dinislage/CouchPotatoServer
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Update libs

Browse Source
This commit is contained in:
Ruud
2013-11-22 16:09:15 +01:00
parent f53364eb6c
commit 88d6148500
119 changed files with 18448 additions and 8923 deletions
Download Patch File Download Diff File Expand all files Collapse all files
libs/guessit/__init__.py
+62 -48
View File
@@ -20,7 +20,7 @@
from __future__ import unicode_literals
__version__ = '0.7-dev'
__version__ = '0.6.2'
__all__ = ['Guess', 'Language',
'guess_file_info', 'guess_video_info',
'guess_movie_info', 'guess_episode_info']
@@ -76,6 +76,7 @@ from guessit.language import Language
from guessit.matcher import IterativeMatcher
from guessit.textutils import clean_string
import logging
import json
log = logging.getLogger(__name__)
@@ -105,17 +106,74 @@ def _guess_filename(filename, filetype):
mtree = IterativeMatcher(filename, filetype=filetype)
m = mtree.matched()
second_pass_opts = []
second_pass_transfo_opts = {}
# if there are multiple possible years found, we assume the first one is
# part of the title, reparse the tree taking this into account
years = set(n.value for n in find_nodes(mtree.match_tree, 'year'))
if len(years) >= 2:
mtree = IterativeMatcher(filename, filetype=filetype,
opts=['skip_first_year'])
second_pass_opts.append('skip_first_year')
to_skip_language_nodes = []
title_nodes = set(n for n in find_nodes(mtree.match_tree, ['title', 'series']))
title_spans = {}
for title_node in title_nodes:
title_spans[title_node.span[0]] = title_node
title_spans[title_node.span[1]] = title_node
for lang_key in ('language', 'subtitleLanguage'):
langs = {}
lang_nodes = set(n for n in find_nodes(mtree.match_tree, lang_key))
for lang_node in lang_nodes:
lang = lang_node.guess.get(lang_key, None)
if len(lang_node.value) > 3 and (lang_node.span[0] in title_spans.keys() or lang_node.span[1] in title_spans.keys()):
# Language is next or before title, and is not a language code. Add to skip for 2nd pass.
# if filetype is subtitle and the language appears last, just before
# the extension, then it is likely a subtitle language
parts = clean_string(lang_node.root.value).split()
if m['type'] in ['moviesubtitle', 'episodesubtitle'] and (parts.index(lang_node.value) == len(parts) - 2):
continue
to_skip_language_nodes.append(lang_node)
elif not lang in langs:
langs[lang] = lang_node
else:
# The same language was found. Keep the more confident one, and add others to skip for 2nd pass.
existing_lang_node = langs[lang]
to_skip = None
if existing_lang_node.guess.confidence('language') >= lang_node.guess.confidence('language'):
# lang_node is to remove
to_skip = lang_node
else:
# existing_lang_node is to remove
langs[lang] = lang_node
to_skip = existing_lang_node
to_skip_language_nodes.append(to_skip)
if to_skip_language_nodes:
second_pass_transfo_opts['guess_language'] = (
((), { 'skip': [ { 'node_idx': node.parent.node_idx,
'span': node.span }
for node in to_skip_language_nodes ] }))
if second_pass_opts or second_pass_transfo_opts:
# 2nd pass is needed
log.info("Running 2nd pass with options: %s" % second_pass_opts)
log.info("Transfo options: %s" % second_pass_transfo_opts)
mtree = IterativeMatcher(filename, filetype=filetype,
opts=second_pass_opts,
transfo_opts=second_pass_transfo_opts)
m = mtree.matched()
if 'language' not in m and 'subtitleLanguage' not in m:
if 'language' not in m and 'subtitleLanguage' not in m or 'title' not in m:
return m
# if we found some language, make sure we didn't cut a title or sth...
@@ -123,51 +181,10 @@ def _guess_filename(filename, filetype):
opts=['nolanguage', 'nocountry'])
m2 = mtree2.matched()
if m.get('title') is None:
return m
if m.get('title') != m2.get('title'):
title = next(find_nodes(mtree.match_tree, 'title'))
title2 = next(find_nodes(mtree2.match_tree, 'title'))
langs = list(find_nodes(mtree.match_tree, ['language', 'subtitleLanguage']))
if not langs:
return warning('A weird error happened with language detection')
# find the language that is likely more relevant
for lng in langs:
if lng.value in title2.value:
# if the language was detected as part of a potential title,
# look at this one in particular
lang = lng
break
else:
# pick the first one if we don't have a better choice
lang = langs[0]
# language code are rarely part of a title, and those
# should be handled by the Language exceptions anyway
if len(lang.value) <= 3:
return m
# if filetype is subtitle and the language appears last, just before
# the extension, then it is likely a subtitle language
parts = clean_string(title.root.value).split()
if (m['type'] in ['moviesubtitle', 'episodesubtitle'] and
parts.index(lang.value) == len(parts) - 2):
return m
# if the language was in the middle of the other potential title,
# keep the other title (eg: The Italian Job), except if it is at the
# very beginning, in which case we consider it an error
if m2['title'].startswith(lang.value):
return m
elif lang.value in title2.value:
return m2
# if a node is in an explicit group, then the correct title is probably
# the other one
if title.root.node_at(title.node_idx[:2]).is_explicit():
@@ -175,9 +192,6 @@ def _guess_filename(filename, filetype):
elif title2.root.node_at(title2.node_idx[:2]).is_explicit():
return m
return warning('Not sure of the title because of the language position')
return m
libs/guessit/__main__.py
+18 -7
View File
@@ -24,16 +24,19 @@ from guessit import u
from guessit import slogging, guess_file_info
from optparse import OptionParser
import logging
import sys
import os
import locale
def detect_filename(filename, filetype, info=['filename']):
def detect_filename(filename, filetype, info=['filename'], advanced = False):
filename = u(filename)
print('For:', filename)
print('GuessIt found:', guess_file_info(filename, filetype, info).nice_string())
print('GuessIt found:', guess_file_info(filename, filetype, info).nice_string(advanced))
def run_demo(episodes=True, movies=True):
def run_demo(episodes=True, movies=True, advanced=False):
# NOTE: tests should not be added here but rather in the tests/ folder
# this is just intended as a quick example
if episodes:
@@ -50,7 +53,7 @@ def run_demo(episodes=True, movies=True):
for f in testeps:
print('-'*80)
detect_filename(f, filetype='episode')
detect_filename(f, filetype='episode', advanced=advanced)
if movies:
@@ -77,12 +80,17 @@ def run_demo(episodes=True, movies=True):
for f in testmovies:
print('-'*80)
detect_filename(f, filetype = 'movie')
detect_filename(f, filetype = 'movie', advanced = advanced)
def main():
slogging.setupLogging()
# see http://bugs.python.org/issue2128
if sys.version_info.major < 3 and os.name == 'nt':
for i, a in enumerate(sys.argv):
sys.argv[i] = a.decode(locale.getpreferredencoding())
parser = OptionParser(usage = 'usage: %prog [options] file1 [file2...]')
parser.add_option('-v', '--verbose', action='store_true', dest='verbose', default=False,
help = 'display debug output')
@@ -92,6 +100,8 @@ def main():
'them, comma-separated')
parser.add_option('-t', '--type', dest = 'filetype', default = 'autodetect',
help = 'the suggested file type: movie, episode or autodetect')
parser.add_option('-a', '--advanced', dest = 'advanced', action='store_true', default = False,
help = 'display advanced information for filename guesses, as json output')
parser.add_option('-d', '--demo', action='store_true', dest='demo', default=False,
help = 'run a few builtin tests instead of analyzing a file')
@@ -100,13 +110,14 @@ def main():
logging.getLogger('guessit').setLevel(logging.DEBUG)
if options.demo:
run_demo(episodes=True, movies=True)
run_demo(episodes=True, movies=True, advanced=options.advanced)
else:
if args:
for filename in args:
detect_filename(filename,
filetype = options.filetype,
info = options.info.split(','))
info = options.info.split(','),
advanced = options.advanced)
else:
parser.print_help()
libs/guessit/fileutils.py
+4 -2
View File
@@ -44,13 +44,14 @@ def split_path(path):
result = []
while True:
head, tail = os.path.split(path)
headlen = len(head)
# on Unix systems, the root folder is '/'
if head == '/' and tail == '':
if head and head == '/'*headlen and tail == '':
return ['/'] + result
# on Windows, the root folder is a drive letter (eg: 'C:\') or for shares \\
if ((len(head) == 3 and head[1:] == ':\\') or (len(head) == 2 and head == '\\\\')) and tail == '':
if ((headlen == 3 and head[1:] == ':\\') or (headlen == 2 and head == '\\\\')) and tail == '':
return [head] + result
if head == '' and tail == '':
@@ -61,6 +62,7 @@ def split_path(path):
path = head
continue
# otherwise, add the last path fragment and keep splitting
result = [tail] + result
path = head
libs/guessit/guess.py
+46 -19
View File
@@ -41,15 +41,21 @@ class Guess(UnicodeMixin, dict):
confidence = kwargs.pop('confidence')
except KeyError:
confidence = 0
try:
raw = kwargs.pop('raw')
except KeyError:
raw = None
dict.__init__(self, *args, **kwargs)
self._confidence = {}
self._raw = {}
for prop in self:
self._confidence[prop] = confidence
def to_dict(self):
self._raw[prop] = raw
def to_dict(self, advanced=False):
data = dict(self)
for prop, value in data.items():
if isinstance(value, datetime.date):
@@ -58,46 +64,65 @@ class Guess(UnicodeMixin, dict):
data[prop] = u(value)
elif isinstance(value, list):
data[prop] = [u(x) for x in value]
if advanced:
data[prop] = {"value": data[prop], "raw": self.raw(prop), "confidence": self.confidence(prop)}
return data
def nice_string(self):
data = self.to_dict()
parts = json.dumps(data, indent=4).split('\n')
for i, p in enumerate(parts):
if p[:5] != ' "':
continue
prop = p.split('"')[1]
parts[i] = (' [%.2f] "' % self.confidence(prop)) + p[5:]
return '\n'.join(parts)
def nice_string(self, advanced=False):
if advanced:
data = self.to_dict(advanced)
return json.dumps(data, indent=4)
else:
data = self.to_dict()
parts = json.dumps(data, indent=4).split('\n')
for i, p in enumerate(parts):
if p[:5] != ' "':
continue
prop = p.split('"')[1]
parts[i] = (' [%.2f] "' % self.confidence(prop)) + p[5:]
return '\n'.join(parts)
def __unicode__(self):
return u(self.to_dict())
def confidence(self, prop):
return self._confidence.get(prop, -1)
def raw(self, prop):
return self._raw.get(prop, None)
def set(self, prop, value, confidence=None):
def set(self, prop, value, confidence=None, raw=None):
self[prop] = value
if confidence is not None:
self._confidence[prop] = confidence
if raw is not None:
self._raw[prop] = raw
def set_confidence(self, prop, value):
self._confidence[prop] = value
def set_raw(self, prop, value):
self._raw[prop] = value
def update(self, other, confidence=None):
def update(self, other, confidence=None, raw=None):
dict.update(self, other)
if isinstance(other, Guess):
for prop in other:
self._confidence[prop] = other.confidence(prop)
self._raw[prop] = other.raw(prop)
if confidence is not None:
for prop in other:
self._confidence[prop] = confidence
if raw is not None:
for prop in other:
self._raw[prop] = raw
def update_highest_confidence(self, other):
"""Update this guess with the values from the given one. In case
there is property present in both, only the one with the highest one
@@ -110,6 +135,7 @@ class Guess(UnicodeMixin, dict):
continue
self[prop] = other[prop]
self._confidence[prop] = other.confidence(prop)
self._raw[prop] = other.raw(prop)
def choose_int(g1, g2):
@@ -181,7 +207,7 @@ def choose_string(g1, g2):
elif v1l in v2l:
return (v1, combined_prob)
# in case of conflict, return the one with highest priority
# in case of conflict, return the one with highest confidence
else:
if c1 > c2:
return (v1, c1 - c2)
@@ -288,7 +314,8 @@ def merge_all(guesses, append=None):
result.set(prop, result.get(prop, []) + [g[prop]],
# TODO: what to do with confidence here? maybe an
# arithmetic mean...
confidence=g.confidence(prop))
confidence=g.confidence(prop),
raw=g.raw(prop))
del g[prop]
libs/guessit/language.py
+11 -1
View File
@@ -296,7 +296,7 @@ UNDETERMINED = Language('und')
ALL_LANGUAGES = frozenset(Language(lng) for lng in lng_all_names) - frozenset([UNDETERMINED])
ALL_LANGUAGES_NAMES = lng_all_names
def search_language(string, lang_filter=None):
def search_language(string, lang_filter=None, skip=None):
"""Looks for language patterns, and if found return the language object,
its group span and an associated confidence.
@@ -345,6 +345,16 @@ def search_language(string, lang_filter=None):
if pos != -1:
end = pos + len(lang)
# skip if span in in skip list
while skip and (pos - 1, end - 1) in skip:
pos = slow.find(lang, end)
if pos == -1:
continue
end = pos + len(lang)
if pos == -1:
continue
# make sure our word is always surrounded by separators
if slow[pos - 1] not in sep or slow[end] not in sep:
continue
libs/guessit/matcher.py
+37 -14
View File
@@ -21,14 +21,14 @@
from __future__ import unicode_literals
from guessit import PY3, u, base_text_type
from guessit.matchtree import MatchTree
from guessit.textutils import normalize_unicode
from guessit.textutils import normalize_unicode, clean_string
import logging
log = logging.getLogger(__name__)
class IterativeMatcher(object):
def __init__(self, filename, filetype='autodetect', opts=None):
def __init__(self, filename, filetype='autodetect', opts=None, transfo_opts=None):
"""An iterative matcher tries to match different patterns that appear
in the filename.
@@ -38,7 +38,8 @@ class IterativeMatcher(object):
a movie.
The recognized 'filetype' values are:
[ autodetect, subtitle, movie, moviesubtitle, episode, episodesubtitle ]
[ autodetect, subtitle, info, movie, moviesubtitle, movieinfo, episode,
episodesubtitle, episodeinfo ]
The IterativeMatcher works mainly in 2 steps:
@@ -61,15 +62,20 @@ class IterativeMatcher(object):
it corresponds to a video codec, denoted by the letter'v' in the 4th line.
(for more info, see guess.matchtree.to_string)
Second, it tries to merge all this information into a single object
containing all the found properties, and does some (basic) conflict
resolution when they arise.
Second, it tries to merge all this information into a single object
containing all the found properties, and does some (basic) conflict
resolution when they arise.
When you create the Matcher, you can pass it:
- a list 'opts' of option names, that act as global flags
- a dict 'transfo_opts' of { transfo_name: (transfo_args, transfo_kwargs) }
with which to call the transfo.process() function.
"""
valid_filetypes = ('autodetect', 'subtitle', 'video',
'movie', 'moviesubtitle',
'episode', 'episodesubtitle')
valid_filetypes = ('autodetect', 'subtitle', 'info', 'video',
'movie', 'moviesubtitle', 'movieinfo',
'episode', 'episodesubtitle', 'episodeinfo')
if filetype not in valid_filetypes:
raise ValueError("filetype needs to be one of %s" % valid_filetypes)
if not PY3 and not isinstance(filename, unicode):
@@ -80,10 +86,22 @@ class IterativeMatcher(object):
if opts is None:
opts = []
elif isinstance(opts, base_text_type):
opts = opts.split()
if not isinstance(opts, list):
raise ValueError('opts must be a list of option names! Received: type=%s val=%s',
type(opts), opts)
if transfo_opts is None:
transfo_opts = {}
if not isinstance(transfo_opts, dict):
raise ValueError('transfo_opts must be a dict of { transfo_name: (args, kwargs) }. '+
'Received: type=%s val=%s', type(transfo_opts), transfo_opts)
self.match_tree = MatchTree(filename)
# sanity check: make sure we don't process a (mostly) empty string
if clean_string(filename) == '':
return
mtree = self.match_tree
mtree.guess.set('type', filetype, confidence=1.0)
@@ -91,7 +109,11 @@ class IterativeMatcher(object):
transfo = __import__('guessit.transfo.' + transfo_name,
globals=globals(), locals=locals(),
fromlist=['process'], level=0)
transfo.process(mtree, *args, **kwargs)
default_args, default_kwargs = transfo_opts.get(transfo_name, ((), {}))
all_args = args or default_args
all_kwargs = dict(default_kwargs)
all_kwargs.update(kwargs) # keep all kwargs merged together
transfo.process(mtree, *all_args, **all_kwargs)
# 1- first split our path into dirs + basename + ext
apply_transfo('split_path_components')
@@ -111,7 +133,7 @@ class IterativeMatcher(object):
# - language before episodes_rexps
# - properties before language (eg: he-aac vs hebrew)
# - release_group before properties (eg: XviD-?? vs xvid)
if mtree.guess['type'] in ('episode', 'episodesubtitle'):
if mtree.guess['type'] in ('episode', 'episodesubtitle', 'episodeinfo'):
strategy = [ 'guess_date', 'guess_website', 'guess_release_group',
'guess_properties', 'guess_language',
'guess_video_rexps',
@@ -124,6 +146,7 @@ class IterativeMatcher(object):
if 'nolanguage' in opts:
strategy.remove('guess_language')
for name in strategy:
apply_transfo(name)
@@ -143,7 +166,7 @@ class IterativeMatcher(object):
# 5- try to identify the remaining unknown groups by looking at their
# position relative to other known elements
if mtree.guess['type'] in ('episode', 'episodesubtitle'):
if mtree.guess['type'] in ('episode', 'episodesubtitle', 'episodeinfo'):
apply_transfo('guess_episode_info_from_position')
else:
apply_transfo('guess_movie_title_from_position')
libs/guessit/patterns.py
+15 -9
View File
@@ -25,6 +25,8 @@ import re
subtitle_exts = [ 'srt', 'idx', 'sub', 'ssa' ]
info_exts = [ 'nfo' ]
video_exts = ['3g2', '3gp', '3gp2', 'asf', 'avi', 'divx', 'flv', 'm4v', 'mk2',
'mka', 'mkv', 'mov', 'mp4', 'mp4a', 'mpeg', 'mpg', 'ogg', 'ogm',
'ogv', 'qt', 'ra', 'ram', 'rm', 'ts', 'wav', 'webm', 'wma', 'wmv']
@@ -32,7 +34,7 @@ video_exts = ['3g2', '3gp', '3gp2', 'asf', 'avi', 'divx', 'flv', 'm4v', 'mk2',
group_delimiters = [ '()', '[]', '{}' ]
# separator character regexp
sep = r'[][)(}{+ /\._-]' # regexp art, hehe :D
sep = r'[][,)(}{+ /\._-]' # regexp art, hehe :D
# character used to represent a deleted char (when matching groups)
deleted = '_'
@@ -49,7 +51,7 @@ episode_rexps = [ # ... Season 2 ...
#(r'[Ss](?P<season>[0-9]{1,3})[^0-9]?(?P<bonusNumber>(?:-?[xX-][0-9]{1,3})+)[^0-9]', 1.0, (0, -1)),
# ... 2x13 ...
(r'[^0-9](?P<season>[0-9]{1,2})[^0-9]?(?P<episodeNumber>(?:-?[xX][0-9]{1,3})+)[^0-9]', 1.0, (1, -1)),
(r'[^0-9](?P<season>[0-9]{1,2})[^0-9 .-]?(?P<episodeNumber>(?:-?[xX][0-9]{1,3})+)[^0-9]', 1.0, (1, -1)),
# ... s02 ...
#(sep + r's(?P<season>[0-9]{1,2})' + sep, 0.6, (1, -1)),
@@ -122,9 +124,12 @@ prop_multi = { 'format': { 'DVD': [ 'DVD', 'DVD-Rip', 'VIDEO-TS', 'DVDivX' ],
'VHS': [ 'VHS' ],
'WEB-DL': [ 'WEB-DL' ] },
'is3D': { True: [ '3D' ] },
'screenSize': { '480p': [ '480[pi]?' ],
'720p': [ '720[pi]?' ],
'1080p': [ '1080[pi]?' ] },
'1080i': [ '1080i' ],
'1080p': [ '1080p', '1080[^i]' ] },
'videoCodec': { 'XviD': [ 'Xvid' ],
'DivX': [ 'DVDivX', 'DivX' ],
@@ -140,7 +145,7 @@ prop_multi = { 'format': { 'DVD': [ 'DVD', 'DVD-Rip', 'VIDEO-TS', 'DVDivX' ],
'DTS': [ 'DTS' ],
'AAC': [ 'He-AAC', 'AAC-He', 'AAC' ] },
'audioChannels': { '5.1': [ r'5\.1', 'DD5[\._ ]1', '5ch' ] },
'audioChannels': { '5.1': [ r'5\.1', 'DD5[._ ]1', '5ch' ] },
'episodeFormat': { 'Minisode': [ 'Minisodes?' ] }
@@ -170,7 +175,7 @@ prop_single = { 'releaseGroup': [ 'ESiR', 'WAF', 'SEPTiC', r'\[XCT\]', 'iNT', 'P
}
_dash = '-'
_psep = '[-\. _]?'
_psep = '[-. _]?'
def _to_rexp(prop):
return re.compile(prop.replace(_dash, _psep), re.IGNORECASE)
@@ -237,8 +242,9 @@ def canonical_form(string):
def compute_canonical_form(property_name, value):
"""Return the canonical form of a property given its type if it is a valid
one, None otherwise."""
for canonical_form, rexps in properties_rexps[property_name].items():
for rexp in rexps:
if rexp.match(value):
return canonical_form
if isinstance(value, basestring):
for canonical_form, rexps in properties_rexps[property_name].items():
for rexp in rexps:
if rexp.match(value):
return canonical_form
return None
libs/guessit/slogging.py
+5 -3
View File
@@ -31,14 +31,15 @@ RED_FONT = "\x1B[0;31m"
RESET_FONT = "\x1B[0m"
def setupLogging(colored=True, with_time=False, with_thread=False, filename=None):
def setupLogging(colored=True, with_time=False, with_thread=False, filename=None, with_lineno=False):
"""Set up a nice colored logger as the main application logger."""
class SimpleFormatter(logging.Formatter):
def __init__(self, with_time, with_thread):
self.fmt = (('%(asctime)s ' if with_time else '') +
'%(levelname)-8s ' +
'[%(name)s:%(funcName)s]' +
'[%(name)s:%(funcName)s' +
(':%(lineno)s' if with_lineno else '') + ']' +
('[%(threadName)s]' if with_thread else '') +
' -- %(message)s')
logging.Formatter.__init__(self, self.fmt)
@@ -47,7 +48,8 @@ def setupLogging(colored=True, with_time=False, with_thread=False, filename=None
def __init__(self, with_time, with_thread):
self.fmt = (('%(asctime)s ' if with_time else '') +
'-CC-%(levelname)-8s ' +
BLUE_FONT + '[%(name)s:%(funcName)s]' +
BLUE_FONT + '[%(name)s:%(funcName)s' +
(':%(lineno)s' if with_lineno else '') + ']' +
RESET_FONT + ('[%(threadName)s]' if with_thread else '') +
' -- %(message)s')
libs/guessit/textutils.py
+7 -4
View File
@@ -43,10 +43,13 @@ def strip_brackets(s):
return s
def clean_string(s):
for c in sep[:-2]: # do not remove dashes ('-')
s = s.replace(c, ' ')
parts = s.split()
def clean_string(st):
for c in sep:
# do not remove certain chars
if c in ['-', ',']:
continue
st = st.replace(c, ' ')
parts = st.split()
result = ' '.join(p for p in parts if p != '')
# now also remove dashes on the outer part of the string
libs/guessit/transfo/__init__.py
+15 -7
View File
@@ -28,7 +28,7 @@ log = logging.getLogger(__name__)
def found_property(node, name, confidence):
node.guess = Guess({name: node.clean_value}, confidence=confidence)
node.guess = Guess({name: node.clean_value}, confidence=confidence, raw=node.value)
log.debug('Found with confidence %.2f: %s' % (confidence, node.guess))
@@ -52,11 +52,17 @@ def format_guess(guess):
def find_and_split_node(node, strategy, logger):
string = ' %s ' % node.value # add sentinels
for matcher, confidence in strategy:
for matcher, confidence, args, kwargs in strategy:
all_args = [string]
if getattr(matcher, 'use_node', False):
result, span = matcher(string, node)
all_args.append(node)
if args:
all_args.append(args)
if kwargs:
result, span = matcher(*all_args, **kwargs)
else:
result, span = matcher(string)
result, span = matcher(*all_args)
if result:
# readjust span to compensate for sentinels
@@ -69,7 +75,7 @@ def find_and_split_node(node, strategy, logger):
if confidence is None:
confidence = 1.0
guess = format_guess(Guess(result, confidence=confidence))
guess = format_guess(Guess(result, confidence=confidence, raw=string[span[0] + 1:span[1] + 1]))
msg = 'Found with confidence %.2f: %s' % (confidence, guess)
(logger or log).debug(msg)
@@ -84,10 +90,12 @@ def find_and_split_node(node, strategy, logger):
class SingleNodeGuesser(object):
def __init__(self, guess_func, confidence, logger=None):
def __init__(self, guess_func, confidence, logger, *args, **kwargs):
self.guess_func = guess_func
self.confidence = confidence
self.logger = logger
self.args = args
self.kwargs = kwargs
def process(self, mtree):
# strategy is a list of pairs (guesser, confidence)
@@ -95,7 +103,7 @@ class SingleNodeGuesser(object):
# it will override it, otherwise it will leave the guess confidence
# - if the guesser returns a simple dict as a guess and confidence is
# specified, it will use it, or 1.0 otherwise
strategy = [ (self.guess_func, self.confidence) ]
strategy = [ (self.guess_func, self.confidence, self.args, self.kwargs) ]
for node in mtree.unidentified_leaves():
find_and_split_node(node, strategy, self.logger)
libs/guessit/transfo/guess_country.py
+1 -1
View File
@@ -45,4 +45,4 @@ def process(mtree):
except ValueError:
continue
node.guess = Guess(country=country, confidence=1.0)
node.guess = Guess(country=country, confidence=1.0, raw=c)
libs/guessit/transfo/guess_episodes_rexps.py
+5 -10
View File
@@ -40,27 +40,22 @@ def guess_episodes_rexps(string):
for rexp, confidence, span_adjust in episode_rexps:
match = re.search(rexp, string, re.IGNORECASE)
if match:
guess = Guess(match.groupdict(), confidence=confidence)
span = (match.start() + span_adjust[0],
span = (match.start() + span_adjust[0],
match.end() + span_adjust[1])
# episodes which have a season > 30 are most likely errors
# (Simpsons is at 24!)
if int(guess.get('season', 0)) > 30:
continue
guess = Guess(match.groupdict(), confidence=confidence, raw=string[span[0]:span[1]])
# decide whether we have only a single episode number or an
# episode list
if guess.get('episodeNumber'):
eplist = number_list(guess['episodeNumber'])
guess.set('episodeNumber', eplist[0], confidence=confidence)
guess.set('episodeNumber', eplist[0], confidence=confidence, raw=string[span[0]:span[1]])
if len(eplist) > 1:
guess.set('episodeList', eplist, confidence=confidence)
guess.set('episodeList', eplist, confidence=confidence, raw=string[span[0]:span[1]])
if guess.get('bonusNumber'):
eplist = number_list(guess['bonusNumber'])
guess.set('bonusNumber', eplist[0], confidence=confidence)
guess.set('bonusNumber', eplist[0], confidence=confidence, raw=string[span[0]:span[1]])
return guess, span
libs/guessit/transfo/guess_filetype.py
+25 -2
View File
@@ -20,7 +20,7 @@
from __future__ import unicode_literals
from guessit import Guess
from guessit.patterns import (subtitle_exts, video_exts, episode_rexps,
from guessit.patterns import (subtitle_exts, info_exts, video_exts, episode_rexps,
find_properties, compute_canonical_form)
from guessit.date import valid_year
from guessit.textutils import clean_string
@@ -53,12 +53,16 @@ def guess_filetype(mtree, filetype):
filetype_container[0] = 'episode'
elif filetype_container[0] == 'subtitle':
filetype_container[0] = 'episodesubtitle'
elif filetype_container[0] == 'info':
filetype_container[0] = 'episodeinfo'
def upgrade_movie():
if filetype_container[0] == 'video':
filetype_container[0] = 'movie'
elif filetype_container[0] == 'subtitle':
filetype_container[0] = 'moviesubtitle'
elif filetype_container[0] == 'info':
filetype_container[0] = 'movieinfo'
def upgrade_subtitle():
if 'movie' in filetype_container[0]:
@@ -68,6 +72,14 @@ def guess_filetype(mtree, filetype):
else:
filetype_container[0] = 'subtitle'
def upgrade_info():
if 'movie' in filetype_container[0]:
filetype_container[0] = 'movieinfo'
elif 'episode' in filetype_container[0]:
filetype_container[0] = 'episodeinfo'
else:
filetype_container[0] = 'info'
def upgrade(type='unknown'):
if filetype_container[0] == 'autodetect':
filetype_container[0] = type
@@ -78,6 +90,9 @@ def guess_filetype(mtree, filetype):
if fileext in subtitle_exts:
upgrade_subtitle()
other = { 'container': fileext }
elif fileext in info_exts:
upgrade_info()
other = { 'container': fileext }
elif fileext in video_exts:
upgrade(type='video')
other = { 'container': fileext }
@@ -104,17 +119,20 @@ def guess_filetype(mtree, filetype):
fname = clean_string(filename).lower()
for m in MOVIES:
if m in fname:
log.debug('Found in exception list of movies -> type = movie')
upgrade_movie()
for s in SERIES:
if s in fname:
log.debug('Found in exception list of series -> type = episode')
upgrade_episode()
# now look whether there are some specific hints for episode vs movie
if filetype_container[0] in ('video', 'subtitle'):
if filetype_container[0] in ('video', 'subtitle', 'info'):
# if we have an episode_rexp (eg: s02e13), it is an episode
for rexp, _, _ in episode_rexps:
match = re.search(rexp, filename, re.IGNORECASE)
if match:
log.debug('Found matching regexp: "%s" (string = "%s") -> type = episode', rexp, match.group())
upgrade_episode()
break
@@ -133,24 +151,29 @@ def guess_filetype(mtree, filetype):
possible = False
if possible:
log.debug('Found possible episode number: %s (from string "%s") -> type = episode', epnumber, match.group())
upgrade_episode()
# if we have certain properties characteristic of episodes, it is an ep
for prop, value, _, _ in find_properties(filename):
log.debug('prop: %s = %s' % (prop, value))
if prop == 'episodeFormat':
log.debug('Found characteristic property of episodes: %s = "%s"', prop, value)
upgrade_episode()
break
elif compute_canonical_form('format', value) == 'DVB':
log.debug('Found characteristic property of episodes: %s = "%s"', prop, value)
upgrade_episode()
break
# origin-specific type
if 'tvu.org.ru' in filename:
log.debug('Found characteristic property of episodes: %s = "%s"', prop, value)
upgrade_episode()
# if no episode info found, assume it's a movie
log.debug('Nothing characteristic found, assuming type = movie')
upgrade_movie()
filetype = filetype_container[0]
libs/guessit/transfo/guess_language.py
+18 -6
View File
@@ -22,22 +22,34 @@ from __future__ import unicode_literals
from guessit import Guess
from guessit.transfo import SingleNodeGuesser
from guessit.language import search_language
from guessit.textutils import clean_string, find_words
import logging
log = logging.getLogger(__name__)
def guess_language(string):
language, span, confidence = search_language(string)
def guess_language(string, node, skip=None):
if skip:
relative_skip = []
for entry in skip:
node_idx = entry['node_idx']
span = entry['span']
if node_idx == node.node_idx[:len(node_idx)]:
relative_span = (span[0] - node.offset + 1, span[1] - node.offset + 1)
relative_skip.append(relative_span)
skip = relative_skip
language, span, confidence = search_language(string, skip=skip)
if language:
return (Guess({'language': language},
confidence=confidence),
confidence=confidence,
raw= string[span[0]:span[1]]),
span)
return None, None
guess_language.use_node = True
def process(mtree):
SingleNodeGuesser(guess_language, None, log).process(mtree)
def process(mtree, *args, **kwargs):
SingleNodeGuesser(guess_language, None, log, *args, **kwargs).process(mtree)
# Note: 'language' is promoted to 'subtitleLanguage' in the post_process transfo
libs/guessit/transfo/guess_movie_title_from_position.py
+2 -1
View File
@@ -29,7 +29,8 @@ log = logging.getLogger(__name__)
def process(mtree):
def found_property(node, name, value, confidence):
node.guess = Guess({ name: value },
confidence=confidence)
confidence=confidence,
raw=value)
log.debug('Found with confidence %.2f: %s' % (confidence, node.guess))
def found_title(node, confidence):
libs/guessit/transfo/guess_video_rexps.py
+4 -3
View File
@@ -38,9 +38,10 @@ def guess_video_rexps(string):
# the soonest that we can catch it)
if metadata.get('cdNumberTotal', -1) is None:
del metadata['cdNumberTotal']
return (Guess(metadata, confidence=confidence),
(match.start() + span_adjust[0],
match.end() + span_adjust[1] - 2))
span = (match.start() + span_adjust[0],
match.end() + span_adjust[1] - 2)
return (Guess(metadata, confidence=confidence, raw=string[span[0]:span[1]]),
span)
return None, None
libs/guessit/transfo/guess_weak_episodes_rexps.py
+2 -2
View File
@@ -48,9 +48,9 @@ def guess_weak_episodes_rexps(string, node):
continue
return Guess({ 'season': season,
'episodeNumber': epnum },
confidence=0.6), span
confidence=0.6, raw=string[span[0]:span[1]]), span
else:
return Guess(metadata, confidence=0.3), span
return Guess(metadata, confidence=0.3, raw=string[span[0]:span[1]]), span
return None, None