X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/453698570f26bebd37b39df8537d993b57d77b8b..903490db3cf588364c9ff08098b1a9193438a7f3:/youtube_dl/extractor/dailymotion.py diff --git a/youtube_dl/extractor/dailymotion.py b/youtube_dl/extractor/dailymotion.py index cf5841a..0e7d587 100644 --- a/youtube_dl/extractor/dailymotion.py +++ b/youtube_dl/extractor/dailymotion.py @@ -6,18 +6,17 @@ import json import itertools from .common import InfoExtractor -from .subtitles import SubtitlesInfoExtractor -from ..compat import ( - compat_str, - compat_urllib_request, -) from ..utils import ( + determine_ext, + error_to_compat_str, ExtractorError, int_or_none, - orderedSet, + parse_iso8601, + sanitized_Request, str_to_int, unescapeHTML, + mimetype2ext, ) @@ -25,16 +24,21 @@ class DailymotionBaseInfoExtractor(InfoExtractor): @staticmethod def _build_request(url): """Build a request with the family filter disabled""" - request = compat_urllib_request.Request(url) - request.add_header('Cookie', 'family_filter=off') - request.add_header('Cookie', 'ff=off') + request = sanitized_Request(url) + request.add_header('Cookie', 'family_filter=off; ff=off') return request + def _download_webpage_handle_no_ff(self, url, *args, **kwargs): + request = self._build_request(url) + return self._download_webpage_handle(request, *args, **kwargs) -class DailymotionIE(DailymotionBaseInfoExtractor, SubtitlesInfoExtractor): - """Information Extractor for Dailymotion""" + def _download_webpage_no_ff(self, url, *args, **kwargs): + request = self._build_request(url) + return self._download_webpage(request, *args, **kwargs) - _VALID_URL = r'(?i)(?:https?://)?(?:(www|touch)\.)?dailymotion\.[a-z]{2,3}/(?:(embed|#)/)?video/(?P[^/?_]+)' + +class DailymotionIE(DailymotionBaseInfoExtractor): + _VALID_URL = r'(?i)https?://(?:(www|touch)\.)?dailymotion\.[a-z]{2,3}/(?:(?:(?:embed|swf|#)/)?video|swf)/(?P[^/?_]+)' IE_NAME = 'dailymotion' _FORMATS = [ @@ -45,82 +49,212 @@ class DailymotionIE(DailymotionBaseInfoExtractor, SubtitlesInfoExtractor): ('stream_h264_hd1080_url', 'hd180'), ] - _TESTS = [ - { - 'url': 'http://www.dailymotion.com/video/x33vw9_tutoriel-de-youtubeur-dl-des-video_tech', - 'md5': '392c4b85a60a90dc4792da41ce3144eb', - 'info_dict': { - 'id': 'x33vw9', - 'ext': 'mp4', - 'uploader': 'Amphora Alex and Van .', - 'title': 'Tutoriel de Youtubeur"DL DES VIDEO DE YOUTUBE"', - } + _TESTS = [{ + 'url': 'http://www.dailymotion.com/video/x5kesuj_office-christmas-party-review-jason-bateman-olivia-munn-t-j-miller_news', + 'md5': '074b95bdee76b9e3654137aee9c79dfe', + 'info_dict': { + 'id': 'x5kesuj', + 'ext': 'mp4', + 'title': 'Office Christmas Party Review – Jason Bateman, Olivia Munn, T.J. Miller', + 'description': 'Office Christmas Party Review - Jason Bateman, Olivia Munn, T.J. Miller', + 'thumbnail': r're:^https?:.*\.(?:jpg|png)$', + 'duration': 187, + 'timestamp': 1493651285, + 'upload_date': '20170501', + 'uploader': 'Deadline', + 'uploader_id': 'x1xm8ri', + 'age_limit': 0, + 'view_count': int, + }, + }, { + 'url': 'https://www.dailymotion.com/video/x2iuewm_steam-machine-models-pricing-listed-on-steam-store-ign-news_videogames', + 'md5': '2137c41a8e78554bb09225b8eb322406', + 'info_dict': { + 'id': 'x2iuewm', + 'ext': 'mp4', + 'title': 'Steam Machine Models, Pricing Listed on Steam Store - IGN News', + 'description': 'Several come bundled with the Steam Controller.', + 'thumbnail': r're:^https?:.*\.(?:jpg|png)$', + 'duration': 74, + 'timestamp': 1425657362, + 'upload_date': '20150306', + 'uploader': 'IGN', + 'uploader_id': 'xijv66', + 'age_limit': 0, + 'view_count': int, }, + 'skip': 'video gone', + }, { # Vevo video - { - 'url': 'http://www.dailymotion.com/video/x149uew_katy-perry-roar-official_musi', - 'info_dict': { - 'title': 'Roar (Official)', - 'id': 'USUV71301934', - 'ext': 'mp4', - 'uploader': 'Katy Perry', - 'upload_date': '20130905', - }, - 'params': { - 'skip_download': True, - }, - 'skip': 'VEVO is only available in some countries', + 'url': 'http://www.dailymotion.com/video/x149uew_katy-perry-roar-official_musi', + 'info_dict': { + 'title': 'Roar (Official)', + 'id': 'USUV71301934', + 'ext': 'mp4', + 'uploader': 'Katy Perry', + 'upload_date': '20130905', + }, + 'params': { + 'skip_download': True, }, + 'skip': 'VEVO is only available in some countries', + }, { # age-restricted video - { - 'url': 'http://www.dailymotion.com/video/xyh2zz_leanna-decker-cyber-girl-of-the-year-desires-nude-playboy-plus_redband', - 'md5': '0d667a7b9cebecc3c89ee93099c4159d', - 'info_dict': { - 'id': 'xyh2zz', - 'ext': 'mp4', - 'title': 'Leanna Decker - Cyber Girl Of The Year Desires Nude [Playboy Plus]', - 'uploader': 'HotWaves1012', - 'age_limit': 18, - } - } - ] + 'url': 'http://www.dailymotion.com/video/xyh2zz_leanna-decker-cyber-girl-of-the-year-desires-nude-playboy-plus_redband', + 'md5': '0d667a7b9cebecc3c89ee93099c4159d', + 'info_dict': { + 'id': 'xyh2zz', + 'ext': 'mp4', + 'title': 'Leanna Decker - Cyber Girl Of The Year Desires Nude [Playboy Plus]', + 'uploader': 'HotWaves1012', + 'age_limit': 18, + }, + 'skip': 'video gone', + }, { + # geo-restricted, player v5 + 'url': 'http://www.dailymotion.com/video/xhza0o', + 'only_matching': True, + }, { + # with subtitles + 'url': 'http://www.dailymotion.com/video/x20su5f_the-power-of-nightmares-1-the-rise-of-the-politics-of-fear-bbc-2004_news', + 'only_matching': True, + }, { + 'url': 'http://www.dailymotion.com/swf/video/x3n92nf', + 'only_matching': True, + }, { + 'url': 'http://www.dailymotion.com/swf/x3ss1m_funny-magic-trick-barry-and-stuart_fun', + 'only_matching': True, + }] + + @staticmethod + def _extract_urls(webpage): + # Look for embedded Dailymotion player + matches = re.findall( + r'<(?:(?:embed|iframe)[^>]+?src=|input[^>]+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=)(["\'])(?P(?:https?:)?//(?:www\.)?dailymotion\.com/(?:embed|swf)/video/.+?)\1', webpage) + return list(map(lambda m: unescapeHTML(m[1]), matches)) def _real_extract(self, url): video_id = self._match_id(url) - url = 'http://www.dailymotion.com/video/%s' % video_id - # Retrieve video webpage to extract further information - request = self._build_request(url) - webpage = self._download_webpage(request, video_id) + webpage = self._download_webpage_no_ff( + 'https://www.dailymotion.com/video/%s' % video_id, video_id) + + age_limit = self._rta_search(webpage) - # Extract URL, uploader and title from webpage - self.report_extraction(video_id) + description = self._og_search_description(webpage) or self._html_search_meta( + 'description', webpage, 'description') + + view_count_str = self._search_regex( + (r']+itemprop="interactionCount"[^>]+content="UserPlays:([\s\d,.]+)"', + r'video_views_count[^>]+>\s+([\s\d\,.]+)'), + webpage, 'view count', default=None) + if view_count_str: + view_count_str = re.sub(r'\s', '', view_count_str) + view_count = str_to_int(view_count_str) + comment_count = int_or_none(self._search_regex( + r']+itemprop="interactionCount"[^>]+content="UserComments:(\d+)"', + webpage, 'comment count', default=None)) + + player_v5 = self._search_regex( + [r'buildPlayer\(({.+?})\);\n', # See https://github.com/rg3/youtube-dl/issues/7826 + r'playerV5\s*=\s*dmp\.create\([^,]+?,\s*({.+?})\);', + r'buildPlayer\(({.+?})\);', + r'var\s+config\s*=\s*({.+?});', + # New layout regex (see https://github.com/rg3/youtube-dl/issues/13580) + r'__PLAYER_CONFIG__\s*=\s*({.+?});'], + webpage, 'player v5', default=None) + if player_v5: + player = self._parse_json(player_v5, video_id) + metadata = player['metadata'] + + self._check_error(metadata) + + formats = [] + for quality, media_list in metadata['qualities'].items(): + for media in media_list: + media_url = media.get('url') + if not media_url: + continue + type_ = media.get('type') + if type_ == 'application/vnd.lumberjack.manifest': + continue + ext = mimetype2ext(type_) or determine_ext(media_url) + if ext == 'm3u8': + formats.extend(self._extract_m3u8_formats( + media_url, video_id, 'mp4', preference=-1, + m3u8_id='hls', fatal=False)) + elif ext == 'f4m': + formats.extend(self._extract_f4m_formats( + media_url, video_id, preference=-1, f4m_id='hds', fatal=False)) + else: + f = { + 'url': media_url, + 'format_id': 'http-%s' % quality, + 'ext': ext, + } + m = re.search(r'H264-(?P\d+)x(?P\d+)', media_url) + if m: + f.update({ + 'width': int(m.group('width')), + 'height': int(m.group('height')), + }) + formats.append(f) + self._sort_formats(formats) + + title = metadata['title'] + duration = int_or_none(metadata.get('duration')) + timestamp = int_or_none(metadata.get('created_time')) + thumbnail = metadata.get('poster_url') + uploader = metadata.get('owner', {}).get('screenname') + uploader_id = metadata.get('owner', {}).get('id') + + subtitles = {} + subtitles_data = metadata.get('subtitles', {}).get('data', {}) + if subtitles_data and isinstance(subtitles_data, dict): + for subtitle_lang, subtitle in subtitles_data.items(): + subtitles[subtitle_lang] = [{ + 'ext': determine_ext(subtitle_url), + 'url': subtitle_url, + } for subtitle_url in subtitle.get('urls', [])] + + return { + 'id': video_id, + 'title': title, + 'description': description, + 'thumbnail': thumbnail, + 'duration': duration, + 'timestamp': timestamp, + 'uploader': uploader, + 'uploader_id': uploader_id, + 'age_limit': age_limit, + 'view_count': view_count, + 'comment_count': comment_count, + 'formats': formats, + 'subtitles': subtitles, + } - # It may just embed a vevo video: - m_vevo = re.search( - r'[\w]*)', - webpage) - if m_vevo is not None: - vevo_id = m_vevo.group('id') - self.to_screen('Vevo video detected: %s' % vevo_id) - return self.url_result('vevo:%s' % vevo_id, ie='Vevo') + # vevo embed + vevo_id = self._search_regex( + r'[\w]*)', + webpage, 'vevo embed', default=None) + if vevo_id: + return self.url_result('vevo:%s' % vevo_id, 'Vevo') - age_limit = self._rta_search(webpage) + # fallback old player + embed_page = self._download_webpage_no_ff( + 'https://www.dailymotion.com/embed/video/%s' % video_id, + video_id, 'Downloading embed page') - video_upload_date = None - mobj = re.search(r'
([0-9]{2})-([0-9]{2})-([0-9]{4})
', webpage) - if mobj is not None: - video_upload_date = mobj.group(3) + mobj.group(2) + mobj.group(1) - - embed_url = 'http://www.dailymotion.com/embed/video/%s' % video_id - embed_page = self._download_webpage(embed_url, video_id, - 'Downloading embed page') - info = self._search_regex(r'var info = ({.*?}),$', embed_page, - 'video info', flags=re.MULTILINE) - info = json.loads(info) - if info.get('error') is not None: - msg = 'Couldn\'t get video, Dailymotion says: %s' % info['error']['title'] - raise ExtractorError(msg, expected=True) + timestamp = parse_iso8601(self._html_search_meta( + 'video:release_date', webpage, 'upload date')) + + info = self._parse_json( + self._search_regex( + r'var info = ({.*?}),$', embed_page, + 'video info', flags=re.MULTILINE), + video_id) + + self._check_error(info) formats = [] for (key, format_id) in self._FORMATS: @@ -138,18 +272,10 @@ class DailymotionIE(DailymotionBaseInfoExtractor, SubtitlesInfoExtractor): 'width': width, 'height': height, }) - if not formats: - raise ExtractorError('Unable to extract video URL') + self._sort_formats(formats) # subtitles video_subtitles = self.extract_subtitles(video_id, webpage) - if self._downloader.params.get('listsubtitles', False): - self._list_available_subtitles(video_id, webpage) - return - - view_count = str_to_int(self._search_regex( - r'video_views_count[^>]+>\s+([\d\.,]+)', - webpage, 'view count', fatal=False)) title = self._og_search_title(webpage, default=None) if title is None: @@ -161,25 +287,37 @@ class DailymotionIE(DailymotionBaseInfoExtractor, SubtitlesInfoExtractor): 'id': video_id, 'formats': formats, 'uploader': info['owner.screenname'], - 'upload_date': video_upload_date, + 'timestamp': timestamp, 'title': title, + 'description': description, 'subtitles': video_subtitles, 'thumbnail': info['thumbnail_url'], 'age_limit': age_limit, 'view_count': view_count, + 'duration': info['duration'] } - def _get_available_subtitles(self, video_id, webpage): + def _check_error(self, info): + error = info.get('error') + if info.get('error') is not None: + title = error['title'] + # See https://developer.dailymotion.com/api#access-error + if error.get('code') == 'DM007': + self.raise_geo_restricted(msg=title) + raise ExtractorError( + '%s said: %s' % (self.IE_NAME, title), expected=True) + + def _get_subtitles(self, video_id, webpage): try: sub_list = self._download_webpage( 'https://api.dailymotion.com/video/%s/subtitles?fields=id,language,url' % video_id, video_id, note=False) except ExtractorError as err: - self._downloader.report_warning('unable to download video subtitles: %s' % compat_str(err)) + self._downloader.report_warning('unable to download video subtitles: %s' % error_to_compat_str(err)) return {} info = json.loads(sub_list) if (info['total'] > 0): - sub_lang_list = dict((l['language'], l['url']) for l in info['list']) + sub_lang_list = dict((l['language'], [{'url': l['url'], 'ext': 'srt'}]) for l in info['list']) return sub_lang_list self._downloader.report_warning('video doesn\'t have subtitles') return {} @@ -187,30 +325,41 @@ class DailymotionIE(DailymotionBaseInfoExtractor, SubtitlesInfoExtractor): class DailymotionPlaylistIE(DailymotionBaseInfoExtractor): IE_NAME = 'dailymotion:playlist' - _VALID_URL = r'(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/playlist/(?P.+?)/' + _VALID_URL = r'(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/playlist/(?P[^/?#&]+)' _MORE_PAGES_INDICATOR = r'(?s)
.*?[^/]+)' + _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?!(?:embed|swf|#|video|playlist)/)(?:(?:old/)?user/)?(?P[^/]+)' _PAGE_TEMPLATE = 'http://www.dailymotion.com/user/%s/%s' _TESTS = [{ 'url': 'https://www.dailymotion.com/user/nqtv', @@ -236,12 +385,24 @@ class DailymotionUserIE(DailymotionPlaylistIE): 'title': 'Rémi Gaillard', }, 'playlist_mincount': 100, + }, { + 'url': 'http://www.dailymotion.com/user/UnderProject', + 'info_dict': { + 'id': 'UnderProject', + 'title': 'UnderProject', + }, + 'playlist_mincount': 1800, + 'expected_warnings': [ + 'Stopped at duplicated page', + ], + 'skip': 'Takes too long time', }] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) user = mobj.group('user') - webpage = self._download_webpage(url, user) + webpage = self._download_webpage( + 'https://www.dailymotion.com/user/%s' % user, user) full_user = unescapeHTML(self._html_search_regex( r'' % re.escape(user), webpage, 'user'))