yt_dlp/extractor/bilibili.py

   1 # coding: utf-8
   2
   3 import hashlib
   4 import itertools
   5 import functools
   6 import re
   7 import math
   8
   9 from .common import InfoExtractor, SearchInfoExtractor
  10 from ..compat import (
  11     compat_parse_qs,
  12     compat_urlparse,
  13     compat_urllib_parse_urlparse
  14 )
  15 from ..utils import (
  16     ExtractorError,
  17     int_or_none,
  18     float_or_none,
  19     parse_iso8601,
  20     traverse_obj,
  21     try_get,
  22     smuggle_url,
  23     srt_subtitles_timecode,
  24     str_or_none,
  25     str_to_int,
  26     strip_jsonp,
  27     unified_timestamp,
  28     unsmuggle_url,
  29     urlencode_postdata,
  30     OnDemandPagedList
  31 )
  32
  33
  34 class BiliBiliIE(InfoExtractor):
  35     _VALID_URL = r'''(?x)
  36                     https?://
  37                         (?:(?:www|bangumi)\.)?
  38                         bilibili\.(?:tv|com)/
  39                         (?:
  40                             (?:
  41                                 video/[aA][vV]|
  42                                 anime/(?P<anime_id>\d+)/play\#
  43                             )(?P<id>\d+)|
  44                             (s/)?video/[bB][vV](?P<id_bv>[^/?#&]+)
  45                         )
  46                         (?:/?\?p=(?P<page>\d+))?
  47                     '''
  48
  49     _TESTS = [{
  50         'url': 'http://www.bilibili.com/video/av1074402/',
  51         'md5': '5f7d29e1a2872f3df0cf76b1f87d3788',
  52         'info_dict': {
  53             'id': '1074402',
  54             'ext': 'flv',
  55             'title': '【金坷垃】金泡沫',
  56             'description': 'md5:ce18c2a2d2193f0df2917d270f2e5923',
  57             'duration': 308.067,
  58             'timestamp': 1398012678,
  59             'upload_date': '20140420',
  60             'thumbnail': r're:^https?://.+\.jpg',
  61             'uploader': '菊子桑',
  62             'uploader_id': '156160',
  63         },
  64     }, {
  65         # Tested in BiliBiliBangumiIE
  66         'url': 'http://bangumi.bilibili.com/anime/1869/play#40062',
  67         'only_matching': True,
  68     }, {
  69         # bilibili.tv
  70         'url': 'http://www.bilibili.tv/video/av1074402/',
  71         'only_matching': True,
  72     }, {
  73         'url': 'http://bangumi.bilibili.com/anime/5802/play#100643',
  74         'md5': '3f721ad1e75030cc06faf73587cfec57',
  75         'info_dict': {
  76             'id': '100643',
  77             'ext': 'mp4',
  78             'title': 'CHAOS;CHILD',
  79             '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...',
  80         },
  81         'skip': 'Geo-restricted to China',
  82     }, {
  83         # Title with double quotes
  84         'url': 'http://www.bilibili.com/video/av8903802/',
  85         'info_dict': {
  86             'id': '8903802',
  87             'title': '阿滴英文｜英文歌分享#6 "Closer',
  88             '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',
  89         },
  90         'playlist': [{
  91             'info_dict': {
  92                 'id': '8903802_part1',
  93                 'ext': 'flv',
  94                 'title': '阿滴英文｜英文歌分享#6 "Closer',
  95                 'description': 'md5:3b1b9e25b78da4ef87e9b548b88ee76a',
  96                 'uploader': '阿滴英文',
  97                 'uploader_id': '65880958',
  98                 'timestamp': 1488382634,
  99                 'upload_date': '20170301',
 100             },
 101             'params': {
 102                 'skip_download': True,
 103             },
 104         }, {
 105             'info_dict': {
 106                 'id': '8903802_part2',
 107                 'ext': 'flv',
 108                 'title': '阿滴英文｜英文歌分享#6 "Closer',
 109                 'description': 'md5:3b1b9e25b78da4ef87e9b548b88ee76a',
 110                 'uploader': '阿滴英文',
 111                 'uploader_id': '65880958',
 112                 'timestamp': 1488382634,
 113                 'upload_date': '20170301',
 114             },
 115             'params': {
 116                 'skip_download': True,
 117             },
 118         }]
 119     }, {
 120         # new BV video id format
 121         'url': 'https://www.bilibili.com/video/BV1JE411F741',
 122         'only_matching': True,
 123     }, {
 124         # Anthology
 125         'url': 'https://www.bilibili.com/video/BV1bK411W797',
 126         'info_dict': {
 127             'id': 'BV1bK411W797',
 128             'title': '物语中的人物是如何吐槽自己的OP的'
 129         },
 130         'playlist_count': 17,
 131     }]
 132
 133     _APP_KEY = 'iVGUTjsxvpLeuDCf'
 134     _BILIBILI_KEY = 'aHRmhWMLkdeMuILqORnYZocwMBpMEOdt'
 135
 136     def _report_error(self, result):
 137         if 'message' in result:
 138             raise ExtractorError('%s said: %s' % (self.IE_NAME, result['message']), expected=True)
 139         elif 'code' in result:
 140             raise ExtractorError('%s returns error %d' % (self.IE_NAME, result['code']), expected=True)
 141         else:
 142             raise ExtractorError('Can\'t extract Bangumi episode ID')
 143
 144     def _real_extract(self, url):
 145         url, smuggled_data = unsmuggle_url(url, {})
 146
 147         mobj = self._match_valid_url(url)
 148         video_id = mobj.group('id_bv') or mobj.group('id')
 149
 150         av_id, bv_id = self._get_video_id_set(video_id, mobj.group('id_bv') is not None)
 151         video_id = av_id
 152
 153         anime_id = mobj.group('anime_id')
 154         page_id = mobj.group('page')
 155         webpage = self._download_webpage(url, video_id)
 156
 157         # Bilibili anthologies are similar to playlists but all videos share the same video ID as the anthology itself.
 158         # If the video has no page argument, check to see if it's an anthology
 159         if page_id is None:
 160             if not self.get_param('noplaylist'):
 161                 r = self._extract_anthology_entries(bv_id, video_id, webpage)
 162                 if r is not None:
 163                     self.to_screen('Downloading anthology %s - add --no-playlist to just download video' % video_id)
 164                     return r
 165             else:
 166                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
 167
 168         if 'anime/' not in url:
 169             cid = self._search_regex(
 170                 r'\bcid(?:["\']:|=)(\d+),["\']page(?:["\']:|=)' + str(page_id), webpage, 'cid',
 171                 default=None
 172             ) or self._search_regex(
 173                 r'\bcid(?:["\']:|=)(\d+)', webpage, 'cid',
 174                 default=None
 175             ) or compat_parse_qs(self._search_regex(
 176                 [r'EmbedPlayer\([^)]+,\s*"([^"]+)"\)',
 177                  r'EmbedPlayer\([^)]+,\s*\\"([^"]+)\\"\)',
 178                  r'<iframe[^>]+src="https://secure\.bilibili\.com/secure,([^"]+)"'],
 179                 webpage, 'player parameters'))['cid'][0]
 180         else:
 181             if 'no_bangumi_tip' not in smuggled_data:
 182                 self.to_screen('Downloading episode %s. To download all videos in anime %s, re-run yt-dlp with %s' % (
 183                     video_id, anime_id, compat_urlparse.urljoin(url, '//bangumi.bilibili.com/anime/%s' % anime_id)))
 184             headers = {
 185                 'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8',
 186                 'Referer': url
 187             }
 188             headers.update(self.geo_verification_headers())
 189
 190             js = self._download_json(
 191                 'http://bangumi.bilibili.com/web_api/get_source', video_id,
 192                 data=urlencode_postdata({'episode_id': video_id}),
 193                 headers=headers)
 194             if 'result' not in js:
 195                 self._report_error(js)
 196             cid = js['result']['cid']
 197
 198         headers = {
 199             'Accept': 'application/json',
 200             'Referer': url
 201         }
 202         headers.update(self.geo_verification_headers())
 203
 204         entries = []
 205
 206         RENDITIONS = ('qn=80&quality=80&type=', 'quality=2&type=mp4')
 207         for num, rendition in enumerate(RENDITIONS, start=1):
 208             payload = 'appkey=%s&cid=%s&otype=json&%s' % (self._APP_KEY, cid, rendition)
 209             sign = hashlib.md5((payload + self._BILIBILI_KEY).encode('utf-8')).hexdigest()
 210
 211             video_info = self._download_json(
 212                 'http://interface.bilibili.com/v2/playurl?%s&sign=%s' % (payload, sign),
 213                 video_id, note='Downloading video info page',
 214                 headers=headers, fatal=num == len(RENDITIONS))
 215
 216             if not video_info:
 217                 continue
 218
 219             if 'durl' not in video_info:
 220                 if num < len(RENDITIONS):
 221                     continue
 222                 self._report_error(video_info)
 223
 224             for idx, durl in enumerate(video_info['durl']):
 225                 formats = [{
 226                     'url': durl['url'],
 227                     'filesize': int_or_none(durl['size']),
 228                 }]
 229                 for backup_url in durl.get('backup_url', []):
 230                     formats.append({
 231                         'url': backup_url,
 232                         # backup URLs have lower priorities
 233                         'quality': -2 if 'hd.mp4' in backup_url else -3,
 234                     })
 235
 236                 for a_format in formats:
 237                     a_format.setdefault('http_headers', {}).update({
 238                         'Referer': url,
 239                     })
 240
 241                 self._sort_formats(formats)
 242
 243                 entries.append({
 244                     'id': '%s_part%s' % (video_id, idx),
 245                     'duration': float_or_none(durl.get('length'), 1000),
 246                     'formats': formats,
 247                 })
 248             break
 249
 250         title = self._html_search_regex(
 251             (r'<h1[^>]+\btitle=(["\'])(?P<title>(?:(?!\1).)+)\1',
 252              r'(?s)<h1[^>]*>(?P<title>.+?)</h1>'), webpage, 'title',
 253             group='title')
 254
 255         # Get part title for anthologies
 256         if page_id is not None:
 257             # TODO: The json is already downloaded by _extract_anthology_entries. Don't redownload for each video
 258             part_title = try_get(
 259                 self._download_json(
 260                     f'https://api.bilibili.com/x/player/pagelist?bvid={bv_id}&jsonp=jsonp',
 261                     video_id, note='Extracting videos in anthology'),
 262                 lambda x: x['data'][int(page_id) - 1]['part'])
 263             title = part_title or title
 264
 265         description = self._html_search_meta('description', webpage)
 266         timestamp = unified_timestamp(self._html_search_regex(
 267             r'<time[^>]+datetime="([^"]+)"', webpage, 'upload time',
 268             default=None) or self._html_search_meta(
 269             'uploadDate', webpage, 'timestamp', default=None))
 270         thumbnail = self._html_search_meta(['og:image', 'thumbnailUrl'], webpage)
 271
 272         # TODO 'view_count' requires deobfuscating Javascript
 273         info = {
 274             'id': str(video_id) if page_id is None else '%s_part%s' % (video_id, page_id),
 275             'cid': cid,
 276             'title': title,
 277             'description': description,
 278             'timestamp': timestamp,
 279             'thumbnail': thumbnail,
 280             'duration': float_or_none(video_info.get('timelength'), scale=1000),
 281         }
 282
 283         uploader_mobj = re.search(
 284             r'<a[^>]+href="(?:https?:)?//space\.bilibili\.com/(?P<id>\d+)"[^>]*>\s*(?P<name>[^<]+?)\s*<',
 285             webpage)
 286         if uploader_mobj:
 287             info.update({
 288                 'uploader': uploader_mobj.group('name').strip(),
 289                 'uploader_id': uploader_mobj.group('id'),
 290             })
 291
 292         if not info.get('uploader'):
 293             info['uploader'] = self._html_search_meta(
 294                 'author', webpage, 'uploader', default=None)
 295
 296         top_level_info = {
 297             'tags': traverse_obj(self._download_json(
 298                 f'https://api.bilibili.com/x/tag/archive/tags?aid={video_id}',
 299                 video_id, fatal=False, note='Downloading tags'), ('data', ..., 'tag_name')),
 300         }
 301
 302         entries[0]['subtitles'] = {
 303             'danmaku': [{
 304                 'ext': 'xml',
 305                 'url': f'https://comment.bilibili.com/{cid}.xml',
 306             }]
 307         }
 308
 309         r'''
 310         # Requires https://github.com/m13253/danmaku2ass which is licenced under GPL3
 311         # See https://github.com/animelover1984/youtube-dl
 312
 313         raw_danmaku = self._download_webpage(
 314             f'https://comment.bilibili.com/{cid}.xml', video_id, fatal=False, note='Downloading danmaku comments')
 315         danmaku = NiconicoIE.CreateDanmaku(raw_danmaku, commentType='Bilibili', x=1024, y=576)
 316         entries[0]['subtitles'] = {
 317             'danmaku': [{
 318                 'ext': 'ass',
 319                 'data': danmaku
 320             }]
 321         }
 322         '''
 323
 324         top_level_info['__post_extractor'] = self.extract_comments(video_id)
 325
 326         for entry in entries:
 327             entry.update(info)
 328
 329         if len(entries) == 1:
 330             entries[0].update(top_level_info)
 331             return entries[0]
 332
 333         for idx, entry in enumerate(entries):
 334             entry['id'] = '%s_part%d' % (video_id, (idx + 1))
 335
 336         return {
 337             '_type': 'multi_video',
 338             'id': str(video_id),
 339             'bv_id': bv_id,
 340             'title': title,
 341             'description': description,
 342             'entries': entries,
 343             **info, **top_level_info
 344         }
 345
 346     def _extract_anthology_entries(self, bv_id, video_id, webpage):
 347         title = self._html_search_regex(
 348             (r'<h1[^>]+\btitle=(["\'])(?P<title>(?:(?!\1).)+)\1',
 349              r'(?s)<h1[^>]*>(?P<title>.+?)</h1>'), webpage, 'title',
 350             group='title')
 351         json_data = self._download_json(
 352             f'https://api.bilibili.com/x/player/pagelist?bvid={bv_id}&jsonp=jsonp',
 353             video_id, note='Extracting videos in anthology')
 354
 355         if json_data['data']:
 356             return self.playlist_from_matches(
 357                 json_data['data'], bv_id, title, ie=BiliBiliIE.ie_key(),
 358                 getter=lambda entry: 'https://www.bilibili.com/video/%s?p=%d' % (bv_id, entry['page']))
 359
 360     def _get_video_id_set(self, id, is_bv):
 361         query = {'bvid': id} if is_bv else {'aid': id}
 362         response = self._download_json(
 363             "http://api.bilibili.cn/x/web-interface/view",
 364             id, query=query,
 365             note='Grabbing original ID via API')
 366
 367         if response['code'] == -400:
 368             raise ExtractorError('Video ID does not exist', expected=True, video_id=id)
 369         elif response['code'] != 0:
 370             raise ExtractorError(f'Unknown error occurred during API check (code {response["code"]})',
 371                                  expected=True, video_id=id)
 372         return response['data']['aid'], response['data']['bvid']
 373
 374     def _get_comments(self, video_id, commentPageNumber=0):
 375         for idx in itertools.count(1):
 376             replies = traverse_obj(
 377                 self._download_json(
 378                     f'https://api.bilibili.com/x/v2/reply?pn={idx}&oid={video_id}&type=1&jsonp=jsonp&sort=2&_=1567227301685',
 379                     video_id, note=f'Extracting comments from page {idx}'),
 380                 ('data', 'replies')) or []
 381             for children in map(self._get_all_children, replies):
 382                 yield from children
 383
 384     def _get_all_children(self, reply):
 385         yield {
 386             'author': traverse_obj(reply, ('member', 'uname')),
 387             'author_id': traverse_obj(reply, ('member', 'mid')),
 388             'id': reply.get('rpid'),
 389             'text': traverse_obj(reply, ('content', 'message')),
 390             'timestamp': reply.get('ctime'),
 391             'parent': reply.get('parent') or 'root',
 392         }
 393         for children in map(self._get_all_children, reply.get('replies') or []):
 394             yield from children
 395
 396
 397 class BiliBiliBangumiIE(InfoExtractor):
 398     _VALID_URL = r'https?://bangumi\.bilibili\.com/anime/(?P<id>\d+)'
 399
 400     IE_NAME = 'bangumi.bilibili.com'
 401     IE_DESC = 'BiliBili番剧'
 402
 403     _TESTS = [{
 404         'url': 'http://bangumi.bilibili.com/anime/1869',
 405         'info_dict': {
 406             'id': '1869',
 407             'title': '混沌武士',
 408             'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
 409         },
 410         'playlist_count': 26,
 411     }, {
 412         'url': 'http://bangumi.bilibili.com/anime/1869',
 413         'info_dict': {
 414             'id': '1869',
 415             'title': '混沌武士',
 416             'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
 417         },
 418         'playlist': [{
 419             'md5': '91da8621454dd58316851c27c68b0c13',
 420             'info_dict': {
 421                 'id': '40062',
 422                 'ext': 'mp4',
 423                 'title': '混沌武士',
 424                 '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...',
 425                 'timestamp': 1414538739,
 426                 'upload_date': '20141028',
 427                 'episode': '疾风怒涛 Tempestuous Temperaments',
 428                 'episode_number': 1,
 429             },
 430         }],
 431         'params': {
 432             'playlist_items': '1',
 433         },
 434     }]
 435
 436     @classmethod
 437     def suitable(cls, url):
 438         return False if BiliBiliIE.suitable(url) else super(BiliBiliBangumiIE, cls).suitable(url)
 439
 440     def _real_extract(self, url):
 441         bangumi_id = self._match_id(url)
 442
 443         # Sometimes this API returns a JSONP response
 444         season_info = self._download_json(
 445             'http://bangumi.bilibili.com/jsonp/seasoninfo/%s.ver' % bangumi_id,
 446             bangumi_id, transform_source=strip_jsonp)['result']
 447
 448         entries = [{
 449             '_type': 'url_transparent',
 450             'url': smuggle_url(episode['webplay_url'], {'no_bangumi_tip': 1}),
 451             'ie_key': BiliBiliIE.ie_key(),
 452             'timestamp': parse_iso8601(episode.get('update_time'), delimiter=' '),
 453             'episode': episode.get('index_title'),
 454             'episode_number': int_or_none(episode.get('index')),
 455         } for episode in season_info['episodes']]
 456
 457         entries = sorted(entries, key=lambda entry: entry.get('episode_number'))
 458
 459         return self.playlist_result(
 460             entries, bangumi_id,
 461             season_info.get('bangumi_title'), season_info.get('evaluate'))
 462
 463
 464 class BilibiliChannelIE(InfoExtractor):
 465     _VALID_URL = r'https?://space.bilibili\.com/(?P<id>\d+)'
 466     _API_URL = "https://api.bilibili.com/x/space/arc/search?mid=%s&pn=%d&jsonp=jsonp"
 467     _TESTS = [{
 468         'url': 'https://space.bilibili.com/3985676/video',
 469         'info_dict': {},
 470         'playlist_mincount': 112,
 471     }]
 472
 473     def _entries(self, list_id):
 474         count, max_count = 0, None
 475
 476         for page_num in itertools.count(1):
 477             data = self._download_json(
 478                 self._API_URL % (list_id, page_num), list_id, note=f'Downloading page {page_num}')['data']
 479
 480             max_count = max_count or try_get(data, lambda x: x['page']['count'])
 481
 482             entries = try_get(data, lambda x: x['list']['vlist'])
 483             if not entries:
 484                 return
 485             for entry in entries:
 486                 yield self.url_result(
 487                     'https://www.bilibili.com/video/%s' % entry['bvid'],
 488                     BiliBiliIE.ie_key(), entry['bvid'])
 489
 490             count += len(entries)
 491             if max_count and count >= max_count:
 492                 return
 493
 494     def _real_extract(self, url):
 495         list_id = self._match_id(url)
 496         return self.playlist_result(self._entries(list_id), list_id)
 497
 498
 499 class BilibiliCategoryIE(InfoExtractor):
 500     IE_NAME = 'Bilibili category extractor'
 501     _MAX_RESULTS = 1000000
 502     _VALID_URL = r'https?://www\.bilibili\.com/v/[a-zA-Z]+\/[a-zA-Z]+'
 503     _TESTS = [{
 504         'url': 'https://www.bilibili.com/v/kichiku/mad',
 505         'info_dict': {
 506             'id': 'kichiku: mad',
 507             'title': 'kichiku: mad'
 508         },
 509         'playlist_mincount': 45,
 510         'params': {
 511             'playlistend': 45
 512         }
 513     }]
 514
 515     def _fetch_page(self, api_url, num_pages, query, page_num):
 516         parsed_json = self._download_json(
 517             api_url, query, query={'Search_key': query, 'pn': page_num},
 518             note='Extracting results from page %s of %s' % (page_num, num_pages))
 519
 520         video_list = try_get(parsed_json, lambda x: x['data']['archives'], list)
 521         if not video_list:
 522             raise ExtractorError('Failed to retrieve video list for page %d' % page_num)
 523
 524         for video in video_list:
 525             yield self.url_result(
 526                 'https://www.bilibili.com/video/%s' % video['bvid'], 'BiliBili', video['bvid'])
 527
 528     def _entries(self, category, subcategory, query):
 529         # map of categories : subcategories : RIDs
 530         rid_map = {
 531             'kichiku': {
 532                 'mad': 26,
 533                 'manual_vocaloid': 126,
 534                 'guide': 22,
 535                 'theatre': 216,
 536                 'course': 127
 537             },
 538         }
 539
 540         if category not in rid_map:
 541             raise ExtractorError(
 542                 f'The category {category} isn\'t supported. Supported categories: {list(rid_map.keys())}')
 543         if subcategory not in rid_map[category]:
 544             raise ExtractorError(
 545                 f'The subcategory {subcategory} isn\'t supported for this category. Supported subcategories: {list(rid_map[category].keys())}')
 546         rid_value = rid_map[category][subcategory]
 547
 548         api_url = 'https://api.bilibili.com/x/web-interface/newlist?rid=%d&type=1&ps=20&jsonp=jsonp' % rid_value
 549         page_json = self._download_json(api_url, query, query={'Search_key': query, 'pn': '1'})
 550         page_data = try_get(page_json, lambda x: x['data']['page'], dict)
 551         count, size = int_or_none(page_data.get('count')), int_or_none(page_data.get('size'))
 552         if count is None or not size:
 553             raise ExtractorError('Failed to calculate either page count or size')
 554
 555         num_pages = math.ceil(count / size)
 556
 557         return OnDemandPagedList(functools.partial(
 558             self._fetch_page, api_url, num_pages, query), size)
 559
 560     def _real_extract(self, url):
 561         u = compat_urllib_parse_urlparse(url)
 562         category, subcategory = u.path.split('/')[2:4]
 563         query = '%s: %s' % (category, subcategory)
 564
 565         return self.playlist_result(self._entries(category, subcategory, query), query, query)
 566
 567
 568 class BiliBiliSearchIE(SearchInfoExtractor):
 569     IE_DESC = 'Bilibili video search'
 570     _MAX_RESULTS = 100000
 571     _SEARCH_KEY = 'bilisearch'
 572
 573     def _search_results(self, query):
 574         for page_num in itertools.count(1):
 575             videos = self._download_json(
 576                 'https://api.bilibili.com/x/web-interface/search/type', query,
 577                 note=f'Extracting results from page {page_num}', query={
 578                     'Search_key': query,
 579                     'keyword': query,
 580                     'page': page_num,
 581                     'context': '',
 582                     'order': 'pubdate',
 583                     'duration': 0,
 584                     'tids_2': '',
 585                     '__refresh__': 'true',
 586                     'search_type': 'video',
 587                     'tids': 0,
 588                     'highlight': 1,
 589                 })['data'].get('result') or []
 590             for video in videos:
 591                 yield self.url_result(video['arcurl'], 'BiliBili', str(video['aid']))
 592
 593
 594 class BilibiliAudioBaseIE(InfoExtractor):
 595     def _call_api(self, path, sid, query=None):
 596         if not query:
 597             query = {'sid': sid}
 598         return self._download_json(
 599             'https://www.bilibili.com/audio/music-service-c/web/' + path,
 600             sid, query=query)['data']
 601
 602
 603 class BilibiliAudioIE(BilibiliAudioBaseIE):
 604     _VALID_URL = r'https?://(?:www\.)?bilibili\.com/audio/au(?P<id>\d+)'
 605     _TEST = {
 606         'url': 'https://www.bilibili.com/audio/au1003142',
 607         'md5': 'fec4987014ec94ef9e666d4d158ad03b',
 608         'info_dict': {
 609             'id': '1003142',
 610             'ext': 'm4a',
 611             'title': '【tsukimi】YELLOW / 神山羊',
 612             'artist': 'tsukimi',
 613             'comment_count': int,
 614             'description': 'YELLOW的mp3版！',
 615             'duration': 183,
 616             'subtitles': {
 617                 'origin': [{
 618                     'ext': 'lrc',
 619                 }],
 620             },
 621             'thumbnail': r're:^https?://.+\.jpg',
 622             'timestamp': 1564836614,
 623             'upload_date': '20190803',
 624             'uploader': 'tsukimi-つきみぐー',
 625             'view_count': int,
 626         },
 627     }
 628
 629     def _real_extract(self, url):
 630         au_id = self._match_id(url)
 631
 632         play_data = self._call_api('url', au_id)
 633         formats = [{
 634             'url': play_data['cdns'][0],
 635             'filesize': int_or_none(play_data.get('size')),
 636             'vcodec': 'none'
 637         }]
 638
 639         song = self._call_api('song/info', au_id)
 640         title = song['title']
 641         statistic = song.get('statistic') or {}
 642
 643         subtitles = None
 644         lyric = song.get('lyric')
 645         if lyric:
 646             subtitles = {
 647                 'origin': [{
 648                     'url': lyric,
 649                 }]
 650             }
 651
 652         return {
 653             'id': au_id,
 654             'title': title,
 655             'formats': formats,
 656             'artist': song.get('author'),
 657             'comment_count': int_or_none(statistic.get('comment')),
 658             'description': song.get('intro'),
 659             'duration': int_or_none(song.get('duration')),
 660             'subtitles': subtitles,
 661             'thumbnail': song.get('cover'),
 662             'timestamp': int_or_none(song.get('passtime')),
 663             'uploader': song.get('uname'),
 664             'view_count': int_or_none(statistic.get('play')),
 665         }
 666
 667
 668 class BilibiliAudioAlbumIE(BilibiliAudioBaseIE):
 669     _VALID_URL = r'https?://(?:www\.)?bilibili\.com/audio/am(?P<id>\d+)'
 670     _TEST = {
 671         'url': 'https://www.bilibili.com/audio/am10624',
 672         'info_dict': {
 673             'id': '10624',
 674             'title': '每日新曲推荐（每日11:00更新）',
 675             'description': '每天11:00更新，为你推送最新音乐',
 676         },
 677         'playlist_count': 19,
 678     }
 679
 680     def _real_extract(self, url):
 681         am_id = self._match_id(url)
 682
 683         songs = self._call_api(
 684             'song/of-menu', am_id, {'sid': am_id, 'pn': 1, 'ps': 100})['data']
 685
 686         entries = []
 687         for song in songs:
 688             sid = str_or_none(song.get('id'))
 689             if not sid:
 690                 continue
 691             entries.append(self.url_result(
 692                 'https://www.bilibili.com/audio/au' + sid,
 693                 BilibiliAudioIE.ie_key(), sid))
 694
 695         if entries:
 696             album_data = self._call_api('menu/info', am_id) or {}
 697             album_title = album_data.get('title')
 698             if album_title:
 699                 for entry in entries:
 700                     entry['album'] = album_title
 701                 return self.playlist_result(
 702                     entries, am_id, album_title, album_data.get('intro'))
 703
 704         return self.playlist_result(entries, am_id)
 705
 706
 707 class BiliBiliPlayerIE(InfoExtractor):
 708     _VALID_URL = r'https?://player\.bilibili\.com/player\.html\?.*?\baid=(?P<id>\d+)'
 709     _TEST = {
 710         'url': 'http://player.bilibili.com/player.html?aid=92494333&cid=157926707&page=1',
 711         'only_matching': True,
 712     }
 713
 714     def _real_extract(self, url):
 715         video_id = self._match_id(url)
 716         return self.url_result(
 717             'http://www.bilibili.tv/video/av%s/' % video_id,
 718             ie=BiliBiliIE.ie_key(), video_id=video_id)
 719
 720
 721 class BiliIntlBaseIE(InfoExtractor):
 722     _API_URL = 'https://api.bili{}/intl/gateway{}'
 723
 724     def _call_api(self, type, endpoint, id):
 725         return self._download_json(self._API_URL.format(type, endpoint), id)['data']
 726
 727     def json2srt(self, json):
 728         data = '\n\n'.join(
 729             f'{i + 1}\n{srt_subtitles_timecode(line["from"])} --> {srt_subtitles_timecode(line["to"])}\n{line["content"]}'
 730             for i, line in enumerate(json['body']))
 731         return data
 732
 733     def _get_subtitles(self, type, ep_id):
 734         sub_json = self._call_api(type, f'/m/subtitle?ep_id={ep_id}&platform=web', ep_id)
 735         subtitles = {}
 736         for sub in sub_json.get('subtitles', []):
 737             sub_url = sub.get('url')
 738             if not sub_url:
 739                 continue
 740             sub_data = self._download_json(sub_url, ep_id, fatal=False)
 741             if not sub_data:
 742                 continue
 743             subtitles.setdefault(sub.get('key', 'en'), []).append({
 744                 'ext': 'srt',
 745                 'data': self.json2srt(sub_data)
 746             })
 747         return subtitles
 748
 749     def _get_formats(self, type, ep_id):
 750         video_json = self._call_api(type, f'/web/playurl?ep_id={ep_id}&platform=web', ep_id)
 751         if not video_json:
 752             self.raise_login_required(method='cookies')
 753         video_json = video_json['playurl']
 754         formats = []
 755         for vid in video_json.get('video', []):
 756             video_res = vid.get('video_resource') or {}
 757             video_info = vid.get('stream_info') or {}
 758             if not video_res.get('url'):
 759                 continue
 760             formats.append({
 761                 'url': video_res['url'],
 762                 'ext': 'mp4',
 763                 'format_note': video_info.get('desc_words'),
 764                 'width': video_res.get('width'),
 765                 'height': video_res.get('height'),
 766                 'vbr': video_res.get('bandwidth'),
 767                 'acodec': 'none',
 768                 'vcodec': video_res.get('codecs'),
 769                 'filesize': video_res.get('size'),
 770             })
 771         for aud in video_json.get('audio_resource', []):
 772             if not aud.get('url'):
 773                 continue
 774             formats.append({
 775                 'url': aud['url'],
 776                 'ext': 'mp4',
 777                 'abr': aud.get('bandwidth'),
 778                 'acodec': aud.get('codecs'),
 779                 'vcodec': 'none',
 780                 'filesize': aud.get('size'),
 781             })
 782
 783         self._sort_formats(formats)
 784         return formats
 785
 786     def _extract_ep_info(self, type, episode_data, ep_id):
 787         return {
 788             'id': ep_id,
 789             'title': episode_data.get('long_title') or episode_data['title'],
 790             'thumbnail': episode_data.get('cover'),
 791             'episode_number': str_to_int(episode_data.get('title')),
 792             'formats': self._get_formats(type, ep_id),
 793             'subtitles': self._get_subtitles(type, ep_id),
 794             'extractor_key': BiliIntlIE.ie_key(),
 795         }
 796
 797
 798 class BiliIntlIE(BiliIntlBaseIE):
 799     _VALID_URL = r'https?://(?:www\.)?bili(?P<type>bili\.tv|intl.com)/(?:[a-z]{2}/)?play/(?P<season_id>\d+)/(?P<id>\d+)'
 800     _TESTS = [{
 801         'url': 'https://www.bilibili.tv/en/play/34613/341736',
 802         'info_dict': {
 803             'id': '341736',
 804             'ext': 'mp4',
 805             'title': 'The First Night',
 806             'thumbnail': 'https://i0.hdslb.com/bfs/intl/management/91e30e5521235d9b163339a26a0b030ebda54310.png',
 807             'episode_number': 2,
 808         },
 809         'params': {
 810             'format': 'bv',
 811         },
 812     }, {
 813         'url': 'https://www.biliintl.com/en/play/34613/341736',
 814         'info_dict': {
 815             'id': '341736',
 816             'ext': 'mp4',
 817             'title': 'The First Night',
 818             'thumbnail': 'https://i0.hdslb.com/bfs/intl/management/91e30e5521235d9b163339a26a0b030ebda54310.png',
 819             'episode_number': 2,
 820         },
 821         'params': {
 822             'format': 'bv',
 823         },
 824     }]
 825
 826     def _real_extract(self, url):
 827         type, season_id, id = self._match_valid_url(url).groups()
 828         data_json = self._call_api(type, f'/web/view/ogv_collection?season_id={season_id}', id)
 829         episode_data = next(
 830             episode for episode in data_json.get('episodes', [])
 831             if str(episode.get('ep_id')) == id)
 832         return self._extract_ep_info(type, episode_data, id)
 833
 834
 835 class BiliIntlSeriesIE(BiliIntlBaseIE):
 836     _VALID_URL = r'https?://(?:www\.)?bili(?P<type>bili\.tv|intl.com)/(?:[a-z]{2}/)?play/(?P<id>\d+)$'
 837     _TESTS = [{
 838         'url': 'https://www.bilibili.tv/en/play/34613',
 839         'playlist_mincount': 15,
 840         'info_dict': {
 841             'id': '34613',
 842         },
 843         'params': {
 844             'skip_download': True,
 845             'format': 'bv',
 846         },
 847     }, {
 848         'url': 'https://www.biliintl.com/en/play/34613',
 849         'playlist_mincount': 15,
 850         'info_dict': {
 851             'id': '34613',
 852         },
 853         'params': {
 854             'skip_download': True,
 855             'format': 'bv',
 856         },
 857     }]
 858
 859     def _entries(self, id, type):
 860         data_json = self._call_api(type, f'/web/view/ogv_collection?season_id={id}', id)
 861         for episode in data_json.get('episodes', []):
 862             episode_id = str(episode.get('ep_id'))
 863             yield self._extract_ep_info(type, episode, episode_id)
 864
 865     def _real_extract(self, url):
 866         type, id = self._match_valid_url(url).groups()
 867         return self.playlist_result(self._entries(id, type), playlist_id=id)