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

Closes #4352

yt_dlp/YoutubeDL.py

@@ -3455,11 +3455,9 @@ class YoutubeDL:
         if fn is None:
             return False
 
-        vid_id = self._make_archive_id(info_dict)
-        if not vid_id:
-            return False  # Incomplete video information
-
-        return vid_id in self.archive
+        vid_ids = [self._make_archive_id(info_dict)]
+        vid_ids.extend(info_dict.get('_old_archive_ids', []))
+        return any(id_ in self.archive for id_ in vid_ids)
 
     def record_download_archive(self, info_dict):
         fn = self.params.get('download_archive')

yt_dlp/extractor/common.py

@@ -334,6 +334,7 @@ class InfoExtractor:
                     'private', 'premium_only', 'subscriber_only', 'needs_auth',
                     'unlisted' or 'public'. Use 'InfoExtractor._availability'
                     to set it
+    _old_archive_ids: A list of old archive ids needed for backward compatibility
     __post_extractor: A function to be called just before the metadata is
                     written to either disk, logger or console. The function
                     must return a dict which will be added to the info_dict.

yt_dlp/extractor/funimation.py

@@ -249,7 +249,8 @@ class FunimationIE(FunimationBaseIE):
         self._sort_formats(formats, ('lang', 'source'))
 
         return {
-            'id': initial_experience_id if only_initial_experience else episode_id,
+            'id': episode_id,
+            '_old_archive_ids': [initial_experience_id],
             'display_id': display_id,
             'duration': duration,
             'title': episode['episodeTitle'],

yt_dlp/extractor/genericembeds.py

@@ -22,6 +22,9 @@ class HTML5MediaEmbedIE(InfoExtractor):
             entry.update({
                 'id': f'{video_id}-{num}',
                 'title': f'{title} ({num})',
+                '_old_archive_ids': [
+                    f'Generic {f"{video_id}-{num}" if len(entries) > 1 else video_id}',
+                ],
             })
             self._sort_formats(entry['formats'])
             yield entry

yt_dlp/extractor/twitch.py

@@ -1162,8 +1162,11 @@ class TwitchClipsIE(TwitchBaseIE):
                 })
             thumbnails.append(thumb)
 
+        old_id = self._search_regex(r'%7C(\d+)(?:-\d+)?.mp4', formats[-1]['url'], 'old id', default=None)
+
         return {
             'id': clip.get('id') or video_id,
+            '_old_archive_ids': [f'{self.ie_key()} {old_id}'] if old_id else None,
             'display_id': video_id,
             'title': clip.get('title') or video_id,
             'formats': formats,