diff options
Diffstat (limited to 'youtube_dl/extractor')
| -rw-r--r-- | youtube_dl/extractor/bokecc.py | 2 | ||||
| -rw-r--r-- | youtube_dl/extractor/cloudy.py | 2 | ||||
| -rw-r--r-- | youtube_dl/extractor/common.py | 156 | ||||
| -rw-r--r-- | youtube_dl/extractor/extractors.py | 18 | ||||
| -rw-r--r-- | youtube_dl/extractor/itv.py | 17 | ||||
| -rw-r--r-- | youtube_dl/extractor/mixcloud.py | 30 | ||||
| -rw-r--r-- | youtube_dl/extractor/orf.py | 1103 | ||||
| -rw-r--r-- | youtube_dl/extractor/palcomp3.py | 9 | ||||
| -rw-r--r-- | youtube_dl/extractor/senateisvp.py | 2 | ||||
| -rw-r--r-- | youtube_dl/extractor/vidlii.py | 59 | ||||
| -rw-r--r-- | youtube_dl/extractor/yandexmusic.py | 23 | ||||
| -rw-r--r-- | youtube_dl/extractor/youtube.py | 803 |
12 files changed, 1464 insertions, 760 deletions
diff --git a/youtube_dl/extractor/bokecc.py b/youtube_dl/extractor/bokecc.py index 6017e8344..4b8bef391 100644 --- a/youtube_dl/extractor/bokecc.py +++ b/youtube_dl/extractor/bokecc.py @@ -32,7 +32,7 @@ class BokeCCBaseIE(InfoExtractor): class BokeCCIE(BokeCCBaseIE): - _IE_DESC = 'CC视频' + IE_DESC = 'CC视频' _VALID_URL = r'https?://union\.bokecc\.com/playvideo\.bo\?(?P<query>.*)' _TESTS = [{ diff --git a/youtube_dl/extractor/cloudy.py b/youtube_dl/extractor/cloudy.py index 85ca20ecc..d39a9a5c2 100644 --- a/youtube_dl/extractor/cloudy.py +++ b/youtube_dl/extractor/cloudy.py @@ -9,7 +9,7 @@ from ..utils import ( class CloudyIE(InfoExtractor): - _IE_DESC = 'cloudy.ec' + IE_DESC = 'cloudy.ec' _VALID_URL = r'https?://(?:www\.)?cloudy\.ec/(?:v/|embed\.php\?.*?\bid=)(?P<id>[A-Za-z0-9]+)' _TESTS = [{ 'url': 'https://www.cloudy.ec/v/af511e2527aac', diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py index 7fae9e57b..a64fcfccc 100644 --- a/youtube_dl/extractor/common.py +++ b/youtube_dl/extractor/common.py @@ -422,6 +422,8 @@ class InfoExtractor(object): _GEO_COUNTRIES = None _GEO_IP_BLOCKS = None _WORKING = True + # supply this in public subclasses: used in supported sites list, etc + # IE_DESC = 'short description of IE' def __init__(self, downloader=None): """Constructor. Receives an optional downloader.""" @@ -503,7 +505,7 @@ class InfoExtractor(object): if not self._x_forwarded_for_ip: # Geo bypass mechanism is explicitly disabled by user - if not self._downloader.params.get('geo_bypass', True): + if not self.get_param('geo_bypass', True): return if not geo_bypass_context: @@ -525,7 +527,7 @@ class InfoExtractor(object): # Explicit IP block specified by user, use it right away # regardless of whether extractor is geo bypassable or not - ip_block = self._downloader.params.get('geo_bypass_ip_block', None) + ip_block = self.get_param('geo_bypass_ip_block', None) # Otherwise use random IP block from geo bypass context but only # if extractor is known as geo bypassable @@ -536,8 +538,8 @@ class InfoExtractor(object): if ip_block: self._x_forwarded_for_ip = GeoUtils.random_ipv4(ip_block) - if self._downloader.params.get('verbose', False): - self._downloader.to_screen( + if self.get_param('verbose', False): + self.to_screen( '[debug] Using fake IP %s as X-Forwarded-For.' % self._x_forwarded_for_ip) return @@ -546,7 +548,7 @@ class InfoExtractor(object): # Explicit country code specified by user, use it right away # regardless of whether extractor is geo bypassable or not - country = self._downloader.params.get('geo_bypass_country', None) + country = self.get_param('geo_bypass_country', None) # Otherwise use random country code from geo bypass context but # only if extractor is known as geo bypassable @@ -557,8 +559,8 @@ class InfoExtractor(object): if country: self._x_forwarded_for_ip = GeoUtils.random_ipv4(country) - if self._downloader.params.get('verbose', False): - self._downloader.to_screen( + if self.get_param('verbose', False): + self.to_screen( '[debug] Using fake IP %s (%s) as X-Forwarded-For.' % (self._x_forwarded_for_ip, country.upper())) @@ -584,9 +586,9 @@ class InfoExtractor(object): raise ExtractorError('An extractor error has occurred.', cause=e) def __maybe_fake_ip_and_retry(self, countries): - if (not self._downloader.params.get('geo_bypass_country', None) + if (not self.get_param('geo_bypass_country', None) and self._GEO_BYPASS - and self._downloader.params.get('geo_bypass', True) + and self.get_param('geo_bypass', True) and not self._x_forwarded_for_ip and countries): country_code = random.choice(countries) @@ -696,7 +698,7 @@ class InfoExtractor(object): if fatal: raise ExtractorError(errmsg, sys.exc_info()[2], cause=err) else: - self._downloader.report_warning(errmsg) + self.report_warning(errmsg) return False def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None): @@ -768,11 +770,11 @@ class InfoExtractor(object): webpage_bytes = prefix + webpage_bytes if not encoding: encoding = self._guess_encoding_from_content(content_type, webpage_bytes) - if self._downloader.params.get('dump_intermediate_pages', False): + if self.get_param('dump_intermediate_pages', False): self.to_screen('Dumping request to ' + urlh.geturl()) dump = base64.b64encode(webpage_bytes).decode('ascii') - self._downloader.to_screen(dump) - if self._downloader.params.get('write_pages', False): + self.to_screen(dump) + if self.get_param('write_pages', False): basen = '%s_%s' % (video_id, urlh.geturl()) if len(basen) > 240: h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest() @@ -974,19 +976,9 @@ class InfoExtractor(object): """Print msg to screen, prefixing it with '[ie_name]'""" self._downloader.to_screen(self.__ie_msg(msg)) - def write_debug(self, msg, only_once=False, _cache=[]): + def write_debug(self, msg, only_once=False): '''Log debug message or Print message to stderr''' - if not self.get_param('verbose', False): - return - message = '[debug] ' + self.__ie_msg(msg) - logger = self.get_param('logger') - if logger: - logger.debug(message) - else: - if only_once and hash(message) in _cache: - return - self._downloader.to_stderr(message) - _cache.append(hash(message)) + self._downloader.write_debug(self.__ie_msg(msg), only_once=only_once) # name, default=None, *args, **kwargs def get_param(self, name, *args, **kwargs): @@ -1082,7 +1074,7 @@ class InfoExtractor(object): if mobj: break - if not self._downloader.params.get('no_color') and compat_os_name != 'nt' and sys.stderr.isatty(): + if not self.get_param('no_color') and compat_os_name != 'nt' and sys.stderr.isatty(): _name = '\033[0;34m%s\033[0m' % name else: _name = name @@ -1100,7 +1092,7 @@ class InfoExtractor(object): elif fatal: raise RegexNotFoundError('Unable to extract %s' % _name) else: - self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message()) + self.report_warning('unable to extract %s' % _name + bug_reports_message()) return None def _search_json(self, start_pattern, string, name, video_id, **kwargs): @@ -1169,10 +1161,10 @@ class InfoExtractor(object): def _get_netrc_login_info(self, netrc_machine=None): username = None password = None - netrc_machine = netrc_machine or self._NETRC_MACHINE - if self._downloader.params.get('usenetrc', False): + if self.get_param('usenetrc', False): try: + netrc_machine = netrc_machine or self._NETRC_MACHINE info = netrc.netrc().authenticators(netrc_machine) if info is not None: username = info[0] @@ -1180,8 +1172,8 @@ class InfoExtractor(object): else: raise netrc.NetrcParseError( 'No authenticators for %s' % netrc_machine) - except (IOError, netrc.NetrcParseError) as err: - self._downloader.report_warning( + except (AttributeError, IOError, netrc.NetrcParseError) as err: + self.report_warning( 'parsing .netrc: %s' % error_to_compat_str(err)) return username, password @@ -1218,10 +1210,10 @@ class InfoExtractor(object): """ if self._downloader is None: return None - downloader_params = self._downloader.params - if downloader_params.get('twofactor') is not None: - return downloader_params['twofactor'] + twofactor = self.get_param('twofactor') + if twofactor is not None: + return twofactor return compat_getpass('Type %s and press [Return]: ' % note) @@ -1356,7 +1348,7 @@ class InfoExtractor(object): elif fatal: raise RegexNotFoundError('Unable to extract JSON-LD') else: - self._downloader.report_warning('unable to extract JSON-LD %s' % bug_reports_message()) + self.report_warning('unable to extract JSON-LD %s' % bug_reports_message()) return {} def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None): @@ -1490,14 +1482,18 @@ class InfoExtractor(object): return dict((k, v) for k, v in info.items() if v is not None) def _search_nextjs_data(self, webpage, video_id, **kw): - nkw = dict((k, v) for k, v in kw.items() if k in ('transform_source', 'fatal')) - kw.pop('transform_source', None) - next_data = self._search_regex( - r'''<script[^>]+\bid\s*=\s*('|")__NEXT_DATA__\1[^>]*>(?P<nd>[^<]+)</script>''', - webpage, 'next.js data', group='nd', **kw) - if not next_data: - return {} - return self._parse_json(next_data, video_id, **nkw) + # ..., *, transform_source=None, fatal=True, default=NO_DEFAULT + + # TODO: remove this backward compat + default = kw.get('default', NO_DEFAULT) + if default == '{}': + kw['default'] = {} + kw = compat_kwargs(kw) + + return self._search_json( + r'''<script\s[^>]*?\bid\s*=\s*('|")__NEXT_DATA__\1[^>]*>''', + webpage, 'next.js data', video_id, end_pattern='</script>', + **kw) def _search_nuxt_data(self, webpage, video_id, *args, **kwargs): """Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function""" @@ -1583,7 +1579,7 @@ class InfoExtractor(object): if f.get('vcodec') == 'none': # audio only preference -= 50 - if self._downloader.params.get('prefer_free_formats'): + if self.get_param('prefer_free_formats'): ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus'] else: ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a'] @@ -1595,7 +1591,7 @@ class InfoExtractor(object): else: if f.get('acodec') == 'none': # video only preference -= 40 - if self._downloader.params.get('prefer_free_formats'): + if self.get_param('prefer_free_formats'): ORDER = ['flv', 'mp4', 'webm'] else: ORDER = ['webm', 'flv', 'mp4'] @@ -1661,7 +1657,7 @@ class InfoExtractor(object): """ Either "http:" or "https:", depending on the user's preferences """ return ( 'http:' - if self._downloader.params.get('prefer_insecure', False) + if self.get_param('prefer_insecure', False) else 'https:') def _proto_relative_url(self, url, scheme=None): @@ -3029,7 +3025,6 @@ class InfoExtractor(object): transform_source=transform_source, default=None) def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs): - # allow passing `transform_source` through to _find_jwplayer_data() transform_source = kwargs.pop('transform_source', None) kwfind = compat_kwargs({'transform_source': transform_source}) if transform_source else {} @@ -3167,7 +3162,7 @@ class InfoExtractor(object): # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as # of jwplayer.flash.swf rtmp_url_parts = re.split( - r'((?:mp4|mp3|flv):)', source_url, 1) + r'((?:mp4|mp3|flv):)', source_url, maxsplit=1) if len(rtmp_url_parts) == 3: rtmp_url, prefix, play_path = rtmp_url_parts a_format.update({ @@ -3194,7 +3189,7 @@ class InfoExtractor(object): if fatal: raise ExtractorError(msg) else: - self._downloader.report_warning(msg) + self.report_warning(msg) return res def _float(self, v, name, fatal=False, **kwargs): @@ -3204,7 +3199,7 @@ class InfoExtractor(object): if fatal: raise ExtractorError(msg) else: - self._downloader.report_warning(msg) + self.report_warning(msg) return res def _set_cookie(self, domain, name, value, expire_time=None, port=None, @@ -3213,12 +3208,12 @@ class InfoExtractor(object): 0, name, value, port, port is not None, domain, True, domain.startswith('.'), path, True, secure, expire_time, discard, None, None, rest) - self._downloader.cookiejar.set_cookie(cookie) + self.cookiejar.set_cookie(cookie) def _get_cookies(self, url): """ Return a compat_cookies_SimpleCookie with the cookies for the url """ req = sanitized_Request(url) - self._downloader.cookiejar.add_cookie_header(req) + self.cookiejar.add_cookie_header(req) return compat_cookies_SimpleCookie(req.get_header('Cookie')) def _apply_first_set_cookie_header(self, url_handle, cookie): @@ -3278,8 +3273,8 @@ class InfoExtractor(object): return not any_restricted def extract_subtitles(self, *args, **kwargs): - if (self._downloader.params.get('writesubtitles', False) - or self._downloader.params.get('listsubtitles')): + if (self.get_param('writesubtitles', False) + or self.get_param('listsubtitles')): return self._get_subtitles(*args, **kwargs) return {} @@ -3296,16 +3291,24 @@ class InfoExtractor(object): return ret @classmethod - def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2): - """ Merge two subtitle dictionaries, language by language. """ - ret = dict(subtitle_dict1) - for lang in subtitle_dict2: - ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang]) - return ret + def _merge_subtitles(cls, subtitle_dict1, *subtitle_dicts, **kwargs): + """ Merge subtitle dictionaries, language by language. """ + + # ..., * , target=None + target = kwargs.get('target') + if target is None: + target = dict(subtitle_dict1) + else: + subtitle_dicts = (subtitle_dict1,) + subtitle_dicts + + for subtitle_dict in subtitle_dicts: + for lang in subtitle_dict: + target[lang] = cls._merge_subtitle_items(target.get(lang, []), subtitle_dict[lang]) + return target def extract_automatic_captions(self, *args, **kwargs): - if (self._downloader.params.get('writeautomaticsub', False) - or self._downloader.params.get('listsubtitles')): + if (self.get_param('writeautomaticsub', False) + or self.get_param('listsubtitles')): return self._get_automatic_captions(*args, **kwargs) return {} @@ -3313,9 +3316,9 @@ class InfoExtractor(object): raise NotImplementedError('This method must be implemented by subclasses') def mark_watched(self, *args, **kwargs): - if (self._downloader.params.get('mark_watched', False) + if (self.get_param('mark_watched', False) and (self._get_login_info()[0] is not None - or self._downloader.params.get('cookiefile') is not None)): + or self.get_param('cookiefile') is not None)): self._mark_watched(*args, **kwargs) def _mark_watched(self, *args, **kwargs): @@ -3323,7 +3326,7 @@ class InfoExtractor(object): def geo_verification_headers(self): headers = {} - geo_verification_proxy = self._downloader.params.get('geo_verification_proxy') + geo_verification_proxy = self.get_param('geo_verification_proxy') if geo_verification_proxy: headers['Ytdl-request-proxy'] = geo_verification_proxy return headers @@ -3334,6 +3337,29 @@ class InfoExtractor(object): def _generic_title(self, url): return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0]) + def _yes_playlist(self, playlist_id, video_id, *args, **kwargs): + # smuggled_data=None, *, playlist_label='playlist', video_label='video' + smuggled_data = args[0] if len(args) == 1 else kwargs.get('smuggled_data') + playlist_label = kwargs.get('playlist_label', 'playlist') + video_label = kwargs.get('video_label', 'video') + + if not playlist_id or not video_id: + return not video_id + + no_playlist = (smuggled_data or {}).get('force_noplaylist') + if no_playlist is not None: + return not no_playlist + + video_id = '' if video_id is True else ' ' + video_id + noplaylist = self.get_param('noplaylist') + self.to_screen( + 'Downloading just the {0}{1} because of --no-playlist'.format(video_label, video_id) + if noplaylist else + 'Downloading {0}{1} - add --no-playlist to download just the {2}{3}'.format( + playlist_label, '' if playlist_id is True else ' ' + playlist_id, + video_label, video_id)) + return not noplaylist + class SearchInfoExtractor(InfoExtractor): """ diff --git a/youtube_dl/extractor/extractors.py b/youtube_dl/extractor/extractors.py index 03d035a27..3da5f8020 100644 --- a/youtube_dl/extractor/extractors.py +++ b/youtube_dl/extractor/extractors.py @@ -898,21 +898,13 @@ from .ooyala import ( ) from .ora import OraTVIE from .orf import ( - ORFTVthekIE, - ORFFM4IE, + ORFONIE, + ORFONLiveIE, ORFFM4StoryIE, - ORFOE1IE, - ORFOE3IE, - ORFNOEIE, - ORFWIEIE, - ORFBGLIE, - ORFOOEIE, - ORFSTMIE, - ORFKTNIE, - ORFSBGIE, - ORFTIRIE, - ORFVBGIE, ORFIPTVIE, + ORFPodcastIE, + ORFRadioIE, + ORFRadioCollectionIE, ) from .outsidetv import OutsideTVIE from .packtpub import ( diff --git a/youtube_dl/extractor/itv.py b/youtube_dl/extractor/itv.py index c64af3be6..2510ad887 100644 --- a/youtube_dl/extractor/itv.py +++ b/youtube_dl/extractor/itv.py @@ -35,15 +35,6 @@ from ..utils import ( class ITVBaseIE(InfoExtractor): - def _search_nextjs_data(self, webpage, video_id, **kw): - transform_source = kw.pop('transform_source', None) - fatal = kw.pop('fatal', True) - return self._parse_json( - self._search_regex( - r'''<script\b[^>]+\bid=('|")__NEXT_DATA__\1[^>]*>(?P<js>[^<]+)</script>''', - webpage, 'next.js data', group='js', fatal=fatal, **kw), - video_id, transform_source=transform_source, fatal=fatal) - def __handle_request_webpage_error(self, err, video_id=None, errnote=None, fatal=True): if errnote is False: return False @@ -109,7 +100,9 @@ class ITVBaseIE(InfoExtractor): class ITVIE(ITVBaseIE): _VALID_URL = r'https?://(?:www\.)?itv\.com/(?:(?P<w>watch)|hub)/[^/]+/(?(w)[\w-]+/)(?P<id>\w+)' - _IE_DESC = 'ITVX' + IE_DESC = 'ITVX' + _WORKING = False + _TESTS = [{ 'note': 'Hub URLs redirect to ITVX', 'url': 'https://www.itv.com/hub/liar/2a4547a0012', @@ -270,7 +263,7 @@ class ITVIE(ITVBaseIE): 'ext': determine_ext(href, 'vtt'), }) - next_data = self._search_nextjs_data(webpage, video_id, fatal=False, default='{}') + next_data = self._search_nextjs_data(webpage, video_id, fatal=False, default={}) video_data.update(traverse_obj(next_data, ('props', 'pageProps', ('title', 'episode')), expected_type=dict)[0] or {}) title = traverse_obj(video_data, 'headerTitle', 'episodeTitle') info = self._og_extract(webpage, require_title=not title) @@ -323,7 +316,7 @@ class ITVIE(ITVBaseIE): class ITVBTCCIE(ITVBaseIE): _VALID_URL = r'https?://(?:www\.)?itv\.com/(?!(?:watch|hub)/)(?:[^/]+/)+(?P<id>[^/?#&]+)' - _IE_DESC = 'ITV articles: News, British Touring Car Championship' + IE_DESC = 'ITV articles: News, British Touring Car Championship' _TESTS = [{ 'note': 'British Touring Car Championship', 'url': 'https://www.itv.com/btcc/articles/btcc-2018-all-the-action-from-brands-hatch', diff --git a/youtube_dl/extractor/mixcloud.py b/youtube_dl/extractor/mixcloud.py index 69319857d..2b5e2c15c 100644 --- a/youtube_dl/extractor/mixcloud.py +++ b/youtube_dl/extractor/mixcloud.py @@ -1,3 +1,4 @@ +# coding: utf-8 from __future__ import unicode_literals import itertools @@ -10,7 +11,7 @@ from ..compat import ( compat_ord, compat_str, compat_urllib_parse_unquote, - compat_zip + compat_zip as zip, ) from ..utils import ( int_or_none, @@ -24,7 +25,7 @@ class MixcloudBaseIE(InfoExtractor): def _call_api(self, object_type, object_fields, display_id, username, slug=None): lookup_key = object_type + 'Lookup' return self._download_json( - 'https://www.mixcloud.com/graphql', display_id, query={ + 'https://app.mixcloud.com/graphql', display_id, query={ 'query': '''{ %s(lookup: {username: "%s"%s}) { %s @@ -44,7 +45,7 @@ class MixcloudIE(MixcloudBaseIE): 'ext': 'm4a', 'title': 'Cryptkeeper', 'description': 'After quite a long silence from myself, finally another Drum\'n\'Bass mix with my favourite current dance floor bangers.', - 'uploader': 'Daniel Holbach', + 'uploader': 'dholbach', # was: 'Daniel Holbach', 'uploader_id': 'dholbach', 'thumbnail': r're:https?://.*\.jpg', 'view_count': int, @@ -57,7 +58,7 @@ class MixcloudIE(MixcloudBaseIE): 'id': 'gillespeterson_caribou-7-inch-vinyl-mix-chat', 'ext': 'mp3', 'title': 'Caribou 7 inch Vinyl Mix & Chat', - 'description': 'md5:2b8aec6adce69f9d41724647c65875e8', + 'description': r're:Last week Dan Snaith aka Caribou swung by the Brownswood.{136}', 'uploader': 'Gilles Peterson Worldwide', 'uploader_id': 'gillespeterson', 'thumbnail': 're:https?://.*', @@ -65,6 +66,23 @@ class MixcloudIE(MixcloudBaseIE): 'timestamp': 1422987057, 'upload_date': '20150203', }, + 'params': { + 'skip_download': '404 not found', + }, + }, { + 'url': 'https://www.mixcloud.com/gillespeterson/carnival-m%C3%BAsica-popular-brasileira-mix/', + 'info_dict': { + 'id': 'gillespeterson_carnival-música-popular-brasileira-mix', + 'ext': 'm4a', + 'title': 'Carnival Música Popular Brasileira Mix', + 'description': r're:Gilles was recently in Brazil to play at Boiler Room.{208}', + 'timestamp': 1454347174, + 'upload_date': '20160201', + 'uploader': 'Gilles Peterson Worldwide', + 'uploader_id': 'gillespeterson', + 'thumbnail': 're:https?://.*', + 'view_count': int, + }, }, { 'url': 'https://beta.mixcloud.com/RedLightRadio/nosedrip-15-red-light-radio-01-18-2016/', 'only_matching': True, @@ -76,10 +94,10 @@ class MixcloudIE(MixcloudBaseIE): """Encrypt/Decrypt XOR cipher. Both ways are possible because it's XOR.""" return ''.join([ compat_chr(compat_ord(ch) ^ compat_ord(k)) - for ch, k in compat_zip(ciphertext, itertools.cycle(key))]) + for ch, k in zip(ciphertext, itertools.cycle(key))]) def _real_extract(self, url): - username, slug = re.match(self._VALID_URL, url).groups() + username, slug = self._match_valid_url(url).groups() username, slug = compat_urllib_parse_unquote(username), compat_urllib_parse_unquote(slug) track_id = '%s_%s' % (username, slug) diff --git a/youtube_dl/extractor/orf.py b/youtube_dl/extractor/orf.py index 8d537d7ae..1ee78edbc 100644 --- a/youtube_dl/extractor/orf.py +++ b/youtube_dl/extractor/orf.py @@ -1,407 +1,394 @@ # coding: utf-8 from __future__ import unicode_literals +import base64 +import functools import re from .common import InfoExtractor -from ..compat import compat_str +from .youtube import YoutubeIE from ..utils import ( clean_html, determine_ext, + ExtractorError, float_or_none, - HEADRequest, int_or_none, - orderedSet, - remove_end, - str_or_none, + merge_dicts, + mimetype2ext, + parse_age_limit, + parse_iso8601, strip_jsonp, - unescapeHTML, + txt_or_none, unified_strdate, + update_url_query, url_or_none, ) - - -class ORFTVthekIE(InfoExtractor): - IE_NAME = 'orf:tvthek' - IE_DESC = 'ORF TVthek' - _VALID_URL = r'https?://tvthek\.orf\.at/(?:[^/]+/)+(?P<id>\d+)' +from ..traversal import T, traverse_obj + +k_float_or_none = functools.partial(float_or_none, scale=1000) + + +class ORFRadioBase(InfoExtractor): + STATION_INFO = { + 'fm4': ('fm4', 'fm4', 'orffm4'), + 'noe': ('noe', 'oe2n', 'orfnoe'), + 'wien': ('wie', 'oe2w', 'orfwie'), + 'burgenland': ('bgl', 'oe2b', 'orfbgl'), + 'ooe': ('ooe', 'oe2o', 'orfooe'), + 'steiermark': ('stm', 'oe2st', 'orfstm'), + 'kaernten': ('ktn', 'oe2k', 'orfktn'), + 'salzburg': ('sbg', 'oe2s', 'orfsbg'), + 'tirol': ('tir', 'oe2t', 'orftir'), + 'vorarlberg': ('vbg', 'oe2v', 'orfvbg'), + 'oe3': ('oe3', 'oe3', 'orfoe3'), + 'oe1': ('oe1', 'oe1', 'orfoe1'), + } + _ID_NAMES = ('id', 'guid', 'program') + + @classmethod + def _get_item_id(cls, data): + return traverse_obj(data, *cls._ID_NAMES, expected_type=txt_or_none) + + @classmethod + def _get_api_payload(cls, data, expected_id, in_payload=False): + if expected_id not in traverse_obj(data, ('payload',)[:1 if in_payload else 0] + (cls._ID_NAMES, T(txt_or_none))): + raise ExtractorError('Unexpected API data result', video_id=expected_id) + return data['payload'] + + @staticmethod + def _extract_podcast_upload(data): + return traverse_obj(data, { + 'url': ('enclosures', 0, 'url'), + 'ext': ('enclosures', 0, 'type', T(mimetype2ext)), + 'filesize': ('enclosures', 0, 'length', T(int_or_none)), + 'title': ('title', T(txt_or_none)), + 'description': ('description', T(clean_html)), + 'timestamp': (('published', 'postDate'), T(parse_iso8601)), + 'duration': ('duration', T(k_float_or_none)), + 'series': ('podcast', 'title'), + 'uploader': ((('podcast', 'author'), 'station'), T(txt_or_none)), + 'uploader_id': ('podcast', 'channel', T(txt_or_none)), + }, get_all=False) + + @classmethod + def _entries(cls, data, station, item_type=None): + if item_type in ('upload', 'podcast-episode'): + yield merge_dicts({ + 'id': cls._get_item_id(data), + 'ext': 'mp3', + 'vcodec': 'none', + }, cls._extract_podcast_upload(data), rev=True) + return + + loop_station = cls.STATION_INFO[station][1] + for info in traverse_obj(data, ((('streams', Ellipsis), 'stream'), T(lambda v: v if v['loopStreamId'] else None))): + item_id = info['loopStreamId'] + host = info.get('host') or 'loopstream01.apa.at' + yield merge_dicts({ + 'id': item_id.replace('.mp3', ''), + 'ext': 'mp3', + 'url': update_url_query('https://{0}/'.format(host), { + 'channel': loop_station, + 'id': item_id, + }), + 'vcodec': 'none', + # '_old_archive_ids': [make_archive_id(old_ie, video_id)], + }, traverse_obj(data, { + 'title': ('title', T(txt_or_none)), + 'description': ('subtitle', T(clean_html)), + 'uploader': 'station', + 'series': ('programTitle', T(txt_or_none)), + }), traverse_obj(info, { + 'duration': (('duration', + (None, T(lambda x: x['end'] - x['start']))), + T(k_float_or_none), any), + 'timestamp': (('start', 'startISO'), T(parse_iso8601), any), + })) + + +class ORFRadioIE(ORFRadioBase): + IE_NAME = 'orf:sound' + _STATION_RE = '|'.join(map(re.escape, ORFRadioBase.STATION_INFO.keys())) + + _VALID_URL = ( + r'https?://sound\.orf\.at/radio/(?P<station>{0})/sendung/(?P<id>\d+)(?:/(?P<show>\w+))?'.format(_STATION_RE), + r'https?://(?P<station>{0})\.orf\.at/player/(?P<date>\d{{8}})/(?P<id>\d+)'.format(_STATION_RE), + ) _TESTS = [{ - 'url': 'http://tvthek.orf.at/program/Aufgetischt/2745173/Aufgetischt-Mit-der-Steirischen-Tafelrunde/8891389', + 'url': 'https://sound.orf.at/radio/ooe/sendung/37802/guten-morgen-oberoesterreich-am-feiertag', + 'info_dict': { + 'id': '37802', + 'title': 'Guten Morgen Oberösterreich am Feiertag', + 'description': 'Oberösterreichs meistgehörte regionale Frühsendung.\nRegionale Nachrichten zu jeder halben Stunde.\nModeration: Wolfgang Lehner\nNachrichten: Stephan Schnabl', + }, 'playlist': [{ - 'md5': '2942210346ed779588f428a92db88712', + 'md5': 'f9ff8517dd681b642a2c900e2c9e6085', 'info_dict': { - 'id': '8896777', - 'ext': 'mp4', - 'title': 'Aufgetischt: Mit der Steirischen Tafelrunde', - 'description': 'md5:c1272f0245537812d4e36419c207b67d', - 'duration': 2668, - 'upload_date': '20141208', - }, + 'id': '2024-05-30_0559_tl_66_7DaysThu1_443862', + 'ext': 'mp3', + 'title': 'Guten Morgen Oberösterreich am Feiertag', + 'description': 'Oberösterreichs meistgehörte regionale Frühsendung.\nRegionale Nachrichten zu jeder halben Stunde.\nModeration: Wolfgang Lehner\nNachrichten: Stephan Schnabl', + 'timestamp': 1717041587, + 'upload_date': '20240530', + 'uploader': 'ooe', + 'duration': 14413.0, + } }], - 'skip': 'Blocked outside of Austria / Germany', + 'skip': 'Shows from ORF Sound are only available for 30 days.' }, { - 'url': 'http://tvthek.orf.at/topic/Im-Wandel-der-Zeit/8002126/Best-of-Ingrid-Thurnher/7982256', + 'url': 'https://oe1.orf.at/player/20240531/758136', + 'md5': '2397717aaf3ae9c22a4f090ee3b8d374', 'info_dict': { - 'id': '7982259', - 'ext': 'mp4', - 'title': 'Best of Ingrid Thurnher', - 'upload_date': '20140527', - 'description': 'Viele Jahre war Ingrid Thurnher das "Gesicht" der ZIB 2. Vor ihrem Wechsel zur ZIB 2 im Jahr 1995 moderierte sie unter anderem "Land und Leute", "Österreich-Bild" und "Niederösterreich heute".', - }, - 'params': { - 'skip_download': True, # rtsp downloads + 'id': '2024-05-31_1905_tl_51_7DaysFri35_2413387', + 'ext': 'mp3', + 'title': '"Who Cares?"', + 'description': 'Europas größte Netzkonferenz re:publica 2024', + 'timestamp': 1717175100, + 'upload_date': '20240531', + 'uploader': 'oe1', + 'duration': 1500, }, - 'skip': 'Blocked outside of Austria / Germany', - }, { - 'url': 'http://tvthek.orf.at/topic/Fluechtlingskrise/10463081/Heimat-Fremde-Heimat/13879132/Senioren-betreuen-Migrantenkinder/13879141', - 'only_matching': True, - }, { - 'url': 'http://tvthek.orf.at/profile/Universum/35429', - 'only_matching': True, + 'skip': 'Shows from ORF Sound are only available for 30 days.' }] def _real_extract(self, url): - playlist_id = self._match_id(url) - webpage = self._download_webpage(url, playlist_id) + m = self._match_valid_url(url) + station, show_id = m.group('station', 'id') + api_station, _, _ = self.STATION_INFO[station] + if 'date' in m.groupdict(): + data = self._download_json( + 'https://audioapi.orf.at/{0}/json/4.0/broadcast/{1}/{2}?_o={3}.orf.at'.format( + api_station, show_id, m.group('date'), station), show_id) + show_id = data['id'] + else: + data = self._download_json( + 'https://audioapi.orf.at/{0}/api/json/5.0/broadcast/{1}?_o=sound.orf.at'.format( + api_station, show_id), show_id) - data_jsb = self._parse_json( - self._search_regex( - r'<div[^>]+class=(["\']).*?VideoPlaylist.*?\1[^>]+data-jsb=(["\'])(?P<json>.+?)\2', - webpage, 'playlist', group='json'), - playlist_id, transform_source=unescapeHTML)['playlist']['videos'] + data = self._get_api_payload(data, show_id, in_payload=True) - entries = [] - for sd in data_jsb: - video_id, title = sd.get('id'), sd.get('title') - if not video_id or not title: - continue - video_id = compat_str(video_id) - formats = [] - for fd in sd['sources']: - src = url_or_none(fd.get('src')) - if not src: - continue - format_id_list = [] - for key in ('delivery', 'quality', 'quality_string'): - value = fd.get(key) - if value: - format_id_list.append(value) - format_id = '-'.join(format_id_list) - ext = determine_ext(src) - if ext == 'm3u8': - m3u8_formats = self._extract_m3u8_formats( - src, video_id, 'mp4', m3u8_id=format_id, fatal=False) - if any('/geoprotection' in f['url'] for f in m3u8_formats): - self.raise_geo_restricted() - formats.extend(m3u8_formats) - elif ext == 'f4m': - formats.extend(self._extract_f4m_formats( - src, video_id, f4m_id=format_id, fatal=False)) - elif ext == 'mpd': - formats.extend(self._extract_mpd_formats( - src, video_id, mpd_id=format_id, fatal=False)) - else: - formats.append({ - 'format_id': format_id, - 'url': src, - 'protocol': fd.get('protocol'), - }) - - # Check for geoblocking. - # There is a property is_geoprotection, but that's always false - geo_str = sd.get('geoprotection_string') - if geo_str: - try: - http_url = next( - f['url'] - for f in formats - if re.match(r'^https?://.*\.mp4$', f['url'])) - except StopIteration: - pass - else: - req = HEADRequest(http_url) - self._request_webpage( - req, video_id, - note='Testing for geoblocking', - errnote=(( - 'This video seems to be blocked outside of %s. ' - 'You may want to try the streaming-* formats.') - % geo_str), - fatal=False) - - self._check_formats(formats, video_id) - self._sort_formats(formats) + # site sends ISO8601 GMT date-times with separate TZ offset, ignored + # TODO: should `..._date` be calculated relative to TZ? - subtitles = {} - for sub in sd.get('subtitles', []): - sub_src = sub.get('src') - if not sub_src: - continue - subtitles.setdefault(sub.get('lang', 'de-AT'), []).append({ - 'url': sub_src, - }) - - upload_date = unified_strdate(sd.get('created_date')) + return merge_dicts( + {'_type': 'multi_video'}, + self.playlist_result( + self._entries(data, station), show_id, + txt_or_none(data.get('title')), + clean_html(data.get('subtitle')))) - thumbnails = [] - preview = sd.get('preview_image_url') - if preview: - thumbnails.append({ - 'id': 'preview', - 'url': preview, - 'preference': 0, - }) - image = sd.get('image_full_url') - if not image and len(data_jsb) == 1: - image = self._og_search_thumbnail(webpage) - if image: - thumbnails.append({ - 'id': 'full', - 'url': image, - 'preference': 1, - }) - entries.append({ - '_type': 'video', - 'id': video_id, - 'title': title, - 'formats': formats, - 'subtitles': subtitles, - 'description': sd.get('description'), - 'duration': int_or_none(sd.get('duration_in_seconds')), - 'upload_date': upload_date, - 'thumbnails': thumbnails, - }) - - return { - '_type': 'playlist', - 'entries': entries, - 'id': playlist_id, - } - - -class ORFRadioIE(InfoExtractor): - def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - show_date = mobj.group('date') - show_id = mobj.group('show') +class ORFRadioCollectionIE(ORFRadioBase): + IE_NAME = 'orf:collection' + _VALID_URL = r'https?://sound\.orf\.at/collection/(?P<coll_id>\d+)(?:/(?P<item_id>\d+))?' - data = self._download_json( - 'http://audioapi.orf.at/%s/api/json/current/broadcast/%s/%s' - % (self._API_STATION, show_id, show_date), show_id) - - entries = [] - for info in data['streams']: - loop_stream_id = str_or_none(info.get('loopStreamId')) - if not loop_stream_id: - continue - title = str_or_none(data.get('title')) - if not title: - continue - start = int_or_none(info.get('start'), scale=1000) - end = int_or_none(info.get('end'), scale=1000) - duration = end - start if end and start else None - entries.append({ - 'id': loop_stream_id.replace('.mp3', ''), - 'url': 'https://loopstream01.apa.at/?channel=%s&id=%s' % (self._LOOP_STATION, loop_stream_id), - 'title': title, - 'description': clean_html(data.get('subtitle')), - 'duration': duration, - 'timestamp': start, + _TESTS = [{ + 'url': 'https://sound.orf.at/collection/4/61908/was-das-uberschreiten-des-15-limits-bedeutet', + 'info_dict': { + 'id': '2577582', + }, + 'playlist': [{ + 'md5': '5789cec7d75575ff58d19c0428c80eb3', + 'info_dict': { + 'id': '2024-06-06_1659_tl_54_7DaysThu6_153926', 'ext': 'mp3', - 'series': data.get('programTitle'), - }) - - return { - '_type': 'playlist', - 'id': show_id, - 'title': data.get('title'), - 'description': clean_html(data.get('subtitle')), - 'entries': entries, - } - - -class ORFFM4IE(ORFRadioIE): - IE_NAME = 'orf:fm4' - IE_DESC = 'radio FM4' - _VALID_URL = r'https?://(?P<station>fm4)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>4\w+)' - _API_STATION = 'fm4' - _LOOP_STATION = 'fm4' - - _TEST = { - 'url': 'http://fm4.orf.at/player/20170107/4CC', - 'md5': '2b0be47375432a7ef104453432a19212', + 'title': 'Klimakrise: Was das Überschreiten des 1,5°-Limits bedeutet', + 'timestamp': 1717686674, + 'upload_date': '20240606', + 'uploader': 'fm4', + }, + }], + 'skip': 'Shows from ORF Sound are only available for 30 days.' + }, { + # persistent playlist (FM4 Highlights) + 'url': 'https://sound.orf.at/collection/4/', 'info_dict': { - 'id': '2017-01-07_2100_tl_54_7DaysSat18_31295', - 'ext': 'mp3', - 'title': 'Solid Steel Radioshow', - 'description': 'Die Mixshow von Coldcut und Ninja Tune.', - 'duration': 3599, - 'timestamp': 1483819257, - 'upload_date': '20170107', + 'id': '4', }, - 'skip': 'Shows from ORF radios are only available for 7 days.', - 'only_matching': True, - } - - -class ORFNOEIE(ORFRadioIE): - IE_NAME = 'orf:noe' - IE_DESC = 'Radio Niederösterreich' - _VALID_URL = r'https?://(?P<station>noe)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>\w+)' - _API_STATION = 'noe' - _LOOP_STATION = 'oe2n' - - _TEST = { - 'url': 'https://noe.orf.at/player/20200423/NGM', - 'only_matching': True, - } - - -class ORFWIEIE(ORFRadioIE): - IE_NAME = 'orf:wien' - IE_DESC = 'Radio Wien' - _VALID_URL = r'https?://(?P<station>wien)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>\w+)' - _API_STATION = 'wie' - _LOOP_STATION = 'oe2w' - - _TEST = { - 'url': 'https://wien.orf.at/player/20200423/WGUM', - 'only_matching': True, - } - - -class ORFBGLIE(ORFRadioIE): - IE_NAME = 'orf:burgenland' - IE_DESC = 'Radio Burgenland' - _VALID_URL = r'https?://(?P<station>burgenland)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>\w+)' - _API_STATION = 'bgl' - _LOOP_STATION = 'oe2b' - - _TEST = { - 'url': 'https://burgenland.orf.at/player/20200423/BGM', - 'only_matching': True, - } - - -class ORFOOEIE(ORFRadioIE): - IE_NAME = 'orf:oberoesterreich' - IE_DESC = 'Radio Oberösterreich' - _VALID_URL = r'https?://(?P<station>ooe)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>\w+)' - _API_STATION = 'ooe' - _LOOP_STATION = 'oe2o' + 'playlist_mincount': 10, + 'playlist_maxcount': 13, + }] - _TEST = { - 'url': 'https://ooe.orf.at/player/20200423/OGMO', - 'only_matching': True, - } + def _real_extract(self, url): + coll_id, item_id = self._match_valid_url(url).group('coll_id', 'item_id') + data = self._download_json( + 'https://collector.orf.at/api/frontend/collections/{0}?_o=sound.orf.at'.format( + coll_id), coll_id) + data = self._get_api_payload(data, coll_id, in_payload=True) + + def yield_items(): + for item in traverse_obj(data, ( + 'content', 'items', lambda _, v: any(k in v['target']['params'] for k in self._ID_NAMES))): + if item_id is None or item_id == txt_or_none(item.get('id')): + target = item['target'] + typed_item_id = self._get_item_id(target['params']) + station = target['params'].get('station') + item_type = target.get('type') + if typed_item_id and (station or item_type): + yield station, typed_item_id, item_type + if item_id is not None: + break + else: + if item_id is not None: + raise ExtractorError('Item not found in collection', + video_id=coll_id, expected=True) + + def item_playlist(station, typed_item_id, item_type): + if item_type == 'upload': + item_data = self._download_json('https://audioapi.orf.at/radiothek/api/2.0/upload/{0}?_o=sound.orf.at'.format( + typed_item_id), typed_item_id) + elif item_type == 'podcast-episode': + item_data = self._download_json('https://audioapi.orf.at/radiothek/api/2.0/episode/{0}?_o=sound.orf.at'.format( + typed_item_id), typed_item_id) + else: + api_station, _, _ = self.STATION_INFO[station] + item_data = self._download_json( + 'https://audioapi.orf.at/{0}/api/json/5.0/{1}/{2}?_o=sound.orf.at'.format( + api_station, item_type or 'broadcastitem', typed_item_id), typed_item_id) + item_data = self._get_api_payload(item_data, typed_item_id, in_payload=True) -class ORFSTMIE(ORFRadioIE): - IE_NAME = 'orf:steiermark' - IE_DESC = 'Radio Steiermark' - _VALID_URL = r'https?://(?P<station>steiermark)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>\w+)' - _API_STATION = 'stm' - _LOOP_STATION = 'oe2st' + return merge_dicts( + {'_type': 'multi_video'}, + self.playlist_result( + self._entries(item_data, station, item_type), typed_item_id, + txt_or_none(data.get('title')), + clean_html(data.get('subtitle')))) - _TEST = { - 'url': 'https://steiermark.orf.at/player/20200423/STGMS', - 'only_matching': True, - } + def yield_item_entries(): + for station, typed_id, item_type in yield_items(): + yield item_playlist(station, typed_id, item_type) + if item_id is not None: + # coll_id = '/'.join((coll_id, item_id)) + return next(yield_item_entries()) -class ORFKTNIE(ORFRadioIE): - IE_NAME = 'orf:kaernten' - IE_DESC = 'Radio Kärnten' - _VALID_URL = r'https?://(?P<station>kaernten)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>\w+)' - _API_STATION = 'ktn' - _LOOP_STATION = 'oe2k' + return self.playlist_result(yield_item_entries(), coll_id, data.get('title')) - _TEST = { - 'url': 'https://kaernten.orf.at/player/20200423/KGUMO', - 'only_matching': True, - } +class ORFPodcastIE(ORFRadioBase): + IE_NAME = 'orf:podcast' + _STATION_RE = '|'.join(map(re.escape, (x[0] for x in ORFRadioBase.STATION_INFO.values()))) + '|tv' + _VALID_URL = r'https?://sound\.orf\.at/podcast/(?P<station>{0})/(?P<show>[\w-]+)/(?P<id>[\w-]+)'.format(_STATION_RE) + _TESTS = [{ + 'url': 'https://sound.orf.at/podcast/stm/der-kraeutertipp-von-christine-lackner/rotklee', + 'md5': '1f2bab2ba90c2ce0c2754196ea78b35f', + 'info_dict': { + 'id': 'der-kraeutertipp-von-christine-lackner/rotklee', + 'ext': 'mp3', + 'title': 'Rotklee', + 'description': 'In der Natur weit verbreitet - in der Medizin längst anerkennt: Rotklee. Dieser Podcast begleitet die Sendung "Radio Steiermark am Vormittag", Radio Steiermark, 28. Mai 2024.', + 'timestamp': 1716891761, + 'upload_date': '20240528', + 'uploader_id': 'stm_kraeutertipp', + 'uploader': 'ORF Radio Steiermark', + 'duration': 101, + 'series': 'Der Kräutertipp von Christine Lackner', + }, + 'skip': 'ORF podcasts are only available for a limited time' + }] -class ORFSBGIE(ORFRadioIE): - IE_NAME = 'orf:salzburg' - IE_DESC = 'Radio Salzburg' - _VALID_URL = r'https?://(?P<station>salzburg)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>\w+)' - _API_STATION = 'sbg' - _LOOP_STATION = 'oe2s' + _ID_NAMES = ('slug', 'guid') - _TEST = { - 'url': 'https://salzburg.orf.at/player/20200423/SGUM', - 'only_matching': True, - } + def _real_extract(self, url): + station, show, show_id = self._match_valid_url(url).group('station', 'show', 'id') + data = self._download_json( + 'https://audioapi.orf.at/radiothek/api/2.0/podcast/{0}/{1}/{2}'.format( + station, show, show_id), show_id) + data = self._get_api_payload(data, show_id, in_payload=True) + return merge_dicts({ + 'id': '/'.join((show, show_id)), + 'ext': 'mp3', + 'vcodec': 'none', + }, self._extract_podcast_upload(data), rev=True) -class ORFTIRIE(ORFRadioIE): - IE_NAME = 'orf:tirol' - IE_DESC = 'Radio Tirol' - _VALID_URL = r'https?://(?P<station>tirol)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>\w+)' - _API_STATION = 'tir' - _LOOP_STATION = 'oe2t' - _TEST = { - 'url': 'https://tirol.orf.at/player/20200423/TGUMO', - 'only_matching': True, - } +class ORFIPTVBase(InfoExtractor): + _TITLE_STRIP_RE = '' + def _extract_video(self, video_id, webpage, fatal=False): -class ORFVBGIE(ORFRadioIE): - IE_NAME = 'orf:vorarlberg' - IE_DESC = 'Radio Vorarlberg' - _VALID_URL = r'https?://(?P<station>vorarlberg)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>\w+)' - _API_STATION = 'vbg' - _LOOP_STATION = 'oe2v' + data = self._download_json( + 'http://bits.orf.at/filehandler/static-api/json/current/data.json?file=%s' % video_id, + video_id)[0] - _TEST = { - 'url': 'https://vorarlberg.orf.at/player/20200423/VGUM', - 'only_matching': True, - } + video = traverse_obj(data, ( + 'sources', ('default', 'q8c'), + T(lambda x: x if x['loadBalancerUrl'] else None), + any)) + load_balancer_url = video['loadBalancerUrl'] -class ORFOE3IE(ORFRadioIE): - IE_NAME = 'orf:oe3' - IE_DESC = 'Radio Österreich 3' - _VALID_URL = r'https?://(?P<station>oe3)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>\w+)' - _API_STATION = 'oe3' - _LOOP_STATION = 'oe3' + try: + rendition = self._download_json( + load_balancer_url, video_id, transform_source=strip_jsonp) + except ExtractorError: + rendition = None + + if not rendition: + rendition = { + 'redirect': { + 'smil': re.sub( + r'(/)jsonp(/.+\.)mp4$', r'\1dash\2smil/manifest.mpd', + load_balancer_url), + }, + } - _TEST = { - 'url': 'https://oe3.orf.at/player/20200424/3WEK', - 'only_matching': True, - } + f = traverse_obj(video, { + 'abr': ('audioBitrate', T(int_or_none)), + 'vbr': ('bitrate', T(int_or_none)), + 'fps': ('videoFps', T(int_or_none)), + 'width': ('videoWidth', T(int_or_none)), + 'height': ('videoHeight', T(int_or_none)), + }) + formats = [] + for format_id, format_url in traverse_obj(rendition, ( + 'redirect', T(dict.items), Ellipsis)): + if format_id == 'rtmp': + ff = f.copy() + ff.update({ + 'url': format_url, + 'format_id': format_id, + }) + formats.append(ff) + elif determine_ext(format_url) == 'f4m': + formats.extend(self._extract_f4m_formats( + format_url, video_id, f4m_id=format_id)) + elif determine_ext(format_url) == 'm3u8': + formats.extend(self._extract_m3u8_formats( + format_url, video_id, 'mp4', m3u8_id=format_id, + entry_protocol='m3u8_native')) + elif determine_ext(format_url) == 'mpd': + formats.extend(self._extract_mpd_formats( + format_url, video_id, mpd_id=format_id)) -class ORFOE1IE(ORFRadioIE): - IE_NAME = 'orf:oe1' - IE_DESC = 'Radio Österreich 1' - _VALID_URL = r'https?://(?P<station>oe1)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>\w+)' - _API_STATION = 'oe1' - _LOOP_STATION = 'oe1' + if formats or fatal: + self._sort_formats(formats) + else: + return - _TEST = { - 'url': 'http://oe1.orf.at/player/20170108/456544', - 'md5': '34d8a6e67ea888293741c86a099b745b', - 'info_dict': { - 'id': '2017-01-08_0759_tl_51_7DaysSun6_256141', - 'ext': 'mp3', - 'title': 'Morgenjournal', - 'duration': 609, - 'timestamp': 1483858796, - 'upload_date': '20170108', - }, - 'skip': 'Shows from ORF radios are only available for 7 days.' - } + return merge_dicts({ + 'id': video_id, + 'title': re.sub(self._TITLE_STRIP_RE, '', self._og_search_title(webpage)), + 'description': self._og_search_description(webpage), + 'upload_date': unified_strdate(self._html_search_meta( + 'dc.date', webpage, 'upload date', fatal=False)), + 'formats': formats, + }, traverse_obj(data, { + 'duration': ('duration', T(k_float_or_none)), + 'thumbnail': ('sources', 'default', 'preview', T(url_or_none)), + }), rev=True) -class ORFIPTVIE(InfoExtractor): +class ORFIPTVIE(ORFIPTVBase): IE_NAME = 'orf:iptv' IE_DESC = 'iptv.ORF.at' + _WORKING = False # URLs redirect to orf.at/ _VALID_URL = r'https?://iptv\.orf\.at/(?:#/)?stories/(?P<id>\d+)' + _TITLE_STRIP_RE = r'\s+-\s+iptv\.ORF\.at\S*$' _TEST = { 'url': 'http://iptv.orf.at/stories/2275236/', @@ -426,74 +413,32 @@ class ORFIPTVIE(InfoExtractor): video_id = self._search_regex( r'data-video(?:id)?="(\d+)"', webpage, 'video id') - data = self._download_json( - 'http://bits.orf.at/filehandler/static-api/json/current/data.json?file=%s' % video_id, - video_id)[0] - - duration = float_or_none(data['duration'], 1000) + return self._extract_video(video_id, webpage) - video = data['sources']['default'] - load_balancer_url = video['loadBalancerUrl'] - abr = int_or_none(video.get('audioBitrate')) - vbr = int_or_none(video.get('bitrate')) - fps = int_or_none(video.get('videoFps')) - width = int_or_none(video.get('videoWidth')) - height = int_or_none(video.get('videoHeight')) - thumbnail = video.get('preview') - - rendition = self._download_json( - load_balancer_url, video_id, transform_source=strip_jsonp) - - f = { - 'abr': abr, - 'vbr': vbr, - 'fps': fps, - 'width': width, - 'height': height, - } - formats = [] - for format_id, format_url in rendition['redirect'].items(): - if format_id == 'rtmp': - ff = f.copy() - ff.update({ - 'url': format_url, - 'format_id': format_id, - }) - formats.append(ff) - elif determine_ext(format_url) == 'f4m': - formats.extend(self._extract_f4m_formats( - format_url, video_id, f4m_id=format_id)) - elif determine_ext(format_url) == 'm3u8': - formats.extend(self._extract_m3u8_formats( - format_url, video_id, 'mp4', m3u8_id=format_id)) - else: - continue - self._sort_formats(formats) - - title = remove_end(self._og_search_title(webpage), ' - iptv.ORF.at') - description = self._og_search_description(webpage) - upload_date = unified_strdate(self._html_search_meta( - 'dc.date', webpage, 'upload date')) - - return { - 'id': video_id, - 'title': title, - 'description': description, - 'duration': duration, - 'thumbnail': thumbnail, - 'upload_date': upload_date, - 'formats': formats, - } - - -class ORFFM4StoryIE(InfoExtractor): +class ORFFM4StoryIE(ORFIPTVBase): IE_NAME = 'orf:fm4:story' IE_DESC = 'fm4.orf.at stories' _VALID_URL = r'https?://fm4\.orf\.at/stories/(?P<id>\d+)' + _TITLE_STRIP_RE = r'\s+-\s+fm4\.ORF\.at\s*$' - _TEST = { + _TESTS = [{ + 'url': 'https://fm4.orf.at/stories/3041554/', + 'add_ie': ['Youtube'], + 'info_dict': { + 'id': '3041554', + 'title': 'Is The EU Green Deal In Mortal Danger?', + }, + 'playlist_count': 4, + 'params': { + 'format': 'bestvideo', + }, + }, { 'url': 'http://fm4.orf.at/stories/2865738/', + 'info_dict': { + 'id': '2865738', + 'title': 'Manu Delago und Inner Tongue live', + }, 'playlist': [{ 'md5': 'e1c2c706c45c7b34cf478bbf409907ca', 'info_dict': { @@ -510,83 +455,311 @@ class ORFFM4StoryIE(InfoExtractor): 'info_dict': { 'id': '547798', 'ext': 'flv', - 'title': 'Manu Delago und Inner Tongue live (2)', + 'title': 'Manu Delago und Inner Tongue https://vod-ww.mdn.ors.at/cms-worldwide_episodes_nas/_definst_/nas/cms-worldwide_episodes/online/14228823_0005.smil/chunklist_b992000_vo.m3u8live (2)', 'duration': 1504.08, 'thumbnail': r're:^https?://.*\.jpg$', 'upload_date': '20170913', 'description': 'Manu Delago und Inner Tongue haben bei der FM4 Soundpark Session live alles gegeben. Hier gibt es Fotos und die gesamte Session als Video.', }, }], - } + 'skip': 'Videos gone', + }] def _real_extract(self, url): story_id = self._match_id(url) webpage = self._download_webpage(url, story_id) entries = [] - all_ids = orderedSet(re.findall(r'data-video(?:id)?="(\d+)"', webpage)) - for idx, video_id in enumerate(all_ids): - data = self._download_json( - 'http://bits.orf.at/filehandler/static-api/json/current/data.json?file=%s' % video_id, - video_id)[0] + seen_ids = set() + for idx, video_id in enumerate(re.findall(r'data-video(?:id)?="(\d+)"', webpage)): + if video_id in seen_ids: + continue + seen_ids.add(video_id) + entry = self._extract_video(video_id, webpage, fatal=False) + if not entry: + continue + + if idx >= 1: + # Titles are duplicates, make them unique + entry['title'] = '%s (%d)' % (entry['title'], idx) - duration = float_or_none(data['duration'], 1000) + entries.append(entry) - video = data['sources']['q8c'] - load_balancer_url = video['loadBalancerUrl'] - abr = int_or_none(video.get('audioBitrate')) - vbr = int_or_none(video.get('bitrate')) - fps = int_or_none(video.get('videoFps')) - width = int_or_none(video.get('videoWidth')) - height = int_or_none(video.get('videoHeight')) - thumbnail = video.get('preview') + seen_ids = set() + for yt_id in re.findall( + r'data-id\s*=\s*["\']([\w-]+)[^>]+\bclass\s*=\s*["\']youtube\b', + webpage): + if yt_id in seen_ids: + continue + seen_ids.add(yt_id) + if YoutubeIE.suitable(yt_id): + entries.append(self.url_result(yt_id, ie='Youtube', video_id=yt_id)) + + return self.playlist_result( + entries, story_id, + re.sub(self._TITLE_STRIP_RE, '', self._og_search_title(webpage, default='') or None)) + + +class ORFONBase(InfoExtractor): + _ENC_PFX = '3dSlfek03nsLKdj4Jsd' + _API_PATH = 'episode' + + def _call_api(self, video_id, **kwargs): + encrypted_id = base64.b64encode('{0}{1}'.format( + self._ENC_PFX, video_id).encode('utf-8')).decode('ascii') + return self._download_json( + 'https://api-tvthek.orf.at/api/v4.3/public/{0}/encrypted/{1}'.format( + self._API_PATH, encrypted_id), + video_id, **kwargs) + + @classmethod + def _parse_metadata(cls, api_json): + return traverse_obj(api_json, { + 'id': ('id', T(int), T(txt_or_none)), + 'age_limit': ('age_classification', T(parse_age_limit)), + 'duration': ((('exact_duration', T(k_float_or_none)), + ('duration_second', T(float_or_none))),), + 'title': (('title', 'headline'), T(txt_or_none)), + 'description': (('description', 'teaser_text'), T(txt_or_none)), + # 'media_type': ('video_type', T(txt_or_none)), + 'thumbnail': ('_embedded', 'image', 'public_urls', 'highlight_teaser', 'url', T(url_or_none)), + 'timestamp': (('date', 'episode_date'), T(parse_iso8601)), + 'release_timestamp': ('release_date', T(parse_iso8601)), + # 'modified_timestamp': ('updated_at', T(parse_iso8601)), + }, get_all=False) + + def _extract_video(self, video_id, segment_id): + # Not a segmented episode: return single video + # Segmented episode without valid segment id: return entire playlist + # Segmented episode with valid segment id and yes-playlist: return entire playlist + # Segmented episode with valid segment id and no-playlist: return single video corresponding to segment id + # If a multi_video playlist would be returned, but an unsegmented source exists, that source is chosen instead. + + api_json = self._call_api(video_id) + + if traverse_obj(api_json, 'is_drm_protected'): + self.report_drm(video_id) + + # updates formats, subtitles + def extract_sources(src_json, video_id): + for manifest_type in traverse_obj(src_json, ('sources', T(dict.keys), Ellipsis)): + for manifest_url in traverse_obj(src_json, ('sources', manifest_type, Ellipsis, 'src', T(url_or_none))): + if manifest_type == 'hls': + fmts, subs = self._extract_m3u8_formats( + manifest_url, video_id, fatal=False, m3u8_id='hls', + ext='mp4', entry_protocol='m3u8_native'), {} + for f in fmts: + if '_vo.' in f['url']: + f['acodec'] = 'none' + elif manifest_type == 'dash': + fmts, subs = self._extract_mpd_formats_and_subtitles( + manifest_url, video_id, fatal=False, mpd_id='dash') + else: + continue + formats.extend(fmts) + self._merge_subtitles(subs, target=subtitles) + + formats, subtitles = [], {} + if segment_id is None: + extract_sources(api_json, video_id) + if not formats: + segments = traverse_obj(api_json, ( + '_embedded', 'segments', lambda _, v: v['id'])) + if len(segments) > 1 and segment_id is not None: + if not self._yes_playlist(video_id, segment_id, playlist_label='collection', video_label='segment'): + segments = [next(s for s in segments if txt_or_none(s['id']) == segment_id)] + + entries = [] + for seg in segments: + formats, subtitles = [], {} + extract_sources(seg, segment_id) + self._sort_formats(formats) + entries.append(merge_dicts({ + 'formats': formats, + 'subtitles': subtitles, + }, self._parse_metadata(seg), rev=True)) + result = merge_dicts( + {'_type': 'multi_video' if len(entries) > 1 else 'playlist'}, + self._parse_metadata(api_json), + self.playlist_result(entries, video_id)) + # not yet processed in core for playlist/multi + self._downloader._fill_common_fields(result) + return result + else: + self._sort_formats(formats) - rendition = self._download_json( - load_balancer_url, video_id, transform_source=strip_jsonp) + for sub_url in traverse_obj(api_json, ( + '_embedded', 'subtitle', + ('xml_url', 'sami_url', 'stl_url', 'ttml_url', 'srt_url', 'vtt_url'), + T(url_or_none))): + self._merge_subtitles({'de': [{'url': sub_url}]}, target=subtitles) - f = { - 'abr': abr, - 'vbr': vbr, - 'fps': fps, - 'width': width, - 'height': height, - } + return merge_dicts({ + 'id': video_id, + 'formats': formats, + 'subtitles': subtitles, + # '_old_archive_ids': [self._downloader._make_archive_id({'ie_key': 'ORFTVthek', 'id': video_id})], + }, self._parse_metadata(api_json), rev=True) - formats = [] - for format_id, format_url in rendition['redirect'].items(): - if format_id == 'rtmp': - ff = f.copy() - ff.update({ - 'url': format_url, - 'format_id': format_id, - }) - formats.append(ff) - elif determine_ext(format_url) == 'f4m': - formats.extend(self._extract_f4m_formats( - format_url, video_id, f4m_id=format_id)) - elif determine_ext(format_url) == 'm3u8': - formats.extend(self._extract_m3u8_formats( - format_url, video_id, 'mp4', m3u8_id=format_id)) - else: - continue - self._sort_formats(formats) + def _real_extract(self, url): + video_id, segment_id = self._match_valid_url(url).group('id', 'segment') + webpage = self._download_webpage(url, video_id) - title = remove_end(self._og_search_title(webpage), ' - fm4.ORF.at') - if idx >= 1: - # Titles are duplicates, make them unique - title += ' (' + str(idx + 1) + ')' - description = self._og_search_description(webpage) - upload_date = unified_strdate(self._html_search_meta( - 'dc.date', webpage, 'upload date')) - - entries.append({ - 'id': video_id, - 'title': title, - 'description': description, - 'duration': duration, - 'thumbnail': thumbnail, - 'upload_date': upload_date, - 'formats': formats, - }) - - return self.playlist_result(entries) + # ORF doesn't like 410 or 404 + if self._search_regex(r'<div\b[^>]*>\s*(Nicht mehr verfügbar)\s*</div>', webpage, 'Availability', default=False): + raise ExtractorError('Content is no longer available', expected=True, video_id=video_id) + + return merge_dicts({ + 'id': video_id, + 'title': self._html_search_meta(['og:title', 'twitter:title'], webpage, default=None), + 'description': self._html_search_meta( + ['description', 'og:description', 'twitter:description'], webpage, default=None), + }, self._search_json_ld(webpage, video_id, default={}), + self._extract_video(video_id, segment_id), + rev=True) + + +class ORFONIE(ORFONBase): + IE_NAME = 'orf:on' + _VALID_URL = r'https?://on\.orf\.at/video/(?P<id>\d+)(?:/(?P<segment>\d+))?' + _TESTS = [{ + 'url': 'https://on.orf.at/video/14210000/school-of-champions-48', + 'info_dict': { + 'id': '14210000', + 'ext': 'mp4', + 'duration': 2651.08, + 'thumbnail': 'https://api-tvthek.orf.at/assets/segments/0167/98/thumb_16697671_segments_highlight_teaser.jpeg', + 'title': 'School of Champions (4/8)', + 'description': r're:(?s)Luca hat sein ganzes Leben in den Bergen Südtirols verbracht und ist bei seiner Mutter aufgewachsen, .{1029} Leo$', + # 'media_type': 'episode', + 'timestamp': 1706558922, + 'upload_date': '20240129', + 'release_timestamp': 1706472362, + 'release_date': '20240128', + # 'modified_timestamp': 1712756663, + # 'modified_date': '20240410', + # '_old_archive_ids': ['orftvthek 14210000'], + }, + 'params': { + 'format': 'bestvideo', + }, + 'skip': 'Available until 2024-08-12', + }, { + 'url': 'https://on.orf.at/video/3220355', + 'md5': '925a93b2b9a37da5c9b979d7cf71aa2e', + 'info_dict': { + 'id': '3220355', + 'ext': 'mp4', + 'duration': 445.04, + 'thumbnail': 'https://api-tvthek.orf.at/assets/segments/0002/60/thumb_159573_segments_highlight_teaser.png', + 'title': '50 Jahre Burgenland: Der Festumzug', + 'description': r're:(?s)Aus allen Landesteilen zogen festlich geschmückte Wagen und Musikkapellen .{270} Jenakowitsch$', + # 'media_type': 'episode', + 'timestamp': 52916400, + 'upload_date': '19710905', + 'release_timestamp': 52916400, + 'release_date': '19710905', + # 'modified_timestamp': 1498536049, + # 'modified_date': '20170627', + # '_old_archive_ids': ['orftvthek 3220355'], + }, + }, { + # Video with multiple segments selecting the second segment + 'url': 'https://on.orf.at/video/14226549/15639808/jugendbande-einbrueche-aus-langeweile', + 'md5': 'fc151bba8c05ea77ab5693617e4a33d3', + 'info_dict': { + 'id': '15639808', + 'ext': 'mp4', + 'duration': 97.707, + 'thumbnail': 'https://api-tvthek.orf.at/assets/segments/0175/43/thumb_17442704_segments_highlight_teaser.jpg', + 'title': 'Jugendbande: Einbrüche aus Langeweile', + 'description': r're:Jugendbande: Einbrüche aus Langeweile \| Neuer Kinder- und .{259} Wanda$', + # 'media_type': 'segment', + 'timestamp': 1715792400, + 'upload_date': '20240515', + # 'modified_timestamp': 1715794394, + # 'modified_date': '20240515', + # '_old_archive_ids': ['orftvthek 15639808'], + }, + 'params': { + 'noplaylist': True, + 'format': 'bestvideo', + }, + 'skip': 'Available until 2024-06-14', + }, { + # Video with multiple segments and no combined version + 'url': 'https://on.orf.at/video/14227864/formel-1-grosser-preis-von-monaco-2024', + 'info_dict': { + '_type': 'multi_video', + 'id': '14227864', + 'duration': 18410.52, + 'thumbnail': 'https://api-tvthek.orf.at/assets/segments/0176/04/thumb_17503881_segments_highlight_teaser.jpg', + 'title': 'Formel 1: Großer Preis von Monaco 2024', + 'description': 'md5:aeeb010710ccf70ce28ccb4482243d4f', + # 'media_type': 'episode', + 'timestamp': 1716721200, + 'upload_date': '20240526', + 'release_timestamp': 1716721802, + 'release_date': '20240526', + # 'modified_timestamp': 1716884702, + # 'modified_date': '20240528', + }, + 'playlist_count': 42, + 'skip': 'Gone: Nicht mehr verfügbar', + }, { + # Video with multiple segments, but with combined version + 'url': 'https://on.orf.at/video/14228172', + 'info_dict': { + 'id': '14228172', + 'ext': 'mp4', + 'duration': 3294.878, + 'thumbnail': 'https://api-tvthek.orf.at/assets/segments/0176/29/thumb_17528242_segments_highlight_teaser.jpg', + 'title': 'Willkommen Österreich mit Stermann & Grissemann', + 'description': r're:Zum Saisonfinale freuen sich die urlaubsreifen Gastgeber Stermann und .{1863} Geschichten\.$', + # 'media_type': 'episode', + 'timestamp': 1716926584, + 'upload_date': '20240528', + 'release_timestamp': 1716919202, + 'release_date': '20240528', + # 'modified_timestamp': 1716968045, + # 'modified_date': '20240529', + # '_old_archive_ids': ['orftvthek 14228172'], + }, + 'params': { + 'format': 'bestvideo', + }, + 'skip': 'Gone: Nicht mehr verfügbar', + }] + + +class ORFONLiveIE(ORFONBase): + _ENC_PFX = '8876324jshjd7293ktd' + _API_PATH = 'livestream' + _VALID_URL = r'https?://on\.orf\.at/livestream/(?P<id>\d+)(?:/(?P<segment>\d+))?' + _TESTS = [{ + 'url': 'https://on.orf.at/livestream/14320204/pressekonferenz-neos-zu-aktuellen-entwicklungen', + 'info_dict': { + 'id': '14320204', + 'ext': 'mp4', + 'title': 'Pressekonferenz: Neos zu aktuellen Entwicklungen', + 'description': r're:(?s)Neos-Chefin Beate Meinl-Reisinger informi.{598}ng\."', + 'timestamp': 1716886335, + 'upload_date': '20240528', + # 'modified_timestamp': 1712756663, + # 'modified_date': '20240410', + # '_old_archive_ids': ['orftvthek 14210000'], + }, + 'params': { + 'format': 'bestvideo', + }, + }] + + @classmethod + def _parse_metadata(cls, api_json): + return merge_dicts( + super(ORFONLiveIE, cls)._parse_metadata(api_json), + traverse_obj(api_json, { + 'timestamp': ('updated_at', T(parse_iso8601)), + 'release_timestamp': ('start', T(parse_iso8601)), + 'is_live': True, + })) diff --git a/youtube_dl/extractor/palcomp3.py b/youtube_dl/extractor/palcomp3.py index fb29d83f9..60f7a4d48 100644 --- a/youtube_dl/extractor/palcomp3.py +++ b/youtube_dl/extractor/palcomp3.py @@ -8,7 +8,7 @@ from ..compat import compat_str from ..utils import ( int_or_none, str_or_none, - try_get, + traverse_obj, ) @@ -109,7 +109,7 @@ class PalcoMP3ArtistIE(PalcoMP3BaseIE): } name''' - @ classmethod + @classmethod def suitable(cls, url): return False if re.match(PalcoMP3IE._VALID_URL, url) else super(PalcoMP3ArtistIE, cls).suitable(url) @@ -118,7 +118,8 @@ class PalcoMP3ArtistIE(PalcoMP3BaseIE): artist = self._call_api(artist_slug, self._ARTIST_FIELDS_TMPL)['artist'] def entries(): - for music in (try_get(artist, lambda x: x['musics']['nodes'], list) or []): + for music in traverse_obj(artist, ( + 'musics', 'nodes', lambda _, m: m['musicID'])): yield self._parse_music(music) return self.playlist_result( @@ -137,7 +138,7 @@ class PalcoMP3VideoIE(PalcoMP3BaseIE): 'title': 'Maiara e Maraisa - Você Faz Falta Aqui - DVD Ao Vivo Em Campo Grande', 'description': 'md5:7043342c09a224598e93546e98e49282', 'upload_date': '20161107', - 'uploader_id': 'maiaramaraisaoficial', + 'uploader_id': '@maiaramaraisaoficial', 'uploader': 'Maiara e Maraisa', } }] diff --git a/youtube_dl/extractor/senateisvp.py b/youtube_dl/extractor/senateisvp.py index db5ef8b57..b8ac58713 100644 --- a/youtube_dl/extractor/senateisvp.py +++ b/youtube_dl/extractor/senateisvp.py @@ -47,7 +47,7 @@ class SenateISVPIE(InfoExtractor): ['vetaff', '76462', 'http://vetaff-f.akamaihd.net'], ['arch', '', 'http://ussenate-f.akamaihd.net/'] ] - _IE_NAME = 'senate.gov' + IE_NAME = 'senate.gov' _VALID_URL = r'https?://(?:www\.)?senate\.gov/isvp/?\?(?P<qs>.+)' _TESTS = [{ 'url': 'http://www.senate.gov/isvp/?comm=judiciary&type=live&stt=&filename=judiciary031715&auto_play=false&wmode=transparent&poster=http%3A%2F%2Fwww.judiciary.senate.gov%2Fthemes%2Fjudiciary%2Fimages%2Fvideo-poster-flash-fit.png', diff --git a/youtube_dl/extractor/vidlii.py b/youtube_dl/extractor/vidlii.py index f4774256b..47f328e87 100644 --- a/youtube_dl/extractor/vidlii.py +++ b/youtube_dl/extractor/vidlii.py @@ -4,6 +4,7 @@ from __future__ import unicode_literals import re from .common import InfoExtractor + from ..utils import ( float_or_none, get_element_by_id, @@ -11,6 +12,7 @@ from ..utils import ( strip_or_none, unified_strdate, urljoin, + str_to_int, ) @@ -36,6 +38,26 @@ class VidLiiIE(InfoExtractor): 'tags': ['Vidlii', 'Jan', 'Videogames'], } }, { + # HD + 'url': 'https://www.vidlii.com/watch?v=2Ng8Abj2Fkl', + 'md5': '450e7da379c884788c3a4fa02a3ce1a4', + 'info_dict': { + 'id': '2Ng8Abj2Fkl', + 'ext': 'mp4', + 'title': 'test', + 'description': 'md5:cc55a86032a7b6b3cbfd0f6b155b52e9', + 'thumbnail': 'https://www.vidlii.com/usfi/thmp/2Ng8Abj2Fkl.jpg', + 'uploader': 'VidLii', + 'uploader_url': 'https://www.vidlii.com/user/VidLii', + 'upload_date': '20200927', + 'duration': 5, + 'view_count': int, + 'comment_count': int, + 'average_rating': float, + 'categories': ['Film & Animation'], + 'tags': list, + }, + }, { 'url': 'https://www.vidlii.com/embed?v=tJluaH4BJ3v&a=0', 'only_matching': True, }] @@ -46,11 +68,32 @@ class VidLiiIE(InfoExtractor): webpage = self._download_webpage( 'https://www.vidlii.com/watch?v=%s' % video_id, video_id) - video_url = self._search_regex( - r'src\s*:\s*(["\'])(?P<url>(?:https?://)?(?:(?!\1).)+)\1', webpage, - 'video url', group='url') + formats = [] + + def add_format(format_url, height=None): + height = int(self._search_regex(r'(\d+)\.mp4', + format_url, 'height', default=360)) + + formats.append({ + 'url': format_url, + 'format_id': '%dp' % height if height else None, + 'height': height, + }) + + sources = re.findall( + r'src\s*:\s*(["\'])(?P<url>(?:https?://)?(?:(?!\1).)+)\1', + webpage) + + formats = [] + if len(sources) > 1: + add_format(sources[1][1]) + self._check_formats(formats, video_id) + if len(sources) > 0: + add_format(sources[0][1]) + + self._sort_formats(formats) - title = self._search_regex( + title = self._html_search_regex( (r'<h1>([^<]+)</h1>', r'<title>([^<]+) - VidLii<'), webpage, 'title') @@ -82,9 +125,9 @@ class VidLiiIE(InfoExtractor): default=None) or self._search_regex( r'duration\s*:\s*(\d+)', webpage, 'duration', fatal=False)) - view_count = int_or_none(self._search_regex( - (r'<strong>(\d+)</strong> views', - r'Views\s*:\s*<strong>(\d+)</strong>'), + view_count = str_to_int(self._html_search_regex( + (r'<strong>([\d,.]+)</strong> views', + r'Views\s*:\s*<strong>([\d,.]+)</strong>'), webpage, 'view count', fatal=False)) comment_count = int_or_none(self._search_regex( @@ -109,7 +152,7 @@ class VidLiiIE(InfoExtractor): return { 'id': video_id, - 'url': video_url, + 'formats': formats, 'title': title, 'description': description, 'thumbnail': thumbnail, diff --git a/youtube_dl/extractor/yandexmusic.py b/youtube_dl/extractor/yandexmusic.py index 84969f8e1..8da5b430f 100644 --- a/youtube_dl/extractor/yandexmusic.py +++ b/youtube_dl/extractor/yandexmusic.py @@ -106,6 +106,25 @@ class YandexMusicTrackIE(YandexMusicBaseIE): }, { 'url': 'http://music.yandex.com/album/540508/track/4878838', 'only_matching': True, + }, { + 'url': 'https://music.yandex.ru/album/16302456/track/85430762', + 'md5': '11b8d50ab03b57738deeaadf661a0a48', + 'info_dict': { + 'id': '85430762', + 'ext': 'mp3', + 'abr': 128, + 'title': 'Haddadi Von Engst, Phonic Youth, Super Flu - Til The End (Super Flu Remix)', + 'filesize': int, + 'duration': 431.14, + 'track': 'Til The End (Super Flu Remix)', + 'album': 'Til The End', + 'album_artist': 'Haddadi Von Engst, Phonic Youth', + 'artist': 'Haddadi Von Engst, Phonic Youth, Super Flu', + 'release_year': 2021, + 'genre': 'house', + 'disc_number': 1, + 'track_number': 2, + } }] def _real_extract(self, url): @@ -116,10 +135,14 @@ class YandexMusicTrackIE(YandexMusicBaseIE): 'track', tld, url, track_id, 'Downloading track JSON', {'track': '%s:%s' % (track_id, album_id)})['track'] track_title = track['title'] + track_version = track.get('version') + if track_version: + track_title = '%s (%s)' % (track_title, track_version) download_data = self._download_json( 'https://music.yandex.ru/api/v2.1/handlers/track/%s:%s/web-album_track-track-track-main/download/m' % (track_id, album_id), track_id, 'Downloading track location url JSON', + query={'hq': 1}, headers={'X-Retpath-Y': url}) fd_data = self._download_json( diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index 90c16e172..b31798729 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -3,11 +3,14 @@ from __future__ import unicode_literals import collections +import hashlib import itertools import json import os.path import random import re +import string +import time import traceback from .common import InfoExtractor, SearchInfoExtractor @@ -24,11 +27,14 @@ from ..compat import ( ) from ..jsinterp import JSInterpreter from ..utils import ( + bug_reports_message, clean_html, dict_get, error_to_compat_str, ExtractorError, + filter_dict, float_or_none, + get_first, extract_attributes, get_element_by_attribute, int_or_none, @@ -43,6 +49,7 @@ from ..utils import ( parse_duration, parse_qs, qualities, + remove_end, remove_start, smuggle_url, str_or_none, @@ -60,11 +67,13 @@ from ..utils import ( url_or_none, urlencode_postdata, urljoin, + variadic, ) class YoutubeBaseInfoExtractor(InfoExtractor): """Provide base functions for Youtube extractors""" + _LOGIN_URL = 'https://accounts.google.com/ServiceLogin' _TWOFACTOR_URL = 'https://accounts.google.com/signin/challenge' @@ -78,9 +87,66 @@ class YoutubeBaseInfoExtractor(InfoExtractor): _PLAYLIST_ID_RE = r'(?:(?:PL|LL|EC|UU|FL|RD|UL|TL|PU|OLAK5uy_)[0-9A-Za-z-_]{10,}|RDMM)' + _INNERTUBE_CLIENTS = { + 'ios': { + 'INNERTUBE_CONTEXT': { + 'client': { + 'clientName': 'IOS', + 'clientVersion': '20.10.4', + 'deviceMake': 'Apple', + 'deviceModel': 'iPhone16,2', + 'userAgent': 'com.google.ios.youtube/20.10.4 (iPhone16,2; U; CPU iOS 18_3_2 like Mac OS X;)', + 'osName': 'iPhone', + 'osVersion': '18.3.2.22D82', + }, + }, + 'INNERTUBE_CONTEXT_CLIENT_NAME': 5, + 'REQUIRE_JS_PLAYER': False, + 'REQUIRE_PO_TOKEN': True, + }, + # mweb has 'ultralow' formats + # See: https://github.com/yt-dlp/yt-dlp/pull/557 + 'mweb': { + 'INNERTUBE_CONTEXT': { + 'client': { + 'clientName': 'MWEB', + 'clientVersion': '2.20250311.03.00', + # mweb previously did not require PO Token with this UA + 'userAgent': 'Mozilla/5.0 (iPad; CPU OS 16_7_10 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/16.6 Mobile/15E148 Safari/604.1,gzip(gfe)', + }, + }, + 'INNERTUBE_CONTEXT_CLIENT_NAME': 2, + 'REQUIRE_PO_TOKEN': True, + 'SUPPORTS_COOKIES': True, + }, + 'tv': { + 'INNERTUBE_CONTEXT': { + 'client': { + 'clientName': 'TVHTML5', + 'clientVersion': '7.20250312.16.00', + 'userAgent': 'Mozilla/5.0 (ChromiumStylePlatform) Cobalt/Version', + }, + }, + 'INNERTUBE_CONTEXT_CLIENT_NAME': 7, + 'SUPPORTS_COOKIES': True, + }, + 'web': { + 'INNERTUBE_CONTEXT': { + 'client': { + 'clientName': 'WEB', + 'clientVersion': '2.20250312.04.00', + }, + }, + 'INNERTUBE_CONTEXT_CLIENT_NAME': 1, + 'REQUIRE_PO_TOKEN': True, + 'SUPPORTS_COOKIES': True, + }, + } + def _login(self): """ Attempt to log in to YouTube. + True is returned if successful or skipped. False is returned if login failed. @@ -136,7 +202,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor): [2, 1, None, 1, 'https://accounts.google.com/ServiceLogin?passive=true&continue=https%3A%2F%2Fwww.youtube.com%2Fsignin%3Fnext%3D%252F%26action_handle_signin%3Dtrue%26hl%3Den%26app%3Ddesktop%26feature%3Dsign_in_button&hl=en&service=youtube&uilel=3&requestPath=%2FServiceLogin&Page=PasswordSeparationSignIn', None, [], 4], - 1, [None, None, []], None, None, None, True + 1, [None, None, []], None, None, None, True, ], username, ] @@ -158,7 +224,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor): None, 1, None, [1, None, None, None, [password, None, True]], [ None, None, [2, 1, None, 1, 'https://accounts.google.com/ServiceLogin?passive=true&continue=https%3A%2F%2Fwww.youtube.com%2Fsignin%3Fnext%3D%252F%26action_handle_signin%3Dtrue%26hl%3Den%26app%3Ddesktop%26feature%3Dsign_in_button&hl=en&service=youtube&uilel=3&requestPath=%2FServiceLogin&Page=PasswordSeparationSignIn', None, [], 4], - 1, [None, None, []], None, None, None, True + 1, [None, None, []], None, None, None, True, ]] challenge_results = req( @@ -211,7 +277,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor): user_hash, None, 2, None, [ 9, None, None, None, None, None, None, None, - [None, tfa_code, True, 2] + [None, tfa_code, True, 2], ]] tfa_results = req( @@ -277,32 +343,57 @@ class YoutubeBaseInfoExtractor(InfoExtractor): if not self._login(): return - _DEFAULT_API_DATA = { - 'context': { - 'client': { - 'clientName': 'WEB', - 'clientVersion': '2.20201021.03.00', - } - }, - } + _DEFAULT_API_DATA = {'context': _INNERTUBE_CLIENTS['web']['INNERTUBE_CONTEXT']} _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, fatal=True, headers=None): + _SAPISID = None + + def _generate_sapisidhash_header(self, origin='https://www.youtube.com'): + time_now = round(time.time()) + if self._SAPISID is None: + yt_cookies = self._get_cookies('https://www.youtube.com') + # Sometimes SAPISID cookie isn't present but __Secure-3PAPISID is. + # See: https://github.com/yt-dlp/yt-dlp/issues/393 + sapisid_cookie = dict_get( + yt_cookies, ('__Secure-3PAPISID', 'SAPISID')) + 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( + '{0} {1} {2}'.format(time_now, self._SAPISID, origin).encode('utf-8')).hexdigest() + return 'SAPISIDHASH {0}_{1}'.format(time_now, sapisidhash) + + def _call_api(self, ep, query, video_id, fatal=True, headers=None, + note='Downloading API JSON'): data = self._DEFAULT_API_DATA.copy() data.update(query) real_headers = {'content-type': 'application/json'} if headers: real_headers.update(headers) + # was: 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8' + api_key = self.get_param('youtube_innertube_key') return self._download_json( 'https://www.youtube.com/youtubei/v1/%s' % ep, video_id=video_id, - note='Downloading API JSON', errnote='Unable to download API page', + note=note, errnote='Unable to download API page', data=json.dumps(data).encode('utf8'), fatal=fatal, - headers=real_headers, - query={'key': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8'}) + headers=real_headers, query=filter_dict({ + 'key': api_key, + 'prettyPrint': 'false', + })) def _extract_yt_initial_data(self, video_id, webpage): return self._parse_json( @@ -311,6 +402,22 @@ class YoutubeBaseInfoExtractor(InfoExtractor): self._YT_INITIAL_DATA_RE), webpage, 'yt initial data'), video_id) + def _extract_visitor_data(self, *args): + """ + Extract visitorData from an API response or ytcfg + + Appears to be used to track session state + """ + visitor_data = self.get_param('youtube_visitor_data') + if visitor_data: + return visitor_data + + return get_first( + args, (('VISITOR_DATA', + ('INNERTUBE_CONTEXT', 'client', 'visitorData'), + ('responseContext', 'visitorData')), + T(compat_str))) + def _extract_ytcfg(self, video_id, webpage): return self._parse_json( self._search_regex( @@ -350,13 +457,33 @@ class YoutubeBaseInfoExtractor(InfoExtractor): 'uploader': uploader, } + @staticmethod + def _extract_thumbnails(data, *path_list, **kw_final_key): + """ + Extract thumbnails from thumbnails dict + @param path_list: path list to level that contains 'thumbnails' key + """ + final_key = kw_final_key.get('final_key', 'thumbnails') + + return traverse_obj(data, (( + tuple(variadic(path) + (final_key, Ellipsis) + for path in path_list or [()])), { + 'url': ('url', T(url_or_none), + # 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 + T(lambda u: update_url(u, query=None) if u and 'maxresdefault' in u else u)), + 'height': ('height', T(int_or_none)), + 'width': ('width', T(int_or_none)), + }, T(lambda t: t if t.get('url') else None))) + def _search_results(self, query, params): data = { 'context': { 'client': { 'clientName': 'WEB', 'clientVersion': '2.20201021.03.00', - } + }, }, 'query': query, } @@ -364,11 +491,15 @@ class YoutubeBaseInfoExtractor(InfoExtractor): data['params'] = params for page_num in itertools.count(1): search = self._download_json( - 'https://www.youtube.com/youtubei/v1/search?key=AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8', + 'https://www.youtube.com/youtubei/v1/search', video_id='query "%s"' % query, note='Downloading page %s' % page_num, errnote='Unable to download API page', fatal=False, data=json.dumps(data).encode('utf8'), + query={ + # 'key': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8', + 'prettyPrint': 'false', + }, headers={'content-type': 'application/json'}) if not search: break @@ -433,7 +564,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor): # (HTML, videodetails, metadata, renderers) 'name': ('content', 'author', (('ownerChannelName', None), 'title'), ['text']), 'url': ('href', 'ownerProfileUrl', 'vanityChannelUrl', - ['navigationEndpoint', 'browseEndpoint', 'canonicalBaseUrl']) + ['navigationEndpoint', 'browseEndpoint', 'canonicalBaseUrl']), } if any((videodetails, metadata, renderers)): result = ( @@ -559,9 +690,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'invidious': '|'.join(_INVIDIOUS_SITES), } _PLAYER_INFO_RE = ( - r'/s/player/(?P<id>[a-zA-Z0-9_-]{8,})/player', - r'/(?P<id>[a-zA-Z0-9_-]{8,})/player(?:_ias\.vflset(?:/[a-zA-Z]{2,3}_[a-zA-Z]{2,3})?|-plasma-ias-(?:phone|tablet)-[a-z]{2}_[A-Z]{2}\.vflset)/base\.js$', - r'\b(?P<id>vfl[a-zA-Z0-9_-]+)\b.*?\.js$', + r'/s/player/(?P<id>[a-zA-Z0-9_-]{8,})/(?:tv-)?player', + r'/(?P<id>[a-zA-Z0-9_-]{8,})/player(?:_ias(?:_tce)?\.vflset(?:/[a-zA-Z]{2,3}_[a-zA-Z]{2,3})?|-plasma-ias-(?:phone|tablet)-[a-z]{2}_[A-Z]{2}\.vflset)/base\.js$', + r'\b(?P<id>vfl[a-zA-Z0-9_-]{6,})\b.*?\.js$', ) _SUBTITLE_FORMATS = ('json3', 'srv1', 'srv2', 'srv3', 'ttml', 'vtt') @@ -642,7 +773,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/8KVIDEO', 'description': '', 'uploader': '8KVIDEO', - 'title': 'UHDTV TEST 8K VIDEO.mp4' + 'title': 'UHDTV TEST 8K VIDEO.mp4', }, 'params': { 'youtube_include_dash_manifest': True, @@ -682,7 +813,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'uploader_url': r're:https?://(?:www\.)?youtube\.com/@theamazingatheist', 'title': 'Burning Everyone\'s Koran', 'description': 'SUBSCRIBE: http://www.youtube.com/saturninefilms \r\n\r\nEven Obama has taken a stand against freedom on this issue: http://www.huffingtonpost.com/2010/09/09/obama-gma-interview-quran_n_710282.html', - } + }, }, # Age-gated videos { @@ -810,7 +941,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): }, 'expected_warnings': [ 'DASH manifest missing', - ] + ], }, # Olympics (https://github.com/ytdl-org/youtube-dl/issues/4431) { @@ -1454,6 +1585,15 @@ class YoutubeIE(YoutubeBaseInfoExtractor): '397': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'}, } + _PLAYER_JS_VARIANT_MAP = ( + ('main', 'player_ias.vflset/en_US/base.js'), + ('tce', 'player_ias_tce.vflset/en_US/base.js'), + ('tv', 'tv-player-ias.vflset/tv-player-ias.js'), + ('tv_es6', 'tv-player-es6.vflset/tv-player-es6.js'), + ('phone', 'player-plasma-ias-phone-en_US.vflset/base.js'), + ('tablet', 'player-plasma-ias-tablet-en_US.vflset/base.js'), + ) + @classmethod def suitable(cls, url): if parse_qs(url).get('list', [None])[0]: @@ -1493,46 +1633,97 @@ class YoutubeIE(YoutubeBaseInfoExtractor): """ Return a string representation of a signature """ return '.'.join(compat_str(len(part)) for part in example_sig.split('.')) - @classmethod - def _extract_player_info(cls, player_url): - for player_re in cls._PLAYER_INFO_RE: - id_m = re.search(player_re, player_url) - if id_m: - break - else: - raise ExtractorError('Cannot identify player %r' % player_url) - return id_m.group('id') + def _extract_player_info(self, player_url): + try: + return self._search_regex( + self._PLAYER_INFO_RE, player_url, 'player info', group='id') + except ExtractorError as e: + raise ExtractorError( + 'Cannot identify player %r' % (player_url,), cause=e) - def _load_player(self, video_id, player_url, fatal=True, player_id=None): - if not player_id: + def _player_js_cache_key(self, player_url, extra_id=None, _cache={}): + if player_url not in _cache: player_id = self._extract_player_info(player_url) - if player_id not in self._code_cache: + player_path = remove_start( + compat_urllib_parse.urlparse(player_url).path, + '/s/player/{0}/'.format(player_id)) + variant = next((k for k, v in self._PLAYER_JS_VARIANT_MAP + if v == player_path), None) + if not variant: + variant = next( + (k for k, v in self._PLAYER_JS_VARIANT_MAP + if re.match(re.escape(v).replace('en_US', r'\w+') + '$', player_path)), + None) + if not variant: + self.write_debug( + 'Unable to determine player JS variant\n' + ' player = {0}'.format(player_url), only_once=True) + variant = re.sub(r'[^a-zA-Z0-9]', '_', remove_end(player_path, '.js')) + _cache[player_url] = join_nonempty(player_id, variant) + + if extra_id: + extra_id = '-'.join((_cache[player_url], extra_id)) + assert os.path.basename(extra_id) == extra_id + return extra_id + return _cache[player_url] + + def _load_player(self, video_id, player_url, fatal=True): + player_js_key = self._player_js_cache_key(player_url) + if player_js_key not in self._code_cache: code = self._download_webpage( player_url, video_id, fatal=fatal, - note='Downloading player ' + player_id, - errnote='Download of %s failed' % player_url) + note='Downloading player {0}'.format(player_js_key), + errnote='Download of {0} failed'.format(player_url)) if code: - self._code_cache[player_id] = code - return self._code_cache[player_id] if fatal else self._code_cache.get(player_id) + self._code_cache[player_js_key] = code + return self._code_cache.get(player_js_key) + + def _load_player_data_from_cache(self, name, player_url, extra_id=None): + cache_id = ('youtube-{0}'.format(name), self._player_js_cache_key(player_url, extra_id)) + data = self._player_cache.get(cache_id) + if data: + return data + + data = self.cache.load(*cache_id, min_ver='2025.04.07') + if data: + self._player_cache[cache_id] = data + return data + + def _store_player_data_to_cache(self, name, player_url, data, extra_id=None): + cache_id = ('youtube-{0}'.format(name), self._player_js_cache_key(player_url, extra_id)) + + if cache_id not in self._player_cache: + self.cache.store(cache_id[0], cache_id[1], data) + self._player_cache[cache_id] = data + + def _remove_player_data_from_cache(self, name, player_url, extra_id=None): + cache_id = ('youtube-{0}'.format(name), self._player_js_cache_key(player_url, extra_id)) + + if cache_id in self._player_cache: + self.cache.clear(*cache_id) + self._player_cache.pop(cache_id, None) def _extract_signature_function(self, video_id, player_url, example_sig): - player_id = self._extract_player_info(player_url) + # player_id = self._extract_player_info(player_url) # Read from filesystem cache - func_id = 'js_{0}_{1}'.format( - player_id, self._signature_cache_id(example_sig)) - assert os.path.basename(func_id) == func_id - - self.write_debug('Extracting signature function {0}'.format(func_id)) - cache_spec, code = self.cache.load('youtube-sigfuncs', func_id), None + extra_id = self._signature_cache_id(example_sig) + self.write_debug('Extracting signature function {0}-{1}'.format(player_url, extra_id)) + cache_spec, code = self._load_player_data_from_cache( + 'sigfuncs', player_url, extra_id=extra_id), None if not cache_spec: - code = self._load_player(video_id, player_url, player_id) - if code: - res = self._parse_sig_js(code) - test_string = ''.join(map(compat_chr, range(len(example_sig)))) - cache_spec = [ord(c) for c in res(test_string)] - self.cache.store('youtube-sigfuncs', func_id, cache_spec) + code = self._load_player(video_id, player_url) + if code: + res = self._parse_sig_js(code) + test_string = ''.join(map(compat_chr, range(len(example_sig)))) + cache_spec = [ord(c) for c in res(test_string)] + self._store_player_data_to_cache( + 'sigfuncs', player_url, cache_spec, extra_id=extra_id) + else: + self.report_warning( + 'Failed to compute signature function {0}-{1}'.format( + player_url, extra_id)) return lambda s: ''.join(s[i] for i in cache_spec) @@ -1578,26 +1769,54 @@ class YoutubeIE(YoutubeBaseInfoExtractor): ' return %s\n') % (signature_id_tuple, expr_code) self.to_screen('Extracted signature function:\n' + code) + def _extract_sig_fn(self, jsi, funcname): + var_ay = self._search_regex( + r'''(?x) + (?:\*/|\{|\n|^)\s*(?:'[^']+'\s*;\s*) + (var\s*[\w$]+\s*=\s*(?: + ('|")(?:\\\2|(?!\2).)+\2\s*\.\s*split\(\s*('|")\W+\3\s*\)| + \[\s*(?:('|")(?:\\\4|(?!\4).)*\4\s*(?:(?=\])|,\s*))+\] + ))(?=\s*[,;]) + ''', jsi.code, 'useful values', default='') + + sig_fn = jsi.extract_function_code(funcname) + + if var_ay: + sig_fn = (sig_fn[0], ';\n'.join((var_ay, sig_fn[1]))) + + return sig_fn + def _parse_sig_js(self, jscode): + # Examples where `sig` is funcname: + # sig=function(a){a=a.split(""); ... ;return a.join("")}; + # ;c&&(c=sig(decodeURIComponent(c)),a.set(b,encodeURIComponent(c)));return a}; + # {var l=f,m=h.sp,n=sig(decodeURIComponent(h.s));l.set(m,encodeURIComponent(n))} + # sig=function(J){J=J.split(""); ... ;return J.join("")}; + # ;N&&(N=sig(decodeURIComponent(N)),J.set(R,encodeURIComponent(N)));return J}; + # {var H=u,k=f.sp,v=sig(decodeURIComponent(f.s));H.set(k,encodeURIComponent(v))} funcname = self._search_regex( - (r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(', - r'\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(', - r'\bm=(?P<sig>[a-zA-Z0-9$]{2,})\(decodeURIComponent\(h\.s\)\)', - r'\bc&&\(c=(?P<sig>[a-zA-Z0-9$]{2,})\(decodeURIComponent\(c\)\)', - r'(?:\b|[^a-zA-Z0-9$])(?P<sig>[a-zA-Z0-9$]{2,})\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)(?:;[a-zA-Z0-9$]{2}\.[a-zA-Z0-9$]{2}\(a,\d+\))?', - r'(?P<sig>[a-zA-Z0-9$]+)\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)', + (r'\b(?P<var>[\w$]+)&&\((?P=var)=(?P<sig>[\w$]{2,})\(decodeURIComponent\((?P=var)\)\)', + r'(?P<sig>[\w$]+)\s*=\s*function\(\s*(?P<arg>[\w$]+)\s*\)\s*{\s*(?P=arg)\s*=\s*(?P=arg)\.split\(\s*""\s*\)\s*;\s*[^}]+;\s*return\s+(?P=arg)\.join\(\s*""\s*\)', + r'(?:\b|[^\w$])(?P<sig>[\w$]{2,})\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)(?:;[\w$]{2}\.[\w$]{2}\(a,\d+\))?', + # Old patterns + r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[\w$]+)\(', + r'\b[\w]+\s*&&\s*[\w]+\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[\w$]+)\(', + r'\bm=(?P<sig>[\w$]{2,})\(decodeURIComponent\(h\.s\)\)', # Obsolete patterns - r'("|\')signature\1\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(', - r'\.sig\|\|(?P<sig>[a-zA-Z0-9$]+)\(', - r'yt\.akamaized\.net/\)\s*\|\|\s*.*?\s*[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?:encodeURIComponent\s*\()?\s*(?P<sig>[a-zA-Z0-9$]+)\(', - r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(', - r'\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(', - r'\bc\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\('), + r'("|\')signature\1\s*,\s*(?P<sig>[\w$]+)\(', + r'\.sig\|\|(?P<sig>[\w$]+)\(', + r'yt\.akamaized\.net/\)\s*\|\|\s*.*?\s*[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?:encodeURIComponent\s*\()?\s*(?P<sig>[\w$]+)\(', + r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?P<sig>[\w$]+)\(', + r'\bc\s*&&\s*[\w]+\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[\w$]+)\('), jscode, 'Initial JS player signature function name', group='sig') jsi = JSInterpreter(jscode) - initial_function = jsi.extract_function(funcname) - return lambda s: initial_function([s]) + + initial_function = self._extract_sig_fn(jsi, funcname) + + func = jsi.extract_function_from_code(*initial_function) + + return lambda s: func([s]) def _cached(self, func, *cache_id): def inner(*args, **kwargs): @@ -1636,7 +1855,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): try: jsi, player_id, func_code = self._extract_n_function_code(video_id, player_url) except ExtractorError as e: - raise ExtractorError('Unable to extract nsig jsi, player_id, func_codefunction code', cause=e) + raise ExtractorError('Unable to extract nsig function code', cause=e) if self.get_param('youtube_print_sig_code'): self.to_screen('Extracted nsig function from {0}:\n{1}\n'.format( player_id, func_code[1])) @@ -1647,7 +1866,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): except JSInterpreter.Exception as e: self.report_warning( '%s (%s %s)' % ( - 'Unable to decode n-parameter: download likely to be throttled', + 'Unable to decode n-parameter: expect download to be blocked or throttled', error_to_compat_str(e), traceback.format_exc()), video_id=video_id) @@ -1657,41 +1876,103 @@ class YoutubeIE(YoutubeBaseInfoExtractor): return ret def _extract_n_function_name(self, jscode): + func_name, idx = None, None + + def generic_n_function_search(func_name=None): + return self._search_regex( + r'''(?xs) + (?:(?<=[^\w$])|^) # instead of \b, which ignores $ + (?P<name>%s)\s*=\s*function\((?!\d)[a-zA-Z\d_$]+\) + \s*\{(?:(?!};).)+?(?: + ["']enhanced_except_ | + return\s*(?P<q>"|')[a-zA-Z\d-]+_w8_(?P=q)\s*\+\s*[\w$]+ + ) + ''' % (func_name or r'(?!\d)[a-zA-Z\d_$]+',), jscode, + 'Initial JS player n function name', group='name', + default=None if func_name else NO_DEFAULT) + + # these special cases are redundant and probably obsolete (2025-04): + # they make the tests run ~10% faster without fallback warnings + r""" func_name, idx = self._search_regex( - r'\.get\("n"\)\)&&\(b=(?P<nfunc>[a-zA-Z_$][\w$]*)(?:\[(?P<idx>\d+)\])?\([\w$]+\)', - jscode, 'Initial JS player n function name', group=('nfunc', 'idx')) + # (y=NuD(),Mw(k),q=k.Z[y]||null)&&(q=narray[idx](q),k.set(y,q),k.V||NuD(''))}}; + # (R="nn"[+J.Z],mW(J),N=J.K[R]||null)&&(N=narray[idx](N),J.set(R,N))}}; + # or: (b=String.fromCharCode(110),c=a.get(b))&&c=narray[idx](c) + # or: (b="nn"[+a.D],c=a.get(b))&&(c=narray[idx](c) + # or: (PL(a),b=a.j.n||null)&&(b=narray[idx](b) + # or: (b="nn"[+a.D],vL(a),c=a.j[b]||null)&&(c=narray[idx](c),a.set(b,c),narray.length||nfunc("") + # old: (b=a.get("n"))&&(b=narray[idx](b)(?P<c>[a-z])\s*=\s*[a-z]\s* + # older: (b=a.get("n"))&&(b=nfunc(b) + r'''(?x) + # (expr, ..., + \((?:(?:\s*[\w$]+\s*=)?(?:[\w$"+\.\s(\[]+(?:[)\]]\s*)?),)* + # b=... + (?P<b>[\w$]+)\s*=\s*(?!(?P=b)[^\w$])[\w$]+\s*(?:(?: + \.\s*[\w$]+ | + \[\s*[\w$]+\s*\] | + \.\s*get\s*\(\s*[\w$"]+\s*\) + )\s*){,2}(?:\s*\|\|\s*null(?=\s*\)))?\s* + \)\s*&&\s*\( # ...)&&( + # b = nfunc, b = narray[idx] + (?P=b)\s*=\s*(?P<nfunc>[\w$]+)\s* + (?:\[\s*(?P<idx>[\w$]+)\s*\]\s*)? + # (...) + \(\s*[\w$]+\s*\) + ''', jscode, 'Initial JS player n function name', group=('nfunc', 'idx'), + default=(None, None)) + """ + + if not func_name: + # nfunc=function(x){...}|function nfunc(x); ... + # ... var y=[nfunc]|y[idx]=nfunc); + # obvious REs hang, so use a two-stage tactic + for m in re.finditer(r'''(?x) + [\n;]var\s(?:(?:(?!,).)+,|\s)*?(?!\d)[\w$]+(?:\[(?P<idx>\d+)\])?\s*=\s* + (?(idx)|\[\s*)(?P<nfunc>(?!\d)[\w$]+)(?(idx)|\s*\]) + \s*?[;\n] + ''', jscode): + fn = self._search_regex( + r'[;,]\s*(function\s+)?({0})(?(1)|\s*=\s*function)\s*\((?!\d)[\w$]+\)\s*\{1}(?!\s*return\s)'.format( + re.escape(m.group('nfunc')), '{'), + jscode, 'Initial JS player n function name (2)', group=2, default=None) + if fn: + func_name = fn + idx = m.group('idx') + if generic_n_function_search(func_name): + # don't look any further + break + + # thx bashonly: yt-dlp/yt-dlp/pull/10611 + if not func_name: + self.report_warning('Falling back to generic n function search', only_once=True) + return generic_n_function_search() + if not idx: return func_name - return self._parse_json(self._search_regex( - r'var {0}\s*=\s*(\[.+?\])\s*[,;]'.format(re.escape(func_name)), jscode, - 'Initial JS player n function list ({0}.{1})'.format(func_name, idx)), - func_name, transform_source=js_to_json)[int(idx)] + return self._search_json( + r'(?<![\w-])var\s(?:(?:(?!,).)+,|\s)*?{0}\s*='.format(re.escape(func_name)), jscode, + 'Initial JS player n function list ({0}.{1})'.format(func_name, idx), + func_name, contains_pattern=r'\[.+\]', end_pattern='[,;]', + transform_source=js_to_json)[int(idx)] def _extract_n_function_code(self, video_id, player_url): player_id = self._extract_player_info(player_url) - func_code = self.cache.load('youtube-nsig', player_id) + func_code = self._load_player_data_from_cache('nsig', player_url) jscode = func_code or self._load_player(video_id, player_url) jsi = JSInterpreter(jscode) if func_code: return jsi, player_id, func_code - func_name = self._extract_n_function_name(jscode) + return self._extract_n_function_code_jsi(video_id, jsi, player_id, player_url) - # For redundancy - func_code = self._search_regex( - r'''(?xs)%s\s*=\s*function\s*\((?P<var>[\w$]+)\)\s* - # NB: The end of the regex is intentionally kept strict - {(?P<code>.+?}\s*return\ [\w$]+.join\(""\))};''' % func_name, - jscode, 'nsig function', group=('var', 'code'), default=None) - if func_code: - func_code = ([func_code[0]], func_code[1]) - else: - self.write_debug('Extracting nsig function with jsinterp') - func_code = jsi.extract_function_code(func_name) + def _extract_n_function_code_jsi(self, video_id, jsi, player_id=None, player_url=None): + func_name = self._extract_n_function_name(jsi.code) - self.cache.store('youtube-nsig', player_id, func_code) + func_code = self._extract_sig_fn(jsi, func_name) + if player_url: + self._store_player_data_to_cache('nsig', player_url, func_code) return jsi, player_id, func_code def _extract_n_function_from_code(self, jsi, func_code): @@ -1699,13 +1980,13 @@ class YoutubeIE(YoutubeBaseInfoExtractor): def extract_nsig(s): try: - ret = func([s]) + ret = func([s], kwargs={'_ytdl_do_not_return': s}) except JSInterpreter.Exception: raise except Exception as e: raise JSInterpreter.Exception(traceback.format_exc(), cause=e) - if ret.startswith('enhanced_except_'): + if ret.startswith('enhanced_except_') or ret.endswith(s): raise JSInterpreter.Exception('Signature function returned an exception') return ret @@ -1724,7 +2005,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor): n_param = n_param[-1] n_response = decrypt_nsig(n_param)(n_param, video_id, player_url) if n_response is None: - # give up if descrambling failed + # give up and forget cached data if descrambling failed + self._remove_player_data_from_cache('nsig', player_url) break fmt['url'] = update_url_query(fmt['url'], {'n': n_response}) @@ -1735,18 +2017,28 @@ class YoutubeIE(YoutubeBaseInfoExtractor): Required to tell API what sig/player version is in use. """ sts = traverse_obj(ytcfg, 'STS', expected_type=int) - if not sts: - # Attempt to extract from player - if player_url is None: - error_msg = 'Cannot extract signature timestamp without player_url.' - if fatal: - raise ExtractorError(error_msg) - self.report_warning(error_msg) - return - code = self._load_player(video_id, player_url, fatal=fatal) - sts = int_or_none(self._search_regex( - r'(?:signatureTimestamp|sts)\s*:\s*(?P<sts>[0-9]{5})', code or '', - 'JS player signature timestamp', group='sts', fatal=fatal)) + if sts: + return sts + + if not player_url: + error_msg = 'Cannot extract signature timestamp without player url' + if fatal: + raise ExtractorError(error_msg) + self.report_warning(error_msg) + return None + + sts = self._load_player_data_from_cache('sts', player_url) + if sts: + return sts + + # Attempt to extract from player + code = self._load_player(video_id, player_url, fatal=fatal) + sts = int_or_none(self._search_regex( + r'(?:signatureTimestamp|sts)\s*:\s*(?P<sts>[0-9]{5})', code or '', + 'JS player signature timestamp', group='sts', fatal=fatal)) + if sts: + self._store_player_data_to_cache('sts', player_url, sts) + return sts def _mark_watched(self, video_id, player_response): @@ -1758,8 +2050,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor): # cpn generation algorithm is reverse engineered from base.js. # In fact it works even with dummy cpn. - CPN_ALPHABET = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_' - cpn = ''.join(CPN_ALPHABET[random.randint(0, 256) & 63] for _ in range(0, 16)) + CPN_ALPHABET = string.ascii_letters + string.digits + '-_' + cpn = ''.join(CPN_ALPHABET[random.randint(0, 256) & 63] for _ in range(16)) # more consistent results setting it to right before the end qs = parse_qs(playback_url) @@ -1819,8 +2111,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): mobj = re.match(cls._VALID_URL, url, re.VERBOSE) if mobj is None: raise ExtractorError('Invalid URL: %s' % url) - video_id = mobj.group(2) - return video_id + return mobj.group(2) def _extract_chapters_from_json(self, data, video_id, duration): chapters_list = try_get( @@ -1881,9 +2172,89 @@ class YoutubeIE(YoutubeBaseInfoExtractor): player_response = self._extract_yt_initial_variable( webpage, self._YT_INITIAL_PLAYER_RESPONSE_RE, video_id, 'initial player response') - if not player_response: + is_live = traverse_obj(player_response, ('videoDetails', 'isLive')) + + if False and not player_response: player_response = self._call_api( 'player', {'videoId': video_id}, video_id) + if True or not player_response: + origin = 'https://www.youtube.com' + pb_context = {'html5Preference': 'HTML5_PREF_WANTS'} + + player_url = self._extract_player_url(webpage) + ytcfg = self._extract_ytcfg(video_id, webpage or '') + sts = self._extract_signature_timestamp(video_id, player_url, ytcfg) + if sts: + pb_context['signatureTimestamp'] = sts + + client_names = traverse_obj(self._INNERTUBE_CLIENTS, ( + T(dict.items), lambda _, k_v: not k_v[1].get('REQUIRE_PO_TOKEN'), + 0))[:1] + if 'web' not in client_names: + # webpage links won't download: ignore links and playability + player_response = filter_dict( + player_response or {}, + lambda k, _: k not in ('streamingData', 'playabilityStatus')) + + if is_live and 'ios' not in client_names: + client_names.append('ios') + + headers = { + 'Sec-Fetch-Mode': 'navigate', + 'Origin': origin, + 'X-Goog-Visitor-Id': self._extract_visitor_data(ytcfg) or '', + } + auth = self._generate_sapisidhash_header(origin) + if auth is not None: + headers['Authorization'] = auth + headers['X-Origin'] = origin + + for client in traverse_obj(self._INNERTUBE_CLIENTS, (client_names, T(dict))): + + query = { + 'playbackContext': { + 'contentPlaybackContext': pb_context, + }, + 'contentCheckOk': True, + 'racyCheckOk': True, + 'context': { + 'client': merge_dicts( + traverse_obj(client, ('INNERTUBE_CONTEXT', 'client')), { + 'hl': 'en', + 'timeZone': 'UTC', + 'utcOffsetMinutes': 0, + }), + }, + 'videoId': video_id, + } + + api_headers = merge_dicts(headers, traverse_obj(client, { + 'X-YouTube-Client-Name': 'INNERTUBE_CONTEXT_CLIENT_NAME', + 'X-YouTube-Client-Version': ( + 'INNERTUBE_CONTEXT', 'client', 'clientVersion'), + 'User-Agent': ( + 'INNERTUBE_CONTEXT', 'client', 'userAgent'), + })) + + api_player_response = self._call_api( + 'player', query, video_id, fatal=False, headers=api_headers, + note=join_nonempty( + 'Downloading', traverse_obj(query, ( + 'context', 'client', 'clientName')), + 'API JSON', delim=' ')) + + hls = traverse_obj( + (player_response, api_player_response), + (Ellipsis, 'streamingData', 'hlsManifestUrl', T(url_or_none))) + if len(hls) == 2 and not hls[0] and hls[1]: + player_response['streamingData']['hlsManifestUrl'] = hls[1] + else: + video_details = merge_dicts(*traverse_obj( + (player_response, api_player_response), + (Ellipsis, 'videoDetails', T(dict)))) + player_response.update(filter_dict( + api_player_response or {}, cndn=lambda k, _: k != 'captions')) + player_response['videoDetails'] = video_details def is_agegated(playability): if not isinstance(playability, dict): @@ -1932,7 +2303,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): headers = { 'X-YouTube-Client-Name': '85', 'X-YouTube-Client-Version': '2.0', - 'Origin': 'https://www.youtube.com' + 'Origin': 'https://www.youtube.com', } video_info = self._call_api('player', query, video_id, fatal=False, headers=headers) @@ -1961,8 +2332,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor): return ''.join([r['text'] for r in runs if isinstance(r.get('text'), compat_str)]) search_meta = ( - lambda x: self._html_search_meta(x, webpage, default=None)) \ - if webpage else lambda x: None + (lambda x: self._html_search_meta(x, webpage, default=None)) + if webpage else lambda _: None) video_details = player_response.get('videoDetails') or {} microformat = try_get( @@ -2026,6 +2397,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): itag_qualities = {} q = qualities(['tiny', 'small', 'medium', 'large', 'hd720', 'hd1080', 'hd1440', 'hd2160', 'hd2880', 'highres']) CHUNK_SIZE = 10 << 20 + is_live = video_details.get('isLive') streaming_data = player_response.get('streamingData') or {} streaming_formats = streaming_data.get('formats') or [] @@ -2034,7 +2406,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): def build_fragments(f): return LazyList({ 'url': update_url_query(f['url'], { - 'range': '{0}-{1}'.format(range_start, min(range_start + CHUNK_SIZE - 1, f['filesize'])) + 'range': '{0}-{1}'.format(range_start, min(range_start + CHUNK_SIZE - 1, f['filesize'])), }) } for range_start in range(0, f['filesize'], CHUNK_SIZE)) @@ -2133,7 +2505,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'protocol': 'http_dash_segments', 'fragments': build_fragments(dct), } if dct['filesize'] else { - 'downloader_options': {'http_chunk_size': CHUNK_SIZE} # No longer useful? + 'downloader_options': {'http_chunk_size': CHUNK_SIZE}, # No longer useful? }) formats.append(dct) @@ -2170,7 +2542,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor): hls_manifest_url = streaming_data.get('hlsManifestUrl') if hls_manifest_url: for f in self._extract_m3u8_formats( - hls_manifest_url, video_id, 'mp4', fatal=False): + hls_manifest_url, video_id, 'mp4', + entry_protocol='m3u8_native', live=is_live, fatal=False): if process_manifest_format( f, 'hls', None, self._search_regex( r'/itag/(\d+)', f['url'], 'itag', default=None)): @@ -2190,12 +2563,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor): formats.append(f) playable_formats = [f for f in formats if not f.get('has_drm')] - if formats and not playable_formats: - # If there are no formats that definitely don't have DRM, all have DRM - self.report_drm(video_id) - formats[:] = playable_formats - - if not formats: + if formats: + if not playable_formats: + # If there are no formats that definitely don't have DRM, all have DRM + self.report_drm(video_id) + formats[:] = playable_formats + else: if streaming_data.get('licenseInfos'): raise ExtractorError( 'This video is DRM protected.', expected=True) @@ -2276,8 +2649,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor): # Strictly de-prioritize damaged formats f['preference'] = -10 - is_live = video_details.get('isLive') - owner_profile_url = self._yt_urljoin(self._extract_author_var( webpage, 'url', videodetails=video_details, metadata=microformat)) @@ -2311,9 +2682,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'is_live': is_live, } - pctr = try_get( - player_response, - lambda x: x['captions']['playerCaptionsTracklistRenderer'], dict) + pctr = traverse_obj( + (player_response, api_player_response), + (Ellipsis, 'captions', 'playerCaptionsTracklistRenderer', T(dict))) if pctr: def process_language(container, base_url, lang_code, query): lang_subs = [] @@ -2327,31 +2698,35 @@ class YoutubeIE(YoutubeBaseInfoExtractor): }) container[lang_code] = lang_subs - subtitles = {} - for caption_track in (pctr.get('captionTracks') or []): - base_url = caption_track.get('baseUrl') - if not base_url: - continue - if caption_track.get('kind') != 'asr': - lang_code = caption_track.get('languageCode') - if not lang_code: + def process_subtitles(): + subtitles = {} + for caption_track in traverse_obj(pctr, ( + Ellipsis, 'captionTracks', lambda _, v: ( + v.get('baseUrl') and v.get('languageCode')))): + base_url = self._yt_urljoin(caption_track['baseUrl']) + if not base_url: continue - process_language( - subtitles, base_url, lang_code, {}) - continue - automatic_captions = {} - for translation_language in (pctr.get('translationLanguages') or []): - translation_language_code = translation_language.get('languageCode') - if not translation_language_code: + lang_code = caption_track['languageCode'] + if caption_track.get('kind') != 'asr': + process_language( + subtitles, base_url, lang_code, {}) continue + automatic_captions = {} process_language( - automatic_captions, base_url, translation_language_code, - {'tlang': translation_language_code}) - info['automatic_captions'] = automatic_captions - info['subtitles'] = subtitles + automatic_captions, base_url, lang_code, {}) + for translation_language in traverse_obj(pctr, ( + Ellipsis, 'translationLanguages', lambda _, v: v.get('languageCode'))): + translation_language_code = translation_language['languageCode'] + process_language( + automatic_captions, base_url, translation_language_code, + {'tlang': translation_language_code}) + info['automatic_captions'] = automatic_captions + info['subtitles'] = subtitles + + process_subtitles() parsed_url = compat_urllib_parse_urlparse(url) - for component in [parsed_url.fragment, parsed_url.query]: + for component in (parsed_url.fragment, parsed_url.query): query = compat_parse_qs(component) for k, v in query.items(): for d_k, s_ks in [('start', ('start', 't')), ('end', ('end',))]: @@ -2581,7 +2956,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): 'title': 'Super Cooper Shorts - Shorts', 'uploader': 'Super Cooper Shorts', 'uploader_id': '@SuperCooperShorts', - } + }, }, { # Channel that does not have a Shorts tab. Test should just download videos on Home tab instead 'url': 'https://www.youtube.com/@emergencyawesome/shorts', @@ -2635,7 +3010,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): 'description': 'md5:609399d937ea957b0f53cbffb747a14c', 'uploader': 'ThirstForScience', 'uploader_id': '@ThirstForScience', - } + }, }, { 'url': 'https://www.youtube.com/c/ChristophLaimer/playlists', 'only_matching': True, @@ -2934,7 +3309,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): 'uploader': '3Blue1Brown', 'uploader_id': '@3blue1brown', 'channel_id': 'UCYO_jab_esuFRV4b17AJtAw', - } + }, }] @classmethod @@ -2959,8 +3334,12 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): expected_type=txt_or_none) def _grid_entries(self, grid_renderer): - for item in grid_renderer['items']: - if not isinstance(item, dict): + for item in traverse_obj(grid_renderer, ('items', Ellipsis, T(dict))): + lockup_view_model = traverse_obj(item, ('lockupViewModel', T(dict))) + if lockup_view_model: + entry = self._extract_lockup_view_model(lockup_view_model) + if entry: + yield entry continue renderer = self._extract_grid_item_renderer(item) if not isinstance(renderer, dict): @@ -3044,6 +3423,39 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): continue yield self._extract_video(renderer) + def _extract_lockup_view_model(self, view_model): + content_id = view_model.get('contentId') + if not content_id: + return + content_type = view_model.get('contentType') + if content_type not in ('LOCKUP_CONTENT_TYPE_PLAYLIST', 'LOCKUP_CONTENT_TYPE_PODCAST'): + self.report_warning( + 'Unsupported lockup view model content type "{0}"{1}'.format(content_type, bug_reports_message()), only_once=True) + return + return merge_dicts(self.url_result( + update_url_query('https://www.youtube.com/playlist', {'list': content_id}), + ie=YoutubeTabIE.ie_key(), video_id=content_id), { + 'title': traverse_obj(view_model, ( + 'metadata', 'lockupMetadataViewModel', 'title', 'content', T(compat_str))), + 'thumbnails': self._extract_thumbnails(view_model, ( + 'contentImage', 'collectionThumbnailViewModel', 'primaryThumbnail', + 'thumbnailViewModel', 'image'), final_key='sources'), + }) + + def _extract_shorts_lockup_view_model(self, view_model): + content_id = traverse_obj(view_model, ( + 'onTap', 'innertubeCommand', 'reelWatchEndpoint', 'videoId', + T(lambda v: v if YoutubeIE.suitable(v) else None))) + if not content_id: + return + return merge_dicts(self.url_result( + content_id, ie=YoutubeIE.ie_key(), video_id=content_id), { + 'title': traverse_obj(view_model, ( + 'overlayMetadata', 'primaryText', 'content', T(compat_str))), + 'thumbnails': self._extract_thumbnails( + view_model, 'thumbnail', final_key='sources'), + }) + def _video_entry(self, video_renderer): video_id = video_renderer.get('videoId') if video_id: @@ -3090,10 +3502,9 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): yield entry def _rich_grid_entries(self, contents): - for content in contents: - content = traverse_obj( - content, ('richItemRenderer', 'content'), - expected_type=dict) or {} + for content in traverse_obj( + contents, (Ellipsis, 'richItemRenderer', 'content'), + expected_type=dict): video_renderer = traverse_obj( content, 'videoRenderer', 'reelItemRenderer', expected_type=dict) @@ -3101,6 +3512,12 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): entry = self._video_entry(video_renderer) if entry: yield entry + # shorts item + shorts_lockup_view_model = content.get('shortsLockupViewModel') + if shorts_lockup_view_model: + entry = self._extract_shorts_lockup_view_model(shorts_lockup_view_model) + if entry: + yield entry # playlist renderer = traverse_obj( content, 'playlistRenderer', expected_type=dict) or {} @@ -3139,23 +3556,15 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): 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'], - dict) - if not continuation_ep: - continue - continuation = try_get( - continuation_ep, lambda x: x['continuationCommand']['token'], compat_str) + for command in traverse_obj(renderer, ( + ('contents', 'items', 'rows'), Ellipsis, 'continuationItemRenderer', + ('continuationEndpoint', ('button', 'buttonRenderer', 'command')), + (('commandExecutorCommand', 'commands', Ellipsis), None), T(dict))): + continuation = traverse_obj(command, ('continuationCommand', 'token', T(compat_str))) if not continuation: continue - ctp = continuation_ep.get('clickTrackingParams') - return YoutubeTabIE._build_continuation_query(continuation, ctp) + ctp = command.get('clickTrackingParams') + return cls._build_continuation_query(continuation, ctp) def _entries(self, tab, item_id, webpage): tab_content = try_get(tab, lambda x: x['content'], dict) @@ -3204,6 +3613,13 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): entry = self._video_entry(renderer) if entry: yield entry + renderer = isr_content.get('richGridRenderer') + if renderer: + for from_ in self._rich_grid_entries( + traverse_obj(renderer, ('contents', Ellipsis, T(dict)))): + yield from_ + continuation = self._extract_continuation(renderer) + continue if not continuation: continuation = self._extract_continuation(is_renderer) @@ -3213,8 +3629,9 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): rich_grid_renderer = tab_content.get('richGridRenderer') if not rich_grid_renderer: return - for entry in self._rich_grid_entries(rich_grid_renderer.get('contents') or []): - yield entry + for from_ in self._rich_grid_entries( + traverse_obj(rich_grid_renderer, ('contents', Ellipsis, T(dict)))): + yield from_ continuation = self._extract_continuation(rich_grid_renderer) @@ -3232,7 +3649,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): 'client': { 'clientName': 'WEB', 'clientVersion': client_version, - } + }, } visitor_data = try_get(context, lambda x: x['client']['visitorData'], compat_str) @@ -3248,10 +3665,10 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): if not continuation: break if visitor_data: - headers['x-goog-visitor-id'] = visitor_data + headers['X-Goog-Visitor-Id'] = visitor_data data['continuation'] = continuation['continuation'] data['clickTracking'] = { - 'clickTrackingParams': continuation['itct'] + 'clickTrackingParams': continuation['itct'], } count = 0 retries = 3 @@ -3260,8 +3677,12 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): # Downloading page may result in intermittent 5xx HTTP error # that is usually worked around with a retry response = self._download_json( - 'https://www.youtube.com/youtubei/v1/browse?key=AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8', + 'https://www.youtube.com/youtubei/v1/browse', None, 'Downloading page %d%s' % (page_num, ' (retry #%d)' % count if count else ''), + query={ + # 'key': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8', + 'prettyPrint': 'false', + }, headers=headers, data=json.dumps(data).encode('utf8')) break except ExtractorError as e: @@ -3430,10 +3851,23 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): def _real_extract(self, url): item_id = self._match_id(url) url = update_url(url, netloc='www.youtube.com') - # Handle both video/playlist URLs qs = parse_qs(url) - video_id = qs.get('v', [None])[0] - playlist_id = qs.get('list', [None])[0] + + def qs_get(key, default=None): + return qs.get(key, [default])[-1] + + # Go around for /feeds/videos.xml?playlist_id={pl_id} + if item_id == 'feeds' and '/feeds/videos.xml?' in url: + playlist_id = qs_get('playlist_id') + if playlist_id: + return self.url_result( + update_url_query('https://www.youtube.com/playlist', { + 'list': playlist_id, + }), ie=self.ie_key(), video_id=playlist_id) + + # Handle both video/playlist URLs + video_id = qs_get('v') + playlist_id = qs_get('list') 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) @@ -3510,7 +3944,7 @@ class YoutubePlaylistIE(InfoExtractor): 'uploader': 'milan', 'uploader_id': '@milan5503', 'channel_id': 'UCEI1-PVPcYXjB73Hfelbmaw', - } + }, }, { 'url': 'http://www.youtube.com/embed/_xDOZElKyNU?list=PLsyOSbh5bs16vubvKePAQ1x3PhKavfBIl', 'playlist_mincount': 455, @@ -3520,7 +3954,7 @@ class YoutubePlaylistIE(InfoExtractor): 'uploader': 'LBK', 'uploader_id': '@music_king', 'channel_id': 'UC21nz3_MesPLqtDqwdvnoxA', - } + }, }, { 'url': 'TLGGrESM50VT6acwMjAyMjAxNw', 'only_matching': True, @@ -3631,7 +4065,7 @@ class YoutubeSearchIE(SearchInfoExtractor, YoutubeBaseInfoExtractor): 'info_dict': { 'id': 'youtube-dl test video', 'title': 'youtube-dl test video', - } + }, }] def _get_n_results(self, query, n): @@ -3651,7 +4085,7 @@ class YoutubeSearchDateIE(YoutubeSearchIE): 'info_dict': { 'id': 'youtube-dl test video', 'title': 'youtube-dl test video', - } + }, }] @@ -3666,7 +4100,7 @@ class YoutubeSearchURLIE(YoutubeBaseInfoExtractor): 'id': 'youtube-dl test video', 'title': 'youtube-dl test video', }, - 'params': {'playlistend': 5} + 'params': {'playlistend': 5}, }, { 'url': 'https://www.youtube.com/results?q=test&sp=EgQIBBgB', 'only_matching': True, @@ -3682,6 +4116,7 @@ class YoutubeSearchURLIE(YoutubeBaseInfoExtractor): class YoutubeFeedsInfoExtractor(YoutubeTabIE): """ Base class for feed extractors + Subclasses must define the _FEED_NAME property. """ _LOGIN_REQUIRED = True |