X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fmtv.py;h=fedd5f46bba67eb5cb4304d38aa0bdffa537e11d;hb=a6211d237b4e7051ca018cc09440502561fedaa7;hp=d887583e6600806dcb8d21bde241b1169a53e6b0;hpb=9f0ee2a3883ec6f6fdccba90085cb925aaa2f617;p=youtube-dl diff --git a/youtube_dl/extractor/mtv.py b/youtube_dl/extractor/mtv.py index d887583e6..fedd5f46b 100644 --- a/youtube_dl/extractor/mtv.py +++ b/youtube_dl/extractor/mtv.py @@ -1,21 +1,28 @@ +# coding: utf-8 from __future__ import unicode_literals import re from .common import InfoExtractor from ..compat import ( - compat_urllib_parse, compat_str, + compat_xpath, ) from ..utils import ( ExtractorError, find_xpath_attr, fix_xml_ampersands, + float_or_none, HEADRequest, + RegexNotFoundError, sanitized_Request, + strip_or_none, + timeconvert, + try_get, unescapeHTML, + update_url_query, url_basename, - RegexNotFoundError, + xpath_text, ) @@ -31,14 +38,10 @@ class MTVServicesInfoExtractor(InfoExtractor): def _id_from_uri(uri): return uri.split(':')[-1] - # This was originally implemented for ComedyCentral, but it also works here @staticmethod - def _transform_rtmp_url(rtmp_video_url): - m = re.match(r'^rtmpe?://.*?/(?Pgsp\..+?/.*)$', rtmp_video_url) - if not m: - return rtmp_video_url - base = 'http://viacommtvstrmfs.fplive.net/' - return base + m.group('finalid') + def _remove_template_parameter(url): + # Remove the templates, like &device={device} + return re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', url) def _get_feed_url(self, uri): return self._FEED_URL @@ -48,8 +51,7 @@ class MTVServicesInfoExtractor(InfoExtractor): thumb_node = itemdoc.find(search_path) if thumb_node is None: return None - else: - return thumb_node.attrib['url'] + return thumb_node.get('url') or thumb_node.text or None def _extract_mobile_video_formats(self, mtvn_id): webpage_url = self._MOBILE_TEMPLATE % mtvn_id @@ -66,7 +68,7 @@ class MTVServicesInfoExtractor(InfoExtractor): url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1) return [{'url': url, 'ext': 'mp4'}] - def _extract_video_formats(self, mdoc, mtvn_id): + def _extract_video_formats(self, mdoc, mtvn_id, video_id): if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4|copyright_error\.flv(?:\?geo\b.+?)?)$', mdoc.find('.//src').text) is not None: if mtvn_id is not None and self._MOBILE_TEMPLATE is not None: self.to_screen('The normal version is not available from your ' @@ -77,21 +79,35 @@ class MTVServicesInfoExtractor(InfoExtractor): formats = [] for rendition in mdoc.findall('.//rendition'): - try: - _, _, ext = rendition.attrib['type'].partition('/') - rtmp_video_url = rendition.find('./src').text - if rtmp_video_url.endswith('siteunavail.png'): - continue - formats.append({ - 'ext': ext, - 'url': self._transform_rtmp_url(rtmp_video_url), - 'format_id': rendition.get('bitrate'), - 'width': int(rendition.get('width')), - 'height': int(rendition.get('height')), - }) - except (KeyError, TypeError): - raise ExtractorError('Invalid rendition field.') - self._sort_formats(formats) + if rendition.get('method') == 'hls': + hls_url = rendition.find('./src').text + formats.extend(self._extract_m3u8_formats( + hls_url, video_id, ext='mp4', entry_protocol='m3u8_native', + m3u8_id='hls', fatal=False)) + else: + # fms + try: + _, _, ext = rendition.attrib['type'].partition('/') + rtmp_video_url = rendition.find('./src').text + if 'error_not_available.swf' in rtmp_video_url: + raise ExtractorError( + '%s said: video is not available' % self.IE_NAME, + expected=True) + if rtmp_video_url.endswith('siteunavail.png'): + continue + formats.extend([{ + 'ext': 'flv' if rtmp_video_url.startswith('rtmp') else ext, + 'url': rtmp_video_url, + 'format_id': '-'.join(filter(None, [ + 'rtmp' if rtmp_video_url.startswith('rtmp') else None, + rendition.get('bitrate')])), + 'width': int(rendition.get('width')), + 'height': int(rendition.get('height')), + }]) + except (KeyError, TypeError): + raise ExtractorError('Invalid rendition field.') + if formats: + self._sort_formats(formats) return formats def _extract_subtitles(self, mdoc, mtvn_id): @@ -100,25 +116,36 @@ class MTVServicesInfoExtractor(InfoExtractor): if transcript.get('kind') != 'captions': continue lang = transcript.get('srclang') - subtitles[lang] = [{ - 'url': compat_str(typographic.get('src')), - 'ext': typographic.get('format') - } for typographic in transcript.findall('./typographic')] + for typographic in transcript.findall('./typographic'): + sub_src = typographic.get('src') + if not sub_src: + continue + ext = typographic.get('format') + if ext == 'cea-608': + ext = 'scc' + subtitles.setdefault(lang, []).append({ + 'url': compat_str(sub_src), + 'ext': ext + }) return subtitles - def _get_video_info(self, itemdoc): + def _get_video_info(self, itemdoc, use_hls=True): uri = itemdoc.find('guid').text video_id = self._id_from_uri(uri) self.report_extraction(video_id) - mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url'] - # Remove the templates, like &device={device} - mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url) + content_el = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))) + mediagen_url = self._remove_template_parameter(content_el.attrib['url']) + mediagen_url = mediagen_url.replace('device={device}', '') if 'acceptMethods' not in mediagen_url: mediagen_url += '&' if '?' in mediagen_url else '?' - mediagen_url += 'acceptMethods=fms' + mediagen_url += 'acceptMethods=' + mediagen_url += 'hls' if use_hls else 'fms' - mediagen_doc = self._download_xml(mediagen_url, video_id, - 'Downloading video urls') + mediagen_doc = self._download_xml( + mediagen_url, video_id, 'Downloading video urls', fatal=False) + + if mediagen_doc is False: + return None item = mediagen_doc.find('./video/item') if item is not None and item.get('type') == 'text': @@ -128,11 +155,9 @@ class MTVServicesInfoExtractor(InfoExtractor): message += item.text raise ExtractorError(message, expected=True) - description_node = itemdoc.find('description') - if description_node is not None: - description = description_node.text.strip() - else: - description = None + description = strip_or_none(xpath_text(itemdoc, 'description')) + + timestamp = timeconvert(xpath_text(itemdoc, 'pubDate')) title_el = None if title_el is None: @@ -140,9 +165,9 @@ class MTVServicesInfoExtractor(InfoExtractor): itemdoc, './/{http://search.yahoo.com/mrss/}category', 'scheme', 'urn:mtvn:video_title') if title_el is None: - title_el = itemdoc.find('.//{http://search.yahoo.com/mrss/}title') + title_el = itemdoc.find(compat_xpath('.//{http://search.yahoo.com/mrss/}title')) if title_el is None: - title_el = itemdoc.find('.//title') or itemdoc.find('./title') + title_el = itemdoc.find(compat_xpath('.//title')) if title_el.text is None: title_el = None @@ -158,35 +183,77 @@ class MTVServicesInfoExtractor(InfoExtractor): if mtvn_id_node is not None: mtvn_id = mtvn_id_node.text + formats = self._extract_video_formats(mediagen_doc, mtvn_id, video_id) + + # Some parts of complete video may be missing (e.g. missing Act 3 in + # http://www.southpark.de/alle-episoden/s14e01-sexual-healing) + if not formats: + return None + + self._sort_formats(formats) + return { 'title': title, - 'formats': self._extract_video_formats(mediagen_doc, mtvn_id), + 'formats': formats, 'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id), 'id': video_id, 'thumbnail': self._get_thumbnail_url(uri, itemdoc), 'description': description, + 'duration': float_or_none(content_el.attrib.get('duration')), + 'timestamp': timestamp, } - def _get_videos_info(self, uri): + def _get_feed_query(self, uri): + data = {'uri': uri} + if self._LANG: + data['lang'] = self._LANG + return data + + def _get_videos_info(self, uri, use_hls=True): video_id = self._id_from_uri(uri) feed_url = self._get_feed_url(uri) - data = compat_urllib_parse.urlencode({'uri': uri}) - info_url = feed_url + '?' - if self._LANG: - info_url += 'lang=%s&' % self._LANG - info_url += data - return self._get_videos_info_from_url(info_url, video_id) + info_url = update_url_query(feed_url, self._get_feed_query(uri)) + return self._get_videos_info_from_url(info_url, video_id, use_hls) - def _get_videos_info_from_url(self, url, video_id): + def _get_videos_info_from_url(self, url, video_id, use_hls=True): idoc = self._download_xml( url, video_id, 'Downloading info', transform_source=fix_xml_ampersands) + + title = xpath_text(idoc, './channel/title') + description = xpath_text(idoc, './channel/description') + + entries = [] + for item in idoc.findall('.//item'): + info = self._get_video_info(item, use_hls) + if info: + entries.append(info) + return self.playlist_result( - [self._get_video_info(item) for item in idoc.findall('.//item')]) + entries, playlist_title=title, playlist_description=description) - def _real_extract(self, url): - title = url_basename(url) - webpage = self._download_webpage(url, title) + def _extract_triforce_mgid(self, webpage, data_zone=None, video_id=None): + triforce_feed = self._parse_json(self._search_regex( + r'triforceManifestFeed\s*=\s*({.+?})\s*;\s*\n', webpage, + 'triforce feed', default='{}'), video_id, fatal=False) + + data_zone = self._search_regex( + r'data-zone=(["\'])(?P.+?_lc_promo.*?)\1', webpage, + 'data zone', default=data_zone, group='zone') + + feed_url = try_get( + triforce_feed, lambda x: x['manifest']['zones'][data_zone]['feed'], + compat_str) + if not feed_url: + return + + feed = self._download_json(feed_url, video_id, fatal=False) + if not feed: + return + + return try_get(feed, lambda x: x['result']['data']['id'], compat_str) + + def _extract_mgid(self, webpage): try: # the url can be http://media.mtvnservices.com/fb/{mgid}.swf # or http://media.mtvnservices.com/{mgid} @@ -199,15 +266,24 @@ class MTVServicesInfoExtractor(InfoExtractor): if mgid is None or ':' not in mgid: mgid = self._search_regex( - [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'], + [r'data-mgid="(.*?)"', r'swfobject\.embedSWF\(".*?(mgid:.*?)"'], webpage, 'mgid', default=None) if not mgid: sm4_embed = self._html_search_meta( 'sm4:video:embed', webpage, 'sm4 embed', default='') mgid = self._search_regex( - r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid') + r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid', default=None) + + if not mgid: + mgid = self._extract_triforce_mgid(webpage) + + return mgid + def _real_extract(self, url): + title = url_basename(url) + webpage = self._download_webpage(url, title) + mgid = self._extract_mgid(webpage) videos_info = self._get_videos_info(mgid) return videos_info @@ -225,6 +301,8 @@ class MTVServicesEmbeddedIE(MTVServicesInfoExtractor): 'ext': 'mp4', 'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds', 'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.', + 'timestamp': 1400126400, + 'upload_date': '20140515', }, } @@ -237,13 +315,9 @@ class MTVServicesEmbeddedIE(MTVServicesInfoExtractor): def _get_feed_url(self, uri): video_id = self._id_from_uri(uri) - site_id = uri.replace(video_id, '') - config_url = ('http://media.mtvnservices.com/pmt/e1/players/{0}/' - 'context4/context5/config.xml'.format(site_id)) - config_doc = self._download_xml(config_url, video_id) - feed_node = config_doc.find('.//feed') - feed_url = feed_node.text.strip().split('?')[0] - return feed_url + config = self._download_json( + 'http://media.mtvnservices.com/pmt/e1/access/index.html?uri=%s&configtype=edge' % uri, video_id) + return self._remove_template_parameter(config['feedWithQueryParams']) def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) @@ -252,6 +326,57 @@ class MTVServicesEmbeddedIE(MTVServicesInfoExtractor): class MTVIE(MTVServicesInfoExtractor): + IE_NAME = 'mtv' + _VALID_URL = r'https?://(?:www\.)?mtv\.com/(?:video-clips|(?:full-)?episodes)/(?P[^/?#.]+)' + _FEED_URL = 'http://www.mtv.com/feeds/mrss/' + + _TESTS = [{ + 'url': 'http://www.mtv.com/video-clips/vl8qof/unlocking-the-truth-trailer', + 'md5': '1edbcdf1e7628e414a8c5dcebca3d32b', + 'info_dict': { + 'id': '5e14040d-18a4-47c4-a582-43ff602de88e', + 'ext': 'mp4', + 'title': 'Unlocking The Truth|July 18, 2016|1|101|Trailer', + 'description': '"Unlocking the Truth" premieres August 17th at 11/10c.', + 'timestamp': 1468846800, + 'upload_date': '20160718', + }, + }, { + 'url': 'http://www.mtv.com/full-episodes/94tujl/unlocking-the-truth-gates-of-hell-season-1-ep-101', + 'only_matching': True, + }, { + 'url': 'http://www.mtv.com/episodes/g8xu7q/teen-mom-2-breaking-the-wall-season-7-ep-713', + 'only_matching': True, + }] + + +class MTVJapanIE(MTVServicesInfoExtractor): + IE_NAME = 'mtvjapan' + _VALID_URL = r'https?://(?:www\.)?mtvjapan\.com/videos/(?P[0-9a-z]+)' + + _TEST = { + 'url': 'http://www.mtvjapan.com/videos/prayht/fresh-info-cadillac-escalade', + 'info_dict': { + 'id': 'bc01da03-6fe5-4284-8880-f291f4e368f5', + 'ext': 'mp4', + 'title': '【Fresh Info】Cadillac ESCALADE Sport Edition', + }, + 'params': { + 'skip_download': True, + }, + } + _GEO_COUNTRIES = ['JP'] + _FEED_URL = 'http://feeds.mtvnservices.com/od/feed/intl-mrss-player-feed' + + def _get_feed_query(self, uri): + return { + 'arcEp': 'mtvjapan.com', + 'mgid': uri, + } + + +class MTVVideoIE(MTVServicesInfoExtractor): + IE_NAME = 'mtv:video' _VALID_URL = r'''(?x)^https?:// (?:(?:www\.)?mtv\.com/videos/.+?/(?P[0-9]+)/[^/]+$| m\.mtv\.com/videos/video\.rbml\?.*?id=(?P[^&]+))''' @@ -267,6 +392,8 @@ class MTVIE(MTVServicesInfoExtractor): 'ext': 'mp4', 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"', 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.', + 'timestamp': 1352610000, + 'upload_date': '20121111', }, }, ] @@ -293,50 +420,38 @@ class MTVIE(MTVServicesInfoExtractor): return self._get_videos_info(uri) -class MTVIggyIE(MTVServicesInfoExtractor): - IE_NAME = 'mtviggy.com' - _VALID_URL = r'https?://www\.mtviggy\.com/videos/.+' - _TEST = { - 'url': 'http://www.mtviggy.com/videos/arcade-fire-behind-the-scenes-at-the-biggest-music-experiment-yet/', - 'info_dict': { - 'id': '984696', - 'ext': 'mp4', - 'title': 'Arcade Fire: Behind the Scenes at the Biggest Music Experiment Yet', - } - } - _FEED_URL = 'http://all.mtvworldverticals.com/feed-xml/' - - class MTVDEIE(MTVServicesInfoExtractor): IE_NAME = 'mtv.de' - _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:artists|shows|news)/(?:[^/]+/)*(?P\d+)-[^/#?]+/*(?:[#?].*)?$' + _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:musik/videoclips|folgen|news)/(?P[0-9a-z]+)' _TESTS = [{ - 'url': 'http://www.mtv.de/artists/10571-cro/videos/61131-traum', + 'url': 'http://www.mtv.de/musik/videoclips/2gpnv7/Traum', 'info_dict': { - 'id': 'music_video-a50bc5f0b3aa4b3190aa', + 'id': 'd5d472bc-f5b7-11e5-bffd-a4badb20dab5', 'ext': 'mp4', - 'title': 'MusicVideo_cro-traum', - 'description': 'Cro - Traum', + 'title': 'Traum', + 'description': 'Traum', }, 'params': { # rtmp download 'skip_download': True, }, + 'skip': 'Blocked at Travis CI', }, { # mediagen URL without query (e.g. http://videos.mtvnn.com/mediagen/e865da714c166d18d6f80893195fcb97) - 'url': 'http://www.mtv.de/shows/933-teen-mom-2/staffeln/5353/folgen/63565-enthullungen', + 'url': 'http://www.mtv.de/folgen/6b1ylu/teen-mom-2-enthuellungen-S5-F1', 'info_dict': { - 'id': 'local_playlist-f5ae778b9832cc837189', + 'id': '1e5a878b-31c5-11e7-a442-0e40cf2fc285', 'ext': 'mp4', - 'title': 'Episode_teen-mom-2_shows_season-5_episode-1_full-episode_part1', + 'title': 'Teen Mom 2', + 'description': 'md5:dc65e357ef7e1085ed53e9e9d83146a7', }, 'params': { # rtmp download 'skip_download': True, }, + 'skip': 'Blocked at Travis CI', }, { - # single video in pagePlaylist with different id - 'url': 'http://www.mtv.de/news/77491-mtv-movies-spotlight-pixels-teil-3', + 'url': 'http://www.mtv.de/news/glolix/77491-mtv-movies-spotlight--pixels--teil-3', 'info_dict': { 'id': 'local_playlist-4e760566473c4c8c5344', 'ext': 'mp4', @@ -347,23 +462,13 @@ class MTVDEIE(MTVServicesInfoExtractor): # rtmp download 'skip_download': True, }, + 'skip': 'Das Video kann zur Zeit nicht abgespielt werden.', }] + _GEO_COUNTRIES = ['DE'] + _FEED_URL = 'http://feeds.mtvnservices.com/od/feed/intl-mrss-player-feed' - def _real_extract(self, url): - video_id = self._match_id(url) - - webpage = self._download_webpage(url, video_id) - - playlist = self._parse_json( - self._search_regex( - r'window\.pagePlaylist\s*=\s*(\[.+?\]);\n', webpage, 'page playlist'), - video_id) - - # news pages contain single video in playlist with different id - if len(playlist) == 1: - return self._get_videos_info_from_url(playlist[0]['mrss'], video_id) - - for item in playlist: - item_id = item.get('id') - if item_id and compat_str(item_id) == video_id: - return self._get_videos_info_from_url(item['mrss'], video_id) + def _get_feed_query(self, uri): + return { + 'arcEp': 'mtv.de', + 'mgid': uri, + }