_ 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     compat_urllib_request,
   9     ExtractorError,
  10     find_xpath_attr,
  11     fix_xml_ampersands,
  12     unescapeHTML,
  13     url_basename,
  14     RegexNotFoundError,
  15 )
  16
  17
  18 def _media_xml_tag(tag):
  19     return '{http://search.yahoo.com/mrss/}%s' % tag
  20
  21
  22 class MTVServicesInfoExtractor(InfoExtractor):
  23     _MOBILE_TEMPLATE = None
  24     @staticmethod
  25     def _id_from_uri(uri):
  26         return uri.split(':')[-1]
  27
  28     # This was originally implemented for ComedyCentral, but it also works here
  29     @staticmethod
  30     def _transform_rtmp_url(rtmp_video_url):
  31         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  32         if not m:
  33             return rtmp_video_url
  34         base = 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=1+_pxI0=Ripod-h264+_pxL0=undefined+_pxM0=+_pxK=18639+_pxE=mp4/44620/mtvnorigin/'
  35         return base + m.group('finalid')
  36
  37     def _get_thumbnail_url(self, uri, itemdoc):
  38         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  39         thumb_node = itemdoc.find(search_path)
  40         if thumb_node is None:
  41             return None
  42         else:
  43             return thumb_node.attrib['url']
  44
  45     def _extract_mobile_video_formats(self, mtvn_id):
  46         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
  47         req = compat_urllib_request.Request(webpage_url)
  48         # Otherwise we get a webpage that would execute some javascript
  49         req.add_header('Youtubedl-user-agent', 'curl/7')
  50         webpage = self._download_webpage(req, mtvn_id,
  51             'Downloading mobile page')
  52         url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
  53         return [{'url': url,'ext': 'mp4',}]
  54
  55     def _extract_video_formats(self, mdoc, mtvn_id):
  56         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4)$', mdoc.find('.//src').text) is not None:
  57             if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
  58                 self._downloader.report_warning('The normal version is not '
  59                     'available from your country, trying with the mobile version')
  60                 return self._extract_mobile_video_formats(mtvn_id)
  61             raise ExtractorError('This video is not available from your country.',
  62                 expected=True)
  63
  64         formats = []
  65         for rendition in mdoc.findall('.//rendition'):
  66             try:
  67                 _, _, ext = rendition.attrib['type'].partition('/')
  68                 rtmp_video_url = rendition.find('./src').text
  69                 formats.append({'ext': ext,
  70                                 'url': self._transform_rtmp_url(rtmp_video_url),
  71                                 'format_id': rendition.get('bitrate'),
  72                                 'width': int(rendition.get('width')),
  73                                 'height': int(rendition.get('height')),
  74                                 })
  75             except (KeyError, TypeError):
  76                 raise ExtractorError('Invalid rendition field.')
  77         return formats
  78
  79     def _get_video_info(self, itemdoc):
  80         uri = itemdoc.find('guid').text
  81         video_id = self._id_from_uri(uri)
  82         self.report_extraction(video_id)
  83         mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
  84         # Remove the templates, like &device={device}
  85         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
  86         if 'acceptMethods' not in mediagen_url:
  87             mediagen_url += '&acceptMethods=fms'
  88
  89         mediagen_doc = self._download_xml(mediagen_url, video_id,
  90             'Downloading video urls')
  91
  92         description_node = itemdoc.find('description')
  93         if description_node is not None:
  94             description = description_node.text.strip()
  95         else:
  96             description = None
  97
  98         title_el = None
  99         if title_el is None:
 100             title_el = find_xpath_attr(
 101                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 102                 'scheme', 'urn:mtvn:video_title')
 103         if title_el is None:
 104             title_el = itemdoc.find('.//{http://search.yahoo.com/mrss/}title')
 105         if title_el is None:
 106             title_el = itemdoc.find('.//title')
 107             if title_el.text is None:
 108                 title_el = None
 109
 110         title = title_el.text
 111         if title is None:
 112             raise ExtractorError('Could not find video title')
 113         title = title.strip()
 114
 115         # This a short id that's used in the webpage urls
 116         mtvn_id = None
 117         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 118                 'scheme', 'urn:mtvn:id')
 119         if mtvn_id_node is not None:
 120             mtvn_id = mtvn_id_node.text
 121
 122         return {
 123             'title': title,
 124             'formats': self._extract_video_formats(mediagen_doc, mtvn_id),
 125             'id': video_id,
 126             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 127             'description': description,
 128         }
 129
 130     def _get_videos_info(self, uri):
 131         video_id = self._id_from_uri(uri)
 132         data = compat_urllib_parse.urlencode({'uri': uri})
 133
 134         idoc = self._download_xml(
 135             self._FEED_URL + '?' + data, video_id,
 136             'Downloading info', transform_source=fix_xml_ampersands)
 137         return [self._get_video_info(item) for item in idoc.findall('.//item')]
 138
 139     def _real_extract(self, url):
 140         title = url_basename(url)
 141         webpage = self._download_webpage(url, title)
 142         try:
 143             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 144             # or http://media.mtvnservices.com/{mgid}
 145             og_url = self._og_search_video_url(webpage)
 146             mgid = url_basename(og_url)
 147             if mgid.endswith('.swf'):
 148                 mgid = mgid[:-4]
 149         except RegexNotFoundError:
 150             mgid = self._search_regex(
 151                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
 152                 webpage, u'mgid')
 153         return self._get_videos_info(mgid)
 154
 155
 156 class MTVIE(MTVServicesInfoExtractor):
 157     _VALID_URL = r'''(?x)^https?://
 158         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 159            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 160
 161     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 162
 163     _TESTS = [
 164         {
 165             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 166             'file': '853555.mp4',
 167             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 168             'info_dict': {
 169                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 170                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 171             },
 172         },
 173         {
 174             'add_ie': ['Vevo'],
 175             'url': 'http://www.mtv.com/videos/taylor-swift/916187/everything-has-changed-ft-ed-sheeran.jhtml',
 176             'file': 'USCJY1331283.mp4',
 177             'md5': '73b4e7fcadd88929292fe52c3ced8caf',
 178             'info_dict': {
 179                 'title': 'Everything Has Changed',
 180                 'upload_date': '20130606',
 181                 'uploader': 'Taylor Swift',
 182             },
 183             'skip': 'VEVO is only available in some countries',
 184         },
 185     ]
 186
 187     def _get_thumbnail_url(self, uri, itemdoc):
 188         return 'http://mtv.mtvnimages.com/uri/' + uri
 189
 190     def _real_extract(self, url):
 191         mobj = re.match(self._VALID_URL, url)
 192         video_id = mobj.group('videoid')
 193         uri = mobj.groupdict().get('mgid')
 194         if uri is None:
 195             webpage = self._download_webpage(url, video_id)
 196
 197             # Some videos come from Vevo.com
 198             m_vevo = re.search(r'isVevoVideo = true;.*?vevoVideoId = "(.*?)";',
 199                                webpage, re.DOTALL)
 200             if m_vevo:
 201                 vevo_id = m_vevo.group(1);
 202                 self.to_screen('Vevo video detected: %s' % vevo_id)
 203                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 204
 205             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 206         return self._get_videos_info(uri)
 207
 208
 209 class MTVIggyIE(MTVServicesInfoExtractor):
 210     IE_NAME = 'mtviggy.com'
 211     _VALID_URL = r'https?://www\.mtviggy\.com/videos/.+'
 212     _TEST = {
 213         'url': 'http://www.mtviggy.com/videos/arcade-fire-behind-the-scenes-at-the-biggest-music-experiment-yet/',
 214         'info_dict': {
 215             'id': '984696',
 216             'ext': 'mp4',
 217             'title': 'Arcade Fire: Behind the Scenes at the Biggest Music Experiment Yet',
 218         }
 219     }
 220     _FEED_URL = 'http://all.mtvworldverticals.com/feed-xml/'