yt_dlp/extractor/tiktok.py

   1 import functools
   2 import itertools
   3 import json
   4 import random
   5 import re
   6 import string
   7 import time
   8 import urllib.parse
   9 import uuid
  10
  11 from .common import InfoExtractor
  12 from ..networking import HEADRequest
  13 from ..utils import (
  14     ExtractorError,
  15     UnsupportedError,
  16     UserNotLive,
  17     determine_ext,
  18     filter_dict,
  19     format_field,
  20     int_or_none,
  21     join_nonempty,
  22     merge_dicts,
  23     mimetype2ext,
  24     parse_qs,
  25     qualities,
  26     srt_subtitles_timecode,
  27     str_or_none,
  28     traverse_obj,
  29     try_call,
  30     try_get,
  31     url_or_none,
  32     urlencode_postdata,
  33 )
  34
  35
  36 class TikTokBaseIE(InfoExtractor):
  37     _UPLOADER_URL_FORMAT = 'https://www.tiktok.com/@%s'
  38     _WEBPAGE_HOST = 'https://www.tiktok.com/'
  39     QUALITIES = ('360p', '540p', '720p', '1080p')
  40
  41     _APP_INFO_DEFAULTS = {
  42         # unique "install id"
  43         'iid': None,
  44         # TikTok (KR/PH/TW/TH/VN) = trill, TikTok (rest of world) = musical_ly, Douyin = aweme
  45         'app_name': 'musical_ly',
  46         'app_version': '35.1.3',
  47         'manifest_app_version': '2023501030',
  48         # "app id": aweme = 1128, trill = 1180, musical_ly = 1233, universal = 0
  49         'aid': '0',
  50     }
  51     _APP_INFO_POOL = None
  52     _APP_INFO = None
  53     _APP_USER_AGENT = None
  54
  55     @functools.cached_property
  56     def _KNOWN_APP_INFO(self):
  57         # If we have a genuine device ID, we may not need any IID
  58         default = [''] if self._KNOWN_DEVICE_ID else []
  59         return self._configuration_arg('app_info', default, ie_key=TikTokIE)
  60
  61     @functools.cached_property
  62     def _KNOWN_DEVICE_ID(self):
  63         return self._configuration_arg('device_id', [None], ie_key=TikTokIE)[0]
  64
  65     @functools.cached_property
  66     def _DEVICE_ID(self):
  67         return self._KNOWN_DEVICE_ID or str(random.randint(7250000000000000000, 7351147085025500000))
  68
  69     @functools.cached_property
  70     def _API_HOSTNAME(self):
  71         return self._configuration_arg(
  72             'api_hostname', ['api16-normal-c-useast1a.tiktokv.com'], ie_key=TikTokIE)[0]
  73
  74     def _get_next_app_info(self):
  75         if self._APP_INFO_POOL is None:
  76             defaults = {
  77                 key: self._configuration_arg(key, [default], ie_key=TikTokIE)[0]
  78                 for key, default in self._APP_INFO_DEFAULTS.items()
  79                 if key != 'iid'
  80             }
  81             self._APP_INFO_POOL = [
  82                 {**defaults, **dict(
  83                     (k, v) for k, v in zip(self._APP_INFO_DEFAULTS, app_info.split('/')) if v
  84                 )} for app_info in self._KNOWN_APP_INFO
  85             ]
  86
  87         if not self._APP_INFO_POOL:
  88             return False
  89
  90         self._APP_INFO = self._APP_INFO_POOL.pop(0)
  91
  92         app_name = self._APP_INFO['app_name']
  93         version = self._APP_INFO['manifest_app_version']
  94         if app_name == 'musical_ly':
  95             package = f'com.zhiliaoapp.musically/{version}'
  96         else:  # trill, aweme
  97             package = f'com.ss.android.ugc.{app_name}/{version}'
  98         self._APP_USER_AGENT = f'{package} (Linux; U; Android 13; en_US; Pixel 7; Build/TD1A.220804.031; Cronet/58.0.2991.0)'
  99
 100         return True
 101
 102     @staticmethod
 103     def _create_url(user_id, video_id):
 104         return f'https://www.tiktok.com/@{user_id or "_"}/video/{video_id}'
 105
 106     def _get_sigi_state(self, webpage, display_id):
 107         return self._search_json(
 108             r'<script[^>]+\bid="(?:SIGI_STATE|sigi-persisted-data)"[^>]*>', webpage,
 109             'sigi state', display_id, end_pattern=r'</script>', default={})
 110
 111     def _get_universal_data(self, webpage, display_id):
 112         return traverse_obj(self._search_json(
 113             r'<script[^>]+\bid="__UNIVERSAL_DATA_FOR_REHYDRATION__"[^>]*>', webpage,
 114             'universal data', display_id, end_pattern=r'</script>', default={}),
 115             ('__DEFAULT_SCOPE__', {dict})) or {}
 116
 117     def _call_api_impl(self, ep, video_id, query=None, data=None, headers=None, fatal=True,
 118                        note='Downloading API JSON', errnote='Unable to download API page'):
 119         self._set_cookie(self._API_HOSTNAME, 'odin_tt', ''.join(random.choices('0123456789abcdef', k=160)))
 120         webpage_cookies = self._get_cookies(self._WEBPAGE_HOST)
 121         if webpage_cookies.get('sid_tt'):
 122             self._set_cookie(self._API_HOSTNAME, 'sid_tt', webpage_cookies['sid_tt'].value)
 123         return self._download_json(
 124             f'https://{self._API_HOSTNAME}/aweme/v1/{ep}/', video_id=video_id,
 125             fatal=fatal, note=note, errnote=errnote, headers={
 126                 'User-Agent': self._APP_USER_AGENT,
 127                 'Accept': 'application/json',
 128                 **(headers or {}),
 129             }, query=query, data=data)
 130
 131     def _build_api_query(self, query):
 132         return filter_dict({
 133             **query,
 134             'device_platform': 'android',
 135             'os': 'android',
 136             'ssmix': 'a',
 137             '_rticket': int(time.time() * 1000),
 138             'cdid': str(uuid.uuid4()),
 139             'channel': 'googleplay',
 140             'aid': self._APP_INFO['aid'],
 141             'app_name': self._APP_INFO['app_name'],
 142             'version_code': ''.join(f'{int(v):02d}' for v in self._APP_INFO['app_version'].split('.')),
 143             'version_name': self._APP_INFO['app_version'],
 144             'manifest_version_code': self._APP_INFO['manifest_app_version'],
 145             'update_version_code': self._APP_INFO['manifest_app_version'],
 146             'ab_version': self._APP_INFO['app_version'],
 147             'resolution': '1080*2400',
 148             'dpi': 420,
 149             'device_type': 'Pixel 7',
 150             'device_brand': 'Google',
 151             'language': 'en',
 152             'os_api': '29',
 153             'os_version': '13',
 154             'ac': 'wifi',
 155             'is_pad': '0',
 156             'current_region': 'US',
 157             'app_type': 'normal',
 158             'sys_region': 'US',
 159             'last_install_time': int(time.time()) - random.randint(86400, 1123200),
 160             'timezone_name': 'America/New_York',
 161             'residence': 'US',
 162             'app_language': 'en',
 163             'timezone_offset': '-14400',
 164             'host_abi': 'armeabi-v7a',
 165             'locale': 'en',
 166             'ac2': 'wifi5g',
 167             'uoo': '1',
 168             'carrier_region': 'US',
 169             'op_region': 'US',
 170             'build_number': self._APP_INFO['app_version'],
 171             'region': 'US',
 172             'ts': int(time.time()),
 173             'iid': self._APP_INFO.get('iid'),
 174             'device_id': self._DEVICE_ID,
 175             'openudid': ''.join(random.choices('0123456789abcdef', k=16)),
 176         })
 177
 178     def _call_api(self, ep, video_id, query=None, data=None, headers=None, fatal=True,
 179                   note='Downloading API JSON', errnote='Unable to download API page'):
 180         if not self._APP_INFO and not self._get_next_app_info():
 181             message = 'No working app info is available'
 182             if fatal:
 183                 raise ExtractorError(message, expected=True)
 184             else:
 185                 self.report_warning(message)
 186                 return
 187
 188         max_tries = len(self._APP_INFO_POOL) + 1  # _APP_INFO_POOL + _APP_INFO
 189         for count in itertools.count(1):
 190             self.write_debug(str(self._APP_INFO))
 191             real_query = self._build_api_query(query or {})
 192             try:
 193                 return self._call_api_impl(
 194                     ep, video_id, query=real_query, data=data, headers=headers,
 195                     fatal=fatal, note=note, errnote=errnote)
 196             except ExtractorError as e:
 197                 if isinstance(e.cause, json.JSONDecodeError) and e.cause.pos == 0:
 198                     message = str(e.cause or e.msg)
 199                     if not self._get_next_app_info():
 200                         if fatal:
 201                             raise
 202                         else:
 203                             self.report_warning(message)
 204                             return
 205                     self.report_warning(f'{message}. Retrying... (attempt {count} of {max_tries})')
 206                     continue
 207                 raise
 208
 209     def _extract_aweme_app(self, aweme_id):
 210         aweme_detail = traverse_obj(
 211             self._call_api('multi/aweme/detail', aweme_id, data=urlencode_postdata({
 212                 'aweme_ids': f'[{aweme_id}]',
 213                 'request_source': '0',
 214             }), headers={'X-Argus': ''}), ('aweme_details', 0, {dict}))
 215         if not aweme_detail:
 216             raise ExtractorError('Unable to extract aweme detail info', video_id=aweme_id)
 217         return self._parse_aweme_video_app(aweme_detail)
 218
 219     def _extract_web_data_and_status(self, url, video_id, fatal=True):
 220         video_data, status = {}, -1
 221
 222         res = self._download_webpage_handle(url, video_id, fatal=fatal, headers={'User-Agent': 'Mozilla/5.0'})
 223         if res is False:
 224             return video_data, status
 225
 226         webpage, urlh = res
 227         if urllib.parse.urlparse(urlh.url).path == '/login':
 228             message = 'TikTok is requiring login for access to this content'
 229             if fatal:
 230                 self.raise_login_required(message)
 231             self.report_warning(f'{message}. {self._login_hint()}')
 232             return video_data, status
 233
 234         if universal_data := self._get_universal_data(webpage, video_id):
 235             self.write_debug('Found universal data for rehydration')
 236             status = traverse_obj(universal_data, ('webapp.video-detail', 'statusCode', {int})) or 0
 237             video_data = traverse_obj(universal_data, ('webapp.video-detail', 'itemInfo', 'itemStruct', {dict}))
 238
 239         elif sigi_data := self._get_sigi_state(webpage, video_id):
 240             self.write_debug('Found sigi state data')
 241             status = traverse_obj(sigi_data, ('VideoPage', 'statusCode', {int})) or 0
 242             video_data = traverse_obj(sigi_data, ('ItemModule', video_id, {dict}))
 243
 244         elif next_data := self._search_nextjs_data(webpage, video_id, default={}):
 245             self.write_debug('Found next.js data')
 246             status = traverse_obj(next_data, ('props', 'pageProps', 'statusCode', {int})) or 0
 247             video_data = traverse_obj(next_data, ('props', 'pageProps', 'itemInfo', 'itemStruct', {dict}))
 248
 249         elif fatal:
 250             raise ExtractorError('Unable to extract webpage video data')
 251
 252         return video_data, status
 253
 254     def _get_subtitles(self, aweme_detail, aweme_id, user_name):
 255         # TODO: Extract text positioning info
 256
 257         EXT_MAP = {  # From lowest to highest preference
 258             'creator_caption': 'json',
 259             'srt': 'srt',
 260             'webvtt': 'vtt',
 261         }
 262         preference = qualities(tuple(EXT_MAP.values()))
 263
 264         subtitles = {}
 265
 266         # aweme/detail endpoint subs
 267         captions_info = traverse_obj(
 268             aweme_detail, ('interaction_stickers', ..., 'auto_video_caption_info', 'auto_captions', ...), expected_type=dict)
 269         for caption in captions_info:
 270             caption_url = traverse_obj(caption, ('url', 'url_list', ...), expected_type=url_or_none, get_all=False)
 271             if not caption_url:
 272                 continue
 273             caption_json = self._download_json(
 274                 caption_url, aweme_id, note='Downloading captions', errnote='Unable to download captions', fatal=False)
 275             if not caption_json:
 276                 continue
 277             subtitles.setdefault(caption.get('language', 'en'), []).append({
 278                 'ext': 'srt',
 279                 'data': '\n\n'.join(
 280                     f'{i + 1}\n{srt_subtitles_timecode(line["start_time"] / 1000)} --> {srt_subtitles_timecode(line["end_time"] / 1000)}\n{line["text"]}'
 281                     for i, line in enumerate(caption_json['utterances']) if line.get('text')),
 282             })
 283         # feed endpoint subs
 284         if not subtitles:
 285             for caption in traverse_obj(aweme_detail, ('video', 'cla_info', 'caption_infos', ...), expected_type=dict):
 286                 if not caption.get('url'):
 287                     continue
 288                 subtitles.setdefault(caption.get('lang') or 'en', []).append({
 289                     'url': caption['url'],
 290                     'ext': EXT_MAP.get(caption.get('Format')),
 291                 })
 292         # webpage subs
 293         if not subtitles:
 294             if user_name:  # only _parse_aweme_video_app needs to extract the webpage here
 295                 aweme_detail, _ = self._extract_web_data_and_status(
 296                     self._create_url(user_name, aweme_id), aweme_id, fatal=False)
 297             for caption in traverse_obj(aweme_detail, ('video', 'subtitleInfos', lambda _, v: v['Url'])):
 298                 subtitles.setdefault(caption.get('LanguageCodeName') or 'en', []).append({
 299                     'url': caption['Url'],
 300                     'ext': EXT_MAP.get(caption.get('Format')),
 301                 })
 302
 303         # Deprioritize creator_caption json since it can't be embedded or used by media players
 304         for lang, subs_list in subtitles.items():
 305             subtitles[lang] = sorted(subs_list, key=lambda x: preference(x['ext']))
 306
 307         return subtitles
 308
 309     def _parse_url_key(self, url_key):
 310         format_id, codec, res, bitrate = self._search_regex(
 311             r'v[^_]+_(?P<id>(?P<codec>[^_]+)_(?P<res>\d+p)_(?P<bitrate>\d+))', url_key,
 312             'url key', default=(None, None, None, None), group=('id', 'codec', 'res', 'bitrate'))
 313         if not format_id:
 314             return {}, None
 315         return {
 316             'format_id': format_id,
 317             'vcodec': 'h265' if codec == 'bytevc1' else codec,
 318             'tbr': int_or_none(bitrate, scale=1000) or None,
 319             'quality': qualities(self.QUALITIES)(res),
 320         }, res
 321
 322     def _parse_aweme_video_app(self, aweme_detail):
 323         aweme_id = aweme_detail['aweme_id']
 324         video_info = aweme_detail['video']
 325         known_resolutions = {}
 326
 327         def audio_meta(url):
 328             ext = determine_ext(url, default_ext='m4a')
 329             return {
 330                 'format_note': 'Music track',
 331                 'ext': ext,
 332                 'acodec': 'aac' if ext == 'm4a' else ext,
 333                 'vcodec': 'none',
 334                 'width': None,
 335                 'height': None,
 336             } if ext == 'mp3' or '-music-' in url else {}
 337
 338         def extract_addr(addr, add_meta={}):
 339             parsed_meta, res = self._parse_url_key(addr.get('url_key', ''))
 340             is_bytevc2 = parsed_meta.get('vcodec') == 'bytevc2'
 341             if res:
 342                 known_resolutions.setdefault(res, {}).setdefault('height', int_or_none(addr.get('height')))
 343                 known_resolutions[res].setdefault('width', int_or_none(addr.get('width')))
 344                 parsed_meta.update(known_resolutions.get(res, {}))
 345                 add_meta.setdefault('height', int_or_none(res[:-1]))
 346             return [{
 347                 'url': url,
 348                 'filesize': int_or_none(addr.get('data_size')),
 349                 'ext': 'mp4',
 350                 'acodec': 'aac',
 351                 'source_preference': -2 if 'aweme/v1' in url else -1,  # Downloads from API might get blocked
 352                 **add_meta, **parsed_meta,
 353                 # bytevc2 is bytedance's own custom h266/vvc codec, as-of-yet unplayable
 354                 'preference': -100 if is_bytevc2 else -1,
 355                 'format_note': join_nonempty(
 356                     add_meta.get('format_note'), '(API)' if 'aweme/v1' in url else None,
 357                     '(UNPLAYABLE)' if is_bytevc2 else None, delim=' '),
 358                 **audio_meta(url),
 359             } for url in addr.get('url_list') or []]
 360
 361         # Hack: Add direct video links first to prioritize them when removing duplicate formats
 362         formats = []
 363         width = int_or_none(video_info.get('width'))
 364         height = int_or_none(video_info.get('height'))
 365         ratio = try_call(lambda: width / height) or 0.5625
 366         if video_info.get('play_addr'):
 367             formats.extend(extract_addr(video_info['play_addr'], {
 368                 'format_id': 'play_addr',
 369                 'format_note': 'Direct video',
 370                 'vcodec': 'h265' if traverse_obj(
 371                     video_info, 'is_bytevc1', 'is_h265') else 'h264',  # TODO: Check for "direct iOS" videos, like https://www.tiktok.com/@cookierun_dev/video/7039716639834656002
 372                 'width': width,
 373                 'height': height,
 374             }))
 375         if video_info.get('download_addr'):
 376             download_addr = video_info['download_addr']
 377             dl_width = int_or_none(download_addr.get('width'))
 378             formats.extend(extract_addr(download_addr, {
 379                 'format_id': 'download_addr',
 380                 'format_note': 'Download video%s' % (', watermarked' if video_info.get('has_watermark') else ''),
 381                 'vcodec': 'h264',
 382                 'width': dl_width,
 383                 'height': try_call(lambda: int(dl_width / ratio)),  # download_addr['height'] is wrong
 384                 'preference': -2 if video_info.get('has_watermark') else -1,
 385             }))
 386         if video_info.get('play_addr_h264'):
 387             formats.extend(extract_addr(video_info['play_addr_h264'], {
 388                 'format_id': 'play_addr_h264',
 389                 'format_note': 'Direct video',
 390                 'vcodec': 'h264',
 391             }))
 392         if video_info.get('play_addr_bytevc1'):
 393             formats.extend(extract_addr(video_info['play_addr_bytevc1'], {
 394                 'format_id': 'play_addr_bytevc1',
 395                 'format_note': 'Direct video',
 396                 'vcodec': 'h265',
 397             }))
 398
 399         for bitrate in video_info.get('bit_rate', []):
 400             if bitrate.get('play_addr'):
 401                 formats.extend(extract_addr(bitrate['play_addr'], {
 402                     'format_id': bitrate.get('gear_name'),
 403                     'format_note': 'Playback video',
 404                     'tbr': try_get(bitrate, lambda x: x['bit_rate'] / 1000),
 405                     'vcodec': 'h265' if traverse_obj(
 406                         bitrate, 'is_bytevc1', 'is_h265') else 'h264',
 407                     'fps': bitrate.get('FPS'),
 408                 }))
 409
 410         self._remove_duplicate_formats(formats)
 411         auth_cookie = self._get_cookies(self._WEBPAGE_HOST).get('sid_tt')
 412         if auth_cookie:
 413             for f in formats:
 414                 self._set_cookie(urllib.parse.urlparse(f['url']).hostname, 'sid_tt', auth_cookie.value)
 415
 416         thumbnails = []
 417         for cover_id in ('cover', 'ai_dynamic_cover', 'animated_cover', 'ai_dynamic_cover_bak',
 418                          'origin_cover', 'dynamic_cover'):
 419             for cover_url in traverse_obj(video_info, (cover_id, 'url_list', ...)):
 420                 thumbnails.append({
 421                     'id': cover_id,
 422                     'url': cover_url,
 423                 })
 424
 425         stats_info = aweme_detail.get('statistics') or {}
 426         music_info = aweme_detail.get('music') or {}
 427         labels = traverse_obj(aweme_detail, ('hybrid_label', ..., 'text'), expected_type=str)
 428
 429         contained_music_track = traverse_obj(
 430             music_info, ('matched_song', 'title'), ('matched_pgc_sound', 'title'), expected_type=str)
 431         contained_music_author = traverse_obj(
 432             music_info, ('matched_song', 'author'), ('matched_pgc_sound', 'author'), 'author', expected_type=str)
 433
 434         is_generic_og_trackname = music_info.get('is_original_sound') and music_info.get('title') == 'original sound - {}'.format(music_info.get('owner_handle'))
 435         if is_generic_og_trackname:
 436             music_track, music_author = contained_music_track or 'original sound', contained_music_author
 437         else:
 438             music_track, music_author = music_info.get('title'), traverse_obj(music_info, ('author', {str}))
 439
 440         author_info = traverse_obj(aweme_detail, ('author', {
 441             'uploader': ('unique_id', {str}),
 442             'uploader_id': ('uid', {str_or_none}),
 443             'channel': ('nickname', {str}),
 444             'channel_id': ('sec_uid', {str}),
 445         }))
 446
 447         return {
 448             'id': aweme_id,
 449             **traverse_obj(aweme_detail, {
 450                 'title': ('desc', {str}),
 451                 'description': ('desc', {str}),
 452                 'timestamp': ('create_time', {int_or_none}),
 453             }),
 454             **traverse_obj(stats_info, {
 455                 'view_count': 'play_count',
 456                 'like_count': 'digg_count',
 457                 'repost_count': 'share_count',
 458                 'comment_count': 'comment_count',
 459             }, expected_type=int_or_none),
 460             **author_info,
 461             'channel_url': format_field(author_info, 'channel_id', self._UPLOADER_URL_FORMAT, default=None),
 462             'uploader_url': format_field(
 463                 author_info, ['uploader', 'uploader_id'], self._UPLOADER_URL_FORMAT, default=None),
 464             'track': music_track,
 465             'album': str_or_none(music_info.get('album')) or None,
 466             'artists': re.split(r'(?:, | & )', music_author) if music_author else None,
 467             'formats': formats,
 468             'subtitles': self.extract_subtitles(
 469                 aweme_detail, aweme_id, traverse_obj(author_info, 'uploader', 'uploader_id', 'channel_id')),
 470             'thumbnails': thumbnails,
 471             'duration': (traverse_obj(video_info, (
 472                 (None, 'download_addr'), 'duration', {functools.partial(int_or_none, scale=1000)}, any))
 473                 or traverse_obj(music_info, ('duration', {int_or_none}))),
 474             'availability': self._availability(
 475                 is_private='Private' in labels,
 476                 needs_subscription='Friends only' in labels,
 477                 is_unlisted='Followers only' in labels),
 478             '_format_sort_fields': ('quality', 'codec', 'size', 'br'),
 479         }
 480
 481     def _extract_web_formats(self, aweme_detail):
 482         COMMON_FORMAT_INFO = {
 483             'ext': 'mp4',
 484             'vcodec': 'h264',
 485             'acodec': 'aac',
 486         }
 487         video_info = traverse_obj(aweme_detail, ('video', {dict})) or {}
 488         play_width = int_or_none(video_info.get('width'))
 489         play_height = int_or_none(video_info.get('height'))
 490         ratio = try_call(lambda: play_width / play_height) or 0.5625
 491         formats = []
 492
 493         for bitrate_info in traverse_obj(video_info, ('bitrateInfo', lambda _, v: v['PlayAddr']['UrlList'])):
 494             format_info, res = self._parse_url_key(
 495                 traverse_obj(bitrate_info, ('PlayAddr', 'UrlKey', {str})) or '')
 496             # bytevc2 is bytedance's own custom h266/vvc codec, as-of-yet unplayable
 497             is_bytevc2 = format_info.get('vcodec') == 'bytevc2'
 498             format_info.update({
 499                 'format_note': 'UNPLAYABLE' if is_bytevc2 else None,
 500                 'preference': -100 if is_bytevc2 else -1,
 501                 'filesize': traverse_obj(bitrate_info, ('PlayAddr', 'DataSize', {int_or_none})),
 502             })
 503
 504             if dimension := (res and int(res[:-1])):
 505                 if dimension == 540:  # '540p' is actually 576p
 506                     dimension = 576
 507                 if ratio < 1:  # portrait: res/dimension is width
 508                     y = int(dimension / ratio)
 509                     format_info.update({
 510                         'width': dimension,
 511                         'height': y - (y % 2),
 512                     })
 513                 else:  # landscape: res/dimension is height
 514                     x = int(dimension * ratio)
 515                     format_info.update({
 516                         'width': x + (x % 2),
 517                         'height': dimension,
 518                     })
 519
 520             for video_url in traverse_obj(bitrate_info, ('PlayAddr', 'UrlList', ..., {url_or_none})):
 521                 formats.append({
 522                     **COMMON_FORMAT_INFO,
 523                     **format_info,
 524                     'url': self._proto_relative_url(video_url),
 525                 })
 526
 527         # We don't have res string for play formats, but need quality for sorting & de-duplication
 528         play_quality = traverse_obj(formats, (lambda _, v: v['width'] == play_width, 'quality', any))
 529
 530         for play_url in traverse_obj(video_info, ('playAddr', ((..., 'src'), None), {url_or_none})):
 531             formats.append({
 532                 **COMMON_FORMAT_INFO,
 533                 'format_id': 'play',
 534                 'url': self._proto_relative_url(play_url),
 535                 'width': play_width,
 536                 'height': play_height,
 537                 'quality': play_quality,
 538             })
 539
 540         for download_url in traverse_obj(video_info, (('downloadAddr', ('download', 'url')), {url_or_none})):
 541             formats.append({
 542                 **COMMON_FORMAT_INFO,
 543                 'format_id': 'download',
 544                 'url': self._proto_relative_url(download_url),
 545             })
 546
 547         self._remove_duplicate_formats(formats)
 548
 549         for f in traverse_obj(formats, lambda _, v: 'unwatermarked' not in v['url']):
 550             f.update({
 551                 'format_note': join_nonempty(f.get('format_note'), 'watermarked', delim=', '),
 552                 'preference': f.get('preference') or -2,
 553             })
 554
 555         # Is it a slideshow with only audio for download?
 556         if not formats and traverse_obj(aweme_detail, ('music', 'playUrl', {url_or_none})):
 557             audio_url = aweme_detail['music']['playUrl']
 558             ext = traverse_obj(parse_qs(audio_url), (
 559                 'mime_type', -1, {lambda x: x.replace('_', '/')}, {mimetype2ext})) or 'm4a'
 560             formats.append({
 561                 'format_id': 'audio',
 562                 'url': self._proto_relative_url(audio_url),
 563                 'ext': ext,
 564                 'acodec': 'aac' if ext == 'm4a' else ext,
 565                 'vcodec': 'none',
 566             })
 567
 568         return formats
 569
 570     def _parse_aweme_video_web(self, aweme_detail, webpage_url, video_id, extract_flat=False):
 571         author_info = traverse_obj(aweme_detail, (('authorInfo', 'author', None), {
 572             'channel': ('nickname', {str}),
 573             'channel_id': (('authorSecId', 'secUid'), {str}),
 574             'uploader': (('uniqueId', 'author'), {str}),
 575             'uploader_id': (('authorId', 'uid', 'id'), {str_or_none}),
 576         }), get_all=False)
 577
 578         return {
 579             'id': video_id,
 580             'formats': None if extract_flat else self._extract_web_formats(aweme_detail),
 581             'subtitles': None if extract_flat else self.extract_subtitles(aweme_detail, video_id, None),
 582             'http_headers': {'Referer': webpage_url},
 583             **author_info,
 584             'channel_url': format_field(author_info, 'channel_id', self._UPLOADER_URL_FORMAT, default=None),
 585             'uploader_url': format_field(
 586                 author_info, ['uploader', 'uploader_id'], self._UPLOADER_URL_FORMAT, default=None),
 587             **traverse_obj(aweme_detail, ('music', {
 588                 'track': ('title', {str}),
 589                 'album': ('album', {str}, {lambda x: x or None}),
 590                 'artists': ('authorName', {str}, {lambda x: re.split(r'(?:, | & )', x) if x else None}),
 591                 'duration': ('duration', {int_or_none}),
 592             })),
 593             **traverse_obj(aweme_detail, {
 594                 'title': ('desc', {str}),
 595                 'description': ('desc', {str}),
 596                 # audio-only slideshows have a video duration of 0 and an actual audio duration
 597                 'duration': ('video', 'duration', {int_or_none}, {lambda x: x or None}),
 598                 'timestamp': ('createTime', {int_or_none}),
 599             }),
 600             **traverse_obj(aweme_detail, ('stats', {
 601                 'view_count': 'playCount',
 602                 'like_count': 'diggCount',
 603                 'repost_count': 'shareCount',
 604                 'comment_count': 'commentCount',
 605             }), expected_type=int_or_none),
 606             'thumbnails': traverse_obj(aweme_detail, (
 607                 (None, 'video'), ('thumbnail', 'cover', 'dynamicCover', 'originCover'), {
 608                     'url': ({url_or_none}, {self._proto_relative_url}),
 609                 },
 610             )),
 611         }
 612
 613
 614 class TikTokIE(TikTokBaseIE):
 615     _VALID_URL = r'https?://www\.tiktok\.com/(?:embed|@(?P<user_id>[\w\.-]+)?/video)/(?P<id>\d+)'
 616     _EMBED_REGEX = [rf'<(?:script|iframe)[^>]+\bsrc=(["\'])(?P<url>{_VALID_URL})']
 617
 618     _TESTS = [{
 619         'url': 'https://www.tiktok.com/@leenabhushan/video/6748451240264420610',
 620         'md5': '736bb7a466c6f0a6afeb597da1e6f5b7',
 621         'info_dict': {
 622             'id': '6748451240264420610',
 623             'ext': 'mp4',
 624             'title': '#jassmanak #lehanga #leenabhushan',
 625             'description': '#jassmanak #lehanga #leenabhushan',
 626             'duration': 13,
 627             'height': 1024,
 628             'width': 576,
 629             'uploader': 'leenabhushan',
 630             'uploader_id': '6691488002098119685',
 631             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAA_Eb4t1vodM1IuTy_cvp9CY22RAb59xqrO0Xtz9CYQJvgXaDvZxYnZYRzDWhhgJmy',
 632             'creator': 'facestoriesbyleenabh',
 633             'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
 634             'upload_date': '20191016',
 635             'timestamp': 1571246252,
 636             'view_count': int,
 637             'like_count': int,
 638             'repost_count': int,
 639             'comment_count': int,
 640             'artist': 'Ysrbeats',
 641             'album': 'Lehanga',
 642             'track': 'Lehanga',
 643         },
 644         'skip': '404 Not Found',
 645     }, {
 646         'url': 'https://www.tiktok.com/@patroxofficial/video/6742501081818877190?langCountry=en',
 647         'md5': 'f21112672ee4ce05ca390fb6522e1b6f',
 648         'info_dict': {
 649             'id': '6742501081818877190',
 650             'ext': 'mp4',
 651             'title': 'md5:5e2a23877420bb85ce6521dbee39ba94',
 652             'description': 'md5:5e2a23877420bb85ce6521dbee39ba94',
 653             'duration': 27,
 654             'height': 1024,
 655             'width': 576,
 656             'uploader': 'patrox',
 657             'uploader_id': '18702747',
 658             'uploader_url': 'https://www.tiktok.com/@patrox',
 659             'channel_url': 'https://www.tiktok.com/@MS4wLjABAAAAiFnldaILebi5heDoVU6bn4jBWWycX6-9U3xuNPqZ8Ws',
 660             'channel_id': 'MS4wLjABAAAAiFnldaILebi5heDoVU6bn4jBWWycX6-9U3xuNPqZ8Ws',
 661             'channel': 'patroX',
 662             'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
 663             'upload_date': '20190930',
 664             'timestamp': 1569860870,
 665             'view_count': int,
 666             'like_count': int,
 667             'repost_count': int,
 668             'comment_count': int,
 669             'artists': ['Evan Todd', 'Jessica Keenan Wynn', 'Alice Lee', 'Barrett Wilbert Weed', 'Jon Eidson'],
 670             'track': 'Big Fun',
 671         },
 672     }, {
 673         # Banned audio, was available on the app, now works with web too
 674         'url': 'https://www.tiktok.com/@barudakhb_/video/6984138651336838402',
 675         'info_dict': {
 676             'id': '6984138651336838402',
 677             'ext': 'mp4',
 678             'title': 'Balas @yolaaftwsr hayu yu ? #SquadRandom_ 🔥',
 679             'description': 'Balas @yolaaftwsr hayu yu ? #SquadRandom_ 🔥',
 680             'uploader': 'barudakhb_',
 681             'channel': 'md5:29f238c49bc0c176cb3cef1a9cea9fa6',
 682             'uploader_id': '6974687867511718913',
 683             'uploader_url': 'https://www.tiktok.com/@barudakhb_',
 684             'channel_url': 'https://www.tiktok.com/@MS4wLjABAAAAbhBwQC-R1iKoix6jDFsF-vBdfx2ABoDjaZrM9fX6arU3w71q3cOWgWuTXn1soZ7d',
 685             'channel_id': 'MS4wLjABAAAAbhBwQC-R1iKoix6jDFsF-vBdfx2ABoDjaZrM9fX6arU3w71q3cOWgWuTXn1soZ7d',
 686             'track': 'Boka Dance',
 687             'artists': ['md5:29f238c49bc0c176cb3cef1a9cea9fa6'],
 688             'timestamp': 1626121503,
 689             'duration': 18,
 690             'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
 691             'upload_date': '20210712',
 692             'view_count': int,
 693             'like_count': int,
 694             'repost_count': int,
 695             'comment_count': int,
 696         },
 697     }, {
 698         # Sponsored video, only available with feed workaround
 699         'url': 'https://www.tiktok.com/@MS4wLjABAAAATh8Vewkn0LYM7Fo03iec3qKdeCUOcBIouRk1mkiag6h3o_pQu_dUXvZ2EZlGST7_/video/7042692929109986561',
 700         'info_dict': {
 701             'id': '7042692929109986561',
 702             'ext': 'mp4',
 703             'title': 'Slap and Run!',
 704             'description': 'Slap and Run!',
 705             'uploader': 'user440922249',
 706             'channel': 'Slap And Run',
 707             'uploader_id': '7036055384943690754',
 708             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAATh8Vewkn0LYM7Fo03iec3qKdeCUOcBIouRk1mkiag6h3o_pQu_dUXvZ2EZlGST7_',
 709             'channel_id': 'MS4wLjABAAAATh8Vewkn0LYM7Fo03iec3qKdeCUOcBIouRk1mkiag6h3o_pQu_dUXvZ2EZlGST7_',
 710             'track': 'Promoted Music',
 711             'timestamp': 1639754738,
 712             'duration': 30,
 713             'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
 714             'upload_date': '20211217',
 715             'view_count': int,
 716             'like_count': int,
 717             'repost_count': int,
 718             'comment_count': int,
 719         },
 720         'skip': 'This video is unavailable',
 721     }, {
 722         # Video without title and description
 723         'url': 'https://www.tiktok.com/@pokemonlife22/video/7059698374567611694',
 724         'info_dict': {
 725             'id': '7059698374567611694',
 726             'ext': 'mp4',
 727             'title': 'TikTok video #7059698374567611694',
 728             'description': '',
 729             'uploader': 'pokemonlife22',
 730             'channel': 'Pokemon',
 731             'uploader_id': '6820838815978423302',
 732             'uploader_url': 'https://www.tiktok.com/@pokemonlife22',
 733             'channel_url': 'https://www.tiktok.com/@MS4wLjABAAAA0tF1nBwQVVMyrGu3CqttkNgM68Do1OXUFuCY0CRQk8fEtSVDj89HqoqvbSTmUP2W',
 734             'channel_id': 'MS4wLjABAAAA0tF1nBwQVVMyrGu3CqttkNgM68Do1OXUFuCY0CRQk8fEtSVDj89HqoqvbSTmUP2W',
 735             'track': 'original sound',
 736             'timestamp': 1643714123,
 737             'duration': 6,
 738             'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
 739             'upload_date': '20220201',
 740             'artists': ['Pokemon'],
 741             'view_count': int,
 742             'like_count': int,
 743             'repost_count': int,
 744             'comment_count': int,
 745         },
 746     }, {
 747         # hydration JSON is sent in a <script> element
 748         'url': 'https://www.tiktok.com/@denidil6/video/7065799023130643713',
 749         'info_dict': {
 750             'id': '7065799023130643713',
 751             'ext': 'mp4',
 752             'title': '#denidil#денидил',
 753             'description': '#denidil#денидил',
 754             'uploader': 'denidil6',
 755             'uploader_id': '7046664115636405250',
 756             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAAsvMSzFdQ4ikl3uR2TEJwMBbB2yZh2Zxwhx-WCo3rbDpAharE3GQCrFuJArI3C8QJ',
 757             'artist': 'Holocron Music',
 758             'album': 'Wolf Sounds (1 Hour) Enjoy the Company of the Animal That Is the Majestic King of the Night',
 759             'track': 'Wolf Sounds (1 Hour) Enjoy the Company of the Animal That Is the Majestic King of the Night',
 760             'timestamp': 1645134536,
 761             'duration': 26,
 762             'upload_date': '20220217',
 763             'view_count': int,
 764             'like_count': int,
 765             'repost_count': int,
 766             'comment_count': int,
 767         },
 768         'skip': 'This video is unavailable',
 769     }, {
 770         # slideshow audio-only mp3 format
 771         'url': 'https://www.tiktok.com/@_le_cannibale_/video/7139980461132074283',
 772         'info_dict': {
 773             'id': '7139980461132074283',
 774             'ext': 'mp3',
 775             'title': 'TikTok video #7139980461132074283',
 776             'description': '',
 777             'channel': 'Antaura',
 778             'uploader': '_le_cannibale_',
 779             'uploader_id': '6604511138619654149',
 780             'uploader_url': 'https://www.tiktok.com/@_le_cannibale_',
 781             'channel_url': 'https://www.tiktok.com/@MS4wLjABAAAAoShJqaw_5gvy48y3azFeFcT4jeyKWbB0VVYasOCt2tTLwjNFIaDcHAM4D-QGXFOP',
 782             'channel_id': 'MS4wLjABAAAAoShJqaw_5gvy48y3azFeFcT4jeyKWbB0VVYasOCt2tTLwjNFIaDcHAM4D-QGXFOP',
 783             'artists': ['nathan !'],
 784             'track': 'grahamscott canon',
 785             'duration': 10,
 786             'upload_date': '20220905',
 787             'timestamp': 1662406249,
 788             'view_count': int,
 789             'like_count': int,
 790             'repost_count': int,
 791             'comment_count': int,
 792             'thumbnail': r're:^https://.+\.(?:webp|jpe?g)',
 793         },
 794     }, {
 795         # only available via web
 796         'url': 'https://www.tiktok.com/@moxypatch/video/7206382937372134662',
 797         'md5': '4cdefa501ac8ac20bf04986e10916fea',
 798         'info_dict': {
 799             'id': '7206382937372134662',
 800             'ext': 'mp4',
 801             'title': 'md5:1d95c0b96560ca0e8a231af4172b2c0a',
 802             'description': 'md5:1d95c0b96560ca0e8a231af4172b2c0a',
 803             'channel': 'MoxyPatch',
 804             'uploader': 'moxypatch',
 805             'uploader_id': '7039142049363379205',
 806             'uploader_url': 'https://www.tiktok.com/@moxypatch',
 807             'channel_url': 'https://www.tiktok.com/@MS4wLjABAAAAFhqKnngMHJSsifL0w1vFOP5kn3Ndo1ODp0XuIBkNMBCkALTvwILdpu12g3pTtL4V',
 808             'channel_id': 'MS4wLjABAAAAFhqKnngMHJSsifL0w1vFOP5kn3Ndo1ODp0XuIBkNMBCkALTvwILdpu12g3pTtL4V',
 809             'artists': ['your worst nightmare'],
 810             'track': 'original sound',
 811             'upload_date': '20230303',
 812             'timestamp': 1677866781,
 813             'duration': 10,
 814             'view_count': int,
 815             'like_count': int,
 816             'repost_count': int,
 817             'comment_count': int,
 818             'thumbnail': r're:^https://.+',
 819             'thumbnails': 'count:3',
 820         },
 821         'expected_warnings': ['Unable to find video in feed'],
 822     }, {
 823         # 1080p format
 824         'url': 'https://www.tiktok.com/@tatemcrae/video/7107337212743830830',  # FIXME: Web can only get audio
 825         'md5': '982512017a8a917124d5a08c8ae79621',
 826         'info_dict': {
 827             'id': '7107337212743830830',
 828             'ext': 'mp4',
 829             'title': 'new music video 4 don’t come backkkk🧸🖤 i hope u enjoy !! @musicontiktok',
 830             'description': 'new music video 4 don’t come backkkk🧸🖤 i hope u enjoy !! @musicontiktok',
 831             'uploader': 'tatemcrae',
 832             'uploader_id': '86328792343818240',
 833             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAA-0bQT0CqebTRr6I4IkYvMDMKSRSJHLNPBo5HrSklJwyA2psXLSZG5FP-LMNpHnJd',
 834             'channel_id': 'MS4wLjABAAAA-0bQT0CqebTRr6I4IkYvMDMKSRSJHLNPBo5HrSklJwyA2psXLSZG5FP-LMNpHnJd',
 835             'channel': 'tate mcrae',
 836             'artists': ['tate mcrae'],
 837             'track': 'original sound',
 838             'upload_date': '20220609',
 839             'timestamp': 1654805899,
 840             'duration': 150,
 841             'view_count': int,
 842             'like_count': int,
 843             'repost_count': int,
 844             'comment_count': int,
 845             'thumbnail': r're:^https://.+\.webp',
 846         },
 847         'skip': 'Unavailable via feed API, only audio available via web',
 848     }, {
 849         # Slideshow, audio-only m4a format
 850         'url': 'https://www.tiktok.com/@hara_yoimiya/video/7253412088251534594',
 851         'md5': '2ff8fe0174db2dbf49c597a7bef4e47d',
 852         'info_dict': {
 853             'id': '7253412088251534594',
 854             'ext': 'm4a',
 855             'title': 'я ред флаг простите #переписка #щитпост #тревожныйтиппривязанности #рекомендации ',
 856             'description': 'я ред флаг простите #переписка #щитпост #тревожныйтиппривязанности #рекомендации ',
 857             'uploader': 'hara_yoimiya',
 858             'uploader_id': '6582536342634676230',
 859             'uploader_url': 'https://www.tiktok.com/@hara_yoimiya',
 860             'channel_url': 'https://www.tiktok.com/@MS4wLjABAAAAIAlDxriiPWLE-p8p1R_0Bx8qWKfi-7zwmGhzU8Mv25W8sNxjfIKrol31qTczzuLB',
 861             'channel_id': 'MS4wLjABAAAAIAlDxriiPWLE-p8p1R_0Bx8qWKfi-7zwmGhzU8Mv25W8sNxjfIKrol31qTczzuLB',
 862             'channel': 'лампочка(!)',
 863             'artists': ['Øneheart'],
 864             'album': 'watching the stars',
 865             'track': 'watching the stars',
 866             'duration': 60,
 867             'upload_date': '20230708',
 868             'timestamp': 1688816612,
 869             'view_count': int,
 870             'like_count': int,
 871             'comment_count': int,
 872             'repost_count': int,
 873             'thumbnail': r're:^https://.+\.(?:webp|jpe?g)',
 874         },
 875     }, {
 876         # Auto-captions available
 877         'url': 'https://www.tiktok.com/@hankgreen1/video/7047596209028074758',
 878         'only_matching': True,
 879     }]
 880
 881     def _real_extract(self, url):
 882         video_id, user_id = self._match_valid_url(url).group('id', 'user_id')
 883
 884         if self._KNOWN_APP_INFO:
 885             try:
 886                 return self._extract_aweme_app(video_id)
 887             except ExtractorError as e:
 888                 e.expected = True
 889                 self.report_warning(f'{e}; trying with webpage')
 890
 891         url = self._create_url(user_id, video_id)
 892         video_data, status = self._extract_web_data_and_status(url, video_id)
 893
 894         if video_data and status == 0:
 895             return self._parse_aweme_video_web(video_data, url, video_id)
 896         elif status == 10216:
 897             raise ExtractorError('This video is private', expected=True)
 898         raise ExtractorError(f'Video not available, status code {status}', video_id=video_id)
 899
 900
 901 class TikTokUserIE(TikTokBaseIE):
 902     IE_NAME = 'tiktok:user'
 903     _VALID_URL = r'(?:tiktokuser:|https?://(?:www\.)?tiktok\.com/@)(?P<id>[\w.-]+)/?(?:$|[#?])'
 904     _TESTS = [{
 905         'url': 'https://tiktok.com/@corgibobaa?lang=en',
 906         'playlist_mincount': 45,
 907         'info_dict': {
 908             'id': 'MS4wLjABAAAAepiJKgwWhulvCpSuUVsp7sgVVsFJbbNaLeQ6OQ0oAJERGDUIXhb2yxxHZedsItgT',
 909             'title': 'corgibobaa',
 910         },
 911     }, {
 912         'url': 'https://www.tiktok.com/@6820838815978423302',
 913         'playlist_mincount': 5,
 914         'info_dict': {
 915             'id': 'MS4wLjABAAAA0tF1nBwQVVMyrGu3CqttkNgM68Do1OXUFuCY0CRQk8fEtSVDj89HqoqvbSTmUP2W',
 916             'title': '6820838815978423302',
 917         },
 918     }, {
 919         'url': 'https://www.tiktok.com/@meme',
 920         'playlist_mincount': 593,
 921         'info_dict': {
 922             'id': 'MS4wLjABAAAAiKfaDWeCsT3IHwY77zqWGtVRIy9v4ws1HbVi7auP1Vx7dJysU_hc5yRiGywojRD6',
 923             'title': 'meme',
 924         },
 925     }, {
 926         'url': 'tiktokuser:MS4wLjABAAAAM3R2BtjzVT-uAtstkl2iugMzC6AtnpkojJbjiOdDDrdsTiTR75-8lyWJCY5VvDrZ',
 927         'playlist_mincount': 31,
 928         'info_dict': {
 929             'id': 'MS4wLjABAAAAM3R2BtjzVT-uAtstkl2iugMzC6AtnpkojJbjiOdDDrdsTiTR75-8lyWJCY5VvDrZ',
 930         },
 931     }]
 932     _USER_AGENT = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:115.0) Gecko/20100101 Firefox/115.0'
 933     _API_BASE_URL = 'https://www.tiktok.com/api/creator/item_list/'
 934
 935     def _build_web_query(self, sec_uid, cursor):
 936         return {
 937             'aid': '1988',
 938             'app_language': 'en',
 939             'app_name': 'tiktok_web',
 940             'browser_language': 'en-US',
 941             'browser_name': 'Mozilla',
 942             'browser_online': 'true',
 943             'browser_platform': 'Win32',
 944             'browser_version': '5.0 (Windows)',
 945             'channel': 'tiktok_web',
 946             'cookie_enabled': 'true',
 947             'count': '15',
 948             'cursor': cursor,
 949             'device_id': self._DEVICE_ID,
 950             'device_platform': 'web_pc',
 951             'focus_state': 'true',
 952             'from_page': 'user',
 953             'history_len': '2',
 954             'is_fullscreen': 'false',
 955             'is_page_visible': 'true',
 956             'language': 'en',
 957             'os': 'windows',
 958             'priority_region': '',
 959             'referer': '',
 960             'region': 'US',
 961             'screen_height': '1080',
 962             'screen_width': '1920',
 963             'secUid': sec_uid,
 964             'type': '1',  # pagination type: 0 == oldest-to-newest, 1 == newest-to-oldest
 965             'tz_name': 'UTC',
 966             'verifyFp': f'verify_{"".join(random.choices(string.hexdigits, k=7))}',
 967             'webcast_language': 'en',
 968         }
 969
 970     def _entries(self, sec_uid, user_name):
 971         display_id = user_name or sec_uid
 972         seen_ids = set()
 973
 974         cursor = int(time.time() * 1E3)
 975         for page in itertools.count(1):
 976             response = self._download_json(
 977                 self._API_BASE_URL, display_id, f'Downloading page {page}',
 978                 query=self._build_web_query(sec_uid, cursor), headers={'User-Agent': self._USER_AGENT})
 979
 980             for video in traverse_obj(response, ('itemList', lambda _, v: v['id'])):
 981                 video_id = video['id']
 982                 if video_id in seen_ids:
 983                     continue
 984                 seen_ids.add(video_id)
 985                 webpage_url = self._create_url(display_id, video_id)
 986                 yield self.url_result(
 987                     webpage_url, TikTokIE,
 988                     **self._parse_aweme_video_web(video, webpage_url, video_id, extract_flat=True))
 989
 990             old_cursor = cursor
 991             cursor = traverse_obj(
 992                 response, ('itemList', -1, 'createTime', {lambda x: int(x * 1E3)}))
 993             if not cursor or old_cursor == cursor:
 994                 # User may not have posted within this ~1 week lookback, so manually adjust cursor
 995                 cursor = old_cursor - 7 * 86_400_000
 996             # In case 'hasMorePrevious' is wrong, break if we have gone back before TikTok existed
 997             if cursor < 1472706000000 or not traverse_obj(response, 'hasMorePrevious'):
 998                 break
 999
1000     def _get_sec_uid(self, user_url, user_name, msg):
1001         webpage = self._download_webpage(
1002             user_url, user_name, fatal=False, headers={'User-Agent': 'Mozilla/5.0'},
1003             note=f'Downloading {msg} webpage', errnote=f'Unable to download {msg} webpage') or ''
1004         return (traverse_obj(self._get_universal_data(webpage, user_name),
1005                              ('webapp.user-detail', 'userInfo', 'user', 'secUid', {str}))
1006                 or traverse_obj(self._get_sigi_state(webpage, user_name),
1007                                 ('LiveRoom', 'liveRoomUserInfo', 'user', 'secUid', {str}),
1008                                 ('UserModule', 'users', ..., 'secUid', {str}, any)))
1009
1010     def _real_extract(self, url):
1011         user_name, sec_uid = self._match_id(url), None
1012         if mobj := re.fullmatch(r'MS4wLjABAAAA[\w-]{64}', user_name):
1013             user_name, sec_uid = None, mobj.group(0)
1014         else:
1015             sec_uid = (self._get_sec_uid(self._UPLOADER_URL_FORMAT % user_name, user_name, 'user')
1016                        or self._get_sec_uid(self._UPLOADER_URL_FORMAT % f'{user_name}/live', user_name, 'live'))
1017
1018         if not sec_uid:
1019             webpage = self._download_webpage(
1020                 f'https://www.tiktok.com/embed/@{user_name}', user_name,
1021                 note='Downloading user embed page', fatal=False) or ''
1022             data = traverse_obj(self._search_json(
1023                 r'<script[^>]+\bid=[\'"]__FRONTITY_CONNECT_STATE__[\'"][^>]*>',
1024                 webpage, 'data', user_name, default={}),
1025                 ('source', 'data', f'/embed/@{user_name}', {dict}))
1026
1027             for aweme_id in traverse_obj(data, ('videoList', ..., 'id', {str})):
1028                 webpage_url = self._create_url(user_name, aweme_id)
1029                 video_data, _ = self._extract_web_data_and_status(webpage_url, aweme_id, fatal=False)
1030                 sec_uid = self._parse_aweme_video_web(
1031                     video_data, webpage_url, aweme_id, extract_flat=True).get('channel_id')
1032                 if sec_uid:
1033                     break
1034
1035         if not sec_uid:
1036             raise ExtractorError(
1037                 'Unable to extract secondary user ID. If you are able to get the channel_id '
1038                 'from a video posted by this user, try using "tiktokuser:channel_id" as the '
1039                 'input URL (replacing `channel_id` with its actual value)', expected=True)
1040
1041         return self.playlist_result(self._entries(sec_uid, user_name), sec_uid, user_name)
1042
1043
1044 class TikTokBaseListIE(TikTokBaseIE):  # XXX: Conventionally, base classes should end with BaseIE/InfoExtractor
1045     def _entries(self, list_id, display_id):
1046         query = {
1047             self._QUERY_NAME: list_id,
1048             'cursor': 0,
1049             'count': 20,
1050             'type': 5,
1051             'device_id': self._DEVICE_ID,
1052         }
1053
1054         for page in itertools.count(1):
1055             for retry in self.RetryManager():
1056                 try:
1057                     post_list = self._call_api(
1058                         self._API_ENDPOINT, display_id, query=query,
1059                         note=f'Downloading video list page {page}',
1060                         errnote='Unable to download video list')
1061                 except ExtractorError as e:
1062                     if isinstance(e.cause, json.JSONDecodeError) and e.cause.pos == 0:
1063                         retry.error = e
1064                         continue
1065                     raise
1066             for video in post_list.get('aweme_list', []):
1067                 yield {
1068                     **self._parse_aweme_video_app(video),
1069                     'extractor_key': TikTokIE.ie_key(),
1070                     'extractor': 'TikTok',
1071                     'webpage_url': f'https://tiktok.com/@_/video/{video["aweme_id"]}',
1072                 }
1073             if not post_list.get('has_more'):
1074                 break
1075             query['cursor'] = post_list['cursor']
1076
1077     def _real_extract(self, url):
1078         list_id = self._match_id(url)
1079         return self.playlist_result(self._entries(list_id, list_id), list_id)
1080
1081
1082 class TikTokSoundIE(TikTokBaseListIE):
1083     IE_NAME = 'tiktok:sound'
1084     _VALID_URL = r'https?://(?:www\.)?tiktok\.com/music/[\w\.-]+-(?P<id>[\d]+)[/?#&]?'
1085     _WORKING = False
1086     _QUERY_NAME = 'music_id'
1087     _API_ENDPOINT = 'music/aweme'
1088     _TESTS = [{
1089         'url': 'https://www.tiktok.com/music/Build-a-Btch-6956990112127585029?lang=en',
1090         'playlist_mincount': 100,
1091         'info_dict': {
1092             'id': '6956990112127585029',
1093         },
1094         'expected_warnings': ['Retrying'],
1095     }, {
1096         # Actual entries are less than listed video count
1097         'url': 'https://www.tiktok.com/music/jiefei-soap-remix-7036843036118469381',
1098         'playlist_mincount': 2182,
1099         'info_dict': {
1100             'id': '7036843036118469381',
1101         },
1102         'expected_warnings': ['Retrying'],
1103     }]
1104
1105
1106 class TikTokEffectIE(TikTokBaseListIE):
1107     IE_NAME = 'tiktok:effect'
1108     _VALID_URL = r'https?://(?:www\.)?tiktok\.com/sticker/[\w\.-]+-(?P<id>[\d]+)[/?#&]?'
1109     _WORKING = False
1110     _QUERY_NAME = 'sticker_id'
1111     _API_ENDPOINT = 'sticker/aweme'
1112     _TESTS = [{
1113         'url': 'https://www.tiktok.com/sticker/MATERIAL-GWOOORL-1258156',
1114         'playlist_mincount': 100,
1115         'info_dict': {
1116             'id': '1258156',
1117         },
1118         'expected_warnings': ['Retrying'],
1119     }, {
1120         # Different entries between mobile and web, depending on region
1121         'url': 'https://www.tiktok.com/sticker/Elf-Friend-479565',
1122         'only_matching': True,
1123     }]
1124
1125
1126 class TikTokTagIE(TikTokBaseListIE):
1127     IE_NAME = 'tiktok:tag'
1128     _VALID_URL = r'https?://(?:www\.)?tiktok\.com/tag/(?P<id>[^/?#&]+)'
1129     _WORKING = False
1130     _QUERY_NAME = 'ch_id'
1131     _API_ENDPOINT = 'challenge/aweme'
1132     _TESTS = [{
1133         'url': 'https://tiktok.com/tag/hello2018',
1134         'playlist_mincount': 39,
1135         'info_dict': {
1136             'id': '46294678',
1137             'title': 'hello2018',
1138         },
1139         'expected_warnings': ['Retrying'],
1140     }, {
1141         'url': 'https://tiktok.com/tag/fypシ?is_copy_url=0&is_from_webapp=v1',
1142         'only_matching': True,
1143     }]
1144
1145     def _real_extract(self, url):
1146         display_id = self._match_id(url)
1147         webpage = self._download_webpage(url, display_id, headers={
1148             'User-Agent': 'facebookexternalhit/1.1 (+http://www.facebook.com/externalhit_uatext.php)',
1149         })
1150         tag_id = self._html_search_regex(r'snssdk\d*://challenge/detail/(\d+)', webpage, 'tag ID')
1151         return self.playlist_result(self._entries(tag_id, display_id), tag_id, display_id)
1152
1153
1154 class TikTokCollectionIE(TikTokBaseIE):
1155     IE_NAME = 'tiktok:collection'
1156     _VALID_URL = r'https?://www\.tiktok\.com/@(?P<user_id>[\w.-]+)/collection/(?P<title>[^/?#]+)-(?P<id>\d+)/?(?:[?#]|$)'
1157     _TESTS = [{
1158         # playlist should have exactly 9 videos
1159         'url': 'https://www.tiktok.com/@imanoreotwe/collection/count-test-7371330159376370462',
1160         'info_dict': {
1161             'id': '7371330159376370462',
1162             'title': 'imanoreotwe-count-test',
1163         },
1164         'playlist_count': 9,
1165     }, {
1166         # tests returning multiple pages of a large collection
1167         'url': 'https://www.tiktok.com/@imanoreotwe/collection/%F0%9F%98%82-7111887189571160875',
1168         'info_dict': {
1169             'id': '7111887189571160875',
1170             'title': 'imanoreotwe-%F0%9F%98%82',
1171         },
1172         'playlist_mincount': 100,
1173     }]
1174     _API_BASE_URL = 'https://www.tiktok.com/api/collection/item_list/'
1175     _PAGE_COUNT = 30
1176
1177     def _build_web_query(self, collection_id, cursor):
1178         return {
1179             'aid': '1988',
1180             'collectionId': collection_id,
1181             'count': self._PAGE_COUNT,
1182             'cursor': cursor,
1183             'sourceType': '113',
1184         }
1185
1186     def _entries(self, collection_id):
1187         cursor = 0
1188         for page in itertools.count(1):
1189             response = self._download_json(
1190                 self._API_BASE_URL, collection_id, f'Downloading page {page}',
1191                 query=self._build_web_query(collection_id, cursor))
1192
1193             for video in traverse_obj(response, ('itemList', lambda _, v: v['id'])):
1194                 video_id = video['id']
1195                 author = traverse_obj(video, ('author', ('uniqueId', 'secUid', 'id'), {str}, any)) or '_'
1196                 webpage_url = self._create_url(author, video_id)
1197                 yield self.url_result(
1198                     webpage_url, TikTokIE,
1199                     **self._parse_aweme_video_web(video, webpage_url, video_id, extract_flat=True))
1200
1201             if not traverse_obj(response, 'hasMore'):
1202                 break
1203             cursor += self._PAGE_COUNT
1204
1205     def _real_extract(self, url):
1206         collection_id, title, user_name = self._match_valid_url(url).group('id', 'title', 'user_id')
1207
1208         return self.playlist_result(
1209             self._entries(collection_id), collection_id, '-'.join((user_name, title)))
1210
1211
1212 class DouyinIE(TikTokBaseIE):
1213     _VALID_URL = r'https?://(?:www\.)?douyin\.com/video/(?P<id>[0-9]+)'
1214     _TESTS = [{
1215         'url': 'https://www.douyin.com/video/6961737553342991651',
1216         'md5': '9ecce7bc5b302601018ecb2871c63a75',
1217         'info_dict': {
1218             'id': '6961737553342991651',
1219             'ext': 'mp4',
1220             'title': '#杨超越  小小水手带你去远航❤️',
1221             'description': '#杨超越  小小水手带你去远航❤️',
1222             'uploader': '6897520xka',
1223             'uploader_id': '110403406559',
1224             'uploader_url': 'https://www.douyin.com/user/MS4wLjABAAAAEKnfa654JAJ_N5lgZDQluwsxmY0lhfmEYNQBBkwGG98',
1225             'channel_id': 'MS4wLjABAAAAEKnfa654JAJ_N5lgZDQluwsxmY0lhfmEYNQBBkwGG98',
1226             'channel': '杨超越',
1227             'duration': 19,
1228             'timestamp': 1620905839,
1229             'upload_date': '20210513',
1230             'track': '@杨超越创作的原声',
1231             'artists': ['杨超越'],
1232             'view_count': int,
1233             'like_count': int,
1234             'repost_count': int,
1235             'comment_count': int,
1236             'thumbnail': r're:https?://.+\.jpe?g',
1237         },
1238     }, {
1239         'url': 'https://www.douyin.com/video/6982497745948921092',
1240         'md5': '15c5e660b7048af3707304e3cc02bbb5',
1241         'info_dict': {
1242             'id': '6982497745948921092',
1243             'ext': 'mp4',
1244             'title': '这个夏日和小羊@杨超越 一起遇见白色幻想',
1245             'description': '这个夏日和小羊@杨超越 一起遇见白色幻想',
1246             'uploader': '0731chaoyue',
1247             'uploader_id': '408654318141572',
1248             'uploader_url': 'https://www.douyin.com/user/MS4wLjABAAAAZJpnglcjW2f_CMVcnqA_6oVBXKWMpH0F8LIHuUu8-lA',
1249             'channel_id': 'MS4wLjABAAAAZJpnglcjW2f_CMVcnqA_6oVBXKWMpH0F8LIHuUu8-lA',
1250             'channel': '杨超越工作室',
1251             'duration': 42,
1252             'timestamp': 1625739481,
1253             'upload_date': '20210708',
1254             'track': '@杨超越工作室创作的原声',
1255             'artists': ['杨超越工作室'],
1256             'view_count': int,
1257             'like_count': int,
1258             'repost_count': int,
1259             'comment_count': int,
1260             'thumbnail': r're:https?://.+\.jpe?g',
1261         },
1262     }, {
1263         'url': 'https://www.douyin.com/video/6953975910773099811',
1264         'md5': '0e6443758b8355db9a3c34864a4276be',
1265         'info_dict': {
1266             'id': '6953975910773099811',
1267             'ext': 'mp4',
1268             'title': '#一起看海  出现在你的夏日里',
1269             'description': '#一起看海  出现在你的夏日里',
1270             'uploader': '6897520xka',
1271             'uploader_id': '110403406559',
1272             'uploader_url': 'https://www.douyin.com/user/MS4wLjABAAAAEKnfa654JAJ_N5lgZDQluwsxmY0lhfmEYNQBBkwGG98',
1273             'channel_id': 'MS4wLjABAAAAEKnfa654JAJ_N5lgZDQluwsxmY0lhfmEYNQBBkwGG98',
1274             'channel': '杨超越',
1275             'duration': 17,
1276             'timestamp': 1619098692,
1277             'upload_date': '20210422',
1278             'track': '@杨超越创作的原声',
1279             'artists': ['杨超越'],
1280             'view_count': int,
1281             'like_count': int,
1282             'repost_count': int,
1283             'comment_count': int,
1284             'thumbnail': r're:https?://.+\.jpe?g',
1285         },
1286     }, {
1287         'url': 'https://www.douyin.com/video/6950251282489675042',
1288         'md5': 'b4db86aec367ef810ddd38b1737d2fed',
1289         'info_dict': {
1290             'id': '6950251282489675042',
1291             'ext': 'mp4',
1292             'title': '哈哈哈，成功了哈哈哈哈哈哈',
1293             'uploader': '杨超越',
1294             'upload_date': '20210412',
1295             'timestamp': 1618231483,
1296             'uploader_id': '110403406559',
1297             'view_count': int,
1298             'like_count': int,
1299             'repost_count': int,
1300             'comment_count': int,
1301         },
1302         'skip': 'No longer available',
1303     }, {
1304         'url': 'https://www.douyin.com/video/6963263655114722595',
1305         'md5': '1440bcf59d8700f8e014da073a4dfea8',
1306         'info_dict': {
1307             'id': '6963263655114722595',
1308             'ext': 'mp4',
1309             'title': '#哪个爱豆的105度最甜 换个角度看看我哈哈',
1310             'description': '#哪个爱豆的105度最甜 换个角度看看我哈哈',
1311             'uploader': '6897520xka',
1312             'uploader_id': '110403406559',
1313             'uploader_url': 'https://www.douyin.com/user/MS4wLjABAAAAEKnfa654JAJ_N5lgZDQluwsxmY0lhfmEYNQBBkwGG98',
1314             'channel_id': 'MS4wLjABAAAAEKnfa654JAJ_N5lgZDQluwsxmY0lhfmEYNQBBkwGG98',
1315             'channel': '杨超越',
1316             'duration': 15,
1317             'timestamp': 1621261163,
1318             'upload_date': '20210517',
1319             'track': '@杨超越创作的原声',
1320             'artists': ['杨超越'],
1321             'view_count': int,
1322             'like_count': int,
1323             'repost_count': int,
1324             'comment_count': int,
1325             'thumbnail': r're:https?://.+\.jpe?g',
1326         },
1327     }]
1328     _UPLOADER_URL_FORMAT = 'https://www.douyin.com/user/%s'
1329     _WEBPAGE_HOST = 'https://www.douyin.com/'
1330
1331     def _real_extract(self, url):
1332         video_id = self._match_id(url)
1333
1334         detail = traverse_obj(self._download_json(
1335             'https://www.douyin.com/aweme/v1/web/aweme/detail/', video_id,
1336             'Downloading web detail JSON', 'Failed to download web detail JSON',
1337             query={'aweme_id': video_id}, fatal=False), ('aweme_detail', {dict}))
1338         if not detail:
1339             # TODO: Run verification challenge code to generate signature cookies
1340             raise ExtractorError(
1341                 'Fresh cookies (not necessarily logged in) are needed',
1342                 expected=not self._get_cookies(self._WEBPAGE_HOST).get('s_v_web_id'))
1343
1344         return self._parse_aweme_video_app(detail)
1345
1346
1347 class TikTokVMIE(InfoExtractor):
1348     _VALID_URL = r'https?://(?:(?:vm|vt)\.tiktok\.com|(?:www\.)tiktok\.com/t)/(?P<id>\w+)'
1349     IE_NAME = 'vm.tiktok'
1350
1351     _TESTS = [{
1352         'url': 'https://www.tiktok.com/t/ZTRC5xgJp',
1353         'info_dict': {
1354             'id': '7170520270497680683',
1355             'ext': 'mp4',
1356             'title': 'md5:c64f6152330c2efe98093ccc8597871c',
1357             'uploader_id': '6687535061741700102',
1358             'upload_date': '20221127',
1359             'view_count': int,
1360             'like_count': int,
1361             'comment_count': int,
1362             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAAObqu3WCTXxmw2xwZ3iLEHnEecEIw7ks6rxWqOqOhaPja9BI7gqUQnjw8_5FSoDXX',
1363             'album': 'Wave of Mutilation: Best of Pixies',
1364             'thumbnail': r're:https://.+\.webp.*',
1365             'duration': 5,
1366             'timestamp': 1669516858,
1367             'repost_count': int,
1368             'artist': 'Pixies',
1369             'track': 'Where Is My Mind?',
1370             'description': 'md5:c64f6152330c2efe98093ccc8597871c',
1371             'uploader': 'sigmachaddeus',
1372             'creator': 'SigmaChad',
1373         },
1374     }, {
1375         'url': 'https://vm.tiktok.com/ZTR45GpSF/',
1376         'info_dict': {
1377             'id': '7106798200794926362',
1378             'ext': 'mp4',
1379             'title': 'md5:edc3e7ea587847f8537468f2fe51d074',
1380             'uploader_id': '6997695878846268418',
1381             'upload_date': '20220608',
1382             'view_count': int,
1383             'like_count': int,
1384             'comment_count': int,
1385             'thumbnail': r're:https://.+\.webp.*',
1386             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAAdZ_NcPPgMneaGrW0hN8O_J_bwLshwNNERRF5DxOw2HKIzk0kdlLrR8RkVl1ksrMO',
1387             'duration': 29,
1388             'timestamp': 1654680400,
1389             'repost_count': int,
1390             'artist': 'Akihitoko',
1391             'track': 'original sound',
1392             'description': 'md5:edc3e7ea587847f8537468f2fe51d074',
1393             'uploader': 'akihitoko1',
1394             'creator': 'Akihitoko',
1395         },
1396     }, {
1397         'url': 'https://vt.tiktok.com/ZSe4FqkKd',
1398         'only_matching': True,
1399     }]
1400
1401     def _real_extract(self, url):
1402         new_url = self._request_webpage(
1403             HEADRequest(url), self._match_id(url), headers={'User-Agent': 'facebookexternalhit/1.1'}).url
1404         if self.suitable(new_url):  # Prevent infinite loop in case redirect fails
1405             raise UnsupportedError(new_url)
1406         return self.url_result(new_url)
1407
1408
1409 class TikTokLiveIE(TikTokBaseIE):
1410     _VALID_URL = r'''(?x)https?://(?:
1411         (?:www\.)?tiktok\.com/@(?P<uploader>[\w.-]+)/live|
1412         m\.tiktok\.com/share/live/(?P<id>\d+)
1413     )'''
1414     IE_NAME = 'tiktok:live'
1415
1416     _TESTS = [{
1417         'url': 'https://www.tiktok.com/@weathernewslive/live',
1418         'info_dict': {
1419             'id': '7210809319192726273',
1420             'ext': 'mp4',
1421             'title': r're:ウェザーニュースLiVE[\d\s:-]*',
1422             'creator': 'ウェザーニュースLiVE',
1423             'uploader': 'weathernewslive',
1424             'uploader_id': '6621496731283095554',
1425             'uploader_url': 'https://www.tiktok.com/@weathernewslive',
1426             'live_status': 'is_live',
1427             'concurrent_view_count': int,
1428         },
1429         'params': {'skip_download': 'm3u8'},
1430     }, {
1431         'url': 'https://www.tiktok.com/@pilarmagenta/live',
1432         'info_dict': {
1433             'id': '7209423610325322522',
1434             'ext': 'mp4',
1435             'title': str,
1436             'creator': 'Pilarmagenta',
1437             'uploader': 'pilarmagenta',
1438             'uploader_id': '6624846890674683909',
1439             'uploader_url': 'https://www.tiktok.com/@pilarmagenta',
1440             'live_status': 'is_live',
1441             'concurrent_view_count': int,
1442         },
1443         'skip': 'Livestream',
1444     }, {
1445         'url': 'https://m.tiktok.com/share/live/7209423610325322522/?language=en',
1446         'only_matching': True,
1447     }, {
1448         'url': 'https://www.tiktok.com/@iris04201/live',
1449         'only_matching': True,
1450     }]
1451
1452     def _call_api(self, url, param, room_id, uploader, key=None):
1453         response = traverse_obj(self._download_json(
1454             url, room_id, fatal=False, query={
1455                 'aid': '1988',
1456                 param: room_id,
1457             }), (key, {dict}), default={})
1458
1459         # status == 2 if live else 4
1460         if int_or_none(response.get('status')) == 2:
1461             return response
1462         # If room_id is obtained via mobile share URL and cannot be refreshed, do not wait for live
1463         elif not uploader:
1464             raise ExtractorError('This livestream has ended', expected=True)
1465         raise UserNotLive(video_id=uploader)
1466
1467     def _real_extract(self, url):
1468         uploader, room_id = self._match_valid_url(url).group('uploader', 'id')
1469         webpage = self._download_webpage(
1470             url, uploader or room_id, headers={'User-Agent': 'Mozilla/5.0'}, fatal=not room_id)
1471
1472         if webpage:
1473             data = self._get_sigi_state(webpage, uploader or room_id)
1474             room_id = (
1475                 traverse_obj(data, ((
1476                     ('LiveRoom', 'liveRoomUserInfo', 'user'),
1477                     ('UserModule', 'users', ...)), 'roomId', {str}, any))
1478                 or self._search_regex(r'snssdk\d*://live\?room_id=(\d+)', webpage, 'room ID', default=room_id))
1479             uploader = uploader or traverse_obj(
1480                 data, ('LiveRoom', 'liveRoomUserInfo', 'user', 'uniqueId'),
1481                 ('UserModule', 'users', ..., 'uniqueId'), get_all=False, expected_type=str)
1482
1483         if not room_id:
1484             raise UserNotLive(video_id=uploader)
1485
1486         formats = []
1487         live_info = self._call_api(
1488             'https://webcast.tiktok.com/webcast/room/info', 'room_id', room_id, uploader, key='data')
1489
1490         get_quality = qualities(('SD1', 'ld', 'SD2', 'sd', 'HD1', 'hd', 'FULL_HD1', 'uhd', 'ORIGION', 'origin'))
1491         parse_inner = lambda x: self._parse_json(x, None)
1492
1493         for quality, stream in traverse_obj(live_info, (
1494                 'stream_url', 'live_core_sdk_data', 'pull_data', 'stream_data',
1495                 {parse_inner}, 'data', {dict}), default={}).items():
1496
1497             sdk_params = traverse_obj(stream, ('main', 'sdk_params', {parse_inner}, {
1498                 'vcodec': ('VCodec', {str}),
1499                 'tbr': ('vbitrate', {lambda x: int_or_none(x, 1000)}),
1500                 'resolution': ('resolution', {lambda x: re.match(r'(?i)\d+x\d+|\d+p', x).group().lower()}),
1501             }))
1502
1503             flv_url = traverse_obj(stream, ('main', 'flv', {url_or_none}))
1504             if flv_url:
1505                 formats.append({
1506                     'url': flv_url,
1507                     'ext': 'flv',
1508                     'format_id': f'flv-{quality}',
1509                     'quality': get_quality(quality),
1510                     **sdk_params,
1511                 })
1512
1513             hls_url = traverse_obj(stream, ('main', 'hls', {url_or_none}))
1514             if hls_url:
1515                 formats.append({
1516                     'url': hls_url,
1517                     'ext': 'mp4',
1518                     'protocol': 'm3u8_native',
1519                     'format_id': f'hls-{quality}',
1520                     'quality': get_quality(quality),
1521                     **sdk_params,
1522                 })
1523
1524         def get_vcodec(*keys):
1525             return traverse_obj(live_info, (
1526                 'stream_url', *keys, {parse_inner}, 'VCodec', {str}))
1527
1528         for stream in ('hls', 'rtmp'):
1529             stream_url = traverse_obj(live_info, ('stream_url', f'{stream}_pull_url', {url_or_none}))
1530             if stream_url:
1531                 formats.append({
1532                     'url': stream_url,
1533                     'ext': 'mp4' if stream == 'hls' else 'flv',
1534                     'protocol': 'm3u8_native' if stream == 'hls' else 'https',
1535                     'format_id': f'{stream}-pull',
1536                     'vcodec': get_vcodec(f'{stream}_pull_url_params'),
1537                     'quality': get_quality('ORIGION'),
1538                 })
1539
1540         for f_id, f_url in traverse_obj(live_info, ('stream_url', 'flv_pull_url', {dict}), default={}).items():
1541             if not url_or_none(f_url):
1542                 continue
1543             formats.append({
1544                 'url': f_url,
1545                 'ext': 'flv',
1546                 'format_id': f'flv-{f_id}'.lower(),
1547                 'vcodec': get_vcodec('flv_pull_url_params', f_id),
1548                 'quality': get_quality(f_id),
1549             })
1550
1551         # If uploader is a guest on another's livestream, primary endpoint will not have m3u8 URLs
1552         if not traverse_obj(formats, lambda _, v: v['ext'] == 'mp4'):
1553             live_info = merge_dicts(live_info, self._call_api(
1554                 'https://www.tiktok.com/api/live/detail/', 'roomID', room_id, uploader, key='LiveRoomInfo'))
1555             if url_or_none(live_info.get('liveUrl')):
1556                 formats.append({
1557                     'url': live_info['liveUrl'],
1558                     'ext': 'mp4',
1559                     'protocol': 'm3u8_native',
1560                     'format_id': 'hls-fallback',
1561                     'vcodec': 'h264',
1562                     'quality': get_quality('origin'),
1563                 })
1564
1565         uploader = uploader or traverse_obj(live_info, ('ownerInfo', 'uniqueId'), ('owner', 'display_id'))
1566
1567         return {
1568             'id': room_id,
1569             'uploader': uploader,
1570             'uploader_url': format_field(uploader, None, self._UPLOADER_URL_FORMAT) or None,
1571             'is_live': True,
1572             'formats': formats,
1573             '_format_sort_fields': ('quality', 'ext'),
1574             **traverse_obj(live_info, {
1575                 'title': 'title',
1576                 'uploader_id': (('ownerInfo', 'owner'), 'id', {str_or_none}),
1577                 'creator': (('ownerInfo', 'owner'), 'nickname'),
1578                 'concurrent_view_count': (('user_count', ('liveRoomStats', 'userCount')), {int_or_none}),
1579             }, get_all=False),
1580         }