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