[Youtube] Fix private feeds/playlists on multi-channel accounts (#143)

Authored by: colethedj
This commit is contained in:
Matthew 2021-03-05 13:59:14 +00:00 committed by GitHub
parent f3eaa8dd1c
commit d069eca7a3
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -301,7 +301,6 @@ def _call_api(self, ep, query, video_id, fatal=True, headers=None,
auth = self._generate_sapisidhash_header() auth = self._generate_sapisidhash_header()
if auth is not None: if auth is not None:
headers.update({'Authorization': auth, 'X-Origin': 'https://www.youtube.com'}) headers.update({'Authorization': auth, 'X-Origin': 'https://www.youtube.com'})
return self._download_json( return self._download_json(
'https://www.youtube.com/youtubei/v1/%s' % ep, 'https://www.youtube.com/youtubei/v1/%s' % ep,
video_id=video_id, fatal=fatal, note=note, errnote=errnote, video_id=video_id, fatal=fatal, note=note, errnote=errnote,
@ -2704,7 +2703,7 @@ def _extract_continuation(cls, renderer):
ctp = continuation_ep.get('clickTrackingParams') ctp = continuation_ep.get('clickTrackingParams')
return YoutubeTabIE._build_continuation_query(continuation, ctp) return YoutubeTabIE._build_continuation_query(continuation, ctp)
def _entries(self, tab, identity_token, item_id): def _entries(self, tab, item_id, identity_token, account_syncid):
def extract_entries(parent_renderer): # this needs to called again for continuation to work with feeds def extract_entries(parent_renderer): # this needs to called again for continuation to work with feeds
contents = try_get(parent_renderer, lambda x: x['contents'], list) or [] contents = try_get(parent_renderer, lambda x: x['contents'], list) or []
@ -2764,6 +2763,10 @@ def extract_entries(parent_renderer): # this needs to called again for continua
if identity_token: if identity_token:
headers['x-youtube-identity-token'] = identity_token headers['x-youtube-identity-token'] = identity_token
if account_syncid:
headers['X-Goog-PageId'] = account_syncid
headers['X-Goog-AuthUser'] = 0
for page_num in itertools.count(1): for page_num in itertools.count(1):
if not continuation: if not continuation:
break break
@ -2883,7 +2886,7 @@ def _extract_uploader(data):
try_get(owner, lambda x: x['navigationEndpoint']['browseEndpoint']['canonicalBaseUrl'], compat_str)) try_get(owner, lambda x: x['navigationEndpoint']['browseEndpoint']['canonicalBaseUrl'], compat_str))
return {k: v for k, v in uploader.items() if v is not None} return {k: v for k, v in uploader.items() if v is not None}
def _extract_from_tabs(self, item_id, webpage, data, tabs, identity_token): def _extract_from_tabs(self, item_id, webpage, data, tabs):
playlist_id = title = description = channel_url = channel_name = channel_id = None playlist_id = title = description = channel_url = channel_name = channel_id = None
thumbnails_list = tags = [] thumbnails_list = tags = []
@ -2947,7 +2950,10 @@ def _extract_from_tabs(self, item_id, webpage, data, tabs, identity_token):
'channel_id': metadata['uploader_id'], 'channel_id': metadata['uploader_id'],
'channel_url': metadata['uploader_url']}) 'channel_url': metadata['uploader_url']})
return self.playlist_result( return self.playlist_result(
self._entries(selected_tab, identity_token, playlist_id), self._entries(
selected_tab, playlist_id,
self._extract_identity_token(webpage, item_id),
self._extract_account_syncid(data)),
**metadata) **metadata)
def _extract_mix_playlist(self, playlist, playlist_id): def _extract_mix_playlist(self, playlist, playlist_id):
@ -3026,6 +3032,17 @@ def _extract_identity_token(self, webpage, item_id):
r'\bID_TOKEN["\']\s*:\s*["\'](.+?)["\']', webpage, r'\bID_TOKEN["\']\s*:\s*["\'](.+?)["\']', webpage,
'identity token', default=None) 'identity token', default=None)
@staticmethod
def _extract_account_syncid(data):
"""Extract syncId required to download private playlists of secondary channels"""
sync_ids = (
try_get(data, lambda x: x['responseContext']['mainAppWebResponseContext']['datasyncId'], compat_str)
or '').split("||")
if len(sync_ids) >= 2 and sync_ids[1]:
# 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
return sync_ids[0]
def _extract_webpage(self, url, item_id): def _extract_webpage(self, url, item_id):
retries = self._downloader.params.get('extractor_retries', 3) retries = self._downloader.params.get('extractor_retries', 3)
count = -1 count = -1
@ -3085,8 +3102,7 @@ def _real_extract(self, url):
tabs = try_get( tabs = try_get(
data, lambda x: x['contents']['twoColumnBrowseResultsRenderer']['tabs'], list) data, lambda x: x['contents']['twoColumnBrowseResultsRenderer']['tabs'], list)
if tabs: if tabs:
identity_token = self._extract_identity_token(webpage, item_id) return self._extract_from_tabs(item_id, webpage, data, tabs)
return self._extract_from_tabs(item_id, webpage, data, tabs, identity_token)
playlist = try_get( playlist = try_get(
data, lambda x: x['contents']['twoColumnWatchNextResults']['playlist']['playlist'], dict) data, lambda x: x['contents']['twoColumnWatchNextResults']['playlist']['playlist'], dict)