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     compat_str,
  10 )
  11 from ..utils import (
  12     ExtractorError,
  13     find_xpath_attr,
  14     fix_xml_ampersands,
  15     HEADRequest,
  16     unescapeHTML,
  17     url_basename,
  18     RegexNotFoundError,
  19 )
  20
  21
  22 def _media_xml_tag(tag):
  23     return '{http://search.yahoo.com/mrss/}%s' % tag
  24
  25
  26 class MTVServicesInfoExtractor(InfoExtractor):
  27     _MOBILE_TEMPLATE = None
  28     _LANG = None
  29
  30     @staticmethod
  31     def _id_from_uri(uri):
  32         return uri.split(':')[-1]
  33
  34     # This was originally implemented for ComedyCentral, but it also works here
  35     @staticmethod
  36     def _transform_rtmp_url(rtmp_video_url):
  37         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  38         if not m:
  39             return rtmp_video_url
  40         base = 'http://viacommtvstrmfs.fplive.net/'
  41         return base + m.group('finalid')
  42
  43     def _get_feed_url(self, uri):
  44         return self._FEED_URL
  45
  46     def _get_thumbnail_url(self, uri, itemdoc):
  47         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  48         thumb_node = itemdoc.find(search_path)
  49         if thumb_node is None:
  50             return None
  51         else:
  52             return thumb_node.attrib['url']
  53
  54     def _extract_mobile_video_formats(self, mtvn_id):
  55         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
  56         req = compat_urllib_request.Request(webpage_url)
  57         # Otherwise we get a webpage that would execute some javascript
  58         req.add_header('User-Agent', 'curl/7')
  59         webpage = self._download_webpage(req, mtvn_id,
  60                                          'Downloading mobile page')
  61         metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
  62         req = HEADRequest(metrics_url)
  63         response = self._request_webpage(req, mtvn_id, 'Resolving url')
  64         url = response.geturl()
  65         # Transform the url to get the best quality:
  66         url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
  67         return [{'url': url, 'ext': 'mp4'}]
  68
  69     def _extract_video_formats(self, mdoc, mtvn_id):
  70         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4)$', mdoc.find('.//src').text) is not None:
  71             if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
  72                 self.to_screen('The normal version is not available from your '
  73                                'country, trying with the mobile version')
  74                 return self._extract_mobile_video_formats(mtvn_id)
  75             raise ExtractorError('This video is not available from your country.',
  76                                  expected=True)
  77
  78         formats = []
  79         for rendition in mdoc.findall('.//rendition'):
  80             try:
  81                 _, _, ext = rendition.attrib['type'].partition('/')
  82                 rtmp_video_url = rendition.find('./src').text
  83                 if rtmp_video_url.endswith('siteunavail.png'):
  84                     continue
  85                 formats.append({
  86                     'ext': ext,
  87                     'url': self._transform_rtmp_url(rtmp_video_url),
  88                     'format_id': rendition.get('bitrate'),
  89                     'width': int(rendition.get('width')),
  90                     'height': int(rendition.get('height')),
  91                 })
  92             except (KeyError, TypeError):
  93                 raise ExtractorError('Invalid rendition field.')
  94         self._sort_formats(formats)
  95         return formats
  96
  97     def _extract_subtitles(self, mdoc, mtvn_id):
  98         subtitles = {}
  99         for transcript in mdoc.findall('.//transcript'):
 100             if transcript.get('kind') != 'captions':
 101                 continue
 102             lang = transcript.get('srclang')
 103             subtitles[lang] = [{
 104                 'url': compat_str(typographic.get('src')),
 105                 'ext': typographic.get('format')
 106             } for typographic in transcript.findall('./typographic')]
 107         return subtitles
 108
 109     def _get_video_info(self, itemdoc):
 110         uri = itemdoc.find('guid').text
 111         video_id = self._id_from_uri(uri)
 112         self.report_extraction(video_id)
 113         mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
 114         # Remove the templates, like &device={device}
 115         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
 116         if 'acceptMethods' not in mediagen_url:
 117             mediagen_url += '&acceptMethods=fms'
 118
 119         mediagen_doc = self._download_xml(mediagen_url, video_id,
 120                                           'Downloading video urls')
 121
 122         item = mediagen_doc.find('./video/item')
 123         if item is not None and item.get('type') == 'text':
 124             message = '%s returned error: ' % self.IE_NAME
 125             if item.get('code') is not None:
 126                 message += '%s - ' % item.get('code')
 127             message += item.text
 128             raise ExtractorError(message, expected=True)
 129
 130         description_node = itemdoc.find('description')
 131         if description_node is not None:
 132             description = description_node.text.strip()
 133         else:
 134             description = None
 135
 136         title_el = None
 137         if title_el is None:
 138             title_el = find_xpath_attr(
 139                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 140                 'scheme', 'urn:mtvn:video_title')
 141         if title_el is None:
 142             title_el = itemdoc.find('.//{http://search.yahoo.com/mrss/}title')
 143         if title_el is None:
 144             title_el = itemdoc.find('.//title')
 145             if title_el.text is None:
 146                 title_el = None
 147
 148         title = title_el.text
 149         if title is None:
 150             raise ExtractorError('Could not find video title')
 151         title = title.strip()
 152
 153         # This a short id that's used in the webpage urls
 154         mtvn_id = None
 155         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 156                                        'scheme', 'urn:mtvn:id')
 157         if mtvn_id_node is not None:
 158             mtvn_id = mtvn_id_node.text
 159
 160         return {
 161             'title': title,
 162             'formats': self._extract_video_formats(mediagen_doc, mtvn_id),
 163             'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id),
 164             'id': video_id,
 165             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 166             'description': description,
 167         }
 168
 169     def _get_videos_info(self, uri):
 170         video_id = self._id_from_uri(uri)
 171         feed_url = self._get_feed_url(uri)
 172         data = compat_urllib_parse.urlencode({'uri': uri})
 173         info_url = feed_url + '?'
 174         if self._LANG:
 175             info_url += 'lang=%s&' % self._LANG
 176         info_url += data
 177         idoc = self._download_xml(
 178             info_url, video_id,
 179             'Downloading info', transform_source=fix_xml_ampersands)
 180         return self.playlist_result(
 181             [self._get_video_info(item) for item in idoc.findall('.//item')])
 182
 183     def _real_extract(self, url):
 184         title = url_basename(url)
 185         webpage = self._download_webpage(url, title)
 186         try:
 187             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 188             # or http://media.mtvnservices.com/{mgid}
 189             og_url = self._og_search_video_url(webpage)
 190             mgid = url_basename(og_url)
 191             if mgid.endswith('.swf'):
 192                 mgid = mgid[:-4]
 193         except RegexNotFoundError:
 194             mgid = None
 195
 196         if mgid is None or ':' not in mgid:
 197             mgid = self._search_regex(
 198                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
 199                 webpage, 'mgid')
 200
 201         videos_info = self._get_videos_info(mgid)
 202         return videos_info
 203
 204
 205 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
 206     IE_NAME = 'mtvservices:embedded'
 207     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
 208
 209     _TEST = {
 210         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
 211         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
 212         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
 213         'info_dict': {
 214             'id': '1043906',
 215             'ext': 'mp4',
 216             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
 217             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
 218         },
 219     }
 220
 221     def _get_feed_url(self, uri):
 222         video_id = self._id_from_uri(uri)
 223         site_id = uri.replace(video_id, '')
 224         config_url = ('http://media.mtvnservices.com/pmt/e1/players/{0}/'
 225                       'context4/context5/config.xml'.format(site_id))
 226         config_doc = self._download_xml(config_url, video_id)
 227         feed_node = config_doc.find('.//feed')
 228         feed_url = feed_node.text.strip().split('?')[0]
 229         return feed_url
 230
 231     def _real_extract(self, url):
 232         mobj = re.match(self._VALID_URL, url)
 233         mgid = mobj.group('mgid')
 234         return self._get_videos_info(mgid)
 235
 236
 237 class MTVIE(MTVServicesInfoExtractor):
 238     _VALID_URL = r'''(?x)^https?://
 239         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 240            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 241
 242     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 243
 244     _TESTS = [
 245         {
 246             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 247             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 248             'info_dict': {
 249                 'id': '853555',
 250                 'ext': 'mp4',
 251                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 252                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 253             },
 254         },
 255     ]
 256
 257     def _get_thumbnail_url(self, uri, itemdoc):
 258         return 'http://mtv.mtvnimages.com/uri/' + uri
 259
 260     def _real_extract(self, url):
 261         mobj = re.match(self._VALID_URL, url)
 262         video_id = mobj.group('videoid')
 263         uri = mobj.groupdict().get('mgid')
 264         if uri is None:
 265             webpage = self._download_webpage(url, video_id)
 266
 267             # Some videos come from Vevo.com
 268             m_vevo = re.search(
 269                 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
 270             if m_vevo:
 271                 vevo_id = m_vevo.group(1)
 272                 self.to_screen('Vevo video detected: %s' % vevo_id)
 273                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 274
 275             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 276         return self._get_videos_info(uri)
 277
 278
 279 class MTVIggyIE(MTVServicesInfoExtractor):
 280     IE_NAME = 'mtviggy.com'
 281     _VALID_URL = r'https?://www\.mtviggy\.com/videos/.+'
 282     _TEST = {
 283         'url': 'http://www.mtviggy.com/videos/arcade-fire-behind-the-scenes-at-the-biggest-music-experiment-yet/',
 284         'info_dict': {
 285             'id': '984696',
 286             'ext': 'mp4',
 287             'title': 'Arcade Fire: Behind the Scenes at the Biggest Music Experiment Yet',
 288         }
 289     }
 290     _FEED_URL = 'http://all.mtvworldverticals.com/feed-xml/'