youtube_dlc/extractor/mtv.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..compat import (
   8     compat_str,
   9     compat_xpath,
  10     compat_urlparse,
  11 )
  12 from ..utils import (
  13     ExtractorError,
  14     find_xpath_attr,
  15     fix_xml_ampersands,
  16     float_or_none,
  17     HEADRequest,
  18     RegexNotFoundError,
  19     sanitized_Request,
  20     strip_or_none,
  21     timeconvert,
  22     try_get,
  23     unescapeHTML,
  24     update_url_query,
  25     url_basename,
  26     get_domain,
  27     xpath_text,
  28 )
  29
  30
  31 def _media_xml_tag(tag):
  32     return '{http://search.yahoo.com/mrss/}%s' % tag
  33
  34
  35 class MTVServicesInfoExtractor(InfoExtractor):
  36     _MOBILE_TEMPLATE = None
  37     _LANG = None
  38
  39     @staticmethod
  40     def _id_from_uri(uri):
  41         return uri.split(':')[-1]
  42
  43     @staticmethod
  44     def _remove_template_parameter(url):
  45         # Remove the templates, like &device={device}
  46         return re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', url)
  47
  48     def _get_feed_url(self, uri):
  49         return self._FEED_URL
  50
  51     def _get_thumbnail_url(self, uri, itemdoc):
  52         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  53         thumb_node = itemdoc.find(search_path)
  54         if thumb_node is None:
  55             return None
  56         return thumb_node.get('url') or thumb_node.text or None
  57
  58     def _extract_mobile_video_formats(self, mtvn_id):
  59         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
  60         req = sanitized_Request(webpage_url)
  61         # Otherwise we get a webpage that would execute some javascript
  62         req.add_header('User-Agent', 'curl/7')
  63         webpage = self._download_webpage(req, mtvn_id,
  64                                          'Downloading mobile page')
  65         metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
  66         req = HEADRequest(metrics_url)
  67         response = self._request_webpage(req, mtvn_id, 'Resolving url')
  68         url = response.geturl()
  69         # Transform the url to get the best quality:
  70         url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
  71         return [{'url': url, 'ext': 'mp4'}]
  72
  73     def _extract_video_formats(self, mdoc, mtvn_id, video_id):
  74         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4|copyright_error\.flv(?:\?geo\b.+?)?)$', mdoc.find('.//src').text) is not None:
  75             if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
  76                 self.to_screen('The normal version is not available from your '
  77                                'country, trying with the mobile version')
  78                 return self._extract_mobile_video_formats(mtvn_id)
  79             raise ExtractorError('This video is not available from your country.',
  80                                  expected=True)
  81
  82         formats = []
  83         for rendition in mdoc.findall('.//rendition'):
  84             if rendition.get('method') == 'hls':
  85                 hls_url = rendition.find('./src').text
  86                 formats.extend(self._extract_m3u8_formats(
  87                     hls_url, video_id, ext='mp4', entry_protocol='m3u8_native',
  88                     m3u8_id='hls', fatal=False))
  89             else:
  90                 # fms
  91                 try:
  92                     _, _, ext = rendition.attrib['type'].partition('/')
  93                     rtmp_video_url = rendition.find('./src').text
  94                     if 'error_not_available.swf' in rtmp_video_url:
  95                         raise ExtractorError(
  96                             '%s said: video is not available' % self.IE_NAME,
  97                             expected=True)
  98                     if rtmp_video_url.endswith('siteunavail.png'):
  99                         continue
 100                     formats.extend([{
 101                         'ext': 'flv' if rtmp_video_url.startswith('rtmp') else ext,
 102                         'url': rtmp_video_url,
 103                         'format_id': '-'.join(filter(None, [
 104                             'rtmp' if rtmp_video_url.startswith('rtmp') else None,
 105                             rendition.get('bitrate')])),
 106                         'width': int(rendition.get('width')),
 107                         'height': int(rendition.get('height')),
 108                     }])
 109                 except (KeyError, TypeError):
 110                     raise ExtractorError('Invalid rendition field.')
 111         if formats:
 112             self._sort_formats(formats)
 113         return formats
 114
 115     def _extract_subtitles(self, mdoc, mtvn_id):
 116         subtitles = {}
 117         for transcript in mdoc.findall('.//transcript'):
 118             if transcript.get('kind') != 'captions':
 119                 continue
 120             lang = transcript.get('srclang')
 121             for typographic in transcript.findall('./typographic'):
 122                 sub_src = typographic.get('src')
 123                 if not sub_src:
 124                     continue
 125                 ext = typographic.get('format')
 126                 if ext == 'cea-608':
 127                     ext = 'scc'
 128                 subtitles.setdefault(lang, []).append({
 129                     'url': compat_str(sub_src),
 130                     'ext': ext
 131                 })
 132         return subtitles
 133
 134     def _get_video_info(self, itemdoc, use_hls=True):
 135         uri = itemdoc.find('guid').text
 136         video_id = self._id_from_uri(uri)
 137         self.report_extraction(video_id)
 138         content_el = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content')))
 139         mediagen_url = self._remove_template_parameter(content_el.attrib['url'])
 140         mediagen_url = mediagen_url.replace('device={device}', '')
 141         if 'acceptMethods' not in mediagen_url:
 142             mediagen_url += '&' if '?' in mediagen_url else '?'
 143             mediagen_url += 'acceptMethods='
 144             mediagen_url += 'hls' if use_hls else 'fms'
 145
 146         mediagen_doc = self._download_xml(
 147             mediagen_url, video_id, 'Downloading video urls', fatal=False)
 148
 149         if mediagen_doc is False:
 150             return None
 151
 152         item = mediagen_doc.find('./video/item')
 153         if item is not None and item.get('type') == 'text':
 154             message = '%s returned error: ' % self.IE_NAME
 155             if item.get('code') is not None:
 156                 message += '%s - ' % item.get('code')
 157             message += item.text
 158             raise ExtractorError(message, expected=True)
 159
 160         description = strip_or_none(xpath_text(itemdoc, 'description'))
 161
 162         timestamp = timeconvert(xpath_text(itemdoc, 'pubDate'))
 163
 164         title_el = None
 165         if title_el is None:
 166             title_el = find_xpath_attr(
 167                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 168                 'scheme', 'urn:mtvn:video_title')
 169         if title_el is None:
 170             title_el = itemdoc.find(compat_xpath('.//{http://search.yahoo.com/mrss/}title'))
 171         if title_el is None:
 172             title_el = itemdoc.find(compat_xpath('.//title'))
 173             if title_el.text is None:
 174                 title_el = None
 175
 176         title = title_el.text
 177         if title is None:
 178             raise ExtractorError('Could not find video title')
 179         title = title.strip()
 180
 181         # This a short id that's used in the webpage urls
 182         mtvn_id = None
 183         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 184                                        'scheme', 'urn:mtvn:id')
 185         if mtvn_id_node is not None:
 186             mtvn_id = mtvn_id_node.text
 187
 188         formats = self._extract_video_formats(mediagen_doc, mtvn_id, video_id)
 189
 190         # Some parts of complete video may be missing (e.g. missing Act 3 in
 191         # http://www.southpark.de/alle-episoden/s14e01-sexual-healing)
 192         if not formats:
 193             return None
 194
 195         self._sort_formats(formats)
 196
 197         return {
 198             'title': title,
 199             'formats': formats,
 200             'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id),
 201             'id': video_id,
 202             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 203             'description': description,
 204             'duration': float_or_none(content_el.attrib.get('duration')),
 205             'timestamp': timestamp,
 206         }
 207
 208     def _get_feed_query(self, uri):
 209         data = {'uri': uri}
 210         if self._LANG:
 211             data['lang'] = self._LANG
 212         return data
 213
 214     def _get_videos_info(self, uri, use_hls=True):
 215         video_id = self._id_from_uri(uri)
 216         feed_url = self._get_feed_url(uri)
 217         info_url = update_url_query(feed_url, self._get_feed_query(uri))
 218         return self._get_videos_info_from_url(info_url, video_id, use_hls)
 219
 220     def _get_videos_info_from_url(self, url, video_id, use_hls=True):
 221         idoc = self._download_xml(
 222             url, video_id,
 223             'Downloading info', transform_source=fix_xml_ampersands)
 224
 225         title = xpath_text(idoc, './channel/title')
 226         description = xpath_text(idoc, './channel/description')
 227
 228         entries = []
 229         for item in idoc.findall('.//item'):
 230             info = self._get_video_info(item, use_hls)
 231             if info:
 232                 entries.append(info)
 233
 234         return self.playlist_result(
 235             entries, playlist_title=title, playlist_description=description)
 236
 237     def _extract_triforce_mgid(self, webpage, data_zone=None, video_id=None):
 238         triforce_feed = self._parse_json(self._search_regex(
 239             r'triforceManifestFeed\s*=\s*({.+?})\s*;\s*\n', webpage,
 240             'triforce feed', default='{}'), video_id, fatal=False)
 241
 242         data_zone = self._search_regex(
 243             r'data-zone=(["\'])(?P<zone>.+?_lc_promo.*?)\1', webpage,
 244             'data zone', default=data_zone, group='zone')
 245
 246         feed_url = try_get(
 247             triforce_feed, lambda x: x['manifest']['zones'][data_zone]['feed'],
 248             compat_str)
 249         if not feed_url:
 250             return
 251
 252         feed = self._download_json(feed_url, video_id, fatal=False)
 253         if not feed:
 254             return
 255
 256         return try_get(feed, lambda x: x['result']['data']['id'], compat_str)
 257
 258     def _extract_new_triforce_mgid(self, webpage, url='', data_zone=None, video_id=None):
 259         # print(compat_urlparse.urlparse(url).netloc)
 260         domain = get_domain(url)
 261         if domain is None:
 262             raise ExtractorError(
 263                 '[%s] could not get domain' % self.IE_NAME,
 264                 expected=True)
 265         url = url.replace("https://", "http://")
 266         enc_url = compat_urlparse.quote(url, safe='')
 267         _TRIFORCE_V8_TEMPLATE = 'https://%s/feeds/triforce/manifest/v8?url=%s'
 268         triforce_manifest_url = _TRIFORCE_V8_TEMPLATE % (domain, enc_url)
 269
 270         manifest = self._download_json(triforce_manifest_url, video_id, fatal=False)
 271         if manifest.get('manifest').get('type') == 'redirect':
 272             self.to_screen('Found a redirect. Downloading manifest from new location')
 273             new_loc = manifest.get('manifest').get('newLocation')
 274             new_loc = new_loc.replace("https://", "http://")
 275             enc_new_loc = compat_urlparse.quote(new_loc, safe='')
 276             triforce_manifest_new_loc = _TRIFORCE_V8_TEMPLATE % (domain, enc_new_loc)
 277             manifest = self._download_json(triforce_manifest_new_loc, video_id, fatal=False)
 278
 279         item_id = try_get(manifest, lambda x: x['manifest']['reporting']['itemId'], compat_str)
 280         if not item_id:
 281             self.to_screen('Found no id!')
 282             return
 283
 284         # 'episode' can be anything. 'content' is used often as well
 285         _MGID_TEMPLATE = 'mgid:arc:episode:%s:%s'
 286         mgid = _MGID_TEMPLATE % (domain, item_id)
 287
 288         return mgid
 289
 290     def _extract_mgid(self, webpage, url):
 291         try:
 292             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 293             # or http://media.mtvnservices.com/{mgid}
 294             og_url = self._og_search_video_url(webpage)
 295             mgid = url_basename(og_url)
 296             if mgid.endswith('.swf'):
 297                 mgid = mgid[:-4]
 298         except RegexNotFoundError:
 299             mgid = None
 300
 301         if mgid is None or ':' not in mgid:
 302             mgid = self._search_regex(
 303                 [r'data-mgid="(.*?)"', r'swfobject\.embedSWF\(".*?(mgid:.*?)"'],
 304                 webpage, 'mgid', default=None)
 305
 306         if not mgid:
 307             sm4_embed = self._html_search_meta(
 308                 'sm4:video:embed', webpage, 'sm4 embed', default='')
 309             mgid = self._search_regex(
 310                 r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid', default=None)
 311
 312         if not mgid:
 313             mgid = self._extract_new_triforce_mgid(webpage, url)
 314
 315         if not mgid:
 316             mgid = self._extract_triforce_mgid(webpage)
 317
 318         return mgid
 319
 320     def _real_extract(self, url):
 321         title = url_basename(url)
 322         webpage = self._download_webpage(url, title)
 323         mgid = self._extract_mgid(webpage, url)
 324         videos_info = self._get_videos_info(mgid)
 325         return videos_info
 326
 327
 328 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
 329     IE_NAME = 'mtvservices:embedded'
 330     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
 331
 332     _TEST = {
 333         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
 334         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
 335         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
 336         'info_dict': {
 337             'id': '1043906',
 338             'ext': 'mp4',
 339             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
 340             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
 341             'timestamp': 1400126400,
 342             'upload_date': '20140515',
 343         },
 344     }
 345
 346     @staticmethod
 347     def _extract_url(webpage):
 348         mobj = re.search(
 349             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//media.mtvnservices.com/embed/.+?)\1', webpage)
 350         if mobj:
 351             return mobj.group('url')
 352
 353     def _get_feed_url(self, uri):
 354         video_id = self._id_from_uri(uri)
 355         config = self._download_json(
 356             'http://media.mtvnservices.com/pmt/e1/access/index.html?uri=%s&configtype=edge' % uri, video_id)
 357         return self._remove_template_parameter(config['feedWithQueryParams'])
 358
 359     def _real_extract(self, url):
 360         mobj = re.match(self._VALID_URL, url)
 361         mgid = mobj.group('mgid')
 362         return self._get_videos_info(mgid)
 363
 364
 365 class MTVIE(MTVServicesInfoExtractor):
 366     IE_NAME = 'mtv'
 367     _VALID_URL = r'https?://(?:www\.)?mtv\.com/(?:video-clips|(?:full-)?episodes)/(?P<id>[^/?#.]+)'
 368     _FEED_URL = 'http://www.mtv.com/feeds/mrss/'
 369
 370     _TESTS = [{
 371         'url': 'http://www.mtv.com/video-clips/vl8qof/unlocking-the-truth-trailer',
 372         'md5': '1edbcdf1e7628e414a8c5dcebca3d32b',
 373         'info_dict': {
 374             'id': '5e14040d-18a4-47c4-a582-43ff602de88e',
 375             'ext': 'mp4',
 376             'title': 'Unlocking The Truth|July 18, 2016|1|101|Trailer',
 377             'description': '"Unlocking the Truth" premieres August 17th at 11/10c.',
 378             'timestamp': 1468846800,
 379             'upload_date': '20160718',
 380         },
 381     }, {
 382         'url': 'http://www.mtv.com/full-episodes/94tujl/unlocking-the-truth-gates-of-hell-season-1-ep-101',
 383         'only_matching': True,
 384     }, {
 385         'url': 'http://www.mtv.com/episodes/g8xu7q/teen-mom-2-breaking-the-wall-season-7-ep-713',
 386         'only_matching': True,
 387     }]
 388
 389
 390 class MTVJapanIE(MTVServicesInfoExtractor):
 391     IE_NAME = 'mtvjapan'
 392     _VALID_URL = r'https?://(?:www\.)?mtvjapan\.com/videos/(?P<id>[0-9a-z]+)'
 393
 394     _TEST = {
 395         'url': 'http://www.mtvjapan.com/videos/prayht/fresh-info-cadillac-escalade',
 396         'info_dict': {
 397             'id': 'bc01da03-6fe5-4284-8880-f291f4e368f5',
 398             'ext': 'mp4',
 399             'title': '【Fresh Info】Cadillac ESCALADE Sport Edition',
 400         },
 401         'params': {
 402             'skip_download': True,
 403         },
 404     }
 405     _GEO_COUNTRIES = ['JP']
 406     _FEED_URL = 'http://feeds.mtvnservices.com/od/feed/intl-mrss-player-feed'
 407
 408     def _get_feed_query(self, uri):
 409         return {
 410             'arcEp': 'mtvjapan.com',
 411             'mgid': uri,
 412         }
 413
 414
 415 class MTVVideoIE(MTVServicesInfoExtractor):
 416     IE_NAME = 'mtv:video'
 417     _VALID_URL = r'''(?x)^https?://
 418         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 419            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 420
 421     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 422
 423     _TESTS = [
 424         {
 425             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 426             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 427             'info_dict': {
 428                 'id': '853555',
 429                 'ext': 'mp4',
 430                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 431                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 432                 'timestamp': 1352610000,
 433                 'upload_date': '20121111',
 434             },
 435         },
 436     ]
 437
 438     def _get_thumbnail_url(self, uri, itemdoc):
 439         return 'http://mtv.mtvnimages.com/uri/' + uri
 440
 441     def _real_extract(self, url):
 442         mobj = re.match(self._VALID_URL, url)
 443         video_id = mobj.group('videoid')
 444         uri = mobj.groupdict().get('mgid')
 445         if uri is None:
 446             webpage = self._download_webpage(url, video_id)
 447
 448             # Some videos come from Vevo.com
 449             m_vevo = re.search(
 450                 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
 451             if m_vevo:
 452                 vevo_id = m_vevo.group(1)
 453                 self.to_screen('Vevo video detected: %s' % vevo_id)
 454                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 455
 456             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 457         return self._get_videos_info(uri)
 458
 459
 460 class MTVDEIE(MTVServicesInfoExtractor):
 461     IE_NAME = 'mtv.de'
 462     _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:musik/videoclips|folgen|news)/(?P<id>[0-9a-z]+)'
 463     _TESTS = [{
 464         'url': 'http://www.mtv.de/musik/videoclips/2gpnv7/Traum',
 465         'info_dict': {
 466             'id': 'd5d472bc-f5b7-11e5-bffd-a4badb20dab5',
 467             'ext': 'mp4',
 468             'title': 'Traum',
 469             'description': 'Traum',
 470         },
 471         'params': {
 472             # rtmp download
 473             'skip_download': True,
 474         },
 475         'skip': 'Blocked at Travis CI',
 476     }, {
 477         # mediagen URL without query (e.g. http://videos.mtvnn.com/mediagen/e865da714c166d18d6f80893195fcb97)
 478         'url': 'http://www.mtv.de/folgen/6b1ylu/teen-mom-2-enthuellungen-S5-F1',
 479         'info_dict': {
 480             'id': '1e5a878b-31c5-11e7-a442-0e40cf2fc285',
 481             'ext': 'mp4',
 482             'title': 'Teen Mom 2',
 483             'description': 'md5:dc65e357ef7e1085ed53e9e9d83146a7',
 484         },
 485         'params': {
 486             # rtmp download
 487             'skip_download': True,
 488         },
 489         'skip': 'Blocked at Travis CI',
 490     }, {
 491         'url': 'http://www.mtv.de/news/glolix/77491-mtv-movies-spotlight--pixels--teil-3',
 492         'info_dict': {
 493             'id': 'local_playlist-4e760566473c4c8c5344',
 494             'ext': 'mp4',
 495             'title': 'Article_mtv-movies-spotlight-pixels-teil-3_short-clips_part1',
 496             'description': 'MTV Movies Supercut',
 497         },
 498         'params': {
 499             # rtmp download
 500             'skip_download': True,
 501         },
 502         'skip': 'Das Video kann zur Zeit nicht abgespielt werden.',
 503     }]
 504     _GEO_COUNTRIES = ['DE']
 505     _FEED_URL = 'http://feeds.mtvnservices.com/od/feed/intl-mrss-player-feed'
 506
 507     def _get_feed_query(self, uri):
 508         return {
 509             'arcEp': 'mtv.de',
 510             'mgid': uri,
 511         }