mirror of
https://github.com/yt-dlp/yt-dlp.git
synced 2026-04-13 04:09:46 +00:00
[ie/youtube] Refactor cookie SID auth
Update to reflect what browser does
This commit is contained in:
coletdjnz
parent
0b6b7742c2
commit
ed9a93f1a1
@ -627,32 +627,81 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
|
||||
client_context.update({'hl': self._preferred_lang or 'en', 'timeZone': 'UTC', 'utcOffsetMinutes': 0})
|
||||
return context
|
||||
|
||||
_SAPISID = None
|
||||
@staticmethod
|
||||
def _make_sid_authorization(scheme, sid, origin, additional_parts):
|
||||
timestamp = str(round(time.time()))
|
||||
|
||||
def _generate_sapisidhash_header(self, origin='https://www.youtube.com'):
|
||||
time_now = round(time.time())
|
||||
hash_parts = []
|
||||
if additional_parts:
|
||||
hash_parts.append(':'.join(additional_parts.values()))
|
||||
hash_parts.extend([timestamp, sid, origin])
|
||||
sidhash = hashlib.sha1(' '.join(hash_parts).encode()).hexdigest()
|
||||
|
||||
parts = [timestamp, sidhash]
|
||||
if additional_parts:
|
||||
parts.append(''.join(additional_parts.keys()))
|
||||
|
||||
return f'{scheme} {"_".join(parts)}'
|
||||
|
||||
_SAPISID = _3PSAPISID = _1PSAPISID = None
|
||||
|
||||
def _load_sid_cookies(self):
|
||||
|
||||
yt_cookies = self._get_cookies('https://www.youtube.com')
|
||||
if self._SAPISID is None:
|
||||
yt_cookies = self._get_cookies('https://www.youtube.com')
|
||||
# Sometimes SAPISID cookie isn't present but __Secure-3PAPISID is.
|
||||
# YouTube also falls back to __Secure-3PAPISID if SAPISID is missing.
|
||||
# See: https://github.com/yt-dlp/yt-dlp/issues/393
|
||||
|
||||
sapisid_cookie = dict_get(
|
||||
yt_cookies, ('__Secure-3PAPISID', 'SAPISID'))
|
||||
yt_cookies, ('SAPISID', '__Secure-3PAPISID'))
|
||||
if sapisid_cookie and sapisid_cookie.value:
|
||||
self._SAPISID = sapisid_cookie.value
|
||||
self.write_debug('Extracted SAPISID cookie')
|
||||
# SAPISID cookie is required if not already present
|
||||
if not yt_cookies.get('SAPISID'):
|
||||
self.write_debug('Copying __Secure-3PAPISID cookie to SAPISID cookie')
|
||||
self._set_cookie(
|
||||
'.youtube.com', 'SAPISID', self._SAPISID, secure=True, expire_time=time_now + 3600)
|
||||
else:
|
||||
self._SAPISID = False
|
||||
if not self._SAPISID:
|
||||
return None
|
||||
# SAPISIDHASH algorithm from https://stackoverflow.com/a/32065323
|
||||
sapisidhash = hashlib.sha1(
|
||||
f'{time_now} {self._SAPISID} {origin}'.encode()).hexdigest()
|
||||
return f'SAPISIDHASH {time_now}_{sapisidhash}'
|
||||
self.write_debug('Found SAPISID cookie')
|
||||
|
||||
if self._1PSAPISID is None:
|
||||
_1papisid_cookie = yt_cookies.get('__Secure-1PAPISID')
|
||||
if _1papisid_cookie and _1papisid_cookie.value:
|
||||
self._1PSAPISID = _1papisid_cookie.value
|
||||
self.write_debug('Found 1PAPISID cookie')
|
||||
|
||||
if self._3PSAPISID is None:
|
||||
_3papisid_cookie = yt_cookies.get('__Secure-3PAPISID')
|
||||
if _3papisid_cookie and _3papisid_cookie.value:
|
||||
self._3PSAPISID = _3papisid_cookie.value
|
||||
self.write_debug('Found 3PAPISID cookie')
|
||||
|
||||
def _generate_sid_authorization(self, origin='https://www.youtube.com', user_session_id=None):
|
||||
"""
|
||||
Generate API Session ID Authorization for Innertube requests. Assumes all requests are secure (https).
|
||||
@param origin: Origin URL
|
||||
@param user_session_id: Optional User Session ID
|
||||
@return: Authorization header value
|
||||
"""
|
||||
|
||||
authorizations = []
|
||||
additional_parts = {}
|
||||
if user_session_id:
|
||||
additional_parts['u'] = user_session_id
|
||||
|
||||
self._load_sid_cookies()
|
||||
|
||||
if self._SAPISID:
|
||||
authorizations.append(
|
||||
self._make_sid_authorization('SAPISIDHASH', self._SAPISID, origin, additional_parts))
|
||||
|
||||
if self._1PSAPISID:
|
||||
authorizations.append(
|
||||
self._make_sid_authorization('SAPISID1PHASH', self._1PSAPISID, origin, additional_parts))
|
||||
|
||||
if self._3PSAPISID:
|
||||
authorizations.append(
|
||||
self._make_sid_authorization('SAPISID3PHASH', self._3PSAPISID, origin, additional_parts))
|
||||
|
||||
if not authorizations:
|
||||
return
|
||||
|
||||
return ' '.join(authorizations)
|
||||
|
||||
def _call_api(self, ep, query, video_id, fatal=True, headers=None,
|
||||
note='Downloading API JSON', errnote='Unable to download API page',
|
||||
@ -688,26 +737,46 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
|
||||
if session_index is not None:
|
||||
return session_index
|
||||
|
||||
def _data_sync_id_to_delegated_session_id(self, data_sync_id):
|
||||
if not data_sync_id:
|
||||
return
|
||||
# datasyncid is of the form "channel_syncid||user_syncid" for secondary channel
|
||||
# and just "user_syncid||" for primary channel. We only want the channel_syncid
|
||||
channel_syncid, _, user_syncid = data_sync_id.partition('||')
|
||||
if user_syncid:
|
||||
return channel_syncid
|
||||
|
||||
def _extract_account_syncid(self, *args):
|
||||
@staticmethod
|
||||
def _parse_data_sync_id(data_sync_id):
|
||||
"""
|
||||
Extract current session ID required to download private playlists of secondary channels
|
||||
Parse data_sync_id into delegated_session_id and user_session_id.
|
||||
|
||||
data_sync_id is of the form "delegated_session_id||user_session_id" for secondary channel
|
||||
and just "user_session_id||" for primary channel.
|
||||
|
||||
@param data_sync_id: data_sync_id string
|
||||
@return: Tuple of (delegated_session_id, user_session_id)
|
||||
"""
|
||||
if not data_sync_id:
|
||||
return None, None
|
||||
first, _, second = data_sync_id.partition('||')
|
||||
return first if second else None, second if second else first
|
||||
|
||||
def _extract_delegated_session_id(self, *args):
|
||||
"""
|
||||
Extract current delegated session ID required to download private playlists of secondary channels
|
||||
@params response and/or ytcfg
|
||||
@return: delegated session ID
|
||||
"""
|
||||
# ytcfg includes channel_syncid if on secondary channel
|
||||
if delegated_sid := traverse_obj(args, (..., 'DELEGATED_SESSION_ID', {str}, any)):
|
||||
return delegated_sid
|
||||
|
||||
data_sync_id = self._extract_data_sync_id(*args)
|
||||
return self._data_sync_id_to_delegated_session_id(data_sync_id)
|
||||
return self._parse_data_sync_id(data_sync_id)[0]
|
||||
|
||||
def _extract_user_session_id(self, *args):
|
||||
"""
|
||||
Extract current user session ID
|
||||
@params response and/or ytcfg
|
||||
@return: user session ID
|
||||
"""
|
||||
if user_sid := traverse_obj(args, (..., 'USER_SESSION_ID', {str}, any)):
|
||||
return user_sid
|
||||
|
||||
data_sync_id = self._extract_data_sync_id(*args)
|
||||
return self._parse_data_sync_id(data_sync_id)[1]
|
||||
|
||||
def _extract_data_sync_id(self, *args):
|
||||
"""
|
||||
@ -734,7 +803,10 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
|
||||
|
||||
@functools.cached_property
|
||||
def is_authenticated(self):
|
||||
return bool(self._generate_sapisidhash_header())
|
||||
if bool(self._generate_sid_authorization()):
|
||||
self.write_debug('Logged in using cookies')
|
||||
return True
|
||||
return False
|
||||
|
||||
def extract_ytcfg(self, video_id, webpage):
|
||||
if not webpage:
|
||||
@ -746,7 +818,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
|
||||
|
||||
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)
|
||||
account_syncid = account_syncid or self._extract_delegated_session_id(ytcfg)
|
||||
if account_syncid:
|
||||
headers['X-Goog-PageId'] = account_syncid
|
||||
if session_index is None:
|
||||
@ -754,7 +826,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
|
||||
if account_syncid or session_index is not None:
|
||||
headers['X-Goog-AuthUser'] = session_index if session_index is not None else 0
|
||||
|
||||
auth = self._generate_sapisidhash_header(origin)
|
||||
auth = self._generate_sid_authorization(origin, user_session_id=self._extract_user_session_id(ytcfg))
|
||||
if auth is not None:
|
||||
headers['Authorization'] = auth
|
||||
headers['X-Origin'] = origin
|
||||
@ -3837,8 +3909,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
|
||||
visitor_data=visitor_data,
|
||||
session_index=self._extract_session_index(master_ytcfg, player_ytcfg),
|
||||
account_syncid=(
|
||||
self._data_sync_id_to_delegated_session_id(data_sync_id)
|
||||
or self._extract_account_syncid(master_ytcfg, initial_pr, player_ytcfg)
|
||||
self._parse_data_sync_id(data_sync_id)
|
||||
or self._extract_delegated_session_id(master_ytcfg, initial_pr, player_ytcfg)
|
||||
),
|
||||
)
|
||||
|
||||
@ -5441,7 +5513,7 @@ class YoutubeTabBaseInfoExtractor(YoutubeBaseInfoExtractor):
|
||||
return self.playlist_result(
|
||||
self._entries(
|
||||
selected_tab, metadata['id'], ytcfg,
|
||||
self._extract_account_syncid(ytcfg, data),
|
||||
self._extract_delegated_session_id(ytcfg, data),
|
||||
self._extract_visitor_data(data, ytcfg)),
|
||||
**metadata)
|
||||
|
||||
@ -5593,7 +5665,7 @@ class YoutubeTabBaseInfoExtractor(YoutubeBaseInfoExtractor):
|
||||
watch_endpoint = try_get(
|
||||
playlist, lambda x: x['contents'][-1]['playlistPanelVideoRenderer']['navigationEndpoint']['watchEndpoint'])
|
||||
headers = self.generate_api_headers(
|
||||
ytcfg=ytcfg, account_syncid=self._extract_account_syncid(ytcfg, data),
|
||||
ytcfg=ytcfg, account_syncid=self._extract_delegated_session_id(ytcfg, data),
|
||||
visitor_data=self._extract_visitor_data(response, data, ytcfg))
|
||||
query = {
|
||||
'playlistId': playlist_id,
|
||||
@ -5691,7 +5763,7 @@ class YoutubeTabBaseInfoExtractor(YoutubeBaseInfoExtractor):
|
||||
if not is_playlist:
|
||||
return
|
||||
headers = self.generate_api_headers(
|
||||
ytcfg=ytcfg, account_syncid=self._extract_account_syncid(ytcfg, data),
|
||||
ytcfg=ytcfg, account_syncid=self._extract_delegated_session_id(ytcfg, data),
|
||||
visitor_data=self._extract_visitor_data(data, ytcfg))
|
||||
query = {
|
||||
'params': 'wgYCCAA=',
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user