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

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import xml.etree.ElementTree
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     compat_urllib_parse,
   9     ExtractorError,
  10     fix_xml_ampersands,
  11 )
  12
  13 def _media_xml_tag(tag):
  14     return '{http://search.yahoo.com/mrss/}%s' % tag
  15
  16
  17 class MTVServicesInfoExtractor(InfoExtractor):
  18     @staticmethod
  19     def _id_from_uri(uri):
  20         return uri.split(':')[-1]
  21
  22     # This was originally implemented for ComedyCentral, but it also works here
  23     @staticmethod
  24     def _transform_rtmp_url(rtmp_video_url):
  25         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  26         if not m:
  27             return rtmp_video_url
  28         base = 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=1+_pxI0=Ripod-h264+_pxL0=undefined+_pxM0=+_pxK=18639+_pxE=mp4/44620/mtvnorigin/'
  29         return base + m.group('finalid')
  30
  31     def _get_thumbnail_url(self, uri, itemdoc):
  32         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  33         thumb_node = itemdoc.find(search_path)
  34         if thumb_node is None:
  35             return None
  36         else:
  37             return thumb_node.attrib['url']
  38
  39     def _extract_video_formats(self, metadataXml):
  40         if '/error_country_block.swf' in metadataXml:
  41             raise ExtractorError('This video is not available from your country.', expected=True)
  42         mdoc = xml.etree.ElementTree.fromstring(metadataXml.encode('utf-8'))
  43
  44         formats = []
  45         for rendition in mdoc.findall('.//rendition'):
  46             try:
  47                 _, _, ext = rendition.attrib['type'].partition('/')
  48                 rtmp_video_url = rendition.find('./src').text
  49                 formats.append({'ext': ext,
  50                                 'url': self._transform_rtmp_url(rtmp_video_url),
  51                                 'format_id': rendition.get('bitrate'),
  52                                 'width': int(rendition.get('width')),
  53                                 'height': int(rendition.get('height')),
  54                                 })
  55             except (KeyError, TypeError):
  56                 raise ExtractorError('Invalid rendition field.')
  57         return formats
  58
  59     def _get_video_info(self, itemdoc):
  60         uri = itemdoc.find('guid').text
  61         video_id = self._id_from_uri(uri)
  62         self.report_extraction(video_id)
  63         mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
  64         # Remove the templates, like &device={device}
  65         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
  66         if 'acceptMethods' not in mediagen_url:
  67             mediagen_url += '&acceptMethods=fms'
  68         mediagen_page = self._download_webpage(mediagen_url, video_id,
  69                                                'Downloading video urls')
  70
  71         description_node = itemdoc.find('description')
  72         if description_node is not None:
  73             description = description_node.text.strip()
  74         else:
  75             description = None
  76
  77         return {
  78             'title': itemdoc.find('title').text,
  79             'formats': self._extract_video_formats(mediagen_page),
  80             'id': video_id,
  81             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
  82             'description': description,
  83         }
  84
  85     def _get_videos_info(self, uri):
  86         video_id = self._id_from_uri(uri)
  87         data = compat_urllib_parse.urlencode({'uri': uri})
  88
  89         idoc = self._download_xml(
  90             self._FEED_URL + '?' + data, video_id,
  91             'Downloading info', transform_source=fix_xml_ampersands)
  92         return [self._get_video_info(item) for item in idoc.findall('.//item')]
  93
  94
  95 class MTVIE(MTVServicesInfoExtractor):
  96     _VALID_URL = r'''(?x)^https?://
  97         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
  98            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
  99
 100     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 101
 102     _TESTS = [
 103         {
 104             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 105             'file': '853555.mp4',
 106             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 107             'info_dict': {
 108                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 109                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 110             },
 111         },
 112         {
 113             'add_ie': ['Vevo'],
 114             'url': 'http://www.mtv.com/videos/taylor-swift/916187/everything-has-changed-ft-ed-sheeran.jhtml',
 115             'file': 'USCJY1331283.mp4',
 116             'md5': '73b4e7fcadd88929292fe52c3ced8caf',
 117             'info_dict': {
 118                 'title': 'Everything Has Changed',
 119                 'upload_date': '20130606',
 120                 'uploader': 'Taylor Swift',
 121             },
 122             'skip': 'VEVO is only available in some countries',
 123         },
 124     ]
 125
 126     def _get_thumbnail_url(self, uri, itemdoc):
 127         return 'http://mtv.mtvnimages.com/uri/' + uri
 128
 129     def _real_extract(self, url):
 130         mobj = re.match(self._VALID_URL, url)
 131         video_id = mobj.group('videoid')
 132         uri = mobj.groupdict().get('mgid')
 133         if uri is None:
 134             webpage = self._download_webpage(url, video_id)
 135
 136             # Some videos come from Vevo.com
 137             m_vevo = re.search(r'isVevoVideo = true;.*?vevoVideoId = "(.*?)";',
 138                                webpage, re.DOTALL)
 139             if m_vevo:
 140                 vevo_id = m_vevo.group(1);
 141                 self.to_screen('Vevo video detected: %s' % vevo_id)
 142                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 143
 144             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 145         return self._get_videos_info(uri)