[extractor/html5] Separate into own extractor (#4307)

Closes #4291

Authored by: coletdjnz, pukkandan

test/test_http.py

@@ -85,7 +85,7 @@ class TestHTTPS(unittest.TestCase):
 
         ydl = YoutubeDL({'logger': FakeLogger(), 'nocheckcertificate': True})
         r = ydl.extract_info('https://127.0.0.1:%d/video.html' % self.port)
-        self.assertEqual(r['entries'][0]['url'], 'https://127.0.0.1:%d/vid.mp4' % self.port)
+        self.assertEqual(r['url'], 'https://127.0.0.1:%d/vid.mp4' % self.port)
 
 
 class TestClientCert(unittest.TestCase):
@@ -113,7 +113,7 @@ class TestClientCert(unittest.TestCase):
             **params,
         })
         r = ydl.extract_info('https://127.0.0.1:%d/video.html' % self.port)
-        self.assertEqual(r['entries'][0]['url'], 'https://127.0.0.1:%d/vid.mp4' % self.port)
+        self.assertEqual(r['url'], 'https://127.0.0.1:%d/vid.mp4' % self.port)
 
     def test_certificate_combined_nopass(self):
         self._run_test(client_certificate=os.path.join(self.certdir, 'clientwithkey.crt'))

yt_dlp/extractor/_extractors.py

@@ -662,6 +662,7 @@ from .hse import (
     HSEShowIE,
     HSEProductIE,
 )
+from .genericembeds import HTML5MediaEmbedIE
 from .huajiao import HuajiaoIE
 from .huya import HuyaLiveIE
 from .huffpost import HuffPostIE

yt_dlp/extractor/generic.py

@@ -3776,25 +3776,6 @@ class GenericIE(InfoExtractor):
         elif embeds:
             return self.playlist_result(embeds, **info_dict)
 
-        # Look for HTML5 media
-        entries = self._parse_html5_media_entries(url, webpage, video_id, m3u8_id='hls')
-        if entries:
-            self.report_detected('HTML5 media')
-            if len(entries) == 1:
-                entries[0].update({
-                    'id': video_id,
-                    'title': video_title,
-                })
-            else:
-                for num, entry in enumerate(entries, start=1):
-                    entry.update({
-                        'id': f'{video_id}-{num}',
-                        'title': '%s (%d)' % (video_title, num),
-                    })
-            for entry in entries:
-                self._sort_formats(entry['formats'])
-            return self.playlist_result(entries, video_id, video_title)
-
         jwplayer_data = self._find_jwplayer_data(
             webpage, video_id, transform_source=js_to_json)
         if jwplayer_data:

yt_dlp/extractor/genericembeds.py

@@ -0,0 +1,27 @@
+from .common import InfoExtractor
+
+
+class HTML5MediaEmbedIE(InfoExtractor):
+    _VALID_URL = False
+    IE_NAME = 'html5'
+    _WEBPAGE_TESTS = [
+        {
+            'url': 'https://html.com/media/',
+            'info_dict': {
+                'title': 'HTML5 Media',
+                'description': 'md5:933b2d02ceffe7a7a0f3c8326d91cc2a',
+            },
+            'playlist_count': 2
+        }
+    ]
+
+    def _extract_from_webpage(self, url, webpage):
+        video_id, title = self._generic_id(url), self._generic_title(url)
+        entries = self._parse_html5_media_entries(url, webpage, video_id, m3u8_id='hls') or []
+        for num, entry in enumerate(entries, start=1):
+            entry.update({
+                'id': f'{video_id}-{num}',
+                'title': f'{title} ({num})',
+            })
+            self._sort_formats(entry['formats'])
+            yield entry