yt_dlp/extractor/bilibili.py

   1 # coding: utf-8
   2
   3 import base64
   4 import hashlib
   5 import itertools
   6 import functools
   7 import re
   8 import math
   9
  10 from .common import InfoExtractor, SearchInfoExtractor
  11 from ..compat import (
  12     compat_parse_qs,
  13     compat_urlparse,
  14     compat_urllib_parse_urlparse
  15 )
  16 from ..utils import (
  17     ExtractorError,
  18     int_or_none,
  19     float_or_none,
  20     mimetype2ext,
  21     parse_iso8601,
  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': '5f7d29e1a2872f3df0cf76b1f87d3788',
  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         },
  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_part1',
  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         'url': 'http://www.bilibili.com/video/av8903802/',
  84         'info_dict': {
  85             'id': '8903802_part1',
  86             'ext': 'mp4',
  87             'title': '阿滴英文｜英文歌分享#6 "Closer',
  88             'upload_date': '20170301',
  89             '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',
  90             'timestamp': 1488382634,
  91             'uploader_id': '65880958',
  92             'uploader': '阿滴英文',
  93         },
  94         'params': {
  95             'skip_download': True,
  96         },
  97     }, {
  98         # new BV video id format
  99         'url': 'https://www.bilibili.com/video/BV1JE411F741',
 100         'only_matching': True,
 101     }, {
 102         # Anthology
 103         'url': 'https://www.bilibili.com/video/BV1bK411W797',
 104         'info_dict': {
 105             'id': 'BV1bK411W797',
 106             'title': '物语中的人物是如何吐槽自己的OP的'
 107         },
 108         'playlist_count': 17,
 109     }]
 110
 111     _APP_KEY = 'iVGUTjsxvpLeuDCf'
 112     _BILIBILI_KEY = 'aHRmhWMLkdeMuILqORnYZocwMBpMEOdt'
 113
 114     def _report_error(self, result):
 115         if 'message' in result:
 116             raise ExtractorError('%s said: %s' % (self.IE_NAME, result['message']), expected=True)
 117         elif 'code' in result:
 118             raise ExtractorError('%s returns error %d' % (self.IE_NAME, result['code']), expected=True)
 119         else:
 120             raise ExtractorError('Can\'t extract Bangumi episode ID')
 121
 122     def _real_extract(self, url):
 123         url, smuggled_data = unsmuggle_url(url, {})
 124
 125         mobj = self._match_valid_url(url)
 126         video_id = mobj.group('id_bv') or mobj.group('id')
 127
 128         av_id, bv_id = self._get_video_id_set(video_id, mobj.group('id_bv') is not None)
 129         video_id = av_id
 130
 131         info = {}
 132         anime_id = mobj.group('anime_id')
 133         page_id = mobj.group('page')
 134         webpage = self._download_webpage(url, video_id)
 135
 136         # Bilibili anthologies are similar to playlists but all videos share the same video ID as the anthology itself.
 137         # If the video has no page argument, check to see if it's an anthology
 138         if page_id is None:
 139             if not self.get_param('noplaylist'):
 140                 r = self._extract_anthology_entries(bv_id, video_id, webpage)
 141                 if r is not None:
 142                     self.to_screen('Downloading anthology %s - add --no-playlist to just download video' % video_id)
 143                     return r
 144             else:
 145                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
 146
 147         if 'anime/' not in url:
 148             cid = self._search_regex(
 149                 r'\bcid(?:["\']:|=)(\d+),["\']page(?:["\']:|=)' + str(page_id), webpage, 'cid',
 150                 default=None
 151             ) or self._search_regex(
 152                 r'\bcid(?:["\']:|=)(\d+)', webpage, 'cid',
 153                 default=None
 154             ) or compat_parse_qs(self._search_regex(
 155                 [r'EmbedPlayer\([^)]+,\s*"([^"]+)"\)',
 156                  r'EmbedPlayer\([^)]+,\s*\\"([^"]+)\\"\)',
 157                  r'<iframe[^>]+src="https://secure\.bilibili\.com/secure,([^"]+)"'],
 158                 webpage, 'player parameters'))['cid'][0]
 159         else:
 160             if 'no_bangumi_tip' not in smuggled_data:
 161                 self.to_screen('Downloading episode %s. To download all videos in anime %s, re-run yt-dlp with %s' % (
 162                     video_id, anime_id, compat_urlparse.urljoin(url, '//bangumi.bilibili.com/anime/%s' % anime_id)))
 163             headers = {
 164                 'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8',
 165                 'Referer': url
 166             }
 167             headers.update(self.geo_verification_headers())
 168
 169             js = self._download_json(
 170                 'http://bangumi.bilibili.com/web_api/get_source', video_id,
 171                 data=urlencode_postdata({'episode_id': video_id}),
 172                 headers=headers)
 173             if 'result' not in js:
 174                 self._report_error(js)
 175             cid = js['result']['cid']
 176
 177         headers = {
 178             'Accept': 'application/json',
 179             'Referer': url
 180         }
 181         headers.update(self.geo_verification_headers())
 182
 183         video_info = self._parse_json(
 184             self._search_regex(r'window.__playinfo__\s*=\s*({.+?})</script>', webpage, 'video info', default=None) or '{}',
 185             video_id, fatal=False)
 186         video_info = video_info.get('data') or {}
 187
 188         durl = traverse_obj(video_info, ('dash', 'video'))
 189         audios = traverse_obj(video_info, ('dash', 'audio')) or []
 190         entries = []
 191
 192         RENDITIONS = ('qn=80&quality=80&type=', 'quality=2&type=mp4')
 193         for num, rendition in enumerate(RENDITIONS, start=1):
 194             payload = 'appkey=%s&cid=%s&otype=json&%s' % (self._APP_KEY, cid, rendition)
 195             sign = hashlib.md5((payload + self._BILIBILI_KEY).encode('utf-8')).hexdigest()
 196             if not video_info:
 197                 video_info = self._download_json(
 198                     'http://interface.bilibili.com/v2/playurl?%s&sign=%s' % (payload, sign),
 199                     video_id, note='Downloading video info page',
 200                     headers=headers, fatal=num == len(RENDITIONS))
 201                 if not video_info:
 202                     continue
 203
 204             if not durl and 'durl' not in video_info:
 205                 if num < len(RENDITIONS):
 206                     continue
 207                 self._report_error(video_info)
 208
 209             formats = []
 210             for idx, durl in enumerate(durl or video_info['durl']):
 211                 formats.append({
 212                     'url': durl.get('baseUrl') or durl.get('base_url') or durl.get('url'),
 213                     'ext': mimetype2ext(durl.get('mimeType') or durl.get('mime_type')),
 214                     'fps': int_or_none(durl.get('frameRate') or durl.get('frame_rate')),
 215                     'width': int_or_none(durl.get('width')),
 216                     'height': int_or_none(durl.get('height')),
 217                     'vcodec': durl.get('codecs'),
 218                     'acodec': 'none' if audios else None,
 219                     'tbr': float_or_none(durl.get('bandwidth'), scale=1000),
 220                     'filesize': int_or_none(durl.get('size')),
 221                 })
 222                 for backup_url in traverse_obj(durl, 'backup_url', expected_type=list) or []:
 223                     formats.append({
 224                         'url': backup_url,
 225                         'quality': -2 if 'hd.mp4' in backup_url else -3,
 226                     })
 227
 228             for audio in audios:
 229                 formats.append({
 230                     'url': audio.get('baseUrl') or audio.get('base_url') or audio.get('url'),
 231                     'ext': mimetype2ext(audio.get('mimeType') or audio.get('mime_type')),
 232                     'fps': int_or_none(audio.get('frameRate') or audio.get('frame_rate')),
 233                     'width': int_or_none(audio.get('width')),
 234                     'height': int_or_none(audio.get('height')),
 235                     'acodec': audio.get('codecs'),
 236                     'vcodec': 'none',
 237                     'tbr': float_or_none(audio.get('bandwidth'), scale=1000),
 238                     'filesize': int_or_none(audio.get('size'))
 239                 })
 240                 for backup_url in traverse_obj(audio, 'backup_url', expected_type=list) or []:
 241                     formats.append({
 242                         'url': backup_url,
 243                         # backup URLs have lower priorities
 244                         'quality': -3,
 245                     })
 246
 247             info.update({
 248                 'id': video_id,
 249                 'duration': float_or_none(durl.get('length'), 1000),
 250                 'formats': formats,
 251                 'http_headers': {
 252                     'Referer': url,
 253                 },
 254             })
 255             break
 256
 257         self._sort_formats(formats)
 258
 259         title = self._html_search_regex((
 260             r'<h1[^>]+title=(["\'])(?P<content>[^"\']+)',
 261             r'(?s)<h1[^>]*>(?P<content>.+?)</h1>',
 262             self._meta_regex('title')
 263         ), webpage, 'title', group='content', fatal=False)
 264
 265         # Get part title for anthologies
 266         if page_id is not None:
 267             # TODO: The json is already downloaded by _extract_anthology_entries. Don't redownload for each video.
 268             part_info = traverse_obj(self._download_json(
 269                 f'https://api.bilibili.com/x/player/pagelist?bvid={bv_id}&jsonp=jsonp',
 270                 video_id, note='Extracting videos in anthology'), 'data', expected_type=list)
 271             title = title if len(part_info) == 1 else traverse_obj(part_info, (int(page_id) - 1, 'part')) or title
 272
 273         description = self._html_search_meta('description', webpage)
 274         timestamp = unified_timestamp(self._html_search_regex(
 275             r'<time[^>]+datetime="([^"]+)"', webpage, 'upload time',
 276             default=None) or self._html_search_meta(
 277             'uploadDate', webpage, 'timestamp', default=None))
 278         thumbnail = self._html_search_meta(['og:image', 'thumbnailUrl'], webpage)
 279
 280         # TODO 'view_count' requires deobfuscating Javascript
 281         info.update({
 282             'id': f'{video_id}_part{page_id or 1}',
 283             'cid': cid,
 284             'title': title,
 285             'description': description,
 286             'timestamp': timestamp,
 287             'thumbnail': thumbnail,
 288             'duration': float_or_none(video_info.get('timelength'), scale=1000),
 289         })
 290
 291         uploader_mobj = re.search(
 292             r'<a[^>]+href="(?:https?:)?//space\.bilibili\.com/(?P<id>\d+)"[^>]*>\s*(?P<name>[^<]+?)\s*<',
 293             webpage)
 294         if uploader_mobj:
 295             info.update({
 296                 'uploader': uploader_mobj.group('name').strip(),
 297                 'uploader_id': uploader_mobj.group('id'),
 298             })
 299
 300         if not info.get('uploader'):
 301             info['uploader'] = self._html_search_meta(
 302                 'author', webpage, 'uploader', default=None)
 303
 304         top_level_info = {
 305             'tags': traverse_obj(self._download_json(
 306                 f'https://api.bilibili.com/x/tag/archive/tags?aid={video_id}',
 307                 video_id, fatal=False, note='Downloading tags'), ('data', ..., 'tag_name')),
 308         }
 309
 310         info['subtitles'] = {
 311             'danmaku': [{
 312                 'ext': 'xml',
 313                 'url': f'https://comment.bilibili.com/{cid}.xml',
 314             }]
 315         }
 316
 317         r'''
 318         # Requires https://github.com/m13253/danmaku2ass which is licenced under GPL3
 319         # See https://github.com/animelover1984/youtube-dl
 320
 321         raw_danmaku = self._download_webpage(
 322             f'https://comment.bilibili.com/{cid}.xml', video_id, fatal=False, note='Downloading danmaku comments')
 323         danmaku = NiconicoIE.CreateDanmaku(raw_danmaku, commentType='Bilibili', x=1024, y=576)
 324         entries[0]['subtitles'] = {
 325             'danmaku': [{
 326                 'ext': 'ass',
 327                 'data': danmaku
 328             }]
 329         }
 330         '''
 331
 332         top_level_info['__post_extractor'] = self.extract_comments(video_id)
 333
 334         for entry in entries:
 335             entry.update(info)
 336
 337         if len(entries) == 1:
 338             entries[0].update(top_level_info)
 339             return entries[0]
 340
 341         for idx, entry in enumerate(entries):
 342             entry['id'] = '%s_part%d' % (video_id, (idx + 1))
 343
 344         return {
 345             'id': str(video_id),
 346             'bv_id': bv_id,
 347             'title': title,
 348             'description': description,
 349             **info, **top_level_info
 350         }
 351
 352     def _extract_anthology_entries(self, bv_id, video_id, webpage):
 353         title = self._html_search_regex(
 354             (r'<h1[^>]+\btitle=(["\'])(?P<title>(?:(?!\1).)+)\1',
 355              r'(?s)<h1[^>]*>(?P<title>.+?)</h1>',
 356              r'<title>(?P<title>.+?)</title>'), webpage, 'title',
 357             group='title')
 358         json_data = self._download_json(
 359             f'https://api.bilibili.com/x/player/pagelist?bvid={bv_id}&jsonp=jsonp',
 360             video_id, note='Extracting videos in anthology')
 361
 362         if json_data['data']:
 363             return self.playlist_from_matches(
 364                 json_data['data'], bv_id, title, ie=BiliBiliIE.ie_key(),
 365                 getter=lambda entry: 'https://www.bilibili.com/video/%s?p=%d' % (bv_id, entry['page']))
 366
 367     def _get_video_id_set(self, id, is_bv):
 368         query = {'bvid': id} if is_bv else {'aid': id}
 369         response = self._download_json(
 370             "http://api.bilibili.cn/x/web-interface/view",
 371             id, query=query,
 372             note='Grabbing original ID via API')
 373
 374         if response['code'] == -400:
 375             raise ExtractorError('Video ID does not exist', expected=True, video_id=id)
 376         elif response['code'] != 0:
 377             raise ExtractorError(f'Unknown error occurred during API check (code {response["code"]})',
 378                                  expected=True, video_id=id)
 379         return response['data']['aid'], response['data']['bvid']
 380
 381     def _get_comments(self, video_id, commentPageNumber=0):
 382         for idx in itertools.count(1):
 383             replies = traverse_obj(
 384                 self._download_json(
 385                     f'https://api.bilibili.com/x/v2/reply?pn={idx}&oid={video_id}&type=1&jsonp=jsonp&sort=2&_=1567227301685',
 386                     video_id, note=f'Extracting comments from page {idx}', fatal=False),
 387                 ('data', 'replies'))
 388             if not replies:
 389                 return
 390             for children in map(self._get_all_children, replies):
 391                 yield from children
 392
 393     def _get_all_children(self, reply):
 394         yield {
 395             'author': traverse_obj(reply, ('member', 'uname')),
 396             'author_id': traverse_obj(reply, ('member', 'mid')),
 397             'id': reply.get('rpid'),
 398             'text': traverse_obj(reply, ('content', 'message')),
 399             'timestamp': reply.get('ctime'),
 400             'parent': reply.get('parent') or 'root',
 401         }
 402         for children in map(self._get_all_children, reply.get('replies') or []):
 403             yield from children
 404
 405
 406 class BiliBiliBangumiIE(InfoExtractor):
 407     _VALID_URL = r'https?://bangumi\.bilibili\.com/anime/(?P<id>\d+)'
 408
 409     IE_NAME = 'bangumi.bilibili.com'
 410     IE_DESC = 'BiliBili番剧'
 411
 412     _TESTS = [{
 413         'url': 'http://bangumi.bilibili.com/anime/1869',
 414         'info_dict': {
 415             'id': '1869',
 416             'title': '混沌武士',
 417             'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
 418         },
 419         'playlist_count': 26,
 420     }, {
 421         'url': 'http://bangumi.bilibili.com/anime/1869',
 422         'info_dict': {
 423             'id': '1869',
 424             'title': '混沌武士',
 425             'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
 426         },
 427         'playlist': [{
 428             'md5': '91da8621454dd58316851c27c68b0c13',
 429             'info_dict': {
 430                 'id': '40062',
 431                 'ext': 'mp4',
 432                 'title': '混沌武士',
 433                 '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...',
 434                 'timestamp': 1414538739,
 435                 'upload_date': '20141028',
 436                 'episode': '疾风怒涛 Tempestuous Temperaments',
 437                 'episode_number': 1,
 438             },
 439         }],
 440         'params': {
 441             'playlist_items': '1',
 442         },
 443     }]
 444
 445     @classmethod
 446     def suitable(cls, url):
 447         return False if BiliBiliIE.suitable(url) else super(BiliBiliBangumiIE, cls).suitable(url)
 448
 449     def _real_extract(self, url):
 450         bangumi_id = self._match_id(url)
 451
 452         # Sometimes this API returns a JSONP response
 453         season_info = self._download_json(
 454             'http://bangumi.bilibili.com/jsonp/seasoninfo/%s.ver' % bangumi_id,
 455             bangumi_id, transform_source=strip_jsonp)['result']
 456
 457         entries = [{
 458             '_type': 'url_transparent',
 459             'url': smuggle_url(episode['webplay_url'], {'no_bangumi_tip': 1}),
 460             'ie_key': BiliBiliIE.ie_key(),
 461             'timestamp': parse_iso8601(episode.get('update_time'), delimiter=' '),
 462             'episode': episode.get('index_title'),
 463             'episode_number': int_or_none(episode.get('index')),
 464         } for episode in season_info['episodes']]
 465
 466         entries = sorted(entries, key=lambda entry: entry.get('episode_number'))
 467
 468         return self.playlist_result(
 469             entries, bangumi_id,
 470             season_info.get('bangumi_title'), season_info.get('evaluate'))
 471
 472
 473 class BilibiliChannelIE(InfoExtractor):
 474     _VALID_URL = r'https?://space.bilibili\.com/(?P<id>\d+)'
 475     _API_URL = "https://api.bilibili.com/x/space/arc/search?mid=%s&pn=%d&jsonp=jsonp"
 476     _TESTS = [{
 477         'url': 'https://space.bilibili.com/3985676/video',
 478         'info_dict': {},
 479         'playlist_mincount': 112,
 480     }]
 481
 482     def _entries(self, list_id):
 483         count, max_count = 0, None
 484
 485         for page_num in itertools.count(1):
 486             data = self._download_json(
 487                 self._API_URL % (list_id, page_num), list_id, note=f'Downloading page {page_num}')['data']
 488
 489             max_count = max_count or traverse_obj(data, ('page', 'count'))
 490
 491             entries = traverse_obj(data, ('list', 'vlist'))
 492             if not entries:
 493                 return
 494             for entry in entries:
 495                 yield self.url_result(
 496                     'https://www.bilibili.com/video/%s' % entry['bvid'],
 497                     BiliBiliIE.ie_key(), entry['bvid'])
 498
 499             count += len(entries)
 500             if max_count and count >= max_count:
 501                 return
 502
 503     def _real_extract(self, url):
 504         list_id = self._match_id(url)
 505         return self.playlist_result(self._entries(list_id), list_id)
 506
 507
 508 class BilibiliCategoryIE(InfoExtractor):
 509     IE_NAME = 'Bilibili category extractor'
 510     _MAX_RESULTS = 1000000
 511     _VALID_URL = r'https?://www\.bilibili\.com/v/[a-zA-Z]+\/[a-zA-Z]+'
 512     _TESTS = [{
 513         'url': 'https://www.bilibili.com/v/kichiku/mad',
 514         'info_dict': {
 515             'id': 'kichiku: mad',
 516             'title': 'kichiku: mad'
 517         },
 518         'playlist_mincount': 45,
 519         'params': {
 520             'playlistend': 45
 521         }
 522     }]
 523
 524     def _fetch_page(self, api_url, num_pages, query, page_num):
 525         parsed_json = self._download_json(
 526             api_url, query, query={'Search_key': query, 'pn': page_num},
 527             note='Extracting results from page %s of %s' % (page_num, num_pages))
 528
 529         video_list = traverse_obj(parsed_json, ('data', 'archives'), expected_type=list)
 530         if not video_list:
 531             raise ExtractorError('Failed to retrieve video list for page %d' % page_num)
 532
 533         for video in video_list:
 534             yield self.url_result(
 535                 'https://www.bilibili.com/video/%s' % video['bvid'], 'BiliBili', video['bvid'])
 536
 537     def _entries(self, category, subcategory, query):
 538         # map of categories : subcategories : RIDs
 539         rid_map = {
 540             'kichiku': {
 541                 'mad': 26,
 542                 'manual_vocaloid': 126,
 543                 'guide': 22,
 544                 'theatre': 216,
 545                 'course': 127
 546             },
 547         }
 548
 549         if category not in rid_map:
 550             raise ExtractorError(
 551                 f'The category {category} isn\'t supported. Supported categories: {list(rid_map.keys())}')
 552         if subcategory not in rid_map[category]:
 553             raise ExtractorError(
 554                 f'The subcategory {subcategory} isn\'t supported for this category. Supported subcategories: {list(rid_map[category].keys())}')
 555         rid_value = rid_map[category][subcategory]
 556
 557         api_url = 'https://api.bilibili.com/x/web-interface/newlist?rid=%d&type=1&ps=20&jsonp=jsonp' % rid_value
 558         page_json = self._download_json(api_url, query, query={'Search_key': query, 'pn': '1'})
 559         page_data = traverse_obj(page_json, ('data', 'page'), expected_type=dict)
 560         count, size = int_or_none(page_data.get('count')), int_or_none(page_data.get('size'))
 561         if count is None or not size:
 562             raise ExtractorError('Failed to calculate either page count or size')
 563
 564         num_pages = math.ceil(count / size)
 565
 566         return OnDemandPagedList(functools.partial(
 567             self._fetch_page, api_url, num_pages, query), size)
 568
 569     def _real_extract(self, url):
 570         u = compat_urllib_parse_urlparse(url)
 571         category, subcategory = u.path.split('/')[2:4]
 572         query = '%s: %s' % (category, subcategory)
 573
 574         return self.playlist_result(self._entries(category, subcategory, query), query, query)
 575
 576
 577 class BiliBiliSearchIE(SearchInfoExtractor):
 578     IE_DESC = 'Bilibili video search'
 579     _MAX_RESULTS = 100000
 580     _SEARCH_KEY = 'bilisearch'
 581
 582     def _search_results(self, query):
 583         for page_num in itertools.count(1):
 584             videos = self._download_json(
 585                 'https://api.bilibili.com/x/web-interface/search/type', query,
 586                 note=f'Extracting results from page {page_num}', query={
 587                     'Search_key': query,
 588                     'keyword': query,
 589                     'page': page_num,
 590                     'context': '',
 591                     'order': 'pubdate',
 592                     'duration': 0,
 593                     'tids_2': '',
 594                     '__refresh__': 'true',
 595                     'search_type': 'video',
 596                     'tids': 0,
 597                     'highlight': 1,
 598                 })['data'].get('result') or []
 599             for video in videos:
 600                 yield self.url_result(video['arcurl'], 'BiliBili', str(video['aid']))
 601
 602
 603 class BilibiliAudioBaseIE(InfoExtractor):
 604     def _call_api(self, path, sid, query=None):
 605         if not query:
 606             query = {'sid': sid}
 607         return self._download_json(
 608             'https://www.bilibili.com/audio/music-service-c/web/' + path,
 609             sid, query=query)['data']
 610
 611
 612 class BilibiliAudioIE(BilibiliAudioBaseIE):
 613     _VALID_URL = r'https?://(?:www\.)?bilibili\.com/audio/au(?P<id>\d+)'
 614     _TEST = {
 615         'url': 'https://www.bilibili.com/audio/au1003142',
 616         'md5': 'fec4987014ec94ef9e666d4d158ad03b',
 617         'info_dict': {
 618             'id': '1003142',
 619             'ext': 'm4a',
 620             'title': '【tsukimi】YELLOW / 神山羊',
 621             'artist': 'tsukimi',
 622             'comment_count': int,
 623             'description': 'YELLOW的mp3版！',
 624             'duration': 183,
 625             'subtitles': {
 626                 'origin': [{
 627                     'ext': 'lrc',
 628                 }],
 629             },
 630             'thumbnail': r're:^https?://.+\.jpg',
 631             'timestamp': 1564836614,
 632             'upload_date': '20190803',
 633             'uploader': 'tsukimi-つきみぐー',
 634             'view_count': int,
 635         },
 636     }
 637
 638     def _real_extract(self, url):
 639         au_id = self._match_id(url)
 640
 641         play_data = self._call_api('url', au_id)
 642         formats = [{
 643             'url': play_data['cdns'][0],
 644             'filesize': int_or_none(play_data.get('size')),
 645             'vcodec': 'none'
 646         }]
 647
 648         song = self._call_api('song/info', au_id)
 649         title = song['title']
 650         statistic = song.get('statistic') or {}
 651
 652         subtitles = None
 653         lyric = song.get('lyric')
 654         if lyric:
 655             subtitles = {
 656                 'origin': [{
 657                     'url': lyric,
 658                 }]
 659             }
 660
 661         return {
 662             'id': au_id,
 663             'title': title,
 664             'formats': formats,
 665             'artist': song.get('author'),
 666             'comment_count': int_or_none(statistic.get('comment')),
 667             'description': song.get('intro'),
 668             'duration': int_or_none(song.get('duration')),
 669             'subtitles': subtitles,
 670             'thumbnail': song.get('cover'),
 671             'timestamp': int_or_none(song.get('passtime')),
 672             'uploader': song.get('uname'),
 673             'view_count': int_or_none(statistic.get('play')),
 674         }
 675
 676
 677 class BilibiliAudioAlbumIE(BilibiliAudioBaseIE):
 678     _VALID_URL = r'https?://(?:www\.)?bilibili\.com/audio/am(?P<id>\d+)'
 679     _TEST = {
 680         'url': 'https://www.bilibili.com/audio/am10624',
 681         'info_dict': {
 682             'id': '10624',
 683             'title': '每日新曲推荐（每日11:00更新）',
 684             'description': '每天11:00更新，为你推送最新音乐',
 685         },
 686         'playlist_count': 19,
 687     }
 688
 689     def _real_extract(self, url):
 690         am_id = self._match_id(url)
 691
 692         songs = self._call_api(
 693             'song/of-menu', am_id, {'sid': am_id, 'pn': 1, 'ps': 100})['data']
 694
 695         entries = []
 696         for song in songs:
 697             sid = str_or_none(song.get('id'))
 698             if not sid:
 699                 continue
 700             entries.append(self.url_result(
 701                 'https://www.bilibili.com/audio/au' + sid,
 702                 BilibiliAudioIE.ie_key(), sid))
 703
 704         if entries:
 705             album_data = self._call_api('menu/info', am_id) or {}
 706             album_title = album_data.get('title')
 707             if album_title:
 708                 for entry in entries:
 709                     entry['album'] = album_title
 710                 return self.playlist_result(
 711                     entries, am_id, album_title, album_data.get('intro'))
 712
 713         return self.playlist_result(entries, am_id)
 714
 715
 716 class BiliBiliPlayerIE(InfoExtractor):
 717     _VALID_URL = r'https?://player\.bilibili\.com/player\.html\?.*?\baid=(?P<id>\d+)'
 718     _TEST = {
 719         'url': 'http://player.bilibili.com/player.html?aid=92494333&cid=157926707&page=1',
 720         'only_matching': True,
 721     }
 722
 723     def _real_extract(self, url):
 724         video_id = self._match_id(url)
 725         return self.url_result(
 726             'http://www.bilibili.tv/video/av%s/' % video_id,
 727             ie=BiliBiliIE.ie_key(), video_id=video_id)
 728
 729
 730 class BiliIntlBaseIE(InfoExtractor):
 731     _API_URL = 'https://api.bilibili.tv/intl/gateway'
 732     _NETRC_MACHINE = 'biliintl'
 733
 734     def _call_api(self, endpoint, *args, **kwargs):
 735         json = self._download_json(self._API_URL + endpoint, *args, **kwargs)
 736         if json.get('code'):
 737             if json['code'] in (10004004, 10004005, 10023006):
 738                 self.raise_login_required()
 739             elif json['code'] == 10004001:
 740                 self.raise_geo_restricted()
 741             else:
 742                 if json.get('message') and str(json['code']) != json['message']:
 743                     errmsg = f'{kwargs.get("errnote", "Unable to download JSON metadata")}: {self.IE_NAME} said: {json["message"]}'
 744                 else:
 745                     errmsg = kwargs.get('errnote', 'Unable to download JSON metadata')
 746                 if kwargs.get('fatal'):
 747                     raise ExtractorError(errmsg)
 748                 else:
 749                     self.report_warning(errmsg)
 750         return json.get('data')
 751
 752     def json2srt(self, json):
 753         data = '\n\n'.join(
 754             f'{i + 1}\n{srt_subtitles_timecode(line["from"])} --> {srt_subtitles_timecode(line["to"])}\n{line["content"]}'
 755             for i, line in enumerate(json['body']) if line.get('content'))
 756         return data
 757
 758     def _get_subtitles(self, ep_id):
 759         sub_json = self._call_api(f'/web/v2/subtitle?episode_id={ep_id}&platform=web', ep_id)
 760         subtitles = {}
 761         for sub in sub_json.get('subtitles') or []:
 762             sub_url = sub.get('url')
 763             if not sub_url:
 764                 continue
 765             sub_data = self._download_json(
 766                 sub_url, ep_id, errnote='Unable to download subtitles', fatal=False,
 767                 note='Downloading subtitles%s' % f' for {sub["lang"]}' if sub.get('lang') else '')
 768             if not sub_data:
 769                 continue
 770             subtitles.setdefault(sub.get('lang_key', 'en'), []).append({
 771                 'ext': 'srt',
 772                 'data': self.json2srt(sub_data)
 773             })
 774         return subtitles
 775
 776     def _get_formats(self, ep_id):
 777         video_json = self._call_api(f'/web/playurl?ep_id={ep_id}&platform=web', ep_id,
 778                                     note='Downloading video formats', errnote='Unable to download video formats')
 779         video_json = video_json['playurl']
 780         formats = []
 781         for vid in video_json.get('video') or []:
 782             video_res = vid.get('video_resource') or {}
 783             video_info = vid.get('stream_info') or {}
 784             if not video_res.get('url'):
 785                 continue
 786             formats.append({
 787                 'url': video_res['url'],
 788                 'ext': 'mp4',
 789                 'format_note': video_info.get('desc_words'),
 790                 'width': video_res.get('width'),
 791                 'height': video_res.get('height'),
 792                 'vbr': video_res.get('bandwidth'),
 793                 'acodec': 'none',
 794                 'vcodec': video_res.get('codecs'),
 795                 'filesize': video_res.get('size'),
 796             })
 797         for aud in video_json.get('audio_resource') or []:
 798             if not aud.get('url'):
 799                 continue
 800             formats.append({
 801                 'url': aud['url'],
 802                 'ext': 'mp4',
 803                 'abr': aud.get('bandwidth'),
 804                 'acodec': aud.get('codecs'),
 805                 'vcodec': 'none',
 806                 'filesize': aud.get('size'),
 807             })
 808
 809         self._sort_formats(formats)
 810         return formats
 811
 812     def _extract_ep_info(self, episode_data, ep_id):
 813         return {
 814             'id': ep_id,
 815             'title': episode_data.get('title_display') or episode_data['title'],
 816             'thumbnail': episode_data.get('cover'),
 817             'episode_number': int_or_none(self._search_regex(
 818                 r'^E(\d+)(?:$| - )', episode_data.get('title_display'), 'episode number', default=None)),
 819             'formats': self._get_formats(ep_id),
 820             'subtitles': self._get_subtitles(ep_id),
 821             'extractor_key': BiliIntlIE.ie_key(),
 822         }
 823
 824     def _perform_login(self, username, password):
 825         try:
 826             from Cryptodome.PublicKey import RSA
 827             from Cryptodome.Cipher import PKCS1_v1_5
 828         except ImportError:
 829             try:
 830                 from Crypto.PublicKey import RSA
 831                 from Crypto.Cipher import PKCS1_v1_5
 832             except ImportError:
 833                 raise ExtractorError('pycryptodomex not found. Please install', expected=True)
 834
 835         key_data = self._download_json(
 836             'https://passport.bilibili.tv/x/intl/passport-login/web/key?lang=en-US', None,
 837             note='Downloading login key', errnote='Unable to download login key')['data']
 838
 839         public_key = RSA.importKey(key_data['key'])
 840         password_hash = PKCS1_v1_5.new(public_key).encrypt((key_data['hash'] + password).encode('utf-8'))
 841         login_post = self._download_json(
 842             'https://passport.bilibili.tv/x/intl/passport-login/web/login/password?lang=en-US', None, data=urlencode_postdata({
 843                 'username': username,
 844                 'password': base64.b64encode(password_hash).decode('ascii'),
 845                 'keep_me': 'true',
 846                 's_locale': 'en_US',
 847                 'isTrusted': 'true'
 848             }), note='Logging in', errnote='Unable to log in')
 849         if login_post.get('code'):
 850             if login_post.get('message'):
 851                 raise ExtractorError(f'Unable to log in: {self.IE_NAME} said: {login_post["message"]}', expected=True)
 852             else:
 853                 raise ExtractorError('Unable to log in')
 854
 855
 856 class BiliIntlIE(BiliIntlBaseIE):
 857     _VALID_URL = r'https?://(?:www\.)?bili(?:bili\.tv|intl\.com)/(?:[a-z]{2}/)?play/(?P<season_id>\d+)/(?P<id>\d+)'
 858     _TESTS = [{
 859         # Bstation page
 860         'url': 'https://www.bilibili.tv/en/play/34613/341736',
 861         'info_dict': {
 862             'id': '341736',
 863             'ext': 'mp4',
 864             'title': 'E2 - The First Night',
 865             'thumbnail': r're:^https://pic\.bstarstatic\.com/ogv/.+\.png$',
 866             'episode_number': 2,
 867         }
 868     }, {
 869         # Non-Bstation page
 870         'url': 'https://www.bilibili.tv/en/play/1033760/11005006',
 871         'info_dict': {
 872             'id': '11005006',
 873             'ext': 'mp4',
 874             'title': 'E3 - Who?',
 875             'thumbnail': r're:^https://pic\.bstarstatic\.com/ogv/.+\.png$',
 876             'episode_number': 3,
 877         }
 878     }, {
 879         # Subtitle with empty content
 880         'url': 'https://www.bilibili.tv/en/play/1005144/10131790',
 881         'info_dict': {
 882             'id': '10131790',
 883             'ext': 'mp4',
 884             'title': 'E140 - Two Heartbeats: Kabuto\'s Trap',
 885             'thumbnail': r're:^https://pic\.bstarstatic\.com/ogv/.+\.png$',
 886             'episode_number': 140,
 887         },
 888         'skip': 'According to the copyright owner\'s request, you may only watch the video after you log in.'
 889     }, {
 890         'url': 'https://www.biliintl.com/en/play/34613/341736',
 891         'only_matching': True,
 892     }]
 893
 894     def _real_extract(self, url):
 895         season_id, video_id = self._match_valid_url(url).groups()
 896         webpage = self._download_webpage(url, video_id)
 897         # Bstation layout
 898         initial_data = self._parse_json(self._search_regex(
 899             r'window\.__INITIAL_DATA__\s*=\s*({.+?});', webpage,
 900             'preload state', default='{}'), video_id, fatal=False) or {}
 901         episode_data = traverse_obj(initial_data, ('OgvVideo', 'epDetail'), expected_type=dict)
 902
 903         if not episode_data:
 904             # Non-Bstation layout, read through episode list
 905             season_json = self._call_api(f'/web/v2/ogv/play/episodes?season_id={season_id}&platform=web', video_id)
 906             episode_data = next(
 907                 episode for episode in traverse_obj(season_json, ('sections', ..., 'episodes', ...), expected_type=dict)
 908                 if str(episode.get('episode_id')) == video_id)
 909         return self._extract_ep_info(episode_data, video_id)
 910
 911
 912 class BiliIntlSeriesIE(BiliIntlBaseIE):
 913     _VALID_URL = r'https?://(?:www\.)?bili(?:bili\.tv|intl\.com)/(?:[a-z]{2}/)?play/(?P<id>\d+)$'
 914     _TESTS = [{
 915         'url': 'https://www.bilibili.tv/en/play/34613',
 916         'playlist_mincount': 15,
 917         'info_dict': {
 918             'id': '34613',
 919             'title': 'Fly Me to the Moon',
 920             'description': 'md5:a861ee1c4dc0acfad85f557cc42ac627',
 921             'categories': ['Romance', 'Comedy', 'Slice of life'],
 922             'thumbnail': r're:^https://pic\.bstarstatic\.com/ogv/.+\.png$',
 923             'view_count': int,
 924         },
 925         'params': {
 926             'skip_download': True,
 927         },
 928     }, {
 929         'url': 'https://www.biliintl.com/en/play/34613',
 930         'only_matching': True,
 931     }]
 932
 933     def _entries(self, series_id):
 934         series_json = self._call_api(f'/web/v2/ogv/play/episodes?season_id={series_id}&platform=web', series_id)
 935         for episode in traverse_obj(series_json, ('sections', ..., 'episodes', ...), expected_type=dict, default=[]):
 936             episode_id = str(episode.get('episode_id'))
 937             yield self._extract_ep_info(episode, episode_id)
 938
 939     def _real_extract(self, url):
 940         series_id = self._match_id(url)
 941         series_info = self._call_api(f'/web/v2/ogv/play/season_info?season_id={series_id}&platform=web', series_id).get('season') or {}
 942         return self.playlist_result(
 943             self._entries(series_id), series_id, series_info.get('title'), series_info.get('description'),
 944             categories=traverse_obj(series_info, ('styles', ..., 'title'), expected_type=str_or_none),
 945             thumbnail=url_or_none(series_info.get('horizontal_cover')), view_count=parse_count(series_info.get('view')))