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

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     compat_urllib_parse,
   8     ExtractorError,
   9     fix_xml_ampersands,
  10     url_basename,
  11     RegexNotFoundError,
  12 )
  13
  14 def _media_xml_tag(tag):
  15     return '{http://search.yahoo.com/mrss/}%s' % tag
  16
  17
  18 class MTVServicesInfoExtractor(InfoExtractor):
  19     @staticmethod
  20     def _id_from_uri(uri):
  21         return uri.split(':')[-1]
  22
  23     # This was originally implemented for ComedyCentral, but it also works here
  24     @staticmethod
  25     def _transform_rtmp_url(rtmp_video_url):
  26         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  27         if not m:
  28             return rtmp_video_url
  29         base = 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=1+_pxI0=Ripod-h264+_pxL0=undefined+_pxM0=+_pxK=18639+_pxE=mp4/44620/mtvnorigin/'
  30         return base + m.group('finalid')
  31
  32     def _get_thumbnail_url(self, uri, itemdoc):
  33         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  34         thumb_node = itemdoc.find(search_path)
  35         if thumb_node is None:
  36             return None
  37         else:
  38             return thumb_node.attrib['url']
  39
  40     def _extract_video_formats(self, mdoc):
  41         if re.match(r'.*/error_country_block\.swf$', mdoc.find('.//src').text) is not None:
  42             raise ExtractorError('This video is not available from your country.', expected=True)
  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'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
  66         if 'acceptMethods' not in mediagen_url:
  67             mediagen_url += '&acceptMethods=fms'
  68         mediagen_doc = self._download_xml(mediagen_url, video_id,
  69             'Downloading video urls')
  70
  71         description_node = itemdoc.find('description')
  72         if description_node is not None:
  73             description = description_node.text.strip()
  74         else:
  75             description = None
  76
  77         return {
  78             'title': itemdoc.find('title').text,
  79             'formats': self._extract_video_formats(mediagen_doc),
  80             'id': video_id,
  81             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
  82             'description': description,
  83         }
  84
  85     def _get_videos_info(self, uri):
  86         video_id = self._id_from_uri(uri)
  87         data = compat_urllib_parse.urlencode({'uri': uri})
  88
  89         idoc = self._download_xml(
  90             self._FEED_URL + '?' + data, video_id,
  91             'Downloading info', transform_source=fix_xml_ampersands)
  92         return [self._get_video_info(item) for item in idoc.findall('.//item')]
  93
  94     def _real_extract(self, url):
  95         title = url_basename(url)
  96         webpage = self._download_webpage(url, title)
  97         try:
  98             # the url is in the format http://media.mtvnservices.com/fb/{mgid}.swf
  99             fb_url = self._og_search_video_url(webpage)
 100             mgid = url_basename(fb_url).rpartition('.')[0]
 101         except RegexNotFoundError:
 102             mgid = self._search_regex(r'data-mgid="(.*?)"', webpage, u'mgid')
 103         return self._get_videos_info(mgid)
 104
 105
 106 class MTVIE(MTVServicesInfoExtractor):
 107     _VALID_URL = r'''(?x)^https?://
 108         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 109            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 110
 111     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 112
 113     _TESTS = [
 114         {
 115             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 116             'file': '853555.mp4',
 117             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 118             'info_dict': {
 119                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 120                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 121             },
 122         },
 123         {
 124             'add_ie': ['Vevo'],
 125             'url': 'http://www.mtv.com/videos/taylor-swift/916187/everything-has-changed-ft-ed-sheeran.jhtml',
 126             'file': 'USCJY1331283.mp4',
 127             'md5': '73b4e7fcadd88929292fe52c3ced8caf',
 128             'info_dict': {
 129                 'title': 'Everything Has Changed',
 130                 'upload_date': '20130606',
 131                 'uploader': 'Taylor Swift',
 132             },
 133             'skip': 'VEVO is only available in some countries',
 134         },
 135     ]
 136
 137     def _get_thumbnail_url(self, uri, itemdoc):
 138         return 'http://mtv.mtvnimages.com/uri/' + uri
 139
 140     def _real_extract(self, url):
 141         mobj = re.match(self._VALID_URL, url)
 142         video_id = mobj.group('videoid')
 143         uri = mobj.groupdict().get('mgid')
 144         if uri is None:
 145             webpage = self._download_webpage(url, video_id)
 146
 147             # Some videos come from Vevo.com
 148             m_vevo = re.search(r'isVevoVideo = true;.*?vevoVideoId = "(.*?)";',
 149                                webpage, re.DOTALL)
 150             if m_vevo:
 151                 vevo_id = m_vevo.group(1);
 152                 self.to_screen('Vevo video detected: %s' % vevo_id)
 153                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 154
 155             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 156         return self._get_videos_info(uri)
 157
 158
 159 class MTVIggyIE(MTVServicesInfoExtractor):
 160     IE_NAME = 'mtviggy.com'
 161     _VALID_URL = r'https?://www\.mtviggy\.com/videos/.+'
 162     _TEST = {
 163         'url': 'http://www.mtviggy.com/videos/arcade-fire-behind-the-scenes-at-the-biggest-music-experiment-yet/',
 164         'info_dict': {
 165             'id': '984696',
 166             'ext': 'mp4',
 167             'title': 'Short',
 168         }
 169     }
 170     _FEED_URL = 'http://all.mtvworldverticals.com/feed-xml/'