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, video_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             if rendition.get('method') == 'hls':
  92                 hls_url = rendition.find('./src').text
  93                 formats.extend(self._extract_m3u8_formats(hls_url, video_id, ext='mp4'))
  94             else:
  95                 # fms
  96                 try:
  97                     _, _, ext = rendition.attrib['type'].partition('/')
  98                     rtmp_video_url = rendition.find('./src').text
  99                     if rtmp_video_url.endswith('siteunavail.png'):
 100                         continue
 101                     new_urls = self._transform_rtmp_url(rtmp_video_url)
 102                     formats.extend([{
 103                         'ext': 'flv' if new_url.startswith('rtmp') else ext,
 104                         'url': new_url,
 105                         'format_id': '-'.join(filter(None, [kind, rendition.get('bitrate')])),
 106                         'width': int(rendition.get('width')),
 107                         'height': int(rendition.get('height')),
 108                     } for kind, new_url in new_urls.items()])
 109                 except (KeyError, TypeError):
 110                     raise ExtractorError('Invalid rendition field.')
 111         self._sort_formats(formats)
 112         return formats
 113
 114     def _extract_subtitles(self, mdoc, mtvn_id):
 115         subtitles = {}
 116         for transcript in mdoc.findall('.//transcript'):
 117             if transcript.get('kind') != 'captions':
 118                 continue
 119             lang = transcript.get('srclang')
 120             subtitles[lang] = [{
 121                 'url': compat_str(typographic.get('src')),
 122                 'ext': typographic.get('format')
 123             } for typographic in transcript.findall('./typographic')]
 124         return subtitles
 125
 126     def _get_video_info(self, itemdoc, use_hls=True):
 127         uri = itemdoc.find('guid').text
 128         video_id = self._id_from_uri(uri)
 129         self.report_extraction(video_id)
 130         content_el = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content')))
 131         mediagen_url = self._remove_template_parameter(content_el.attrib['url'])
 132         mediagen_url = mediagen_url.replace('device={device}', '')
 133         if 'acceptMethods' not in mediagen_url:
 134             mediagen_url += '&' if '?' in mediagen_url else '?'
 135             mediagen_url += 'acceptMethods='
 136             mediagen_url += 'hls' if use_hls else 'fms'
 137
 138         mediagen_doc = self._download_xml(mediagen_url, video_id,
 139                                           'Downloading video urls')
 140
 141         item = mediagen_doc.find('./video/item')
 142         if item is not None and item.get('type') == 'text':
 143             message = '%s returned error: ' % self.IE_NAME
 144             if item.get('code') is not None:
 145                 message += '%s - ' % item.get('code')
 146             message += item.text
 147             raise ExtractorError(message, expected=True)
 148
 149         description = strip_or_none(xpath_text(itemdoc, 'description'))
 150
 151         timestamp = timeconvert(xpath_text(itemdoc, 'pubDate'))
 152
 153         title_el = None
 154         if title_el is None:
 155             title_el = find_xpath_attr(
 156                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 157                 'scheme', 'urn:mtvn:video_title')
 158         if title_el is None:
 159             title_el = itemdoc.find(compat_xpath('.//{http://search.yahoo.com/mrss/}title'))
 160         if title_el is None:
 161             title_el = itemdoc.find(compat_xpath('.//title'))
 162             if title_el.text is None:
 163                 title_el = None
 164
 165         title = title_el.text
 166         if title is None:
 167             raise ExtractorError('Could not find video title')
 168         title = title.strip()
 169
 170         # This a short id that's used in the webpage urls
 171         mtvn_id = None
 172         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 173                                        'scheme', 'urn:mtvn:id')
 174         if mtvn_id_node is not None:
 175             mtvn_id = mtvn_id_node.text
 176
 177         formats = self._extract_video_formats(mediagen_doc, mtvn_id, video_id)
 178
 179         return {
 180             'title': title,
 181             'formats': formats,
 182             'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id),
 183             'id': video_id,
 184             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 185             'description': description,
 186             'duration': float_or_none(content_el.attrib.get('duration')),
 187             'timestamp': timestamp,
 188         }
 189
 190     def _get_feed_query(self, uri):
 191         data = {'uri': uri}
 192         if self._LANG:
 193             data['lang'] = self._LANG
 194         return data
 195
 196     def _get_videos_info(self, uri, use_hls=True):
 197         video_id = self._id_from_uri(uri)
 198         feed_url = self._get_feed_url(uri)
 199         info_url = update_url_query(feed_url, self._get_feed_query(uri))
 200         return self._get_videos_info_from_url(info_url, video_id, use_hls)
 201
 202     def _get_videos_info_from_url(self, url, video_id, use_hls=True):
 203         idoc = self._download_xml(
 204             url, video_id,
 205             'Downloading info', transform_source=fix_xml_ampersands)
 206
 207         title = xpath_text(idoc, './channel/title')
 208         description = xpath_text(idoc, './channel/description')
 209
 210         return self.playlist_result(
 211             [self._get_video_info(item, use_hls) for item in idoc.findall('.//item')],
 212             playlist_title=title, playlist_description=description)
 213
 214     def _extract_mgid(self, webpage, default=NO_DEFAULT):
 215         try:
 216             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 217             # or http://media.mtvnservices.com/{mgid}
 218             og_url = self._og_search_video_url(webpage)
 219             mgid = url_basename(og_url)
 220             if mgid.endswith('.swf'):
 221                 mgid = mgid[:-4]
 222         except RegexNotFoundError:
 223             mgid = None
 224
 225         if mgid is None or ':' not in mgid:
 226             mgid = self._search_regex(
 227                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
 228                 webpage, 'mgid', default=None)
 229
 230         if not mgid:
 231             sm4_embed = self._html_search_meta(
 232                 'sm4:video:embed', webpage, 'sm4 embed', default='')
 233             mgid = self._search_regex(
 234                 r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid', default=default)
 235         return mgid
 236
 237     def _real_extract(self, url):
 238         title = url_basename(url)
 239         webpage = self._download_webpage(url, title)
 240         mgid = self._extract_mgid(webpage)
 241         videos_info = self._get_videos_info(mgid)
 242         return videos_info
 243
 244
 245 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
 246     IE_NAME = 'mtvservices:embedded'
 247     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
 248
 249     _TEST = {
 250         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
 251         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
 252         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
 253         'info_dict': {
 254             'id': '1043906',
 255             'ext': 'mp4',
 256             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
 257             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
 258             'timestamp': 1400126400,
 259             'upload_date': '20140515',
 260         },
 261     }
 262
 263     @staticmethod
 264     def _extract_url(webpage):
 265         mobj = re.search(
 266             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//media.mtvnservices.com/embed/.+?)\1', webpage)
 267         if mobj:
 268             return mobj.group('url')
 269
 270     def _get_feed_url(self, uri):
 271         video_id = self._id_from_uri(uri)
 272         config = self._download_json(
 273             'http://media.mtvnservices.com/pmt/e1/access/index.html?uri=%s&configtype=edge' % uri, video_id)
 274         return self._remove_template_parameter(config['feedWithQueryParams'])
 275
 276     def _real_extract(self, url):
 277         mobj = re.match(self._VALID_URL, url)
 278         mgid = mobj.group('mgid')
 279         return self._get_videos_info(mgid)
 280
 281
 282 class MTVIE(MTVServicesInfoExtractor):
 283     IE_NAME = 'mtv'
 284     _VALID_URL = r'https?://(?:www\.)?mtv\.com/(?:video-clips|full-episodes)/(?P<id>[^/?#.]+)'
 285     _FEED_URL = 'http://www.mtv.com/feeds/mrss/'
 286
 287     _TESTS = [{
 288         'url': 'http://www.mtv.com/video-clips/vl8qof/unlocking-the-truth-trailer',
 289         'md5': '1edbcdf1e7628e414a8c5dcebca3d32b',
 290         'info_dict': {
 291             'id': '5e14040d-18a4-47c4-a582-43ff602de88e',
 292             'ext': 'mp4',
 293             'title': 'Unlocking The Truth|July 18, 2016|1|101|Trailer',
 294             'description': '"Unlocking the Truth" premieres August 17th at 11/10c.',
 295             'timestamp': 1468846800,
 296             'upload_date': '20160718',
 297         },
 298     }, {
 299         'url': 'http://www.mtv.com/full-episodes/94tujl/unlocking-the-truth-gates-of-hell-season-1-ep-101',
 300         'only_matching': True,
 301     }]
 302
 303
 304 class MTVVideoIE(MTVServicesInfoExtractor):
 305     IE_NAME = 'mtv:video'
 306     _VALID_URL = r'''(?x)^https?://
 307         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 308            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 309
 310     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 311
 312     _TESTS = [
 313         {
 314             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 315             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 316             'info_dict': {
 317                 'id': '853555',
 318                 'ext': 'mp4',
 319                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 320                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 321                 'timestamp': 1352610000,
 322                 'upload_date': '20121111',
 323             },
 324         },
 325     ]
 326
 327     def _get_thumbnail_url(self, uri, itemdoc):
 328         return 'http://mtv.mtvnimages.com/uri/' + uri
 329
 330     def _real_extract(self, url):
 331         mobj = re.match(self._VALID_URL, url)
 332         video_id = mobj.group('videoid')
 333         uri = mobj.groupdict().get('mgid')
 334         if uri is None:
 335             webpage = self._download_webpage(url, video_id)
 336
 337             # Some videos come from Vevo.com
 338             m_vevo = re.search(
 339                 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
 340             if m_vevo:
 341                 vevo_id = m_vevo.group(1)
 342                 self.to_screen('Vevo video detected: %s' % vevo_id)
 343                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 344
 345             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 346         return self._get_videos_info(uri)
 347
 348
 349 class MTVDEIE(MTVServicesInfoExtractor):
 350     IE_NAME = 'mtv.de'
 351     _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:artists|shows|news)/(?:[^/]+/)*(?P<id>\d+)-[^/#?]+/*(?:[#?].*)?$'
 352     _TESTS = [{
 353         'url': 'http://www.mtv.de/artists/10571-cro/videos/61131-traum',
 354         'info_dict': {
 355             'id': 'music_video-a50bc5f0b3aa4b3190aa',
 356             'ext': 'flv',
 357             'title': 'MusicVideo_cro-traum',
 358             'description': 'Cro - Traum',
 359         },
 360         'params': {
 361             # rtmp download
 362             'skip_download': True,
 363         },
 364         'skip': 'Blocked at Travis CI',
 365     }, {
 366         # mediagen URL without query (e.g. http://videos.mtvnn.com/mediagen/e865da714c166d18d6f80893195fcb97)
 367         'url': 'http://www.mtv.de/shows/933-teen-mom-2/staffeln/5353/folgen/63565-enthullungen',
 368         'info_dict': {
 369             'id': 'local_playlist-f5ae778b9832cc837189',
 370             'ext': 'flv',
 371             'title': 'Episode_teen-mom-2_shows_season-5_episode-1_full-episode_part1',
 372         },
 373         'params': {
 374             # rtmp download
 375             'skip_download': True,
 376         },
 377         'skip': 'Blocked at Travis CI',
 378     }, {
 379         'url': 'http://www.mtv.de/news/77491-mtv-movies-spotlight-pixels-teil-3',
 380         'info_dict': {
 381             'id': 'local_playlist-4e760566473c4c8c5344',
 382             'ext': 'mp4',
 383             'title': 'Article_mtv-movies-spotlight-pixels-teil-3_short-clips_part1',
 384             'description': 'MTV Movies Supercut',
 385         },
 386         'params': {
 387             # rtmp download
 388             'skip_download': True,
 389         },
 390         'skip': 'Das Video kann zur Zeit nicht abgespielt werden.',
 391     }]
 392
 393     def _real_extract(self, url):
 394         video_id = self._match_id(url)
 395
 396         webpage = self._download_webpage(url, video_id)
 397
 398         playlist = self._parse_json(
 399             self._search_regex(
 400                 r'window\.pagePlaylist\s*=\s*(\[.+?\]);\n', webpage, 'page playlist'),
 401             video_id)
 402
 403         def _mrss_url(item):
 404             return item['mrss'] + item.get('mrssvars', '')
 405
 406         # news pages contain single video in playlist with different id
 407         if len(playlist) == 1:
 408             return self._get_videos_info_from_url(_mrss_url(playlist[0]), video_id)
 409
 410         for item in playlist:
 411             item_id = item.get('id')
 412             if item_id and compat_str(item_id) == video_id:
 413                 return self._get_videos_info_from_url(_mrss_url(item), video_id)