youtube_dl/extractor/tudou.py

   1 # coding: utf-8
   2
   3 from __future__ import unicode_literals
   4
   5 import re
   6 import json
   7
   8 from .common import InfoExtractor
   9
  10
  11 class TudouIE(InfoExtractor):
  12     _VALID_URL = r'https?://(?:www\.)?tudou\.com/(?:listplay|programs(?:/view)?|albumplay)/.*?/(?P<id>[^/?#]+?)(?:\.html)?/?(?:$|[?#])'
  13     _TESTS = [{
  14         'url': 'http://www.tudou.com/listplay/zzdE77v6Mmo/2xN2duXMxmw.html',
  15         'md5': '140a49ed444bd22f93330985d8475fcb',
  16         'info_dict': {
  17             'id': '159448201',
  18             'ext': 'f4v',
  19             'title': '卡马乔国足开大脚长传冲吊集锦',
  20             'thumbnail': 're:^https?://.*\.jpg$',
  21         }
  22     }, {
  23         'url': 'http://www.tudou.com/programs/view/ajX3gyhL0pc/',
  24         'info_dict': {
  25             'id': '117049447',
  26             'ext': 'f4v',
  27             'title': 'La Sylphide-Bolshoi-Ekaterina Krysanova & Vyacheslav Lopatin 2012',
  28             'thumbnail': 're:^https?://.*\.jpg$',
  29         }
  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         m = re.search(r'vcode:\s*[\'"](.+?)[\'"]', webpage)
  47         if m and m.group(1):
  48             return {
  49                 '_type': 'url',
  50                 'url': 'youku:' + m.group(1),
  51                 'ie_key': 'Youku'
  52             }
  53
  54         title = self._search_regex(
  55             r",kw:\s*['\"](.+?)[\"']", webpage, 'title')
  56         thumbnail_url = self._search_regex(
  57             r",pic:\s*[\"'](.+?)[\"']", webpage, 'thumbnail URL', fatal=False)
  58
  59         player_url = self._search_regex(
  60             r"playerUrl\s*:\s*['\"](.+?\.swf)[\"']",
  61             webpage, 'player URL', default=self._PLAYER_URL)
  62
  63         segs_json = self._search_regex(r'segs: \'(.*)\'', webpage, 'segments')
  64         segments = json.loads(segs_json)
  65         # It looks like the keys are the arguments that have to be passed as
  66         # the hd field in the request url, we pick the higher
  67         # Also, filter non-number qualities (see issue #3643).
  68         quality = sorted(filter(lambda k: k.isdigit(), segments.keys()),
  69                          key=lambda k: int(k))[-1]
  70         parts = segments[quality]
  71         result = []
  72         len_parts = len(parts)
  73         if len_parts > 1:
  74             self.to_screen('%s: found %s parts' % (video_id, len_parts))
  75         for part in parts:
  76             part_id = part['k']
  77             final_url = self._url_for_id(part_id, quality)
  78             ext = (final_url.split('?')[0]).split('.')[-1]
  79             part_info = {
  80                 'id': '%s' % part_id,
  81                 'url': final_url,
  82                 'ext': ext,
  83                 'title': title,
  84                 'thumbnail': thumbnail_url,
  85                 'http_headers': {
  86                     'Referer': player_url,
  87                 },
  88             }
  89             result.append(part_info)
  90
  91         return {
  92             '_type': 'multi_video',
  93             'entries': result,
  94             'id': video_id,
  95             'title': title,
  96         }