_ Git - youtube-dl/blob - youtube_dl/extractor/mtv.py

   1 import re
   2 import xml.etree.ElementTree
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     compat_urllib_parse,
   7     ExtractorError,
   8 )
   9
  10
  11 class MTVIE(InfoExtractor):
  12     _VALID_URL = r'^https?://(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$'
  13
  14     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
  15
  16     _TESTS = [
  17         {
  18             u'url': u'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
  19             u'file': u'853555.mp4',
  20             u'md5': u'850f3f143316b1e71fa56a4edfd6e0f8',
  21             u'info_dict': {
  22                 u'title': u'Taylor Swift - "Ours (VH1 Storytellers)"',
  23                 u'description': u'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
  24             },
  25         },
  26         {
  27             u'url': u'http://www.mtv.com/videos/taylor-swift/916187/everything-has-changed-ft-ed-sheeran.jhtml',
  28             u'file': u'USCJY1331283.mp4',
  29             u'md5': u'73b4e7fcadd88929292fe52c3ced8caf',
  30             u'info_dict': {
  31                 u'title': u'Everything Has Changed',
  32                 u'upload_date': u'20130606',
  33                 u'uploader': u'Taylor Swift',
  34             },
  35             u'skip': u'VEVO is only available in some countries',
  36         },
  37     ]
  38
  39     @staticmethod
  40     def _id_from_uri(uri):
  41         return uri.split(':')[-1]
  42
  43     # This was originally implemented for ComedyCentral, but it also works here
  44     @staticmethod
  45     def _transform_rtmp_url(rtmp_video_url):
  46         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  47         if not m:
  48             raise ExtractorError(u'Cannot transform RTMP url')
  49         base = 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=1+_pxI0=Ripod-h264+_pxL0=undefined+_pxM0=+_pxK=18639+_pxE=mp4/44620/mtvnorigin/'
  50         return base + m.group('finalid')
  51
  52     def _extract_video_url(self, metadataXml):
  53         if '/error_country_block.swf' in metadataXml:
  54             raise ExtractorError(u'This video is not available from your country.', expected=True)
  55         mdoc = xml.etree.ElementTree.fromstring(metadataXml.encode('utf-8'))
  56         renditions = mdoc.findall('.//rendition')
  57
  58         # For now, always pick the highest quality.
  59         rendition = renditions[-1]
  60
  61         try:
  62             _,_,ext = rendition.attrib['type'].partition('/')
  63             format = ext + '-' + rendition.attrib['width'] + 'x' + rendition.attrib['height'] + '_' + rendition.attrib['bitrate']
  64             rtmp_video_url = rendition.find('./src').text
  65         except KeyError:
  66             raise ExtractorError('Invalid rendition field.')
  67         video_url = self._transform_rtmp_url(rtmp_video_url)
  68         return {'ext': ext, 'url': video_url, 'format': format}
  69
  70     def _get_video_info(self, itemdoc):
  71         uri = itemdoc.find('guid').text
  72         video_id = self._id_from_uri(uri)
  73         self.report_extraction(video_id)
  74         media_namespace = {'media': 'http://search.yahoo.com/mrss/'}
  75         mediagen_url = itemdoc.find('media:group/media:content', media_namespace).attrib['url']
  76         if 'acceptMethods' not in mediagen_url:
  77             mediagen_url += '&acceptMethods=fms'
  78         mediagen_page = self._download_webpage(mediagen_url, video_id,
  79                                                u'Downloading video urls')
  80         video_info = self._extract_video_url(mediagen_page)
  81
  82         description_node = itemdoc.find('description')
  83         if description_node is not None:
  84             description = description_node.text
  85         else:
  86             description = None
  87         video_info.update({'title': itemdoc.find('title').text,
  88                            'id': video_id,
  89                            'thumbnail': 'http://mtv.mtvnimages.com/uri/' + uri,
  90                            'description': description,
  91                            })
  92         return video_info
  93
  94     def _get_videos_info(self, uri):
  95         video_id = self._id_from_uri(uri)
  96         data = compat_urllib_parse.urlencode({'uri': uri})
  97         infoXml = self._download_webpage(self._FEED_URL +'?' + data, video_id,
  98                                          u'Downloading info')
  99         idoc = xml.etree.ElementTree.fromstring(infoXml.encode('utf-8'))
 100         return [self._get_video_info(item) for item in idoc.findall('.//item')]
 101
 102     def _real_extract(self, url):
 103         mobj = re.match(self._VALID_URL, url)
 104         video_id = mobj.group('videoid')
 105
 106         webpage = self._download_webpage(url, video_id)
 107
 108         # Some videos come from Vevo.com
 109         m_vevo = re.search(r'isVevoVideo = true;.*?vevoVideoId = "(.*?)";',
 110                            webpage, re.DOTALL)
 111         if m_vevo:
 112             vevo_id = m_vevo.group(1);
 113             self.to_screen(u'Vevo video detected: %s' % vevo_id)
 114             return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 115
 116         uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, u'uri')
 117         return self._get_videos_info(uri)