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