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     compat_urllib_request,
   9     ExtractorError,
  10     find_xpath_attr,
  11     fix_xml_ampersands,
  12     HEADRequest,
  13     unescapeHTML,
  14     url_basename,
  15     RegexNotFoundError,
  16 )
  17
  18
  19 def _media_xml_tag(tag):
  20     return '{http://search.yahoo.com/mrss/}%s' % tag
  21
  22
  23 class MTVServicesInfoExtractor(InfoExtractor):
  24     _MOBILE_TEMPLATE = None
  25     @staticmethod
  26     def _id_from_uri(uri):
  27         return uri.split(':')[-1]
  28
  29     # This was originally implemented for ComedyCentral, but it also works here
  30     @staticmethod
  31     def _transform_rtmp_url(rtmp_video_url):
  32         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  33         if not m:
  34             return rtmp_video_url
  35         base = 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=1+_pxI0=Ripod-h264+_pxL0=undefined+_pxM0=+_pxK=18639+_pxE=mp4/44620/mtvnorigin/'
  36         return base + m.group('finalid')
  37
  38     def _get_thumbnail_url(self, uri, itemdoc):
  39         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  40         thumb_node = itemdoc.find(search_path)
  41         if thumb_node is None:
  42             return None
  43         else:
  44             return thumb_node.attrib['url']
  45
  46     def _extract_mobile_video_formats(self, mtvn_id):
  47         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
  48         req = compat_urllib_request.Request(webpage_url)
  49         # Otherwise we get a webpage that would execute some javascript
  50         req.add_header('Youtubedl-user-agent', 'curl/7')
  51         webpage = self._download_webpage(req, mtvn_id,
  52             'Downloading mobile page')
  53         metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
  54         req = HEADRequest(metrics_url)
  55         response = self._request_webpage(req, mtvn_id, 'Resolving url')
  56         url = response.geturl()
  57         # Transform the url to get the best quality:
  58         url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
  59         return [{'url': url,'ext': 'mp4'}]
  60
  61     def _extract_video_formats(self, mdoc, mtvn_id):
  62         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4)$', mdoc.find('.//src').text) is not None:
  63             if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
  64                 self.to_screen('The normal version is not available from your '
  65                     'country, trying with the mobile version')
  66                 return self._extract_mobile_video_formats(mtvn_id)
  67             raise ExtractorError('This video is not available from your country.',
  68                 expected=True)
  69
  70         formats = []
  71         for rendition in mdoc.findall('.//rendition'):
  72             try:
  73                 _, _, ext = rendition.attrib['type'].partition('/')
  74                 rtmp_video_url = rendition.find('./src').text
  75                 formats.append({'ext': ext,
  76                                 'url': self._transform_rtmp_url(rtmp_video_url),
  77                                 'format_id': rendition.get('bitrate'),
  78                                 'width': int(rendition.get('width')),
  79                                 'height': int(rendition.get('height')),
  80                                 })
  81             except (KeyError, TypeError):
  82                 raise ExtractorError('Invalid rendition field.')
  83         self._sort_formats(formats)
  84         return formats
  85
  86     def _get_video_info(self, itemdoc):
  87         uri = itemdoc.find('guid').text
  88         video_id = self._id_from_uri(uri)
  89         self.report_extraction(video_id)
  90         mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
  91         # Remove the templates, like &device={device}
  92         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
  93         if 'acceptMethods' not in mediagen_url:
  94             mediagen_url += '&acceptMethods=fms'
  95
  96         mediagen_doc = self._download_xml(mediagen_url, video_id,
  97             'Downloading video urls')
  98
  99         description_node = itemdoc.find('description')
 100         if description_node is not None:
 101             description = description_node.text.strip()
 102         else:
 103             description = None
 104
 105         title_el = None
 106         if title_el is None:
 107             title_el = find_xpath_attr(
 108                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 109                 'scheme', 'urn:mtvn:video_title')
 110         if title_el is None:
 111             title_el = itemdoc.find('.//{http://search.yahoo.com/mrss/}title')
 112         if title_el is None:
 113             title_el = itemdoc.find('.//title')
 114             if title_el.text is None:
 115                 title_el = None
 116
 117         title = title_el.text
 118         if title is None:
 119             raise ExtractorError('Could not find video title')
 120         title = title.strip()
 121
 122         # This a short id that's used in the webpage urls
 123         mtvn_id = None
 124         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 125                 'scheme', 'urn:mtvn:id')
 126         if mtvn_id_node is not None:
 127             mtvn_id = mtvn_id_node.text
 128
 129         return {
 130             'title': title,
 131             'formats': self._extract_video_formats(mediagen_doc, mtvn_id),
 132             'id': video_id,
 133             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 134             'description': description,
 135         }
 136
 137     def _get_videos_info(self, uri):
 138         video_id = self._id_from_uri(uri)
 139         data = compat_urllib_parse.urlencode({'uri': uri})
 140
 141         idoc = self._download_xml(
 142             self._FEED_URL + '?' + data, video_id,
 143             'Downloading info', transform_source=fix_xml_ampersands)
 144         return [self._get_video_info(item) for item in idoc.findall('.//item')]
 145
 146     def _real_extract(self, url):
 147         title = url_basename(url)
 148         webpage = self._download_webpage(url, title)
 149         try:
 150             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 151             # or http://media.mtvnservices.com/{mgid}
 152             og_url = self._og_search_video_url(webpage)
 153             mgid = url_basename(og_url)
 154             if mgid.endswith('.swf'):
 155                 mgid = mgid[:-4]
 156         except RegexNotFoundError:
 157             mgid = self._search_regex(
 158                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
 159                 webpage, u'mgid')
 160         return self._get_videos_info(mgid)
 161
 162
 163 class MTVIE(MTVServicesInfoExtractor):
 164     _VALID_URL = r'''(?x)^https?://
 165         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 166            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 167
 168     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 169
 170     _TESTS = [
 171         {
 172             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 173             'file': '853555.mp4',
 174             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 175             'info_dict': {
 176                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 177                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 178             },
 179         },
 180         {
 181             'add_ie': ['Vevo'],
 182             'url': 'http://www.mtv.com/videos/taylor-swift/916187/everything-has-changed-ft-ed-sheeran.jhtml',
 183             'file': 'USCJY1331283.mp4',
 184             'md5': '73b4e7fcadd88929292fe52c3ced8caf',
 185             'info_dict': {
 186                 'title': 'Everything Has Changed',
 187                 'upload_date': '20130606',
 188                 'uploader': 'Taylor Swift',
 189             },
 190             'skip': 'VEVO is only available in some countries',
 191         },
 192     ]
 193
 194     def _get_thumbnail_url(self, uri, itemdoc):
 195         return 'http://mtv.mtvnimages.com/uri/' + uri
 196
 197     def _real_extract(self, url):
 198         mobj = re.match(self._VALID_URL, url)
 199         video_id = mobj.group('videoid')
 200         uri = mobj.groupdict().get('mgid')
 201         if uri is None:
 202             webpage = self._download_webpage(url, video_id)
 203
 204             # Some videos come from Vevo.com
 205             m_vevo = re.search(r'isVevoVideo = true;.*?vevoVideoId = "(.*?)";',
 206                                webpage, re.DOTALL)
 207             if m_vevo:
 208                 vevo_id = m_vevo.group(1);
 209                 self.to_screen('Vevo video detected: %s' % vevo_id)
 210                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 211
 212             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 213         return self._get_videos_info(uri)
 214
 215
 216 class MTVIggyIE(MTVServicesInfoExtractor):
 217     IE_NAME = 'mtviggy.com'
 218     _VALID_URL = r'https?://www\.mtviggy\.com/videos/.+'
 219     _TEST = {
 220         'url': 'http://www.mtviggy.com/videos/arcade-fire-behind-the-scenes-at-the-biggest-music-experiment-yet/',
 221         'info_dict': {
 222             'id': '984696',
 223             'ext': 'mp4',
 224             'title': 'Arcade Fire: Behind the Scenes at the Biggest Music Experiment Yet',
 225         }
 226     }
 227     _FEED_URL = 'http://all.mtvworldverticals.com/feed-xml/'