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_urllib_parse,
   8     compat_str,
   9 )
  10 from ..utils import (
  11     ExtractorError,
  12     find_xpath_attr,
  13     fix_xml_ampersands,
  14     HEADRequest,
  15     sanitized_Request,
  16     unescapeHTML,
  17     url_basename,
  18     RegexNotFoundError,
  19 )
  20
  21
  22 def _media_xml_tag(tag):
  23     return '{http://search.yahoo.com/mrss/}%s' % tag
  24
  25
  26 class MTVServicesInfoExtractor(InfoExtractor):
  27     _MOBILE_TEMPLATE = None
  28     _LANG = None
  29
  30     @staticmethod
  31     def _id_from_uri(uri):
  32         return uri.split(':')[-1]
  33
  34     # This was originally implemented for ComedyCentral, but it also works here
  35     @staticmethod
  36     def _transform_rtmp_url(rtmp_video_url):
  37         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  38         if not m:
  39             return rtmp_video_url
  40         base = 'http://viacommtvstrmfs.fplive.net/'
  41         return base + m.group('finalid')
  42
  43     def _get_feed_url(self, uri):
  44         return self._FEED_URL
  45
  46     def _get_thumbnail_url(self, uri, itemdoc):
  47         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  48         thumb_node = itemdoc.find(search_path)
  49         if thumb_node is None:
  50             return None
  51         else:
  52             return thumb_node.attrib['url']
  53
  54     def _extract_mobile_video_formats(self, mtvn_id):
  55         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
  56         req = sanitized_Request(webpage_url)
  57         # Otherwise we get a webpage that would execute some javascript
  58         req.add_header('User-Agent', 'curl/7')
  59         webpage = self._download_webpage(req, mtvn_id,
  60                                          'Downloading mobile page')
  61         metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
  62         req = HEADRequest(metrics_url)
  63         response = self._request_webpage(req, mtvn_id, 'Resolving url')
  64         url = response.geturl()
  65         # Transform the url to get the best quality:
  66         url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
  67         return [{'url': url, 'ext': 'mp4'}]
  68
  69     def _extract_video_formats(self, mdoc, mtvn_id):
  70         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4|copyright_error\.flv(?:\?geo\b.+?)?)$', mdoc.find('.//src').text) is not None:
  71             if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
  72                 self.to_screen('The normal version is not available from your '
  73                                'country, trying with the mobile version')
  74                 return self._extract_mobile_video_formats(mtvn_id)
  75             raise ExtractorError('This video is not available from your country.',
  76                                  expected=True)
  77
  78         formats = []
  79         for rendition in mdoc.findall('.//rendition'):
  80             try:
  81                 _, _, ext = rendition.attrib['type'].partition('/')
  82                 rtmp_video_url = rendition.find('./src').text
  83                 if rtmp_video_url.endswith('siteunavail.png'):
  84                     continue
  85                 formats.append({
  86                     'ext': ext,
  87                     'url': self._transform_rtmp_url(rtmp_video_url),
  88                     'format_id': rendition.get('bitrate'),
  89                     'width': int(rendition.get('width')),
  90                     'height': int(rendition.get('height')),
  91                 })
  92             except (KeyError, TypeError):
  93                 raise ExtractorError('Invalid rendition field.')
  94         self._sort_formats(formats)
  95         return formats
  96
  97     def _extract_subtitles(self, mdoc, mtvn_id):
  98         subtitles = {}
  99         for transcript in mdoc.findall('.//transcript'):
 100             if transcript.get('kind') != 'captions':
 101                 continue
 102             lang = transcript.get('srclang')
 103             subtitles[lang] = [{
 104                 'url': compat_str(typographic.get('src')),
 105                 'ext': typographic.get('format')
 106             } for typographic in transcript.findall('./typographic')]
 107         return subtitles
 108
 109     def _get_video_info(self, itemdoc):
 110         uri = itemdoc.find('guid').text
 111         video_id = self._id_from_uri(uri)
 112         self.report_extraction(video_id)
 113         mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
 114         # Remove the templates, like &device={device}
 115         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
 116         if 'acceptMethods' not in mediagen_url:
 117             mediagen_url += '&' if '?' in mediagen_url else '?'
 118             mediagen_url += 'acceptMethods=fms'
 119
 120         mediagen_doc = self._download_xml(mediagen_url, video_id,
 121                                           'Downloading video urls')
 122
 123         item = mediagen_doc.find('./video/item')
 124         if item is not None and item.get('type') == 'text':
 125             message = '%s returned error: ' % self.IE_NAME
 126             if item.get('code') is not None:
 127                 message += '%s - ' % item.get('code')
 128             message += item.text
 129             raise ExtractorError(message, expected=True)
 130
 131         description_node = itemdoc.find('description')
 132         if description_node is not None:
 133             description = description_node.text.strip()
 134         else:
 135             description = None
 136
 137         title_el = None
 138         if title_el is None:
 139             title_el = find_xpath_attr(
 140                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 141                 'scheme', 'urn:mtvn:video_title')
 142         if title_el is None:
 143             title_el = itemdoc.find('.//{http://search.yahoo.com/mrss/}title')
 144         if title_el is None:
 145             title_el = itemdoc.find('.//title') or itemdoc.find('./title')
 146             if title_el.text is None:
 147                 title_el = None
 148
 149         title = title_el.text
 150         if title is None:
 151             raise ExtractorError('Could not find video title')
 152         title = title.strip()
 153
 154         # This a short id that's used in the webpage urls
 155         mtvn_id = None
 156         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 157                                        'scheme', 'urn:mtvn:id')
 158         if mtvn_id_node is not None:
 159             mtvn_id = mtvn_id_node.text
 160
 161         return {
 162             'title': title,
 163             'formats': self._extract_video_formats(mediagen_doc, mtvn_id),
 164             'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id),
 165             'id': video_id,
 166             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 167             'description': description,
 168         }
 169
 170     def _get_feed_query(self, uri):
 171         data = {'uri': uri}
 172         if self._LANG:
 173             data['lang'] = self._LANG
 174         return compat_urllib_parse.urlencode(data)
 175
 176     def _get_videos_info(self, uri):
 177         video_id = self._id_from_uri(uri)
 178         feed_url = self._get_feed_url(uri)
 179         info_url = feed_url + '?' + self._get_feed_query(uri)
 180         return self._get_videos_info_from_url(info_url, video_id)
 181
 182     def _get_videos_info_from_url(self, url, video_id):
 183         idoc = self._download_xml(
 184             url, video_id,
 185             'Downloading info', transform_source=fix_xml_ampersands)
 186         return self.playlist_result(
 187             [self._get_video_info(item) for item in idoc.findall('.//item')])
 188
 189     def _extract_mgid(self, webpage):
 190         try:
 191             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 192             # or http://media.mtvnservices.com/{mgid}
 193             og_url = self._og_search_video_url(webpage)
 194             mgid = url_basename(og_url)
 195             if mgid.endswith('.swf'):
 196                 mgid = mgid[:-4]
 197         except RegexNotFoundError:
 198             mgid = None
 199
 200         if mgid is None or ':' not in mgid:
 201             mgid = self._search_regex(
 202                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
 203                 webpage, 'mgid', default=None)
 204
 205         if not mgid:
 206             sm4_embed = self._html_search_meta(
 207                 'sm4:video:embed', webpage, 'sm4 embed', default='')
 208             mgid = self._search_regex(
 209                 r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid')
 210         return mgid
 211
 212     def _real_extract(self, url):
 213         title = url_basename(url)
 214         webpage = self._download_webpage(url, title)
 215         mgid = self._extract_mgid(webpage)
 216         videos_info = self._get_videos_info(mgid)
 217         return videos_info
 218
 219
 220 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
 221     IE_NAME = 'mtvservices:embedded'
 222     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
 223
 224     _TEST = {
 225         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
 226         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
 227         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
 228         'info_dict': {
 229             'id': '1043906',
 230             'ext': 'mp4',
 231             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
 232             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
 233         },
 234     }
 235
 236     @staticmethod
 237     def _extract_url(webpage):
 238         mobj = re.search(
 239             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//media.mtvnservices.com/embed/.+?)\1', webpage)
 240         if mobj:
 241             return mobj.group('url')
 242
 243     def _get_feed_url(self, uri):
 244         video_id = self._id_from_uri(uri)
 245         site_id = uri.replace(video_id, '')
 246         config_url = ('http://media.mtvnservices.com/pmt/e1/players/{0}/'
 247                       'context4/context5/config.xml'.format(site_id))
 248         config_doc = self._download_xml(config_url, video_id)
 249         feed_node = config_doc.find('.//feed')
 250         feed_url = feed_node.text.strip().split('?')[0]
 251         return feed_url
 252
 253     def _real_extract(self, url):
 254         mobj = re.match(self._VALID_URL, url)
 255         mgid = mobj.group('mgid')
 256         return self._get_videos_info(mgid)
 257
 258
 259 class MTVIE(MTVServicesInfoExtractor):
 260     _VALID_URL = r'''(?x)^https?://
 261         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 262            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 263
 264     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 265
 266     _TESTS = [
 267         {
 268             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 269             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 270             'info_dict': {
 271                 'id': '853555',
 272                 'ext': 'mp4',
 273                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 274                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 275             },
 276         },
 277     ]
 278
 279     def _get_thumbnail_url(self, uri, itemdoc):
 280         return 'http://mtv.mtvnimages.com/uri/' + uri
 281
 282     def _real_extract(self, url):
 283         mobj = re.match(self._VALID_URL, url)
 284         video_id = mobj.group('videoid')
 285         uri = mobj.groupdict().get('mgid')
 286         if uri is None:
 287             webpage = self._download_webpage(url, video_id)
 288
 289             # Some videos come from Vevo.com
 290             m_vevo = re.search(
 291                 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
 292             if m_vevo:
 293                 vevo_id = m_vevo.group(1)
 294                 self.to_screen('Vevo video detected: %s' % vevo_id)
 295                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 296
 297             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 298         return self._get_videos_info(uri)
 299
 300
 301 class MTVIggyIE(MTVServicesInfoExtractor):
 302     IE_NAME = 'mtviggy.com'
 303     _VALID_URL = r'https?://www\.mtviggy\.com/videos/.+'
 304     _TEST = {
 305         'url': 'http://www.mtviggy.com/videos/arcade-fire-behind-the-scenes-at-the-biggest-music-experiment-yet/',
 306         'info_dict': {
 307             'id': '984696',
 308             'ext': 'mp4',
 309             'title': 'Arcade Fire: Behind the Scenes at the Biggest Music Experiment Yet',
 310         }
 311     }
 312     _FEED_URL = 'http://all.mtvworldverticals.com/feed-xml/'
 313
 314
 315 class MTVDEIE(MTVServicesInfoExtractor):
 316     IE_NAME = 'mtv.de'
 317     _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:artists|shows|news)/(?:[^/]+/)*(?P<id>\d+)-[^/#?]+/*(?:[#?].*)?$'
 318     _TESTS = [{
 319         'url': 'http://www.mtv.de/artists/10571-cro/videos/61131-traum',
 320         'info_dict': {
 321             'id': 'music_video-a50bc5f0b3aa4b3190aa',
 322             'ext': 'mp4',
 323             'title': 'MusicVideo_cro-traum',
 324             'description': 'Cro - Traum',
 325         },
 326         'params': {
 327             # rtmp download
 328             'skip_download': True,
 329         },
 330     }, {
 331         # mediagen URL without query (e.g. http://videos.mtvnn.com/mediagen/e865da714c166d18d6f80893195fcb97)
 332         'url': 'http://www.mtv.de/shows/933-teen-mom-2/staffeln/5353/folgen/63565-enthullungen',
 333         'info_dict': {
 334             'id': 'local_playlist-f5ae778b9832cc837189',
 335             'ext': 'mp4',
 336             'title': 'Episode_teen-mom-2_shows_season-5_episode-1_full-episode_part1',
 337         },
 338         'params': {
 339             # rtmp download
 340             'skip_download': True,
 341         },
 342     }, {
 343         # single video in pagePlaylist with different id
 344         'url': 'http://www.mtv.de/news/77491-mtv-movies-spotlight-pixels-teil-3',
 345         'info_dict': {
 346             'id': 'local_playlist-4e760566473c4c8c5344',
 347             'ext': 'mp4',
 348             'title': 'Article_mtv-movies-spotlight-pixels-teil-3_short-clips_part1',
 349             'description': 'MTV Movies Supercut',
 350         },
 351         'params': {
 352             # rtmp download
 353             'skip_download': True,
 354         },
 355     }]
 356
 357     def _real_extract(self, url):
 358         video_id = self._match_id(url)
 359
 360         webpage = self._download_webpage(url, video_id)
 361
 362         playlist = self._parse_json(
 363             self._search_regex(
 364                 r'window\.pagePlaylist\s*=\s*(\[.+?\]);\n', webpage, 'page playlist'),
 365             video_id)
 366
 367         # news pages contain single video in playlist with different id
 368         if len(playlist) == 1:
 369             return self._get_videos_info_from_url(playlist[0]['mrss'], video_id)
 370
 371         for item in playlist:
 372             item_id = item.get('id')
 373             if item_id and compat_str(item_id) == video_id:
 374                 return self._get_videos_info_from_url(item['mrss'], video_id)