_ 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'))
  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         self._sort_formats(formats)
 110         return formats
 111
 112     def _extract_subtitles(self, mdoc, mtvn_id):
 113         subtitles = {}
 114         for transcript in mdoc.findall('.//transcript'):
 115             if transcript.get('kind') != 'captions':
 116                 continue
 117             lang = transcript.get('srclang')
 118             subtitles[lang] = [{
 119                 'url': compat_str(typographic.get('src')),
 120                 'ext': typographic.get('format')
 121             } for typographic in transcript.findall('./typographic')]
 122         return subtitles
 123
 124     def _get_video_info(self, itemdoc, use_hls=True):
 125         uri = itemdoc.find('guid').text
 126         video_id = self._id_from_uri(uri)
 127         self.report_extraction(video_id)
 128         content_el = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content')))
 129         mediagen_url = self._remove_template_parameter(content_el.attrib['url'])
 130         mediagen_url = mediagen_url.replace('device={device}', '')
 131         if 'acceptMethods' not in mediagen_url:
 132             mediagen_url += '&' if '?' in mediagen_url else '?'
 133             mediagen_url += 'acceptMethods='
 134             mediagen_url += 'hls' if use_hls else 'fms'
 135
 136         mediagen_doc = self._download_xml(mediagen_url, video_id,
 137                                           'Downloading video urls')
 138
 139         item = mediagen_doc.find('./video/item')
 140         if item is not None and item.get('type') == 'text':
 141             message = '%s returned error: ' % self.IE_NAME
 142             if item.get('code') is not None:
 143                 message += '%s - ' % item.get('code')
 144             message += item.text
 145             raise ExtractorError(message, expected=True)
 146
 147         description = strip_or_none(xpath_text(itemdoc, 'description'))
 148
 149         timestamp = timeconvert(xpath_text(itemdoc, 'pubDate'))
 150
 151         title_el = None
 152         if title_el is None:
 153             title_el = find_xpath_attr(
 154                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 155                 'scheme', 'urn:mtvn:video_title')
 156         if title_el is None:
 157             title_el = itemdoc.find(compat_xpath('.//{http://search.yahoo.com/mrss/}title'))
 158         if title_el is None:
 159             title_el = itemdoc.find(compat_xpath('.//title'))
 160             if title_el.text is None:
 161                 title_el = None
 162
 163         title = title_el.text
 164         if title is None:
 165             raise ExtractorError('Could not find video title')
 166         title = title.strip()
 167
 168         # This a short id that's used in the webpage urls
 169         mtvn_id = None
 170         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 171                                        'scheme', 'urn:mtvn:id')
 172         if mtvn_id_node is not None:
 173             mtvn_id = mtvn_id_node.text
 174
 175         formats = self._extract_video_formats(mediagen_doc, mtvn_id, video_id)
 176
 177         return {
 178             'title': title,
 179             'formats': formats,
 180             'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id),
 181             'id': video_id,
 182             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 183             'description': description,
 184             'duration': float_or_none(content_el.attrib.get('duration')),
 185             'timestamp': timestamp,
 186         }
 187
 188     def _get_feed_query(self, uri):
 189         data = {'uri': uri}
 190         if self._LANG:
 191             data['lang'] = self._LANG
 192         return data
 193
 194     def _get_videos_info(self, uri, use_hls=True):
 195         video_id = self._id_from_uri(uri)
 196         feed_url = self._get_feed_url(uri)
 197         info_url = update_url_query(feed_url, self._get_feed_query(uri))
 198         return self._get_videos_info_from_url(info_url, video_id, use_hls)
 199
 200     def _get_videos_info_from_url(self, url, video_id, use_hls=True):
 201         idoc = self._download_xml(
 202             url, video_id,
 203             'Downloading info', transform_source=fix_xml_ampersands)
 204
 205         title = xpath_text(idoc, './channel/title')
 206         description = xpath_text(idoc, './channel/description')
 207
 208         return self.playlist_result(
 209             [self._get_video_info(item, use_hls) for item in idoc.findall('.//item')],
 210             playlist_title=title, playlist_description=description)
 211
 212     def _extract_triforce_mgid(self, webpage, data_zone=None, video_id=None):
 213         triforce_feed = self._parse_json(self._search_regex(
 214             r'triforceManifestFeed\s*=\s*({.+?})\s*;\s*\n', webpage,
 215             'triforce feed', default='{}'), video_id, fatal=False)
 216
 217         data_zone = self._search_regex(
 218             r'data-zone=(["\'])(?P<zone>.+?_lc_promo.*?)\1', webpage,
 219             'data zone', default=data_zone, group='zone')
 220
 221         feed_url = try_get(
 222             triforce_feed, lambda x: x['manifest']['zones'][data_zone]['feed'],
 223             compat_str)
 224         if not feed_url:
 225             return
 226
 227         feed = self._download_json(feed_url, video_id, fatal=False)
 228         if not feed:
 229             return
 230
 231         return try_get(feed, lambda x: x['result']['data']['id'], compat_str)
 232
 233     def _extract_mgid(self, webpage):
 234         try:
 235             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 236             # or http://media.mtvnservices.com/{mgid}
 237             og_url = self._og_search_video_url(webpage)
 238             mgid = url_basename(og_url)
 239             if mgid.endswith('.swf'):
 240                 mgid = mgid[:-4]
 241         except RegexNotFoundError:
 242             mgid = None
 243
 244         if mgid is None or ':' not in mgid:
 245             mgid = self._search_regex(
 246                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
 247                 webpage, 'mgid', default=None)
 248
 249         if not mgid:
 250             sm4_embed = self._html_search_meta(
 251                 'sm4:video:embed', webpage, 'sm4 embed', default='')
 252             mgid = self._search_regex(
 253                 r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid', default=None)
 254
 255         if not mgid:
 256             mgid = self._extract_triforce_mgid(webpage)
 257
 258         return mgid
 259
 260     def _real_extract(self, url):
 261         title = url_basename(url)
 262         webpage = self._download_webpage(url, title)
 263         mgid = self._extract_mgid(webpage)
 264         videos_info = self._get_videos_info(mgid)
 265         return videos_info
 266
 267
 268 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
 269     IE_NAME = 'mtvservices:embedded'
 270     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
 271
 272     _TEST = {
 273         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
 274         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
 275         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
 276         'info_dict': {
 277             'id': '1043906',
 278             'ext': 'mp4',
 279             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
 280             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
 281             'timestamp': 1400126400,
 282             'upload_date': '20140515',
 283         },
 284     }
 285
 286     @staticmethod
 287     def _extract_url(webpage):
 288         mobj = re.search(
 289             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//media.mtvnservices.com/embed/.+?)\1', webpage)
 290         if mobj:
 291             return mobj.group('url')
 292
 293     def _get_feed_url(self, uri):
 294         video_id = self._id_from_uri(uri)
 295         config = self._download_json(
 296             'http://media.mtvnservices.com/pmt/e1/access/index.html?uri=%s&configtype=edge' % uri, video_id)
 297         return self._remove_template_parameter(config['feedWithQueryParams'])
 298
 299     def _real_extract(self, url):
 300         mobj = re.match(self._VALID_URL, url)
 301         mgid = mobj.group('mgid')
 302         return self._get_videos_info(mgid)
 303
 304
 305 class MTVIE(MTVServicesInfoExtractor):
 306     IE_NAME = 'mtv'
 307     _VALID_URL = r'https?://(?:www\.)?mtv\.com/(?:video-clips|full-episodes)/(?P<id>[^/?#.]+)'
 308     _FEED_URL = 'http://www.mtv.com/feeds/mrss/'
 309
 310     _TESTS = [{
 311         'url': 'http://www.mtv.com/video-clips/vl8qof/unlocking-the-truth-trailer',
 312         'md5': '1edbcdf1e7628e414a8c5dcebca3d32b',
 313         'info_dict': {
 314             'id': '5e14040d-18a4-47c4-a582-43ff602de88e',
 315             'ext': 'mp4',
 316             'title': 'Unlocking The Truth|July 18, 2016|1|101|Trailer',
 317             'description': '"Unlocking the Truth" premieres August 17th at 11/10c.',
 318             'timestamp': 1468846800,
 319             'upload_date': '20160718',
 320         },
 321     }, {
 322         'url': 'http://www.mtv.com/full-episodes/94tujl/unlocking-the-truth-gates-of-hell-season-1-ep-101',
 323         'only_matching': True,
 324     }]
 325
 326
 327 class MTVVideoIE(MTVServicesInfoExtractor):
 328     IE_NAME = 'mtv:video'
 329     _VALID_URL = r'''(?x)^https?://
 330         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 331            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 332
 333     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 334
 335     _TESTS = [
 336         {
 337             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 338             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 339             'info_dict': {
 340                 'id': '853555',
 341                 'ext': 'mp4',
 342                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 343                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 344                 'timestamp': 1352610000,
 345                 'upload_date': '20121111',
 346             },
 347         },
 348     ]
 349
 350     def _get_thumbnail_url(self, uri, itemdoc):
 351         return 'http://mtv.mtvnimages.com/uri/' + uri
 352
 353     def _real_extract(self, url):
 354         mobj = re.match(self._VALID_URL, url)
 355         video_id = mobj.group('videoid')
 356         uri = mobj.groupdict().get('mgid')
 357         if uri is None:
 358             webpage = self._download_webpage(url, video_id)
 359
 360             # Some videos come from Vevo.com
 361             m_vevo = re.search(
 362                 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
 363             if m_vevo:
 364                 vevo_id = m_vevo.group(1)
 365                 self.to_screen('Vevo video detected: %s' % vevo_id)
 366                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 367
 368             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 369         return self._get_videos_info(uri)
 370
 371
 372 class MTVDEIE(MTVServicesInfoExtractor):
 373     IE_NAME = 'mtv.de'
 374     _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:artists|shows|news)/(?:[^/]+/)*(?P<id>\d+)-[^/#?]+/*(?:[#?].*)?$'
 375     _TESTS = [{
 376         'url': 'http://www.mtv.de/artists/10571-cro/videos/61131-traum',
 377         'info_dict': {
 378             'id': 'music_video-a50bc5f0b3aa4b3190aa',
 379             'ext': 'flv',
 380             'title': 'MusicVideo_cro-traum',
 381             'description': 'Cro - Traum',
 382         },
 383         'params': {
 384             # rtmp download
 385             'skip_download': True,
 386         },
 387         'skip': 'Blocked at Travis CI',
 388     }, {
 389         # mediagen URL without query (e.g. http://videos.mtvnn.com/mediagen/e865da714c166d18d6f80893195fcb97)
 390         'url': 'http://www.mtv.de/shows/933-teen-mom-2/staffeln/5353/folgen/63565-enthullungen',
 391         'info_dict': {
 392             'id': 'local_playlist-f5ae778b9832cc837189',
 393             'ext': 'flv',
 394             'title': 'Episode_teen-mom-2_shows_season-5_episode-1_full-episode_part1',
 395         },
 396         'params': {
 397             # rtmp download
 398             'skip_download': True,
 399         },
 400         'skip': 'Blocked at Travis CI',
 401     }, {
 402         'url': 'http://www.mtv.de/news/77491-mtv-movies-spotlight-pixels-teil-3',
 403         'info_dict': {
 404             'id': 'local_playlist-4e760566473c4c8c5344',
 405             'ext': 'mp4',
 406             'title': 'Article_mtv-movies-spotlight-pixels-teil-3_short-clips_part1',
 407             'description': 'MTV Movies Supercut',
 408         },
 409         'params': {
 410             # rtmp download
 411             'skip_download': True,
 412         },
 413         'skip': 'Das Video kann zur Zeit nicht abgespielt werden.',
 414     }]
 415
 416     def _real_extract(self, url):
 417         video_id = self._match_id(url)
 418
 419         webpage = self._download_webpage(url, video_id)
 420
 421         playlist = self._parse_json(
 422             self._search_regex(
 423                 r'window\.pagePlaylist\s*=\s*(\[.+?\]);\n', webpage, 'page playlist'),
 424             video_id)
 425
 426         def _mrss_url(item):
 427             return item['mrss'] + item.get('mrssvars', '')
 428
 429         # news pages contain single video in playlist with different id
 430         if len(playlist) == 1:
 431             return self._get_videos_info_from_url(_mrss_url(playlist[0]), video_id)
 432
 433         for item in playlist:
 434             item_id = item.get('id')
 435             if item_id and compat_str(item_id) == video_id:
 436                 return self._get_videos_info_from_url(_mrss_url(item), video_id)