[common] recognize hls manifests that contain video only formats(#11394)

[youtube-dl.git] / youtube_dl / extractor / common.py

diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py
index 50841f0cf936b4c00cc6e384bbe62cc204c7bf4c..6ae946569095a4e6e82a8ff5f04580aa3229ee3a 100644 (file)
--- a/youtube_dl/extractor/common.py
+++ b/youtube_dl/extractor/common.py
@@ -886,7 +886,7 @@ class InfoExtractor(object):
                         'url': e.get('contentUrl'),
                         'title': unescapeHTML(e.get('name')),
                         'description': unescapeHTML(e.get('description')),
-                        'thumbnail': e.get('thumbnailUrl'),
+                        'thumbnail': e.get('thumbnailUrl') or e.get('thumbnailURL'),
                         'duration': parse_duration(e.get('duration')),
                         'timestamp': unified_timestamp(e.get('uploadDate')),
                         'filesize': float_or_none(e.get('contentSize')),
@@ -1224,6 +1224,7 @@ class InfoExtractor(object):
                 'protocol': entry_protocol,
                 'preference': preference,
             }]
+        audio_groups = set()
         last_info = {}
         last_media = {}
         for line in m3u8_doc.splitlines():
@@ -1239,15 +1240,18 @@ class InfoExtractor(object):
                         for v in (media.get('GROUP-ID'), media.get('NAME')):
                             if v:
                                 format_id.append(v)
-                        formats.append({
+                        f = {
                             'format_id': '-'.join(format_id),
                             'url': format_url(media_url),
                             'language': media.get('LANGUAGE'),
-                            'vcodec': 'none' if media_type == 'AUDIO' else None,
                             'ext': ext,
                             'protocol': entry_protocol,
                             'preference': preference,
-                        })
+                        }
+                        if media_type == 'AUDIO':
+                            f['vcodec'] = 'none'
+                            audio_groups.add(media['GROUP-ID'])
+                        formats.append(f)
                     else:
                         # When there is no URI in EXT-X-MEDIA let this tag's
                         # data be used by regular URI lines below
@@ -1280,9 +1284,10 @@ class InfoExtractor(object):
                 }
                 resolution = last_info.get('RESOLUTION')
                 if resolution:
-                    width_str, height_str = resolution.split('x')
-                    f['width'] = int(width_str)
-                    f['height'] = int(height_str)
+                    mobj = re.search(r'(?P<width>\d+)[xX](?P<height>\d+)', resolution)
+                    if mobj:
+                        f['width'] = int(mobj.group('width'))
+                        f['height'] = int(mobj.group('height'))
                 # Unified Streaming Platform
                 mobj = re.search(
                     r'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f['url'])
@@ -1294,6 +1299,9 @@ class InfoExtractor(object):
                         'abr': abr,
                     })
                 f.update(parse_codecs(last_info.get('CODECS')))
+                if last_info.get('AUDIO') in audio_groups:
+                    # TODO: update acodec for for audio only formats with the same GROUP-ID
+                    f['acodec'] = 'none'
                 formats.append(f)
                 last_info = {}
                 last_media = {}
@@ -1702,7 +1710,7 @@ class InfoExtractor(object):
                                 representation_ms_info['fragments'] = [{
                                     'url': media_template % {
                                         'Number': segment_number,
-                                        'Bandwidth': representation_attrib.get('bandwidth'),
+                                        'Bandwidth': int_or_none(representation_attrib.get('bandwidth')),
                                     },
                                     'duration': segment_duration,
                                 } for segment_number in range(
@@ -1720,7 +1728,7 @@ class InfoExtractor(object):
                                 def add_segment_url():
                                     segment_url = media_template % {
                                         'Time': segment_time,
-                                        'Bandwidth': representation_attrib.get('bandwidth'),
+                                        'Bandwidth': int_or_none(representation_attrib.get('bandwidth')),
                                         'Number': segment_number,
                                     }
                                     representation_ms_info['fragments'].append({