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'&[^=]*?={.*?}(?=(&|$))', '', url)
  44
  45     def _get_feed_url(self, uri):
  46         return self._FEED_URL
  47
  48     def _get_thumbnail_url(self, uri, itemdoc):
  49         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  50         thumb_node = itemdoc.find(search_path)
  51         if thumb_node is None:
  52             return None
  53         return thumb_node.get('url') or thumb_node.text or None
  54
  55     def _extract_mobile_video_formats(self, mtvn_id):
  56         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
  57         req = sanitized_Request(webpage_url)
  58         # Otherwise we get a webpage that would execute some javascript
  59         req.add_header('User-Agent', 'curl/7')
  60         webpage = self._download_webpage(req, mtvn_id,
  61                                          'Downloading mobile page')
  62         metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
  63         req = HEADRequest(metrics_url)
  64         response = self._request_webpage(req, mtvn_id, 'Resolving url')
  65         url = response.geturl()
  66         # Transform the url to get the best quality:
  67         url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
  68         return [{'url': url, 'ext': 'mp4'}]
  69
  70     def _extract_video_formats(self, mdoc, mtvn_id, video_id):
  71         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4|copyright_error\.flv(?:\?geo\b.+?)?)$', mdoc.find('.//src').text) is not None:
  72             if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
  73                 self.to_screen('The normal version is not available from your '
  74                                'country, trying with the mobile version')
  75                 return self._extract_mobile_video_formats(mtvn_id)
  76             raise ExtractorError('This video is not available from your country.',
  77                                  expected=True)
  78
  79         formats = []
  80         for rendition in mdoc.findall('.//rendition'):
  81             if rendition.get('method') == 'hls':
  82                 hls_url = rendition.find('./src').text
  83                 formats.extend(self._extract_m3u8_formats(
  84                     hls_url, video_id, ext='mp4', entry_protocol='m3u8_native',
  85                     m3u8_id='hls', fatal=False))
  86             else:
  87                 # fms
  88                 try:
  89                     _, _, ext = rendition.attrib['type'].partition('/')
  90                     rtmp_video_url = rendition.find('./src').text
  91                     if 'error_not_available.swf' in rtmp_video_url:
  92                         raise ExtractorError(
  93                             '%s said: video is not available' % self.IE_NAME,
  94                             expected=True)
  95                     if rtmp_video_url.endswith('siteunavail.png'):
  96                         continue
  97                     formats.extend([{
  98                         'ext': 'flv' if rtmp_video_url.startswith('rtmp') else ext,
  99                         'url': rtmp_video_url,
 100                         'format_id': '-'.join(filter(None, [
 101                             'rtmp' if rtmp_video_url.startswith('rtmp') else None,
 102                             rendition.get('bitrate')])),
 103                         'width': int(rendition.get('width')),
 104                         'height': int(rendition.get('height')),
 105                     }])
 106                 except (KeyError, TypeError):
 107                     raise ExtractorError('Invalid rendition field.')
 108         if formats:
 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             for typographic in transcript.findall('./typographic'):
 119                 sub_src = typographic.get('src')
 120                 if not sub_src:
 121                     continue
 122                 ext = typographic.get('format')
 123                 if ext == 'cea-608':
 124                     ext = 'scc'
 125                 subtitles.setdefault(lang, []).append({
 126                     'url': compat_str(sub_src),
 127                     'ext': ext
 128                 })
 129         return subtitles
 130
 131     def _get_video_info(self, itemdoc, use_hls=True):
 132         uri = itemdoc.find('guid').text
 133         video_id = self._id_from_uri(uri)
 134         self.report_extraction(video_id)
 135         content_el = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content')))
 136         mediagen_url = self._remove_template_parameter(content_el.attrib['url'])
 137         mediagen_url = mediagen_url.replace('device={device}', '')
 138         if 'acceptMethods' not in mediagen_url:
 139             mediagen_url += '&' if '?' in mediagen_url else '?'
 140             mediagen_url += 'acceptMethods='
 141             mediagen_url += 'hls' if use_hls else 'fms'
 142
 143         mediagen_doc = self._download_xml(
 144             mediagen_url, video_id, 'Downloading video urls', fatal=False)
 145
 146         if mediagen_doc is False:
 147             return None
 148
 149         item = mediagen_doc.find('./video/item')
 150         if item is not None and item.get('type') == 'text':
 151             message = '%s returned error: ' % self.IE_NAME
 152             if item.get('code') is not None:
 153                 message += '%s - ' % item.get('code')
 154             message += item.text
 155             raise ExtractorError(message, expected=True)
 156
 157         description = strip_or_none(xpath_text(itemdoc, 'description'))
 158
 159         timestamp = timeconvert(xpath_text(itemdoc, 'pubDate'))
 160
 161         title_el = None
 162         if title_el is None:
 163             title_el = find_xpath_attr(
 164                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 165                 'scheme', 'urn:mtvn:video_title')
 166         if title_el is None:
 167             title_el = itemdoc.find(compat_xpath('.//{http://search.yahoo.com/mrss/}title'))
 168         if title_el is None:
 169             title_el = itemdoc.find(compat_xpath('.//title'))
 170             if title_el.text is None:
 171                 title_el = None
 172
 173         title = title_el.text
 174         if title is None:
 175             raise ExtractorError('Could not find video title')
 176         title = title.strip()
 177
 178         # This a short id that's used in the webpage urls
 179         mtvn_id = None
 180         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 181                                        'scheme', 'urn:mtvn:id')
 182         if mtvn_id_node is not None:
 183             mtvn_id = mtvn_id_node.text
 184
 185         formats = self._extract_video_formats(mediagen_doc, mtvn_id, video_id)
 186
 187         # Some parts of complete video may be missing (e.g. missing Act 3 in
 188         # http://www.southpark.de/alle-episoden/s14e01-sexual-healing)
 189         if not formats:
 190             return None
 191
 192         self._sort_formats(formats)
 193
 194         return {
 195             'title': title,
 196             'formats': formats,
 197             'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id),
 198             'id': video_id,
 199             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 200             'description': description,
 201             'duration': float_or_none(content_el.attrib.get('duration')),
 202             'timestamp': timestamp,
 203         }
 204
 205     def _get_feed_query(self, uri):
 206         data = {'uri': uri}
 207         if self._LANG:
 208             data['lang'] = self._LANG
 209         return data
 210
 211     def _get_videos_info(self, uri, use_hls=True):
 212         video_id = self._id_from_uri(uri)
 213         feed_url = self._get_feed_url(uri)
 214         info_url = update_url_query(feed_url, self._get_feed_query(uri))
 215         return self._get_videos_info_from_url(info_url, video_id, use_hls)
 216
 217     def _get_videos_info_from_url(self, url, video_id, use_hls=True):
 218         idoc = self._download_xml(
 219             url, video_id,
 220             'Downloading info', transform_source=fix_xml_ampersands)
 221
 222         title = xpath_text(idoc, './channel/title')
 223         description = xpath_text(idoc, './channel/description')
 224
 225         entries = []
 226         for item in idoc.findall('.//item'):
 227             info = self._get_video_info(item, use_hls)
 228             if info:
 229                 entries.append(info)
 230
 231         return self.playlist_result(
 232             entries, playlist_title=title, playlist_description=description)
 233
 234     def _extract_triforce_mgid(self, webpage, data_zone=None, video_id=None):
 235         triforce_feed = self._parse_json(self._search_regex(
 236             r'triforceManifestFeed\s*=\s*({.+?})\s*;\s*\n', webpage,
 237             'triforce feed', default='{}'), video_id, fatal=False)
 238
 239         data_zone = self._search_regex(
 240             r'data-zone=(["\'])(?P<zone>.+?_lc_promo.*?)\1', webpage,
 241             'data zone', default=data_zone, group='zone')
 242
 243         feed_url = try_get(
 244             triforce_feed, lambda x: x['manifest']['zones'][data_zone]['feed'],
 245             compat_str)
 246         if not feed_url:
 247             return
 248
 249         feed = self._download_json(feed_url, video_id, fatal=False)
 250         if not feed:
 251             return
 252
 253         return try_get(feed, lambda x: x['result']['data']['id'], compat_str)
 254
 255     def _extract_mgid(self, webpage):
 256         try:
 257             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 258             # or http://media.mtvnservices.com/{mgid}
 259             og_url = self._og_search_video_url(webpage)
 260             mgid = url_basename(og_url)
 261             if mgid.endswith('.swf'):
 262                 mgid = mgid[:-4]
 263         except RegexNotFoundError:
 264             mgid = None
 265
 266         if mgid is None or ':' not in mgid:
 267             mgid = self._search_regex(
 268                 [r'data-mgid="(.*?)"', r'swfobject\.embedSWF\(".*?(mgid:.*?)"'],
 269                 webpage, 'mgid', default=None)
 270
 271         if not mgid:
 272             sm4_embed = self._html_search_meta(
 273                 'sm4:video:embed', webpage, 'sm4 embed', default='')
 274             mgid = self._search_regex(
 275                 r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid', default=None)
 276
 277         if not mgid:
 278             mgid = self._extract_triforce_mgid(webpage)
 279
 280         return mgid
 281
 282     def _real_extract(self, url):
 283         title = url_basename(url)
 284         webpage = self._download_webpage(url, title)
 285         mgid = self._extract_mgid(webpage)
 286         videos_info = self._get_videos_info(mgid)
 287         return videos_info
 288
 289
 290 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
 291     IE_NAME = 'mtvservices:embedded'
 292     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
 293
 294     _TEST = {
 295         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
 296         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
 297         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
 298         'info_dict': {
 299             'id': '1043906',
 300             'ext': 'mp4',
 301             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
 302             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
 303             'timestamp': 1400126400,
 304             'upload_date': '20140515',
 305         },
 306     }
 307
 308     @staticmethod
 309     def _extract_url(webpage):
 310         mobj = re.search(
 311             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//media.mtvnservices.com/embed/.+?)\1', webpage)
 312         if mobj:
 313             return mobj.group('url')
 314
 315     def _get_feed_url(self, uri):
 316         video_id = self._id_from_uri(uri)
 317         config = self._download_json(
 318             'http://media.mtvnservices.com/pmt/e1/access/index.html?uri=%s&configtype=edge' % uri, video_id)
 319         return self._remove_template_parameter(config['feedWithQueryParams'])
 320
 321     def _real_extract(self, url):
 322         mobj = re.match(self._VALID_URL, url)
 323         mgid = mobj.group('mgid')
 324         return self._get_videos_info(mgid)
 325
 326
 327 class MTVIE(MTVServicesInfoExtractor):
 328     IE_NAME = 'mtv'
 329     _VALID_URL = r'https?://(?:www\.)?mtv\.com/(?:video-clips|(?:full-)?episodes)/(?P<id>[^/?#.]+)'
 330     _FEED_URL = 'http://www.mtv.com/feeds/mrss/'
 331
 332     _TESTS = [{
 333         'url': 'http://www.mtv.com/video-clips/vl8qof/unlocking-the-truth-trailer',
 334         'md5': '1edbcdf1e7628e414a8c5dcebca3d32b',
 335         'info_dict': {
 336             'id': '5e14040d-18a4-47c4-a582-43ff602de88e',
 337             'ext': 'mp4',
 338             'title': 'Unlocking The Truth|July 18, 2016|1|101|Trailer',
 339             'description': '"Unlocking the Truth" premieres August 17th at 11/10c.',
 340             'timestamp': 1468846800,
 341             'upload_date': '20160718',
 342         },
 343     }, {
 344         'url': 'http://www.mtv.com/full-episodes/94tujl/unlocking-the-truth-gates-of-hell-season-1-ep-101',
 345         'only_matching': True,
 346     }, {
 347         'url': 'http://www.mtv.com/episodes/g8xu7q/teen-mom-2-breaking-the-wall-season-7-ep-713',
 348         'only_matching': True,
 349     }]
 350
 351
 352 class MTV81IE(InfoExtractor):
 353     IE_NAME = 'mtv81'
 354     _VALID_URL = r'https?://(?:www\.)?mtv81\.com/videos/(?P<id>[^/?#.]+)'
 355
 356     _TEST = {
 357         'url': 'http://www.mtv81.com/videos/artist-to-watch/the-godfather-of-japanese-hip-hop-segment-1/',
 358         'md5': '1edbcdf1e7628e414a8c5dcebca3d32b',
 359         'info_dict': {
 360             'id': '5e14040d-18a4-47c4-a582-43ff602de88e',
 361             'ext': 'mp4',
 362             'title': 'Unlocking The Truth|July 18, 2016|1|101|Trailer',
 363             'description': '"Unlocking the Truth" premieres August 17th at 11/10c.',
 364             'timestamp': 1468846800,
 365             'upload_date': '20160718',
 366         },
 367     }
 368
 369     def _extract_mgid(self, webpage):
 370         return self._search_regex(
 371             r'getTheVideo\((["\'])(?P<id>mgid:.+?)\1', webpage,
 372             'mgid', group='id')
 373
 374     def _real_extract(self, url):
 375         video_id = self._match_id(url)
 376         webpage = self._download_webpage(url, video_id)
 377         mgid = self._extract_mgid(webpage)
 378         return self.url_result('http://media.mtvnservices.com/embed/%s' % mgid)
 379
 380
 381 class MTVVideoIE(MTVServicesInfoExtractor):
 382     IE_NAME = 'mtv:video'
 383     _VALID_URL = r'''(?x)^https?://
 384         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 385            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 386
 387     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 388
 389     _TESTS = [
 390         {
 391             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 392             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 393             'info_dict': {
 394                 'id': '853555',
 395                 'ext': 'mp4',
 396                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 397                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 398                 'timestamp': 1352610000,
 399                 'upload_date': '20121111',
 400             },
 401         },
 402     ]
 403
 404     def _get_thumbnail_url(self, uri, itemdoc):
 405         return 'http://mtv.mtvnimages.com/uri/' + uri
 406
 407     def _real_extract(self, url):
 408         mobj = re.match(self._VALID_URL, url)
 409         video_id = mobj.group('videoid')
 410         uri = mobj.groupdict().get('mgid')
 411         if uri is None:
 412             webpage = self._download_webpage(url, video_id)
 413
 414             # Some videos come from Vevo.com
 415             m_vevo = re.search(
 416                 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
 417             if m_vevo:
 418                 vevo_id = m_vevo.group(1)
 419                 self.to_screen('Vevo video detected: %s' % vevo_id)
 420                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 421
 422             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 423         return self._get_videos_info(uri)
 424
 425
 426 class MTVDEIE(MTVServicesInfoExtractor):
 427     IE_NAME = 'mtv.de'
 428     _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:artists|shows|news)/(?:[^/]+/)*(?P<id>\d+)-[^/#?]+/*(?:[#?].*)?$'
 429     _TESTS = [{
 430         'url': 'http://www.mtv.de/artists/10571-cro/videos/61131-traum',
 431         'info_dict': {
 432             'id': 'music_video-a50bc5f0b3aa4b3190aa',
 433             'ext': 'flv',
 434             'title': 'MusicVideo_cro-traum',
 435             'description': 'Cro - Traum',
 436         },
 437         'params': {
 438             # rtmp download
 439             'skip_download': True,
 440         },
 441         'skip': 'Blocked at Travis CI',
 442     }, {
 443         # mediagen URL without query (e.g. http://videos.mtvnn.com/mediagen/e865da714c166d18d6f80893195fcb97)
 444         'url': 'http://www.mtv.de/shows/933-teen-mom-2/staffeln/5353/folgen/63565-enthullungen',
 445         'info_dict': {
 446             'id': 'local_playlist-f5ae778b9832cc837189',
 447             'ext': 'flv',
 448             'title': 'Episode_teen-mom-2_shows_season-5_episode-1_full-episode_part1',
 449         },
 450         'params': {
 451             # rtmp download
 452             'skip_download': True,
 453         },
 454         'skip': 'Blocked at Travis CI',
 455     }, {
 456         'url': 'http://www.mtv.de/news/77491-mtv-movies-spotlight-pixels-teil-3',
 457         'info_dict': {
 458             'id': 'local_playlist-4e760566473c4c8c5344',
 459             'ext': 'mp4',
 460             'title': 'Article_mtv-movies-spotlight-pixels-teil-3_short-clips_part1',
 461             'description': 'MTV Movies Supercut',
 462         },
 463         'params': {
 464             # rtmp download
 465             'skip_download': True,
 466         },
 467         'skip': 'Das Video kann zur Zeit nicht abgespielt werden.',
 468     }]
 469
 470     def _real_extract(self, url):
 471         video_id = self._match_id(url)
 472
 473         webpage = self._download_webpage(url, video_id)
 474
 475         playlist = self._parse_json(
 476             self._search_regex(
 477                 r'window\.pagePlaylist\s*=\s*(\[.+?\]);\n', webpage, 'page playlist'),
 478             video_id)
 479
 480         def _mrss_url(item):
 481             return item['mrss'] + item.get('mrssvars', '')
 482
 483         # news pages contain single video in playlist with different id
 484         if len(playlist) == 1:
 485             return self._get_videos_info_from_url(_mrss_url(playlist[0]), video_id)
 486
 487         for item in playlist:
 488             item_id = item.get('id')
 489             if item_id and compat_str(item_id) == video_id:
 490                 return self._get_videos_info_from_url(_mrss_url(item), video_id)