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_urllib_parse_urlencode,
   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     sanitized_Request,
  18     strip_or_none,
  19     timeconvert,
  20     unescapeHTML,
  21     url_basename,
  22     RegexNotFoundError,
  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     # This was originally implemented for ComedyCentral, but it also works here
  40     @classmethod
  41     def _transform_rtmp_url(cls, rtmp_video_url):
  42         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  43         if not m:
  44             return {'rtmp': rtmp_video_url}
  45         base = 'http://viacommtvstrmfs.fplive.net/'
  46         return {'http': base + m.group('finalid')}
  47
  48     def _get_feed_url(self, uri):
  49         return self._FEED_URL
  50
  51     def _get_thumbnail_url(self, uri, itemdoc):
  52         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  53         thumb_node = itemdoc.find(search_path)
  54         if thumb_node is None:
  55             return None
  56         else:
  57             return thumb_node.attrib['url']
  58
  59     def _extract_mobile_video_formats(self, mtvn_id):
  60         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
  61         req = sanitized_Request(webpage_url)
  62         # Otherwise we get a webpage that would execute some javascript
  63         req.add_header('User-Agent', 'curl/7')
  64         webpage = self._download_webpage(req, mtvn_id,
  65                                          'Downloading mobile page')
  66         metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
  67         req = HEADRequest(metrics_url)
  68         response = self._request_webpage(req, mtvn_id, 'Resolving url')
  69         url = response.geturl()
  70         # Transform the url to get the best quality:
  71         url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
  72         return [{'url': url, 'ext': 'mp4'}]
  73
  74     def _extract_video_formats(self, mdoc, mtvn_id):
  75         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4|copyright_error\.flv(?:\?geo\b.+?)?)$', mdoc.find('.//src').text) is not None:
  76             if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
  77                 self.to_screen('The normal version is not available from your '
  78                                'country, trying with the mobile version')
  79                 return self._extract_mobile_video_formats(mtvn_id)
  80             raise ExtractorError('This video is not available from your country.',
  81                                  expected=True)
  82
  83         formats = []
  84         for rendition in mdoc.findall('.//rendition'):
  85             try:
  86                 _, _, ext = rendition.attrib['type'].partition('/')
  87                 rtmp_video_url = rendition.find('./src').text
  88                 if rtmp_video_url.endswith('siteunavail.png'):
  89                     continue
  90                 new_urls = self._transform_rtmp_url(rtmp_video_url)
  91                 formats.extend([{
  92                     'ext': 'flv' if new_url.startswith('rtmp') else ext,
  93                     'url': new_url,
  94                     'format_id': '-'.join(filter(None, [kind, rendition.get('bitrate')])),
  95                     'width': int(rendition.get('width')),
  96                     'height': int(rendition.get('height')),
  97                 } for kind, new_url in new_urls.items()])
  98             except (KeyError, TypeError):
  99                 raise ExtractorError('Invalid rendition field.')
 100         self._sort_formats(formats)
 101         return formats
 102
 103     def _extract_subtitles(self, mdoc, mtvn_id):
 104         subtitles = {}
 105         for transcript in mdoc.findall('.//transcript'):
 106             if transcript.get('kind') != 'captions':
 107                 continue
 108             lang = transcript.get('srclang')
 109             subtitles[lang] = [{
 110                 'url': compat_str(typographic.get('src')),
 111                 'ext': typographic.get('format')
 112             } for typographic in transcript.findall('./typographic')]
 113         return subtitles
 114
 115     def _get_video_info(self, itemdoc):
 116         uri = itemdoc.find('guid').text
 117         video_id = self._id_from_uri(uri)
 118         self.report_extraction(video_id)
 119         content_el = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content')))
 120         mediagen_url = content_el.attrib['url']
 121         # Remove the templates, like &device={device}
 122         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
 123         if 'acceptMethods' not in mediagen_url:
 124             mediagen_url += '&' if '?' in mediagen_url else '?'
 125             mediagen_url += 'acceptMethods=fms'
 126
 127         mediagen_doc = self._download_xml(mediagen_url, video_id,
 128                                           'Downloading video urls')
 129
 130         item = mediagen_doc.find('./video/item')
 131         if item is not None and item.get('type') == 'text':
 132             message = '%s returned error: ' % self.IE_NAME
 133             if item.get('code') is not None:
 134                 message += '%s - ' % item.get('code')
 135             message += item.text
 136             raise ExtractorError(message, expected=True)
 137
 138         description = strip_or_none(xpath_text(itemdoc, 'description'))
 139
 140         timestamp = timeconvert(xpath_text(itemdoc, 'pubDate'))
 141
 142         title_el = None
 143         if title_el is None:
 144             title_el = find_xpath_attr(
 145                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 146                 'scheme', 'urn:mtvn:video_title')
 147         if title_el is None:
 148             title_el = itemdoc.find(compat_xpath('.//{http://search.yahoo.com/mrss/}title'))
 149         if title_el is None:
 150             title_el = itemdoc.find(compat_xpath('.//title'))
 151             if title_el.text is None:
 152                 title_el = None
 153
 154         title = title_el.text
 155         if title is None:
 156             raise ExtractorError('Could not find video title')
 157         title = title.strip()
 158
 159         # This a short id that's used in the webpage urls
 160         mtvn_id = None
 161         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 162                                        'scheme', 'urn:mtvn:id')
 163         if mtvn_id_node is not None:
 164             mtvn_id = mtvn_id_node.text
 165
 166         return {
 167             'title': title,
 168             'formats': self._extract_video_formats(mediagen_doc, mtvn_id),
 169             'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id),
 170             'id': video_id,
 171             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 172             'description': description,
 173             'duration': float_or_none(content_el.attrib.get('duration')),
 174             'timestamp': timestamp,
 175         }
 176
 177     def _get_feed_query(self, uri):
 178         data = {'uri': uri}
 179         if self._LANG:
 180             data['lang'] = self._LANG
 181         return compat_urllib_parse_urlencode(data)
 182
 183     def _get_videos_info(self, uri):
 184         video_id = self._id_from_uri(uri)
 185         feed_url = self._get_feed_url(uri)
 186         info_url = feed_url + '?' + self._get_feed_query(uri)
 187         return self._get_videos_info_from_url(info_url, video_id)
 188
 189     def _get_videos_info_from_url(self, url, video_id):
 190         idoc = self._download_xml(
 191             url, video_id,
 192             'Downloading info', transform_source=fix_xml_ampersands)
 193
 194         title = xpath_text(idoc, './channel/title')
 195         description = xpath_text(idoc, './channel/description')
 196
 197         return self.playlist_result(
 198             [self._get_video_info(item) for item in idoc.findall('.//item')],
 199             playlist_title=title, playlist_description=description)
 200
 201     def _extract_mgid(self, webpage):
 202         try:
 203             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 204             # or http://media.mtvnservices.com/{mgid}
 205             og_url = self._og_search_video_url(webpage)
 206             mgid = url_basename(og_url)
 207             if mgid.endswith('.swf'):
 208                 mgid = mgid[:-4]
 209         except RegexNotFoundError:
 210             mgid = None
 211
 212         if mgid is None or ':' not in mgid:
 213             mgid = self._search_regex(
 214                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
 215                 webpage, 'mgid', default=None)
 216
 217         if not mgid:
 218             sm4_embed = self._html_search_meta(
 219                 'sm4:video:embed', webpage, 'sm4 embed', default='')
 220             mgid = self._search_regex(
 221                 r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid')
 222         return mgid
 223
 224     def _real_extract(self, url):
 225         title = url_basename(url)
 226         webpage = self._download_webpage(url, title)
 227         mgid = self._extract_mgid(webpage)
 228         videos_info = self._get_videos_info(mgid)
 229         return videos_info
 230
 231
 232 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
 233     IE_NAME = 'mtvservices:embedded'
 234     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
 235
 236     _TEST = {
 237         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
 238         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
 239         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
 240         'info_dict': {
 241             'id': '1043906',
 242             'ext': 'mp4',
 243             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
 244             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
 245             'timestamp': 1400126400,
 246             'upload_date': '20140515',
 247         },
 248     }
 249
 250     @staticmethod
 251     def _extract_url(webpage):
 252         mobj = re.search(
 253             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//media.mtvnservices.com/embed/.+?)\1', webpage)
 254         if mobj:
 255             return mobj.group('url')
 256
 257     def _get_feed_url(self, uri):
 258         video_id = self._id_from_uri(uri)
 259         site_id = uri.replace(video_id, '')
 260         config_url = ('http://media.mtvnservices.com/pmt/e1/players/{0}/'
 261                       'context4/context5/config.xml'.format(site_id))
 262         config_doc = self._download_xml(config_url, video_id)
 263         feed_node = config_doc.find('.//feed')
 264         feed_url = feed_node.text.strip().split('?')[0]
 265         return feed_url
 266
 267     def _real_extract(self, url):
 268         mobj = re.match(self._VALID_URL, url)
 269         mgid = mobj.group('mgid')
 270         return self._get_videos_info(mgid)
 271
 272
 273 class MTVIE(MTVServicesInfoExtractor):
 274     _VALID_URL = r'''(?x)^https?://
 275         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 276            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 277
 278     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 279
 280     _TESTS = [
 281         {
 282             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 283             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 284             'info_dict': {
 285                 'id': '853555',
 286                 'ext': 'mp4',
 287                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 288                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 289                 'timestamp': 1352610000,
 290                 'upload_date': '20121111',
 291             },
 292         },
 293     ]
 294
 295     def _get_thumbnail_url(self, uri, itemdoc):
 296         return 'http://mtv.mtvnimages.com/uri/' + uri
 297
 298     def _real_extract(self, url):
 299         mobj = re.match(self._VALID_URL, url)
 300         video_id = mobj.group('videoid')
 301         uri = mobj.groupdict().get('mgid')
 302         if uri is None:
 303             webpage = self._download_webpage(url, video_id)
 304
 305             # Some videos come from Vevo.com
 306             m_vevo = re.search(
 307                 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
 308             if m_vevo:
 309                 vevo_id = m_vevo.group(1)
 310                 self.to_screen('Vevo video detected: %s' % vevo_id)
 311                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 312
 313             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 314         return self._get_videos_info(uri)
 315
 316
 317 class MTVDEIE(MTVServicesInfoExtractor):
 318     IE_NAME = 'mtv.de'
 319     _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:artists|shows|news)/(?:[^/]+/)*(?P<id>\d+)-[^/#?]+/*(?:[#?].*)?$'
 320     _TESTS = [{
 321         'url': 'http://www.mtv.de/artists/10571-cro/videos/61131-traum',
 322         'info_dict': {
 323             'id': 'music_video-a50bc5f0b3aa4b3190aa',
 324             'ext': 'flv',
 325             'title': 'MusicVideo_cro-traum',
 326             'description': 'Cro - Traum',
 327         },
 328         'params': {
 329             # rtmp download
 330             'skip_download': True,
 331         },
 332         'skip': 'Blocked at Travis CI',
 333     }, {
 334         # mediagen URL without query (e.g. http://videos.mtvnn.com/mediagen/e865da714c166d18d6f80893195fcb97)
 335         'url': 'http://www.mtv.de/shows/933-teen-mom-2/staffeln/5353/folgen/63565-enthullungen',
 336         'info_dict': {
 337             'id': 'local_playlist-f5ae778b9832cc837189',
 338             'ext': 'flv',
 339             'title': 'Episode_teen-mom-2_shows_season-5_episode-1_full-episode_part1',
 340         },
 341         'params': {
 342             # rtmp download
 343             'skip_download': True,
 344         },
 345         'skip': 'Blocked at Travis CI',
 346     }, {
 347         'url': 'http://www.mtv.de/news/77491-mtv-movies-spotlight-pixels-teil-3',
 348         'info_dict': {
 349             'id': 'local_playlist-4e760566473c4c8c5344',
 350             'ext': 'mp4',
 351             'title': 'Article_mtv-movies-spotlight-pixels-teil-3_short-clips_part1',
 352             'description': 'MTV Movies Supercut',
 353         },
 354         'params': {
 355             # rtmp download
 356             'skip_download': True,
 357         },
 358         'skip': 'Das Video kann zur Zeit nicht abgespielt werden.',
 359     }]
 360
 361     def _real_extract(self, url):
 362         video_id = self._match_id(url)
 363
 364         webpage = self._download_webpage(url, video_id)
 365
 366         playlist = self._parse_json(
 367             self._search_regex(
 368                 r'window\.pagePlaylist\s*=\s*(\[.+?\]);\n', webpage, 'page playlist'),
 369             video_id)
 370
 371         def _mrss_url(item):
 372             return item['mrss'] + item.get('mrssvars', '')
 373
 374         # news pages contain single video in playlist with different id
 375         if len(playlist) == 1:
 376             return self._get_videos_info_from_url(_mrss_url(playlist[0]), video_id)
 377
 378         for item in playlist:
 379             item_id = item.get('id')
 380             if item_id and compat_str(item_id) == video_id:
 381                 return self._get_videos_info_from_url(_mrss_url(item), video_id)