[extractor] Detect more subtitle codecs in MPD manifests (#2174)

Authored by: fstirlitz

yt_dlp/extractor/common.py

@@ -2712,11 +2712,15 @@ class InfoExtractor(object):
                     mime_type = representation_attrib['mimeType']
                     content_type = representation_attrib.get('contentType', mime_type.split('/')[0])
 
-                    codecs = representation_attrib.get('codecs', '')
+                    codecs = parse_codecs(representation_attrib.get('codecs', ''))
                     if content_type not in ('video', 'audio', 'text'):
                         if mime_type == 'image/jpeg':
                             content_type = mime_type
-                        elif codecs.split('.')[0] == 'stpp':
+                        elif codecs['vcodec'] != 'none':
+                            content_type = 'video'
+                        elif codecs['acodec'] != 'none':
+                            content_type = 'audio'
+                        elif codecs.get('tcodec', 'none') != 'none':
                             content_type = 'text'
                         elif mimetype2ext(mime_type) in ('tt', 'dfxp', 'ttml', 'xml', 'json'):
                             content_type = 'text'
@@ -2762,8 +2766,8 @@ class InfoExtractor(object):
                             'format_note': 'DASH %s' % content_type,
                             'filesize': filesize,
                             'container': mimetype2ext(mime_type) + '_dash',
+                            **codecs
                         }
-                        f.update(parse_codecs(codecs))
                     elif content_type == 'text':
                         f = {
                             'ext': mimetype2ext(mime_type),

yt_dlp/utils.py

@@ -3196,7 +3196,7 @@ def parse_codecs(codecs_str):
         return {}
     split_codecs = list(filter(None, map(
         str.strip, codecs_str.strip().strip(',').split(','))))
-    vcodec, acodec, hdr = None, None, None
+    vcodec, acodec, tcodec, hdr = None, None, None, None
     for full_codec in split_codecs:
         parts = full_codec.split('.')
         codec = parts[0].replace('0', '')
@@ -3213,13 +3213,17 @@ def parse_codecs(codecs_str):
         elif codec in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
             if not acodec:
                 acodec = full_codec
+        elif codec in ('stpp', 'wvtt',):
+            if not tcodec:
+                tcodec = full_codec
         else:
             write_string('WARNING: Unknown codec %s\n' % full_codec, sys.stderr)
-    if vcodec or acodec:
+    if vcodec or acodec or tcodec:
         return {
             'vcodec': vcodec or 'none',
             'acodec': acodec or 'none',
             'dynamic_range': hdr,
+            **({'tcodec': tcodec} if tcodec is not None else {}),
         }
     elif len(split_codecs) == 2:
         return {