}
},
'INNERTUBE_CONTEXT_CLIENT_NAME': 2
- }
+ },
+ # This client can access age restricted videos (unless the uploader has disabled the 'allow embedding' option)
+ # See: https://github.com/zerodytrash/YouTube-Internal-Clients
+ 'tv_embedded': {
+ 'INNERTUBE_API_KEY': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8',
+ 'INNERTUBE_CONTEXT': {
+ 'client': {
+ 'clientName': 'TVHTML5_SIMPLY_EMBEDDED_PLAYER',
+ 'clientVersion': '2.0',
+ },
+ },
+ 'INNERTUBE_CONTEXT_CLIENT_NAME': 85
+ },
}
+def _split_innertube_client(client_name):
+ variant, *base = client_name.rsplit('.', 1)
+ if base:
+ return variant, base[0], variant
+ base, *variant = client_name.split('_', 1)
+ return client_name, base, variant[0] if variant else None
+
+
def build_innertube_clients():
THIRD_PARTY = {
- 'embedUrl': 'https://google.com', # Can be any valid URL
+ 'embedUrl': 'https://www.youtube.com/', # Can be any valid URL
}
- BASE_CLIENTS = ('android', 'web', 'ios', 'mweb')
+ BASE_CLIENTS = ('android', 'web', 'tv', 'ios', 'mweb')
priority = qualities(BASE_CLIENTS[::-1])
for client, ytcfg in tuple(INNERTUBE_CLIENTS.items()):
ytcfg.setdefault('REQUIRE_JS_PLAYER', True)
ytcfg['INNERTUBE_CONTEXT']['client'].setdefault('hl', 'en')
- base_client, *variant = client.split('_')
+ _, base_client, variant = _split_innertube_client(client)
ytcfg['priority'] = 10 * priority(base_client)
if not variant:
- INNERTUBE_CLIENTS[f'{client}_agegate'] = agegate_ytcfg = copy.deepcopy(ytcfg)
- agegate_ytcfg['INNERTUBE_CONTEXT']['client']['clientScreen'] = 'EMBED'
- agegate_ytcfg['INNERTUBE_CONTEXT']['thirdParty'] = THIRD_PARTY
- agegate_ytcfg['priority'] -= 1
- elif variant == ['embedded']:
+ INNERTUBE_CLIENTS[f'{client}_embedscreen'] = embedscreen = copy.deepcopy(ytcfg)
+ embedscreen['INNERTUBE_CONTEXT']['client']['clientScreen'] = 'EMBED'
+ embedscreen['INNERTUBE_CONTEXT']['thirdParty'] = THIRD_PARTY
+ embedscreen['priority'] -= 3
+ elif variant == 'embedded':
ytcfg['INNERTUBE_CONTEXT']['thirdParty'] = THIRD_PARTY
ytcfg['priority'] -= 2
else:
def _real_initialize(self):
self._initialize_pref()
self._initialize_consent()
+ self._check_login_required()
+
+ def _check_login_required(self):
if (self._LOGIN_REQUIRED
and self.get_param('cookiefile') is None
and self.get_param('cookiesfrombrowser') is None):
headers['X-Origin'] = origin
return {h: v for h, v in headers.items() if v is not None}
+ def _download_ytcfg(self, client, video_id):
+ url = {
+ 'web': 'https://www.youtube.com',
+ 'web_music': 'https://music.youtube.com',
+ 'web_embedded': f'https://www.youtube.com/embed/{video_id}?html5=1'
+ }.get(client)
+ if not url:
+ return {}
+ webpage = self._download_webpage(
+ url, video_id, fatal=False, note=f'Downloading {client.replace("_", " ").strip()} client config')
+ return self.extract_ytcfg(video_id, webpage) or {}
+
@staticmethod
def _build_api_continuation_query(continuation, ctp=None):
query = {
return None
def _extract_time_text(self, renderer, *path_list):
+ """@returns (timestamp, time_text)"""
text = self._get_text(renderer, *path_list) or ''
dt = self.extract_relative_time(text)
timestamp = None
description = self._get_text(renderer, 'descriptionSnippet')
duration = parse_duration(self._get_text(
renderer, 'lengthText', ('thumbnailOverlays', ..., 'thumbnailOverlayTimeStatusRenderer', 'text')))
+ if duration is None:
+ duration = parse_duration(self._search_regex(
+ r'(?i)(ago)(?!.*\1)\s+(?P<duration>[a-z0-9 ,]+?)(?:\s+[\d,]+\s+views)?(?:\s+-\s+play\s+short)?$',
+ traverse_obj(renderer, ('title', 'accessibility', 'accessibilityData', 'label'), default='', expected_type=str),
+ video_id, default=None, group='duration'))
+
view_count = self._get_count(renderer, 'viewCountText')
uploader = self._get_text(renderer, 'ownerText', 'shortBylineText')
channel_id = traverse_obj(
- renderer, ('shortBylineText', 'runs', ..., 'navigationEndpoint', 'browseEndpoint', 'browseId'), expected_type=str, get_all=False)
+ renderer, ('shortBylineText', 'runs', ..., 'navigationEndpoint', 'browseEndpoint', 'browseId'),
+ expected_type=str, get_all=False)
timestamp, time_text = self._extract_time_text(renderer, 'publishedTimeText')
scheduled_timestamp = str_to_int(traverse_obj(renderer, ('upcomingEventData', 'startTime'), get_all=False))
overlay_style = traverse_obj(
- renderer, ('thumbnailOverlays', ..., 'thumbnailOverlayTimeStatusRenderer', 'style'), get_all=False, expected_type=str)
+ renderer, ('thumbnailOverlays', ..., 'thumbnailOverlayTimeStatusRenderer', 'style'),
+ get_all=False, expected_type=str)
badges = self._extract_badges(renderer)
thumbnails = self._extract_thumbnails(renderer, 'thumbnail')
navigation_url = urljoin('https://www.youtube.com/', traverse_obj(
- renderer, ('navigationEndpoint', 'commandMetadata', 'webCommandMetadata', 'url'), expected_type=str))
+ renderer, ('navigationEndpoint', 'commandMetadata', 'webCommandMetadata', 'url'),
+ expected_type=str)) or ''
url = f'https://www.youtube.com/watch?v={video_id}'
- if overlay_style == 'SHORTS' or (navigation_url and '/shorts/' in navigation_url):
+ if overlay_style == 'SHORTS' or '/shorts/' in navigation_url:
url = f'https://www.youtube.com/shorts/{video_id}'
return {
'uploader': uploader,
'channel_id': channel_id,
'thumbnails': thumbnails,
- 'upload_date': strftime_or_none(timestamp, '%Y%m%d') if self._configuration_arg('approximate_date', ie_key='youtubetab') else None,
+ 'upload_date': (strftime_or_none(timestamp, '%Y%m%d')
+ if self._configuration_arg('approximate_date', ie_key='youtubetab')
+ else None),
'live_status': ('is_upcoming' if scheduled_timestamp is not None
else 'was_live' if 'streamed' in time_text.lower()
else 'is_live' if overlay_style is not None and overlay_style == 'LIVE' or 'live now' in badges
return orderedSet(requested_clients)
- def _extract_player_ytcfg(self, client, video_id):
- url = {
- 'web_music': 'https://music.youtube.com',
- 'web_embedded': f'https://www.youtube.com/embed/{video_id}?html5=1'
- }.get(client)
- if not url:
- return {}
- webpage = self._download_webpage(url, video_id, fatal=False, note='Downloading %s config' % client.replace('_', ' ').strip())
- return self.extract_ytcfg(video_id, webpage) or {}
-
def _extract_player_responses(self, clients, video_id, webpage, master_ytcfg):
initial_pr = None
if webpage:
prs = []
def append_client(*client_names):
- """ Append the first client name that exists """
+ """ Append the first client name that exists but not already used """
for client_name in client_names:
- if client_name in INNERTUBE_CLIENTS:
- if client_name not in all_clients:
+ actual_client = _split_innertube_client(client_name)[0]
+ if actual_client in INNERTUBE_CLIENTS:
+ if actual_client not in all_clients:
clients.append(client_name)
- all_clients.add(client_name)
- return
+ all_clients.add(actual_client)
+ return
# Android player_response does not have microFormats which are needed for
# extraction of some data. So we return the initial_pr with formats
tried_iframe_fallback = False
player_url = None
while clients:
- client = clients.pop()
+ client, base_client, variant = _split_innertube_client(clients.pop())
player_ytcfg = master_ytcfg if client == 'web' else {}
- if 'configs' not in self._configuration_arg('player_skip'):
- player_ytcfg = self._extract_player_ytcfg(client, video_id) or player_ytcfg
+ if 'configs' not in self._configuration_arg('player_skip') and client != 'web':
+ player_ytcfg = self._download_ytcfg(client, video_id) or player_ytcfg
player_url = player_url or self._extract_player_url(master_ytcfg, player_ytcfg, webpage=webpage)
require_js_player = self._get_default_ytcfg(client).get('REQUIRE_JS_PLAYER')
prs.append(pr)
# creator clients can bypass AGE_VERIFICATION_REQUIRED if logged in
- if client.endswith('_agegate') and self._is_unplayable(pr) and self.is_authenticated:
- append_client(client.replace('_agegate', '_creator'))
+ if variant == 'embedded' and self._is_unplayable(pr) and self.is_authenticated:
+ append_client(f'{base_client}_creator')
elif self._is_agegated(pr):
- append_client(f'{client}_embedded', f'{client.replace("_embedded", "")}_agegate')
+ if variant == 'tv_embedded':
+ append_client(f'{base_client}_embedded')
+ elif not variant:
+ append_client(f'tv_embedded.{base_client}', f'{base_client}_embedded')
if last_error:
if not len(prs):
streaming_formats = traverse_obj(streaming_data, (..., ('formats', 'adaptiveFormats'), ...), default=[])
for fmt in streaming_formats:
- if fmt.get('targetDurationSec') or fmt.get('drmFamilies'):
+ if fmt.get('targetDurationSec'):
continue
itag = str_or_none(fmt.get('itag'))
'fps': int_or_none(fmt.get('fps')) or None,
'height': height,
'quality': q(quality),
+ 'has_drm': bool(fmt.get('drmFamilies')),
'tbr': tbr,
'url': fmt_url,
'width': int_or_none(fmt.get('width')),
subtitles, automatic_captions = {}, {}
for lang_code, caption_track in captions.items():
base_url = caption_track.get('baseUrl')
+ orig_lang = parse_qs(base_url).get('lang', [None])[-1]
if not base_url:
continue
lang_name = self._get_text(caption_track, 'name', max_runs=1)
for trans_code, trans_name in translation_languages.items():
if not trans_code:
continue
+ orig_trans_code = trans_code
if caption_track.get('kind') != 'asr':
+ if 'translated_subs' in self._configuration_arg('skip'):
+ continue
trans_code += f'-{lang_code}'
trans_name += format_field(lang_name, template=' from %s')
# Add an "-orig" label to the original language so that it can be distinguished.
# The subs are returned without "-orig" as well for compatibility
- if lang_code == f'a-{trans_code}':
+ if lang_code == f'a-{orig_trans_code}':
process_language(
automatic_captions, base_url, f'{trans_code}-orig', f'{trans_name} (Original)', {})
# Setting tlang=lang returns damaged subtitles.
- # Not using lang_code == f'a-{trans_code}' here for future-proofing
- orig_lang = parse_qs(base_url).get('lang', [None])[-1]
process_language(automatic_captions, base_url, trans_code, trans_name,
- {} if orig_lang == trans_code else {'tlang': trans_code})
+ {} if orig_lang == orig_trans_code else {'tlang': trans_code})
info['automatic_captions'] = automatic_captions
info['subtitles'] = subtitles
def _extract_basic_item_renderer(item):
# Modified from _extract_grid_item_renderer
known_basic_renderers = (
- 'playlistRenderer', 'videoRenderer', 'channelRenderer', 'showRenderer'
+ 'playlistRenderer', 'videoRenderer', 'channelRenderer', 'showRenderer', 'reelItemRenderer'
)
for key, renderer in item.items():
if not isinstance(renderer, dict):
if video_id:
return self._extract_video(video_renderer)
+ def _hashtag_tile_entry(self, hashtag_tile_renderer):
+ url = urljoin('https://youtube.com', traverse_obj(
+ hashtag_tile_renderer, ('onTapCommand', 'commandMetadata', 'webCommandMetadata', 'url')))
+ if url:
+ return self.url_result(
+ url, ie=YoutubeTabIE.ie_key(), title=self._get_text(hashtag_tile_renderer, 'hashtag'))
+
def _post_thread_entries(self, post_thread_renderer):
post_renderer = try_get(
post_thread_renderer, lambda x: x['post']['backstagePostRenderer'], dict)
known_renderers = {
'playlistVideoListRenderer': self._playlist_entries,
'gridRenderer': self._grid_entries,
- 'shelfRenderer': lambda x: self._shelf_entries(x),
+ 'reelShelfRenderer': self._grid_entries,
+ 'shelfRenderer': self._shelf_entries,
'musicResponsiveListItemRenderer': lambda x: [self._music_reponsive_list_entry(x)],
'backstagePostThreadRenderer': self._post_thread_entries,
'videoRenderer': lambda x: [self._video_entry(x)],
'playlistRenderer': lambda x: self._grid_entries({'items': [{'playlistRenderer': x}]}),
'channelRenderer': lambda x: self._grid_entries({'items': [{'channelRenderer': x}]}),
+ 'hashtagTileRenderer': lambda x: [self._hashtag_tile_entry(x)]
}
for key, renderer in isr_content.items():
if key not in known_renderers:
if fatal:
raise ExtractorError('Unable to find selected tab')
- @classmethod
- def _extract_uploader(cls, data):
+ def _extract_uploader(self, data):
uploader = {}
- renderer = cls._extract_sidebar_info_renderer(data, 'playlistSidebarSecondaryInfoRenderer') or {}
+ renderer = self._extract_sidebar_info_renderer(data, 'playlistSidebarSecondaryInfoRenderer') or {}
owner = try_get(
renderer, lambda x: x['videoOwner']['videoOwnerRenderer']['title']['runs'][0], dict)
if owner:
- uploader['uploader'] = owner.get('text')
+ owner_text = owner.get('text')
+ uploader['uploader'] = self._search_regex(
+ r'^by (.+) and \d+ others?$', owner_text, 'uploader', default=owner_text)
uploader['uploader_id'] = try_get(
owner, lambda x: x['navigationEndpoint']['browseEndpoint']['browseId'], compat_str)
uploader['uploader_url'] = urljoin(
})
primary_thumbnails = self._extract_thumbnails(
- primary_sidebar_renderer, ('thumbnailRenderer', 'playlistVideoThumbnailRenderer', 'thumbnail'))
+ primary_sidebar_renderer, ('thumbnailRenderer', ('playlistVideoThumbnailRenderer', 'playlistCustomThumbnailRenderer'), 'thumbnail'))
if playlist_id is None:
playlist_id = item_id
check_get_keys='contents', fatal=False, ytcfg=ytcfg,
note='Downloading API JSON with unavailable videos')
+ @property
+ def skip_webpage(self):
+ return 'webpage' in self._configuration_arg('skip', ie_key=YoutubeTabIE.ie_key())
+
def _extract_webpage(self, url, item_id, fatal=True):
retries = self.get_param('extractor_retries', 3)
count = -1
return webpage, data
+ def _report_playlist_authcheck(self, ytcfg, fatal=True):
+ """Use if failed to extract ytcfg (and data) from initial webpage"""
+ if not ytcfg and self.is_authenticated:
+ msg = 'Playlists that require authentication may not extract correctly without a successful webpage download'
+ if 'authcheck' not in self._configuration_arg('skip', ie_key=YoutubeTabIE.ie_key()) and fatal:
+ raise ExtractorError(
+ f'{msg}. If you are not downloading private content, or '
+ 'your cookies are only for the first account and channel,'
+ ' pass "--extractor-args youtubetab:skip=authcheck" to skip this check',
+ expected=True)
+ self.report_warning(msg, only_once=True)
+
def _extract_data(self, url, item_id, ytcfg=None, fatal=True, webpage_fatal=False, default_client='web'):
data = None
- if 'webpage' not in self._configuration_arg('skip'):
+ if not self.skip_webpage:
webpage, data = self._extract_webpage(url, item_id, fatal=webpage_fatal)
ytcfg = ytcfg or self.extract_ytcfg(item_id, webpage)
# Reject webpage data if redirected to home page without explicitly requesting
raise ExtractorError(msg, expected=True)
self.report_warning(msg, only_once=True)
if not data:
- if not ytcfg and self.is_authenticated:
- msg = 'Playlists that require authentication may not extract correctly without a successful webpage download.'
- if 'authcheck' not in self._configuration_arg('skip') and fatal:
- raise ExtractorError(
- msg + ' If you are not downloading private content, or your cookies are only for the first account and channel,'
- ' pass "--extractor-args youtubetab:skip=authcheck" to skip this check',
- expected=True)
- self.report_warning(msg, only_once=True)
+ self._report_playlist_authcheck(ytcfg, fatal=fatal)
data = self._extract_tab_endpoint(url, item_id, ytcfg, fatal=fatal, default_client=default_client)
return data, ytcfg
('contents', 'tabbedSearchResultsRenderer', 'tabs', 0, 'tabRenderer', 'content', 'sectionListRenderer', 'contents'),
('continuationContents', ),
)
+ display_id = f'query "{query}"'
check_get_keys = tuple(set(keys[0] for keys in content_keys))
+ ytcfg = self._download_ytcfg(default_client, display_id) if not self.skip_webpage else {}
+ self._report_playlist_authcheck(ytcfg, fatal=False)
continuation_list = [None]
+ search = None
for page_num in itertools.count(1):
data.update(continuation_list[0] or {})
+ headers = self.generate_api_headers(
+ ytcfg=ytcfg, visitor_data=self._extract_visitor_data(search), default_client=default_client)
search = self._extract_response(
- item_id='query "%s" page %s' % (query, page_num), ep='search', query=data,
- default_client=default_client, check_get_keys=check_get_keys)
+ item_id=f'{display_id} page {page_num}', ep='search', query=data,
+ default_client=default_client, check_get_keys=check_get_keys, ytcfg=ytcfg, headers=headers)
slr_contents = traverse_obj(search, *content_keys)
yield from self._extract_entries({'contents': list(variadic(slr_contents))}, continuation_list)
if not continuation_list[0]:
'note': 'non-standard redirect to regional channel',
'url': 'https://www.youtube.com/channel/UCwVVpHQ2Cs9iGJfpdFngePQ',
'only_matching': True
+ }, {
+ 'note': 'collaborative playlist (uploader name in the form "by <uploader> and x other(s)")',
+ 'url': 'https://www.youtube.com/playlist?list=PLx-_-Kk4c89oOHEDQAojOXzEzemXxoqx6',
+ 'info_dict': {
+ 'id': 'PLx-_-Kk4c89oOHEDQAojOXzEzemXxoqx6',
+ 'modified_date': '20220407',
+ 'channel_url': 'https://www.youtube.com/channel/UCKcqXmCcyqnhgpA5P0oHH_Q',
+ 'tags': [],
+ 'uploader_id': 'UCKcqXmCcyqnhgpA5P0oHH_Q',
+ 'uploader': 'pukkandan',
+ 'availability': 'unlisted',
+ 'channel_id': 'UCKcqXmCcyqnhgpA5P0oHH_Q',
+ 'channel': 'pukkandan',
+ 'description': 'Test for collaborative playlist',
+ 'title': 'yt-dlp test - collaborative playlist',
+ 'uploader_url': 'https://www.youtube.com/channel/UCKcqXmCcyqnhgpA5P0oHH_Q',
+ },
+ 'playlist_mincount': 2
}]
@classmethod
'id': 'python',
'title': 'python',
}
-
+ }, {
+ 'url': 'https://www.youtube.com/results?search_query=%23cats',
+ 'playlist_mincount': 1,
+ 'info_dict': {
+ 'id': '#cats',
+ 'title': '#cats',
+ 'entries': [{
+ 'url': r're:https://(www\.)?youtube\.com/hashtag/cats',
+ 'title': '#cats',
+ }],
+ },
}, {
'url': 'https://www.youtube.com/results?q=test&sp=EgQIBBgB',
'only_matching': True,
Subclasses must define the _FEED_NAME property.
"""
_LOGIN_REQUIRED = True
- _TESTS = []
+
+ def _real_initialize(self):
+ YoutubeBaseInfoExtractor._check_login_required(self)
@property
def IE_NAME(self):
projects / yt-dlp.git / blobdiff