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