[yt-dlp.git] / yt_dlp / extractor / tiktok.py

# coding: utf-8
from __future__ import unicode_literals

import itertools
import random
import string
import time
import json

from .common import InfoExtractor
from ..utils import (
    ExtractorError,
    int_or_none,
    str_or_none,
    traverse_obj,
    try_get,
    qualities,
)


class TikTokBaseIE(InfoExtractor):
    _APP_VERSION = '20.9.3'
    _MANIFEST_APP_VERSION = '291'
    QUALITIES = ('360p', '540p', '720p')

    def _call_api(self, ep, query, video_id, fatal=True,
                  note='Downloading API JSON', errnote='Unable to download API page'):
        real_query = {
            **query,
            'version_name': self._APP_VERSION,
            'version_code': self._MANIFEST_APP_VERSION,
            'build_number': self._APP_VERSION,
            'manifest_version_code': self._MANIFEST_APP_VERSION,
            'update_version_code': self._MANIFEST_APP_VERSION,
            'openudid': ''.join(random.choice('0123456789abcdef') for i in range(16)),
            'uuid': ''.join([random.choice(string.digits) for num in range(16)]),
            '_rticket': int(time.time() * 1000),
            'ts': int(time.time()),
            'device_brand': 'Google',
            'device_type': 'Pixel 4',
            'device_platform': 'android',
            'resolution': '1080*1920',
            'dpi': 420,
            'os_version': '10',
            'os_api': '29',
            'carrier_region': 'US',
            'sys_region': 'US',
            'region': 'US',
            'app_name': 'trill',
            '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': 1180,
            'ssmix': 'a',
            'as': 'a1qwert123',
            'cp': 'cbfhckdckkde1',
        }
        self._set_cookie('.tiktokv.com', 'odin_tt', ''.join(random.choice('0123456789abcdef') for i in range(160)))
        return self._download_json(
            'https://api-t2.tiktokv.com/aweme/v1/%s/' % ep, video_id=video_id,
            fatal=fatal, note=note, errnote=errnote, headers={
                'User-Agent': f'com.ss.android.ugc.trill/{self._MANIFEST_APP_VERSION} (Linux; U; Android 10; en_US; Pixel 4; Build/QQ3A.200805.001; Cronet/58.0.2991.0)',
                'Accept': 'application/json',
            }, query=real_query)

    def _parse_aweme_video(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<id>(?P<codec>[^_]+)_(?P<res>\d+p)_(?P<bitrate>\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 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'))
                known_resolutions[res].setdefault('width', add_meta.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(filter(None, (
                    add_meta.get('format_note'), '(API)' if 'aweme/v1' in url else '')))
            } 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',  # Always h264?
                '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',
                }))

        self._remove_duplicate_formats(formats)
        self._sort_formats(formats, ('quality', 'codec', 'size', 'br'))

        thumbnails = []
        for cover_id in ('cover', 'ai_dynamic_cover', 'animated_cover', 'ai_dynamic_cover_bak',
                         'origin_cover', 'dynamic_cover'):
            cover = video_info.get(cover_id)
            if cover:
                for cover_url in cover['url_list']:
                    thumbnails.append({
                        'id': cover_id,
                        'url': cover_url,
                    })

        stats_info = aweme_detail.get('statistics', {})
        author_info = aweme_detail.get('author', {})
        music_info = aweme_detail.get('music', {})
        user_id = str_or_none(author_info.get('nickname'))

        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,
            'title': aweme_detail['desc'],
            'description': aweme_detail['desc'],
            'view_count': int_or_none(stats_info.get('play_count')),
            'like_count': int_or_none(stats_info.get('digg_count')),
            'repost_count': int_or_none(stats_info.get('share_count')),
            'comment_count': int_or_none(stats_info.get('comment_count')),
            'uploader': str_or_none(author_info.get('unique_id')),
            'creator': user_id,
            'uploader_id': str_or_none(author_info.get('uid')),
            'uploader_url': f'https://www.tiktok.com/@{user_id}' if user_id else None,
            'track': music_track,
            'album': str_or_none(music_info.get('album')) or None,
            'artist': music_author,
            'timestamp': int_or_none(aweme_detail.get('create_time')),
            'formats': formats,
            'thumbnails': thumbnails,
            'duration': int_or_none(traverse_obj(video_info, 'duration', ('download_addr', 'duration')), scale=1000)
        }


class TikTokIE(TikTokBaseIE):
    _VALID_URL = r'https?://www\.tiktok\.com/@[\w\.-]+/video/(?P<id>\d+)'

    _TESTS = [{
        'url': 'https://www.tiktok.com/@leenabhushan/video/6748451240264420610',
        'md5': '34a7543afd5a151b0840ba6736fb633b',
        'info_dict': {
            'id': '6748451240264420610',
            'ext': 'mp4',
            'title': '#jassmanak #lehanga #leenabhushan',
            'description': '#jassmanak #lehanga #leenabhushan',
            'duration': 13,
            'height': 1280,
            'width': 720,
            'uploader': 'leenabhushan',
            'uploader_id': '6691488002098119685',
            'uploader_url': 'https://www.tiktok.com/@leenabhushan',
            '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,
        }
    }, {
        'url': 'https://www.tiktok.com/@patroxofficial/video/6742501081818877190?langCountry=en',
        'md5': '06b9800d47d5fe51a19e322dd86e61c9',
        '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/@patrox',
            '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,
        }
    }, {
        # Promoted content/ad
        'url': 'https://www.tiktok.com/@MS4wLjABAAAAAR29F6J2Ktu0Daw03BJyXPNoRQ-W7U5a0Mn3lVCq2rQhjOd_WNLclHUoFgwX8Eno/video/6932675057474981122',
        'only_matching': True,
    }]

    def _extract_aweme(self, props_data, webpage, url):
        video_info = try_get(
            props_data, lambda x: x['pageProps']['itemInfo']['itemStruct'], dict)
        author_info = try_get(
            props_data, lambda x: x['pageProps']['itemInfo']['itemStruct']['author'], dict) or {}
        music_info = try_get(
            props_data, lambda x: x['pageProps']['itemInfo']['itemStruct']['music'], dict) or {}
        stats_info = try_get(props_data, lambda x: x['pageProps']['itemInfo']['itemStruct']['stats'], dict) or {}

        user_id = str_or_none(author_info.get('uniqueId'))
        download_url = try_get(video_info, (lambda x: x['video']['playAddr'],
                                            lambda x: x['video']['downloadAddr']))
        height = try_get(video_info, lambda x: x['video']['height'], int)
        width = try_get(video_info, lambda x: x['video']['width'], int)
        thumbnails = [{
            'url': video_info.get('thumbnail') or self._og_search_thumbnail(webpage),
            'width': width,
            'height': height
        }]
        tracker = try_get(props_data, lambda x: x['initialProps']['$wid'])

        return {
            'id': str_or_none(video_info.get('id')),
            'url': download_url,
            'ext': 'mp4',
            'height': height,
            'width': width,
            'title': video_info.get('desc') or self._og_search_title(webpage),
            'duration': try_get(video_info, lambda x: x['video']['duration'], int),
            'view_count': int_or_none(stats_info.get('playCount')),
            'like_count': int_or_none(stats_info.get('diggCount')),
            'repost_count': int_or_none(stats_info.get('shareCount')),
            'comment_count': int_or_none(stats_info.get('commentCount')),
            'timestamp': try_get(video_info, lambda x: int(x['createTime']), int),
            'creator': str_or_none(author_info.get('nickname')),
            'uploader': user_id,
            'uploader_id': str_or_none(author_info.get('id')),
            'uploader_url': f'https://www.tiktok.com/@{user_id}',
            'track': str_or_none(music_info.get('title')),
            'album': str_or_none(music_info.get('album')) or None,
            'artist': str_or_none(music_info.get('authorName')),
            'thumbnails': thumbnails,
            'description': str_or_none(video_info.get('desc')),
            'webpage_url': self._og_search_url(webpage),
            'http_headers': {
                'Referer': url,
                'Cookie': 'tt_webid=%s; tt_webid_v2=%s' % (tracker, tracker),
            }
        }

    def _extract_aweme_app(self, aweme_id):
        aweme_detail = self._call_api('aweme/detail', {'aweme_id': aweme_id}, aweme_id,
                                      note='Downloading video details', errnote='Unable to download video details')['aweme_detail']
        return self._parse_aweme_video(aweme_detail)

    def _real_extract(self, url):
        video_id = self._match_id(url)

        try:
            return self._extract_aweme_app(video_id)
        except ExtractorError as e:
            self.report_warning(f'{e}; Retrying with webpage')

        # If we only call once, we get a 403 when downlaoding the video.
        self._download_webpage(url, video_id)
        webpage = self._download_webpage(url, video_id, note='Downloading video webpage')
        json_string = self._search_regex(
            r'id=\"__NEXT_DATA__\"\s+type=\"application\/json\"\s*[^>]+>\s*(?P<json_string_ld>[^<]+)',
            webpage, 'json_string', group='json_string_ld')
        json_data = self._parse_json(json_string, video_id)
        props_data = try_get(json_data, lambda x: x['props'], expected_type=dict)

        # Chech statusCode for success
        status = props_data.get('pageProps').get('statusCode')
        if status == 0:
            return self._extract_aweme(props_data, webpage, url)
        elif status == 10216:
            raise ExtractorError('This video is private', expected=True)

        raise ExtractorError('Video not available', video_id=video_id)


class TikTokUserIE(TikTokBaseIE):
    IE_NAME = 'tiktok:user'
    _VALID_URL = r'https?://(?:www\.)?tiktok\.com/@(?P<id>[\w\.-]+)/?(?:$|[#?])'
    _TESTS = [{
        'url': 'https://tiktok.com/@corgibobaa?lang=en',
        'playlist_mincount': 45,
        'info_dict': {
            'id': '6935371178089399301',
        },
    }, {
        'url': 'https://www.tiktok.com/@meme',
        'playlist_mincount': 593,
        'info_dict': {
            'id': '79005827461758976',
        },
    }]

    r'''  # TODO: Fix by adding _signature to api_url
    def _entries(self, webpage, user_id, username):
        secuid = self._search_regex(r'\"secUid\":\"(?P<secUid>[^\"]+)', webpage, username)
        verifyfp_cookie = self._get_cookies('https://www.tiktok.com').get('s_v_web_id')
        if not verifyfp_cookie:
            raise ExtractorError('Improper cookies (missing s_v_web_id).', expected=True)
        api_url = f'https://m.tiktok.com/api/post/item_list/?aid=1988&cookie_enabled=true&count=30&verifyFp={verifyfp_cookie.value}&secUid={secuid}&cursor='
        cursor = '0'
        for page in itertools.count():
            data_json = self._download_json(api_url + cursor, username, note='Downloading Page %d' % page)
            for video in data_json.get('itemList', []):
                video_id = video['id']
                video_url = f'https://www.tiktok.com/@{user_id}/video/{video_id}'
                yield self._url_result(video_url, 'TikTok', video_id, str_or_none(video.get('desc')))
            if not data_json.get('hasMore'):
                break
            cursor = data_json['cursor']
    '''

    def _entries_api(self, webpage, user_id, username):
        query = {
            'user_id': user_id,
            'count': 21,
            'max_cursor': 0,
            'min_cursor': 0,
            'retry_type': 'no_retry',
            'device_id': ''.join(random.choice(string.digits) for i in range(19)),  # Some endpoints don't like randomized device_id, so it isn't directly set in _call_api.
        }

        max_retries = self.get_param('extractor_retries', 3)
        for page in itertools.count(1):
            for retries in itertools.count():
                try:
                    post_list = self._call_api('aweme/post', query, username,
                                               note='Downloading user video list page %d%s' % (page, f' (attempt {retries})' if retries != 0 else ''),
                                               errnote='Unable to download user video list')
                except ExtractorError as e:
                    if isinstance(e.cause, json.JSONDecodeError) and e.cause.pos == 0 and retries != max_retries:
                        self.report_warning('%s. Retrying...' % str(e.cause or e.msg))
                        continue
                    raise
                break
            for video in post_list.get('aweme_list', []):
                yield {
                    **self._parse_aweme_video(video),
                    'ie_key': TikTokIE.ie_key(),
                    'extractor': 'TikTok',
                }
            if not post_list.get('has_more'):
                break
            query['max_cursor'] = post_list['max_cursor']

    def _real_extract(self, url):
        user_id = self._match_id(url)
        webpage = self._download_webpage(url, user_id, headers={
            'User-Agent': 'facebookexternalhit/1.1 (+http://www.facebook.com/externalhit_uatext.php)'
        })
        own_id = self._html_search_regex(r'snssdk\d*://user/profile/(\d+)', webpage, 'user ID')
        return self.playlist_result(self._entries_api(webpage, own_id, user_id), user_id)
Commit	Line	Data
1ead840d KS	1	# coding: utf-8
1ead840d KS	2	from __future__ import unicode_literals
f7f18f90 A	3
f7f18f90 A	4	import itertools
bd9ff55b M	5	import random
	6	import string
	7	import time
0fd6661e	8	import json
1ead840d KS	9
	10	from .common import InfoExtractor
	11	from ..utils import (
ce18a19b	12	ExtractorError,
1ead840d KS	13	int_or_none,
1ead840d KS	14	str_or_none,
bd9ff55b M	15	traverse_obj,
	16	try_get,
	17	qualities,
1ead840d KS	18	)
	19
	20
0fd6661e	21	class TikTokBaseIE(InfoExtractor):
bd9ff55b M	22	_APP_VERSION = '20.9.3'
	23	_MANIFEST_APP_VERSION = '291'
	24	QUALITIES = ('360p', '540p', '720p')
ce18a19b	25
0fd6661e M	26	def _call_api(self, ep, query, video_id, fatal=True,
	27	note='Downloading API JSON', errnote='Unable to download API page'):
	28	real_query = {
	29	**query,
bd9ff55b M	30	'version_name': self._APP_VERSION,
	31	'version_code': self._MANIFEST_APP_VERSION,
	32	'build_number': self._APP_VERSION,
	33	'manifest_version_code': self._MANIFEST_APP_VERSION,
	34	'update_version_code': self._MANIFEST_APP_VERSION,
	35	'openudid': ''.join(random.choice('0123456789abcdef') for i in range(16)),
	36	'uuid': ''.join([random.choice(string.digits) for num in range(16)]),
	37	'_rticket': int(time.time() * 1000),
	38	'ts': int(time.time()),
	39	'device_brand': 'Google',
	40	'device_type': 'Pixel 4',
	41	'device_platform': 'android',
	42	'resolution': '1080*1920',
	43	'dpi': 420,
	44	'os_version': '10',
	45	'os_api': '29',
	46	'carrier_region': 'US',
	47	'sys_region': 'US',
	48	'region': 'US',
	49	'app_name': 'trill',
	50	'app_language': 'en',
	51	'language': 'en',
	52	'timezone_name': 'America/New_York',
	53	'timezone_offset': '-14400',
	54	'channel': 'googleplay',
	55	'ac': 'wifi',
	56	'mcc_mnc': '310260',
	57	'is_my_cn': 0,
	58	'aid': 1180,
	59	'ssmix': 'a',
	60	'as': 'a1qwert123',
	61	'cp': 'cbfhckdckkde1',
	62	}
bd9ff55b	63	self._set_cookie('.tiktokv.com', 'odin_tt', ''.join(random.choice('0123456789abcdef') for i in range(160)))
0fd6661e M	64	return self._download_json(
	65	'https://api-t2.tiktokv.com/aweme/v1/%s/' % ep, video_id=video_id,
	66	fatal=fatal, note=note, errnote=errnote, headers={
bd9ff55b	67	'User-Agent': f'com.ss.android.ugc.trill/{self._MANIFEST_APP_VERSION} (Linux; U; Android 10; en_US; Pixel 4; Build/QQ3A.200805.001; Cronet/58.0.2991.0)',
0fd6661e M	68	'Accept': 'application/json',
	69	}, query=real_query)
	70
	71	def _parse_aweme_video(self, aweme_detail):
	72	aweme_id = aweme_detail['aweme_id']
bd9ff55b M	73	video_info = aweme_detail['video']
	74
	75	def parse_url_key(url_key):
	76	format_id, codec, res, bitrate = self._search_regex(
	77	r'v[^_]+_(?P<id>(?P<codec>[^_]+)_(?P<res>\d+p)_(?P<bitrate>\d+))', url_key,
	78	'url key', default=(None, None, None, None), group=('id', 'codec', 'res', 'bitrate'))
	79	if not format_id:
	80	return {}, None
	81	return {
	82	'format_id': format_id,
	83	'vcodec': 'h265' if codec == 'bytevc1' else codec,
	84	'tbr': int_or_none(bitrate, scale=1000) or None,
	85	'quality': qualities(self.QUALITIES)(res),
	86	}, res
	87
	88	known_resolutions = {}
	89
	90	def extract_addr(addr, add_meta={}):
	91	parsed_meta, res = parse_url_key(addr.get('url_key', ''))
	92	if res:
	93	known_resolutions.setdefault(res, {}).setdefault('height', add_meta.get('height'))
	94	known_resolutions[res].setdefault('width', add_meta.get('width'))
	95	parsed_meta.update(known_resolutions.get(res, {}))
	96	add_meta.setdefault('height', int_or_none(res[:-1]))
	97	return [{
	98	'url': url,
	99	'filesize': int_or_none(addr.get('data_size')),
	100	'ext': 'mp4',
	101	'acodec': 'aac',
0fd6661e M	102	'source_preference': -2 if 'aweme/v1' in url else -1, # Downloads from API might get blocked
	103	add_meta, parsed_meta,
	104	'format_note': ' '.join(filter(None, (
	105	add_meta.get('format_note'), '(API)' if 'aweme/v1' in url else '')))
bd9ff55b M	106	} for url in addr.get('url_list') or []]
	107
	108	# Hack: Add direct video links first to prioritize them when removing duplicate formats
	109	formats = []
	110	if video_info.get('play_addr'):
	111	formats.extend(extract_addr(video_info['play_addr'], {
	112	'format_id': 'play_addr',
	113	'format_note': 'Direct video',
	114	'vcodec': 'h265' if traverse_obj(
	115	video_info, 'is_bytevc1', 'is_h265') else 'h264', # Always h264?
	116	'width': video_info.get('width'),
	117	'height': video_info.get('height'),
	118	}))
	119	if video_info.get('download_addr'):
	120	formats.extend(extract_addr(video_info['download_addr'], {
	121	'format_id': 'download_addr',
	122	'format_note': 'Download video%s' % (', watermarked' if video_info.get('has_watermark') else ''),
	123	'vcodec': 'h264',
	124	'width': video_info.get('width'),
	125	'height': video_info.get('height'),
0fd6661e	126	'preference': -2 if video_info.get('has_watermark') else -1,
bd9ff55b M	127	}))
	128	if video_info.get('play_addr_h264'):
	129	formats.extend(extract_addr(video_info['play_addr_h264'], {
	130	'format_id': 'play_addr_h264',
	131	'format_note': 'Direct video',
	132	'vcodec': 'h264',
	133	}))
	134	if video_info.get('play_addr_bytevc1'):
	135	formats.extend(extract_addr(video_info['play_addr_bytevc1'], {
	136	'format_id': 'play_addr_bytevc1',
	137	'format_note': 'Direct video',
	138	'vcodec': 'h265',
	139	}))
	140
	141	for bitrate in video_info.get('bit_rate', []):
	142	if bitrate.get('play_addr'):
	143	formats.extend(extract_addr(bitrate['play_addr'], {
	144	'format_id': bitrate.get('gear_name'),
	145	'format_note': 'Playback video',
	146	'tbr': try_get(bitrate, lambda x: x['bit_rate'] / 1000),
	147	'vcodec': 'h265' if traverse_obj(
	148	bitrate, 'is_bytevc1', 'is_h265') else 'h264',
	149	}))
	150
	151	self._remove_duplicate_formats(formats)
0fd6661e	152	self._sort_formats(formats, ('quality', 'codec', 'size', 'br'))
bd9ff55b M	153
	154	thumbnails = []
	155	for cover_id in ('cover', 'ai_dynamic_cover', 'animated_cover', 'ai_dynamic_cover_bak',
	156	'origin_cover', 'dynamic_cover'):
	157	cover = video_info.get(cover_id)
	158	if cover:
	159	for cover_url in cover['url_list']:
	160	thumbnails.append({
	161	'id': cover_id,
	162	'url': cover_url,
	163	})
	164
	165	stats_info = aweme_detail.get('statistics', {})
	166	author_info = aweme_detail.get('author', {})
	167	music_info = aweme_detail.get('music', {})
	168	user_id = str_or_none(author_info.get('nickname'))
	169
	170	contained_music_track = traverse_obj(
	171	music_info, ('matched_song', 'title'), ('matched_pgc_sound', 'title'), expected_type=str)
	172	contained_music_author = traverse_obj(
	173	music_info, ('matched_song', 'author'), ('matched_pgc_sound', 'author'), 'author', expected_type=str)
	174
	175	is_generic_og_trackname = music_info.get('is_original_sound') and music_info.get('title') == 'original sound - %s' % music_info.get('owner_handle')
	176	if is_generic_og_trackname:
	177	music_track, music_author = contained_music_track or 'original sound', contained_music_author
	178	else:
	179	music_track, music_author = music_info.get('title'), music_info.get('author')
	180
	181	return {
	182	'id': aweme_id,
	183	'title': aweme_detail['desc'],
	184	'description': aweme_detail['desc'],
	185	'view_count': int_or_none(stats_info.get('play_count')),
	186	'like_count': int_or_none(stats_info.get('digg_count')),
	187	'repost_count': int_or_none(stats_info.get('share_count')),
	188	'comment_count': int_or_none(stats_info.get('comment_count')),
	189	'uploader': str_or_none(author_info.get('unique_id')),
	190	'creator': user_id,
	191	'uploader_id': str_or_none(author_info.get('uid')),
	192	'uploader_url': f'https://www.tiktok.com/@{user_id}' if user_id else None,
	193	'track': music_track,
	194	'album': str_or_none(music_info.get('album')) or None,
	195	'artist': music_author,
	196	'timestamp': int_or_none(aweme_detail.get('create_time')),
	197	'formats': formats,
	198	'thumbnails': thumbnails,
	199	'duration': int_or_none(traverse_obj(video_info, 'duration', ('download_addr', 'duration')), scale=1000)
	200	}
	201
0fd6661e M	202
	203	class TikTokIE(TikTokBaseIE):
	204	_VALID_URL = r'https?://www\.tiktok\.com/@[\w\.-]+/video/(?P<id>\d+)'
	205
	206	_TESTS = [{
	207	'url': 'https://www.tiktok.com/@leenabhushan/video/6748451240264420610',
	208	'md5': '34a7543afd5a151b0840ba6736fb633b',
	209	'info_dict': {
	210	'id': '6748451240264420610',
	211	'ext': 'mp4',
	212	'title': '#jassmanak #lehanga #leenabhushan',
	213	'description': '#jassmanak #lehanga #leenabhushan',
	214	'duration': 13,
	215	'height': 1280,
	216	'width': 720,
	217	'uploader': 'leenabhushan',
	218	'uploader_id': '6691488002098119685',
	219	'uploader_url': 'https://www.tiktok.com/@leenabhushan',
	220	'creator': 'facestoriesbyleenabh',
	221	'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
	222	'upload_date': '20191016',
	223	'timestamp': 1571246252,
	224	'view_count': int,
	225	'like_count': int,
	226	'repost_count': int,
	227	'comment_count': int,
	228	}
	229	}, {
	230	'url': 'https://www.tiktok.com/@patroxofficial/video/6742501081818877190?langCountry=en',
	231	'md5': '06b9800d47d5fe51a19e322dd86e61c9',
	232	'info_dict': {
	233	'id': '6742501081818877190',
	234	'ext': 'mp4',
	235	'title': 'md5:5e2a23877420bb85ce6521dbee39ba94',
	236	'description': 'md5:5e2a23877420bb85ce6521dbee39ba94',
	237	'duration': 27,
	238	'height': 960,
	239	'width': 540,
	240	'uploader': 'patrox',
	241	'uploader_id': '18702747',
	242	'uploader_url': 'https://www.tiktok.com/@patrox',
	243	'creator': 'patroX',
	244	'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
	245	'upload_date': '20190930',
	246	'timestamp': 1569860870,
	247	'view_count': int,
	248	'like_count': int,
	249	'repost_count': int,
	250	'comment_count': int,
	251	}
	252	}, {
	253	# Promoted content/ad
	254	'url': 'https://www.tiktok.com/@MS4wLjABAAAAAR29F6J2Ktu0Daw03BJyXPNoRQ-W7U5a0Mn3lVCq2rQhjOd_WNLclHUoFgwX8Eno/video/6932675057474981122',
	255	'only_matching': True,
	256	}]
	257
	258	def _extract_aweme(self, props_data, webpage, url):
	259	video_info = try_get(
	260	props_data, lambda x: x['pageProps']['itemInfo']['itemStruct'], dict)
	261	author_info = try_get(
	262	props_data, lambda x: x['pageProps']['itemInfo']['itemStruct']['author'], dict) or {}
	263	music_info = try_get(
	264	props_data, lambda x: x['pageProps']['itemInfo']['itemStruct']['music'], dict) or {}
	265	stats_info = try_get(props_data, lambda x: x['pageProps']['itemInfo']['itemStruct']['stats'], dict) or {}
266
267	user_id = str_or_none(author_info.get('uniqueId'))
268	download_url = try_get(video_info, (lambda x: x['video']['playAddr'],
269	lambda x: x['video']['downloadAddr']))
270	height = try_get(video_info, lambda x: x['video']['height'], int)
271	width = try_get(video_info, lambda x: x['video']['width'], int)
272	thumbnails = [{
273	'url': video_info.get('thumbnail') or self._og_search_thumbnail(webpage),
274	'width': width,
275	'height': height
276	}]
277	tracker = try_get(props_data, lambda x: x['initialProps']['$wid'])
278
279	return {
280	'id': str_or_none(video_info.get('id')),
281	'url': download_url,
282	'ext': 'mp4',
283	'height': height,
284	'width': width,
285	'title': video_info.get('desc') or self._og_search_title(webpage),
286	'duration': try_get(video_info, lambda x: x['video']['duration'], int),
287	'view_count': int_or_none(stats_info.get('playCount')),
288	'like_count': int_or_none(stats_info.get('diggCount')),
289	'repost_count': int_or_none(stats_info.get('shareCount')),
290	'comment_count': int_or_none(stats_info.get('commentCount')),
291	'timestamp': try_get(video_info, lambda x: int(x['createTime']), int),
292	'creator': str_or_none(author_info.get('nickname')),
293	'uploader': user_id,
294	'uploader_id': str_or_none(author_info.get('id')),
295	'uploader_url': f'https://www.tiktok.com/@{user_id}',
296	'track': str_or_none(music_info.get('title')),
297	'album': str_or_none(music_info.get('album')) or None,
298	'artist': str_or_none(music_info.get('authorName')),
299	'thumbnails': thumbnails,
300	'description': str_or_none(video_info.get('desc')),
301	'webpage_url': self._og_search_url(webpage),
302	'http_headers': {
303	'Referer': url,
304	'Cookie': 'tt_webid=%s; tt_webid_v2=%s' % (tracker, tracker),
305	}
306	}
307
308	def _extract_aweme_app(self, aweme_id):
309	aweme_detail = self._call_api('aweme/detail', {'aweme_id': aweme_id}, aweme_id,
310	note='Downloading video details', errnote='Unable to download video details')['aweme_detail']
311	return self._parse_aweme_video(aweme_detail)
312
ce18a19b S	313	def _real_extract(self, url):
ce18a19b S	314	video_id = self._match_id(url)
ce18a19b	315
bd9ff55b M	316	try:
	317	return self._extract_aweme_app(video_id)
	318	except ExtractorError as e:
	319	self.report_warning(f'{e}; Retrying with webpage')
	320
7bbc0bbc	321	# If we only call once, we get a 403 when downlaoding the video.
61e76c1e	322	self._download_webpage(url, video_id)
6fb11ca8	323	webpage = self._download_webpage(url, video_id, note='Downloading video webpage')
4b6d03ed	324	json_string = self._search_regex(
6255e567 AG	325	r'id=\"__NEXT_DATA__\"\s+type=\"application\/json\"\s[^>]+>\s(?P<json_string_ld>[^<]+)',
6255e567 AG	326	webpage, 'json_string', group='json_string_ld')
4b6d03ed	327	json_data = self._parse_json(json_string, video_id)
4f5a0ad8	328	props_data = try_get(json_data, lambda x: x['props'], expected_type=dict)
ce18a19b	329
4b6d03ed	330	# Chech statusCode for success
1418a043	331	status = props_data.get('pageProps').get('statusCode')
1418a043	332	if status == 0:
4f5a0ad8	333	return self._extract_aweme(props_data, webpage, url)
1418a043	334	elif status == 10216:
1418a043	335	raise ExtractorError('This video is private', expected=True)
4b6d03ed	336
6fb11ca8	337	raise ExtractorError('Video not available', video_id=video_id)
f7f18f90 A	338
f7f18f90 A	339
0fd6661e	340	class TikTokUserIE(TikTokBaseIE):
f7f18f90	341	IE_NAME = 'tiktok:user'
0fd6661e	342	_VALID_URL = r'https?://(?:www\.)?tiktok\.com/@(?P<id>[\w\.-]+)/?(?:$\|[#?])'
f7f18f90	343	_TESTS = [{
526d74ec	344	'url': 'https://tiktok.com/@corgibobaa?lang=en',
f7f18f90 A	345	'playlist_mincount': 45,
	346	'info_dict': {
	347	'id': '6935371178089399301',
	348	},
f7f18f90 A	349	}, {
	350	'url': 'https://www.tiktok.com/@meme',
	351	'playlist_mincount': 593,
	352	'info_dict': {
	353	'id': '79005827461758976',
	354	},
f7f18f90 A	355	}]
f7f18f90 A	356
0fd6661e M	357	r''' # TODO: Fix by adding _signature to api_url
	358	def _entries(self, webpage, user_id, username):
	359	secuid = self._search_regex(r'\"secUid\":\"(?P<secUid>[^\"]+)', webpage, username)
f7f18f90 A	360	verifyfp_cookie = self._get_cookies('https://www.tiktok.com').get('s_v_web_id')
	361	if not verifyfp_cookie:
	362	raise ExtractorError('Improper cookies (missing s_v_web_id).', expected=True)
	363	api_url = f'https://m.tiktok.com/api/post/item_list/?aid=1988&cookie_enabled=true&count=30&verifyFp={verifyfp_cookie.value}&secUid={secuid}&cursor='
	364	cursor = '0'
	365	for page in itertools.count():
0fd6661e	366	data_json = self._download_json(api_url + cursor, username, note='Downloading Page %d' % page)
f7f18f90 A	367	for video in data_json.get('itemList', []):
	368	video_id = video['id']
	369	video_url = f'https://www.tiktok.com/@{user_id}/video/{video_id}'
bd9ff55b	370	yield self._url_result(video_url, 'TikTok', video_id, str_or_none(video.get('desc')))
526d74ec	371	if not data_json.get('hasMore'):
f7f18f90 A	372	break
f7f18f90 A	373	cursor = data_json['cursor']
0fd6661e M	374	'''
	375
	376	def _entries_api(self, webpage, user_id, username):
	377	query = {
	378	'user_id': user_id,
	379	'count': 21,
	380	'max_cursor': 0,
	381	'min_cursor': 0,
	382	'retry_type': 'no_retry',
	383	'device_id': ''.join(random.choice(string.digits) for i in range(19)), # Some endpoints don't like randomized device_id, so it isn't directly set in _call_api.
	384	}
	385
	386	max_retries = self.get_param('extractor_retries', 3)
	387	for page in itertools.count(1):
	388	for retries in itertools.count():
	389	try:
	390	post_list = self._call_api('aweme/post', query, username,
	391	note='Downloading user video list page %d%s' % (page, f' (attempt {retries})' if retries != 0 else ''),
	392	errnote='Unable to download user video list')
	393	except ExtractorError as e:
	394	if isinstance(e.cause, json.JSONDecodeError) and e.cause.pos == 0 and retries != max_retries:
	395	self.report_warning('%s. Retrying...' % str(e.cause or e.msg))
	396	continue
	397	raise
	398	break
	399	for video in post_list.get('aweme_list', []):
	400	yield {
	401	**self._parse_aweme_video(video),
	402	'ie_key': TikTokIE.ie_key(),
	403	'extractor': 'TikTok',
	404	}
	405	if not post_list.get('has_more'):
	406	break
	407	query['max_cursor'] = post_list['max_cursor']
f7f18f90 A	408
	409	def _real_extract(self, url):
	410	user_id = self._match_id(url)
0fd6661e M	411	webpage = self._download_webpage(url, user_id, headers={
	412	'User-Agent': 'facebookexternalhit/1.1 (+http://www.facebook.com/externalhit_uatext.php)'
	413	})
	414	own_id = self._html_search_regex(r'snssdk\d*://user/profile/(\d+)', webpage, 'user ID')
	415	return self.playlist_result(self._entries_api(webpage, own_id, user_id), user_id)