Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/mtv.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .subtitles import SubtitlesInfoExtractor
   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(SubtitlesInfoExtractor):
  27     _MOBILE_TEMPLATE = None
  28
  29     @staticmethod
  30     def _id_from_uri(uri):
  31         return uri.split(':')[-1]
  32
  33     # This was originally implemented for ComedyCentral, but it also works here
  34     @staticmethod
  35     def _transform_rtmp_url(rtmp_video_url):
  36         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
  37         if not m:
  38             return rtmp_video_url
  39         base = 'http://viacommtvstrmfs.fplive.net/'
  40         return base + m.group('finalid')
  41
  42     def _get_feed_url(self, uri):
  43         return self._FEED_URL
  44
  45     def _get_thumbnail_url(self, uri, itemdoc):
  46         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
  47         thumb_node = itemdoc.find(search_path)
  48         if thumb_node is None:
  49             return None
  50         else:
  51             return thumb_node.attrib['url']
  52
  53     def _extract_mobile_video_formats(self, mtvn_id):
  54         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
  55         req = compat_urllib_request.Request(webpage_url)
  56         # Otherwise we get a webpage that would execute some javascript
  57         req.add_header('User-Agent', 'curl/7')
  58         webpage = self._download_webpage(req, mtvn_id,
  59                                          'Downloading mobile page')
  60         metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
  61         req = HEADRequest(metrics_url)
  62         response = self._request_webpage(req, mtvn_id, 'Resolving url')
  63         url = response.geturl()
  64         # Transform the url to get the best quality:
  65         url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
  66         return [{'url': url, 'ext': 'mp4'}]
  67
  68     def _extract_video_formats(self, mdoc, mtvn_id):
  69         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4)$', mdoc.find('.//src').text) is not None:
  70             if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
  71                 self.to_screen('The normal version is not available from your '
  72                                'country, trying with the mobile version')
  73                 return self._extract_mobile_video_formats(mtvn_id)
  74             raise ExtractorError('This video is not available from your country.',
  75                                  expected=True)
  76
  77         formats = []
  78         for rendition in mdoc.findall('.//rendition'):
  79             try:
  80                 _, _, ext = rendition.attrib['type'].partition('/')
  81                 rtmp_video_url = rendition.find('./src').text
  82                 if rtmp_video_url.endswith('siteunavail.png'):
  83                     continue
  84                 formats.append({
  85                     'ext': ext,
  86                     'url': self._transform_rtmp_url(rtmp_video_url),
  87                     'format_id': rendition.get('bitrate'),
  88                     'width': int(rendition.get('width')),
  89                     'height': int(rendition.get('height')),
  90                 })
  91             except (KeyError, TypeError):
  92                 raise ExtractorError('Invalid rendition field.')
  93         self._sort_formats(formats)
  94         return formats
  95
  96     def _extract_subtitles(self, mdoc, mtvn_id):
  97         subtitles = {}
  98         FORMATS = {
  99             'scc': 'cea-608',
 100             'eia-608': 'cea-608',
 101             'xml': 'ttml',
 102         }
 103         subtitles_format = FORMATS.get(
 104             self._downloader.params.get('subtitlesformat'), 'ttml')
 105         for transcript in mdoc.findall('.//transcript'):
 106             if transcript.get('kind') != 'captions':
 107                 continue
 108             lang = transcript.get('srclang')
 109             for typographic in transcript.findall('./typographic'):
 110                 captions_format = typographic.get('format')
 111                 if captions_format == subtitles_format:
 112                     subtitles[lang] = compat_str(typographic.get('src'))
 113                     break
 114         if self._downloader.params.get('listsubtitles', False):
 115             self._list_available_subtitles(mtvn_id, subtitles)
 116         return self.extract_subtitles(mtvn_id, subtitles)
 117
 118     def _get_video_info(self, itemdoc):
 119         uri = itemdoc.find('guid').text
 120         video_id = self._id_from_uri(uri)
 121         self.report_extraction(video_id)
 122         mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
 123         # Remove the templates, like &device={device}
 124         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
 125         if 'acceptMethods' not in mediagen_url:
 126             mediagen_url += '&acceptMethods=fms'
 127
 128         mediagen_doc = self._download_xml(mediagen_url, video_id,
 129                                           'Downloading video urls')
 130
 131         description_node = itemdoc.find('description')
 132         if description_node is not None:
 133             description = description_node.text.strip()
 134         else:
 135             description = None
 136
 137         title_el = None
 138         if title_el is None:
 139             title_el = find_xpath_attr(
 140                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
 141                 'scheme', 'urn:mtvn:video_title')
 142         if title_el is None:
 143             title_el = itemdoc.find('.//{http://search.yahoo.com/mrss/}title')
 144         if title_el is None:
 145             title_el = itemdoc.find('.//title')
 146             if title_el.text is None:
 147                 title_el = None
 148
 149         title = title_el.text
 150         if title is None:
 151             raise ExtractorError('Could not find video title')
 152         title = title.strip()
 153
 154         # This a short id that's used in the webpage urls
 155         mtvn_id = None
 156         mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
 157                                        'scheme', 'urn:mtvn:id')
 158         if mtvn_id_node is not None:
 159             mtvn_id = mtvn_id_node.text
 160
 161         return {
 162             'title': title,
 163             'formats': self._extract_video_formats(mediagen_doc, mtvn_id),
 164             'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id),
 165             'id': video_id,
 166             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
 167             'description': description,
 168         }
 169
 170     def _get_videos_info(self, uri):
 171         video_id = self._id_from_uri(uri)
 172         feed_url = self._get_feed_url(uri)
 173         data = compat_urllib_parse.urlencode({'uri': uri})
 174         idoc = self._download_xml(
 175             feed_url + '?' + data, video_id,
 176             'Downloading info', transform_source=fix_xml_ampersands)
 177         return self.playlist_result(
 178             [self._get_video_info(item) for item in idoc.findall('.//item')])
 179
 180     def _real_extract(self, url):
 181         title = url_basename(url)
 182         webpage = self._download_webpage(url, title)
 183         try:
 184             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
 185             # or http://media.mtvnservices.com/{mgid}
 186             og_url = self._og_search_video_url(webpage)
 187             mgid = url_basename(og_url)
 188             if mgid.endswith('.swf'):
 189                 mgid = mgid[:-4]
 190         except RegexNotFoundError:
 191             mgid = None
 192
 193         if mgid is None or ':' not in mgid:
 194             mgid = self._search_regex(
 195                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
 196                 webpage, 'mgid')
 197
 198         videos_info = self._get_videos_info(mgid)
 199         if self._downloader.params.get('listsubtitles', False):
 200             return
 201         return videos_info
 202
 203
 204 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
 205     IE_NAME = 'mtvservices:embedded'
 206     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
 207
 208     _TEST = {
 209         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
 210         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
 211         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
 212         'info_dict': {
 213             'id': '1043906',
 214             'ext': 'mp4',
 215             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
 216             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
 217         },
 218     }
 219
 220     def _get_feed_url(self, uri):
 221         video_id = self._id_from_uri(uri)
 222         site_id = uri.replace(video_id, '')
 223         config_url = ('http://media.mtvnservices.com/pmt/e1/players/{0}/'
 224                       'context4/context5/config.xml'.format(site_id))
 225         config_doc = self._download_xml(config_url, video_id)
 226         feed_node = config_doc.find('.//feed')
 227         feed_url = feed_node.text.strip().split('?')[0]
 228         return feed_url
 229
 230     def _real_extract(self, url):
 231         mobj = re.match(self._VALID_URL, url)
 232         mgid = mobj.group('mgid')
 233         return self._get_videos_info(mgid)
 234
 235
 236 class MTVIE(MTVServicesInfoExtractor):
 237     _VALID_URL = r'''(?x)^https?://
 238         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
 239            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
 240
 241     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
 242
 243     _TESTS = [
 244         {
 245             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
 246             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
 247             'info_dict': {
 248                 'id': '853555',
 249                 'ext': 'mp4',
 250                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
 251                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
 252             },
 253         },
 254     ]
 255
 256     def _get_thumbnail_url(self, uri, itemdoc):
 257         return 'http://mtv.mtvnimages.com/uri/' + uri
 258
 259     def _real_extract(self, url):
 260         mobj = re.match(self._VALID_URL, url)
 261         video_id = mobj.group('videoid')
 262         uri = mobj.groupdict().get('mgid')
 263         if uri is None:
 264             webpage = self._download_webpage(url, video_id)
 265
 266             # Some videos come from Vevo.com
 267             m_vevo = re.search(
 268                 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
 269             if m_vevo:
 270                 vevo_id = m_vevo.group(1)
 271                 self.to_screen('Vevo video detected: %s' % vevo_id)
 272                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
 273
 274             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
 275         return self._get_videos_info(uri)
 276
 277
 278 class MTVIggyIE(MTVServicesInfoExtractor):
 279     IE_NAME = 'mtviggy.com'
 280     _VALID_URL = r'https?://www\.mtviggy\.com/videos/.+'
 281     _TEST = {
 282         'url': 'http://www.mtviggy.com/videos/arcade-fire-behind-the-scenes-at-the-biggest-music-experiment-yet/',
 283         'info_dict': {
 284             'id': '984696',
 285             'ext': 'mp4',
 286             'title': 'Arcade Fire: Behind the Scenes at the Biggest Music Experiment Yet',
 287         }
 288     }
 289     _FEED_URL = 'http://all.mtvworldverticals.com/feed-xml/'