X-Git-Url: http://git.bitcoin.ninja/index.cgi?p=youtube-dl;a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fyoutube.py;h=29773877e8c44ef479ce39697755c815ed739c44;hp=480f403dadf91a355a6e91fac8c46945a04d1dc1;hb=c3e543893bfba7faa7c13e53fbe6b60f936b81f1;hpb=c89267d31ad99eb5b1a87cd354de5280a2a087b1 diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index 480f403da..29773877e 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -16,6 +16,7 @@ from ..jsinterp import JSInterpreter from ..swfinterp import SWFInterpreter from ..compat import ( compat_chr, + compat_kwargs, compat_parse_qs, compat_urllib_parse_unquote, compat_urllib_parse_unquote_plus, @@ -36,16 +37,18 @@ from ..utils import ( orderedSet, parse_codecs, parse_duration, + qualities, remove_quotes, remove_start, - sanitized_Request, smuggle_url, + str_or_none, str_to_int, try_get, unescapeHTML, unified_strdate, unsmuggle_url, uppercase_escape, + url_or_none, urlencode_postdata, ) @@ -54,12 +57,16 @@ class YoutubeBaseInfoExtractor(InfoExtractor): """Provide base functions for Youtube extractors""" _LOGIN_URL = 'https://accounts.google.com/ServiceLogin' _TWOFACTOR_URL = 'https://accounts.google.com/signin/challenge' - _PASSWORD_CHALLENGE_URL = 'https://accounts.google.com/signin/challenge/sl/password' + + _LOOKUP_URL = 'https://accounts.google.com/_/signin/sl/lookup' + _CHALLENGE_URL = 'https://accounts.google.com/_/signin/sl/challenge' + _TFA_URL = 'https://accounts.google.com/_/signin/challenge?hl=en&TL={0}' + _NETRC_MACHINE = 'youtube' # If True it will raise an error if no login info is provided _LOGIN_REQUIRED = False - _PLAYLIST_ID_RE = r'(?:PL|LL|EC|UU|FL|RD|UL|TL)[0-9A-Za-z-_]{10,}' + _PLAYLIST_ID_RE = r'(?:PL|LL|EC|UU|FL|RD|UL|TL|OLAK5uy_)[0-9A-Za-z-_]{10,}' def _set_language(self): self._set_cookie( @@ -80,10 +87,10 @@ class YoutubeBaseInfoExtractor(InfoExtractor): If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised. """ - (username, password) = self._get_login_info() + username, password = self._get_login_info() # No authentication to be performed if username is None: - if self._LOGIN_REQUIRED: + if self._LOGIN_REQUIRED and self._downloader.params.get('cookiefile') is None: raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True) return True @@ -96,74 +103,170 @@ class YoutubeBaseInfoExtractor(InfoExtractor): login_form = self._hidden_inputs(login_page) - login_form.update({ - 'checkConnection': 'youtube', - 'Email': username, - 'Passwd': password, - }) + def req(url, f_req, note, errnote): + data = login_form.copy() + data.update({ + 'pstMsg': 1, + 'checkConnection': 'youtube', + 'checkedDomains': 'youtube', + 'hl': 'en', + 'deviceinfo': '[null,null,null,[],null,"US",null,null,[],"GlifWebSignIn",null,[null,null,[]]]', + 'f.req': json.dumps(f_req), + 'flowName': 'GlifWebSignIn', + 'flowEntry': 'ServiceLogin', + }) + return self._download_json( + url, None, note=note, errnote=errnote, + transform_source=lambda s: re.sub(r'^[^[]*', '', s), + fatal=False, + data=urlencode_postdata(data), headers={ + 'Content-Type': 'application/x-www-form-urlencoded;charset=utf-8', + 'Google-Accounts-XSRF': 1, + }) - login_results = self._download_webpage( - self._PASSWORD_CHALLENGE_URL, None, - note='Logging in', errnote='unable to log in', fatal=False, - data=urlencode_postdata(login_form)) - if login_results is False: - return False + def warn(message): + self._downloader.report_warning(message) + + lookup_req = [ + username, + None, [], None, 'US', None, None, 2, False, 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 + ], + username, + ] - error_msg = self._html_search_regex( - r'<[^>]+id="errormsg_0_Passwd"[^>]*>([^<]+)<', - login_results, 'error message', default=None) - if error_msg: - raise ExtractorError('Unable to login: %s' % error_msg, expected=True) + lookup_results = req( + self._LOOKUP_URL, lookup_req, + 'Looking up account info', 'Unable to look up account info') - if re.search(r'id="errormsg_0_Passwd"', login_results) is not None: - raise ExtractorError('Please use your account password and a two-factor code instead of an application-specific password.', expected=True) + if lookup_results is False: + return False - # Two-Factor - # TODO add SMS and phone call support - these require making a request and then prompting the user + user_hash = try_get(lookup_results, lambda x: x[0][2], compat_str) + if not user_hash: + warn('Unable to extract user hash') + return False - if re.search(r'(?i)]+id="challenge"', login_results) is not None: - tfa_code = self._get_tfa_info('2-step verification code') + challenge_req = [ + user_hash, + 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 + ]] - if not tfa_code: - self._downloader.report_warning( - 'Two-factor authentication required. Provide it either interactively or with --twofactor ' - '(Note that only TOTP (Google Authenticator App) codes work at this time.)') - return False + challenge_results = req( + self._CHALLENGE_URL, challenge_req, + 'Logging in', 'Unable to log in') - tfa_code = remove_start(tfa_code, 'G-') + if challenge_results is False: + return - tfa_form_strs = self._form_hidden_inputs('challenge', login_results) + login_res = try_get(challenge_results, lambda x: x[0][5], list) + if login_res: + login_msg = try_get(login_res, lambda x: x[5], compat_str) + warn( + 'Unable to login: %s' % 'Invalid password' + if login_msg == 'INCORRECT_ANSWER_ENTERED' else login_msg) + return False - tfa_form_strs.update({ - 'Pin': tfa_code, - 'TrustDevice': 'on', - }) + res = try_get(challenge_results, lambda x: x[0][-1], list) + if not res: + warn('Unable to extract result entry') + return False - tfa_data = urlencode_postdata(tfa_form_strs) + login_challenge = try_get(res, lambda x: x[0][0], list) + if login_challenge: + challenge_str = try_get(login_challenge, lambda x: x[2], compat_str) + if challenge_str == 'TWO_STEP_VERIFICATION': + # SEND_SUCCESS - TFA code has been successfully sent to phone + # QUOTA_EXCEEDED - reached the limit of TFA codes + status = try_get(login_challenge, lambda x: x[5], compat_str) + if status == 'QUOTA_EXCEEDED': + warn('Exceeded the limit of TFA codes, try later') + return False + + tl = try_get(challenge_results, lambda x: x[1][2], compat_str) + if not tl: + warn('Unable to extract TL') + return False + + tfa_code = self._get_tfa_info('2-step verification code') + + if not tfa_code: + warn( + 'Two-factor authentication required. Provide it either interactively or with --twofactor ' + '(Note that only TOTP (Google Authenticator App) codes work at this time.)') + return False + + tfa_code = remove_start(tfa_code, 'G-') + + tfa_req = [ + user_hash, None, 2, None, + [ + 9, None, None, None, None, None, None, None, + [None, tfa_code, True, 2] + ]] + + tfa_results = req( + self._TFA_URL.format(tl), tfa_req, + 'Submitting TFA code', 'Unable to submit TFA code') + + if tfa_results is False: + return False + + tfa_res = try_get(tfa_results, lambda x: x[0][5], list) + if tfa_res: + tfa_msg = try_get(tfa_res, lambda x: x[5], compat_str) + warn( + 'Unable to finish TFA: %s' % 'Invalid TFA code' + if tfa_msg == 'INCORRECT_ANSWER_ENTERED' else tfa_msg) + return False + + check_cookie_url = try_get( + tfa_results, lambda x: x[0][-1][2], compat_str) + else: + CHALLENGES = { + 'LOGIN_CHALLENGE': "This device isn't recognized. For your security, Google wants to make sure it's really you.", + 'USERNAME_RECOVERY': 'Please provide additional information to aid in the recovery process.', + 'REAUTH': "There is something unusual about your activity. For your security, Google wants to make sure it's really you.", + } + challenge = CHALLENGES.get( + challenge_str, + '%s returned error %s.' % (self.IE_NAME, challenge_str)) + warn('%s\nGo to https://accounts.google.com/, login and solve a challenge.' % challenge) + return False + else: + check_cookie_url = try_get(res, lambda x: x[2], compat_str) - tfa_req = sanitized_Request(self._TWOFACTOR_URL, tfa_data) - tfa_results = self._download_webpage( - tfa_req, None, - note='Submitting TFA code', errnote='unable to submit tfa', fatal=False) + if not check_cookie_url: + warn('Unable to extract CheckCookie URL') + return False - if tfa_results is False: - return False + check_cookie_results = self._download_webpage( + check_cookie_url, None, 'Checking cookie', fatal=False) - if re.search(r'(?i)]+id="challenge"', tfa_results) is not None: - self._downloader.report_warning('Two-factor code expired or invalid. Please try again, or use a one-use backup code instead.') - return False - if re.search(r'(?i)]+id="gaia_loginform"', tfa_results) is not None: - self._downloader.report_warning('unable to log in - did the page structure change?') - return False - if re.search(r'smsauth-interstitial-reviewsettings', tfa_results) is not None: - self._downloader.report_warning('Your Google account has a security notice. Please log in on your web browser, resolve the notice, and try again.') - return False + if check_cookie_results is False: + return False - if re.search(r'(?i)]+id="gaia_loginform"', login_results) is not None: - self._downloader.report_warning('unable to log in: bad username or password') + if 'https://myaccount.google.com/' not in check_cookie_results: + warn('Unable to log in') return False + return True + def _download_webpage_handle(self, *args, **kwargs): + query = kwargs.get('query', {}).copy() + query['disable_polymer'] = 'true' + kwargs['query'] = query + return super(YoutubeBaseInfoExtractor, self)._download_webpage_handle( + *args, **compat_kwargs(kwargs)) + def _real_initialize(self): if self._downloader is None: return @@ -245,8 +348,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor): (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/| (?:www\.)?deturl\.com/www\.youtube\.com/| (?:www\.)?pwnyoutube\.com/| + (?:www\.)?hooktube\.com/| (?:www\.)?yourepeat\.com/| tube\.majestyc\.net/| + (?:www\.)?invidio\.us/| youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains (?:.*?\#/)? # handle anchor (#/) redirect urls (?: # the various things that can precede the ID: @@ -390,12 +495,15 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'uploader': 'Philipp Hagemeister', 'uploader_id': 'phihag', 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/phihag', + 'channel_id': 'UCLqxVugv74EIW3VWh2NOa3Q', + 'channel_url': r're:https?://(?:www\.)?youtube\.com/channel/UCLqxVugv74EIW3VWh2NOa3Q', 'upload_date': '20121002', 'license': 'Standard YouTube License', 'description': 'test chars: "\'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de .', 'categories': ['Science & Technology'], 'tags': ['youtube-dl'], 'duration': 10, + 'view_count': int, 'like_count': int, 'dislike_count': int, 'start_time': 1, @@ -421,6 +529,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/IconaPop', 'license': 'Standard YouTube License', 'creator': 'Icona Pop', + 'track': 'I Love It (feat. Charli XCX)', + 'artist': 'Icona Pop', } }, { @@ -439,6 +549,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/justintimberlakeVEVO', 'license': 'Standard YouTube License', 'creator': 'Justin Timberlake', + 'track': 'Tunnel Vision', + 'artist': 'Justin Timberlake', 'age_limit': 18, } }, @@ -474,6 +586,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'categories': ['Science & Technology'], 'tags': ['youtube-dl'], 'duration': 10, + 'view_count': int, 'like_count': int, 'dislike_count': int, }, @@ -508,7 +621,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'id': 'IB3lcPjvWLA', 'ext': 'm4a', 'title': 'Afrojack, Spree Wilson - The Spark ft. Spree Wilson', - 'description': 'md5:12e7067fa6735a77bdcbb58cb1187d2d', + 'description': 'md5:1900ed86ee514927b9e00fbead6969a5', 'duration': 244, 'uploader': 'AfrojackVEVO', 'uploader_id': 'AfrojackVEVO', @@ -549,7 +662,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'ext': 'mp4', 'duration': 219, 'upload_date': '20100909', - 'uploader': 'The Amazing Atheist', + 'uploader': 'TJ Kirk', 'uploader_id': 'TheAmazingAtheist', 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/TheAmazingAtheist', 'license': 'Standard YouTube License', @@ -579,10 +692,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'url': 'https://www.youtube.com/watch?v=6kLq3WMV1nU', 'info_dict': { 'id': '6kLq3WMV1nU', - 'ext': 'mp4', + 'ext': 'webm', 'title': 'Dedication To My Ex (Miss That) (Lyric Video)', 'description': 'md5:33765bb339e1b47e7e72b5490139bb41', - 'duration': 247, + 'duration': 246, 'uploader': 'LloydVEVO', 'uploader_id': 'LloydVEVO', 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/LloydVEVO', @@ -592,6 +705,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): }, }, # video_info is None (https://github.com/rg3/youtube-dl/issues/4421) + # YouTube Red ad is not captured for creator { 'url': '__2ABJjxzNo', 'info_dict': { @@ -643,7 +757,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'uploader_id': 'AllenMeow', 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/AllenMeow', 'description': 'made by Wacom from Korea | 字幕&加油添醋 by TY\'s Allen | 感謝heylisa00cavey1001同學熱情提供梗及翻譯', - 'uploader': '孫艾倫', + 'uploader': '孫ᄋᄅ', 'license': 'Standard YouTube License', 'title': '[A-made] 變態妍字幕版 太妍 我就是這樣的人', }, @@ -670,7 +784,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'url': 'https://www.youtube.com/watch?v=FIl7x6_3R5Y', 'info_dict': { 'id': 'FIl7x6_3R5Y', - 'ext': 'mp4', + 'ext': 'webm', 'title': 'md5:7b81415841e02ecd4313668cde88737a', 'description': 'md5:116377fd2963b81ec4ce64b542173306', 'duration': 220, @@ -679,8 +793,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/dorappi2000', 'uploader': 'dorappi2000', 'license': 'Standard YouTube License', - 'formats': 'mincount:32', + 'formats': 'mincount:31', }, + 'skip': 'not actual anymore', }, # DASH manifest with segment_list { @@ -795,7 +910,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'id': 'lsguqyKfVQg', 'ext': 'mp4', 'title': '{dark walk}; Loki/AC/Dishonored; collab w/Elflover21', - 'alt_title': 'Dark Walk', + 'alt_title': 'Dark Walk - Position Music', 'description': 'md5:8085699c11dc3f597ce0410b0dcbb34a', 'duration': 133, 'upload_date': '20151119', @@ -803,7 +918,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/IronSoulElf', 'uploader': 'IronSoulElf', 'license': 'Standard YouTube License', - 'creator': 'Todd Haberman, Daniel Law Heath & Aaron Kaplan', + 'creator': 'Todd Haberman, Daniel Law Heath and Aaron Kaplan', + 'track': 'Dark Walk - Position Music', + 'artist': 'Todd Haberman, Daniel Law Heath and Aaron Kaplan', }, 'params': { 'skip_download': True, @@ -860,7 +977,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'description': 'md5:dda0d780d5a6e120758d1711d062a867', 'duration': 4060, 'upload_date': '20151119', - 'uploader': 'Bernie 2016', + 'uploader': 'Bernie Sanders', 'uploader_id': 'UCH1dpzjCEiGAt8CXkryhkZg', 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCH1dpzjCEiGAt8CXkryhkZg', 'license': 'Creative Commons Attribution license (reuse allowed)', @@ -895,6 +1012,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'params': { 'skip_download': True, }, + 'skip': 'This video is not available.', }, { # YouTube Red video with episode data @@ -903,7 +1021,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'id': 'iqKdEhx-dD4', 'ext': 'mp4', 'title': 'Isolation - Mind Field (Ep 1)', - 'description': 'md5:8013b7ddea787342608f63a13ddc9492', + 'description': 'md5:25b78d2f64ae81719f5c96319889b736', 'duration': 2085, 'upload_date': '20170118', 'uploader': 'Vsauce', @@ -921,6 +1039,26 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'Skipping DASH manifest', ], }, + { + # The following content has been identified by the YouTube community + # as inappropriate or offensive to some audiences. + 'url': 'https://www.youtube.com/watch?v=6SJNVb0GnPI', + 'info_dict': { + 'id': '6SJNVb0GnPI', + 'ext': 'mp4', + 'title': 'Race Differences in Intelligence', + 'description': 'md5:5d161533167390427a1f8ee89a1fc6f1', + 'duration': 965, + 'upload_date': '20140124', + 'uploader': 'New Century Foundation', + 'uploader_id': 'UCEJYpZGqgUob0zVVEaLhvVg', + 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCEJYpZGqgUob0zVVEaLhvVg', + 'license': 'Standard YouTube License', + }, + 'params': { + 'skip_download': True, + }, + }, { # itag 212 'url': '1t24XAntNCY', @@ -935,6 +1073,15 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'url': 'https://www.youtube.com/watch?v=MuAGGZNfUkU&list=RDMM', 'only_matching': True, }, + { + 'url': 'https://invidio.us/watch?v=BaW_jenozKc', + 'only_matching': True, + }, + { + # DRM protected + 'url': 'https://www.youtube.com/watch?v=s7_qI6_mIXc', + 'only_matching': True, + } ] def __init__(self, *args, **kwargs): @@ -963,7 +1110,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): def _extract_signature_function(self, video_id, player_url, example_sig): id_m = re.match( - r'.*?-(?P[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player(?:-new)?|(?:/[a-z]{2}_[A-Z]{2})?/base)?\.(?P[a-z]+)$', + r'.*?-(?P[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player(?:-new)?|(?:/[a-z]{2,3}_[A-Z]{2})?/base)?\.(?P[a-z]+)$', player_url) if not id_m: raise ExtractorError('Cannot identify player %r' % player_url) @@ -1049,7 +1196,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor): def _parse_sig_js(self, jscode): funcname = self._search_regex( (r'(["\'])signature\1\s*,\s*(?P[a-zA-Z0-9$]+)\(', - r'\.sig\|\|(?P[a-zA-Z0-9$]+)\('), + r'\.sig\|\|(?P[a-zA-Z0-9$]+)\(', + r'yt\.akamaized\.net/\)\s*\|\|\s*.*?\s*c\s*&&\s*d\.set\([^,]+\s*,\s*(?P[a-zA-Z0-9$]+)\(', + r'\bc\s*&&\s*d\.set\([^,]+\s*,\s*(?P[a-zA-Z0-9$]+)\(', + r'\bc\s*&&\s*d\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P[a-zA-Z0-9$]+)\('), jscode, 'Initial JS player signature function name', group='sig') jsi = JSInterpreter(jscode) @@ -1188,42 +1338,65 @@ class YoutubeIE(YoutubeBaseInfoExtractor): sub_lang_list[sub_lang] = sub_formats return sub_lang_list + def make_captions(sub_url, sub_langs): + parsed_sub_url = compat_urllib_parse_urlparse(sub_url) + caption_qs = compat_parse_qs(parsed_sub_url.query) + captions = {} + for sub_lang in sub_langs: + sub_formats = [] + for ext in self._SUBTITLE_FORMATS: + caption_qs.update({ + 'tlang': [sub_lang], + 'fmt': [ext], + }) + sub_url = compat_urlparse.urlunparse(parsed_sub_url._replace( + query=compat_urllib_parse_urlencode(caption_qs, True))) + sub_formats.append({ + 'url': sub_url, + 'ext': ext, + }) + captions[sub_lang] = sub_formats + return captions + + # New captions format as of 22.06.2017 + player_response = args.get('player_response') + if player_response and isinstance(player_response, compat_str): + player_response = self._parse_json( + player_response, video_id, fatal=False) + if player_response: + renderer = player_response['captions']['playerCaptionsTracklistRenderer'] + base_url = renderer['captionTracks'][0]['baseUrl'] + sub_lang_list = [] + for lang in renderer['translationLanguages']: + lang_code = lang.get('languageCode') + if lang_code: + sub_lang_list.append(lang_code) + return make_captions(base_url, sub_lang_list) + # Some videos don't provide ttsurl but rather caption_tracks and # caption_translation_languages (e.g. 20LmZk1hakA) + # Does not used anymore as of 22.06.2017 caption_tracks = args['caption_tracks'] caption_translation_languages = args['caption_translation_languages'] caption_url = compat_parse_qs(caption_tracks.split(',')[0])['u'][0] - parsed_caption_url = compat_urllib_parse_urlparse(caption_url) - caption_qs = compat_parse_qs(parsed_caption_url.query) - - sub_lang_list = {} + sub_lang_list = [] for lang in caption_translation_languages.split(','): lang_qs = compat_parse_qs(compat_urllib_parse_unquote_plus(lang)) sub_lang = lang_qs.get('lc', [None])[0] - if not sub_lang: - continue - sub_formats = [] - for ext in self._SUBTITLE_FORMATS: - caption_qs.update({ - 'tlang': [sub_lang], - 'fmt': [ext], - }) - sub_url = compat_urlparse.urlunparse(parsed_caption_url._replace( - query=compat_urllib_parse_urlencode(caption_qs, True))) - sub_formats.append({ - 'url': sub_url, - 'ext': ext, - }) - sub_lang_list[sub_lang] = sub_formats - return sub_lang_list + if sub_lang: + sub_lang_list.append(sub_lang) + return make_captions(caption_url, sub_lang_list) # An extractor error can be raise by the download process if there are # no automatic captions but there are subtitles - except (KeyError, ExtractorError): + except (KeyError, IndexError, ExtractorError): self._downloader.report_warning(err_msg) return {} - def _mark_watched(self, video_id, video_info): - playback_url = video_info.get('videostats_playback_base_url', [None])[0] + def _mark_watched(self, video_id, video_info, player_response): + playback_url = url_or_none(try_get( + player_response, + lambda x: x['playbackTracking']['videostatsPlaybackUrl']['baseUrl']) or try_get( + video_info, lambda x: x['videostats_playback_base_url'][0])) if not playback_url: return parsed_playback_url = compat_urlparse.urlparse(playback_url) @@ -1245,6 +1418,43 @@ class YoutubeIE(YoutubeBaseInfoExtractor): playback_url, video_id, 'Marking watched', 'Unable to mark watched', fatal=False) + @staticmethod + def _extract_urls(webpage): + # Embedded YouTube player + entries = [ + unescapeHTML(mobj.group('url')) + for mobj in re.finditer(r'''(?x) + (?: + ]+?src=| + data-video-url=| + ]+?src=| + embedSWF\(?:\s*| + ]+data=| + new\s+SWFObject\( + ) + (["\']) + (?P(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/ + (?:embed|v|p)/[0-9A-Za-z_-]{11}.*?) + \1''', webpage)] + + # lazyYT YouTube embed + entries.extend(list(map( + unescapeHTML, + re.findall(r'class="lazyYT" data-youtube-id="([^"]+)"', webpage)))) + + # Wordpress "YouTube Video Importer" plugin + matches = re.findall(r'''(?x)]+ + class=(?P[\'"])[^\'"]*\byvii_single_video_player\b[^\'"]*(?P=q1)[^>]+ + data-video_id=(?P[\'"])([^\'"]+)(?P=q2)''', webpage) + entries.extend(m[-1] for m in matches) + + return entries + + @staticmethod + def _extract_url(webpage): + urls = YoutubeIE._extract_urls(webpage) + return urls[0] if urls else None + @classmethod def extract_id(cls, url): mobj = re.match(cls._VALID_URL, url, re.VERBOSE) @@ -1257,6 +1467,41 @@ class YoutubeIE(YoutubeBaseInfoExtractor): url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id return self._download_webpage(url, video_id, note='Searching for annotations.', errnote='Unable to download video annotations.') + @staticmethod + def _extract_chapters(description, duration): + if not description: + return None + chapter_lines = re.findall( + r'(?:^|)([^<]*]+onclick=["\']yt\.www\.watch\.player\.seekTo[^>]+>(\d{1,2}:\d{1,2}(?::\d{1,2})?)[^>]*)(?=$|)', + description) + if not chapter_lines: + return None + chapters = [] + for next_num, (chapter_line, time_point) in enumerate( + chapter_lines, start=1): + start_time = parse_duration(time_point) + if start_time is None: + continue + if start_time > duration: + break + end_time = (duration if next_num == len(chapter_lines) + else parse_duration(chapter_lines[next_num][1])) + if end_time is None: + continue + if end_time > duration: + end_time = duration + if start_time > end_time: + break + chapter_title = re.sub( + r']+>[^<]+', '', chapter_line).strip(' \t-') + chapter_title = re.sub(r'\s+', ' ', chapter_title) + chapters.append({ + 'start_time': start_time, + 'end_time': end_time, + 'title': chapter_title, + }) + return chapters + def _real_extract(self, url): url, smuggled_data = unsmuggle_url(url, {}) @@ -1300,9 +1545,16 @@ class YoutubeIE(YoutubeBaseInfoExtractor): if dash_mpd and dash_mpd[0] not in dash_mpds: dash_mpds.append(dash_mpd[0]) + is_live = None + view_count = None + + def extract_view_count(v_info): + return int_or_none(try_get(v_info, lambda x: x['view_count'][0])) + + player_response = {} + # Get video info embed_webpage = None - is_live = None if re.search(r'player-age-gate-content">', video_webpage) is not None: age_gate = True # We simulate the access to the video from www.youtube.com/v/{video_id} @@ -1325,11 +1577,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor): else: age_gate = False video_info = None + sts = None # Try looking directly into the video webpage ytplayer_config = self._get_ytplayer_config(video_id, video_webpage) if ytplayer_config: args = ytplayer_config['args'] - if args.get('url_encoded_fmt_stream_map'): + if args.get('url_encoded_fmt_stream_map') or args.get('hlsvp'): # Convert to the same format returned by compat_parse_qs video_info = dict((k, [v]) for k, v in args.items()) add_dash_mpd(video_info) @@ -1341,6 +1594,13 @@ class YoutubeIE(YoutubeBaseInfoExtractor): args['ypc_vid'], YoutubeIE.ie_key(), video_id=args['ypc_vid']) if args.get('livestream') == '1' or args.get('live_playback') == 1: is_live = True + sts = ytplayer_config.get('sts') + if not player_response: + pl_response = str_or_none(args.get('player_response')) + if pl_response: + pl_response = self._parse_json(pl_response, video_id, fatal=False) + if isinstance(pl_response, dict): + player_response = pl_response if not video_info or self._downloader.params.get('youtube_include_dash_manifest', True): # We also try looking in get_video_info since it may contain different dashmpd # URL that points to a DASH manifest with possibly different itag set (some itags @@ -1349,17 +1609,33 @@ class YoutubeIE(YoutubeBaseInfoExtractor): # The general idea is to take a union of itags of both DASH manifests (for example # video with such 'manifest behavior' see https://github.com/rg3/youtube-dl/issues/6093) self.report_video_info_webpage_download(video_id) - for el_type in ['&el=info', '&el=embedded', '&el=detailpage', '&el=vevo', '']: - video_info_url = ( - '%s://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en' - % (proto, video_id, el_type)) + for el in ('info', 'embedded', 'detailpage', 'vevo', ''): + query = { + 'video_id': video_id, + 'ps': 'default', + 'eurl': '', + 'gl': 'US', + 'hl': 'en', + } + if el: + query['el'] = el + if sts: + query['sts'] = sts video_info_webpage = self._download_webpage( - video_info_url, + '%s://www.youtube.com/get_video_info' % proto, video_id, note=False, - errnote='unable to download video info webpage') + errnote='unable to download video info webpage', + fatal=False, query=query) + if not video_info_webpage: + continue get_video_info = compat_parse_qs(video_info_webpage) - if get_video_info.get('use_cipher_signature') != ['True']: - add_dash_mpd(get_video_info) + if not player_response: + pl_response = get_video_info.get('player_response', [None])[0] + if isinstance(pl_response, dict): + player_response = pl_response + add_dash_mpd(get_video_info) + if view_count is None: + view_count = extract_view_count(get_video_info) if not video_info: video_info = get_video_info if 'token' in get_video_info: @@ -1375,6 +1651,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor): if 'token' not in video_info: video_info = get_video_info break + + def extract_unavailable_message(): + return self._html_search_regex( + r'(?s)]+id="unavailable-message"[^>]*>(.+?)', + video_webpage, 'unavailable message', default=None) + if 'token' not in video_info: if 'reason' in video_info: if 'The uploader has not made this video available in your country.' in video_info['reason']: @@ -1383,25 +1665,49 @@ class YoutubeIE(YoutubeBaseInfoExtractor): countries = regions_allowed.split(',') if regions_allowed else None self.raise_geo_restricted( msg=video_info['reason'][0], countries=countries) + reason = video_info['reason'][0] + if 'Invalid parameters' in reason: + unavailable_message = extract_unavailable_message() + if unavailable_message: + reason = unavailable_message raise ExtractorError( - 'YouTube said: %s' % video_info['reason'][0], + 'YouTube said: %s' % reason, expected=True, video_id=video_id) else: raise ExtractorError( '"token" parameter not in video info for unknown reason', video_id=video_id) + if video_info.get('license_info'): + raise ExtractorError('This video is DRM protected.', expected=True) + + video_details = try_get( + player_response, lambda x: x['videoDetails'], dict) or {} + # title if 'title' in video_info: video_title = video_info['title'][0] + elif 'title' in player_response: + video_title = video_details['title'] else: self._downloader.report_warning('Unable to extract video title') video_title = '_' # description - video_description = get_element_by_id("eow-description", video_webpage) + description_original = video_description = get_element_by_id("eow-description", video_webpage) if video_description: - video_description = re.sub(r'''(?x) + + def replace_url(m): + redir_url = compat_urlparse.urljoin(url, m.group(1)) + parsed_redir_url = compat_urllib_parse_urlparse(redir_url) + if re.search(r'^(?:www\.)?(?:youtube(?:-nocookie)?\.com|youtu\.be)$', parsed_redir_url.netloc) and parsed_redir_url.path == '/redirect': + qs = compat_parse_qs(parsed_redir_url.query) + q = qs.get('q') + if q and q[0]: + return q[0] + return redir_url + + description_original = video_description = re.sub(r'''(?x) ]*> [^<]+\.{3}\s* - ''', r'\1', video_description) + ''', replace_url, video_description) video_description = clean_html(video_description) else: fd_mobj = re.search(r'', - video_webpage) - if mobj is not None: - video_uploader_id = mobj.group('uploader_id') - video_uploader_url = mobj.group('uploader_url') - else: - self._downloader.report_warning('unable to extract uploader nickname') - - # thumbnail image - # We try first to get a high quality image: - m_thumb = re.search(r'', - video_webpage, re.DOTALL) - if m_thumb is not None: - video_thumbnail = m_thumb.group(1) - elif 'thumbnail_url' not in video_info: - self._downloader.report_warning('unable to extract video thumbnail') - video_thumbnail = None - else: # don't panic if we can't find it - video_thumbnail = compat_urllib_parse_unquote_plus(video_info['thumbnail_url'][0]) - - # upload date - upload_date = self._html_search_meta( - 'datePublished', video_webpage, 'upload date', default=None) - if not upload_date: - upload_date = self._search_regex( - [r'(?s)id="eow-date.*?>(.*?)', - r'id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live|Started) on (.+?)'], - video_webpage, 'upload date', default=None) - if upload_date: - upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split()) - upload_date = unified_strdate(upload_date) - - video_license = self._html_search_regex( - r']+class="title"[^>]*>\s*License\s*\s*]*>\s*
  • (.+?)]+class="title"[^>]*>\s*Music\s*\s*]*>\s*
  • (?P.+?) by (?P<creator>.+?)(?:\(.+?\))?</li', - video_webpage) - if m_music: - video_alt_title = remove_quotes(unescapeHTML(m_music.group('title'))) - video_creator = clean_html(m_music.group('creator')) - else: - video_alt_title = video_creator = None - - m_episode = re.search( - r'<div[^>]+id="watch7-headline"[^>]*>\s*<span[^>]*>.*?>(?P<series>[^<]+)</a></b>\s*S(?P<season>\d+)\s*•\s*E(?P<episode>\d+)</span>', - video_webpage) - if m_episode: - series = m_episode.group('series') - season_number = int(m_episode.group('season')) - episode_number = int(m_episode.group('episode')) - else: - series = season_number = episode_number = None - - m_cat_container = self._search_regex( - r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>', - video_webpage, 'categories', default=None) - if m_cat_container: - category = self._html_search_regex( - r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category', - default=None) - video_categories = None if category is None else [category] - else: - video_categories = None - - video_tags = [ - unescapeHTML(m.group('content')) - for m in re.finditer(self._meta_regex('og:video:tag'), video_webpage)] - - def _extract_count(count_name): - return str_to_int(self._search_regex( - r'-%s-button[^>]+><span[^>]+class="yt-uix-button-content"[^>]*>([\d,]+)</span>' - % re.escape(count_name), - video_webpage, count_name, default=None)) - - like_count = _extract_count('like') - dislike_count = _extract_count('dislike') - - # subtitles - video_subtitles = self.extract_subtitles(video_id, video_webpage) - automatic_captions = self.extract_automatic_captions(video_id, video_webpage) - - video_duration = try_get( - video_info, lambda x: int_or_none(x['length_seconds'][0])) - if not video_duration: - video_duration = parse_duration(self._html_search_meta( - 'duration', video_webpage, 'video duration')) - - # annotations - video_annotations = None - if self._downloader.params.get('writeannotations', False): - video_annotations = self._extract_annotations(video_id) + def _extract_filesize(media_url): + return int_or_none(self._search_regex( + r'\bclen[=/](\d+)', media_url, 'filesize', default=None)) if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'): self.report_rtmp_download() @@ -1566,7 +1776,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'url': video_info['conn'][0], 'player_url': player_url, }] - elif len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1: + elif not is_live and (len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1): encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts', [''])[0] if 'rtmpe%3Dyes' in encoded_url_map: raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True) @@ -1583,6 +1793,26 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'width': int_or_none(width_height[0]), 'height': int_or_none(width_height[1]), } + q = qualities(['small', 'medium', 'hd720']) + streaming_formats = try_get(player_response, lambda x: x['streamingData']['formats'], list) + if streaming_formats: + for fmt in streaming_formats: + itag = str_or_none(fmt.get('itag')) + if not itag: + continue + quality = fmt.get('quality') + quality_label = fmt.get('qualityLabel') or quality + formats_spec[itag] = { + 'asr': int_or_none(fmt.get('audioSampleRate')), + 'filesize': int_or_none(fmt.get('contentLength')), + 'format_note': quality_label, + 'fps': int_or_none(fmt.get('fps')), + 'height': int_or_none(fmt.get('height')), + 'quality': q(quality), + # bitrate for itag 43 is always 2147483647 + 'tbr': float_or_none(fmt.get('averageBitrate') or fmt.get('bitrate'), 1000) if itag != '43' else None, + 'width': int_or_none(fmt.get('width')), + } formats = [] for url_data_str in encoded_url_map.split(','): url_data = compat_parse_qs(url_data_str) @@ -1591,12 +1821,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor): format_id = url_data['itag'][0] url = url_data['url'][0] - if 'sig' in url_data: - url += '&signature=' + url_data['sig'][0] - elif 's' in url_data: - encrypted_sig = url_data['s'][0] + if 's' in url_data or self._downloader.params.get('youtube_include_dash_manifest', True): ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")' - jsplayer_url_json = self._search_regex( ASSETS_RE, embed_webpage if age_gate else video_webpage, @@ -1617,6 +1843,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor): video_webpage, 'age gate player URL') player_url = json.loads(player_url_json) + if 'sig' in url_data: + url += '&signature=' + url_data['sig'][0] + elif 's' in url_data: + encrypted_sig = url_data['s'][0] + if self._downloader.params.get('verbose'): if player_url is None: player_version = 'unknown' @@ -1630,7 +1861,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): else: player_version = self._search_regex( [r'html5player-([^/]+?)(?:/html5player(?:-new)?)?\.js', - r'(?:www|player)-([^/]+)(?:/[a-z]{2}_[A-Z]{2})?/base\.js'], + r'(?:www|player(?:_ias)?)-([^/]+)(?:/[a-z]{2,3}_[A-Z]{2})?/base\.js'], player_url, 'html5 player', fatal=False) player_desc = 'html5 player %s' % player_version @@ -1661,13 +1892,19 @@ class YoutubeIE(YoutubeBaseInfoExtractor): mobj = re.search(r'^(?P<width>\d+)[xX](?P<height>\d+)$', url_data.get('size', [''])[0]) width, height = (int(mobj.group('width')), int(mobj.group('height'))) if mobj else (None, None) + filesize = int_or_none(url_data.get( + 'clen', [None])[0]) or _extract_filesize(url) + + quality = url_data.get('quality', [None])[0] + more_fields = { - 'filesize': int_or_none(url_data.get('clen', [None])[0]), + 'filesize': filesize, 'tbr': float_or_none(url_data.get('bitrate', [None])[0], 1000), 'width': width, 'height': height, 'fps': int_or_none(url_data.get('fps', [None])[0]), - 'format_note': url_data.get('quality_label', [None])[0] or url_data.get('quality', [None])[0], + 'format_note': url_data.get('quality_label', [None])[0] or quality, + 'quality': q(quality), } for key, value in more_fields.items(): if value: @@ -1688,32 +1925,184 @@ class YoutubeIE(YoutubeBaseInfoExtractor): break if codecs: dct.update(parse_codecs(codecs)) + if dct.get('acodec') == 'none' or dct.get('vcodec') == 'none': + dct['downloader_options'] = { + # Youtube throttles chunks >~10M + 'http_chunk_size': 10485760, + } formats.append(dct) - elif video_info.get('hlsvp'): - manifest_url = video_info['hlsvp'][0] - formats = [] - m3u8_formats = self._extract_m3u8_formats( - manifest_url, video_id, 'mp4', fatal=False) - for a_format in m3u8_formats: - itag = self._search_regex( - r'/itag/(\d+)/', a_format['url'], 'itag', default=None) - if itag: - a_format['format_id'] = itag - if itag in self._formats: - dct = self._formats[itag].copy() - dct.update(a_format) - a_format = dct - a_format['player_url'] = player_url - # Accept-Encoding header causes failures in live streams on Youtube and Youtube Gaming - a_format.setdefault('http_headers', {})['Youtubedl-no-compression'] = 'True' - formats.append(a_format) else: - unavailable_message = self._html_search_regex( - r'(?s)<h1[^>]+id="unavailable-message"[^>]*>(.+?)</h1>', - video_webpage, 'unavailable message', default=None) - if unavailable_message: - raise ExtractorError(unavailable_message, expected=True) - raise ExtractorError('no conn, hlsvp or url_encoded_fmt_stream_map information found in video info') + manifest_url = ( + url_or_none(try_get( + player_response, + lambda x: x['streamingData']['hlsManifestUrl'], + compat_str)) or + url_or_none(try_get( + video_info, lambda x: x['hlsvp'][0], compat_str))) + if manifest_url: + formats = [] + m3u8_formats = self._extract_m3u8_formats( + manifest_url, video_id, 'mp4', fatal=False) + for a_format in m3u8_formats: + itag = self._search_regex( + r'/itag/(\d+)/', a_format['url'], 'itag', default=None) + if itag: + a_format['format_id'] = itag + if itag in self._formats: + dct = self._formats[itag].copy() + dct.update(a_format) + a_format = dct + a_format['player_url'] = player_url + # Accept-Encoding header causes failures in live streams on Youtube and Youtube Gaming + a_format.setdefault('http_headers', {})['Youtubedl-no-compression'] = 'True' + formats.append(a_format) + else: + error_message = clean_html(video_info.get('reason', [None])[0]) + if not error_message: + error_message = extract_unavailable_message() + if error_message: + raise ExtractorError(error_message, expected=True) + raise ExtractorError('no conn, hlsvp, hlsManifestUrl or url_encoded_fmt_stream_map information found in video info') + + # uploader + video_uploader = try_get( + video_info, lambda x: x['author'][0], + compat_str) or str_or_none(video_details.get('author')) + if video_uploader: + video_uploader = compat_urllib_parse_unquote_plus(video_uploader) + else: + self._downloader.report_warning('unable to extract uploader name') + + # uploader_id + video_uploader_id = None + video_uploader_url = None + mobj = re.search( + r'<link itemprop="url" href="(?P<uploader_url>https?://www\.youtube\.com/(?:user|channel)/(?P<uploader_id>[^"]+))">', + video_webpage) + if mobj is not None: + video_uploader_id = mobj.group('uploader_id') + video_uploader_url = mobj.group('uploader_url') + else: + self._downloader.report_warning('unable to extract uploader nickname') + + channel_id = self._html_search_meta( + 'channelId', video_webpage, 'channel id') + channel_url = 'http://www.youtube.com/channel/%s' % channel_id if channel_id else None + + # thumbnail image + # We try first to get a high quality image: + m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">', + video_webpage, re.DOTALL) + if m_thumb is not None: + video_thumbnail = m_thumb.group(1) + elif 'thumbnail_url' not in video_info: + self._downloader.report_warning('unable to extract video thumbnail') + video_thumbnail = None + else: # don't panic if we can't find it + video_thumbnail = compat_urllib_parse_unquote_plus(video_info['thumbnail_url'][0]) + + # upload date + upload_date = self._html_search_meta( + 'datePublished', video_webpage, 'upload date', default=None) + if not upload_date: + upload_date = self._search_regex( + [r'(?s)id="eow-date.*?>(.*?)</span>', + r'(?:id="watch-uploader-info".*?>.*?|["\']simpleText["\']\s*:\s*["\'])(?:Published|Uploaded|Streamed live|Started) on (.+?)[<"\']'], + video_webpage, 'upload date', default=None) + upload_date = unified_strdate(upload_date) + + video_license = self._html_search_regex( + r'<h4[^>]+class="title"[^>]*>\s*License\s*</h4>\s*<ul[^>]*>\s*<li>(.+?)</li', + video_webpage, 'license', default=None) + + m_music = re.search( + r'''(?x) + <h4[^>]+class="title"[^>]*>\s*Music\s*</h4>\s* + <ul[^>]*>\s* + <li>(?P<title>.+?) + by (?P<creator>.+?) + (?: + \(.+?\)| + <a[^>]* + (?: + \bhref=["\']/red[^>]*>| # drop possible + >\s*Listen ad-free with YouTube Red # YouTube Red ad + ) + .*? + )?</li + ''', + video_webpage) + if m_music: + video_alt_title = remove_quotes(unescapeHTML(m_music.group('title'))) + video_creator = clean_html(m_music.group('creator')) + else: + video_alt_title = video_creator = None + + def extract_meta(field): + return self._html_search_regex( + r'<h4[^>]+class="title"[^>]*>\s*%s\s*</h4>\s*<ul[^>]*>\s*<li>(.+?)</li>\s*' % field, + video_webpage, field, default=None) + + track = extract_meta('Song') + artist = extract_meta('Artist') + + m_episode = re.search( + r'<div[^>]+id="watch7-headline"[^>]*>\s*<span[^>]*>.*?>(?P<series>[^<]+)</a></b>\s*S(?P<season>\d+)\s*•\s*E(?P<episode>\d+)</span>', + video_webpage) + if m_episode: + series = unescapeHTML(m_episode.group('series')) + season_number = int(m_episode.group('season')) + episode_number = int(m_episode.group('episode')) + else: + series = season_number = episode_number = None + + m_cat_container = self._search_regex( + r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>', + video_webpage, 'categories', default=None) + if m_cat_container: + category = self._html_search_regex( + r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category', + default=None) + video_categories = None if category is None else [category] + else: + video_categories = None + + video_tags = [ + unescapeHTML(m.group('content')) + for m in re.finditer(self._meta_regex('og:video:tag'), video_webpage)] + + def _extract_count(count_name): + return str_to_int(self._search_regex( + r'-%s-button[^>]+><span[^>]+class="yt-uix-button-content"[^>]*>([\d,]+)</span>' + % re.escape(count_name), + video_webpage, count_name, default=None)) + + like_count = _extract_count('like') + dislike_count = _extract_count('dislike') + + if view_count is None: + view_count = str_to_int(self._search_regex( + r'<[^>]+class=["\']watch-view-count[^>]+>\s*([\d,\s]+)', video_webpage, + 'view count', default=None)) + + # subtitles + video_subtitles = self.extract_subtitles(video_id, video_webpage) + automatic_captions = self.extract_automatic_captions(video_id, video_webpage) + + video_duration = try_get( + video_info, lambda x: int_or_none(x['length_seconds'][0])) + if not video_duration: + video_duration = int_or_none(video_details.get('lengthSeconds')) + if not video_duration: + video_duration = parse_duration(self._html_search_meta( + 'duration', video_webpage, 'video duration')) + + # annotations + video_annotations = None + if self._downloader.params.get('writeannotations', False): + video_annotations = self._extract_annotations(video_id) + + chapters = self._extract_chapters(description_original, video_duration) # Look for the DASH manifest if self._downloader.params.get('youtube_include_dash_manifest', True): @@ -1731,6 +2120,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor): for df in self._extract_mpd_formats( mpd_url, video_id, fatal=dash_mpd_fatal, formats_dict=self._formats): + if not df.get('filesize'): + df['filesize'] = _extract_filesize(df['url']) # Do not overwrite DASH format found in some previous DASH manifest if df['format_id'] not in dash_formats: dash_formats[df['format_id']] = df @@ -1769,18 +2160,20 @@ class YoutubeIE(YoutubeBaseInfoExtractor): self._sort_formats(formats) - self.mark_watched(video_id, video_info) + self.mark_watched(video_id, video_info, player_response) return { 'id': video_id, 'uploader': video_uploader, 'uploader_id': video_uploader_id, 'uploader_url': video_uploader_url, + 'channel_id': channel_id, + 'channel_url': channel_url, 'upload_date': upload_date, 'license': video_license, - 'creator': video_creator, + 'creator': video_creator or artist, 'title': video_title, - 'alt_title': video_alt_title, + 'alt_title': video_alt_title or track, 'thumbnail': video_thumbnail, 'description': video_description, 'categories': video_categories, @@ -1790,6 +2183,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'duration': video_duration, 'age_limit': 18 if age_gate else 0, 'annotations': video_annotations, + 'chapters': chapters, 'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id, 'view_count': view_count, 'like_count': like_count, @@ -1802,49 +2196,22 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'series': series, 'season_number': season_number, 'episode_number': episode_number, + 'track': track, + 'artist': artist, } -class YoutubeSharedVideoIE(InfoExtractor): - _VALID_URL = r'(?:https?:)?//(?:www\.)?youtube\.com/shared\?.*\bci=(?P<id>[0-9A-Za-z_-]{11})' - IE_NAME = 'youtube:shared' - - _TEST = { - 'url': 'https://www.youtube.com/shared?ci=1nEzmT-M4fU', - 'info_dict': { - 'id': 'uPDB5I9wfp8', - 'ext': 'webm', - 'title': 'Pocoyo: 90 minutos de episódios completos Português para crianças - PARTE 3', - 'description': 'md5:d9e4d9346a2dfff4c7dc4c8cec0f546d', - 'upload_date': '20160219', - 'uploader': 'Pocoyo - Português (BR)', - 'uploader_id': 'PocoyoBrazil', - }, - 'add_ie': ['Youtube'], - 'params': { - # There are already too many Youtube downloads - 'skip_download': True, - }, - } - - def _real_extract(self, url): - video_id = self._match_id(url) - - webpage = self._download_webpage(url, video_id) - - real_video_id = self._html_search_meta( - 'videoId', webpage, 'YouTube video id', fatal=True) - - return self.url_result(real_video_id, YoutubeIE.ie_key()) - - class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor): IE_DESC = 'YouTube.com playlists' _VALID_URL = r"""(?x)(?: (?:https?://)? (?:\w+\.)? (?: - youtube\.com/ + (?: + youtube\.com| + invidio\.us + ) + / (?: (?:course|view_play_list|my_playlists|artist|playlist|watch|embed/(?:videoseries|[0-9A-Za-z_-]{11})) \? (?:.*?[&;])*? (?:p|a|list)= @@ -1853,7 +2220,7 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor): youtu\.be/[0-9A-Za-z_-]{11}\?.*?\blist= ) ( - (?:PL|LL|EC|UU|FL|RD|UL|TL)?[0-9A-Za-z-_]{10,} + (?:PL|LL|EC|UU|FL|RD|UL|TL|OLAK5uy_)?[0-9A-Za-z-_]{10,} # Top tracks, they can also include dots |(?:MC)[\w\.]* ) @@ -1861,7 +2228,7 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor): | (%(playlist_id)s) )""" % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE} - _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s&disable_polymer=true' + _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s' _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)(?:[^>]+>(?P<title>[^<]+))?' IE_NAME = 'youtube:playlist' _TESTS = [{ @@ -1957,6 +2324,7 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor): 'description': 'md5:507cdcb5a49ac0da37a920ece610be80', 'categories': ['People & Blogs'], 'tags': list, + 'view_count': int, 'like_count': int, 'dislike_count': int, }, @@ -1991,6 +2359,13 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor): }, { 'url': 'TLGGrESM50VT6acwMjAyMjAxNw', 'only_matching': True, + }, { + # music album playlist + 'url': 'OLAK5uy_m4xAFdmMC5rX3Ji3g93pQe3hqLZw_9LhM', + 'only_matching': True, + }, { + 'url': 'https://invidio.us/playlist?list=PLDIoUOhQQPlXr63I_vwF9GD8sAKh77dWU', + 'only_matching': True, }] def _real_initialize(self): @@ -2057,6 +2432,19 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor): r'(?s)<h1 class="pl-header-title[^"]*"[^>]*>\s*(.*?)\s*</h1>', page, 'title', default=None) + _UPLOADER_BASE = r'class=["\']pl-header-details[^>]+>\s*<li>\s*<a[^>]+\bhref=' + uploader = self._search_regex( + r'%s["\']/(?:user|channel)/[^>]+>([^<]+)' % _UPLOADER_BASE, + page, 'uploader', default=None) + mobj = re.search( + r'%s(["\'])(?P<path>/(?:user|channel)/(?P<uploader_id>.+?))\1' % _UPLOADER_BASE, + page) + if mobj: + uploader_id = mobj.group('uploader_id') + uploader_url = compat_urlparse.urljoin(url, mobj.group('path')) + else: + uploader_id = uploader_url = None + has_videos = True if not playlist_title: @@ -2067,8 +2455,15 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor): except StopIteration: has_videos = False - return has_videos, self.playlist_result( + playlist = self.playlist_result( self._entries(page, playlist_id), playlist_id, playlist_title) + playlist.update({ + 'uploader': uploader, + 'uploader_id': uploader_id, + 'uploader_url': uploader_url, + }) + + return has_videos, playlist def _check_download_just_video(self, url, playlist_id): # Check if it's a video-specific URL @@ -2113,7 +2508,7 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor): class YoutubeChannelIE(YoutubePlaylistBaseInfoExtractor): IE_DESC = 'YouTube.com channels' - _VALID_URL = r'https?://(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/(?P<id>[0-9A-Za-z_-]+)' + _VALID_URL = r'https?://(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com|(?:www\.)?invidio\.us)/channel/(?P<id>[0-9A-Za-z_-]+)' _TEMPLATE_URL = 'https://www.youtube.com/channel/%s/videos' _VIDEO_RE = r'(?:title="(?P<title>[^"]+)"[^>]+)?href="/watch\?v=(?P<id>[0-9A-Za-z_-]+)&?' IE_NAME = 'youtube:channel' @@ -2134,6 +2529,9 @@ class YoutubeChannelIE(YoutubePlaylistBaseInfoExtractor): 'id': 'UUs0ifCMCm1icqRbqhUINa0w', 'title': 'Uploads from Deus Ex', }, + }, { + 'url': 'https://invidio.us/channel/UC23qupoDRn9YOAVzeoxjOQA', + 'only_matching': True, }] @classmethod @@ -2204,7 +2602,7 @@ class YoutubeChannelIE(YoutubePlaylistBaseInfoExtractor): class YoutubeUserIE(YoutubeChannelIE): IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)' - _VALID_URL = r'(?:(?:https?://(?:\w+\.)?youtube\.com/(?:(?P<user>user|c)/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)(?P<id>[A-Za-z0-9_-]+)' + _VALID_URL = r'(?:(?:https?://(?:\w+\.)?youtube\.com/(?:(?P<user>user|c)/)?(?!(?:attribution_link|watch|results|shared)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)(?P<id>[A-Za-z0-9_-]+)' _TEMPLATE_URL = 'https://www.youtube.com/%s/%s/videos' IE_NAME = 'youtube:user' @@ -2297,10 +2695,11 @@ class YoutubeLiveIE(YoutubeBaseInfoExtractor): webpage = self._download_webpage(url, channel_id, fatal=False) if webpage: page_type = self._og_search_property( - 'type', webpage, 'page type', default=None) + 'type', webpage, 'page type', default='') video_id = self._html_search_meta( 'videoId', webpage, 'video id', default=None) - if page_type == 'video' and video_id and re.match(r'^[0-9A-Za-z_-]{11}$', video_id): + if page_type.startswith('video') and video_id and re.match( + r'^[0-9A-Za-z_-]{11}$', video_id): return self.url_result(video_id, YoutubeIE.ie_key()) return self.url_result(base_url) @@ -2335,7 +2734,11 @@ class YoutubePlaylistsIE(YoutubePlaylistsBaseInfoExtractor): }] -class YoutubeSearchIE(SearchInfoExtractor, YoutubePlaylistIE): +class YoutubeSearchBaseInfoExtractor(YoutubePlaylistBaseInfoExtractor): + _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})(?:[^"]*"[^>]+\btitle="(?P<title>[^"]+))?' + + +class YoutubeSearchIE(SearchInfoExtractor, YoutubeSearchBaseInfoExtractor): IE_DESC = 'YouTube.com searches' # there doesn't appear to be a real limit, for example if you search for # 'python' you get more than 8.000.000 results @@ -2369,8 +2772,7 @@ class YoutubeSearchIE(SearchInfoExtractor, YoutubePlaylistIE): raise ExtractorError( '[youtube] No video results', expected=True) - new_videos = self._ids_to_results(orderedSet(re.findall( - r'href="/watch\?v=(.{11})', html_content))) + new_videos = list(self._process_page(html_content)) videos += new_videos if not new_videos or len(videos) > limit: break @@ -2393,11 +2795,10 @@ class YoutubeSearchDateIE(YoutubeSearchIE): _EXTRA_QUERY_ARGS = {'search_sort': 'video_date_uploaded'} -class YoutubeSearchURLIE(YoutubePlaylistBaseInfoExtractor): +class YoutubeSearchURLIE(YoutubeSearchBaseInfoExtractor): IE_DESC = 'YouTube.com search URLs' IE_NAME = 'youtube:search_url' _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?(?:search_query|q)=(?P<query>[^&]+)(?:[&]|$)' - _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})(?:[^"]*"[^>]+\btitle="(?P<title>[^"]+))?' _TESTS = [{ 'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video', 'playlist_mincount': 5, @@ -2449,10 +2850,7 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): def _real_initialize(self): self._login() - def _real_extract(self, url): - page = self._download_webpage( - 'https://www.youtube.com/feed/%s' % self._FEED_NAME, self._PLAYLIST_TITLE) - + def _entries(self, page): # The extraction process is the same as for playlists, but the regex # for the video ids doesn't contain an index ids = [] @@ -2463,12 +2861,15 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): # 'recommended' feed has infinite 'load more' and each new portion spins # the same videos in (sometimes) slightly different order, so we'll check # for unicity and break when portion has no new videos - new_ids = filter(lambda video_id: video_id not in ids, orderedSet(matches)) + new_ids = list(filter(lambda video_id: video_id not in ids, orderedSet(matches))) if not new_ids: break ids.extend(new_ids) + for entry in self._ids_to_results(new_ids): + yield entry + mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html) if not mobj: break @@ -2480,8 +2881,12 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): content_html = more['content_html'] more_widget_html = more['load_more_widget_html'] + def _real_extract(self, url): + page = self._download_webpage( + 'https://www.youtube.com/feed/%s' % self._FEED_NAME, + self._PLAYLIST_TITLE) return self.playlist_result( - self._ids_to_results(ids), playlist_title=self._PLAYLIST_TITLE) + self._entries(page), playlist_title=self._PLAYLIST_TITLE) class YoutubeWatchLaterIE(YoutubePlaylistIE):