ))
if [ "$#" -lt 1 ]; then
echo "$usage"
return;
fi
find . -type f -name "${2:-*}" -print0 | \
xargs -0 egrep --color=always -sn ${case} "$1" 2&#62;&#38;- | more
}
function swap()
{ # Swap 2 filenames around, if they exist (from Uzi's bashrc).
local TMPFILE=tmp.$$
[ $# -ne 2 ] &#38;&#38; echo "swap: 2 arguments needed" &#38;&#38; return 1
[ ! -e $1 ] &#38;&#38; echo "swap: $1 does not exist" &#38;&#38; return 1
[ ! -e $2 ] &#38;&#38; echo "swap: $2 does not exist" &#38;&#38; return 1
mv "$1" $TMPFILE
mv "$2" "$1"
mv $TMPFILE "$2"
}
function extract() # Handy Extract Program
{
if [ -f $1 ] ; then
case $1 in
*.tar.bz2) tar xvjf $1 ;;
*.tar.gz) tar xvzf $1 ;;
*.bz2) bunzip2 $1 ;;
*.rar) unrar x $1 ;;
*.gz) gunzip $1 ;;
*.tar) tar xvf $1 ;;
*.tbz2) tar xvjf $1 ;;
*.tgz) tar xvzf $1 ;;
*.zip) unzip $1 ;;
*.Z) uncompress $1 ;;
*.7z) 7z x $1 ;;
*) echo "'$1' cannot be extracted via &#62;extract&#60;" ;;
esac
else
echo "'$1' is not a valid file!"
fi
}
# Creates an archive (*.tar.gz) from given directory.
function maketar() { tar cvzf "${1%%/}.tar.gz" "${1%%/}/"; }
# Create a ZIP archive of a file or folder.
function makezip() { zip -r "${1%%/}.zip" "$1" ; }
# Make your directories and files access rights sane.
function sanitize() { chmod -R u=rwX,g=rX,o= "$@" ;}
#-------------------------------------------------------------
# Process/system related functions:
#-------------------------------------------------------------
function my_ps() { ps $@ -u $USER -o pid,%cpu,%mem,bsdtime,command ; }
function pp() { my_ps f | awk '!/awk/ &#38;&#38; $0~var' var=${1:-".*"} ; }
function killps() # kill by process name
{
local pid pname sig="-TERM" # default signal
if [ "$#" -lt 1 ] || [ "$#" -gt 2 ]; then
echo "Usage: killps [-SIGNAL] pattern"
return;
fi
if [ $# = 2 ]; then sig=$1 ; fi
for pid in $(my_ps| awk '!/awk/ &#38;&#38; $0~pat { print $1 }' pat=${!#} )
do
pname=$(my_ps | awk '$1~var { print $5 }' var=$pid )
if ask "Kill process $pid &#60;$pname&#62; with signal $sig?"
then kill $sig $pid
fi
done
}
function mydf() # Pretty-print of 'df' output.
{ # Inspired by 'dfc' utility.
for fs ; do
if [ ! -d $fs ]
then
echo -e $fs" :No such file or directory" ; continue
fi
local info=( $(command df -P $fs | awk 'END{ print $2,$3,$5 }') )
local free=( $(command df -Pkh $fs | awk 'END{ print $4 }') )
local nbstars=$(( 20 * ${info[1]} / ${info[0]} ))
local out="["
for ((j=0;j&#60;20;j++)); do
if [ ${j} -lt ${nbstars} ]; then
out=$out"*"
else
out=$out"-"
fi
done
out=${info[2]}" "$out"] ("$free" free on "$fs")"
echo -e $out
done
}
function my_ip() # Get IP adress on ethernet.
{
MY_IP=$(/sbin/ifconfig eth0 | awk '/inet/ { print $2 } ' |
sed -e s/addr://)
echo ${MY_IP:-"Not connected"}
}
function ii() # Get current host related info.
{
echo -e "\nYou are logged on ${BRed}$HOST"
echo -e "\n${BRed}Additionnal information:$NC " ; uname -a
echo -e "\n${BRed}Users logged on:$NC " ; w -hs |
cut -d " " -f1 | sort | uniq
echo -e "\n${BRed}Current date :$NC " ; date
echo -e "\n${BRed}Machine stats :$NC " ; uptime
echo -e "\n${BRed}Memory stats :$NC " ; free
echo -e "\n${BRed}Diskspace :$NC " ; mydf / $HOME
echo -e "\n${BRed}Local IP Address :$NC" ; my_ip
echo -e "\n${BRed}Open connections :$NC "; netstat -pan --inet;
echo
}
#-------------------------------------------------------------
# Misc utilities:
#-------------------------------------------------------------
function repeat() # Repeat n times command.
{
local i max
max=$1; shift;
for ((i=1; i &#60;= max ; i++)); do # --&#62; C-like syntax
eval "$@";
done
}
function ask() # See 'killps' for example of use.
{
echo -n "$@" '[y/n] ' ; read ans
case "$ans" in
y*|Y*) return 0 ;;
*) return 1 ;;
esac
}
function corename() # Get name of app that created a corefile.
{
for file ; do
echo -n $file : ; gdb --core=$file --batch | head -1
done
}
#=========================================================================
#
# PROGRAMMABLE COMPLETION SECTION
# Most are taken from the bash 2.05 documentation and from Ian McDonald's
# 'Bash completion' package (http://www.caliban.org/bash/#completion)
# You will in fact need bash more recent then 3.0 for some features.
#
# Note that most linux distributions now provide many completions
# 'out of the box' - however, you might need to make your own one day,
# so I kept those here as examples.
#=========================================================================
if [ "${BASH_VERSION%.*}" \&#60; "3.0" ]; then
echo "You will need to upgrade to version 3.0 for full \
programmable completion features"
return
fi
shopt -s extglob # Necessary.
complete -A hostname rsh rcp telnet rlogin ftp ping disk
complete -A export printenv
complete -A variable export local readonly unset
complete -A enabled builtin
complete -A alias alias unalias
complete -A function function
complete -A user su mail finger
complete -A helptopic help # Currently same as builtins.
complete -A shopt shopt
complete -A stopped -P '%' bg
complete -A job -P '%' fg jobs disown
complete -A directory mkdir rmdir
complete -A directory -o default cd
# Compression
complete -f -o default -X '*.+(zip|ZIP)' zip
complete -f -o default -X '!*.+(zip|ZIP)' unzip
complete -f -o default -X '*.+(z|Z)' compress
complete -f -o default -X '!*.+(z|Z)' uncompress
complete -f -o default -X '*.+(gz|GZ)' gzip
complete -f -o default -X '!*.+(gz|GZ)' gunzip
complete -f -o default -X '*.+(bz2|BZ2)' bzip2
complete -f -o default -X '!*.+(bz2|BZ2)' bunzip2
complete -f -o default -X '!*.+(zip|ZIP|z|Z|gz|GZ|bz2|BZ2)' extract
# Documents - Postscript,pdf,dvi.....
complete -f -o default -X '!*.+(ps|PS)' gs ghostview ps2pdf ps2ascii
complete -f -o default -X \
'!*.+(dvi|DVI)' dvips dvipdf xdvi dviselect dvitype
complete -f -o default -X '!*.+(pdf|PDF)' acroread pdf2ps
complete -f -o default -X '!*.@(@(?(e)ps|?(E)PS|pdf|PDF)?\
(.gz|.GZ|.bz2|.BZ2|.Z))' gv ggv
complete -f -o default -X '!*.texi*' makeinfo texi2dvi texi2html texi2pdf
complete -f -o default -X '!*.tex' tex latex slitex
complete -f -o default -X '!*.lyx' lyx
complete -f -o default -X '!*.+(htm*|HTM*)' lynx html2ps
complete -f -o default -X \
'!*.+(doc|DOC|xls|XLS|ppt|PPT|sx?|SX?|csv|CSV|od?|OD?|ott|OTT)' soffice
# Multimedia
complete -f -o default -X \
'!*.+(gif|GIF|jp*g|JP*G|bmp|BMP|xpm|XPM|png|PNG)' xv gimp ee gqview
complete -f -o default -X '!*.+(mp3|MP3)' mpg123 mpg321
complete -f -o default -X '!*.+(ogg|OGG)' ogg123
complete -f -o default -X \
'!*.@(mp[23]|MP[23]|ogg|OGG|wav|WAV|pls|\
m3u|xm|mod|s[3t]m|it|mtm|ult|flac)' xmms
complete -f -o default -X '!*.@(mp?(e)g|MP?(E)G|wma|avi|AVI|\
asf|vob|VOB|bin|dat|vcd|ps|pes|fli|viv|rm|ram|yuv|mov|MOV|qt|\
QT|wmv|mp3|MP3|ogg|OGG|ogm|OGM|mp4|MP4|wav|WAV|asx|ASX)' xine
complete -f -o default -X '!*.pl' perl perl5
# This is a 'universal' completion function - it works when commands have
#+ a so-called 'long options' mode , ie: 'ls --all' instead of 'ls -a'
# Needs the '-o' option of grep
#+ (try the commented-out version if not available).
# First, remove '=' from completion word separators
#+ (this will allow completions like 'ls --color=auto' to work correctly).
COMP_WORDBREAKS=${COMP_WORDBREAKS/=/}
_get_longopts()
{
#$1 --help | sed -e '/--/!d' -e 's/.*--\([^[:space:].,]*\).*/--\1/'| \
#grep ^"$2" |sort -u ;
$1 --help | grep -o -e "--[^[:space:].,]*" | grep -e "$2" |sort -u
}
_longopts()
{
local cur
cur=${COMP_WORDS[COMP_CWORD]}
case "${cur:-*}" in
-*) ;;
*) return ;;
esac
case "$1" in
\~*) eval cmd="$1" ;;
*) cmd="$1" ;;
esac
COMPREPLY=( $(_get_longopts ${1} ${cur} ) )
}
complete -o default -F _longopts configure bash
complete -o default -F _longopts wget id info a2ps ls recode
_tar()
{
local cur ext regex tar untar
COMPREPLY=()
cur=${COMP_WORDS[COMP_CWORD]}
# If we want an option, return the possible long options.
case "$cur" in
-*) COMPREPLY=( $(_get_longopts $1 $cur ) ); return 0;;
esac
if [ $COMP_CWORD -eq 1 ]; then
COMPREPLY=( $( compgen -W 'c t x u r d A' -- $cur ) )
return 0
fi
case "${COMP_WORDS[1]}" in
?(-)c*f)
COMPREPLY=( $( compgen -f $cur ) )
return 0
;;
+([^Izjy])f)
ext='tar'
regex=$ext
;;
*z*f)
ext='tar.gz'
regex='t\(ar\.\)\(gz\|Z\)'
;;
*[Ijy]*f)
ext='t?(ar.)bz?(2)'
regex='t\(ar\.\)bz2\?'
;;
*)
COMPREPLY=( $( compgen -f $cur ) )
return 0
;;
esac
if [[ "$COMP_LINE" == tar*.$ext' '* ]]; then
# Complete on files in tar file.
#
# Get name of tar file from command line.
tar=$( echo "$COMP_LINE" | \
sed -e 's|^.* \([^ ]*'$regex'\) .*$|\1|' )
# Devise how to untar and list it.
untar=t${COMP_WORDS[1]//[^Izjyf]/}
COMPREPLY=( $( compgen -W "$( echo $( tar $untar $tar \
2&#62;/dev/null ) )" -- "$cur" ) )
return 0
else
# File completion on relevant files.
COMPREPLY=( $( compgen -G $cur\*.$ext ) )
fi
return 0
}
complete -F _tar -o default tar
_make()
{
local mdef makef makef_dir="." makef_inc gcmd cur prev i;
COMPREPLY=();
cur=${COMP_WORDS[COMP_CWORD]};
prev=${COMP_WORDS[COMP_CWORD-1]};
case "$prev" in
-*f)
COMPREPLY=($(compgen -f $cur ));
return 0
;;
esac;
case "$cur" in
-*)
COMPREPLY=($(_get_longopts $1 $cur ));
return 0
;;
esac;
# ... make reads
# GNUmakefile,
# then makefile
# then Makefile ...
if [ -f ${makef_dir}/GNUmakefile ]; then
makef=${makef_dir}/GNUmakefile
elif [ -f ${makef_dir}/makefile ]; then
makef=${makef_dir}/makefile
elif [ -f ${makef_dir}/Makefile ]; then
makef=${makef_dir}/Makefile
else
makef=${makef_dir}/*.mk # Local convention.
fi
# Before we scan for targets, see if a Makefile name was
#+ specified with -f.
for (( i=0; i &#60; ${#COMP_WORDS[@]}; i++ )); do
if [[ ${COMP_WORDS[i]} == -f ]]; then
# eval for tilde expansion
eval makef=${COMP_WORDS[i+1]}
break
fi
done
[ ! -f $makef ] &#38;&#38; return 0
# Deal with included Makefiles.
makef_inc=$( grep -E '^-?include' $makef |
sed -e "s,^.* ,"$makef_dir"/," )
for file in $makef_inc; do
[ -f $file ] &#38;&#38; makef="$makef $file"
done
# If we have a partial word to complete, restrict completions
#+ to matches of that word.
if [ -n "$cur" ]; then gcmd='grep "^$cur"' ; else gcmd=cat ; fi
COMPREPLY=( $( awk -F':' '/^[a-zA-Z0-9][^$#\/\t=]*:([^=]|$)/ \
{split($1,A,/ /);for(i in A)print A[i]}' \
$makef 2&#62;/dev/null | eval $gcmd ))
}
complete -F _make -X '+($*|*.[cho])' make gmake pmake
_killall()
{
local cur prev
COMPREPLY=()
cur=${COMP_WORDS[COMP_CWORD]}
# Get a list of processes
#+ (the first sed evaluation
#+ takes care of swapped out processes, the second
#+ takes care of getting the basename of the process).
COMPREPLY=( $( ps -u $USER -o comm | \
sed -e '1,1d' -e 's#[]\[]##g' -e 's#^.*/##'| \
awk '{if ($0 ~ /^'$cur'/) print $0}' ))
return 0
}
complete -F _killall killall killps
# Local Variables:
# mode:shell-script
# sh-shell:bash
# End:</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>And, here is a snippet from Andrzej Szelachowski's instructive
<TT
CLASS="FILENAME"
>.bash_profile</TT
> file.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="BASHPROF"
></A
><P
><B
>Example M-2. <TT
CLASS="FILENAME"
>.bash_profile</TT
> file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># From Andrzej Szelachowski's ~/.bash_profile:
# Note that a variable may require special treatment
#+ if it will be exported.
DARKGRAY='\e[1;30m'
LIGHTRED='\e[1;31m'
GREEN='\e[32m'
YELLOW='\e[1;33m'
LIGHTBLUE='\e[1;34m'
NC='\e[m'
PCT="\`if [[ \$EUID -eq 0 ]]; then T='$LIGHTRED' ; else T='$LIGHTBLUE'; fi;
echo \$T \`"
# For "literal" command substitution to be assigned to a variable,
#+ use escapes and double quotes:
#+ PCT="\` ... \`" . . .
# Otherwise, the value of PCT variable is assigned only once,
#+ when the variable is exported/read from .bash_profile,
#+ and it will not change afterwards even if the user ID changes.
PS1="\n$GREEN[\w] \n$DARKGRAY($PCT\t$DARKGRAY)-($PCT\u$DARKGRAY)-($PCT\!
$DARKGRAY)$YELLOW-&#62; $NC"
# Escape a variables whose value changes:
# if [[ \$EUID -eq 0 ]],
# Otherwise the value of the EUID variable will be assigned only once,
#+ as above.
# When a variable is assigned, it should be called escaped:
#+ echo \$T,
# Otherwise the value of the T variable is taken from the moment the PCT
#+ variable is exported/read from .bash_profile.
# So, in this example it would be null.
# When a variable's value contains a semicolon it should be strong quoted:
# T='$LIGHTRED',
# Otherwise, the semicolon will be interpreted as a command separator.
# Variables PCT and PS1 can be merged into a new PS1 variable:
PS1="\`if [[ \$EUID -eq 0 ]]; then PCT='$LIGHTRED';
else PCT='$LIGHTBLUE'; fi;
echo '\n$GREEN[\w] \n$DARKGRAY('\$PCT'\t$DARKGRAY)-\
('\$PCT'\u$DARKGRAY)-('\$PCT'\!$DARKGRAY)$YELLOW-&#62; $NC'\`"
# The trick is to use strong quoting for parts of old PS1 variable.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="DOSBATCH"
></A
>Appendix N. Converting DOS Batch Files to Shell Scripts</H1
><P
><A
NAME="DOSBATCH1"
></A
></P
><P
>Quite a number of programmers learned scripting on a PC running
DOS. Even the crippled DOS batch file language allowed writing some
fairly powerful scripts and applications, though they often required
extensive kludges and workarounds. Occasionally, the need still
arises to convert an old DOS batch file to a UNIX shell script. This
is generally not difficult, as DOS batch file operators are only a
limited subset of the equivalent shell scripting ones.</P
><DIV
CLASS="TABLE"
><A
NAME="AEN24336"
></A
><P
><B
>Table N-1. Batch file keywords / variables / operators, and their shell equivalents</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Batch File Operator</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Shell Script Equivalent</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Meaning</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>%</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>$</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>command-line parameter prefix</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>/</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>-</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>command option flag</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>\</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>/</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>directory path separator</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>==</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>=</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>(equal-to) string comparison test</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>!==!</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>!=</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>(not equal-to) string comparison test</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>|</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>|</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>pipe</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>@</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>set <TT
CLASS="OPTION"
>+v</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>do not echo current command</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>*</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>*</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>filename <SPAN
CLASS="QUOTE"
>"wild card"</SPAN
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#62;</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&#62;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>file redirection (overwrite)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#62;&#62;</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&#62;&#62;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>file redirection (append)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>&#60;</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>&#60;</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>redirect <TT
CLASS="FILENAME"
>stdin</TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>%VAR%</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>$VAR</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>environmental variable</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>REM</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>#</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>comment</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>NOT</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>!</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>negate following test</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>NUL</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="FILENAME"
>/dev/null</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><SPAN
CLASS="QUOTE"
>"black hole"</SPAN
> for burying command output</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>ECHO</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>echo</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>echo (many more option in Bash)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>ECHO.</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>echo</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>echo blank line</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>ECHO OFF</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>set <TT
CLASS="OPTION"
>+v</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>do not echo command(s) following</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>FOR %%VAR IN (LIST) DO</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>for var in [list]; do</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><SPAN
CLASS="QUOTE"
>"for"</SPAN
> loop</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>:LABEL</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>none (unnecessary)</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>label</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>GOTO</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>none (use a function)</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>jump to another location in the script</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>PAUSE</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>sleep</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>pause or wait an interval</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>CHOICE</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>case or select</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>menu choice</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>IF</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>if</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>if-test</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>IF EXIST <TT
CLASS="REPLACEABLE"
><I
>FILENAME</I
></TT
></TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>if [ -e filename ]</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>test if file exists</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>IF !%N==!</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>if [ -z "$N" ]</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>if replaceable parameter <SPAN
CLASS="QUOTE"
>"N"</SPAN
> not present</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>CALL</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>source or . (dot operator)</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><SPAN
CLASS="QUOTE"
>"include"</SPAN
> another script</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>COMMAND /C</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>source or . (dot operator)</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><SPAN
CLASS="QUOTE"
>"include"</SPAN
> another script (same as
CALL)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>SET</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>export</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>set an environmental variable</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>SHIFT</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>shift</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>left shift command-line argument list</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>SGN</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>-lt or -gt</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>sign (of integer)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>ERRORLEVEL</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>$?</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>exit status</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>CON</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="FILENAME"
>stdin</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><SPAN
CLASS="QUOTE"
>"console"</SPAN
> (<TT
CLASS="FILENAME"
>stdin</TT
>)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>PRN</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="FILENAME"
>/dev/lp0</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>(generic) printer device</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>LPT1</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="FILENAME"
>/dev/lp0</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>first printer device</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>COM1</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="FILENAME"
>/dev/ttyS0</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>first serial port</TD
></TR
></TBODY
></TABLE
></DIV
><P
><A
NAME="DOSUNIXEQUIV"
></A
></P
><P
>Batch files usually contain DOS commands. These must be
translated into their UNIX equivalents in order to convert a
batch file into a shell script.</P
><DIV
CLASS="TABLE"
><A
NAME="AEN24545"
></A
><P
><B
>Table N-2. DOS commands and their UNIX equivalents</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>DOS Command</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>UNIX Equivalent</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Effect</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>ASSIGN</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>ln</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>link file or directory</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>ATTRIB</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>chmod</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>change file permissions</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>CD</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>cd</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>change directory</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>CHDIR</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>cd</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>change directory</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>CLS</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>clear</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>clear screen</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>COMP</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>diff, comm, cmp</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>file compare</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>COPY</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>cp</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>file copy</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>Ctl-C</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Ctl-C</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>break (signal)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>Ctl-Z</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Ctl-D</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>EOF (end-of-file)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>DEL</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>rm</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>delete file(s)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>DELTREE</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>rm -rf</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>delete directory recursively</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>DIR</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>ls -l</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>directory listing</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>ERASE</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>rm</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>delete file(s)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>EXIT</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>exit</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>exit current process</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>FC</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>comm, cmp</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>file compare</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>FIND</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>grep</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>find strings in files</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>MD</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>mkdir</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>make directory</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>MKDIR</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>mkdir</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>make directory</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>MORE</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>more</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>text file paging filter</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>MOVE</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>mv</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>move</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>PATH</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>$PATH</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>path to executables</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>REN</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>mv</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>rename (move)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>RENAME</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>mv</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>rename (move)</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>RD</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>rmdir</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>remove directory</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>RMDIR</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>rmdir</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>remove directory</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>SORT</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>sort</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>sort file</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>TIME</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>date</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>display system time</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>TYPE</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>cat</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>output file to <TT
CLASS="FILENAME"
>stdout</TT
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>XCOPY</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>cp</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>(extended) file copy</TD
></TR
></TBODY
></TABLE
></DIV
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>Virtually all UNIX and shell operators and commands have
many more options and enhancements than their DOS and batch file
counterparts. Many DOS batch files rely on auxiliary utilities,
such as <B
CLASS="COMMAND"
>ask.com</B
>, a crippled counterpart to
<A
HREF="#READREF"
>read</A
>.</P
><P
>DOS supports only a very limited and incompatible subset of
filename <A
HREF="#GLOBBINGREF"
>wild-card expansion</A
>,
recognizing just the <SPAN
CLASS="TOKEN"
>*</SPAN
> and <SPAN
CLASS="TOKEN"
>?</SPAN
>
characters.</P
></TD
></TR
></TABLE
></DIV
><P
>Converting a DOS batch file into a shell script is generally
straightforward, and the result ofttimes reads better than the
original.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="VIEWDAT"
></A
><P
><B
>Example N-1. VIEWDATA.BAT: DOS Batch File</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>REM VIEWDATA
REM INSPIRED BY AN EXAMPLE IN "DOS POWERTOOLS"
REM BY PAUL SOMERSON
@ECHO OFF
IF !%1==! GOTO VIEWDATA
REM IF NO COMMAND-LINE ARG...
FIND "%1" C:\BOZO\BOOKLIST.TXT
GOTO EXIT0
REM PRINT LINE WITH STRING MATCH, THEN EXIT.
:VIEWDATA
TYPE C:\BOZO\BOOKLIST.TXT | MORE
REM SHOW ENTIRE FILE, 1 PAGE AT A TIME.
:EXIT0</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
> The script conversion is somewhat of an improvement.
<A
NAME="AEN24713"
HREF="#FTN.AEN24713"
><SPAN
CLASS="footnote"
>[152]</SPAN
></A
>
</P
><DIV
CLASS="EXAMPLE"
><A
NAME="VIEWDATA"
></A
><P
><B
>Example N-2. <I
CLASS="FIRSTTERM"
>viewdata.sh</I
>: Shell Script Conversion
of VIEWDATA.BAT</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# viewdata.sh
# Conversion of VIEWDATA.BAT to shell script.
DATAFILE=/home/bozo/datafiles/book-collection.data
ARGNO=1
# @ECHO OFF Command unnecessary here.
if [ $# -lt "$ARGNO" ] # IF !%1==! GOTO VIEWDATA
then
less $DATAFILE # TYPE C:\MYDIR\BOOKLIST.TXT | MORE
else
grep "$1" $DATAFILE # FIND "%1" C:\MYDIR\BOOKLIST.TXT
fi
exit 0 # :EXIT0
# GOTOs, labels, smoke-and-mirrors, and flimflam unnecessary.
# The converted script is short, sweet, and clean,
#+ which is more than can be said for the original.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><P
>Ted Davis' <A
HREF="http://www.maem.umr.edu/batch/"
TARGET="_top"
>Shell
Scripts on the PC</A
> site had a set of comprehensive
tutorials on the old-fashioned art of batch file
programming. Unfortunately the page has vanished without a
trace.</P
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="EXERCISES"
></A
>Appendix O. Exercises</H1
><P
>The exercises that follow test and extend your knowledge
of scripting. Think of them as a challenge, as an entertaining way
to take you further along the stony path toward UNIX wizardry.</P
><P
> <P
CLASS="LITERALLAYOUT"
>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;On&nbsp;a&nbsp;dingy&nbsp;side&nbsp;street&nbsp;in&nbsp;a&nbsp;run-down&nbsp;section&nbsp;of&nbsp;Hoboken,&nbsp;New&nbsp;Jersey,<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;there&nbsp;sits&nbsp;a&nbsp;nondescript&nbsp;squat&nbsp;two-story&nbsp;brick&nbsp;building&nbsp;with&nbsp;an&nbsp;inscription<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;incised&nbsp;on&nbsp;a&nbsp;marble&nbsp;plate&nbsp;in&nbsp;its&nbsp;wall:<br>
<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;<TT
CLASS="COMPUTEROUTPUT"
>Bash Scripting Hall of Fame</TT
>.<br>
<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;Inside,&nbsp;among&nbsp;various&nbsp;dusty&nbsp;uninteresting&nbsp;exhibits&nbsp;is&nbsp;a&nbsp;corroding,<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;cobweb-festooned&nbsp;brass&nbsp;plaque&nbsp;inscribed&nbsp;with&nbsp;a&nbsp;short,&nbsp;very&nbsp;short<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;list&nbsp;of&nbsp;those&nbsp;few&nbsp;persons&nbsp;who&nbsp;have&nbsp;successfully&nbsp;mastered&nbsp;the&nbsp;material<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;in&nbsp;the&nbsp;<I
CLASS="FIRSTTERM"
>Advanced Bash Scripting Guide</I
>,&nbsp;as&nbsp;evidenced&nbsp;by&nbsp;their&nbsp;performance<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;on&nbsp;the&nbsp;following&nbsp;Exercise&nbsp;sections.<br>
<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;(Alas,&nbsp;the&nbsp;author&nbsp;of&nbsp;the&nbsp;<I
CLASS="FIRSTTERM"
>ABS Guide</I
>&nbsp;is&nbsp;not&nbsp;represented&nbsp;among&nbsp;the&nbsp;exhibits.<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;This&nbsp;is&nbsp;possibly&nbsp;due&nbsp;to&nbsp;malicious&nbsp;rumors&nbsp;about&nbsp;<A
HREF="#NOCREDS"
>lack of credentials</A
>&nbsp;and<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;<A
HREF="#KTOUR0"
>deficient scripting skills</A
>.)<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</P
>
</P
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="SCRIPTANALYSIS"
></A
>O.1. Analyzing Scripts</H1
><P
>Examine the following script. Run it, then explain what it
does. Annotate the script and rewrite it in a more compact and
elegant manner.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
MAX=10000
for((nr=1; nr&#60;$MAX; nr++))
do
let "t1 = nr % 5"
if [ "$t1" -ne 3 ]
then
continue
fi
let "t2 = nr % 7"
if [ "$t2" -ne 4 ]
then
continue
fi
let "t3 = nr % 9"
if [ "$t3" -ne 5 ]
then
continue
fi
break # What happens when you comment out this line? Why?
done
echo "Number = $nr"
exit 0</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>---</P
><P
>Explain what the following script does. It is really just
a parameterized command-line pipe.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
DIRNAME=/usr/bin
FILETYPE="shell script"
LOGFILE=logfile
file "$DIRNAME"/* | fgrep "$FILETYPE" | tee $LOGFILE | wc -l
exit 0</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>---</P
><P
>Examine and explain the following script. For hints, you
might refer to the listings for <A
HREF="#FINDREF"
>find</A
> and <A
HREF="#STATREF"
>stat</A
>.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Author: Nathan Coulter
# This code is released to the public domain.
# The author gave permission to use this code snippet in the ABS Guide.
find -maxdepth 1 -type f -printf '%f\000' | {
while read -d $'\000'; do
mv "$REPLY" "$(date -d "$(stat -c '%y' "$REPLY") " '+%Y%m%d%H%M%S'
)-$REPLY"
done
}
# Warning: Test-drive this script in a "scratch" directory.
# It will somehow affect all the files there.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>---</P
><P
>A reader sent in the following code snippet.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>while read LINE
do
echo $LINE
done &#60; `tail -f /var/log/messages`</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>He wished to write a script tracking changes to the system log
file, <TT
CLASS="FILENAME"
>/var/log/messages</TT
>. Unfortunately,
the above code block hangs and does nothing
useful. Why? Fix this so it does work. (Hint:
rather than <A
HREF="#REDIRREF"
>redirecting the
<TT
CLASS="FILENAME"
>stdin</TT
> of the loop</A
>, try a <A
HREF="#PIPEREF"
>pipe</A
>.)</P
><P
>---</P
><P
>Analyze the following <SPAN
CLASS="QUOTE"
>"one-liner"</SPAN
> (here
split into two lines for clarity) contributed by Rory
Winston:</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>export SUM=0; for f in $(find src -name "*.java");
do export SUM=$(($SUM + $(wc -l $f | awk '{ print $1 }'))); done; echo $SUM</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Hint: First, break the script up into bite-sized
sections. Then, carefully examine its use of <A
HREF="#DBLPARENS"
>double-parentheses</A
> arithmetic,
the <A
HREF="#EXPORTREF"
>export</A
> command,
the <A
HREF="#FINDREF"
>find</A
> command, the
<A
HREF="#WCREF"
>wc</A
> command, and <A
HREF="#AWKREF"
>awk</A
>.</P
><P
>---</P
><P
>Analyze <A
HREF="#LIFESLOW"
>Example A-10</A
>, and reorganize it in a
simplified and more logical style. See how many of the variables
can be eliminated, and try to optimize the script to speed up
its execution time.</P
><P
>Alter the script so that it accepts any ordinary ASCII
text file as input for its initial <SPAN
CLASS="QUOTE"
>"generation"</SPAN
>. The
script will read the first <TT
CLASS="PARAMETER"
><I
>$ROW*$COL</I
></TT
>
characters, and set the occurrences of vowels as
<SPAN
CLASS="QUOTE"
>"living"</SPAN
> cells. Hint: be sure to translate the
spaces in the input file to underscore characters.</P
></DIV
><DIV
CLASS="SECT1"
><HR><H1
CLASS="SECT1"
><A
NAME="WRITINGSCRIPTS"
></A
>O.2. Writing Scripts</H1
><P
><A
NAME="WRITINGSCRIPTS1"
></A
></P
><P
>Write a script to carry out each of the following tasks.</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="EXEASY1"
></A
>EASY</B
></P
><DL
><DT
><B
CLASS="COMMAND"
>Self-reproducing Script</B
></DT
><DD
><P
>Write a script that backs itself up, that is, copies
itself to a file named <TT
CLASS="FILENAME"
>backup.sh</TT
>.</P
><P
>Hint: Use the <A
HREF="#CATREF"
>cat</A
> command
and the appropriate <A
HREF="#SCRNAMEPARAM"
>positional
parameter</A
>.</P
></DD
><DT
><B
CLASS="COMMAND"
>Home Directory Listing</B
></DT
><DD
><P
>Perform a recursive directory listing on the user's home
directory and save the information to a file. Compress
the file, have the script prompt the user to insert
a USB flash drive, then press <B
CLASS="KEYCAP"
>ENTER</B
>.
Finally, save the file to the flash drive after making
certain the flash drive has properly mounted by parsing
the output of <A
HREF="#DFREF"
>df</A
>. Note that
the flash drive must be <I
CLASS="FIRSTTERM"
>unmounted</I
>
before it is removed.</P
></DD
><DT
><B
CLASS="COMMAND"
>Converting <A
HREF="#FORLOOPREF1"
>for</A
>
loops to <A
HREF="#WHILELOOPREF"
>while</A
> and <A
HREF="#UNTILLOOPREF"
>until</A
> loops</B
></DT
><DD
><P
>Convert the <I
CLASS="FIRSTTERM"
>for loops</I
> in <A
HREF="#EX22"
>Example 11-1</A
> to <I
CLASS="FIRSTTERM"
>while
loops</I
>. Hint: store the data in an <A
HREF="#ARRAYREF"
>array</A
> and step through the array
elements.</P
><P
>Having already done the <SPAN
CLASS="QUOTE"
>"heavy lifting,"</SPAN
>
now convert the loops in the example to <I
CLASS="FIRSTTERM"
> until
loops</I
>.</P
></DD
><DT
><B
CLASS="COMMAND"
>Changing the line spacing of a text file</B
></DT
><DD
><P
>Write a script that reads each line of a target file, then
writes the line back to <TT
CLASS="FILENAME"
>stdout</TT
>, but with
an extra blank line following. This has the effect of
<EM
>double-spacing</EM
> the file.</P
><P
>Include all necessary code to check whether the script
gets the necessary command-line argument (a filename),
and whether the specified file exists.</P
><P
>When the script runs correctly, modify it to
<EM
>triple-space</EM
> the target file.</P
><P
>Finally, write a script to remove all blank lines from
the target file, <EM
>single-spacing</EM
> it.</P
></DD
><DT
><B
CLASS="COMMAND"
>Backwards Listing</B
></DT
><DD
><P
>Write a script that echoes itself to
<TT
CLASS="FILENAME"
>stdout</TT
>, but
<EM
>backwards</EM
>.</P
></DD
><DT
><B
CLASS="COMMAND"
>Automatically Decompressing Files</B
></DT
><DD
><P
>Given a list of filenames as input, this script
queries each target file (parsing the output of the
<A
HREF="#FILEREF"
>file</A
> command) for
the type of compression used on it. Then the script
automatically invokes the appropriate decompression command
(<B
CLASS="COMMAND"
>gunzip</B
>, <B
CLASS="COMMAND"
>bunzip2</B
>,
<B
CLASS="COMMAND"
>unzip</B
>, <B
CLASS="COMMAND"
>uncompress</B
>,
or whatever). If a target file is not compressed, the
script emits a warning message, but takes no other action
on that particular file.</P
></DD
><DT
><B
CLASS="COMMAND"
>Unique System ID</B
></DT
><DD
><P
>Generate a <SPAN
CLASS="QUOTE"
>"unique"</SPAN
> 6-digit hexadecimal
identifier for your computer. Do <EM
>not</EM
>
use the flawed <A
HREF="#HOSTIDREF"
>hostid</A
>
command. Hint: <B
CLASS="COMMAND"
><A
HREF="#MD5SUMREF"
>md5sum</A
>
<A
HREF="#DATAFILESREF1"
><TT
CLASS="FILENAME"
>/etc/passwd</TT
></A
></B
>,
then select the first 6 digits of output.</P
></DD
><DT
><B
CLASS="COMMAND"
>Backup</B
></DT
><DD
><P
>Archive as a <SPAN
CLASS="QUOTE"
>"tarball"</SPAN
>
(<TT
CLASS="FILENAME"
>*.tar.gz</TT
> file) all the files
in your home directory tree
(<TT
CLASS="FILENAME"
>/home/your-name</TT
>) that have
been modified in the last 24 hours. Hint: use <A
HREF="#FINDREF"
>find</A
>.</P
><P
>Optional: you may use this as the basis of a
<I
CLASS="FIRSTTERM"
>backup</I
> script.</P
></DD
><DT
><B
CLASS="COMMAND"
>Checking whether a process is still running</B
></DT
><DD
><P
>Given a <A
HREF="#PROCESSIDREF"
>process ID</A
>
(<I
CLASS="FIRSTTERM"
>PID</I
>) as an argument, this script
will check, at user-specified intervals, whether
the given process is still running. You may use
the <A
HREF="#PPSSREF"
>ps</A
> and <A
HREF="#SLEEPREF"
>sleep</A
> commands.</P
></DD
><DT
><B
CLASS="COMMAND"
>Primes</B
></DT
><DD
><P
>Print (to <TT
CLASS="FILENAME"
>stdout</TT
>) all
prime numbers between 60000 and 63000. The output
should be nicely formatted in columns (hint:
use <A
HREF="#PRINTFREF"
>printf</A
>).</P
></DD
><DT
><B
CLASS="COMMAND"
>Lottery Numbers</B
></DT
><DD
><P
>One type of lottery involves picking five
different numbers, in the range of 1 - 50. Write a
script that generates five pseudorandom numbers in this
range, <EM
>with no duplicates</EM
>. The
script will give the option of echoing the numbers to
<TT
CLASS="FILENAME"
>stdout</TT
> or saving them to a file,
along with the date and time the particular number set
was generated. (If your script consistently generates
<EM
>winning</EM
> lottery numbers, then you
can retire on the proceeds and leave shell scripting to
those of us who have to work for a living.)</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="EXMEDIUM1"
></A
>INTERMEDIATE</B
></P
><DL
><DT
><B
CLASS="COMMAND"
>Integer or String</B
></DT
><DD
><P
>Write a script <A
HREF="#FUNCTIONREF"
>function</A
>
that determines if an argument passed to it is an integer
or a string. The function will return TRUE (0) if
passed an integer, and FALSE (1) if passed a string.</P
><P
>Hint: What does the following expression return
when <TT
CLASS="VARNAME"
>$1</TT
> is <EM
>not</EM
>
an integer?</P
><P
><TT
CLASS="VARNAME"
>expr $1 + 0</TT
></P
></DD
><DT
><B
CLASS="COMMAND"
><A
HREF="#ASCIIDEF"
>ASCII</A
>
to Integer</B
></DT
><DD
><P
>The <I
CLASS="FIRSTTERM"
>atoi</I
> function in
<B
CLASS="COMMAND"
>C</B
> converts a string character to
an integer. Write a shell script function that performs
the same operation. Likewise, write a shell script function
that does the inverse, mirroring the <B
CLASS="COMMAND"
>C</B
>
<I
CLASS="FIRSTTERM"
>itoa</I
> function which converts an
integer into an ASCII character.</P
></DD
><DT
><B
CLASS="COMMAND"
>Managing Disk Space</B
></DT
><DD
><P
>List, one at a time, all files larger than 100K in
the <TT
CLASS="FILENAME"
>/home/username</TT
>
directory tree. Give the user the option to delete or
compress the file, then proceed to show the next one. Write
to a logfile the names of all deleted files and the
deletion times.</P
></DD
><DT
><B
CLASS="COMMAND"
>Banner</B
></DT
><DD
><P
>Simulate the functionality of the deprecated <A
HREF="#BANNERREF"
>banner</A
> command in a script.</P
></DD
><DT
><B
CLASS="COMMAND"
>Removing Inactive Accounts</B
></DT
><DD
><P
>Inactive accounts on a network server waste disk space and may
become a security risk. Write an administrative script
(to be invoked by <I
CLASS="FIRSTTERM"
>root</I
> or the <A
HREF="#CRONREF"
>cron daemon</A
>) that checks
for and deletes user accounts that have not been accessed
within the last 90 days.</P
></DD
><DT
><B
CLASS="COMMAND"
>Enforcing Disk Quotas</B
></DT
><DD
><P
>Write a script for a multi-user system that checks users'
disk usage. If a user surpasses a preset limit
(500 MB, for example) in her <TT
CLASS="FILENAME"
>/home/username</TT
>
directory, then the script automatically sends her a
<SPAN
CLASS="QUOTE"
>"pigout"</SPAN
> warning e-mail.</P
><P
>The
script will use the <A
HREF="#DUREF"
>du</A
>
and <A
HREF="#COMMMAIL1"
>mail</A
> commands. As
an option, it will allow setting and enforcing quotas
using the <A
HREF="#QUOTAREF"
>quota</A
> and <A
HREF="#SETQUOTAREF"
>setquota</A
> commands.</P
></DD
><DT
><B
CLASS="COMMAND"
>Logged in User Information</B
></DT
><DD
><P
>For all logged in users, show their real names and the time
and date of their last login.</P
><P
>Hint: use <A
HREF="#WHOREF"
>who</A
>,
<A
HREF="#LASTLOGREF"
>lastlog</A
>,
and parse <A
HREF="#DATAFILESREF1"
><TT
CLASS="FILENAME"
>/etc/passwd</TT
></A
>.</P
></DD
><DT
><B
CLASS="COMMAND"
>Safe Delete</B
></DT
><DD
><P
>Implement, as a script, a <SPAN
CLASS="QUOTE"
>"safe"</SPAN
> delete
command, <TT
CLASS="FILENAME"
>sdel.sh</TT
>. Filenames passed as
command-line arguments to this script are not deleted,
but instead <A
HREF="#GZIPREF"
>gzipped</A
>
if not already compressed (use <A
HREF="#FILEREF"
>file</A
> to check), then moved
to a <TT
CLASS="FILENAME"
>~/TRASH</TT
>
directory. Upon invocation, the script checks the <TT
CLASS="FILENAME"
>~/TRASH</TT
> directory for files
older than 48 hours and <A
HREF="#RMREF"
>permanently
deletes</A
> them. (An better alternative might be to
have a second script handle this, periodically invoked
by the <A
HREF="#CRONREF"
>cron daemon</A
>.)</P
><P
><EM
>Extra credit:</EM
> Write the script
so it can handle files and directories <A
HREF="#RMRECURS"
>recursively</A
>. This would give it
the capability of <SPAN
CLASS="QUOTE"
>"safely deleting"</SPAN
> entire
directory structures.</P
></DD
><DT
><B
CLASS="COMMAND"
>Making Change</B
></DT
><DD
><P
>What is the most efficient way to make change for $1.68,
using only coins in common circulations (up to 25c)? It's
6 quarters, 1 dime, a nickel, and three cents.</P
><P
>Given any arbitrary command-line input in dollars and
cents ($*.??), calculate the change, using the minimum
number of coins. If your home country is not the United
States, you may use your local currency units instead. The
script will need to parse the command-line input, then
change it to multiples of the smallest monetary unit (cents
or whatever). Hint: look at <A
HREF="#EX61"
>Example 24-8</A
>.</P
></DD
><DT
><B
CLASS="COMMAND"
>Quadratic Equations</B
></DT
><DD
><P
>Solve a <I
CLASS="FIRSTTERM"
>quadratic</I
> equation of the form
<TT
CLASS="PARAMETER"
><I
>Ax^2 + Bx + C = 0</I
></TT
>. Have a script take
as arguments the coefficients, <TT
CLASS="USERINPUT"
><B
>A</B
></TT
>,
<TT
CLASS="USERINPUT"
><B
>B</B
></TT
>, and <TT
CLASS="USERINPUT"
><B
>C</B
></TT
>,
and return the solutions to five decimal places.</P
><P
>Hint: pipe the coefficients to <A
HREF="#BCREF"
>bc</A
>, using the well-known formula,
<TT
CLASS="PARAMETER"
><I
>x = ( -B +/- sqrt( B^2 - 4AC ) ) / 2A</I
></TT
>.</P
></DD
><DT
><B
CLASS="COMMAND"
>Table of Logarithms</B
></DT
><DD
><P
>Using the <A
HREF="#BCREF"
>bc</A
> and <A
HREF="#PRINTFREF"
>printf</A
> commands, print out a
nicely-formatted table of eight-place natural logarithms
in the interval between 0.00 and 100.00, in steps of
.01.</P
><P
>Hint: <I
CLASS="FIRSTTERM"
>bc</I
> requires the
<TT
CLASS="OPTION"
>-l</TT
> option to load the math library.</P
></DD
><DT
><B
CLASS="COMMAND"
>Unicode Table</B
></DT
><DD
><P
>Using <A
HREF="#ASCIISH"
>Example T-1</A
> as a template,
write a script that prints to a file a complete
<A
HREF="#UNICODEREF"
>Unicode</A
> table.</P
><P
>Hint: Use the <TT
CLASS="OPTION"
>-e</TT
> option to
<A
HREF="#ECHOREF"
>echo</A
>:
<B
CLASS="COMMAND"
>echo -e '\uXXXX'</B
>, where
<TT
CLASS="REPLACEABLE"
><I
>XXXX</I
></TT
>
is the Unicode numerical character designation.
This requires <A
HREF="#BASH42"
>version 4.2</A
>
or later of Bash.</P
></DD
><DT
><B
CLASS="COMMAND"
>Sum of Matching Numbers</B
></DT
><DD
><P
>Find the sum of all five-digit numbers (in the range
10000 - 99999) containing <EM
>exactly two</EM
>
out of the following set of digits: { 4, 5, 6 }. These may
repeat within the same number, and if so, they count once
for each occurrence.</P
><P
>Some examples of <I
CLASS="FIRSTTERM"
>matching numbers</I
> are
42057, 74638, and 89515.</P
></DD
><DT
><B
CLASS="COMMAND"
>Lucky Numbers</B
></DT
><DD
><P
>A <I
CLASS="FIRSTTERM"
>lucky number</I
> is one whose
individual digits add up to 7, in successive additions. For
example, 62431 is a <I
CLASS="FIRSTTERM"
>lucky number</I
>
(6 + 2 + 4 + 3 + 1 = 16, 1 + 6 = 7). Find all the
<I
CLASS="FIRSTTERM"
>lucky numbers</I
> between 1000 and
10000.</P
></DD
><DT
><B
CLASS="COMMAND"
>Craps</B
></DT
><DD
><P
>Borrowing the ASCII graphics from <A
HREF="#PETALS"
>Example A-40</A
>,
write a script that plays the well-known gambling game of
<I
CLASS="FIRSTTERM"
>craps</I
>. The script will accept bets
from one or more players, roll the dice, and keep track of
wins and losses, as well as of each player's bankroll.</P
></DD
><DT
><B
CLASS="COMMAND"
>Tic-tac-toe</B
></DT
><DD
><P
>Write a script that plays the child's game of
<I
CLASS="FIRSTTERM"
>tic-tac-toe</I
> against a human
player. The script will let the human choose whether
to take the first move. The script will follow
an optimal strategy, and therefore never lose. To simplify
matters, you may use ASCII graphics:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> o | x |
----------
| x |
----------
| o |
Your move, human (row, column)?</PRE
></FONT
></TD
></TR
></TABLE
></P
></DD
><DT
><B
CLASS="COMMAND"
>Alphabetizing a String</B
></DT
><DD
><P
>Alphabetize (in ASCII order) an arbitrary string
read from the command-line.</P
></DD
><DT
><B
CLASS="COMMAND"
>Parsing</B
></DT
><DD
><P
>Parse <A
HREF="#DATAFILESREF1"
><TT
CLASS="FILENAME"
>/etc/passwd</TT
></A
>,
and output its contents in nice, easy-to-read tabular
form.</P
></DD
><DT
><B
CLASS="COMMAND"
>Logging Logins</B
></DT
><DD
><P
>Parse <TT
CLASS="FILENAME"
>/var/log/messages</TT
> to
produce a nicely formatted file of user logins and login
times. The script may need to run as
<I
CLASS="FIRSTTERM"
>root</I
>. (Hint: Search for the string
<SPAN
CLASS="QUOTE"
>"LOGIN."</SPAN
>)</P
></DD
><DT
><B
CLASS="COMMAND"
>Pretty-Printing a Data File</B
></DT
><DD
><P
>Certain database and spreadsheet packages use
save-files with the fields separated by commas, commonly
referred to as <I
CLASS="FIRSTTERM"
>comma-separated values</I
>
or CSVs. Other applications often need to parse these
files.</P
><P
>Given a data file with comma-separated
<A
HREF="#FIELDREF"
>fields</A
>, of the form:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>Jones,Bill,235 S. Williams St.,Denver,CO,80221,(303) 244-7989
Smith,Tom,404 Polk Ave.,Los Angeles,CA,90003,(213) 879-5612
...</PRE
></FONT
></TD
></TR
></TABLE
>
Reformat the data and print it out to
<TT
CLASS="FILENAME"
>stdout</TT
> in labeled, evenly-spaced columns.</P
></DD
><DT
><B
CLASS="COMMAND"
>Justification</B
></DT
><DD
><P
>Given ASCII text input either from
<TT
CLASS="FILENAME"
>stdin</TT
> or a file, adjust
the word spacing to right-justify each line to a
user-specified line-width, then send the output to
<TT
CLASS="FILENAME"
>stdout</TT
>.</P
></DD
><DT
><B
CLASS="COMMAND"
>Mailing List</B
></DT
><DD
><P
>Using the <A
HREF="#COMMMAIL1"
>mail</A
> command,
write a script that manages a simple mailing list. The
script automatically e-mails the monthly company newsletter,
read from a specified text file, and sends it to all the
addresses on the mailing list, which the script reads from
another specified file.</P
></DD
><DT
><B
CLASS="COMMAND"
>Generating Passwords</B
></DT
><DD
><P
>Generate pseudorandom 8-character passwords, using
characters in the ranges [0-9], [A-Z], [a-z]. Each password
must contain at least two digits.</P
></DD
><DT
><B
CLASS="COMMAND"
>Monitoring a User</B
></DT
><DD
><P
>You suspect that one particular user on the network
has been abusing her privileges and possibly attempting to
hack the system. Write a script to automatically monitor
and log her activities when she's signed on. The log file
will save entries for the previous week, and delete those
entries more than seven days old.</P
><P
>You may use <A
HREF="#LASTREF"
>last</A
>,
<A
HREF="#LASTLOGREF"
>lastlog</A
>, and <A
HREF="#LASTCOMMREF"
>lastcomm</A
> to aid your
surveillance of the suspected fiend.</P
></DD
><DT
><B
CLASS="COMMAND"
>Checking for Broken Links</B
></DT
><DD
><P
>Using <A
HREF="#LYNXREF"
>lynx</A
> with the
<TT
CLASS="OPTION"
>-traversal</TT
> option, write a script that
checks a Web site for broken links.</P
></DD
></DL
></DIV
><P
></P
><DIV
CLASS="VARIABLELIST"
><P
><B
><A
NAME="EXDIFFICULT1"
></A
>DIFFICULT</B
></P
><DL
><DT
><B
CLASS="COMMAND"
>Testing Passwords</B
></DT
><DD
><P
>Write a script to check and validate passwords. The object
is to flag <SPAN
CLASS="QUOTE"
>"weak"</SPAN
> or easily guessed password
candidates.</P
><P
>A trial password will be input to the script as a
command-line parameter. To be considered acceptable,
a password must meet the following minimum qualifications:
<P
></P
><UL
><LI
><P
>Minimum length of 8 characters</P
></LI
><LI
><P
>Must contain at least one numeric character</P
></LI
><LI
><P
>Must contain at least one of the following
non-alphabetic characters: <SPAN
CLASS="TOKEN"
>@</SPAN
>,
<SPAN
CLASS="TOKEN"
>#</SPAN
>, <SPAN
CLASS="TOKEN"
>$</SPAN
>, <SPAN
CLASS="TOKEN"
>%</SPAN
>,
<SPAN
CLASS="TOKEN"
>&#38;</SPAN
>, <SPAN
CLASS="TOKEN"
>*</SPAN
>, <SPAN
CLASS="TOKEN"
>+</SPAN
>,
<SPAN
CLASS="TOKEN"
>-</SPAN
>, <SPAN
CLASS="TOKEN"
>=</SPAN
></P
></LI
></UL
></P
><P
>Optional:
<P
></P
><UL
><LI
><P
>Do a dictionary check on every sequence of at least
four consecutive alphabetic characters in the password under
test. This will eliminate passwords containing embedded
<SPAN
CLASS="QUOTE"
>"words"</SPAN
> found in a standard dictionary.</P
></LI
><LI
><P
>Enable the script to check all the passwords on your
system. These do not reside in
<A
HREF="#DATAFILESREF1"
><TT
CLASS="FILENAME"
>/etc/passwd</TT
></A
>.</P
></LI
></UL
></P
><P
>This exercise tests mastery of <A
HREF="#REGEXREF"
>Regular Expressions</A
>.</P
></DD
><DT
><B
CLASS="COMMAND"
>Cross Reference</B
></DT
><DD
><P
>Write a script that generates a
<I
CLASS="FIRSTTERM"
>cross-reference</I
>
(<I
CLASS="FIRSTTERM"
>concordance</I
>) on a target file.
The output will be a listing of all word occurrences in
the target file, along with the line numbers in which
each word occurs. Traditionally, <I
CLASS="FIRSTTERM"
>linked
list</I
> constructs would be used in such
applications. Therefore, you should investigate <A
HREF="#ARRAYREF"
>arrays</A
> in the course of
this exercise. <A
HREF="#WF"
>Example 16-12</A
> is probably
<EM
>not</EM
> a good place to start.</P
></DD
><DT
><A
NAME="NEWTONSQRT"
></A
><B
CLASS="COMMAND"
>Square Root</B
></DT
><DD
><P
>Write a script to calculate square roots of numbers
using <I
CLASS="FIRSTTERM"
>Newton's Method</I
>.</P
><P
>The algorithm for this, expressed as a snippet of Bash
<A
HREF="#PSEUDOCODEREF"
>pseudo-code</A
> is:</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
># (Isaac) Newton's Method for speedy extraction
#+ of square roots.
guess = $argument
# $argument is the number to find the square root of.
# $guess is each successive calculated "guess" -- or trial solution --
#+ of the square root.
# Our first "guess" at a square root is the argument itself.
oldguess = 0
# $oldguess is the previous $guess.
tolerance = .000001
# To how close a tolerance we wish to calculate.
loopcnt = 0
# Let's keep track of how many times through the loop.
# Some arguments will require more loop iterations than others.
while [ ABS( $guess $oldguess ) -gt $tolerance ]
# ^^^^^^^^^^^^^^^^^^^^^^^ Fix up syntax, of course.
# "ABS" is a (floating point) function to find the absolute value
#+ of the difference between the two terms.
# So, as long as difference between current and previous
#+ trial solution (guess) exceeds the tolerance, keep looping.
do
oldguess = $guess # Update $oldguess to previous $guess.
# =======================================================
guess = ( $oldguess + ( $argument / $oldguess ) ) / 2.0
# = 1/2 ( ($oldguess **2 + $argument) / $oldguess )
# equivalent to:
# = 1/2 ( $oldguess + $argument / $oldguess )
# that is, "averaging out" the trial solution and
#+ the proportion of argument deviation
#+ (in effect, splitting the error in half).
# This converges on an accurate solution
#+ with surprisingly few loop iterations . . .
#+ for arguments &#62; $tolerance, of course.
# =======================================================
(( loopcnt++ )) # Update loop counter.
done</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>It's a simple enough recipe, and
<EM
>seems</EM
> at first glance easy enough to
convert into a working Bash script. The problem, though,
is that Bash has <A
HREF="#NOFLOATINGPOINT"
>no native
support for floating point numbers</A
>. So, the script
writer needs to use <A
HREF="#BCREF"
>bc</A
> or
possibly <A
HREF="#AWKREF"
>awk</A
> to convert the
numbers and do the calculations. It could get rather messy
. . .</P
></DD
><DT
><B
CLASS="COMMAND"
>Logging File Accesses</B
></DT
><DD
><P
>Log all accesses to the files in <TT
CLASS="FILENAME"
>/etc</TT
> during the course of
a single day. This information should include the filename,
user name, and access time. If any alterations to the
files take place, that will be flagged. Write this data
as tabular (tab-separated) formatted records in a logfile.</P
></DD
><DT
><B
CLASS="COMMAND"
>Monitoring Processes</B
></DT
><DD
><P
>Write a script to continually monitor all running
processes and to keep track of how many child processes each
parent spawns. If a process spawns more than five children,
then the script sends an e-mail to the system administrator
(or <I
CLASS="FIRSTTERM"
>root</I
>) with all relevant
information, including the time, PID of the parent, PIDs
of the children, etc. The script appends a report to a log
file every ten minutes. </P
></DD
><DT
><B
CLASS="COMMAND"
>Strip Comments</B
></DT
><DD
><P
>Strip all comments from a shell script whose name
is specified on the command-line. Note that the initial
<A
HREF="#SHABANGREF"
>#! line</A
> must not be
stripped out.</P
></DD
><DT
><B
CLASS="COMMAND"
>Strip HTML Tags</B
></DT
><DD
><P
>Strip all the HTML tags from a specified HTML file, then
reformat it into lines between 60 and 75 characters
in length. Reset paragraph and block spacing, as
appropriate, and convert HTML tables to their approximate
text equivalent.</P
></DD
><DT
><B
CLASS="COMMAND"
>XML Conversion</B
></DT
><DD
><P
>Convert an XML file to both HTML and text format.</P
><P
>Optional: A script that converts Docbook/SGML to XML.</P
></DD
><DT
><A
NAME="CSPAMMERS"
></A
><B
CLASS="COMMAND"
>Chasing Spammers</B
></DT
><DD
><P
> Write a script that analyzes a spam e-mail by doing
DNS lookups on the IP addresses in the headers to identify
the relay hosts as well as the originating ISP. The
script will forward the unaltered spam message to the
responsible ISPs. Of course, it will be necessary to
filter out <EM
>your own ISP's IP address</EM
>,
so you don't end up complaining about yourself.</P
><P
>As necessary, use the appropriate <A
HREF="#COMMUNINFO1"
>network analysis commands</A
>.</P
><P
>For some ideas, see <A
HREF="#ISSPAMMER"
>Example 16-41</A
> and <A
HREF="#ISSPAMMER2"
>Example A-28</A
>.</P
><P
>Optional: Write a script that searches through a list of
e-mail messages and deletes the spam according to specified
filters.</P
></DD
><DT
><B
CLASS="COMMAND"
>Creating man pages</B
></DT
><DD
><P
>Write a script that automates the process of creating
<A
HREF="#MANREF"
>man pages</A
>.</P
><P
>Given a text file which contains information to be
formatted into a <I
CLASS="FIRSTTERM"
>man page</I
>, the
script will read the file, then invoke the appropriate
<A
HREF="#GROFFREF"
>groff</A
> commands to
output the corresponding <I
CLASS="FIRSTTERM"
>man page</I
>
to <TT
CLASS="FILENAME"
>stdout</TT
>. The text file contains
blocks of information under the standard <I
CLASS="FIRSTTERM"
>man
page</I
> headings, i.e., NAME, SYNOPSIS,
DESCRIPTION, etc.</P
><P
><A
HREF="#MANED"
>Example A-39</A
> is an instructive first step.</P
></DD
><DT
><B
CLASS="COMMAND"
>Hex Dump</B
></DT
><DD
><P
>Do a hex(adecimal) dump on a binary file
specified as an argument to the script. The output should
be in neat tabular <A
HREF="#FIELDREF"
>fields</A
>,
with the first field showing the address, each of the
next 8 fields a 4-byte hex number, and the final field
the ASCII equivalent of the previous 8 fields.</P
><P
>The obvious followup to this is to extend the hex
dump script into a disassembler. Using a lookup table,
or some other clever gimmick, convert the hex values into
80x86 op codes.</P
></DD
><DT
><B
CLASS="COMMAND"
>Emulating a Shift Register</B
></DT
><DD
><P
>Using <A
HREF="#STACKEX"
>Example 27-15</A
> as an inspiration,
write a script that emulates a 64-bit shift register as
an <A
HREF="#ARRAYREF"
>array</A
>. Implement
functions to <I
CLASS="FIRSTTERM"
>load</I
> the register,
<I
CLASS="FIRSTTERM"
>shift left</I
>, <I
CLASS="FIRSTTERM"
>shift
right</I
>, and <I
CLASS="FIRSTTERM"
>rotate</I
>
it. Finally, write a function that interprets the register
contents as eight 8-bit ASCII characters.</P
></DD
><DT
><B
CLASS="COMMAND"
>Calculating Determinants</B
></DT
><DD
><P
>Write a script that calculates
determinants
<A
NAME="AEN25254"
HREF="#FTN.AEN25254"
><SPAN
CLASS="footnote"
>[153]</SPAN
></A
>
by <A
HREF="#RECURSIONREF0"
>recursively</A
> expanding the
<I
CLASS="FIRSTTERM"
>minors</I
>. Use a 4 x 4 determinant as
a test case.</P
></DD
><DT
><B
CLASS="COMMAND"
>Hidden Words</B
></DT
><DD
><P
>Write a <SPAN
CLASS="QUOTE"
>"word-find"</SPAN
> puzzle generator,
a script that hides 10 input words in a 10 x 10 array
of random letters. The words may be hidden across, down,
or diagonally.</P
><P
>Optional: Write a script that <EM
>solves</EM
>
word-find puzzles. To keep this from becoming too difficult,
the solution script will find only horizontal and vertical
words. (Hint: Treat each row and column as a string, and
search for substrings.)</P
></DD
><DT
><B
CLASS="COMMAND"
>Anagramming</B
></DT
><DD
><P
> Anagram 4-letter input. For example, the
anagrams of <EM
>word</EM
> are:
<EM
>do or rod row word</EM
>. You may use
<TT
CLASS="FILENAME"
>/usr/share/dict/linux.words</TT
> as the
reference list.</P
></DD
><DT
><B
CLASS="COMMAND"
>Word Ladders</B
></DT
><DD
><P
>A <SPAN
CLASS="QUOTE"
>"word ladder"</SPAN
> is a sequence of words,
with each successive word in the sequence differing from
the previous one by a single letter.</P
><P
>For example, to <SPAN
CLASS="QUOTE"
>"ladder"</SPAN
> from
<EM
>mark</EM
> to
<EM
>vase</EM
>:</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>mark --&#62; park --&#62; part --&#62; past --&#62; vast --&#62; vase
^ ^ ^ ^ ^</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Write a script that solves word ladder puzzles. Given
a starting and an ending word, the script will list all
intermediate steps in the <SPAN
CLASS="QUOTE"
>"ladder."</SPAN
> Note
that <EM
>all</EM
> words in the sequence must
be legitimate dictionary words.</P
></DD
><DT
><B
CLASS="COMMAND"
>Fog Index</B
></DT
><DD
><P
>The <SPAN
CLASS="QUOTE"
>"fog index"</SPAN
> of a passage of text
estimates its reading difficulty, as a number corresponding
roughly to a school grade level. For example, a passage
with a fog index of 12 should be comprehensible to anyone
with 12 years of schooling.</P
><P
>The Gunning version of the fog index uses the following
algorithm.</P
><P
></P
><OL
TYPE="1"
><LI
><P
>Choose a section of the text at least
100 words in length.</P
></LI
><LI
><P
>Count the number of sentences (a portion of
a sentence truncated by the boundary of the text section
counts as one).</P
></LI
><LI
><P
>Find the average number of words per
sentence.</P
><P
>AVE_WDS_SEN = TOTAL_WORDS / SENTENCES</P
></LI
><LI
><P
>Count the number of <SPAN
CLASS="QUOTE"
>"difficult"</SPAN
>
words in the segment -- those containing at least
3 syllables. Divide this quantity by total words to
get the proportion of difficult words.</P
><P
>PRO_DIFF_WORDS = LONG_WORDS / TOTAL_WORDS</P
></LI
><LI
><P
>The Gunning fog index is the sum of the above two
quantities, multiplied by 0.4, then rounded to the
nearest integer.</P
><P
>G_FOG_INDEX = int ( 0.4 * ( AVE_WDS_SEN + PRO_DIFF_WORDS ) )</P
></LI
></OL
><P
>Step 4 is by far the most difficult portion of the
exercise. There exist various algorithms for estimating
the syllable count of a word. A rule-of-thumb formula
might consider the number of letters in a word and the
vowel-consonant mix.</P
><P
>A strict interpretation of the Gunning fog index does
not count compound words and proper nouns as
<SPAN
CLASS="QUOTE"
>"difficult"</SPAN
> words, but this would enormously
complicate the script.</P
></DD
><DT
><B
CLASS="COMMAND"
>Calculating PI using Buffon's Needle</B
></DT
><DD
><P
>The Eighteenth Century French mathematician de Buffon
came up with a novel experiment. Repeatedly drop a needle
of length <TT
CLASS="REPLACEABLE"
><I
>n</I
></TT
> onto a wooden floor
composed of long and narrow parallel boards. The cracks
separating the equal-width floorboards are a fixed distance
<TT
CLASS="REPLACEABLE"
><I
>d</I
></TT
> apart. Keep track of the
total drops and the number of times the needle intersects
a crack on the floor. The ratio of these two quantities
turns out to be a fractional multiple of PI.</P
><P
>In the spirit of <A
HREF="#CANNON"
>Example 16-50</A
>, write a
script that runs a Monte Carlo simulation of
<I
CLASS="FIRSTTERM"
>Buffon's Needle</I
>. To simplify matters,
set the needle length equal to the distance between the
cracks, <TT
CLASS="PARAMETER"
><I
>n = d</I
></TT
>.</P
><P
>Hint: there are actually two critical variables:
the distance from the center of the needle to the nearest
crack, and the inclination angle of the needle to that crack.
You may use <A
HREF="#BCREF"
>bc</A
> to handle
the calculations.</P
></DD
><DT
><B
CLASS="COMMAND"
>Playfair Cipher</B
></DT
><DD
><P
>Implement the Playfair (Wheatstone) Cipher in a
script.</P
><P
>The Playfair Cipher encrypts text by substitution
of <I
CLASS="FIRSTTERM"
>digrams</I
> (2-letter groupings).
It is traditional to use a 5 x 5 letter scrambled-alphabet
<I
CLASS="FIRSTTERM"
>key square</I
> for the encryption and
decryption.</P
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
> C O D E S
A B F G H
I K L M N
P Q R T U
V W X Y Z
Each letter of the alphabet appears once, except "I" also represents
"J". The arbitrarily chosen key word, "CODES" comes first, then all
the rest of the alphabet, in order from left to right, skipping letters
already used.
To encrypt, separate the plaintext message into digrams (2-letter
groups). If a group has two identical letters, delete the second, and
form a new group. If there is a single letter left over at the end,
insert a "null" character, typically an "X."
THIS IS A TOP SECRET MESSAGE
TH IS IS AT OP SE CR ET ME SA GE
For each digram, there are three possibilities.
-----------------------------------------------
1) Both letters will be on the same row of the key square:
For each letter, substitute the one immediately to the right, in that
row. If necessary, wrap around left to the beginning of the row.
or
2) Both letters will be in the same column of the key square:
For each letter, substitute the one immediately below it, in that
row. If necessary, wrap around to the top of the column.
or
3) Both letters will form the corners of a rectangle within the key square:
For each letter, substitute the one on the other corner the rectangle
which lies on the same row.
The "TH" digram falls under case #3.
G H
M N
T U (Rectangle with "T" and "H" at corners)
T --&#62; U
H --&#62; G
The "SE" digram falls under case #1.
C O D E S (Row containing "S" and "E")
S --&#62; C (wraps around left to beginning of row)
E --&#62; S
=========================================================================
To decrypt encrypted text, reverse the above procedure under cases #1
and #2 (move in opposite direction for substitution). Under case #3,
just take the remaining two corners of the rectangle.
Helen Fouche Gaines' classic work, ELEMENTARY CRYPTANALYSIS (1939), gives a
fairly detailed description of the Playfair Cipher and its solution methods.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>This script will have three main sections</P
><P
></P
><OL
TYPE="I"
><LI
><P
>Generating the <I
CLASS="FIRSTTERM"
>key square</I
>,
based on a user-input keyword.</P
></LI
><LI
><P
>Encrypting a <I
CLASS="FIRSTTERM"
>plaintext</I
>
message.</P
></LI
><LI
><P
>Decrypting encrypted
text.</P
></LI
></OL
><P
>The script will make extensive use of <A
HREF="#ARRAYREF"
>arrays</A
> and <A
HREF="#FUNCTIONREF"
>functions</A
>.
You may use <A
HREF="#GRONSFELD"
>Example A-56</A
> as an
inspiration.</P
></DD
></DL
></DIV
><P
>--</P
><P
>Please do not send the author your solutions to these
exercises. There are more appropriate ways to impress him with
your cleverness, such as submitting bugfixes and suggestions
for improving the book.</P
></DIV
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="REVISIONHISTORY"
></A
>Appendix P. Revision History</H1
><TABLE
BORDER="0"
BGCOLOR="#6495ED"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SYNOPSIS"
> This document first appeared as a 60-page HOWTO in the late spring
of 2000. Since then, it has gone through quite a number of updates
and revisions. This book could not have been written without the
assistance of the Linux community, and especially of the volunteers
of the <A
HREF="http://www.tldp.org"
TARGET="_top"
>Linux Documentation Project</A
>.
</PRE
></FONT
></TD
></TR
></TABLE
><P
>Here is the e-mail to the LDP requesting permission to submit
version 0.1.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>From thegrendel@theriver.com Sat Jun 10 09:05:33 2000 -0700
Date: Sat, 10 Jun 2000 09:05:28 -0700 (MST)
From: "M. Leo Cooper" &#60;thegrendel@theriver.com&#62;
X-Sender: thegrendel@localhost
To: ldp-discuss@lists.linuxdoc.org
Subject: Permission to submit HOWTO
Dear HOWTO Coordinator,
I am working on and would like to submit to the LDP a HOWTO on the subject
of "Bash Scripting" (shell scripting, using 'bash'). As it happens,
I have been writing this document, off and on, for about the last eight
months or so, and I could produce a first draft in ASCII text format in
a matter of just a few more days.
I began writing this out of frustration at being unable to find a
decent book on shell scripting. I managed to locate some pretty good
articles on various aspects of scripting, but nothing like a complete,
beginning-to-end tutorial. Well, in keeping with my philosophy, if all
else fails, do it yourself.
As it stands, this proposed "Bash-Scripting HOWTO" would serve as a
combination tutorial and reference, with the heavier emphasis on the
tutorial. It assumes Linux experience, but only a very basic level
of programming skills. Interspersed with the text are 79 illustrative
example scripts of varying complexity, all liberally commented. There
are even exercises for the reader.
At this stage, I'm up to 18,000+ words (124k), and that's over 50 pages of
text (whew!).
I haven't mentioned that I've previously authored an LDP HOWTO, the
"Software-Building HOWTO", which I wrote in Linuxdoc/SGML. I don't know
if I could handle Docbook/SGML, and I'm glad you have volunteers to do
the conversion. You people seem to have gotten on a more organized basis
these last few months. Working with Greg Hankins and Tim Bynum was nice,
but a professional team is even nicer.
Anyhow, please advise.
Mendel Cooper
thegrendel@theriver.com</PRE
></FONT
></TD
></TR
></TABLE
></P
><DIV
CLASS="TABLE"
><A
NAME="AEN25364"
></A
><P
><B
>Table P-1. Revision History</B
></P
><TABLE
BORDER="0"
CLASS="CALSTABLE"
><THEAD
><TR
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Release</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Date</TH
><TH
ALIGN="LEFT"
VALIGN="TOP"
>Comments</TH
></TR
></THEAD
><TBODY
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
>0.1</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>14 Jun 2000</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Initial release.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>0.2</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>30 Oct 2000</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Bugs fixed, plus much additional material and more
example scripts.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>0.3</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>12 Feb 2001</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Major update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>0.4</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>08 Jul 2001</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Complete revision and expansion of the book.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>0.5</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>03 Sep 2001</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Major update: Bugfixes, material added,
sections reorganized.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>1.0</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>14 Oct 2001</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Stable release: Bugfixes, reorganization, material
added.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>1.1</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>06 Jan 2002</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Bugfixes, material and scripts added.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>1.2</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>31 Mar 2002</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>Bugfixes, material and scripts added.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>1.3</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>02 Jun 2002</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>TANGERINE release: A few bugfixes, much more material and
scripts added.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>1.4</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>16 Jun 2002</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>MANGO release: A number of typos fixed, more
material and scripts.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>1.5</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>13 Jul 2002</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>PAPAYA release: A few bugfixes, much more material and
scripts added.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>1.6</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>29 Sep 2002</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>POMEGRANATE release: Bugfixes, more material,
one more script.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>1.7</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>05 Jan 2003</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>COCONUT release: A couple of bugfixes, more material,
one more script.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>1.8</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>10 May 2003</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>BREADFRUIT release: A number of bugfixes, more scripts and
material.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>1.9</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>21 Jun 2003</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>PERSIMMON release: Bugfixes, and more material.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>2.0</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>24 Aug 2003</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>GOOSEBERRY release: Major update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>2.1</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>14 Sep 2003</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>HUCKLEBERRY release: Bugfixes, and more material.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>2.2</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>31 Oct 2003</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>CRANBERRY release: Major update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>2.3</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>03 Jan 2004</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>STRAWBERRY release: Bugfixes and more material.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>2.4</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>25 Jan 2004</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>MUSKMELON release: Bugfixes.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>2.5</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>15 Feb 2004</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>STARFRUIT release: Bugfixes and more material.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>2.6</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>15 Mar 2004</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>SALAL release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>2.7</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>18 Apr 2004</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>MULBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>2.8</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>11 Jul 2004</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>ELDERBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>3.0</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>03 Oct 2004</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>LOGANBERRY release: Major update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>3.1</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>14 Nov 2004</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>BAYBERRY release: Bugfix update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>3.2</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>06 Feb 2005</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>BLUEBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>3.3</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>20 Mar 2005</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>RASPBERRY release: Bugfixes, much material added.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>3.4</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>08 May 2005</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>TEABERRY release: Bugfixes, stylistic revisions.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>3.5</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>05 Jun 2005</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>BOXBERRY release: Bugfixes, some material added.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>3.6</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>28 Aug 2005</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>POKEBERRY release: Bugfixes, some material added.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>3.7</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>23 Oct 2005</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>WHORTLEBERRY release: Bugfixes, some material added.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>3.8</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>26 Feb 2006</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>BLAEBERRY release: Bugfixes, some material added.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>3.9</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>15 May 2006</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>SPICEBERRY release: Bugfixes, some material added.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>4.0</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>18 Jun 2006</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>WINTERBERRY release: Major reorganization.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>4.1</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>08 Oct 2006</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>WAXBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>4.2</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>10 Dec 2006</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>SPARKLEBERRY release: Important update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>4.3</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>29 Apr 2007</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>INKBERRY release: Bugfixes, material added.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>5.0</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>24 Jun 2007</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>SERVICEBERRY release: Major update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>5.1</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>10 Nov 2007</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>LINGONBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>5.2</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>16 Mar 2008</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>SILVERBERRY release: Important update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>5.3</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>11 May 2008</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>GOLDENBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>5.4</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>21 Jul 2008</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>ANGLEBERRY release: Major update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>5.5</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>23 Nov 2008</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>FARKLEBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>5.6</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>26 Jan 2009</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>WORCESTERBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>6.0</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>23 Mar 2009</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>THIMBLEBERRY release: Major update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>6.1</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>30 Sep 2009</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>BUFFALOBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>6.2</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>17 Mar 2010</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>ROWANBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>6.3</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>30 Apr 2011</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>SWOZZLEBERRY release: Major update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>6.4</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>30 Aug 2011</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>VORTEXBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>6.5</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>05 Apr 2012</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>TUNGSTENBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>6.6</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>27 Nov 2012</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>YTTERBIUMBERRY release: Minor update.</TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
><TT
CLASS="OPTION"
>10</TT
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>10 Mar 2014</TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
>YTTERBIUMBERRY release: License change.</TD
></TR
></TBODY
></TABLE
></DIV
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="MIRRORSITES"
></A
>Appendix Q. Download and Mirror Sites</H1
><P
><A
NAME="WHERE_TARBALL"
></A
></P
><P
>The latest update of this document, as an archived,
<A
HREF="#BZIPREF"
>bzip2-ed</A
>
<SPAN
CLASS="QUOTE"
>"tarball"</SPAN
> including both the SGML source
and rendered HTML, may be downloaded from the <A
HREF="http://bash.deta.in/abs-guide-latest.tar.bz2"
TARGET="_top"
>author's
home site</A
>).
A <A
HREF="http://bash.deta.in/abs-guide.pdf"
TARGET="_top"
> pdf version</A
> is also available (<A
HREF="http://www.mediafire.com/file/xi34ape1bifcnlb/abs-guide.pdf"
TARGET="_top"
>mirror
site</A
>).
There is likewise an <A
HREF="http://bash.deta.in/abs-guide.epub"
TARGET="_top"
> epub version</A
>, courtesy of Craig Barnes and Michael Satke.
The <A
HREF="http://bash.deta.in/Change.log"
TARGET="_top"
>change
log</A
> gives a detailed revision history.
The <EM
>ABS Guide</EM
> even has <A
HREF="http://freecode.com/projects/advancedbashscriptingguide/"
TARGET="_top"
> its own <TT
CLASS="FILENAME"
>freshmeat.net/freecode</TT
> page</A
>
to keep track of major updates, user comments, and popularity
ratings for the project.</P
><P
>The legacy hosting site for this document is the <A
HREF="http://www.tldp.org/LDP/abs/"
TARGET="_top"
>Linux Documentation Project</A
>,
which maintains many other Guides and HOWTOs as well.</P
><P
>Many thanks to Ronny Bangsund for donating <A
HREF="http://bash.deta.in/"
TARGET="_top"
>server space</A
> to host
this project.</P
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="TODOLIST"
></A
>Appendix R. To Do List</H1
><P
></P
><UL
><LI
><P
>A comprehensive survey of <A
HREF="#BASHCOMPAT"
>incompatibilities</A
> between
Bash and the classic <A
HREF="#BASHDEF"
>Bourne
shell</A
>.</P
></LI
><LI
><P
>Same as above, but for the Korn shell
(<I
CLASS="FIRSTTERM"
>ksh</I
>).</P
></LI
></UL
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="COPYRIGHT"
></A
>Appendix S. Copyright</H1
><P
>The <I
CLASS="CITETITLE"
>Advanced Bash Scripting
Guide</I
> is herewith granted to the PUBLIC DOMAIN.
This has the following implications and consequences.
</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>A. All previous releases of the Advanced Bash Scripting Guide
are as well granted to the Public Domain.
A1. All printed editions, whether authorized by the author or not,
are as well granted to the Public Domain. This legally overrides
any stated intention or wishes of the publishers. Any statement
of copyright is void and invalid.
THERE ARE NO EXCEPTIONS TO THIS.
A2. Any release of the Advanced Bash Scripting Guide, whether in
electronic or print form is granted to the Public Domain by the
express directive of the author and previous copyright holder, Mendel
Cooper. No other person(s) or entities have ever held a valid copyright.
B. As a Public Domain document, unlimited copying and distribution rights
are granted. There can be NO restrictions. If anyone has published or will
in the future publish an original or modified version of this document,
then only additional original material may be copyrighted. The core
work will remain in the Public Domain.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>By law, distributors and publishers (including on-line
publishers) are prohibited from imposing any conditions,
strictures, or provisions on this document, any previous versions,
or any derivative versions. The author asserts
that he has <EM
>not</EM
> entered into any contractual
obligations that would alter the foregoing declarations.</P
><P
>Essentially, you may freely distribute this book or any
derivative thereof in electronic or printed form. If you have previously
purchased or are in possession of a printed copy of a current or
previous edition, you have the LEGAL RIGHT to copy and/or redistribute
it, regardless of any copyright notice. Any copyright notice is
void.</P
><P
><EM
>Additionally, the author wishes to state his intention
that:</EM
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>If you copy or distribute this book, kindly DO NOT
use the materials within, or any portion thereof, in a patent or copyright
lawsuit against the Open Source community, its developers, its
distributors, or against any of its associated software or documentation
including, but not limited to, the Linux kernel, Open Office, Samba,
and Wine. Kindly DO NOT use any of the materials within
this book in testimony or depositions as a plaintiff's "expert witness" in
any lawsuit against the Open Source community, any of its developers, its
distributors, or any of its associated software or documentation.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>A Public Domain license essentially does not restrict ANY
legitimate distribution or use of this book. The author especially
encourages its (royalty-free!) use for classroom and instructional
purposes.</P
><P
> To date, limited print rights (Lulu edition) have been granted
to one individual and to <EM
>no one else</EM
>. Neither
that individual nor Lulu holds or ever has held a valid copyright.</P
><DIV
CLASS="WARNING"
><P
></P
><TABLE
CLASS="WARNING"
WIDTH="100%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/warning.gif"
HSPACE="5"
ALT="Warning"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>It has come to the attention of the author that
<EM
>unauthorized</EM
> electronic and print
editions of this book are being sold commercially on <SPAN
CLASS="TRADEMARK"
>itunes</SPAN
>&reg;, <EM
>amazon.com</EM
>
and elsewhere. These are illegal and pirated editions produced
without the author's permission, and readers of this book are
strongly urged not to purchase them. In fact, these pirated editions are
now legal, but necessarily fall into the Public Domain, and any
copyright notices contained within them are invalid and void.</P
></TD
></TR
></TABLE
></DIV
><P
>The author produced this book in a manner consistent with the
spirit of the <A
HREF="http://www.tldp.org/manifesto.html"
TARGET="_top"
>LDP
Manifesto</A
>.</P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN25689"
></A
><P
></P
><P
>Linux is a trademark registered to Linus Torvalds.</P
><P
>Fedora is a trademark registered to Red Hat.</P
><P
>Unix and UNIX are trademarks registered to the Open Group.</P
><P
>MS Windows is a trademark registered to the Microsoft Corp.</P
><P
>Solaris is a trademark registered to Oracle, Inc.</P
><P
>OSX is a trademark registered to Apple, Inc.</P
><P
>Yahoo is a trademark registered to Yahoo, Inc.</P
><P
>Pentium is a trademark registered to Intel, Inc.</P
><P
>Thinkpad is a trademark registered to Lenovo, Inc.</P
><P
>Scrabble is a trademark registered to Hasbro, Inc.</P
><P
>Librie, PRS-500, and PRS-505 are trademarks registered to
Sony, Inc.</P
><P
>All other commercial trademarks mentioned in the body of this work
are registered to their respective owners.</P
><P
></P
></DIV
></TD
></TR
></TABLE
><P
>Hyun Jin Cha has done a <A
HREF="http://kldp.org/HOWTO/html/Adv-Bash-Scr-HOWTO/index.html"
TARGET="_top"
>Korean
translation</A
> of version 1.0.11 of this book. Spanish,
Portuguese, <A
HREF="http://abs.traduc.org/"
TARGET="_top"
>French</A
>, German, <A
HREF="http://it.tldp.org/guide/abs/index.html"
TARGET="_top"
>Italian</A
>,
<A
HREF="http://gazette.linux.ru.net/rus/articles/index-abs-guide.html"
TARGET="_top"
>Russian</A
>,
<A
HREF="http://premekvihan.net/bash"
TARGET="_top"
>Czech</A
>, <A
HREF="http://www.linuxsir.org/bbs/showthread.php?t=256887"
TARGET="_top"
>Chinese</A
>,
Indonesian, Dutch, Romanian, Bulgarian, and Turkish translations are also
available or in progress. If you wish to translate this document
into another language, please feel free to do so, subject to
the terms stated above. The author wishes to be notified of such
efforts.</P
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN25709"
></A
><P
></P
><P
>Those generous readers desiring to
make a donation to the author may contribute a small
amount via Paypal to my e-mail address,
<TT
CLASS="EMAIL"
>&#60;<A
HREF="mailto:thegrendel.abs@gmail.com"
>thegrendel.abs@gmail.com</A
>&#62;</TT
>.
(An <TT
CLASS="USERINPUT"
><B
>Honor Roll of Supporters</B
></TT
>
is given at the beginning of the <A
HREF="http://bash.deta.in/Change.log"
TARGET="_top"
>Change Log</A
>.)
This is <EM
>not</EM
> a requirement.
The <I
CLASS="FIRSTTERM"
>ABS Guide</I
> is a free and freely
distributed document for the use and enjoyment of the Linux
community. However, in these difficult times, showing support
for voluntary projects and especially to authors of limited
means is more critically important than ever.</P
><P
></P
></DIV
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="APPENDIX"
><HR><H1
><A
NAME="ASCIITABLE"
></A
>Appendix T. ASCII Table</H1
><P
>Traditionally, a book of this sort has an <A
HREF="#ASCIIDEF"
>ASCII</A
> Table appendix.
This book does not. Instead, here are several short
scripts, each of which generates a complete ASCII table.</P
><DIV
CLASS="EXAMPLE"
><A
NAME="ASCIISH"
></A
><P
><B
>Example T-1. A script that generates an ASCII table</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ascii.sh
# ver. 0.2, reldate 26 Aug 2008
# Patched by ABS Guide author.
# Original script by Sebastian Arming.
# Used with permission (thanks!).
exec &#62;ASCII.txt # Save stdout to file,
#+ as in the example scripts
#+ reassign-stdout.sh and upperconv.sh.
MAXNUM=256
COLUMNS=5
OCT=8
OCTSQU=64
LITTLESPACE=-3
BIGSPACE=-5
i=1 # Decimal counter
o=1 # Octal counter
while [ "$i" -lt "$MAXNUM" ]; do # We don't have to count past 400 octal.
paddi=" $i"
echo -n "${paddi: $BIGSPACE} " # Column spacing.
paddo="00$o"
# echo -ne "\\${paddo: $LITTLESPACE}" # Original.
echo -ne "\\0${paddo: $LITTLESPACE}" # Fixup.
# ^
echo -n " "
if (( i % $COLUMNS == 0)); then # New line.
echo
fi
((i++, o++))
# The octal notation for 8 is 10, and 64 decimal is 100 octal.
(( i % $OCT == 0)) &#38;&#38; ((o+=2))
(( i % $OCTSQU == 0)) &#38;&#38; ((o+=20))
done
exit $?
# Compare this script with the "pr-asc.sh" example.
# This one handles "unprintable" characters.
# Exercise:
# Rewrite this script to use decimal numbers, rather than octal.</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ASCII2SH"
></A
><P
><B
>Example T-2. Another ASCII table script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Script author: Joseph Steinhauser
# Lightly edited by ABS Guide author, but not commented.
# Used in ABS Guide with permission.
#-------------------------------------------------------------------------
#-- File: ascii.sh Print ASCII chart, base 10/8/16 (JETS-2012)
#-------------------------------------------------------------------------
#-- Usage: ascii [oct|dec|hex|help|8|10|16]
#--
#-- This script prints out a summary of ASCII char codes from Zero to 127.
#-- Numeric values may be printed in Base10, Octal, or Hex.
#--
#-- Format Based on: /usr/share/lib/pub/ascii with base-10 as default.
#-- For more detail, man ascii . . .
#-------------------------------------------------------------------------
[ -n "$BASH_VERSION" ] &#38;&#38; shopt -s extglob
case "$1" in
oct|[Oo]?([Cc][Tt])|8) Obase=Octal; Numy=3o;;
hex|[Hh]?([Ee][Xx])|16|[Xx]) Obase=Hex; Numy=2X;;
help|?(-)[h?]) sed -n '2,/^[ ]*$/p' $0;exit;;
code|[Cc][Oo][Dd][Ee])sed -n '/case/,$p' $0;exit;;
*) Obase=Decimal
esac # CODE is actually shorter than the chart!
printf "\t\t## $Obase ASCII Chart ##\n\n"; FM1="|%0${Numy:-3d}"; LD=-1
AB="nul soh stx etx eot enq ack bel bs tab nl vt np cr so si dle"
AD="dc1 dc2 dc3 dc4 nak syn etb can em sub esc fs gs rs us sp"
for TOK in $AB $AD; do ABR[$((LD+=1))]=$TOK; done;
ABR[127]=del
IDX=0
while [ $IDX -le 127 ] &#38;&#38; CHR="${ABR[$IDX]}"
do ((${#CHR}))&#38;&#38; FM2='%-3s'|| FM2=`printf '\\\\%o ' $IDX`
printf "$FM1 $FM2" "$IDX" $CHR; (( (IDX+=1)%8))||echo '|'
done
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="ASCII3SH"
></A
><P
><B
>Example T-3. A third ASCII table script, using
<I
CLASS="FIRSTTERM"
>awk</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# ASCII table script, using awk.
# Author: Joseph Steinhauser
# Used in ABS Guide with permission.
#-------------------------------------------------------------------------
#-- File: ascii Print ASCII chart, base 10/8/16 (JETS-2010)
#-------------------------------------------------------------------------
#-- Usage: ascii [oct|dec|hex|help|8|10|16]
#--
#-- This script prints a summary of ASCII char codes from Zero to 127.
#-- Numeric values may be printed in Base10, Octal, or Hex (Base16).
#--
#-- Format Based on: /usr/share/lib/pub/ascii with base-10 as default.
#-- For more detail, man ascii
#-------------------------------------------------------------------------
[ -n "$BASH_VERSION" ] &#38;&#38; shopt -s extglob
case "$1" in
oct|[Oo]?([Cc][Tt])|8) Obase=Octal; Numy=3o;;
hex|[Hh]?([Ee][Xx])|16|[Xx]) Obase=Hex; Numy=2X;;
help|?(-)[h?]) sed -n '2,/^[ ]*$/p' $0;exit;;
code|[Cc][Oo][Dd][Ee])sed -n '/case/,$p' $0;exit;;
*) Obase=Decimal
esac
export Obase # CODE is actually shorter than the chart!
awk 'BEGIN{print "\n\t\t## "ENVIRON["Obase"]" ASCII Chart ##\n"
ab="soh,stx,etx,eot,enq,ack,bel,bs,tab,nl,vt,np,cr,so,si,dle,"
ad="dc1,dc2,dc3,dc4,nak,syn,etb,can,em,sub,esc,fs,gs,rs,us,sp"
split(ab ad,abr,",");abr[0]="nul";abr[127]="del";
fm1="|%0'"${Numy:- 4d}"' %-3s"
for(idx=0;idx&#60;128;idx++){fmt=fm1 (++colz%8?"":"|\n")
printf(fmt,idx,(idx in abr)?abr[idx]:sprintf("%c",idx))} }'
exit $?</PRE
></FONT
></TD
></TR
></TABLE
></DIV
></DIV
><DIV
CLASS="INDEX"
><HR><H1
><A
NAME="XREFINDEX"
></A
>Index</H1
><P
>This index / glossary / quick-reference lists many of the
important topics covered in the text. Terms are arranged
in <EM
>approximate</EM
> ASCII sorting order,
<EM
>modified as necessary</EM
> for enhanced
clarity.</P
><P
>Note that <EM
>commands</EM
> are indexed
in <A
HREF="#PART4A"
>Part 4</A
>.</P
><P
>* * *</P
><P
><B
CLASS="COMMAND"
>^ </B
> (caret)
<P
></P
><UL
><LI
><P
> <A
HREF="#BEGLINEREF"
>Beginning-of-line</A
>,
in a <A
HREF="#REGEXREF"
>Regular Expression</A
>
</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>^</B
></P
><P
><B
CLASS="COMMAND"
>^^</B
></P
><P
><A
HREF="#CASEMODPARAMSUB"
>Uppercase
conversion</A
> in <I
CLASS="FIRSTTERM"
>parameter
substitution</I
></P
></LI
></UL
>
</P
><P
><B
CLASS="COMMAND"
>~ </B
> <I
CLASS="FIRSTTERM"
>Tilde</I
>
<P
></P
><UL
><LI
><P
> <B
CLASS="COMMAND"
>~ </B
>
<A
HREF="#TILDEREF"
>home directory</A
>, corresponds
to <A
HREF="#HOMEDIRREF"
><TT
CLASS="VARNAME"
>$HOME</TT
></A
>
</P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>~/ </B
>
<A
HREF="#TILDEREF"
> <I
CLASS="FIRSTTERM"
>Current user's</I
>
home directory</A
>
</P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>~+ </B
>
<A
HREF="#WORKINGDIRREF"
><EM
>Current</EM
>
working directory</A
>
</P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>~- </B
>
<A
HREF="#PREVWORKINGDIR"
><EM
>Previous</EM
>
working directory</A
>
</P
></LI
></UL
>
</P
><P
><B
CLASS="COMMAND"
>= </B
> <I
CLASS="FIRSTTERM"
>Equals</I
> sign
<P
></P
><UL
><LI
><P
> <B
CLASS="COMMAND"
>= </B
>
<A
HREF="#EQREF"
>Variable assignment</A
> operator
</P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>= </B
>
<A
HREF="#SCOMPARISON1"
>String comparison</A
> operator
</P
><P
> <B
CLASS="COMMAND"
>== </B
>
<A
HREF="#SCOMPARISON2"
>String comparison</A
> operator
</P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>=~ </B
>
<I
CLASS="FIRSTTERM"
>Regular Expression</I
>
<A
HREF="#REGEXMATCHREF"
>match</A
> operator
</P
><P
><A
HREF="#FINDSPLIT0"
><I
CLASS="FIRSTTERM"
>Example
script</I
></A
></P
></LI
></UL
>
</P
><P
><B
CLASS="COMMAND"
>&#60; </B
> Left angle bracket
<P
></P
><UL
><LI
><P
>Is-less-than</P
><P
><A
HREF="#LTREF"
>String comparison</A
></P
><P
> <A
HREF="#INTLT"
>Integer comparison</A
> within
<A
HREF="#DBLPARENS"
>double parentheses</A
>
</P
></LI
><LI
><P
>Redirection</P
><P
> <B
CLASS="COMMAND"
>&#60; </B
>
<A
HREF="#IOREDIRECTIONREF2"
><TT
CLASS="FILENAME"
>stdin</TT
></A
>
</P
><P
> <B
CLASS="COMMAND"
>&#60;&#60; </B
>
<A
HREF="#HEREDOCRRREF"
><I
CLASS="FIRSTTERM"
>Here document</I
></A
>
</P
><P
> <B
CLASS="COMMAND"
>&#60;&#60;&#60; </B
>
<A
HREF="#HERESTRINGREF"
><I
CLASS="FIRSTTERM"
>Here string</I
></A
>
</P
><P
> <B
CLASS="COMMAND"
>&#60;&#62; </B
>
<A
HREF="#REDIRRW"
>Opening a file</A
> for
<EM
>both</EM
> reading and writing
</P
></LI
></UL
>
</P
><P
><B
CLASS="COMMAND"
>&#62; </B
> Right angle bracket
<P
></P
><UL
><LI
><P
>Is-greater-than</P
><P
><A
HREF="#GTREF"
>String comparison</A
></P
><P
><A
HREF="#INTGT"
>Integer comparison</A
>,
within <I
CLASS="FIRSTTERM"
>double parentheses</I
>
</P
></LI
><LI
><P
>Redirection</P
><P
><B
CLASS="COMMAND"
>&#62; </B
>
<A
HREF="#IOREDIRECTIONREF"
>Redirect
<TT
CLASS="FILENAME"
>stdout</TT
></A
> to a file
</P
><P
> <B
CLASS="COMMAND"
>&#62;&#62; </B
>
<A
HREF="#IOREDIRECTIONREF"
>Redirect
<TT
CLASS="FILENAME"
>stdout</TT
></A
> to a file,
but <I
CLASS="FIRSTTERM"
>append</I
>
</P
><P
><B
CLASS="COMMAND"
>i&#62;&#38;j </B
>
<A
HREF="#IOREDIRECTIONREF1"
>Redirect
<I
CLASS="FIRSTTERM"
>file descriptor</I
>
<TT
CLASS="FILENAME"
>i</TT
></A
> to
<I
CLASS="FIRSTTERM"
>file descriptor</I
>
<TT
CLASS="FILENAME"
>j</TT
></P
><P
> <B
CLASS="COMMAND"
>&#62;&#38;j </B
>
<A
HREF="#IOREDIRECTIONREF1"
>Redirect
<TT
CLASS="FILENAME"
>stdout</TT
></A
> to
<I
CLASS="FIRSTTERM"
>file descriptor</I
>
<TT
CLASS="FILENAME"
>j</TT
></P
><P
> <B
CLASS="COMMAND"
>&#62;&#38;2 </B
>
<A
HREF="#REDIROUTERROR2"
>Redirect
<TT
CLASS="FILENAME"
>stdout</TT
></A
> of a command to
<TT
CLASS="FILENAME"
>stderr</TT
></P
><P
> <B
CLASS="COMMAND"
>2&#62;&#38;1 </B
>
<A
HREF="#IOREDIRECTIONREF1"
>Redirect
<TT
CLASS="FILENAME"
>stderr</TT
></A
>
to <TT
CLASS="FILENAME"
>stdout</TT
></P
><P
> <B
CLASS="COMMAND"
>&#38;&#62; </B
>
<A
HREF="#REDIROUTERROR"
>Redirect
<EM
>both</EM
> <TT
CLASS="FILENAME"
>stdout</TT
>
and <TT
CLASS="FILENAME"
>stderr</TT
></A
> of a command to a
file</P
><P
><B
CLASS="COMMAND"
>:&#62; <TT
CLASS="FILENAME"
>file</TT
></B
>
<A
HREF="#IOREDIRECTIONREF"
>Truncate file</A
>
to zero length</P
></LI
></UL
>
</P
><P
><B
CLASS="COMMAND"
>| </B
>
<A
HREF="#PIPEREF"
>Pipe</A
>, a device for passing the output
of a command to another command or to the shell</P
><P
><B
CLASS="COMMAND"
>|| </B
>
<A
HREF="#ORREF"
>Logical OR test operator</A
></P
><P
><B
CLASS="COMMAND"
>- </B
> (dash)
<P
></P
><UL
><LI
><P
> <A
HREF="#DEFPARAM1"
>Prefix to <I
CLASS="FIRSTTERM"
>default
parameter</I
></A
>, in <I
CLASS="FIRSTTERM"
>parameter
substitution</I
></P
></LI
><LI
><P
> <A
HREF="#DASHREF"
>Prefix to <I
CLASS="FIRSTTERM"
>option
flag</I
></A
></P
></LI
><LI
><P
> <A
HREF="#DASHREF2"
>Indicating
<I
CLASS="FIRSTTERM"
>redirection</I
></A
> from
<TT
CLASS="FILENAME"
>stdin</TT
> or <TT
CLASS="FILENAME"
>stdout</TT
>
</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-- </B
> (double-dash)</P
><P
><A
HREF="#DOUBLEDASHREF"
>Prefix to
<I
CLASS="FIRSTTERM"
>long</I
> command options</A
></P
><P
> <A
HREF="#PLUSPLUSREF"
><I
CLASS="FIRSTTERM"
>C-style</I
>
variable decrement</A
> within <A
HREF="#DBLPARENSREF"
>double
parentheses</A
>
</P
></LI
></UL
>
</P
><P
><B
CLASS="COMMAND"
>; </B
> (semicolon)
<P
></P
><UL
><LI
><P
><A
HREF="#SEMICOLONREF"
>As command
separator</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>\; </B
>
<A
HREF="#FINDREF0"
><I
CLASS="FIRSTTERM"
>Escaped</I
> semicolon</A
>,
terminates a <A
HREF="#FINDREF"
>find</A
>
command</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>;; </B
>
<A
HREF="#DOUBLESEMICOLON"
>Double-semicolon</A
>, terminator in a
<A
HREF="#CASEESAC1"
>case</A
> option</P
><P
>Required when ...</P
><P
><A
HREF="#NEEDSEMICOLON"
><I
CLASS="FIRSTTERM"
>do</I
>
keyword is on the first line of
<I
CLASS="FIRSTTERM"
>loop</I
></A
></P
><P
><A
HREF="#OMITSEMICOLON"
> terminating <I
CLASS="FIRSTTERM"
>curly-bracketed</I
>
code block</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>;;&#38;</B
> <B
CLASS="COMMAND"
>;&#38;</B
>
<A
HREF="#NCTERM"
>Terminators</A
> in
a <I
CLASS="FIRSTTERM"
>case</I
> option (<A
HREF="#BASH4REF"
>version 4+</A
> of Bash).</P
></LI
></UL
></P
><P
><B
CLASS="COMMAND"
>:</B
> Colon
<P
></P
><UL
><LI
><P
> <B
CLASS="COMMAND"
>:&#62; <TT
CLASS="FILENAME"
>filename</TT
></B
>
<A
HREF="#IOREDIRECTIONREF"
>Truncate file</A
>
to zero length</P
></LI
><LI
><P
><A
HREF="#NULLREF"
><I
CLASS="FIRSTTERM"
>null</I
>
command</A
>, equivalent to the
<A
HREF="#TRUEREF"
>true</A
> Bash
builtin</P
></LI
><LI
><P
>Used in an <A
HREF="#ANONHEREDOC0"
>anonymous
here document</A
></P
></LI
><LI
><P
>Used in an <A
HREF="#COLONINFUNCTION"
>otherwise
empty function</A
></P
></LI
><LI
><P
>Used as a <A
HREF="#FSTRANGEREF"
>function
name</A
></P
></LI
></UL
></P
><P
><B
CLASS="COMMAND"
>! </B
>
<A
HREF="#NOTREF"
>Negation operator</A
>, inverts <A
HREF="#NEGCOND"
>exit status</A
> of a test or command
<P
></P
><UL
><LI
><P
><B
CLASS="COMMAND"
>!= </B
>
<A
HREF="#NOTEQUAL"
>not-equal-to</A
>
String comparison operator</P
></LI
></UL
></P
><P
><B
CLASS="COMMAND"
>? </B
> (question mark)
<P
></P
><UL
><LI
><P
> <A
HREF="#QUEXREGEX"
>Match zero or one characters</A
>,
in an <A
HREF="#EXTREGEX"
> Extended Regular
Expression</A
></P
></LI
><LI
><P
><A
HREF="#QUEXWC"
>Single-character
<I
CLASS="FIRSTTERM"
>wild card</I
></A
>,
in <A
HREF="#GLOBBINGREF"
>globbing</A
></P
></LI
><LI
><P
>In a
<A
HREF="#CSTRINARY"
><I
CLASS="FIRSTTERM"
>C</I
>-style
Trinary operator</A
></P
></LI
></UL
></P
><P
><B
CLASS="COMMAND"
>// </B
>
<A
HREF="#DOUBLESLASHREF"
>Double forward slash</A
>,
behavior of <A
HREF="#CDREF"
>cd</A
> command toward</P
><P
><B
CLASS="COMMAND"
>. </B
> (dot / period)
<P
></P
><UL
><LI
><P
> <B
CLASS="COMMAND"
>. </B
>
<A
HREF="#DOTREF"
>Load a file</A
> (into a script),
equivalent to <A
HREF="#SOURCEREF"
>source</A
> command
</P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>. </B
>
<A
HREF="#REGEXDOT"
>Match single character</A
>,
in a <A
HREF="#REGEXREF"
>Regular Expression</A
>
</P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>. </B
>
<A
HREF="#DOTDIRECTORY"
>Current working
directory</A
>
</P
><P
> <B
CLASS="COMMAND"
>./ </B
>
<A
HREF="#CURRENTWDREF"
>Current working
directory</A
>
</P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>.. </B
>
<A
HREF="#DOTDIRECTORY"
><I
CLASS="FIRSTTERM"
>Parent</I
>
directory</A
>
</P
></LI
></UL
>
</P
><P
><B
CLASS="COMMAND"
>' ... '</B
>
(single quotes)
<A
HREF="#SNGLQUO"
><I
CLASS="FIRSTTERM"
>strong</I
>
quoting</A
></P
><P
><B
CLASS="COMMAND"
>" ... " </B
>
(double quotes)
<A
HREF="#DBLQUO"
><I
CLASS="FIRSTTERM"
>weak</I
>
quoting</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#QUOTINGBSL"
><I
CLASS="FIRSTTERM"
>Double-quoting</I
>
the <I
CLASS="FIRSTTERM"
>backslash</I
> (<B
CLASS="COMMAND"
>\</B
>)
character</A
></P
></LI
></UL
>
</P
><P
><B
CLASS="COMMAND"
>,</B
>
<P
></P
><UL
><LI
><P
> <A
HREF="#COMMAOP"
>Comma operator</A
>
</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>,</B
></P
><P
><B
CLASS="COMMAND"
>,,</B
></P
><P
><A
HREF="#CASEMODPARAMSUB"
>Lowercase
conversion</A
> in <I
CLASS="FIRSTTERM"
>parameter
substitution</I
></P
></LI
></UL
>
</P
><P
><B
CLASS="COMMAND"
>() </B
> Parentheses
<P
></P
><UL
><LI
><P
><B
CLASS="COMMAND"
>( ... ) </B
>
<A
HREF="#PARENSREF"
>Command group</A
>;
starts a
<A
HREF="#SUBSHELLSREF"
>subshell</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>( ... ) </B
>
<A
HREF="#PARENGRPS"
>Enclose group</A
>
of <I
CLASS="FIRSTTERM"
>Extended Regular
Expressions</I
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>&#62;( ... )</B
></P
><P
><B
CLASS="COMMAND"
>&#60;( ... ) </B
>
<A
HREF="#PROCESSSUBREF"
>Process substitution</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
> ... ) </B
>
<A
HREF="#CASEPAREN"
>Terminates test-condition</A
>
in <I
CLASS="FIRSTTERM"
>case</I
> construct</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>(( ... )) </B
>
<A
HREF="#DBLPARENSREF"
>Double parentheses</A
>,
in arithmetic expansion</P
></LI
></UL
></P
><P
><B
CLASS="COMMAND"
>[ </B
> <A
HREF="#LEFTBRACKET"
>Left bracket</A
>,
<I
CLASS="FIRSTTERM"
>test</I
> construct</P
><P
><B
CLASS="COMMAND"
>[ ]</B
>Brackets
<P
></P
><UL
><LI
><P
><A
HREF="#BRACKARRAY"
><I
CLASS="FIRSTTERM"
>Array</I
>
element</A
></P
></LI
><LI
><P
><A
HREF="#BRACKETSREF"
>Enclose character set
to match</A
> in a <I
CLASS="FIRSTTERM"
>Regular
Expression</I
></P
></LI
><LI
><P
><A
HREF="#BRACKTEST"
><I
CLASS="FIRSTTERM"
>Test</I
> construct</A
></P
></LI
></UL
></P
><P
><B
CLASS="COMMAND"
>[[ ... ]]</B
>
<A
HREF="#DBLBRACKETS"
>Double brackets</A
>,
extended <I
CLASS="FIRSTTERM"
>test</I
> construct</P
><P
><B
CLASS="COMMAND"
>$ </B
>
<A
HREF="#DOLLARSIGNREF"
><I
CLASS="FIRSTTERM"
>Anchor</I
></A
>,
in a <A
HREF="#REGEXREF"
>Regular Expression</A
></P
><P
><B
CLASS="COMMAND"
>$ </B
>
<A
HREF="#VARSUBN"
>Prefix to a variable name</A
></P
><P
><B
CLASS="COMMAND"
>$( ... ) </B
>
<A
HREF="#COMMANDSUBREF0"
>Command substitution</A
>,
setting a variable with output of a command,
using parentheses notation</P
><P
><B
CLASS="COMMAND"
>` ... ` </B
>
<A
HREF="#BACKQUOTESREF"
>Command substitution</A
>,
using <A
HREF="#BACKTICKSREF"
>backquotes</A
>
notation</P
><P
><B
CLASS="COMMAND"
>$[ ... ]</B
>
<A
HREF="#BRACKETARITH"
>Integer expansion</A
>
(deprecated)</P
><P
><B
CLASS="COMMAND"
>${ ... }</B
> Variable manipulation / evaluation
<P
></P
><UL
><LI
><P
><B
CLASS="COMMAND"
>${var}</B
>
<A
HREF="#PSSUB1"
>Value of a variable</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>${#var}</B
>
<A
HREF="#PSOREX1"
>Length of a variable</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>${#@}</B
></P
><P
><B
CLASS="COMMAND"
>${#*}</B
>
<A
HREF="#NUMPOSPARAM"
>Number of
<I
CLASS="FIRSTTERM"
>positional
parameters</I
></A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>${parameter?err_msg}</B
>
<A
HREF="#QERRMSG"
>Parameter-unset
message</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>${parameter-default}</B
></P
><P
><B
CLASS="COMMAND"
>${parameter:-default}</B
></P
><P
><B
CLASS="COMMAND"
>${parameter=default}</B
></P
><P
><B
CLASS="COMMAND"
>${parameter:=default}</B
>
<A
HREF="#DEFPARAM1"
>Set default
parameter</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>${parameter+alt_value}</B
></P
><P
><B
CLASS="COMMAND"
>${parameter:+alt_value}</B
></P
><P
><A
HREF="#PARAMALTV"
>Alternate value</A
>
of parameter, if set</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>${!var}</B
></P
><P
><A
HREF="#IVR2"
>Indirect referencing of a variable</A
>,
new notation</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>${!#}</B
></P
><P
><A
HREF="#LASTARGREF"
>Final <I
CLASS="FIRSTTERM"
>positional
parameter</I
></A
>.
(This is an <I
CLASS="FIRSTTERM"
>indirect reference</I
> to
<A
HREF="#CLACOUNTREF"
>$#</A
>.)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>${!varprefix*}</B
></P
><P
><B
CLASS="COMMAND"
>${!varprefix@}</B
></P
><P
><A
HREF="#VARPREFIXM"
>Match
<EM
>names</EM
></A
> of all previously declared
variables beginning with <TT
CLASS="VARNAME"
>varprefix</TT
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>${string:position}</B
></P
><P
><B
CLASS="COMMAND"
>${string:position:length}</B
>
<A
HREF="#SUBSTREXTR01"
>Substring
extraction</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>${var#Pattern}</B
></P
><P
><B
CLASS="COMMAND"
>${var##Pattern}</B
>
<A
HREF="#PSOREX2"
>Substring
removal</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>${var%Pattern}</B
></P
><P
><B
CLASS="COMMAND"
>${var%%Pattern}</B
>
<A
HREF="#PCTPATREF"
>Substring
removal</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>${string/substring/replacement}</B
></P
><P
><B
CLASS="COMMAND"
>${string//substring/replacement}</B
></P
><P
><B
CLASS="COMMAND"
>${string/#substring/replacement}</B
></P
><P
><B
CLASS="COMMAND"
>${string/%substring/replacement}</B
>
<A
HREF="#SUBSTRREPL00"
>Substring
replacement</A
></P
></LI
></UL
></P
><P
><B
CLASS="COMMAND"
>$' ... '</B
>
<A
HREF="#STRQ"
>String expansion</A
>,
using <I
CLASS="FIRSTTERM"
>escaped</I
> characters.</P
><P
><B
CLASS="COMMAND"
>\ </B
>
<A
HREF="#ESCP"
>Escape</A
> the character following
<P
></P
><UL
><LI
><P
><B
CLASS="COMMAND"
>\&#60; ... \&#62; </B
>
<A
HREF="#ANGLEBRAC"
>Angle brackets</A
>,
<I
CLASS="FIRSTTERM"
>escaped</I
>,
word boundary in a <A
HREF="#REGEXREF"
>Regular
Expression</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>\{ N \}</B
>
<A
HREF="#ESCPCB"
><SPAN
CLASS="QUOTE"
>"Curly"</SPAN
> brackets</A
>,
<I
CLASS="FIRSTTERM"
>escaped</I
>,
number of character sets to match in an <A
HREF="#EXTREGEX"
>Extended RE</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>\; </B
>
<A
HREF="#FINDREF0"
><I
CLASS="FIRSTTERM"
>Semicolon</I
></A
>,
<I
CLASS="FIRSTTERM"
>escaped</I
>,
terminates a <A
HREF="#FINDREF"
>find</A
>
command</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>\$$ </B
> <A
HREF="#IVRREF"
>Indirect
reverencing of a variable</A
>, old-style notation</P
></LI
><LI
><P
><A
HREF="#ESCNEWLINE"
>Escaping
a <I
CLASS="FIRSTTERM"
>newline</I
></A
>,
to write a multi-line command</P
></LI
></UL
></P
><P
><B
CLASS="COMMAND"
>&#38; </B
>
<P
></P
><UL
><LI
><P
> <B
CLASS="COMMAND"
>&#38;&#62; </B
>
<A
HREF="#REDIROUTERROR"
>Redirect
<EM
>both</EM
> <TT
CLASS="FILENAME"
>stdout</TT
>
and <TT
CLASS="FILENAME"
>stderr</TT
></A
> of a command to a
file</P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>&#62;&#38;j </B
>
<A
HREF="#IOREDIRECTIONREF1"
>Redirect
<TT
CLASS="FILENAME"
>stdout</TT
></A
> to
<I
CLASS="FIRSTTERM"
>file descriptor</I
>
<I
CLASS="FIRSTTERM"
>j</I
></P
><P
> <B
CLASS="COMMAND"
>&#62;&#38;2 </B
>
<A
HREF="#REDIROUTERROR2"
>Redirect
<TT
CLASS="FILENAME"
>stdout</TT
></A
> of a command to
<TT
CLASS="FILENAME"
>stderr</TT
></P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>i&#62;&#38;j </B
>
<A
HREF="#IOREDIRECTIONREF1"
>Redirect
<I
CLASS="FIRSTTERM"
>file descriptor</I
></A
>
<I
CLASS="FIRSTTERM"
>i</I
> to
<I
CLASS="FIRSTTERM"
>file descriptor</I
>
<I
CLASS="FIRSTTERM"
>j</I
></P
><P
> <B
CLASS="COMMAND"
>2&#62;&#38;1 </B
>
<A
HREF="#IOREDIRECTIONREF1"
>Redirect
<TT
CLASS="FILENAME"
>stderr</TT
></A
> to
<TT
CLASS="FILENAME"
>stdout</TT
></P
></LI
><LI
><P
><A
HREF="#CFD"
>Closing <I
CLASS="FIRSTTERM"
>file
descriptors</I
></A
></P
><P
><B
CLASS="COMMAND"
>n&#60;&#38;-</B
>
Close input file descriptor
<I
CLASS="FIRSTTERM"
>n</I
></P
><P
><B
CLASS="COMMAND"
>0&#60;&#38;-</B
>,
<B
CLASS="COMMAND"
>&#60;&#38;-</B
>
Close <TT
CLASS="FILENAME"
>stdin</TT
></P
><P
><B
CLASS="COMMAND"
>n&#62;&#38;-</B
>
Close output file descriptor
<I
CLASS="FIRSTTERM"
>n</I
></P
><P
><B
CLASS="COMMAND"
>1&#62;&#38;-</B
>,
<B
CLASS="COMMAND"
>&#62;&#38;-</B
>
Close <TT
CLASS="FILENAME"
>stdout</TT
></P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>&#38;&#38; </B
>
<A
HREF="#LOGICALAND"
>Logical AND
test operator</A
></P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>Command &#38;</B
>
<A
HREF="#BGJOB"
>Run job in
<I
CLASS="FIRSTTERM"
>background</I
></A
></P
></LI
></UL
></P
><P
><B
CLASS="COMMAND"
># </B
>
<A
HREF="#HASHMARKREF"
>Hashmark</A
>,
special symbol beginning a script
<I
CLASS="FIRSTTERM"
>comment</I
></P
><P
><B
CLASS="COMMAND"
>#!</B
>
<A
HREF="#SHABANGREF"
>Sha-bang</A
>,
special string starting a
<A
HREF="#WHATSASCRIPT"
>shell script</A
></P
><P
><B
CLASS="COMMAND"
>* </B
> Asterisk
<P
></P
><UL
><LI
><P
> <A
HREF="#ASTERISKREF"
><I
CLASS="FIRSTTERM"
>Wild
card</I
></A
>,
in <A
HREF="#GLOBBINGREF"
>globbing</A
>
</P
></LI
><LI
><P
> <A
HREF="#ASTERISKREF2"
>Any number of characters</A
>
in a <A
HREF="#REGEXREF"
>Regular Expression</A
>
</P
></LI
><LI
><P
> <B
CLASS="COMMAND"
> ** </B
>
<A
HREF="#EXPONENTIATIONREF"
>Exponentiation</A
>,
arithmetic operator
</P
></LI
><LI
><P
> <B
CLASS="COMMAND"
> ** </B
>
Extended <I
CLASS="FIRSTTERM"
>globbing</I
>
<A
HREF="#GLOBSTARREF"
>file-match operator</A
>
</P
></LI
></UL
>
</P
><P
><B
CLASS="COMMAND"
>% </B
> Percent sign
<P
></P
><UL
><LI
><P
> <A
HREF="#MODULOREF"
>Modulo</A
>, division-remainder
arithmetic operation
</P
></LI
><LI
><P
> <A
HREF="#PCTPATREF"
>Substring removal</A
>
(pattern matching) operator</P
></LI
></UL
>
</P
><P
><B
CLASS="COMMAND"
>+ </B
> Plus sign
<P
></P
><UL
><LI
><P
><A
HREF="#PLUSREF"
><I
CLASS="FIRSTTERM"
>Character
match</I
></A
>, in an <A
HREF="#EXTREGEX"
>extended Regular
Expression</A
></P
></LI
><LI
><P
><A
HREF="#PARAMALTV"
>Prefix to
<I
CLASS="FIRSTTERM"
>alternate parameter</I
></A
>, in
<I
CLASS="FIRSTTERM"
>parameter substitution</I
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
> ++ </B
>
<A
HREF="#PLUSPLUSREF"
><I
CLASS="FIRSTTERM"
>C-style</I
>
variable increment</A
>, within
<A
HREF="#DBLPARENSREF"
>double
parentheses</A
></P
></LI
></UL
></P
><P
>* * *</P
><P
><I
CLASS="FIRSTTERM"
>Shell Variables</I
></P
><P
><B
CLASS="COMMAND"
>$_ </B
>
<A
HREF="#UNDERSCOREREF"
>Last argument to previous
command</A
></P
><P
><B
CLASS="COMMAND"
>$- </B
>
<A
HREF="#FLPREF"
>Flags passed to script</A
>, using <A
HREF="#SETREF"
>set</A
></P
><P
><B
CLASS="COMMAND"
>$! </B
>
<A
HREF="#PIDVARREF"
><I
CLASS="FIRSTTERM"
>Process ID</I
>
of last background job</A
></P
><P
><B
CLASS="COMMAND"
>$? </B
>
<A
HREF="#EXSREF"
><I
CLASS="FIRSTTERM"
>Exit
status</I
> of a command</A
></P
><P
><B
CLASS="COMMAND"
>$@ </B
>
All the <I
CLASS="FIRSTTERM"
>positional parameters</I
>, <A
HREF="#APPREF2"
>as <EM
>separate</EM
>
words</A
></P
><P
><B
CLASS="COMMAND"
> $* </B
>
All the <I
CLASS="FIRSTTERM"
>positional parameters</I
>, <A
HREF="#APPREF"
>as a <EM
>single</EM
>
word</A
></P
><P
><B
CLASS="COMMAND"
>$$ </B
>
<A
HREF="#PROCESSIDREF"
>Process ID</A
> of the
script</P
><P
><B
CLASS="COMMAND"
>$# </B
>
<A
HREF="#CLACOUNTREF"
>Number of arguments passed</A
> to a
<A
HREF="#FUNCTIONREF"
>function</A
>, or to the script
itself</P
><P
><B
CLASS="COMMAND"
>$0 </B
>
<A
HREF="#SCRNAMEPARAM"
>Filename of the script</A
></P
><P
><B
CLASS="COMMAND"
>$1 </B
>
<A
HREF="#POSPARAMREF1"
>First argument passed to script</A
></P
><P
><B
CLASS="COMMAND"
>$9 </B
>
<A
HREF="#POSPARAMREF1"
>Ninth argument passed to script</A
></P
><P
><A
HREF="#SPECSHVARTAB"
><B
CLASS="COMMAND"
>Table</B
></A
>
of <I
CLASS="FIRSTTERM"
>shell variables</I
></P
><P
>* * * * * *</P
><P
><B
CLASS="COMMAND"
>-a </B
>
<A
HREF="#COMPOUNDAND"
>Logical AND</A
>
compound comparison test</P
><P
>Address database, <A
HREF="#EX30"
>script
example</A
></P
><P
><I
CLASS="FIRSTTERM"
>Advanced Bash Scripting Guide</I
>,
<A
HREF="#WHERE_TARBALL"
>where to download</A
></P
><P
><A
HREF="#ALIASREF"
>Alias</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#UNALIASREF"
>Removing an
<I
CLASS="FIRSTTERM"
>alias</I
></A
>, using
<I
CLASS="FIRSTTERM"
>unalias</I
></P
></LI
></UL
></P
><P
><A
HREF="#AGRAM2"
>Anagramming</A
></P
><P
><A
HREF="#LCONS1"
><I
CLASS="FIRSTTERM"
>And</I
> list</A
>
<P
></P
><UL
><LI
><P
> <A
HREF="#ANDDEFAULT"
>To supply default command-line
argument</A
></P
></LI
></UL
></P
><P
><A
HREF="#LOGOPS1"
><I
CLASS="FIRSTTERM"
>And</I
> logical
operator</A
> <B
CLASS="COMMAND"
>&#38;&#38;</B
></P
><P
><A
HREF="#ANGLEBRAC"
>Angle brackets</A
>,
<I
CLASS="FIRSTTERM"
>escaped</I
>,
<B
CLASS="COMMAND"
>\&#60; . . . \&#62; </B
>
word boundary in a <A
HREF="#REGEXREF"
>Regular
Expression</A
></P
><P
><A
HREF="#ANONHEREDOC0"
>Anonymous
<I
CLASS="FIRSTTERM"
>here document</I
></A
>, using
<B
CLASS="COMMAND"
>:</B
></P
><P
><A
HREF="#FAARCHIVING1"
>Archiving</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#RPMREF"
>rpm</A
></P
></LI
><LI
><P
><A
HREF="#TARREF"
>tar</A
></P
></LI
></UL
></P
><P
><A
HREF="#ARITHEXPREF"
>Arithmetic expansion</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#ARXS"
> <I
CLASS="FIRSTTERM"
>exit status</I
> of</A
></P
></LI
><LI
><P
><A
HREF="#ARITHEXPVAR1"
> variations of</A
></P
></LI
></UL
></P
><P
><A
HREF="#AROPS1"
>Arithmetic operators</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#ARITHOPSCOMB"
> combination operators</A
>, <I
CLASS="FIRSTTERM"
>C</I
>-style</P
><P
><B
CLASS="COMMAND"
> += </B
>
<B
CLASS="COMMAND"
> -= </B
>
<B
CLASS="COMMAND"
> *= </B
>
<B
CLASS="COMMAND"
> /= </B
>
<B
CLASS="COMMAND"
> %= </B
>
</P
><DIV
CLASS="NOTE"
><P
></P
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="../images/note.gif"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
HREF="#PLUSEQSTR"
>In certain contexts</A
>,
<B
CLASS="COMMAND"
>+=</B
> can also function as a <I
CLASS="FIRSTTERM"
>string
concatenation</I
> operator.</P
></TD
></TR
></TABLE
></DIV
></LI
></UL
></P
><P
><A
HREF="#ARRAYREF"
>Arrays</A
></P
><P
></P
><UL
><LI
><P
><A
HREF="#ASSOCARR"
>Associative arrays</A
>
</P
><P
><A
HREF="#ASSOCARRTST"
>more efficient</A
>
than conventional arrays</P
></LI
><LI
><P
><A
HREF="#ARRAYREF"
>Bracket notation</A
>
</P
></LI
><LI
><P
><A
HREF="#ARRAYAPPEND0"
>Concatenating</A
>,
<I
CLASS="FIRSTTERM"
>example script</I
></P
></LI
><LI
><P
><A
HREF="#COPYARRAY0"
>Copying</A
>
</P
></LI
><LI
><P
><A
HREF="#ARRAYDECLARE"
>Declaring</A
></P
><P
><TT
CLASS="VARNAME"
>declare -a
array_name</TT
></P
></LI
><LI
><P
><A
HREF="#ARRAYINDIR"
>Embedded
arrays</A
></P
></LI
><LI
><P
><A
HREF="#EMPTYARRAY0"
>Empty
arrays, empty elements</A
>, <I
CLASS="FIRSTTERM"
>example
script</I
></P
></LI
><LI
><P
><A
HREF="#ARRAYINDIR"
>Indirect references</A
>
</P
></LI
><LI
><P
><A
HREF="#ARRAYINIT0"
>Initialization</A
></P
><P
><TT
CLASS="VARNAME"
>array=( element1 element2 ... elementN)</TT
></P
><P
><A
HREF="#ARRAYASSIGN0"
><I
CLASS="FIRSTTERM"
>Example
script</I
></A
></P
><P
>Using <A
HREF="#ARRAYINITCS"
>command
substitution</A
></P
></LI
><LI
><P
><A
HREF="#ARRAYINITCS"
>Loading a
file</A
> into an array</P
></LI
><LI
><P
><A
HREF="#ARRAYMULTIDIM"
>Multidimensional</A
>,
simulating</P
></LI
><LI
><P
><A
HREF="#ARRAYNEST"
>Nesting and
embedding</A
></P
></LI
><LI
><P
><A
HREF="#ARRAYNOTATION"
>Notation
and usage</A
></P
></LI
><LI
><P
><A
HREF="#ARRAYNUMELEMENTS"
>Number of elements in</A
></P
><P
><TT
CLASS="VARNAME"
>${#array_name[@]}</TT
></P
><P
><TT
CLASS="VARNAME"
>${#array_name[*]}</TT
></P
></LI
><LI
><P
><A
HREF="#ARRAYSYNTAX"
>Operations</A
></P
></LI
><LI
><P
><A
HREF="#PASSARRAY"
>Passing
an <I
CLASS="FIRSTTERM"
>array</I
></A
> to a
function</P
></LI
><LI
><P
>As <A
HREF="#RETARRAY"
> <I
CLASS="FIRSTTERM"
>return value</I
> from
a function</A
></P
></LI
><LI
><P
>Special properties,
<A
HREF="#ARRAYSPECIALPROPS"
>example
script</A
></P
></LI
><LI
><P
>String operations,
<A
HREF="#ARRAYSTRINGOPS"
>example
script</A
></P
></LI
><LI
><P
><A
HREF="#ARRAYUNSET"
><I
CLASS="FIRSTTERM"
>unset</I
> deletes array
elements</A
></P
></LI
></UL
><P
><A
HREF="#READARROW"
>Arrow keys</A
>, detecting</P
><P
>ASCII
<P
></P
><UL
><LI
><P
><A
HREF="#ASCIIDEF"
>Definition</A
></P
></LI
><LI
><P
><A
HREF="#ASCIITABLE"
> Scripts for generating ASCII table</A
></P
></LI
></UL
>
</P
><P
><A
HREF="#AWK"
>awk</A
> field-oriented text
processing language
<P
></P
><UL
><LI
><P
><A
HREF="#AWKRANDOMREF"
> <TT
CLASS="VARNAME"
>rand()</TT
></A
>,
random function</P
></LI
><LI
><P
><A
HREF="#AWKSTRINGMANIP2"
>String
manipulation</A
></P
></LI
><LI
><P
><A
HREF="#EXPORTAWK"
>Using
<I
CLASS="FIRSTTERM"
>export</I
></A
> to pass a variable to an
embedded <I
CLASS="FIRSTTERM"
>awk</I
> script</P
></LI
></UL
></P
><P
>* * *</P
><P
>Backlight, <A
HREF="#BACKLIGHT"
>setting the
brightness</A
></P
><P
><A
HREF="#BACKTICKSREF"
>Backquotes</A
>,
used in <A
HREF="#BACKQUOTESREF"
>command
substitution</A
></P
><P
><A
HREF="#BASE0"
>Base conversion</A
>,
<I
CLASS="FIRSTTERM"
>example script</I
></P
><P
><A
HREF="#BASHDEF"
>Bash</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#BASH3GOTCHA"
>Bad scripting
practices</A
></P
></LI
><LI
><P
><A
HREF="#BASICSREV0"
>Basics reviewed</A
>,
<I
CLASS="FIRSTTERM"
>script example</I
></P
></LI
><LI
><P
><A
HREF="#CLOPTS"
>Command-line
options</A
></P
><P
><A
HREF="#OPTIONSTABLE"
><B
CLASS="COMMAND"
>Table</B
></A
></P
></LI
><LI
><P
> <A
HREF="#BASHCOMPAT"
>Features that classic
<I
CLASS="FIRSTTERM"
>Bourne</I
> shell lacks</A
></P
></LI
><LI
><P
><A
HREF="#INTERNALVARIABLES"
>Internal variables</A
></P
></LI
><LI
><P
><A
HREF="#BASH2REF"
> Version 2</A
></P
></LI
><LI
><P
><A
HREF="#BASH3REF"
> Version 3</A
></P
></LI
><LI
><P
><A
HREF="#BASH4REF"
>Version 4</A
></P
><P
><A
HREF="#BASH41"
>Version 4.1</A
></P
><P
><A
HREF="#BASH42"
>Version 4.2</A
></P
></LI
></UL
></P
><P
><A
HREF="#SAMPLE-BASHRC"
>.bashrc</A
></P
><P
><A
HREF="#BASHSUBSHELLREF"
><TT
CLASS="VARNAME"
>$BASH_SUBSHELL</TT
></A
></P
><P
><A
HREF="#BASICCOMMANDS1"
>Basic commands</A
>, external</P
><P
><A
HREF="#DOSBATCH1"
>Batch files</A
>,
<I
CLASS="FIRSTTERM"
>DOS</I
></P
><P
><A
HREF="#BATCHPROCREF"
>Batch processing</A
></P
><P
><A
HREF="#BCREF"
>bc</A
>, calculator utility
<P
></P
><UL
><LI
><P
><A
HREF="#BCHEREDOC"
>In a <I
CLASS="FIRSTTERM"
>here
document</I
></A
></P
></LI
><LI
><P
><A
HREF="#BCTEMPLATE"
>Template</A
>
for calculating a script variable</P
></LI
></UL
></P
><P
><A
HREF="#BIBLIO"
>Bibliography</A
></P
><P
><A
HREF="#BISONREF"
>Bison</A
> utility</P
><P
><A
HREF="#BITWSOPS1"
>Bitwise operators</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#BASE64"
>Example script</A
>
</P
></LI
></UL
>
</P
><P
><A
HREF="#BLOCKDEVREF"
>Block devices</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#BLOCKDEVTEST"
>testing
for</A
></P
></LI
></UL
>
</P
><P
><A
HREF="#CODEBLOCKREF"
>Blocks of code</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#NODODONE"
>Iterating / looping</A
></P
></LI
><LI
><P
><A
HREF="#BLOCKIO"
>Redirection</A
></P
><P
><I
CLASS="FIRSTTERM"
>Script example</I
>:
<A
HREF="#BLOCKIO2"
>Redirecting output of a a code
block</A
></P
></LI
></UL
></P
><P
><A
HREF="#BFS"
>Bootable flash drives</A
>, creating</P
><P
><A
HREF="#BRACEEXPREF"
>Brace expansion</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#BRACEEXPREF33"
>Extended</A
>,
<TT
CLASS="REPLACEABLE"
><I
>{a..z}</I
></TT
></P
></LI
><LI
><P
><A
HREF="#BRACEEXPREF3"
>Parameterizing</A
>
</P
></LI
><LI
><P
>With <A
HREF="#BRACEEXPREF4"
>increment and
zero-padding</A
> (new feature in Bash, <A
HREF="#BASH4REF"
>version
4</A
>)</P
></LI
></UL
></P
><P
>Brackets, <B
CLASS="COMMAND"
>[ ]</B
>
<P
></P
><UL
><LI
><P
><A
HREF="#BRACKARRAY"
><I
CLASS="FIRSTTERM"
>Array</I
>
element</A
></P
></LI
><LI
><P
><A
HREF="#BRACKETSREF"
>Enclose character set
to match</A
> in a <I
CLASS="FIRSTTERM"
>Regular
Expression</I
></P
></LI
><LI
><P
><A
HREF="#BRACKTEST"
><I
CLASS="FIRSTTERM"
>Test</I
> construct</A
></P
></LI
></UL
></P
><P
>Brackets, <I
CLASS="FIRSTTERM"
>curly</I
>, <B
CLASS="COMMAND"
>{}</B
>,
used in</P
><P
></P
><UL
><LI
><P
><A
HREF="#CODEBLOCKREF"
>Code
block</A
></P
></LI
><LI
><P
><A
HREF="#CURLYBRACKETSREF"
> <I
CLASS="FIRSTTERM"
>find</I
></A
></P
></LI
><LI
><P
><A
HREF="#ESCPCB"
><I
CLASS="FIRSTTERM"
>Extended Regular
Expressions</I
></A
></P
></LI
><LI
><P
><A
HREF="#BRACKETNOTATION"
><I
CLASS="FIRSTTERM"
>Positional
parameters</I
></A
></P
></LI
><LI
><P
><A
HREF="#XARGSCURLYREF"
><I
CLASS="FIRSTTERM"
>xargs</I
></A
></P
></LI
></UL
><P
><A
HREF="#BRKCONT1"
>break</A
>
<I
CLASS="FIRSTTERM"
>loop</I
> control command
<P
></P
><UL
><LI
><P
><A
HREF="#BREAKPARAM"
>Parameter</A
>
(optional)</P
></LI
></UL
></P
><P
><A
HREF="#BUILTINREF"
>Builtins</A
> in
<I
CLASS="FIRSTTERM"
>Bash</I
>
<P
></P
><UL
><LI
><P
><A
HREF="#BLTINFRK"
>Do not fork a subprocess</A
>
</P
></LI
></UL
></P
><P
>* * *</P
><P
><A
HREF="#CASEESAC1"
><I
CLASS="FIRSTTERM"
>case</I
>
construct</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#CASECL"
>Command-line parameters</A
>,
handling</P
></LI
><LI
><P
><A
HREF="#CSGLOB"
>Globbing</A
>,
filtering strings with</P
></LI
></UL
></P
><P
><A
HREF="#CATREF"
>cat</A
>,
con<I
CLASS="FIRSTTERM"
>cat</I
>entate file(s)
<P
></P
><UL
><LI
><P
><A
HREF="#CATABUSE"
>Abuse of</A
></P
></LI
><LI
><P
><A
HREF="#CATSCRIPTREF"
><I
CLASS="FIRSTTERM"
>cat</I
>
scripts</A
></P
></LI
><LI
><P
><A
HREF="#CATLESSEFF"
>Less efficient than
redirecting <TT
CLASS="FILENAME"
>stdin</TT
></A
></P
></LI
><LI
><P
><A
HREF="#READPIPEREF"
>Piping the output
of</A
>, to a <A
HREF="#READREF"
>read</A
></P
></LI
><LI
><P
><A
HREF="#CATUSES"
>Uses of</A
></P
></LI
></UL
></P
><P
><A
HREF="#CHARDEVREF"
>Character devices</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#CHARDEVTEST"
>testing
for</A
></P
></LI
></UL
>
</P
><P
><A
HREF="#CHECKSUMREF"
>Checksum</A
></P
><P
><A
HREF="#CHILDREF"
>Child processes</A
></P
><P
><A
HREF="#NULLREF"
>Colon</A
>, <B
CLASS="COMMAND"
>: </B
>,
equivalent to the <A
HREF="#TRUEREF"
>true</A
> Bash
builtin</P
><P
><A
HREF="#COLORIZINGREF"
>Colorizing scripts</A
>
<P
></P
><UL
><LI
><P
>Cycling through the background colors, <A
HREF="#SHOWALLC"
>example script</A
></P
></LI
><LI
><P
><A
HREF="#COLORIZTABLE"
><B
CLASS="COMMAND"
>Table</B
></A
> of
color escape sequences</P
></LI
><LI
><P
><A
HREF="#COLORIZTEMPL"
>Template</A
>,
colored text on colored background</P
></LI
></UL
></P
><P
><A
HREF="#COMMAOP"
>Comma operator</A
>,
linking commands or operations</P
><P
><A
HREF="#BASH-OPTIONS"
>Command-line options</A
></P
><P
><A
HREF="#CNFH"
>command_not_found_handle ()</A
>
<I
CLASS="FIRSTTERM"
>builtin</I
> error-handling function
(<A
HREF="#BASH4REF"
>version 4+</A
> of Bash)</P
><P
><A
HREF="#COMMANDSUBREF"
>Command substitution</A
>
<P
></P
><UL
><LI
><P
> <A
HREF="#CSPARENS"
><B
CLASS="COMMAND"
>$( ... )</B
></A
>,
preferred notation</P
></LI
><LI
><P
><A
HREF="#BACKQUOTESREF"
><I
CLASS="FIRSTTERM"
>Backquotes</I
></A
></P
></LI
><LI
><P
><A
HREF="#CSTOOLSET"
>Extending the
<I
CLASS="FIRSTTERM"
>Bash</I
> toolset</A
></P
></LI
><LI
><P
><A
HREF="#CSSUBSH"
>Invokes a
<I
CLASS="FIRSTTERM"
>subshell</I
></A
></P
></LI
><LI
><P
><A
HREF="#CSNEST"
>Nesting</A
></P
></LI
><LI
><P
><A
HREF="#CSTRNL"
>Removes trailing newlines</A
>
</P
></LI
><LI
><P
><A
HREF="#CSVL"
>Setting variable from loop output</A
>
</P
></LI
><LI
><P
><A
HREF="#CSWS"
>Word
splitting</A
></P
></LI
></UL
></P
><P
><A
HREF="#COMMENTH"
>Comment headers</A
>,
special purpose</P
><P
>Commenting out blocks of code
<P
></P
><UL
><LI
><P
>Using an <A
HREF="#CBLOCK1"
><I
CLASS="FIRSTTERM"
>anonymous</I
> here
document</A
></P
></LI
><LI
><P
>Using an <A
HREF="#COMOUTBL"
> <I
CLASS="FIRSTTERM"
>if-then</I
> construct</A
></P
></LI
></UL
></P
><P
><A
HREF="#COMMUNICATIONS"
>Communications and
hosts</A
></P
><P
><A
HREF="#CCOMPARISON1"
>Compound comparison</A
>
operators</P
><P
><A
HREF="#FACOMPRESSION1"
>Compression utilities</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#BZIPREF"
>bzip2</A
></P
></LI
><LI
><P
><A
HREF="#COMPRESSREF"
>compress</A
></P
></LI
><LI
><P
><A
HREF="#GZIPREF"
>gzip</A
></P
></LI
><LI
><P
><A
HREF="#ZIPREF"
>zip</A
></P
></LI
></UL
></P
><P
><A
HREF="#BRKCONT1"
>continue</A
> loop control command</P
><P
><A
HREF="#CONTROLCHARREF"
>Control characters</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#CTLCREF"
>Control-C</A
>,
<I
CLASS="FIRSTTERM"
>break</I
></P
></LI
><LI
><P
><A
HREF="#CTLDREF"
>Control-D</A
>,
terminate / log out / erase</P
></LI
><LI
><P
><A
HREF="#CTLGREF"
>Control-G</A
>,
<TT
CLASS="USERINPUT"
><B
>BEL</B
></TT
>
(<I
CLASS="FIRSTTERM"
>beep</I
>)</P
></LI
><LI
><P
><A
HREF="#CTLHREF"
>Control-H</A
>,
<I
CLASS="FIRSTTERM"
>rubout</I
></P
></LI
><LI
><P
><A
HREF="#CTLJREF"
>Control-J</A
>,
<I
CLASS="FIRSTTERM"
>newline</I
></P
></LI
><LI
><P
><A
HREF="#CTLMREF"
>Control-M</A
>,
carriage return</P
></LI
></UL
></P
><P
><A
HREF="#COPROCREF"
>Coprocesses</A
></P
><P
><A
HREF="#CRONREF"
>cron</A
>, scheduling
<I
CLASS="FIRSTTERM"
>daemon</I
></P
><P
><A
HREF="#CSTYLE"
><I
CLASS="FIRSTTERM"
>C</I
>-style syntax
</A
>, for handling variables</P
><P
><A
HREF="#CWSOLVER"
>Crossword puzzle solver</A
></P
><P
><A
HREF="#GRONSFELD"
>Cryptography</A
></P
><P
>Curly brackets {}
<P
></P
><UL
><LI
><P
><A
HREF="#CURLYBRACKETSREF"
>in
<I
CLASS="FIRSTTERM"
>find</I
> command</A
></P
></LI
><LI
><P
><A
HREF="#ESCPCB"
>in an
<I
CLASS="FIRSTTERM"
>Extended Regular
Expression</I
></A
></P
></LI
><LI
><P
><A
HREF="#XARGSCURLYREF"
>in
<I
CLASS="FIRSTTERM"
>xargs</I
></A
></P
></LI
></UL
></P
><P
>* * *</P
><P
><A
HREF="#DAEMONREF"
>Daemons</A
>, in UNIX-type OS</P
><P
><A
HREF="#DATEREF"
>date</A
></P
><P
><A
HREF="#DCREF"
>dc</A
>, calculator utility</P
><P
><A
HREF="#DDREF"
>dd</A
>, <I
CLASS="FIRSTTERM"
>data
duplicator</I
> command
<P
></P
><UL
><LI
><P
><A
HREF="#DDCONVERSIONS"
>Conversions</A
></P
></LI
><LI
><P
><A
HREF="#DDCOPY"
>Copying raw data</A
>
to/from devices</P
></LI
><LI
><P
><A
HREF="#DDFDEL"
>File deletion</A
>,
<I
CLASS="FIRSTTERM"
>secure</I
></P
></LI
><LI
><P
><A
HREF="#DDKEYSTROKES"
>Keystrokes</A
>, capturing</P
></LI
><LI
><P
><A
HREF="#DDOPTIONS"
>Options</A
></P
></LI
><LI
><P
><A
HREF="#DDRANDOM"
>Random access</A
> on a data
stream</P
></LI
><LI
><P
><EM
>Raspberry Pi</EM
>,
<A
HREF="#RPSDCARD01"
>script for preparing a bootable SD
card</A
>
</P
></LI
><LI
><P
><A
HREF="#DDSWAP"
>Swapfiles</A
>, initializing</P
></LI
><LI
><P
><A
HREF="#DDLINK"
>Thread on
<I
CLASS="FIRSTTERM"
>www.linuxquestions.org</I
></A
></P
></LI
></UL
></P
><P
><A
HREF="#DEBUGGING"
>Debugging scripts</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#DEBUGTOOLS"
>Tools</A
></P
></LI
><LI
><P
><A
HREF="#DEBUGTRAP"
><I
CLASS="FIRSTTERM"
>Trapping</I
> at
exit</A
></P
></LI
><LI
><P
><A
HREF="#TRAPREF1"
><I
CLASS="FIRSTTERM"
>Trapping</I
>
signals</A
></P
></LI
></UL
></P
><P
><A
HREF="#NUMCONSTANTS"
>Decimal number</A
>,
Bash interprets numbers as</P
><P
><A
HREF="#DECLARE1REF"
>declare</A
> builtin</P
><P
></P
><UL
><LI
><P
><A
HREF="#DECLAREOPSREF1"
>options</A
></P
><P
><A
HREF="#DECLARECASEMOD"
>case-modification</A
>
options (<A
HREF="#BASH4REF"
>version 4+</A
> of Bash)
</P
></LI
></UL
><P
><A
HREF="#DEFPARAM"
>Default parameters</A
></P
><P
><A
HREF="#DEVPROCREF"
><TT
CLASS="FILENAME"
>/dev</TT
></A
> directory
<P
></P
><UL
><LI
><P
><A
HREF="#DEVNULLREF"
><TT
CLASS="FILENAME"
>/dev/null</TT
></A
>
pseudo-device file</P
></LI
><LI
><P
><A
HREF="#URANDOMREF"
><TT
CLASS="FILENAME"
>/dev/urandom</TT
></A
>
pseudo-device file, generating pseudorandom numbers
with</P
></LI
><LI
><P
><A
HREF="#ZEROSREF1"
><TT
CLASS="FILENAME"
>/dev/zero</TT
></A
>,
pseudo-device file</P
></LI
></UL
></P
><P
><A
HREF="#DEVFILEREF"
>Device file</A
></P
><P
><A
HREF="#DIALOGREF"
><I
CLASS="FIRSTTERM"
>dialog</I
></A
>,
utility for generating <I
CLASS="FIRSTTERM"
>dialog</I
> boxes in
a script</P
><P
><A
HREF="#DIRSTACKREF"
><TT
CLASS="VARNAME"
>$DIRSTACK</TT
></A
>
<I
CLASS="FIRSTTERM"
>directory stack</I
></P
><P
><A
HREF="#DISABLEDCOMMREF"
>Disabled commands</A
>,
in <I
CLASS="FIRSTTERM"
>restricted shells</I
></P
><P
><A
HREF="#DOINREF"
>do</A
> keyword,
begins execution of commands within a <A
HREF="#LOOPREF00"
>loop</A
></P
><P
><A
HREF="#DOINREF"
>done</A
> keyword,
terminates a loop</P
><P
><A
HREF="#DOSBATCH1"
><I
CLASS="FIRSTTERM"
>DOS</I
> batch
files</A
>, converting to shell scripts</P
><P
><A
HREF="#DOSUNIXEQUIV"
><I
CLASS="FIRSTTERM"
>DOS</I
>
commands</A
>, UNIX equivalents of
(<B
CLASS="COMMAND"
>table</B
>)</P
><P
><A
HREF="#DOTFILESREF"
><I
CLASS="FIRSTTERM"
>dot files</I
></A
>,
<SPAN
CLASS="QUOTE"
>"hidden"</SPAN
> setup and configuration files</P
><P
><A
HREF="#DBLBRACKETS"
>Double brackets</A
>
<B
CLASS="COMMAND"
>[[ ... ]]</B
> <A
HREF="#IFTHEN"
>test</A
>
construct</P
><P
></P
><UL
><LI
><P
>and <A
HREF="#DBLBRAEV"
>evaluation of
<I
CLASS="FIRSTTERM"
>octal/hex</I
> constants</A
></P
></LI
></UL
><P
><A
HREF="#DBLPARENSREF"
>Double parentheses</A
>
<B
CLASS="COMMAND"
>(( ... )) </B
> arithmetic expansion/evaluation
construct</P
><P
><A
HREF="#DBLQUO"
>Double quotes</A
>
<B
CLASS="COMMAND"
>" ... "</B
> <I
CLASS="FIRSTTERM"
>weak</I
> quoting
<P
></P
><UL
><LI
><P
><A
HREF="#QUOTINGBSL"
><I
CLASS="FIRSTTERM"
>Double-quoting</I
>
the <I
CLASS="FIRSTTERM"
>backslash</I
> (<B
CLASS="COMMAND"
>\</B
>)
character</A
></P
></LI
></UL
>
</P
><P
><A
HREF="#DOUBLESPACE"
>Double-spacing a text
file</A
>, using <A
HREF="#SEDREF"
>sed</A
></P
><P
>* * *</P
><P
><B
CLASS="COMMAND"
>-e </B
>
<A
HREF="#RTIF"
>File exists</A
> test</P
><P
><A
HREF="#ECHOREF"
>echo</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#ECHOGREPREF"
>Feeding commands down
a <I
CLASS="FIRSTTERM"
>pipe</I
></A
></P
></LI
><LI
><P
><A
HREF="#ECHOCS"
>Setting a variable</A
>
using <A
HREF="#COMMANDSUBREF"
>command
substitution</A
></P
></LI
><LI
><P
><A
HREF="#BINECHO"
><TT
CLASS="FILENAME"
>/bin/echo</TT
></A
>,
external <I
CLASS="FIRSTTERM"
>echo</I
> command</P
></LI
></UL
></P
><P
><A
HREF="#ELIFREF1"
>elif</A
>,
Contraction of <I
CLASS="FIRSTTERM"
>else</I
>
and <A
HREF="#IFTHEN"
>if</A
></P
><P
><A
HREF="#ELSEREF"
>else</A
></P
><P
>Encrypting files, using <A
HREF="#OPENSSLREF"
>openssl</A
></P
><P
><A
HREF="#CASEESAC1"
>esac</A
>, keyword terminating
<I
CLASS="FIRSTTERM"
>case</I
> construct</P
><P
><A
HREF="#ENVREF"
><I
CLASS="FIRSTTERM"
>Environmental</I
>
variables</A
></P
><P
><A
HREF="#EQUALREF"
> -eq </A
>,
<I
CLASS="FIRSTTERM"
>is-equal-to</I
> <A
HREF="#ICOMPARISON1"
>integer
comparison</A
> test</P
><P
><A
HREF="#PRIMES0"
>Eratosthenes,
Sieve of</A
>, algorithm for generating prime numbers</P
><P
><A
HREF="#SPM"
>Escaped characters</A
>,
special meanings of
<P
></P
><UL
><LI
><P
>Within <A
HREF="#STRQ"
>$' ... '</A
>
string expansion</P
></LI
><LI
><P
><A
HREF="#UNICODEREF2"
>Used with
<I
CLASS="FIRSTTERM"
>Unicode</I
> characters</A
></P
></LI
></UL
>
</P
><P
><A
HREF="#FSTABREF"
><TT
CLASS="FILENAME"
>/etc/fstab</TT
></A
>
(filesystem mount) file</P
><P
><A
HREF="#DATAFILESREF1"
><TT
CLASS="FILENAME"
>/etc/passwd</TT
></A
>
(user account) file</P
><P
><A
HREF="#EUIDREF"
><TT
CLASS="VARNAME"
>$EUID</TT
></A
>,
<I
CLASS="FIRSTTERM"
>Effective user ID</I
></P
><P
><A
HREF="#EVALREF"
>eval</A
>,
Combine and <I
CLASS="FIRSTTERM"
>evaluate</I
> expression(s),
with variable expansion
<P
></P
><UL
><LI
><P
><A
HREF="#EVALEFF"
>Effects
of</A
>, <I
CLASS="FIRSTTERM"
>Example script</I
></P
></LI
><LI
><P
><A
HREF="#EVALFORCED"
>Forces
<I
CLASS="FIRSTTERM"
>reevaluation</I
></A
> of
arguments</P
></LI
><LI
><P
>And <A
HREF="#EVALINDREF"
>indirect
references</A
></P
></LI
><LI
><P
><A
HREF="#EVALRISK"
>Risk of
using</A
></P
></LI
><LI
><P
><A
HREF="#SAMORSE"
>Using
<I
CLASS="FIRSTTERM"
>eval</I
> to convert <I
CLASS="FIRSTTERM"
>array</I
>
elements into a command list</A
></P
></LI
><LI
><P
><A
HREF="#ARRCHOICE0"
>Using
<I
CLASS="FIRSTTERM"
>eval</I
> to select among
variables</A
></P
></LI
></UL
></P
><P
><A
HREF="#DBLBRAEV"
>Evaluation of
<I
CLASS="FIRSTTERM"
>octal/hex</I
> constants within
[[ ... ]]</A
></P
><P
><A
HREF="#USINGEXECREF"
>exec</A
> command,
using in <A
HREF="#IOREDIRREF"
>redirection</A
></P
><P
><A
HREF="#EXERCISES"
>Exercises</A
></P
><P
>Exit and Exit status
<P
></P
><UL
><LI
><P
><A
HREF="#EXITCOMMANDREF"
>exit</A
>
command</P
></LI
><LI
><P
><A
HREF="#EXITSTATUSREF"
>Exit status</A
>
(<I
CLASS="FIRSTTERM"
>exit code</I
>, <I
CLASS="FIRSTTERM"
>return</I
>
status of a command)</P
><P
><A
HREF="#EXITCODESREF"
><B
CLASS="COMMAND"
>Table</B
></A
>,
<I
CLASS="FIRSTTERM"
>Exit
codes</I
> with special meanings</P
><P
> <A
HREF="#GOTCHAEXITVALANAMALIES"
>Anomalous</A
>
</P
><P
><A
HREF="#EXCOOR"
>Out of range</A
></P
><P
><A
HREF="#PIPEEX"
><I
CLASS="FIRSTTERM"
>Pipe</I
></A
>
exit status</P
><P
><A
HREF="#EXITRETURN1"
> Specified by a <I
CLASS="FIRSTTERM"
>function
return</I
></A
></P
><P
><A
HREF="#EXITSUCCESS"
><I
CLASS="FIRSTTERM"
>Successful</I
></A
>,
<B
CLASS="COMMAND"
>0</B
></P
><P
><A
HREF="#SYSEXITSREF"
><TT
CLASS="FILENAME"
>/usr/include/sysexits.h</TT
></A
>,
system file listing C/C++ standard exit codes</P
></LI
></UL
></P
><P
><A
HREF="#EXPORTREF2"
>Export</A
>,
to make available variables to <A
HREF="#CHILDREF"
>child processes</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#EXPORTAWK"
>Passing a variable
to an embedded <I
CLASS="FIRSTTERM"
>awk</I
>
script</A
></P
></LI
></UL
></P
><P
><A
HREF="#EXPRREF"
>expr</A
>,
<I
CLASS="FIRSTTERM"
>Expression</I
> evaluator
<P
></P
><UL
><LI
><P
><A
HREF="#EXPEXTRSUB"
>Substring
extraction</A
></P
></LI
><LI
><P
><A
HREF="#SUBSTRINGINDEX2"
>Substring
<I
CLASS="FIRSTTERM"
>index</I
> (numerical position in
string)</A
></P
></LI
><LI
><P
><A
HREF="#EXPRMATCH"
>Substring
matching</A
></P
></LI
></UL
></P
><P
><A
HREF="#EXTREGEX"
>Extended <I
CLASS="FIRSTTERM"
>Regular
Expressions</I
></A
>
<P
></P
><UL
><LI
><P
><B
CLASS="COMMAND"
>? </B
> (question mark)
<A
HREF="#QUEXREGEX"
>Match zero / one characters</A
>
</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>( ... ) </B
>
<A
HREF="#PARENGRPS"
>Group
of expressions</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>\{ N \}</B
>
<A
HREF="#ESCPCB"
><SPAN
CLASS="QUOTE"
>"Curly"</SPAN
> brackets</A
>,
<I
CLASS="FIRSTTERM"
>escaped</I
>,
number of character sets to match</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>+ </B
>
<A
HREF="#PLUSREF"
><I
CLASS="FIRSTTERM"
>Character
match</I
></A
></P
></LI
></UL
></P
><P
>* * *</P
><P
><A
HREF="#FACTORREF"
>factor</A
>, decomposes an
integer into its prime factors
<P
></P
><UL
><LI
><P
>Application: <A
HREF="#PRIMES2"
>Generating
prime numbers</A
></P
></LI
></UL
></P
><P
><A
HREF="#FALSEREF"
>false</A
>,
returns <I
CLASS="FIRSTTERM"
>unsuccessful</I
> (1) <A
HREF="#EXITSTATUSREF"
>exit status</A
></P
><P
><A
HREF="#FIELDREF"
>Field</A
>, a group of characters
that comprises an item of data</P
><P
><A
HREF="#FILEARCHIV"
>Files / Archiving</A
></P
><P
><A
HREF="#FDREF"
>File descriptors</A
></P
><P
></P
><UL
><LI
><P
><A
HREF="#CFD"
>Closing</A
></P
><P
><B
CLASS="COMMAND"
>n&#60;&#38;-</B
>
Close input file descriptor <I
CLASS="FIRSTTERM"
>n</I
></P
><P
><B
CLASS="COMMAND"
>0&#60;&#38;-</B
>,
<B
CLASS="COMMAND"
>&#60;&#38;-</B
>
Close <TT
CLASS="FILENAME"
>stdin</TT
></P
><P
><B
CLASS="COMMAND"
>n&#62;&#38;-</B
>
Close output file descriptor <I
CLASS="FIRSTTERM"
>n</I
></P
><P
><B
CLASS="COMMAND"
>1&#62;&#38;-</B
>,
<B
CLASS="COMMAND"
>&#62;&#38;-</B
>
Close <TT
CLASS="FILENAME"
>stdout</TT
></P
></LI
><LI
><P
><A
HREF="#FDREF1"
>File handles in
<I
CLASS="FIRSTTERM"
>C</I
></A
>, similarity to</P
></LI
></UL
><P
><A
HREF="#OPENSSLREF"
>File encryption</A
></P
><P
><A
HREF="#FINDREF"
>find</A
>
<P
></P
><UL
><LI
><P
><B
CLASS="COMMAND"
>{} </B
>
<A
HREF="#CURLYBRACKETSREF"
>Curly
brackets</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>\; </B
>
<A
HREF="#FINDREF0"
><I
CLASS="FIRSTTERM"
>Escaped</I
>
semicolon</A
></P
></LI
></UL
></P
><P
><A
HREF="#FILTERDEF"
>Filter</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#FILTERDASH"
>Using <SPAN
CLASS="TOKEN"
>-</SPAN
>
with file-processing utility as a filter</A
></P
></LI
><LI
><P
><A
HREF="#FILTEROUTP"
> Feeding output of a filter back to <EM
>same</EM
>
filter</A
></P
></LI
></UL
></P
><P
><A
HREF="#NOFLOATINGPOINT"
>Floating point numbers</A
>,
Bash does not recognize</P
><P
><A
HREF="#FOLDREF"
>fold</A
>, a filter to wrap lines of
text</P
><P
><A
HREF="#FORKREF"
>Forking</A
> a <I
CLASS="FIRSTTERM"
>child</I
>
process</P
><P
><A
HREF="#FORLOOPREF1"
><I
CLASS="FIRSTTERM"
>for</I
>
loops</A
></P
><P
><A
HREF="#FUNCTIONREF"
>Functions</A
></P
><P
></P
><UL
><LI
><P
><A
HREF="#PASSEDARGS"
>Arguments
passed</A
> referred to by position</P
></LI
><LI
><P
><A
HREF="#CAPTURERETVAL"
>Capturing
the return value</A
> of a function
using <A
HREF="#ECHOREF"
>echo</A
></P
></LI
><LI
><P
><A
HREF="#COLONFNAME"
><I
CLASS="FIRSTTERM"
>Colon</I
></A
>
as function name</P
></LI
><LI
><P
><A
HREF="#FUNCTDEFMUST"
>Definition
must precede</A
> first call to function</P
></LI
><LI
><P
><A
HREF="#EXITRETURN1"
>Exit
status</A
></P
></LI
><LI
><P
><A
HREF="#LOCALREF1"
>Local variables</A
></P
><P
>and <A
HREF="#LOCVARRECUR"
>recursion</A
></P
></LI
><LI
><P
><A
HREF="#PASSARRAY"
>Passing
an <I
CLASS="FIRSTTERM"
>array</I
></A
> to a
function</P
></LI
><LI
><P
><A
HREF="#FUNCPOINTERS"
>Passing pointers</A
>
to a function</P
></LI
><LI
><P
><A
HREF="#PASSEDARGS"
>Positional
parameters</A
></P
></LI
><LI
><P
><A
HREF="#RECURSIONREF0"
>Recursion</A
></P
></LI
><LI
><P
><A
HREF="#REDSTDINFUNC1"
>Redirecting
<TT
CLASS="FILENAME"
>stdin</TT
></A
>
of a function</P
></LI
><LI
><P
><A
HREF="#RETURNREF"
>return</A
></P
><P
>Multiple <I
CLASS="FIRSTTERM"
>return values</I
> from
a function,
<A
HREF="#STDDEV"
>example script</A
></P
><P
><A
HREF="#RETARRAY"
> Returning an <I
CLASS="FIRSTTERM"
>array</I
></A
> from
a function</P
><P
><A
HREF="#RVT"
><I
CLASS="FIRSTTERM"
>Return</I
>
range limits</A
>, workarounds</P
></LI
><LI
><P
><A
HREF="#FSHIFTREF"
><I
CLASS="FIRSTTERM"
>Shift</I
>
arguments passed</A
> to a function</P
></LI
><LI
><P
><A
HREF="#FSTRANGEREF"
>Unusual function
names</A
></P
></LI
></UL
><P
>* * *</P
><P
>Games and amusements
<P
></P
><UL
><LI
><P
><A
HREF="#AGRAM"
>Anagrams</A
></P
></LI
><LI
><P
><A
HREF="#AGRAM2"
>Anagrams</A
>, again</P
></LI
><LI
><P
><A
HREF="#BINGO"
>Bingo Number Generator</A
></P
></LI
><LI
><P
><A
HREF="#CWSOLVER"
>Crossword puzzle
solver</A
></P
></LI
><LI
><P
><A
HREF="#CRYPTOQUOTE"
>Crypto-Quotes</A
></P
></LI
><LI
><P
><A
HREF="#CARDS"
>Dealing a deck of cards</A
></P
></LI
><LI
><P
><A
HREF="#FIFTEEN"
>Fifteen Puzzle</A
></P
></LI
><LI
><P
><A
HREF="#HORSERACE"
>Horse race</A
></P
></LI
><LI
><P
><A
HREF="#KTOUR"
>Knight's Tour</A
></P
></LI
><LI
><P
><A
HREF="#LIFESLOW"
><SPAN
CLASS="QUOTE"
>"Life"</SPAN
>
game</A
></P
></LI
><LI
><P
><A
HREF="#MSQUARE"
>Magic Squares</A
></P
></LI
><LI
><P
><A
HREF="#MUSICSCR"
>Music-playing
script</A
></P
></LI
><LI
><P
><A
HREF="#NIM"
>Nim</A
></P
></LI
><LI
><P
><A
HREF="#BROWNIAN"
>Pachinko</A
></P
></LI
><LI
><P
><A
HREF="#QKY"
>Perquackey</A
></P
></LI
><LI
><P
><A
HREF="#PETALS"
>Petals Around the Rose</A
></P
></LI
><LI
><P
><A
HREF="#BASHPODDER"
>Podcasting</A
></P
></LI
><LI
><P
><A
HREF="#POEM"
>Poem</A
></P
></LI
><LI
><P
><A
HREF="#SPEECH00"
>Speech
generation</A
></P
></LI
><LI
><P
><A
HREF="#HANOI"
>Towers of Hanoi</A
></P
><P
><A
HREF="#HANOI2"
>Graphic
version</A
></P
><P
><A
HREF="#HANOI2A"
>Alternate graphic
version</A
></P
></LI
></UL
></P
><P
><A
HREF="#GETOPTY"
>getopt</A
>,
<I
CLASS="FIRSTTERM"
>external</I
> command for parsing script
<I
CLASS="FIRSTTERM"
>command-line</I
> arguments
<P
></P
><UL
><LI
><P
><A
HREF="#GETOPTSIMPLE1"
>Emulated
in a script</A
></P
></LI
></UL
></P
><P
><A
HREF="#GETOPTSX"
>getopts</A
>,
Bash <I
CLASS="FIRSTTERM"
>builtin</I
> for parsing script
<I
CLASS="FIRSTTERM"
>command-line</I
> arguments
<P
></P
><UL
><LI
><P
><A
HREF="#GETOPTSOPT"
><TT
CLASS="VARNAME"
>$OPTIND</TT
> /
<TT
CLASS="VARNAME"
>$OPTARG</TT
></A
></P
></LI
></UL
></P
><P
><A
HREF="#SCOPEREF"
>Global</A
>
variable</P
><P
><A
HREF="#GLOBBINGREF2"
>Globbing</A
>,
filename expansion
<P
></P
><UL
><LI
><P
><A
HREF="#HANDLINGFNAMES"
>Handling
filenames correctly</A
></P
></LI
><LI
><P
><A
HREF="#ASTERISKREF"
><I
CLASS="FIRSTTERM"
>Wild
cards</I
></A
></P
></LI
><LI
><P
><A
HREF="#WDOTFILEWC"
>Will not match
<TT
CLASS="FILENAME"
>dot files</TT
></A
></P
></LI
></UL
></P
><P
><A
HREF="#GOLDENRATIO"
>Golden Ratio</A
>
(<I
CLASS="FIRSTTERM"
>Phi</I
>)</P
><P
><A
HREF="#GE0REF"
> -ge </A
>,
<I
CLASS="FIRSTTERM"
>greater-than or equal</I
>
<A
HREF="#ICOMPARISON1"
>integer comparison</A
> test</P
><P
><A
HREF="#GT0REF"
> -gt </A
>,
<I
CLASS="FIRSTTERM"
>greater-than</I
> <A
HREF="#ICOMPARISON1"
>integer
comparison</A
> test</P
><P
><A
HREF="#GROFFREF"
><I
CLASS="FIRSTTERM"
>groff</I
></A
>,
text markup and formatting language</P
><P
><A
HREF="#GRONSFELD"
>Gronsfeld cipher</A
></P
><P
><A
HREF="#GROUPSREF"
><TT
CLASS="VARNAME"
>$GROUPS</TT
></A
>,
<I
CLASS="FIRSTTERM"
>Groups</I
> user belongs to</P
><P
><A
HREF="#GZIPREF"
>gzip</A
>, compression utility</P
><P
>* * *</P
><P
><A
HREF="#HASHREF"
>Hashing</A
>, creating lookup keys
in a table
<P
></P
><UL
><LI
><P
><A
HREF="#HASHEX2_0"
><I
CLASS="FIRSTTERM"
>Example
script</I
></A
></P
></LI
></UL
></P
><P
><A
HREF="#HEADREF"
>head</A
>, <I
CLASS="FIRSTTERM"
>echo</I
>
to <TT
CLASS="FILENAME"
>stdout</TT
> lines at the beginning of a text
file</P
><P
><A
HREF="#HELPREF"
>help</A
>, gives usage summary of
a Bash <A
HREF="#BUILTINREF"
>builtin</A
></P
><P
><A
HREF="#HEREDOCREF"
><I
CLASS="FIRSTTERM"
>Here</I
>
documents</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#ANONHEREDOC0"
><I
CLASS="FIRSTTERM"
>Anonymous</I
>
here documents</A
>, using <B
CLASS="COMMAND"
>:</B
></P
><P
><A
HREF="#CBLOCK1"
>Commenting out</A
>
blocks of code</P
><P
><A
HREF="#HSELFDOC"
>Self-documenting</A
>
scripts</P
></LI
><LI
><P
><A
HREF="#BCHEREDOC"
><I
CLASS="FIRSTTERM"
>bc</I
> in a <I
CLASS="FIRSTTERM"
>here
document</I
></A
></P
></LI
><LI
><P
><A
HREF="#CATSCRIPTREF"
><I
CLASS="FIRSTTERM"
>cat</I
>
scripts</A
></P
></LI
><LI
><P
><A
HREF="#HERECS"
>Command
substitution</A
></P
></LI
><LI
><P
><A
HREF="#EXSCRIPTREF"
><I
CLASS="FIRSTTERM"
>ex</I
>
scripts</A
></P
></LI
><LI
><P
><A
HREF="#HEREFUNC"
><I
CLASS="FIRSTTERM"
>Function</I
></A
>,
supplying input to</P
></LI
><LI
><P
><A
HREF="#HERESTRINGSREF"
><I
CLASS="FIRSTTERM"
>Here</I
>
strings</A
></P
><P
>Calculating the <A
HREF="#GOLDENRATIO"
>Golden
Ratio</A
></P
><P
><A
HREF="#HSPRE"
>Prepending text</A
></P
><P
><A
HREF="#HSLOOP"
>As the <TT
CLASS="FILENAME"
>stdin</TT
> of a
<I
CLASS="FIRSTTERM"
>loop</I
></A
></P
><P
><A
HREF="#HSREAD"
>Using
<I
CLASS="FIRSTTERM"
>read</I
></A
></P
></LI
><LI
><P
><A
HREF="#LIMITSTRINGREF"
><I
CLASS="FIRSTTERM"
>Limit</I
>
string</A
></P
><P
><A
HREF="#EXCLLS"
><SPAN
CLASS="TOKEN"
>!</SPAN
> as a
<I
CLASS="FIRSTTERM"
>limit string</I
></A
></P
><P
><A
HREF="#INDENTEDLS"
>Closing <I
CLASS="FIRSTTERM"
>limit
string</I
></A
> may not be indented</P
><P
><A
HREF="#LIMITSTRDASH"
>Dash option</A
>
to limit string, <TT
CLASS="OPTION"
>&#60;&#60;-LimitString</TT
>
</P
></LI
><LI
><P
><A
HREF="#HERELIT"
>Literal
text output</A
>,
for generating program code</P
></LI
><LI
><P
><A
HREF="#HEREPARAMSUB"
>Parameter
substitution</A
></P
><P
><A
HREF="#HEREESC"
>Disabling</A
>
<I
CLASS="FIRSTTERM"
>parameter substitution</I
></P
></LI
><LI
><P
><A
HREF="#HEREPASSP"
>Passing
parameters</A
></P
></LI
><LI
><P
><A
HREF="#HERETEMP"
>Temporary
files</A
></P
></LI
><LI
><P
><A
HREF="#VIHERE"
>Using
<I
CLASS="FIRSTTERM"
>vi</I
> non-interactively</A
></P
></LI
></UL
></P
><P
><A
HREF="#HISTCOMMANDS"
>History commands</A
></P
><P
><A
HREF="#HOMEDIRREF"
><TT
CLASS="VARNAME"
>$HOME</TT
></A
>,
<I
CLASS="FIRSTTERM"
>user's home directory</I
></P
><P
><A
HREF="#HOMEWORK"
>Homework assignment solver</A
></P
><P
><A
HREF="#HOSTNAMEREF"
><TT
CLASS="VARNAME"
>$HOSTNAME</TT
></A
>,
system <I
CLASS="FIRSTTERM"
>host name</I
></P
><P
>* * *</P
><P
><A
HREF="#RCSREF"
><TT
CLASS="VARNAME"
>$Id</TT
>
parameter</A
>, in <I
CLASS="FIRSTTERM"
>rcs</I
> (Revision Control
System)</P
><P
><A
HREF="#IFTHEN"
>if [ condition ]; then ...</A
>
<I
CLASS="FIRSTTERM"
>test</I
> construct
<P
></P
><UL
><LI
><P
><A
HREF="#IFGREPREF"
>if-grep</A
>,
<I
CLASS="FIRSTTERM"
>if</I
> and <A
HREF="#GREPREF"
>grep</A
>
in combination</P
><P
><A
HREF="#IFGREPFIX"
>Fixup</A
>
for <I
CLASS="FIRSTTERM"
>if-grep</I
> test</P
></LI
></UL
></P
><P
><A
HREF="#IFSREF"
><TT
CLASS="VARNAME"
>$IFS</TT
></A
>,
<I
CLASS="FIRSTTERM"
>Internal field separator</I
> variable
<P
></P
><UL
><LI
><P
><A
HREF="#IFSWS"
>Defaults to
<I
CLASS="FIRSTTERM"
>whitespace</I
></A
></P
></LI
></UL
></P
><P
><A
HREF="#ICOMPARISON1"
>Integer comparison
operators</A
></P
><P
><A
HREF="#DOINREF"
>in</A
>,
<I
CLASS="FIRSTTERM"
>keyword</I
> preceding
<TT
CLASS="VARNAME"
>[list]</TT
> in a <I
CLASS="FIRSTTERM"
>for</I
>
loop</P
><P
><A
HREF="#INITTABREF"
>Initialization table</A
>,
<TT
CLASS="FILENAME"
>/etc/inittab</TT
></P
><P
><A
HREF="#CODEBLOCKREF"
>Inline group</A
>,
i.e., code block</P
><P
><A
HREF="#IITEST"
>Interactive script</A
>, test for</P
><P
><A
HREF="#IOREDIRREF"
>I/O redirection</A
></P
><P
><A
HREF="#IVRREF"
>Indirect referencing of variables</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#IVR2"
>New notation</A
>, introduced
in <A
HREF="#BASH2REF"
>version 2</A
> of Bash (<A
HREF="#VARREFNEW"
> example script</A
>)</P
></LI
></UL
></P
><P
><A
HREF="#IPTABLESREF"
>iptables</A
>,
packet filtering and firewall utility
<P
></P
><UL
><LI
><P
><A
HREF="#IPTABLES01"
>Usage
example</A
></P
></LI
><LI
><P
><A
HREF="#IPTABLES02"
>Example
script</A
></P
></LI
></UL
></P
><P
><A
HREF="#ITERATIONREF"
>Iteration</A
></P
><P
>* * *</P
><P
><A
HREF="#JOBIDTABLE0"
>Job IDs</A
>, table</P
><P
><A
HREF="#JOTREF"
>jot</A
>,
Emit a sequence of integers. Equivalent to <A
HREF="#SEQREF"
>seq</A
>.
<P
></P
><UL
><LI
><P
><A
HREF="#JOTRANDOM"
>Random sequence
generation</A
></P
></LI
></UL
></P
><P
><A
HREF="#JABH"
>Just another Bash hacker!</A
></P
><P
>* * *</P
><P
><A
HREF="#KEYWORDREF"
>Keywords</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#MISSINGKEYWORD"
>error</A
>,
if missing</P
></LI
></UL
></P
><P
><A
HREF="#KILLREF"
>kill</A
>,
terminate a process by <A
HREF="#PROCESSIDDEF"
>process ID</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#ZOMBIEREF"
>Options</A
> (<TT
CLASS="OPTION"
>-l</TT
>,
<TT
CLASS="OPTION"
>-9</TT
>)</P
></LI
></UL
></P
><P
><A
HREF="#KILLALLREF"
>killall</A
>,
terminate a process <EM
>by name</EM
></P
><P
><A
HREF="#KILLALL2REF"
><I
CLASS="FIRSTTERM"
>killall
script</I
></A
> in <TT
CLASS="FILENAME"
>/etc/rc.d/init.d</TT
></P
><P
>* * *</P
><P
><A
HREF="#LASTPIPEREF"
>lastpipe</A
> shell
option</P
><P
><A
HREF="#LE0REF"
> -le </A
>,
<I
CLASS="FIRSTTERM"
>less-than or equal</I
>
<A
HREF="#ICOMPARISON1"
>integer comparison</A
> test</P
><P
><A
HREF="#LETREF"
>let</A
>,
setting and carrying out arithmetic operations on variables
<P
></P
><UL
><LI
><P
><I
CLASS="FIRSTTERM"
>C-style</I
>
<A
HREF="#EX46"
>increment and decrement
operators</A
></P
></LI
></UL
></P
><P
><A
HREF="#LIMITSTRINGREF"
>Limit string</A
>,
in a <A
HREF="#HEREDOCREF"
>here document</A
></P
><P
><A
HREF="#LINENOREF"
><TT
CLASS="VARNAME"
>$LINENO</TT
></A
>,
variable indicating the <I
CLASS="FIRSTTERM"
>line number</I
> where
it appears in a script</P
><P
><A
HREF="#LINKREF"
>Link</A
>, file
(using <I
CLASS="FIRSTTERM"
>ln</I
> command)
<P
></P
><UL
><LI
><P
><A
HREF="#LINKMINVOK"
>Invoking script with multiple names</A
>,
using <I
CLASS="FIRSTTERM"
>ln</I
></P
></LI
><LI
><P
><A
HREF="#SYMLINKREF"
><I
CLASS="FIRSTTERM"
>symbolic</I
>
links</A
>, <I
CLASS="FIRSTTERM"
>ln -s </I
></P
></LI
></UL
></P
><P
><A
HREF="#LISTCONSREF"
>List constructs</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#LCONS1"
><I
CLASS="FIRSTTERM"
>And</I
>
list</A
></P
></LI
><LI
><P
><A
HREF="#ORLISTREF"
><I
CLASS="FIRSTTERM"
>Or</I
>
list</A
></P
></LI
></UL
></P
><P
><A
HREF="#LOCALREF1"
>Local variables</A
>
<P
></P
><UL
><LI
><P
>and <A
HREF="#LOCVARRECUR"
>recursion</A
></P
></LI
></UL
></P
><P
><A
HREF="#LOCALIZATION"
>Localization</A
></P
><P
><A
HREF="#LOGOPS1"
>Logical operators</A
>
(<TT
CLASS="VARNAME"
>&#38;&#38;</TT
>, <TT
CLASS="VARNAME"
>||</TT
>,
etc.)</P
><P
><A
HREF="#LOGOUTFILEREF1"
>Logout file</A
>,
the <TT
CLASS="FILENAME"
>~/.bash_logout</TT
> file</P
><P
><A
HREF="#ISOMOUNTREF0"
>Loopback device</A
>,
mounting a file on a <A
HREF="#BLOCKDEVREF"
>block
device</A
></P
><P
><A
HREF="#LOOPS1"
>Loops</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#BRKCONT1"
>break</A
>
loop control command</P
></LI
><LI
><P
><A
HREF="#BRKCONT1"
>continue</A
> loop
control command</P
></LI
><LI
><P
><I
CLASS="FIRSTTERM"
>C</I
>-style
loop within <A
HREF="#DBLPARENSREF"
>double
parentheses</A
></P
><P
><A
HREF="#LOOPCSTYLE"
><I
CLASS="FIRSTTERM"
>for</I
> loop</A
></P
><P
><A
HREF="#WLOOPCSTYLE"
><I
CLASS="FIRSTTERM"
>while</I
> loop</A
></P
></LI
><LI
><P
><A
HREF="#DOINREF"
>do</A
> (keyword),
begins execution of commands within a loop</P
></LI
><LI
><P
><A
HREF="#DOINREF"
>done</A
> (keyword),
terminates a loop</P
></LI
><LI
><P
><A
HREF="#FORLOOPREF1"
><I
CLASS="FIRSTTERM"
>for</I
>
loops</A
></P
><P
><TT
CLASS="REPLACEABLE"
><I
>for</I
></TT
> <TT
CLASS="VARNAME"
>arg</TT
>
<TT
CLASS="REPLACEABLE"
><I
>in</I
></TT
> <TT
CLASS="VARNAME"
>[list]</TT
>;
<TT
CLASS="REPLACEABLE"
><I
>do</I
></TT
></P
><P
><A
HREF="#LOOPCS"
><I
CLASS="FIRSTTERM"
>Command
substitution</I
> to generate
<TT
CLASS="VARNAME"
>[list]</TT
></A
></P
><P
><A
HREF="#LIGLOB"
>Filename expansion
in <TT
CLASS="VARNAME"
>[list]</TT
></A
></P
><P
><A
HREF="#MULTPARAML"
>Multiple parameters
in each <TT
CLASS="VARNAME"
>[list]</TT
> element</A
></P
><P
><A
HREF="#OMITLIST"
>Omitting
<TT
CLASS="VARNAME"
>[list]</TT
></A
>, defaults to
<A
HREF="#POSPARAMREF"
>positional parameters</A
></P
><P
><A
HREF="#PARAMLI"
>Parameterizing
<TT
CLASS="VARNAME"
>[list]</TT
></A
></P
><P
><A
HREF="#LOOPREDIR"
>Redirection</A
></P
></LI
><LI
><P
><A
HREF="#DOINREF"
>in</A
>,
(keyword) preceding [list] in a
<I
CLASS="FIRSTTERM"
>for</I
> loop</P
></LI
><LI
><P
><A
HREF="#NESTEDLOOPS"
>Nested
loops</A
></P
></LI
><LI
><P
><A
HREF="#BGLOOP0"
>Running a
loop <EM
>in the background</EM
></A
>,
<I
CLASS="FIRSTTERM"
>script example</I
></P
></LI
><LI
><P
>Semicolon required, when <I
CLASS="FIRSTTERM"
>do</I
> is on first line
of loop</P
><P
><A
HREF="#NEEDSEMICOLON"
><I
CLASS="FIRSTTERM"
>for</I
>
loop</A
></P
><P
><A
HREF="#WHILENEEDSEMI"
><I
CLASS="FIRSTTERM"
>while</I
>
loop</A
></P
></LI
><LI
><P
><A
HREF="#UNTILLOOPREF"
>until</A
> loop</P
><P
><TT
CLASS="REPLACEABLE"
><I
>until [ condition-is-true ]; do</I
></TT
>
</P
></LI
><LI
><P
><A
HREF="#WHILELOOPREF"
>while</A
>
loop</P
><P
><TT
CLASS="REPLACEABLE"
><I
>while [ condition ]; do</I
></TT
></P
><P
><A
HREF="#WHILEFUNC"
>Function call</A
>
inside test brackets</P
><P
><A
HREF="#WHMULTCOND"
>Multiple conditions</A
></P
><P
><A
HREF="#WHILENOBRACKETS"
>Omitting <I
CLASS="FIRSTTERM"
>test
brackets</I
></A
></P
><P
><A
HREF="#WHREDIR"
>Redirection</A
></P
><P
><A
HREF="#WHILEREADREF2"
> <I
CLASS="FIRSTTERM"
>while read</I
></A
> construct</P
></LI
><LI
><P
><A
HREF="#CHOOSELOOP"
>Which type of loop to
use</A
></P
></LI
></UL
></P
><P
>Loopback devices
<P
></P
><UL
><LI
><P
><A
HREF="#LOOPBACKREF"
>In <TT
CLASS="FILENAME"
>/dev</TT
> directory</A
></P
></LI
><LI
><P
><A
HREF="#ISOMOUNTREF0"
>Mounting an ISO image</A
></P
></LI
></UL
></P
><P
><A
HREF="#LT0REF"
> -lt </A
>,
<I
CLASS="FIRSTTERM"
>less-than</I
> <A
HREF="#ICOMPARISON1"
>integer
comparison</A
> test</P
><P
>* * *</P
><P
><A
HREF="#M4REF"
>m4</A
>, macro processing language</P
><P
><A
HREF="#MACHTYPEREF"
><TT
CLASS="VARNAME"
>$MACHTYPE</TT
></A
>,
<I
CLASS="FIRSTTERM"
>Machine type</I
></P
><P
><A
HREF="#MAGNUMREF"
>Magic number</A
>,
marker at the head of a file indicating the file type</P
><P
><A
HREF="#MAKEFILEREF"
><TT
CLASS="FILENAME"
>Makefile</TT
></A
>,
file containing the list of dependencies used by <A
HREF="#MAKEREF"
>make</A
> command</P
><P
><A
HREF="#MANREF"
>man</A
>, <I
CLASS="FIRSTTERM"
>manual
page</I
> (lookup)
<P
></P
><UL
><LI
><P
><A
HREF="#MANED"
><I
CLASS="FIRSTTERM"
>Man page</I
>
editor</A
> (script)</P
></LI
></UL
>
</P
><P
><A
HREF="#MAPFILEREF"
>mapfile</A
> builtin,
loads an array with a text file</P
><P
><A
HREF="#MATHC"
>Math commands</A
></P
><P
><A
HREF="#METAMEANINGREF"
>Meta-meaning</A
></P
><P
><A
HREF="#SAMORSE"
>Morse code training</A
> script</P
><P
><A
HREF="#MODULOREF"
>Modulo</A
>, arithmetic
<I
CLASS="FIRSTTERM"
>remainder</I
> operator
<P
></P
><UL
><LI
><P
>Application: <A
HREF="#PRIMES1"
>Generating
prime numbers</A
></P
></LI
></UL
></P
><P
><A
HREF="#MONTHLYPMT0"
>Mortgage calculations</A
>,
<I
CLASS="FIRSTTERM"
>example script</I
></P
><P
>* * *</P
><P
><B
CLASS="COMMAND"
>-n </B
>
<A
HREF="#STRINGNOTNULL"
>String not
<I
CLASS="FIRSTTERM"
>null</I
></A
> test</P
><P
><A
HREF="#NAMEDPIPEREF"
>Named pipe</A
>, a temporary
FIFO buffer
<P
></P
><UL
><LI
><P
><A
HREF="#ZFIFO"
><I
CLASS="FIRSTTERM"
>Example
script</I
></A
></P
></LI
></UL
></P
><P
><A
HREF="#NCREF"
>nc</A
>, <I
CLASS="FIRSTTERM"
>netcat</I
>,
a network toolkit for TCP and UDP ports</P
><P
><A
HREF="#NEQUALREF"
>-ne</A
>,
<I
CLASS="FIRSTTERM"
>not-equal-to</I
> <A
HREF="#ICOMPARISON1"
>integer comparison</A
> test</P
><P
><A
HREF="#NOTREF"
>Negation operator</A
>,
<B
CLASS="COMMAND"
>!</B
>, reverses the sense of a <A
HREF="#IFTHEN"
>test</A
></P
><P
><A
HREF="#NETSTATREF"
>netstat</A
>, Network
statistics</P
><P
><A
HREF="#NETWORKPROGRAMMING"
>Network programming</A
></P
><P
><A
HREF="#NLREF"
>nl</A
>, a filter to number lines of
text</P
><P
><A
HREF="#NOCLOBBERREF"
><I
CLASS="FIRSTTERM"
>Noclobber</I
></A
>,
<TT
CLASS="OPTION"
>-C</TT
> option to Bash to prevent overwriting
of files</P
><P
><A
HREF="#LOGOPS1"
><I
CLASS="FIRSTTERM"
>NOT</I
> logical
operator</A
>, <B
CLASS="COMMAND"
>!</B
></P
><P
><A
HREF="#NULLVAR"
><I
CLASS="FIRSTTERM"
>null</I
> variable
assignment</A
>, avoiding</P
><P
>* * *</P
><P
><B
CLASS="COMMAND"
>-o </B
>
<A
HREF="#COMPOUNDOR"
>Logical OR</A
>
compound comparison test</P
><P
>Obfuscation
<P
></P
><UL
><LI
><P
><A
HREF="#COLONFNAME"
><I
CLASS="FIRSTTERM"
>Colon</I
></A
>
as function name</P
></LI
><LI
><P
><A
HREF="#HOMEWORK"
>Homework assignment</A
></P
></LI
><LI
><P
><A
HREF="#JABH"
>Just another Bash
hacker!</A
></P
></LI
></UL
>
</P
><P
><A
HREF="#OCTALREF"
>octal</A
>, base-8 numbers</P
><P
><A
HREF="#ODREF"
>od</A
>, <I
CLASS="FIRSTTERM"
>octal
dump</I
></P
><P
><A
HREF="#OLDPWD"
><TT
CLASS="VARNAME"
>$OLDPWD</TT
></A
>
Previous working directory</P
><P
><A
HREF="#OPENSSLREF"
>openssl</A
> encryption
utility</P
><P
>Operator
<P
></P
><UL
><LI
><P
><A
HREF="#OPERATORDEF"
>Definition
of</A
></P
></LI
><LI
><P
><A
HREF="#OPPRECEDENCE1"
>Precedence</A
></P
></LI
></UL
></P
><P
><A
HREF="#OPTIONSREF"
>Options</A
>,
passed to shell or script on command line or by <A
HREF="#SETREF"
>set</A
> command</P
><P
><A
HREF="#ORLISTREF"
><I
CLASS="FIRSTTERM"
>Or</I
>
list</A
></P
><P
><A
HREF="#ORREF"
><I
CLASS="FIRSTTERM"
>Or</I
> logical
operator</A
>, <B
CLASS="COMMAND"
>||</B
></P
><P
>* * *</P
><P
><A
HREF="#PARAMSUBREF"
>Parameter
substitution</A
>
<P
></P
><UL
><LI
><P
><I
CLASS="FIRSTTERM"
>${parameter+alt_value}</I
></P
><P
><I
CLASS="FIRSTTERM"
>${parameter:+alt_value}</I
></P
><P
><A
HREF="#PARAMALTV"
>Alternate value</A
>
of parameter, if set</P
></LI
><LI
><P
><I
CLASS="FIRSTTERM"
>${parameter-default}</I
></P
><P
><I
CLASS="FIRSTTERM"
>${parameter:-default}</I
></P
><P
><I
CLASS="FIRSTTERM"
>${parameter=default}</I
></P
><P
><I
CLASS="FIRSTTERM"
>${parameter:=default}</I
></P
><P
><A
HREF="#DEFPARAM1"
>Default
parameters</A
></P
></LI
><LI
><P
><I
CLASS="FIRSTTERM"
>${!varprefix*}</I
></P
><P
><I
CLASS="FIRSTTERM"
>${!varprefix@}</I
></P
><P
><A
HREF="#VARPREFIXM"
>Parameter
<EM
>name</EM
> match</A
></P
></LI
><LI
><P
><I
CLASS="FIRSTTERM"
>${parameter?err_msg}</I
></P
><P
><A
HREF="#QERRMSG"
>Parameter-unset message</A
></P
></LI
><LI
><P
><I
CLASS="FIRSTTERM"
>${parameter}</I
></P
><P
><A
HREF="#PSSUB1"
>Value of
<I
CLASS="FIRSTTERM"
>parameter</I
></A
></P
></LI
><LI
><P
><A
HREF="#CASEMODPARAMSUB"
> <I
CLASS="FIRSTTERM"
>Case modification</I
></A
>
(<A
HREF="#BASH4REF"
>version 4+</A
> of
Bash).</P
></LI
><LI
><P
><A
HREF="#PW0"
><I
CLASS="FIRSTTERM"
>Script
example</I
></A
></P
></LI
><LI
><P
><A
HREF="#PARSUBTAB"
><B
CLASS="COMMAND"
>Table</B
></A
>
of <I
CLASS="FIRSTTERM"
>parameter
substitution</I
></P
></LI
></UL
></P
><P
><A
HREF="#PARCHILDPROBREF"
>Parent / child process
problem</A
>, a <I
CLASS="FIRSTTERM"
>child</I
> process cannot
<A
HREF="#EXPORTREF"
>export</A
> variables to a <A
HREF="#FORKREF"
>parent process</A
></P
><P
>Parentheses
<P
></P
><UL
><LI
><P
><A
HREF="#PARENSREF"
>Command
group</A
></P
></LI
><LI
><P
><A
HREF="#PARENGRPS"
>Enclose group</A
>
of <I
CLASS="FIRSTTERM"
>Extended Regular
Expressions</I
></P
></LI
><LI
><P
><A
HREF="#DBLPARENSREF"
>Double parentheses</A
>,
in arithmetic expansion</P
></LI
></UL
></P
><P
><A
HREF="#PATHREF"
><TT
CLASS="VARNAME"
>$PATH</TT
></A
>,
the <I
CLASS="FIRSTTERM"
>path</I
> (location of system
binaries)
<P
></P
><UL
><LI
><P
>Appending directories to <TT
CLASS="VARNAME"
>$PATH</TT
>
<A
HREF="#PATHAPPEND"
>using the <TT
CLASS="VARNAME"
>+=</TT
>
operator</A
>.</P
></LI
></UL
></P
><P
><A
HREF="#PATHNAMEREF"
>Pathname</A
>,
a <TT
CLASS="FILENAME"
>filename</TT
> that incorporates the complete
<I
CLASS="FIRSTTERM"
>path</I
> of a given file.
<P
></P
><UL
><LI
><P
><A
HREF="#PATHMANAGEMENT"
>Parsing
<I
CLASS="FIRSTTERM"
>pathnames</I
></A
></P
></LI
></UL
></P
><P
><A
HREF="#PERLREF"
>Perl</A
>, programming language
<P
></P
><UL
><LI
><P
><A
HREF="#BASHANDPERL0"
>Combined</A
> in the
same file with a <I
CLASS="FIRSTTERM"
>Bash</I
> script</P
></LI
><LI
><P
><A
HREF="#PERLEMB"
>Embedded</A
> in a
<I
CLASS="FIRSTTERM"
>Bash</I
> script</P
></LI
></UL
></P
><P
><A
HREF="#QKY"
><EM
>Perquackey</EM
>-type
anagramming game</A
> (<EM
>Quackey</EM
> script)</P
><P
><A
HREF="#PETALS"
><EM
>Petals Around the
Rose</EM
></A
></P
><P
><A
HREF="#PROCESSIDDEF"
>PID</A
>,
<I
CLASS="FIRSTTERM"
>Process ID</I
>, an identification
number assigned to a running process.</P
><P
><A
HREF="#PIPEREF"
>Pipe</A
>, <B
CLASS="COMMAND"
>| </B
>,
a device for passing the output of a command to another command
or to the shell
<P
></P
><UL
><LI
><P
><A
HREF="#CATABUSE"
>Avoiding
unnecessary commands</A
> in a
<I
CLASS="FIRSTTERM"
>pipe</I
></P
></LI
><LI
><P
><A
HREF="#COMMINPIPE"
><I
CLASS="FIRSTTERM"
>Comments</I
> embedded
within</A
></P
></LI
><LI
><P
><A
HREF="#PIPEEX"
>Exit status</A
>
of a pipe</P
></LI
><LI
><P
><A
HREF="#PIPEFAILREF"
>Pipefail</A
>,
<I
CLASS="FIRSTTERM"
>set -o pipefail</I
>
option to indicate <A
HREF="#EXITSTATUSREF"
>exit status</A
>
within a <I
CLASS="FIRSTTERM"
>pipe</I
></P
></LI
><LI
><P
><A
HREF="#PIPESTATUSREF"
><TT
CLASS="VARNAME"
>$PIPESTATUS</TT
></A
>,
<I
CLASS="FIRSTTERM"
>exit status</I
> of last executed
pipe</P
></LI
><LI
><P
><A
HREF="#UCREF"
>Piping output of a command</A
>
to a script</P
></LI
><LI
><P
><A
HREF="#CATLESSEFF"
>Redirecting
<TT
CLASS="FILENAME"
>stdin</TT
></A
>, rather than using
<A
HREF="#CATREF"
>cat</A
> in a
<I
CLASS="FIRSTTERM"
>pipe</I
></P
></LI
></UL
></P
><P
><A
HREF="#GOTCHAS"
>Pitfalls</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#DASHNREDR"
><B
CLASS="COMMAND"
>-</B
>
(dash) is <EM
>not</EM
> redirection
operator</A
></P
></LI
><LI
><P
><A
HREF="#DOUBLESLASHREF"
><B
CLASS="COMMAND"
>//
</B
> (double forward slash)</A
>, behavior of <A
HREF="#CDREF"
>cd</A
> command toward</P
></LI
><LI
><P
><A
HREF="#BINSH"
><SPAN
CLASS="TOKEN"
>#!/bin/sh</SPAN
></A
>
script header disables <A
HREF="#BASHCOMPAT"
>extended <I
CLASS="FIRSTTERM"
>Bash</I
>
features</A
></P
></LI
><LI
><P
><A
HREF="#CATABUSE"
>Abuse of
<I
CLASS="FIRSTTERM"
>cat</I
></A
></P
></LI
><LI
><P
><A
HREF="#CGIREF"
><I
CLASS="FIRSTTERM"
>CGI</I
>
programming</A
>, using scripts for</P
></LI
><LI
><P
>Closing <I
CLASS="FIRSTTERM"
>limit string</I
>
in a <I
CLASS="FIRSTTERM"
>here document</I
>,
<A
HREF="#INDENTEDLS"
>indenting</A
></P
></LI
><LI
><P
><A
HREF="#DOSNEWLINES"
>DOS-type newlines
(<SPAN
CLASS="TOKEN"
>\r\n</SPAN
>)</A
> crash a script</P
></LI
><LI
><P
><A
HREF="#QUOTINGBSL"
><I
CLASS="FIRSTTERM"
>Double-quoting</I
>
the <I
CLASS="FIRSTTERM"
>backslash</I
> (<B
CLASS="COMMAND"
>\</B
>)
character</A
></P
></LI
><LI
><P
><A
HREF="#EVALRISK"
>eval</A
>, risk of
using</P
></LI
><LI
><P
><A
HREF="#EXECPERM"
>Execute permission
lacking</A
> for commands within a script</P
></LI
><LI
><P
> <I
CLASS="FIRSTTERM"
>Exit status</I
>,
<A
HREF="#GOTCHAEXITVALANAMALIES"
>anomalous</A
>
</P
></LI
><LI
><P
> <I
CLASS="FIRSTTERM"
>Exit status</I
>
<A
HREF="#ARXS1"
>of arithmetic expression <EM
>not</EM
>
equivalent to an <I
CLASS="FIRSTTERM"
>error code</I
></A
>
</P
></LI
><LI
><P
><A
HREF="#PARCHILDPROBREF"
><I
CLASS="FIRSTTERM"
>Export</I
>
problem</A
>, <I
CLASS="FIRSTTERM"
>child</I
> process
to <I
CLASS="FIRSTTERM"
>parent</I
> process</P
></LI
><LI
><P
><A
HREF="#LATEVERF"
>Extended
<I
CLASS="FIRSTTERM"
>Bash</I
> features</A
> not
available</P
></LI
><LI
><P
><A
HREF="#FAILQUOTE"
>Failing to
<I
CLASS="FIRSTTERM"
>quote</I
> variables</A
>
within <I
CLASS="FIRSTTERM"
>test</I
> brackets</P
></LI
><LI
><P
><A
HREF="#GNUREF"
><I
CLASS="FIRSTTERM"
>GNU</I
>
command set</A
>, in cross-platform scripts</P
></LI
><LI
><P
><I
CLASS="FIRSTTERM"
>let</I
> misuse:
<A
HREF="#LETBAD"
>attempting to set string variables</A
>
</P
></LI
><LI
><P
><A
HREF="#RVTCAUTION2"
>Multiple echo
statements</A
> in a <A
HREF="#RVT"
>function whose
output is captured</A
></P
></LI
><LI
><P
><A
HREF="#NULLVAR"
><I
CLASS="FIRSTTERM"
>null</I
>
variable assignment</A
></P
></LI
><LI
><P
><A
HREF="#NUMSTRCOMPNE"
>Numerical and string
comparison operators</A
> <EM
>not</EM
>
equivalent</P
><P
><A
HREF="#EQDIF"
><B
CLASS="COMMAND"
>=</B
>
and <B
CLASS="COMMAND"
>-eq</B
></A
> <EM
>not</EM
>
interchangeable</P
></LI
><LI
><P
><A
HREF="#OMITSEMICOLON"
>Omitting
terminal <I
CLASS="FIRSTTERM"
>semicolon</I
></A
>,
in a <I
CLASS="FIRSTTERM"
>curly-bracketed</I
> <A
HREF="#CODEBLOCKREF"
>code block</A
></P
></LI
><LI
><P
>Piping</P
><P
><A
HREF="#PIPELOOP"
><I
CLASS="FIRSTTERM"
>echo</I
>
to a loop</A
></P
><P
><A
HREF="#BADREAD0"
><I
CLASS="FIRSTTERM"
>echo</I
>
to <I
CLASS="FIRSTTERM"
>read</I
></A
> (however, this problem
<A
HREF="#GOODREAD0"
>can be circumvented</A
>)</P
><P
><A
HREF="#PTAILGREP"
><I
CLASS="FIRSTTERM"
>tail</I
>
<TT
CLASS="OPTION"
>-f</TT
> to <I
CLASS="FIRSTTERM"
>grep</I
></A
></P
></LI
><LI
><P
>Preserving <I
CLASS="FIRSTTERM"
>whitespace</I
>
within a variable, <A
HREF="#VARSPLITTING"
>unintended
consequences</A
></P
></LI
><LI
><P
><A
HREF="#SUIDSCR"
><I
CLASS="FIRSTTERM"
>suid</I
>
commands inside a script</A
></P
></LI
><LI
><P
><A
HREF="#UNDOCF"
>Undocumented
<I
CLASS="FIRSTTERM"
>Bash</I
> features</A
>,
danger of</P
></LI
><LI
><P
>Updates to <I
CLASS="FIRSTTERM"
>Bash</I
>
<A
HREF="#UPDATEBREAKS"
>breaking older
scripts</A
></P
></LI
><LI
><P
><A
HREF="#UNINITVAR"
>Uninitialized
variables</A
></P
></LI
><LI
><P
><A
HREF="#INAPPVN"
>Variable names</A
>,
inappropriate</P
></LI
><LI
><P
><A
HREF="#VARSUBSH"
>Variables in a
<I
CLASS="FIRSTTERM"
>subshell</I
></A
>, <I
CLASS="FIRSTTERM"
>scope</I
>
limited</P
></LI
><LI
><P
><A
HREF="#BADREAD0"
>Subshell in <I
CLASS="FIRSTTERM"
>while-read</I
>
loop</A
></P
></LI
><LI
><P
><A
HREF="#WSBAD"
>Whitespace</A
>,
misuse of</P
></LI
></UL
></P
><P
>Pointers</P
><P
></P
><UL
><LI
><P
><A
HREF="#FDREF1"
>and file descriptors</A
></P
></LI
><LI
><P
><A
HREF="#FUNCPOINTERS"
>and functions</A
></P
></LI
><LI
><P
><A
HREF="#IRRREF"
>and <I
CLASS="FIRSTTERM"
>indirect
references</I
></A
></P
></LI
><LI
><P
><A
HREF="#POINTERREF"
>and
<I
CLASS="FIRSTTERM"
>variables</I
></A
></P
></LI
></UL
><P
><A
HREF="#PORTABILITYISSUES"
>Portability issues</A
>
in shell scripting
<P
></P
><UL
><LI
><P
><A
HREF="#SETPUM"
>Setting <I
CLASS="FIRSTTERM"
>path</I
>
and <I
CLASS="FIRSTTERM"
>umask</I
></A
></P
></LI
><LI
><P
><A
HREF="#TESTSUITE0"
>A <I
CLASS="FIRSTTERM"
>test suite</I
>
script</A
> (Bash versus classic Bourne
shell)</P
></LI
><LI
><P
><A
HREF="#WHATISREF3"
>Using <I
CLASS="FIRSTTERM"
>whatis</I
></A
>
</P
></LI
></UL
></P
><P
><A
HREF="#POSPARAMREF1"
>Positional parameters</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#APPREF2"
><TT
CLASS="VARNAME"
>$@</TT
></A
>,
as <I
CLASS="FIRSTTERM"
>separate</I
> words</P
></LI
><LI
><P
><A
HREF="#APPREF"
><TT
CLASS="VARNAME"
>$*</TT
></A
>,
as a <I
CLASS="FIRSTTERM"
>single</I
> word</P
></LI
><LI
><P
><A
HREF="#PASSEDARGS"
>in
functions</A
></P
></LI
></UL
></P
><P
><A
HREF="#POSIX2REF"
><SPAN
CLASS="ACRONYM"
>POSIX</SPAN
></A
>,
<I
CLASS="FIRSTTERM"
>Portable Operating System Interface /
UNIX</I
>
<P
></P
><UL
><LI
><P
><A
HREF="#POSIX3REF"
><TT
CLASS="OPTION"
>--posix</TT
>
option</A
></P
></LI
><LI
><P
><A
HREF="#POSIX3REF"
>1003.2 standard</A
></P
></LI
><LI
><P
><A
HREF="#POSIXREF"
>Character
classes</A
></P
></LI
></UL
></P
><P
><A
HREF="#PPIDREF"
><TT
CLASS="VARNAME"
>$PPID</TT
></A
>,
<I
CLASS="FIRSTTERM"
>process ID</I
> of parent process</P
><P
><A
HREF="#OPPRECEDENCE1"
>Precedence</A
>, operator</P
><P
><A
HREF="#PREPENDREF"
><I
CLASS="FIRSTTERM"
>Prepending</I
></A
>
lines at head of a file, <I
CLASS="FIRSTTERM"
>script
example</I
></P
><P
>Prime numbers
<P
></P
><UL
><LI
><P
>Generating primes
<A
HREF="#PRIMES2"
>using the <I
CLASS="FIRSTTERM"
>factor</I
>
command</A
></P
></LI
><LI
><P
>Generating primes
<A
HREF="#PRIMES1"
>using the <I
CLASS="FIRSTTERM"
>modulo</I
>
operator</A
></P
></LI
><LI
><P
>Sieve of Eratosthenes, <A
HREF="#PRIMES0"
>example script</A
></P
></LI
></UL
></P
><P
><A
HREF="#PRINTFREF"
>printf</A
>,
<I
CLASS="FIRSTTERM"
>formatted print</I
> command</P
><P
><A
HREF="#PROCREF2"
><TT
CLASS="FILENAME"
>/proc</TT
></A
> directory
<P
></P
><UL
><LI
><P
><A
HREF="#PROCRUNNING"
>Running processes</A
>,
files describing</P
></LI
><LI
><P
><A
HREF="#PROCWARNING"
>Writing to
files in <TT
CLASS="FILENAME"
>/proc</TT
></A
>,
<EM
>warning</EM
></P
></LI
></UL
></P
><P
><A
HREF="#PROCESSREF"
>Process</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#CHILDREF2"
>Child
process</A
></P
></LI
><LI
><P
><A
HREF="#PARENTREF"
>Parent
process</A
></P
></LI
><LI
><P
><A
HREF="#PROCESSIDDEF"
>Process
ID</A
> (PID)</P
></LI
></UL
></P
><P
><A
HREF="#PROCESSSUBREF"
>Process substitution</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#PCC2DIR"
>To compare contents of
directories</A
></P
></LI
><LI
><P
><A
HREF="#PSFDSTDIN"
>To supply <TT
CLASS="FILENAME"
>stdin</TT
> of a
command</A
></P
></LI
><LI
><P
><A
HREF="#COMMANDSPARENS1"
>Template</A
></P
></LI
><LI
><P
><A
HREF="#GOODREAD0"
><I
CLASS="FIRSTTERM"
>while-read</I
> loop
without a
<I
CLASS="FIRSTTERM"
>subshell</I
></A
></P
></LI
></UL
></P
><P
><A
HREF="#TABEXPANSION"
>Programmable completion</A
>
(tab expansion)</P
><P
>Prompt
<P
></P
><UL
><LI
><P
><A
HREF="#PS1REF"
><TT
CLASS="VARNAME"
>$PS1</TT
></A
>, <I
CLASS="FIRSTTERM"
>Main
prompt</I
>, seen at command
line</P
></LI
><LI
><P
><A
HREF="#SECPROMPTREF"
><TT
CLASS="VARNAME"
>$PS2</TT
></A
>,
Secondary prompt</P
></LI
></UL
></P
><P
><A
HREF="#PSEUDOCODEREF"
>Pseudo-code</A
>,
as problem-solving method</P
><P
><A
HREF="#PWDREF"
><TT
CLASS="VARNAME"
>$PWD</TT
></A
>,
Current working directory</P
><P
>* * *</P
><P
><A
HREF="#QKY"
>Quackey</A
>, a
<EM
>Perquackey</EM
>-type anagramming game (script)</P
><P
>Question mark, <B
CLASS="COMMAND"
>? </B
>
<P
></P
><UL
><LI
><P
><A
HREF="#QUEXREGEX"
>Character
match</A
> in an Extended <I
CLASS="FIRSTTERM"
>Regular
Expression</I
></P
></LI
><LI
><P
><A
HREF="#QUEXWC"
>Single-character
<I
CLASS="FIRSTTERM"
>wild card</I
></A
>,
in <A
HREF="#GLOBBINGREF"
>globbing</A
></P
></LI
><LI
><P
>In a
<A
HREF="#CSTRINARY"
><I
CLASS="FIRSTTERM"
>C</I
>-style
Trinary (ternary) operator</A
></P
></LI
></UL
></P
><P
><A
HREF="#QUOTINGDEF"
>Quoting</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#QUOTINGREF"
>Character
string</A
></P
></LI
><LI
><P
><A
HREF="#QUOTINGVAR"
>Variables</A
></P
><P
><A
HREF="#FAILQUOTE"
>within <I
CLASS="FIRSTTERM"
>test</I
>
brackets</A
></P
></LI
><LI
><P
><A
HREF="#WSQUO"
><I
CLASS="FIRSTTERM"
>Whitespace</I
></A
>,
using <I
CLASS="FIRSTTERM"
>quoting</I
> to
preserve</P
></LI
></UL
></P
><P
>* * *</P
><P
>Random numbers
<P
></P
><UL
><LI
><P
><A
HREF="#URANDOMREF"
><TT
CLASS="FILENAME"
>/dev/urandom</TT
></A
>
</P
></LI
><LI
><P
><A
HREF="#AWKRANDOMREF"
> <TT
CLASS="VARNAME"
>rand()</TT
></A
>,
random function in <A
HREF="#AWKREF"
>awk</A
></P
></LI
><LI
><P
><A
HREF="#RANDOMVAR01"
><TT
CLASS="VARNAME"
>$RANDOM</TT
></A
>, Bash
function that returns a pseudorandom integer</P
></LI
><LI
><P
><A
HREF="#DATERANDREF"
>Random sequence
generation</A
>, using <A
HREF="#DATEREF"
>date</A
>
command</P
></LI
><LI
><P
><A
HREF="#JOTRANDOM"
>Random sequence
generation</A
>, using <A
HREF="#JOTREF"
>jot</A
></P
></LI
><LI
><P
><A
HREF="#RANDSTRING0"
>Random string</A
>,
generating</P
></LI
></UL
></P
><P
>Raspberry Pi (single-board computer)
<P
></P
><UL
><LI
><P
><A
HREF="#RPSDCARD01"
>Script for preparing a bootable SD card</A
>
</P
></LI
></UL
></P
><P
><A
HREF="#RCSREF"
>rcs</A
></P
><P
><A
HREF="#READREF"
>read</A
>, set value of a
variable from <A
HREF="#STDINOUTDEF"
><TT
CLASS="FILENAME"
>stdin</TT
></A
>
<P
></P
><UL
><LI
><P
><A
HREF="#READARROW"
>Detecting
<I
CLASS="FIRSTTERM"
>arrow</I
> keys</A
></P
></LI
><LI
><P
><A
HREF="#READOPTIONS"
>Options</A
></P
></LI
><LI
><P
><A
HREF="#READPIPEREF"
>Piping
output of <I
CLASS="FIRSTTERM"
>cat</I
></A
>
to <I
CLASS="FIRSTTERM"
>read</I
></P
></LI
><LI
><P
><A
HREF="#HSREAD"
><SPAN
CLASS="QUOTE"
>"Prepending"</SPAN
>
text</A
></P
></LI
><LI
><P
><A
HREF="#BADREAD0"
> Problems piping <I
CLASS="FIRSTTERM"
>echo</I
></A
>
to <I
CLASS="FIRSTTERM"
>read</I
></P
></LI
><LI
><P
><A
HREF="#READREDIR0"
>Redirection from a
file</A
> to <I
CLASS="FIRSTTERM"
>read</I
></P
></LI
><LI
><P
><A
HREF="#REPLYREF"
><TT
CLASS="VARNAME"
>$REPLY</TT
></A
>,
default <I
CLASS="FIRSTTERM"
>read</I
> variable</P
></LI
><LI
><P
><A
HREF="#READTIMED"
>Timed
input</A
></P
></LI
><LI
><P
><A
HREF="#WHILEREADREF2"
> <I
CLASS="FIRSTTERM"
>while read</I
></A
>
construct</P
></LI
></UL
></P
><P
><A
HREF="#READLINEREF"
>readline</A
> library</P
><P
><A
HREF="#RECURSIONREF"
>Recursion</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#RECURSIONDEMO0"
>Demonstration of</A
></P
></LI
><LI
><P
><A
HREF="#FACTORIALREF"
>Factorial</A
></P
></LI
><LI
><P
><A
HREF="#FIBOREF"
>Fibonacci
sequence</A
></P
></LI
><LI
><P
><A
HREF="#LOCVARRECUR"
>Local
variables</A
></P
></LI
><LI
><P
><A
HREF="#SCRIPTRECURSION"
>Script
calling itself recursively</A
></P
></LI
><LI
><P
><A
HREF="#HANOIREF"
>Towers
of Hanoi</A
></P
></LI
></UL
></P
><P
>Redirection
<P
></P
><UL
><LI
><P
><A
HREF="#REDIRREF"
>Code
blocks</A
></P
></LI
><LI
><P
><A
HREF="#USINGEXECREF"
>exec
&#60;<TT
CLASS="FILENAME"
>filename</TT
></A
>,</P
><P
>to reassign
<A
HREF="#FDREF"
>file descriptors</A
></P
></LI
><LI
><P
><A
HREF="#IOREDIRINTRO"
>Introductory-level
explanation</A
> of <I
CLASS="FIRSTTERM"
>I/O
redirection</I
></P
></LI
><LI
><P
><A
HREF="#IOREDIRECTIONREF2"
> Open a file</A
> for <EM
>both</EM
>
reading and writing</P
><P
><TT
CLASS="FILENAME"
>&#60;&#62;filename</TT
>
</P
></LI
><LI
><P
><A
HREF="#READREDIR0"
><I
CLASS="FIRSTTERM"
>read</I
>
input redirected</A
> from a file</P
></LI
><LI
><P
> <A
HREF="#IOREDIRECTIONREF1"
> <TT
CLASS="FILENAME"
>stderr</TT
> to
<TT
CLASS="FILENAME"
>stdout</TT
></A
></P
><P
><TT
CLASS="FILENAME"
>2&#62;&#38;1 </TT
></P
></LI
><LI
><P
><A
HREF="#COXEX"
><TT
CLASS="FILENAME"
>stdin</TT
> /
<TT
CLASS="FILENAME"
>stdout</TT
></A
>,
using <B
CLASS="COMMAND"
> - </B
></P
></LI
><LI
><P
><A
HREF="#REDSTDINFUNC1"
> <TT
CLASS="FILENAME"
>stdin</TT
>of a
<I
CLASS="FIRSTTERM"
>function</I
></A
></P
></LI
><LI
><P
><A
HREF="#IOREDIRECTIONREF"
> <TT
CLASS="FILENAME"
>stdout</TT
> to a file</A
></P
><P
><TT
CLASS="REPLACEABLE"
><I
>&#62;</I
></TT
> ... <TT
CLASS="REPLACEABLE"
><I
>&#62;&#62;</I
></TT
>
</P
></LI
><LI
><P
> <A
HREF="#IOREDIRECTIONREF1"
> <TT
CLASS="FILENAME"
>stdout</TT
> to
<I
CLASS="FIRSTTERM"
>file descriptor</I
></A
>
<I
CLASS="FIRSTTERM"
>j</I
></P
><P
><TT
CLASS="FILENAME"
>&#62;&#38;j </TT
></P
></LI
><LI
><P
> <A
HREF="#IOREDIRECTIONREF1"
> file descriptor<TT
CLASS="FILENAME"
>i</TT
> to
<I
CLASS="FIRSTTERM"
>file descriptor</I
></A
>
<I
CLASS="FIRSTTERM"
>j</I
></P
><P
><TT
CLASS="FILENAME"
>i&#62;&#38;j </TT
></P
></LI
><LI
><P
> <A
HREF="#REDIROUTERROR2"
><TT
CLASS="FILENAME"
>stdout</TT
>
of a command</A
> to <TT
CLASS="FILENAME"
>stderr</TT
></P
><P
><TT
CLASS="FILENAME"
>&#62;&#38;2</TT
></P
></LI
><LI
><P
> <A
HREF="#REDIROUTERROR"
><TT
CLASS="FILENAME"
>stdout</TT
>
<EM
>and</EM
> <TT
CLASS="FILENAME"
>stderr</TT
>
of a command</A
> to a file</P
><P
><TT
CLASS="FILENAME"
>&#38;&#62; </TT
></P
></LI
><LI
><P
> <A
HREF="#TEEREF"
>tee</A
>, redirect to a file
output of command(s) partway through a <A
HREF="#PIPEREF"
>pipe</A
></P
></LI
></UL
></P
><P
><A
HREF="#REFCARDS"
>Reference Cards</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#MISCTAB"
>Miscellaneous
constructs</A
></P
></LI
><LI
><P
><A
HREF="#PARSUBTAB"
>Parameter
substitution/expansion</A
></P
></LI
><LI
><P
><A
HREF="#SPECSHVARTAB"
>Special shell
variables</A
></P
></LI
><LI
><P
><A
HREF="#STRINGOPSTAB"
>String
operations</A
></P
></LI
><LI
><P
>Test operators</P
><P
><A
HREF="#BINCOMPTAB"
>Binary comparison</A
></P
><P
><A
HREF="#FILESTAB"
>Files</A
></P
></LI
></UL
></P
><P
><A
HREF="#REGEXREF"
><I
CLASS="FIRSTTERM"
>Regular
Expressions</I
></A
>
<P
></P
><UL
><LI
><P
><B
CLASS="COMMAND"
>^ </B
> (caret)
<A
HREF="#BEGLINEREF"
>Beginning-of-line</A
>
</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>$ </B
> (dollar sign)
<A
HREF="#DOLLARSIGNREF"
><I
CLASS="FIRSTTERM"
>Anchor</I
></A
>
</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>. </B
> (dot)
<A
HREF="#REGEXDOT"
>Match single
character</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>* </B
> (asterisk)
<A
HREF="#ASTERISKREF2"
>Any number of
characters</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>[ ]</B
> (brackets) <A
HREF="#BRACKETSREF"
>Enclose character set
to match</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>\ </B
> (backslash) <A
HREF="#REGEXBS"
>Escape</A
>, interpret following
character literally</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>\&#60; ... \&#62; </B
>
(angle brackets, <I
CLASS="FIRSTTERM"
>escaped</I
>)
<A
HREF="#ANGLEBRAC"
>Word boundary</A
></P
></LI
><LI
><P
><A
HREF="#EXTREGEX"
>Extended</A
>
REs</P
><P
><B
CLASS="COMMAND"
>+ </B
>
<A
HREF="#PLUSREF"
><I
CLASS="FIRSTTERM"
>Character
match</I
></A
></P
><P
><B
CLASS="COMMAND"
>\{ \} </B
> <A
HREF="#ESCPCB"
>Escaped
<SPAN
CLASS="QUOTE"
>"curly"</SPAN
> brackets</A
></P
><P
><B
CLASS="COMMAND"
>[: :] </B
> <A
HREF="#POSIXREF"
>POSIX character classes</A
></P
></LI
></UL
></P
><P
><A
HREF="#REPLYREF"
><TT
CLASS="VARNAME"
>$REPLY</TT
></A
>,
Default value associated with <A
HREF="#READREF"
>read</A
>
command</P
><P
><A
HREF="#RESTRICTEDSHREF"
>Restricted shell</A
>,
shell (or script) with certain commands disabled</P
><P
><A
HREF="#RETURNREF"
>return</A
>,
command that terminates a <A
HREF="#FUNCTIONREF"
>function</A
></P
><P
><A
HREF="#RUNPARTSREF"
>run-parts</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#RUNPARTSREF2"
>Running scripts
in sequence</A
>, without user intervention</P
></LI
></UL
></P
><P
>* * *</P
><P
><A
HREF="#SCOPEREF"
>Scope</A
> of a variable,
definition</P
><P
><A
HREF="#INVOCATIONOPTIONSREF"
>Script options</A
>,
set at command line</P
><P
><A
HREF="#LIBROUTINES"
>Scripting routines</A
>,
library of useful definitions and <A
HREF="#FUNCTIONREF"
>functions</A
></P
><P
><A
HREF="#SECPROMPTREF"
>Secondary prompt</A
>,
<B
CLASS="COMMAND"
><TT
CLASS="VARNAME"
>$PS2</TT
></B
></P
><P
><A
HREF="#SECURITYISSUES"
>Security issues</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#NMAPREF"
>nmap</A
>,
<I
CLASS="FIRSTTERM"
>network mapper</I
>
/ port scanner</P
></LI
><LI
><P
><A
HREF="#SUDOREF"
>sudo</A
></P
></LI
><LI
><P
><A
HREF="#SUIDSCR"
><I
CLASS="FIRSTTERM"
>suid</I
>
commands inside a script</A
></P
></LI
><LI
><P
><A
HREF="#INFECTEDSCRIPTS1"
>Viruses,
trojans, and worms</A
> in scripts</P
></LI
><LI
><P
><A
HREF="#SECURITYTIPS1"
>Writing
secure scripts</A
></P
></LI
></UL
></P
><P
><A
HREF="#SEDREF"
>sed</A
>,
pattern-based programming language
<P
></P
><UL
><LI
><P
><A
HREF="#SEDBASICTABLE"
><B
CLASS="COMMAND"
>Table</B
></A
>,
basic operators</P
></LI
><LI
><P
><A
HREF="#SEDOPTABLE"
><B
CLASS="COMMAND"
>Table</B
></A
>,
examples of operators</P
></LI
></UL
></P
><P
><A
HREF="#SELECTREF"
>select</A
>,
construct for menu building
<P
></P
><UL
><LI
><P
><A
HREF="#INLISTOMIT"
><TT
CLASS="USERINPUT"
><B
>in
<TT
CLASS="REPLACEABLE"
><I
>list</I
></TT
></B
></TT
> omitted</A
>
</P
></LI
></UL
></P
><P
><A
HREF="#SEMAPHOREREF"
>Semaphore</A
></P
><P
><A
HREF="#NEEDSEMICOLON"
>Semicolon required</A
>,
when <A
HREF="#DOINREF"
>do</A
>
<I
CLASS="FIRSTTERM"
>keyword</I
> is on first line of <A
HREF="#FORLOOPREF1"
>loop</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#OMITSEMICOLON"
>When
terminating <I
CLASS="FIRSTTERM"
>curly-bracketed</I
>
code block</A
></P
></LI
></UL
></P
><P
><A
HREF="#SEQREF"
>seq</A
>,
Emit a sequence of integers. Equivalent to <A
HREF="#JOTREF"
>jot</A
>.</P
><P
><A
HREF="#SETREF"
>set</A
>,
Change value of internal script variables
<P
></P
><UL
><LI
><P
><A
HREF="#UNDVARERR"
>set -u</A
>,
Abort script with error message if attempting to use
an <I
CLASS="FIRSTTERM"
>undeclared</I
> variable.</P
></LI
></UL
>
</P
><P
><A
HREF="#WHATSASCRIPT"
>Shell script</A
>,
definition of</P
><P
><A
HREF="#SHWRAPPER"
>Shell wrapper</A
>,
script embedding a command or utility</P
><P
><A
HREF="#SHIFTREF"
>shift</A
>, reassigning
<I
CLASS="FIRSTTERM"
>positional parameters</I
></P
><P
><A
HREF="#SHLVLREF"
><TT
CLASS="VARNAME"
>$SHLVL</TT
></A
>,
<I
CLASS="FIRSTTERM"
>shell level</I
>, depth to which the shell
(or script) is nested</P
><P
><A
HREF="#SHOPTREF"
>shopt</A
>,
change <I
CLASS="FIRSTTERM"
>shell options</I
></P
><P
><A
HREF="#SIGNALD"
>Signal</A
>,
a message sent to a process</P
><P
>Simulations
<P
></P
><UL
><LI
><P
><A
HREF="#BROWNIANREF"
>Brownian
motion</A
></P
></LI
><LI
><P
><A
HREF="#BROWNIANREF"
>Galton
board</A
></P
></LI
><LI
><P
><A
HREF="#HORSERACEREF"
>Horserace</A
></P
></LI
><LI
><P
><A
HREF="#LIFEREF"
><EM
>Life</EM
></A
>,
game of</P
></LI
><LI
><P
><A
HREF="#CANNONREF"
>PI</A
>,
approximating by firing cannonballs</P
></LI
><LI
><P
><A
HREF="#STACKEX0"
>Pushdown
<I
CLASS="FIRSTTERM"
>stack</I
></A
></P
></LI
></UL
></P
><P
><A
HREF="#SNGLQUO"
>Single quotes</A
>
(<B
CLASS="COMMAND"
>' ... '</B
>) <I
CLASS="FIRSTTERM"
>strong</I
> <A
HREF="#QUOTINGREF"
>quoting</A
></P
><P
><A
HREF="#SOCKETREF"
>Socket</A
>, a communication
node associated with an I/O port</P
><P
>Sorting
<P
></P
><UL
><LI
><P
><A
HREF="#BUBBLESORT"
>Bubble
sort</A
></P
></LI
><LI
><P
><A
HREF="#INSERTIONSORT0"
>Insertion
sort</A
></P
></LI
></UL
></P
><P
><A
HREF="#SOURCEREF"
>source</A
>,
execute a script or, within a script, import a file
<P
></P
><UL
><LI
><P
><A
HREF="#SOURCEPARAMS"
>Passing positional
parameters</A
></P
></LI
></UL
></P
><P
><I
CLASS="FIRSTTERM"
>Spam</I
>, dealing with
<P
></P
><UL
><LI
><P
><A
HREF="#SPAMLOOKUP_0"
><I
CLASS="FIRSTTERM"
>Example
script</I
></A
></P
></LI
><LI
><P
><A
HREF="#ISSPAMMER_0"
><I
CLASS="FIRSTTERM"
>Example
script</I
></A
></P
></LI
><LI
><P
><A
HREF="#ISSPAMMER2_0"
><I
CLASS="FIRSTTERM"
>Example
script</I
></A
></P
></LI
><LI
><P
><A
HREF="#WHX0"
><I
CLASS="FIRSTTERM"
>Example
script</I
></A
></P
></LI
></UL
></P
><P
><A
HREF="#SCHARLIST1"
>Special characters</A
></P
><P
>Stack
<P
></P
><UL
><LI
><P
><A
HREF="#STACKDEFREF"
>Definition</A
></P
></LI
><LI
><P
>Emulating a <I
CLASS="FIRSTTERM"
>push-down stack</I
>,
<A
HREF="#STACKEX0"
>example script</A
></P
></LI
></UL
></P
><P
>Standard Deviation, <A
HREF="#STDDEV"
>example script</A
></P
><P
><A
HREF="#FILESREF1"
>Startup files</A
>, Bash</P
><P
><A
HREF="#STDINOUTDEF"
><TT
CLASS="FILENAME"
>stdin</TT
>
and <TT
CLASS="FILENAME"
>stdout</TT
></A
></P
><P
><A
HREF="#STOPWATCH"
>Stopwatch</A
>,
example script</P
><P
>Strings
<P
></P
><UL
><LI
><P
> <B
CLASS="COMMAND"
>=~ </B
>
<A
HREF="#REGEXMATCHREF"
>String match operator</A
>
</P
></LI
><LI
><P
><A
HREF="#SCOMPARISON1"
>Comparison</A
></P
></LI
><LI
><P
><A
HREF="#PSOREX1"
>Length</A
></P
><P
><TT
CLASS="REPLACEABLE"
><I
>${#string}</I
></TT
></P
></LI
><LI
><P
><A
HREF="#STRINGMANIP"
>Manipulation</A
></P
></LI
><LI
><P
><A
HREF="#AWKSTRINGMANIP2"
>Manipulation</A
>,
using <A
HREF="#AWKREF"
>awk</A
></P
></LI
><LI
><P
><A
HREF="#STRINGNOTNULL"
><I
CLASS="FIRSTTERM"
>Null</I
>
string</A
>, testing for</P
></LI
><LI
><P
><A
HREF="#PROTECTLITERAL0"
>Protecting strings</A
>
from expansion and/or reinterpretation,
<I
CLASS="FIRSTTERM"
>script example</I
></P
><P
><A
HREF="#UNPROTECTLITERAL0"
><I
CLASS="FIRSTTERM"
>Unprotecting</I
>
strings</A
>,
<I
CLASS="FIRSTTERM"
>script example</I
></P
></LI
><LI
><P
><I
CLASS="FIRSTTERM"
>strchr()</I
>, <A
HREF="#SUBSTRINGINDEX2"
>equivalent of</A
></P
></LI
><LI
><P
><I
CLASS="FIRSTTERM"
>strlen()</I
>, <A
HREF="#STRLEN"
>equivalent of</A
></P
></LI
><LI
><P
><A
HREF="#STRINGSREF"
>strings</A
> command,
find printable strings in a binary or data file</P
></LI
><LI
><P
>Substring extraction</P
><P
><A
HREF="#SUBSTREXTR01"
>${string:position}</A
></P
><P
><A
HREF="#SUBSTREXTR02"
>${string:position:length}</A
></P
><P
><A
HREF="#EXPEXTRSUB"
>Using
<I
CLASS="FIRSTTERM"
>expr</I
></A
>
</P
></LI
><LI
><P
><A
HREF="#SUBSTRINGINDEX2"
>Substring
<I
CLASS="FIRSTTERM"
>index</I
></A
> (numerical position in
string)</P
></LI
><LI
><P
><A
HREF="#EXPRPAREN"
>Substring
<I
CLASS="FIRSTTERM"
>matching</I
></A
>, using
<A
HREF="#EXPRREF"
>expr</A
></P
></LI
><LI
><P
><A
HREF="#PSOREX1"
>Substring
<I
CLASS="FIRSTTERM"
>removal</I
></A
></P
><P
><A
HREF="#PSOREXSH"
>${var#Pattern}</A
></P
><P
><A
HREF="#PSOREXLO"
>${var##Pattern}</A
></P
><P
><A
HREF="#PCTREP1"
>${var%Pattern}</A
></P
><P
><A
HREF="#PCTREP2"
>${var%%Pattern}</A
></P
></LI
><LI
><P
>Substring replacement</P
><P
><A
HREF="#SUBSTRREPL00"
>${string/substring/replacement}</A
></P
><P
><A
HREF="#SUBSTRREPL01"
>${string//substring/replacement}</A
></P
><P
><A
HREF="#SUBSTRREPL02"
>${string/#substring/replacement}</A
></P
><P
><A
HREF="#SUBSTRREPL03"
>${string/%substring/replacement}</A
>
</P
><P
><A
HREF="#DAYSBETWEEN0"
><I
CLASS="FIRSTTERM"
>Script
example</I
></A
></P
></LI
><LI
><P
><A
HREF="#STRINGOPSTAB"
><B
CLASS="COMMAND"
>Table</B
></A
>
of <I
CLASS="FIRSTTERM"
>string/substring</I
> manipulation and
extraction operators</P
></LI
></UL
></P
><P
><A
HREF="#SNGLQUO"
><I
CLASS="FIRSTTERM"
>Strong</I
> quoting</A
>
<B
CLASS="COMMAND"
>' ... '</B
></P
><P
><A
HREF="#SCRSTYLE"
>Stylesheet</A
>
for writing scripts</P
><P
><A
HREF="#SUBSHELLSREF"
>Subshell</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#SUBSHELLPARENS1"
>Command list within
parentheses</A
></P
></LI
><LI
><P
><A
HREF="#SUBSHNLEVREF"
>Variables</A
>,
<TT
CLASS="VARNAME"
>$BASH_SUBSHELL</TT
> and
<TT
CLASS="VARNAME"
>$SHLVL</TT
></P
></LI
><LI
><P
>Variables in a <I
CLASS="FIRSTTERM"
>subshell</I
></P
><P
><A
HREF="#VARSUBSH"
><I
CLASS="FIRSTTERM"
>scope</I
>
limited</A
>, but ...</P
><P
> ... <A
HREF="#SUBSHTMP"
>can be accessed outside
the subshell?</A
>
</P
></LI
></UL
></P
><P
><A
HREF="#SUREF"
>su</A
> <I
CLASS="FIRSTTERM"
>Substitute
user</I
>, log on as a different user or as
<I
CLASS="FIRSTTERM"
>root</I
></P
><P
><A
HREF="#SUIDREF"
>suid</A
> (<I
CLASS="FIRSTTERM"
>set user
id</I
>) file flag
<P
></P
><UL
><LI
><P
><A
HREF="#SUIDSCR"
><I
CLASS="FIRSTTERM"
>suid</I
>
commands inside a script</A
>, not advisable</P
></LI
></UL
></P
><P
><A
HREF="#SYMLINKREF"
>Symbolic links</A
></P
><P
><A
HREF="#SWAPFILEREF"
>Swapfiles</A
></P
><P
>* * *</P
><P
><A
HREF="#TABEXPANSION"
>Tab completion</A
></P
><P
>Table lookup, <A
HREF="#RESISTOR"
>script
example</A
></P
><P
><A
HREF="#TAILREF"
>tail</A
>, <I
CLASS="FIRSTTERM"
>echo</I
>
to <TT
CLASS="FILENAME"
>stdout</TT
> lines at the (tail) end of a text
file</P
><P
><A
HREF="#TARREF"
>tar</A
>, archiving utility</P
><P
><A
HREF="#TEEREF"
>tee</A
>, redirect to a file
output of command(s) partway through a <A
HREF="#PIPEREF"
>pipe</A
></P
><P
><A
HREF="#TERMINALSSYS1"
>Terminals</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#SETSERIALREF"
>setserial</A
></P
></LI
><LI
><P
><A
HREF="#SETTERMREF"
>setterm</A
></P
></LI
><LI
><P
><A
HREF="#STTYREF"
>stty</A
></P
></LI
><LI
><P
><A
HREF="#TPUTREF"
>tput</A
></P
></LI
><LI
><P
><A
HREF="#WALLREF"
>wall</A
></P
></LI
></UL
></P
><P
><I
CLASS="FIRSTTERM"
>test</I
> command
<P
></P
><UL
><LI
><P
><A
HREF="#TTESTREF"
>Bash
<I
CLASS="FIRSTTERM"
>builtin</I
></A
></P
></LI
><LI
><P
><A
HREF="#USRBINTEST"
>external
command</A
>, <TT
CLASS="FILENAME"
>/usr/bin/test</TT
>
(equivalent to <TT
CLASS="FILENAME"
>/usr/bin/[</TT
>)</P
></LI
></UL
></P
><P
><A
HREF="#TESTCONSTRUCTS1"
>Test constructs</A
></P
><P
>Test operators
<P
></P
><UL
><LI
><P
><B
CLASS="COMMAND"
>-a </B
>
<A
HREF="#COMPOUNDAND"
>Logical AND</A
>
compound comparison</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-e </B
>
<A
HREF="#RTIF"
>File exists</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-eq </B
>
<A
HREF="#EQUALREF"
>is-equal-to</A
> (integer
comparison)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-f </B
>
<A
HREF="#REGULARFILE"
>File is a <I
CLASS="FIRSTTERM"
>regular</I
>
file</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-ge </B
>
<A
HREF="#GE0REF"
>greater-than or equal</A
>
(integer comparison)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-gt </B
>
<A
HREF="#GT0REF"
>greater-than</A
> (integer
comparison)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-le </B
>
<A
HREF="#LE0REF"
>less-than or equal</A
>
(integer comparison)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-lt </B
>
<A
HREF="#LT0REF"
>less-than</A
>
(integer comparison)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-n </B
>
<A
HREF="#STRINGNOTNULL"
>not-zero-length</A
>
(string comparison)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-ne </B
>
<A
HREF="#NEQUALREF"
>not-equal-to</A
>
(integer comparison)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-o </B
>
<A
HREF="#COMPOUNDOR"
>Logical OR</A
>
compound comparison</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-u </B
>
<A
HREF="#SUIDREF"
><I
CLASS="FIRSTTERM"
>suid</I
>
flag set</A
>, file test</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>-z </B
>
<A
HREF="#STRINGNULL"
>is-zero-length</A
>
(string comparison)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>= </B
>
<A
HREF="#SCOMPARISON1"
>is-equal-to</A
>
(string comparison)</P
><P
><B
CLASS="COMMAND"
>== </B
>
<A
HREF="#SCOMPARISON2"
>is-equal-to</A
>
(string comparison)</P
></LI
><LI
><P
> <B
CLASS="COMMAND"
>&#60; </B
>
<A
HREF="#LTREF"
>less-than</A
>
(string comparison)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>&#60; </B
>
<A
HREF="#INTLT"
>less-than</A
>,
(integer comparison, within
<A
HREF="#DBLPARENS"
>double parentheses</A
>)
</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>&#60;= </B
>
<A
HREF="#LTEQ"
>less-than-or-equal</A
>,
(integer comparison, within
<I
CLASS="FIRSTTERM"
>double parentheses</I
>)
</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>&#62; </B
>
<A
HREF="#GTREF"
>greater-than</A
>
(string comparison)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>&#62; </B
>
<A
HREF="#INTGT"
>greater-than</A
>,
(integer comparison, within
<I
CLASS="FIRSTTERM"
>double parentheses</I
>)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>&#62;= </B
>
<A
HREF="#GTEQ"
>greater-than-or-equal</A
>,
(integer comparison, within
<I
CLASS="FIRSTTERM"
>double parentheses</I
>)
</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>|| </B
>
<A
HREF="#ORREF"
>Logical OR </A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>&#38;&#38; </B
>
<A
HREF="#LOGICALAND"
>Logical AND</A
></P
></LI
><LI
><P
><B
CLASS="COMMAND"
>! </B
>
<A
HREF="#NOTREF"
>Negation operator</A
>, inverts <A
HREF="#EXITSTATUSREF"
>exit status</A
> of a
test</P
><P
><B
CLASS="COMMAND"
>!= </B
>
<A
HREF="#NOTEQUAL"
>not-equal-to</A
>
(string comparison)</P
></LI
><LI
><P
><B
CLASS="COMMAND"
>Tables</B
> of <I
CLASS="FIRSTTERM"
>test</I
>
operators</P
><P
><A
HREF="#BINCOMPTAB"
>Binary comparison</A
></P
><P
><A
HREF="#FILESTAB"
>File</A
></P
></LI
></UL
></P
><P
><A
HREF="#TEXTPROC"
>Text and text file
processing</A
></P
><P
><A
HREF="#TIMEDATE"
>Time / Date</A
></P
><P
>Timed input
<P
></P
><UL
><LI
><P
><A
HREF="#READTIMED"
>Using
<I
CLASS="FIRSTTERM"
>read -t</I
></A
></P
></LI
><LI
><P
><A
HREF="#STTYTO"
>Using
<I
CLASS="FIRSTTERM"
>stty</I
></A
></P
></LI
><LI
><P
><A
HREF="#TIMINGLOOP"
>Using
timing loop</A
></P
></LI
><LI
><P
><A
HREF="#TMOUTREF"
>Using
<TT
CLASS="VARNAME"
>$TMOUT</TT
></A
></P
></LI
></UL
></P
><P
><A
HREF="#ASSORTEDTIPS"
>Tips and hints</A
>
for Bash scripts
<P
></P
><UL
><LI
><P
>Array, <A
HREF="#RETARRAY"
>as
<I
CLASS="FIRSTTERM"
>return value</I
> from
a function</A
></P
><P
><I
CLASS="FIRSTTERM"
>Associative</I
> array
<A
HREF="#ASSOCARRTST"
>more efficient</A
>
than a numerically-indexed array</P
></LI
><LI
><P
><A
HREF="#CAPTURERETVAL"
>Capturing
the return value</A
> of a function, using
<I
CLASS="FIRSTTERM"
>echo</I
></P
></LI
><LI
><P
><A
HREF="#CGISCRIPT"
><I
CLASS="FIRSTTERM"
>CGI</I
>
programming</A
>, using scripts for</P
></LI
><LI
><P
>Comment blocks</P
><P
>Using <A
HREF="#CBLOCK1"
><I
CLASS="FIRSTTERM"
>anonymous
here documents</I
></A
></P
><P
>Using <A
HREF="#COMOUTBL"
> <I
CLASS="FIRSTTERM"
>if-then</I
>
constructs</A
></P
></LI
><LI
><P
><A
HREF="#COMMENTH"
>Comment headers</A
>,
special purpose</P
></LI
><LI
><P
><A
HREF="#CSTYLE"
><I
CLASS="FIRSTTERM"
>C</I
>-style syntax </A
>,
for manipulating variables</P
></LI
><LI
><P
><A
HREF="#DOUBLESPACE"
>Double-spacing a text
file</A
></P
></LI
><LI
><P
>Filenames prefixed with a dash,
<A
HREF="#DASHREM"
>removing</A
></P
></LI
><LI
><P
><A
HREF="#FILTEROUTP"
>Filter</A
>,
feeding output back to <EM
>same</EM
>
filter</P
></LI
><LI
><P
>Function<A
HREF="#RVT"
> <I
CLASS="FIRSTTERM"
>return</I
> value
workarounds</A
></P
></LI
><LI
><P
><A
HREF="#IFGREPFIX"
><I
CLASS="FIRSTTERM"
>if-grep</I
>
test fixup</A
></P
></LI
><LI
><P
><A
HREF="#LIBROUTINES"
>Library</A
>
of useful definitions and
<I
CLASS="FIRSTTERM"
>functions</I
></P
></LI
><LI
><P
><A
HREF="#NULLVAR"
><I
CLASS="FIRSTTERM"
>null</I
>
variable assignment</A
>, avoiding</P
></LI
><LI
><P
><A
HREF="#PASSARRAY"
>Passing
an <I
CLASS="FIRSTTERM"
>array</I
></A
> to a
function</P
></LI
><LI
><P
><TT
CLASS="VARNAME"
>$PATH</TT
>,
appending to, <A
HREF="#PATHAPPEND"
>using the
<TT
CLASS="VARNAME"
>+=</TT
> operator</A
>.</P
></LI
><LI
><P
><A
HREF="#PREPENDREF"
><I
CLASS="FIRSTTERM"
>Prepending</I
></A
>
lines at head of a file</P
></LI
><LI
><P
><A
HREF="#PROGRESSBAR"
>Progress
bar</A
> template</P
></LI
><LI
><P
><A
HREF="#PSEUDOCODEREF"
>Pseudo-code</A
></P
></LI
><LI
><P
><A
HREF="#RCSREF"
>rcs</A
></P
></LI
><LI
><P
><A
HREF="#DEVNULLREDIRECT"
>Redirecting a <I
CLASS="FIRSTTERM"
>test</I
>
to <TT
CLASS="FILENAME"
>/dev/null</TT
></A
> to suppress output</P
></LI
><LI
><P
><A
HREF="#RUNPARTSREF2"
>Running scripts in
sequence</A
> without user intervention, using <A
HREF="#RUNPARTSREF"
>run-parts</A
></P
></LI
><LI
><P
>Script<A
HREF="#SCRIPTASEMB"
> as embedded
command</A
></P
></LI
><LI
><P
>Script <I
CLASS="FIRSTTERM"
>portability</I
></P
><P
><A
HREF="#SETPUM"
>Setting <I
CLASS="FIRSTTERM"
>path</I
>
and <I
CLASS="FIRSTTERM"
>umask</I
></A
></P
><P
><A
HREF="#WHATISREF3"
>Using <I
CLASS="FIRSTTERM"
>whatis</I
></A
>
</P
></LI
><LI
><P
><A
HREF="#SETVAREMB"
>Setting script variable</A
>
to a block of embedded <I
CLASS="FIRSTTERM"
>sed</I
> or
<I
CLASS="FIRSTTERM"
>awk</I
> code</P
></LI
><LI
><P
>Speeding up script execution by <A
HREF="#LCALL"
>disabling <I
CLASS="FIRSTTERM"
>unicode</I
></A
>
</P
></LI
><LI
><P
>Subshell variable,
<A
HREF="#SUBSHTMP"
>accessing outside the subshell</A
>
</P
></LI
><LI
><P
><A
HREF="#INTPARAM"
>Testing a variable</A
>
to see if it contains only digits</P
></LI
><LI
><P
><A
HREF="#DEVNULLREDIRECT"
>Testing whether
a command exists</A
>, using <A
HREF="#TYPEREF"
>type</A
></P
></LI
><LI
><P
><A
HREF="#TRACKINGSCR"
>Tracking script
usage</A
></P
></LI
><LI
><P
><A
HREF="#GOODREAD0"
><I
CLASS="FIRSTTERM"
>while-read</I
> loop
without a
<I
CLASS="FIRSTTERM"
>subshell</I
></A
></P
></LI
><LI
><P
><A
HREF="#WIDGETREF"
>Widgets</A
>, invoking from a
script</P
></LI
></UL
></P
><P
><A
HREF="#TMOUTREF"
><TT
CLASS="VARNAME"
>$TMOUT</TT
></A
>,
Timeout interval</P
><P
><A
HREF="#TOKENREF"
>Token</A
>, a symbol that may
expand to a <A
HREF="#KEYWORDREF"
>keyword</A
> or
command</P
><P
><A
HREF="#TPUTREF"
>tput</A
>, terminal-control
command</P
><P
><A
HREF="#TRREF"
>tr</A
>, character translation filter
<P
></P
><UL
><LI
><P
><A
HREF="#TRD2U"
>DOS to Unix text file
conversion</A
></P
></LI
><LI
><P
><A
HREF="#TROPTIONS"
>Options</A
></P
></LI
><LI
><P
><A
HREF="#SOUNDEX0"
>Soundex</A
>, <I
CLASS="FIRSTTERM"
>example
script</I
></P
></LI
><LI
><P
><A
HREF="#TRVARIANTS"
>Variants</A
></P
></LI
></UL
></P
><P
><A
HREF="#TRAPREF1"
><I
CLASS="FIRSTTERM"
>Trap</I
></A
>,
specifying an action upon receipt of a <A
HREF="#SIGNALD"
>signal</A
></P
><P
><I
CLASS="FIRSTTERM"
>Trinary (ternary)</I
> operator,
<I
CLASS="FIRSTTERM"
>C</I
>-style,
<TT
CLASS="USERINPUT"
><B
>var&#62;10?88:99</B
></TT
>
<P
></P
><UL
><LI
><P
><A
HREF="#CSTRINARY"
>in
<I
CLASS="FIRSTTERM"
>double-parentheses</I
>
construct</A
></P
></LI
><LI
><P
><A
HREF="#EX46"
>in <I
CLASS="FIRSTTERM"
>let</I
>
construct</A
></P
></LI
></UL
></P
><P
><A
HREF="#TRUEREF"
>true</A
>,
returns <I
CLASS="FIRSTTERM"
>successful</I
> (0) <A
HREF="#EXITSTATUSREF"
>exit status</A
></P
><P
><A
HREF="#DECLARE1REF"
>typeset</A
> builtin
<P
></P
><UL
><LI
><P
><A
HREF="#DECLAREOPSREF1"
>options</A
></P
></LI
></UL
></P
><P
>* * *</P
><P
><A
HREF="#UIDREF"
><TT
CLASS="VARNAME"
>$UID</TT
></A
>,
User ID number</P
><P
><A
HREF="#UNALIASREF"
>unalias</A
>,
to remove an <A
HREF="#ALIASREF"
>alias</A
></P
><P
><A
HREF="#UNAMEREF"
>uname</A
>,
output system information</P
><P
><A
HREF="#UNICODEREF"
>Unicode</A
>, encoding standard
for representing letters and symbols
<P
></P
><UL
><LI
><P
><A
HREF="#LCALL"
>Disabling <I
CLASS="FIRSTTERM"
>unicode</I
></A
>
to optimize script</P
></LI
></UL
></P
><P
><A
HREF="#UNINITVAR"
>Uninitialized variables</A
> </P
><P
><A
HREF="#UNIQREF"
>uniq</A
>,
filter to remove duplicate lines from a sorted file</P
><P
><A
HREF="#UNSETREF"
>unset</A
>,
delete a shell variable</P
><P
><A
HREF="#UNTILLOOPREF"
>until</A
> loop</P
><P
><I
CLASS="FIRSTTERM"
>until [ condition-is-true ]; do</I
></P
><P
>* * *</P
><P
><I
CLASS="FIRSTTERM"
>Variables</I
>
<P
></P
><UL
><LI
><P
><A
HREF="#ARRAYOPSVARS"
>Array operations on
</A
></P
></LI
><LI
><P
><A
HREF="#ASNOP1"
>Assignment</A
></P
><P
><A
HREF="#EX15_0"
><I
CLASS="FIRSTTERM"
>Script
example</I
></A
></P
><P
><A
HREF="#EX16_0"
><I
CLASS="FIRSTTERM"
>Script
example</I
></A
></P
><P
><A
HREF="#VARUNSETTING"
><I
CLASS="FIRSTTERM"
>Script
example</I
></A
></P
></LI
><LI
><P
><A
HREF="#INTERNALVARIABLES"
><I
CLASS="FIRSTTERM"
>Bash</I
>
internal variables</A
></P
></LI
><LI
><P
><A
HREF="#SETVAREMB"
>Block of <I
CLASS="FIRSTTERM"
>sed</I
> or
<I
CLASS="FIRSTTERM"
>awk</I
> code</A
>, setting a variable
to</P
></LI
><LI
><P
> <I
CLASS="FIRSTTERM"
>C-style</I
>
<A
HREF="#PLUSPLUSREF"
> increment/decrement/trinary operations</A
></P
></LI
><LI
><P
> <A
HREF="#SETREF"
>Change value of internal script variables</A
>
using <I
CLASS="FIRSTTERM"
>set</I
></P
></LI
><LI
><P
><A
HREF="#DECLARE1REF"
>declare</A
>,
to modify the properties of variables</P
></LI
><LI
><P
><A
HREF="#UNSETREF"
>Deleting a shell variable</A
>
using <I
CLASS="FIRSTTERM"
>unset</I
></P
></LI
><LI
><P
><A
HREF="#ENVREF"
>Environmental</A
></P
></LI
><LI
><P
><A
HREF="#EXPREPL1"
>Expansion / Substring
replacement</A
> operators</P
></LI
><LI
><P
><A
HREF="#IVRREF"
>Indirect
referencing</A
></P
><P
><TT
CLASS="REPLACEABLE"
><I
>eval variable1=\$$variable2</I
></TT
></P
><P
><A
HREF="#IVR2"
>Newer notation</A
></P
><P
><TT
CLASS="REPLACEABLE"
><I
>${!variable}</I
></TT
></P
></LI
><LI
><P
><A
HREF="#INTVARREF"
>Integer</A
></P
></LI
><LI
><P
><A
HREF="#BVUNTYPED"
>Integer / string</A
> (variables are
untyped)</P
></LI
><LI
><P
><A
HREF="#PSOREX1"
>Length</A
></P
><P
><TT
CLASS="REPLACEABLE"
><I
>${#var}</I
></TT
></P
></LI
><LI
><P
><A
HREF="#LVALUEREF"
>Lvalue</A
></P
></LI
><LI
><P
><A
HREF="#PSSUB1"
>Manipulating and
expanding</A
></P
></LI
><LI
><P
><A
HREF="#VARNAMEVAL"
><I
CLASS="FIRSTTERM"
>Name</I
>
and <I
CLASS="FIRSTTERM"
>value</I
> of a variable</A
>,
distinguishing between</P
></LI
><LI
><P
><A
HREF="#STRINGNOTNULL"
><I
CLASS="FIRSTTERM"
>Null</I
>
string</A
>, testing for</P
></LI
><LI
><P
><A
HREF="#NULLVAR"
><I
CLASS="FIRSTTERM"
>Null</I
>
variable assignment</A
>, avoiding</P
></LI
><LI
><P
><A
HREF="#QUOTINGVAR"
>Quoting</A
></P
><P
><A
HREF="#FAILQUOTE"
>within <I
CLASS="FIRSTTERM"
>test</I
>
brackets</A
></P
><P
><A
HREF="#WSQUO"
>to preserve
<I
CLASS="FIRSTTERM"
>whitespace</I
></A
></P
></LI
><LI
><P
><A
HREF="#LVALUEREF"
>rvalue</A
></P
></LI
><LI
><P
><A
HREF="#VARUNSETTING"
>Setting to <I
CLASS="FIRSTTERM"
>null</I
>
value</A
></P
></LI
><LI
><P
><A
HREF="#PARVIS"
>In
<I
CLASS="FIRSTTERM"
>subshell</I
></A
>
not visible to parent shell</P
></LI
><LI
><P
>Testing a variable <A
HREF="#INTPARAM"
> if it contains only digits</A
></P
></LI
><LI
><P
><A
HREF="#TYPINGREF"
>Typing</A
>,
restricting the properties of a variable</P
></LI
><LI
><P
><A
HREF="#UNDVARERR"
>Undeclared</A
>,
error message</P
></LI
><LI
><P
><A
HREF="#UNINITVAR1"
>Uninitialized</A
>
</P
></LI
><LI
><P
><A
HREF="#VARSPLITTING"
>Unquoted
variable</A
>,
<I
CLASS="FIRSTTERM"
>splitting</I
></P
></LI
><LI
><P
><A
HREF="#UNSETREF"
>Unsetting</A
></P
></LI
><LI
><P
><A
HREF="#BVUNTYPED"
>Untyped</A
></P
></LI
></UL
></P
><P
>* * *</P
><P
><A
HREF="#WAITREF"
>wait</A
>,
suspend script execution
<P
></P
><UL
><LI
><P
><A
HREF="#WAITHANG"
>To remedy
script hang</A
></P
></LI
></UL
></P
><P
><A
HREF="#DBLQUO"
><I
CLASS="FIRSTTERM"
>Weak</I
> quoting</A
>
<B
CLASS="COMMAND"
>" ... "</B
></P
><P
><A
HREF="#WHILELOOPREF"
>while</A
> loop</P
><P
><I
CLASS="FIRSTTERM"
>while [ condition ]; do</I
>
<P
></P
><UL
><LI
><P
><A
HREF="#WHLOOPC"
>C-style syntax</A
>
</P
></LI
><LI
><P
><A
HREF="#WHILEFUNC"
>Calling a
<I
CLASS="FIRSTTERM"
>function</I
> within
<I
CLASS="FIRSTTERM"
>test</I
> brackets</A
></P
></LI
><LI
><P
><A
HREF="#WHMULTCOND"
>Multiple
conditions</A
></P
></LI
><LI
><P
><A
HREF="#WHILENOBRACKETS"
>Omitting
<I
CLASS="FIRSTTERM"
>test</I
> brackets</A
></P
></LI
><LI
><P
><A
HREF="#WHILEREADREF2"
> <I
CLASS="FIRSTTERM"
>while read</I
></A
>
construct
</P
><P
><A
HREF="#GOODREAD0"
>Avoiding a
<I
CLASS="FIRSTTERM"
>subshell</I
></A
></P
></LI
></UL
></P
><P
><A
HREF="#WHITESPACEREF"
>Whitespace</A
>,
spaces, tabs, and newline characters
<P
></P
><UL
><LI
><P
><A
HREF="#IFSWS"
><TT
CLASS="VARNAME"
>$IFS</TT
> defaults
to</A
></P
></LI
><LI
><P
><A
HREF="#WSBAD"
>Inappropriate use
of</A
></P
></LI
><LI
><P
><A
HREF="#INDENTEDLS"
>Preceding closing
<I
CLASS="FIRSTTERM"
>limit string</I
></A
> in a <I
CLASS="FIRSTTERM"
>here
document</I
>, error</P
></LI
><LI
><P
><A
HREF="#WSBCOMM"
>Preceding script
comments</A
></P
></LI
><LI
><P
><A
HREF="#WSQUO"
><I
CLASS="FIRSTTERM"
>Quoting</I
></A
>,
to preserve <I
CLASS="FIRSTTERM"
>whitespace</I
> within strings
or variables</P
></LI
><LI
><P
><A
HREF="#WSPOSIX"
>[:space:]</A
>,
<I
CLASS="FIRSTTERM"
>POSIX</I
> character class</P
></LI
></UL
></P
><P
><A
HREF="#WHOREF"
>who</A
>, information about logged on users
<P
></P
><UL
><LI
><P
><A
HREF="#WREF"
>w</A
></P
></LI
><LI
><P
><A
HREF="#WHOAMIREF"
>whoami</A
></P
></LI
><LI
><P
><A
HREF="#LOGNAMEREF"
>logname</A
></P
></LI
></UL
></P
><P
><A
HREF="#WIDGETREF"
>Widgets</A
></P
><P
><A
HREF="#WILDCARDDEF"
>Wild card</A
> characters
<P
></P
><UL
><LI
><P
><A
HREF="#ASTERISKREF"
>Asterisk
*</A
></P
></LI
><LI
><P
>In <A
HREF="#LIGLOB"
> <TT
CLASS="REPLACEABLE"
><I
>[list]</I
></TT
> constructs</A
>
</P
></LI
><LI
><P
><A
HREF="#WILDCARDQU"
>Question mark
?</A
></P
></LI
><LI
><P
><A
HREF="#WDOTFILEWC"
>Will not match
<TT
CLASS="FILENAME"
>dot files</TT
></A
></P
></LI
></UL
></P
><P
>Word splitting
<P
></P
><UL
><LI
><P
><A
HREF="#WSPLITREF"
>Definition</A
></P
></LI
><LI
><P
><A
HREF="#CSWS"
>Resulting from <I
CLASS="FIRSTTERM"
>command
substitution</I
></A
></P
></LI
></UL
></P
><P
><A
HREF="#SHWRAPPER"
>Wrapper</A
>, shell</P
><P
>* * *</P
><P
><A
HREF="#XARGSREF"
>xargs</A
>,
Filter for grouping arguments
<P
></P
><UL
><LI
><P
><A
HREF="#XARGSCURLYREF"
> Curly brackets</A
></P
></LI
><LI
><P
><A
HREF="#XARGSLIMARGS"
> Limiting arguments passed</A
></P
></LI
><LI
><P
><A
HREF="#XARGSLIMARGS"
>Options</A
></P
></LI
><LI
><P
>Processes arguments <A
HREF="#XARGSONEATATIME"
>one at a time</A
></P
></LI
><LI
><P
><A
HREF="#XARGSWS"
> Whitespace</A
>, handling</P
></LI
></UL
></P
><P
>* * *</P
><P
><A
HREF="#YESREF"
>yes</A
>
<P
></P
><UL
><LI
><P
><A
HREF="#YESEMU"
>Emulation</A
></P
></LI
></UL
></P
><P
>* * *</P
><P
><B
CLASS="COMMAND"
>-z </B
>
<A
HREF="#STRINGNULL"
>String is
<I
CLASS="FIRSTTERM"
>null</I
></A
></P
><P
><A
HREF="#ZOMBIEREF"
><I
CLASS="FIRSTTERM"
>Zombie</I
></A
>,
a process that has terminated, but not yet been <A
HREF="#KILLREF"
>killed</A
> by its
<A
HREF="#PARENTREF"
>parent</A
></P
></DIV
></DIV
><H3
CLASS="FOOTNOTES"
>Notes</H3
><TABLE
BORDER="0"
CLASS="FOOTNOTES"
WIDTH="100%"
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN62"
HREF="#AEN62"
><SPAN
CLASS="footnote"
>[1]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>These are referred to as <A
HREF="#BUILTINREF"
>builtins</A
>, features internal to the
shell.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN87"
HREF="#AEN87"
><SPAN
CLASS="footnote"
>[2]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Although <A
HREF="#RECURSIONREF0"
>recursion
<EM
>is</EM
> possible in a shell script</A
>,
it tends to be slow and its implementation is often
an <A
HREF="#FIBOREF"
>ugly kludge</A
>.
</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN139"
HREF="#AEN139"
><SPAN
CLASS="footnote"
>[3]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>An <I
CLASS="FIRSTTERM"
>acronym</I
>
is an <EM
>ersatz</EM
> word formed by pasting
together the initial letters of the words into a tongue-tripping
phrase. This morally corrupt and pernicious practice
deserves appropriately severe punishment. Public
flogging suggests itself.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN147"
HREF="#AEN147"
><SPAN
CLASS="footnote"
>[4]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Many of the features of <I
CLASS="FIRSTTERM"
>ksh88</I
>,
and even a few from the updated <I
CLASS="FIRSTTERM"
>ksh93</I
>
have been merged into Bash.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN157"
HREF="#AEN157"
><SPAN
CLASS="footnote"
>[5]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>By convention, user-written shell scripts
that are Bourne shell compliant generally take a name with a
<TT
CLASS="FILENAME"
>.sh</TT
> extension. System scripts, such as
those found in <TT
CLASS="FILENAME"
>/etc/rc.d</TT
>,
do not necessarily conform to this nomenclature.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN205"
HREF="#AEN205"
><SPAN
CLASS="footnote"
>[6]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>More commonly seen in the literature as
<I
CLASS="FIRSTTERM"
>she-bang</I
> or <I
CLASS="FIRSTTERM"
>sh-bang</I
>.
This derives from the concatenation of the tokens
<I
CLASS="FIRSTTERM"
>sharp</I
> (<SPAN
CLASS="TOKEN"
>#</SPAN
>) and
<I
CLASS="FIRSTTERM"
>bang</I
> (<SPAN
CLASS="TOKEN"
>!</SPAN
>).</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN214"
HREF="#AEN214"
><SPAN
CLASS="footnote"
>[7]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Some flavors of UNIX (those based on 4.2 BSD)
allegedly take a four-byte magic number, requiring
a blank after the <SPAN
CLASS="TOKEN"
>!</SPAN
> --
<TT
CLASS="USERINPUT"
><B
>#! /bin/sh</B
></TT
>. <A
HREF="http://www.in-ulm.de/~mascheck/various/shebang/#details"
TARGET="_top"
> According to Sven Mascheck</A
> this is probably a myth.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN226"
HREF="#AEN226"
><SPAN
CLASS="footnote"
>[8]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The <SPAN
CLASS="TOKEN"
>#!</SPAN
> line in a shell script
will be the first thing the command interpreter
(<B
CLASS="COMMAND"
>sh</B
> or <B
CLASS="COMMAND"
>bash</B
>)
sees. Since this line begins with a <SPAN
CLASS="TOKEN"
>#</SPAN
>,
it will be correctly interpreted as a comment when the
command interpreter finally executes the script. The
line has already served its purpose - calling the command
interpreter.</P
><P
>If, in fact, the script includes an
<EM
>extra</EM
> <SPAN
CLASS="TOKEN"
>#!</SPAN
> line, then
<B
CLASS="COMMAND"
>bash</B
> will interpret it as a comment.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
echo "Part 1 of script."
a=1
#!/bin/bash
# This does *not* launch a new script.
echo "Part 2 of script."
echo $a # Value of $a stays at 1.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN242"
HREF="#AEN242"
><SPAN
CLASS="footnote"
>[9]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This allows some cute tricks.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/rm
# Self-deleting script.
# Nothing much seems to happen when you run this... except that the file disappears.
WHATEVER=85
echo "This line will never print (betcha!)."
exit $WHATEVER # Doesn't matter. The script will not exit here.
# Try an echo $? after script termination.
# You'll get a 0, not a 85.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>Also, try starting a <TT
CLASS="FILENAME"
>README</TT
> file with a
<TT
CLASS="USERINPUT"
><B
>#!/bin/more</B
></TT
>, and making it executable.
The result is a self-listing documentation file. (A <A
HREF="#HEREDOCREF"
>here document</A
> using
<A
HREF="#CATREF"
>cat</A
> is possibly a better alternative
-- see <A
HREF="#EX71"
>Example 19-3</A
>).</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN256"
HREF="#AEN256"
><SPAN
CLASS="footnote"
>[10]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="POSIX2REF"
></A
><STRONG
>P</STRONG
>ortable
<STRONG
>O</STRONG
>perating
<STRONG
>S</STRONG
>ystem <EM
>I</EM
>nterface, an attempt to
standardize UNI<STRONG
>X</STRONG
>-like
OSes. The POSIX specifications are listed on the <A
HREF="http://www.opengroup.org/onlinepubs/007904975/toc.htm"
TARGET="_top"
>Open
Group site</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN269"
HREF="#AEN269"
><SPAN
CLASS="footnote"
>[11]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>To avoid this possibility, a script may begin
with a <A
HREF="#ENVV2REF"
>#!/bin/env bash</A
>
<I
CLASS="FIRSTTERM"
>sha-bang</I
> line. This may be
useful on UNIX machines where <I
CLASS="FIRSTTERM"
>bash</I
>
is not located in <TT
CLASS="FILENAME"
>/bin</TT
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN279"
HREF="#AEN279"
><SPAN
CLASS="footnote"
>[12]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>If <I
CLASS="FIRSTTERM"
>Bash</I
> is your default
shell, then the <SPAN
CLASS="TOKEN"
>#!</SPAN
> isn't necessary at the
beginning of a script. However, if launching a script from
a different shell, such as <I
CLASS="FIRSTTERM"
>tcsh</I
>,
then you <EM
>will</EM
> need the
<SPAN
CLASS="TOKEN"
>#!</SPAN
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN300"
HREF="#AEN300"
><SPAN
CLASS="footnote"
>[13]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Caution: invoking a <I
CLASS="FIRSTTERM"
>Bash</I
>
script by <TT
CLASS="USERINPUT"
><B
>sh scriptname</B
></TT
> turns off
Bash-specific extensions, and the script may therefore fail
to execute.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN315"
HREF="#AEN315"
><SPAN
CLASS="footnote"
>[14]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>A script needs <I
CLASS="FIRSTTERM"
>read</I
>, as
well as execute permission for it to run, since the shell
needs to be able to read it.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN327"
HREF="#AEN327"
><SPAN
CLASS="footnote"
>[15]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Why not simply invoke the script with
<TT
CLASS="USERINPUT"
><B
>scriptname</B
></TT
>? If the directory you
are in (<A
HREF="#PWDREF"
>$PWD</A
>) is where
<TT
CLASS="FILENAME"
>scriptname</TT
> is located, why doesn't
this work? This fails because, for security reasons, the
current directory (<TT
CLASS="FILENAME"
>./</TT
>)
is not by default included in a user's <A
HREF="#PATHREF"
>$PATH</A
>. It is therefore necessary to
explicitly invoke the script in the current directory with
a <TT
CLASS="USERINPUT"
><B
>./scriptname</B
></TT
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN612"
HREF="#AEN612"
><SPAN
CLASS="footnote"
>[16]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="OPERATORDEF"
></A
>An
<I
CLASS="FIRSTTERM"
>operator</I
> is an agent that carries
out an <I
CLASS="FIRSTTERM"
>operation</I
>. Some examples
are the common <A
HREF="#AROPS1"
>arithmetic
operators</A
>, <B
CLASS="COMMAND"
>+ - * /</B
>. In
Bash, there is some overlap between the concepts
of <I
CLASS="FIRSTTERM"
>operator</I
> and <A
HREF="#KEYWORDREF"
>keyword</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN888"
HREF="#AEN888"
><SPAN
CLASS="footnote"
>[17]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This is more commonly known
as the <I
CLASS="FIRSTTERM"
>ternary</I
>
operator. Unfortunately, <I
CLASS="FIRSTTERM"
>ternary</I
>
is an ugly word. It doesn't roll off
the tongue, and it doesn't elucidate. It
obfuscates. <I
CLASS="FIRSTTERM"
>Trinary</I
> is by far
the more elegant usage.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN1001"
HREF="#AEN1001"
><SPAN
CLASS="footnote"
>[18]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="ASCIIDEF"
></A
></P
><P
><B
CLASS="COMMAND"
>A</B
>merican
<B
CLASS="COMMAND"
>S</B
>tandard
<B
CLASS="COMMAND"
>C</B
>ode
for
<B
CLASS="COMMAND"
>I</B
>nformation
<B
CLASS="COMMAND"
>I</B
>nterchange.
This is a system for encoding text characters
(alphabetic, numeric, and a limited set of symbols)
as 7-bit numbers that can be stored and manipulated by
computers. Many of the ASCII characters are
represented on a standard keyboard.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN1071"
HREF="#AEN1071"
><SPAN
CLASS="footnote"
>[19]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="PROCESSIDDEF"
></A
></P
><P
>A <I
CLASS="FIRSTTERM"
>PID</I
>, or
<I
CLASS="FIRSTTERM"
>process ID</I
>, is a number assigned
to a running process. The <I
CLASS="FIRSTTERM"
>PID</I
>s
of running processes may be viewed with a <A
HREF="#PPSSREF"
>ps</A
> command.
</P
><P
><A
NAME="PROCESSREF"
></A
></P
><P
> <TT
CLASS="USERINPUT"
><B
>Definition:</B
></TT
> A
<I
CLASS="FIRSTTERM"
>process</I
> is a currently
executing command (or program), sometimes referred
to as a <I
CLASS="FIRSTTERM"
>job</I
>. </P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN1124"
HREF="#AEN1124"
><SPAN
CLASS="footnote"
>[20]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The shell does the <I
CLASS="FIRSTTERM"
>brace
expansion</I
>. The command itself acts upon the
<EM
>result</EM
> of the expansion.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN1199"
HREF="#AEN1199"
><SPAN
CLASS="footnote"
>[21]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Exception: a code block in braces as
part of a pipe <EM
>may</EM
> run as a
<A
HREF="#SUBSHELLSREF"
>subshell</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>ls | { read firstline; read secondline; }
# Error. The code block in braces runs as a subshell,
#+ so the output of "ls" cannot be passed to variables within the block.
echo "First line is $firstline; second line is $secondline" # Won't work.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN1564"
HREF="#AEN1564"
><SPAN
CLASS="footnote"
>[22]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
> Even as in olden times a
<I
CLASS="FIRSTTERM"
>philtre</I
> denoted a potion alleged
to have magical transformative powers, so does a UNIX
<I
CLASS="FIRSTTERM"
>filter</I
> transform its target in
(roughly) analogous fashion. (The coder who comes up with a
<SPAN
CLASS="QUOTE"
>"love philtre"</SPAN
> that runs on a Linux machine
will likely win accolades and honors.)</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN2107"
HREF="#AEN2107"
><SPAN
CLASS="footnote"
>[23]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Bash stores a list of commands
previously issued from the command-line
in a <I
CLASS="FIRSTTERM"
>buffer</I
>, or
memory space, for recall with the <A
HREF="#BUILTINREF"
>builtin</A
>
<I
CLASS="FIRSTTERM"
>history</I
>
commands.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN2198"
HREF="#AEN2198"
><SPAN
CLASS="footnote"
>[24]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>A linefeed (<I
CLASS="FIRSTTERM"
>newline</I
>)
is also a whitespace character. This explains
why a <I
CLASS="FIRSTTERM"
>blank line</I
>,
consisting only of a linefeed, is considered
whitespace.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN2258"
HREF="#AEN2258"
><SPAN
CLASS="footnote"
>[25]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="LVALUEREF"
></A
>Technically, the
<I
CLASS="FIRSTTERM"
>name</I
> of a variable is called an
<I
CLASS="FIRSTTERM"
>lvalue</I
>, meaning that it appears
on the <EM
>left</EM
> side of an assignment
statment, as in <TT
CLASS="USERINPUT"
><B
>VARIABLE=23</B
></TT
>.
A variable's <I
CLASS="FIRSTTERM"
>value</I
> is
an <I
CLASS="FIRSTTERM"
>rvalue</I
>, meaning that
it appears on the <EM
>right</EM
>
side of an assignment statement, as in
<TT
CLASS="USERINPUT"
><B
>VAR2=$VARIABLE</B
></TT
>.</P
><P
><A
NAME="POINTERREF"
></A
>A variable's
<I
CLASS="FIRSTTERM"
>name</I
> is, in fact,
a <I
CLASS="FIRSTTERM"
>reference</I
>, a
<I
CLASS="FIRSTTERM"
>pointer</I
> to the memory
location(s) where the actual data associated with
that variable is kept.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN2450"
HREF="#AEN2450"
><SPAN
CLASS="footnote"
>[26]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Note that <A
HREF="#PASSEDARGS"
><I
CLASS="FIRSTTERM"
>functions</I
>
also take positional parameters</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN2464"
HREF="#AEN2464"
><SPAN
CLASS="footnote"
>[27]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="ARG0"
></A
>The process calling the
script sets the <TT
CLASS="VARNAME"
>$0</TT
> parameter. By
convention, this parameter is the name of the script. See
the <A
HREF="#MANREF"
>manpage</A
> (manual page)
for <B
CLASS="COMMAND"
>execv</B
>.</P
><P
>From the <I
CLASS="FIRSTTERM"
>command-line</I
>, however,
<TT
CLASS="VARNAME"
>$0</TT
> is the name of the shell.
<TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo $0</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>bash</TT
>
<TT
CLASS="PROMPT"
>tcsh% </TT
><TT
CLASS="USERINPUT"
><B
>echo $0</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>tcsh</TT
></PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN2501"
HREF="#AEN2501"
><SPAN
CLASS="footnote"
>[28]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>If the the script is <A
HREF="#SOURCEREF"
>sourced</A
> or <A
HREF="#SYMLINKREF"
>symlinked</A
>, then
this will not work. It is safer to check <A
HREF="#BASHSOURCEREF"
>$BASH_Source</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN2609"
HREF="#AEN2609"
><SPAN
CLASS="footnote"
>[29]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Unless there is a file named
<TT
CLASS="FILENAME"
>first</TT
> in the current working directory. Yet
another reason to <I
CLASS="FIRSTTERM"
>quote</I
>. (Thank you, Harald
Koenig, for pointing this out.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN2630"
HREF="#AEN2630"
><SPAN
CLASS="footnote"
>[30]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="QUOTINGBSL"
></A
></P
><P
>Encapsulating <SPAN
CLASS="QUOTE"
>"!"</SPAN
> within double
quotes gives an error when used <EM
>from the command
line</EM
>. This is interpreted as a <A
HREF="#HISTCOMMANDS"
>history command</A
>. Within a script,
though, this problem does not occur, since the Bash history
mechanism is disabled then.</P
><P
>Of more concern is the <EM
>apparently</EM
>
inconsistent behavior of <TT
CLASS="REPLACEABLE"
><I
>\</I
></TT
>
within double quotes, and especially following an
<B
CLASS="COMMAND"
>echo -e</B
> command.</P
><P
> <TABLE
BORDER="1"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="SCREEN"
><TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo hello\!</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>hello!</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "hello\!"</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>hello\!</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo \</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>&#62;</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "\"</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>&#62;</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo \a</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>a</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "\a"</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>\a</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo x\ty</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>xty</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "x\ty"</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>x\ty</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo -e x\ty</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>xty</TT
>
<TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo -e "x\ty"</B
></TT
>
<TT
CLASS="COMPUTEROUTPUT"
>x y</TT
>
</PRE
></FONT
></TD
></TR
></TABLE
>
</P
><P
>Double quotes following an <I
CLASS="FIRSTTERM"
>echo</I
>
<EM
>sometimes</EM
> escape
<TT
CLASS="REPLACEABLE"
><I
>\</I
></TT
>. Moreover, the
<TT
CLASS="OPTION"
>-e</TT
> option to <I
CLASS="FIRSTTERM"
>echo</I
>
causes the <SPAN
CLASS="QUOTE"
>"\t"</SPAN
> to be interpreted as a
<I
CLASS="FIRSTTERM"
>tab</I
>.</P
><P
>(Thank you, Wayne Pollock, for pointing this out, and Geoff
Lee and Daniel Barclay for explaining it.) </P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN2688"
HREF="#AEN2688"
><SPAN
CLASS="footnote"
>[31]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="WSPLITREF"
></A
><SPAN
CLASS="QUOTE"
>"Word
splitting,"</SPAN
> in this context, means dividing
a character string into separate and discrete
arguments.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN2981"
HREF="#AEN2981"
><SPAN
CLASS="footnote"
>[32]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>In those instances when there is no <A
HREF="#RETURNREF"
>return</A
>
terminating the function.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN3140"
HREF="#AEN3140"
><SPAN
CLASS="footnote"
>[33]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="TOKENREF"
></A
>A
<I
CLASS="FIRSTTERM"
>token</I
> is a symbol or short
string with a special meaning attached to it (a <A
HREF="#METAMEANINGREF"
>meta-meaning</A
>). In Bash,
certain tokens, such as <B
CLASS="COMMAND"
>[</B
> and <A
HREF="#DOTREF"
>. (dot-command)</A
>, may expand to
<I
CLASS="FIRSTTERM"
>keywords</I
> and commands.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN3289"
HREF="#AEN3289"
><SPAN
CLASS="footnote"
>[34]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
> Per the 1913 edition of <EM
>Webster's
Dictionary</EM
>:
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>Deprecate
...
To pray against, as an evil;
to seek to avert by prayer;
to desire the removal of;
to seek deliverance from;
to express deep regret for;
to disapprove of strongly.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN3400"
HREF="#AEN3400"
><SPAN
CLASS="footnote"
>[35]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Be aware that <I
CLASS="FIRSTTERM"
>suid</I
>
binaries may open security holes. The
<I
CLASS="FIRSTTERM"
>suid</I
> flag has no effect on
shell scripts.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN3423"
HREF="#AEN3423"
><SPAN
CLASS="footnote"
>[36]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>On Linux systems, the sticky
bit is no longer used for files, only on
directories.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN3669"
HREF="#AEN3669"
><SPAN
CLASS="footnote"
>[37]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>As S.C. points out, in a compound test,
even quoting the string variable might not
suffice. <TT
CLASS="USERINPUT"
><B
>[ -n "$string" -o "$a" = "$b" ]</B
></TT
>
may cause an error with some versions of Bash if
<TT
CLASS="VARNAME"
>$string</TT
> is empty. The safe way
is to append an extra character to possibly empty variables,
<TT
CLASS="USERINPUT"
><B
>[ "x$string" != x -o "x$a" = "x$b" ]</B
></TT
>
(the <SPAN
CLASS="QUOTE"
>"x's"</SPAN
> cancel out).</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN3907"
HREF="#AEN3907"
><SPAN
CLASS="footnote"
>[38]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>In a different context, <B
CLASS="COMMAND"
>+=</B
> can
serve as a <I
CLASS="FIRSTTERM"
>string concatenation</I
>
operator. This can be useful for <A
HREF="#PATHAPPEND"
>modifying <I
CLASS="FIRSTTERM"
>environmental
variables</I
></A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN4242"
HREF="#AEN4242"
><SPAN
CLASS="footnote"
>[39]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><I
CLASS="FIRSTTERM"
>Side effects</I
>
are, of course, unintended -- and usually undesirable --
consequences.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN4290"
HREF="#AEN4290"
><SPAN
CLASS="footnote"
>[40]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><I
CLASS="FIRSTTERM"
>Precedence</I
>, in this context,
has approximately the same meaning as
<I
CLASS="FIRSTTERM"
>priority</I
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN4671"
HREF="#AEN4671"
><SPAN
CLASS="footnote"
>[41]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="STACKDEFREF"
></A
>
A <I
CLASS="FIRSTTERM"
>stack register</I
>
is a set of consecutive memory locations, such that
the values stored (<I
CLASS="FIRSTTERM"
>pushed</I
>)
are retrieved (<I
CLASS="FIRSTTERM"
>popped</I
>)
in <EM
>reverse</EM
> order. The last
value stored is the first retrieved. This is
sometimes called a <TT
CLASS="REPLACEABLE"
><I
>LIFO</I
></TT
>
(<I
CLASS="FIRSTTERM"
>last-in-first-out</I
>) or
<I
CLASS="FIRSTTERM"
>pushdown</I
> stack.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN5154"
HREF="#AEN5154"
><SPAN
CLASS="footnote"
>[42]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The PID of the currently running script is
<TT
CLASS="VARNAME"
>$$</TT
>, of course.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN5320"
HREF="#AEN5320"
><SPAN
CLASS="footnote"
>[43]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
> Somewhat analogous to <A
HREF="#RECURSIONREF"
>recursion</A
>, in this context
<I
CLASS="FIRSTTERM"
>nesting</I
> refers to a pattern
embedded within a larger pattern. One of the definitions
of <I
CLASS="FIRSTTERM"
>nest</I
>, according to the 1913
edition of <EM
>Webster's Dictionary</EM
>,
illustrates this beautifully: <SPAN
CLASS="QUOTE"
>"<EM
>A collection of
boxes, cases, or the like, of graduated size, each put
within the one next larger.</EM
>"</SPAN
>
</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN5479"
HREF="#AEN5479"
><SPAN
CLASS="footnote"
>[44]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The words <SPAN
CLASS="QUOTE"
>"argument"</SPAN
>
and <SPAN
CLASS="QUOTE"
>"parameter"</SPAN
> are often used
interchangeably. In the context of this document, they
have the same precise meaning: <EM
>a variable passed
to a script or function.</EM
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN5654"
HREF="#AEN5654"
><SPAN
CLASS="footnote"
>[45]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Within a script, inside a subshell,
<TT
CLASS="VARNAME"
>$$</TT
> <A
HREF="#BASHPIDREF"
>returns
the PID of the script</A
>, not the
subshell.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN5685"
HREF="#AEN5685"
><SPAN
CLASS="footnote"
>[46]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="TYPINGREF"
></A
>In this context,
<I
CLASS="FIRSTTERM"
>typing</I
>
a variable means to classify it and restrict its properties.
For example, a variable <I
CLASS="FIRSTTERM"
>declared</I
>
or <I
CLASS="FIRSTTERM"
>typed</I
> as an integer
is no longer available for <A
HREF="#STRINGOPSTAB"
>string
operations</A
>.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>declare -i intvar
intvar=23
echo "$intvar" # 23
intvar=stringval
echo "$intvar" # 0</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN5817"
HREF="#AEN5817"
><SPAN
CLASS="footnote"
>[47]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>True <SPAN
CLASS="QUOTE"
>"randomness,"</SPAN
> insofar as
it exists at all, can only be found in certain incompletely
understood natural phenomena, such as radioactive
decay. Computers only <I
CLASS="FIRSTTERM"
>simulate</I
>
randomness, and computer-generated sequences of
<SPAN
CLASS="QUOTE"
>"random"</SPAN
> numbers are therefore referred to as
<I
CLASS="FIRSTTERM"
>pseudorandom</I
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN5857"
HREF="#AEN5857"
><SPAN
CLASS="footnote"
>[48]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The <I
CLASS="FIRSTTERM"
>seed</I
> of a
computer-generated pseudorandom number series
can be considered an identification label. For
example, think of the pseudorandom series with a
seed of <EM
>23</EM
> as <TT
CLASS="REPLACEABLE"
><I
>Series
#23</I
></TT
>.</P
><P
>A property of a pseurandom number series is the length of
the cycle before it starts repeating itself. A good pseurandom
generator will produce series with very long cycles.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN5987"
HREF="#AEN5987"
><SPAN
CLASS="footnote"
>[49]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This applies to either command-line
arguments or parameters passed to a <A
HREF="#FUNCTIONREF"
>function</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN6164"
HREF="#AEN6164"
><SPAN
CLASS="footnote"
>[50]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Note that
<TT
CLASS="REPLACEABLE"
><I
>$substring</I
></TT
> and
<TT
CLASS="REPLACEABLE"
><I
>$replacement</I
></TT
> may refer to
either <I
CLASS="FIRSTTERM"
>literal strings</I
> or
<I
CLASS="FIRSTTERM"
>variables</I
>, depending on
context. See the first usage example.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN6310"
HREF="#AEN6310"
><SPAN
CLASS="footnote"
>[51]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>If $parameter is null in a
non-interactive script, it will terminate with a <A
HREF="#EXITCODESREF"
><SPAN
CLASS="RETURNVALUE"
>127</SPAN
>
exit status</A
> (the Bash error code for
<SPAN
CLASS="QUOTE"
>"command not found"</SPAN
>).</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN6560"
HREF="#AEN6560"
><SPAN
CLASS="footnote"
>[52]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="ITERATIONREF"
></A
><I
CLASS="FIRSTTERM"
>Iteration</I
>:
Repeated execution of a command or group of commands, usually --
but not always, <I
CLASS="FIRSTTERM"
>while</I
> a given condition
holds, or <I
CLASS="FIRSTTERM"
>until</I
> a given condition is
met.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN6981"
HREF="#AEN6981"
><SPAN
CLASS="footnote"
>[53]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>These are shell <A
HREF="#BUILTINREF"
>builtins</A
>,
whereas other loop commands, such as <A
HREF="#WHILELOOPREF"
>while</A
> and <A
HREF="#CASEESAC1"
>case</A
>, are <A
HREF="#KEYWORDREF"
>keywords</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN7087"
HREF="#AEN7087"
><SPAN
CLASS="footnote"
>[54]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Pattern-match lines may also <EM
>start</EM
>
with a <B
CLASS="COMMAND"
>(</B
> left paren to give the layout
a more structured appearance.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>case $( arch ) in # $( arch ) returns machine architecture.
( i386 ) echo "80386-based machine";;
# ^ ^
( i486 ) echo "80486-based machine";;
( i586 ) echo "Pentium-based machine";;
( i686 ) echo "Pentium2+-based machine";;
( * ) echo "Other type of machine";;
esac</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN7205"
HREF="#AEN7205"
><SPAN
CLASS="footnote"
>[55]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>For purposes of <I
CLASS="FIRSTTERM"
>command
substitution</I
>, a <B
CLASS="COMMAND"
>command</B
>
may be an external system command, an internal scripting
<A
HREF="#BUILTINREF"
>builtin</A
>, or even <A
HREF="#RVT"
>a script function</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN7211"
HREF="#AEN7211"
><SPAN
CLASS="footnote"
>[56]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>In a more technically correct sense,
<I
CLASS="FIRSTTERM"
>command substitution</I
> extracts the
<TT
CLASS="FILENAME"
>stdout</TT
> of a command, then assigns
it to a variable using the <SPAN
CLASS="TOKEN"
>=</SPAN
>
operator.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN7308"
HREF="#AEN7308"
><SPAN
CLASS="footnote"
>[57]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
> In fact, nesting with backticks is also possible,
but only by escaping the inner backticks, as John
Default points out.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>word_count=` wc -w \`echo * | awk '{print $8}'\` `</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN8607"
HREF="#AEN8607"
><SPAN
CLASS="footnote"
>[58]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>As Nathan Coulter points out, "while forking a
process is a low-cost operation, executing a new program in
the newly-forked child process adds more
overhead."</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN8650"
HREF="#AEN8650"
><SPAN
CLASS="footnote"
>[59]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>An exception to this is the <A
HREF="#TIMREF"
>time</A
> command, listed in the
official Bash documentation as a keyword (<SPAN
CLASS="QUOTE"
>"reserved
word"</SPAN
>).</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN9009"
HREF="#AEN9009"
><SPAN
CLASS="footnote"
>[60]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Note that <I
CLASS="FIRSTTERM"
>let</I
>
<A
HREF="#LETBAD"
>cannot be used
for setting <I
CLASS="FIRSTTERM"
>string</I
>
variables.</A
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN9199"
HREF="#AEN9199"
><SPAN
CLASS="footnote"
>[61]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>To <I
CLASS="FIRSTTERM"
>Export</I
>
information is to make it available in a more general context.
See also <A
HREF="#SCOPEREF"
>scope</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN9289"
HREF="#AEN9289"
><SPAN
CLASS="footnote"
>[62]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>An <I
CLASS="FIRSTTERM"
>option</I
> is an
argument that acts as a flag, switching script behaviors
on or off. The argument associated with a particular
option indicates the behavior that the option (flag)
switches on or off.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN9393"
HREF="#AEN9393"
><SPAN
CLASS="footnote"
>[63]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Technically, an
<B
CLASS="COMMAND"
>exit</B
> only terminates the
process (or shell) in which it is running,
<EM
>not</EM
> the <I
CLASS="FIRSTTERM"
>parent
process</I
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN9425"
HREF="#AEN9425"
><SPAN
CLASS="footnote"
>[64]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Unless the <B
CLASS="COMMAND"
>exec</B
> is used
to <A
HREF="#USINGEXECREF"
>reassign file
descriptors</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN9591"
HREF="#AEN9591"
><SPAN
CLASS="footnote"
>[65]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="HASHREF"
></A
></P
><P
><I
CLASS="FIRSTTERM"
>Hashing</I
> is a method of
creating lookup keys for data stored in a table. The
<EM
>data items themselves</EM
> are
<SPAN
CLASS="QUOTE"
>"scrambled"</SPAN
> to create keys, using one of
a number of simple mathematical
<I
CLASS="FIRSTTERM"
>algorithms</I
> (methods, or
recipes).</P
><P
>An advantage of <I
CLASS="FIRSTTERM"
>hashing</I
> is that
it is fast. A disadvantage is that
<I
CLASS="FIRSTTERM"
>collisions</I
> -- where a single key
maps to more than one data item -- are possible.</P
><P
>For examples of hashing see <A
HREF="#HASHLIB"
>Example A-20</A
> and
<A
HREF="#HASHEXAMPLE"
>Example A-21</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN9621"
HREF="#AEN9621"
><SPAN
CLASS="footnote"
>[66]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="READLINEREF"
></A
>The
<I
CLASS="FIRSTTERM"
>readline</I
> library is what
Bash uses for reading input in an
interactive shell.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN9753"
HREF="#AEN9753"
><SPAN
CLASS="footnote"
>[67]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This only applies to <I
CLASS="FIRSTTERM"
>child
processes</I
>, of course.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN9928"
HREF="#AEN9928"
><SPAN
CLASS="footnote"
>[68]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The C source for a number of loadable builtins is
typically found in the <TT
CLASS="FILENAME"
>/usr/share/doc/bash-?.??/functions</TT
>
directory.</P
><P
>Note that the <TT
CLASS="OPTION"
>-f</TT
> option to
<B
CLASS="COMMAND"
>enable</B
> is not <A
HREF="#PORTABILITYISSUES"
>portable</A
> to all
systems.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN9949"
HREF="#AEN9949"
><SPAN
CLASS="footnote"
>[69]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The same effect as
<B
CLASS="COMMAND"
>autoload</B
> can be achieved with <A
HREF="#DECLAREREF"
>typeset -fu</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN10025"
HREF="#AEN10025"
><SPAN
CLASS="footnote"
>[70]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The <TT
CLASS="OPTION"
>-v</TT
> option also orders the
sort by <EM
>upper- and lowercase prefixed</EM
>
filenames.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN10228"
HREF="#AEN10228"
><SPAN
CLASS="footnote"
>[71]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="DOTFILESREF"
></A
></P
><P
><I
CLASS="FIRSTTERM"
>Dotfiles</I
> are files whose
names begin with a <I
CLASS="FIRSTTERM"
>dot</I
>, such as
<TT
CLASS="FILENAME"
>~/.Xdefaults</TT
>. Such filenames do
not appear in a normal <B
CLASS="COMMAND"
>ls</B
> listing
(although an <B
CLASS="COMMAND"
>ls -a</B
> will show
them), and they cannot be deleted by an accidental
<B
CLASS="COMMAND"
>rm -rf *</B
>. Dotfiles are generally
used as setup and configuration files in a user's
home directory.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN10301"
HREF="#AEN10301"
><SPAN
CLASS="footnote"
>[72]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This particular feature may not yet be
implemented in the version of the ext2/ext3 filesystem
installed on your system. Check the documentation for
your Linux distro.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN10465"
HREF="#AEN10465"
><SPAN
CLASS="footnote"
>[73]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>And even when <I
CLASS="FIRSTTERM"
>xargs</I
> is
not strictly necessary, it can speed up execution of a command
involving <A
HREF="#BATCHPROCREF"
>batch-processing</A
> of multiple
files.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN11502"
HREF="#AEN11502"
><SPAN
CLASS="footnote"
>[74]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This is only true of the GNU version of
<B
CLASS="COMMAND"
>tr</B
>, not the generic version often found on
commercial UNIX systems.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN11885"
HREF="#AEN11885"
><SPAN
CLASS="footnote"
>[75]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>An <I
CLASS="FIRSTTERM"
>archive</I
>,
in the sense discussed here, is simply a set of related
files stored in a single location.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN11896"
HREF="#AEN11896"
><SPAN
CLASS="footnote"
>[76]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
> A <TT
CLASS="REPLACEABLE"
><I
>tar czvf ArchiveName.tar.gz *</I
></TT
>
<EM
>will</EM
> include dotfiles in
subdirectories <EM
>below</EM
> the current
working directory. This is an undocumented GNU
<B
CLASS="COMMAND"
>tar</B
> <SPAN
CLASS="QUOTE"
>"feature."</SPAN
>
</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN12840"
HREF="#AEN12840"
><SPAN
CLASS="footnote"
>[77]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The checksum may be expressed as a
<I
CLASS="FIRSTTERM"
>hexadecimal</I
> number, or to some
other base.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN12849"
HREF="#AEN12849"
><SPAN
CLASS="footnote"
>[78]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>For even <EM
>better</EM
>
security, use the <I
CLASS="FIRSTTERM"
>sha256sum</I
>,
<I
CLASS="FIRSTTERM"
>sha512</I
>, and
<I
CLASS="FIRSTTERM"
>sha1pass</I
>
commands.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN12969"
HREF="#AEN12969"
><SPAN
CLASS="footnote"
>[79]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This is a symmetric block cipher, used to
encrypt files on a single system or local network,
as opposed to the <I
CLASS="FIRSTTERM"
>public key</I
>
cipher class, of which <I
CLASS="FIRSTTERM"
>pgp</I
> is a
well-known example.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN13030"
HREF="#AEN13030"
><SPAN
CLASS="footnote"
>[80]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Creates a temporary
<I
CLASS="FIRSTTERM"
>directory</I
> when invoked with the
<TT
CLASS="OPTION"
>-d</TT
> option.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN13320"
HREF="#AEN13320"
><SPAN
CLASS="footnote"
>[81]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="DAEMONREF"
></A
></P
><P
>A <I
CLASS="FIRSTTERM"
>daemon</I
> is a background
process not attached to a terminal session. Daemons
perform designated services either at specified times
or explicitly triggered by certain events.</P
><P
>The word <SPAN
CLASS="QUOTE"
>"daemon"</SPAN
> means ghost in
Greek, and there is certainly something mysterious,
almost supernatural, about the way UNIX daemons
wander about behind the scenes, silently carrying
out their appointed tasks.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN14105"
HREF="#AEN14105"
><SPAN
CLASS="footnote"
>[82]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This is actually a script adapted from
the Debian Linux distribution.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN14214"
HREF="#AEN14214"
><SPAN
CLASS="footnote"
>[83]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The <I
CLASS="FIRSTTERM"
>print queue</I
> is
the group of jobs <SPAN
CLASS="QUOTE"
>"waiting in line"</SPAN
> to be
printed.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN14218"
HREF="#AEN14218"
><SPAN
CLASS="footnote"
>[84]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Large mechanical <I
CLASS="FIRSTTERM"
>line
printers</I
> printed a single
line of type at a time onto joined
sheets of <I
CLASS="FIRSTTERM"
>greenbar</I
>
paper, to the accompaniment of <A
HREF="http://www.columbia.edu/cu/computinghistory/1403.html"
TARGET="_top"
>a
great deal of noise</A
>. The hardcopy
thusly printed was referred to as a
<I
CLASS="FIRSTTERM"
>printout</I
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN14280"
HREF="#AEN14280"
><SPAN
CLASS="footnote"
>[85]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>For an excellent overview of this
topic, see Andy Vaught's article, <A
HREF="http://www2.linuxjournal.com/lj-issues/issue41/2156.html"
TARGET="_top"
>Introduction
to Named Pipes</A
>, in the September, 1997 issue of
<A
HREF="http://www.linuxjournal.com"
TARGET="_top"
><I
CLASS="CITETITLE"
>Linux
Journal</I
></A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN14318"
HREF="#AEN14318"
><SPAN
CLASS="footnote"
>[86]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><SPAN
CLASS="ACRONYM"
>EBCDIC</SPAN
> (pronounced
<SPAN
CLASS="QUOTE"
>"ebb-sid-ick"</SPAN
>) is an acronym for Extended
Binary Coded Decimal Interchange Code, an obsolete
IBM data format. A bizarre application of
the <TT
CLASS="OPTION"
>conv=ebcdic</TT
> option of
<B
CLASS="COMMAND"
>dd</B
> is as a quick 'n easy, but not
very secure text file encoder.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>cat $file | dd conv=swab,ebcdic &#62; $file_encrypted
# Encode (looks like gibberish).
# Might as well switch bytes (swab), too, for a little extra obscurity.
cat $file_encrypted | dd conv=swab,ascii &#62; $file_plaintext
# Decode.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN14523"
HREF="#AEN14523"
><SPAN
CLASS="footnote"
>[87]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>A <I
CLASS="FIRSTTERM"
>macro</I
> is a
symbolic constant that expands into a command string
or a set of operations on parameters. Simply put,
it's a shortcut or abbreviation.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN14695"
HREF="#AEN14695"
><SPAN
CLASS="footnote"
>[88]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This is the case on a Linux machine or a UNIX
system with disk quotas.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN14727"
HREF="#AEN14727"
><SPAN
CLASS="footnote"
>[89]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The <B
CLASS="COMMAND"
>userdel</B
> command
will fail if the particular user being deleted is
still logged on.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN16255"
HREF="#AEN16255"
><SPAN
CLASS="footnote"
>[90]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>For more detail on burning CDRs, see Alex
Withers' article, <A
HREF="http://www2.linuxjournal.com/lj-issues/issue66/3335.html"
TARGET="_top"
>Creating
CDs</A
>, in the October, 1999 issue of <A
HREF="http://www.linuxjournal.com"
TARGET="_top"
><I
CLASS="CITETITLE"
>Linux
Journal</I
></A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN16504"
HREF="#AEN16504"
><SPAN
CLASS="footnote"
>[91]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The <TT
CLASS="OPTION"
>-c</TT
> option to <A
HREF="#MKE2FSREF"
>mke2fs</A
> also invokes a check for bad
blocks.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN16659"
HREF="#AEN16659"
><SPAN
CLASS="footnote"
>[92]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Since only <I
CLASS="FIRSTTERM"
>root</I
>
has write permission in the <TT
CLASS="FILENAME"
>/var/lock</TT
> directory,
a user script cannot set a lock file there.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN16748"
HREF="#AEN16748"
><SPAN
CLASS="footnote"
>[93]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Operators of single-user Linux systems
generally prefer something simpler for backups, such
as <B
CLASS="COMMAND"
>tar</B
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN16782"
HREF="#AEN16782"
><SPAN
CLASS="footnote"
>[94]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>As of the <A
HREF="#BASH4REF"
>version
4 update</A
> of Bash, the <TT
CLASS="OPTION"
>-f</TT
>
and <TT
CLASS="OPTION"
>-c</TT
> options take a block size
of 512 when in <A
HREF="#POSIX2REF"
>POSIX</A
>
mode. Additionally, there are two new options:
<TT
CLASS="OPTION"
>-b</TT
> for <A
HREF="#SOCKETREF"
>socket</A
> buffer size, and
<TT
CLASS="OPTION"
>-T</TT
> for the limit on the number of
<I
CLASS="FIRSTTERM"
>threads</I
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN16847"
HREF="#AEN16847"
><SPAN
CLASS="footnote"
>[95]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>NAND is the logical
<I
CLASS="FIRSTTERM"
>not-and</I
> operator. Its effect
is somewhat similar to subtraction.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN16975"
HREF="#AEN16975"
><SPAN
CLASS="footnote"
>[96]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>In Bash and other Bourne shell derivatives, it is
possible to set variables in a single command's environment.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>var1=value1 var2=value2 commandXXX
# $var1 and $var2 set in the environment of 'commandXXX' only.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN17079"
HREF="#AEN17079"
><SPAN
CLASS="footnote"
>[97]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The <I
CLASS="FIRSTTERM"
>killall</I
> system
script should not be confused with the <A
HREF="#KILLALLREF"
>killall</A
> command in <TT
CLASS="FILENAME"
>/usr/bin</TT
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN17134"
HREF="#AEN17134"
><SPAN
CLASS="footnote"
>[98]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="METAMEANINGREF"
></A
>A
<I
CLASS="FIRSTTERM"
>meta-meaning</I
> is the meaning of a
term or expression on a higher level of abstraction. For
example, the <I
CLASS="FIRSTTERM"
>literal</I
> meaning
of <I
CLASS="FIRSTTERM"
>regular expression</I
> is an
ordinary expression that conforms to accepted usage. The
<I
CLASS="FIRSTTERM"
>meta-meaning</I
> is drastically different,
as discussed at length in this chapter.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN17189"
HREF="#AEN17189"
><SPAN
CLASS="footnote"
>[99]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Since <A
HREF="#SEDREF"
>sed</A
>, <A
HREF="#AWKREF"
>awk</A
>, and <A
HREF="#GREPREF"
>grep</A
> process single lines, there
will usually not be a newline to match. In those cases where
there is a newline in a multiple line expression, the dot
will match the newline.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
sed -e 'N;s/.*/[&#38;]/' &#60;&#60; EOF # Here Document
line1
line2
EOF
# OUTPUT:
# [line1
# line2]
echo
awk '{ $0=$1 "\n" $2; if (/line.1/) {print}}' &#60;&#60; EOF
line 1
line 2
EOF
# OUTPUT:
# line
# 1
# Thanks, S.C.
exit 0</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN17572"
HREF="#AEN17572"
><SPAN
CLASS="footnote"
>[100]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><I
CLASS="FIRSTTERM"
>Filename expansion</I
>
means expanding filename patterns or templates
containing special characters. For example,
<TT
CLASS="FILENAME"
>example.???</TT
> might expand
to <TT
CLASS="FILENAME"
>example.001</TT
> and/or
<TT
CLASS="FILENAME"
>example.txt</TT
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN17581"
HREF="#AEN17581"
><SPAN
CLASS="footnote"
>[101]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="WILDCARDDEF"
></A
>A <I
CLASS="FIRSTTERM"
>wild
card</I
> character, analogous to a wild card in poker,
can represent (almost) any other character.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN17592"
HREF="#AEN17592"
><SPAN
CLASS="footnote"
>[102]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
> Filename expansion <EM
>can</EM
>
match dotfiles, but only if the pattern explicitly includes the dot
as a literal character.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>~/[.]bashrc # Will not expand to ~/.bashrc
~/?bashrc # Neither will this.
# Wild cards and metacharacters will NOT
#+ expand to a dot in globbing.
~/.[b]ashrc # Will expand to ~/.bashrc
~/.ba?hrc # Likewise.
~/.bashr* # Likewise.
# Setting the "dotglob" option turns this off.
# Thanks, S.C.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN17822"
HREF="#AEN17822"
><SPAN
CLASS="footnote"
>[103]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Except, as Dennis Benzinger points out,
if <A
HREF="#LIMITSTRDASH"
>using
<B
CLASS="COMMAND"
>&#60;&#60;-</B
> to suppress
tabs</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN17884"
HREF="#AEN17884"
><SPAN
CLASS="footnote"
>[104]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>By convention in UNIX and Linux, data streams
and peripherals (<A
HREF="#DEVFILEREF"
>device files</A
>)
are treated as files, in a fashion analogous to ordinary
files.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN17894"
HREF="#AEN17894"
><SPAN
CLASS="footnote"
>[105]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="FDREF1"
></A
>A <I
CLASS="FIRSTTERM"
>file
descriptor</I
> is simply a number that
the operating system assigns to an open file
to keep track of it. Consider it a simplified
type of file pointer. It is analogous
to a <I
CLASS="FIRSTTERM"
>file handle</I
> in
<B
CLASS="COMMAND"
>C</B
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN17906"
HREF="#AEN17906"
><SPAN
CLASS="footnote"
>[106]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Using <TT
CLASS="REPLACEABLE"
><I
>file
descriptor 5</I
></TT
> might cause problems.
When Bash creates a child process, as with <A
HREF="#EXECREF"
>exec</A
>, the child inherits
fd 5 (see Chet Ramey's archived e-mail, <A
HREF="http://groups.google.com/group/gnu.bash.bug/browse_thread/thread/13955daafded3b5c/18c17050087f9f37"
TARGET="_top"
> SUBJECT: RE: File descriptor 5 is held open</A
>).
Best leave this particular fd alone.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN18102"
HREF="#AEN18102"
><SPAN
CLASS="footnote"
>[107]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>An external command invoked with an <A
HREF="#EXECREF"
>exec</A
> does <EM
>not</EM
>
(usually) fork off a subprocess / subshell.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN18244"
HREF="#AEN18244"
><SPAN
CLASS="footnote"
>[108]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This has the same effect as a
<A
HREF="#NAMEDPIPEREF"
>named pipe</A
> (temp
file), and, in fact, named pipes were at one time used
in process substitution.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN18474"
HREF="#AEN18474"
><SPAN
CLASS="footnote"
>[109]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The <B
CLASS="COMMAND"
>return</B
> command is a
Bash <A
HREF="#BUILTINREF"
>builtin</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN18568"
HREF="#AEN18568"
><SPAN
CLASS="footnote"
>[110]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>However, as Thomas Braunberger points out, a local
variable declared in a function <EM
>is also visible
to functions called by the parent
function.</EM
></P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
function1 ()
{
local func1var=20
echo "Within function1, \$func1var = $func1var."
function2
}
function2 ()
{
echo "Within function2, \$func1var = $func1var."
}
function1
exit 0
# Output of the script:
# Within function1, $func1var = 20.
# Within function2, $func1var = 20.</PRE
></FONT
></TD
></TR
></TABLE
></P
><P
>This is documented in the Bash manual:</P
><P
><SPAN
CLASS="QUOTE"
>"Local can only be used within a function;
it makes the variable name have a visible scope
restricted to that function <EM
>and its
children</EM
>."</SPAN
> [emphasis added]
<EM
>The ABS Guide author considers this behavior
to be a bug.</EM
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN18607"
HREF="#AEN18607"
><SPAN
CLASS="footnote"
>[111]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Otherwise known as
<I
CLASS="FIRSTTERM"
>redundancy</I
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN18610"
HREF="#AEN18610"
><SPAN
CLASS="footnote"
>[112]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Otherwise known as
<I
CLASS="FIRSTTERM"
>tautology</I
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN18614"
HREF="#AEN18614"
><SPAN
CLASS="footnote"
>[113]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Otherwise known as a
<I
CLASS="FIRSTTERM"
>metaphor</I
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN18617"
HREF="#AEN18617"
><SPAN
CLASS="footnote"
>[114]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Otherwise known as a
<I
CLASS="FIRSTTERM"
>recursive function</I
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN18632"
HREF="#AEN18632"
><SPAN
CLASS="footnote"
>[115]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Too many levels of recursion may crash a
script with a segfault.
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="100%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>#!/bin/bash
# Warning: Running this script could possibly lock up your system!
# If you're lucky, it will segfault before using up all available memory.
recursive_function ()
{
echo "$1" # Makes the function do something, and hastens the segfault.
(( $1 &#60; $2 )) &#38;&#38; recursive_function $(( $1 + 1 )) $2;
# As long as 1st parameter is less than 2nd,
#+ increment 1st and recurse.
}
recursive_function 1 50000 # Recurse 50,000 levels!
# Most likely segfaults (depending on stack size, set by ulimit -m).
# Recursion this deep might cause even a C program to segfault,
#+ by using up all the memory allotted to the stack.
echo "This will probably not print."
exit 0 # This script will not exit normally.
# Thanks, St<53>phane Chazelas.</PRE
></FONT
></TD
></TR
></TABLE
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN18669"
HREF="#AEN18669"
><SPAN
CLASS="footnote"
>[116]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
> ... as the first word of a command string.
Obviously, an alias is only meaningful at the
<EM
>beginning</EM
> of a command.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN18676"
HREF="#AEN18676"
><SPAN
CLASS="footnote"
>[117]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>However, aliases do seem to expand positional
parameters.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN19045"
HREF="#AEN19045"
><SPAN
CLASS="footnote"
>[118]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>The entries in <TT
CLASS="FILENAME"
>/dev</TT
>
provide mount points for physical and virtual devices. These
entries use very little drive space.</P
><P
>Some devices, such as <TT
CLASS="FILENAME"
>/dev/null</TT
>,
<TT
CLASS="FILENAME"
>/dev/zero</TT
>,
and <TT
CLASS="FILENAME"
>/dev/urandom</TT
> are virtual. They
are not actual physical devices and exist only in
software.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN19065"
HREF="#AEN19065"
><SPAN
CLASS="footnote"
>[119]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="BLOCKDEVREF"
></A
>A <I
CLASS="FIRSTTERM"
>block
device</I
> reads and/or writes data in chunks,
or <I
CLASS="FIRSTTERM"
>blocks</I
>, in contrast to a <A
NAME="CHARDEVREF"
></A
><I
CLASS="FIRSTTERM"
>character device</I
>,
which acesses data in <I
CLASS="FIRSTTERM"
>character</I
>
units. Examples of block devices are hard drives, CDROM
drives, and flash drives. Examples of character devices are
keyboards, modems, sound cards.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN19093"
HREF="#AEN19093"
><SPAN
CLASS="footnote"
>[120]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Of course, the mount point
<TT
CLASS="FILENAME"
>/mnt/flashdrive</TT
> must exist. If not,
then, as <I
CLASS="FIRSTTERM"
>root</I
>, <B
CLASS="COMMAND"
>mkdir
/mnt/flashdrive</B
>.</P
><P
>To actually mount the drive, use the following command:
<B
CLASS="COMMAND"
>mount /mnt/flashdrive</B
></P
><P
>Newer Linux distros automount flash drives in the
<TT
CLASS="FILENAME"
>/media</TT
>
directory without user intervention.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN19180"
HREF="#AEN19180"
><SPAN
CLASS="footnote"
>[121]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Certain system commands, such as
<A
HREF="#PROCINFOREF"
>procinfo</A
>,
<A
HREF="#FREEREF"
>free</A
>,
<A
HREF="#VMSTATREF"
>vmstat</A
>,
<A
HREF="#LSDEVREF"
>lsdev</A
>,
and <A
HREF="#UPTIMEREF"
>uptime</A
>
do this as well.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN19460"
HREF="#AEN19460"
><SPAN
CLASS="footnote"
>[122]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>By convention, <TT
CLASS="REPLACEABLE"
><I
>signal
0</I
></TT
> is assigned to <A
HREF="#EXITCOMMANDREF"
>exit</A
>. </P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN19993"
HREF="#AEN19993"
><SPAN
CLASS="footnote"
>[123]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Setting the <A
HREF="#SUIDREF"
>suid</A
>
permission on the script itself has no effect in Linux
and most other UNIX flavors.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN20041"
HREF="#AEN20041"
><SPAN
CLASS="footnote"
>[124]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>In this context, <SPAN
CLASS="QUOTE"
>"magic
numbers"</SPAN
> have an entirely different meaning than
the <A
HREF="#MAGNUMREF"
>magic numbers</A
> used
to designate file types.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN20130"
HREF="#AEN20130"
><SPAN
CLASS="footnote"
>[125]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Quite a number of Linux utilities are, in fact,
shell wrappers. Some examples are
<TT
CLASS="FILENAME"
>/usr/bin/pdf2ps</TT
>,
<TT
CLASS="FILENAME"
>/usr/bin/batch</TT
>, and
<TT
CLASS="FILENAME"
>/usr/bin/xmkmf</TT
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN20259"
HREF="#AEN20259"
><SPAN
CLASS="footnote"
>[126]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><SPAN
CLASS="ACRONYM"
>ANSI</SPAN
> is, of course, the
acronym for the American National Standards
Institute. This august body establishes and maintains
various technical and industrial standards.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN20452"
HREF="#AEN20452"
><SPAN
CLASS="footnote"
>[127]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This usually means liberal use of
<A
HREF="#FUNCTIONREF"
>functions</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN20748"
HREF="#AEN20748"
><SPAN
CLASS="footnote"
>[128]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>See Marius van Oers' article, <A
HREF="http://www.virusbtn.com/magazine/archives/200204/malshell.xml"
TARGET="_top"
>Unix
Shell Scripting Malware</A
>, and also the
<A
HREF="#DENNINGREF"
><EM
>Denning</EM
>
reference</A
> in the
<I
CLASS="FIRSTTERM"
>bibliography</I
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN20799"
HREF="#AEN20799"
><SPAN
CLASS="footnote"
>[129]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Or, better yet, <A
HREF="#ENVV2REF"
>#!/bin/env sh</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN21025"
HREF="#AEN21025"
><SPAN
CLASS="footnote"
>[130]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>To be more specific, Bash 4+ has
<EM
>limited</EM
> support for associative
arrays. It's a bare-bones implementation,
and it lacks the much of the functionality of such
arrays in other programming languages. Note, however,
that <A
HREF="#ASSOCARRTST"
>associative arrays in
Bash seem to execute faster and more efficiently than
numerically-indexed arrays</A
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN21068"
HREF="#AEN21068"
><SPAN
CLASS="footnote"
>[131]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Copyright 1995-2009 by Chester Ramey.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN21096"
HREF="#AEN21096"
><SPAN
CLASS="footnote"
>[132]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This only works with <A
HREF="#PIPEREF"
>pipes</A
> and certain other
<I
CLASS="FIRSTTERM"
>special</I
> files.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN21101"
HREF="#AEN21101"
><SPAN
CLASS="footnote"
>[133]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>But only in conjunction with
<A
HREF="#READLINEREF"
>readline</A
>, i.e.,
from the command-line.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN21179"
HREF="#AEN21179"
><SPAN
CLASS="footnote"
>[134]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>And while you're at it, consider fixing
the notorious <A
HREF="#PIPEREADREF0"
>piped read</A
>
problem.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN21281"
HREF="#AEN21281"
><SPAN
CLASS="footnote"
>[135]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This is the notorious <EM
>flog
it to death</EM
> technique that works so well
with slow learners, eccentrics, odd ducks, fools and
geniuses.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN21292"
HREF="#AEN21292"
><SPAN
CLASS="footnote"
>[136]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>In fact, he has no credentials or special
qualifications. He's a school dropout with no formal credentials
or professional experience whatsoever. None. Zero. Nada. Aside
from the <EM
>ABS Guide</EM
>, his major claim to
fame is a First Place in the sack race at the Colfax Elementary
School Field Day in June, 1958.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN21295"
HREF="#AEN21295"
><SPAN
CLASS="footnote"
>[137]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Those who can, do. Those who can't . . . get an
MCSE.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN21318"
HREF="#AEN21318"
><SPAN
CLASS="footnote"
>[138]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Sometimes it seems as if he has spent
his entire life flouting conventional wisdom and defying the
sonorous Voice of Authority: <SPAN
CLASS="QUOTE"
>"<EM
>Hey, you
can't do that!</EM
>"</SPAN
></P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN21329"
HREF="#AEN21329"
><SPAN
CLASS="footnote"
>[139]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Well, if you <EM
>absolutely</EM
>
insist, you can try modifying <A
HREF="#HOMEWORK"
>Example A-44</A
> to suit
your purposes.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN21579"
HREF="#AEN21579"
><SPAN
CLASS="footnote"
>[140]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>It was hard to resist the obvious pun. No slight
intended, since the book is a pretty decent introduction to
the basic concepts of shell scripting.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN23174"
HREF="#AEN23174"
><SPAN
CLASS="footnote"
>[141]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><I
CLASS="FIRSTTERM"
>Sed</I
> executes without
user intervention.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN23185"
HREF="#AEN23185"
><SPAN
CLASS="footnote"
>[142]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>If no address range is specified, the default
is <EM
>all</EM
> lines.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN23443"
HREF="#AEN23443"
><SPAN
CLASS="footnote"
>[143]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Its name derives from the initials of its authors,
<B
CLASS="COMMAND"
>A</B
>ho, <B
CLASS="COMMAND"
>W</B
>einberg, and
<B
CLASS="COMMAND"
>K</B
>ernighan.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN23629"
HREF="#AEN23629"
><SPAN
CLASS="footnote"
>[144]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
><A
NAME="EXCOOR"
></A
>Out of range exit values
can result in unexpected exit codes. An exit value
greater than <SPAN
CLASS="ERRORCODE"
>255</SPAN
> returns an
exit code <A
HREF="#MODULOREF"
>modulo</A
>
<SPAN
CLASS="ERRORCODE"
>256</SPAN
>. For example, <I
CLASS="FIRSTTERM"
>exit
3809</I
> gives an exit code of <SPAN
CLASS="ERRORCODE"
>225</SPAN
>
(3809 % 256 = 225).</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN23647"
HREF="#AEN23647"
><SPAN
CLASS="footnote"
>[145]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>An update of <TT
CLASS="FILENAME"
>/usr/include/sysexits.h</TT
>
allocates previously unused exit codes from <SPAN
CLASS="RETURNVALUE"
>64
- 78</SPAN
>. It may be anticipated that the range of
unallotted exit codes will be further restricted in the future.
The author of this document will <EM
>not</EM
> do
fixups on the scripting examples to conform to the changing
standard. This should not cause any problems, since there
is no overlap or conflict in usage of exit codes between
compiled C/C++ binaries and shell scripts.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN23892"
HREF="#AEN23892"
><SPAN
CLASS="footnote"
>[146]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This does not apply to <B
CLASS="COMMAND"
>csh</B
>,
<B
CLASS="COMMAND"
>tcsh</B
>, and other shells not related to or
descended from the classic Bourne shell
(<B
CLASS="COMMAND"
>sh</B
>).</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN23937"
HREF="#AEN23937"
><SPAN
CLASS="footnote"
>[147]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>In older versions of UNIX, passwords
<EM
>were</EM
> stored in
<TT
CLASS="FILENAME"
>/etc/passwd</TT
>, and that explains
the name of the file.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN23969"
HREF="#AEN23969"
><SPAN
CLASS="footnote"
>[148]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Some early UNIX systems had a fast, small-capacity fixed
disk (containing <TT
CLASS="FILENAME"
>/</TT
>,
the root partition), and a second drive which
was larger, but slower (containing <TT
CLASS="FILENAME"
>/usr</TT
> and other
partitions). The most frequently used programs and
utilities therefore resided on the small-but-fast
drive, in <TT
CLASS="FILENAME"
>/bin</TT
>,
and the others on the slower drive, in <TT
CLASS="FILENAME"
>/usr/bin</TT
>.</P
><P
>This likewise accounts for the split between
<TT
CLASS="FILENAME"
>/sbin</TT
> and
<TT
CLASS="FILENAME"
>/usr/sbin</TT
>,
<TT
CLASS="FILENAME"
>/lib</TT
> and <TT
CLASS="FILENAME"
>/usr/lib</TT
>, etc.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN24082"
HREF="#AEN24082"
><SPAN
CLASS="footnote"
>[149]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>This works only from the <I
CLASS="FIRSTTERM"
>command
line</I
>, of course, and not within a
script.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN24160"
HREF="#AEN24160"
><SPAN
CLASS="footnote"
>[150]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Normally the default parameter completion files reside
in either the <TT
CLASS="FILENAME"
>/etc/profile.d</TT
>
directory or in <TT
CLASS="FILENAME"
>/etc/bash_completion</TT
>. These autoload on
system startup. So, after writing a useful completion script, you
might wish to move it (as <I
CLASS="FIRSTTERM"
>root</I
>, of course)
to one of these directories.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN24173"
HREF="#AEN24173"
><SPAN
CLASS="footnote"
>[151]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>It has been extensively documented that
programmers are willing to put in long hours of effort in
order to save ten minutes of <SPAN
CLASS="QUOTE"
>"unnecessary"</SPAN
>
labor. This is known as
<I
CLASS="FIRSTTERM"
>optimization</I
>.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN24713"
HREF="#AEN24713"
><SPAN
CLASS="footnote"
>[152]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>Various readers have suggested modifications
of the above batch file to prettify it and make it more
compact and efficient. In the opinion of the <EM
>ABS
Guide</EM
> author, this is wasted effort. A Bash script
can access a DOS filesystem, or even an NTFS partition (with
the help of <A
HREF="http://www.ntfs-3g.org"
TARGET="_top"
>ntfs-3g</A
>)
to do batch or scripted operations.</P
></TD
></TR
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN25254"
HREF="#AEN25254"
><SPAN
CLASS="footnote"
>[153]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
>For all you clever types who failed intermediate algebra,
a <I
CLASS="FIRSTTERM"
>determinant</I
> is a numerical value
associated with a multidimensional
<I
CLASS="FIRSTTERM"
>matrix</I
> (<A
HREF="#ARRAYREF"
>array</A
> of numbers).
<TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><FONT
COLOR="#000000"
><PRE
CLASS="PROGRAMLISTING"
>For the simple case of a 2 x 2 determinant:
|a b|
|b a|
The solution is a*a - b*b, where "a" and "b" represent numbers.</PRE
></FONT
></TD
></TR
></TABLE
>
</P
></TD
></TR
></TABLE
></BODY
></HTML
>