You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
bazarr/libs/guessit/rules/properties/title.py

350 lines
13 KiB

#!/usr/bin/env python
# -*- coding: utf-8 -*-
"""
title property
"""
from rebulk import Rebulk, Rule, AppendMatch, RemoveMatch, AppendTags
from rebulk.formatters import formatters
from .film import FilmTitleRule
from .language import (
SubtitlePrefixLanguageRule,
SubtitleSuffixLanguageRule,
SubtitleExtensionRule,
NON_SPECIFIC_LANGUAGES
)
from ..common import seps, title_seps
from ..common.comparators import marker_sorted
from ..common.expected import build_expected_function
from ..common.formatters import cleanup, reorder_title
from ..common.pattern import is_disabled
from ..common.validators import seps_surround
def title(config): # pylint:disable=unused-argument
"""
Builder for rebulk object.
:param config: rule configuration
:type config: dict
:return: Created Rebulk object
:rtype: Rebulk
"""
rebulk = Rebulk(disabled=lambda context: is_disabled(context, 'title'))
rebulk.rules(TitleFromPosition, PreferTitleWithYear)
expected_title = build_expected_function('expected_title')
rebulk.functional(expected_title, name='title', tags=['expected', 'title'],
validator=seps_surround,
formatter=formatters(cleanup, reorder_title),
conflict_solver=lambda match, other: other,
disabled=lambda context: not context.get('expected_title'))
return rebulk
class TitleBaseRule(Rule):
"""
Add title match in existing matches
"""
# pylint:disable=no-self-use,unused-argument
consequence = [AppendMatch, RemoveMatch]
def __init__(self, match_name, match_tags=None, alternative_match_name=None):
super().__init__()
self.match_name = match_name
self.match_tags = match_tags
self.alternative_match_name = alternative_match_name
def hole_filter(self, hole, matches):
"""
Filter holes for titles.
:param hole:
:type hole:
:param matches:
:type matches:
:return:
:rtype:
"""
return True
def filepart_filter(self, filepart, matches):
"""
Filter filepart for titles.
:param filepart:
:type filepart:
:param matches:
:type matches:
:return:
:rtype:
"""
return True
def holes_process(self, holes, matches):
"""
process holes
:param holes:
:type holes:
:param matches:
:type matches:
:return:
:rtype:
"""
cropped_holes = []
group_markers = matches.markers.named('group')
for group_marker in group_markers:
path_marker = matches.markers.at_match(group_marker, predicate=lambda m: m.name == 'path', index=0)
if path_marker and path_marker.span == group_marker.span:
group_markers.remove(group_marker)
for hole in holes:
cropped_holes.extend(hole.crop(group_markers))
return cropped_holes
@staticmethod
def is_ignored(match):
"""
Ignore matches when scanning for title (hole).
Full word language and countries won't be ignored if they are uppercase.
"""
return not (len(match) > 3 and match.raw.isupper()) and match.name in ('language', 'country', 'episode_details')
def should_keep(self, match, to_keep, matches, filepart, hole, starting):
"""
Check if this match should be accepted when ending or starting a hole.
:param match:
:type match:
:param to_keep:
:type to_keep: list[Match]
:param matches:
:type matches: Matches
:param hole: the filepart match
:type hole: Match
:param hole: the hole match
:type hole: Match
:param starting: true if match is starting the hole
:type starting: bool
:return:
:rtype:
"""
if match.name in ('language', 'country'):
# Keep language if exactly matching the hole.
if len(hole.value) == len(match.raw):
return True
# Keep language if other languages exists in the filepart.
outside_matches = filepart.crop(hole)
other_languages = []
for outside in outside_matches:
other_languages.extend(matches.range(outside.start, outside.end,
lambda c_match: c_match.name == match.name and
c_match not in to_keep and
c_match.value not in NON_SPECIFIC_LANGUAGES))
if not other_languages and (not starting or len(match.raw) <= 3):
return True
return False
def should_remove(self, match, matches, filepart, hole, context):
"""
Check if this match should be removed after beeing ignored.
:param match:
:param matches:
:param filepart:
:param hole:
:return:
"""
if context.get('type') == 'episode' and match.name == 'episode_details':
return match.start >= hole.start and match.end <= hole.end
return True
def check_titles_in_filepart(self, filepart, matches, context): # pylint:disable=inconsistent-return-statements
"""
Find title in filepart (ignoring language)
"""
# pylint:disable=too-many-locals,too-many-branches,too-many-statements
start, end = filepart.span
holes = matches.holes(start, end + 1, formatter=formatters(cleanup, reorder_title),
ignore=self.is_ignored,
predicate=lambda m: m.value)
holes = self.holes_process(holes, matches)
for hole in holes:
if not hole or (self.hole_filter and not self.hole_filter(hole, matches)):
continue
to_remove = []
to_keep = []
ignored_matches = matches.range(hole.start, hole.end, self.is_ignored)
if ignored_matches:
for ignored_match in reversed(ignored_matches):
# pylint:disable=undefined-loop-variable, cell-var-from-loop
trailing = matches.chain_before(hole.end, seps, predicate=lambda m: m == ignored_match)
if trailing:
should_keep = self.should_keep(ignored_match, to_keep, matches, filepart, hole, False)
if should_keep:
# pylint:disable=unpacking-non-sequence
try:
append, crop = should_keep
except TypeError:
append, crop = should_keep, should_keep
if append:
to_keep.append(ignored_match)
if crop:
hole.end = ignored_match.start
for ignored_match in ignored_matches:
if ignored_match not in to_keep:
starting = matches.chain_after(hole.start, seps,
predicate=lambda m: m == ignored_match)
if starting:
should_keep = self.should_keep(ignored_match, to_keep, matches, filepart, hole, True)
if should_keep:
# pylint:disable=unpacking-non-sequence
try:
append, crop = should_keep
except TypeError:
append, crop = should_keep, should_keep
if append:
to_keep.append(ignored_match)
if crop:
hole.start = ignored_match.end
for match in ignored_matches:
if self.should_remove(match, matches, filepart, hole, context):
to_remove.append(match)
for keep_match in to_keep:
if keep_match in to_remove:
to_remove.remove(keep_match)
if hole and hole.value:
hole.name = self.match_name
hole.tags = self.match_tags
if self.alternative_match_name:
# Split and keep values that can be a title
titles = hole.split(title_seps, lambda m: m.value)
for title_match in list(titles[1:]):
previous_title = titles[titles.index(title_match) - 1]
separator = matches.input_string[previous_title.end:title_match.start]
if len(separator) == 1 and separator == '-' \
and previous_title.raw[-1] not in seps \
and title_match.raw[0] not in seps:
titles[titles.index(title_match) - 1].end = title_match.end
titles.remove(title_match)
else:
title_match.name = self.alternative_match_name
else:
titles = [hole]
return titles, to_remove
def when(self, matches, context):
ret = []
to_remove = []
if matches.named(self.match_name, lambda match: 'expected' in match.tags):
return False
fileparts = [filepart for filepart in list(marker_sorted(matches.markers.named('path'), matches))
if not self.filepart_filter or self.filepart_filter(filepart, matches)]
# Priorize fileparts containing the year
years_fileparts = []
for filepart in fileparts:
year_match = matches.range(filepart.start, filepart.end, lambda match: match.name == 'year', 0)
if year_match:
years_fileparts.append(filepart)
for filepart in fileparts:
try:
years_fileparts.remove(filepart)
except ValueError:
pass
titles = self.check_titles_in_filepart(filepart, matches, context)
if titles:
titles, to_remove_c = titles
ret.extend(titles)
to_remove.extend(to_remove_c)
break
# Add title match in all fileparts containing the year.
for filepart in years_fileparts:
titles = self.check_titles_in_filepart(filepart, matches, context)
if titles:
# pylint:disable=unbalanced-tuple-unpacking
titles, to_remove_c = titles
ret.extend(titles)
to_remove.extend(to_remove_c)
if ret or to_remove:
return ret, to_remove
return False
class TitleFromPosition(TitleBaseRule):
"""
Add title match in existing matches
"""
dependency = [FilmTitleRule, SubtitlePrefixLanguageRule, SubtitleSuffixLanguageRule, SubtitleExtensionRule]
properties = {'title': [None], 'alternative_title': [None]}
def __init__(self):
super().__init__('title', ['title'], 'alternative_title')
def enabled(self, context):
return not is_disabled(context, 'alternative_title')
class PreferTitleWithYear(Rule):
"""
Prefer title where filepart contains year.
"""
dependency = TitleFromPosition
consequence = [RemoveMatch, AppendTags(['equivalent-ignore'])]
properties = {'title': [None]}
def when(self, matches, context):
with_year_in_group = []
with_year = []
titles = matches.named('title')
for title_match in titles:
filepart = matches.markers.at_match(title_match, lambda marker: marker.name == 'path', 0)
if filepart:
year_match = matches.range(filepart.start, filepart.end, lambda match: match.name == 'year', 0)
if year_match:
group = matches.markers.at_match(year_match, lambda m: m.name == 'group')
if group:
with_year_in_group.append(title_match)
else:
with_year.append(title_match)
to_tag = []
if with_year_in_group:
title_values = {title_match.value for title_match in with_year_in_group}
to_tag.extend(with_year_in_group)
elif with_year:
title_values = {title_match.value for title_match in with_year}
to_tag.extend(with_year)
else:
title_values = {title_match.value for title_match in titles}
to_remove = []
for title_match in titles:
if title_match.value not in title_values:
to_remove.append(title_match)
if to_remove or to_tag:
return to_remove, to_tag
return False