import base64 import calendar import copy import datetime import functools import hashlib import itertools import json import math import os.path import random import re import sys import threading import time import traceback from .common import InfoExtractor, SearchInfoExtractor from ..compat import ( compat_chr, compat_HTTPError, compat_parse_qs, compat_str, compat_urllib_parse_unquote_plus, compat_urllib_parse_urlencode, compat_urllib_parse_urlparse, compat_urlparse, ) from ..jsinterp import JSInterpreter from ..utils import ( NO_DEFAULT, ExtractorError, bug_reports_message, classproperty, clean_html, datetime_from_str, dict_get, error_to_compat_str, float_or_none, format_field, get_first, int_or_none, is_html, join_nonempty, js_to_json, mimetype2ext, network_exceptions, orderedSet, parse_codecs, parse_count, parse_duration, parse_iso8601, parse_qs, qualities, remove_end, remove_start, smuggle_url, str_or_none, str_to_int, strftime_or_none, traverse_obj, try_get, unescapeHTML, unified_strdate, unified_timestamp, unsmuggle_url, update_url_query, url_or_none, urljoin, variadic, ) # any clients starting with _ cannot be explicity requested by the user INNERTUBE_CLIENTS = { 'web': { 'INNERTUBE_API_KEY': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8', 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'WEB', 'clientVersion': '2.20211221.00.00', } }, 'INNERTUBE_CONTEXT_CLIENT_NAME': 1 }, 'web_embedded': { 'INNERTUBE_API_KEY': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8', 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'WEB_EMBEDDED_PLAYER', 'clientVersion': '1.20211215.00.01', }, }, 'INNERTUBE_CONTEXT_CLIENT_NAME': 56 }, 'web_music': { 'INNERTUBE_API_KEY': 'AIzaSyC9XL3ZjWddXya6X74dJoCTL-WEYFDNX30', 'INNERTUBE_HOST': 'music.youtube.com', 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'WEB_REMIX', 'clientVersion': '1.20211213.00.00', } }, 'INNERTUBE_CONTEXT_CLIENT_NAME': 67, }, 'web_creator': { 'INNERTUBE_API_KEY': 'AIzaSyBUPetSUmoZL-OhlxA7wSac5XinrygCqMo', 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'WEB_CREATOR', 'clientVersion': '1.20211220.02.00', } }, 'INNERTUBE_CONTEXT_CLIENT_NAME': 62, }, 'android': { 'INNERTUBE_API_KEY': 'AIzaSyA8eiZmM1FaDVjRy-df2KTyQ_vz_yYM39w', 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'ANDROID', 'clientVersion': '16.49', } }, 'INNERTUBE_CONTEXT_CLIENT_NAME': 3, 'REQUIRE_JS_PLAYER': False }, 'android_embedded': { 'INNERTUBE_API_KEY': 'AIzaSyCjc_pVEDi4qsv5MtC2dMXzpIaDoRFLsxw', 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'ANDROID_EMBEDDED_PLAYER', 'clientVersion': '16.49', }, }, 'INNERTUBE_CONTEXT_CLIENT_NAME': 55, 'REQUIRE_JS_PLAYER': False }, 'android_music': { 'INNERTUBE_API_KEY': 'AIzaSyAOghZGza2MQSZkY_zfZ370N-PUdXEo8AI', 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'ANDROID_MUSIC', 'clientVersion': '4.57', } }, 'INNERTUBE_CONTEXT_CLIENT_NAME': 21, 'REQUIRE_JS_PLAYER': False }, 'android_creator': { 'INNERTUBE_API_KEY': 'AIzaSyD_qjV8zaaUMehtLkrKFgVeSX_Iqbtyws8', 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'ANDROID_CREATOR', 'clientVersion': '21.47', }, }, 'INNERTUBE_CONTEXT_CLIENT_NAME': 14, 'REQUIRE_JS_PLAYER': False }, # iOS clients have HLS live streams. Setting device model to get 60fps formats. # See: https://github.com/TeamNewPipe/NewPipeExtractor/issues/680#issuecomment-1002724558 'ios': { 'INNERTUBE_API_KEY': 'AIzaSyB-63vPrdThhKuerbB2N_l7Kwwcxj6yUAc', 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'IOS', 'clientVersion': '16.46', 'deviceModel': 'iPhone14,3', } }, 'INNERTUBE_CONTEXT_CLIENT_NAME': 5, 'REQUIRE_JS_PLAYER': False }, 'ios_embedded': { 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'IOS_MESSAGES_EXTENSION', 'clientVersion': '16.46', 'deviceModel': 'iPhone14,3', }, }, 'INNERTUBE_CONTEXT_CLIENT_NAME': 66, 'REQUIRE_JS_PLAYER': False }, 'ios_music': { 'INNERTUBE_API_KEY': 'AIzaSyBAETezhkwP0ZWA02RsqT1zu78Fpt0bC_s', 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'IOS_MUSIC', 'clientVersion': '4.57', }, }, 'INNERTUBE_CONTEXT_CLIENT_NAME': 26, 'REQUIRE_JS_PLAYER': False }, 'ios_creator': { 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'IOS_CREATOR', 'clientVersion': '21.47', }, }, 'INNERTUBE_CONTEXT_CLIENT_NAME': 15, 'REQUIRE_JS_PLAYER': False }, # mweb has 'ultralow' formats # See: https://github.com/yt-dlp/yt-dlp/pull/557 'mweb': { 'INNERTUBE_API_KEY': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8', 'INNERTUBE_CONTEXT': { 'client': { 'clientName': 'MWEB', 'clientVersion': '2.20211221.01.00', } }, '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://www.youtube.com/', # Can be any valid URL } BASE_CLIENTS = ('android', 'web', 'tv', 'ios', 'mweb') priority = qualities(BASE_CLIENTS[::-1]) for client, ytcfg in tuple(INNERTUBE_CLIENTS.items()): ytcfg.setdefault('INNERTUBE_API_KEY', 'AIzaSyDCU8hByM-4DrUqRUYnGn-3llEO78bcxq8') ytcfg.setdefault('INNERTUBE_HOST', 'www.youtube.com') ytcfg.setdefault('REQUIRE_JS_PLAYER', True) ytcfg['INNERTUBE_CONTEXT']['client'].setdefault('hl', 'en') _, base_client, variant = _split_innertube_client(client) ytcfg['priority'] = 10 * priority(base_client) if not variant: 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: ytcfg['priority'] -= 3 build_innertube_clients() class YoutubeBaseInfoExtractor(InfoExtractor): """Provide base functions for Youtube extractors""" _RESERVED_NAMES = ( r'channel|c|user|playlist|watch|w|v|embed|e|watch_popup|clip|' r'shorts|movies|results|search|shared|hashtag|trending|explore|feed|feeds|' r'browse|oembed|get_video_info|iframe_api|s/player|' r'storefront|oops|index|account|reporthistory|t/terms|about|upload|signin|logout') _PLAYLIST_ID_RE = r'(?:(?:PL|LL|EC|UU|FL|RD|UL|TL|PU|OLAK5uy_)[0-9A-Za-z-_]{10,}|RDMM|WL|LL|LM)' # _NETRC_MACHINE = 'youtube' # If True it will raise an error if no login info is provided _LOGIN_REQUIRED = False _INVIDIOUS_SITES = ( # invidious-redirect websites r'(?:www\.)?redirect\.invidious\.io', r'(?:(?:www|dev)\.)?invidio\.us', # Invidious instances taken from https://github.com/iv-org/documentation/blob/master/docs/instances.md r'(?:www\.)?invidious\.pussthecat\.org', r'(?:www\.)?invidious\.zee\.li', r'(?:www\.)?invidious\.ethibox\.fr', r'(?:www\.)?invidious\.3o7z6yfxhbw7n3za4rss6l434kmv55cgw2vuziwuigpwegswvwzqipyd\.onion', r'(?:www\.)?osbivz6guyeahrwp2lnwyjk2xos342h4ocsxyqrlaopqjuhwn2djiiyd\.onion', r'(?:www\.)?u2cvlit75owumwpy4dj2hsmvkq7nvrclkpht7xgyye2pyoxhpmclkrad\.onion', # youtube-dl invidious instances list r'(?:(?:www|no)\.)?invidiou\.sh', r'(?:(?:www|fi)\.)?invidious\.snopyta\.org', r'(?:www\.)?invidious\.kabi\.tk', r'(?:www\.)?invidious\.mastodon\.host', r'(?:www\.)?invidious\.zapashcanon\.fr', r'(?:www\.)?(?:invidious(?:-us)?|piped)\.kavin\.rocks', r'(?:www\.)?invidious\.tinfoil-hat\.net', r'(?:www\.)?invidious\.himiko\.cloud', r'(?:www\.)?invidious\.reallyancient\.tech', r'(?:www\.)?invidious\.tube', r'(?:www\.)?invidiou\.site', r'(?:www\.)?invidious\.site', r'(?:www\.)?invidious\.xyz', r'(?:www\.)?invidious\.nixnet\.xyz', r'(?:www\.)?invidious\.048596\.xyz', r'(?:www\.)?invidious\.drycat\.fr', r'(?:www\.)?inv\.skyn3t\.in', r'(?:www\.)?tube\.poal\.co', r'(?:www\.)?tube\.connect\.cafe', r'(?:www\.)?vid\.wxzm\.sx', r'(?:www\.)?vid\.mint\.lgbt', r'(?:www\.)?vid\.puffyan\.us', r'(?:www\.)?yewtu\.be', r'(?:www\.)?yt\.elukerio\.org', r'(?:www\.)?yt\.lelux\.fi', r'(?:www\.)?invidious\.ggc-project\.de', r'(?:www\.)?yt\.maisputain\.ovh', r'(?:www\.)?ytprivate\.com', r'(?:www\.)?invidious\.13ad\.de', r'(?:www\.)?invidious\.toot\.koeln', r'(?:www\.)?invidious\.fdn\.fr', r'(?:www\.)?watch\.nettohikari\.com', r'(?:www\.)?invidious\.namazso\.eu', r'(?:www\.)?invidious\.silkky\.cloud', r'(?:www\.)?invidious\.exonip\.de', r'(?:www\.)?invidious\.riverside\.rocks', r'(?:www\.)?invidious\.blamefran\.net', r'(?:www\.)?invidious\.moomoo\.de', r'(?:www\.)?ytb\.trom\.tf', r'(?:www\.)?yt\.cyberhost\.uk', r'(?:www\.)?kgg2m7yk5aybusll\.onion', r'(?:www\.)?qklhadlycap4cnod\.onion', r'(?:www\.)?axqzx4s6s54s32yentfqojs3x5i7faxza6xo3ehd4bzzsg2ii4fv2iid\.onion', r'(?:www\.)?c7hqkpkpemu6e7emz5b4vyz7idjgdvgaaa3dyimmeojqbgpea3xqjoid\.onion', r'(?:www\.)?fz253lmuao3strwbfbmx46yu7acac2jz27iwtorgmbqlkurlclmancad\.onion', r'(?:www\.)?invidious\.l4qlywnpwqsluw65ts7md3khrivpirse744un3x7mlskqauz5pyuzgqd\.onion', r'(?:www\.)?owxfohz4kjyv25fvlqilyxast7inivgiktls3th44jhk3ej3i7ya\.b32\.i2p', r'(?:www\.)?4l2dgddgsrkf2ous66i6seeyi6etzfgrue332grh2n7madpwopotugyd\.onion', r'(?:www\.)?w6ijuptxiku4xpnnaetxvnkc5vqcdu7mgns2u77qefoixi63vbvnpnqd\.onion', r'(?:www\.)?kbjggqkzv65ivcqj6bumvp337z6264huv5kpkwuv6gu5yjiskvan7fad\.onion', r'(?:www\.)?grwp24hodrefzvjjuccrkw3mjq4tzhaaq32amf33dzpmuxe7ilepcmad\.onion', r'(?:www\.)?hpniueoejy4opn7bc4ftgazyqjoeqwlvh2uiku2xqku6zpoa4bf5ruid\.onion', # piped instances from https://github.com/TeamPiped/Piped/wiki/Instances r'(?:www\.)?piped\.kavin\.rocks', r'(?:www\.)?piped\.silkky\.cloud', r'(?:www\.)?piped\.tokhmi\.xyz', r'(?:www\.)?piped\.moomoo\.me', r'(?:www\.)?il\.ax', r'(?:www\.)?piped\.syncpundit\.com', r'(?:www\.)?piped\.mha\.fi', r'(?:www\.)?piped\.mint\.lgbt', r'(?:www\.)?piped\.privacy\.com\.de', ) def _initialize_consent(self): cookies = self._get_cookies('https://www.youtube.com/') if cookies.get('__Secure-3PSID'): return consent_id = None consent = cookies.get('CONSENT') if consent: if 'YES' in consent.value: return consent_id = self._search_regex( r'PENDING\+(\d+)', consent.value, 'consent', default=None) if not consent_id: consent_id = random.randint(100, 999) self._set_cookie('.youtube.com', 'CONSENT', 'YES+cb.20210328-17-p0.en+FX+%s' % consent_id) def _initialize_pref(self): cookies = self._get_cookies('https://www.youtube.com/') pref_cookie = cookies.get('PREF') pref = {} if pref_cookie: try: pref = dict(compat_urlparse.parse_qsl(pref_cookie.value)) except ValueError: self.report_warning('Failed to parse user PREF cookie' + bug_reports_message()) pref.update({'hl': 'en', 'tz': 'UTC'}) self._set_cookie('.youtube.com', name='PREF', value=compat_urllib_parse_urlencode(pref)) 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): self.raise_login_required('Login details are needed to download this content', method='cookies') _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|= 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] @staticmethod def _extract_visitor_data(*args): """ Extracts visitorData from an API response or ytcfg Appears to be used to track session state """ return get_first( args, [('VISITOR_DATA', ('INNERTUBE_CONTEXT', 'client', 'visitorData'), ('responseContext', 'visitorData'))], expected_type=str) @property def is_authenticated(self): return bool(self._generate_sapisidhash_header()) def extract_ytcfg(self, video_id, webpage): if not webpage: return {} return self._parse_json( self._search_regex( 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'): origin = 'https://' + (api_hostname if api_hostname else self._get_innertube_host(default_client)) headers = { 'X-YouTube-Client-Name': compat_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) } if session_index is None: session_index = self._extract_session_index(ytcfg) 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) if auth is not None: headers['Authorization'] = auth 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 = { 'continuation': continuation } # TODO: Inconsistency with clickTrackingParams. # Currently we have a fixed ctp contained within context (from ytcfg) # and a ctp in root query for continuation. if ctp: query['clickTracking'] = {'clickTrackingParams': ctp} return query @classmethod def _extract_next_continuation_data(cls, renderer): next_continuation = try_get( renderer, (lambda x: x['continuations'][0]['nextContinuationData'], lambda x: x['continuation']['reloadContinuationData']), dict) if not next_continuation: return continuation = next_continuation.get('continuation') if not continuation: return ctp = next_continuation.get('clickTrackingParams') return cls._build_api_continuation_query(continuation, ctp) @classmethod def _extract_continuation_ep_data(cls, continuation_ep: dict): if isinstance(continuation_ep, dict): continuation = try_get( continuation_ep, lambda x: x['continuationCommand']['token'], compat_str) if not continuation: return ctp = continuation_ep.get('clickTrackingParams') return cls._build_api_continuation_query(continuation, ctp) @classmethod def _extract_continuation(cls, renderer): next_continuation = cls._extract_next_continuation_data(renderer) if next_continuation: return next_continuation contents = [] for key in ('contents', 'items'): contents.extend(try_get(renderer, lambda x: x[key], list) or []) for content in contents: if not isinstance(content, dict): continue continuation_ep = try_get( content, (lambda x: x['continuationItemRenderer']['continuationEndpoint'], lambda x: x['continuationItemRenderer']['button']['buttonRenderer']['command']), dict) continuation = cls._extract_continuation_ep_data(continuation_ep) if continuation: return continuation @classmethod def _extract_alerts(cls, data): for alert_dict in try_get(data, lambda x: x['alerts'], list) or []: if not isinstance(alert_dict, dict): continue for alert in alert_dict.values(): alert_type = alert.get('type') if not alert_type: continue message = cls._get_text(alert, 'text') if message: yield alert_type, message def _report_alerts(self, alerts, expected=True, fatal=True, only_once=False): errors = [] warnings = [] for alert_type, alert_message in alerts: if alert_type.lower() == 'error' and fatal: errors.append([alert_type, alert_message]) else: warnings.append([alert_type, alert_message]) for alert_type, alert_message in (warnings + errors[:-1]): self.report_warning(f'YouTube said: {alert_type} - {alert_message}', only_once=only_once) if errors: raise ExtractorError('YouTube said: %s' % errors[-1][1], expected=expected) def _extract_and_report_alerts(self, data, *args, **kwargs): return self._report_alerts(self._extract_alerts(data), *args, **kwargs) def _extract_badges(self, renderer: dict): badges = set() for badge in try_get(renderer, lambda x: x['badges'], list) or []: label = try_get(badge, lambda x: x['metadataBadgeRenderer']['label'], compat_str) if label: badges.add(label.lower()) return badges @staticmethod def _get_text(data, *path_list, max_runs=None): for path in path_list or [None]: if path is None: obj = [data] else: obj = traverse_obj(data, path, default=[]) if not any(key is ... or isinstance(key, (list, tuple)) for key in variadic(path)): obj = [obj] for item in obj: text = try_get(item, lambda x: x['simpleText'], compat_str) if text: return text runs = try_get(item, lambda x: x['runs'], list) or [] if not runs and isinstance(item, list): runs = item runs = runs[:min(len(runs), max_runs or len(runs))] text = ''.join(traverse_obj(runs, (..., 'text'), expected_type=str, default=[])) if text: return text def _get_count(self, data, *path_list): count_text = self._get_text(data, *path_list) or '' count = parse_count(count_text) if count is None: count = str_to_int( self._search_regex(r'^([\d,]+)', re.sub(r'\s', '', count_text), 'count', default=None)) return count @staticmethod def _extract_thumbnails(data, *path_list): """ Extract thumbnails from thumbnails dict @param path_list: path list to level that contains 'thumbnails' key """ thumbnails = [] for path in path_list or [()]: for thumbnail in traverse_obj(data, (*variadic(path), 'thumbnails', ...), default=[]): thumbnail_url = url_or_none(thumbnail.get('url')) if not thumbnail_url: continue # Sometimes youtube gives a wrong thumbnail URL. See: # https://github.com/yt-dlp/yt-dlp/issues/233 # https://github.com/ytdl-org/youtube-dl/issues/28023 if 'maxresdefault' in thumbnail_url: thumbnail_url = thumbnail_url.split('?')[0] thumbnails.append({ 'url': thumbnail_url, 'height': int_or_none(thumbnail.get('height')), 'width': int_or_none(thumbnail.get('width')), }) return thumbnails @staticmethod def extract_relative_time(relative_time_text): """ Extracts a relative time from string and converts to dt object e.g. 'streamed 6 days ago', '5 seconds ago (edited)', 'updated today' """ mobj = re.search(r'(?Ptoday|yesterday|now)|(?P