mirror of
https://github.com/yt-dlp/yt-dlp.git
synced 2024-11-12 18:04:18 +01:00
[ie/youtube] Support logging in with OAuth (#11001)
See: https://github.com/yt-dlp/yt-dlp/wiki/Extractors#logging-in-with-oauth Authored by: coletdjnz
This commit is contained in:
parent
46fe60ff19
commit
b8635c1d47
1 changed files with 262 additions and 45 deletions
|
@ -22,7 +22,7 @@
|
|||
from .common import InfoExtractor, SearchInfoExtractor
|
||||
from .openload import PhantomJSwrapper
|
||||
from ..jsinterp import JSInterpreter
|
||||
from ..networking.exceptions import HTTPError, network_exceptions
|
||||
from ..networking.exceptions import HTTPError, TransportError, network_exceptions
|
||||
from ..utils import (
|
||||
NO_DEFAULT,
|
||||
ExtractorError,
|
||||
|
@ -55,6 +55,7 @@
|
|||
str_or_none,
|
||||
str_to_int,
|
||||
strftime_or_none,
|
||||
time_seconds,
|
||||
traverse_obj,
|
||||
try_call,
|
||||
try_get,
|
||||
|
@ -515,6 +516,8 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
|
|||
_YT_HANDLE_RE = r'@[\w.-]{3,30}' # https://support.google.com/youtube/answer/11585688?hl=en
|
||||
_YT_CHANNEL_UCID_RE = r'UC[\w-]{22}'
|
||||
|
||||
_NETRC_MACHINE = 'youtube'
|
||||
|
||||
def ucid_or_none(self, ucid):
|
||||
return self._search_regex(rf'^({self._YT_CHANNEL_UCID_RE})$', ucid, 'UC-id', default=None)
|
||||
|
||||
|
@ -573,9 +576,213 @@ def _real_initialize(self):
|
|||
self._initialize_consent()
|
||||
self._check_login_required()
|
||||
|
||||
def _perform_login(self, username, password):
|
||||
auth_type, _, user = (username or '').partition('+')
|
||||
|
||||
if auth_type != 'oauth':
|
||||
raise ExtractorError(self._youtube_login_hint, expected=True)
|
||||
|
||||
self._initialize_oauth(user, password)
|
||||
|
||||
'''
|
||||
OAuth 2.0 Device Authorization Grant flow, used by the YouTube TV client (youtube.com/tv).
|
||||
|
||||
For more information regarding OAuth 2.0 and the Device Authorization Grant flow in general, see:
|
||||
- https://developers.google.com/identity/protocols/oauth2/limited-input-device
|
||||
- https://accounts.google.com/.well-known/openid-configuration
|
||||
- https://www.rfc-editor.org/rfc/rfc8628
|
||||
- https://www.rfc-editor.org/rfc/rfc6749
|
||||
|
||||
Note: The official client appears to use a proxied version of the oauth2 endpoints on youtube.com/o/oauth2,
|
||||
which applies some modifications to the response (such as returning errors as 200 OK).
|
||||
Since the client works with the standard API, we will use that as it is well-documented.
|
||||
'''
|
||||
|
||||
_OAUTH_PROFILE = None
|
||||
_OAUTH_ACCESS_TOKEN_CACHE = {}
|
||||
_OAUTH_DISPLAY_ID = 'oauth'
|
||||
|
||||
# YouTube TV (TVHTML5) client. You can find these at youtube.com/tv
|
||||
_OAUTH_CLIENT_ID = '861556708454-d6dlm3lh05idd8npek18k6be8ba3oc68.apps.googleusercontent.com'
|
||||
_OAUTH_CLIENT_SECRET = 'SboVhoG9s0rNafixCSGGKXAT'
|
||||
_OAUTH_SCOPE = 'http://gdata.youtube.com https://www.googleapis.com/auth/youtube-paid-content'
|
||||
|
||||
# From https://accounts.google.com/.well-known/openid-configuration
|
||||
# Technically, these should be fetched dynamically and not hard-coded.
|
||||
# However, as these endpoints rarely change, we can risk saving an extra request for every invocation.
|
||||
_OAUTH_DEVICE_AUTHORIZATION_ENDPOINT = 'https://oauth2.googleapis.com/device/code'
|
||||
_OAUTH_TOKEN_ENDPOINT = 'https://oauth2.googleapis.com/token'
|
||||
|
||||
@property
|
||||
def _oauth_cache_key(self):
|
||||
return f'oauth_refresh_token_{self._OAUTH_PROFILE}'
|
||||
|
||||
def _read_oauth_error_response(self, response):
|
||||
return traverse_obj(
|
||||
self._webpage_read_content(response, self._OAUTH_TOKEN_ENDPOINT, self._OAUTH_DISPLAY_ID, fatal=False),
|
||||
({json.loads}, 'error', {str}))
|
||||
|
||||
def _set_oauth_info(self, token_response):
|
||||
YoutubeBaseInfoExtractor._OAUTH_ACCESS_TOKEN_CACHE.setdefault(self._OAUTH_PROFILE, {}).update({
|
||||
'access_token': token_response['access_token'],
|
||||
'token_type': token_response['token_type'],
|
||||
'expiry': time_seconds(
|
||||
seconds=traverse_obj(token_response, ('expires_in', {float_or_none}), default=300) - 10),
|
||||
})
|
||||
refresh_token = traverse_obj(token_response, ('refresh_token', {str}))
|
||||
if refresh_token:
|
||||
self.cache.store(self._NETRC_MACHINE, self._oauth_cache_key, refresh_token)
|
||||
YoutubeBaseInfoExtractor._OAUTH_ACCESS_TOKEN_CACHE[self._OAUTH_PROFILE]['refresh_token'] = refresh_token
|
||||
|
||||
def _initialize_oauth(self, user, refresh_token):
|
||||
self._OAUTH_PROFILE = user or 'default'
|
||||
|
||||
if self._OAUTH_PROFILE in YoutubeBaseInfoExtractor._OAUTH_ACCESS_TOKEN_CACHE:
|
||||
self.write_debug(f'{self._OAUTH_DISPLAY_ID}: Using cached access token for profile "{self._OAUTH_PROFILE}"')
|
||||
return
|
||||
|
||||
YoutubeBaseInfoExtractor._OAUTH_ACCESS_TOKEN_CACHE[self._OAUTH_PROFILE] = {}
|
||||
|
||||
if refresh_token:
|
||||
refresh_token = refresh_token.strip('\'') or None
|
||||
|
||||
# Allow refresh token passed to initialize cache
|
||||
if refresh_token:
|
||||
self.cache.store(self._NETRC_MACHINE, self._oauth_cache_key, refresh_token)
|
||||
|
||||
refresh_token = refresh_token or self.cache.load(self._NETRC_MACHINE, self._oauth_cache_key)
|
||||
if refresh_token:
|
||||
YoutubeBaseInfoExtractor._OAUTH_ACCESS_TOKEN_CACHE[self._OAUTH_PROFILE]['refresh_token'] = refresh_token
|
||||
try:
|
||||
token_response = self._refresh_token(refresh_token)
|
||||
except ExtractorError as e:
|
||||
error_msg = str(e.orig_msg).replace('Failed to refresh access token: ', '')
|
||||
self.report_warning(f'{self._OAUTH_DISPLAY_ID}: Failed to refresh access token: {error_msg}')
|
||||
token_response = self._oauth_authorize
|
||||
else:
|
||||
token_response = self._oauth_authorize
|
||||
|
||||
self._set_oauth_info(token_response)
|
||||
self.write_debug(f'{self._OAUTH_DISPLAY_ID}: Logged in using profile "{self._OAUTH_PROFILE}"')
|
||||
|
||||
def _refresh_token(self, refresh_token):
|
||||
try:
|
||||
token_response = self._download_json(
|
||||
self._OAUTH_TOKEN_ENDPOINT,
|
||||
video_id=self._OAUTH_DISPLAY_ID,
|
||||
note='Refreshing access token',
|
||||
data=json.dumps({
|
||||
'client_id': self._OAUTH_CLIENT_ID,
|
||||
'client_secret': self._OAUTH_CLIENT_SECRET,
|
||||
'refresh_token': refresh_token,
|
||||
'grant_type': 'refresh_token',
|
||||
}).encode(),
|
||||
headers={'Content-Type': 'application/json'})
|
||||
except ExtractorError as e:
|
||||
if isinstance(e.cause, HTTPError):
|
||||
error = self._read_oauth_error_response(e.cause.response)
|
||||
if error == 'invalid_grant':
|
||||
# RFC6749 § 5.2
|
||||
raise ExtractorError(
|
||||
'Failed to refresh access token: Refresh token is invalid, revoked, or expired (invalid_grant)',
|
||||
expected=True, video_id=self._OAUTH_DISPLAY_ID)
|
||||
raise ExtractorError(
|
||||
f'Failed to refresh access token: Authorization server returned error {error}',
|
||||
video_id=self._OAUTH_DISPLAY_ID)
|
||||
raise
|
||||
return token_response
|
||||
|
||||
@property
|
||||
def _oauth_authorize(self):
|
||||
code_response = self._download_json(
|
||||
self._OAUTH_DEVICE_AUTHORIZATION_ENDPOINT,
|
||||
video_id=self._OAUTH_DISPLAY_ID,
|
||||
note='Initializing authorization flow',
|
||||
data=json.dumps({
|
||||
'client_id': self._OAUTH_CLIENT_ID,
|
||||
'scope': self._OAUTH_SCOPE,
|
||||
}).encode(),
|
||||
headers={'Content-Type': 'application/json'})
|
||||
|
||||
verification_url = traverse_obj(code_response, ('verification_url', {str}))
|
||||
user_code = traverse_obj(code_response, ('user_code', {str}))
|
||||
if not verification_url or not user_code:
|
||||
raise ExtractorError(
|
||||
'Authorization server did not provide verification_url or user_code', video_id=self._OAUTH_DISPLAY_ID)
|
||||
|
||||
# note: The whitespace is intentional
|
||||
self.to_screen(
|
||||
f'{self._OAUTH_DISPLAY_ID}: To give yt-dlp access to your account, '
|
||||
f'go to {verification_url} and enter code {user_code}')
|
||||
|
||||
# RFC8628 § 3.5: default poll interval is 5 seconds if not provided
|
||||
poll_interval = traverse_obj(code_response, ('interval', {int}), default=5)
|
||||
|
||||
for retry in self.RetryManager():
|
||||
while True:
|
||||
try:
|
||||
token_response = self._download_json(
|
||||
self._OAUTH_TOKEN_ENDPOINT,
|
||||
video_id=self._OAUTH_DISPLAY_ID,
|
||||
note=False,
|
||||
errnote='Failed to request access token',
|
||||
data=json.dumps({
|
||||
'client_id': self._OAUTH_CLIENT_ID,
|
||||
'client_secret': self._OAUTH_CLIENT_SECRET,
|
||||
'device_code': code_response['device_code'],
|
||||
'grant_type': 'urn:ietf:params:oauth:grant-type:device_code',
|
||||
}).encode(),
|
||||
headers={'Content-Type': 'application/json'})
|
||||
except ExtractorError as e:
|
||||
if isinstance(e.cause, TransportError):
|
||||
retry.error = e
|
||||
break
|
||||
elif isinstance(e.cause, HTTPError):
|
||||
error = self._read_oauth_error_response(e.cause.response)
|
||||
if not error:
|
||||
retry.error = e
|
||||
break
|
||||
|
||||
if error == 'authorization_pending':
|
||||
time.sleep(poll_interval)
|
||||
continue
|
||||
elif error == 'expired_token':
|
||||
raise ExtractorError(
|
||||
'Authorization timed out', expected=True, video_id=self._OAUTH_DISPLAY_ID)
|
||||
elif error == 'access_denied':
|
||||
raise ExtractorError(
|
||||
'You denied access to an account', expected=True, video_id=self._OAUTH_DISPLAY_ID)
|
||||
elif error == 'slow_down':
|
||||
# RFC8628 § 3.5: add 5 seconds to the poll interval
|
||||
poll_interval += 5
|
||||
time.sleep(poll_interval)
|
||||
continue
|
||||
else:
|
||||
raise ExtractorError(
|
||||
f'Authorization server returned an error when fetching access token: {error}',
|
||||
video_id=self._OAUTH_DISPLAY_ID)
|
||||
raise
|
||||
|
||||
return token_response
|
||||
|
||||
def _update_oauth(self):
|
||||
token = YoutubeBaseInfoExtractor._OAUTH_ACCESS_TOKEN_CACHE.get(self._OAUTH_PROFILE)
|
||||
if token is None or token['expiry'] > time.time():
|
||||
return
|
||||
|
||||
self._set_oauth_info(self._refresh_token(token['refresh_token']))
|
||||
|
||||
@property
|
||||
def _youtube_login_hint(self):
|
||||
return ('Use --username=oauth[+PROFILE] --password="" to log in using oauth, '
|
||||
f'or else u{self._login_hint(method="cookies")[1:]}. '
|
||||
'See https://github.com/yt-dlp/yt-dlp/wiki/Extractors#logging-in-with-oauth for more on how to use oauth. '
|
||||
'See https://github.com/yt-dlp/yt-dlp/wiki/Extractors#exporting-youtube-cookies for help with cookies')
|
||||
|
||||
def _check_login_required(self):
|
||||
if self._LOGIN_REQUIRED and not self._cookies_passed:
|
||||
self.raise_login_required('Login details are needed to download this content', method='cookies')
|
||||
if self._LOGIN_REQUIRED and not self.is_authenticated:
|
||||
self.raise_login_required(
|
||||
f'Login details are needed to download this content. {self._youtube_login_hint}', method=None)
|
||||
|
||||
_YT_INITIAL_DATA_RE = r'(?:window\s*\[\s*["\']ytInitialData["\']\s*\]|ytInitialData)\s*='
|
||||
_YT_INITIAL_PLAYER_RESPONSE_RE = r'ytInitialPlayerResponse\s*='
|
||||
|
@ -674,17 +881,6 @@ def _extract_session_index(*data):
|
|||
if session_index is not None:
|
||||
return session_index
|
||||
|
||||
# Deprecated?
|
||||
def _extract_identity_token(self, ytcfg=None, webpage=None):
|
||||
if ytcfg:
|
||||
token = try_get(ytcfg, lambda x: x['ID_TOKEN'], str)
|
||||
if token:
|
||||
return token
|
||||
if webpage:
|
||||
return self._search_regex(
|
||||
r'\bID_TOKEN["\']\s*:\s*["\'](.+?)["\']', webpage,
|
||||
'identity token', default=None, fatal=False)
|
||||
|
||||
def _data_sync_id_to_delegated_session_id(self, data_sync_id):
|
||||
if not data_sync_id:
|
||||
return
|
||||
|
@ -731,7 +927,7 @@ def _extract_visitor_data(self, *args):
|
|||
|
||||
@functools.cached_property
|
||||
def is_authenticated(self):
|
||||
return bool(self._generate_sapisidhash_header())
|
||||
return self._OAUTH_PROFILE or bool(self._generate_sapisidhash_header())
|
||||
|
||||
def extract_ytcfg(self, video_id, webpage):
|
||||
if not webpage:
|
||||
|
@ -741,21 +937,21 @@ def extract_ytcfg(self, video_id, webpage):
|
|||
r'ytcfg\.set\s*\(\s*({.+?})\s*\)\s*;', webpage, 'ytcfg',
|
||||
default='{}'), video_id, fatal=False) or {}
|
||||
|
||||
def generate_api_headers(
|
||||
self, *, ytcfg=None, account_syncid=None, session_index=None,
|
||||
visitor_data=None, identity_token=None, api_hostname=None, default_client='web'):
|
||||
def _generate_oauth_headers(self):
|
||||
self._update_oauth()
|
||||
oauth_token = YoutubeBaseInfoExtractor._OAUTH_ACCESS_TOKEN_CACHE.get(self._OAUTH_PROFILE)
|
||||
if not oauth_token:
|
||||
return {}
|
||||
|
||||
origin = 'https://' + (self._select_api_hostname(api_hostname, default_client))
|
||||
headers = {
|
||||
'X-YouTube-Client-Name': str(
|
||||
self._ytcfg_get_safe(ytcfg, lambda x: x['INNERTUBE_CONTEXT_CLIENT_NAME'], default_client=default_client)),
|
||||
'X-YouTube-Client-Version': self._extract_client_version(ytcfg, default_client),
|
||||
'Origin': origin,
|
||||
'X-Youtube-Identity-Token': identity_token or self._extract_identity_token(ytcfg),
|
||||
'X-Goog-PageId': account_syncid or self._extract_account_syncid(ytcfg),
|
||||
'X-Goog-Visitor-Id': visitor_data or self._extract_visitor_data(ytcfg),
|
||||
'User-Agent': self._ytcfg_get_safe(ytcfg, lambda x: x['INNERTUBE_CONTEXT']['client']['userAgent'], default_client=default_client),
|
||||
return {
|
||||
'Authorization': f'{oauth_token["token_type"]} {oauth_token["access_token"]}',
|
||||
}
|
||||
|
||||
def _generate_cookie_auth_headers(self, *, ytcfg=None, account_syncid=None, session_index=None, origin=None, **kwargs):
|
||||
headers = {}
|
||||
account_syncid = account_syncid or self._extract_account_syncid(ytcfg)
|
||||
if account_syncid:
|
||||
headers['X-Goog-PageId'] = account_syncid
|
||||
if session_index is None:
|
||||
session_index = self._extract_session_index(ytcfg)
|
||||
if account_syncid or session_index is not None:
|
||||
|
@ -765,8 +961,29 @@ def generate_api_headers(
|
|||
if auth is not None:
|
||||
headers['Authorization'] = auth
|
||||
headers['X-Origin'] = origin
|
||||
|
||||
return headers
|
||||
|
||||
def generate_api_headers(
|
||||
self, *, ytcfg=None, account_syncid=None, session_index=None,
|
||||
visitor_data=None, api_hostname=None, default_client='web', **kwargs):
|
||||
|
||||
origin = 'https://' + (self._select_api_hostname(api_hostname, default_client))
|
||||
headers = {
|
||||
'X-YouTube-Client-Name': str(
|
||||
self._ytcfg_get_safe(ytcfg, lambda x: x['INNERTUBE_CONTEXT_CLIENT_NAME'], default_client=default_client)),
|
||||
'X-YouTube-Client-Version': self._extract_client_version(ytcfg, default_client),
|
||||
'Origin': origin,
|
||||
'X-Goog-Visitor-Id': visitor_data or self._extract_visitor_data(ytcfg),
|
||||
'User-Agent': self._ytcfg_get_safe(ytcfg, lambda x: x['INNERTUBE_CONTEXT']['client']['userAgent'], default_client=default_client),
|
||||
**self._generate_oauth_headers(),
|
||||
**self._generate_cookie_auth_headers(ytcfg=ytcfg, account_syncid=account_syncid, session_index=session_index, origin=origin),
|
||||
}
|
||||
return filter_dict(headers)
|
||||
|
||||
def _generate_webpage_headers(self):
|
||||
return self._generate_oauth_headers()
|
||||
|
||||
def _download_ytcfg(self, client, video_id):
|
||||
url = {
|
||||
'web': 'https://www.youtube.com',
|
||||
|
@ -776,7 +993,8 @@ def _download_ytcfg(self, client, video_id):
|
|||
if not url:
|
||||
return {}
|
||||
webpage = self._download_webpage(
|
||||
url, video_id, fatal=False, note=f'Downloading {client.replace("_", " ").strip()} client config')
|
||||
url, video_id, fatal=False, note=f'Downloading {client.replace("_", " ").strip()} client config',
|
||||
headers=self._generate_webpage_headers())
|
||||
return self.extract_ytcfg(video_id, webpage) or {}
|
||||
|
||||
@staticmethod
|
||||
|
@ -3041,7 +3259,8 @@ def _load_player(self, video_id, player_url, fatal=True):
|
|||
code = self._download_webpage(
|
||||
player_url, video_id, fatal=fatal,
|
||||
note='Downloading player ' + player_id,
|
||||
errnote=f'Download of {player_url} failed')
|
||||
errnote=f'Download of {player_url} failed',
|
||||
headers=self._generate_webpage_headers())
|
||||
if code:
|
||||
self._code_cache[player_id] = code
|
||||
return self._code_cache.get(player_id)
|
||||
|
@ -3324,7 +3543,8 @@ def _mark_watched(self, video_id, player_responses):
|
|||
|
||||
self._download_webpage(
|
||||
url, video_id, f'Marking {label}watched',
|
||||
'Unable to mark watched', fatal=False)
|
||||
'Unable to mark watched', fatal=False,
|
||||
headers=self._generate_webpage_headers())
|
||||
|
||||
@classmethod
|
||||
def _extract_from_webpage(cls, url, webpage):
|
||||
|
@ -4305,7 +4525,7 @@ def _download_player_responses(self, url, smuggled_data, video_id, webpage_url):
|
|||
if pp:
|
||||
query['pp'] = pp
|
||||
webpage = self._download_webpage(
|
||||
webpage_url, video_id, fatal=False, query=query)
|
||||
webpage_url, video_id, fatal=False, query=query, headers=self._generate_webpage_headers())
|
||||
|
||||
master_ytcfg = self.extract_ytcfg(video_id, webpage) or self._get_default_ytcfg()
|
||||
|
||||
|
@ -5593,7 +5813,7 @@ def _extract_webpage(self, url, item_id, fatal=True):
|
|||
webpage, data = None, None
|
||||
for retry in self.RetryManager(fatal=fatal):
|
||||
try:
|
||||
webpage = self._download_webpage(url, item_id, note='Downloading webpage')
|
||||
webpage = self._download_webpage(url, item_id, note='Downloading webpage', headers=self._generate_webpage_headers())
|
||||
data = self.extract_yt_initial_data(item_id, webpage or '', fatal=fatal) or {}
|
||||
except ExtractorError as e:
|
||||
if isinstance(e.cause, network_exceptions):
|
||||
|
@ -6967,7 +7187,7 @@ def _real_extract(self, url, smuggled_data):
|
|||
raise ExtractorError('Unable to recognize tab page')
|
||||
|
||||
|
||||
class YoutubePlaylistIE(InfoExtractor):
|
||||
class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
|
||||
IE_DESC = 'YouTube playlists'
|
||||
_VALID_URL = r'''(?x)(?:
|
||||
(?:https?://)?
|
||||
|
@ -7081,7 +7301,7 @@ def _real_extract(self, url):
|
|||
return self.url_result(url, ie=YoutubeTabIE.ie_key(), video_id=playlist_id)
|
||||
|
||||
|
||||
class YoutubeYtBeIE(InfoExtractor):
|
||||
class YoutubeYtBeIE(YoutubeBaseInfoExtractor):
|
||||
IE_DESC = 'youtu.be'
|
||||
_VALID_URL = rf'https?://youtu\.be/(?P<id>[0-9A-Za-z_-]{{11}})/*?.*?\blist=(?P<playlist_id>{YoutubeBaseInfoExtractor._PLAYLIST_ID_RE})'
|
||||
_TESTS = [{
|
||||
|
@ -7132,7 +7352,7 @@ def _real_extract(self, url):
|
|||
}), ie=YoutubeTabIE.ie_key(), video_id=playlist_id)
|
||||
|
||||
|
||||
class YoutubeLivestreamEmbedIE(InfoExtractor):
|
||||
class YoutubeLivestreamEmbedIE(YoutubeBaseInfoExtractor):
|
||||
IE_DESC = 'YouTube livestream embeds'
|
||||
_VALID_URL = r'https?://(?:\w+\.)?youtube\.com/embed/live_stream/?\?(?:[^#]+&)?channel=(?P<id>[^&#]+)'
|
||||
_TESTS = [{
|
||||
|
@ -7147,7 +7367,7 @@ def _real_extract(self, url):
|
|||
ie=YoutubeTabIE.ie_key(), video_id=channel_id)
|
||||
|
||||
|
||||
class YoutubeYtUserIE(InfoExtractor):
|
||||
class YoutubeYtUserIE(YoutubeBaseInfoExtractor):
|
||||
IE_DESC = 'YouTube user videos; "ytuser:" prefix'
|
||||
IE_NAME = 'youtube:user'
|
||||
_VALID_URL = r'ytuser:(?P<id>.+)'
|
||||
|
@ -7434,7 +7654,7 @@ def _real_extract(self, url):
|
|||
return self.playlist_result(self._search_results(query, params, default_client='web_music'), title, title)
|
||||
|
||||
|
||||
class YoutubeFeedsInfoExtractor(InfoExtractor):
|
||||
class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
|
||||
"""
|
||||
Base class for feed extractors
|
||||
Subclasses must re-define the _FEED_NAME property.
|
||||
|
@ -7442,9 +7662,6 @@ class YoutubeFeedsInfoExtractor(InfoExtractor):
|
|||
_LOGIN_REQUIRED = True
|
||||
_FEED_NAME = 'feeds'
|
||||
|
||||
def _real_initialize(self):
|
||||
YoutubeBaseInfoExtractor._check_login_required(self)
|
||||
|
||||
@classproperty
|
||||
def IE_NAME(cls):
|
||||
return f'youtube:{cls._FEED_NAME}'
|
||||
|
@ -7454,7 +7671,7 @@ def _real_extract(self, url):
|
|||
f'https://www.youtube.com/feed/{self._FEED_NAME}', ie=YoutubeTabIE.ie_key())
|
||||
|
||||
|
||||
class YoutubeWatchLaterIE(InfoExtractor):
|
||||
class YoutubeWatchLaterIE(YoutubeBaseInfoExtractor):
|
||||
IE_NAME = 'youtube:watchlater'
|
||||
IE_DESC = 'Youtube watch later list; ":ytwatchlater" keyword (requires cookies)'
|
||||
_VALID_URL = r':ytwatchlater'
|
||||
|
@ -7508,7 +7725,7 @@ class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
|
|||
}]
|
||||
|
||||
|
||||
class YoutubeShortsAudioPivotIE(InfoExtractor):
|
||||
class YoutubeShortsAudioPivotIE(YoutubeBaseInfoExtractor):
|
||||
IE_DESC = 'YouTube Shorts audio pivot (Shorts using audio of a given video)'
|
||||
IE_NAME = 'youtube:shorts:pivot:audio'
|
||||
_VALID_URL = r'https?://(?:www\.)?youtube\.com/source/(?P<id>[\w-]{11})/shorts'
|
||||
|
@ -7532,7 +7749,7 @@ def _real_extract(self, url):
|
|||
ie=YoutubeTabIE)
|
||||
|
||||
|
||||
class YoutubeTruncatedURLIE(InfoExtractor):
|
||||
class YoutubeTruncatedURLIE(YoutubeBaseInfoExtractor):
|
||||
IE_NAME = 'youtube:truncated_url'
|
||||
IE_DESC = False # Do not list
|
||||
_VALID_URL = r'''(?x)
|
||||
|
@ -7691,7 +7908,7 @@ def _real_extract(self, url):
|
|||
return self.url_result(redirect_url)
|
||||
|
||||
|
||||
class YoutubeTruncatedIDIE(InfoExtractor):
|
||||
class YoutubeTruncatedIDIE(YoutubeBaseInfoExtractor):
|
||||
IE_NAME = 'youtube:truncated_id'
|
||||
IE_DESC = False # Do not list
|
||||
_VALID_URL = r'https?://(?:www\.)?youtube\.com/watch\?v=(?P<id>[0-9A-Za-z_-]{1,10})$'
|
||||
|
|
Loading…
Reference in a new issue