Update libs
This commit is contained in:
Ruud
+62
-48
@@ -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
|
||||
|
||||
|
||||
|
||||
Reference in New Issue
Block a user