@ -27,6 +27,7 @@ from ..compat import (
)
from . . jsinterp import JSInterpreter
from . . utils import (
bug_reports_message ,
clean_html ,
dict_get ,
error_to_compat_str ,
@ -48,6 +49,7 @@ from ..utils import (
parse_duration ,
parse_qs ,
qualities ,
remove_end ,
remove_start ,
smuggle_url ,
str_or_none ,
@ -65,6 +67,7 @@ from ..utils import (
url_or_none ,
urlencode_postdata ,
urljoin ,
variadic ,
)
@ -89,12 +92,12 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
' INNERTUBE_CONTEXT ' : {
' client ' : {
' clientName ' : ' IOS ' ,
' clientVersion ' : ' 19.45 .4' ,
' clientVersion ' : ' 20.10 .4' ,
' deviceMake ' : ' Apple ' ,
' deviceModel ' : ' iPhone16,2 ' ,
' userAgent ' : ' com.google.ios.youtube/ 19.45.4 (iPhone16,2; U; CPU iOS 18_1_0 like Mac OS X;)' ,
' userAgent ' : ' com.google.ios.youtube/ 20.10.4 (iPhone16,2; U; CPU iOS 18_3_2 like Mac OS X;)' ,
' osName ' : ' iPhone ' ,
' osVersion ' : ' 18. 1.0.22B83 ' ,
' osVersion ' : ' 18. 3.2.22D82 ' ,
} ,
} ,
' INNERTUBE_CONTEXT_CLIENT_NAME ' : 5 ,
@ -107,7 +110,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
' INNERTUBE_CONTEXT ' : {
' client ' : {
' clientName ' : ' MWEB ' ,
' clientVersion ' : ' 2.202 41202.07 .00' ,
' clientVersion ' : ' 2.202 50311.03 .00' ,
# mweb previously did not require PO Token with this UA
' userAgent ' : ' Mozilla/5.0 (iPad; CPU OS 16_7_10 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/16.6 Mobile/15E148 Safari/604.1,gzip(gfe) ' ,
} ,
@ -120,7 +123,8 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
' INNERTUBE_CONTEXT ' : {
' client ' : {
' clientName ' : ' TVHTML5 ' ,
' clientVersion ' : ' 7.20241201.18.00 ' ,
' clientVersion ' : ' 7.20250312.16.00 ' ,
' userAgent ' : ' Mozilla/5.0 (ChromiumStylePlatform) Cobalt/Version ' ,
} ,
} ,
' INNERTUBE_CONTEXT_CLIENT_NAME ' : 7 ,
@ -130,7 +134,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
' INNERTUBE_CONTEXT ' : {
' client ' : {
' clientName ' : ' WEB ' ,
' clientVersion ' : ' 2.202 41126.01 .00' ,
' clientVersion ' : ' 2.202 50312.04 .00' ,
} ,
} ,
' INNERTUBE_CONTEXT_CLIENT_NAME ' : 1 ,
@ -339,14 +343,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
if not self . _login ( ) :
return
_DEFAULT_API_DATA = {
' context ' : {
' client ' : {
' clientName ' : ' WEB ' ,
' clientVersion ' : ' 2.20201021.03.00 ' ,
} ,
} ,
}
_DEFAULT_API_DATA = { ' context ' : _INNERTUBE_CLIENTS [ ' web ' ] [ ' INNERTUBE_CONTEXT ' ] }
_YT_INITIAL_DATA_RE = r ' (?:window \ s* \ [ \ s*[ " \' ]ytInitialData[ " \' ] \ s* \ ]|ytInitialData) \ s*= \ s*( { .+?}) \ s*; '
_YT_INITIAL_PLAYER_RESPONSE_RE = r ' ytInitialPlayerResponse \ s*= \ s*( { .+?}) \ s*; '
@ -460,6 +457,26 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
' uploader ' : uploader ,
}
@staticmethod
def _extract_thumbnails ( data , * path_list , * * kw_final_key ) :
"""
Extract thumbnails from thumbnails dict
@param path_list : path list to level that contains ' thumbnails ' key
"""
final_key = kw_final_key . get ( ' final_key ' , ' thumbnails ' )
return traverse_obj ( data , ( (
tuple ( variadic ( path ) + ( final_key , Ellipsis )
for path in path_list or [ ( ) ] ) ) , {
' url ' : ( ' url ' , T ( url_or_none ) ,
# Sometimes youtube gives a wrong thumbnail URL. See:
# https://github.com/yt-dlp/yt-dlp/issues/233
# https://github.com/ytdl-org/youtube-dl/issues/28023
T ( lambda u : update_url ( u , query = None ) if u and ' maxresdefault ' in u else u ) ) ,
' height ' : ( ' height ' , T ( int_or_none ) ) ,
' width ' : ( ' width ' , T ( int_or_none ) ) ,
} , T ( lambda t : t if t . get ( ' url ' ) else None ) ) )
def _search_results ( self , query , params ) :
data = {
' context ' : {
@ -474,11 +491,15 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
data [ ' params ' ] = params
for page_num in itertools . count ( 1 ) :
search = self . _download_json (
' https://www.youtube.com/youtubei/v1/search ?key=AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8 ' ,
' https://www.youtube.com/youtubei/v1/search ' ,
video_id = ' query " %s " ' % query ,
note = ' Downloading page %s ' % page_num ,
errnote = ' Unable to download API page ' , fatal = False ,
data = json . dumps ( data ) . encode ( ' utf8 ' ) ,
query = {
# 'key': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8',
' prettyPrint ' : ' false ' ,
} ,
headers = { ' content-type ' : ' application/json ' } )
if not search :
break
@ -669,9 +690,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
' invidious ' : ' | ' . join ( _INVIDIOUS_SITES ) ,
}
_PLAYER_INFO_RE = (
r ' /s/player/(?P<id>[a-zA-Z0-9_-] { 8,})/ player' ,
r ' /(?P<id>[a-zA-Z0-9_-] { 8,})/player(?:_ias \ .vflset(?:/[a-zA-Z] { 2,3}_[a-zA-Z] { 2,3})?|-plasma-ias-(?:phone|tablet)-[a-z] {2} _[A-Z] {2} \ .vflset)/base \ .js$ ' ,
r ' \ b(?P<id>vfl[a-zA-Z0-9_-] + )\ b.*? \ .js$ ' ,
r ' /s/player/(?P<id>[a-zA-Z0-9_-] { 8,})/ (?:tv-)? player' ,
r ' /(?P<id>[a-zA-Z0-9_-] { 8,})/player(?:_ias (?:_tce)? \ .vflset(?:/[a-zA-Z] { 2,3}_[a-zA-Z] { 2,3})?|-plasma-ias-(?:phone|tablet)-[a-z] {2} _[A-Z] {2} \ .vflset)/base \ .js$ ' ,
r ' \ b(?P<id>vfl[a-zA-Z0-9_-] { 6,} \ b.*? \ .js$ ' ,
)
_SUBTITLE_FORMATS = ( ' json3 ' , ' srv1 ' , ' srv2 ' , ' srv3 ' , ' ttml ' , ' vtt ' )
@ -1564,6 +1585,15 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
' 397 ' : { ' acodec ' : ' none ' , ' vcodec ' : ' av01.0.05M.08 ' } ,
}
_PLAYER_JS_VARIANT_MAP = (
( ' main ' , ' player_ias.vflset/en_US/base.js ' ) ,
( ' tce ' , ' player_ias_tce.vflset/en_US/base.js ' ) ,
( ' tv ' , ' tv-player-ias.vflset/tv-player-ias.js ' ) ,
( ' tv_es6 ' , ' tv-player-es6.vflset/tv-player-es6.js ' ) ,
( ' phone ' , ' player-plasma-ias-phone-en_US.vflset/base.js ' ) ,
( ' tablet ' , ' player-plasma-ias-tablet-en_US.vflset/base.js ' ) ,
)
@classmethod
def suitable ( cls , url ) :
if parse_qs ( url ) . get ( ' list ' , [ None ] ) [ 0 ] :
@ -1603,46 +1633,97 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
""" Return a string representation of a signature """
return ' . ' . join ( compat_str ( len ( part ) ) for part in example_sig . split ( ' . ' ) )
@classmethod
def _extract_player_info ( cls , player_url ) :
for player_re in cls . _PLAYER_INFO_RE :
id_m = re . search ( player_re , player_url )
if id_m :
break
else :
raise ExtractorError ( ' Cannot identify player %r ' % player_url )
return id_m . group ( ' id ' )
def _extract_player_info ( self , player_url ) :
try :
return self . _search_regex (
self . _PLAYER_INFO_RE , player_url , ' player info ' , group = ' id ' )
except ExtractorError as e :
raise ExtractorError (
' Cannot identify player %r ' % ( player_url , ) , cause = e )
def _ load_player( self , video_id , player_url , fatal = True , player_id = None ) :
if not player_id :
def _player_js_cache_key ( self , player_url , extra_id = None , _cache = { } ) :
if player_url not in _cache :
player_id = self . _extract_player_info ( player_url )
if player_id not in self . _code_cache :
player_path = remove_start (
compat_urllib_parse . urlparse ( player_url ) . path ,
' /s/player/ {0} / ' . format ( player_id ) )
variant = next ( ( k for k , v in self . _PLAYER_JS_VARIANT_MAP
if v == player_path ) , None )
if not variant :
variant = next (
( k for k , v in self . _PLAYER_JS_VARIANT_MAP
if re . match ( re . escape ( v ) . replace ( ' en_US ' , r ' \ w+ ' ) + ' $ ' , player_path ) ) ,
None )
if not variant :
self . write_debug (
' Unable to determine player JS variant \n '
' player = {0} ' . format ( player_url ) , only_once = True )
variant = re . sub ( r ' [^a-zA-Z0-9] ' , ' _ ' , remove_end ( player_path , ' .js ' ) )
_cache [ player_url ] = join_nonempty ( player_id , variant )
if extra_id :
extra_id = ' - ' . join ( ( _cache [ player_url ] , extra_id ) )
assert os . path . basename ( extra_id ) == extra_id
return extra_id
return _cache [ player_url ]
def _load_player ( self , video_id , player_url , fatal = True ) :
player_js_key = self . _player_js_cache_key ( player_url )
if player_js_key not in self . _code_cache :
code = self . _download_webpage (
player_url , video_id , fatal = fatal ,
note = ' Downloading player ' + player_id ,
errnote = ' Download of %s failed ' % player_url )
note = ' Downloading player {0} ' . format ( player_js_key ) ,
errnote = ' Download of {0} failed ' . format ( player_url ) )
if code :
self . _code_cache [ player_id ] = code
return self . _code_cache [ player_id ] if fatal else self . _code_cache . get ( player_id )
self . _code_cache [ player_js_key ] = code
return self . _code_cache . get ( player_js_key )
def _load_player_data_from_cache ( self , name , player_url , extra_id = None ) :
cache_id = ( ' youtube- {0} ' . format ( name ) , self . _player_js_cache_key ( player_url , extra_id ) )
data = self . _player_cache . get ( cache_id )
if data :
return data
data = self . cache . load ( * cache_id , min_ver = ' 2025.04.07 ' )
if data :
self . _player_cache [ cache_id ] = data
return data
def _store_player_data_to_cache ( self , name , player_url , data , extra_id = None ) :
cache_id = ( ' youtube- {0} ' . format ( name ) , self . _player_js_cache_key ( player_url , extra_id ) )
if cache_id not in self . _player_cache :
self . cache . store ( cache_id [ 0 ] , cache_id [ 1 ] , data )
self . _player_cache [ cache_id ] = data
def _remove_player_data_from_cache ( self , name , player_url , extra_id = None ) :
cache_id = ( ' youtube- {0} ' . format ( name ) , self . _player_js_cache_key ( player_url , extra_id ) )
if cache_id in self . _player_cache :
self . cache . clear ( * cache_id )
self . _player_cache . pop ( cache_id , None )
def _extract_signature_function ( self , video_id , player_url , example_sig ) :
player_id = self . _extract_player_info ( player_url )
# player_id = self._extract_player_info(player_url
# Read from filesystem cache
func_id = ' js_ {0} _ {1} ' . format (
player_id , self . _signature_cache_id ( example_sig ) )
assert os . path . basename ( func_id ) == func_id
self . write_debug ( ' Extracting signature function {0} ' . format ( func_id ) )
cache_spec , code = self . cache . load ( ' youtube-sigfuncs ' , func_id ) , None
extra_id = self . _signature_cache_id ( example_sig )
self . write_debug ( ' Extracting signature function {0} - {1} ' . format ( player_url , extra_id ) )
cache_spec , code = self . _load_player_data_from_cache (
' sigfuncs ' , player_url , extra_id = extra_id ) , None
if not cache_spec :
code = self . _load_player ( video_id , player_url , player_id )
if code :
res = self . _parse_sig_js ( code )
test_string = ' ' . join ( map ( compat_chr , range ( len ( example_sig ) ) ) )
cache_spec = [ ord ( c ) for c in res ( test_string ) ]
self . cache . store ( ' youtube-sigfuncs ' , func_id , cache_spec )
code = self . _load_player ( video_id , player_url )
if code :
res = self . _parse_sig_js ( code )
test_string = ' ' . join ( map ( compat_chr , range ( len ( example_sig ) ) ) )
cache_spec = [ ord ( c ) for c in res ( test_string ) ]
self . _store_player_data_to_cache (
' sigfuncs ' , player_url , cache_spec , extra_id = extra_id )
else :
self . report_warning (
' Failed to compute signature function {0} - {1} ' . format (
player_url , extra_id ) )
return lambda s : ' ' . join ( s [ i ] for i in cache_spec )
@ -1688,6 +1769,23 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
' return %s \n ' ) % ( signature_id_tuple , expr_code )
self . to_screen ( ' Extracted signature function: \n ' + code )
def _extract_sig_fn ( self , jsi , funcname ) :
var_ay = self . _search_regex (
r ''' (?x)
( ? : \* / | \{ | \n | ^ ) \s * ( ? : ' [^ ' ] + ' \ s*; \ s*)
( var \s * [ \w $ ] + \s * = \s * ( ? :
( ' | " )(?: \\ \2 |(?! \2 ).)+ \2 \ s* \ . \ s*split \ ( \ s*( ' | " ) \ W+ \3 \ s* \ )|
\[ \s * ( ? : ( ' | " )(?: \\ \4 |(?! \4 ).)* \4 \ s*(?:(?= \ ])|, \ s*))+ \ ]
) ) ( ? = \s * [ , ; ] )
''' , jsi.code, ' useful values ' , default= ' ' )
sig_fn = jsi . extract_function_code ( funcname )
if var_ay :
sig_fn = ( sig_fn [ 0 ] , ' ; \n ' . join ( ( var_ay , sig_fn [ 1 ] ) ) )
return sig_fn
def _parse_sig_js ( self , jscode ) :
# Examples where `sig` is funcname:
# sig=function(a){a=a.split(""); ... ;return a.join("")};
@ -1713,8 +1811,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
jscode , ' Initial JS player signature function name ' , group = ' sig ' )
jsi = JSInterpreter ( jscode )
initial_function = jsi . extract_function ( funcname )
return lambda s : initial_function ( [ s ] )
initial_function = self . _extract_sig_fn ( jsi , funcname )
func = jsi . extract_function_from_code ( * initial_function )
return lambda s : func ( [ s ] )
def _cached ( self , func , * cache_id ) :
def inner ( * args , * * kwargs ) :
@ -1774,6 +1876,24 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
return ret
def _extract_n_function_name ( self , jscode ) :
func_name , idx = None , None
def generic_n_function_search ( func_name = None ) :
return self . _search_regex (
r ''' (?xs)
( ? : ( ? < = [ ^ \w $ ] ) | ^ ) # instead of \b, which ignores $
( ? P < name > % s ) \s * = \s * function \( ( ? ! \d ) [ a - zA - Z \d_ $ ] + \)
\s * \{ ( ? : ( ? ! } ; ) . ) + ? ( ? :
[ " ' ]enhanced_except_ |
return \s * ( ? P < q > " | ' )[a-zA-Z \ d-]+_w8_(?P=q) \ s* \ + \ s*[ \ w$]+
)
''' % (func_name or r ' (?! \ d)[a-zA-Z \ d_$]+ ' ,), jscode,
' Initial JS player n function name ' , group = ' name ' ,
default = None if func_name else NO_DEFAULT )
# these special cases are redundant and probably obsolete (2025-04):
# they make the tests run ~10% faster without fallback warnings
r """
func_name , idx = self . _search_regex (
# (y=NuD(),Mw(k),q=k.Z[y]||null)&&(q=narray[idx](q),k.set(y,q),k.V||NuD(''))}};
# (R="nn"[+J.Z],mW(J),N=J.K[R]||null)&&(N=narray[idx](N),J.set(R,N))}};
@ -1800,41 +1920,59 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
\( \s * [ \w $ ] + \s * \)
''' , jscode, ' Initial JS player n function name ' , group=( ' nfunc ' , ' idx ' ),
default = ( None , None ) )
"""
if not func_name :
# nfunc=function(x){...}|function nfunc(x); ...
# ... var y=[nfunc]|y[idx]=nfunc);
# obvious REs hang, so use a two-stage tactic
for m in re . finditer ( r ''' (?x)
[ \n ; ] var \s ( ? : ( ? : ( ? ! , ) . ) + , | \s ) * ? ( ? ! \d ) [ \w $ ] + ( ? : \[ ( ? P < idx > \d + ) \] ) ? \s * = \s *
( ? ( idx ) | \[ \s * ) ( ? P < nfunc > ( ? ! \d ) [ \w $ ] + ) ( ? ( idx ) | \s * \] )
\s * ? [ ; \n ]
''' , jscode):
fn = self . _search_regex (
r ' [;,] \ s*(function \ s+)?( {0} )(?(1)| \ s*= \ s*function) \ s* \ ((?! \ d)[ \ w$]+ \ ) \ s* \ {1} (?! \ s*return \ s) ' . format (
re . escape ( m . group ( ' nfunc ' ) ) , ' { ' ) ,
jscode , ' Initial JS player n function name (2) ' , group = 2 , default = None )
if fn :
func_name = fn
idx = m . group ( ' idx ' )
if generic_n_function_search ( func_name ) :
# don't look any further
break
# thx bashonly: yt-dlp/yt-dlp/pull/10611
if not func_name :
self . report_warning ( ' Falling back to generic n function search ' )
return self . _search_regex (
r ''' (?xs)
( ? : ( ? < = [ ^ \w $ ] ) | ^ ) # instead of \b, which ignores $
( ? P < name > ( ? ! \d ) [ a - zA - Z \d_ $ ] + ) \s * = \s * function \( ( ? ! \d ) [ a - zA - Z \d_ $ ] + \)
\s * \{ ( ? : ( ? ! } ; ) . ) + ? ( ? :
[ " ' ]enhanced_except_ |
return \s * ( ? P < q > " | ' )[a-zA-Z \ d-]+_w8_(?P=q) \ s* \ + \ s*[ \ w$]+
)
''' , jscode, ' Initial JS player n function name ' , group= ' name ' )
self . report_warning ( ' Falling back to generic n function search ' , only_once = True )
return generic_n_function_search ( )
if not idx :
return func_name
return self . _search_json (
r ' var \ s+ {0} \ s*= ' . format ( re . escape ( func_name ) ) , jscode ,
r ' (?<![ \ w-])var \ s(?:(?:(?!,).)+,| \ s)*? {0} \ s*= ' . format ( re . escape ( func_name ) ) , jscode ,
' Initial JS player n function list ( {0} . {1} ) ' . format ( func_name , idx ) ,
func_name , contains_pattern = r ' \ [[ \ s \ S]+ \ ] ' , end_pattern = ' [,;] ' ,
func_name , contains_pattern = r ' \ [.+ \ ] ' , end_pattern = ' [,;] ' ,
transform_source = js_to_json ) [ int ( idx ) ]
def _extract_n_function_code ( self , video_id , player_url ) :
player_id = self . _extract_player_info ( player_url )
func_code = self . cache. load ( ' youtube-nsig ' , player_id )
func_code = self . _load_player_data_from_cache( ' nsig ' , player_url )
jscode = func_code or self . _load_player ( video_id , player_url )
jsi = JSInterpreter ( jscode )
if func_code :
return jsi , player_id , func_code
func_name = self . _extract_n_function_name ( jscode )
return self . _extract_n_function_code_jsi ( video_id , jsi , player_id , player_url )
func_code = jsi . extract_function_code ( func_name )
def _extract_n_function_code_jsi ( self , video_id , jsi , player_id = None , player_url = None ) :
func_name = self . _extract_n_function_name ( jsi . code )
self . cache . store ( ' youtube-nsig ' , player_id , func_code )
func_code = self . _extract_sig_fn ( jsi , func_name )
if player_url :
self . _store_player_data_to_cache ( ' nsig ' , player_url , func_code )
return jsi , player_id , func_code
def _extract_n_function_from_code ( self , jsi , func_code ) :
@ -1867,7 +2005,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
n_param = n_param [ - 1 ]
n_response = decrypt_nsig ( n_param ) ( n_param , video_id , player_url )
if n_response is None :
# give up if descrambling failed
# give up and forget cached data if descrambling failed
self . _remove_player_data_from_cache ( ' nsig ' , player_url )
break
fmt [ ' url ' ] = update_url_query ( fmt [ ' url ' ] , { ' n ' : n_response } )
@ -1878,18 +2017,28 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
Required to tell API what sig / player version is in use .
"""
sts = traverse_obj ( ytcfg , ' STS ' , expected_type = int )
if not sts :
# Attempt to extract from player
if player_url is None :
error_msg = ' Cannot extract signature timestamp without player_url. '
if fatal :
raise ExtractorError ( error_msg )
self . report_warning ( error_msg )
return
code = self . _load_player ( video_id , player_url , fatal = fatal )
sts = int_or_none ( self . _search_regex (
r ' (?:signatureTimestamp|sts) \ s*: \ s*(?P<sts>[0-9] {5} ) ' , code or ' ' ,
' JS player signature timestamp ' , group = ' sts ' , fatal = fatal ) )
if sts :
return sts
if not player_url :
error_msg = ' Cannot extract signature timestamp without player url '
if fatal :
raise ExtractorError ( error_msg )
self . report_warning ( error_msg )
return None
sts = self . _load_player_data_from_cache ( ' sts ' , player_url )
if sts :
return sts
# Attempt to extract from player
code = self . _load_player ( video_id , player_url , fatal = fatal )
sts = int_or_none ( self . _search_regex (
r ' (?:signatureTimestamp|sts) \ s*: \ s*(?P<sts>[0-9] {5} ) ' , code or ' ' ,
' JS player signature timestamp ' , group = ' sts ' , fatal = fatal ) )
if sts :
self . _store_player_data_to_cache ( ' sts ' , player_url , sts )
return sts
def _mark_watched ( self , video_id , player_response ) :
@ -2103,7 +2252,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
video_details = merge_dicts ( * traverse_obj (
( player_response , api_player_response ) ,
( Ellipsis , ' videoDetails ' , T ( dict ) ) ) )
player_response . update ( api_player_response or { } )
player_response . update ( filter_dict (
api_player_response or { } , cndn = lambda k , _ : k != ' captions ' ) )
player_response [ ' videoDetails ' ] = video_details
def is_agegated ( playability ) :
@ -2533,8 +2683,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
}
pctr = traverse_obj (
player_response ,
( ' captions ' , ' playerCaptionsTracklistRenderer ' , T ( dict ) ) )
( player_response , api_player_response ) ,
( Ellipsis , ' captions ' , ' playerCaptionsTracklistRenderer ' , T ( dict ) ) )
if pctr :
def process_language ( container , base_url , lang_code , query ) :
lang_subs = [ ]
@ -2551,20 +2701,21 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
def process_subtitles ( ) :
subtitles = { }
for caption_track in traverse_obj ( pctr , (
' captionTracks ' , lambda _ , v : v . get ( ' baseUrl ' ) ) ) :
Ellipsis , ' captionTracks ' , lambda _ , v : (
v . get ( ' baseUrl ' ) and v . get ( ' languageCode ' ) ) ) ) :
base_url = self . _yt_urljoin ( caption_track [ ' baseUrl ' ] )
if not base_url :
continue
lang_code = caption_track [ ' languageCode ' ]
if caption_track . get ( ' kind ' ) != ' asr ' :
lang_code = caption_track . get ( ' languageCode ' )
if not lang_code :
continue
process_language (
subtitles , base_url , lang_code , { } )
continue
automatic_captions = { }
process_language (
automatic_captions , base_url , lang_code , { } )
for translation_language in traverse_obj ( pctr , (
' translationLanguages ' , lambda _ , v : v . get ( ' languageCode ' ) ) ) :
Ellipsis , ' translationLanguages ' , lambda _ , v : v . get ( ' languageCode ' ) ) ) :
translation_language_code = translation_language [ ' languageCode ' ]
process_language (
automatic_captions , base_url , translation_language_code ,
@ -3183,8 +3334,12 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
expected_type = txt_or_none )
def _grid_entries ( self , grid_renderer ) :
for item in grid_renderer [ ' items ' ] :
if not isinstance ( item , dict ) :
for item in traverse_obj ( grid_renderer , ( ' items ' , Ellipsis , T ( dict ) ) ) :
lockup_view_model = traverse_obj ( item , ( ' lockupViewModel ' , T ( dict ) ) )
if lockup_view_model :
entry = self . _extract_lockup_view_model ( lockup_view_model )
if entry :
yield entry
continue
renderer = self . _extract_grid_item_renderer ( item )
if not isinstance ( renderer , dict ) :
@ -3268,6 +3423,39 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
continue
yield self . _extract_video ( renderer )
def _extract_lockup_view_model ( self , view_model ) :
content_id = view_model . get ( ' contentId ' )
if not content_id :
return
content_type = view_model . get ( ' contentType ' )
if content_type not in ( ' LOCKUP_CONTENT_TYPE_PLAYLIST ' , ' LOCKUP_CONTENT_TYPE_PODCAST ' ) :
self . report_warning (
' Unsupported lockup view model content type " {0} " {1} ' . format ( content_type , bug_reports_message ( ) ) , only_once = True )
return
return merge_dicts ( self . url_result (
update_url_query ( ' https://www.youtube.com/playlist ' , { ' list ' : content_id } ) ,
ie = YoutubeTabIE . ie_key ( ) , video_id = content_id ) , {
' title ' : traverse_obj ( view_model , (
' metadata ' , ' lockupMetadataViewModel ' , ' title ' , ' content ' , T ( compat_str ) ) ) ,
' thumbnails ' : self . _extract_thumbnails ( view_model , (
' contentImage ' , ' collectionThumbnailViewModel ' , ' primaryThumbnail ' ,
' thumbnailViewModel ' , ' image ' ) , final_key = ' sources ' ) ,
} )
def _extract_shorts_lockup_view_model ( self , view_model ) :
content_id = traverse_obj ( view_model , (
' onTap ' , ' innertubeCommand ' , ' reelWatchEndpoint ' , ' videoId ' ,
T ( lambda v : v if YoutubeIE . suitable ( v ) else None ) ) )
if not content_id :
return
return merge_dicts ( self . url_result (
content_id , ie = YoutubeIE . ie_key ( ) , video_id = content_id ) , {
' title ' : traverse_obj ( view_model , (
' overlayMetadata ' , ' primaryText ' , ' content ' , T ( compat_str ) ) ) ,
' thumbnails ' : self . _extract_thumbnails (
view_model , ' thumbnail ' , final_key = ' sources ' ) ,
} )
def _video_entry ( self , video_renderer ) :
video_id = video_renderer . get ( ' videoId ' )
if video_id :
@ -3314,10 +3502,9 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
yield entry
def _rich_grid_entries ( self , contents ) :
for content in contents :
content = traverse_obj (
content , ( ' richItemRenderer ' , ' content ' ) ,
expected_type = dict ) or { }
for content in traverse_obj (
contents , ( Ellipsis , ' richItemRenderer ' , ' content ' ) ,
expected_type = dict ) :
video_renderer = traverse_obj (
content , ' videoRenderer ' , ' reelItemRenderer ' ,
expected_type = dict )
@ -3325,6 +3512,12 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
entry = self . _video_entry ( video_renderer )
if entry :
yield entry
# shorts item
shorts_lockup_view_model = content . get ( ' shortsLockupViewModel ' )
if shorts_lockup_view_model :
entry = self . _extract_shorts_lockup_view_model ( shorts_lockup_view_model )
if entry :
yield entry
# playlist
renderer = traverse_obj (
content , ' playlistRenderer ' , expected_type = dict ) or { }
@ -3363,23 +3556,15 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
next_continuation = cls . _extract_next_continuation_data ( renderer )
if next_continuation :
return next_continuation
contents = [ ]
for key in ( ' contents ' , ' items ' ) :
contents . extend ( try_get ( renderer , lambda x : x [ key ] , list ) or [ ] )
for content in contents :
if not isinstance ( content , dict ) :
continue
continuation_ep = try_get (
content , lambda x : x [ ' continuationItemRenderer ' ] [ ' continuationEndpoint ' ] ,
dict )
if not continuation_ep :
continue
continuation = try_get (
continuation_ep , lambda x : x [ ' continuationCommand ' ] [ ' token ' ] , compat_str )
for command in traverse_obj ( renderer , (
( ' contents ' , ' items ' , ' rows ' ) , Ellipsis , ' continuationItemRenderer ' ,
( ' continuationEndpoint ' , ( ' button ' , ' buttonRenderer ' , ' command ' ) ) ,
( ( ' commandExecutorCommand ' , ' commands ' , Ellipsis ) , None ) , T ( dict ) ) ) :
continuation = traverse_obj ( command , ( ' continuationCommand ' , ' token ' , T ( compat_str ) ) )
if not continuation :
continue
ctp = co ntinuation_ep . get ( ' clickTrackingParams ' )
return YoutubeTabIE . _build_continuation_query ( continuation , ctp )
ctp = command . get ( ' clickTrackingParams ' )
return cls . _build_continuation_query ( continuation , ctp )
def _entries ( self , tab , item_id , webpage ) :
tab_content = try_get ( tab , lambda x : x [ ' content ' ] , dict )
@ -3428,6 +3613,13 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
entry = self . _video_entry ( renderer )
if entry :
yield entry
renderer = isr_content . get ( ' richGridRenderer ' )
if renderer :
for from_ in self . _rich_grid_entries (
traverse_obj ( renderer , ( ' contents ' , Ellipsis , T ( dict ) ) ) ) :
yield from_
continuation = self . _extract_continuation ( renderer )
continue
if not continuation :
continuation = self . _extract_continuation ( is_renderer )
@ -3437,8 +3629,9 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
rich_grid_renderer = tab_content . get ( ' richGridRenderer ' )
if not rich_grid_renderer :
return
for entry in self . _rich_grid_entries ( rich_grid_renderer . get ( ' contents ' ) or [ ] ) :
yield entry
for from_ in self . _rich_grid_entries (
traverse_obj ( rich_grid_renderer , ( ' contents ' , Ellipsis , T ( dict ) ) ) ) :
yield from_
continuation = self . _extract_continuation ( rich_grid_renderer )
@ -3484,8 +3677,12 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
# Downloading page may result in intermittent 5xx HTTP error
# that is usually worked around with a retry
response = self . _download_json (
' https://www.youtube.com/youtubei/v1/browse ?key=AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8 ' ,
' https://www.youtube.com/youtubei/v1/browse ' ,
None , ' Downloading page %d %s ' % ( page_num , ' (retry # %d ) ' % count if count else ' ' ) ,
query = {
# 'key': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8',
' prettyPrint ' : ' false ' ,
} ,
headers = headers , data = json . dumps ( data ) . encode ( ' utf8 ' ) )
break
except ExtractorError as e :
Connor Mason
1 month ago