[tiktok] fix extraction

youtube_dl/extractor/extractors.py

@@ -1151,10 +1151,7 @@ from .thisamericanlife import ThisAmericanLifeIE
 from .thisav import ThisAVIE
 from .thisoldhouse import ThisOldHouseIE
 from .threeqsdn import ThreeQSDNIE
-from .tiktok import (
+from .tiktok import TikTokIE
-    TikTokIE,
-    TikTokUserIE,
-)
 from .tinypic import TinyPicIE
 from .tmz import (
     TMZIE,

youtube_dl/extractor/tiktok.py

@@ -1,66 +1,128 @@
 # coding: utf-8
 from __future__ import unicode_literals
+from datetime import datetime
+import re
 
 from .common import InfoExtractor
 from ..utils import (
-    compat_str,
     ExtractorError,
     int_or_none,
     str_or_none,
-    try_get,
+    try_get
-    url_or_none,
 )
 
 
 class TikTokBaseIE(InfoExtractor):
-    def _extract_aweme(self, data):
+    def _video_info(self, video_info):
-        video = data['video']
+        return {
-        description = str_or_none(try_get(data, lambda x: x['desc']))
+            'id': str_or_none(video_info.get('id')),
-        width = int_or_none(try_get(data, lambda x: video['width']))
+            'thumbnail': try_get(video_info, lambda x: x['covers'][0], str) or try_get(video_info, lambda x: x['video']['videoMeta']['cover'][0], str),
-        height = int_or_none(try_get(data, lambda x: video['height']))
+            'video_url': try_get(video_info, lambda x: x['video']['urls'][0], str) or video_info.get('video', {}).get('urls', [None])[0],
+            'width': try_get(video_info, lambda x: x['video']['videoMeta']['width'], int) or try_get(video_info, lambda x: x['width'], int),
+            'height': try_get(video_info, lambda x: x['video']['videoMeta']['height'], int) or try_get(video_info, lambda x: x['height'], int),
+            'duration': try_get(video_info, lambda x: x['video']['videoMeta']['duration'], int),
+            'description': str_or_none(video_info.get('text')),
+            'comment_count': int_or_none(video_info.get('commentCount')),
+            'like_count': int_or_none(video_info.get('diggCount')),
+            'repost_count': int_or_none(video_info.get('shareCount')),
+            'timestamp': str_or_none(video_info.get('createTime')),
+            'track_id': str_or_none(video_info.get('musicId'))
+        }
 
-        format_urls = set()
+    def _author_info(self, author_info):
-        formats = []
+        return {
-        for format_id in (
+            'uploader': str_or_none(author_info.get('uniqueId')),
-                'play_addr_lowbr', 'play_addr', 'play_addr_h264',
+            'creator': str_or_none(author_info.get('nickName')),
-                'download_addr'):
+            'uploader_id': str_or_none(author_info.get('userId')),
-            for format in try_get(
+            'channel_id': str_or_none(author_info.get('userId'))
-                    video, lambda x: x[format_id]['url_list'], list) or []:
+        }
-                format_url = url_or_none(format)
+
-                if not format_url:
+    def _track_info(self, track_info):
-                    continue
+        return {
-                if format_url in format_urls:
+            'track': str_or_none(track_info.get('musicName')),
-                    continue
+            'track_id': str_or_none(track_info.get('musicId')),
-                format_urls.add(format_url)
+            'artist': str_or_none(track_info.get('authorName'))
-                formats.append({
+        }
-                    'url': format_url,
+
-                    'ext': 'mp4',
+    def _share_info(self, share_info):
-                    'height': height,
+        return {
-                    'width': width,
+            'title': str_or_none(share_info.get('title')),
+            'description': str_or_none(share_info.get('desc')),
+            'image': try_get(share_info, lambda x: x['image'], dict),
+            'width': try_get(share_info, lambda x: x['image']['width'], int),
+            'height': try_get(share_info, lambda x: x['image']['height'], int),
+        }
+
+    def _extract_aweme(self, video_data, webpage):
+        video_info_data = try_get(
+            video_data, lambda x: x['videoData']['itemInfos'], dict)
+        author_info_data = try_get(
+            video_data, lambda x: x['videoData']['authorInfos'], dict)
+        track_info_data = try_get(
+            video_data, lambda x: x['videoData']['musicInfos'], dict)
+        share_info_data = try_get(video_data, lambda x: x['shareMeta'], dict)
+
+        video_info = self._video_info(video_info_data)
+        author_info = self._author_info(author_info_data)
+        track_info = self._track_info(track_info_data)
+        share_info = self._share_info(share_info_data)
+
+        timestamp = int(video_info.get('timestamp')) or 0
+        date = str_or_none(datetime.fromtimestamp(
+            timestamp).strftime('%Y%m%d'))
+
+        thumbnails = []
+        thumbnails.append({
+            'url': video_info.get('thumbnail') or self._og_search_thumbnail(webpage),
+            'width': video_info.get('width'),
+            'height': video_info.get('height')
         })
-        self._sort_formats(formats)
 
-        thumbnail = url_or_none(try_get(
+        description = video_info.get(
-            video, lambda x: x['cover']['url_list'][0], compat_str))
+            'description') or share_info.get('description')
-        uploader = try_get(data, lambda x: x['author']['nickname'], compat_str)
+        if description is None:
-        timestamp = int_or_none(data.get('create_time'))
+            tags = []
-        comment_count = int_or_none(data.get('comment_count')) or int_or_none(
+        else:
-            try_get(data, lambda x: x['statistics']['comment_count']))
+            tags = re.findall(r"#(\w+)", description)
-        repost_count = int_or_none(try_get(
-            data, lambda x: x['statistics']['share_count']))
 
-        aweme_id = data['aweme_id']
+        formats = []
+        formats.append({
+            'url': video_info.get('video_url') or self._og_search_video_url(webpage),
+            'ext': 'mp4',
+            'height': video_info.get('height'),
+            'width': video_info.get('width'),
+        })
 
         return {
-            'id': aweme_id,
+            'artist': track_info.get('artist'),
-            'title': uploader or aweme_id,
+            'channel_id': author_info.get('channel_id'),
+            'channel_url': 'https://www.tiktok.com/@{}'.format(author_info.get('uploader')),
+            'comment_count': video_info.get('comment_count'),
+            'creator': author_info.get('creator'),
             'description': description,
-            'thumbnail': thumbnail,
+            'duration': video_info.get('duration'),
-            'uploader': uploader,
-            'timestamp': timestamp,
-            'comment_count': comment_count,
-            'repost_count': repost_count,
             'formats': formats,
+            'height': video_info.get('height'),
+            'id': video_info.get('id'),
+            'like_count': video_info.get('like_count'),
+            'playlist_title': share_info.get('title'),
+            'playlist_uploader': author_info.get('uploader'),
+            'playlist_uploader_id': author_info.get('uploader_id'),
+            'repost_count': video_info.get('repost_count'),
+            'release_date': date,
+            'tags': tags,
+            'thumbnail': video_info.get('thumbnail'),
+            'thumbnails': thumbnails,
+            'timestamp': int(video_info.get('timestamp')),
+            'title': share_info.get('title') or self._og_search_title(webpage),
+            'track': track_info.get('track'),
+            'track_id': track_info.get('track_id'),
+            'upload_date': date,
+            'uploader': author_info.get('uploader'),
+            'uploader_id': author_info.get('uploader_id'),
+            'uploader_url': 'https://www.tiktok.com/@{}'.format(author_info.get('uploader')),
+            'webpage_url': self._og_search_url(webpage),
+            'width': video_info.get('width')
         }
 
 
@@ -68,71 +130,123 @@ class TikTokIE(TikTokBaseIE):
     _VALID_URL = r'''(?x)
                  https?://
                  (?:
-                                (?:m\.)?tiktok\.com/v|
+                    (?:www|m)\.
-                                (?:www\.)?tiktok\.com/share/video
+                    (?:tiktok.com)\/
+                    (@(?P<username>[\w\.]+))?
+                    (?:v|video|embed|trending)?(?:\/)?
+                    (?:video)?(?:\/)?
+                    (?:\?shareId=)?
                  )
-                            /(?P<id>\d+)
+                 (?P<id>[\d]{6,})
+                 (?:\.html)?
+                 (?:\?.*)?
+                 $
                  '''
+
     _TESTS = [{
-        'url': 'https://m.tiktok.com/v/6606727368545406213.html',
+        'url': 'https://www.tiktok.com/@leenabhushan/video/6748451240264420610',
-        'md5': 'd584b572e92fcd48888051f238022420',
+        'md5': '34a7543afd5a151b0840ba6736fb633b',
         'info_dict': {
-            'id': '6606727368545406213',
+            'id': '6748451240264420610',
             'ext': 'mp4',
-            'title': 'Zureeal',
+            'title': 'facestoriesbyleenabh on TikTok',
-            'description': '#bowsette#mario#cosplay#uk#lgbt#gaming#asian#bowsettecosplay',
+            'description': 'md5:a9f6c0c44a1ff2249cae610372d0ae95',
-            'thumbnail': r're:^https?://.*~noop.image',
+            'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
-            'uploader': 'Zureeal',
+            'uploader': 'leenabhushan',
-            'timestamp': 1538248586,
+            'timestamp': 1571246252,
-            'upload_date': '20180929',
+            'upload_date': '20191016',
             'comment_count': int,
             'repost_count': int,
+            'like_count': int,
+            'playlist_title': 'facestoriesbyleenabh on TikTok',
+            'playlist_uploader': 'leenabhushan',
+            'playlist_uploader_id': '6691488002098119685',
+            'artist': 'Jass Manak',
+            'channel_id': '6691488002098119685',
+            'channel_url': 'https://www.tiktok.com/@leenabhushan',
+            'creator': 'facestoriesbyleenabh',
+            'duration': 13,
+            'formats': list,
+            'height': 1280,
+            'release_date': '20191016',
+            'tags': list,
+            'thumbnails': list,
+            'track': 'Lehanga',
+            'track_id': '6716465478027447045',
+            'uploader_id': '6691488002098119685',
+            'uploader_url': r're:https://www.tiktok.com/@leenabhushan',
+            'webpage_url': r're:https://www.tiktok.com/@leenabhushan/(video/)?6748451240264420610',
+            'width': 720,
         }
     }, {
-        'url': 'https://www.tiktok.com/share/video/6606727368545406213',
+        'url': 'https://www.tiktok.com/@patroxofficial/video/6742501081818877190?langCountry=en',
-        'only_matching': True,
+        'md5': '06b9800d47d5fe51a19e322dd86e61c9',
+        'info_dict': {
+            'artist': 'Evan Todd, Jessica Keenan Wynn, Alice Lee, Barrett Wilbert Weed & Jon Eidson',
+            'channel_id': '18702747',
+            'channel_url': 'https://www.tiktok.com/@patroxofficial',
+            'comment_count': int,
+            'creator': 'patroX',
+            'description': 'md5:5e2a23877420bb85ce6521dbee39ba94',
+            'duration': 27,
+            'ext': 'mp4',
+            'formats': list,
+            'height': 960,
+            'id': '6742501081818877190',
+            'like_count': int,
+            'playlist_title': 'patroX on TikTok',
+            'playlist_uploader_id': '18702747',
+            'playlist_uploader': 'patroxofficial',
+            'release_date': '20190930',
+            'repost_count': int,
+            'tags': list,
+            'thumbnail': r're:^https?://[\w\/\.\-]+(~[\w\-]+\.image)?',
+            'thumbnails': list,
+            'timestamp': 1569860870,
+            'title': 'patroX on TikTok',
+            'track_id': '209649576000286720',
+            'track': 'Big Fun',
+            'upload_date': '20190930',
+            'uploader_id': '18702747',
+            'uploader_url': r're:https://www.tiktok.com/@patroxofficial',
+            'uploader': 'patroxofficial',
+            'webpage_url': r're:https://www.tiktok.com/@patroxofficial/(video/)?6742501081818877190',
+            'width': 540,
+        }
+    }, {
+        'url': 'https://m.tiktok.com/v/6749869095467945218.html',
+        'only_matching': True
+    }, {
+        'url': 'https://www.tiktok.com/@cchelseameow/video/6751181801206729990',
+        'only_matching': True
+    }, {
+        'url': 'https://www.tiktok.com/embed/6567659045795758085',
+        'only_matching': True
+    }, {
+        'url': 'https://www.tiktok.com/trending?shareId=6744531482393545985',
+        'only_matching': True
+    }, {
+        'url': 'https://www.tiktok.com/@leenabhushan/video/6748451240264420610?enter_from=h5_m',
+        'only_matching': True
     }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        webpage = self._download_webpage(
-            'https://m.tiktok.com/v/%s.html' % video_id, video_id)
-        data = self._parse_json(self._search_regex(
-            r'\bdata\s*=\s*({.+?})\s*;', webpage, 'data'), video_id)
-        return self._extract_aweme(data)
 
+        headers = {
+            'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:72.0) Gecko/20100101 Firefox/72.0',
+            'Referer': url
+        }
+        webpage = self._download_webpage(url, video_id, headers=headers, note='Downloading video webpage')
+        json_string = self._search_regex(
+            r'id=\"__NEXT_DATA__\"\s+type=\"application\/json\">\s*(?P<json_string>[^<]+)',
+            webpage, 'json_string', group='json_string')
+        json_data = self._parse_json(json_string, video_id)
+        video_data = try_get(json_data, lambda x: x['props']['pageProps'], expected_type=dict)
 
-class TikTokUserIE(TikTokBaseIE):
+        # Chech statusCode for success
-    _VALID_URL = r'''(?x)
+        if video_data.get('statusCode') == 0:
-                        https?://
+            return self._extract_aweme(video_data, webpage)
-                            (?:
+
-                                (?:m\.)?tiktok\.com/h5/share/usr|
+        raise ExtractorError("Video not available", video_id=video_id)
-                                (?:www\.)?tiktok\.com/share/user
-                            )
-                            /(?P<id>\d+)
-                    '''
-    _TESTS = [{
-        'url': 'https://m.tiktok.com/h5/share/usr/188294915489964032.html',
-        'info_dict': {
-            'id': '188294915489964032',
-        },
-        'playlist_mincount': 24,
-    }, {
-        'url': 'https://www.tiktok.com/share/user/188294915489964032',
-        'only_matching': True,
-    }]
 
-    def _real_extract(self, url):
-        user_id = self._match_id(url)
-        data = self._download_json(
-            'https://m.tiktok.com/h5/share/usr/list/%s/' % user_id, user_id,
-            query={'_signature': '_'})
-        entries = []
-        for aweme in data['aweme_list']:
-            try:
-                entry = self._extract_aweme(aweme)
-            except ExtractorError:
-                continue
-            entry['extractor_key'] = TikTokIE.ie_key()
-            entries.append(entry)
-        return self.playlist_result(entries, user_id)