yt_dlp/extractor/tiktok.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import itertools
   5 import random
   6 import string
   7 import time
   8 import json
   9
  10 from .common import InfoExtractor
  11 from ..compat import compat_urllib_parse_unquote
  12 from ..utils import (
  13     ExtractorError,
  14     int_or_none,
  15     join_nonempty,
  16     str_or_none,
  17     traverse_obj,
  18     try_get,
  19     url_or_none,
  20     qualities,
  21 )
  22
  23
  24 class TikTokBaseIE(InfoExtractor):
  25     _APP_VERSION = '20.9.3'
  26     _MANIFEST_APP_VERSION = '291'
  27     _APP_NAME = 'trill'
  28     _AID = 1180
  29     _API_HOSTNAME = 'api-t2.tiktokv.com'
  30     _UPLOADER_URL_FORMAT = 'https://www.tiktok.com/@%s'
  31     QUALITIES = ('360p', '540p', '720p')
  32
  33     def _call_api(self, ep, query, video_id, fatal=True,
  34                   note='Downloading API JSON', errnote='Unable to download API page'):
  35         real_query = {
  36             **query,
  37             'version_name': self._APP_VERSION,
  38             'version_code': self._MANIFEST_APP_VERSION,
  39             'build_number': self._APP_VERSION,
  40             'manifest_version_code': self._MANIFEST_APP_VERSION,
  41             'update_version_code': self._MANIFEST_APP_VERSION,
  42             'openudid': ''.join(random.choice('0123456789abcdef') for _ in range(16)),
  43             'uuid': ''.join([random.choice(string.digits) for _ in range(16)]),
  44             '_rticket': int(time.time() * 1000),
  45             'ts': int(time.time()),
  46             'device_brand': 'Google',
  47             'device_type': 'Pixel 4',
  48             'device_platform': 'android',
  49             'resolution': '1080*1920',
  50             'dpi': 420,
  51             'os_version': '10',
  52             'os_api': '29',
  53             'carrier_region': 'US',
  54             'sys_region': 'US',
  55             'region': 'US',
  56             'app_name': self._APP_NAME,
  57             'app_language': 'en',
  58             'language': 'en',
  59             'timezone_name': 'America/New_York',
  60             'timezone_offset': '-14400',
  61             'channel': 'googleplay',
  62             'ac': 'wifi',
  63             'mcc_mnc': '310260',
  64             'is_my_cn': 0,
  65             'aid': self._AID,
  66             'ssmix': 'a',
  67             'as': 'a1qwert123',
  68             'cp': 'cbfhckdckkde1',
  69         }
  70         self._set_cookie(self._API_HOSTNAME, 'odin_tt', ''.join(random.choice('0123456789abcdef') for _ in range(160)))
  71         return self._download_json(
  72             'https://%s/aweme/v1/%s/' % (self._API_HOSTNAME, ep), video_id=video_id,
  73             fatal=fatal, note=note, errnote=errnote, headers={
  74                 'User-Agent': f'com.ss.android.ugc.trill/{self._MANIFEST_APP_VERSION} (Linux; U; Android 10; en_US; Pixel 4; Build/QQ3A.200805.001; Cronet/58.0.2991.0)',
  75                 'Accept': 'application/json',
  76             }, query=real_query)
  77
  78     def _parse_aweme_video_app(self, aweme_detail):
  79         aweme_id = aweme_detail['aweme_id']
  80         video_info = aweme_detail['video']
  81
  82         def parse_url_key(url_key):
  83             format_id, codec, res, bitrate = self._search_regex(
  84                 r'v[^_]+_(?P<id>(?P<codec>[^_]+)_(?P<res>\d+p)_(?P<bitrate>\d+))', url_key,
  85                 'url key', default=(None, None, None, None), group=('id', 'codec', 'res', 'bitrate'))
  86             if not format_id:
  87                 return {}, None
  88             return {
  89                 'format_id': format_id,
  90                 'vcodec': 'h265' if codec == 'bytevc1' else codec,
  91                 'tbr': int_or_none(bitrate, scale=1000) or None,
  92                 'quality': qualities(self.QUALITIES)(res),
  93             }, res
  94
  95         known_resolutions = {}
  96
  97         def extract_addr(addr, add_meta={}):
  98             parsed_meta, res = parse_url_key(addr.get('url_key', ''))
  99             if res:
 100                 known_resolutions.setdefault(res, {}).setdefault('height', add_meta.get('height'))
 101                 known_resolutions[res].setdefault('width', add_meta.get('width'))
 102                 parsed_meta.update(known_resolutions.get(res, {}))
 103                 add_meta.setdefault('height', int_or_none(res[:-1]))
 104             return [{
 105                 'url': url,
 106                 'filesize': int_or_none(addr.get('data_size')),
 107                 'ext': 'mp4',
 108                 'acodec': 'aac',
 109                 'source_preference': -2 if 'aweme/v1' in url else -1,  # Downloads from API might get blocked
 110                 **add_meta, **parsed_meta,
 111                 'format_note': join_nonempty(
 112                     add_meta.get('format_note'), '(API)' if 'aweme/v1' in url else None, delim=' ')
 113             } for url in addr.get('url_list') or []]
 114
 115         # Hack: Add direct video links first to prioritize them when removing duplicate formats
 116         formats = []
 117         if video_info.get('play_addr'):
 118             formats.extend(extract_addr(video_info['play_addr'], {
 119                 'format_id': 'play_addr',
 120                 'format_note': 'Direct video',
 121                 'vcodec': 'h265' if traverse_obj(
 122                     video_info, 'is_bytevc1', 'is_h265') else 'h264',  # Always h264?
 123                 'width': video_info.get('width'),
 124                 'height': video_info.get('height'),
 125             }))
 126         if video_info.get('download_addr'):
 127             formats.extend(extract_addr(video_info['download_addr'], {
 128                 'format_id': 'download_addr',
 129                 'format_note': 'Download video%s' % (', watermarked' if video_info.get('has_watermark') else ''),
 130                 'vcodec': 'h264',
 131                 'width': video_info.get('width'),
 132                 'height': video_info.get('height'),
 133                 'preference': -2 if video_info.get('has_watermark') else -1,
 134             }))
 135         if video_info.get('play_addr_h264'):
 136             formats.extend(extract_addr(video_info['play_addr_h264'], {
 137                 'format_id': 'play_addr_h264',
 138                 'format_note': 'Direct video',
 139                 'vcodec': 'h264',
 140             }))
 141         if video_info.get('play_addr_bytevc1'):
 142             formats.extend(extract_addr(video_info['play_addr_bytevc1'], {
 143                 'format_id': 'play_addr_bytevc1',
 144                 'format_note': 'Direct video',
 145                 'vcodec': 'h265',
 146             }))
 147
 148         for bitrate in video_info.get('bit_rate', []):
 149             if bitrate.get('play_addr'):
 150                 formats.extend(extract_addr(bitrate['play_addr'], {
 151                     'format_id': bitrate.get('gear_name'),
 152                     'format_note': 'Playback video',
 153                     'tbr': try_get(bitrate, lambda x: x['bit_rate'] / 1000),
 154                     'vcodec': 'h265' if traverse_obj(
 155                         bitrate, 'is_bytevc1', 'is_h265') else 'h264',
 156                     'fps': bitrate.get('FPS'),
 157                 }))
 158
 159         self._remove_duplicate_formats(formats)
 160         self._sort_formats(formats, ('quality', 'codec', 'size', 'br'))
 161
 162         thumbnails = []
 163         for cover_id in ('cover', 'ai_dynamic_cover', 'animated_cover', 'ai_dynamic_cover_bak',
 164                          'origin_cover', 'dynamic_cover'):
 165             cover = video_info.get(cover_id)
 166             if cover:
 167                 for cover_url in cover['url_list']:
 168                     thumbnails.append({
 169                         'id': cover_id,
 170                         'url': cover_url,
 171                     })
 172
 173         stats_info = aweme_detail.get('statistics', {})
 174         author_info = aweme_detail.get('author', {})
 175         music_info = aweme_detail.get('music', {})
 176         user_url = self._UPLOADER_URL_FORMAT % (traverse_obj(author_info,
 177                                                              'sec_uid', 'id', 'uid', 'unique_id',
 178                                                              expected_type=str_or_none, get_all=False))
 179
 180         contained_music_track = traverse_obj(
 181             music_info, ('matched_song', 'title'), ('matched_pgc_sound', 'title'), expected_type=str)
 182         contained_music_author = traverse_obj(
 183             music_info, ('matched_song', 'author'), ('matched_pgc_sound', 'author'), 'author', expected_type=str)
 184
 185         is_generic_og_trackname = music_info.get('is_original_sound') and music_info.get('title') == 'original sound - %s' % music_info.get('owner_handle')
 186         if is_generic_og_trackname:
 187             music_track, music_author = contained_music_track or 'original sound', contained_music_author
 188         else:
 189             music_track, music_author = music_info.get('title'), music_info.get('author')
 190
 191         return {
 192             'id': aweme_id,
 193             'title': aweme_detail['desc'],
 194             'description': aweme_detail['desc'],
 195             'view_count': int_or_none(stats_info.get('play_count')),
 196             'like_count': int_or_none(stats_info.get('digg_count')),
 197             'repost_count': int_or_none(stats_info.get('share_count')),
 198             'comment_count': int_or_none(stats_info.get('comment_count')),
 199             'uploader': str_or_none(author_info.get('unique_id')),
 200             'creator': str_or_none(author_info.get('nickname')),
 201             'uploader_id': str_or_none(author_info.get('uid')),
 202             'uploader_url': user_url,
 203             'track': music_track,
 204             'album': str_or_none(music_info.get('album')) or None,
 205             'artist': music_author,
 206             'timestamp': int_or_none(aweme_detail.get('create_time')),
 207             'formats': formats,
 208             'thumbnails': thumbnails,
 209             'duration': int_or_none(traverse_obj(video_info, 'duration', ('download_addr', 'duration')), scale=1000)
 210         }
 211
 212     def _parse_aweme_video_web(self, aweme_detail, webpage_url):
 213         video_info = aweme_detail['video']
 214         author_info = traverse_obj(aweme_detail, 'author', 'authorInfo', default={})
 215         music_info = aweme_detail.get('music') or {}
 216         stats_info = aweme_detail.get('stats') or {}
 217         user_url = self._UPLOADER_URL_FORMAT % (traverse_obj(author_info,
 218                                                              'secUid', 'id', 'uid', 'uniqueId',
 219                                                              expected_type=str_or_none, get_all=False))
 220
 221         formats = []
 222         play_url = video_info.get('playAddr')
 223         width = video_info.get('width')
 224         height = video_info.get('height')
 225         if isinstance(play_url, str):
 226             formats = [{
 227                 'url': self._proto_relative_url(play_url),
 228                 'ext': 'mp4',
 229                 'width': width,
 230                 'height': height,
 231             }]
 232         elif isinstance(play_url, list):
 233             formats = [{
 234                 'url': self._proto_relative_url(url),
 235                 'ext': 'mp4',
 236                 'width': width,
 237                 'height': height,
 238             } for url in traverse_obj(play_url, (..., 'src'), expected_type=url_or_none, default=[]) if url]
 239
 240         download_url = url_or_none(video_info.get('downloadAddr')) or traverse_obj(video_info, ('download', 'url'), expected_type=url_or_none)
 241         if download_url:
 242             formats.append({
 243                 'format_id': 'download',
 244                 'url': self._proto_relative_url(download_url),
 245                 'ext': 'mp4',
 246                 'width': width,
 247                 'height': height,
 248             })
 249         self._remove_duplicate_formats(formats)
 250         self._sort_formats(formats)
 251
 252         thumbnails = []
 253         for thumbnail_name in ('thumbnail', 'cover', 'dynamicCover', 'originCover'):
 254             if aweme_detail.get(thumbnail_name):
 255                 thumbnails = [{
 256                     'url': self._proto_relative_url(aweme_detail[thumbnail_name]),
 257                     'width': width,
 258                     'height': height
 259                 }]
 260
 261         return {
 262             'id': traverse_obj(aweme_detail, 'id', 'awemeId', expected_type=str_or_none),
 263             'title': aweme_detail.get('desc'),
 264             'duration': try_get(aweme_detail, lambda x: x['video']['duration'], int),
 265             'view_count': int_or_none(stats_info.get('playCount')),
 266             'like_count': int_or_none(stats_info.get('diggCount')),
 267             'repost_count': int_or_none(stats_info.get('shareCount')),
 268             'comment_count': int_or_none(stats_info.get('commentCount')),
 269             'timestamp': int_or_none(aweme_detail.get('createTime')),
 270             'creator': str_or_none(author_info.get('nickname')),
 271             'uploader': str_or_none(author_info.get('uniqueId')),
 272             'uploader_id': str_or_none(author_info.get('id')),
 273             'uploader_url': user_url,
 274             'track': str_or_none(music_info.get('title')),
 275             'album': str_or_none(music_info.get('album')) or None,
 276             'artist': str_or_none(music_info.get('authorName')),
 277             'formats': formats,
 278             'thumbnails': thumbnails,
 279             'description': str_or_none(aweme_detail.get('desc')),
 280             'http_headers': {
 281                 'Referer': webpage_url
 282             }
 283         }
 284
 285
 286 class TikTokIE(TikTokBaseIE):
 287     _VALID_URL = r'https?://www\.tiktok\.com/@[\w\.-]+/video/(?P<id>\d+)'
 288
 289     _TESTS = [{
 290         'url': 'https://www.tiktok.com/@leenabhushan/video/6748451240264420610',
 291         'md5': '736bb7a466c6f0a6afeb597da1e6f5b7',
 292         'info_dict': {
 293             'id': '6748451240264420610',
 294             'ext': 'mp4',
 295             'title': '#jassmanak #lehanga #leenabhushan',
 296             'description': '#jassmanak #lehanga #leenabhushan',
 297             'duration': 13,
 298             'height': 1024,
 299             'width': 576,
 300             'uploader': 'leenabhushan',
 301             'uploader_id': '6691488002098119685',
 302             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAA_Eb4t1vodM1IuTy_cvp9CY22RAb59xqrO0Xtz9CYQJvgXaDvZxYnZYRzDWhhgJmy',
 303             'creator': 'facestoriesbyleenabh',
 304             'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
 305             'upload_date': '20191016',
 306             'timestamp': 1571246252,
 307             'view_count': int,
 308             'like_count': int,
 309             'repost_count': int,
 310             'comment_count': int,
 311         }
 312     }, {
 313         'url': 'https://www.tiktok.com/@patroxofficial/video/6742501081818877190?langCountry=en',
 314         'md5': '6f3cf8cdd9b28cb8363fe0a9a160695b',
 315         'info_dict': {
 316             'id': '6742501081818877190',
 317             'ext': 'mp4',
 318             'title': 'md5:5e2a23877420bb85ce6521dbee39ba94',
 319             'description': 'md5:5e2a23877420bb85ce6521dbee39ba94',
 320             'duration': 27,
 321             'height': 960,
 322             'width': 540,
 323             'uploader': 'patrox',
 324             'uploader_id': '18702747',
 325             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAAiFnldaILebi5heDoVU6bn4jBWWycX6-9U3xuNPqZ8Ws',
 326             'creator': 'patroX',
 327             'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
 328             'upload_date': '20190930',
 329             'timestamp': 1569860870,
 330             'view_count': int,
 331             'like_count': int,
 332             'repost_count': int,
 333             'comment_count': int,
 334         }
 335     }, {
 336         # Promoted content/ad
 337         'url': 'https://www.tiktok.com/@MS4wLjABAAAAAR29F6J2Ktu0Daw03BJyXPNoRQ-W7U5a0Mn3lVCq2rQhjOd_WNLclHUoFgwX8Eno/video/6932675057474981122',
 338         'only_matching': True,
 339     }]
 340
 341     def _extract_aweme_app(self, aweme_id):
 342         aweme_detail = self._call_api('aweme/detail', {'aweme_id': aweme_id}, aweme_id,
 343                                       note='Downloading video details', errnote='Unable to download video details')['aweme_detail']
 344         return self._parse_aweme_video_app(aweme_detail)
 345
 346     def _real_extract(self, url):
 347         video_id = self._match_id(url)
 348
 349         try:
 350             return self._extract_aweme_app(video_id)
 351         except ExtractorError as e:
 352             self.report_warning(f'{e}; Retrying with webpage')
 353
 354         # If we only call once, we get a 403 when downlaoding the video.
 355         self._download_webpage(url, video_id)
 356         webpage = self._download_webpage(url, video_id, note='Downloading video webpage')
 357         json_string = self._search_regex(
 358             r'id=\"__NEXT_DATA__\"\s+type=\"application\/json\"\s*[^>]+>\s*(?P<json_string_ld>[^<]+)',
 359             webpage, 'json_string', group='json_string_ld')
 360         json_data = self._parse_json(json_string, video_id)
 361         props_data = try_get(json_data, lambda x: x['props'], expected_type=dict)
 362
 363         # Chech statusCode for success
 364         status = props_data.get('pageProps').get('statusCode')
 365         if status == 0:
 366             return self._parse_aweme_video_web(props_data['pageProps']['itemInfo']['itemStruct'], url)
 367         elif status == 10216:
 368             raise ExtractorError('This video is private', expected=True)
 369
 370         raise ExtractorError('Video not available', video_id=video_id)
 371
 372
 373 class TikTokUserIE(TikTokBaseIE):
 374     IE_NAME = 'tiktok:user'
 375     _VALID_URL = r'https?://(?:www\.)?tiktok\.com/@(?P<id>[\w\.-]+)/?(?:$|[#?])'
 376     _TESTS = [{
 377         'url': 'https://tiktok.com/@corgibobaa?lang=en',
 378         'playlist_mincount': 45,
 379         'info_dict': {
 380             'id': '6935371178089399301',
 381             'title': 'corgibobaa',
 382         },
 383         'expected_warnings': ['Retrying']
 384     }, {
 385         'url': 'https://www.tiktok.com/@meme',
 386         'playlist_mincount': 593,
 387         'info_dict': {
 388             'id': '79005827461758976',
 389             'title': 'meme',
 390         },
 391         'expected_warnings': ['Retrying']
 392     }]
 393
 394     r'''  # TODO: Fix by adding _signature to api_url
 395     def _entries(self, webpage, user_id, username):
 396         secuid = self._search_regex(r'\"secUid\":\"(?P<secUid>[^\"]+)', webpage, username)
 397         verifyfp_cookie = self._get_cookies('https://www.tiktok.com').get('s_v_web_id')
 398         if not verifyfp_cookie:
 399             raise ExtractorError('Improper cookies (missing s_v_web_id).', expected=True)
 400         api_url = f'https://m.tiktok.com/api/post/item_list/?aid=1988&cookie_enabled=true&count=30&verifyFp={verifyfp_cookie.value}&secUid={secuid}&cursor='
 401         cursor = '0'
 402         for page in itertools.count():
 403             data_json = self._download_json(api_url + cursor, username, note='Downloading Page %d' % page)
 404             for video in data_json.get('itemList', []):
 405                 video_id = video['id']
 406                 video_url = f'https://www.tiktok.com/@{user_id}/video/{video_id}'
 407                 yield self._url_result(video_url, 'TikTok', video_id, str_or_none(video.get('desc')))
 408             if not data_json.get('hasMore'):
 409                 break
 410             cursor = data_json['cursor']
 411     '''
 412
 413     def _entries_api(self, webpage, user_id, username):
 414         query = {
 415             'user_id': user_id,
 416             'count': 21,
 417             'max_cursor': 0,
 418             'min_cursor': 0,
 419             'retry_type': 'no_retry',
 420             'device_id': ''.join(random.choice(string.digits) for _ in range(19)),  # Some endpoints don't like randomized device_id, so it isn't directly set in _call_api.
 421         }
 422
 423         max_retries = self.get_param('extractor_retries', 3)
 424         for page in itertools.count(1):
 425             for retries in itertools.count():
 426                 try:
 427                     post_list = self._call_api('aweme/post', query, username,
 428                                                note='Downloading user video list page %d%s' % (page, f' (attempt {retries})' if retries != 0 else ''),
 429                                                errnote='Unable to download user video list')
 430                 except ExtractorError as e:
 431                     if isinstance(e.cause, json.JSONDecodeError) and e.cause.pos == 0 and retries != max_retries:
 432                         self.report_warning('%s. Retrying...' % str(e.cause or e.msg))
 433                         continue
 434                     raise
 435                 break
 436             for video in post_list.get('aweme_list', []):
 437                 yield {
 438                     **self._parse_aweme_video_app(video),
 439                     'ie_key': TikTokIE.ie_key(),
 440                     'extractor': 'TikTok',
 441                 }
 442             if not post_list.get('has_more'):
 443                 break
 444             query['max_cursor'] = post_list['max_cursor']
 445
 446     def _real_extract(self, url):
 447         user_name = self._match_id(url)
 448         webpage = self._download_webpage(url, user_name, headers={
 449             'User-Agent': 'facebookexternalhit/1.1 (+http://www.facebook.com/externalhit_uatext.php)'
 450         })
 451         user_id = self._html_search_regex(r'snssdk\d*://user/profile/(\d+)', webpage, 'user ID')
 452         return self.playlist_result(self._entries_api(webpage, user_id, user_name), user_id, user_name)
 453
 454
 455 class DouyinIE(TikTokIE):
 456     _VALID_URL = r'https?://(?:www\.)?douyin\.com/video/(?P<id>[0-9]+)'
 457     _TESTS = [{
 458         'url': 'https://www.douyin.com/video/6961737553342991651',
 459         'md5': '10523312c8b8100f353620ac9dc8f067',
 460         'info_dict': {
 461             'id': '6961737553342991651',
 462             'ext': 'mp4',
 463             'title': '#杨超越  小小水手带你去远航❤️',
 464             'uploader': '杨超越',
 465             'upload_date': '20210513',
 466             'timestamp': 1620905839,
 467             'uploader_id': '110403406559',
 468             'view_count': int,
 469             'like_count': int,
 470             'repost_count': int,
 471             'comment_count': int,
 472         }
 473     }, {
 474         'url': 'https://www.douyin.com/video/6982497745948921092',
 475         'md5': 'd78408c984b9b5102904cf6b6bc2d712',
 476         'info_dict': {
 477             'id': '6982497745948921092',
 478             'ext': 'mp4',
 479             'title': '这个夏日和小羊@杨超越 一起遇见白色幻想',
 480             'uploader': '杨超越工作室',
 481             'upload_date': '20210708',
 482             'timestamp': 1625739481,
 483             'uploader_id': '408654318141572',
 484             'view_count': int,
 485             'like_count': int,
 486             'repost_count': int,
 487             'comment_count': int,
 488         }
 489     }, {
 490         'url': 'https://www.douyin.com/video/6953975910773099811',
 491         'md5': '72e882e24f75064c218b76c8b713c185',
 492         'info_dict': {
 493             'id': '6953975910773099811',
 494             'ext': 'mp4',
 495             'title': '#一起看海  出现在你的夏日里',
 496             'uploader': '杨超越',
 497             'upload_date': '20210422',
 498             'timestamp': 1619098692,
 499             'uploader_id': '110403406559',
 500             'view_count': int,
 501             'like_count': int,
 502             'repost_count': int,
 503             'comment_count': int,
 504         }
 505     }, {
 506         'url': 'https://www.douyin.com/video/6950251282489675042',
 507         'md5': 'b4db86aec367ef810ddd38b1737d2fed',
 508         'info_dict': {
 509             'id': '6950251282489675042',
 510             'ext': 'mp4',
 511             'title': '哈哈哈，成功了哈哈哈哈哈哈',
 512             'uploader': '杨超越',
 513             'upload_date': '20210412',
 514             'timestamp': 1618231483,
 515             'uploader_id': '110403406559',
 516             'view_count': int,
 517             'like_count': int,
 518             'repost_count': int,
 519             'comment_count': int,
 520         }
 521     }, {
 522         'url': 'https://www.douyin.com/video/6963263655114722595',
 523         'md5': '1abe1c477d05ee62efb40bf2329957cf',
 524         'info_dict': {
 525             'id': '6963263655114722595',
 526             'ext': 'mp4',
 527             'title': '#哪个爱豆的105度最甜 换个角度看看我哈哈',
 528             'uploader': '杨超越',
 529             'upload_date': '20210517',
 530             'timestamp': 1621261163,
 531             'uploader_id': '110403406559',
 532             'view_count': int,
 533             'like_count': int,
 534             'repost_count': int,
 535             'comment_count': int,
 536         }
 537     }]
 538     _APP_VERSION = '9.6.0'
 539     _MANIFEST_APP_VERSION = '960'
 540     _APP_NAME = 'aweme'
 541     _AID = 1128
 542     _API_HOSTNAME = 'aweme.snssdk.com'
 543     _UPLOADER_URL_FORMAT = 'https://www.douyin.com/user/%s'
 544
 545     def _real_extract(self, url):
 546         video_id = self._match_id(url)
 547
 548         try:
 549             return self._extract_aweme_app(video_id)
 550         except ExtractorError as e:
 551             self.report_warning(f'{e}; Retrying with webpage')
 552
 553         webpage = self._download_webpage(url, video_id)
 554         render_data_json = self._search_regex(
 555             r'<script [^>]*\bid=[\'"]RENDER_DATA[\'"][^>]*>(%7B.+%7D)</script>',
 556             webpage, 'render data', default=None)
 557         if not render_data_json:
 558             # TODO: Run verification challenge code to generate signature cookies
 559             raise ExtractorError('Fresh cookies (not necessarily logged in) are needed')
 560
 561         render_data = self._parse_json(
 562             render_data_json, video_id, transform_source=compat_urllib_parse_unquote)
 563         return self._parse_aweme_video_web(
 564             traverse_obj(render_data, (..., 'aweme', 'detail'), get_all=False), url)