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

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import json
   6
   7 from .common import InfoExtractor
   8 from ..utils import (
   9     compat_urllib_parse,
  10     orderedSet,
  11     compat_urllib_parse_urlparse,
  12     compat_urlparse,
  13 )
  14
  15
  16 class CondeNastIE(InfoExtractor):
  17     """
  18     Condé Nast is a media group, some of its sites use a custom HTML5 player
  19     that works the same in all of them.
  20     """
  21
  22     # The keys are the supported sites and the values are the name to be shown
  23     # to the user and in the extractor description.
  24     _SITES = {
  25         'wired': 'WIRED',
  26         'gq': 'GQ',
  27         'vogue': 'Vogue',
  28         'glamour': 'Glamour',
  29         'wmagazine': 'W Magazine',
  30         'vanityfair': 'Vanity Fair',
  31         'cnevids': 'Condé Nast',
  32     }
  33
  34     _VALID_URL = r'http://(video|www|player)\.(?P<site>%s)\.com/(?P<type>watch|series|video|embed)/(?P<id>[^/?#]+)' % '|'.join(_SITES.keys())
  35     IE_DESC = 'Condé Nast media group: %s' % ', '.join(sorted(_SITES.values()))
  36
  37     _TEST = {
  38         'url': 'http://video.wired.com/watch/3d-printed-speakers-lit-with-led',
  39         'md5': '1921f713ed48aabd715691f774c451f7',
  40         'info_dict': {
  41             'id': '5171b343c2b4c00dd0c1ccb3',
  42             'ext': 'mp4',
  43             'title': '3D Printed Speakers Lit With LED',
  44             'description': '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.',
  45         }
  46     }
  47
  48     def _extract_series(self, url, webpage):
  49         title = self._html_search_regex(r'<div class="cne-series-info">.*?<h1>(.+?)</h1>',
  50                                         webpage, 'series title', flags=re.DOTALL)
  51         url_object = compat_urllib_parse_urlparse(url)
  52         base_url = '%s://%s' % (url_object.scheme, url_object.netloc)
  53         m_paths = re.finditer(r'<p class="cne-thumb-title">.*?<a href="(/watch/.+?)["\?]',
  54                               webpage, flags=re.DOTALL)
  55         paths = orderedSet(m.group(1) for m in m_paths)
  56         build_url = lambda path: compat_urlparse.urljoin(base_url, path)
  57         entries = [self.url_result(build_url(path), 'CondeNast') for path in paths]
  58         return self.playlist_result(entries, playlist_title=title)
  59
  60     def _extract_video(self, webpage, url_type):
  61         if url_type != 'embed':
  62             description = self._html_search_regex(
  63                 [
  64                     r'<div class="cne-video-description">(.+?)</div>',
  65                     r'<div class="video-post-content">(.+?)</div>',
  66                 ],
  67                 webpage, 'description', fatal=False, flags=re.DOTALL)
  68         else:
  69             description = None
  70         params = self._search_regex(r'var params = {(.+?)}[;,]', webpage,
  71                                     'player params', flags=re.DOTALL)
  72         video_id = self._search_regex(r'videoId: [\'"](.+?)[\'"]', params, 'video id')
  73         player_id = self._search_regex(r'playerId: [\'"](.+?)[\'"]', params, 'player id')
  74         target = self._search_regex(r'target: [\'"](.+?)[\'"]', params, 'target')
  75         data = compat_urllib_parse.urlencode({'videoId': video_id,
  76                                               'playerId': player_id,
  77                                               'target': target,
  78                                               })
  79         base_info_url = self._search_regex(r'url = [\'"](.+?)[\'"][,;]',
  80                                            webpage, 'base info url',
  81                                            default='http://player.cnevids.com/player/loader.js?')
  82         info_url = base_info_url + data
  83         info_page = self._download_webpage(info_url, video_id,
  84                                            'Downloading video info')
  85         video_info = self._search_regex(r'var video = ({.+?});', info_page, 'video info')
  86         video_info = json.loads(video_info)
  87
  88         formats = [{
  89             'format_id': '%s-%s' % (fdata['type'].split('/')[-1], fdata['quality']),
  90             'url': fdata['src'],
  91             'ext': fdata['type'].split('/')[-1],
  92             'quality': 1 if fdata['quality'] == 'high' else 0,
  93         } for fdata in video_info['sources'][0]]
  94         self._sort_formats(formats)
  95
  96         return {
  97             'id': video_id,
  98             'formats': formats,
  99             'title': video_info['title'],
 100             'thumbnail': video_info['poster_frame'],
 101             'description': description,
 102         }
 103
 104     def _real_extract(self, url):
 105         mobj = re.match(self._VALID_URL, url)
 106         site = mobj.group('site')
 107         url_type = mobj.group('type')
 108         item_id = mobj.group('id')
 109
 110         self.to_screen('Extracting from %s with the Condé Nast extractor' % self._SITES[site])
 111         webpage = self._download_webpage(url, item_id)
 112
 113         if url_type == 'series':
 114             return self._extract_series(url, webpage)
 115         else:
 116             return self._extract_video(webpage, url_type)