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