yt_dlp/extractor/bilibili.py

   1 import base64
   2 import hashlib
   3 import itertools
   4 import functools
   5 import re
   6 import math
   7
   8 from .common import InfoExtractor, SearchInfoExtractor
   9 from ..compat import (
  10     compat_parse_qs,
  11     compat_urlparse,
  12     compat_urllib_parse_urlparse
  13 )
  14 from ..utils import (
  15     ExtractorError,
  16     filter_dict,
  17     int_or_none,
  18     float_or_none,
  19     mimetype2ext,
  20     parse_iso8601,
  21     qualities,
  22     traverse_obj,
  23     parse_count,
  24     smuggle_url,
  25     srt_subtitles_timecode,
  26     str_or_none,
  27     strip_jsonp,
  28     unified_timestamp,
  29     unsmuggle_url,
  30     urlencode_postdata,
  31     url_or_none,
  32     OnDemandPagedList
  33 )
  34
  35
  36 class BiliBiliIE(InfoExtractor):
  37     _VALID_URL = r'''(?x)
  38                     https?://
  39                         (?:(?:www|bangumi)\.)?
  40                         bilibili\.(?:tv|com)/
  41                         (?:
  42                             (?:
  43                                 video/[aA][vV]|
  44                                 anime/(?P<anime_id>\d+)/play\#
  45                             )(?P<id>\d+)|
  46                             (s/)?video/[bB][vV](?P<id_bv>[^/?#&]+)
  47                         )
  48                         (?:/?\?p=(?P<page>\d+))?
  49                     '''
  50
  51     _TESTS = [{
  52         'url': 'http://www.bilibili.com/video/av1074402/',
  53         'md5': '7ac275ec84a99a6552c5d229659a0fe1',
  54         'info_dict': {
  55             'id': '1074402_part1',
  56             'ext': 'mp4',
  57             'title': '【金坷垃】金泡沫',
  58             'uploader_id': '156160',
  59             'uploader': '菊子桑',
  60             'upload_date': '20140420',
  61             'description': 'md5:ce18c2a2d2193f0df2917d270f2e5923',
  62             'timestamp': 1398012678,
  63             'tags': ['顶上去报复社会', '该来的总会来的', '金克拉是检验歌曲的唯一标准', '坷垃教主', '金坷垃', '邓紫棋', '治愈系坷垃'],
  64             'bv_id': 'BV11x411K7CN',
  65             'cid': '1554319',
  66             'thumbnail': 'http://i2.hdslb.com/bfs/archive/c79a8cf0347cd7a897c53a2f756e96aead128e8c.jpg',
  67             'duration': 308.36,
  68         },
  69     }, {
  70         # Tested in BiliBiliBangumiIE
  71         'url': 'http://bangumi.bilibili.com/anime/1869/play#40062',
  72         'only_matching': True,
  73     }, {
  74         # bilibili.tv
  75         'url': 'http://www.bilibili.tv/video/av1074402/',
  76         'only_matching': True,
  77     }, {
  78         'url': 'http://bangumi.bilibili.com/anime/5802/play#100643',
  79         'md5': '3f721ad1e75030cc06faf73587cfec57',
  80         'info_dict': {
  81             'id': '100643_part1',
  82             'ext': 'mp4',
  83             'title': 'CHAOS;CHILD',
  84             'description': '如果你是神明，并且能够让妄想成为现实。那你会进行怎么样的妄想？是淫靡的世界？独裁社会？毁灭性的制裁？还是……2015年，涩谷。从6年前发生的大灾害“涩谷地震”之后复兴了的这个街区里新设立的私立高中...',
  85         },
  86         'skip': 'Geo-restricted to China',
  87     }, {
  88         'url': 'http://www.bilibili.com/video/av8903802/',
  89         'info_dict': {
  90             'id': '8903802_part1',
  91             'ext': 'mp4',
  92             'title': '阿滴英文｜英文歌分享#6 "Closer',
  93             'upload_date': '20170301',
  94             'description': '滴妹今天唱Closer給你聽! 有史以来，被推最多次也是最久的歌曲，其实歌词跟我原本想像差蛮多的，不过还是好听！ 微博@阿滴英文',
  95             'timestamp': 1488382634,
  96             'uploader_id': '65880958',
  97             'uploader': '阿滴英文',
  98             'thumbnail': 'http://i2.hdslb.com/bfs/archive/49267ce20bc246be6304bf369a3ded0256854c23.jpg',
  99             'cid': '14694589',
 100             'duration': 554.117,
 101             'bv_id': 'BV13x41117TL',
 102             'tags': ['人文', '英语', '文化', '公开课', '阿滴英文'],
 103         },
 104         'params': {
 105             'skip_download': True,
 106         },
 107     }, {
 108         # new BV video id format
 109         'url': 'https://www.bilibili.com/video/BV1JE411F741',
 110         'only_matching': True,
 111     }, {
 112         # Anthology
 113         'url': 'https://www.bilibili.com/video/BV1bK411W797',
 114         'info_dict': {
 115             'id': 'BV1bK411W797',
 116             'title': '物语中的人物是如何吐槽自己的OP的'
 117         },
 118         'playlist_count': 17,
 119     }, {
 120         # Correct matching of single and double quotes in title
 121         'url': 'https://www.bilibili.com/video/BV1NY411E7Rx/',
 122         'info_dict': {
 123             'id': '255513412_part1',
 124             'ext': 'mp4',
 125             'title': 'Vid"eo" Te\'st',
 126             'cid': '570602418',
 127             'thumbnail': 'http://i2.hdslb.com/bfs/archive/0c0de5a90b6d5b991b8dcc6cde0afbf71d564791.jpg',
 128             'upload_date': '20220408',
 129             'timestamp': 1649436552,
 130             'description': 'Vid"eo" Te\'st',
 131             'uploader_id': '1630758804',
 132             'bv_id': 'BV1NY411E7Rx',
 133             'duration': 60.394,
 134             'uploader': 'bili_31244483705',
 135             'tags': ['VLOG'],
 136         },
 137         'params': {
 138             'skip_download': True,
 139         },
 140     }]
 141
 142     _APP_KEY = 'iVGUTjsxvpLeuDCf'
 143     _BILIBILI_KEY = 'aHRmhWMLkdeMuILqORnYZocwMBpMEOdt'
 144
 145     def _report_error(self, result):
 146         if 'message' in result:
 147             raise ExtractorError('%s said: %s' % (self.IE_NAME, result['message']), expected=True)
 148         elif 'code' in result:
 149             raise ExtractorError('%s returns error %d' % (self.IE_NAME, result['code']), expected=True)
 150         else:
 151             raise ExtractorError('Can\'t extract Bangumi episode ID')
 152
 153     def _real_extract(self, url):
 154         url, smuggled_data = unsmuggle_url(url, {})
 155
 156         mobj = self._match_valid_url(url)
 157         video_id = mobj.group('id_bv') or mobj.group('id')
 158
 159         av_id, bv_id = self._get_video_id_set(video_id, mobj.group('id_bv') is not None)
 160         video_id = av_id
 161
 162         info = {}
 163         anime_id = mobj.group('anime_id')
 164         page_id = mobj.group('page')
 165         webpage = self._download_webpage(url, video_id)
 166
 167         # Bilibili anthologies are similar to playlists but all videos share the same video ID as the anthology itself.
 168         # If the video has no page argument, check to see if it's an anthology
 169         if page_id is None:
 170             if not self.get_param('noplaylist'):
 171                 r = self._extract_anthology_entries(bv_id, video_id, webpage)
 172                 if r is not None:
 173                     self.to_screen('Downloading anthology %s - add --no-playlist to just download video' % video_id)
 174                     return r
 175             else:
 176                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
 177
 178         if 'anime/' not in url:
 179             cid = self._search_regex(
 180                 r'\bcid(?:["\']:|=)(\d+),["\']page(?:["\']:|=)' + str(page_id), webpage, 'cid',
 181                 default=None
 182             ) or self._search_regex(
 183                 r'\bcid(?:["\']:|=)(\d+)', webpage, 'cid',
 184                 default=None
 185             ) or compat_parse_qs(self._search_regex(
 186                 [r'EmbedPlayer\([^)]+,\s*"([^"]+)"\)',
 187                  r'EmbedPlayer\([^)]+,\s*\\"([^"]+)\\"\)',
 188                  r'<iframe[^>]+src="https://secure\.bilibili\.com/secure,([^"]+)"'],
 189                 webpage, 'player parameters'))['cid'][0]
 190         else:
 191             if 'no_bangumi_tip' not in smuggled_data:
 192                 self.to_screen('Downloading episode %s. To download all videos in anime %s, re-run yt-dlp with %s' % (
 193                     video_id, anime_id, compat_urlparse.urljoin(url, '//bangumi.bilibili.com/anime/%s' % anime_id)))
 194             headers = {
 195                 'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8',
 196                 'Referer': url
 197             }
 198             headers.update(self.geo_verification_headers())
 199
 200             js = self._download_json(
 201                 'http://bangumi.bilibili.com/web_api/get_source', video_id,
 202                 data=urlencode_postdata({'episode_id': video_id}),
 203                 headers=headers)
 204             if 'result' not in js:
 205                 self._report_error(js)
 206             cid = js['result']['cid']
 207
 208         headers = {
 209             'Accept': 'application/json',
 210             'Referer': url
 211         }
 212         headers.update(self.geo_verification_headers())
 213
 214         video_info = self._parse_json(
 215             self._search_regex(r'window.__playinfo__\s*=\s*({.+?})</script>', webpage, 'video info', default=None) or '{}',
 216             video_id, fatal=False)
 217         video_info = video_info.get('data') or {}
 218
 219         durl = traverse_obj(video_info, ('dash', 'video'))
 220         audios = traverse_obj(video_info, ('dash', 'audio')) or []
 221         entries = []
 222
 223         RENDITIONS = ('qn=80&quality=80&type=', 'quality=2&type=mp4')
 224         for num, rendition in enumerate(RENDITIONS, start=1):
 225             payload = 'appkey=%s&cid=%s&otype=json&%s' % (self._APP_KEY, cid, rendition)
 226             sign = hashlib.md5((payload + self._BILIBILI_KEY).encode('utf-8')).hexdigest()
 227             if not video_info:
 228                 video_info = self._download_json(
 229                     'http://interface.bilibili.com/v2/playurl?%s&sign=%s' % (payload, sign),
 230                     video_id, note='Downloading video info page',
 231                     headers=headers, fatal=num == len(RENDITIONS))
 232                 if not video_info:
 233                     continue
 234
 235             if not durl and 'durl' not in video_info:
 236                 if num < len(RENDITIONS):
 237                     continue
 238                 self._report_error(video_info)
 239
 240             formats = []
 241             for idx, durl in enumerate(durl or video_info['durl']):
 242                 formats.append({
 243                     'url': durl.get('baseUrl') or durl.get('base_url') or durl.get('url'),
 244                     'ext': mimetype2ext(durl.get('mimeType') or durl.get('mime_type')),
 245                     'fps': int_or_none(durl.get('frameRate') or durl.get('frame_rate')),
 246                     'width': int_or_none(durl.get('width')),
 247                     'height': int_or_none(durl.get('height')),
 248                     'vcodec': durl.get('codecs'),
 249                     'acodec': 'none' if audios else None,
 250                     'tbr': float_or_none(durl.get('bandwidth'), scale=1000),
 251                     'filesize': int_or_none(durl.get('size')),
 252                 })
 253                 for backup_url in traverse_obj(durl, 'backup_url', expected_type=list) or []:
 254                     formats.append({
 255                         'url': backup_url,
 256                         'quality': -2 if 'hd.mp4' in backup_url else -3,
 257                     })
 258
 259             for audio in audios:
 260                 formats.append({
 261                     'url': audio.get('baseUrl') or audio.get('base_url') or audio.get('url'),
 262                     'ext': mimetype2ext(audio.get('mimeType') or audio.get('mime_type')),
 263                     'fps': int_or_none(audio.get('frameRate') or audio.get('frame_rate')),
 264                     'width': int_or_none(audio.get('width')),
 265                     'height': int_or_none(audio.get('height')),
 266                     'acodec': audio.get('codecs'),
 267                     'vcodec': 'none',
 268                     'tbr': float_or_none(audio.get('bandwidth'), scale=1000),
 269                     'filesize': int_or_none(audio.get('size'))
 270                 })
 271                 for backup_url in traverse_obj(audio, 'backup_url', expected_type=list) or []:
 272                     formats.append({
 273                         'url': backup_url,
 274                         # backup URLs have lower priorities
 275                         'quality': -3,
 276                     })
 277
 278             info.update({
 279                 'id': video_id,
 280                 'duration': float_or_none(durl.get('length'), 1000),
 281                 'formats': formats,
 282                 'http_headers': {
 283                     'Referer': url,
 284                 },
 285             })
 286             break
 287
 288         self._sort_formats(formats)
 289
 290         title = self._html_search_regex((
 291             r'<h1[^>]+title=(["])(?P<content>[^"]+)',
 292             r'<h1[^>]+title=([\'])(?P<content>[^\']+)',
 293             r'(?s)<h1[^>]*>(?P<content>.+?)</h1>',
 294             self._meta_regex('title')
 295         ), webpage, 'title', group='content', fatal=False)
 296
 297         # Get part title for anthologies
 298         if page_id is not None:
 299             # TODO: The json is already downloaded by _extract_anthology_entries. Don't redownload for each video.
 300             part_info = traverse_obj(self._download_json(
 301                 f'https://api.bilibili.com/x/player/pagelist?bvid={bv_id}&jsonp=jsonp',
 302                 video_id, note='Extracting videos in anthology'), 'data', expected_type=list)
 303             title = title if len(part_info) == 1 else traverse_obj(part_info, (int(page_id) - 1, 'part')) or title
 304
 305         description = self._html_search_meta('description', webpage)
 306         timestamp = unified_timestamp(self._html_search_regex(
 307             r'<time[^>]+datetime="([^"]+)"', webpage, 'upload time',
 308             default=None) or self._html_search_meta(
 309             'uploadDate', webpage, 'timestamp', default=None))
 310         thumbnail = self._html_search_meta(['og:image', 'thumbnailUrl'], webpage)
 311
 312         # TODO 'view_count' requires deobfuscating Javascript
 313         info.update({
 314             'id': f'{video_id}_part{page_id or 1}',
 315             'cid': cid,
 316             'title': title,
 317             'description': description,
 318             'timestamp': timestamp,
 319             'thumbnail': thumbnail,
 320             'duration': float_or_none(video_info.get('timelength'), scale=1000),
 321         })
 322
 323         uploader_mobj = re.search(
 324             r'<a[^>]+href="(?:https?:)?//space\.bilibili\.com/(?P<id>\d+)"[^>]*>\s*(?P<name>[^<]+?)\s*<',
 325             webpage)
 326         if uploader_mobj:
 327             info.update({
 328                 'uploader': uploader_mobj.group('name').strip(),
 329                 'uploader_id': uploader_mobj.group('id'),
 330             })
 331
 332         if not info.get('uploader'):
 333             info['uploader'] = self._html_search_meta(
 334                 'author', webpage, 'uploader', default=None)
 335
 336         top_level_info = {
 337             'tags': traverse_obj(self._download_json(
 338                 f'https://api.bilibili.com/x/tag/archive/tags?aid={video_id}',
 339                 video_id, fatal=False, note='Downloading tags'), ('data', ..., 'tag_name')),
 340         }
 341
 342         info['subtitles'] = {
 343             'danmaku': [{
 344                 'ext': 'xml',
 345                 'url': f'https://comment.bilibili.com/{cid}.xml',
 346             }]
 347         }
 348
 349         r'''
 350         # Requires https://github.com/m13253/danmaku2ass which is licenced under GPL3
 351         # See https://github.com/animelover1984/youtube-dl
 352
 353         raw_danmaku = self._download_webpage(
 354             f'https://comment.bilibili.com/{cid}.xml', video_id, fatal=False, note='Downloading danmaku comments')
 355         danmaku = NiconicoIE.CreateDanmaku(raw_danmaku, commentType='Bilibili', x=1024, y=576)
 356         entries[0]['subtitles'] = {
 357             'danmaku': [{
 358                 'ext': 'ass',
 359                 'data': danmaku
 360             }]
 361         }
 362         '''
 363
 364         top_level_info['__post_extractor'] = self.extract_comments(video_id)
 365
 366         for entry in entries:
 367             entry.update(info)
 368
 369         if len(entries) == 1:
 370             entries[0].update(top_level_info)
 371             return entries[0]
 372
 373         for idx, entry in enumerate(entries):
 374             entry['id'] = '%s_part%d' % (video_id, (idx + 1))
 375
 376         return {
 377             'id': str(video_id),
 378             'bv_id': bv_id,
 379             'title': title,
 380             'description': description,
 381             **info, **top_level_info
 382         }
 383
 384     def _extract_anthology_entries(self, bv_id, video_id, webpage):
 385         title = self._html_search_regex(
 386             (r'<h1[^>]+\btitle=(["\'])(?P<title>(?:(?!\1).)+)\1',
 387              r'(?s)<h1[^>]*>(?P<title>.+?)</h1>',
 388              r'<title>(?P<title>.+?)</title>'), webpage, 'title',
 389             group='title')
 390         json_data = self._download_json(
 391             f'https://api.bilibili.com/x/player/pagelist?bvid={bv_id}&jsonp=jsonp',
 392             video_id, note='Extracting videos in anthology')
 393
 394         if json_data['data']:
 395             return self.playlist_from_matches(
 396                 json_data['data'], bv_id, title, ie=BiliBiliIE.ie_key(),
 397                 getter=lambda entry: 'https://www.bilibili.com/video/%s?p=%d' % (bv_id, entry['page']))
 398
 399     def _get_video_id_set(self, id, is_bv):
 400         query = {'bvid': id} if is_bv else {'aid': id}
 401         response = self._download_json(
 402             "http://api.bilibili.cn/x/web-interface/view",
 403             id, query=query,
 404             note='Grabbing original ID via API')
 405
 406         if response['code'] == -400:
 407             raise ExtractorError('Video ID does not exist', expected=True, video_id=id)
 408         elif response['code'] != 0:
 409             raise ExtractorError(f'Unknown error occurred during API check (code {response["code"]})',
 410                                  expected=True, video_id=id)
 411         return response['data']['aid'], response['data']['bvid']
 412
 413     def _get_comments(self, video_id, commentPageNumber=0):
 414         for idx in itertools.count(1):
 415             replies = traverse_obj(
 416                 self._download_json(
 417                     f'https://api.bilibili.com/x/v2/reply?pn={idx}&oid={video_id}&type=1&jsonp=jsonp&sort=2&_=1567227301685',
 418                     video_id, note=f'Extracting comments from page {idx}', fatal=False),
 419                 ('data', 'replies'))
 420             if not replies:
 421                 return
 422             for children in map(self._get_all_children, replies):
 423                 yield from children
 424
 425     def _get_all_children(self, reply):
 426         yield {
 427             'author': traverse_obj(reply, ('member', 'uname')),
 428             'author_id': traverse_obj(reply, ('member', 'mid')),
 429             'id': reply.get('rpid'),
 430             'text': traverse_obj(reply, ('content', 'message')),
 431             'timestamp': reply.get('ctime'),
 432             'parent': reply.get('parent') or 'root',
 433         }
 434         for children in map(self._get_all_children, reply.get('replies') or []):
 435             yield from children
 436
 437
 438 class BiliBiliBangumiIE(InfoExtractor):
 439     _VALID_URL = r'https?://bangumi\.bilibili\.com/anime/(?P<id>\d+)'
 440
 441     IE_NAME = 'bangumi.bilibili.com'
 442     IE_DESC = 'BiliBili番剧'
 443
 444     _TESTS = [{
 445         'url': 'http://bangumi.bilibili.com/anime/1869',
 446         'info_dict': {
 447             'id': '1869',
 448             'title': '混沌武士',
 449             'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
 450         },
 451         'playlist_count': 26,
 452     }, {
 453         'url': 'http://bangumi.bilibili.com/anime/1869',
 454         'info_dict': {
 455             'id': '1869',
 456             'title': '混沌武士',
 457             'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
 458         },
 459         'playlist': [{
 460             'md5': '91da8621454dd58316851c27c68b0c13',
 461             'info_dict': {
 462                 'id': '40062',
 463                 'ext': 'mp4',
 464                 'title': '混沌武士',
 465                 'description': '故事发生在日本的江户时代。风是一个小酒馆的打工女。一日，酒馆里来了一群恶霸，虽然他们的举动令风十分不满，但是毕竟风只是一届女流，无法对他们采取什么行动，只能在心里嘟哝。这时，酒家里又进来了个“不良份子...',
 466                 'timestamp': 1414538739,
 467                 'upload_date': '20141028',
 468                 'episode': '疾风怒涛 Tempestuous Temperaments',
 469                 'episode_number': 1,
 470             },
 471         }],
 472         'params': {
 473             'playlist_items': '1',
 474         },
 475     }]
 476
 477     @classmethod
 478     def suitable(cls, url):
 479         return False if BiliBiliIE.suitable(url) else super(BiliBiliBangumiIE, cls).suitable(url)
 480
 481     def _real_extract(self, url):
 482         bangumi_id = self._match_id(url)
 483
 484         # Sometimes this API returns a JSONP response
 485         season_info = self._download_json(
 486             'http://bangumi.bilibili.com/jsonp/seasoninfo/%s.ver' % bangumi_id,
 487             bangumi_id, transform_source=strip_jsonp)['result']
 488
 489         entries = [{
 490             '_type': 'url_transparent',
 491             'url': smuggle_url(episode['webplay_url'], {'no_bangumi_tip': 1}),
 492             'ie_key': BiliBiliIE.ie_key(),
 493             'timestamp': parse_iso8601(episode.get('update_time'), delimiter=' '),
 494             'episode': episode.get('index_title'),
 495             'episode_number': int_or_none(episode.get('index')),
 496         } for episode in season_info['episodes']]
 497
 498         entries = sorted(entries, key=lambda entry: entry.get('episode_number'))
 499
 500         return self.playlist_result(
 501             entries, bangumi_id,
 502             season_info.get('bangumi_title'), season_info.get('evaluate'))
 503
 504
 505 class BilibiliChannelIE(InfoExtractor):
 506     _VALID_URL = r'https?://space.bilibili\.com/(?P<id>\d+)'
 507     _API_URL = "https://api.bilibili.com/x/space/arc/search?mid=%s&pn=%d&jsonp=jsonp"
 508     _TESTS = [{
 509         'url': 'https://space.bilibili.com/3985676/video',
 510         'info_dict': {},
 511         'playlist_mincount': 112,
 512     }]
 513
 514     def _entries(self, list_id):
 515         count, max_count = 0, None
 516
 517         for page_num in itertools.count(1):
 518             data = self._download_json(
 519                 self._API_URL % (list_id, page_num), list_id, note=f'Downloading page {page_num}')['data']
 520
 521             max_count = max_count or traverse_obj(data, ('page', 'count'))
 522
 523             entries = traverse_obj(data, ('list', 'vlist'))
 524             if not entries:
 525                 return
 526             for entry in entries:
 527                 yield self.url_result(
 528                     'https://www.bilibili.com/video/%s' % entry['bvid'],
 529                     BiliBiliIE.ie_key(), entry['bvid'])
 530
 531             count += len(entries)
 532             if max_count and count >= max_count:
 533                 return
 534
 535     def _real_extract(self, url):
 536         list_id = self._match_id(url)
 537         return self.playlist_result(self._entries(list_id), list_id)
 538
 539
 540 class BilibiliCategoryIE(InfoExtractor):
 541     IE_NAME = 'Bilibili category extractor'
 542     _MAX_RESULTS = 1000000
 543     _VALID_URL = r'https?://www\.bilibili\.com/v/[a-zA-Z]+\/[a-zA-Z]+'
 544     _TESTS = [{
 545         'url': 'https://www.bilibili.com/v/kichiku/mad',
 546         'info_dict': {
 547             'id': 'kichiku: mad',
 548             'title': 'kichiku: mad'
 549         },
 550         'playlist_mincount': 45,
 551         'params': {
 552             'playlistend': 45
 553         }
 554     }]
 555
 556     def _fetch_page(self, api_url, num_pages, query, page_num):
 557         parsed_json = self._download_json(
 558             api_url, query, query={'Search_key': query, 'pn': page_num},
 559             note='Extracting results from page %s of %s' % (page_num, num_pages))
 560
 561         video_list = traverse_obj(parsed_json, ('data', 'archives'), expected_type=list)
 562         if not video_list:
 563             raise ExtractorError('Failed to retrieve video list for page %d' % page_num)
 564
 565         for video in video_list:
 566             yield self.url_result(
 567                 'https://www.bilibili.com/video/%s' % video['bvid'], 'BiliBili', video['bvid'])
 568
 569     def _entries(self, category, subcategory, query):
 570         # map of categories : subcategories : RIDs
 571         rid_map = {
 572             'kichiku': {
 573                 'mad': 26,
 574                 'manual_vocaloid': 126,
 575                 'guide': 22,
 576                 'theatre': 216,
 577                 'course': 127
 578             },
 579         }
 580
 581         if category not in rid_map:
 582             raise ExtractorError(
 583                 f'The category {category} isn\'t supported. Supported categories: {list(rid_map.keys())}')
 584         if subcategory not in rid_map[category]:
 585             raise ExtractorError(
 586                 f'The subcategory {subcategory} isn\'t supported for this category. Supported subcategories: {list(rid_map[category].keys())}')
 587         rid_value = rid_map[category][subcategory]
 588
 589         api_url = 'https://api.bilibili.com/x/web-interface/newlist?rid=%d&type=1&ps=20&jsonp=jsonp' % rid_value
 590         page_json = self._download_json(api_url, query, query={'Search_key': query, 'pn': '1'})
 591         page_data = traverse_obj(page_json, ('data', 'page'), expected_type=dict)
 592         count, size = int_or_none(page_data.get('count')), int_or_none(page_data.get('size'))
 593         if count is None or not size:
 594             raise ExtractorError('Failed to calculate either page count or size')
 595
 596         num_pages = math.ceil(count / size)
 597
 598         return OnDemandPagedList(functools.partial(
 599             self._fetch_page, api_url, num_pages, query), size)
 600
 601     def _real_extract(self, url):
 602         u = compat_urllib_parse_urlparse(url)
 603         category, subcategory = u.path.split('/')[2:4]
 604         query = '%s: %s' % (category, subcategory)
 605
 606         return self.playlist_result(self._entries(category, subcategory, query), query, query)
 607
 608
 609 class BiliBiliSearchIE(SearchInfoExtractor):
 610     IE_DESC = 'Bilibili video search'
 611     _MAX_RESULTS = 100000
 612     _SEARCH_KEY = 'bilisearch'
 613
 614     def _search_results(self, query):
 615         for page_num in itertools.count(1):
 616             videos = self._download_json(
 617                 'https://api.bilibili.com/x/web-interface/search/type', query,
 618                 note=f'Extracting results from page {page_num}', query={
 619                     'Search_key': query,
 620                     'keyword': query,
 621                     'page': page_num,
 622                     'context': '',
 623                     'duration': 0,
 624                     'tids_2': '',
 625                     '__refresh__': 'true',
 626                     'search_type': 'video',
 627                     'tids': 0,
 628                     'highlight': 1,
 629                 })['data'].get('result')
 630             if not videos:
 631                 break
 632             for video in videos:
 633                 yield self.url_result(video['arcurl'], 'BiliBili', str(video['aid']))
 634
 635
 636 class BilibiliAudioBaseIE(InfoExtractor):
 637     def _call_api(self, path, sid, query=None):
 638         if not query:
 639             query = {'sid': sid}
 640         return self._download_json(
 641             'https://www.bilibili.com/audio/music-service-c/web/' + path,
 642             sid, query=query)['data']
 643
 644
 645 class BilibiliAudioIE(BilibiliAudioBaseIE):
 646     _VALID_URL = r'https?://(?:www\.)?bilibili\.com/audio/au(?P<id>\d+)'
 647     _TEST = {
 648         'url': 'https://www.bilibili.com/audio/au1003142',
 649         'md5': 'fec4987014ec94ef9e666d4d158ad03b',
 650         'info_dict': {
 651             'id': '1003142',
 652             'ext': 'm4a',
 653             'title': '【tsukimi】YELLOW / 神山羊',
 654             'artist': 'tsukimi',
 655             'comment_count': int,
 656             'description': 'YELLOW的mp3版！',
 657             'duration': 183,
 658             'subtitles': {
 659                 'origin': [{
 660                     'ext': 'lrc',
 661                 }],
 662             },
 663             'thumbnail': r're:^https?://.+\.jpg',
 664             'timestamp': 1564836614,
 665             'upload_date': '20190803',
 666             'uploader': 'tsukimi-つきみぐー',
 667             'view_count': int,
 668         },
 669     }
 670
 671     def _real_extract(self, url):
 672         au_id = self._match_id(url)
 673
 674         play_data = self._call_api('url', au_id)
 675         formats = [{
 676             'url': play_data['cdns'][0],
 677             'filesize': int_or_none(play_data.get('size')),
 678             'vcodec': 'none'
 679         }]
 680
 681         for a_format in formats:
 682             a_format.setdefault('http_headers', {}).update({
 683                 'Referer': url,
 684             })
 685
 686         song = self._call_api('song/info', au_id)
 687         title = song['title']
 688         statistic = song.get('statistic') or {}
 689
 690         subtitles = None
 691         lyric = song.get('lyric')
 692         if lyric:
 693             subtitles = {
 694                 'origin': [{
 695                     'url': lyric,
 696                 }]
 697             }
 698
 699         return {
 700             'id': au_id,
 701             'title': title,
 702             'formats': formats,
 703             'artist': song.get('author'),
 704             'comment_count': int_or_none(statistic.get('comment')),
 705             'description': song.get('intro'),
 706             'duration': int_or_none(song.get('duration')),
 707             'subtitles': subtitles,
 708             'thumbnail': song.get('cover'),
 709             'timestamp': int_or_none(song.get('passtime')),
 710             'uploader': song.get('uname'),
 711             'view_count': int_or_none(statistic.get('play')),
 712         }
 713
 714
 715 class BilibiliAudioAlbumIE(BilibiliAudioBaseIE):
 716     _VALID_URL = r'https?://(?:www\.)?bilibili\.com/audio/am(?P<id>\d+)'
 717     _TEST = {
 718         'url': 'https://www.bilibili.com/audio/am10624',
 719         'info_dict': {
 720             'id': '10624',
 721             'title': '每日新曲推荐（每日11:00更新）',
 722             'description': '每天11:00更新，为你推送最新音乐',
 723         },
 724         'playlist_count': 19,
 725     }
 726
 727     def _real_extract(self, url):
 728         am_id = self._match_id(url)
 729
 730         songs = self._call_api(
 731             'song/of-menu', am_id, {'sid': am_id, 'pn': 1, 'ps': 100})['data']
 732
 733         entries = []
 734         for song in songs:
 735             sid = str_or_none(song.get('id'))
 736             if not sid:
 737                 continue
 738             entries.append(self.url_result(
 739                 'https://www.bilibili.com/audio/au' + sid,
 740                 BilibiliAudioIE.ie_key(), sid))
 741
 742         if entries:
 743             album_data = self._call_api('menu/info', am_id) or {}
 744             album_title = album_data.get('title')
 745             if album_title:
 746                 for entry in entries:
 747                     entry['album'] = album_title
 748                 return self.playlist_result(
 749                     entries, am_id, album_title, album_data.get('intro'))
 750
 751         return self.playlist_result(entries, am_id)
 752
 753
 754 class BiliBiliPlayerIE(InfoExtractor):
 755     _VALID_URL = r'https?://player\.bilibili\.com/player\.html\?.*?\baid=(?P<id>\d+)'
 756     _TEST = {
 757         'url': 'http://player.bilibili.com/player.html?aid=92494333&cid=157926707&page=1',
 758         'only_matching': True,
 759     }
 760
 761     def _real_extract(self, url):
 762         video_id = self._match_id(url)
 763         return self.url_result(
 764             'http://www.bilibili.tv/video/av%s/' % video_id,
 765             ie=BiliBiliIE.ie_key(), video_id=video_id)
 766
 767
 768 class BiliIntlBaseIE(InfoExtractor):
 769     _API_URL = 'https://api.bilibili.tv/intl/gateway'
 770     _NETRC_MACHINE = 'biliintl'
 771
 772     def _call_api(self, endpoint, *args, **kwargs):
 773         json = self._download_json(self._API_URL + endpoint, *args, **kwargs)
 774         if json.get('code'):
 775             if json['code'] in (10004004, 10004005, 10023006):
 776                 self.raise_login_required()
 777             elif json['code'] == 10004001:
 778                 self.raise_geo_restricted()
 779             else:
 780                 if json.get('message') and str(json['code']) != json['message']:
 781                     errmsg = f'{kwargs.get("errnote", "Unable to download JSON metadata")}: {self.IE_NAME} said: {json["message"]}'
 782                 else:
 783                     errmsg = kwargs.get('errnote', 'Unable to download JSON metadata')
 784                 if kwargs.get('fatal'):
 785                     raise ExtractorError(errmsg)
 786                 else:
 787                     self.report_warning(errmsg)
 788         return json.get('data')
 789
 790     def json2srt(self, json):
 791         data = '\n\n'.join(
 792             f'{i + 1}\n{srt_subtitles_timecode(line["from"])} --> {srt_subtitles_timecode(line["to"])}\n{line["content"]}'
 793             for i, line in enumerate(traverse_obj(json, (
 794                 'body', lambda _, l: l['content'] and l['from'] and l['to']))))
 795         return data
 796
 797     def _get_subtitles(self, *, ep_id=None, aid=None):
 798         sub_json = self._call_api(
 799             '/web/v2/subtitle', ep_id or aid, fatal=False,
 800             note='Downloading subtitles list', errnote='Unable to download subtitles list',
 801             query=filter_dict({
 802                 'platform': 'web',
 803                 's_locale': 'en_US',
 804                 'episode_id': ep_id,
 805                 'aid': aid,
 806             })) or {}
 807         subtitles = {}
 808         for sub in sub_json.get('subtitles') or []:
 809             sub_url = sub.get('url')
 810             if not sub_url:
 811                 continue
 812             sub_data = self._download_json(
 813                 sub_url, ep_id or aid, errnote='Unable to download subtitles', fatal=False,
 814                 note='Downloading subtitles%s' % f' for {sub["lang"]}' if sub.get('lang') else '')
 815             if not sub_data:
 816                 continue
 817             subtitles.setdefault(sub.get('lang_key', 'en'), []).append({
 818                 'ext': 'srt',
 819                 'data': self.json2srt(sub_data)
 820             })
 821         return subtitles
 822
 823     def _get_formats(self, *, ep_id=None, aid=None):
 824         video_json = self._call_api(
 825             '/web/playurl', ep_id or aid, note='Downloading video formats',
 826             errnote='Unable to download video formats', query=filter_dict({
 827                 'platform': 'web',
 828                 'ep_id': ep_id,
 829                 'aid': aid,
 830             }))
 831         video_json = video_json['playurl']
 832         formats = []
 833         for vid in video_json.get('video') or []:
 834             video_res = vid.get('video_resource') or {}
 835             video_info = vid.get('stream_info') or {}
 836             if not video_res.get('url'):
 837                 continue
 838             formats.append({
 839                 'url': video_res['url'],
 840                 'ext': 'mp4',
 841                 'format_note': video_info.get('desc_words'),
 842                 'width': video_res.get('width'),
 843                 'height': video_res.get('height'),
 844                 'vbr': video_res.get('bandwidth'),
 845                 'acodec': 'none',
 846                 'vcodec': video_res.get('codecs'),
 847                 'filesize': video_res.get('size'),
 848             })
 849         for aud in video_json.get('audio_resource') or []:
 850             if not aud.get('url'):
 851                 continue
 852             formats.append({
 853                 'url': aud['url'],
 854                 'ext': 'mp4',
 855                 'abr': aud.get('bandwidth'),
 856                 'acodec': aud.get('codecs'),
 857                 'vcodec': 'none',
 858                 'filesize': aud.get('size'),
 859             })
 860
 861         self._sort_formats(formats)
 862         return formats
 863
 864     def _extract_video_info(self, video_data, *, ep_id=None, aid=None):
 865         return {
 866             'id': ep_id or aid,
 867             'title': video_data.get('title_display') or video_data.get('title'),
 868             'thumbnail': video_data.get('cover'),
 869             'episode_number': int_or_none(self._search_regex(
 870                 r'^E(\d+)(?:$| - )', video_data.get('title_display') or '', 'episode number', default=None)),
 871             'formats': self._get_formats(ep_id=ep_id, aid=aid),
 872             'subtitles': self._get_subtitles(ep_id=ep_id, aid=aid),
 873             'extractor_key': BiliIntlIE.ie_key(),
 874         }
 875
 876     def _perform_login(self, username, password):
 877         try:
 878             from Cryptodome.PublicKey import RSA
 879             from Cryptodome.Cipher import PKCS1_v1_5
 880         except ImportError:
 881             try:
 882                 from Crypto.PublicKey import RSA
 883                 from Crypto.Cipher import PKCS1_v1_5
 884             except ImportError:
 885                 raise ExtractorError('pycryptodomex not found. Please install', expected=True)
 886
 887         key_data = self._download_json(
 888             'https://passport.bilibili.tv/x/intl/passport-login/web/key?lang=en-US', None,
 889             note='Downloading login key', errnote='Unable to download login key')['data']
 890
 891         public_key = RSA.importKey(key_data['key'])
 892         password_hash = PKCS1_v1_5.new(public_key).encrypt((key_data['hash'] + password).encode('utf-8'))
 893         login_post = self._download_json(
 894             'https://passport.bilibili.tv/x/intl/passport-login/web/login/password?lang=en-US', None, data=urlencode_postdata({
 895                 'username': username,
 896                 'password': base64.b64encode(password_hash).decode('ascii'),
 897                 'keep_me': 'true',
 898                 's_locale': 'en_US',
 899                 'isTrusted': 'true'
 900             }), note='Logging in', errnote='Unable to log in')
 901         if login_post.get('code'):
 902             if login_post.get('message'):
 903                 raise ExtractorError(f'Unable to log in: {self.IE_NAME} said: {login_post["message"]}', expected=True)
 904             else:
 905                 raise ExtractorError('Unable to log in')
 906
 907
 908 class BiliIntlIE(BiliIntlBaseIE):
 909     _VALID_URL = r'https?://(?:www\.)?bili(?:bili\.tv|intl\.com)/(?:[a-z]{2}/)?(play/(?P<season_id>\d+)/(?P<ep_id>\d+)|video/(?P<aid>\d+))'
 910     _TESTS = [{
 911         # Bstation page
 912         'url': 'https://www.bilibili.tv/en/play/34613/341736',
 913         'info_dict': {
 914             'id': '341736',
 915             'ext': 'mp4',
 916             'title': 'E2 - The First Night',
 917             'thumbnail': r're:^https://pic\.bstarstatic\.com/ogv/.+\.png$',
 918             'episode_number': 2,
 919         }
 920     }, {
 921         # Non-Bstation page
 922         'url': 'https://www.bilibili.tv/en/play/1033760/11005006',
 923         'info_dict': {
 924             'id': '11005006',
 925             'ext': 'mp4',
 926             'title': 'E3 - Who?',
 927             'thumbnail': r're:^https://pic\.bstarstatic\.com/ogv/.+\.png$',
 928             'episode_number': 3,
 929         }
 930     }, {
 931         # Subtitle with empty content
 932         'url': 'https://www.bilibili.tv/en/play/1005144/10131790',
 933         'info_dict': {
 934             'id': '10131790',
 935             'ext': 'mp4',
 936             'title': 'E140 - Two Heartbeats: Kabuto\'s Trap',
 937             'thumbnail': r're:^https://pic\.bstarstatic\.com/ogv/.+\.png$',
 938             'episode_number': 140,
 939         },
 940         'skip': 'According to the copyright owner\'s request, you may only watch the video after you log in.'
 941     }, {
 942         'url': 'https://www.biliintl.com/en/play/34613/341736',
 943         'only_matching': True,
 944     }, {
 945         # User-generated content (as opposed to a series licensed from a studio)
 946         'url': 'https://bilibili.tv/en/video/2019955076',
 947         'only_matching': True,
 948     }, {
 949         # No language in URL
 950         'url': 'https://www.bilibili.tv/video/2019955076',
 951         'only_matching': True,
 952     }]
 953
 954     def _real_extract(self, url):
 955         season_id, ep_id, aid = self._match_valid_url(url).group('season_id', 'ep_id', 'aid')
 956         video_id = ep_id or aid
 957         webpage = self._download_webpage(url, video_id)
 958         # Bstation layout
 959         initial_data = (
 960             self._search_json(r'window\.__INITIAL_(?:DATA|STATE)__\s*=', webpage, 'preload state', video_id, default={})
 961             or self._search_nuxt_data(webpage, video_id, '__initialState', fatal=False, traverse=None))
 962         video_data = traverse_obj(
 963             initial_data, ('OgvVideo', 'epDetail'), ('UgcVideo', 'videoData'), ('ugc', 'archive'), expected_type=dict)
 964
 965         if season_id and not video_data:
 966             # Non-Bstation layout, read through episode list
 967             season_json = self._call_api(f'/web/v2/ogv/play/episodes?season_id={season_id}&platform=web', video_id)
 968             video_data = traverse_obj(season_json,
 969                                       ('sections', ..., 'episodes', lambda _, v: str(v['episode_id']) == ep_id),
 970                                       expected_type=dict, get_all=False)
 971         return self._extract_video_info(video_data or {}, ep_id=ep_id, aid=aid)
 972
 973
 974 class BiliIntlSeriesIE(BiliIntlBaseIE):
 975     _VALID_URL = r'https?://(?:www\.)?bili(?:bili\.tv|intl\.com)/(?:[a-z]{2}/)?play/(?P<id>\d+)$'
 976     _TESTS = [{
 977         'url': 'https://www.bilibili.tv/en/play/34613',
 978         'playlist_mincount': 15,
 979         'info_dict': {
 980             'id': '34613',
 981             'title': 'Fly Me to the Moon',
 982             'description': 'md5:a861ee1c4dc0acfad85f557cc42ac627',
 983             'categories': ['Romance', 'Comedy', 'Slice of life'],
 984             'thumbnail': r're:^https://pic\.bstarstatic\.com/ogv/.+\.png$',
 985             'view_count': int,
 986         },
 987         'params': {
 988             'skip_download': True,
 989         },
 990     }, {
 991         'url': 'https://www.biliintl.com/en/play/34613',
 992         'only_matching': True,
 993     }]
 994
 995     def _entries(self, series_id):
 996         series_json = self._call_api(f'/web/v2/ogv/play/episodes?season_id={series_id}&platform=web', series_id)
 997         for episode in traverse_obj(series_json, ('sections', ..., 'episodes', ...), expected_type=dict, default=[]):
 998             episode_id = str(episode.get('episode_id'))
 999             yield self._extract_video_info(episode, ep_id=episode_id)
1000
1001     def _real_extract(self, url):
1002         series_id = self._match_id(url)
1003         series_info = self._call_api(f'/web/v2/ogv/play/season_info?season_id={series_id}&platform=web', series_id).get('season') or {}
1004         return self.playlist_result(
1005             self._entries(series_id), series_id, series_info.get('title'), series_info.get('description'),
1006             categories=traverse_obj(series_info, ('styles', ..., 'title'), expected_type=str_or_none),
1007             thumbnail=url_or_none(series_info.get('horizontal_cover')), view_count=parse_count(series_info.get('view')))
1008
1009
1010 class BiliLiveIE(InfoExtractor):
1011     _VALID_URL = r'https?://live.bilibili.com/(?P<id>\d+)'
1012
1013     _TESTS = [{
1014         'url': 'https://live.bilibili.com/196',
1015         'info_dict': {
1016             'id': '33989',
1017             'description': "周六杂谈回，其他时候随机游戏。 | \n录播：@下播型泛式录播组。 | \n直播通知群（全员禁言）：666906670，902092584，59971⑧481 （功能一样，别多加）",
1018             'ext': 'flv',
1019             'title': "太空狼人杀联动，不被爆杀就算赢",
1020             'thumbnail': "https://i0.hdslb.com/bfs/live/new_room_cover/e607bc1529057ef4b332e1026e62cf46984c314d.jpg",
1021             'timestamp': 1650802769,
1022         },
1023         'skip': 'not live'
1024     }, {
1025         'url': 'https://live.bilibili.com/196?broadcast_type=0&is_room_feed=1?spm_id_from=333.999.space_home.strengthen_live_card.click',
1026         'only_matching': True
1027     }]
1028
1029     _FORMATS = {
1030         80: {'format_id': 'low', 'format_note': '流畅'},
1031         150: {'format_id': 'high_res', 'format_note': '高清'},
1032         250: {'format_id': 'ultra_high_res', 'format_note': '超清'},
1033         400: {'format_id': 'blue_ray', 'format_note': '蓝光'},
1034         10000: {'format_id': 'source', 'format_note': '原画'},
1035         20000: {'format_id': '4K', 'format_note': '4K'},
1036         30000: {'format_id': 'dolby', 'format_note': '杜比'},
1037     }
1038
1039     _quality = staticmethod(qualities(list(_FORMATS)))
1040
1041     def _call_api(self, path, room_id, query):
1042         api_result = self._download_json(f'https://api.live.bilibili.com/{path}', room_id, query=query)
1043         if api_result.get('code') != 0:
1044             raise ExtractorError(api_result.get('message') or 'Unable to download JSON metadata')
1045         return api_result.get('data') or {}
1046
1047     def _parse_formats(self, qn, fmt):
1048         for codec in fmt.get('codec') or []:
1049             if codec.get('current_qn') != qn:
1050                 continue
1051             for url_info in codec['url_info']:
1052                 yield {
1053                     'url': f'{url_info["host"]}{codec["base_url"]}{url_info["extra"]}',
1054                     'ext': fmt.get('format_name'),
1055                     'vcodec': codec.get('codec_name'),
1056                     'quality': self._quality(qn),
1057                     **self._FORMATS[qn],
1058                 }
1059
1060     def _real_extract(self, url):
1061         room_id = self._match_id(url)
1062         room_data = self._call_api('room/v1/Room/get_info', room_id, {'id': room_id})
1063         if room_data.get('live_status') == 0:
1064             raise ExtractorError('Streamer is not live', expected=True)
1065
1066         formats = []
1067         for qn in self._FORMATS.keys():
1068             stream_data = self._call_api('xlive/web-room/v2/index/getRoomPlayInfo', room_id, {
1069                 'room_id': room_id,
1070                 'qn': qn,
1071                 'codec': '0,1',
1072                 'format': '0,2',
1073                 'mask': '0',
1074                 'no_playurl': '0',
1075                 'platform': 'web',
1076                 'protocol': '0,1',
1077             })
1078             for fmt in traverse_obj(stream_data, ('playurl_info', 'playurl', 'stream', ..., 'format', ...)) or []:
1079                 formats.extend(self._parse_formats(qn, fmt))
1080         self._sort_formats(formats)
1081
1082         return {
1083             'id': room_id,
1084             'title': room_data.get('title'),
1085             'description': room_data.get('description'),
1086             'thumbnail': room_data.get('user_cover'),
1087             'timestamp': stream_data.get('live_time'),
1088             'formats': formats,
1089             'http_headers': {
1090                 'Referer': url,
1091             },
1092         }