[ie/youtube] Skip formats if nsig decoding fails (#10223)

Ref: https://github.com/ytdl-org/youtube-dl/issues/32815

Authored by: bashonly
This commit is contained in:
bashonly 2024-06-21 18:19:59 -05:00 committed by GitHub
parent 96472d72f2
commit 800ec085cc
No known key found for this signature in database
GPG key ID: B5690EEEBB952194

View file

@ -3847,6 +3847,13 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
itag_qualities[itag] = quality
if height:
res_qualities[height] = quality
is_default = audio_track.get('audioIsDefault')
is_descriptive = 'descriptive' in (audio_track.get('displayName') or '').lower()
language_code = audio_track.get('id', '').split('.')[0]
if language_code and is_default:
original_language = language_code
# FORMAT_STREAM_TYPE_OTF(otf=1) requires downloading the init fragment
# (adding `&sq=0` to the URL) and parsing emsg box to determine the
# number of fragment that would subsequently requested with (`&sq=N`)
@ -3872,7 +3879,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
continue
query = parse_qs(fmt_url)
throttled = False
if query.get('n'):
try:
decrypt_nsig = self._cached(self._decrypt_nsig, 'nsig', query['n'][0])
@ -3886,22 +3892,16 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
f'to workaround the issue. {PhantomJSwrapper.INSTALL_HINT}\n')
if player_url:
self.report_warning(
f'nsig extraction failed: You may experience throttling for some formats\n{phantomjs_hint}'
f'nsig extraction failed: Some formats may be missing\n{phantomjs_hint}'
f' n = {query["n"][0]} ; player = {player_url}', video_id=video_id, only_once=True)
self.write_debug(e, only_once=True)
else:
self.report_warning(
'Cannot decrypt nsig without player_url: You may experience throttling for some formats',
'Cannot decrypt nsig without player_url: Some formats may be missing',
video_id=video_id, only_once=True)
throttled = True
continue
tbr = float_or_none(fmt.get('averageBitrate') or fmt.get('bitrate'), 1000)
is_default = audio_track.get('audioIsDefault')
is_descriptive = 'descriptive' in (audio_track.get('displayName') or '').lower()
language_code = audio_track.get('id', '').split('.')[0]
if language_code and is_default:
original_language = language_code
format_duration = traverse_obj(fmt, ('approxDurationMs', {lambda x: float_or_none(x, 1000)}))
# Some formats may have much smaller duration than others (possibly damaged during encoding)
# E.g. 2-nOtRESiUc Ref: https://github.com/yt-dlp/yt-dlp/issues/2823
@ -3932,12 +3932,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
name, fmt.get('isDrc') and 'DRC',
try_get(fmt, lambda x: x['projectionType'].replace('RECTANGULAR', '').lower()),
try_get(fmt, lambda x: x['spatialAudioType'].replace('SPATIAL_AUDIO_TYPE_', '').lower()),
throttled and 'THROTTLED', is_damaged and 'DAMAGED', is_broken and 'BROKEN',
is_damaged and 'DAMAGED', is_broken and 'BROKEN',
(self.get_param('verbose') or all_formats) and client_name,
delim=', '),
# Format 22 is likely to be damaged. See https://github.com/yt-dlp/yt-dlp/issues/3372
'source_preference': ((-10 if throttled else -5 if itag == '22' else -1)
+ (100 if 'Premium' in name else 0)),
'source_preference': (-5 if itag == '22' else -1) + (100 if 'Premium' in name else 0),
'fps': fps if fps > 1 else None, # For some formats, fps is wrongly returned as 1
'audio_channels': fmt.get('audioChannels'),
'height': height,
@ -4357,7 +4356,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
'playable_in_embed': get_first(playability_statuses, 'playableInEmbed'),
'live_status': live_status,
'release_timestamp': live_start_time,
'_format_sort_fields': ( # source_preference is lower for throttled/potentially damaged formats
'_format_sort_fields': ( # source_preference is lower for potentially damaged formats
'quality', 'res', 'fps', 'hdr:12', 'source', 'vcodec:vp9.2', 'channels', 'acodec', 'lang', 'proto'),
}