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 ..compat import (
   7     compat_urllib_parse,
   8     compat_urllib_request,
   9 )
  10 from ..utils import (
  11     ExtractorError,
  12     find_xpath_attr,
  13     fix_xml_ampersands,
  14     HEADRequest,
  15     unescapeHTML,
  16     url_basename,
  17     RegexNotFoundError,
  18 )
  19
  20
  21 def _media_xml_tag(tag):
  22     return '{http://search.yahoo.com/mrss/}%s' % tag
  23
  24
  25 class MTVServicesInfoExtractor(InfoExtractor):
  26     _MOBILE_TEMPLATE = None
  27
  28     @staticmethod
  29     def _id_from_uri(uri):
  30         return uri.split(':')[-1]
  31
  32     # This was originally implemented for ComedyCentral, but it also works here
  33     @staticmethod
  34     def _transform_rtmp_url(rtmp_video_url):
  35         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  36         if not m:
  37             return rtmp_video_url
  38         base = 'http://viacommtvstrmfs.fplive.net/'
  39         return base + m.group('finalid')
  40
  41     def _get_feed_url(self, uri):
  42         return self._FEED_URL
  43
  44     def _get_thumbnail_url(self, uri, itemdoc):
  45         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  46         thumb_node = itemdoc.find(search_path)
  47         if thumb_node is None:
  48             return None
  49         else:
  50             return thumb_node.attrib['url']
  51
  52     def _extract_mobile_video_formats(self, mtvn_id):
  53         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
  54         req = compat_urllib_request.Request(webpage_url)
  55         # Otherwise we get a webpage that would execute some javascript
  56         req.add_header('Youtubedl-user-agent', 'curl/7')
  57         webpage = self._download_webpage(req, mtvn_id,
  58                                          'Downloading mobile page')
  59         metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
  60         req = HEADRequest(metrics_url)
  61         response = self._request_webpage(req, mtvn_id, 'Resolving url')
  62         url = response.geturl()
  63         # Transform the url to get the best quality:
  64         url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
  65         return [{'url': url, 'ext': 'mp4'}]
  66
  67     def _extract_video_formats(self, mdoc, mtvn_id):
  68         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4)$', mdoc.find('.//src').text) is not None:
  69             if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
  70                 self.to_screen('The normal version is not available from your '
  71                                'country, trying with the mobile version')
  72                 return self._extract_mobile_video_formats(mtvn_id)
  73             raise ExtractorError('This video is not available from your country.',
  74                                  expected=True)
  75
  76         formats = []
  77         for rendition in mdoc.findall('.//rendition'):
  78             try:
  79                 _, _, ext = rendition.attrib['type'].partition('/')
  80                 rtmp_video_url = rendition.find('./src').text
  81                 formats.append({'ext': ext,
  82                                 'url': self._transform_rtmp_url(rtmp_video_url),
  83                                 'format_id': rendition.get('bitrate'),
  84                                 'width': int(rendition.get('width')),
  85                                 'height': int(rendition.get('height')),
  86                                 })
  87             except (KeyError, TypeError):
  88                 raise ExtractorError('Invalid rendition field.')
  89         self._sort_formats(formats)
  90         return formats
  91
  92     def _get_video_info(self, itemdoc):
  93         uri = itemdoc.find('guid').text
  94         video_id = self._id_from_uri(uri)
  95         self.report_extraction(video_id)
  96         mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
  97         # Remove the templates, like &device={device}
  98         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
  99         if 'acceptMethods' not in mediagen_url:
 100             mediagen_url += '&acceptMethods=fms'
 101
 102         mediagen_doc = self._download_xml(mediagen_url, video_id,
 103                                           'Downloading video urls')
 104
 105         description_node = itemdoc.find('description')
 106         if description_node is not None:
 107             description = description_node.text.strip()
 108         else:
 109             description = None
 110
 111         title_el = None
 112         if title_el is None:
 113             title_el = find_xpath_attr(
 114                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 115                 'scheme', 'urn:mtvn:video_title')
 116         if title_el is None:
 117             title_el = itemdoc.find('.//{http://search.yahoo.com/mrss/}title')
 118         if title_el is None:
 119             title_el = itemdoc.find('.//title')
 120             if title_el.text is None:
 121                 title_el = None
 122
 123         title = title_el.text
 124         if title is None:
 125             raise ExtractorError('Could not find video title')
 126         title = title.strip()
 127
 128         # This a short id that's used in the webpage urls
 129         mtvn_id = None
 130         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 131                                        'scheme', 'urn:mtvn:id')
 132         if mtvn_id_node is not None:
 133             mtvn_id = mtvn_id_node.text
 134
 135         return {
 136             'title': title,
 137             'formats': self._extract_video_formats(mediagen_doc, mtvn_id),
 138             'id': video_id,
 139             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 140             'description': description,
 141         }
 142
 143     def _get_videos_info(self, uri):
 144         video_id = self._id_from_uri(uri)
 145         feed_url = self._get_feed_url(uri)
 146         data = compat_urllib_parse.urlencode({'uri': uri})
 147         idoc = self._download_xml(
 148             feed_url + '?' + data, video_id,
 149             'Downloading info', transform_source=fix_xml_ampersands)
 150         return self.playlist_result(
 151             [self._get_video_info(item) for item in idoc.findall('.//item')])
 152
 153     def _real_extract(self, url):
 154         title = url_basename(url)
 155         webpage = self._download_webpage(url, title)
 156         try:
 157             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 158             # or http://media.mtvnservices.com/{mgid}
 159             og_url = self._og_search_video_url(webpage)
 160             mgid = url_basename(og_url)
 161             if mgid.endswith('.swf'):
 162                 mgid = mgid[:-4]
 163         except RegexNotFoundError:
 164             mgid = None
 165
 166         if mgid is None or ':' not in mgid:
 167             mgid = self._search_regex(
 168                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
 169                 webpage, 'mgid')
 170         return self._get_videos_info(mgid)
 171
 172
 173 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
 174     IE_NAME = 'mtvservices:embedded'
 175     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
 176
 177     _TEST = {
 178         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
 179         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
 180         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
 181         'info_dict': {
 182             'id': '1043906',
 183             'ext': 'mp4',
 184             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
 185             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
 186         },
 187     }
 188
 189     def _get_feed_url(self, uri):
 190         video_id = self._id_from_uri(uri)
 191         site_id = uri.replace(video_id, '')
 192         config_url = ('http://media.mtvnservices.com/pmt/e1/players/{0}/'
 193                       'context4/context5/config.xml'.format(site_id))
 194         config_doc = self._download_xml(config_url, video_id)
 195         feed_node = config_doc.find('.//feed')
 196         feed_url = feed_node.text.strip().split('?')[0]
 197         return feed_url
 198
 199     def _real_extract(self, url):
 200         mobj = re.match(self._VALID_URL, url)
 201         mgid = mobj.group('mgid')
 202         return self._get_videos_info(mgid)
 203
 204
 205 class MTVIE(MTVServicesInfoExtractor):
 206     _VALID_URL = r'''(?x)^https?://
 207         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 208            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 209
 210     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 211
 212     _TESTS = [
 213         {
 214             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 215             'file': '853555.mp4',
 216             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 217             'info_dict': {
 218                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 219                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 220             },
 221         },
 222         {
 223             'add_ie': ['Vevo'],
 224             'url': 'http://www.mtv.com/videos/taylor-swift/916187/everything-has-changed-ft-ed-sheeran.jhtml',
 225             'file': 'USCJY1331283.mp4',
 226             'md5': '73b4e7fcadd88929292fe52c3ced8caf',
 227             'info_dict': {
 228                 'title': 'Everything Has Changed',
 229                 'upload_date': '20130606',
 230                 'uploader': 'Taylor Swift',
 231             },
 232             'skip': 'VEVO is only available in some countries',
 233         },
 234     ]
 235
 236     def _get_thumbnail_url(self, uri, itemdoc):
 237         return 'http://mtv.mtvnimages.com/uri/' + uri
 238
 239     def _real_extract(self, url):
 240         mobj = re.match(self._VALID_URL, url)
 241         video_id = mobj.group('videoid')
 242         uri = mobj.groupdict().get('mgid')
 243         if uri is None:
 244             webpage = self._download_webpage(url, video_id)
 245
 246             # Some videos come from Vevo.com
 247             m_vevo = re.search(r'isVevoVideo = true;.*?vevoVideoId = "(.*?)";',
 248                                webpage, re.DOTALL)
 249             if m_vevo:
 250                 vevo_id = m_vevo.group(1)
 251                 self.to_screen('Vevo video detected: %s' % vevo_id)
 252                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 253
 254             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 255         return self._get_videos_info(uri)
 256
 257
 258 class MTVIggyIE(MTVServicesInfoExtractor):
 259     IE_NAME = 'mtviggy.com'
 260     _VALID_URL = r'https?://www\.mtviggy\.com/videos/.+'
 261     _TEST = {
 262         'url': 'http://www.mtviggy.com/videos/arcade-fire-behind-the-scenes-at-the-biggest-music-experiment-yet/',
 263         'info_dict': {
 264             'id': '984696',
 265             'ext': 'mp4',
 266             'title': 'Arcade Fire: Behind the Scenes at the Biggest Music Experiment Yet',
 267         }
 268     }
 269     _FEED_URL = 'http://all.mtvworldverticals.com/feed-xml/'