Merge 3091daa8345566b6b7b26d9b505e6fde01059333 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/newgrounds.py

@@ -5,7 +5,6 @@ import re
 from .common import InfoExtractor
 from ..utils import (
     extract_attributes,
-    int_or_none,
     parse_duration,
     parse_filesize,
     unified_timestamp,
@@ -27,15 +26,15 @@ class NewgroundsIE(InfoExtractor):
             'duration': 143,
         },
     }, {
-        'url': 'https://www.newgrounds.com/portal/view/673111',
+        'url': 'https://www.newgrounds.com/portal/view/850292',
-        'md5': '3394735822aab2478c31b1004fe5e5bc',
+        'md5': 'bb7cacf45e1b4d648e2dac2d79284d67',
         'info_dict': {
-            'id': '673111',
+            'id': '850292',
             'ext': 'mp4',
-            'title': 'Dancin',
+            'title': 'Timeless (2021)',
-            'uploader': 'Squirrelman82',
+            'uploader': 'Kevuhn',
-            'timestamp': 1460256780,
+            'timestamp': 1657896960,
-            'upload_date': '20160410',
+            'upload_date': '20220715',
         },
     }, {
         # source format unavailable, additional mp4 formats
@@ -44,7 +43,7 @@ class NewgroundsIE(InfoExtractor):
             'id': '689400',
             'ext': 'mp4',
             'title': 'ZTV News Episode 8',
-            'uploader': 'BennettTheSage',
+            'uploader': 'ZONE-SAMA',
             'timestamp': 1487965140,
             'upload_date': '20170224',
         },
@@ -58,40 +57,46 @@ class NewgroundsIE(InfoExtractor):
 
         webpage = self._download_webpage(url, media_id)
 
-        title = self._html_search_regex(
+        paths = url.split('/')
-            r'<title>([^>]+)</title>', webpage, 'title')
+        if paths[-3] == 'audio':
+            isAudio = True
+        else:
+            isAudio = False
 
+        if isAudio:
             media_url = self._parse_json(self._search_regex(
                 r'"url"\s*:\s*("[^"]+"),', webpage, ''), media_id)
 
+            uploader = self._html_search_regex(
+                (r'(?s)<h4[^>]*>(.+?)</h4>.*?<em>\s*Author\s*</em>',
+                 r'(?:Author|Writer)\s*<a[^>]+>([^<]+)'), webpage, 'uploader',
+                fatal=False)
+
             formats = [{
                 'url': media_url,
                 'format_id': 'source',
                 'quality': 1,
             }]
 
-        max_resolution = int_or_none(self._search_regex(
+        else:
-            r'max_resolution["\']\s*:\s*(\d+)', webpage, 'max resolution',
+            media_url = 'https://www.newgrounds.com/portal/video/' + media_id
-            default=None))
+            media = self._download_json(media_url, media_id, headers={'X-Requested-With': 'XMLHttpRequest'})
-        if max_resolution:
+
-            url_base = media_url.rpartition('.')[0]
+            uploader = media['author']
-            for resolution in (360, 720, 1080):
+
-                if resolution > max_resolution:
+            sources = media['sources']
-                    break
+
+            formats = []
+            for source in sources:
+                for i in range(len(sources[source])):
                     formats.append({
-                    'url': '%s.%dp.mp4' % (url_base, resolution),
+                        'url': sources[source][i]['src'],
-                    'format_id': '%dp' % resolution,
+                        'format_id': source,
-                    'height': resolution,
+                        'height': int(source[:-2])  # 1080p -> 1080
                     })
 
-        self._check_formats(formats, media_id)
+        title = self._html_search_regex(
-        self._sort_formats(formats)
+            r'<title>([^>]+)</title>', webpage, 'title')
-
-        uploader = self._html_search_regex(
-            (r'(?s)<h4[^>]*>(.+?)</h4>.*?<em>\s*Author\s*</em>',
-             r'(?:Author|Writer)\s*<a[^>]+>([^<]+)'), webpage, 'uploader',
-            fatal=False)
-
         timestamp = unified_timestamp(self._html_search_regex(
             (r'<dt>\s*Uploaded\s*</dt>\s*<dd>([^<]+</dd>\s*<dd>[^<]+)',
              r'<dt>\s*Uploaded\s*</dt>\s*<dd>([^<]+)'), webpage, 'timestamp',
@@ -99,7 +104,6 @@ class NewgroundsIE(InfoExtractor):
         duration = parse_duration(self._search_regex(
             r'(?s)<dd>\s*Song\s*</dd>\s*<dd>.+?</dd>\s*<dd>([^<]+)', webpage,
             'duration', default=None))
-
         filesize_approx = parse_filesize(self._html_search_regex(
             r'(?s)<dd>\s*Song\s*</dd>\s*<dd>(.+?)</dd>', webpage, 'filesize',
             default=None))
@@ -109,6 +113,9 @@ class NewgroundsIE(InfoExtractor):
         if '<dd>Song' in webpage:
             formats[0]['vcodec'] = 'none'
 
+        self._check_formats(formats, media_id)
+        self._sort_formats(formats)
+
         return {
             'id': media_id,
             'title': title,

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']:
+        for item in traverse_obj(grid_renderer, ('items', Ellipsis, T(dict))):
-            if not isinstance(item, 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: