_ 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     find_xpath_attr,
  11     fix_xml_ampersands,
  12 )
  13
  14
  15 def _media_xml_tag(tag):
  16     return '{http://search.yahoo.com/mrss/}%s' % tag
  17
  18
  19 class MTVServicesInfoExtractor(InfoExtractor):
  20     @staticmethod
  21     def _id_from_uri(uri):
  22         return uri.split(':')[-1]
  23
  24     # This was originally implemented for ComedyCentral, but it also works here
  25     @staticmethod
  26     def _transform_rtmp_url(rtmp_video_url):
  27         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  28         if not m:
  29             return rtmp_video_url
  30         base = 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=1+_pxI0=Ripod-h264+_pxL0=undefined+_pxM0=+_pxK=18639+_pxE=mp4/44620/mtvnorigin/'
  31         return base + m.group('finalid')
  32
  33     def _get_thumbnail_url(self, uri, itemdoc):
  34         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  35         thumb_node = itemdoc.find(search_path)
  36         if thumb_node is None:
  37             return None
  38         else:
  39             return thumb_node.attrib['url']
  40
  41     def _extract_video_formats(self, metadataXml):
  42         if '/error_country_block.swf' in metadataXml:
  43             raise ExtractorError('This video is not available from your country.', expected=True)
  44         mdoc = xml.etree.ElementTree.fromstring(metadataXml.encode('utf-8'))
  45
  46         formats = []
  47         for rendition in mdoc.findall('.//rendition'):
  48             try:
  49                 _, _, ext = rendition.attrib['type'].partition('/')
  50                 rtmp_video_url = rendition.find('./src').text
  51                 formats.append({'ext': ext,
  52                                 'url': self._transform_rtmp_url(rtmp_video_url),
  53                                 'format_id': rendition.get('bitrate'),
  54                                 'width': int(rendition.get('width')),
  55                                 'height': int(rendition.get('height')),
  56                                 })
  57             except (KeyError, TypeError):
  58                 raise ExtractorError('Invalid rendition field.')
  59         return formats
  60
  61     def _get_video_info(self, itemdoc):
  62         uri = itemdoc.find('guid').text
  63         video_id = self._id_from_uri(uri)
  64         self.report_extraction(video_id)
  65         mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
  66         # Remove the templates, like &device={device}
  67         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
  68         if 'acceptMethods' not in mediagen_url:
  69             mediagen_url += '&acceptMethods=fms'
  70         mediagen_page = self._download_webpage(mediagen_url, video_id,
  71                                                'Downloading video urls')
  72
  73         description_node = itemdoc.find('description')
  74         if description_node is not None:
  75             description = description_node.text.strip()
  76         else:
  77             description = None
  78
  79         title_el = None
  80         if title_el is None:
  81             title_el = find_xpath_attr(
  82                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
  83                 'scheme', 'urn:mtvn:video_title')
  84         if title_el is None:
  85             title_el = itemdoc.find('.//{http://search.yahoo.com/mrss/}title')
  86         if title_el is None:
  87             title_el = itemdoc.find('.//title')
  88         title = title_el.text
  89         if title is None:
  90             raise ExtractorError('Could not find video title')
  91
  92         return {
  93             'title': title,
  94             'formats': self._extract_video_formats(mediagen_page),
  95             'id': video_id,
  96             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
  97             'description': description,
  98         }
  99
 100     def _get_videos_info(self, uri):
 101         video_id = self._id_from_uri(uri)
 102         data = compat_urllib_parse.urlencode({'uri': uri})
 103
 104         idoc = self._download_xml(
 105             self._FEED_URL + '?' + data, video_id,
 106             'Downloading info', transform_source=fix_xml_ampersands)
 107         return [self._get_video_info(item) for item in idoc.findall('.//item')]
 108
 109
 110 class MTVIE(MTVServicesInfoExtractor):
 111     _VALID_URL = r'''(?x)^https?://
 112         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 113            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 114
 115     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 116
 117     _TESTS = [
 118         {
 119             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 120             'file': '853555.mp4',
 121             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 122             'info_dict': {
 123                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 124                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 125             },
 126         },
 127         {
 128             'add_ie': ['Vevo'],
 129             'url': 'http://www.mtv.com/videos/taylor-swift/916187/everything-has-changed-ft-ed-sheeran.jhtml',
 130             'file': 'USCJY1331283.mp4',
 131             'md5': '73b4e7fcadd88929292fe52c3ced8caf',
 132             'info_dict': {
 133                 'title': 'Everything Has Changed',
 134                 'upload_date': '20130606',
 135                 'uploader': 'Taylor Swift',
 136             },
 137             'skip': 'VEVO is only available in some countries',
 138         },
 139     ]
 140
 141     def _get_thumbnail_url(self, uri, itemdoc):
 142         return 'http://mtv.mtvnimages.com/uri/' + uri
 143
 144     def _real_extract(self, url):
 145         mobj = re.match(self._VALID_URL, url)
 146         video_id = mobj.group('videoid')
 147         uri = mobj.groupdict().get('mgid')
 148         if uri is None:
 149             webpage = self._download_webpage(url, video_id)
 150
 151             # Some videos come from Vevo.com
 152             m_vevo = re.search(r'isVevoVideo = true;.*?vevoVideoId = "(.*?)";',
 153                                webpage, re.DOTALL)
 154             if m_vevo:
 155                 vevo_id = m_vevo.group(1);
 156                 self.to_screen('Vevo video detected: %s' % vevo_id)
 157                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 158
 159             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 160         return self._get_videos_info(uri)