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 def _media_xml_tag(tag):
  11     return '{http://search.yahoo.com/mrss/}%s' % tag
  12
  13 class MTVIE(InfoExtractor):
  14     _VALID_URL = r'^https?://(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$'
  15
  16     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
  17
  18     _TESTS = [
  19         {
  20             u'url': u'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
  21             u'file': u'853555.mp4',
  22             u'md5': u'850f3f143316b1e71fa56a4edfd6e0f8',
  23             u'info_dict': {
  24                 u'title': u'Taylor Swift - "Ours (VH1 Storytellers)"',
  25                 u'description': u'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
  26             },
  27         },
  28         {
  29             u'url': u'http://www.mtv.com/videos/taylor-swift/916187/everything-has-changed-ft-ed-sheeran.jhtml',
  30             u'file': u'USCJY1331283.mp4',
  31             u'md5': u'73b4e7fcadd88929292fe52c3ced8caf',
  32             u'info_dict': {
  33                 u'title': u'Everything Has Changed',
  34                 u'upload_date': u'20130606',
  35                 u'uploader': u'Taylor Swift',
  36             },
  37             u'skip': u'VEVO is only available in some countries',
  38         },
  39     ]
  40
  41     @staticmethod
  42     def _id_from_uri(uri):
  43         return uri.split(':')[-1]
  44
  45     # This was originally implemented for ComedyCentral, but it also works here
  46     @staticmethod
  47     def _transform_rtmp_url(rtmp_video_url):
  48         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  49         if not m:
  50             raise ExtractorError(u'Cannot transform RTMP url')
  51         base = 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=1+_pxI0=Ripod-h264+_pxL0=undefined+_pxM0=+_pxK=18639+_pxE=mp4/44620/mtvnorigin/'
  52         return base + m.group('finalid')
  53
  54     def _get_thumbnail_url(self, uri, itemdoc):
  55         return 'http://mtv.mtvnimages.com/uri/' + uri
  56
  57     def _extract_video_formats(self, metadataXml):
  58         if '/error_country_block.swf' in metadataXml:
  59             raise ExtractorError(u'This video is not available from your country.', expected=True)
  60         mdoc = xml.etree.ElementTree.fromstring(metadataXml.encode('utf-8'))
  61         renditions = mdoc.findall('.//rendition')
  62
  63         formats = []
  64         for rendition in mdoc.findall('.//rendition'):
  65             try:
  66                 _, _, ext = rendition.attrib['type'].partition('/')
  67                 rtmp_video_url = rendition.find('./src').text
  68                 formats.append({'ext': ext,
  69                                 'url': self._transform_rtmp_url(rtmp_video_url),
  70                                 'format_id': rendition.get('bitrate'),
  71                                 'width': int(rendition.get('width')),
  72                                 'height': int(rendition.get('height')),
  73                                 })
  74             except (KeyError, TypeError):
  75                 raise ExtractorError('Invalid rendition field.')
  76         return formats
  77
  78     def _get_video_info(self, itemdoc):
  79         uri = itemdoc.find('guid').text
  80         video_id = self._id_from_uri(uri)
  81         self.report_extraction(video_id)
  82         mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
  83         if 'acceptMethods' not in mediagen_url:
  84             mediagen_url += '&acceptMethods=fms'
  85         mediagen_page = self._download_webpage(mediagen_url, video_id,
  86                                                u'Downloading video urls')
  87
  88         description_node = itemdoc.find('description')
  89         if description_node is not None:
  90             description = description_node.text.strip()
  91         else:
  92             description = None
  93
  94         info = {
  95             'title': itemdoc.find('title').text,
  96             'formats': self._extract_video_formats(mediagen_page),
  97             'id': video_id,
  98             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
  99             'description': description,
 100         }
 101
 102         # TODO: Remove when #980 has been merged
 103         info.update(info['formats'][-1])
 104
 105         return info
 106
 107     def _get_videos_info(self, uri):
 108         video_id = self._id_from_uri(uri)
 109         data = compat_urllib_parse.urlencode({'uri': uri})
 110         infoXml = self._download_webpage(self._FEED_URL +'?' + data, video_id,
 111                                          u'Downloading info')
 112         idoc = xml.etree.ElementTree.fromstring(infoXml.encode('utf-8'))
 113         return [self._get_video_info(item) for item in idoc.findall('.//item')]
 114
 115     def _real_extract(self, url):
 116         mobj = re.match(self._VALID_URL, url)
 117         video_id = mobj.group('videoid')
 118
 119         webpage = self._download_webpage(url, video_id)
 120
 121         # Some videos come from Vevo.com
 122         m_vevo = re.search(r'isVevoVideo = true;.*?vevoVideoId = "(.*?)";',
 123                            webpage, re.DOTALL)
 124         if m_vevo:
 125             vevo_id = m_vevo.group(1);
 126             self.to_screen(u'Vevo video detected: %s' % vevo_id)
 127             return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 128
 129         uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, u'uri')
 130         return self._get_videos_info(uri)