X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/a6abd8dc822fb70852437ab5a77ced5f633739f8..76d85602f8a22ca3817c3a86f4f0e8969c0b02a9:/youtube_dl/extractor/youtube.py diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index e4b26b8..687e0b4 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -11,35 +11,44 @@ import time import traceback from .common import InfoExtractor, SearchInfoExtractor -from .subtitles import SubtitlesInfoExtractor from ..jsinterp import JSInterpreter from ..swfinterp import SWFInterpreter from ..compat import ( compat_chr, compat_parse_qs, compat_urllib_parse, + compat_urllib_parse_unquote, + compat_urllib_parse_unquote_plus, + compat_urllib_parse_urlparse, compat_urllib_request, compat_urlparse, compat_str, ) from ..utils import ( clean_html, + encode_dict, ExtractorError, + float_or_none, get_element_by_attribute, get_element_by_id, int_or_none, - OnDemandPagedList, orderedSet, + parse_duration, + remove_start, + smuggle_url, + str_to_int, unescapeHTML, unified_strdate, + unsmuggle_url, uppercase_escape, + ISO3166Utils, ) class YoutubeBaseInfoExtractor(InfoExtractor): """Provide base functions for Youtube extractors""" _LOGIN_URL = 'https://accounts.google.com/ServiceLogin' - _TWOFACTOR_URL = 'https://accounts.google.com/SecondFactor' + _TWOFACTOR_URL = 'https://accounts.google.com/signin/challenge' _NETRC_MACHINE = 'youtube' # If True it will raise an error if no login info is provided _LOGIN_REQUIRED = False @@ -50,6 +59,11 @@ class YoutubeBaseInfoExtractor(InfoExtractor): # YouTube sets the expire time to about two months expire_time=time.time() + 2 * 30 * 24 * 3600) + def _ids_to_results(self, ids): + return [ + self.url_result(vid_id, 'Youtube', video_id=vid_id) + for vid_id in ids] + def _login(self): """ Attempt to log in to YouTube. @@ -98,10 +112,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor): 'hl': 'en_US', } - # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode - # chokes on unicode - login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in login_form_strs.items()) - login_data = compat_urllib_parse.urlencode(login_form).encode('ascii') + login_data = compat_urllib_parse.urlencode(encode_dict(login_form_strs)).encode('ascii') req = compat_urllib_request.Request(self._LOGIN_URL, login_data) login_results = self._download_webpage( @@ -116,42 +127,25 @@ class YoutubeBaseInfoExtractor(InfoExtractor): # Two-Factor # TODO add SMS and phone call support - these require making a request and then prompting the user - if re.search(r'(?i)]* id="gaia_secondfactorform"', login_results) is not None: - tfa_code = self._get_tfa_info() + if re.search(r'(?i)]* id="challenge"', login_results) is not None: + tfa_code = self._get_tfa_info('2-step verification code') - if tfa_code is None: - self._downloader.report_warning('Two-factor authentication required. Provide it with --twofactor ') - self._downloader.report_warning('(Note that only TOTP (Google Authenticator App) codes work at this time.)') + 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 - # Unlike the first login form, secTok and timeStmp are both required for the TFA form - - match = re.search(r'id="secTok"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U) - if match is None: - self._downloader.report_warning('Failed to get secTok - did the page structure change?') - secTok = match.group(1) - match = re.search(r'id="timeStmp"\n\s+value=\'(.+)\'/>', login_results, re.M | re.U) - if match is None: - self._downloader.report_warning('Failed to get timeStmp - did the page structure change?') - timeStmp = match.group(1) - - tfa_form_strs = { - 'continue': 'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1', - 'smsToken': '', - 'smsUserPin': tfa_code, - 'smsVerifyPin': 'Verify', - - 'PersistentCookie': 'yes', - 'checkConnection': '', - 'checkedDomains': 'youtube', - 'pstMsg': '1', - 'secTok': secTok, - 'timeStmp': timeStmp, - 'service': 'youtube', - 'hl': 'en_US', - } - tfa_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in tfa_form_strs.items()) - tfa_data = compat_urllib_parse.urlencode(tfa_form).encode('ascii') + tfa_code = remove_start(tfa_code, 'G-') + + tfa_form_strs = self._form_hidden_inputs('challenge', login_results) + + tfa_form_strs.update({ + 'Pin': tfa_code, + 'TrustDevice': 'on', + }) + + tfa_data = compat_urllib_parse.urlencode(encode_dict(tfa_form_strs)).encode('ascii') tfa_req = compat_urllib_request.Request(self._TWOFACTOR_URL, tfa_data) tfa_results = self._download_webpage( @@ -161,8 +155,8 @@ class YoutubeBaseInfoExtractor(InfoExtractor): if tfa_results is False: return False - if re.search(r'(?i)]* id="gaia_secondfactorform"', tfa_results) is not None: - self._downloader.report_warning('Two-factor code expired. Please try again, or use a one-use backup code instead.') + 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?') @@ -184,7 +178,53 @@ class YoutubeBaseInfoExtractor(InfoExtractor): return -class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): +class YoutubePlaylistBaseInfoExtractor(InfoExtractor): + # Extract the video ids from the playlist pages + def _entries(self, page, playlist_id): + more_widget_html = content_html = page + for page_num in itertools.count(1): + for video_id, video_title in self.extract_videos_from_page(content_html): + yield self.url_result( + video_id, 'Youtube', video_id=video_id, + video_title=video_title) + + mobj = re.search(r'data-uix-load-more-href="/?(?P[^"]+)"', more_widget_html) + if not mobj: + break + + more = self._download_json( + 'https://youtube.com/%s' % mobj.group('more'), playlist_id, + 'Downloading page #%s' % page_num, + transform_source=uppercase_escape) + content_html = more['content_html'] + if not content_html.strip(): + # Some webpages show a "Load more" button but they don't + # have more videos + break + more_widget_html = more['load_more_widget_html'] + + def extract_videos_from_page(self, page): + ids_in_page = [] + titles_in_page = [] + for mobj in re.finditer(self._VIDEO_RE, page): + # The link with index 0 is not the first video of the playlist (not sure if still actual) + if 'index' in mobj.groupdict() and mobj.group('id') == '0': + continue + video_id = mobj.group('id') + video_title = unescapeHTML(mobj.group('title')) + if video_title: + video_title = video_title.strip() + try: + idx = ids_in_page.index(video_id) + if video_title and not titles_in_page[idx]: + titles_in_page[idx] = video_title + except ValueError: + ids_in_page.append(video_id) + titles_in_page.append(video_title) + return zip(ids_in_page, titles_in_page) + + +class YoutubeIE(YoutubeBaseInfoExtractor): IE_DESC = 'YouTube.com' _VALID_URL = r"""(?x)^ ( @@ -201,11 +241,14 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): |(?: # or the v= param in all its forms (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx) (?:\?|\#!?) # the params delimiter ? or # or #! - (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx) + (?:.*?&)?? # any other preceding param (like /?s=tuff&v=xxxx) v= ) )) - |youtu\.be/ # just youtu.be/xxxx + |(?: + youtu\.be| # just youtu.be/xxxx + vid\.plus # or vid.plus/xxxx + )/ |(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId= ) )? # all until now is optional -> you can pass the naked ID @@ -230,6 +273,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): '44': {'ext': 'webm', 'width': 854, 'height': 480}, '45': {'ext': 'webm', 'width': 1280, 'height': 720}, '46': {'ext': 'webm', 'width': 1920, 'height': 1080}, + '59': {'ext': 'mp4', 'width': 854, 'height': 480}, + '78': {'ext': 'mp4', 'width': 854, 'height': 480}, # 3d videos @@ -269,13 +314,13 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'vcodec': 'none', 'abr': 256, 'preference': -50, 'container': 'm4a_dash'}, # Dash webm - '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}, - '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}, - '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}, - '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}, - '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}, - '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}, - '278': {'ext': 'webm', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'container': 'webm', 'vcodec': 'VP9'}, + '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40}, + '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40}, + '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40}, + '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40}, + '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40}, + '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'vp8', 'preference': -40}, + '278': {'ext': 'webm', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'container': 'webm', 'vcodec': 'vp9'}, '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, @@ -285,11 +330,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, - '302': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'}, - '303': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'}, - '308': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'}, - '313': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'VP9'}, - '315': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'VP9'}, + '302': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'vp9'}, + '303': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'vp9'}, + '308': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'vp9'}, + '313': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'vcodec': 'vp9'}, + '315': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40, 'fps': 60, 'vcodec': 'vp9'}, # Dash webm audio '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 128, 'preference': -50}, @@ -307,7 +352,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): IE_NAME = 'youtube' _TESTS = [ { - 'url': 'http://www.youtube.com/watch?v=BaW_jenozKc', + 'url': 'http://www.youtube.com/watch?v=BaW_jenozKcj&t=1s&end=9', 'info_dict': { 'id': 'BaW_jenozKc', 'ext': 'mp4', @@ -317,8 +362,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'upload_date': '20121002', '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'], 'like_count': int, 'dislike_count': int, + 'start_time': 1, + 'end_time': 9, } }, { @@ -329,7 +377,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'ext': 'mp4', 'upload_date': '20120506', 'title': 'Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]', - 'description': 'md5:fea86fda2d5a5784273df5c7cc994d9f', + 'description': 'md5:782e8651347686cba06e58f71ab51773', + 'tags': ['Icona Pop i love it', 'sweden', 'pop music', 'big beat records', 'big beat', 'charli', + 'xcx', 'charli xcx', 'girls', 'hbo', 'i love it', "i don't care", 'icona', 'pop', + 'iconic ep', 'iconic', 'love', 'it'], 'uploader': 'Icona Pop', 'uploader_id': 'IconaPop', } @@ -345,6 +396,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'description': 'md5:64249768eec3bc4276236606ea996373', 'uploader': 'justintimberlakeVEVO', 'uploader_id': 'justintimberlakeVEVO', + 'age_limit': 18, } }, { @@ -360,6 +412,26 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'uploader_id': 'setindia' } }, + { + 'url': 'http://www.youtube.com/watch?v=BaW_jenozKcj&v=UxxajLWwzqY', + 'note': 'Use the first video ID in the URL', + 'info_dict': { + 'id': 'BaW_jenozKc', + 'ext': 'mp4', + 'title': 'youtube-dl test video "\'/\\ä↭𝕐', + 'uploader': 'Philipp Hagemeister', + 'uploader_id': 'phihag', + 'upload_date': '20121002', + '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'], + 'like_count': int, + 'dislike_count': int, + }, + 'params': { + 'skip_download': True, + }, + }, { 'url': 'http://www.youtube.com/watch?v=a9LDPn-MO4I', 'note': '256k DASH audio (format 141) via DASH manifest', @@ -401,7 +473,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'id': 'nfWlot6h_JM', 'ext': 'm4a', 'title': 'Taylor Swift - Shake It Off', - 'description': 'md5:2acfda1b285bdd478ccec22f9918199d', + 'description': 'md5:95f66187cd7c8b2c13eb78e1223b63c3', 'uploader': 'TaylorSwiftVEVO', 'uploader_id': 'TaylorSwiftVEVO', 'upload_date': '20140818', @@ -435,6 +507,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'uploader': 'The Witcher', 'uploader_id': 'WitcherGame', 'upload_date': '20140605', + 'age_limit': 18, }, }, # Age-gate video with encrypted signature @@ -448,6 +521,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'uploader': 'LloydVEVO', 'uploader_id': 'LloydVEVO', 'upload_date': '20110629', + 'age_limit': 18, }, }, # video_info is None (https://github.com/rg3/youtube-dl/issues/4421) @@ -472,7 +546,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'info_dict': { 'id': 'lqQg6PlCWgI', 'ext': 'mp4', - 'upload_date': '20120731', + 'upload_date': '20120724', 'uploader_id': 'olympic', 'description': 'HO09 - Women - GER-AUS - Hockey - 31 July 2012 - London 2012 Olympic Games', 'uploader': 'Olympics', @@ -495,6 +569,111 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'uploader': '孫艾倫', 'title': '[A-made] 變態妍字幕版 太妍 我就是這樣的人', }, + }, + # url_encoded_fmt_stream_map is empty string + { + 'url': 'qEJwOuvDf7I', + 'info_dict': { + 'id': 'qEJwOuvDf7I', + 'ext': 'webm', + 'title': 'Обсуждение судебной практики по выборам 14 сентября 2014 года в Санкт-Петербурге', + 'description': '', + 'upload_date': '20150404', + 'uploader_id': 'spbelect', + 'uploader': 'Наблюдатели Петербурга', + }, + 'params': { + 'skip_download': 'requires avconv', + } + }, + # Extraction from multiple DASH manifests (https://github.com/rg3/youtube-dl/pull/6097) + { + 'url': 'https://www.youtube.com/watch?v=FIl7x6_3R5Y', + 'info_dict': { + 'id': 'FIl7x6_3R5Y', + 'ext': 'mp4', + 'title': 'md5:7b81415841e02ecd4313668cde88737a', + 'description': 'md5:116377fd2963b81ec4ce64b542173306', + 'upload_date': '20150625', + 'uploader_id': 'dorappi2000', + 'uploader': 'dorappi2000', + 'formats': 'mincount:33', + }, + }, + # DASH manifest with segment_list + { + 'url': 'https://www.youtube.com/embed/CsmdDsKjzN8', + 'md5': '8ce563a1d667b599d21064e982ab9e31', + 'info_dict': { + 'id': 'CsmdDsKjzN8', + 'ext': 'mp4', + 'upload_date': '20150501', # According to '[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player)?\.(?P[a-z]+)$', + r'.*?-(?P[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player(?:-new)?|/base)?\.(?P[a-z]+)$', player_url) if not id_m: raise ExtractorError('Cannot identify player %r' % player_url) @@ -540,26 +719,30 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): if cache_spec is not None: return lambda s: ''.join(s[i] for i in cache_spec) + download_note = ( + 'Downloading player %s' % player_url + if self._downloader.params.get('verbose') else + 'Downloading %s player %s' % (player_type, player_id) + ) if player_type == 'js': code = self._download_webpage( player_url, video_id, - note='Downloading %s player %s' % (player_type, player_id), + note=download_note, errnote='Download of %s failed' % player_url) res = self._parse_sig_js(code) elif player_type == 'swf': urlh = self._request_webpage( player_url, video_id, - note='Downloading %s player %s' % (player_type, player_id), + note=download_note, errnote='Download of %s failed' % player_url) code = urlh.read() res = self._parse_sig_swf(code) else: assert False, 'Invalid player type %r' % player_type - if cache_spec is None: - test_string = ''.join(map(compat_chr, range(len(example_sig)))) - cache_res = res(test_string) - cache_spec = [ord(c) for c in cache_res] + test_string = ''.join(map(compat_chr, range(len(example_sig)))) + cache_res = res(test_string) + cache_spec = [ord(c) for c in cache_res] self._downloader.cache.store('youtube-sigfuncs', func_id, cache_spec) return res @@ -643,7 +826,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): raise ExtractorError( 'Signature extraction failed: ' + tb, cause=e) - def _get_available_subtitles(self, video_id, webpage): + def _get_subtitles(self, video_id, webpage): try: subs_doc = self._download_xml( 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id, @@ -657,23 +840,27 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): lang = track.attrib['lang_code'] if lang in sub_lang_list: continue - params = compat_urllib_parse.urlencode({ - 'lang': lang, - 'v': video_id, - 'fmt': self._downloader.params.get('subtitlesformat', 'srt'), - 'name': track.attrib['name'].encode('utf-8'), - }) - url = 'https://www.youtube.com/api/timedtext?' + params - sub_lang_list[lang] = url + sub_formats = [] + for ext in ['sbv', 'vtt', 'srt']: + params = compat_urllib_parse.urlencode({ + 'lang': lang, + 'v': video_id, + 'fmt': ext, + 'name': track.attrib['name'].encode('utf-8'), + }) + sub_formats.append({ + 'url': 'https://www.youtube.com/api/timedtext?' + params, + 'ext': ext, + }) + sub_lang_list[lang] = sub_formats if not sub_lang_list: self._downloader.report_warning('video doesn\'t have subtitles') return {} return sub_lang_list - def _get_available_automatic_caption(self, video_id, webpage): + def _get_automatic_captions(self, video_id, webpage): """We need the webpage for getting the captions url, pass it as an argument to speed up the process.""" - sub_format = self._downloader.params.get('subtitlesformat', 'srt') self.to_screen('%s: Looking for automatic captions' % video_id) mobj = re.search(r';ytplayer.config = ({.*?});', webpage) err_msg = 'Couldn\'t find automatic captions for %s' % video_id @@ -703,14 +890,20 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): sub_lang_list = {} for lang_node in caption_list.findall('target'): sub_lang = lang_node.attrib['lang_code'] - params = compat_urllib_parse.urlencode({ - 'lang': original_lang, - 'tlang': sub_lang, - 'fmt': sub_format, - 'ts': timestamp, - 'kind': caption_kind, - }) - sub_lang_list[sub_lang] = caption_url + '&' + params + sub_formats = [] + for ext in ['sbv', 'vtt', 'srt']: + params = compat_urllib_parse.urlencode({ + 'lang': original_lang, + 'tlang': sub_lang, + 'fmt': ext, + 'ts': timestamp, + 'kind': caption_kind, + }) + sub_formats.append({ + 'url': caption_url + '&' + params, + 'ext': ext, + }) + sub_lang_list[sub_lang] = sub_formats return sub_lang_list # An extractor error can be raise by the download process if there are # no automatic captions but there are subtitles @@ -746,55 +939,95 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): return self._download_webpage(url, video_id, note='Searching for annotations.', errnote='Unable to download video annotations.') def _parse_dash_manifest( - self, video_id, dash_manifest_url, player_url, age_gate): + self, video_id, dash_manifest_url, player_url, age_gate, fatal=True): def decrypt_sig(mobj): s = mobj.group(1) dec_s = self._decrypt_signature(s, video_id, player_url, age_gate) return '/signature/%s' % dec_s - dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url) + dash_manifest_url = re.sub(r'/s/([a-fA-F0-9\.]+)', decrypt_sig, dash_manifest_url) dash_doc = self._download_xml( dash_manifest_url, video_id, note='Downloading DASH manifest', - errnote='Could not download DASH manifest') + errnote='Could not download DASH manifest', + fatal=fatal) + + if dash_doc is False: + return [] formats = [] - for r in dash_doc.findall('.//{urn:mpeg:DASH:schema:MPD:2011}Representation'): - url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL') - if url_el is None: - continue - format_id = r.attrib['id'] - video_url = url_el.text - filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength')) - f = { - 'format_id': format_id, - 'url': video_url, - 'width': int_or_none(r.attrib.get('width')), - 'height': int_or_none(r.attrib.get('height')), - 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000), - 'asr': int_or_none(r.attrib.get('audioSamplingRate')), - 'filesize': filesize, - 'fps': int_or_none(r.attrib.get('frameRate')), - } - try: - existing_format = next( - fo for fo in formats - if fo['format_id'] == format_id) - except StopIteration: - f.update(self._formats.get(format_id, {}).items()) - formats.append(f) - else: - existing_format.update(f) + for a in dash_doc.findall('.//{urn:mpeg:DASH:schema:MPD:2011}AdaptationSet'): + mime_type = a.attrib.get('mimeType') + for r in a.findall('{urn:mpeg:DASH:schema:MPD:2011}Representation'): + url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL') + if url_el is None: + continue + if mime_type == 'text/vtt': + # TODO implement WebVTT downloading + pass + elif mime_type.startswith('audio/') or mime_type.startswith('video/'): + segment_list = r.find('{urn:mpeg:DASH:schema:MPD:2011}SegmentList') + format_id = r.attrib['id'] + video_url = url_el.text + filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength')) + f = { + 'format_id': format_id, + 'url': video_url, + 'width': int_or_none(r.attrib.get('width')), + 'height': int_or_none(r.attrib.get('height')), + 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000), + 'asr': int_or_none(r.attrib.get('audioSamplingRate')), + 'filesize': filesize, + 'fps': int_or_none(r.attrib.get('frameRate')), + } + if segment_list is not None: + f.update({ + 'initialization_url': segment_list.find('{urn:mpeg:DASH:schema:MPD:2011}Initialization').attrib['sourceURL'], + 'segment_urls': [segment.attrib.get('media') for segment in segment_list.findall('{urn:mpeg:DASH:schema:MPD:2011}SegmentURL')], + 'protocol': 'http_dash_segments', + }) + try: + existing_format = next( + fo for fo in formats + if fo['format_id'] == format_id) + except StopIteration: + full_info = self._formats.get(format_id, {}).copy() + full_info.update(f) + codecs = r.attrib.get('codecs') + if codecs: + if full_info.get('acodec') == 'none' and 'vcodec' not in full_info: + full_info['vcodec'] = codecs + elif full_info.get('vcodec') == 'none' and 'acodec' not in full_info: + full_info['acodec'] = codecs + formats.append(full_info) + else: + existing_format.update(f) + else: + self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type) return formats def _real_extract(self, url): + url, smuggled_data = unsmuggle_url(url, {}) + proto = ( 'http' if self._downloader.params.get('prefer_insecure', False) else 'https') + start_time = None + end_time = None + parsed_url = compat_urllib_parse_urlparse(url) + for component in [parsed_url.fragment, parsed_url.query]: + query = compat_parse_qs(component) + if start_time is None and 't' in query: + start_time = parse_duration(query['t'][0]) + if start_time is None and 'start' in query: + start_time = parse_duration(query['start'][0]) + if end_time is None and 'end' in query: + end_time = parse_duration(query['end'][0]) + # Extract original video URL from URL with redirection, like age verification, using next_url parameter mobj = re.search(self._NEXT_URL_RE, url) if mobj: - url = proto + '://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/') + url = proto + '://www.youtube.com/' + compat_urllib_parse_unquote(mobj.group(1)).lstrip('/') video_id = self.extract_id(url) # Get video webpage @@ -808,8 +1041,16 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): else: player_url = None + dash_mpds = [] + + def add_dash_mpd(video_info): + dash_mpd = video_info.get('dashmpd') + if dash_mpd and dash_mpd[0] not in dash_mpds: + dash_mpds.append(dash_mpd[0]) + # 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} @@ -828,24 +1069,31 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): note='Refetching age-gated info webpage', errnote='unable to download video info webpage') video_info = compat_parse_qs(video_info_webpage) + add_dash_mpd(video_info) else: age_gate = False - try: - # Try looking directly into the video webpage - mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage) - if not mobj: - raise ValueError('Could not find ytplayer.config') # caught below + video_info = None + # Try looking directly into the video webpage + mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage) + if mobj: json_code = uppercase_escape(mobj.group(1)) ytplayer_config = json.loads(json_code) args = ytplayer_config['args'] - # Convert to the same format returned by compat_parse_qs - video_info = dict((k, [v]) for k, v in args.items()) - if 'url_encoded_fmt_stream_map' not in args: - raise ValueError('No stream_map present') # caught below - except ValueError: - # We fallback to the get_video_info pages (used by the embed page) + if args.get('url_encoded_fmt_stream_map'): + # 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) + if args.get('livestream') == '1' or args.get('live_playback') == 1: + is_live = True + 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 + # are missing from DASH manifest pointed by webpage's dashmpd, some - from DASH + # manifest pointed by get_video_info's dashmpd). + # 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=embedded', '&el=detailpage', '&el=vevo', '']: + 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)) @@ -853,11 +1101,32 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): video_info_url, video_id, note=False, errnote='unable to download video info webpage') - video_info = compat_parse_qs(video_info_webpage) - if 'token' in video_info: + 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 video_info: + video_info = get_video_info + if 'token' in get_video_info: + # Different get_video_info requests may report different results, e.g. + # some may report video unavailability, but some may serve it without + # any complaint (see https://github.com/rg3/youtube-dl/issues/7362, + # the original webpage as well as el=info and el=embedded get_video_info + # requests report video unavailability due to geo restriction while + # el=detailpage succeeds and returns valid data). This is probably + # due to YouTube measures against IP ranges of hosting providers. + # Working around by preferring the first succeeded video_info containing + # the token if no such video_info yet was found. + if 'token' not in video_info: + video_info = get_video_info break 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']: + regions_allowed = self._html_search_meta('regionsAllowed', video_webpage, default=None) + if regions_allowed: + raise ExtractorError('YouTube said: This video is available in %s only' % ( + ', '.join(map(ISO3166Utils.short2full, regions_allowed.split(',')))), + expected=True) raise ExtractorError( 'YouTube said: %s' % video_info['reason'][0], expected=True, video_id=video_id) @@ -866,6 +1135,55 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): '"token" parameter not in video info for unknown reason', video_id=video_id) + # title + if 'title' in video_info: + video_title = video_info['title'][0] + else: + self._downloader.report_warning('Unable to extract video title') + video_title = '_' + + # description + video_description = get_element_by_id("eow-description", video_webpage) + if video_description: + video_description = re.sub(r'''(?x) + + [^<]+ + + ''', r'\1', video_description) + video_description = clean_html(video_description) + else: + fd_mobj = re.search(r'', @@ -908,18 +1219,19 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 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]) + video_thumbnail = compat_urllib_parse_unquote_plus(video_info['thumbnail_url'][0]) # upload date - upload_date = None - mobj = re.search(r'(?s)id="eow-date.*?>(.*?)', video_webpage) - if mobj is None: - mobj = re.search( - r'(?s)id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live) on (.*?)', - video_webpage) - if mobj is not None: - upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split()) - upload_date = unified_strdate(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) m_cat_container = self._search_regex( r'(?s)]*>\s*Category\s*\s*]*>(.*?)', @@ -932,48 +1244,28 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): else: video_categories = None - # description - video_description = get_element_by_id("eow-description", video_webpage) - if video_description: - video_description = re.sub(r'''(?x) - - [^<]+ - - ''', r'\1', video_description) - video_description = clean_html(video_description) - else: - fd_mobj = re.search(r']*>.*?([\d,]+)\s*' % re.escape(count_name), - video_webpage, count_name, default=None) - if count is not None: - return int(count.replace(',', '')) - return None + return str_to_int(self._search_regex( + r'-%s-button[^>]+>]+class="yt-uix-button-content"[^>]*>([\d,]+)' + % 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) - - if self._downloader.params.get('listsubtitles', False): - self._list_available_subtitles(video_id, video_webpage) - return + automatic_captions = self.extract_automatic_captions(video_id, video_webpage) if 'length_seconds' not in video_info: self._downloader.report_warning('unable to extract video duration') video_duration = None else: - video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0])) + video_duration = int(compat_urllib_parse_unquote_plus(video_info['length_seconds'][0])) # annotations video_annotations = None @@ -1005,7 +1297,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 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) - url_map = {} + formats = [] for url_data_str in encoded_url_map.split(','): url_data = compat_parse_qs(url_data_str) if 'itag' not in url_data or 'url' not in url_data: @@ -1051,7 +1343,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): player_desc = 'flash player %s' % player_version else: player_version = self._search_regex( - r'html5player-([^/]+?)(?:/html5player)?\.js', + [r'html5player-([^/]+?)(?:/html5player(?:-new)?)?\.js', r'(?:www|player)-([^/]+)/base\.js'], player_url, 'html5 player', fatal=False) player_desc = 'html5 player %s' % player_version @@ -1065,8 +1357,50 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): url += '&signature=' + signature if 'ratebypass' not in url: url += '&ratebypass=yes' - url_map[format_id] = url - formats = _map_to_format_list(url_map) + + # Some itags are not included in DASH manifest thus corresponding formats will + # lack metadata (see https://github.com/rg3/youtube-dl/pull/5993). + # Trying to extract metadata from url_encoded_fmt_stream_map entry. + mobj = re.search(r'^(?P\d+)[xX](?P\d+)$', url_data.get('size', [''])[0]) + width, height = (int(mobj.group('width')), int(mobj.group('height'))) if mobj else (None, None) + dct = { + 'format_id': format_id, + 'url': url, + 'player_url': player_url, + 'filesize': int_or_none(url_data.get('clen', [None])[0]), + 'tbr': float_or_none(url_data.get('bitrate', [None])[0], 1000), + 'width': width, + 'height': height, + 'fps': int_or_none(url_data.get('fps', [None])[0]), + 'format_note': url_data.get('quality_label', [None])[0] or url_data.get('quality', [None])[0], + } + type_ = url_data.get('type', [None])[0] + if type_: + type_split = type_.split(';') + kind_ext = type_split[0].split('/') + if len(kind_ext) == 2: + kind, ext = kind_ext + dct['ext'] = ext + if kind in ('audio', 'video'): + codecs = None + for mobj in re.finditer( + r'(?P[a-zA-Z_-]+)=(?P["\']?)(?P.+?)(?P=quote)(?:;|$)', type_): + if mobj.group('key') == 'codecs': + codecs = mobj.group('val') + break + if codecs: + codecs = codecs.split(',') + if len(codecs) == 2: + acodec, vcodec = codecs[0], codecs[1] + else: + acodec, vcodec = (codecs[0], 'none') if kind == 'audio' else ('none', codecs[0]) + dct.update({ + 'acodec': acodec, + 'vcodec': vcodec, + }) + if format_id in self._formats: + dct.update(self._formats[format_id]) + formats.append(dct) elif video_info.get('hlsvp'): manifest_url = video_info['hlsvp'][0] url_map = self._extract_from_m3u8(manifest_url, video_id) @@ -1076,23 +1410,32 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): # Look for the DASH manifest if self._downloader.params.get('youtube_include_dash_manifest', True): - dash_mpd = video_info.get('dashmpd') - if dash_mpd: - dash_manifest_url = dash_mpd[0] + dash_mpd_fatal = True + for dash_manifest_url in dash_mpds: + dash_formats = {} try: - dash_formats = self._parse_dash_manifest( - video_id, dash_manifest_url, player_url, age_gate) + for df in self._parse_dash_manifest( + video_id, dash_manifest_url, player_url, age_gate, dash_mpd_fatal): + # 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 + # Additional DASH manifests may end up in HTTP Error 403 therefore + # allow them to fail without bug report message if we already have + # some DASH manifest succeeded. This is temporary workaround to reduce + # burst of bug reports until we figure out the reason and whether it + # can be fixed at all. + dash_mpd_fatal = False except (ExtractorError, KeyError) as e: self.report_warning( 'Skipping DASH manifest: %r' % e, video_id) - else: - # Hide the formats we found through non-DASH - dash_keys = set(df['format_id'] for df in dash_formats) - for f in formats: - if f['format_id'] in dash_keys: - f['format_id'] = 'nondash-%s' % f['format_id'] - f['preference'] = f.get('preference', 0) - 10000 - formats.extend(dash_formats) + if dash_formats: + # Remove the formats we found through non-DASH, they + # contain less info and it can be wrong, because we use + # fixed values (for example the resolution). See + # https://github.com/rg3/youtube-dl/issues/5774 for an + # example. + formats = [f for f in formats if f['format_id'] not in dash_formats.keys()] + formats.extend(dash_formats.values()) # Check for malformed aspect ratio stretched_m = re.search( @@ -1115,7 +1458,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'thumbnail': video_thumbnail, 'description': video_description, 'categories': video_categories, + 'tags': video_tags, 'subtitles': video_subtitles, + 'automatic_captions': automatic_captions, 'duration': video_duration, 'age_limit': 18 if age_gate else 0, 'annotations': video_annotations, @@ -1123,11 +1468,15 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'view_count': view_count, 'like_count': like_count, 'dislike_count': dislike_count, + 'average_rating': float_or_none(video_info.get('avg_rating', [None])[0]), 'formats': formats, + 'is_live': is_live, + 'start_time': start_time, + 'end_time': end_time, } -class YoutubePlaylistIE(YoutubeBaseInfoExtractor): +class YoutubePlaylistIE(YoutubeBaseInfoExtractor, YoutubePlaylistBaseInfoExtractor): IE_DESC = 'YouTube.com playlists' _VALID_URL = r"""(?x)(?: (?:https?://)? @@ -1139,16 +1488,16 @@ class YoutubePlaylistIE(YoutubeBaseInfoExtractor): | p/ ) ( - (?:PL|LL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,} + (?:PL|LL|EC|UU|FL|RD|UL)?[0-9A-Za-z-_]{10,} # Top tracks, they can also include dots |(?:MC)[\w\.]* ) .* | - ((?:PL|LL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,}) + ((?:PL|LL|EC|UU|FL|RD|UL)[0-9A-Za-z-_]{10,}) )""" _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s' - _VIDEO_RE = r'href="\s*/watch\?v=(?P[0-9A-Za-z_-]{11})&[^"]*?index=(?P\d+)' + _VIDEO_RE = r'href="\s*/watch\?v=(?P[0-9A-Za-z_-]{11})&[^"]*?index=(?P\d+)(?:[^>]+>(?P[^<]+))?' IE_NAME = 'youtube:playlist' _TESTS = [{ 'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re', @@ -1224,13 +1573,8 @@ class YoutubePlaylistIE(YoutubeBaseInfoExtractor): def _real_initialize(self): self._login() - def _ids_to_results(self, ids): - return [ - self.url_result(vid_id, 'Youtube', video_id=vid_id) - for vid_id in ids] - def _extract_mix(self, playlist_id): - # The mixes are generated from a a single video + # The mixes are generated from a single video # the id of the playlist is just 'RD' + video_id url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id) webpage = self._download_webpage( @@ -1249,6 +1593,33 @@ class YoutubePlaylistIE(YoutubeBaseInfoExtractor): return self.playlist_result(url_results, playlist_id, title) + def _extract_playlist(self, playlist_id): + url = self._TEMPLATE_URL % playlist_id + page = self._download_webpage(url, playlist_id) + + for match in re.findall(r'<div class="yt-alert-message">([^<]+)</div>', page): + match = match.strip() + # Check if the playlist exists or is private + if re.match(r'[^<]*(The|This) playlist (does not exist|is private)[^<]*', match): + raise ExtractorError( + 'The playlist doesn\'t exist or is private, use --username or ' + '--netrc to access it.', + expected=True) + elif re.match(r'[^<]*Invalid parameters[^<]*', match): + raise ExtractorError( + 'Invalid parameters. Maybe URL is incorrect.', + expected=True) + elif re.match(r'[^<]*Choose your language[^<]*', match): + continue + else: + self.report_warning('Youtube gives an alert message: ' + match) + + playlist_title = self._html_search_regex( + r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>', + page, 'title') + + return self.playlist_result(self._entries(page, playlist_id), playlist_id, playlist_title) + def _real_extract(self, url): # Extract playlist id mobj = re.match(self._VALID_URL, url) @@ -1266,80 +1637,64 @@ class YoutubePlaylistIE(YoutubeBaseInfoExtractor): else: self.to_screen('Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id)) - if playlist_id.startswith('RD'): + if playlist_id.startswith('RD') or playlist_id.startswith('UL'): # Mixes require a custom extraction process return self._extract_mix(playlist_id) - url = self._TEMPLATE_URL % playlist_id - page = self._download_webpage(url, playlist_id) - more_widget_html = content_html = page - - # Check if the playlist exists or is private - if re.search(r'<div class="yt-alert-message">[^<]*?(The|This) playlist (does not exist|is private)[^<]*?</div>', page) is not None: - raise ExtractorError( - 'The playlist doesn\'t exist or is private, use --username or ' - '--netrc to access it.', - expected=True) - - # Extract the video ids from the playlist pages - ids = [] + return self._extract_playlist(playlist_id) - for page_num in itertools.count(1): - matches = re.finditer(self._VIDEO_RE, content_html) - # We remove the duplicates and the link with index 0 - # (it's not the first video of the playlist) - new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0') - ids.extend(new_ids) - mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html) - if not mobj: - break - - more = self._download_json( - 'https://youtube.com/%s' % mobj.group('more'), playlist_id, - 'Downloading page #%s' % page_num, - transform_source=uppercase_escape) - content_html = more['content_html'] - if not content_html.strip(): - # Some webpages show a "Load more" button but they don't - # have more videos - break - more_widget_html = more['load_more_widget_html'] - - playlist_title = self._html_search_regex( - r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>', - page, 'title') - - url_results = self._ids_to_results(ids) - return self.playlist_result(url_results, playlist_id, playlist_title) - - -class YoutubeChannelIE(InfoExtractor): +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_-]+)' + _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' _TESTS = [{ 'note': 'paginated channel', 'url': 'https://www.youtube.com/channel/UCKfVa3S1e4PHvxWcwyMMg8w', 'playlist_mincount': 91, 'info_dict': { - 'id': 'UCKfVa3S1e4PHvxWcwyMMg8w', + 'id': 'UUKfVa3S1e4PHvxWcwyMMg8w', + 'title': 'Uploads from lex will', } + }, { + 'note': 'Age restricted channel', + # from https://www.youtube.com/user/DeusExOfficial + 'url': 'https://www.youtube.com/channel/UCs0ifCMCm1icqRbqhUINa0w', + 'playlist_mincount': 64, + 'info_dict': { + 'id': 'UUs0ifCMCm1icqRbqhUINa0w', + 'title': 'Uploads from Deus Ex', + }, }] - def extract_videos_from_page(self, page): - ids_in_page = [] - for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page): - if mobj.group(1) not in ids_in_page: - ids_in_page.append(mobj.group(1)) - return ids_in_page - def _real_extract(self, url): channel_id = self._match_id(url) - video_ids = [] - url = 'https://www.youtube.com/channel/%s/videos' % channel_id - channel_page = self._download_webpage(url, channel_id) + url = self._TEMPLATE_URL % channel_id + + # Channel by page listing is restricted to 35 pages of 30 items, i.e. 1050 videos total (see #5778) + # Workaround by extracting as a playlist if managed to obtain channel playlist URL + # otherwise fallback on channel by page extraction + channel_page = self._download_webpage( + url + '?view=57', channel_id, + 'Downloading channel page', fatal=False) + if channel_page is False: + channel_playlist_id = False + else: + channel_playlist_id = self._html_search_meta( + 'channelId', channel_page, 'channel id', default=None) + if not channel_playlist_id: + channel_playlist_id = self._search_regex( + r'data-(?:channel-external-|yt)id="([^"]+)"', + channel_page, 'channel id', default=None) + if channel_playlist_id and channel_playlist_id.startswith('UC'): + playlist_id = 'UU' + channel_playlist_id[2:] + return self.url_result( + compat_urlparse.urljoin(url, '/playlist?list=%s' % playlist_id), 'YoutubePlaylist') + + channel_page = self._download_webpage(url, channel_id, 'Downloading page #1') autogenerated = re.search(r'''(?x) class="[^"]*?(?: channel-header-autogenerated-label| @@ -1349,43 +1704,20 @@ class YoutubeChannelIE(InfoExtractor): if autogenerated: # The videos are contained in a single page # the ajax pages can't be used, they are empty - video_ids = self.extract_videos_from_page(channel_page) entries = [ - self.url_result(video_id, 'Youtube', video_id=video_id) - for video_id in video_ids] + self.url_result( + video_id, 'Youtube', video_id=video_id, + video_title=video_title) + for video_id, video_title in self.extract_videos_from_page(channel_page)] return self.playlist_result(entries, channel_id) - def _entries(): - more_widget_html = content_html = channel_page - for pagenum in itertools.count(1): + return self.playlist_result(self._entries(channel_page, channel_id), channel_id) - ids_in_page = self.extract_videos_from_page(content_html) - for video_id in ids_in_page: - yield self.url_result( - video_id, 'Youtube', video_id=video_id) - mobj = re.search( - r'data-uix-load-more-href="/?(?P<more>[^"]+)"', - more_widget_html) - if not mobj: - break - - more = self._download_json( - 'https://youtube.com/%s' % mobj.group('more'), channel_id, - 'Downloading page #%s' % (pagenum + 1), - transform_source=uppercase_escape) - content_html = more['content_html'] - more_widget_html = more['load_more_widget_html'] - - return self.playlist_result(_entries(), channel_id) - - -class YoutubeUserIE(InfoExtractor): +class YoutubeUserIE(YoutubeChannelIE): IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)' _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)(?P<id>[A-Za-z0-9_-]+)' - _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/users/%s' - _GDATA_PAGE_SIZE = 50 - _GDATA_URL = 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json' + _TEMPLATE_URL = 'https://www.youtube.com/user/%s/videos' IE_NAME = 'youtube:user' _TESTS = [{ @@ -1409,95 +1741,57 @@ class YoutubeUserIE(InfoExtractor): else: return super(YoutubeUserIE, cls).suitable(url) - def _real_extract(self, url): - username = self._match_id(url) - - # Download video ids using YouTube Data API. Result size per - # query is limited (currently to 50 videos) so we need to query - # page by page until there are no video ids - it means we got - # all of them. - - def download_page(pagenum): - start_index = pagenum * self._GDATA_PAGE_SIZE + 1 - - gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index) - page = self._download_webpage( - gdata_url, username, - 'Downloading video ids from %d to %d' % ( - start_index, start_index + self._GDATA_PAGE_SIZE)) - - try: - response = json.loads(page) - except ValueError as err: - raise ExtractorError('Invalid JSON in API response: ' + compat_str(err)) - if 'entry' not in response['feed']: - return - - # Extract video identifiers - entries = response['feed']['entry'] - for entry in entries: - title = entry['title']['$t'] - video_id = entry['id']['$t'].split('/')[-1] - yield { - '_type': 'url', - 'url': video_id, - 'ie_key': 'Youtube', - 'id': video_id, - 'title': title, - } - url_results = OnDemandPagedList(download_page, self._GDATA_PAGE_SIZE) - - return self.playlist_result(url_results, playlist_title=username) - -class YoutubeSearchIE(SearchInfoExtractor): +class YoutubeSearchIE(SearchInfoExtractor, YoutubePlaylistIE): IE_DESC = 'YouTube.com searches' - _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc' - _MAX_RESULTS = 1000 + # there doesn't appear to be a real limit, for example if you search for + # 'python' you get more than 8.000.000 results + _MAX_RESULTS = float('inf') IE_NAME = 'youtube:search' _SEARCH_KEY = 'ytsearch' + _EXTRA_QUERY_ARGS = {} + _TESTS = [] def _get_n_results(self, query, n): """Get a specified number of results for a query""" - video_ids = [] - pagenum = 0 + videos = [] limit = n - PAGE_SIZE = 50 - while (PAGE_SIZE * pagenum) < limit: - result_url = self._API_URL % ( - compat_urllib_parse.quote_plus(query.encode('utf-8')), - (PAGE_SIZE * pagenum) + 1) - data_json = self._download_webpage( + for pagenum in itertools.count(1): + url_query = { + 'search_query': query.encode('utf-8'), + 'page': pagenum, + 'spf': 'navigate', + } + url_query.update(self._EXTRA_QUERY_ARGS) + result_url = 'https://www.youtube.com/results?' + compat_urllib_parse.urlencode(url_query) + data = self._download_json( result_url, video_id='query "%s"' % query, - note='Downloading page %s' % (pagenum + 1), + note='Downloading page %s' % pagenum, errnote='Unable to download API page') - data = json.loads(data_json) - api_response = data['data'] + html_content = data[1]['body']['content'] - if 'items' not in api_response: + if 'class="search-message' in html_content: raise ExtractorError( '[youtube] No video results', expected=True) - new_ids = list(video['id'] for video in api_response['items']) - video_ids += new_ids - - limit = min(n, api_response['totalItems']) - pagenum += 1 + new_videos = self._ids_to_results(orderedSet(re.findall( + r'href="/watch\?v=(.{11})', html_content))) + videos += new_videos + if not new_videos or len(videos) > limit: + break - if len(video_ids) > n: - video_ids = video_ids[:n] - videos = [self.url_result(video_id, 'Youtube', video_id=video_id) - for video_id in video_ids] + if len(videos) > n: + videos = videos[:n] return self.playlist_result(videos, query) class YoutubeSearchDateIE(YoutubeSearchIE): IE_NAME = YoutubeSearchIE.IE_NAME + ':date' - _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published' _SEARCH_KEY = 'ytsearchdate' IE_DESC = 'YouTube.com searches, newest videos first' + _EXTRA_QUERY_ARGS = {'search_sort': 'video_date_uploaded'} class YoutubeSearchURLIE(InfoExtractor): @@ -1514,14 +1808,14 @@ class YoutubeSearchURLIE(InfoExtractor): def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) - query = compat_urllib_parse.unquote_plus(mobj.group('query')) + query = compat_urllib_parse_unquote_plus(mobj.group('query')) webpage = self._download_webpage(url, query) result_code = self._search_regex( - r'(?s)<ol class="item-section"(.*?)</ol>', webpage, 'result HTML') + r'(?s)<ol[^>]+class="item-section"(.*?)</ol>', webpage, 'result HTML') part_codes = re.findall( - r'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code) + r'(?s)<h3[^>]+class="[^"]*yt-lockup-title[^"]*"[^>]*>(.*?)</h3>', result_code) entries = [] for part_code in part_codes: part_title = self._html_search_regex( @@ -1548,8 +1842,8 @@ class YoutubeShowIE(InfoExtractor): _VALID_URL = r'https?://www\.youtube\.com/show/(?P<id>[^?#]*)' IE_NAME = 'youtube:show' _TESTS = [{ - 'url': 'http://www.youtube.com/show/airdisasters', - 'playlist_mincount': 3, + 'url': 'https://www.youtube.com/show/airdisasters', + 'playlist_mincount': 5, 'info_dict': { 'id': 'airdisasters', 'title': 'Air Disasters', @@ -1560,7 +1854,7 @@ class YoutubeShowIE(InfoExtractor): mobj = re.match(self._VALID_URL, url) playlist_id = mobj.group('id') webpage = self._download_webpage( - url, playlist_id, 'Downloading show webpage') + 'https://www.youtube.com/show/%s/playlists' % playlist_id, playlist_id, 'Downloading show webpage') # There's one playlist for each season of the show m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage)) self.to_screen('%s: Found %s seasons' % (playlist_id, len(m_seasons))) @@ -1581,20 +1875,10 @@ class YoutubeShowIE(InfoExtractor): class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): """ - Base class for extractors that fetch info from - http://www.youtube.com/feed_ajax + Base class for feed extractors Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties. """ _LOGIN_REQUIRED = True - # use action_load_personal_feed instead of action_load_system_feed - _PERSONAL_FEED = False - - @property - def _FEED_TEMPLATE(self): - action = 'action_load_system_feed' - if self._PERSONAL_FEED: - action = 'action_load_personal_feed' - return 'https://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME) @property def IE_NAME(self): @@ -1604,51 +1888,49 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): self._login() def _real_extract(self, url): - feed_entries = [] - paging = 0 - for i in itertools.count(1): - info = self._download_json( - self._FEED_TEMPLATE % paging, - '%s feed' % self._FEED_NAME, - 'Downloading page %s' % i, - transform_source=uppercase_escape) - feed_html = info.get('feed_html') or info.get('content_html') - load_more_widget_html = info.get('load_more_widget_html') or feed_html - m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html) - ids = orderedSet(m.group(1) for m in m_ids) - feed_entries.extend( - self.url_result(video_id, 'Youtube', video_id=video_id) - for video_id in ids) - mobj = re.search( - r'data-uix-load-more-href="/?[^"]+paging=(?P<paging>\d+)', - load_more_widget_html) - if mobj is None: + page = self._download_webpage( + 'https://www.youtube.com/feed/%s' % self._FEED_NAME, self._PLAYLIST_TITLE) + + # The extraction process is the same as for playlists, but the regex + # for the video ids doesn't contain an index + ids = [] + more_widget_html = content_html = page + for page_num in itertools.count(1): + matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html) + + # '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)) + if not new_ids: break - paging = mobj.group('paging') - return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE) + ids.extend(new_ids) -class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor): - IE_DESC = 'YouTube.com recommended videos, ":ytrec" for short (requires authentication)' - _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?' - _FEED_NAME = 'recommended' - _PLAYLIST_TITLE = 'Youtube Recommended videos' + mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html) + if not mobj: + break + + more = self._download_json( + 'https://youtube.com/%s' % mobj.group('more'), self._PLAYLIST_TITLE, + 'Downloading page #%s' % page_num, + transform_source=uppercase_escape) + content_html = more['content_html'] + more_widget_html = more['load_more_widget_html'] + return self.playlist_result( + self._ids_to_results(ids), playlist_title=self._PLAYLIST_TITLE) -class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor): + +class YoutubeWatchLaterIE(YoutubePlaylistIE): + IE_NAME = 'youtube:watchlater' IE_DESC = 'Youtube watch later list, ":ytwatchlater" for short (requires authentication)' - _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater' - _FEED_NAME = 'watch_later' - _PLAYLIST_TITLE = 'Youtube Watch Later' - _PERSONAL_FEED = True + _VALID_URL = r'https?://www\.youtube\.com/(?:feed/watch_later|playlist\?list=WL)|:ytwatchlater' + _TESTS = [] # override PlaylistIE tests -class YoutubeHistoryIE(YoutubeFeedsInfoExtractor): - IE_DESC = 'Youtube watch history, ":ythistory" for short (requires authentication)' - _VALID_URL = 'https?://www\.youtube\.com/feed/history|:ythistory' - _FEED_NAME = 'history' - _PERSONAL_FEED = True - _PLAYLIST_TITLE = 'Youtube Watch History' + def _real_extract(self, url): + return self._extract_playlist('WL') class YoutubeFavouritesIE(YoutubeBaseInfoExtractor): @@ -1663,42 +1945,25 @@ class YoutubeFavouritesIE(YoutubeBaseInfoExtractor): return self.url_result(playlist_id, 'YoutubePlaylist') -class YoutubeSubscriptionsIE(YoutubePlaylistIE): - IE_NAME = 'youtube:subscriptions' - IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)' - _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?' - _TESTS = [] - - def _real_extract(self, url): - title = 'Youtube Subscriptions' - page = self._download_webpage('https://www.youtube.com/feed/subscriptions', title) - - # The extraction process is the same as for playlists, but the regex - # for the video ids doesn't contain an index - ids = [] - more_widget_html = content_html = page +class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor): + IE_DESC = 'YouTube.com recommended videos, ":ytrec" for short (requires authentication)' + _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?' + _FEED_NAME = 'recommended' + _PLAYLIST_TITLE = 'Youtube Recommended videos' - for page_num in itertools.count(1): - matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html) - new_ids = orderedSet(matches) - ids.extend(new_ids) - mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html) - if not mobj: - break +class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor): + IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)' + _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?' + _FEED_NAME = 'subscriptions' + _PLAYLIST_TITLE = 'Youtube Subscriptions' - more = self._download_json( - 'https://youtube.com/%s' % mobj.group('more'), title, - 'Downloading page #%s' % page_num, - transform_source=uppercase_escape) - content_html = more['content_html'] - more_widget_html = more['load_more_widget_html'] - return { - '_type': 'playlist', - 'title': title, - 'entries': self._ids_to_results(ids), - } +class YoutubeHistoryIE(YoutubeFeedsInfoExtractor): + IE_DESC = 'Youtube watch history, ":ythistory" for short (requires authentication)' + _VALID_URL = 'https?://www\.youtube\.com/feed/history|:ythistory' + _FEED_NAME = 'history' + _PLAYLIST_TITLE = 'Youtube History' class YoutubeTruncatedURLIE(InfoExtractor): @@ -1712,6 +1977,7 @@ class YoutubeTruncatedURLIE(InfoExtractor): annotation_id=annotation_[^&]+| x-yt-cl=[0-9]+| hl=[^&]*| + t=[0-9]+ )? | attribution_link\?a=[^&]+ @@ -1734,6 +2000,9 @@ class YoutubeTruncatedURLIE(InfoExtractor): }, { 'url': 'https://www.youtube.com/watch?hl=en-GB', 'only_matching': True, + }, { + 'url': 'https://www.youtube.com/watch?t=2372', + 'only_matching': True, }] def _real_extract(self, url):