youtube_dl/extractor/tudou.py

   1 # coding: utf-8
   2
   3 from __future__ import unicode_literals
   4
   5 from .common import InfoExtractor
   6
   7
   8 class TudouIE(InfoExtractor):
   9     _VALID_URL = r'https?://(?:www\.)?tudou\.com/(?:listplay|programs(?:/view)?|albumplay)/([^/]+/)*(?P<id>[^/?#]+?)(?:\.html)?/?(?:$|[?#])'
  10     _TESTS = [{
  11         'url': 'http://www.tudou.com/listplay/zzdE77v6Mmo/2xN2duXMxmw.html',
  12         'md5': '140a49ed444bd22f93330985d8475fcb',
  13         'info_dict': {
  14             'id': '159448201',
  15             'ext': 'f4v',
  16             'title': '卡马乔国足开大脚长传冲吊集锦',
  17             'thumbnail': 're:^https?://.*\.jpg$',
  18         }
  19     }, {
  20         'url': 'http://www.tudou.com/programs/view/ajX3gyhL0pc/',
  21         'info_dict': {
  22             'id': '117049447',
  23             'ext': 'f4v',
  24             'title': 'La Sylphide-Bolshoi-Ekaterina Krysanova & Vyacheslav Lopatin 2012',
  25             'thumbnail': 're:^https?://.*\.jpg$',
  26         }
  27     }, {
  28         'url': 'http://www.tudou.com/albumplay/cJAHGih4yYg.html',
  29         'only_matching': True,
  30     }]
  31
  32     _PLAYER_URL = 'http://js.tudouui.com/bin/lingtong/PortalPlayer_177.swf'
  33
  34     def _url_for_id(self, id, quality=None):
  35         info_url = "http://v2.tudou.com/f?id=" + str(id)
  36         if quality:
  37             info_url += '&hd' + quality
  38         webpage = self._download_webpage(info_url, id, "Opening the info webpage")
  39         final_url = self._html_search_regex('>(.+?)</f>', webpage, 'video url')
  40         return final_url
  41
  42     def _real_extract(self, url):
  43         video_id = self._match_id(url)
  44         webpage = self._download_webpage(url, video_id)
  45
  46         youku_vcode = self._search_regex(
  47             r'vcode:\s*[\'"](.+?)[\'"]', webpage, 'youku vcode', default=None)
  48         if youku_vcode:
  49             return self.url_result('youku:' + youku_vcode, ie='Youku')
  50
  51         title = self._search_regex(
  52             r",kw:\s*['\"](.+?)[\"']", webpage, 'title')
  53         thumbnail_url = self._search_regex(
  54             r",pic:\s*[\"'](.+?)[\"']", webpage, 'thumbnail URL', fatal=False)
  55
  56         player_url = self._search_regex(
  57             r"playerUrl\s*:\s*['\"](.+?\.swf)[\"']",
  58             webpage, 'player URL', default=self._PLAYER_URL)
  59
  60         segments = self._parse_json(self._search_regex(
  61             r'segs: \'(.*)\'', webpage, 'segments'), video_id)
  62         # It looks like the keys are the arguments that have to be passed as
  63         # the hd field in the request url, we pick the higher
  64         # Also, filter non-number qualities (see issue #3643).
  65         quality = sorted(filter(lambda k: k.isdigit(), segments.keys()),
  66                          key=lambda k: int(k))[-1]
  67         parts = segments[quality]
  68         result = []
  69         len_parts = len(parts)
  70         if len_parts > 1:
  71             self.to_screen('%s: found %s parts' % (video_id, len_parts))
  72         for part in parts:
  73             part_id = part['k']
  74             final_url = self._url_for_id(part_id, quality)
  75             ext = (final_url.split('?')[0]).split('.')[-1]
  76             part_info = {
  77                 'id': '%s' % part_id,
  78                 'url': final_url,
  79                 'ext': ext,
  80                 'title': title,
  81                 'thumbnail': thumbnail_url,
  82                 'http_headers': {
  83                     'Referer': player_url,
  84                 },
  85             }
  86             result.append(part_info)
  87
  88         return {
  89             '_type': 'multi_video',
  90             'entries': result,
  91             'id': video_id,
  92             'title': title,
  93         }