X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/bddc9fc577d16b1428924bf8a5c37ef1d9295f14..03342304420e5daeb428ffdcc7bbd2bbfecfa61a:/youtube_dl/extractor/mtv.py?ds=inline diff --git a/youtube_dl/extractor/mtv.py b/youtube_dl/extractor/mtv.py index a801c81..ed06836 100644 --- a/youtube_dl/extractor/mtv.py +++ b/youtube_dl/extractor/mtv.py @@ -1,72 +1,377 @@ +from __future__ import unicode_literals + import re -import socket -import xml.etree.ElementTree from .common import InfoExtractor -from ..utils import ( - compat_http_client, +from ..compat import ( + compat_urllib_parse, compat_str, - compat_urllib_error, - compat_urllib_request, - +) +from ..utils import ( ExtractorError, + find_xpath_attr, + fix_xml_ampersands, + float_or_none, + HEADRequest, + sanitized_Request, + unescapeHTML, + url_basename, + RegexNotFoundError, ) -class MTVIE(InfoExtractor): - _VALID_URL = r'^(?Phttps?://)?(?:www\.)?mtv\.com/videos/[^/]+/(?P[0-9]+)/[^/]+$' - _WORKING = False +def _media_xml_tag(tag): + return '{http://search.yahoo.com/mrss/}%s' % tag - def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - if mobj is None: - raise ExtractorError(u'Invalid URL: %s' % url) - if not mobj.group('proto'): - url = 'http://' + url - video_id = mobj.group('videoid') - webpage = self._download_webpage(url, video_id) +class MTVServicesInfoExtractor(InfoExtractor): + _MOBILE_TEMPLATE = None + _LANG = None + + @staticmethod + def _id_from_uri(uri): + return uri.split(':')[-1] + + # This was originally implemented for ComedyCentral, but it also works here + @staticmethod + def _transform_rtmp_url(rtmp_video_url): + m = re.match(r'^rtmpe?://.*?/(?Pgsp\..+?/.*)$', rtmp_video_url) + if not m: + return rtmp_video_url + base = 'http://viacommtvstrmfs.fplive.net/' + return base + m.group('finalid') + + def _get_feed_url(self, uri): + return self._FEED_URL - #song_name = self._html_search_regex(r'', - # webpage, u'song name', fatal=False) + def _get_thumbnail_url(self, uri, itemdoc): + search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail')) + thumb_node = itemdoc.find(search_path) + if thumb_node is None: + return None + else: + return thumb_node.attrib['url'] - video_title = self._html_search_regex(r'', - webpage, u'title') + def _extract_mobile_video_formats(self, mtvn_id): + webpage_url = self._MOBILE_TEMPLATE % mtvn_id + req = sanitized_Request(webpage_url) + # Otherwise we get a webpage that would execute some javascript + req.add_header('User-Agent', 'curl/7') + webpage = self._download_webpage(req, mtvn_id, + 'Downloading mobile page') + metrics_url = unescapeHTML(self._search_regex(r'', - webpage, u'mtvn_uri', fatal=False) + def _extract_video_formats(self, mdoc, mtvn_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 ' + 'country, trying with the mobile version') + return self._extract_mobile_video_formats(mtvn_id) + raise ExtractorError('This video is not available from your country.', + expected=True) - content_id = self._search_regex(r'MTVN.Player.defaultPlaylistId = ([0-9]+);', - webpage, u'content id', fatal=False) + 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) + return formats - videogen_url = 'http://www.mtv.com/player/includes/mediaGen.jhtml?uri=' + mtvn_uri + '&id=' + content_id + '&vid=' + video_id + '&ref=www.mtvn.com&viewUri=' + mtvn_uri + def _extract_subtitles(self, mdoc, mtvn_id): + subtitles = {} + for transcript in mdoc.findall('.//transcript'): + 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')] + return subtitles + + def _get_video_info(self, itemdoc): + uri = itemdoc.find('guid').text + video_id = self._id_from_uri(uri) self.report_extraction(video_id) - request = compat_urllib_request.Request(videogen_url) - try: - metadataXml = compat_urllib_request.urlopen(request).read() - except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: - raise ExtractorError(u'Unable to download video metadata: %s' % compat_str(err)) + content_el = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))) + mediagen_url = content_el.attrib['url'] + # Remove the templates, like &device={device} + mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url) + if 'acceptMethods' not in mediagen_url: + mediagen_url += '&' if '?' in mediagen_url else '?' + mediagen_url += 'acceptMethods=fms' - mdoc = xml.etree.ElementTree.fromstring(metadataXml) - renditions = mdoc.findall('.//rendition') + mediagen_doc = self._download_xml(mediagen_url, video_id, + 'Downloading video urls') - # For now, always pick the highest quality. - rendition = renditions[-1] + item = mediagen_doc.find('./video/item') + if item is not None and item.get('type') == 'text': + message = '%s returned error: ' % self.IE_NAME + if item.get('code') is not None: + message += '%s - ' % item.get('code') + message += item.text + raise ExtractorError(message, expected=True) - try: - _,_,ext = rendition.attrib['type'].partition('/') - format = ext + '-' + rendition.attrib['width'] + 'x' + rendition.attrib['height'] + '_' + rendition.attrib['bitrate'] - video_url = rendition.find('./src').text - except KeyError: - raise ExtractorError('Invalid rendition field.') + description_node = itemdoc.find('description') + if description_node is not None: + description = description_node.text.strip() + else: + description = None - info = { + title_el = None + if title_el is None: + title_el = find_xpath_attr( + 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') + if title_el is None: + title_el = itemdoc.find('.//title') or itemdoc.find('./title') + if title_el.text is None: + title_el = None + + title = title_el.text + if title is None: + raise ExtractorError('Could not find video title') + title = title.strip() + + # This a short id that's used in the webpage urls + mtvn_id = None + mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category', + 'scheme', 'urn:mtvn:id') + if mtvn_id_node is not None: + mtvn_id = mtvn_id_node.text + + return { + 'title': title, + 'formats': self._extract_video_formats(mediagen_doc, mtvn_id), + 'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id), 'id': video_id, - 'url': video_url, - 'upload_date': None, - 'title': video_title, - 'ext': ext, - 'format': format, + 'thumbnail': self._get_thumbnail_url(uri, itemdoc), + 'description': description, + 'duration': float_or_none(content_el.attrib.get('duration')), } - return [info] + def _get_feed_query(self, uri): + data = {'uri': uri} + if self._LANG: + data['lang'] = self._LANG + return compat_urllib_parse.urlencode(data) + + def _get_videos_info(self, uri): + video_id = self._id_from_uri(uri) + feed_url = self._get_feed_url(uri) + info_url = feed_url + '?' + self._get_feed_query(uri) + return self._get_videos_info_from_url(info_url, video_id) + + def _get_videos_info_from_url(self, url, video_id): + idoc = self._download_xml( + url, video_id, + 'Downloading info', transform_source=fix_xml_ampersands) + return self.playlist_result( + [self._get_video_info(item) for item in idoc.findall('.//item')]) + + def _extract_mgid(self, webpage): + try: + # the url can be http://media.mtvnservices.com/fb/{mgid}.swf + # or http://media.mtvnservices.com/{mgid} + og_url = self._og_search_video_url(webpage) + mgid = url_basename(og_url) + if mgid.endswith('.swf'): + mgid = mgid[:-4] + except RegexNotFoundError: + mgid = None + + if mgid is None or ':' not in mgid: + mgid = self._search_regex( + [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') + 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 + + +class MTVServicesEmbeddedIE(MTVServicesInfoExtractor): + IE_NAME = 'mtvservices:embedded' + _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P.+?)(\?|/|$)' + + _TEST = { + # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/ + 'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906', + 'md5': 'cb349b21a7897164cede95bd7bf3fbb9', + 'info_dict': { + 'id': '1043906', + 'ext': 'mp4', + 'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds', + 'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.', + }, + } + + @staticmethod + def _extract_url(webpage): + mobj = re.search( + r']+?src=(["\'])(?P(?:https?:)?//media.mtvnservices.com/embed/.+?)\1', webpage) + if mobj: + return mobj.group('url') + + def _get_feed_url(self, uri): + video_id = self._id_from_uri(uri) + site_id = uri.replace(video_id, '') + config_url = ('http://media.mtvnservices.com/pmt/e1/players/{0}/' + 'context4/context5/config.xml'.format(site_id)) + config_doc = self._download_xml(config_url, video_id) + feed_node = config_doc.find('.//feed') + feed_url = feed_node.text.strip().split('?')[0] + return feed_url + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + mgid = mobj.group('mgid') + return self._get_videos_info(mgid) + + +class MTVIE(MTVServicesInfoExtractor): + _VALID_URL = r'''(?x)^https?:// + (?:(?:www\.)?mtv\.com/videos/.+?/(?P[0-9]+)/[^/]+$| + m\.mtv\.com/videos/video\.rbml\?.*?id=(?P[^&]+))''' + + _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/' + + _TESTS = [ + { + 'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml', + 'md5': '850f3f143316b1e71fa56a4edfd6e0f8', + 'info_dict': { + 'id': '853555', + 'ext': 'mp4', + 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"', + 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.', + }, + }, + ] + + def _get_thumbnail_url(self, uri, itemdoc): + return 'http://mtv.mtvnimages.com/uri/' + uri + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + video_id = mobj.group('videoid') + uri = mobj.groupdict().get('mgid') + if uri is None: + webpage = self._download_webpage(url, video_id) + + # Some videos come from Vevo.com + m_vevo = re.search( + r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage) + if m_vevo: + vevo_id = m_vevo.group(1) + self.to_screen('Vevo video detected: %s' % vevo_id) + return self.url_result('vevo:%s' % vevo_id, ie='Vevo') + + uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri') + 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+)-[^/#?]+/*(?:[#?].*)?$' + _TESTS = [{ + 'url': 'http://www.mtv.de/artists/10571-cro/videos/61131-traum', + 'info_dict': { + 'id': 'music_video-a50bc5f0b3aa4b3190aa', + 'ext': 'mp4', + 'title': 'MusicVideo_cro-traum', + 'description': 'Cro - Traum', + }, + 'params': { + # rtmp download + 'skip_download': True, + }, + }, { + # 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', + 'info_dict': { + 'id': 'local_playlist-f5ae778b9832cc837189', + 'ext': 'mp4', + 'title': 'Episode_teen-mom-2_shows_season-5_episode-1_full-episode_part1', + }, + 'params': { + # rtmp download + 'skip_download': True, + }, + }, { + # single video in pagePlaylist with different id + 'url': 'http://www.mtv.de/news/77491-mtv-movies-spotlight-pixels-teil-3', + 'info_dict': { + 'id': 'local_playlist-4e760566473c4c8c5344', + 'ext': 'mp4', + 'title': 'Article_mtv-movies-spotlight-pixels-teil-3_short-clips_part1', + 'description': 'MTV Movies Supercut', + }, + 'params': { + # rtmp download + 'skip_download': True, + }, + }] + + 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)