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

   1 # coding: utf-8
   2
   3 import re
   4 import json
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     compat_urllib_parse,
   9     orderedSet,
  10     compat_urllib_parse_urlparse,
  11     compat_urlparse,
  12 )
  13
  14
  15 class CondeNastIE(InfoExtractor):
  16     """
  17     Condé Nast is a media group, some of its sites use a custom HTML5 player
  18     that works the same in all of them.
  19     """
  20
  21     # The keys are the supported sites and the values are the name to be shown
  22     # to the user and in the extractor description.
  23     _SITES = {'wired': u'WIRED',
  24               'gq': u'GQ',
  25               'vogue': u'Vogue',
  26               'glamour': u'Glamour',
  27               'wmagazine': u'W Magazine',
  28               'vanityfair': u'Vanity Fair',
  29               }
  30
  31     _VALID_URL = r'http://(video|www).(?P<site>%s).com/(?P<type>watch|series|video)/(?P<id>.+)' % '|'.join(_SITES.keys())
  32     IE_DESC = u'Condé Nast media group: %s' % ', '.join(sorted(_SITES.values()))
  33
  34     _TEST = {
  35         u'url': u'http://video.wired.com/watch/3d-printed-speakers-lit-with-led',
  36         u'file': u'5171b343c2b4c00dd0c1ccb3.mp4',
  37         u'md5': u'1921f713ed48aabd715691f774c451f7',
  38         u'info_dict': {
  39             u'title': u'3D Printed Speakers Lit With LED',
  40             u'description': u'Check out these beautiful 3D printed LED speakers.  You can\'t actually buy them, but LumiGeek is working on a board that will let you make you\'re own.',
  41         }
  42     }
  43
  44     def _extract_series(self, url, webpage):
  45         title = self._html_search_regex(r'<div class="cne-series-info">.*?<h1>(.+?)</h1>',
  46                                         webpage, u'series title', flags=re.DOTALL)
  47         url_object = compat_urllib_parse_urlparse(url)
  48         base_url = '%s://%s' % (url_object.scheme, url_object.netloc)
  49         m_paths = re.finditer(r'<p class="cne-thumb-title">.*?<a href="(/watch/.+?)["\?]',
  50                               webpage, flags=re.DOTALL)
  51         paths = orderedSet(m.group(1) for m in m_paths)
  52         build_url = lambda path: compat_urlparse.urljoin(base_url, path)
  53         entries = [self.url_result(build_url(path), 'CondeNast') for path in paths]
  54         return self.playlist_result(entries, playlist_title=title)
  55
  56     def _extract_video(self, webpage):
  57         description = self._html_search_regex([r'<div class="cne-video-description">(.+?)</div>',
  58                                                r'<div class="video-post-content">(.+?)</div>',
  59                                                ],
  60                                               webpage, u'description',
  61                                               fatal=False, flags=re.DOTALL)
  62         params = self._search_regex(r'var params = {(.+?)}[;,]', webpage,
  63                                     u'player params', flags=re.DOTALL)
  64         video_id = self._search_regex(r'videoId: [\'"](.+?)[\'"]', params, u'video id')
  65         player_id = self._search_regex(r'playerId: [\'"](.+?)[\'"]', params, u'player id')
  66         target = self._search_regex(r'target: [\'"](.+?)[\'"]', params, u'target')
  67         data = compat_urllib_parse.urlencode({'videoId': video_id,
  68                                               'playerId': player_id,
  69                                               'target': target,
  70                                               })
  71         base_info_url = self._search_regex(r'url = [\'"](.+?)[\'"][,;]',
  72                                            webpage, u'base info url',
  73                                            default='http://player.cnevids.com/player/loader.js?')
  74         info_url = base_info_url + data
  75         info_page = self._download_webpage(info_url, video_id,
  76                                            u'Downloading video info')
  77         video_info = self._search_regex(r'var video = ({.+?});', info_page, u'video info')
  78         video_info = json.loads(video_info)
  79
  80         def _formats_sort_key(f):
  81             type_ord = 1 if f['type'] == 'video/mp4' else 0
  82             quality_ord = 1 if f['quality'] == 'high' else 0
  83             return (quality_ord, type_ord)
  84         best_format = sorted(video_info['sources'][0], key=_formats_sort_key)[-1]
  85
  86         return {'id': video_id,
  87                 'url': best_format['src'],
  88                 'ext': best_format['type'].split('/')[-1],
  89                 'title': video_info['title'],
  90                 'thumbnail': video_info['poster_frame'],
  91                 'description': description,
  92                 }
  93
  94     def _real_extract(self, url):
  95         mobj = re.match(self._VALID_URL, url)
  96         site = mobj.group('site')
  97         url_type = mobj.group('type')
  98         id = mobj.group('id')
  99
 100         self.to_screen(u'Extracting from %s with the Condé Nast extractor' % self._SITES[site])
 101         webpage = self._download_webpage(url, id)
 102
 103         if url_type == 'series':
 104             return self._extract_series(url, webpage)
 105         else:
 106             return self._extract_video(webpage)