import itertools import json import random import re import string import time from .common import InfoExtractor from ..compat import compat_urllib_parse_unquote, compat_urllib_parse_urlparse from ..networking import HEADRequest from ..utils import ( ExtractorError, LazyList, UnsupportedError, UserNotLive, determine_ext, format_field, get_element_by_id, get_first, int_or_none, join_nonempty, merge_dicts, qualities, remove_start, srt_subtitles_timecode, str_or_none, traverse_obj, try_call, try_get, url_or_none, ) class TikTokBaseIE(InfoExtractor): _APP_VERSIONS = [('26.1.3', '260103'), ('26.1.2', '260102'), ('26.1.1', '260101'), ('25.6.2', '250602')] _WORKING_APP_VERSION = None _APP_NAME = 'trill' _AID = 1180 _UPLOADER_URL_FORMAT = 'https://www.tiktok.com/@%s' _WEBPAGE_HOST = 'https://www.tiktok.com/' QUALITIES = ('360p', '540p', '720p', '1080p') @property def _API_HOSTNAME(self): return self._configuration_arg( 'api_hostname', ['api16-normal-c-useast1a.tiktokv.com'], ie_key=TikTokIE)[0] @staticmethod def _create_url(user_id, video_id): return f'https://www.tiktok.com/@{user_id or "_"}/video/{video_id}' def _get_sigi_state(self, webpage, display_id): return self._parse_json(get_element_by_id( 'SIGI_STATE|sigi-persisted-data', webpage, escape_value=False), display_id) def _call_api_impl(self, ep, query, manifest_app_version, video_id, fatal=True, note='Downloading API JSON', errnote='Unable to download API page'): self._set_cookie(self._API_HOSTNAME, 'odin_tt', ''.join(random.choices('0123456789abcdef', k=160))) webpage_cookies = self._get_cookies(self._WEBPAGE_HOST) if webpage_cookies.get('sid_tt'): self._set_cookie(self._API_HOSTNAME, 'sid_tt', webpage_cookies['sid_tt'].value) return self._download_json( 'https://%s/aweme/v1/%s/' % (self._API_HOSTNAME, ep), video_id=video_id, fatal=fatal, note=note, errnote=errnote, headers={ 'User-Agent': f'com.ss.android.ugc.{self._APP_NAME}/{manifest_app_version} (Linux; U; Android 13; en_US; Pixel 7; Build/TD1A.220804.031; Cronet/58.0.2991.0)', 'Accept': 'application/json', }, query=query) def _build_api_query(self, query, app_version, manifest_app_version): return { **query, 'version_name': app_version, 'version_code': manifest_app_version, 'build_number': app_version, 'manifest_version_code': manifest_app_version, 'update_version_code': manifest_app_version, 'openudid': ''.join(random.choices('0123456789abcdef', k=16)), 'uuid': ''.join(random.choices(string.digits, k=16)), '_rticket': int(time.time() * 1000), 'ts': int(time.time()), 'device_brand': 'Google', 'device_type': 'Pixel 7', 'device_platform': 'android', 'resolution': '1080*2400', 'dpi': 420, 'os_version': '13', 'os_api': '29', 'carrier_region': 'US', 'sys_region': 'US', 'region': 'US', 'app_name': self._APP_NAME, 'app_language': 'en', 'language': 'en', 'timezone_name': 'America/New_York', 'timezone_offset': '-14400', 'channel': 'googleplay', 'ac': 'wifi', 'mcc_mnc': '310260', 'is_my_cn': 0, 'aid': self._AID, 'ssmix': 'a', 'as': 'a1qwert123', 'cp': 'cbfhckdckkde1', } def _call_api(self, ep, query, video_id, fatal=True, note='Downloading API JSON', errnote='Unable to download API page'): if not self._WORKING_APP_VERSION: app_version = self._configuration_arg('app_version', [''], ie_key=TikTokIE.ie_key())[0] manifest_app_version = self._configuration_arg('manifest_app_version', [''], ie_key=TikTokIE.ie_key())[0] if app_version and manifest_app_version: self._WORKING_APP_VERSION = (app_version, manifest_app_version) self.write_debug('Imported app version combo from extractor arguments') elif app_version or manifest_app_version: self.report_warning('Only one of the two required version params are passed as extractor arguments', only_once=True) if self._WORKING_APP_VERSION: app_version, manifest_app_version = self._WORKING_APP_VERSION real_query = self._build_api_query(query, app_version, manifest_app_version) return self._call_api_impl(ep, real_query, manifest_app_version, video_id, fatal, note, errnote) for count, (app_version, manifest_app_version) in enumerate(self._APP_VERSIONS, start=1): real_query = self._build_api_query(query, app_version, manifest_app_version) try: res = self._call_api_impl(ep, real_query, manifest_app_version, video_id, fatal, note, errnote) self._WORKING_APP_VERSION = (app_version, manifest_app_version) return res except ExtractorError as e: if isinstance(e.cause, json.JSONDecodeError) and e.cause.pos == 0: if count == len(self._APP_VERSIONS): if fatal: raise e else: self.report_warning(str(e.cause or e.msg)) return self.report_warning('%s. Retrying... (attempt %s of %s)' % (str(e.cause or e.msg), count, len(self._APP_VERSIONS))) continue raise e def _extract_aweme_app(self, aweme_id): feed_list = self._call_api( 'feed', {'aweme_id': aweme_id}, aweme_id, note='Downloading video feed', errnote='Unable to download video feed').get('aweme_list') or [] aweme_detail = next((aweme for aweme in feed_list if str(aweme.get('aweme_id')) == aweme_id), None) if not aweme_detail: raise ExtractorError('Unable to find video in feed', video_id=aweme_id) return self._parse_aweme_video_app(aweme_detail) def _get_subtitles(self, aweme_detail, aweme_id): # TODO: Extract text positioning info subtitles = {} # aweme/detail endpoint subs captions_info = traverse_obj( aweme_detail, ('interaction_stickers', ..., 'auto_video_caption_info', 'auto_captions', ...), expected_type=dict) for caption in captions_info: caption_url = traverse_obj(caption, ('url', 'url_list', ...), expected_type=url_or_none, get_all=False) if not caption_url: continue caption_json = self._download_json( caption_url, aweme_id, note='Downloading captions', errnote='Unable to download captions', fatal=False) if not caption_json: continue subtitles.setdefault(caption.get('language', 'en'), []).append({ 'ext': 'srt', 'data': '\n\n'.join( f'{i + 1}\n{srt_subtitles_timecode(line["start_time"] / 1000)} --> {srt_subtitles_timecode(line["end_time"] / 1000)}\n{line["text"]}' for i, line in enumerate(caption_json['utterances']) if line.get('text')) }) # feed endpoint subs if not subtitles: for caption in traverse_obj(aweme_detail, ('video', 'cla_info', 'caption_infos', ...), expected_type=dict): if not caption.get('url'): continue subtitles.setdefault(caption.get('lang') or 'en', []).append({ 'ext': remove_start(caption.get('caption_format'), 'web'), 'url': caption['url'], }) # webpage subs if not subtitles: for caption in traverse_obj(aweme_detail, ('video', 'subtitleInfos', ...), expected_type=dict): if not caption.get('Url'): continue subtitles.setdefault(caption.get('LanguageCodeName') or 'en', []).append({ 'ext': remove_start(caption.get('Format'), 'web'), 'url': caption['Url'], }) return subtitles def _parse_aweme_video_app(self, aweme_detail): aweme_id = aweme_detail['aweme_id'] video_info = aweme_detail['video'] def parse_url_key(url_key): format_id, codec, res, bitrate = self._search_regex( r'v[^_]+_(?P(?P[^_]+)_(?P\d+p)_(?P\d+))', url_key, 'url key', default=(None, None, None, None), group=('id', 'codec', 'res', 'bitrate')) if not format_id: return {}, None return { 'format_id': format_id, 'vcodec': 'h265' if codec == 'bytevc1' else codec, 'tbr': int_or_none(bitrate, scale=1000) or None, 'quality': qualities(self.QUALITIES)(res), }, res known_resolutions = {} def audio_meta(url): ext = determine_ext(url, default_ext='m4a') return { 'format_note': 'Music track', 'ext': ext, 'acodec': 'aac' if ext == 'm4a' else ext, 'vcodec': 'none', 'width': None, 'height': None, } if ext == 'mp3' or '-music-' in url else {} def extract_addr(addr, add_meta={}): parsed_meta, res = parse_url_key(addr.get('url_key', '')) if res: known_resolutions.setdefault(res, {}).setdefault('height', add_meta.get('height') or addr.get('height')) known_resolutions[res].setdefault('width', add_meta.get('width') or addr.get('width')) parsed_meta.update(known_resolutions.get(res, {})) add_meta.setdefault('height', int_or_none(res[:-1])) return [{ 'url': url, 'filesize': int_or_none(addr.get('data_size')), 'ext': 'mp4', 'acodec': 'aac', 'source_preference': -2 if 'aweme/v1' in url else -1, # Downloads from API might get blocked **add_meta, **parsed_meta, 'format_note': join_nonempty( add_meta.get('format_note'), '(API)' if 'aweme/v1' in url else None, delim=' '), **audio_meta(url), } for url in addr.get('url_list') or []] # Hack: Add direct video links first to prioritize them when removing duplicate formats formats = [] if video_info.get('play_addr'): formats.extend(extract_addr(video_info['play_addr'], { 'format_id': 'play_addr', 'format_note': 'Direct video', 'vcodec': 'h265' if traverse_obj( video_info, 'is_bytevc1', 'is_h265') else 'h264', # TODO: Check for "direct iOS" videos, like https://www.tiktok.com/@cookierun_dev/video/7039716639834656002 'width': video_info.get('width'), 'height': video_info.get('height'), })) if video_info.get('download_addr'): formats.extend(extract_addr(video_info['download_addr'], { 'format_id': 'download_addr', 'format_note': 'Download video%s' % (', watermarked' if video_info.get('has_watermark') else ''), 'vcodec': 'h264', 'width': video_info.get('width'), 'height': video_info.get('height'), 'preference': -2 if video_info.get('has_watermark') else -1, })) if video_info.get('play_addr_h264'): formats.extend(extract_addr(video_info['play_addr_h264'], { 'format_id': 'play_addr_h264', 'format_note': 'Direct video', 'vcodec': 'h264', })) if video_info.get('play_addr_bytevc1'): formats.extend(extract_addr(video_info['play_addr_bytevc1'], { 'format_id': 'play_addr_bytevc1', 'format_note': 'Direct video', 'vcodec': 'h265', })) for bitrate in video_info.get('bit_rate', []): if bitrate.get('play_addr'): formats.extend(extract_addr(bitrate['play_addr'], { 'format_id': bitrate.get('gear_name'), 'format_note': 'Playback video', 'tbr': try_get(bitrate, lambda x: x['bit_rate'] / 1000), 'vcodec': 'h265' if traverse_obj( bitrate, 'is_bytevc1', 'is_h265') else 'h264', 'fps': bitrate.get('FPS'), })) self._remove_duplicate_formats(formats) auth_cookie = self._get_cookies(self._WEBPAGE_HOST).get('sid_tt') if auth_cookie: for f in formats: self._set_cookie(compat_urllib_parse_urlparse(f['url']).hostname, 'sid_tt', auth_cookie.value) thumbnails = [] for cover_id in ('cover', 'ai_dynamic_cover', 'animated_cover', 'ai_dynamic_cover_bak', 'origin_cover', 'dynamic_cover'): for cover_url in traverse_obj(video_info, (cover_id, 'url_list', ...)): thumbnails.append({ 'id': cover_id, 'url': cover_url, }) stats_info = aweme_detail.get('statistics') or {} author_info = aweme_detail.get('author') or {} music_info = aweme_detail.get('music') or {} user_url = self._UPLOADER_URL_FORMAT % (traverse_obj(author_info, 'sec_uid', 'id', 'uid', 'unique_id', expected_type=str_or_none, get_all=False)) labels = traverse_obj(aweme_detail, ('hybrid_label', ..., 'text'), expected_type=str) contained_music_track = traverse_obj( music_info, ('matched_song', 'title'), ('matched_pgc_sound', 'title'), expected_type=str) contained_music_author = traverse_obj( music_info, ('matched_song', 'author'), ('matched_pgc_sound', 'author'), 'author', expected_type=str) is_generic_og_trackname = music_info.get('is_original_sound') and music_info.get('title') == 'original sound - %s' % music_info.get('owner_handle') if is_generic_og_trackname: music_track, music_author = contained_music_track or 'original sound', contained_music_author else: music_track, music_author = music_info.get('title'), music_info.get('author') return { 'id': aweme_id, 'extractor_key': TikTokIE.ie_key(), 'extractor': TikTokIE.IE_NAME, 'webpage_url': self._create_url(author_info.get('uid'), aweme_id), **traverse_obj(aweme_detail, { 'title': ('desc', {str}), 'description': ('desc', {str}), 'timestamp': ('create_time', {int_or_none}), }), **traverse_obj(stats_info, { 'view_count': 'play_count', 'like_count': 'digg_count', 'repost_count': 'share_count', 'comment_count': 'comment_count', }, expected_type=int_or_none), **traverse_obj(author_info, { 'uploader': 'unique_id', 'uploader_id': 'uid', 'creator': 'nickname', 'channel_id': 'sec_uid', }, expected_type=str_or_none), 'uploader_url': user_url, 'track': music_track, 'album': str_or_none(music_info.get('album')) or None, 'artist': music_author or None, 'formats': formats, 'subtitles': self.extract_subtitles(aweme_detail, aweme_id), 'thumbnails': thumbnails, 'duration': int_or_none(traverse_obj(video_info, 'duration', ('download_addr', 'duration')), scale=1000), 'availability': self._availability( is_private='Private' in labels, needs_subscription='Friends only' in labels, is_unlisted='Followers only' in labels), '_format_sort_fields': ('quality', 'codec', 'size', 'br'), } def _parse_aweme_video_web(self, aweme_detail, webpage_url, video_id): video_info = aweme_detail['video'] author_info = traverse_obj(aweme_detail, 'authorInfo', 'author', expected_type=dict, default={}) music_info = aweme_detail.get('music') or {} stats_info = aweme_detail.get('stats') or {} channel_id = traverse_obj(author_info or aweme_detail, (('authorSecId', 'secUid'), {str}), get_all=False) user_url = self._UPLOADER_URL_FORMAT % channel_id if channel_id else None formats = [] width = int_or_none(video_info.get('width')) height = int_or_none(video_info.get('height')) for play_url in traverse_obj(video_info, ('playAddr', ((..., 'src'), None), {url_or_none})): formats.append({ 'url': self._proto_relative_url(play_url), 'ext': 'mp4', 'width': width, 'height': height, }) for download_url in traverse_obj(video_info, (('downloadAddr', ('download', 'url')), {url_or_none})): formats.append({ 'format_id': 'download', 'url': self._proto_relative_url(download_url), 'ext': 'mp4', 'width': width, 'height': height, }) self._remove_duplicate_formats(formats) thumbnails = [] for thumb_url in traverse_obj(aweme_detail, ( (None, 'video'), ('thumbnail', 'cover', 'dynamicCover', 'originCover'), {url_or_none})): thumbnails.append({ 'url': self._proto_relative_url(thumb_url), 'width': width, 'height': height, }) return { 'id': video_id, **traverse_obj(aweme_detail, { 'title': ('desc', {str}), 'description': ('desc', {str}), 'duration': ('video', 'duration', {int_or_none}), 'timestamp': ('createTime', {int_or_none}), }), **traverse_obj(author_info or aweme_detail, { 'creator': ('nickname', {str}), 'uploader': (('uniqueId', 'author'), {str}), 'uploader_id': (('authorId', 'uid', 'id'), {str_or_none}), }, get_all=False), **traverse_obj(stats_info, { 'view_count': 'playCount', 'like_count': 'diggCount', 'repost_count': 'shareCount', 'comment_count': 'commentCount', }, expected_type=int_or_none), **traverse_obj(music_info, { 'track': 'title', 'album': ('album', {lambda x: x or None}), 'artist': 'authorName', }, expected_type=str), 'channel_id': channel_id, 'uploader_url': user_url, 'formats': formats, 'thumbnails': thumbnails, 'http_headers': { 'Referer': webpage_url, } } class TikTokIE(TikTokBaseIE): _VALID_URL = r'https?://www\.tiktok\.com/(?:embed|@(?P[\w\.-]+)?/video)/(?P\d+)' _EMBED_REGEX = [rf'<(?:script|iframe)[^>]+\bsrc=(["\'])(?P{_VALID_URL})'] _TESTS = [{ 'url': 'https://www.tiktok.com/@leenabhushan/video/6748451240264420610', 'md5': '736bb7a466c6f0a6afeb597da1e6f5b7', 'info_dict': { 'id': '6748451240264420610', 'ext': 'mp4', 'title': '#jassmanak #lehanga #leenabhushan', 'description': '#jassmanak #lehanga #leenabhushan', 'duration': 13, 'height': 1024, 'width': 576, 'uploader': 'leenabhushan', 'uploader_id': '6691488002098119685', 'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAA_Eb4t1vodM1IuTy_cvp9CY22RAb59xqrO0Xtz9CYQJvgXaDvZxYnZYRzDWhhgJmy', 'creator': 'facestoriesbyleenabh', 'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?', 'upload_date': '20191016', 'timestamp': 1571246252, 'view_count': int, 'like_count': int, 'repost_count': int, 'comment_count': int, 'artist': 'Ysrbeats', 'album': 'Lehanga', 'track': 'Lehanga', }, 'skip': '404 Not Found', }, { 'url': 'https://www.tiktok.com/@patroxofficial/video/6742501081818877190?langCountry=en', 'md5': '6f3cf8cdd9b28cb8363fe0a9a160695b', 'info_dict': { 'id': '6742501081818877190', 'ext': 'mp4', 'title': 'md5:5e2a23877420bb85ce6521dbee39ba94', 'description': 'md5:5e2a23877420bb85ce6521dbee39ba94', 'duration': 27, 'height': 960, 'width': 540, 'uploader': 'patrox', 'uploader_id': '18702747', 'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAAiFnldaILebi5heDoVU6bn4jBWWycX6-9U3xuNPqZ8Ws', 'channel_id': 'MS4wLjABAAAAiFnldaILebi5heDoVU6bn4jBWWycX6-9U3xuNPqZ8Ws', 'creator': 'patroX', 'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?', 'upload_date': '20190930', 'timestamp': 1569860870, 'view_count': int, 'like_count': int, 'repost_count': int, 'comment_count': int, 'artist': 'Evan Todd, Jessica Keenan Wynn, Alice Lee, Barrett Wilbert Weed & Jon Eidson', 'track': 'Big Fun', }, }, { # Banned audio, only available on the app 'url': 'https://www.tiktok.com/@barudakhb_/video/6984138651336838402', 'info_dict': { 'id': '6984138651336838402', 'ext': 'mp4', 'title': 'Balas @yolaaftwsr hayu yu ? #SquadRandom_ 🔥', 'description': 'Balas @yolaaftwsr hayu yu ? #SquadRandom_ 🔥', 'uploader': 'barudakhb_', 'creator': 'md5:29f238c49bc0c176cb3cef1a9cea9fa6', 'uploader_id': '6974687867511718913', 'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAAbhBwQC-R1iKoix6jDFsF-vBdfx2ABoDjaZrM9fX6arU3w71q3cOWgWuTXn1soZ7d', 'channel_id': 'MS4wLjABAAAAbhBwQC-R1iKoix6jDFsF-vBdfx2ABoDjaZrM9fX6arU3w71q3cOWgWuTXn1soZ7d', 'track': 'Boka Dance', 'artist': 'md5:29f238c49bc0c176cb3cef1a9cea9fa6', 'timestamp': 1626121503, 'duration': 18, 'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?', 'upload_date': '20210712', 'view_count': int, 'like_count': int, 'repost_count': int, 'comment_count': int, }, }, { # Sponsored video, only available with feed workaround 'url': 'https://www.tiktok.com/@MS4wLjABAAAATh8Vewkn0LYM7Fo03iec3qKdeCUOcBIouRk1mkiag6h3o_pQu_dUXvZ2EZlGST7_/video/7042692929109986561', 'info_dict': { 'id': '7042692929109986561', 'ext': 'mp4', 'title': 'Slap and Run!', 'description': 'Slap and Run!', 'uploader': 'user440922249', 'creator': 'Slap And Run', 'uploader_id': '7036055384943690754', 'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAATh8Vewkn0LYM7Fo03iec3qKdeCUOcBIouRk1mkiag6h3o_pQu_dUXvZ2EZlGST7_', 'channel_id': 'MS4wLjABAAAATh8Vewkn0LYM7Fo03iec3qKdeCUOcBIouRk1mkiag6h3o_pQu_dUXvZ2EZlGST7_', 'track': 'Promoted Music', 'timestamp': 1639754738, 'duration': 30, 'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?', 'upload_date': '20211217', 'view_count': int, 'like_count': int, 'repost_count': int, 'comment_count': int, }, 'params': {'skip_download': True}, # XXX: unable to download video data: HTTP Error 403: Forbidden }, { # Video without title and description 'url': 'https://www.tiktok.com/@pokemonlife22/video/7059698374567611694', 'info_dict': { 'id': '7059698374567611694', 'ext': 'mp4', 'title': 'TikTok video #7059698374567611694', 'description': '', 'uploader': 'pokemonlife22', 'creator': 'Pokemon', 'uploader_id': '6820838815978423302', 'uploader_url': 'https://www.tiktok.com/@MS4wLjABAAAA0tF1nBwQVVMyrGu3CqttkNgM68Do1OXUFuCY0CRQk8fEtSVDj89HqoqvbSTmUP2W', 'channel_id': 'MS4wLjABAAAA0tF1nBwQVVMyrGu3CqttkNgM68Do1OXUFuCY0CRQk8fEtSVDj89HqoqvbSTmUP2W', 'track': 'original sound', 'timestamp': 1643714123, 'duration': 6, 'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?', 'upload_date': '20220201', 'artist': 'Pokemon', 'view_count': int, 'like_count': int, 'repost_count': int, 'comment_count': int, }, }, { # hydration JSON is sent in a