[extractor/common] Recognize `src` attribute from HTML5 media elements (#3899)

Authored by: Lesmiscore

test/test_InfoExtractor.py

@@ -502,6 +502,24 @@ class TestInfoExtractor(unittest.TestCase):
                 }],
             })
 
+        # from https://0000.studio/
+        # with type attribute but without extension in URL
+        expect_dict(
+            self,
+            self.ie._parse_html5_media_entries(
+                'https://0000.studio',
+                r'''
+                <video src="https://d1ggyt9m8pwf3g.cloudfront.net/protected/ap-northeast-1:1864af40-28d5-492b-b739-b32314b1a527/archive/clip/838db6a7-8973-4cd6-840d-8517e4093c92"
+                    controls="controls" type="video/mp4" preload="metadata" autoplay="autoplay" playsinline class="object-contain">
+                </video>
+                ''', None)[0],
+            {
+                'formats': [{
+                    'url': 'https://d1ggyt9m8pwf3g.cloudfront.net/protected/ap-northeast-1:1864af40-28d5-492b-b739-b32314b1a527/archive/clip/838db6a7-8973-4cd6-840d-8517e4093c92',
+                    'ext': 'mp4',
+                }],
+            })
+
     def test_extract_jwplayer_data_realworld(self):
         # from http://www.suffolk.edu/sjc/
         expect_dict(

yt_dlp/extractor/common.py

@@ -3197,7 +3197,8 @@ class InfoExtractor:
                 return f
             return {}
 
-        def _media_formats(src, cur_media_type, type_info={}):
+        def _media_formats(src, cur_media_type, type_info=None):
+            type_info = type_info or {}
             full_url = absolute_url(src)
             ext = type_info.get('ext') or determine_ext(full_url)
             if ext == 'm3u8':
@@ -3215,6 +3216,7 @@ class InfoExtractor:
                 formats = [{
                     'url': full_url,
                     'vcodec': 'none' if cur_media_type == 'audio' else None,
+                    'ext': ext,
                 }]
             return is_plain_url, formats
 
@@ -3241,7 +3243,8 @@ class InfoExtractor:
             media_attributes = extract_attributes(media_tag)
             src = strip_or_none(media_attributes.get('src'))
             if src:
-                _, formats = _media_formats(src, media_type)
+                f = parse_content_type(media_attributes.get('type'))
+                _, formats = _media_formats(src, media_type, f)
                 media_info['formats'].extend(formats)
             media_info['thumbnail'] = absolute_url(media_attributes.get('poster'))
             if media_content: