_ Git - youtube-dl/blob - youtube_dl/extractor/mtv.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..compat import (
   7     compat_str,
   8     compat_xpath,
   9 )
  10 from ..utils import (
  11     ExtractorError,
  12     find_xpath_attr,
  13     fix_xml_ampersands,
  14     float_or_none,
  15     HEADRequest,
  16     RegexNotFoundError,
  17     sanitized_Request,
  18     strip_or_none,
  19     timeconvert,
  20     try_get,
  21     unescapeHTML,
  22     update_url_query,
  23     url_basename,
  24     xpath_text,
  25 )
  26
  27
  28 def _media_xml_tag(tag):
  29     return '{http://search.yahoo.com/mrss/}%s' % tag
  30
  31
  32 class MTVServicesInfoExtractor(InfoExtractor):
  33     _MOBILE_TEMPLATE = None
  34     _LANG = None
  35
  36     @staticmethod
  37     def _id_from_uri(uri):
  38         return uri.split(':')[-1]
  39
  40     @staticmethod
  41     def _remove_template_parameter(url):
  42         # Remove the templates, like &device={device}
  43         return re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', url)
  44
  45     def _get_feed_url(self, uri):
  46         return self._FEED_URL
  47
  48     def _get_thumbnail_url(self, uri, itemdoc):
  49         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  50         thumb_node = itemdoc.find(search_path)
  51         if thumb_node is None:
  52             return None
  53         else:
  54             return thumb_node.attrib['url']
  55
  56     def _extract_mobile_video_formats(self, mtvn_id):
  57         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
  58         req = sanitized_Request(webpage_url)
  59         # Otherwise we get a webpage that would execute some javascript
  60         req.add_header('User-Agent', 'curl/7')
  61         webpage = self._download_webpage(req, mtvn_id,
  62                                          'Downloading mobile page')
  63         metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
  64         req = HEADRequest(metrics_url)
  65         response = self._request_webpage(req, mtvn_id, 'Resolving url')
  66         url = response.geturl()
  67         # Transform the url to get the best quality:
  68         url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
  69         return [{'url': url, 'ext': 'mp4'}]
  70
  71     def _extract_video_formats(self, mdoc, mtvn_id, video_id):
  72         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4|copyright_error\.flv(?:\?geo\b.+?)?)$', mdoc.find('.//src').text) is not None:
  73             if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
  74                 self.to_screen('The normal version is not available from your '
  75                                'country, trying with the mobile version')
  76                 return self._extract_mobile_video_formats(mtvn_id)
  77             raise ExtractorError('This video is not available from your country.',
  78                                  expected=True)
  79
  80         formats = []
  81         for rendition in mdoc.findall('.//rendition'):
  82             if rendition.get('method') == 'hls':
  83                 hls_url = rendition.find('./src').text
  84                 formats.extend(self._extract_m3u8_formats(
  85                     hls_url, video_id, ext='mp4', entry_protocol='m3u8_native',
  86                     m3u8_id='hls', fatal=False))
  87             else:
  88                 # fms
  89                 try:
  90                     _, _, ext = rendition.attrib['type'].partition('/')
  91                     rtmp_video_url = rendition.find('./src').text
  92                     if 'error_not_available.swf' in rtmp_video_url:
  93                         raise ExtractorError(
  94                             '%s said: video is not available' % self.IE_NAME,
  95                             expected=True)
  96                     if rtmp_video_url.endswith('siteunavail.png'):
  97                         continue
  98                     formats.extend([{
  99                         'ext': 'flv' if rtmp_video_url.startswith('rtmp') else ext,
 100                         'url': rtmp_video_url,
 101                         'format_id': '-'.join(filter(None, [
 102                             'rtmp' if rtmp_video_url.startswith('rtmp') else None,
 103                             rendition.get('bitrate')])),
 104                         'width': int(rendition.get('width')),
 105                         'height': int(rendition.get('height')),
 106                     }])
 107                 except (KeyError, TypeError):
 108                     raise ExtractorError('Invalid rendition field.')
 109         if formats:
 110             self._sort_formats(formats)
 111         return formats
 112
 113     def _extract_subtitles(self, mdoc, mtvn_id):
 114         subtitles = {}
 115         for transcript in mdoc.findall('.//transcript'):
 116             if transcript.get('kind') != 'captions':
 117                 continue
 118             lang = transcript.get('srclang')
 119             subtitles[lang] = [{
 120                 'url': compat_str(typographic.get('src')),
 121                 'ext': typographic.get('format')
 122             } for typographic in transcript.findall('./typographic')]
 123         return subtitles
 124
 125     def _get_video_info(self, itemdoc, use_hls=True):
 126         uri = itemdoc.find('guid').text
 127         video_id = self._id_from_uri(uri)
 128         self.report_extraction(video_id)
 129         content_el = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content')))
 130         mediagen_url = self._remove_template_parameter(content_el.attrib['url'])
 131         mediagen_url = mediagen_url.replace('device={device}', '')
 132         if 'acceptMethods' not in mediagen_url:
 133             mediagen_url += '&' if '?' in mediagen_url else '?'
 134             mediagen_url += 'acceptMethods='
 135             mediagen_url += 'hls' if use_hls else 'fms'
 136
 137         mediagen_doc = self._download_xml(
 138             mediagen_url, video_id, 'Downloading video urls', fatal=False)
 139
 140         if mediagen_doc is False:
 141             return None
 142
 143         item = mediagen_doc.find('./video/item')
 144         if item is not None and item.get('type') == 'text':
 145             message = '%s returned error: ' % self.IE_NAME
 146             if item.get('code') is not None:
 147                 message += '%s - ' % item.get('code')
 148             message += item.text
 149             raise ExtractorError(message, expected=True)
 150
 151         description = strip_or_none(xpath_text(itemdoc, 'description'))
 152
 153         timestamp = timeconvert(xpath_text(itemdoc, 'pubDate'))
 154
 155         title_el = None
 156         if title_el is None:
 157             title_el = find_xpath_attr(
 158                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 159                 'scheme', 'urn:mtvn:video_title')
 160         if title_el is None:
 161             title_el = itemdoc.find(compat_xpath('.//{http://search.yahoo.com/mrss/}title'))
 162         if title_el is None:
 163             title_el = itemdoc.find(compat_xpath('.//title'))
 164             if title_el.text is None:
 165                 title_el = None
 166
 167         title = title_el.text
 168         if title is None:
 169             raise ExtractorError('Could not find video title')
 170         title = title.strip()
 171
 172         # This a short id that's used in the webpage urls
 173         mtvn_id = None
 174         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 175                                        'scheme', 'urn:mtvn:id')
 176         if mtvn_id_node is not None:
 177             mtvn_id = mtvn_id_node.text
 178
 179         formats = self._extract_video_formats(mediagen_doc, mtvn_id, video_id)
 180
 181         # Some parts of complete video may be missing (e.g. missing Act 3 in
 182         # http://www.southpark.de/alle-episoden/s14e01-sexual-healing)
 183         if not formats:
 184             return None
 185
 186         self._sort_formats(formats)
 187
 188         return {
 189             'title': title,
 190             'formats': formats,
 191             'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id),
 192             'id': video_id,
 193             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 194             'description': description,
 195             'duration': float_or_none(content_el.attrib.get('duration')),
 196             'timestamp': timestamp,
 197         }
 198
 199     def _get_feed_query(self, uri):
 200         data = {'uri': uri}
 201         if self._LANG:
 202             data['lang'] = self._LANG
 203         return data
 204
 205     def _get_videos_info(self, uri, use_hls=True):
 206         video_id = self._id_from_uri(uri)
 207         feed_url = self._get_feed_url(uri)
 208         info_url = update_url_query(feed_url, self._get_feed_query(uri))
 209         return self._get_videos_info_from_url(info_url, video_id, use_hls)
 210
 211     def _get_videos_info_from_url(self, url, video_id, use_hls=True):
 212         idoc = self._download_xml(
 213             url, video_id,
 214             'Downloading info', transform_source=fix_xml_ampersands)
 215
 216         title = xpath_text(idoc, './channel/title')
 217         description = xpath_text(idoc, './channel/description')
 218
 219         entries = []
 220         for item in idoc.findall('.//item'):
 221             info = self._get_video_info(item, use_hls)
 222             if info:
 223                 entries.append(info)
 224
 225         return self.playlist_result(
 226             entries, playlist_title=title, playlist_description=description)
 227
 228     def _extract_triforce_mgid(self, webpage, data_zone=None, video_id=None):
 229         triforce_feed = self._parse_json(self._search_regex(
 230             r'triforceManifestFeed\s*=\s*({.+?})\s*;\s*\n', webpage,
 231             'triforce feed', default='{}'), video_id, fatal=False)
 232
 233         data_zone = self._search_regex(
 234             r'data-zone=(["\'])(?P<zone>.+?_lc_promo.*?)\1', webpage,
 235             'data zone', default=data_zone, group='zone')
 236
 237         feed_url = try_get(
 238             triforce_feed, lambda x: x['manifest']['zones'][data_zone]['feed'],
 239             compat_str)
 240         if not feed_url:
 241             return
 242
 243         feed = self._download_json(feed_url, video_id, fatal=False)
 244         if not feed:
 245             return
 246
 247         return try_get(feed, lambda x: x['result']['data']['id'], compat_str)
 248
 249     def _extract_mgid(self, webpage):
 250         try:
 251             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 252             # or http://media.mtvnservices.com/{mgid}
 253             og_url = self._og_search_video_url(webpage)
 254             mgid = url_basename(og_url)
 255             if mgid.endswith('.swf'):
 256                 mgid = mgid[:-4]
 257         except RegexNotFoundError:
 258             mgid = None
 259
 260         if mgid is None or ':' not in mgid:
 261             mgid = self._search_regex(
 262                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
 263                 webpage, 'mgid', default=None)
 264
 265         if not mgid:
 266             sm4_embed = self._html_search_meta(
 267                 'sm4:video:embed', webpage, 'sm4 embed', default='')
 268             mgid = self._search_regex(
 269                 r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid', default=None)
 270
 271         if not mgid:
 272             mgid = self._extract_triforce_mgid(webpage)
 273
 274         return mgid
 275
 276     def _real_extract(self, url):
 277         title = url_basename(url)
 278         webpage = self._download_webpage(url, title)
 279         mgid = self._extract_mgid(webpage)
 280         videos_info = self._get_videos_info(mgid)
 281         return videos_info
 282
 283
 284 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
 285     IE_NAME = 'mtvservices:embedded'
 286     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
 287
 288     _TEST = {
 289         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
 290         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
 291         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
 292         'info_dict': {
 293             'id': '1043906',
 294             'ext': 'mp4',
 295             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
 296             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
 297             'timestamp': 1400126400,
 298             'upload_date': '20140515',
 299         },
 300     }
 301
 302     @staticmethod
 303     def _extract_url(webpage):
 304         mobj = re.search(
 305             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//media.mtvnservices.com/embed/.+?)\1', webpage)
 306         if mobj:
 307             return mobj.group('url')
 308
 309     def _get_feed_url(self, uri):
 310         video_id = self._id_from_uri(uri)
 311         config = self._download_json(
 312             'http://media.mtvnservices.com/pmt/e1/access/index.html?uri=%s&configtype=edge' % uri, video_id)
 313         return self._remove_template_parameter(config['feedWithQueryParams'])
 314
 315     def _real_extract(self, url):
 316         mobj = re.match(self._VALID_URL, url)
 317         mgid = mobj.group('mgid')
 318         return self._get_videos_info(mgid)
 319
 320
 321 class MTVIE(MTVServicesInfoExtractor):
 322     IE_NAME = 'mtv'
 323     _VALID_URL = r'https?://(?:www\.)?mtv\.com/(?:video-clips|(?:full-)?episodes)/(?P<id>[^/?#.]+)'
 324     _FEED_URL = 'http://www.mtv.com/feeds/mrss/'
 325
 326     _TESTS = [{
 327         'url': 'http://www.mtv.com/video-clips/vl8qof/unlocking-the-truth-trailer',
 328         'md5': '1edbcdf1e7628e414a8c5dcebca3d32b',
 329         'info_dict': {
 330             'id': '5e14040d-18a4-47c4-a582-43ff602de88e',
 331             'ext': 'mp4',
 332             'title': 'Unlocking The Truth|July 18, 2016|1|101|Trailer',
 333             'description': '"Unlocking the Truth" premieres August 17th at 11/10c.',
 334             'timestamp': 1468846800,
 335             'upload_date': '20160718',
 336         },
 337     }, {
 338         'url': 'http://www.mtv.com/full-episodes/94tujl/unlocking-the-truth-gates-of-hell-season-1-ep-101',
 339         'only_matching': True,
 340     }, {
 341         'url': 'http://www.mtv.com/episodes/g8xu7q/teen-mom-2-breaking-the-wall-season-7-ep-713',
 342         'only_matching': True,
 343     }]
 344
 345
 346 class MTV81IE(InfoExtractor):
 347     IE_NAME = 'mtv81'
 348     _VALID_URL = r'https?://(?:www\.)?mtv81\.com/videos/(?P<id>[^/?#.]+)'
 349
 350     _TEST = {
 351         'url': 'http://www.mtv81.com/videos/artist-to-watch/the-godfather-of-japanese-hip-hop-segment-1/',
 352         'md5': '1edbcdf1e7628e414a8c5dcebca3d32b',
 353         'info_dict': {
 354             'id': '5e14040d-18a4-47c4-a582-43ff602de88e',
 355             'ext': 'mp4',
 356             'title': 'Unlocking The Truth|July 18, 2016|1|101|Trailer',
 357             'description': '"Unlocking the Truth" premieres August 17th at 11/10c.',
 358             'timestamp': 1468846800,
 359             'upload_date': '20160718',
 360         },
 361     }
 362
 363     def _extract_mgid(self, webpage):
 364         return self._search_regex(
 365             r'getTheVideo\((["\'])(?P<id>mgid:.+?)\1', webpage,
 366             'mgid', group='id')
 367
 368     def _real_extract(self, url):
 369         video_id = self._match_id(url)
 370         webpage = self._download_webpage(url, video_id)
 371         mgid = self._extract_mgid(webpage)
 372         return self.url_result('http://media.mtvnservices.com/embed/%s' % mgid)
 373
 374
 375 class MTVVideoIE(MTVServicesInfoExtractor):
 376     IE_NAME = 'mtv:video'
 377     _VALID_URL = r'''(?x)^https?://
 378         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 379            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 380
 381     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 382
 383     _TESTS = [
 384         {
 385             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 386             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 387             'info_dict': {
 388                 'id': '853555',
 389                 'ext': 'mp4',
 390                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 391                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 392                 'timestamp': 1352610000,
 393                 'upload_date': '20121111',
 394             },
 395         },
 396     ]
 397
 398     def _get_thumbnail_url(self, uri, itemdoc):
 399         return 'http://mtv.mtvnimages.com/uri/' + uri
 400
 401     def _real_extract(self, url):
 402         mobj = re.match(self._VALID_URL, url)
 403         video_id = mobj.group('videoid')
 404         uri = mobj.groupdict().get('mgid')
 405         if uri is None:
 406             webpage = self._download_webpage(url, video_id)
 407
 408             # Some videos come from Vevo.com
 409             m_vevo = re.search(
 410                 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
 411             if m_vevo:
 412                 vevo_id = m_vevo.group(1)
 413                 self.to_screen('Vevo video detected: %s' % vevo_id)
 414                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 415
 416             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 417         return self._get_videos_info(uri)
 418
 419
 420 class MTVDEIE(MTVServicesInfoExtractor):
 421     IE_NAME = 'mtv.de'
 422     _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:artists|shows|news)/(?:[^/]+/)*(?P<id>\d+)-[^/#?]+/*(?:[#?].*)?$'
 423     _TESTS = [{
 424         'url': 'http://www.mtv.de/artists/10571-cro/videos/61131-traum',
 425         'info_dict': {
 426             'id': 'music_video-a50bc5f0b3aa4b3190aa',
 427             'ext': 'flv',
 428             'title': 'MusicVideo_cro-traum',
 429             'description': 'Cro - Traum',
 430         },
 431         'params': {
 432             # rtmp download
 433             'skip_download': True,
 434         },
 435         'skip': 'Blocked at Travis CI',
 436     }, {
 437         # mediagen URL without query (e.g. http://videos.mtvnn.com/mediagen/e865da714c166d18d6f80893195fcb97)
 438         'url': 'http://www.mtv.de/shows/933-teen-mom-2/staffeln/5353/folgen/63565-enthullungen',
 439         'info_dict': {
 440             'id': 'local_playlist-f5ae778b9832cc837189',
 441             'ext': 'flv',
 442             'title': 'Episode_teen-mom-2_shows_season-5_episode-1_full-episode_part1',
 443         },
 444         'params': {
 445             # rtmp download
 446             'skip_download': True,
 447         },
 448         'skip': 'Blocked at Travis CI',
 449     }, {
 450         'url': 'http://www.mtv.de/news/77491-mtv-movies-spotlight-pixels-teil-3',
 451         'info_dict': {
 452             'id': 'local_playlist-4e760566473c4c8c5344',
 453             'ext': 'mp4',
 454             'title': 'Article_mtv-movies-spotlight-pixels-teil-3_short-clips_part1',
 455             'description': 'MTV Movies Supercut',
 456         },
 457         'params': {
 458             # rtmp download
 459             'skip_download': True,
 460         },
 461         'skip': 'Das Video kann zur Zeit nicht abgespielt werden.',
 462     }]
 463
 464     def _real_extract(self, url):
 465         video_id = self._match_id(url)
 466
 467         webpage = self._download_webpage(url, video_id)
 468
 469         playlist = self._parse_json(
 470             self._search_regex(
 471                 r'window\.pagePlaylist\s*=\s*(\[.+?\]);\n', webpage, 'page playlist'),
 472             video_id)
 473
 474         def _mrss_url(item):
 475             return item['mrss'] + item.get('mrssvars', '')
 476
 477         # news pages contain single video in playlist with different id
 478         if len(playlist) == 1:
 479             return self._get_videos_info_from_url(_mrss_url(playlist[0]), video_id)
 480
 481         for item in playlist:
 482             item_id = item.get('id')
 483             if item_id and compat_str(item_id) == video_id:
 484                 return self._get_videos_info_from_url(_mrss_url(item), video_id)