_ Git - youtube-dl/blob - youtube_dl/extractor/condenast.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..compat import (
   8     compat_urllib_parse,
   9     compat_urllib_parse_urlparse,
  10     compat_urlparse,
  11 )
  12 from ..utils import (
  13     orderedSet,
  14 )
  15
  16
  17 class CondeNastIE(InfoExtractor):
  18     """
  19     Condé Nast is a media group, some of its sites use a custom HTML5 player
  20     that works the same in all of them.
  21     """
  22
  23     # The keys are the supported sites and the values are the name to be shown
  24     # to the user and in the extractor description.
  25     _SITES = {
  26         'allure': 'Allure',
  27         'architecturaldigest': 'Architectural Digest',
  28         'arstechnica': 'Ars Technica',
  29         'bonappetit': 'Bon Appetit',
  30         'brides': 'Brides',
  31         'cnevids': 'Condé Nast',
  32         'cntraveler': 'Condé Nast Traveler',
  33         'details': 'Details',
  34         'epicurious': 'Epicurious',
  35         'glamour': 'Glamour',
  36         'golfdigest': 'Golf Digest',
  37         'gq': 'GQ',
  38         'newyorker': 'The New Yorker',
  39         'self': 'SELF',
  40         'teenvogue': 'Teen Vogue',
  41         'vanityfair': 'Vanity Fair',
  42         'vogue': 'Vogue',
  43         'wired': 'WIRED',
  44         'wmagazine': 'W Magazine',
  45     }
  46
  47     _VALID_URL = r'http://(video|www|player)\.(?P<site>%s)\.com/(?P<type>watch|series|video|embed)/(?P<id>[^/?#]+)' % '|'.join(_SITES.keys())
  48     IE_DESC = 'Condé Nast media group: %s' % ', '.join(sorted(_SITES.values()))
  49
  50     EMBED_URL = r'(?:https?:)?//player\.(?P<site>%s)\.com/(?P<type>embed)/.+?' % '|'.join(_SITES.keys())
  51
  52     _TEST = {
  53         'url': 'http://video.wired.com/watch/3d-printed-speakers-lit-with-led',
  54         'md5': '1921f713ed48aabd715691f774c451f7',
  55         'info_dict': {
  56             'id': '5171b343c2b4c00dd0c1ccb3',
  57             'ext': 'mp4',
  58             'title': '3D Printed Speakers Lit With LED',
  59             '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.',
  60         }
  61     }
  62
  63     def _extract_series(self, url, webpage):
  64         title = self._html_search_regex(r'<div class="cne-series-info">.*?<h1>(.+?)</h1>',
  65                                         webpage, 'series title', flags=re.DOTALL)
  66         url_object = compat_urllib_parse_urlparse(url)
  67         base_url = '%s://%s' % (url_object.scheme, url_object.netloc)
  68         m_paths = re.finditer(r'<p class="cne-thumb-title">.*?<a href="(/watch/.+?)["\?]',
  69                               webpage, flags=re.DOTALL)
  70         paths = orderedSet(m.group(1) for m in m_paths)
  71         build_url = lambda path: compat_urlparse.urljoin(base_url, path)
  72         entries = [self.url_result(build_url(path), 'CondeNast') for path in paths]
  73         return self.playlist_result(entries, playlist_title=title)
  74
  75     def _extract_video(self, webpage, url_type):
  76         if url_type != 'embed':
  77             description = self._html_search_regex(
  78                 [
  79                     r'<div class="cne-video-description">(.+?)</div>',
  80                     r'<div class="video-post-content">(.+?)</div>',
  81                 ],
  82                 webpage, 'description', fatal=False, flags=re.DOTALL)
  83         else:
  84             description = None
  85         params = self._search_regex(r'var params = {(.+?)}[;,]', webpage,
  86                                     'player params', flags=re.DOTALL)
  87         video_id = self._search_regex(r'videoId: [\'"](.+?)[\'"]', params, 'video id')
  88         player_id = self._search_regex(r'playerId: [\'"](.+?)[\'"]', params, 'player id')
  89         target = self._search_regex(r'target: [\'"](.+?)[\'"]', params, 'target')
  90         data = compat_urllib_parse.urlencode({'videoId': video_id,
  91                                               'playerId': player_id,
  92                                               'target': target,
  93                                               })
  94         base_info_url = self._search_regex(r'url = [\'"](.+?)[\'"][,;]',
  95                                            webpage, 'base info url',
  96                                            default='http://player.cnevids.com/player/loader.js?')
  97         info_url = base_info_url + data
  98         info_page = self._download_webpage(info_url, video_id,
  99                                            'Downloading video info')
 100         video_info = self._search_regex(r'var\s*video\s*=\s*({.+?});', info_page, 'video info')
 101         video_info = self._parse_json(video_info, video_id)
 102
 103         formats = [{
 104             'format_id': '%s-%s' % (fdata['type'].split('/')[-1], fdata['quality']),
 105             'url': fdata['src'],
 106             'ext': fdata['type'].split('/')[-1],
 107             'quality': 1 if fdata['quality'] == 'high' else 0,
 108         } for fdata in video_info['sources'][0]]
 109         self._sort_formats(formats)
 110
 111         return {
 112             'id': video_id,
 113             'formats': formats,
 114             'title': video_info['title'],
 115             'thumbnail': video_info['poster_frame'],
 116             'description': description,
 117         }
 118
 119     def _real_extract(self, url):
 120         mobj = re.match(self._VALID_URL, url)
 121         site = mobj.group('site')
 122         url_type = mobj.group('type')
 123         item_id = mobj.group('id')
 124
 125         self.to_screen('Extracting from %s with the Condé Nast extractor' % self._SITES[site])
 126         webpage = self._download_webpage(url, item_id)
 127
 128         if url_type == 'series':
 129             return self._extract_series(url, webpage)
 130         else:
 131             return self._extract_video(webpage, url_type)