[youtube] Skip WebVTT in DASH manifest (#5297)

youtube_dl/extractor/youtube.py

@@ -788,33 +788,41 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             errnote='Could not download DASH manifest')
 
         formats = []
-        for r in dash_doc.findall('.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
+        for a in dash_doc.findall('.//{urn:mpeg:DASH:schema:MPD:2011}AdaptationSet'):
-            url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
+            mime_type = a.attrib.get('mimeType')
-            if url_el is None:
+            for r in a.findall('{urn:mpeg:DASH:schema:MPD:2011}Representation'):
-                continue
+                url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
-            format_id = r.attrib['id']
+                if url_el is None:
-            video_url = url_el.text
+                    continue
-            filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
+                if mime_type == 'text/vtt':
-            f = {
+                    # TODO implement WebVTT downloading
-                'format_id': format_id,
+                    pass
-                'url': video_url,
+                elif mime_type.startswith('audio/') or mime_type.startswith('video/'):
-                'width': int_or_none(r.attrib.get('width')),
+                    format_id = r.attrib['id']
-                'height': int_or_none(r.attrib.get('height')),
+                    video_url = url_el.text
-                'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
+                    filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
-                'asr': int_or_none(r.attrib.get('audioSamplingRate')),
+                    f = {
-                'filesize': filesize,
+                        'format_id': format_id,
-                'fps': int_or_none(r.attrib.get('frameRate')),
+                        'url': video_url,
-            }
+                        'width': int_or_none(r.attrib.get('width')),
-            try:
+                        'height': int_or_none(r.attrib.get('height')),
-                existing_format = next(
+                        'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
-                    fo for fo in formats
+                        'asr': int_or_none(r.attrib.get('audioSamplingRate')),
-                    if fo['format_id'] == format_id)
+                        'filesize': filesize,
-            except StopIteration:
+                        'fps': int_or_none(r.attrib.get('frameRate')),
-                full_info = self._formats.get(format_id, {}).copy()
+                    }
-                full_info.update(f)
+                    try:
-                formats.append(full_info)
+                        existing_format = next(
-            else:
+                            fo for fo in formats
-                existing_format.update(f)
+                            if fo['format_id'] == format_id)
+                    except StopIteration:
+                        full_info = self._formats.get(format_id, {}).copy()
+                        full_info.update(f)
+                        formats.append(full_info)
+                    else:
+                        existing_format.update(f)
+                else:
+                    self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
         return formats
 
     def _real_extract(self, url):