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     unescapeHTML,
  21     update_url_query,
  22     url_basename,
  23     xpath_text,
  24 )
  25
  26
  27 def _media_xml_tag(tag):
  28     return '{http://search.yahoo.com/mrss/}%s' % tag
  29
  30
  31 class MTVServicesInfoExtractor(InfoExtractor):
  32     _MOBILE_TEMPLATE = None
  33     _LANG = None
  34
  35     @staticmethod
  36     def _id_from_uri(uri):
  37         return uri.split(':')[-1]
  38
  39     @staticmethod
  40     def _remove_template_parameter(url):
  41         # Remove the templates, like &device={device}
  42         return re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', url)
  43
  44     # This was originally implemented for ComedyCentral, but it also works here
  45     @classmethod
  46     def _transform_rtmp_url(cls, rtmp_video_url):
  47         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  48         if not m:
  49             return {'rtmp': rtmp_video_url}
  50         base = 'http://viacommtvstrmfs.fplive.net/'
  51         return {'http': base + m.group('finalid')}
  52
  53     def _get_feed_url(self, uri):
  54         return self._FEED_URL
  55
  56     def _get_thumbnail_url(self, uri, itemdoc):
  57         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  58         thumb_node = itemdoc.find(search_path)
  59         if thumb_node is None:
  60             return None
  61         else:
  62             return thumb_node.attrib['url']
  63
  64     def _extract_mobile_video_formats(self, mtvn_id):
  65         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
  66         req = sanitized_Request(webpage_url)
  67         # Otherwise we get a webpage that would execute some javascript
  68         req.add_header('User-Agent', 'curl/7')
  69         webpage = self._download_webpage(req, mtvn_id,
  70                                          'Downloading mobile page')
  71         metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
  72         req = HEADRequest(metrics_url)
  73         response = self._request_webpage(req, mtvn_id, 'Resolving url')
  74         url = response.geturl()
  75         # Transform the url to get the best quality:
  76         url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
  77         return [{'url': url, 'ext': 'mp4'}]
  78
  79     def _extract_video_formats(self, mdoc, mtvn_id):
  80         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4|copyright_error\.flv(?:\?geo\b.+?)?)$', mdoc.find('.//src').text) is not None:
  81             if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
  82                 self.to_screen('The normal version is not available from your '
  83                                'country, trying with the mobile version')
  84                 return self._extract_mobile_video_formats(mtvn_id)
  85             raise ExtractorError('This video is not available from your country.',
  86                                  expected=True)
  87
  88         formats = []
  89         for rendition in mdoc.findall('.//rendition'):
  90             try:
  91                 _, _, ext = rendition.attrib['type'].partition('/')
  92                 rtmp_video_url = rendition.find('./src').text
  93                 if rtmp_video_url.endswith('siteunavail.png'):
  94                     continue
  95                 new_urls = self._transform_rtmp_url(rtmp_video_url)
  96                 formats.extend([{
  97                     'ext': 'flv' if new_url.startswith('rtmp') else ext,
  98                     'url': new_url,
  99                     'format_id': '-'.join(filter(None, [kind, rendition.get('bitrate')])),
 100                     'width': int(rendition.get('width')),
 101                     'height': int(rendition.get('height')),
 102                 } for kind, new_url in new_urls.items()])
 103             except (KeyError, TypeError):
 104                 raise ExtractorError('Invalid rendition field.')
 105         self._sort_formats(formats)
 106         return formats
 107
 108     def _extract_subtitles(self, mdoc, mtvn_id):
 109         subtitles = {}
 110         for transcript in mdoc.findall('.//transcript'):
 111             if transcript.get('kind') != 'captions':
 112                 continue
 113             lang = transcript.get('srclang')
 114             subtitles[lang] = [{
 115                 'url': compat_str(typographic.get('src')),
 116                 'ext': typographic.get('format')
 117             } for typographic in transcript.findall('./typographic')]
 118         return subtitles
 119
 120     def _get_video_info(self, itemdoc):
 121         uri = itemdoc.find('guid').text
 122         video_id = self._id_from_uri(uri)
 123         self.report_extraction(video_id)
 124         content_el = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content')))
 125         mediagen_url = self._remove_template_parameter(content_el.attrib['url'])
 126         if 'acceptMethods' not in mediagen_url:
 127             mediagen_url += '&' if '?' in mediagen_url else '?'
 128             mediagen_url += 'acceptMethods=fms'
 129
 130         mediagen_doc = self._download_xml(mediagen_url, video_id,
 131                                           'Downloading video urls')
 132
 133         item = mediagen_doc.find('./video/item')
 134         if item is not None and item.get('type') == 'text':
 135             message = '%s returned error: ' % self.IE_NAME
 136             if item.get('code') is not None:
 137                 message += '%s - ' % item.get('code')
 138             message += item.text
 139             raise ExtractorError(message, expected=True)
 140
 141         description = strip_or_none(xpath_text(itemdoc, 'description'))
 142
 143         timestamp = timeconvert(xpath_text(itemdoc, 'pubDate'))
 144
 145         title_el = None
 146         if title_el is None:
 147             title_el = find_xpath_attr(
 148                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 149                 'scheme', 'urn:mtvn:video_title')
 150         if title_el is None:
 151             title_el = itemdoc.find(compat_xpath('.//{http://search.yahoo.com/mrss/}title'))
 152         if title_el is None:
 153             title_el = itemdoc.find(compat_xpath('.//title'))
 154             if title_el.text is None:
 155                 title_el = None
 156
 157         title = title_el.text
 158         if title is None:
 159             raise ExtractorError('Could not find video title')
 160         title = title.strip()
 161
 162         # This a short id that's used in the webpage urls
 163         mtvn_id = None
 164         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 165                                        'scheme', 'urn:mtvn:id')
 166         if mtvn_id_node is not None:
 167             mtvn_id = mtvn_id_node.text
 168
 169         return {
 170             'title': title,
 171             'formats': self._extract_video_formats(mediagen_doc, mtvn_id),
 172             'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id),
 173             'id': video_id,
 174             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 175             'description': description,
 176             'duration': float_or_none(content_el.attrib.get('duration')),
 177             'timestamp': timestamp,
 178         }
 179
 180     def _get_feed_query(self, uri):
 181         data = {'uri': uri}
 182         if self._LANG:
 183             data['lang'] = self._LANG
 184         return data
 185
 186     def _get_videos_info(self, uri):
 187         video_id = self._id_from_uri(uri)
 188         feed_url = self._get_feed_url(uri)
 189         info_url = update_url_query(feed_url, self._get_feed_query(uri))
 190         return self._get_videos_info_from_url(info_url, video_id)
 191
 192     def _get_videos_info_from_url(self, url, video_id):
 193         idoc = self._download_xml(
 194             url, video_id,
 195             'Downloading info', transform_source=fix_xml_ampersands)
 196
 197         title = xpath_text(idoc, './channel/title')
 198         description = xpath_text(idoc, './channel/description')
 199
 200         return self.playlist_result(
 201             [self._get_video_info(item) for item in idoc.findall('.//item')],
 202             playlist_title=title, playlist_description=description)
 203
 204     def _extract_mgid(self, webpage):
 205         try:
 206             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 207             # or http://media.mtvnservices.com/{mgid}
 208             og_url = self._og_search_video_url(webpage)
 209             mgid = url_basename(og_url)
 210             if mgid.endswith('.swf'):
 211                 mgid = mgid[:-4]
 212         except RegexNotFoundError:
 213             mgid = None
 214
 215         if mgid is None or ':' not in mgid:
 216             mgid = self._search_regex(
 217                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
 218                 webpage, 'mgid', default=None)
 219
 220         if not mgid:
 221             sm4_embed = self._html_search_meta(
 222                 'sm4:video:embed', webpage, 'sm4 embed', default='')
 223             mgid = self._search_regex(
 224                 r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid')
 225         return mgid
 226
 227     def _real_extract(self, url):
 228         title = url_basename(url)
 229         webpage = self._download_webpage(url, title)
 230         mgid = self._extract_mgid(webpage)
 231         videos_info = self._get_videos_info(mgid)
 232         return videos_info
 233
 234
 235 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
 236     IE_NAME = 'mtvservices:embedded'
 237     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
 238
 239     _TEST = {
 240         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
 241         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
 242         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
 243         'info_dict': {
 244             'id': '1043906',
 245             'ext': 'mp4',
 246             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
 247             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
 248             'timestamp': 1400126400,
 249             'upload_date': '20140515',
 250         },
 251     }
 252
 253     @staticmethod
 254     def _extract_url(webpage):
 255         mobj = re.search(
 256             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//media.mtvnservices.com/embed/.+?)\1', webpage)
 257         if mobj:
 258             return mobj.group('url')
 259
 260     def _get_feed_url(self, uri):
 261         video_id = self._id_from_uri(uri)
 262         config = self._download_json(
 263             'http://media.mtvnservices.com/pmt/e1/access/index.html?uri=%s&configtype=edge' % uri, video_id)
 264         return self._remove_template_parameter(config['feedWithQueryParams'])
 265
 266     def _real_extract(self, url):
 267         mobj = re.match(self._VALID_URL, url)
 268         mgid = mobj.group('mgid')
 269         return self._get_videos_info(mgid)
 270
 271
 272 class MTVIE(MTVServicesInfoExtractor):
 273     IE_NAME = 'mtv'
 274     _VALID_URL = r'https?://(?:www\.)?mtv\.com/(?:video-clips|full-episodes)/(?P<id>[^/?#.]+)'
 275     _FEED_URL = 'http://www.mtv.com/feeds/mrss/'
 276
 277     _TESTS = [{
 278         'url': 'http://www.mtv.com/video-clips/vl8qof/unlocking-the-truth-trailer',
 279         'md5': '1edbcdf1e7628e414a8c5dcebca3d32b',
 280         'info_dict': {
 281             'id': '5e14040d-18a4-47c4-a582-43ff602de88e',
 282             'ext': 'mp4',
 283             'title': 'Unlocking The Truth|July 18, 2016|1|101|Trailer',
 284             'description': '"Unlocking the Truth" premieres August 17th at 11/10c.',
 285             'timestamp': 1468846800,
 286             'upload_date': '20160718',
 287         },
 288     }, {
 289         'url': 'http://www.mtv.com/full-episodes/94tujl/unlocking-the-truth-gates-of-hell-season-1-ep-101',
 290         'only_matching': True,
 291     }]
 292
 293
 294 class MTVVideoIE(MTVServicesInfoExtractor):
 295     IE_NAME = 'mtv:video'
 296     _VALID_URL = r'''(?x)^https?://
 297         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 298            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 299
 300     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 301
 302     _TESTS = [
 303         {
 304             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 305             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 306             'info_dict': {
 307                 'id': '853555',
 308                 'ext': 'mp4',
 309                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 310                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 311                 'timestamp': 1352610000,
 312                 'upload_date': '20121111',
 313             },
 314         },
 315     ]
 316
 317     def _get_thumbnail_url(self, uri, itemdoc):
 318         return 'http://mtv.mtvnimages.com/uri/' + uri
 319
 320     def _real_extract(self, url):
 321         mobj = re.match(self._VALID_URL, url)
 322         video_id = mobj.group('videoid')
 323         uri = mobj.groupdict().get('mgid')
 324         if uri is None:
 325             webpage = self._download_webpage(url, video_id)
 326
 327             # Some videos come from Vevo.com
 328             m_vevo = re.search(
 329                 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
 330             if m_vevo:
 331                 vevo_id = m_vevo.group(1)
 332                 self.to_screen('Vevo video detected: %s' % vevo_id)
 333                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 334
 335             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 336         return self._get_videos_info(uri)
 337
 338
 339 class MTVDEIE(MTVServicesInfoExtractor):
 340     IE_NAME = 'mtv.de'
 341     _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:artists|shows|news)/(?:[^/]+/)*(?P<id>\d+)-[^/#?]+/*(?:[#?].*)?$'
 342     _TESTS = [{
 343         'url': 'http://www.mtv.de/artists/10571-cro/videos/61131-traum',
 344         'info_dict': {
 345             'id': 'music_video-a50bc5f0b3aa4b3190aa',
 346             'ext': 'flv',
 347             'title': 'MusicVideo_cro-traum',
 348             'description': 'Cro - Traum',
 349         },
 350         'params': {
 351             # rtmp download
 352             'skip_download': True,
 353         },
 354         'skip': 'Blocked at Travis CI',
 355     }, {
 356         # mediagen URL without query (e.g. http://videos.mtvnn.com/mediagen/e865da714c166d18d6f80893195fcb97)
 357         'url': 'http://www.mtv.de/shows/933-teen-mom-2/staffeln/5353/folgen/63565-enthullungen',
 358         'info_dict': {
 359             'id': 'local_playlist-f5ae778b9832cc837189',
 360             'ext': 'flv',
 361             'title': 'Episode_teen-mom-2_shows_season-5_episode-1_full-episode_part1',
 362         },
 363         'params': {
 364             # rtmp download
 365             'skip_download': True,
 366         },
 367         'skip': 'Blocked at Travis CI',
 368     }, {
 369         'url': 'http://www.mtv.de/news/77491-mtv-movies-spotlight-pixels-teil-3',
 370         'info_dict': {
 371             'id': 'local_playlist-4e760566473c4c8c5344',
 372             'ext': 'mp4',
 373             'title': 'Article_mtv-movies-spotlight-pixels-teil-3_short-clips_part1',
 374             'description': 'MTV Movies Supercut',
 375         },
 376         'params': {
 377             # rtmp download
 378             'skip_download': True,
 379         },
 380         'skip': 'Das Video kann zur Zeit nicht abgespielt werden.',
 381     }]
 382
 383     def _real_extract(self, url):
 384         video_id = self._match_id(url)
 385
 386         webpage = self._download_webpage(url, video_id)
 387
 388         playlist = self._parse_json(
 389             self._search_regex(
 390                 r'window\.pagePlaylist\s*=\s*(\[.+?\]);\n', webpage, 'page playlist'),
 391             video_id)
 392
 393         def _mrss_url(item):
 394             return item['mrss'] + item.get('mrssvars', '')
 395
 396         # news pages contain single video in playlist with different id
 397         if len(playlist) == 1:
 398             return self._get_videos_info_from_url(_mrss_url(playlist[0]), video_id)
 399
 400         for item in playlist:
 401             item_id = item.get('id')
 402             if item_id and compat_str(item_id) == video_id:
 403                 return self._get_videos_info_from_url(_mrss_url(item), video_id)