diff options
Diffstat (limited to 'youtube_dl/extractor/youtube.py')
| -rw-r--r-- | youtube_dl/extractor/youtube.py | 230 |
1 files changed, 121 insertions, 109 deletions
diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index 143fac98a..1bf9cb7d4 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -74,14 +74,8 @@ class YoutubeBaseInfoExtractor(InfoExtractor): self._downloader.report_warning(u'unable to fetch login page: %s' % compat_str(err)) return False - galx = None - dsh = None - match = re.search(re.compile(r'<input.+?name="GALX".+?value="(.+?)"', re.DOTALL), login_page) - if match: - galx = match.group(1) - match = re.search(re.compile(r'<input.+?name="dsh".+?value="(.+?)"', re.DOTALL), login_page) - if match: - dsh = match.group(1) + galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"', + login_page, u'Login GALX parameter') # Log in login_form_strs = { @@ -95,7 +89,6 @@ class YoutubeBaseInfoExtractor(InfoExtractor): u'checkConnection': u'', u'checkedDomains': u'youtube', u'dnConn': u'', - u'dsh': dsh, u'pstMsg': u'0', u'rmShown': u'1', u'secTok': u'', @@ -146,10 +139,10 @@ class YoutubeBaseInfoExtractor(InfoExtractor): class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): IE_DESC = u'YouTube.com' - _VALID_URL = r"""^ + _VALID_URL = r"""(?x)^ ( - (?:https?://)? # http(s):// (optional) - (?:(?:(?:(?:\w+\.)?youtube(?:-nocookie)?\.com/| + (?:https?://|//)? # http(s):// or protocol-independent URL (optional) + (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/| tube\.majestyc\.net/| youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains (?:.*?\#/)? # handle anchor (#/) redirect urls @@ -236,11 +229,13 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): '136': 'mp4', '137': 'mp4', '138': 'mp4', - '139': 'mp4', - '140': 'mp4', - '141': 'mp4', '160': 'mp4', + # Dash mp4 audio + '139': 'm4a', + '140': 'm4a', + '141': 'm4a', + # Dash webm '171': 'webm', '172': 'webm', @@ -345,18 +340,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): } }, { - u"url": u"http://www.youtube.com/watch?v=1ltcDfZMA3U", - u"file": u"1ltcDfZMA3U.flv", - u"note": u"Test VEVO video (#897)", - u"info_dict": { - u"upload_date": u"20070518", - u"title": u"Maps - It Will Find You", - u"description": u"Music video by Maps performing It Will Find You.", - u"uploader": u"MuteUSA", - u"uploader_id": u"MuteUSA" - } - }, - { u"url": u"http://www.youtube.com/watch?v=UxxajLWwzqY", u"file": u"UxxajLWwzqY.mp4", u"note": u"Test generic use_cipher_signature video (#897)", @@ -380,6 +363,18 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): u"uploader_id": u"justintimberlakeVEVO" } }, + { + u"url": u"//www.YouTube.com/watch?v=yZIXLfi8CZQ", + u"file": u"yZIXLfi8CZQ.mp4", + u"note": u"Embed-only video (#1746)", + u"info_dict": { + u"upload_date": u"20120608", + u"title": u"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012", + u"description": u"md5:09b78bd971f1e3e289601dfba15ca4f7", + u"uploader": u"SET India", + u"uploader_id": u"setindia" + } + }, ] @@ -387,7 +382,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): def suitable(cls, url): """Receives a URL and returns True if suitable for this IE.""" if YoutubePlaylistIE.suitable(url): return False - return re.match(cls._VALID_URL, url, re.VERBOSE) is not None + return re.match(cls._VALID_URL, url) is not None def __init__(self, *args, **kwargs): super(YoutubeIE, self).__init__(*args, **kwargs) @@ -1036,6 +1031,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): """Turn the encrypted s field into a working signature""" if player_url is not None: + if player_url.startswith(u'//'): + player_url = u'https:' + player_url try: player_id = (player_url, len(s)) if player_id not in self._player_cache: @@ -1099,7 +1096,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): else: raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s))) - def _get_available_subtitles(self, video_id): + def _get_available_subtitles(self, video_id, webpage): try: sub_list = self._download_webpage( 'http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id, @@ -1115,8 +1112,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): params = compat_urllib_parse.urlencode({ 'lang': lang, 'v': video_id, - 'fmt': self._downloader.params.get('subtitlesformat'), - 'name': l[0], + 'fmt': self._downloader.params.get('subtitlesformat', 'srt'), + 'name': l[0].encode('utf-8'), }) url = u'http://www.youtube.com/api/timedtext?' + params sub_lang_list[lang] = url @@ -1128,7 +1125,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): def _get_available_automatic_caption(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') + sub_format = self._downloader.params.get('subtitlesformat', 'srt') self.to_screen(u'%s: Looking for automatic captions' % video_id) mobj = re.search(r';ytplayer.config = ({.*?});', webpage) err_msg = u'Couldn\'t find automatic captions for %s' % video_id @@ -1150,7 +1147,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): list_page = self._download_webpage(list_url, video_id) caption_list = xml.etree.ElementTree.fromstring(list_page.encode('utf-8')) original_lang_node = caption_list.find('track') - if not original_lang_node or original_lang_node.attrib.get('kind') != 'asr' : + if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' : self._downloader.report_warning(u'Video doesn\'t have automatic captions') return {} original_lang = original_lang_node.attrib['lang_code'] @@ -1287,7 +1284,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): # We simulate the access to the video from www.youtube.com/v/{video_id} # this can be viewed without login into Youtube data = compat_urllib_parse.urlencode({'video_id': video_id, - 'el': 'embedded', + 'el': 'player_embedded', 'gl': 'US', 'hl': 'en', 'eurl': 'https://youtube.googleapis.com/v/' + video_id, @@ -1316,6 +1313,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): else: raise ExtractorError(u'"token" parameter not in video info for unknown reason') + if 'view_count' in video_info: + view_count = int(video_info['view_count'][0]) + else: + view_count = None + # Check for "rental" videos if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info: raise ExtractorError(u'"rental" videos not supported') @@ -1403,32 +1405,29 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): # this signatures are encrypted if 'url_encoded_fmt_stream_map' not in args: raise ValueError(u'No stream_map present') # caught below - m_s = re.search(r'[&,]s=', args['url_encoded_fmt_stream_map']) + re_signature = re.compile(r'[&,]s=') + m_s = re_signature.search(args['url_encoded_fmt_stream_map']) if m_s is not None: self.to_screen(u'%s: Encrypted signatures detected.' % video_id) video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']] - m_s = re.search(r'[&,]s=', args.get('adaptive_fmts', u'')) + m_s = re_signature.search(args.get('adaptive_fmts', u'')) if m_s is not None: - if 'url_encoded_fmt_stream_map' in video_info: - video_info['url_encoded_fmt_stream_map'][0] += ',' + args['adaptive_fmts'] - else: - video_info['url_encoded_fmt_stream_map'] = [args['adaptive_fmts']] - elif 'adaptive_fmts' in video_info: - if 'url_encoded_fmt_stream_map' in video_info: - video_info['url_encoded_fmt_stream_map'][0] += ',' + video_info['adaptive_fmts'][0] + if 'adaptive_fmts' in video_info: + video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts'] else: - video_info['url_encoded_fmt_stream_map'] = video_info['adaptive_fmts'] + video_info['adaptive_fmts'] = [args['adaptive_fmts']] except ValueError: pass if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'): self.report_rtmp_download() video_url_list = [(None, video_info['conn'][0])] - elif 'url_encoded_fmt_stream_map' in video_info and len(video_info['url_encoded_fmt_stream_map']) >= 1: - if 'rtmpe%3Dyes' in video_info['url_encoded_fmt_stream_map'][0]: + elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 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) url_map = {} - for url_data_str in video_info['url_encoded_fmt_stream_map'][0].split(','): + for url_data_str in encoded_url_map.split(','): url_data = compat_parse_qs(url_data_str) if 'itag' in url_data and 'url' in url_data: url = url_data['url'][0] @@ -1481,13 +1480,13 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info') results = [] - for format_param, video_real_url in video_url_list: + for itag, video_real_url in video_url_list: # Extension - video_extension = self._video_extensions.get(format_param, 'flv') + video_extension = self._video_extensions.get(itag, 'flv') - video_format = '{0} - {1}{2}'.format(format_param if format_param else video_extension, - self._video_dimensions.get(format_param, '???'), - ' ('+self._special_itags[format_param]+')' if format_param in self._special_itags else '') + video_format = '{0} - {1}{2}'.format(itag if itag else video_extension, + self._video_dimensions.get(itag, '???'), + ' ('+self._special_itags[itag]+')' if itag in self._special_itags else '') results.append({ 'id': video_id, @@ -1498,17 +1497,20 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'title': video_title, 'ext': video_extension, 'format': video_format, + 'format_id': itag, 'thumbnail': video_thumbnail, 'description': video_description, 'player_url': player_url, 'subtitles': video_subtitles, 'duration': video_duration, 'age_limit': 18 if age_gate else 0, - 'annotations': video_annotations + 'annotations': video_annotations, + 'webpage_url': 'https://www.youtube.com/watch?v=%s' % video_id, + 'view_count': view_count, }) return results -class YoutubePlaylistIE(InfoExtractor): +class YoutubePlaylistIE(YoutubeBaseInfoExtractor): IE_DESC = u'YouTube.com playlists' _VALID_URL = r"""(?: (?:https?://)? @@ -1524,8 +1526,9 @@ class YoutubePlaylistIE(InfoExtractor): | ((?:PL|EC|UU|FL)[0-9A-Za-z-_]{10,}) )""" - _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/playlists/%s?max-results=%i&start-index=%i&v=2&alt=json&safeSearch=none' - _MAX_RESULTS = 50 + _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s&page=%s' + _MORE_PAGES_INDICATOR = r'data-link-type="next"' + _VIDEO_RE = r'href="/watch\?v=([0-9A-Za-z_-]{11})&' IE_NAME = u'youtube:playlist' @classmethod @@ -1533,6 +1536,9 @@ class YoutubePlaylistIE(InfoExtractor): """Receives a URL and returns True if suitable for this IE.""" return re.match(cls._VALID_URL, url, re.VERBOSE) is not None + def _real_initialize(self): + self._login() + def _real_extract(self, url): # Extract playlist id mobj = re.match(self._VALID_URL, url, re.VERBOSE) @@ -1546,51 +1552,33 @@ class YoutubePlaylistIE(InfoExtractor): video_id = query_dict['v'][0] if self._downloader.params.get('noplaylist'): self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id) - return self.url_result('https://www.youtube.com/watch?v=' + video_id, 'Youtube') + return self.url_result(video_id, 'Youtube', video_id=video_id) else: self.to_screen(u'Downloading playlist PL%s - add --no-playlist to just download video %s' % (playlist_id, video_id)) - # Download playlist videos from API - videos = [] + # Extract the video ids from the playlist pages + ids = [] for page_num in itertools.count(1): - start_index = self._MAX_RESULTS * (page_num - 1) + 1 - if start_index >= 1000: - self._downloader.report_warning(u'Max number of results reached') - break - url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, start_index) + url = self._TEMPLATE_URL % (playlist_id, page_num) page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num) + # The ids are duplicated + new_ids = orderedSet(re.findall(self._VIDEO_RE, page)) + ids.extend(new_ids) - try: - response = json.loads(page) - except ValueError as err: - raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err)) - - if 'feed' not in response: - raise ExtractorError(u'Got a malformed response from YouTube API') - playlist_title = response['feed']['title']['$t'] - if 'entry' not in response['feed']: - # Number of videos is a multiple of self._MAX_RESULTS + if re.search(self._MORE_PAGES_INDICATOR, page) is None: break - for entry in response['feed']['entry']: - index = entry['yt$position']['$t'] - if 'media$group' in entry and 'yt$videoid' in entry['media$group']: - videos.append(( - index, - 'https://www.youtube.com/watch?v=' + entry['media$group']['yt$videoid']['$t'] - )) - - videos = [v[1] for v in sorted(videos)] + playlist_title = self._og_search_title(page) - url_results = [self.url_result(vurl, 'Youtube') for vurl in videos] - return [self.playlist_result(url_results, playlist_id, playlist_title)] + url_results = [self.url_result(vid_id, 'Youtube', video_id=vid_id) + for vid_id in ids] + return self.playlist_result(url_results, playlist_id, playlist_title) class YoutubeChannelIE(InfoExtractor): IE_DESC = u'YouTube.com channels' _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)" - _TEMPLATE_URL = 'http://www.youtube.com/channel/%s/videos?sort=da&flow=list&view=0&page=%s&gl=US&hl=en' _MORE_PAGES_INDICATOR = 'yt-uix-load-more' _MORE_PAGES_URL = 'http://www.youtube.com/c4_browse_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s' IE_NAME = u'youtube:channel' @@ -1611,36 +1599,37 @@ class YoutubeChannelIE(InfoExtractor): # Download channel page channel_id = mobj.group(1) video_ids = [] - pagenum = 1 - - url = self._TEMPLATE_URL % (channel_id, pagenum) - page = self._download_webpage(url, channel_id, - u'Downloading page #%s' % pagenum) - - # Extract video identifiers - ids_in_page = self.extract_videos_from_page(page) - video_ids.extend(ids_in_page) + url = 'https://www.youtube.com/channel/%s/videos' % channel_id + channel_page = self._download_webpage(url, channel_id) + if re.search(r'channel-header-autogenerated-label', channel_page) is not None: + autogenerated = True + else: + autogenerated = False - # Download any subsequent channel pages using the json-based channel_ajax query - if self._MORE_PAGES_INDICATOR in page: + 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) + else: + # Download all channel pages using the json-based channel_ajax query for pagenum in itertools.count(1): url = self._MORE_PAGES_URL % (pagenum, channel_id) page = self._download_webpage(url, channel_id, u'Downloading page #%s' % pagenum) - + page = json.loads(page) - + ids_in_page = self.extract_videos_from_page(page['content_html']) video_ids.extend(ids_in_page) - - if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']: + + if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']: break self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids))) - urls = ['http://www.youtube.com/watch?v=%s' % id for id in video_ids] - url_entries = [self.url_result(eurl, 'Youtube') for eurl in urls] - return [self.playlist_result(url_entries, channel_id)] + url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id) + for video_id in video_ids] + return self.playlist_result(url_entries, channel_id) class YoutubeUserIE(InfoExtractor): @@ -1704,9 +1693,11 @@ class YoutubeUserIE(InfoExtractor): if len(ids_in_page) < self._GDATA_PAGE_SIZE: break - urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids] - url_results = [self.url_result(rurl, 'Youtube') for rurl in urls] - return [self.playlist_result(url_results, playlist_title = username)] + url_results = [ + self.url_result(video_id, 'Youtube', video_id=video_id) + for video_id in video_ids] + return self.playlist_result(url_results, playlist_title=username) + class YoutubeSearchIE(SearchInfoExtractor): IE_DESC = u'YouTube.com searches' @@ -1747,9 +1738,14 @@ class YoutubeSearchIE(SearchInfoExtractor): if len(video_ids) > n: video_ids = video_ids[:n] - videos = [self.url_result('http://www.youtube.com/watch?v=%s' % id, 'Youtube') for id in video_ids] + videos = [self.url_result(video_id, 'Youtube', video_id=video_id) + for video_id in video_ids] return self.playlist_result(videos, query) +class YoutubeSearchDateIE(YoutubeSearchIE): + _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 = u'YouTube.com searches, newest videos first' class YoutubeShowIE(InfoExtractor): IE_DESC = u'YouTube.com (multi-season) shows' @@ -1803,7 +1799,9 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): feed_html = info['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(id, 'Youtube') for id in ids) + feed_entries.extend( + self.url_result(video_id, 'Youtube', video_id=video_id) + for video_id in ids) if info['paging'] is None: break return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE) @@ -1828,6 +1826,20 @@ class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor): _PAGING_STEP = 100 _PERSONAL_FEED = True +class YoutubeHistoryIE(YoutubeFeedsInfoExtractor): + IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)' + _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory' + _FEED_NAME = 'history' + _PERSONAL_FEED = True + _PLAYLIST_TITLE = u'Youtube Watch History' + + def _real_extract(self, url): + webpage = self._download_webpage('https://www.youtube.com/feed/history', u'History') + data_paging = self._search_regex(r'data-paging="(\d+)"', webpage, u'data-paging') + # The step is actually a ridiculously big number (like 1374343569725646) + self._PAGING_STEP = int(data_paging) + return super(YoutubeHistoryIE, self)._real_extract(url) + class YoutubeFavouritesIE(YoutubeBaseInfoExtractor): IE_NAME = u'youtube:favorites' IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)' |