X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/38bb9b1b0a044cabaf5691553815e334cd2e9213..e5a601f21d25b8806d04be9671c8772e331e3bda:/youtube_dl/extractor/dailymotion.py?ds=sidebyside diff --git a/youtube_dl/extractor/dailymotion.py b/youtube_dl/extractor/dailymotion.py index 1a41c0d..2e6226e 100644 --- a/youtube_dl/extractor/dailymotion.py +++ b/youtube_dl/extractor/dailymotion.py @@ -7,14 +7,13 @@ import itertools from .common import InfoExtractor -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, ) @@ -24,15 +23,21 @@ class DailymotionBaseInfoExtractor(InfoExtractor): @staticmethod def _build_request(url): """Build a request with the family filter disabled""" - request = compat_urllib_request.Request(url) + 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) + + def _download_webpage_no_ff(self, url, *args, **kwargs): + request = self._build_request(url) + return self._download_webpage(request, *args, **kwargs) -class DailymotionIE(DailymotionBaseInfoExtractor): - """Information Extractor for Dailymotion""" - _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/(?P[^/?_]+)' IE_NAME = 'dailymotion' _FORMATS = [ @@ -50,10 +55,17 @@ class DailymotionIE(DailymotionBaseInfoExtractor): 'info_dict': { 'id': 'x2iuewm', 'ext': 'mp4', - 'uploader': 'IGN', 'title': 'Steam Machine Models, Pricing Listed on Steam Store - IGN News', - 'upload_date': '20150306', + 'description': 'Several come bundled with the Steam Controller.', + 'thumbnail': 're:^https?:.*\.(?:jpg|png)$', 'duration': 74, + 'timestamp': 1425657362, + 'upload_date': '20150306', + 'uploader': 'IGN', + 'uploader_id': 'xijv66', + 'age_limit': 0, + 'view_count': int, + 'comment_count': int, } }, # Vevo video @@ -82,46 +94,141 @@ class DailymotionIE(DailymotionBaseInfoExtractor): 'uploader': 'HotWaves1012', 'age_limit': 18, } + }, + # 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, } ] def _real_extract(self, url): video_id = self._match_id(url) - url = 'https://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) - # Extract URL, uploader and title from webpage - self.report_extraction(video_id) + age_limit = self._rta_search(webpage) + + 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', fatal=False) + 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', fatal=False)) + + 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\(({.+?})\);'], + 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 = determine_ext(media_url) + if type_ == 'application/x-mpegURL' or ext == 'm3u8': + formats.extend(self._extract_m3u8_formats( + media_url, video_id, 'mp4', preference=-1, + m3u8_id='hls', fatal=False)) + elif type_ == 'application/f4m' or 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, + } + 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( + # vevo embed + vevo_id = self._search_regex( 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') + 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'', webpage) - if mobj is not None: - video_upload_date = mobj.group(1) + mobj.group(2) + mobj.group(3) - - embed_url = 'https://www.dailymotion.com/embed/video/%s' % video_id - embed_request = self._build_request(embed_url) - embed_page = self._download_webpage( - embed_request, 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: @@ -139,16 +246,11 @@ class DailymotionIE(DailymotionBaseInfoExtractor): '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) - 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: title = self._html_search_regex( @@ -159,8 +261,9 @@ class DailymotionIE(DailymotionBaseInfoExtractor): '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, @@ -168,13 +271,18 @@ class DailymotionIE(DailymotionBaseInfoExtractor): 'duration': info['duration'] } + def _check_error(self, info): + if info.get('error') is not None: + raise ExtractorError( + '%s said: %s' % (self.IE_NAME, info['error']['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): @@ -199,18 +307,26 @@ class DailymotionPlaylistIE(DailymotionBaseInfoExtractor): }] def _extract_entries(self, id): - video_ids = [] + video_ids = set() + processed_urls = set() for pagenum in itertools.count(1): - request = self._build_request(self._PAGE_TEMPLATE % (id, pagenum)) - webpage = self._download_webpage(request, - id, 'Downloading page %s' % pagenum) + page_url = self._PAGE_TEMPLATE % (id, pagenum) + webpage, urlh = self._download_webpage_handle_no_ff( + page_url, id, 'Downloading page %s' % pagenum) + if urlh.geturl() in processed_urls: + self.report_warning('Stopped at duplicated page %s, which is the same as %s' % ( + page_url, urlh.geturl()), id) + break + + processed_urls.add(urlh.geturl()) - video_ids.extend(re.findall(r'data-xid="(.+?)"', webpage)) + for video_id in re.findall(r'data-xid="(.+?)"', webpage): + if video_id not in video_ids: + yield self.url_result('http://www.dailymotion.com/video/%s' % video_id, 'Dailymotion') + video_ids.add(video_id) if re.search(self._MORE_PAGES_INDICATOR, webpage) is None: break - return [self.url_result('http://www.dailymotion.com/video/%s' % video_id, 'Dailymotion') - for video_id in orderedSet(video_ids)] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) @@ -227,7 +343,7 @@ class DailymotionPlaylistIE(DailymotionBaseInfoExtractor): class DailymotionUserIE(DailymotionPlaylistIE): IE_NAME = 'dailymotion:user' - _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?:(?:old/)?user/)?(?P[^/]+)$' + _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', @@ -236,6 +352,17 @@ 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): @@ -272,13 +399,13 @@ class DailymotionCloudIE(DailymotionBaseInfoExtractor): }] @classmethod - def _extract_dmcloud_url(self, webpage): - mobj = re.search(r']+src=[\'"](%s)[\'"]' % self._VALID_EMBED_URL, webpage) + def _extract_dmcloud_url(cls, webpage): + mobj = re.search(r']+src=[\'"](%s)[\'"]' % cls._VALID_EMBED_URL, webpage) if mobj: return mobj.group(1) mobj = re.search( - r']+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=[\'"](%s)[\'"]' % self._VALID_EMBED_URL, + r']+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=[\'"](%s)[\'"]' % cls._VALID_EMBED_URL, webpage) if mobj: return mobj.group(1) @@ -286,8 +413,7 @@ class DailymotionCloudIE(DailymotionBaseInfoExtractor): def _real_extract(self, url): video_id = self._match_id(url) - request = self._build_request(url) - webpage = self._download_webpage(request, video_id) + webpage = self._download_webpage_no_ff(url, video_id) title = self._html_search_regex(r'([^>]+)', webpage, 'title')