[extractor, test] Basic framework for embed tests (#4307)

and split download tests so they can be more easily run in CI

Authored by: coletdjnz

yt_dlp/extractor/common.py

@@ -3665,11 +3665,18 @@ class InfoExtractor:
             t['name'] = cls.ie_key()
             yield t
 
+    @classmethod
+    def get_webpage_testcases(cls):
+        tests = getattr(cls, '_WEBPAGE_TESTS', [])
+        for t in tests:
+            t['name'] = cls.ie_key()
+        return tests
+
     @classproperty
     def age_limit(cls):
         """Get age limit from the testcases"""
         return max(traverse_obj(
-            tuple(cls.get_testcases(include_onlymatching=False)),
+            (*cls.get_testcases(include_onlymatching=False), *cls.get_webpage_testcases()),
             (..., (('playlist', 0), None), 'info_dict', 'age_limit')) or [0])
 
     @classmethod
@@ -3844,7 +3851,10 @@ class InfoExtractor:
     def extract_from_webpage(cls, ydl, url, webpage):
         ie = (cls if isinstance(cls._extract_from_webpage, types.MethodType)
               else ydl.get_info_extractor(cls.ie_key()))
-        yield from ie._extract_from_webpage(url, webpage) or []
+        for info in ie._extract_from_webpage(url, webpage) or []:
+            # url = None since we do not want to set (webpage/original)_url
+            ydl.add_default_extra_info(info, ie, None)
+            yield info
 
     @classmethod
     def _extract_from_webpage(cls, url, webpage):

yt_dlp/extractor/generic.py

@@ -933,21 +933,6 @@ class GenericIE(InfoExtractor):
                 'skip_download': True,
             }
         },
-        # YouTube <object> embed
-        {
-            'url': 'http://www.improbable.com/2017/04/03/untrained-modern-youths-and-ancient-masters-in-selfie-portraits/',
-            'md5': '516718101ec834f74318df76259fb3cc',
-            'info_dict': {
-                'id': 'msN87y-iEx0',
-                'ext': 'webm',
-                'title': 'Feynman: Mirrors FUN TO IMAGINE 6',
-                'upload_date': '20080526',
-                'description': 'md5:0ffc78ea3f01b2e2c247d5f8d1d3c18d',
-                'uploader': 'Christopher Sykes',
-                'uploader_id': 'ChristopherJSykes',
-            },
-            'add_ie': ['Youtube'],
-        },
         # Camtasia studio
         {
             'url': 'http://www.ll.mit.edu/workshops/education/videocourses/antennas/lecture1/video/',

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