Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/crunchyroll.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import json
   6 import zlib
   7
   8 from hashlib import sha1
   9 from math import pow, sqrt, floor
  10 from .common import InfoExtractor
  11 from .vrv import VRVIE
  12 from ..compat import (
  13     compat_b64decode,
  14     compat_etree_Element,
  15     compat_etree_fromstring,
  16     compat_str,
  17     compat_urllib_parse_urlencode,
  18     compat_urllib_request,
  19     compat_urlparse,
  20 )
  21 from ..utils import (
  22     ExtractorError,
  23     bytes_to_intlist,
  24     extract_attributes,
  25     float_or_none,
  26     intlist_to_bytes,
  27     int_or_none,
  28     lowercase_escape,
  29     merge_dicts,
  30     remove_end,
  31     sanitized_Request,
  32     urlencode_postdata,
  33     xpath_text,
  34 )
  35 from ..aes import (
  36     aes_cbc_decrypt,
  37 )
  38
  39
  40 class CrunchyrollBaseIE(InfoExtractor):
  41     _LOGIN_URL = 'https://www.crunchyroll.com/login'
  42     _LOGIN_FORM = 'login_form'
  43     _NETRC_MACHINE = 'crunchyroll'
  44
  45     def _call_rpc_api(self, method, video_id, note=None, data=None):
  46         data = data or {}
  47         data['req'] = 'RpcApi' + method
  48         data = compat_urllib_parse_urlencode(data).encode('utf-8')
  49         return self._download_xml(
  50             'https://www.crunchyroll.com/xml/',
  51             video_id, note, fatal=False, data=data, headers={
  52                 'Content-Type': 'application/x-www-form-urlencoded',
  53             })
  54
  55     def _login(self):
  56         username, password = self._get_login_info()
  57         if username is None:
  58             return
  59
  60         login_page = self._download_webpage(
  61             self._LOGIN_URL, None, 'Downloading login page')
  62
  63         def is_logged(webpage):
  64             return 'href="/logout"' in webpage
  65
  66         # Already logged in
  67         if is_logged(login_page):
  68             return
  69
  70         login_form_str = self._search_regex(
  71             r'(?P<form><form[^>]+?id=(["\'])%s\2[^>]*>)' % self._LOGIN_FORM,
  72             login_page, 'login form', group='form')
  73
  74         post_url = extract_attributes(login_form_str).get('action')
  75         if not post_url:
  76             post_url = self._LOGIN_URL
  77         elif not post_url.startswith('http'):
  78             post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url)
  79
  80         login_form = self._form_hidden_inputs(self._LOGIN_FORM, login_page)
  81
  82         login_form.update({
  83             'login_form[name]': username,
  84             'login_form[password]': password,
  85         })
  86
  87         response = self._download_webpage(
  88             post_url, None, 'Logging in', 'Wrong login info',
  89             data=urlencode_postdata(login_form),
  90             headers={'Content-Type': 'application/x-www-form-urlencoded'})
  91
  92         # Successful login
  93         if is_logged(response):
  94             return
  95
  96         error = self._html_search_regex(
  97             '(?s)<ul[^>]+class=["\']messages["\'][^>]*>(.+?)</ul>',
  98             response, 'error message', default=None)
  99         if error:
 100             raise ExtractorError('Unable to login: %s' % error, expected=True)
 101
 102         raise ExtractorError('Unable to log in')
 103
 104     def _real_initialize(self):
 105         self._login()
 106
 107     @staticmethod
 108     def _add_skip_wall(url):
 109         parsed_url = compat_urlparse.urlparse(url)
 110         qs = compat_urlparse.parse_qs(parsed_url.query)
 111         # Always force skip_wall to bypass maturity wall, namely 18+ confirmation message:
 112         # > This content may be inappropriate for some people.
 113         # > Are you sure you want to continue?
 114         # since it's not disabled by default in crunchyroll account's settings.
 115         # See https://github.com/ytdl-org/youtube-dl/issues/7202.
 116         qs['skip_wall'] = ['1']
 117         return compat_urlparse.urlunparse(
 118             parsed_url._replace(query=compat_urllib_parse_urlencode(qs, True)))
 119
 120
 121 class CrunchyrollIE(CrunchyrollBaseIE, VRVIE):
 122     IE_NAME = 'crunchyroll'
 123     _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.(?:com|fr)/(?:media(?:-|/\?id=)|(?:[^/]*/){1,2}[^/?&]*?)(?P<video_id>[0-9]+))(?:[/?&]|$)'
 124     _TESTS = [{
 125         'url': 'http://www.crunchyroll.com/wanna-be-the-strongest-in-the-world/episode-1-an-idol-wrestler-is-born-645513',
 126         'info_dict': {
 127             'id': '645513',
 128             'ext': 'mp4',
 129             'title': 'Wanna be the Strongest in the World Episode 1 – An Idol-Wrestler is Born!',
 130             'description': 'md5:2d17137920c64f2f49981a7797d275ef',
 131             'thumbnail': r're:^https?://.*\.jpg$',
 132             'uploader': 'Yomiuri Telecasting Corporation (YTV)',
 133             'upload_date': '20131013',
 134             'url': 're:(?!.*&amp)',
 135         },
 136         'params': {
 137             # rtmp
 138             'skip_download': True,
 139         },
 140         'skip': 'Video gone',
 141     }, {
 142         'url': 'http://www.crunchyroll.com/media-589804/culture-japan-1',
 143         'info_dict': {
 144             'id': '589804',
 145             'ext': 'flv',
 146             'title': 'Culture Japan Episode 1 – Rebuilding Japan after the 3.11',
 147             'description': 'md5:2fbc01f90b87e8e9137296f37b461c12',
 148             'thumbnail': r're:^https?://.*\.jpg$',
 149             'uploader': 'Danny Choo Network',
 150             'upload_date': '20120213',
 151         },
 152         'params': {
 153             # rtmp
 154             'skip_download': True,
 155         },
 156         'skip': 'Video gone',
 157     }, {
 158         'url': 'http://www.crunchyroll.com/rezero-starting-life-in-another-world-/episode-5-the-morning-of-our-promise-is-still-distant-702409',
 159         'info_dict': {
 160             'id': '702409',
 161             'ext': 'mp4',
 162             'title': compat_str,
 163             'description': compat_str,
 164             'thumbnail': r're:^https?://.*\.jpg$',
 165             'uploader': 'Re:Zero Partners',
 166             'timestamp': 1462098900,
 167             'upload_date': '20160501',
 168         },
 169         'params': {
 170             # m3u8 download
 171             'skip_download': True,
 172         },
 173     }, {
 174         'url': 'http://www.crunchyroll.com/konosuba-gods-blessing-on-this-wonderful-world/episode-1-give-me-deliverance-from-this-judicial-injustice-727589',
 175         'info_dict': {
 176             'id': '727589',
 177             'ext': 'mp4',
 178             'title': compat_str,
 179             'description': compat_str,
 180             'thumbnail': r're:^https?://.*\.jpg$',
 181             'uploader': 'Kadokawa Pictures Inc.',
 182             'timestamp': 1484130900,
 183             'upload_date': '20170111',
 184             'series': compat_str,
 185             'season': "KONOSUBA -God's blessing on this wonderful world! 2",
 186             'season_number': 2,
 187             'episode': 'Give Me Deliverance From This Judicial Injustice!',
 188             'episode_number': 1,
 189         },
 190         'params': {
 191             # m3u8 download
 192             'skip_download': True,
 193         },
 194     }, {
 195         'url': 'http://www.crunchyroll.fr/girl-friend-beta/episode-11-goodbye-la-mode-661697',
 196         'only_matching': True,
 197     }, {
 198         # geo-restricted (US), 18+ maturity wall, non-premium available
 199         'url': 'http://www.crunchyroll.com/cosplay-complex-ova/episode-1-the-birth-of-the-cosplay-club-565617',
 200         'only_matching': True,
 201     }, {
 202         # A description with double quotes
 203         'url': 'http://www.crunchyroll.com/11eyes/episode-1-piros-jszaka-red-night-535080',
 204         'info_dict': {
 205             'id': '535080',
 206             'ext': 'mp4',
 207             'title': compat_str,
 208             'description': compat_str,
 209             'uploader': 'Marvelous AQL Inc.',
 210             'timestamp': 1255512600,
 211             'upload_date': '20091014',
 212         },
 213         'params': {
 214             # Just test metadata extraction
 215             'skip_download': True,
 216         },
 217     }, {
 218         # make sure we can extract an uploader name that's not a link
 219         'url': 'http://www.crunchyroll.com/hakuoki-reimeiroku/episode-1-dawn-of-the-divine-warriors-606899',
 220         'info_dict': {
 221             'id': '606899',
 222             'ext': 'mp4',
 223             'title': 'Hakuoki Reimeiroku Episode 1 – Dawn of the Divine Warriors',
 224             'description': 'Ryunosuke was left to die, but Serizawa-san asked him a simple question "Do you want to live?"',
 225             'uploader': 'Geneon Entertainment',
 226             'upload_date': '20120717',
 227         },
 228         'params': {
 229             # just test metadata extraction
 230             'skip_download': True,
 231         },
 232         'skip': 'Video gone',
 233     }, {
 234         # A video with a vastly different season name compared to the series name
 235         'url': 'http://www.crunchyroll.com/nyarko-san-another-crawling-chaos/episode-1-test-590532',
 236         'info_dict': {
 237             'id': '590532',
 238             'ext': 'mp4',
 239             'title': compat_str,
 240             'description': compat_str,
 241             'uploader': 'TV TOKYO',
 242             'timestamp': 1330956000,
 243             'upload_date': '20120305',
 244             'series': 'Nyarko-san: Another Crawling Chaos',
 245             'season': 'Haiyoru! Nyaruani (ONA)',
 246         },
 247         'params': {
 248             # Just test metadata extraction
 249             'skip_download': True,
 250         },
 251     }, {
 252         'url': 'http://www.crunchyroll.com/media-723735',
 253         'only_matching': True,
 254     }, {
 255         'url': 'https://www.crunchyroll.com/en-gb/mob-psycho-100/episode-2-urban-legends-encountering-rumors-780921',
 256         'only_matching': True,
 257     }]
 258
 259     _FORMAT_IDS = {
 260         '360': ('60', '106'),
 261         '480': ('61', '106'),
 262         '720': ('62', '106'),
 263         '1080': ('80', '108'),
 264     }
 265
 266     def _download_webpage(self, url_or_request, *args, **kwargs):
 267         request = (url_or_request if isinstance(url_or_request, compat_urllib_request.Request)
 268                    else sanitized_Request(url_or_request))
 269         # Accept-Language must be set explicitly to accept any language to avoid issues
 270         # similar to https://github.com/ytdl-org/youtube-dl/issues/6797.
 271         # Along with IP address Crunchyroll uses Accept-Language to guess whether georestriction
 272         # should be imposed or not (from what I can see it just takes the first language
 273         # ignoring the priority and requires it to correspond the IP). By the way this causes
 274         # Crunchyroll to not work in georestriction cases in some browsers that don't place
 275         # the locale lang first in header. However allowing any language seems to workaround the issue.
 276         request.add_header('Accept-Language', '*')
 277         return super(CrunchyrollBaseIE, self)._download_webpage(request, *args, **kwargs)
 278
 279     def _decrypt_subtitles(self, data, iv, id):
 280         data = bytes_to_intlist(compat_b64decode(data))
 281         iv = bytes_to_intlist(compat_b64decode(iv))
 282         id = int(id)
 283
 284         def obfuscate_key_aux(count, modulo, start):
 285             output = list(start)
 286             for _ in range(count):
 287                 output.append(output[-1] + output[-2])
 288             # cut off start values
 289             output = output[2:]
 290             output = list(map(lambda x: x % modulo + 33, output))
 291             return output
 292
 293         def obfuscate_key(key):
 294             num1 = int(floor(pow(2, 25) * sqrt(6.9)))
 295             num2 = (num1 ^ key) << 5
 296             num3 = key ^ num1
 297             num4 = num3 ^ (num3 >> 3) ^ num2
 298             prefix = intlist_to_bytes(obfuscate_key_aux(20, 97, (1, 2)))
 299             shaHash = bytes_to_intlist(sha1(prefix + str(num4).encode('ascii')).digest())
 300             # Extend 160 Bit hash to 256 Bit
 301             return shaHash + [0] * 12
 302
 303         key = obfuscate_key(id)
 304
 305         decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv))
 306         return zlib.decompress(decrypted_data)
 307
 308     def _convert_subtitles_to_srt(self, sub_root):
 309         output = ''
 310
 311         for i, event in enumerate(sub_root.findall('./events/event'), 1):
 312             start = event.attrib['start'].replace('.', ',')
 313             end = event.attrib['end'].replace('.', ',')
 314             text = event.attrib['text'].replace('\\N', '\n')
 315             output += '%d\n%s --> %s\n%s\n\n' % (i, start, end, text)
 316         return output
 317
 318     def _convert_subtitles_to_ass(self, sub_root):
 319         output = ''
 320
 321         def ass_bool(strvalue):
 322             assvalue = '0'
 323             if strvalue == '1':
 324                 assvalue = '-1'
 325             return assvalue
 326
 327         output = '[Script Info]\n'
 328         output += 'Title: %s\n' % sub_root.attrib['title']
 329         output += 'ScriptType: v4.00+\n'
 330         output += 'WrapStyle: %s\n' % sub_root.attrib['wrap_style']
 331         output += 'PlayResX: %s\n' % sub_root.attrib['play_res_x']
 332         output += 'PlayResY: %s\n' % sub_root.attrib['play_res_y']
 333         output += """
 334 [V4+ Styles]
 335 Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, OutlineColour, BackColour, Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, Spacing, Angle, BorderStyle, Outline, Shadow, Alignment, MarginL, MarginR, MarginV, Encoding
 336 """
 337         for style in sub_root.findall('./styles/style'):
 338             output += 'Style: ' + style.attrib['name']
 339             output += ',' + style.attrib['font_name']
 340             output += ',' + style.attrib['font_size']
 341             output += ',' + style.attrib['primary_colour']
 342             output += ',' + style.attrib['secondary_colour']
 343             output += ',' + style.attrib['outline_colour']
 344             output += ',' + style.attrib['back_colour']
 345             output += ',' + ass_bool(style.attrib['bold'])
 346             output += ',' + ass_bool(style.attrib['italic'])
 347             output += ',' + ass_bool(style.attrib['underline'])
 348             output += ',' + ass_bool(style.attrib['strikeout'])
 349             output += ',' + style.attrib['scale_x']
 350             output += ',' + style.attrib['scale_y']
 351             output += ',' + style.attrib['spacing']
 352             output += ',' + style.attrib['angle']
 353             output += ',' + style.attrib['border_style']
 354             output += ',' + style.attrib['outline']
 355             output += ',' + style.attrib['shadow']
 356             output += ',' + style.attrib['alignment']
 357             output += ',' + style.attrib['margin_l']
 358             output += ',' + style.attrib['margin_r']
 359             output += ',' + style.attrib['margin_v']
 360             output += ',' + style.attrib['encoding']
 361             output += '\n'
 362
 363         output += """
 364 [Events]
 365 Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
 366 """
 367         for event in sub_root.findall('./events/event'):
 368             output += 'Dialogue: 0'
 369             output += ',' + event.attrib['start']
 370             output += ',' + event.attrib['end']
 371             output += ',' + event.attrib['style']
 372             output += ',' + event.attrib['name']
 373             output += ',' + event.attrib['margin_l']
 374             output += ',' + event.attrib['margin_r']
 375             output += ',' + event.attrib['margin_v']
 376             output += ',' + event.attrib['effect']
 377             output += ',' + event.attrib['text']
 378             output += '\n'
 379
 380         return output
 381
 382     def _extract_subtitles(self, subtitle):
 383         sub_root = compat_etree_fromstring(subtitle)
 384         return [{
 385             'ext': 'srt',
 386             'data': self._convert_subtitles_to_srt(sub_root),
 387         }, {
 388             'ext': 'ass',
 389             'data': self._convert_subtitles_to_ass(sub_root),
 390         }]
 391
 392     def _get_subtitles(self, video_id, webpage):
 393         subtitles = {}
 394         for sub_id, sub_name in re.findall(r'\bssid=([0-9]+)"[^>]+?\btitle="([^"]+)', webpage):
 395             sub_doc = self._call_rpc_api(
 396                 'Subtitle_GetXml', video_id,
 397                 'Downloading subtitles for ' + sub_name, data={
 398                     'subtitle_script_id': sub_id,
 399                 })
 400             if not isinstance(sub_doc, compat_etree_Element):
 401                 continue
 402             sid = sub_doc.get('id')
 403             iv = xpath_text(sub_doc, 'iv', 'subtitle iv')
 404             data = xpath_text(sub_doc, 'data', 'subtitle data')
 405             if not sid or not iv or not data:
 406                 continue
 407             subtitle = self._decrypt_subtitles(data, iv, sid).decode('utf-8')
 408             lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
 409             if not lang_code:
 410                 continue
 411             subtitles[lang_code] = self._extract_subtitles(subtitle)
 412         return subtitles
 413
 414     def _real_extract(self, url):
 415         mobj = re.match(self._VALID_URL, url)
 416         video_id = mobj.group('video_id')
 417
 418         if mobj.group('prefix') == 'm':
 419             mobile_webpage = self._download_webpage(url, video_id, 'Downloading mobile webpage')
 420             webpage_url = self._search_regex(r'<link rel="canonical" href="([^"]+)" />', mobile_webpage, 'webpage_url')
 421         else:
 422             webpage_url = 'http://www.' + mobj.group('url')
 423
 424         webpage = self._download_webpage(
 425             self._add_skip_wall(webpage_url), video_id,
 426             headers=self.geo_verification_headers())
 427         note_m = self._html_search_regex(
 428             r'<div class="showmedia-trailer-notice">(.+?)</div>',
 429             webpage, 'trailer-notice', default='')
 430         if note_m:
 431             raise ExtractorError(note_m)
 432
 433         mobj = re.search(r'Page\.messaging_box_controller\.addItems\(\[(?P<msg>{.+?})\]\)', webpage)
 434         if mobj:
 435             msg = json.loads(mobj.group('msg'))
 436             if msg.get('type') == 'error':
 437                 raise ExtractorError('crunchyroll returned error: %s' % msg['message_body'], expected=True)
 438
 439         if 'To view this, please log in to verify you are 18 or older.' in webpage:
 440             self.raise_login_required()
 441
 442         media = self._parse_json(self._search_regex(
 443             r'vilos\.config\.media\s*=\s*({.+?});',
 444             webpage, 'vilos media', default='{}'), video_id)
 445         media_metadata = media.get('metadata') or {}
 446
 447         language = self._search_regex(
 448             r'(?:vilos\.config\.player\.language|LOCALE)\s*=\s*(["\'])(?P<lang>(?:(?!\1).)+)\1',
 449             webpage, 'language', default=None, group='lang')
 450
 451         video_title = self._html_search_regex(
 452             (r'(?s)<h1[^>]*>((?:(?!<h1).)*?<(?:span[^>]+itemprop=["\']title["\']|meta[^>]+itemprop=["\']position["\'])[^>]*>(?:(?!<h1).)+?)</h1>',
 453              r'<title>(.+?),\s+-\s+.+? Crunchyroll'),
 454             webpage, 'video_title', default=None)
 455         if not video_title:
 456             video_title = re.sub(r'^Watch\s+', '', self._og_search_description(webpage))
 457         video_title = re.sub(r' {2,}', ' ', video_title)
 458         video_description = (self._parse_json(self._html_search_regex(
 459             r'<script[^>]*>\s*.+?\[media_id=%s\].+?({.+?"description"\s*:.+?})\);' % video_id,
 460             webpage, 'description', default='{}'), video_id) or media_metadata).get('description')
 461         if video_description:
 462             video_description = lowercase_escape(video_description.replace(r'\r\n', '\n'))
 463         video_uploader = self._html_search_regex(
 464             # try looking for both an uploader that's a link and one that's not
 465             [r'<a[^>]+href="/publisher/[^"]+"[^>]*>([^<]+)</a>', r'<div>\s*Publisher:\s*<span>\s*(.+?)\s*</span>\s*</div>'],
 466             webpage, 'video_uploader', default=False)
 467
 468         formats = []
 469         for stream in media.get('streams', []):
 470             audio_lang = stream.get('audio_lang')
 471             hardsub_lang = stream.get('hardsub_lang')
 472             vrv_formats = self._extract_vrv_formats(
 473                 stream.get('url'), video_id, stream.get('format'),
 474                 audio_lang, hardsub_lang)
 475             for f in vrv_formats:
 476                 if not hardsub_lang:
 477                     f['preference'] = 1
 478                 language_preference = 0
 479                 if audio_lang == language:
 480                     language_preference += 1
 481                 if hardsub_lang == language:
 482                     language_preference += 1
 483                 if language_preference:
 484                     f['language_preference'] = language_preference
 485             formats.extend(vrv_formats)
 486         if not formats:
 487             available_fmts = []
 488             for a, fmt in re.findall(r'(<a[^>]+token=["\']showmedia\.([0-9]{3,4})p["\'][^>]+>)', webpage):
 489                 attrs = extract_attributes(a)
 490                 href = attrs.get('href')
 491                 if href and '/freetrial' in href:
 492                     continue
 493                 available_fmts.append(fmt)
 494             if not available_fmts:
 495                 for p in (r'token=["\']showmedia\.([0-9]{3,4})p"', r'showmedia\.([0-9]{3,4})p'):
 496                     available_fmts = re.findall(p, webpage)
 497                     if available_fmts:
 498                         break
 499             if not available_fmts:
 500                 available_fmts = self._FORMAT_IDS.keys()
 501             video_encode_ids = []
 502
 503             for fmt in available_fmts:
 504                 stream_quality, stream_format = self._FORMAT_IDS[fmt]
 505                 video_format = fmt + 'p'
 506                 stream_infos = []
 507                 streamdata = self._call_rpc_api(
 508                     'VideoPlayer_GetStandardConfig', video_id,
 509                     'Downloading media info for %s' % video_format, data={
 510                         'media_id': video_id,
 511                         'video_format': stream_format,
 512                         'video_quality': stream_quality,
 513                         'current_page': url,
 514                     })
 515                 if isinstance(streamdata, compat_etree_Element):
 516                     stream_info = streamdata.find('./{default}preload/stream_info')
 517                     if stream_info is not None:
 518                         stream_infos.append(stream_info)
 519                 stream_info = self._call_rpc_api(
 520                     'VideoEncode_GetStreamInfo', video_id,
 521                     'Downloading stream info for %s' % video_format, data={
 522                         'media_id': video_id,
 523                         'video_format': stream_format,
 524                         'video_encode_quality': stream_quality,
 525                     })
 526                 if isinstance(stream_info, compat_etree_Element):
 527                     stream_infos.append(stream_info)
 528                 for stream_info in stream_infos:
 529                     video_encode_id = xpath_text(stream_info, './video_encode_id')
 530                     if video_encode_id in video_encode_ids:
 531                         continue
 532                     video_encode_ids.append(video_encode_id)
 533
 534                     video_file = xpath_text(stream_info, './file')
 535                     if not video_file:
 536                         continue
 537                     if video_file.startswith('http'):
 538                         formats.extend(self._extract_m3u8_formats(
 539                             video_file, video_id, 'mp4', entry_protocol='m3u8_native',
 540                             m3u8_id='hls', fatal=False))
 541                         continue
 542
 543                     video_url = xpath_text(stream_info, './host')
 544                     if not video_url:
 545                         continue
 546                     metadata = stream_info.find('./metadata')
 547                     format_info = {
 548                         'format': video_format,
 549                         'height': int_or_none(xpath_text(metadata, './height')),
 550                         'width': int_or_none(xpath_text(metadata, './width')),
 551                     }
 552
 553                     if '.fplive.net/' in video_url:
 554                         video_url = re.sub(r'^rtmpe?://', 'http://', video_url.strip())
 555                         parsed_video_url = compat_urlparse.urlparse(video_url)
 556                         direct_video_url = compat_urlparse.urlunparse(parsed_video_url._replace(
 557                             netloc='v.lvlt.crcdn.net',
 558                             path='%s/%s' % (remove_end(parsed_video_url.path, '/'), video_file.split(':')[-1])))
 559                         if self._is_valid_url(direct_video_url, video_id, video_format):
 560                             format_info.update({
 561                                 'format_id': 'http-' + video_format,
 562                                 'url': direct_video_url,
 563                             })
 564                             formats.append(format_info)
 565                             continue
 566
 567                     format_info.update({
 568                         'format_id': 'rtmp-' + video_format,
 569                         'url': video_url,
 570                         'play_path': video_file,
 571                         'ext': 'flv',
 572                     })
 573                     formats.append(format_info)
 574         self._sort_formats(formats, ('preference', 'language_preference', 'height', 'width', 'tbr', 'fps'))
 575
 576         metadata = self._call_rpc_api(
 577             'VideoPlayer_GetMediaMetadata', video_id,
 578             note='Downloading media info', data={
 579                 'media_id': video_id,
 580             })
 581
 582         subtitles = {}
 583         for subtitle in media.get('subtitles', []):
 584             subtitle_url = subtitle.get('url')
 585             if not subtitle_url:
 586                 continue
 587             subtitles.setdefault(subtitle.get('language', 'enUS'), []).append({
 588                 'url': subtitle_url,
 589                 'ext': subtitle.get('format', 'ass'),
 590             })
 591         if not subtitles:
 592             subtitles = self.extract_subtitles(video_id, webpage)
 593
 594         # webpage provide more accurate data than series_title from XML
 595         series = self._html_search_regex(
 596             r'(?s)<h\d[^>]+\bid=["\']showmedia_about_episode_num[^>]+>(.+?)</h\d',
 597             webpage, 'series', fatal=False)
 598
 599         season = episode = episode_number = duration = thumbnail = None
 600
 601         if isinstance(metadata, compat_etree_Element):
 602             season = xpath_text(metadata, 'series_title')
 603             episode = xpath_text(metadata, 'episode_title')
 604             episode_number = int_or_none(xpath_text(metadata, 'episode_number'))
 605             duration = float_or_none(media_metadata.get('duration'), 1000)
 606             thumbnail = xpath_text(metadata, 'episode_image_url')
 607
 608         if not episode:
 609             episode = media_metadata.get('title')
 610         if not episode_number:
 611             episode_number = int_or_none(media_metadata.get('episode_number'))
 612         if not thumbnail:
 613             thumbnail = media_metadata.get('thumbnail', {}).get('url')
 614
 615         season_number = int_or_none(self._search_regex(
 616             r'(?s)<h\d[^>]+id=["\']showmedia_about_episode_num[^>]+>.+?</h\d>\s*<h4>\s*Season (\d+)',
 617             webpage, 'season number', default=None))
 618
 619         info = self._search_json_ld(webpage, video_id, default={})
 620
 621         return merge_dicts({
 622             'id': video_id,
 623             'title': video_title,
 624             'description': video_description,
 625             'duration': duration,
 626             'thumbnail': thumbnail,
 627             'uploader': video_uploader,
 628             'series': series,
 629             'season': season,
 630             'season_number': season_number,
 631             'episode': episode,
 632             'episode_number': episode_number,
 633             'subtitles': subtitles,
 634             'formats': formats,
 635         }, info)
 636
 637
 638 class CrunchyrollShowPlaylistIE(CrunchyrollBaseIE):
 639     IE_NAME = 'crunchyroll:playlist'
 640     _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.com/(?!(?:news|anime-news|library|forum|launchcalendar|lineup|store|comics|freetrial|login|media-\d+))(?P<id>[\w\-]+))/?(?:\?|$)'
 641
 642     _TESTS = [{
 643         'url': 'http://www.crunchyroll.com/a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 644         'info_dict': {
 645             'id': 'a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 646             'title': 'A Bridge to the Starry Skies - Hoshizora e Kakaru Hashi'
 647         },
 648         'playlist_count': 13,
 649     }, {
 650         # geo-restricted (US), 18+ maturity wall, non-premium available
 651         'url': 'http://www.crunchyroll.com/cosplay-complex-ova',
 652         'info_dict': {
 653             'id': 'cosplay-complex-ova',
 654             'title': 'Cosplay Complex OVA'
 655         },
 656         'playlist_count': 3,
 657         'skip': 'Georestricted',
 658     }, {
 659         # geo-restricted (US), 18+ maturity wall, non-premium will be available since 2015.11.14
 660         'url': 'http://www.crunchyroll.com/ladies-versus-butlers?skip_wall=1',
 661         'only_matching': True,
 662     }]
 663
 664     def _real_extract(self, url):
 665         show_id = self._match_id(url)
 666
 667         webpage = self._download_webpage(
 668             self._add_skip_wall(url), show_id,
 669             headers=self.geo_verification_headers())
 670         title = self._html_search_meta('name', webpage, default=None)
 671
 672         episode_paths = re.findall(
 673             r'(?s)<li id="showview_videos_media_(\d+)"[^>]+>.*?<a href="([^"]+)"',
 674             webpage)
 675         entries = [
 676             self.url_result('http://www.crunchyroll.com' + ep, 'Crunchyroll', ep_id)
 677             for ep_id, ep in episode_paths
 678         ]
 679         entries.reverse()
 680
 681         return {
 682             '_type': 'playlist',
 683             'id': show_id,
 684             'title': title,
 685             'entries': entries,
 686         }