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     remove_start,
  27     srt_subtitles_timecode,
  28     str_or_none,
  29     traverse_obj,
  30     try_call,
  31     try_get,
  32     url_or_none,
  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': '34.1.2',
  47         'manifest_app_version': '2023401020',
  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, query, video_id, 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             }, query=query)
 129
 130     def _build_api_query(self, query):
 131         return filter_dict({
 132             **query,
 133             'device_platform': 'android',
 134             'os': 'android',
 135             'ssmix': 'a',
 136             '_rticket': int(time.time() * 1000),
 137             'cdid': str(uuid.uuid4()),
 138             'channel': 'googleplay',
 139             'aid': self._APP_INFO['aid'],
 140             'app_name': self._APP_INFO['app_name'],
 141             'version_code': ''.join(f'{int(v):02d}' for v in self._APP_INFO['app_version'].split('.')),
 142             'version_name': self._APP_INFO['app_version'],
 143             'manifest_version_code': self._APP_INFO['manifest_app_version'],
 144             'update_version_code': self._APP_INFO['manifest_app_version'],
 145             'ab_version': self._APP_INFO['app_version'],
 146             'resolution': '1080*2400',
 147             'dpi': 420,
 148             'device_type': 'Pixel 7',
 149             'device_brand': 'Google',
 150             'language': 'en',
 151             'os_api': '29',
 152             'os_version': '13',
 153             'ac': 'wifi',
 154             'is_pad': '0',
 155             'current_region': 'US',
 156             'app_type': 'normal',
 157             'sys_region': 'US',
 158             'last_install_time': int(time.time()) - random.randint(86400, 1123200),
 159             'timezone_name': 'America/New_York',
 160             'residence': 'US',
 161             'app_language': 'en',
 162             'timezone_offset': '-14400',
 163             'host_abi': 'armeabi-v7a',
 164             'locale': 'en',
 165             'ac2': 'wifi5g',
 166             'uoo': '1',
 167             'carrier_region': 'US',
 168             'op_region': 'US',
 169             'build_number': self._APP_INFO['app_version'],
 170             'region': 'US',
 171             'ts': int(time.time()),
 172             'iid': self._APP_INFO.get('iid'),
 173             'device_id': self._DEVICE_ID,
 174             'openudid': ''.join(random.choices('0123456789abcdef', k=16)),
 175         })
 176
 177     def _call_api(self, ep, query, video_id, fatal=True,
 178                   note='Downloading API JSON', errnote='Unable to download API page'):
 179         if not self._APP_INFO and not self._get_next_app_info():
 180             message = 'No working app info is available'
 181             if fatal:
 182                 raise ExtractorError(message, expected=True)
 183             else:
 184                 self.report_warning(message)
 185                 return
 186
 187         max_tries = len(self._APP_INFO_POOL) + 1  # _APP_INFO_POOL + _APP_INFO
 188         for count in itertools.count(1):
 189             self.write_debug(str(self._APP_INFO))
 190             real_query = self._build_api_query(query)
 191             try:
 192                 return self._call_api_impl(ep, real_query, video_id, fatal, note, errnote)
 193             except ExtractorError as e:
 194                 if isinstance(e.cause, json.JSONDecodeError) and e.cause.pos == 0:
 195                     message = str(e.cause or e.msg)
 196                     if not self._get_next_app_info():
 197                         if fatal:
 198                             raise
 199                         else:
 200                             self.report_warning(message)
 201                             return
 202                     self.report_warning(f'{message}. Retrying... (attempt {count} of {max_tries})')
 203                     continue
 204                 raise
 205
 206     def _extract_aweme_app(self, aweme_id):
 207         feed_list = self._call_api(
 208             'feed', {'aweme_id': aweme_id}, aweme_id, note='Downloading video feed',
 209             errnote='Unable to download video feed').get('aweme_list') or []
 210         aweme_detail = next((aweme for aweme in feed_list if str(aweme.get('aweme_id')) == aweme_id), None)
 211         if not aweme_detail:
 212             raise ExtractorError('Unable to find video in feed', video_id=aweme_id)
 213         return self._parse_aweme_video_app(aweme_detail)
 214
 215     def _extract_web_data_and_status(self, url, video_id, fatal=True):
 216         video_data, status = {}, -1
 217
 218         res = self._download_webpage_handle(url, video_id, fatal=fatal, headers={'User-Agent': 'Mozilla/5.0'})
 219         if res is False:
 220             return video_data, status
 221
 222         webpage, urlh = res
 223         if urllib.parse.urlparse(urlh.url).path == '/login':
 224             message = 'TikTok is requiring login for access to this content'
 225             if fatal:
 226                 self.raise_login_required(message)
 227             self.report_warning(f'{message}. {self._login_hint()}')
 228             return video_data, status
 229
 230         if universal_data := self._get_universal_data(webpage, video_id):
 231             self.write_debug('Found universal data for rehydration')
 232             status = traverse_obj(universal_data, ('webapp.video-detail', 'statusCode', {int})) or 0
 233             video_data = traverse_obj(universal_data, ('webapp.video-detail', 'itemInfo', 'itemStruct', {dict}))
 234
 235         elif sigi_data := self._get_sigi_state(webpage, video_id):
 236             self.write_debug('Found sigi state data')
 237             status = traverse_obj(sigi_data, ('VideoPage', 'statusCode', {int})) or 0
 238             video_data = traverse_obj(sigi_data, ('ItemModule', video_id, {dict}))
 239
 240         elif next_data := self._search_nextjs_data(webpage, video_id, default={}):
 241             self.write_debug('Found next.js data')
 242             status = traverse_obj(next_data, ('props', 'pageProps', 'statusCode', {int})) or 0
 243             video_data = traverse_obj(next_data, ('props', 'pageProps', 'itemInfo', 'itemStruct', {dict}))
 244
 245         elif fatal:
 246             raise ExtractorError('Unable to extract webpage video data')
 247
 248         return video_data, status
 249
 250     def _get_subtitles(self, aweme_detail, aweme_id, user_name):
 251         # TODO: Extract text positioning info
 252         subtitles = {}
 253         # aweme/detail endpoint subs
 254         captions_info = traverse_obj(
 255             aweme_detail, ('interaction_stickers', ..., 'auto_video_caption_info', 'auto_captions', ...), expected_type=dict)
 256         for caption in captions_info:
 257             caption_url = traverse_obj(caption, ('url', 'url_list', ...), expected_type=url_or_none, get_all=False)
 258             if not caption_url:
 259                 continue
 260             caption_json = self._download_json(
 261                 caption_url, aweme_id, note='Downloading captions', errnote='Unable to download captions', fatal=False)
 262             if not caption_json:
 263                 continue
 264             subtitles.setdefault(caption.get('language', 'en'), []).append({
 265                 'ext': 'srt',
 266                 'data': '\n\n'.join(
 267                     f'{i + 1}\n{srt_subtitles_timecode(line["start_time"] / 1000)} --> {srt_subtitles_timecode(line["end_time"] / 1000)}\n{line["text"]}'
 268                     for i, line in enumerate(caption_json['utterances']) if line.get('text')),
 269             })
 270         # feed endpoint subs
 271         if not subtitles:
 272             for caption in traverse_obj(aweme_detail, ('video', 'cla_info', 'caption_infos', ...), expected_type=dict):
 273                 if not caption.get('url'):
 274                     continue
 275                 subtitles.setdefault(caption.get('lang') or 'en', []).append({
 276                     'ext': remove_start(caption.get('caption_format'), 'web'),
 277                     'url': caption['url'],
 278                 })
 279         # webpage subs
 280         if not subtitles:
 281             if user_name:  # only _parse_aweme_video_app needs to extract the webpage here
 282                 aweme_detail, _ = self._extract_web_data_and_status(
 283                     self._create_url(user_name, aweme_id), aweme_id, fatal=False)
 284             for caption in traverse_obj(aweme_detail, ('video', 'subtitleInfos', lambda _, v: v['Url'])):
 285                 subtitles.setdefault(caption.get('LanguageCodeName') or 'en', []).append({
 286                     'ext': remove_start(caption.get('Format'), 'web'),
 287                     'url': caption['Url'],
 288                 })
 289         return subtitles
 290
 291     def _parse_url_key(self, url_key):
 292         format_id, codec, res, bitrate = self._search_regex(
 293             r'v[^_]+_(?P<id>(?P<codec>[^_]+)_(?P<res>\d+p)_(?P<bitrate>\d+))', url_key,
 294             'url key', default=(None, None, None, None), group=('id', 'codec', 'res', 'bitrate'))
 295         if not format_id:
 296             return {}, None
 297         return {
 298             'format_id': format_id,
 299             'vcodec': 'h265' if codec == 'bytevc1' else codec,
 300             'tbr': int_or_none(bitrate, scale=1000) or None,
 301             'quality': qualities(self.QUALITIES)(res),
 302         }, res
 303
 304     def _parse_aweme_video_app(self, aweme_detail):
 305         aweme_id = aweme_detail['aweme_id']
 306         video_info = aweme_detail['video']
 307         known_resolutions = {}
 308
 309         def audio_meta(url):
 310             ext = determine_ext(url, default_ext='m4a')
 311             return {
 312                 'format_note': 'Music track',
 313                 'ext': ext,
 314                 'acodec': 'aac' if ext == 'm4a' else ext,
 315                 'vcodec': 'none',
 316                 'width': None,
 317                 'height': None,
 318             } if ext == 'mp3' or '-music-' in url else {}
 319
 320         def extract_addr(addr, add_meta={}):
 321             parsed_meta, res = self._parse_url_key(addr.get('url_key', ''))
 322             is_bytevc2 = parsed_meta.get('vcodec') == 'bytevc2'
 323             if res:
 324                 known_resolutions.setdefault(res, {}).setdefault('height', int_or_none(addr.get('height')))
 325                 known_resolutions[res].setdefault('width', int_or_none(addr.get('width')))
 326                 parsed_meta.update(known_resolutions.get(res, {}))
 327                 add_meta.setdefault('height', int_or_none(res[:-1]))
 328             return [{
 329                 'url': url,
 330                 'filesize': int_or_none(addr.get('data_size')),
 331                 'ext': 'mp4',
 332                 'acodec': 'aac',
 333                 'source_preference': -2 if 'aweme/v1' in url else -1,  # Downloads from API might get blocked
 334                 **add_meta, **parsed_meta,
 335                 # bytevc2 is bytedance's own custom h266/vvc codec, as-of-yet unplayable
 336                 'preference': -100 if is_bytevc2 else -1,
 337                 'format_note': join_nonempty(
 338                     add_meta.get('format_note'), '(API)' if 'aweme/v1' in url else None,
 339                     '(UNPLAYABLE)' if is_bytevc2 else None, delim=' '),
 340                 **audio_meta(url),
 341             } for url in addr.get('url_list') or []]
 342
 343         # Hack: Add direct video links first to prioritize them when removing duplicate formats
 344         formats = []
 345         width = int_or_none(video_info.get('width'))
 346         height = int_or_none(video_info.get('height'))
 347         ratio = try_call(lambda: width / height) or 0.5625
 348         if video_info.get('play_addr'):
 349             formats.extend(extract_addr(video_info['play_addr'], {
 350                 'format_id': 'play_addr',
 351                 'format_note': 'Direct video',
 352                 'vcodec': 'h265' if traverse_obj(
 353                     video_info, 'is_bytevc1', 'is_h265') else 'h264',  # TODO: Check for "direct iOS" videos, like https://www.tiktok.com/@cookierun_dev/video/7039716639834656002
 354                 'width': width,
 355                 'height': height,
 356             }))
 357         if video_info.get('download_addr'):
 358             download_addr = video_info['download_addr']
 359             dl_width = int_or_none(download_addr.get('width'))
 360             formats.extend(extract_addr(download_addr, {
 361                 'format_id': 'download_addr',
 362                 'format_note': 'Download video%s' % (', watermarked' if video_info.get('has_watermark') else ''),
 363                 'vcodec': 'h264',
 364                 'width': dl_width,
 365                 'height': try_call(lambda: int(dl_width / ratio)),  # download_addr['height'] is wrong
 366                 'preference': -2 if video_info.get('has_watermark') else -1,
 367             }))
 368         if video_info.get('play_addr_h264'):
 369             formats.extend(extract_addr(video_info['play_addr_h264'], {
 370                 'format_id': 'play_addr_h264',
 371                 'format_note': 'Direct video',
 372                 'vcodec': 'h264',
 373             }))
 374         if video_info.get('play_addr_bytevc1'):
 375             formats.extend(extract_addr(video_info['play_addr_bytevc1'], {
 376                 'format_id': 'play_addr_bytevc1',
 377                 'format_note': 'Direct video',
 378                 'vcodec': 'h265',
 379             }))
 380
 381         for bitrate in video_info.get('bit_rate', []):
 382             if bitrate.get('play_addr'):
 383                 formats.extend(extract_addr(bitrate['play_addr'], {
 384                     'format_id': bitrate.get('gear_name'),
 385                     'format_note': 'Playback video',
 386                     'tbr': try_get(bitrate, lambda x: x['bit_rate'] / 1000),
 387                     'vcodec': 'h265' if traverse_obj(
 388                         bitrate, 'is_bytevc1', 'is_h265') else 'h264',
 389                     'fps': bitrate.get('FPS'),
 390                 }))
 391
 392         self._remove_duplicate_formats(formats)
 393         auth_cookie = self._get_cookies(self._WEBPAGE_HOST).get('sid_tt')
 394         if auth_cookie:
 395             for f in formats:
 396                 self._set_cookie(urllib.parse.urlparse(f['url']).hostname, 'sid_tt', auth_cookie.value)
 397
 398         thumbnails = []
 399         for cover_id in ('cover', 'ai_dynamic_cover', 'animated_cover', 'ai_dynamic_cover_bak',
 400                          'origin_cover', 'dynamic_cover'):
 401             for cover_url in traverse_obj(video_info, (cover_id, 'url_list', ...)):
 402                 thumbnails.append({
 403                     'id': cover_id,
 404                     'url': cover_url,
 405                 })
 406
 407         stats_info = aweme_detail.get('statistics') or {}
 408         music_info = aweme_detail.get('music') or {}
 409         labels = traverse_obj(aweme_detail, ('hybrid_label', ..., 'text'), expected_type=str)
 410
 411         contained_music_track = traverse_obj(
 412             music_info, ('matched_song', 'title'), ('matched_pgc_sound', 'title'), expected_type=str)
 413         contained_music_author = traverse_obj(
 414             music_info, ('matched_song', 'author'), ('matched_pgc_sound', 'author'), 'author', expected_type=str)
 415
 416         is_generic_og_trackname = music_info.get('is_original_sound') and music_info.get('title') == 'original sound - {}'.format(music_info.get('owner_handle'))
 417         if is_generic_og_trackname:
 418             music_track, music_author = contained_music_track or 'original sound', contained_music_author
 419         else:
 420             music_track, music_author = music_info.get('title'), traverse_obj(music_info, ('author', {str}))
 421
 422         author_info = traverse_obj(aweme_detail, ('author', {
 423             'uploader': ('unique_id', {str}),
 424             'uploader_id': ('uid', {str_or_none}),
 425             'channel': ('nickname', {str}),
 426             'channel_id': ('sec_uid', {str}),
 427         }))
 428
 429         return {
 430             'id': aweme_id,
 431             **traverse_obj(aweme_detail, {
 432                 'title': ('desc', {str}),
 433                 'description': ('desc', {str}),
 434                 'timestamp': ('create_time', {int_or_none}),
 435             }),
 436             **traverse_obj(stats_info, {
 437                 'view_count': 'play_count',
 438                 'like_count': 'digg_count',
 439                 'repost_count': 'share_count',
 440                 'comment_count': 'comment_count',
 441             }, expected_type=int_or_none),
 442             **author_info,
 443             'channel_url': format_field(author_info, 'channel_id', self._UPLOADER_URL_FORMAT, default=None),
 444             'uploader_url': format_field(
 445                 author_info, ['uploader', 'uploader_id'], self._UPLOADER_URL_FORMAT, default=None),
 446             'track': music_track,
 447             'album': str_or_none(music_info.get('album')) or None,
 448             'artists': re.split(r'(?:, | & )', music_author) if music_author else None,
 449             'formats': formats,
 450             'subtitles': self.extract_subtitles(
 451                 aweme_detail, aweme_id, traverse_obj(author_info, 'uploader', 'uploader_id', 'channel_id')),
 452             'thumbnails': thumbnails,
 453             'duration': (traverse_obj(video_info, (
 454                 (None, 'download_addr'), 'duration', {functools.partial(int_or_none, scale=1000)}, any))
 455                 or traverse_obj(music_info, ('duration', {int_or_none}))),
 456             'availability': self._availability(
 457                 is_private='Private' in labels,
 458                 needs_subscription='Friends only' in labels,
 459                 is_unlisted='Followers only' in labels),
 460             '_format_sort_fields': ('quality', 'codec', 'size', 'br'),
 461         }
 462
 463     def _extract_web_formats(self, aweme_detail):
 464         COMMON_FORMAT_INFO = {
 465             'ext': 'mp4',
 466             'vcodec': 'h264',
 467             'acodec': 'aac',
 468         }
 469         video_info = traverse_obj(aweme_detail, ('video', {dict})) or {}
 470         play_width = int_or_none(video_info.get('width'))
 471         play_height = int_or_none(video_info.get('height'))
 472         ratio = try_call(lambda: play_width / play_height) or 0.5625
 473         formats = []
 474
 475         for bitrate_info in traverse_obj(video_info, ('bitrateInfo', lambda _, v: v['PlayAddr']['UrlList'])):
 476             format_info, res = self._parse_url_key(
 477                 traverse_obj(bitrate_info, ('PlayAddr', 'UrlKey', {str})) or '')
 478             # bytevc2 is bytedance's own custom h266/vvc codec, as-of-yet unplayable
 479             is_bytevc2 = format_info.get('vcodec') == 'bytevc2'
 480             format_info.update({
 481                 'format_note': 'UNPLAYABLE' if is_bytevc2 else None,
 482                 'preference': -100 if is_bytevc2 else -1,
 483                 'filesize': traverse_obj(bitrate_info, ('PlayAddr', 'DataSize', {int_or_none})),
 484             })
 485
 486             if dimension := (res and int(res[:-1])):
 487                 if dimension == 540:  # '540p' is actually 576p
 488                     dimension = 576
 489                 if ratio < 1:  # portrait: res/dimension is width
 490                     y = int(dimension / ratio)
 491                     format_info.update({
 492                         'width': dimension,
 493                         'height': y - (y % 2),
 494                     })
 495                 else:  # landscape: res/dimension is height
 496                     x = int(dimension * ratio)
 497                     format_info.update({
 498                         'width': x + (x % 2),
 499                         'height': dimension,
 500                     })
 501
 502             for video_url in traverse_obj(bitrate_info, ('PlayAddr', 'UrlList', ..., {url_or_none})):
 503                 formats.append({
 504                     **COMMON_FORMAT_INFO,
 505                     **format_info,
 506                     'url': self._proto_relative_url(video_url),
 507                 })
 508
 509         # We don't have res string for play formats, but need quality for sorting & de-duplication
 510         play_quality = traverse_obj(formats, (lambda _, v: v['width'] == play_width, 'quality', any))
 511
 512         for play_url in traverse_obj(video_info, ('playAddr', ((..., 'src'), None), {url_or_none})):
 513             formats.append({
 514                 **COMMON_FORMAT_INFO,
 515                 'format_id': 'play',
 516                 'url': self._proto_relative_url(play_url),
 517                 'width': play_width,
 518                 'height': play_height,
 519                 'quality': play_quality,
 520             })
 521
 522         for download_url in traverse_obj(video_info, (('downloadAddr', ('download', 'url')), {url_or_none})):
 523             formats.append({
 524                 **COMMON_FORMAT_INFO,
 525                 'format_id': 'download',
 526                 'url': self._proto_relative_url(download_url),
 527             })
 528
 529         self._remove_duplicate_formats(formats)
 530
 531         for f in traverse_obj(formats, lambda _, v: 'unwatermarked' not in v['url']):
 532             f.update({
 533                 'format_note': join_nonempty(f.get('format_note'), 'watermarked', delim=', '),
 534                 'preference': f.get('preference') or -2,
 535             })
 536
 537         # Is it a slideshow with only audio for download?
 538         if not formats and traverse_obj(aweme_detail, ('music', 'playUrl', {url_or_none})):
 539             audio_url = aweme_detail['music']['playUrl']
 540             ext = traverse_obj(parse_qs(audio_url), (
 541                 'mime_type', -1, {lambda x: x.replace('_', '/')}, {mimetype2ext})) or 'm4a'
 542             formats.append({
 543                 'format_id': 'audio',
 544                 'url': self._proto_relative_url(audio_url),
 545                 'ext': ext,
 546                 'acodec': 'aac' if ext == 'm4a' else ext,
 547                 'vcodec': 'none',
 548             })
 549
 550         return formats
 551
 552     def _parse_aweme_video_web(self, aweme_detail, webpage_url, video_id, extract_flat=False):
 553         author_info = traverse_obj(aweme_detail, (('authorInfo', 'author', None), {
 554             'channel': ('nickname', {str}),
 555             'channel_id': (('authorSecId', 'secUid'), {str}),
 556             'uploader': (('uniqueId', 'author'), {str}),
 557             'uploader_id': (('authorId', 'uid', 'id'), {str_or_none}),
 558         }), get_all=False)
 559
 560         return {
 561             'id': video_id,
 562             'formats': None if extract_flat else self._extract_web_formats(aweme_detail),
 563             'subtitles': None if extract_flat else self.extract_subtitles(aweme_detail, video_id, None),
 564             'http_headers': {'Referer': webpage_url},
 565             **author_info,
 566             'channel_url': format_field(author_info, 'channel_id', self._UPLOADER_URL_FORMAT, default=None),
 567             'uploader_url': format_field(
 568                 author_info, ['uploader', 'uploader_id'], self._UPLOADER_URL_FORMAT, default=None),
 569             **traverse_obj(aweme_detail, ('music', {
 570                 'track': ('title', {str}),
 571                 'album': ('album', {str}, {lambda x: x or None}),
 572                 'artists': ('authorName', {str}, {lambda x: re.split(r'(?:, | & )', x) if x else None}),
 573                 'duration': ('duration', {int_or_none}),
 574             })),
 575             **traverse_obj(aweme_detail, {
 576                 'title': ('desc', {str}),
 577                 'description': ('desc', {str}),
 578                 # audio-only slideshows have a video duration of 0 and an actual audio duration
 579                 'duration': ('video', 'duration', {int_or_none}, {lambda x: x or None}),
 580                 'timestamp': ('createTime', {int_or_none}),
 581             }),
 582             **traverse_obj(aweme_detail, ('stats', {
 583                 'view_count': 'playCount',
 584                 'like_count': 'diggCount',
 585                 'repost_count': 'shareCount',
 586                 'comment_count': 'commentCount',
 587             }), expected_type=int_or_none),
 588             'thumbnails': traverse_obj(aweme_detail, (
 589                 (None, 'video'), ('thumbnail', 'cover', 'dynamicCover', 'originCover'), {
 590                     'url': ({url_or_none}, {self._proto_relative_url}),
 591                 },
 592             )),
 593         }
 594
 595
 596 class TikTokIE(TikTokBaseIE):
 597     _VALID_URL = r'https?://www\.tiktok\.com/(?:embed|@(?P<user_id>[\w\.-]+)?/video)/(?P<id>\d+)'
 598     _EMBED_REGEX = [rf'<(?:script|iframe)[^>]+\bsrc=(["\'])(?P<url>{_VALID_URL})']
 599
 600     _TESTS = [{
 601         'url': 'https://www.tiktok.com/@leenabhushan/video/6748451240264420610',
 602         'md5': '736bb7a466c6f0a6afeb597da1e6f5b7',
 603         'info_dict': {
 604             'id': '6748451240264420610',
 605             'ext': 'mp4',
 606             'title': '#jassmanak #lehanga #leenabhushan',
 607             'description': '#jassmanak #lehanga #leenabhushan',
 608             'duration': 13,
 609             'height': 1024,
 610             'width': 576,
 611             'uploader': 'leenabhushan',
 612             'uploader_id': '6691488002098119685',
 613             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAA_Eb4t1vodM1IuTy_cvp9CY22RAb59xqrO0Xtz9CYQJvgXaDvZxYnZYRzDWhhgJmy',
 614             'creator': 'facestoriesbyleenabh',
 615             'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
 616             'upload_date': '20191016',
 617             'timestamp': 1571246252,
 618             'view_count': int,
 619             'like_count': int,
 620             'repost_count': int,
 621             'comment_count': int,
 622             'artist': 'Ysrbeats',
 623             'album': 'Lehanga',
 624             'track': 'Lehanga',
 625         },
 626         'skip': '404 Not Found',
 627     }, {
 628         'url': 'https://www.tiktok.com/@patroxofficial/video/6742501081818877190?langCountry=en',
 629         'md5': 'f21112672ee4ce05ca390fb6522e1b6f',
 630         'info_dict': {
 631             'id': '6742501081818877190',
 632             'ext': 'mp4',
 633             'title': 'md5:5e2a23877420bb85ce6521dbee39ba94',
 634             'description': 'md5:5e2a23877420bb85ce6521dbee39ba94',
 635             'duration': 27,
 636             'height': 1024,
 637             'width': 576,
 638             'uploader': 'patrox',
 639             'uploader_id': '18702747',
 640             'uploader_url': 'https://www.tiktok.com/@patrox',
 641             'channel_url': 'https://www.tiktok.com/@MS4wLjABAAAAiFnldaILebi5heDoVU6bn4jBWWycX6-9U3xuNPqZ8Ws',
 642             'channel_id': 'MS4wLjABAAAAiFnldaILebi5heDoVU6bn4jBWWycX6-9U3xuNPqZ8Ws',
 643             'channel': 'patroX',
 644             'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
 645             'upload_date': '20190930',
 646             'timestamp': 1569860870,
 647             'view_count': int,
 648             'like_count': int,
 649             'repost_count': int,
 650             'comment_count': int,
 651             'artists': ['Evan Todd', 'Jessica Keenan Wynn', 'Alice Lee', 'Barrett Wilbert Weed', 'Jon Eidson'],
 652             'track': 'Big Fun',
 653         },
 654     }, {
 655         # Banned audio, was available on the app, now works with web too
 656         'url': 'https://www.tiktok.com/@barudakhb_/video/6984138651336838402',
 657         'info_dict': {
 658             'id': '6984138651336838402',
 659             'ext': 'mp4',
 660             'title': 'Balas @yolaaftwsr hayu yu ? #SquadRandom_ 🔥',
 661             'description': 'Balas @yolaaftwsr hayu yu ? #SquadRandom_ 🔥',
 662             'uploader': 'barudakhb_',
 663             'channel': 'md5:29f238c49bc0c176cb3cef1a9cea9fa6',
 664             'uploader_id': '6974687867511718913',
 665             'uploader_url': 'https://www.tiktok.com/@barudakhb_',
 666             'channel_url': 'https://www.tiktok.com/@MS4wLjABAAAAbhBwQC-R1iKoix6jDFsF-vBdfx2ABoDjaZrM9fX6arU3w71q3cOWgWuTXn1soZ7d',
 667             'channel_id': 'MS4wLjABAAAAbhBwQC-R1iKoix6jDFsF-vBdfx2ABoDjaZrM9fX6arU3w71q3cOWgWuTXn1soZ7d',
 668             'track': 'Boka Dance',
 669             'artists': ['md5:29f238c49bc0c176cb3cef1a9cea9fa6'],
 670             'timestamp': 1626121503,
 671             'duration': 18,
 672             'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
 673             'upload_date': '20210712',
 674             'view_count': int,
 675             'like_count': int,
 676             'repost_count': int,
 677             'comment_count': int,
 678         },
 679     }, {
 680         # Sponsored video, only available with feed workaround
 681         'url': 'https://www.tiktok.com/@MS4wLjABAAAATh8Vewkn0LYM7Fo03iec3qKdeCUOcBIouRk1mkiag6h3o_pQu_dUXvZ2EZlGST7_/video/7042692929109986561',
 682         'info_dict': {
 683             'id': '7042692929109986561',
 684             'ext': 'mp4',
 685             'title': 'Slap and Run!',
 686             'description': 'Slap and Run!',
 687             'uploader': 'user440922249',
 688             'channel': 'Slap And Run',
 689             'uploader_id': '7036055384943690754',
 690             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAATh8Vewkn0LYM7Fo03iec3qKdeCUOcBIouRk1mkiag6h3o_pQu_dUXvZ2EZlGST7_',
 691             'channel_id': 'MS4wLjABAAAATh8Vewkn0LYM7Fo03iec3qKdeCUOcBIouRk1mkiag6h3o_pQu_dUXvZ2EZlGST7_',
 692             'track': 'Promoted Music',
 693             'timestamp': 1639754738,
 694             'duration': 30,
 695             'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
 696             'upload_date': '20211217',
 697             'view_count': int,
 698             'like_count': int,
 699             'repost_count': int,
 700             'comment_count': int,
 701         },
 702         'skip': 'This video is unavailable',
 703     }, {
 704         # Video without title and description
 705         'url': 'https://www.tiktok.com/@pokemonlife22/video/7059698374567611694',
 706         'info_dict': {
 707             'id': '7059698374567611694',
 708             'ext': 'mp4',
 709             'title': 'TikTok video #7059698374567611694',
 710             'description': '',
 711             'uploader': 'pokemonlife22',
 712             'channel': 'Pokemon',
 713             'uploader_id': '6820838815978423302',
 714             'uploader_url': 'https://www.tiktok.com/@pokemonlife22',
 715             'channel_url': 'https://www.tiktok.com/@MS4wLjABAAAA0tF1nBwQVVMyrGu3CqttkNgM68Do1OXUFuCY0CRQk8fEtSVDj89HqoqvbSTmUP2W',
 716             'channel_id': 'MS4wLjABAAAA0tF1nBwQVVMyrGu3CqttkNgM68Do1OXUFuCY0CRQk8fEtSVDj89HqoqvbSTmUP2W',
 717             'track': 'original sound',
 718             'timestamp': 1643714123,
 719             'duration': 6,
 720             'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
 721             'upload_date': '20220201',
 722             'artists': ['Pokemon'],
 723             'view_count': int,
 724             'like_count': int,
 725             'repost_count': int,
 726             'comment_count': int,
 727         },
 728     }, {
 729         # hydration JSON is sent in a <script> element
 730         'url': 'https://www.tiktok.com/@denidil6/video/7065799023130643713',
 731         'info_dict': {
 732             'id': '7065799023130643713',
 733             'ext': 'mp4',
 734             'title': '#denidil#денидил',
 735             'description': '#denidil#денидил',
 736             'uploader': 'denidil6',
 737             'uploader_id': '7046664115636405250',
 738             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAAsvMSzFdQ4ikl3uR2TEJwMBbB2yZh2Zxwhx-WCo3rbDpAharE3GQCrFuJArI3C8QJ',
 739             'artist': 'Holocron Music',
 740             'album': 'Wolf Sounds (1 Hour) Enjoy the Company of the Animal That Is the Majestic King of the Night',
 741             'track': 'Wolf Sounds (1 Hour) Enjoy the Company of the Animal That Is the Majestic King of the Night',
 742             'timestamp': 1645134536,
 743             'duration': 26,
 744             'upload_date': '20220217',
 745             'view_count': int,
 746             'like_count': int,
 747             'repost_count': int,
 748             'comment_count': int,
 749         },
 750         'skip': 'This video is unavailable',
 751     }, {
 752         # slideshow audio-only mp3 format
 753         'url': 'https://www.tiktok.com/@_le_cannibale_/video/7139980461132074283',
 754         'info_dict': {
 755             'id': '7139980461132074283',
 756             'ext': 'mp3',
 757             'title': 'TikTok video #7139980461132074283',
 758             'description': '',
 759             'channel': 'Antaura',
 760             'uploader': '_le_cannibale_',
 761             'uploader_id': '6604511138619654149',
 762             'uploader_url': 'https://www.tiktok.com/@_le_cannibale_',
 763             'channel_url': 'https://www.tiktok.com/@MS4wLjABAAAAoShJqaw_5gvy48y3azFeFcT4jeyKWbB0VVYasOCt2tTLwjNFIaDcHAM4D-QGXFOP',
 764             'channel_id': 'MS4wLjABAAAAoShJqaw_5gvy48y3azFeFcT4jeyKWbB0VVYasOCt2tTLwjNFIaDcHAM4D-QGXFOP',
 765             'artists': ['nathan !'],
 766             'track': 'grahamscott canon',
 767             'duration': 10,
 768             'upload_date': '20220905',
 769             'timestamp': 1662406249,
 770             'view_count': int,
 771             'like_count': int,
 772             'repost_count': int,
 773             'comment_count': int,
 774             'thumbnail': r're:^https://.+\.(?:webp|jpe?g)',
 775         },
 776     }, {
 777         # only available via web
 778         'url': 'https://www.tiktok.com/@moxypatch/video/7206382937372134662',
 779         'md5': '4cdefa501ac8ac20bf04986e10916fea',
 780         'info_dict': {
 781             'id': '7206382937372134662',
 782             'ext': 'mp4',
 783             'title': 'md5:1d95c0b96560ca0e8a231af4172b2c0a',
 784             'description': 'md5:1d95c0b96560ca0e8a231af4172b2c0a',
 785             'channel': 'MoxyPatch',
 786             'uploader': 'moxypatch',
 787             'uploader_id': '7039142049363379205',
 788             'uploader_url': 'https://www.tiktok.com/@moxypatch',
 789             'channel_url': 'https://www.tiktok.com/@MS4wLjABAAAAFhqKnngMHJSsifL0w1vFOP5kn3Ndo1ODp0XuIBkNMBCkALTvwILdpu12g3pTtL4V',
 790             'channel_id': 'MS4wLjABAAAAFhqKnngMHJSsifL0w1vFOP5kn3Ndo1ODp0XuIBkNMBCkALTvwILdpu12g3pTtL4V',
 791             'artists': ['your worst nightmare'],
 792             'track': 'original sound',
 793             'upload_date': '20230303',
 794             'timestamp': 1677866781,
 795             'duration': 10,
 796             'view_count': int,
 797             'like_count': int,
 798             'repost_count': int,
 799             'comment_count': int,
 800             'thumbnail': r're:^https://.+',
 801             'thumbnails': 'count:3',
 802         },
 803         'expected_warnings': ['Unable to find video in feed'],
 804     }, {
 805         # 1080p format
 806         'url': 'https://www.tiktok.com/@tatemcrae/video/7107337212743830830',  # FIXME: Web can only get audio
 807         'md5': '982512017a8a917124d5a08c8ae79621',
 808         'info_dict': {
 809             'id': '7107337212743830830',
 810             'ext': 'mp4',
 811             'title': 'new music video 4 don’t come backkkk🧸🖤 i hope u enjoy !! @musicontiktok',
 812             'description': 'new music video 4 don’t come backkkk🧸🖤 i hope u enjoy !! @musicontiktok',
 813             'uploader': 'tatemcrae',
 814             'uploader_id': '86328792343818240',
 815             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAA-0bQT0CqebTRr6I4IkYvMDMKSRSJHLNPBo5HrSklJwyA2psXLSZG5FP-LMNpHnJd',
 816             'channel_id': 'MS4wLjABAAAA-0bQT0CqebTRr6I4IkYvMDMKSRSJHLNPBo5HrSklJwyA2psXLSZG5FP-LMNpHnJd',
 817             'channel': 'tate mcrae',
 818             'artists': ['tate mcrae'],
 819             'track': 'original sound',
 820             'upload_date': '20220609',
 821             'timestamp': 1654805899,
 822             'duration': 150,
 823             'view_count': int,
 824             'like_count': int,
 825             'repost_count': int,
 826             'comment_count': int,
 827             'thumbnail': r're:^https://.+\.webp',
 828         },
 829         'skip': 'Unavailable via feed API, only audio available via web',
 830     }, {
 831         # Slideshow, audio-only m4a format
 832         'url': 'https://www.tiktok.com/@hara_yoimiya/video/7253412088251534594',
 833         'md5': '2ff8fe0174db2dbf49c597a7bef4e47d',
 834         'info_dict': {
 835             'id': '7253412088251534594',
 836             'ext': 'm4a',
 837             'title': 'я ред флаг простите #переписка #щитпост #тревожныйтиппривязанности #рекомендации ',
 838             'description': 'я ред флаг простите #переписка #щитпост #тревожныйтиппривязанности #рекомендации ',
 839             'uploader': 'hara_yoimiya',
 840             'uploader_id': '6582536342634676230',
 841             'uploader_url': 'https://www.tiktok.com/@hara_yoimiya',
 842             'channel_url': 'https://www.tiktok.com/@MS4wLjABAAAAIAlDxriiPWLE-p8p1R_0Bx8qWKfi-7zwmGhzU8Mv25W8sNxjfIKrol31qTczzuLB',
 843             'channel_id': 'MS4wLjABAAAAIAlDxriiPWLE-p8p1R_0Bx8qWKfi-7zwmGhzU8Mv25W8sNxjfIKrol31qTczzuLB',
 844             'channel': 'лампочка(!)',
 845             'artists': ['Øneheart'],
 846             'album': 'watching the stars',
 847             'track': 'watching the stars',
 848             'duration': 60,
 849             'upload_date': '20230708',
 850             'timestamp': 1688816612,
 851             'view_count': int,
 852             'like_count': int,
 853             'comment_count': int,
 854             'repost_count': int,
 855             'thumbnail': r're:^https://.+\.(?:webp|jpe?g)',
 856         },
 857     }, {
 858         # Auto-captions available
 859         'url': 'https://www.tiktok.com/@hankgreen1/video/7047596209028074758',
 860         'only_matching': True,
 861     }]
 862
 863     def _real_extract(self, url):
 864         video_id, user_id = self._match_valid_url(url).group('id', 'user_id')
 865
 866         if self._KNOWN_APP_INFO:
 867             try:
 868                 return self._extract_aweme_app(video_id)
 869             except ExtractorError as e:
 870                 e.expected = True
 871                 self.report_warning(f'{e}; trying with webpage')
 872
 873         url = self._create_url(user_id, video_id)
 874         video_data, status = self._extract_web_data_and_status(url, video_id)
 875
 876         if video_data and status == 0:
 877             return self._parse_aweme_video_web(video_data, url, video_id)
 878         elif status == 10216:
 879             raise ExtractorError('This video is private', expected=True)
 880         raise ExtractorError(f'Video not available, status code {status}', video_id=video_id)
 881
 882
 883 class TikTokUserIE(TikTokBaseIE):
 884     IE_NAME = 'tiktok:user'
 885     _VALID_URL = r'(?:tiktokuser:|https?://(?:www\.)?tiktok\.com/@)(?P<id>[\w.-]+)/?(?:$|[#?])'
 886     _TESTS = [{
 887         'url': 'https://tiktok.com/@corgibobaa?lang=en',
 888         'playlist_mincount': 45,
 889         'info_dict': {
 890             'id': 'MS4wLjABAAAAepiJKgwWhulvCpSuUVsp7sgVVsFJbbNaLeQ6OQ0oAJERGDUIXhb2yxxHZedsItgT',
 891             'title': 'corgibobaa',
 892         },
 893     }, {
 894         'url': 'https://www.tiktok.com/@6820838815978423302',
 895         'playlist_mincount': 5,
 896         'info_dict': {
 897             'id': 'MS4wLjABAAAA0tF1nBwQVVMyrGu3CqttkNgM68Do1OXUFuCY0CRQk8fEtSVDj89HqoqvbSTmUP2W',
 898             'title': '6820838815978423302',
 899         },
 900     }, {
 901         'url': 'https://www.tiktok.com/@meme',
 902         'playlist_mincount': 593,
 903         'info_dict': {
 904             'id': 'MS4wLjABAAAAiKfaDWeCsT3IHwY77zqWGtVRIy9v4ws1HbVi7auP1Vx7dJysU_hc5yRiGywojRD6',
 905             'title': 'meme',
 906         },
 907     }, {
 908         'url': 'tiktokuser:MS4wLjABAAAAM3R2BtjzVT-uAtstkl2iugMzC6AtnpkojJbjiOdDDrdsTiTR75-8lyWJCY5VvDrZ',
 909         'playlist_mincount': 31,
 910         'info_dict': {
 911             'id': 'MS4wLjABAAAAM3R2BtjzVT-uAtstkl2iugMzC6AtnpkojJbjiOdDDrdsTiTR75-8lyWJCY5VvDrZ',
 912         },
 913     }]
 914     _USER_AGENT = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:115.0) Gecko/20100101 Firefox/115.0'
 915     _API_BASE_URL = 'https://www.tiktok.com/api/creator/item_list/'
 916
 917     def _build_web_query(self, sec_uid, cursor):
 918         return {
 919             'aid': '1988',
 920             'app_language': 'en',
 921             'app_name': 'tiktok_web',
 922             'browser_language': 'en-US',
 923             'browser_name': 'Mozilla',
 924             'browser_online': 'true',
 925             'browser_platform': 'Win32',
 926             'browser_version': '5.0 (Windows)',
 927             'channel': 'tiktok_web',
 928             'cookie_enabled': 'true',
 929             'count': '15',
 930             'cursor': cursor,
 931             'device_id': self._DEVICE_ID,
 932             'device_platform': 'web_pc',
 933             'focus_state': 'true',
 934             'from_page': 'user',
 935             'history_len': '2',
 936             'is_fullscreen': 'false',
 937             'is_page_visible': 'true',
 938             'language': 'en',
 939             'os': 'windows',
 940             'priority_region': '',
 941             'referer': '',
 942             'region': 'US',
 943             'screen_height': '1080',
 944             'screen_width': '1920',
 945             'secUid': sec_uid,
 946             'type': '1',  # pagination type: 0 == oldest-to-newest, 1 == newest-to-oldest
 947             'tz_name': 'UTC',
 948             'verifyFp': f'verify_{"".join(random.choices(string.hexdigits, k=7))}',
 949             'webcast_language': 'en',
 950         }
 951
 952     def _entries(self, sec_uid, user_name):
 953         display_id = user_name or sec_uid
 954         seen_ids = set()
 955
 956         cursor = int(time.time() * 1E3)
 957         for page in itertools.count(1):
 958             response = self._download_json(
 959                 self._API_BASE_URL, display_id, f'Downloading page {page}',
 960                 query=self._build_web_query(sec_uid, cursor), headers={'User-Agent': self._USER_AGENT})
 961
 962             for video in traverse_obj(response, ('itemList', lambda _, v: v['id'])):
 963                 video_id = video['id']
 964                 if video_id in seen_ids:
 965                     continue
 966                 seen_ids.add(video_id)
 967                 webpage_url = self._create_url(display_id, video_id)
 968                 yield self.url_result(
 969                     webpage_url, TikTokIE,
 970                     **self._parse_aweme_video_web(video, webpage_url, video_id, extract_flat=True))
 971
 972             old_cursor = cursor
 973             cursor = traverse_obj(
 974                 response, ('itemList', -1, 'createTime', {lambda x: int(x * 1E3)}))
 975             if not cursor or old_cursor == cursor:
 976                 # User may not have posted within this ~1 week lookback, so manually adjust cursor
 977                 cursor = old_cursor - 7 * 86_400_000
 978             # In case 'hasMorePrevious' is wrong, break if we have gone back before TikTok existed
 979             if cursor < 1472706000000 or not traverse_obj(response, 'hasMorePrevious'):
 980                 break
 981
 982     def _get_sec_uid(self, user_url, user_name, msg):
 983         webpage = self._download_webpage(
 984             user_url, user_name, fatal=False, headers={'User-Agent': 'Mozilla/5.0'},
 985             note=f'Downloading {msg} webpage', errnote=f'Unable to download {msg} webpage') or ''
 986         return (traverse_obj(self._get_universal_data(webpage, user_name),
 987                              ('webapp.user-detail', 'userInfo', 'user', 'secUid', {str}))
 988                 or traverse_obj(self._get_sigi_state(webpage, user_name),
 989                                 ('LiveRoom', 'liveRoomUserInfo', 'user', 'secUid', {str}),
 990                                 ('UserModule', 'users', ..., 'secUid', {str}, any)))
 991
 992     def _real_extract(self, url):
 993         user_name, sec_uid = self._match_id(url), None
 994         if mobj := re.fullmatch(r'MS4wLjABAAAA[\w-]{64}', user_name):
 995             user_name, sec_uid = None, mobj.group(0)
 996         else:
 997             sec_uid = (self._get_sec_uid(self._UPLOADER_URL_FORMAT % user_name, user_name, 'user')
 998                        or self._get_sec_uid(self._UPLOADER_URL_FORMAT % f'{user_name}/live', user_name, 'live'))
 999
1000         if not sec_uid:
1001             webpage = self._download_webpage(
1002                 f'https://www.tiktok.com/embed/@{user_name}', user_name,
1003                 note='Downloading user embed page', fatal=False) or ''
1004             data = traverse_obj(self._search_json(
1005                 r'<script[^>]+\bid=[\'"]__FRONTITY_CONNECT_STATE__[\'"][^>]*>',
1006                 webpage, 'data', user_name, default={}),
1007                 ('source', 'data', f'/embed/@{user_name}', {dict}))
1008
1009             for aweme_id in traverse_obj(data, ('videoList', ..., 'id', {str})):
1010                 webpage_url = self._create_url(user_name, aweme_id)
1011                 video_data, _ = self._extract_web_data_and_status(webpage_url, aweme_id, fatal=False)
1012                 sec_uid = self._parse_aweme_video_web(
1013                     video_data, webpage_url, aweme_id, extract_flat=True).get('channel_id')
1014                 if sec_uid:
1015                     break
1016
1017         if not sec_uid:
1018             raise ExtractorError(
1019                 'Unable to extract secondary user ID. If you are able to get the channel_id '
1020                 'from a video posted by this user, try using "tiktokuser:channel_id" as the '
1021                 'input URL (replacing `channel_id` with its actual value)', expected=True)
1022
1023         return self.playlist_result(self._entries(sec_uid, user_name), sec_uid, user_name)
1024
1025
1026 class TikTokBaseListIE(TikTokBaseIE):  # XXX: Conventionally, base classes should end with BaseIE/InfoExtractor
1027     def _entries(self, list_id, display_id):
1028         query = {
1029             self._QUERY_NAME: list_id,
1030             'cursor': 0,
1031             'count': 20,
1032             'type': 5,
1033             'device_id': self._DEVICE_ID,
1034         }
1035
1036         for page in itertools.count(1):
1037             for retry in self.RetryManager():
1038                 try:
1039                     post_list = self._call_api(
1040                         self._API_ENDPOINT, query, display_id, note=f'Downloading video list page {page}',
1041                         errnote='Unable to download video list')
1042                 except ExtractorError as e:
1043                     if isinstance(e.cause, json.JSONDecodeError) and e.cause.pos == 0:
1044                         retry.error = e
1045                         continue
1046                     raise
1047             for video in post_list.get('aweme_list', []):
1048                 yield {
1049                     **self._parse_aweme_video_app(video),
1050                     'extractor_key': TikTokIE.ie_key(),
1051                     'extractor': 'TikTok',
1052                     'webpage_url': f'https://tiktok.com/@_/video/{video["aweme_id"]}',
1053                 }
1054             if not post_list.get('has_more'):
1055                 break
1056             query['cursor'] = post_list['cursor']
1057
1058     def _real_extract(self, url):
1059         list_id = self._match_id(url)
1060         return self.playlist_result(self._entries(list_id, list_id), list_id)
1061
1062
1063 class TikTokSoundIE(TikTokBaseListIE):
1064     IE_NAME = 'tiktok:sound'
1065     _VALID_URL = r'https?://(?:www\.)?tiktok\.com/music/[\w\.-]+-(?P<id>[\d]+)[/?#&]?'
1066     _WORKING = False
1067     _QUERY_NAME = 'music_id'
1068     _API_ENDPOINT = 'music/aweme'
1069     _TESTS = [{
1070         'url': 'https://www.tiktok.com/music/Build-a-Btch-6956990112127585029?lang=en',
1071         'playlist_mincount': 100,
1072         'info_dict': {
1073             'id': '6956990112127585029',
1074         },
1075         'expected_warnings': ['Retrying'],
1076     }, {
1077         # Actual entries are less than listed video count
1078         'url': 'https://www.tiktok.com/music/jiefei-soap-remix-7036843036118469381',
1079         'playlist_mincount': 2182,
1080         'info_dict': {
1081             'id': '7036843036118469381',
1082         },
1083         'expected_warnings': ['Retrying'],
1084     }]
1085
1086
1087 class TikTokEffectIE(TikTokBaseListIE):
1088     IE_NAME = 'tiktok:effect'
1089     _VALID_URL = r'https?://(?:www\.)?tiktok\.com/sticker/[\w\.-]+-(?P<id>[\d]+)[/?#&]?'
1090     _WORKING = False
1091     _QUERY_NAME = 'sticker_id'
1092     _API_ENDPOINT = 'sticker/aweme'
1093     _TESTS = [{
1094         'url': 'https://www.tiktok.com/sticker/MATERIAL-GWOOORL-1258156',
1095         'playlist_mincount': 100,
1096         'info_dict': {
1097             'id': '1258156',
1098         },
1099         'expected_warnings': ['Retrying'],
1100     }, {
1101         # Different entries between mobile and web, depending on region
1102         'url': 'https://www.tiktok.com/sticker/Elf-Friend-479565',
1103         'only_matching': True,
1104     }]
1105
1106
1107 class TikTokTagIE(TikTokBaseListIE):
1108     IE_NAME = 'tiktok:tag'
1109     _VALID_URL = r'https?://(?:www\.)?tiktok\.com/tag/(?P<id>[^/?#&]+)'
1110     _WORKING = False
1111     _QUERY_NAME = 'ch_id'
1112     _API_ENDPOINT = 'challenge/aweme'
1113     _TESTS = [{
1114         'url': 'https://tiktok.com/tag/hello2018',
1115         'playlist_mincount': 39,
1116         'info_dict': {
1117             'id': '46294678',
1118             'title': 'hello2018',
1119         },
1120         'expected_warnings': ['Retrying'],
1121     }, {
1122         'url': 'https://tiktok.com/tag/fypシ?is_copy_url=0&is_from_webapp=v1',
1123         'only_matching': True,
1124     }]
1125
1126     def _real_extract(self, url):
1127         display_id = self._match_id(url)
1128         webpage = self._download_webpage(url, display_id, headers={
1129             'User-Agent': 'facebookexternalhit/1.1 (+http://www.facebook.com/externalhit_uatext.php)',
1130         })
1131         tag_id = self._html_search_regex(r'snssdk\d*://challenge/detail/(\d+)', webpage, 'tag ID')
1132         return self.playlist_result(self._entries(tag_id, display_id), tag_id, display_id)
1133
1134
1135 class TikTokCollectionIE(TikTokBaseIE):
1136     IE_NAME = 'tiktok:collection'
1137     _VALID_URL = r'https?://www\.tiktok\.com/@(?P<user_id>[\w.-]+)/collection/(?P<title>[^/?#]+)-(?P<id>\d+)/?(?:[?#]|$)'
1138     _TESTS = [{
1139         # playlist should have exactly 9 videos
1140         'url': 'https://www.tiktok.com/@imanoreotwe/collection/count-test-7371330159376370462',
1141         'info_dict': {
1142             'id': '7371330159376370462',
1143             'title': 'imanoreotwe-count-test',
1144         },
1145         'playlist_count': 9,
1146     }, {
1147         # tests returning multiple pages of a large collection
1148         'url': 'https://www.tiktok.com/@imanoreotwe/collection/%F0%9F%98%82-7111887189571160875',
1149         'info_dict': {
1150             'id': '7111887189571160875',
1151             'title': 'imanoreotwe-%F0%9F%98%82',
1152         },
1153         'playlist_mincount': 100,
1154     }]
1155     _API_BASE_URL = 'https://www.tiktok.com/api/collection/item_list/'
1156     _PAGE_COUNT = 30
1157
1158     def _build_web_query(self, collection_id, cursor):
1159         return {
1160             'aid': '1988',
1161             'collectionId': collection_id,
1162             'count': self._PAGE_COUNT,
1163             'cursor': cursor,
1164             'sourceType': '113',
1165         }
1166
1167     def _entries(self, collection_id):
1168         cursor = 0
1169         for page in itertools.count(1):
1170             response = self._download_json(
1171                 self._API_BASE_URL, collection_id, f'Downloading page {page}',
1172                 query=self._build_web_query(collection_id, cursor))
1173
1174             for video in traverse_obj(response, ('itemList', lambda _, v: v['id'])):
1175                 video_id = video['id']
1176                 author = traverse_obj(video, ('author', ('uniqueId', 'secUid', 'id'), {str}, any)) or '_'
1177                 webpage_url = self._create_url(author, video_id)
1178                 yield self.url_result(
1179                     webpage_url, TikTokIE,
1180                     **self._parse_aweme_video_web(video, webpage_url, video_id, extract_flat=True))
1181
1182             if not traverse_obj(response, 'hasMore'):
1183                 break
1184             cursor += self._PAGE_COUNT
1185
1186     def _real_extract(self, url):
1187         collection_id, title, user_name = self._match_valid_url(url).group('id', 'title', 'user_id')
1188
1189         return self.playlist_result(
1190             self._entries(collection_id), collection_id, '-'.join((user_name, title)))
1191
1192
1193 class DouyinIE(TikTokBaseIE):
1194     _VALID_URL = r'https?://(?:www\.)?douyin\.com/video/(?P<id>[0-9]+)'
1195     _TESTS = [{
1196         'url': 'https://www.douyin.com/video/6961737553342991651',
1197         'md5': '9ecce7bc5b302601018ecb2871c63a75',
1198         'info_dict': {
1199             'id': '6961737553342991651',
1200             'ext': 'mp4',
1201             'title': '#杨超越  小小水手带你去远航❤️',
1202             'description': '#杨超越  小小水手带你去远航❤️',
1203             'uploader': '6897520xka',
1204             'uploader_id': '110403406559',
1205             'uploader_url': 'https://www.douyin.com/user/MS4wLjABAAAAEKnfa654JAJ_N5lgZDQluwsxmY0lhfmEYNQBBkwGG98',
1206             'channel_id': 'MS4wLjABAAAAEKnfa654JAJ_N5lgZDQluwsxmY0lhfmEYNQBBkwGG98',
1207             'channel': '杨超越',
1208             'duration': 19,
1209             'timestamp': 1620905839,
1210             'upload_date': '20210513',
1211             'track': '@杨超越创作的原声',
1212             'artists': ['杨超越'],
1213             'view_count': int,
1214             'like_count': int,
1215             'repost_count': int,
1216             'comment_count': int,
1217             'thumbnail': r're:https?://.+\.jpe?g',
1218         },
1219     }, {
1220         'url': 'https://www.douyin.com/video/6982497745948921092',
1221         'md5': '15c5e660b7048af3707304e3cc02bbb5',
1222         'info_dict': {
1223             'id': '6982497745948921092',
1224             'ext': 'mp4',
1225             'title': '这个夏日和小羊@杨超越 一起遇见白色幻想',
1226             'description': '这个夏日和小羊@杨超越 一起遇见白色幻想',
1227             'uploader': '0731chaoyue',
1228             'uploader_id': '408654318141572',
1229             'uploader_url': 'https://www.douyin.com/user/MS4wLjABAAAAZJpnglcjW2f_CMVcnqA_6oVBXKWMpH0F8LIHuUu8-lA',
1230             'channel_id': 'MS4wLjABAAAAZJpnglcjW2f_CMVcnqA_6oVBXKWMpH0F8LIHuUu8-lA',
1231             'channel': '杨超越工作室',
1232             'duration': 42,
1233             'timestamp': 1625739481,
1234             'upload_date': '20210708',
1235             'track': '@杨超越工作室创作的原声',
1236             'artists': ['杨超越工作室'],
1237             'view_count': int,
1238             'like_count': int,
1239             'repost_count': int,
1240             'comment_count': int,
1241             'thumbnail': r're:https?://.+\.jpe?g',
1242         },
1243     }, {
1244         'url': 'https://www.douyin.com/video/6953975910773099811',
1245         'md5': '0e6443758b8355db9a3c34864a4276be',
1246         'info_dict': {
1247             'id': '6953975910773099811',
1248             'ext': 'mp4',
1249             'title': '#一起看海  出现在你的夏日里',
1250             'description': '#一起看海  出现在你的夏日里',
1251             'uploader': '6897520xka',
1252             'uploader_id': '110403406559',
1253             'uploader_url': 'https://www.douyin.com/user/MS4wLjABAAAAEKnfa654JAJ_N5lgZDQluwsxmY0lhfmEYNQBBkwGG98',
1254             'channel_id': 'MS4wLjABAAAAEKnfa654JAJ_N5lgZDQluwsxmY0lhfmEYNQBBkwGG98',
1255             'channel': '杨超越',
1256             'duration': 17,
1257             'timestamp': 1619098692,
1258             'upload_date': '20210422',
1259             'track': '@杨超越创作的原声',
1260             'artists': ['杨超越'],
1261             'view_count': int,
1262             'like_count': int,
1263             'repost_count': int,
1264             'comment_count': int,
1265             'thumbnail': r're:https?://.+\.jpe?g',
1266         },
1267     }, {
1268         'url': 'https://www.douyin.com/video/6950251282489675042',
1269         'md5': 'b4db86aec367ef810ddd38b1737d2fed',
1270         'info_dict': {
1271             'id': '6950251282489675042',
1272             'ext': 'mp4',
1273             'title': '哈哈哈，成功了哈哈哈哈哈哈',
1274             'uploader': '杨超越',
1275             'upload_date': '20210412',
1276             'timestamp': 1618231483,
1277             'uploader_id': '110403406559',
1278             'view_count': int,
1279             'like_count': int,
1280             'repost_count': int,
1281             'comment_count': int,
1282         },
1283         'skip': 'No longer available',
1284     }, {
1285         'url': 'https://www.douyin.com/video/6963263655114722595',
1286         'md5': '1440bcf59d8700f8e014da073a4dfea8',
1287         'info_dict': {
1288             'id': '6963263655114722595',
1289             'ext': 'mp4',
1290             'title': '#哪个爱豆的105度最甜 换个角度看看我哈哈',
1291             'description': '#哪个爱豆的105度最甜 换个角度看看我哈哈',
1292             'uploader': '6897520xka',
1293             'uploader_id': '110403406559',
1294             'uploader_url': 'https://www.douyin.com/user/MS4wLjABAAAAEKnfa654JAJ_N5lgZDQluwsxmY0lhfmEYNQBBkwGG98',
1295             'channel_id': 'MS4wLjABAAAAEKnfa654JAJ_N5lgZDQluwsxmY0lhfmEYNQBBkwGG98',
1296             'channel': '杨超越',
1297             'duration': 15,
1298             'timestamp': 1621261163,
1299             'upload_date': '20210517',
1300             'track': '@杨超越创作的原声',
1301             'artists': ['杨超越'],
1302             'view_count': int,
1303             'like_count': int,
1304             'repost_count': int,
1305             'comment_count': int,
1306             'thumbnail': r're:https?://.+\.jpe?g',
1307         },
1308     }]
1309     _UPLOADER_URL_FORMAT = 'https://www.douyin.com/user/%s'
1310     _WEBPAGE_HOST = 'https://www.douyin.com/'
1311
1312     def _real_extract(self, url):
1313         video_id = self._match_id(url)
1314
1315         detail = traverse_obj(self._download_json(
1316             'https://www.douyin.com/aweme/v1/web/aweme/detail/', video_id,
1317             'Downloading web detail JSON', 'Failed to download web detail JSON',
1318             query={'aweme_id': video_id}, fatal=False), ('aweme_detail', {dict}))
1319         if not detail:
1320             # TODO: Run verification challenge code to generate signature cookies
1321             raise ExtractorError(
1322                 'Fresh cookies (not necessarily logged in) are needed',
1323                 expected=not self._get_cookies(self._WEBPAGE_HOST).get('s_v_web_id'))
1324
1325         return self._parse_aweme_video_app(detail)
1326
1327
1328 class TikTokVMIE(InfoExtractor):
1329     _VALID_URL = r'https?://(?:(?:vm|vt)\.tiktok\.com|(?:www\.)tiktok\.com/t)/(?P<id>\w+)'
1330     IE_NAME = 'vm.tiktok'
1331
1332     _TESTS = [{
1333         'url': 'https://www.tiktok.com/t/ZTRC5xgJp',
1334         'info_dict': {
1335             'id': '7170520270497680683',
1336             'ext': 'mp4',
1337             'title': 'md5:c64f6152330c2efe98093ccc8597871c',
1338             'uploader_id': '6687535061741700102',
1339             'upload_date': '20221127',
1340             'view_count': int,
1341             'like_count': int,
1342             'comment_count': int,
1343             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAAObqu3WCTXxmw2xwZ3iLEHnEecEIw7ks6rxWqOqOhaPja9BI7gqUQnjw8_5FSoDXX',
1344             'album': 'Wave of Mutilation: Best of Pixies',
1345             'thumbnail': r're:https://.+\.webp.*',
1346             'duration': 5,
1347             'timestamp': 1669516858,
1348             'repost_count': int,
1349             'artist': 'Pixies',
1350             'track': 'Where Is My Mind?',
1351             'description': 'md5:c64f6152330c2efe98093ccc8597871c',
1352             'uploader': 'sigmachaddeus',
1353             'creator': 'SigmaChad',
1354         },
1355     }, {
1356         'url': 'https://vm.tiktok.com/ZTR45GpSF/',
1357         'info_dict': {
1358             'id': '7106798200794926362',
1359             'ext': 'mp4',
1360             'title': 'md5:edc3e7ea587847f8537468f2fe51d074',
1361             'uploader_id': '6997695878846268418',
1362             'upload_date': '20220608',
1363             'view_count': int,
1364             'like_count': int,
1365             'comment_count': int,
1366             'thumbnail': r're:https://.+\.webp.*',
1367             'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAAdZ_NcPPgMneaGrW0hN8O_J_bwLshwNNERRF5DxOw2HKIzk0kdlLrR8RkVl1ksrMO',
1368             'duration': 29,
1369             'timestamp': 1654680400,
1370             'repost_count': int,
1371             'artist': 'Akihitoko',
1372             'track': 'original sound',
1373             'description': 'md5:edc3e7ea587847f8537468f2fe51d074',
1374             'uploader': 'akihitoko1',
1375             'creator': 'Akihitoko',
1376         },
1377     }, {
1378         'url': 'https://vt.tiktok.com/ZSe4FqkKd',
1379         'only_matching': True,
1380     }]
1381
1382     def _real_extract(self, url):
1383         new_url = self._request_webpage(
1384             HEADRequest(url), self._match_id(url), headers={'User-Agent': 'facebookexternalhit/1.1'}).url
1385         if self.suitable(new_url):  # Prevent infinite loop in case redirect fails
1386             raise UnsupportedError(new_url)
1387         return self.url_result(new_url)
1388
1389
1390 class TikTokLiveIE(TikTokBaseIE):
1391     _VALID_URL = r'''(?x)https?://(?:
1392         (?:www\.)?tiktok\.com/@(?P<uploader>[\w.-]+)/live|
1393         m\.tiktok\.com/share/live/(?P<id>\d+)
1394     )'''
1395     IE_NAME = 'tiktok:live'
1396
1397     _TESTS = [{
1398         'url': 'https://www.tiktok.com/@weathernewslive/live',
1399         'info_dict': {
1400             'id': '7210809319192726273',
1401             'ext': 'mp4',
1402             'title': r're:ウェザーニュースLiVE[\d\s:-]*',
1403             'creator': 'ウェザーニュースLiVE',
1404             'uploader': 'weathernewslive',
1405             'uploader_id': '6621496731283095554',
1406             'uploader_url': 'https://www.tiktok.com/@weathernewslive',
1407             'live_status': 'is_live',
1408             'concurrent_view_count': int,
1409         },
1410         'params': {'skip_download': 'm3u8'},
1411     }, {
1412         'url': 'https://www.tiktok.com/@pilarmagenta/live',
1413         'info_dict': {
1414             'id': '7209423610325322522',
1415             'ext': 'mp4',
1416             'title': str,
1417             'creator': 'Pilarmagenta',
1418             'uploader': 'pilarmagenta',
1419             'uploader_id': '6624846890674683909',
1420             'uploader_url': 'https://www.tiktok.com/@pilarmagenta',
1421             'live_status': 'is_live',
1422             'concurrent_view_count': int,
1423         },
1424         'skip': 'Livestream',
1425     }, {
1426         'url': 'https://m.tiktok.com/share/live/7209423610325322522/?language=en',
1427         'only_matching': True,
1428     }, {
1429         'url': 'https://www.tiktok.com/@iris04201/live',
1430         'only_matching': True,
1431     }]
1432
1433     def _call_api(self, url, param, room_id, uploader, key=None):
1434         response = traverse_obj(self._download_json(
1435             url, room_id, fatal=False, query={
1436                 'aid': '1988',
1437                 param: room_id,
1438             }), (key, {dict}), default={})
1439
1440         # status == 2 if live else 4
1441         if int_or_none(response.get('status')) == 2:
1442             return response
1443         # If room_id is obtained via mobile share URL and cannot be refreshed, do not wait for live
1444         elif not uploader:
1445             raise ExtractorError('This livestream has ended', expected=True)
1446         raise UserNotLive(video_id=uploader)
1447
1448     def _real_extract(self, url):
1449         uploader, room_id = self._match_valid_url(url).group('uploader', 'id')
1450         webpage = self._download_webpage(
1451             url, uploader or room_id, headers={'User-Agent': 'Mozilla/5.0'}, fatal=not room_id)
1452
1453         if webpage:
1454             data = self._get_sigi_state(webpage, uploader or room_id)
1455             room_id = (traverse_obj(data, ('UserModule', 'users', ..., 'roomId', {str_or_none}), get_all=False)
1456                        or self._search_regex(r'snssdk\d*://live\?room_id=(\d+)', webpage, 'room ID', default=None)
1457                        or room_id)
1458             uploader = uploader or traverse_obj(
1459                 data, ('LiveRoom', 'liveRoomUserInfo', 'user', 'uniqueId'),
1460                 ('UserModule', 'users', ..., 'uniqueId'), get_all=False, expected_type=str)
1461
1462         if not room_id:
1463             raise UserNotLive(video_id=uploader)
1464
1465         formats = []
1466         live_info = self._call_api(
1467             'https://webcast.tiktok.com/webcast/room/info', 'room_id', room_id, uploader, key='data')
1468
1469         get_quality = qualities(('SD1', 'ld', 'SD2', 'sd', 'HD1', 'hd', 'FULL_HD1', 'uhd', 'ORIGION', 'origin'))
1470         parse_inner = lambda x: self._parse_json(x, None)
1471
1472         for quality, stream in traverse_obj(live_info, (
1473                 'stream_url', 'live_core_sdk_data', 'pull_data', 'stream_data',
1474                 {parse_inner}, 'data', {dict}), default={}).items():
1475
1476             sdk_params = traverse_obj(stream, ('main', 'sdk_params', {parse_inner}, {
1477                 'vcodec': ('VCodec', {str}),
1478                 'tbr': ('vbitrate', {lambda x: int_or_none(x, 1000)}),
1479                 'resolution': ('resolution', {lambda x: re.match(r'(?i)\d+x\d+|\d+p', x).group().lower()}),
1480             }))
1481
1482             flv_url = traverse_obj(stream, ('main', 'flv', {url_or_none}))
1483             if flv_url:
1484                 formats.append({
1485                     'url': flv_url,
1486                     'ext': 'flv',
1487                     'format_id': f'flv-{quality}',
1488                     'quality': get_quality(quality),
1489                     **sdk_params,
1490                 })
1491
1492             hls_url = traverse_obj(stream, ('main', 'hls', {url_or_none}))
1493             if hls_url:
1494                 formats.append({
1495                     'url': hls_url,
1496                     'ext': 'mp4',
1497                     'protocol': 'm3u8_native',
1498                     'format_id': f'hls-{quality}',
1499                     'quality': get_quality(quality),
1500                     **sdk_params,
1501                 })
1502
1503         def get_vcodec(*keys):
1504             return traverse_obj(live_info, (
1505                 'stream_url', *keys, {parse_inner}, 'VCodec', {str}))
1506
1507         for stream in ('hls', 'rtmp'):
1508             stream_url = traverse_obj(live_info, ('stream_url', f'{stream}_pull_url', {url_or_none}))
1509             if stream_url:
1510                 formats.append({
1511                     'url': stream_url,
1512                     'ext': 'mp4' if stream == 'hls' else 'flv',
1513                     'protocol': 'm3u8_native' if stream == 'hls' else 'https',
1514                     'format_id': f'{stream}-pull',
1515                     'vcodec': get_vcodec(f'{stream}_pull_url_params'),
1516                     'quality': get_quality('ORIGION'),
1517                 })
1518
1519         for f_id, f_url in traverse_obj(live_info, ('stream_url', 'flv_pull_url', {dict}), default={}).items():
1520             if not url_or_none(f_url):
1521                 continue
1522             formats.append({
1523                 'url': f_url,
1524                 'ext': 'flv',
1525                 'format_id': f'flv-{f_id}'.lower(),
1526                 'vcodec': get_vcodec('flv_pull_url_params', f_id),
1527                 'quality': get_quality(f_id),
1528             })
1529
1530         # If uploader is a guest on another's livestream, primary endpoint will not have m3u8 URLs
1531         if not traverse_obj(formats, lambda _, v: v['ext'] == 'mp4'):
1532             live_info = merge_dicts(live_info, self._call_api(
1533                 'https://www.tiktok.com/api/live/detail/', 'roomID', room_id, uploader, key='LiveRoomInfo'))
1534             if url_or_none(live_info.get('liveUrl')):
1535                 formats.append({
1536                     'url': live_info['liveUrl'],
1537                     'ext': 'mp4',
1538                     'protocol': 'm3u8_native',
1539                     'format_id': 'hls-fallback',
1540                     'vcodec': 'h264',
1541                     'quality': get_quality('origin'),
1542                 })
1543
1544         uploader = uploader or traverse_obj(live_info, ('ownerInfo', 'uniqueId'), ('owner', 'display_id'))
1545
1546         return {
1547             'id': room_id,
1548             'uploader': uploader,
1549             'uploader_url': format_field(uploader, None, self._UPLOADER_URL_FORMAT) or None,
1550             'is_live': True,
1551             'formats': formats,
1552             '_format_sort_fields': ('quality', 'ext'),
1553             **traverse_obj(live_info, {
1554                 'title': 'title',
1555                 'uploader_id': (('ownerInfo', 'owner'), 'id', {str_or_none}),
1556                 'creator': (('ownerInfo', 'owner'), 'nickname'),
1557                 'concurrent_view_count': (('user_count', ('liveRoomStats', 'userCount')), {int_or_none}),
1558             }, get_all=False),
1559         }