_ 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     find_xpath_attr,
  10     fix_xml_ampersands,
  11     url_basename,
  12     RegexNotFoundError,
  13 )
  14
  15
  16 def _media_xml_tag(tag):
  17     return '{http://search.yahoo.com/mrss/}%s' % tag
  18
  19
  20 class MTVServicesInfoExtractor(InfoExtractor):
  21     @staticmethod
  22     def _id_from_uri(uri):
  23         return uri.split(':')[-1]
  24
  25     # This was originally implemented for ComedyCentral, but it also works here
  26     @staticmethod
  27     def _transform_rtmp_url(rtmp_video_url):
  28         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  29         if not m:
  30             return rtmp_video_url
  31         base = 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=1+_pxI0=Ripod-h264+_pxL0=undefined+_pxM0=+_pxK=18639+_pxE=mp4/44620/mtvnorigin/'
  32         return base + m.group('finalid')
  33
  34     def _get_thumbnail_url(self, uri, itemdoc):
  35         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  36         thumb_node = itemdoc.find(search_path)
  37         if thumb_node is None:
  38             return None
  39         else:
  40             return thumb_node.attrib['url']
  41
  42     def _extract_video_formats(self, mdoc):
  43         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4)$', mdoc.find('.//src').text) is not None:
  44             raise ExtractorError('This video is not available from your country.',
  45                 expected=True)
  46
  47         formats = []
  48         for rendition in mdoc.findall('.//rendition'):
  49             try:
  50                 _, _, ext = rendition.attrib['type'].partition('/')
  51                 rtmp_video_url = rendition.find('./src').text
  52                 formats.append({'ext': ext,
  53                                 'url': self._transform_rtmp_url(rtmp_video_url),
  54                                 'format_id': rendition.get('bitrate'),
  55                                 'width': int(rendition.get('width')),
  56                                 'height': int(rendition.get('height')),
  57                                 })
  58             except (KeyError, TypeError):
  59                 raise ExtractorError('Invalid rendition field.')
  60         return formats
  61
  62     def _get_video_info(self, itemdoc):
  63         uri = itemdoc.find('guid').text
  64         video_id = self._id_from_uri(uri)
  65         self.report_extraction(video_id)
  66         mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
  67         # Remove the templates, like &device={device}
  68         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
  69         if 'acceptMethods' not in mediagen_url:
  70             mediagen_url += '&acceptMethods=fms'
  71
  72         mediagen_doc = self._download_xml(mediagen_url, video_id,
  73             'Downloading video urls')
  74
  75         description_node = itemdoc.find('description')
  76         if description_node is not None:
  77             description = description_node.text.strip()
  78         else:
  79             description = None
  80
  81         title_el = None
  82         if title_el is None:
  83             title_el = find_xpath_attr(
  84                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
  85                 'scheme', 'urn:mtvn:video_title')
  86         if title_el is None:
  87             title_el = itemdoc.find('.//{http://search.yahoo.com/mrss/}title')
  88         if title_el is None:
  89             title_el = itemdoc.find('.//title')
  90             if title_el.text is None:
  91                 title_el = None
  92
  93         title = title_el.text
  94         if title is None:
  95             raise ExtractorError('Could not find video title')
  96         title = title.strip()
  97
  98         return {
  99             'title': title,
 100             'formats': self._extract_video_formats(mediagen_doc),
 101             'id': video_id,
 102             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 103             'description': description,
 104         }
 105
 106     def _get_videos_info(self, uri):
 107         video_id = self._id_from_uri(uri)
 108         data = compat_urllib_parse.urlencode({'uri': uri})
 109
 110         idoc = self._download_xml(
 111             self._FEED_URL + '?' + data, video_id,
 112             'Downloading info', transform_source=fix_xml_ampersands)
 113         return [self._get_video_info(item) for item in idoc.findall('.//item')]
 114
 115     def _real_extract(self, url):
 116         title = url_basename(url)
 117         webpage = self._download_webpage(url, title)
 118         try:
 119             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 120             # or http://media.mtvnservices.com/{mgid}
 121             og_url = self._og_search_video_url(webpage)
 122             mgid = url_basename(og_url)
 123             if mgid.endswith('.swf'):
 124                 mgid = mgid[:-4]
 125         except RegexNotFoundError:
 126             mgid = self._search_regex(
 127                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
 128                 webpage, u'mgid')
 129         return self._get_videos_info(mgid)
 130
 131
 132 class MTVIE(MTVServicesInfoExtractor):
 133     _VALID_URL = r'''(?x)^https?://
 134         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 135            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 136
 137     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 138
 139     _TESTS = [
 140         {
 141             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 142             'file': '853555.mp4',
 143             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 144             'info_dict': {
 145                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 146                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 147             },
 148         },
 149         {
 150             'add_ie': ['Vevo'],
 151             'url': 'http://www.mtv.com/videos/taylor-swift/916187/everything-has-changed-ft-ed-sheeran.jhtml',
 152             'file': 'USCJY1331283.mp4',
 153             'md5': '73b4e7fcadd88929292fe52c3ced8caf',
 154             'info_dict': {
 155                 'title': 'Everything Has Changed',
 156                 'upload_date': '20130606',
 157                 'uploader': 'Taylor Swift',
 158             },
 159             'skip': 'VEVO is only available in some countries',
 160         },
 161     ]
 162
 163     def _get_thumbnail_url(self, uri, itemdoc):
 164         return 'http://mtv.mtvnimages.com/uri/' + uri
 165
 166     def _real_extract(self, url):
 167         mobj = re.match(self._VALID_URL, url)
 168         video_id = mobj.group('videoid')
 169         uri = mobj.groupdict().get('mgid')
 170         if uri is None:
 171             webpage = self._download_webpage(url, video_id)
 172
 173             # Some videos come from Vevo.com
 174             m_vevo = re.search(r'isVevoVideo = true;.*?vevoVideoId = "(.*?)";',
 175                                webpage, re.DOTALL)
 176             if m_vevo:
 177                 vevo_id = m_vevo.group(1);
 178                 self.to_screen('Vevo video detected: %s' % vevo_id)
 179                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 180
 181             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 182         return self._get_videos_info(uri)
 183
 184
 185 class MTVIggyIE(MTVServicesInfoExtractor):
 186     IE_NAME = 'mtviggy.com'
 187     _VALID_URL = r'https?://www\.mtviggy\.com/videos/.+'
 188     _TEST = {
 189         'url': 'http://www.mtviggy.com/videos/arcade-fire-behind-the-scenes-at-the-biggest-music-experiment-yet/',
 190         'info_dict': {
 191             'id': '984696',
 192             'ext': 'mp4',
 193             'title': 'Arcade Fire: Behind the Scenes at the Biggest Music Experiment Yet',
 194         }
 195     }
 196     _FEED_URL = 'http://all.mtvworldverticals.com/feed-xml/'