X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/9113dfef91df19343cf76c6274dd0a85258c1004..0fc91537a0580721211b68e8cc4367a33cd6d4f2:/youtube_dl/extractor/youtube.py diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index 0919bef..89c8b7f 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -37,6 +37,7 @@ from ..utils import ( orderedSet, parse_codecs, parse_duration, + qualities, remove_quotes, remove_start, smuggle_url, @@ -84,10 +85,10 @@ 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 @@ -246,9 +247,9 @@ class YoutubeBaseInfoExtractor(InfoExtractor): return True - def _download_webpage(self, *args, **kwargs): + def _download_webpage_handle(self, *args, **kwargs): kwargs.setdefault('query', {})['disable_polymer'] = 'true' - return super(YoutubeBaseInfoExtractor, self)._download_webpage( + return super(YoutubeBaseInfoExtractor, self)._download_webpage_handle( *args, **compat_kwargs(kwargs)) def _real_initialize(self): @@ -509,6 +510,8 @@ 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', } }, { @@ -527,6 +530,8 @@ 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, } }, @@ -596,7 +601,7 @@ 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', @@ -637,7 +642,7 @@ 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', @@ -667,10 +672,10 @@ 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', @@ -732,7 +737,7 @@ 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', }, @@ -759,7 +764,7 @@ 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, @@ -768,8 +773,9 @@ 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 { @@ -884,7 +890,7 @@ 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', @@ -892,7 +898,9 @@ 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, @@ -949,7 +957,7 @@ 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)', @@ -984,6 +992,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'params': { 'skip_download': True, }, + 'skip': 'This video is not available.', }, { # YouTube Red video with episode data @@ -992,7 +1001,7 @@ 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', @@ -1025,7 +1034,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'uploader_id': 'UCEJYpZGqgUob0zVVEaLhvVg', 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCEJYpZGqgUob0zVVEaLhvVg', 'license': 'Standard YouTube License', - 'view_count': int, }, 'params': { 'skip_download': True, @@ -1537,7 +1545,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 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) @@ -1596,6 +1604,12 @@ 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']: @@ -1604,8 +1618,13 @@ 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( @@ -1682,125 +1701,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor): if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info: raise ExtractorError('"rental" videos not supported. See https://github.com/rg3/youtube-dl/issues/359 for more information.', expected=True) - # Start extracting information - self.report_information_extraction(video_id) - - # uploader - if 'author' not in video_info: - raise ExtractorError('Unable to extract uploader name') - video_uploader = compat_urllib_parse_unquote_plus(video_info['author'][0]) - - # uploader_id - video_uploader_id = None - video_uploader_url = None - 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".*?>.*?|["\']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']+class="title"[^>]*>\s*License\s*\s*]*>\s*
  • (.+?)]+class="title"[^>]*>\s*Music\s*\s* - ]*>\s* -
  • (?P.+?) - 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 - - 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) - - chapters = self._extract_chapters(description_original, video_duration) + 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() @@ -1810,7 +1713,7 @@ 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) @@ -1827,6 +1730,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'width': int_or_none(width_height[0]), 'height': int_or_none(width_height[1]), } + q = qualities(['small', 'medium', 'hd720']) formats = [] for url_data_str in encoded_url_map.split(','): url_data = compat_parse_qs(url_data_str) @@ -1906,13 +1810,19 @@ 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_label', [None])[0] or 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': quality, + 'quality': q(quality), } for key, value in more_fields.items(): if value: @@ -1933,6 +1843,11 @@ 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] @@ -1953,13 +1868,140 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 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) + 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 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) + 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') + + # 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 = 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) + + chapters = self._extract_chapters(description_original, video_duration) + # Look for the DASH manifest if self._downloader.params.get('youtube_include_dash_manifest', True): dash_mpd_fatal = True @@ -1976,6 +2018,8 @@ 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 @@ -2023,9 +2067,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'uploader_url': video_uploader_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, @@ -2048,6 +2092,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'series': series, 'season_number': season_number, 'episode_number': episode_number, + 'track': track, + 'artist': artist, } @@ -2437,7 +2483,7 @@ 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' @@ -2530,10 +2576,11 @@ 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) @@ -2568,7 +2615,11 @@ 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 @@ -2602,8 +2653,7 @@ 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 @@ -2626,11 +2676,10 @@ 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, @@ -2682,10 +2731,7 @@ 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 = [] @@ -2696,12 +2742,15 @@ 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 @@ -2713,8 +2762,12 @@ 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):