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/subliminal_patch/providers/supersubtitles.py

564 lines
23 KiB

6 years ago
# coding=utf-8
import io
import logging
import re
import time
from babelfish import language_converters
from subzero.language import Language
from requests import Session
from requests.exceptions import JSONDecodeError
4 years ago
import urllib.parse
from random import randint
6 years ago
from subliminal.subtitle import fix_line_ending
from subliminal_patch.providers import Provider
from subliminal_patch.providers.mixins import ProviderSubtitleArchiveMixin
from subliminal.providers import ParserBeautifulSoup
4 years ago
from bs4.element import Tag, NavigableString
6 years ago
from subliminal.score import get_equivalent_release_groups
from subliminal_patch.subtitle import Subtitle, guess_matches
from subliminal_patch.exceptions import APIThrottled
6 years ago
from subliminal.utils import sanitize, sanitize_release_group
from subliminal.video import Episode, Movie
from zipfile import ZipFile, is_zipfile
6 years ago
from rarfile import RarFile, is_rarfile
4 years ago
from subliminal_patch.utils import sanitize, fix_inconsistent_naming
6 years ago
from guessit import guessit
from .utils import FIRST_THOUSAND_OR_SO_USER_AGENTS as AGENT_LIST
6 years ago
logger = logging.getLogger(__name__)
language_converters.register('supersubtitles = subliminal_patch.converters.supersubtitles:SuperSubtitlesConverter')
4 years ago
def fix_tv_naming(title):
"""Fix TV show titles with inconsistent naming using dictionary, but do not sanitize them.
:param str title: original title.
:return: new title.
:rtype: str
"""
return fix_inconsistent_naming(title, {"DC's Legends of Tomorrow": "Legends of Tomorrow",
4 years ago
"Star Trek: The Next Generation": "Star Trek TNG",
"Loki (aka. Marvel\'s Loki)": "Loki",
"Marvel's": "",
4 years ago
}, True)
def fix_movie_naming(title):
return fix_inconsistent_naming(title, {
}, True)
6 years ago
class SuperSubtitlesSubtitle(Subtitle):
"""SuperSubtitles Subtitle."""
provider_name = 'supersubtitles'
def __str__(self):
subtit = "Subtitle id: " + str(self.subtitle_id) \
+ " Series: " + self.series \
+ " Season: " + str(self.season) \
+ " Episode: " + str(self.episode) \
+ " Version: " + str(self.version) \
+ " Releases: " + str(self.releases) \
+ " DownloadLink: " + str(self.page_link) \
+ " Matches: " + str(self.matches)
if self.year:
subtit = subtit + " Year: " + str(self.year)
return subtit.encode('utf-8')
def __init__(self, language, page_link, subtitle_id, series, season, episode, version,
4 years ago
releases, year, imdb_id, uploader, asked_for_episode=None, asked_for_release_group=None):
6 years ago
super(SuperSubtitlesSubtitle, self).__init__(language, page_link=page_link)
self.subtitle_id = subtitle_id
self.series = series
self.season = season
self.episode = episode
self.version = version
self.releases = releases
self.year = year
4 years ago
self.uploader = uploader
6 years ago
if year:
self.year = int(year)
self.release_info = u" ,".join([u"%s (%s)" % (self.__get_name(), releases[0])] +
(releases[1:] if len(releases) > 1 else []))
6 years ago
self.page_link = page_link
self.asked_for_release_group = asked_for_release_group
self.asked_for_episode = asked_for_episode
self.imdb_id = imdb_id
self.is_pack = True
4 years ago
self.matches = None
6 years ago
def numeric_id(self):
return self.subtitle_id
def __get_name(self):
6 years ago
ep_addon = (" S%02dE%02d" % (self.season, self.episode)) if self.episode else ""
return u"%s%s%s" % (self.series, " (%s)" % self.year if self.year else "", ep_addon)
def __repr__(self):
6 years ago
return '<%s %r [%s]>' % (
self.__class__.__name__, u"%s [%s]" % (self.__get_name(), self.version), self.language)
6 years ago
@property
def id(self):
return str(self.subtitle_id)
def get_matches(self, video):
type_ = "movie" if isinstance(video, Movie) else "episode"
matches = guess_matches(video, guessit(self.release_info, {"type": type_}))
6 years ago
# episode
if isinstance(video, Episode):
# series
if video.series and sanitize(self.series) == sanitize(video.series):
matches.add('series')
# imdb_id
if video.series_imdb_id and self.imdb_id and str(self.imdb_id) == str(video.series_imdb_id):
matches.add('series_imdb_id')
matches.add('series')
matches.add('year')
# year
if 'year' not in matches and 'series' in matches and video.original_series and self.year is None:
6 years ago
matches.add('year')
# movie
elif isinstance(video, Movie):
# title
if video.title and (sanitize(self.series) in (
sanitize(name) for name in [video.title] + video.alternative_titles)):
matches.add('title')
# imdb_id
if video.imdb_id and self.imdb_id == video.imdb_id:
matches.add('imdb_id')
matches.add('title')
matches.add('year')
# year
if video.year and self.year == video.year:
matches.add('year')
# release_group
if video.release_group and self.releases:
video_release_groups = get_equivalent_release_groups(sanitize_release_group(video.release_group))
for release in self.releases:
if any(r in sanitize_release_group(release) for r in video_release_groups):
matches.add('release_group')
if video.resolution and video.resolution in release.lower():
matches.add('resolution')
if video.source and video.source in release.lower():
matches.add('source')
# We don't have to continue in case it is a perfect match
if all(m in matches for m in ['release_group', 'resolution', 'source']):
break
6 years ago
self.matches = matches
return matches
class SuperSubtitlesProvider(Provider, ProviderSubtitleArchiveMixin):
"""SuperSubtitles Provider."""
4 years ago
languages = {Language('hun', 'HU')} | {Language(lang) for lang in [
6 years ago
'hun', 'eng'
]}
video_types = (Episode, Movie)
# https://www.feliratok.eu/?search=&soriSorszam=&nyelv=&sorozatnev=The+Flash+%282014%29&sid=3212&complexsearch=true&knyelv=0&evad=4&epizod1=1&cimke=0&minoseg=0&rlsr=0&tab=all
server_url = 'https://www.feliratok.eu/'
6 years ago
hearing_impaired_verifiable = False
multi_result_throttle = 2 # seconds
4 years ago
def __init__(self):
self.session = None
6 years ago
def initialize(self):
self.session = Session()
self.session.headers = {
'User-Agent': AGENT_LIST[randint(0, len(AGENT_LIST) - 1)],
'Referer': 'https://www.feliratok.eu/index.php'
}
6 years ago
def terminate(self):
self.session.close()
4 years ago
@staticmethod
def get_language(text):
6 years ago
if text == 'Magyar':
return Language.fromsupersubtitles('hu')
if text == 'Angol':
return Language.fromsupersubtitles('en')
return None
def find_imdb_id(self, sub_id):
"""
"""
# TODO: add memoization to this method logic
6 years ago
url = self.server_url + "index.php?tipus=adatlap&azon=a_" + str(sub_id)
# url = https://www.feliratok.eu/index.php?tipus=adatlap&azon=a_1518600916
6 years ago
logger.info('Get IMDB id from URL %s', url)
r = self.session.get(url, timeout=10).content
soup = ParserBeautifulSoup(r, ['lxml'])
links = soup.find_all("a")
for value in links:
if "imdb.com" in str(value):
4 years ago
# <a alt="iMDB" href="http://www.imdb.com/title/tt2357547/" target="_blank"><img alt="iMDB"
# src="img/adatlap/imdb.png"/></a>
imdb_id = re.search(r'(?<=www\.imdb\.com/title/).*(?=/")', str(value))
imdb_id = imdb_id.group() if imdb_id else ''
logger.debug("IMDB ID found: %s", imdb_id)
6 years ago
return imdb_id
return None
def find_id(self, series, year, original_title):
"""
We need to find the id of the series at the following url:
https://www.feliratok.eu/index.php?term=SERIESNAME&nyelv=0&action=autoname
6 years ago
Where SERIESNAME is a searchable string.
The result will be something like this:
4 years ago
[{"name":"DC\u2019s Legends of Tomorrow (2016)","ID":"3725"},{"name":"Miles from Tomorrowland (2015)",
"ID":"3789"},{"name":"No Tomorrow (2016)","ID":"4179"}]
6 years ago
"""
# Search for exact name
url = self.server_url + "index.php?term=" + series + "&nyelv=0&action=autoname"
# url = self.server_url + "index.php?term=" + "fla"+ "&nyelv=0&action=autoname"
logger.info('Get series id from URL %s', url)
r = self.session.get(url, timeout=10)
# r is something like this:
4 years ago
# [{"name":"DC\u2019s Legends of Tomorrow (2016)","ID":"3725"},{"name":"Miles from Tomorrowland (2015)",
# "ID":"3789"},{"name":"No Tomorrow (2016)","ID":"4179"}]
6 years ago
results = r.json()
# check all of the results:
for result in results:
try:
# "name":"Miles from Tomorrowland (2015)","ID":"3789"
4 years ago
result_year = re.search(r"(?<=\()\d\d\d\d(?=\))", result['name'])
result_year = result_year.group() if result_year else ''
6 years ago
except IndexError:
result_year = ""
try:
# "name":"Miles from Tomorrowland (2015)","ID":"3789"
4 years ago
result_title = re.search(r".*(?=\(\d\d\d\d\))", result['name'])
result_title = result_title.group() if result_title else ''
6 years ago
result_id = result['ID']
except IndexError:
continue
for title in (result_title, fix_tv_naming(result_title)):
title = title.strip().replace("<EFBFBD>", "").replace("& ", "").replace(" ", ".")
if not title:
continue
guessable = title.strip() + ".s01e01." + result_year
guess = guessit(guessable, {'type': "episode"})
6 years ago
sanitized_original_title = sanitize(original_title.replace('& ', ''))
guess_title = sanitize(guess['title'])
6 years ago
if sanitized_original_title == guess_title and year and guess['year'] and \
year == guess['year']:
# Return the founded id
return result_id
elif sanitized_original_title == guess_title and not year:
# Return the founded id
return result_id
6 years ago
return None
4 years ago
def query(self, series, languages, video=None):
6 years ago
year = video.year
subtitle = None
if isinstance(video, Episode):
series = video.series
season = video.season
episode = video.episode
4 years ago
# seriesa = series.replace(' ', '+')
6 years ago
# Get ID of series with original name
series_id = self.find_id(series, year, series)
if not series_id:
# If not founded try without ' char
4 years ago
modified_series = urllib.parse.quote_plus(series.replace('\'', ''))
6 years ago
series_id = self.find_id(modified_series, year, series)
if not series_id and modified_series:
# If still not founded try with the longest word is series title
modified_series = modified_series.split('+')
modified_series = max(modified_series, key=len)
series_id = self.find_id(modified_series, year, series)
if not series_id:
return None
subtitle = self.retrieve_series_subtitles(series_id, season, episode, video, languages)
6 years ago
if isinstance(video, Movie):
4 years ago
title = urllib.parse.quote_plus(series)
6 years ago
# https://www.feliratok.eu/index.php?search=The+Hitman%27s+BodyGuard&soriSorszam=&nyelv=&tab=film
6 years ago
url = self.server_url + "index.php?search=" + title + "&soriSorszam=&nyelv=&tab=film"
4 years ago
subtitle = self.process_subs(languages, video, url)
6 years ago
return subtitle
def retrieve_series_subtitles(self, series_id, season, episode, video, languages):
"""
Retrieve subtitles for a given episode
:param series_id: the ID of the series returned by @find_id.
:param season: the season number
:param episode: the episode number
:param video: video details
:param languages: languages to search for
:return: list of subtitles for the given episode
"""
if isinstance(video, Movie):
return None
subtitles = []
logger.info('Getting the list of subtitles for %s', video)
# First, try using every param that we got
episode_subs, season_subs = self.get_subtitle_list(series_id, season, episode, video)
if episode_subs:
sub_list = episode_subs
else:
'''
Sometimes the site is a bit buggy when you are searching for an episode sub that is only present in a
season pack, so we have to make a separate call for that without supplying the episode number
'''
_, sub_list = self.get_subtitle_list(series_id, season, None, video)
series_imdb_id = None
# Convert the list of subtitles for the proper format
for sub in sub_list.values():
'''
Since it is not possible to narrow down the languages in the request, we need to filter out the
inappropriate elements
'''
if sub['language'] in languages:
link = self.server_url + '/index.php?action=letolt&felirat=' + str(sub['id'])
# For episodes we open the series page so all subtitles imdb_id must be the same
if series_imdb_id is None:
series_imdb_id = self.find_imdb_id(sub['id'])
# Let's create a SuperSubtitlesSubtitle instance from the data that we got and add it to the list
subtitles.append(SuperSubtitlesSubtitle(sub['language'], link, sub['id'], sub['name'], sub['season'],
sub['episode'], ', '.join(sub['releases']), sub['releases'],
video.year, series_imdb_id, sub['uploader'], video.episode,
asked_for_release_group=video.release_group))
return subtitles
def get_subtitle_list(self, series_id, season, episode, video):
"""
We can retrieve the list of subtitles for a given show via the following url:
https://www.feliratok.eu/index.php?action=xbmc&sid=SERIES_ID&ev=SEASON&rtol=EPISODE
SERIES_ID is the ID of the show returned by the @find_id method. It is a mandatory parameter.
SEASON is the season number. Optional paramter.
EPISODE is the episode number. Optional parameter (using this param can cause problems).
NOTE: you gonna get back multiple records for the same subtitle, in case it is compatible with multiple releases
"""
# Construct the url
url = self.server_url + "index.php?action=xbmc&sid=" + str(series_id) + "&ev=" + str(season)
# Use the 'rtol' param in case we have a valid episode number
if episode:
url += "&rtol=" + str(episode)
try:
results = self.session.get(url, timeout=10).json()
except JSONDecodeError:
# provider returned improper JSON
results = None
'''
In order to work, the result should be a JSON like this:
{
"10": {
"language":"Angol",
"nev":"The Flash (Season 5) (1080p)",
"baselink":"http://www.feliratok.eu/index.php",
"fnev":"The.Flash.S05.HDTV.WEB.720p.1080p.ENG.zip",
"felirat":"1560706755",
"evad":"5",
"ep":"-1",
"feltolto":"J1GG4",
"pontos_talalat":"111",
"evadpakk":"1"
}, ...
}
'''
subtitle_list = {}
season_pack_list = {}
# Check the results. If a list or a Nonetype is returned, ignore it:
if results and not isinstance(results, list):
for result in results.values():
'''
Gonna get back multiple records for the same subtitle, in case it is compatible with multiple releases,
so we have to group them manually
'''
sub_id = int(result['felirat'])
# 'Nev' is something like this:
# Marvel's The Falcon and the Winter Soldier - 1x05 (WEB.2160p-KOGi)
# or
# Loki (Season 1) (DSNP.WEB-DL.720p-TOMMY)
search_name = re.search(r'^(.*)\s(?:-\s\d+x\d+|(\(Season\s\d+\)))?\s\((.*)\)$', result['nev'])
name = search_name.group(1) if search_name else ''
release = search_name.group(3) if search_name else ''
# In case of 0 it is an episode sub, in other cases, it is a season pack
target = subtitle_list if not int(result['evadpakk']) else season_pack_list
# Check that this sub_id is not already in the list
if sub_id not in target.keys():
target[sub_id] = {
'id': sub_id,
'name': name,
'language': self.get_language(result['language']),
'season': int(result['evad']),
'episode': result['ep'] if not result['evadpakk'] else int(video.episode),
'uploader': result['feltolto'],
'releases': [release],
'fname': result['fnev']
}
else:
target[sub_id]['releases'].append(release)
else:
logger.debug("Invalid results: %s", results)
return subtitle_list, season_pack_list
4 years ago
def process_subs(self, languages, video, url):
if isinstance(video, Episode):
return None
6 years ago
subtitles = []
logger.info('URL for subtitles %s', url)
r = self.session.get(url, timeout=10).content
soup = ParserBeautifulSoup(r, ['lxml'])
tables = soup.find_all("table")
tables = tables[0].find_all("tr")
i = 0
6 years ago
for table in tables:
if "vilagit" in str(table) and i > 1:
asked_for_episode = None
sub_season = None
sub_episode = None
sub_english = table.findAll("div", {"class": "eredeti"})
sub_english_name = re.search(r'(?<=<div class="eredeti">).*?(?=</div>)', str(sub_english))
sub_english_name = sub_english_name.group() if sub_english_name else ''
sub_english_name = sub_english_name.split(' (')[0]
6 years ago
sub_english_name = sub_english_name.replace('&amp;', '&')
4 years ago
sub_version = 'n/a'
if len(str(sub_english).split('(')) > 1:
sub_version = (str(sub_english).split('(')[len(str(sub_english).split('(')) - 1]).split(')')[0]
6 years ago
# <small>Angol</small>
4 years ago
lang = table.find("small")
sub_language = re.search(r"(?<=<small>).*(?=</small>)", str(lang))
sub_language = sub_language.group() if sub_language else ''
sub_language = self.get_language(sub_language)
6 years ago
4 years ago
# <a href="/index.php?action=letolt&amp;fnev=DCs Legends of Tomorrow - 03x11 - Here I Go Again.SVA.
# English.C.orig.Addic7ed.com.srt&amp;felirat=1519162191">
6 years ago
link = str(table.findAll("a")[len(table.findAll("a")) - 1]).replace("amp;", "")
4 years ago
sub_downloadlink = re.search(r'(?<=href="/).*(?=">)', link)
sub_downloadlink = sub_downloadlink.group() if sub_downloadlink else ''
sub_downloadlink = self.server_url + sub_downloadlink
6 years ago
4 years ago
sub_id = re.search(r"(?<=felirat=).*(?=\">)", link)
sub_id = sub_id.group() if sub_id else ''
6 years ago
sub_year = video.year
sub_releases = [s.strip() for s in sub_version.split(',')]
4 years ago
uploader = ''
for item in table.contents[7].contents:
if isinstance(item, Tag):
uploader = item.text.lstrip('\r\n\t\t\t\t\t').rstrip('\r\n\t\t\t\t')
elif isinstance(item, NavigableString):
uploader = item.lstrip('\r\n\t\t\t\t\t').rstrip('\r\n\t\t\t\t')
sub_imdb_id = self.find_imdb_id(sub_id)
6 years ago
4 years ago
subtitle = SuperSubtitlesSubtitle(sub_language, sub_downloadlink, sub_id, sub_english_name.strip(),
sub_season, sub_episode, sub_version, sub_releases, sub_year,
sub_imdb_id, uploader, asked_for_episode,
asked_for_release_group=video.release_group)
if subtitle.language in languages:
subtitles.append(subtitle)
6 years ago
i = i + 1
return subtitles
def list_subtitles(self, video, languages):
4 years ago
titles = []
6 years ago
if isinstance(video, Episode):
titles = [video.series] + video.alternative_series
elif isinstance(video, Movie):
titles = [video.title] + video.alternative_titles
4 years ago
subtitles = []
6 years ago
for title in titles:
4 years ago
subs = self.query(title, languages, video=video)
6 years ago
if subs:
4 years ago
for item in subs:
4 years ago
if isinstance(video, Episode):
fixed_title = fix_tv_naming(item.series)
else:
fixed_title = fix_movie_naming(item.series)
# Check for the original and the fixed titles too
if any(x in (fixed_title.strip(), item.series) for x in titles):
4 years ago
subtitles.append(item)
6 years ago
time.sleep(self.multi_result_throttle)
4 years ago
return subtitles
6 years ago
def download_subtitle(self, subtitle):
# download as a zip
logger.info('Downloading subtitle %r', subtitle.subtitle_id)
r = self.session.get(subtitle.page_link, timeout=10)
r.raise_for_status()
archive_stream = io.BytesIO(r.content)
archive = None
if is_rarfile(archive_stream):
6 years ago
archive = RarFile(archive_stream)
elif is_zipfile(archive_stream):
6 years ago
archive = ZipFile(archive_stream)
else:
subtitle.content = fix_line_ending(r.content)
if archive is not None:
subtitle.content = self.get_subtitle_from_archive(subtitle, archive)