from ..swfinterp import SWFInterpreter
from ..compat import (
compat_chr,
+ compat_HTTPError,
compat_kwargs,
compat_parse_qs,
compat_urllib_parse_unquote,
_TFA_URL = 'https://accounts.google.com/_/signin/challenge?hl=en&TL={0}'
_RESERVED_NAMES = (
- r'course|embed|watch|w|results|storefront|'
- r'shared|index|account|reporthistory|t/terms|about|upload|signin|logout|'
- r'feed/(watch_later|history|subscriptions|library|trending|recommended)')
+ r'embed|e|watch_popup|channel|c|user|playlist|watch|w|v|movies|results|shared|'
+ r'storefront|oops|index|account|reporthistory|t/terms|about|upload|signin|logout|'
+ r'feed/(?:watch_later|history|subscriptions|library|trending|recommended)')
_NETRC_MACHINE = 'youtube'
# If True it will raise an error if no login info is provided
_LOGIN_REQUIRED = False
- _PLAYLIST_ID_RE = r'(?:(?:PL|LL|EC|UU|FL|RD|UL|TL|PU|OLAK5uy_)[0-9A-Za-z-_]{10,}|RDMM)'
-
- _YOUTUBE_CLIENT_HEADERS = {
- 'x-youtube-client-name': '1',
- 'x-youtube-client-version': '1.20200609.04.02',
- }
+ _PLAYLIST_ID_RE = r'(?:(?:PL|LL|EC|UU|FL|RD|UL|TL|PU|OLAK5uy_)[0-9A-Za-z-_]{10,}|RDMM|WL|LL|LM)'
def _set_language(self):
self._set_cookie(
}
_YT_INITIAL_DATA_RE = r'(?:window\s*\[\s*["\']ytInitialData["\']\s*\]|ytInitialData)\s*=\s*({.+?})\s*;'
+ _YT_INITIAL_PLAYER_RESPONSE_RE = r'ytInitialPlayerResponse\s*=\s*({.+?})\s*;'
+ _YT_INITIAL_BOUNDARY_RE = r'(?:var\s+meta|</script|\n)'
def _call_api(self, ep, query, video_id):
data = self._DEFAULT_API_DATA.copy()
def _extract_yt_initial_data(self, video_id, webpage):
return self._parse_json(
self._search_regex(
- (r'%s\s*\n' % self._YT_INITIAL_DATA_RE,
+ (r'%s\s*%s' % (self._YT_INITIAL_DATA_RE, self._YT_INITIAL_BOUNDARY_RE),
self._YT_INITIAL_DATA_RE), webpage, 'yt initial data'),
video_id)
+ def _extract_ytcfg(self, video_id, webpage):
+ return self._parse_json(
+ self._search_regex(
+ r'ytcfg\.set\s*\(\s*({.+?})\s*\)\s*;', webpage, 'ytcfg',
+ default='{}'), video_id, fatal=False)
+
class YoutubeIE(YoutubeBaseInfoExtractor):
IE_DESC = 'YouTube.com'
# Invidious instances taken from https://github.com/omarroth/invidious/wiki/Invidious-Instances
(?:(?:www|dev)\.)?invidio\.us/|
(?:(?:www|no)\.)?invidiou\.sh/|
- (?:(?:www|fi|de)\.)?invidious\.snopyta\.org/|
+ (?:(?:www|fi)\.)?invidious\.snopyta\.org/|
(?:www\.)?invidious\.kabi\.tk/|
(?:www\.)?invidious\.13ad\.de/|
(?:www\.)?invidious\.mastodon\.host/|
+ (?:www\.)?invidious\.zapashcanon\.fr/|
+ (?:www\.)?invidious\.kavin\.rocks/|
+ (?:www\.)?invidious\.tube/|
+ (?:www\.)?invidiou\.site/|
+ (?:www\.)?invidious\.site/|
+ (?:www\.)?invidious\.xyz/|
(?:www\.)?invidious\.nixnet\.xyz/|
(?:www\.)?invidious\.drycat\.fr/|
(?:www\.)?tube\.poal\.co/|
+ (?:www\.)?tube\.connect\.cafe/|
(?:www\.)?vid\.wxzm\.sx/|
+ (?:www\.)?vid\.mint\.lgbt/|
(?:www\.)?yewtu\.be/|
(?:www\.)?yt\.elukerio\.org/|
(?:www\.)?yt\.lelux\.fi/|
},
},
{
- # with '};' inside yt initial data (see https://github.com/ytdl-org/youtube-dl/issues/27093)
+ # with '};' inside yt initial data (see [1])
+ # see [2] for an example with '};' inside ytInitialPlayerResponse
+ # 1. https://github.com/ytdl-org/youtube-dl/issues/27093
+ # 2. https://github.com/ytdl-org/youtube-dl/issues/27216
'url': 'https://www.youtube.com/watch?v=CHqg6qOn4no',
'info_dict': {
'id': 'CHqg6qOn4no',
'skip_download': True,
},
},
+ {
+ # another example of '};' in ytInitialData
+ 'url': 'https://www.youtube.com/watch?v=gVfgbahppCY',
+ 'only_matching': True,
+ },
+ {
+ 'url': 'https://www.youtube.com/watch_popup?v=63RmMXCd_bQ',
+ 'only_matching': True,
+ },
]
def __init__(self, *args, **kwargs):
# https://github.com/ytdl-org/youtube-dl/pull/7599)
r';ytplayer\.config\s*=\s*({.+?});ytplayer',
r';ytplayer\.config\s*=\s*({.+?});',
- r'ytInitialPlayerResponse\s*=\s*({.+?});var meta' # Needed???
)
config = self._search_regex(
patterns, webpage, 'ytplayer.config', default=None)
return self._parse_json(
uppercase_escape(config), video_id, fatal=False)
- def _get_music_metadata_from_yt_initial(self, yt_initial):
- music_metadata = []
- key_map = {
- 'Album': 'album',
- 'Artist': 'artist',
- 'Song': 'track'
- }
- contents = try_get(yt_initial, lambda x: x['contents']['twoColumnWatchNextResults']['results']['results']['contents'])
- if type(contents) is list:
- for content in contents:
- music_track = {}
- if type(content) is not dict:
- continue
- videoSecondaryInfoRenderer = try_get(content, lambda x: x['videoSecondaryInfoRenderer'])
- if type(videoSecondaryInfoRenderer) is not dict:
- continue
- rows = try_get(videoSecondaryInfoRenderer, lambda x: x['metadataRowContainer']['metadataRowContainerRenderer']['rows'])
- if type(rows) is not list:
- continue
- for row in rows:
- metadataRowRenderer = try_get(row, lambda x: x['metadataRowRenderer'])
- if type(metadataRowRenderer) is not dict:
- continue
- key = try_get(metadataRowRenderer, lambda x: x['title']['simpleText'])
- value = try_get(metadataRowRenderer, lambda x: x['contents'][0]['simpleText']) or \
- try_get(metadataRowRenderer, lambda x: x['contents'][0]['runs'][0]['text'])
- if type(key) is not str or type(value) is not str:
- continue
- if key in key_map:
- if key_map[key] in music_track:
- # we've started on a new track
- music_metadata.append(music_track)
- music_track = {}
- music_track[key_map[key]] = value
- if len(music_track.keys()):
- music_metadata.append(music_track)
- return music_metadata
-
- def _get_automatic_captions(self, video_id, webpage):
+ def _get_automatic_captions(self, video_id, player_response, player_config):
"""We need the webpage for getting the captions url, pass it as an
argument to speed up the process."""
self.to_screen('%s: Looking for automatic captions' % video_id)
- player_config = self._get_ytplayer_config(video_id, webpage)
err_msg = 'Couldn\'t find automatic captions for %s' % video_id
- if not player_config:
+ if not (player_response or player_config):
self._downloader.report_warning(err_msg)
return {}
try:
- args = player_config['args']
+ args = player_config.get('args') if player_config else {}
caption_url = args.get('ttsurl')
if caption_url:
timestamp = args['timestamp']
return captions
# New captions format as of 22.06.2017
- player_response = args.get('player_response')
- if player_response and isinstance(player_response, compat_str):
- player_response = self._parse_json(
- player_response, video_id, fatal=False)
- if player_response:
- renderer = player_response['captions']['playerCaptionsTracklistRenderer']
- base_url = renderer['captionTracks'][0]['baseUrl']
- sub_lang_list = []
- for lang in renderer['translationLanguages']:
- lang_code = lang.get('languageCode')
- if lang_code:
- sub_lang_list.append(lang_code)
- return make_captions(base_url, sub_lang_list)
+ if player_response:
+ renderer = player_response['captions']['playerCaptionsTracklistRenderer']
+ base_url = renderer['captionTracks'][0]['baseUrl']
+ sub_lang_list = []
+ for lang in renderer['translationLanguages']:
+ lang_code = lang.get('languageCode')
+ if lang_code:
+ sub_lang_list.append(lang_code)
+ return make_captions(base_url, sub_lang_list)
# Some videos don't provide ttsurl but rather caption_tracks and
# caption_translation_languages (e.g. 20LmZk1hakA)
if embedded_config:
return embedded_config
+ video_info = {}
player_response = {}
+ ytplayer_config = None
+ embed_webpage = None
# Get video info
- video_info = {}
- embed_webpage = None
if (self._og_search_property('restrictions:age', video_webpage, default=None) == '18+'
or re.search(r'player-age-gate-content">', video_webpage) is not None):
cookie_keys = self._get_cookies('https://www.youtube.com').keys()
if not video_info and not player_response:
player_response = extract_player_response(
self._search_regex(
- r'ytInitialPlayerResponse\s*=\s*({.+?})\s*;', video_webpage,
+ (r'%s\s*%s' % (self._YT_INITIAL_PLAYER_RESPONSE_RE, self._YT_INITIAL_BOUNDARY_RE),
+ self._YT_INITIAL_PLAYER_RESPONSE_RE), video_webpage,
'initial player response', default='{}'),
video_id)
if not isinstance(video_info, dict):
video_info = {}
+ playable_in_embed = try_get(
+ player_response, lambda x: x['playabilityStatus']['playableInEmbed'])
+
video_details = try_get(
player_response, lambda x: x['videoDetails'], dict) or {}
# Youtube Music Auto-generated description
release_date = release_year = None
if video_description:
- mobj = re.search(r'(?s)Provided to YouTube by [^\n]+\n+(?P<track>[^·]+)·(?P<artist>[^\n]+)\n+(?P<album>[^\n]+)(?:.+?℗\s*(?P<release_year>\d{4})(?!\d))?(?:.+?Released on\s*:\s*(?P<release_date>\d{4}-\d{2}-\d{2}))?(.+?\nArtist\s*:\s*(?P<clean_artist>[^\n]+))?', video_description)
+ mobj = re.search(r'(?s)(?P<track>[^·\n]+)·(?P<artist>[^\n]+)\n+(?P<album>[^\n]+)(?:.+?℗\s*(?P<release_year>\d{4})(?!\d))?(?:.+?Released on\s*:\s*(?P<release_date>\d{4}-\d{2}-\d{2}))?(.+?\nArtist\s*:\s*(?P<clean_artist>[^\n]+))?.+\nAuto-generated by YouTube\.\s*$', video_description)
if mobj:
if not track:
track = mobj.group('track').strip()
if release_year:
release_year = int(release_year)
- yt_initial = self._get_yt_initial_data(video_id, video_webpage)
- if yt_initial:
- music_metadata = self._get_music_metadata_from_yt_initial(yt_initial)
- if len(music_metadata):
- album = music_metadata[0].get('album')
- artist = music_metadata[0].get('artist')
- track = music_metadata[0].get('track')
+ yt_initial_data = self._extract_yt_initial_data(video_id, video_webpage)
+ contents = try_get(yt_initial_data, lambda x: x['contents']['twoColumnWatchNextResults']['results']['results']['contents'], list) or []
+ for content in contents:
+ rows = try_get(content, lambda x: x['videoSecondaryInfoRenderer']['metadataRowContainer']['metadataRowContainerRenderer']['rows'], list) or []
+ multiple_songs = False
+ for row in rows:
+ if try_get(row, lambda x: x['metadataRowRenderer']['hasDividerLine']) is True:
+ multiple_songs = True
+ break
+ for row in rows:
+ mrr = row.get('metadataRowRenderer') or {}
+ mrr_title = try_get(
+ mrr, lambda x: x['title']['simpleText'], compat_str)
+ mrr_contents = try_get(
+ mrr, lambda x: x['contents'][0], dict) or {}
+ mrr_contents_text = try_get(mrr_contents, [lambda x: x['simpleText'], lambda x: x['runs'][0]['text']], compat_str)
+ if not (mrr_title and mrr_contents_text):
+ continue
+ if mrr_title == 'License':
+ video_license = mrr_contents_text
+ elif not multiple_songs:
+ if mrr_title == 'Album':
+ album = mrr_contents_text
+ elif mrr_title == 'Artist':
+ artist = mrr_contents_text
+ elif mrr_title == 'Song':
+ track = mrr_contents_text
m_episode = re.search(
r'<div[^>]+id="watch7-headline"[^>]*>\s*<span[^>]*>.*?>(?P<series>[^<]+)</a></b>\s*S(?P<season>\d+)\s*•\s*E(?P<episode>\d+)</span>',
# subtitles
video_subtitles = self.extract_subtitles(
video_id, video_webpage, has_live_chat_replay)
- automatic_captions = self.extract_automatic_captions(video_id, video_webpage)
+ automatic_captions = self.extract_automatic_captions(video_id, player_response, ytplayer_config)
video_duration = try_get(
video_info, lambda x: int_or_none(x['length_seconds'][0]))
# annotations
video_annotations = None
if self._downloader.params.get('writeannotations', False):
- xsrf_token = self._search_regex(
- r'([\'"])XSRF_TOKEN\1\s*:\s*([\'"])(?P<xsrf_token>[A-Za-z0-9+/=]+)\2',
- video_webpage, 'xsrf token', group='xsrf_token', fatal=False)
+ xsrf_token = None
+ ytcfg = self._extract_ytcfg(video_id, video_webpage)
+ if ytcfg:
+ xsrf_token = try_get(ytcfg, lambda x: x['XSRF_TOKEN'], compat_str)
+ if not xsrf_token:
+ xsrf_token = self._search_regex(
+ r'([\'"])XSRF_TOKEN\1\s*:\s*([\'"])(?P<xsrf_token>(?:(?!\2).)+)\2',
+ video_webpage, 'xsrf token', group='xsrf_token', fatal=False)
invideo_url = try_get(
player_response, lambda x: x['annotations'][0]['playerAnnotationsUrlsRenderer']['invideoUrl'], compat_str)
if xsrf_token and invideo_url:
- xsrf_field_name = self._search_regex(
- r'([\'"])XSRF_FIELD_NAME\1\s*:\s*([\'"])(?P<xsrf_field_name>\w+)\2',
- video_webpage, 'xsrf field name',
- group='xsrf_field_name', default='session_token')
+ xsrf_field_name = None
+ if ytcfg:
+ xsrf_field_name = try_get(ytcfg, lambda x: x['XSRF_FIELD_NAME'], compat_str)
+ if not xsrf_field_name:
+ xsrf_field_name = self._search_regex(
+ r'([\'"])XSRF_FIELD_NAME\1\s*:\s*([\'"])(?P<xsrf_field_name>\w+)\2',
+ video_webpage, 'xsrf field name',
+ group='xsrf_field_name', default='session_token')
video_annotations = self._download_webpage(
self._proto_relative_url(invideo_url),
video_id, note='Downloading annotations',
'release_date': release_date,
'release_year': release_year,
'subscriber_count': subscriber_count,
+ 'playable_in_embed': playable_in_embed,
}
class YoutubeTabIE(YoutubeBaseInfoExtractor):
IE_DESC = 'YouTube.com tab'
- _VALID_URL = (
- r'https?://(?:\w+\.)?(?:youtube(?:kids)?\.com|invidio\.us)/'
- r'(?:(?!(%s)([/#?]|$))|'
- r'(?:channel|c|user)/|'
- r'(?:playlist|watch)\?.*?\blist=)'
- r'(?P<id>[^/?#&]+)') % YoutubeBaseInfoExtractor._RESERVED_NAMES
+ _VALID_URL = r'''(?x)
+ https?://
+ (?:\w+\.)?
+ (?:
+ youtube(?:kids)?\.com|
+ invidio\.us
+ )/
+ (?:
+ (?:channel|c|user)/|
+ (?P<not_channel>
+ feed/|
+ (?:playlist|watch)\?.*?\blist=
+ )|
+ (?!(?:%s)\b) # Direct URLs
+ )
+ (?P<id>[^/?\#&]+)
+ ''' % YoutubeBaseInfoExtractor._RESERVED_NAMES
IE_NAME = 'youtube:tab'
_TESTS = [{
}, {
'url': 'https://www.youtube.com/c/CommanderVideoHq/live',
'only_matching': True,
- },
- # TODO
- # {
- # 'url': 'https://www.youtube.com/TheYoungTurks/live',
- # 'only_matching': True,
- # }
- ]
+ }, {
+ 'url': 'https://www.youtube.com/feed/trending',
+ 'only_matching': True,
+ }, {
+ # needs auth
+ 'url': 'https://www.youtube.com/feed/library',
+ 'only_matching': True,
+ }, {
+ # needs auth
+ 'url': 'https://www.youtube.com/feed/history',
+ 'only_matching': True,
+ }, {
+ # needs auth
+ 'url': 'https://www.youtube.com/feed/subscriptions',
+ 'only_matching': True,
+ }, {
+ # needs auth
+ 'url': 'https://www.youtube.com/feed/watch_later',
+ 'only_matching': True,
+ }, {
+ # no longer available?
+ 'url': 'https://www.youtube.com/feed/recommended',
+ 'only_matching': True,
+ }, {
+ # inline playlist with not always working continuations
+ 'url': 'https://www.youtube.com/watch?v=UC6u0Tct-Fo&list=PL36D642111D65BE7C',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://www.youtube.com/course?list=ECUl4u3cNGP61MdtwGTqZA0MreSaDybji8',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://www.youtube.com/course',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://www.youtube.com/zsecurity',
+ 'only_matching': True,
+ }, {
+ 'url': 'http://www.youtube.com/NASAgovVideo/videos',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://www.youtube.com/TheYoungTurks/live',
+ 'only_matching': True,
+ }]
+
+ @classmethod
+ def suitable(cls, url):
+ return False if YoutubeIE.suitable(url) else super(
+ YoutubeTabIE, cls).suitable(url)
def _extract_channel_id(self, webpage):
channel_id = self._html_search_meta(
'https://www.youtube.com/channel/%s' % channel_id,
ie=YoutubeTabIE.ie_key(), video_title=title)
- def _shelf_entries_trimmed(self, shelf_renderer):
- renderer = try_get(
- shelf_renderer, lambda x: x['content']['horizontalListRenderer'], dict)
- if not renderer:
+ def _shelf_entries_from_content(self, shelf_renderer):
+ content = shelf_renderer.get('content')
+ if not isinstance(content, dict):
return
- # TODO: add support for nested playlists so each shelf is processed
- # as separate playlist
- # TODO: this includes only first N items
- for entry in self._grid_entries(renderer):
- yield entry
+ renderer = content.get('gridRenderer')
+ if renderer:
+ # TODO: add support for nested playlists so each shelf is processed
+ # as separate playlist
+ # TODO: this includes only first N items
+ for entry in self._grid_entries(renderer):
+ yield entry
+ renderer = content.get('horizontalListRenderer')
+ if renderer:
+ # TODO
+ pass
- def _shelf_entries(self, shelf_renderer):
+ def _shelf_entries(self, shelf_renderer, skip_channels=False):
ep = try_get(
shelf_renderer, lambda x: x['endpoint']['commandMetadata']['webCommandMetadata']['url'],
compat_str)
shelf_url = urljoin('https://www.youtube.com', ep)
- if not shelf_url:
- return
- title = try_get(
- shelf_renderer, lambda x: x['title']['runs'][0]['text'], compat_str)
- yield self.url_result(shelf_url, video_title=title)
+ if shelf_url:
+ # Skipping links to another channels, note that checking for
+ # endpoint.commandMetadata.webCommandMetadata.webPageTypwebPageType == WEB_PAGE_TYPE_CHANNEL
+ # will not work
+ if skip_channels and '/channels?' in shelf_url:
+ return
+ title = try_get(
+ shelf_renderer, lambda x: x['title']['runs'][0]['text'], compat_str)
+ yield self.url_result(shelf_url, video_title=title)
+ # Shelf may not contain shelf URL, fallback to extraction from content
+ for entry in self._shelf_entries_from_content(shelf_renderer):
+ yield entry
def _playlist_entries(self, video_list_renderer):
for content in video_list_renderer['contents']:
continue
yield self._extract_video(renderer)
+ r""" # Not needed in the new implementation
def _itemSection_entries(self, item_sect_renderer):
for content in item_sect_renderer['contents']:
if not isinstance(content, dict):
if not video_id:
continue
yield self._extract_video(renderer)
+ """
def _rich_entries(self, rich_grid_renderer):
renderer = try_get(
- rich_grid_renderer, lambda x: x['content']['videoRenderer'], dict)
+ rich_grid_renderer, lambda x: x['content']['videoRenderer'], dict) or {}
video_id = renderer.get('videoId')
if not video_id:
return
for entry in self._post_thread_entries(renderer):
yield entry
+ @staticmethod
+ def _build_continuation_query(continuation, ctp=None):
+ query = {
+ 'ctoken': continuation,
+ 'continuation': continuation,
+ }
+ if ctp:
+ query['itct'] = ctp
+ return query
+
@staticmethod
def _extract_next_continuation_data(renderer):
next_continuation = try_get(
if not continuation:
return
ctp = next_continuation.get('clickTrackingParams')
- return {
- 'ctoken': continuation,
- 'continuation': continuation,
- 'itct': ctp,
- }
+ return YoutubeTabIE._build_continuation_query(continuation, ctp)
@classmethod
def _extract_continuation(cls, renderer):
if not continuation:
continue
ctp = continuation_ep.get('clickTrackingParams')
- if not ctp:
- continue
- return {
- 'ctoken': continuation,
- 'continuation': continuation,
- 'itct': ctp,
- }
+ return YoutubeTabIE._build_continuation_query(continuation, ctp)
def _entries(self, tab, identity_token):
- def extract_entries(parent_renderer):
- slr_contents = try_get(parent_renderer, lambda x: x['contents'], list) or []
- for slr_content in slr_contents:
- if not isinstance(slr_content, dict):
+ 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 []
+ for content in contents:
+ if not isinstance(content, dict):
continue
- is_renderer = try_get(slr_content, lambda x: x['itemSectionRenderer'], dict)
+ is_renderer = try_get(content, lambda x: x['itemSectionRenderer'], dict)
if not is_renderer:
- renderer = slr_content.get('richItemRenderer')
+ renderer = content.get('richItemRenderer')
if renderer:
for entry in self._rich_entries(renderer):
yield entry
continue
renderer = isr_content.get('shelfRenderer')
if renderer:
- for entry in self._shelf_entries(renderer):
+ is_channels_tab = tab.get('title') == 'Channels'
+ for entry in self._shelf_entries(renderer, not is_channels_tab):
yield entry
- continuation_list[0] = self._extract_continuation(parent_renderer)
continue
renderer = isr_content.get('backstagePostThreadRenderer')
if renderer:
entry = self._video_entry(renderer)
if entry:
yield entry
+
if not continuation_list[0]:
continuation_list[0] = self._extract_continuation(is_renderer)
- if not continuation_list[0]:
- continuation_list[0] = self._extract_continuation(parent_renderer)
+
+ if not continuation_list[0]:
+ continuation_list[0] = self._extract_continuation(parent_renderer)
continuation_list = [None] # Python 2 doesnot support nonlocal
+ tab_content = try_get(tab, lambda x: x['content'], dict)
+ if not tab_content:
+ return
parent_renderer = (
- try_get(tab, lambda x: x['sectionListRenderer'], dict)
- or try_get(tab, lambda x: x['richGridRenderer'], dict) or {})
- if parent_renderer:
- for entry in extract_entries(parent_renderer):
- yield entry
-
+ try_get(tab_content, lambda x: x['sectionListRenderer'], dict)
+ or try_get(tab_content, lambda x: x['richGridRenderer'], dict) or {})
+ for entry in extract_entries(parent_renderer):
+ yield entry
continuation = continuation_list[0]
headers = {
for page_num in itertools.count(1):
if not continuation:
break
- if hasattr(self, '_MAX_PAGES') and page_num > self._MAX_PAGES:
- break
- browse = self._download_json(
- 'https://www.youtube.com/browse_ajax', None,
- 'Downloading page %d' % page_num,
- headers=headers, query=continuation, fatal=False)
+ count = 0
+ retries = 3
+ while count <= retries:
+ try:
+ # Downloading page may result in intermittent 5xx HTTP error
+ # that is usually worked around with a retry
+ browse = self._download_json(
+ 'https://www.youtube.com/browse_ajax', None,
+ 'Downloading page %d%s'
+ % (page_num, ' (retry #%d)' % count if count else ''),
+ headers=headers, query=continuation)
+ break
+ except ExtractorError as e:
+ if isinstance(e.cause, compat_HTTPError) and e.cause.code in (500, 503):
+ count += 1
+ if count <= retries:
+ continue
+ raise
if not browse:
break
response = try_get(browse, lambda x: x[1]['response'], dict)
yield entry
continuation = self._extract_continuation(continuation_renderer)
continue
- continuation_renderer = continuation_contents.get('sectionListContinuation')
+ continuation_renderer = continuation_contents.get('sectionListContinuation') # for feeds
if continuation_renderer:
continuation_list = [None]
for entry in extract_entries(continuation_renderer):
continuation_item = continuation_items[0]
if not isinstance(continuation_item, dict):
continue
- renderer = continuation_item.get('playlistVideoRenderer')
+ renderer = continuation_item.get('playlistVideoRenderer') or continuation_item.get('itemSectionRenderer')
if renderer:
video_list_renderer = {'contents': continuation_items}
for entry in self._playlist_entries(video_list_renderer):
yield entry
continuation = self._extract_continuation(video_list_renderer)
continue
- renderer = continuation_item.get('itemSectionRenderer')
- if renderer:
- for entry in self._itemSection_entries(renderer):
- yield entry
- continuation = self._extract_continuation({'contents': continuation_items})
- continue
break
@staticmethod
selected_tab = self._extract_selected_tab(tabs)
renderer = try_get(
data, lambda x: x['metadata']['channelMetadataRenderer'], dict)
- playlist_id = None
+ playlist_id = title = description = None
if renderer:
channel_title = renderer.get('title') or item_id
tab_title = selected_tab.get('title')
description = None
playlist_id = item_id
if playlist_id is None:
- return None
+ playlist_id = item_id
+ if title is None:
+ title = "Youtube " + playlist_id.title()
playlist = self.playlist_result(
- self._entries(selected_tab['content'], identity_token),
+ self._entries(selected_tab, identity_token),
playlist_id=playlist_id, playlist_title=title,
playlist_description=description)
playlist.update(self._extract_uploader(data))
return playlist
- def _extract_from_playlist(self, item_id, data, playlist):
+ def _extract_from_playlist(self, item_id, url, data, playlist):
title = playlist.get('title') or try_get(
data, lambda x: x['titleText']['simpleText'], compat_str)
playlist_id = playlist.get('playlistId') or item_id
+ # Inline playlist rendition continuation does not always work
+ # at Youtube side, so delegating regular tab-based playlist URL
+ # processing whenever possible.
+ playlist_url = urljoin(url, try_get(
+ playlist, lambda x: x['endpoint']['commandMetadata']['webCommandMetadata']['url'],
+ compat_str))
+ if playlist_url and playlist_url != url:
+ return self.url_result(
+ playlist_url, ie=YoutubeTabIE.ie_key(), video_id=playlist_id,
+ video_title=title)
return self.playlist_result(
self._playlist_entries(playlist), playlist_id=playlist_id,
playlist_title=title)
+ @staticmethod
+ def _extract_alerts(data):
+ for alert_dict in try_get(data, lambda x: x['alerts'], list) or []:
+ if not isinstance(alert_dict, dict):
+ continue
+ for renderer in alert_dict:
+ alert = alert_dict[renderer]
+ alert_type = alert.get('type')
+ if not alert_type:
+ continue
+ message = try_get(alert, lambda x: x['text']['simpleText'], compat_str)
+ if message:
+ yield alert_type, message
+ for run in try_get(alert, lambda x: x['text']['runs'], list) or []:
+ message = try_get(run, lambda x: x['text'], compat_str)
+ if message:
+ yield alert_type, message
+
+ def _extract_identity_token(self, webpage, item_id):
+ ytcfg = self._extract_ytcfg(item_id, webpage)
+ if ytcfg:
+ token = try_get(ytcfg, lambda x: x['ID_TOKEN'], compat_str)
+ if token:
+ return token
+ return self._search_regex(
+ r'\bID_TOKEN["\']\s*:\s*["\'](.+?)["\']', webpage,
+ 'identity token', default=None)
+
def _real_extract(self, url):
item_id = self._match_id(url)
url = compat_urlparse.urlunparse(
compat_urlparse.urlparse(url)._replace(netloc='www.youtube.com'))
+ is_home = re.match(r'(?P<pre>%s)(?P<post>/?(?![^#?]).*$)' % self._VALID_URL, url)
+ if is_home is not None and is_home.group('not_channel') is None and item_id != 'feed':
+ self._downloader.report_warning(
+ 'A channel/user page was given. All the channel\'s videos will be downloaded. '
+ 'To download only the videos in the home page, add a "/featured" to the URL')
+ url = '%s/videos%s' % (is_home.group('pre'), is_home.group('post') or '')
+
# Handle both video/playlist URLs
qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
video_id = qs.get('v', [None])[0]
playlist_id = qs.get('list', [None])[0]
+
+ if is_home is not None and is_home.group('not_channel') is not None and is_home.group('not_channel').startswith('watch') and not video_id:
+ if playlist_id:
+ self._downloader.report_warning('%s is not a valid Youtube URL. Trying to download playlist %s' % (url, playlist_id))
+ url = 'https://www.youtube.com/playlist?list=%s' % playlist_id
+ # return self.url_result(playlist_id, ie=YoutubePlaylistIE.ie_key())
+ else:
+ raise ExtractorError('Unable to recognize tab page')
if video_id and playlist_id:
if self._downloader.params.get('noplaylist'):
self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
return self.url_result(video_id, ie=YoutubeIE.ie_key(), video_id=video_id)
self.to_screen('Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
+
webpage = self._download_webpage(url, item_id)
- identity_token = self._search_regex(
- r'\bID_TOKEN["\']\s*:\s/l*["\'](.+?)["\']', webpage,
- 'identity token', default=None)
+ identity_token = self._extract_identity_token(webpage, item_id)
data = self._extract_yt_initial_data(item_id, webpage)
+ for alert_type, alert_message in self._extract_alerts(data):
+ self._downloader.report_warning('YouTube said: %s - %s' % (alert_type, alert_message))
tabs = try_get(
data, lambda x: x['contents']['twoColumnBrowseResultsRenderer']['tabs'], list)
if tabs:
playlist = try_get(
data, lambda x: x['contents']['twoColumnWatchNextResults']['playlist']['playlist'], dict)
if playlist:
- return self._extract_from_playlist(item_id, data, playlist)
+ return self._extract_from_playlist(item_id, url, data, playlist)
# Fallback to video extraction if no playlist alike page is recognized.
# First check for the current video then try the v attribute of URL query.
video_id = try_get(
(?:
(?:
youtube(?:kids)?\.com|
- invidio\.us|
- youtu\.be
+ invidio\.us
)
/.*?\?.*?\blist=
)?
'uploader_id': 'UC21nz3_MesPLqtDqwdvnoxA',
}
}, {
+ 'url': 'TLGGrESM50VT6acwMjAyMjAxNw',
+ 'only_matching': True,
+ }, {
+ # music album playlist
+ 'url': 'OLAK5uy_m4xAFdmMC5rX3Ji3g93pQe3hqLZw_9LhM',
+ 'only_matching': True,
+ }]
+
+ @classmethod
+ def suitable(cls, url):
+ return False if YoutubeTabIE.suitable(url) else super(
+ YoutubePlaylistIE, cls).suitable(url)
+
+ def _real_extract(self, url):
+ playlist_id = self._match_id(url)
+ qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
+ if not qs:
+ qs = {'list': playlist_id}
+ return self.url_result(
+ update_url_query('https://www.youtube.com/playlist', qs),
+ ie=YoutubeTabIE.ie_key(), video_id=playlist_id)
+
+
+class YoutubeYtBeIE(InfoExtractor):
+ IE_DESC = 'youtu.be'
+ _VALID_URL = r'https?://youtu\.be/(?P<id>[0-9A-Za-z_-]{11})/*?.*?\blist=(?P<playlist_id>%(playlist_id)s)' % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE}
+ _TESTS = [{
'url': 'https://youtu.be/yeWKywCrFtk?list=PL2qgrgXsNUG5ig9cat4ohreBjYLAPC0J5',
'info_dict': {
'id': 'yeWKywCrFtk',
}, {
'url': 'https://youtu.be/uWyaPkt-VOI?list=PL9D9FC436B881BA21',
'only_matching': True,
- }, {
- 'url': 'TLGGrESM50VT6acwMjAyMjAxNw',
- 'only_matching': True,
- }, {
- # music album playlist
- 'url': 'OLAK5uy_m4xAFdmMC5rX3Ji3g93pQe3hqLZw_9LhM',
- 'only_matching': True,
}]
- @classmethod
- def suitable(cls, url):
- return False if YoutubeTabIE.suitable(url) else super(
- YoutubePlaylistIE, cls).suitable(url)
-
def _real_extract(self, url):
- playlist_id = self._match_id(url)
- qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
- if not qs:
- qs = {'list': playlist_id}
+ mobj = re.match(self._VALID_URL, url)
+ video_id = mobj.group('id')
+ playlist_id = mobj.group('playlist_id')
return self.url_result(
- update_url_query('https://www.youtube.com/playlist', qs),
- ie=YoutubeTabIE.ie_key(), video_id=playlist_id)
+ update_url_query('https://www.youtube.com/watch', {
+ 'v': video_id,
+ 'list': playlist_id,
+ 'feature': 'youtu.be',
+ }), ie=YoutubeTabIE.ie_key(), video_id=playlist_id)
class YoutubeYtUserIE(InfoExtractor):
+ IE_DESC = 'YouTube.com user videos, URL or "ytuser" keyword'
_VALID_URL = r'ytuser:(?P<id>.+)'
_TESTS = [{
'url': 'ytuser:phihag',
ie=YoutubeTabIE.ie_key(), video_id=user_id)
+class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
+ IE_NAME = 'youtube:favorites'
+ IE_DESC = 'YouTube.com liked videos, ":ytfav" for short (requires authentication)'
+ _VALID_URL = r':ytfav(?:ou?rite)?s?'
+ _LOGIN_REQUIRED = True
+ _TESTS = [{
+ 'url': ':ytfav',
+ 'only_matching': True,
+ }, {
+ 'url': ':ytfavorites',
+ 'only_matching': True,
+ }]
+
+ def _real_extract(self, url):
+ return self.url_result(
+ 'https://www.youtube.com/playlist?list=LL',
+ ie=YoutubeTabIE.ie_key())
+
+
class YoutubeSearchIE(SearchInfoExtractor, YoutubeBaseInfoExtractor):
IE_DESC = 'YouTube.com searches'
# there doesn't appear to be a real limit, for example if you search for
list)
if not slr_contents:
break
- isr_contents = try_get(
- slr_contents,
- lambda x: x[0]['itemSectionRenderer']['contents'],
- list)
+
+ isr_contents = []
+ continuation_token = None
+ # Youtube sometimes adds promoted content to searches,
+ # changing the index location of videos and token.
+ # So we search through all entries till we find them.
+ for index, isr in enumerate(slr_contents):
+ if not isr_contents:
+ isr_contents = try_get(
+ slr_contents,
+ (lambda x: x[index]['itemSectionRenderer']['contents']),
+ list)
+ for content in isr_contents:
+ if content.get('videoRenderer') is not None:
+ break
+ else:
+ isr_contents = []
+
+ if continuation_token is None:
+ continuation_token = try_get(
+ slr_contents,
+ lambda x: x[index]['continuationItemRenderer']['continuationEndpoint']['continuationCommand'][
+ 'token'],
+ compat_str)
+ if continuation_token is not None and isr_contents:
+ break
+
if not isr_contents:
break
for content in isr_contents:
description = try_get(video, lambda x: x['descriptionSnippet']['runs'][0]['text'], compat_str)
duration = parse_duration(try_get(video, lambda x: x['lengthText']['simpleText'], compat_str))
view_count_text = try_get(video, lambda x: x['viewCountText']['simpleText'], compat_str) or ''
- view_count = int_or_none(self._search_regex(
- r'^(\d+)', re.sub(r'\s', '', view_count_text),
+ view_count = str_to_int(self._search_regex(
+ r'^([\d,]+)', re.sub(r'\s', '', view_count_text),
'view count', default=None))
uploader = try_get(video, lambda x: x['ownerText']['runs'][0]['text'], compat_str)
total += 1
}
if total == n:
return
- token = try_get(
- slr_contents,
- lambda x: x[1]['continuationItemRenderer']['continuationEndpoint']['continuationCommand']['token'],
- compat_str)
- if not token:
+ if not continuation_token:
break
- data['continuation'] = token
+ data['continuation'] = continuation_token
def _get_n_results(self, query, n):
"""Get a specified number of results for a query"""
class YoutubeSearchDateIE(YoutubeSearchIE):
IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
_SEARCH_KEY = 'ytsearchdate'
- IE_DESC = 'YouTube.com searches, newest videos first'
+ IE_DESC = 'YouTube.com searches, newest videos first, "ytsearchdate" keyword'
_SEARCH_PARAMS = 'CAI%3D'
-class YoutubeSearchURLIE(InfoExtractor):
- IE_DESC = 'YouTube.com search URLs'
- IE_NAME = 'youtube:search_url'
- _VALID_URL = r'https?://(?:www\.)?youtube\.com/results/?(?:\?|\?[^#]*?&)(?:sp=(?P<param1>[^&#]+)&(?:[^#]*&)?)?(?:q|search_query)=(?P<query>[^#&]+)(?:[^#]*?&sp=(?P<param2>[^#&]+))?'
+class YoutubeSearchURLIE(YoutubeSearchIE):
+ IE_DESC = 'YouTube.com searches, "ytsearch" keyword'
+ IE_NAME = YoutubeSearchIE.IE_NAME + '_url'
+ _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?(?:search_query|q)=(?:[^&]+)(?:[&]|$)'
# _MAX_RESULTS = 100
_TESTS = [{
'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video',
'only_matching': True,
}]
+ @classmethod
+ def _make_valid_url(cls):
+ return cls._VALID_URL
+
def _real_extract(self, url):
- mobj = re.match(self._VALID_URL, url)
- query = compat_urllib_parse_unquote_plus(mobj.group('query'))
- IE = YoutubeSearchIE(self._downloader)
- IE._SEARCH_PARAMS = mobj.group('param1') or mobj.group('param2')
- if hasattr(self, '_MAX_RESULTS'):
- IE._MAX_RESULTS = self._MAX_RESULTS
- return IE._get_n_results(query, IE._MAX_RESULTS)
+ qs = compat_parse_qs(compat_urllib_parse_urlparse(url).query)
+ query = (qs.get('search_query') or qs.get('q'))[0]
+ self._SEARCH_PARAMS = qs.get('sp', ('',))[0]
+ return self._get_n_results(query, self._MAX_RESULTS)
class YoutubeFeedsInfoExtractor(YoutubeTabIE):
"""
Base class for feed extractors
- Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
+ Subclasses must define the _FEED_NAME property.
"""
_LOGIN_REQUIRED = True
# _MAX_PAGES = 5
def _real_initialize(self):
self._login()
- def _shelf_entries(self, shelf_renderer):
- renderer = try_get(shelf_renderer, lambda x: x['content']['gridRenderer'], dict)
- if not renderer:
- return
- for entry in self._grid_entries(renderer):
- yield entry
-
- def _extract_from_tabs(self, item_id, webpage, data, tabs, identity_token):
- selected_tab = self._extract_selected_tab(tabs)
- return self.playlist_result(
- self._entries(selected_tab['content'], identity_token),
- playlist_title=self._PLAYLIST_TITLE)
-
def _real_extract(self, url):
- item_id = self._FEED_NAME
- url = 'https://www.youtube.com/feed/%s' % self._FEED_NAME
- webpage = self._download_webpage(url, item_id)
- identity_token = self._search_regex(
- r'\bID_TOKEN["\']\s*:\s*["\'](.+?)["\']', webpage,
- 'identity token', default=None)
- data = self._extract_yt_initial_data(item_id, webpage)
- tabs = try_get(
- data, lambda x: x['contents']['twoColumnBrowseResultsRenderer']['tabs'], list)
- if tabs:
- return self._extract_from_tabs(item_id, webpage, data, tabs, identity_token)
- # Failed to recognize
- raise ExtractorError('Unable to recognize feed page')
+ return self.url_result(
+ 'https://www.youtube.com/feed/%s' % self._FEED_NAME,
+ ie=YoutubeTabIE.ie_key())
class YoutubeWatchLaterIE(InfoExtractor):
IE_NAME = 'youtube:watchlater'
- IE_DESC = 'Youtube watch later list, ":ytwatchlater" or "WL" for short (requires authentication)'
- _VALID_URL = r'https?://(?:www\.)?youtube\.com/feed/watch_later|:ytwatchlater|WL'
-
+ IE_DESC = 'Youtube watch later list, ":ytwatchlater" for short (requires authentication)'
+ _VALID_URL = r':ytwatchlater'
_TESTS = [{
- 'url': 'https://www.youtube.com/feed/watch_later',
- 'only_matching': True,
- }, {
'url': ':ytwatchlater',
'only_matching': True,
}]
'https://www.youtube.com/playlist?list=WL', ie=YoutubeTabIE.ie_key())
-class YoutubeFavouritesIE(InfoExtractor):
- IE_NAME = 'youtube:favourites'
- IE_DESC = 'YouTube.com liked videos, ":ytfav" or "LL" for short (requires authentication)'
- _VALID_URL = r':ytfav(?:ou?rite)?s?|LL'
-
- _TESTS = [{
- 'url': ':ytfav',
- 'only_matching': True,
- }]
-
- def _real_extract(self, url):
- return self.url_result(
- 'https://www.youtube.com/playlist?list=LL', ie=YoutubeTabIE.ie_key())
-
-
class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
IE_DESC = 'YouTube.com recommended videos, ":ytrec" for short (requires authentication)'
- _VALID_URL = r'https?://(?:www\.)?youtube\.com(?:/feed/recommended|/?[?#]|/?$)|:ytrec(?:ommended)?'
+ _VALID_URL = r'https?://(?:www\.)?youtube\.com/?(?:[?#]|$)|:ytrec(?:ommended)?'
_FEED_NAME = 'recommended'
- _PLAYLIST_TITLE = 'Youtube Recommended videos'
+ _TESTS = [{
+ 'url': ':ytrec',
+ 'only_matching': True,
+ }, {
+ 'url': ':ytrecommended',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://youtube.com',
+ 'only_matching': True,
+ }]
class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
- IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
- _VALID_URL = r'https?://(?:www\.)?youtube\.com/feed/subscriptions|:ytsub(?:scription)?s?'
+ IE_DESC = 'YouTube.com subscriptions feed, ":ytsubs" for short (requires authentication)'
+ _VALID_URL = r':ytsub(?:scription)?s?'
_FEED_NAME = 'subscriptions'
- _PLAYLIST_TITLE = 'Youtube Subscriptions'
+ _TESTS = [{
+ 'url': ':ytsubs',
+ 'only_matching': True,
+ }, {
+ 'url': ':ytsubscriptions',
+ 'only_matching': True,
+ }]
class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
IE_DESC = 'Youtube watch history, ":ythistory" for short (requires authentication)'
- _VALID_URL = r'https?://(?:www\.)?youtube\.com/feed/history|:ythistory'
+ _VALID_URL = r':ythistory'
_FEED_NAME = 'history'
- _PLAYLIST_TITLE = 'Youtube History'
+ _TESTS = [{
+ 'url': ':ythistory',
+ 'only_matching': True,
+ }]
class YoutubeTruncatedURLIE(InfoExtractor):
projects / yt-dlp.git / blobdiff