[extractor/youtube] Add client name to format_note when -v (#6254)

Authored by: Lesmiscore, pukkandan
This commit is contained in:
Lesmiscore 2023-03-12 02:03:23 +09:00 committed by GitHub
parent 7a6c8a0807
commit c795c39f27
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23

View file

@ -66,6 +66,8 @@
variadic, variadic,
) )
STREAMING_DATA_CLIENT_NAME = '__yt_dlp_client'
# any clients starting with _ cannot be explicitly requested by the user # any clients starting with _ cannot be explicitly requested by the user
INNERTUBE_CLIENTS = { INNERTUBE_CLIENTS = {
'web': { 'web': {
@ -248,6 +250,11 @@ def _split_innertube_client(client_name):
return client_name, base, variant[0] if variant else None return client_name, base, variant[0] if variant else None
def short_client_name(client_name):
main, *parts = _split_innertube_client(client_name)[0].replace('embedscreen', 'e_s').split('_')
return join_nonempty(main[:4], ''.join(x[0] for x in parts)).upper()
def build_innertube_clients(): def build_innertube_clients():
THIRD_PARTY = { THIRD_PARTY = {
'embedUrl': 'https://www.youtube.com/', # Can be any valid URL 'embedUrl': 'https://www.youtube.com/', # Can be any valid URL
@ -3594,6 +3601,12 @@ def append_client(*client_names):
self.report_warning( self.report_warning(
f'Skipping player response from {client} client (got player response for video "{pr_video_id}" instead of "{video_id}")' + bug_reports_message()) f'Skipping player response from {client} client (got player response for video "{pr_video_id}" instead of "{video_id}")' + bug_reports_message())
else: else:
# Save client name for introspection later
name = short_client_name(client)
sd = traverse_obj(pr, ('streamingData', {dict})) or {}
sd[STREAMING_DATA_CLIENT_NAME] = name
for f in traverse_obj(sd, (('formats', 'adaptiveFormats'), ..., {dict})):
f[STREAMING_DATA_CLIENT_NAME] = name
prs.append(pr) prs.append(pr)
# creator clients can bypass AGE_VERIFICATION_REQUIRED if logged in # creator clients can bypass AGE_VERIFICATION_REQUIRED if logged in
@ -3712,6 +3725,8 @@ def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, l
if is_damaged: if is_damaged:
self.report_warning( self.report_warning(
f'{video_id}: Some formats are possibly damaged. They will be deprioritized', only_once=True) f'{video_id}: Some formats are possibly damaged. They will be deprioritized', only_once=True)
client_name = fmt.get(STREAMING_DATA_CLIENT_NAME)
dct = { dct = {
'asr': int_or_none(fmt.get('audioSampleRate')), 'asr': int_or_none(fmt.get('audioSampleRate')),
'filesize': int_or_none(fmt.get('contentLength')), 'filesize': int_or_none(fmt.get('contentLength')),
@ -3723,7 +3738,9 @@ def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, l
fmt.get('isDrc') and 'DRC', fmt.get('isDrc') and 'DRC',
try_get(fmt, lambda x: x['projectionType'].replace('RECTANGULAR', '').lower()), try_get(fmt, lambda x: x['projectionType'].replace('RECTANGULAR', '').lower()),
try_get(fmt, lambda x: x['spatialAudioType'].replace('SPATIAL_AUDIO_TYPE_', '').lower()), try_get(fmt, lambda x: x['spatialAudioType'].replace('SPATIAL_AUDIO_TYPE_', '').lower()),
throttled and 'THROTTLED', is_damaged and 'DAMAGED', delim=', '), throttled and 'THROTTLED', is_damaged and 'DAMAGED',
self.get_param('verbose') and client_name,
delim=', '),
# Format 22 is likely to be damaged. See https://github.com/yt-dlp/yt-dlp/issues/3372 # 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, 'source_preference': -10 if throttled else -5 if itag == '22' else -1,
'fps': int_or_none(fmt.get('fps')) or None, 'fps': int_or_none(fmt.get('fps')) or None,
@ -3784,7 +3801,7 @@ def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, l
elif skip_bad_formats and live_status == 'is_live' and needs_live_processing != 'is_live': elif skip_bad_formats and live_status == 'is_live' and needs_live_processing != 'is_live':
skip_manifests.add('dash') skip_manifests.add('dash')
def process_manifest_format(f, proto, itag): def process_manifest_format(f, proto, client_name, itag):
key = (proto, f.get('language')) key = (proto, f.get('language'))
if key in itags[itag]: if key in itags[itag]:
return False return False
@ -3798,17 +3815,21 @@ def process_manifest_format(f, proto, itag):
f['quality'] = q(itag_qualities.get(try_get(f, lambda f: f['format_id'].split('-')[0]), -1)) f['quality'] = q(itag_qualities.get(try_get(f, lambda f: f['format_id'].split('-')[0]), -1))
if f['quality'] == -1 and f.get('height'): if f['quality'] == -1 and f.get('height'):
f['quality'] = q(res_qualities[min(res_qualities, key=lambda x: abs(x - f['height']))]) f['quality'] = q(res_qualities[min(res_qualities, key=lambda x: abs(x - f['height']))])
if self.get_param('verbose'):
f['format_note'] = join_nonempty(f.get('format_note'), client_name, delim=', ')
return True return True
subtitles = {} subtitles = {}
for sd in streaming_data: for sd in streaming_data:
client_name = sd.get(STREAMING_DATA_CLIENT_NAME)
hls_manifest_url = 'hls' not in skip_manifests and sd.get('hlsManifestUrl') hls_manifest_url = 'hls' not in skip_manifests and sd.get('hlsManifestUrl')
if hls_manifest_url: if hls_manifest_url:
fmts, subs = self._extract_m3u8_formats_and_subtitles( fmts, subs = self._extract_m3u8_formats_and_subtitles(
hls_manifest_url, video_id, 'mp4', fatal=False, live=live_status == 'is_live') hls_manifest_url, video_id, 'mp4', fatal=False, live=live_status == 'is_live')
subtitles = self._merge_subtitles(subs, subtitles) subtitles = self._merge_subtitles(subs, subtitles)
for f in fmts: for f in fmts:
if process_manifest_format(f, 'hls', self._search_regex( if process_manifest_format(f, 'hls', client_name, self._search_regex(
r'/itag/(\d+)', f['url'], 'itag', default=None)): r'/itag/(\d+)', f['url'], 'itag', default=None)):
yield f yield f
@ -3817,7 +3838,7 @@ def process_manifest_format(f, proto, itag):
formats, subs = self._extract_mpd_formats_and_subtitles(dash_manifest_url, video_id, fatal=False) formats, subs = self._extract_mpd_formats_and_subtitles(dash_manifest_url, video_id, fatal=False)
subtitles = self._merge_subtitles(subs, subtitles) # Prioritize HLS subs over DASH subtitles = self._merge_subtitles(subs, subtitles) # Prioritize HLS subs over DASH
for f in formats: for f in formats:
if process_manifest_format(f, 'dash', f['format_id']): if process_manifest_format(f, 'dash', client_name, f['format_id']):
f['filesize'] = int_or_none(self._search_regex( f['filesize'] = int_or_none(self._search_regex(
r'/clen/(\d+)', f.get('fragment_base_url') or f['url'], 'file size', default=None)) r'/clen/(\d+)', f.get('fragment_base_url') or f['url'], 'file size', default=None))
if needs_live_processing: if needs_live_processing: