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