Improve sec_uid extraction and error message

bashonly · Nov 14, 2023 · 33da60a · 33da60a
1 parent 38d4483
commit 33da60a
Showing 1 changed file with 23 additions and 8 deletions.
diff --git a/yt_dlp_plugins/extractor/tt_user.py b/yt_dlp_plugins/extractor/tt_user.py
@@ -3,7 +3,7 @@
 import string
 import time
 
-from yt_dlp.utils import ExtractorError, int_or_none, traverse_obj, try_call
+from yt_dlp.utils import ExtractorError, int_or_none, traverse_obj
 from yt_dlp.extractor.tiktok import TikTokIE, TikTokUserIE
 
 
@@ -98,10 +98,20 @@ def _get_sec_uid(self, user_url, user_name, msg):
         webpage = self._download_webpage(
             user_url, user_name, fatal=False, headers={'User-Agent': 'Mozilla/5.0'},
             note=f'Downloading {msg} webpage', errnote=f'Unable to download {msg} webpage')
-        data = try_call(lambda: self._get_sigi_state(webpage, user_name))
-        return traverse_obj(
-            data, ('LiveRoom', 'liveRoomUserInfo', 'user', 'secUid'),
-            ('UserModule', 'users', ..., 'secUid'), get_all=False, expected_type=str)
+        sec_uid = traverse_obj(self._search_json(
+            r'<script[^>]+\bid="__UNIVERSAL_DATA_FOR_REHYDRATION__"[^>]*>', webpage,
+            'rehydration data', user_name, end_pattern=r'</script>', default={}),
+            ('__DEFAULT_SCOPE__', 'webapp.user-detail', 'userInfo', 'user', 'secUid', {str}))
+        if sec_uid:
+            return sec_uid
+        try:
+            return traverse_obj(
+                self._get_sigi_state(webpage, user_name),
+                ('LiveRoom', 'liveRoomUserInfo', 'user', 'secUid'),
+                ('UserModule', 'users', ..., 'secUid'),
+                get_all=False, expected_type=str)
+        except ExtractorError:
+            return None
 
     def _real_extract(self, url):
         user_name = self._match_id(url)
@@ -118,9 +128,11 @@ def _real_extract(self, url):
 
         if not sec_uid:
             webpage = self._download_webpage(
-                f'https://www.tiktok.com/embed/@{user_name}', user_name, note='Downloading user embed page')
+                f'https://www.tiktok.com/embed/@{user_name}', user_name,
+                note='Downloading user embed page', fatal=False)
             data = traverse_obj(self._search_json(
-                r'<script[^>]+\bid=[\'"]__FRONTITY_CONNECT_STATE__[\'"][^>]*>', webpage, 'data', user_name),
+                r'<script[^>]+\bid=[\'"]__FRONTITY_CONNECT_STATE__[\'"][^>]*>',
+                webpage, 'data', user_name, default={}),
                 ('source', 'data', f'/embed/@{user_name}', {dict}))
 
             for aweme_id in traverse_obj(data, ('videoList', ..., 'id')):
@@ -132,7 +144,10 @@ def _real_extract(self, url):
                     break
 
             if not sec_uid:
-                raise ExtractorError('Could not extract secondary user ID')
+                raise ExtractorError(
+                    'Could not extract secondary user ID. '
+                    'Try using  --extractor-arg "tiktok:sec_uid=ID"  with your command, '
+                    'replacing "ID" with the channel_id of the requested user')
 
         return self.playlist_result(self._entries(sec_uid, user_name), user_name)