187 lines
6.2 KiB

#! /usr/bin/python
# -*- coding: utf8 -*-
from __future__ import absolute_import, division, print_function
import copy
import logging
from tldp.utils import max_mtime, mtime_gt
from tldp.sources import SourceCollection
from tldp.outputs import OutputCollection
logger = logging.getLogger(__name__)
# -- any individual document (source or output) will have a status
# from the following list of status_types
status_types = [
# -- the user probably doesn't usually care (too much) about listing
# every single published document and source document, but is probably
# mostly interested in specific documents grouped by status; so the
# status_classes are just sets of status_types
status_classes = dict(zip(status_types, [[x] for x in status_types]))
status_classes['outputs'] = ['output']
status_classes['sources'] = ['source']
status_classes['problems'] = ['orphan', 'broken', 'stale']
status_classes['work'] = ['new', 'orphan', 'broken', 'stale']
status_classes['orphans'] = ['orphan']
status_classes['orphaned'] = ['orphan']
status_classes['all'] = ['published', 'new', 'orphan', 'broken', 'stale']
class Inventory(object):
'''a container for classifying documents by their status
Every SourceDocument has no more than one matching OutputDirectory.
The Inventory class encodes the logic for identifying the following
different status possibilities for an arbitrary set of SourceDocuments and
The following are possible values for status:
- 'source': a source document before any status detection
- 'output': an output document before any status detection
- 'new': a source document without any matching output stem
- 'published': a pair of source/output documents with matching stems
- 'orphan': an output document without any matching source stem
- 'broken': a published document with missing output files
- 'stale': a published document with new(er) source files
The Inventory object is intended to be used to identify work that needs to
be done on individual source documents to produce up-to-date output
def __repr__(self):
return '<%s: %d published, %d orphan, %d new, %d stale, %d broken>' % (
def __init__(self, pubdir, sourcedirs):
'''construct an Inventory
pubdir: path to the OutputCollection
sourcedirs: a list of directories which could be passed to the
SourceCollection object; essentially a directory containing
SourceDocuments; for example LDP/LDP/howto/linuxdoc and
self.output = OutputCollection(pubdir)
self.source = SourceCollection(sourcedirs)
s = copy.deepcopy(self.source)
o = copy.deepcopy(self.output)
sset = set(s.keys())
oset = set(o.keys())
# -- orphan identification
self.orphan = OutputCollection()
for doc in oset.difference(sset):
self.orphan[doc] = o[doc]
del o[doc]
self.orphan[doc].status = 'orphan'
logger.debug("Identified %d orphan documents: %r.", len(self.orphan),
# -- unpublished ('new') identification
# = SourceCollection()
for doc in sset.difference(oset):[doc] = s[doc]
del s[doc][doc].status = 'new'
logger.debug("Identified %d new documents: %r.", len(,
# -- published identification; source and output should be same size
assert len(s) == len(o)
for stem, odoc in o.items():
sdoc = s[stem]
sdoc.output = odoc
odoc.source = sdoc
sdoc.status = sdoc.output.status = 'published'
self.published = s
logger.debug("Identified %d published documents.", len(self.published))
# -- stale identification
self.stale = SourceCollection()
for stem, sdoc in s.items():
odoc = sdoc.output
mtime = max_mtime(odoc.statinfo)
fset = mtime_gt(mtime, sdoc.statinfo)
if fset:
sdoc.newer = fset
for f in fset:
logger.debug("%s found updated source file %s", stem, f)
odoc.status = sdoc.status = 'stale'
self.stale[stem] = sdoc
logger.debug("Identified %d stale documents: %r.", len(self.stale),
# -- stale identification
self.broken = SourceCollection()
for stem, sdoc in s.items():
if not sdoc.output.iscomplete:
self.broken[stem] = sdoc
sdoc.status = sdoc.output.status = 'broken'
logger.debug("Identified %d broken documents: %r.", len(self.broken),
def getByStatusClass(self, status_class):
desired = status_classes.get(status_class, None)
assert isinstance(desired, list)
collection = SourceCollection()
for status_type in desired:
collection.update(getattr(self, status_type))
return collection
def outputs(self):
return self.getByStatusClass('outputs')
def sources(self):
return self.getByStatusClass('sources')
def problems(self):
return self.getByStatusClass('problems')
def work(self):
return self.getByStatusClass('work')
def orphans(self):
return self.getByStatusClass('orphans')
def orphaned(self):
return self.getByStatusClass('orphaned')
def all(self):
return self.getByStatusClass('all')
# -- end of file