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