[tiktok] Improve extraction and add support for user pages (closes #18135)

youtube_dl/extractor/extractors.py

@@ -1124,7 +1124,10 @@ from .thisamericanlife import ThisAmericanLifeIE
 from .thisav import ThisAVIE
 from .thisoldhouse import ThisOldHouseIE
 from .threeqsdn import ThreeQSDNIE
-from .tiktok import TikTokIE
+from .tiktok import (
+    TikTokIE,
+    TikTokUserIE,
+)
 from .tinypic import TinyPicIE
 from .tmz import (
     TMZIE,

youtube_dl/extractor/tiktok.py

@@ -4,6 +4,7 @@ from __future__ import unicode_literals
 from .common import InfoExtractor
 from ..utils import (
     compat_str,
+    ExtractorError,
     int_or_none,
     str_or_none,
     try_get,
@@ -11,69 +12,106 @@ from ..utils import (
 )
 
 
-class TikTokIE(InfoExtractor):
+class TikTokBaseIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:m\.)?tiktok\.com/v/(?P<id>[0-9]+)'
+    def _extract_aweme(self, data):
+        video = data['video']
+        description = str_or_none(try_get(data, lambda x: x['desc']))
+        width = int_or_none(try_get(data, lambda x: video['width']))
+        height = int_or_none(try_get(data, lambda x: video['height']))
+
+        format_urls = set()
+        formats = []
+        for format_id in (
+                'play_addr_lowbr', 'play_addr', 'play_addr_h264',
+                'download_addr'):
+            for format in try_get(
+                    video, lambda x: x[format_id]['url_list'], list) or []:
+                format_url = url_or_none(format)
+                if not format_url:
+                    continue
+                if format_url in format_urls:
+                    continue
+                format_urls.add(format_url)
+                formats.append({
+                    'url': format_url,
+                    'ext': 'mp4',
+                    'height': height,
+                    'width': width,
+                })
+        self._sort_formats(formats)
+
+        thumbnail = url_or_none(try_get(
+            video, lambda x: x['cover']['url_list'][0], compat_str))
+        uploader = try_get(data, lambda x: x['author']['nickname'], compat_str)
+        timestamp = int_or_none(data.get('create_time'))
+        comment_count = int_or_none(data.get('comment_count')) or int_or_none(
+            try_get(data, lambda x: x['statistics']['comment_count']))
+        repost_count = int_or_none(try_get(
+            data, lambda x: x['statistics']['share_count']))
+
+        aweme_id = data['aweme_id']
+
+        return {
+            'id': aweme_id,
+            'title': uploader or aweme_id,
+            'description': description,
+            'thumbnail': thumbnail,
+            'uploader': uploader,
+            'timestamp': timestamp,
+            'comment_count': comment_count,
+            'repost_count': repost_count,
+            'formats': formats,
+        }
+
+
+class TikTokIE(TikTokBaseIE):
+    _VALID_URL = r'https?://(?:m\.)?tiktok\.com/v/(?P<id>\d+)'
     _TEST = {
         'url': 'https://m.tiktok.com/v/6606727368545406213.html',
         'md5': 'd584b572e92fcd48888051f238022420',
         'info_dict': {
             'id': '6606727368545406213',
             'ext': 'mp4',
-            'title': 'Zureeal on TikTok',
+            'title': 'Zureeal',
-            'thumbnail': r're:^https?://.*~noop.image',
             'description': '#bowsette#mario#cosplay#uk#lgbt#gaming#asian#bowsettecosplay',
+            'thumbnail': r're:^https?://.*~noop.image',
             'uploader': 'Zureeal',
-            'width': 540,
+            'timestamp': 1538248586,
-            'height': 960,
+            'upload_date': '20180929',
+            'comment_count': int,
+            'repost_count': int,
         }
     }
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
+        data = self._parse_json(self._search_regex(
+            r'\bdata\s*=\s*({.+?})\s*;', webpage, 'data'), video_id)
+        return self._extract_aweme(data)
 
-        data = self._parse_json(
-            self._search_regex(
-                r'var\s+data\s*=\s*({.+?});', webpage, 'data'
-            ), video_id)
 
-        title = self._og_search_title(webpage)
+class TikTokUserIE(TikTokBaseIE):
+    _VALID_URL = r'https?://(?:m\.)?tiktok\.com/h5/share/usr/(?P<id>\d+)'
+    _TEST = {
+        'url': 'https://m.tiktok.com/h5/share/usr/188294915489964032.html',
+        'info_dict': {
+            'id': '188294915489964032',
+        },
+        'playlist_mincount': 24,
+    }
 
-        description = str_or_none(try_get(data, lambda x: x['desc']))
+    def _real_extract(self, url):
-        width = int_or_none(try_get(data, lambda x: x['video']['width']))
+        user_id = self._match_id(url)
-        height = int_or_none(try_get(data, lambda x: x['video']['height']))
+        data = self._download_json(
-
+            'https://m.tiktok.com/h5/share/usr/list/%s/' % user_id, user_id,
-        formats = []
+            query={'_signature': '_'})
-
+        entries = []
-        for count, (key, label) in enumerate((('play_addr_lowbr', 'Low'), ('play_addr', 'Normal'), ('download_addr', 'Download')), -2):
+        for aweme in data['aweme_list']:
-            for format in try_get(data, lambda x: x['video'][key]['url_list']):
+            try:
-                format_url = url_or_none(format)
+                entry = self._extract_aweme(aweme)
-                if not format_url:
+            except ExtractorError:
-                    continue
+                continue
-                formats.append({
+            entry['extractor_key'] = TikTokIE.ie_key()
-                    'url': format_url,
+            entries.append(entry)
-                    'ext': 'mp4',
+        return self.playlist_result(entries, user_id)
-                    'height': height,
-                    'width': width,
-                    'format_note': label,
-                    'quality': count
-                })
-
-        self._sort_formats(formats)
-
-        uploader = try_get(data, lambda x: x['author']['nickname'], compat_str)
-
-        thumbnail = url_or_none(
-            try_get(
-                data, lambda x: x['video']['cover']['url_list'][0], compat_str))
-
-        return {
-            'id': video_id,
-            'title': title,
-            'description': description,
-            'uploader': uploader,
-            'formats': formats,
-            'thumbnail': thumbnail,
-            'width': width,
-            'height': height,
-        }