Merge e944ee88955b61d2761b2926b5ecf08605963e00 into 673277e510ebd996b62a2fcc76169bf3cce29910

* [YouTube] Extract playlist items from LOCKUP_VIEW_MODEL_...
* resolves #33073
* thx seproDev (yt-dlp/yt-dlp#11615)

Co-authored-by: sepro <sepro@sepr0.com>

youtube_dl/extractor/soundcloud.py

@@ -233,8 +233,8 @@ class SoundcloudIE(InfoExtractor):
                 'id': '583011102',
                 'ext': 'mp3',
                 'title': 'Mezzo Valzer',
-                'description': 'md5:4138d582f81866a530317bae316e8b61',
-                'uploader': 'Micronie',
+                'description': 'md5:8de664f12895716c8ac6f1da6348eb8e',
+                'uploader': 'Giovanni Sarani',
                 'uploader_id': '3352531',
                 'timestamp': 1551394171,
                 'upload_date': '20190228',
@@ -289,10 +289,13 @@ class SoundcloudIE(InfoExtractor):
         raise ExtractorError('Unable to extract client id')
 
     def _download_json(self, *args, **kwargs):
-        non_fatal = kwargs.get('fatal') is False
-        if non_fatal:
-            del kwargs['fatal']
+        non_fatal = not kwargs.get('fatal', True)
         query = kwargs.get('query', {}).copy()
+        oauth_token = self._get_cookies(self._BASE_URL).get('oauth_token')
+        if oauth_token:
+            if not kwargs.get('headers'):
+                kwargs['headers'] = {}
+            kwargs['headers']['authorization'] = 'OAuth ' + oauth_token.value
         for _ in range(2):
             query['client_id'] = self._CLIENT_ID
             kwargs['query'] = query
@@ -319,6 +322,7 @@ class SoundcloudIE(InfoExtractor):
         track_id = compat_str(info['id'])
         title = info['title']
 
+        duration = info.get('duration')
         format_urls = set()
         formats = []
         query = {'client_id': self._CLIENT_ID}
@@ -335,12 +339,14 @@ class SoundcloudIE(InfoExtractor):
                 if urlh:
                     format_url = urlh.geturl()
                     format_urls.add(format_url)
+                    filesize = int_or_none(urlh.headers.get('Content-Length'))
                     formats.append({
                         'format_id': 'download',
                         'ext': urlhandle_detect_ext(urlh) or 'mp3',
-                        'filesize': int_or_none(urlh.headers.get('Content-Length')),
+                        'filesize': filesize,
                         'url': format_url,
                         'preference': 10,
+                        'abr': int_or_none(filesize, invscale=8, scale=duration),
                     })
 
         def invalid_url(url):
@@ -456,7 +462,7 @@ class SoundcloudIE(InfoExtractor):
             'title': title,
             'description': info.get('description'),
             'thumbnails': thumbnails,
-            'duration': float_or_none(info.get('duration'), 1000),
+            'duration': float_or_none(duration, 1000),
             'webpage_url': info.get('permalink_url'),
             'license': info.get('license'),
             'view_count': extract_count('playback'),

youtube_dl/extractor/youtube.py

@@ -27,6 +27,7 @@ from ..compat import (
 )
 from ..jsinterp import JSInterpreter
 from ..utils import (
+    bug_reports_message,
     clean_html,
     dict_get,
     error_to_compat_str,
@@ -65,6 +66,7 @@ from ..utils import (
     url_or_none,
     urlencode_postdata,
     urljoin,
+    variadic,
 )
 
 
@@ -460,6 +462,26 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'uploader': uploader,
         }
 
+    @staticmethod
+    def _extract_thumbnails(data, *path_list, **kw_final_key):
+        """
+        Extract thumbnails from thumbnails dict
+        @param path_list: path list to level that contains 'thumbnails' key
+        """
+        final_key = kw_final_key.get('final_key', 'thumbnails')
+
+        return traverse_obj(data, ((
+            tuple(variadic(path) + (final_key, Ellipsis)
+                  for path in path_list or [()])), {
+            'url': ('url', T(url_or_none),
+                    # Sometimes youtube gives a wrong thumbnail URL. See:
+                    # https://github.com/yt-dlp/yt-dlp/issues/233
+                    # https://github.com/ytdl-org/youtube-dl/issues/28023
+                    T(lambda u: update_url(u, query=None) if u and 'maxresdefault' in u else u)),
+            'height': ('height', T(int_or_none)),
+            'width': ('width', T(int_or_none)),
+        }, T(lambda t: t if t.get('url') else None)))
+
     def _search_results(self, query, params):
         data = {
             'context': {
@@ -3183,8 +3205,12 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
             expected_type=txt_or_none)
 
     def _grid_entries(self, grid_renderer):
-        for item in grid_renderer['items']:
-            if not isinstance(item, dict):
+        for item in traverse_obj(grid_renderer, ('items', Ellipsis, T(dict))):
+            lockup_view_model = traverse_obj(item, ('lockupViewModel', T(dict)))
+            if lockup_view_model:
+                entry = self._extract_lockup_view_model(lockup_view_model)
+                if entry:
+                    yield entry
                 continue
             renderer = self._extract_grid_item_renderer(item)
             if not isinstance(renderer, dict):
@@ -3268,6 +3294,25 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
                 continue
             yield self._extract_video(renderer)
 
+    def _extract_lockup_view_model(self, view_model):
+        content_id = view_model.get('contentId')
+        if not content_id:
+            return
+        content_type = view_model.get('contentType')
+        if content_type not in ('LOCKUP_CONTENT_TYPE_PLAYLIST', 'LOCKUP_CONTENT_TYPE_PODCAST'):
+            self.report_warning(
+                'Unsupported lockup view model content type "{0}"{1}'.format(content_type, bug_reports_message()), only_once=True)
+            return
+        return merge_dicts(self.url_result(
+            update_url_query('https://www.youtube.com/playlist', {'list': content_id}),
+            ie=YoutubeTabIE.ie_key(), video_id=content_id), {
+                'title': traverse_obj(view_model, (
+                    'metadata', 'lockupMetadataViewModel', 'title', 'content', T(compat_str))),
+                'thumbnails': self._extract_thumbnails(view_model, (
+                    'contentImage', 'collectionThumbnailViewModel', 'primaryThumbnail',
+                    'thumbnailViewModel', 'image'), final_key='sources'),
+        })
+
     def _video_entry(self, video_renderer):
         video_id = video_renderer.get('videoId')
         if video_id: