[extractor] Support multiple archive ids for one video (#4307)

[yt-dlp.git] / yt_dlp / extractor / youtube.py

diff --git a/yt_dlp/extractor/youtube.py b/yt_dlp/extractor/youtube.py
index 33c0e0b582516605fb9793d8289220b8d24523e8..4dc8e79ac1ba22576cb46293052ba88dad835778 100644 (file)
--- a/yt_dlp/extractor/youtube.py
+++ b/yt_dlp/extractor/youtube.py
@@ -2266,6 +2266,42 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         }
     ]
 
+    _WEBPAGE_TESTS = [
+        # YouTube <object> embed
+        {
+            'url': 'http://www.improbable.com/2017/04/03/untrained-modern-youths-and-ancient-masters-in-selfie-portraits/',
+            'md5': '873c81d308b979f0e23ee7e620b312a3',
+            'info_dict': {
+                'id': 'msN87y-iEx0',
+                'ext': 'mp4',
+                'title': 'Feynman: Mirrors FUN TO IMAGINE 6',
+                'upload_date': '20080526',
+                'description': 'md5:873c81d308b979f0e23ee7e620b312a3',
+                'uploader': 'Christopher Sykes',
+                'uploader_id': 'ChristopherJSykes',
+                'age_limit': 0,
+                'tags': ['feynman', 'mirror', 'science', 'physics', 'imagination', 'fun', 'cool', 'puzzle'],
+                'channel_id': 'UCCeo--lls1vna5YJABWAcVA',
+                'playable_in_embed': True,
+                'thumbnail': 'https://i.ytimg.com/vi/msN87y-iEx0/hqdefault.jpg',
+                'like_count': int,
+                'comment_count': int,
+                'channel': 'Christopher Sykes',
+                'live_status': 'not_live',
+                'channel_url': 'https://www.youtube.com/channel/UCCeo--lls1vna5YJABWAcVA',
+                'availability': 'public',
+                'duration': 195,
+                'view_count': int,
+                'categories': ['Science & Technology'],
+                'channel_follower_count': int,
+                'uploader_url': 'http://www.youtube.com/user/ChristopherJSykes',
+            },
+            'params': {
+                'skip_download': True,
+            }
+        },
+    ]
+
     @classmethod
     def suitable(cls, url):
         from ..utils import parse_qs
@@ -3463,13 +3499,6 @@ def feed_entry(name):
             or get_first(microformats, 'lengthSeconds')
             or parse_duration(search_meta('duration'))) or None
 
-        if get_first(video_details, 'isPostLiveDvr'):
-            self.write_debug('Video is in Post-Live Manifestless mode')
-            if (duration or 0) > 4 * 3600:
-                self.report_warning(
-                    'The livestream has not finished processing. Only 4 hours of the video can be currently downloaded. '
-                    'This is a known issue and patches are welcome')
-
         live_broadcast_details, is_live, streaming_data, formats, automatic_captions = \
             self._list_formats(video_id, microformats, video_details, player_responses, player_url)
 
@@ -3562,8 +3591,7 @@ def feed_entry(name):
 
         formats.extend(self._extract_storyboard(player_responses, duration))
 
-        # Source is given priority since formats that throttle are given lower source_preference
-        # When throttling issue is fully fixed, remove this
+        # source_preference is lower for throttled/potentially damaged formats
         self._sort_formats(formats, ('quality', 'res', 'fps', 'hdr:12', 'source', 'codec:vp9.2', 'lang', 'proto'))
 
         info = {
@@ -3601,6 +3629,14 @@ def feed_entry(name):
             'release_timestamp': live_start_time,
         }
 
+        if get_first(video_details, 'isPostLiveDvr'):
+            self.write_debug('Video is in Post-Live Manifestless mode')
+            info['live_status'] = 'post_live'
+            if (duration or 0) > 4 * 3600:
+                self.report_warning(
+                    'The livestream has not finished processing. Only 4 hours of the video can be currently downloaded. '
+                    'This is a known issue and patches are welcome')
+
         subtitles = {}
         pctr = traverse_obj(player_responses, (..., 'captions', 'playerCaptionsTracklistRenderer'), expected_type=dict)
         if pctr: