Raphaël G. Git Repositories - youtubedl/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 def _media_xml_tag(tag):
  11     return '{http://search.yahoo.com/mrss/}%s' % tag
  12
  13
  14 class MTVServicesInfoExtractor(InfoExtractor):
  15     @staticmethod
  16     def _id_from_uri(uri):
  17         return uri.split(':')[-1]
  18
  19     # This was originally implemented for ComedyCentral, but it also works here
  20     @staticmethod
  21     def _transform_rtmp_url(rtmp_video_url):
  22         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  23         if not m:
  24             return rtmp_video_url
  25         base = 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=1+_pxI0=Ripod-h264+_pxL0=undefined+_pxM0=+_pxK=18639+_pxE=mp4/44620/mtvnorigin/'
  26         return base + m.group('finalid')
  27
  28     def _get_thumbnail_url(self, uri, itemdoc):
  29         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  30         thumb_node = itemdoc.find(search_path)
  31         if thumb_node is None:
  32             return None
  33         else:
  34             return thumb_node.attrib['url']
  35
  36     def _extract_video_formats(self, metadataXml):
  37         if '/error_country_block.swf' in metadataXml:
  38             raise ExtractorError(u'This video is not available from your country.', expected=True)
  39         mdoc = xml.etree.ElementTree.fromstring(metadataXml.encode('utf-8'))
  40
  41         formats = []
  42         for rendition in mdoc.findall('.//rendition'):
  43             try:
  44                 _, _, ext = rendition.attrib['type'].partition('/')
  45                 rtmp_video_url = rendition.find('./src').text
  46                 formats.append({'ext': ext,
  47                                 'url': self._transform_rtmp_url(rtmp_video_url),
  48                                 'format_id': rendition.get('bitrate'),
  49                                 'width': int(rendition.get('width')),
  50                                 'height': int(rendition.get('height')),
  51                                 })
  52             except (KeyError, TypeError):
  53                 raise ExtractorError('Invalid rendition field.')
  54         return formats
  55
  56     def _get_video_info(self, itemdoc):
  57         uri = itemdoc.find('guid').text
  58         video_id = self._id_from_uri(uri)
  59         self.report_extraction(video_id)
  60         mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
  61         # Remove the templates, like &device={device}
  62         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', u'', mediagen_url)
  63         if 'acceptMethods' not in mediagen_url:
  64             mediagen_url += '&acceptMethods=fms'
  65         mediagen_page = self._download_webpage(mediagen_url, video_id,
  66                                                u'Downloading video urls')
  67
  68         description_node = itemdoc.find('description')
  69         if description_node is not None:
  70             description = description_node.text.strip()
  71         else:
  72             description = None
  73
  74         return {
  75             'title': itemdoc.find('title').text,
  76             'formats': self._extract_video_formats(mediagen_page),
  77             'id': video_id,
  78             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
  79             'description': description,
  80         }
  81
  82     def _get_videos_info(self, uri):
  83         video_id = self._id_from_uri(uri)
  84         data = compat_urllib_parse.urlencode({'uri': uri})
  85         idoc = self._download_xml(self._FEED_URL +'?' + data, video_id,
  86                                          u'Downloading info')
  87         return [self._get_video_info(item) for item in idoc.findall('.//item')]
  88
  89
  90 class MTVIE(MTVServicesInfoExtractor):
  91     _VALID_URL = r'^https?://(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$'
  92
  93     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
  94
  95     _TESTS = [
  96         {
  97             u'url': u'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
  98             u'file': u'853555.mp4',
  99             u'md5': u'850f3f143316b1e71fa56a4edfd6e0f8',
 100             u'info_dict': {
 101                 u'title': u'Taylor Swift - "Ours (VH1 Storytellers)"',
 102                 u'description': u'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 103             },
 104         },
 105         {
 106             u'add_ie': ['Vevo'],
 107             u'url': u'http://www.mtv.com/videos/taylor-swift/916187/everything-has-changed-ft-ed-sheeran.jhtml',
 108             u'file': u'USCJY1331283.mp4',
 109             u'md5': u'73b4e7fcadd88929292fe52c3ced8caf',
 110             u'info_dict': {
 111                 u'title': u'Everything Has Changed',
 112                 u'upload_date': u'20130606',
 113                 u'uploader': u'Taylor Swift',
 114             },
 115             u'skip': u'VEVO is only available in some countries',
 116         },
 117     ]
 118
 119     def _get_thumbnail_url(self, uri, itemdoc):
 120         return 'http://mtv.mtvnimages.com/uri/' + uri
 121
 122     def _real_extract(self, url):
 123         mobj = re.match(self._VALID_URL, url)
 124         video_id = mobj.group('videoid')
 125
 126         webpage = self._download_webpage(url, video_id)
 127
 128         # Some videos come from Vevo.com
 129         m_vevo = re.search(r'isVevoVideo = true;.*?vevoVideoId = "(.*?)";',
 130                            webpage, re.DOTALL)
 131         if m_vevo:
 132             vevo_id = m_vevo.group(1);
 133             self.to_screen(u'Vevo video detected: %s' % vevo_id)
 134             return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 135
 136         uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, u'uri')
 137         return self._get_videos_info(uri)