[extractor/common] Fix extraction of DASH formats with the same representation id (closes #15111)

This commit is contained in:
Sergey M․ 2017-12-29 23:14:15 +07:00
parent 84f085d4bd
commit 9d6ac71c27
No known key found for this signature in database
GPG key ID: 2C393E0F18A9236D
2 changed files with 19 additions and 10 deletions

View file

@ -493,9 +493,20 @@ jwplayer("mediaplayer").setup({"abouttext":"Visit Indie DB","aboutlink":"http:\/
_TEST_CASES = [
(
# https://github.com/rg3/youtube-dl/issues/13919
# Also tests duplicate representation ids, see
# https://github.com/rg3/youtube-dl/issues/15111
'float_duration',
'http://unknown/manifest.mpd',
[{
'manifest_url': 'http://unknown/manifest.mpd',
'ext': 'm4a',
'format_id': '318597',
'format_note': 'DASH audio',
'protocol': 'http_dash_segments',
'acodec': 'mp4a.40.2',
'vcodec': 'none',
'tbr': 61.587,
}, {
'manifest_url': 'http://unknown/manifest.mpd',
'ext': 'mp4',
'format_id': '318597',

View file

@ -2007,16 +2007,14 @@ class InfoExtractor(object):
f['url'] = initialization_url
f['fragments'].append({location_key(initialization_url): initialization_url})
f['fragments'].extend(representation_ms_info['fragments'])
try:
existing_format = next(
fo for fo in formats
if fo['format_id'] == representation_id)
except StopIteration:
# According to [1, 5.3.5.2, Table 7, page 35] @id of Representation
# is not necessarily unique within a Period thus formats with
# the same `format_id` are quite possible. There are numerous examples
# of such manifests (see https://github.com/rg3/youtube-dl/issues/15111,
# https://github.com/rg3/youtube-dl/issues/13919)
full_info = formats_dict.get(representation_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