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'&[^=]*?={.*?}(?=(&|$))', '', url)
  47
  48     def _get_feed_url(self, uri, url=None):
  49         return self._FEED_URL
  50
  51     def _get_thumbnail_url(self, uri, itemdoc):
  52         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  53         thumb_node = itemdoc.find(search_path)
  54         if thumb_node is None:
  55             return None
  56         return thumb_node.get('url') or thumb_node.text or None
  57
  58     def _extract_mobile_video_formats(self, mtvn_id):
  59         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
  60         req = sanitized_Request(webpage_url)
  61         # Otherwise we get a webpage that would execute some javascript
  62         req.add_header('User-Agent', 'curl/7')
  63         webpage = self._download_webpage(req, mtvn_id,
  64                                          'Downloading mobile page')
  65         metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
  66         req = HEADRequest(metrics_url)
  67         response = self._request_webpage(req, mtvn_id, 'Resolving url')
  68         url = response.geturl()
  69         # Transform the url to get the best quality:
  70         url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
  71         return [{'url': url, 'ext': 'mp4'}]
  72
  73     def _extract_video_formats(self, mdoc, mtvn_id, video_id):
  74         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4|copyright_error\.flv(?:\?geo\b.+?)?)$', 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, url=None):
 215         video_id = self._id_from_uri(uri)
 216         feed_url = self._get_feed_url(uri, url)
 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='', video_id=None):
 259         if url == '':
 260             return
 261         domain = get_domain(url)
 262         if domain is None:
 263             raise ExtractorError(
 264                 '[%s] could not get domain' % self.IE_NAME,
 265                 expected=True)
 266         url = url.replace("https://", "http://")
 267         enc_url = compat_urlparse.quote(url, safe='')
 268         _TRIFORCE_V8_TEMPLATE = 'https://%s/feeds/triforce/manifest/v8?url=%s'
 269         triforce_manifest_url = _TRIFORCE_V8_TEMPLATE % (domain, enc_url)
 270
 271         manifest = self._download_json(triforce_manifest_url, video_id, fatal=False)
 272         if manifest:
 273             if manifest.get('manifest').get('type') == 'redirect':
 274                 self.to_screen('Found a redirect. Downloading manifest from new location')
 275                 new_loc = manifest.get('manifest').get('newLocation')
 276                 new_loc = new_loc.replace("https://", "http://")
 277                 enc_new_loc = compat_urlparse.quote(new_loc, safe='')
 278                 triforce_manifest_new_loc = _TRIFORCE_V8_TEMPLATE % (domain, enc_new_loc)
 279                 manifest = self._download_json(triforce_manifest_new_loc, video_id, fatal=False)
 280
 281         item_id = try_get(manifest, lambda x: x['manifest']['reporting']['itemId'], compat_str)
 282         if not item_id:
 283             self.to_screen('No id found!')
 284             return
 285
 286         # 'episode' can be anything. 'content' is used often as well
 287         _MGID_TEMPLATE = 'mgid:arc:episode:%s:%s'
 288         mgid = _MGID_TEMPLATE % (domain, item_id)
 289
 290         return mgid
 291
 292     def _extract_mgid(self, webpage, url, title=None, data_zone=None):
 293         try:
 294             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 295             # or http://media.mtvnservices.com/{mgid}
 296             og_url = self._og_search_video_url(webpage)
 297             mgid = url_basename(og_url)
 298             if mgid.endswith('.swf'):
 299                 mgid = mgid[:-4]
 300         except RegexNotFoundError:
 301             mgid = None
 302
 303         if not title:
 304             title = url_basename(url)
 305
 306         try:
 307             window_data = self._parse_json(self._search_regex(
 308                 r'(?s)window.__DATA__ = (?P<json>{.+});', webpage,
 309                 'JSON Window Data', default=None, fatal=False, group='json'), title, fatal=False)
 310             main_container = None
 311             for i in range(len(window_data['children'])):
 312                 if window_data['children'][i]['type'] == 'MainContainer':
 313                     main_container = window_data['children'][i]
 314             mgid = main_container['children'][0]['props']['media']['video']['config']['uri']
 315         except (KeyError, IndexError, TypeError):
 316             pass
 317
 318         if mgid is None or ':' not in mgid:
 319             mgid = self._search_regex(
 320                 [r'data-mgid="(.*?)"', r'swfobject\.embedSWF\(".*?(mgid:.*?)"'],
 321                 webpage, 'mgid', default=None)
 322
 323         if not mgid:
 324             sm4_embed = self._html_search_meta(
 325                 'sm4:video:embed', webpage, 'sm4 embed', default='')
 326             mgid = self._search_regex(
 327                 r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid', default=None)
 328
 329         if not mgid:
 330             mgid = self._extract_new_triforce_mgid(webpage, url)
 331
 332         if not mgid:
 333             mgid = self._extract_triforce_mgid(webpage, data_zone)
 334
 335         return mgid
 336
 337     def _real_extract(self, url):
 338         title = url_basename(url)
 339         webpage = self._download_webpage(url, title)
 340         mgid = self._extract_mgid(webpage, url, title=title)
 341         videos_info = self._get_videos_info(mgid, url=url)
 342         return videos_info
 343
 344
 345 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
 346     IE_NAME = 'mtvservices:embedded'
 347     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
 348
 349     _TEST = {
 350         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
 351         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
 352         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
 353         'info_dict': {
 354             'id': '1043906',
 355             'ext': 'mp4',
 356             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
 357             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
 358             'timestamp': 1400126400,
 359             'upload_date': '20140515',
 360         },
 361     }
 362
 363     @staticmethod
 364     def _extract_url(webpage):
 365         mobj = re.search(
 366             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//media.mtvnservices.com/embed/.+?)\1', webpage)
 367         if mobj:
 368             return mobj.group('url')
 369
 370     def _get_feed_url(self, uri):
 371         video_id = self._id_from_uri(uri)
 372         config = self._download_json(
 373             'http://media.mtvnservices.com/pmt/e1/access/index.html?uri=%s&configtype=edge' % uri, video_id)
 374         return self._remove_template_parameter(config['feedWithQueryParams'])
 375
 376     def _real_extract(self, url):
 377         mobj = re.match(self._VALID_URL, url)
 378         mgid = mobj.group('mgid')
 379         return self._get_videos_info(mgid)
 380
 381
 382 class MTVIE(MTVServicesInfoExtractor):
 383     IE_NAME = 'mtv'
 384     _VALID_URL = r'https?://(?:www\.)?mtv\.com/(?:video-clips|(?:full-)?episodes)/(?P<id>[^/?#.]+)'
 385     _FEED_URL = 'http://www.mtv.com/feeds/mrss/'
 386
 387     _TESTS = [{
 388         'url': 'http://www.mtv.com/video-clips/vl8qof/unlocking-the-truth-trailer',
 389         'md5': '1edbcdf1e7628e414a8c5dcebca3d32b',
 390         'info_dict': {
 391             'id': '5e14040d-18a4-47c4-a582-43ff602de88e',
 392             'ext': 'mp4',
 393             'title': 'Unlocking The Truth|July 18, 2016|1|101|Trailer',
 394             'description': '"Unlocking the Truth" premieres August 17th at 11/10c.',
 395             'timestamp': 1468846800,
 396             'upload_date': '20160718',
 397         },
 398     }, {
 399         'url': 'http://www.mtv.com/full-episodes/94tujl/unlocking-the-truth-gates-of-hell-season-1-ep-101',
 400         'only_matching': True,
 401     }, {
 402         'url': 'http://www.mtv.com/episodes/g8xu7q/teen-mom-2-breaking-the-wall-season-7-ep-713',
 403         'only_matching': True,
 404     }]
 405
 406     @staticmethod
 407     def extract_child_with_type(parent, t):
 408         children = parent['children']
 409         return next(c for c in children if c.get('type') == t)
 410
 411     def _extract_mgid(self, webpage):
 412         data = self._parse_json(self._search_regex(
 413             r'__DATA__\s*=\s*({.+?});', webpage, 'data'), None)
 414         main_container = self.extract_child_with_type(data, 'MainContainer')
 415         video_player = self.extract_child_with_type(main_container, 'VideoPlayer')
 416         return video_player['props']['media']['video']['config']['uri']
 417
 418
 419 class MTVJapanIE(MTVServicesInfoExtractor):
 420     IE_NAME = 'mtvjapan'
 421     _VALID_URL = r'https?://(?:www\.)?mtvjapan\.com/videos/(?P<id>[0-9a-z]+)'
 422
 423     _TEST = {
 424         'url': 'http://www.mtvjapan.com/videos/prayht/fresh-info-cadillac-escalade',
 425         'info_dict': {
 426             'id': 'bc01da03-6fe5-4284-8880-f291f4e368f5',
 427             'ext': 'mp4',
 428             'title': '【Fresh Info】Cadillac ESCALADE Sport Edition',
 429         },
 430         'params': {
 431             'skip_download': True,
 432         },
 433     }
 434     _GEO_COUNTRIES = ['JP']
 435     _FEED_URL = 'http://feeds.mtvnservices.com/od/feed/intl-mrss-player-feed'
 436
 437     def _get_feed_query(self, uri):
 438         return {
 439             'arcEp': 'mtvjapan.com',
 440             'mgid': uri,
 441         }
 442
 443
 444 class MTVVideoIE(MTVServicesInfoExtractor):
 445     IE_NAME = 'mtv:video'
 446     _VALID_URL = r'''(?x)^https?://
 447         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 448            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 449
 450     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 451
 452     _TESTS = [
 453         {
 454             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 455             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 456             'info_dict': {
 457                 'id': '853555',
 458                 'ext': 'mp4',
 459                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 460                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 461                 'timestamp': 1352610000,
 462                 'upload_date': '20121111',
 463             },
 464         },
 465     ]
 466
 467     def _get_thumbnail_url(self, uri, itemdoc):
 468         return 'http://mtv.mtvnimages.com/uri/' + uri
 469
 470     def _real_extract(self, url):
 471         mobj = re.match(self._VALID_URL, url)
 472         video_id = mobj.group('videoid')
 473         uri = mobj.groupdict().get('mgid')
 474         if uri is None:
 475             webpage = self._download_webpage(url, video_id)
 476
 477             # Some videos come from Vevo.com
 478             m_vevo = re.search(
 479                 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
 480             if m_vevo:
 481                 vevo_id = m_vevo.group(1)
 482                 self.to_screen('Vevo video detected: %s' % vevo_id)
 483                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 484
 485             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 486         return self._get_videos_info(uri)
 487
 488
 489 class MTVDEIE(MTVServicesInfoExtractor):
 490     IE_NAME = 'mtv.de'
 491     _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:musik/videoclips|folgen|news)/(?P<id>[0-9a-z]+)'
 492     _TESTS = [{
 493         'url': 'http://www.mtv.de/musik/videoclips/2gpnv7/Traum',
 494         'info_dict': {
 495             'id': 'd5d472bc-f5b7-11e5-bffd-a4badb20dab5',
 496             'ext': 'mp4',
 497             'title': 'Traum',
 498             'description': 'Traum',
 499         },
 500         'params': {
 501             # rtmp download
 502             'skip_download': True,
 503         },
 504         'skip': 'Blocked at Travis CI',
 505     }, {
 506         # mediagen URL without query (e.g. http://videos.mtvnn.com/mediagen/e865da714c166d18d6f80893195fcb97)
 507         'url': 'http://www.mtv.de/folgen/6b1ylu/teen-mom-2-enthuellungen-S5-F1',
 508         'info_dict': {
 509             'id': '1e5a878b-31c5-11e7-a442-0e40cf2fc285',
 510             'ext': 'mp4',
 511             'title': 'Teen Mom 2',
 512             'description': 'md5:dc65e357ef7e1085ed53e9e9d83146a7',
 513         },
 514         'params': {
 515             # rtmp download
 516             'skip_download': True,
 517         },
 518         'skip': 'Blocked at Travis CI',
 519     }, {
 520         'url': 'http://www.mtv.de/news/glolix/77491-mtv-movies-spotlight--pixels--teil-3',
 521         'info_dict': {
 522             'id': 'local_playlist-4e760566473c4c8c5344',
 523             'ext': 'mp4',
 524             'title': 'Article_mtv-movies-spotlight-pixels-teil-3_short-clips_part1',
 525             'description': 'MTV Movies Supercut',
 526         },
 527         'params': {
 528             # rtmp download
 529             'skip_download': True,
 530         },
 531         'skip': 'Das Video kann zur Zeit nicht abgespielt werden.',
 532     }]
 533     _GEO_COUNTRIES = ['DE']
 534     _FEED_URL = 'http://feeds.mtvnservices.com/od/feed/intl-mrss-player-feed'
 535
 536     def _get_feed_query(self, uri):
 537         return {
 538             'arcEp': 'mtv.de',
 539             'mgid': uri,
 540         }