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

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import sys
   5 pyvs = sys.version_info[0]
   6 import re
   7 import base64
   8
   9 from .common import InfoExtractor
  10 from ..utils import ExtractorError
  11
  12 class YoukuIE(InfoExtractor):
  13     IE_NAME = 'youku'
  14     _VALID_URL = r'''(?x)
  15         (?:
  16             http://(?:v|player)\.youku\.com/(?:v_show/id_|player\.php/sid/)|
  17             youku:)
  18         (?P<id>[A-Za-z0-9]+)(?:\.html|/v\.swf|)
  19     '''
  20
  21     _TEST = {
  22             'url': 'http://v.youku.com/v_show/id_XMTc1ODE5Njcy.html',
  23             'md5': '5f3af4192eabacc4501508d54a8cabd7',
  24             'info_dict': {
  25                 'id': 'XMTc1ODE5Njcy',
  26                 'title': '★Smile﹗♡ Git Fresh -Booty Music舞蹈.',
  27                 'ext': 'flv'
  28             }
  29     }
  30
  31     def construct_video_urls(self, data1, data2):
  32         # get sid, token
  33         def yk_t(s1, s2):
  34             ls = list(range(256))
  35             t = 0
  36             for i in range(256):
  37                 t = (t + ls[i] + ord(s1[i%len(s1)])) % 256
  38                 ls[i], ls[t] = ls[t], ls[i]
  39             s = '' if pyvs == 3 else b''
  40             x, y = 0, 0
  41             for i in range(len(s2)):
  42                 y = (y + 1) % 256
  43                 x = (x + ls[y]) % 256
  44                 ls[x], ls[y] = ls[y], ls[x]
  45                 if isinstance(s2[i], int):
  46                     s += chr(s2[i] ^ ls[(ls[x]+ls[y]) % 256])
  47                 else:
  48                     s += chr(ord(s2[i]) ^ ls[(ls[x]+ls[y]) % 256])
  49             return s
  50
  51         sid, token = yk_t(
  52             'becaf9be',
  53             base64.b64decode(bytes(data2['ep'], 'ascii')) \
  54                 if pyvs == 3 \
  55                 else base64.b64decode(data2['ep'])
  56         ).split('_')
  57
  58         # get oip
  59         oip = data2['ip']
  60
  61         # get fileid
  62         string_ls = list(
  63             'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ/\:._-1234567890')
  64         shuffled_string_ls = []
  65         seed = data1['seed']
  66         N = len(string_ls)
  67         for ii in range(N):
  68             seed = (seed * 0xd3 + 0x754f) % 0x10000
  69             idx = seed * len(string_ls) // 0x10000
  70             shuffled_string_ls.append(string_ls[idx])
  71             del string_ls[idx]
  72
  73         fileid_dict = {}
  74         for format in data1['streamtypes']:
  75             streamfileid = [
  76                 int(i) for i in data1['streamfileids'][format].strip('*').split('*')]
  77             fileid = ''.join(
  78                 [shuffled_string_ls[i] for i in streamfileid])
  79             fileid_dict[format] = fileid[:8] + '%s' + fileid[10:]
  80
  81         def get_fileid(format, n):
  82             fileid = fileid_dict[format] % hex(int(n))[2:].upper().zfill(2)
  83             return fileid
  84
  85         # get ep
  86         def generate_ep(format, n):
  87             fileid = get_fileid(format, n)
  88             ep_t = yk_t(
  89                 'bf7e5f01',
  90                 bytes('%s_%s_%s' % (sid, fileid, token), 'ascii') \
  91                 if pyvs == 3 \
  92                 else ('%s_%s_%s' % (sid, fileid, token))
  93             )
  94             ep = base64.b64encode(
  95                 bytes(ep_t, 'latin') \
  96                 if pyvs == 3 \
  97                 else ep_t
  98             ).decode()
  99             ep = ep.replace('+', '%2B')
 100             ep = ep.replace('/', '%2F')
 101             ep = ep.replace('=', '%2D')
 102             return ep
 103
 104         # generate video_urls
 105         video_urls_dict = {}
 106         for format in data1['streamtypes']:
 107             video_urls = []
 108             for dt in data1['segs'][format]:
 109                 n = str(int(dt['no']))
 110                 video_url = \
 111                     'http://k.youku.com/player/getFlvPath/' + \
 112                     'sid/' + sid + \
 113                     '_' + str(int(n)+1).zfill(2) + \
 114                     '/st/' + self.parse_ext_l(format) + \
 115                     '/fileid/' + get_fileid(format, n)  + '?' + \
 116                     'K=' + str(dt['k']) + \
 117                     '&hd=' + self.get_hd(format) + \
 118                     '&myp=0' + \
 119                     '&ts=' + str(dt['seconds']) + \
 120                     '&ypp=0&ctype=12&ev=1' + \
 121                     '&token=' + str(token) + \
 122                     '&oip=' + str(oip) + \
 123                     '&ep=' + generate_ep(format, n)
 124                 video_urls.append(video_url)
 125             video_urls_dict[format] = video_urls
 126
 127         return video_urls_dict
 128
 129     def get_hd(self, fm):
 130         hd_id_dict = {
 131             'flv': '0',
 132             'mp4': '1',
 133             'hd2': '2',
 134             'hd3': '3',
 135             '3gp': '0',
 136             '3gphd': '1'
 137         }
 138         return hd_id_dict[fm]
 139
 140     def parse_ext_l(self, fm):
 141         ext_dict = {
 142             'flv': 'flv',
 143             'mp4': 'mp4',
 144             'hd2': 'flv',
 145             'hd3': 'flv',
 146             '3gp': 'flv',
 147             '3gphd': 'mp4',
 148         }
 149         return ext_dict[fm]
 150
 151     def _real_extract(self, url):
 152         mobj = re.match(self._VALID_URL, url)
 153         video_id = mobj.group('id')
 154
 155         # request basic data
 156         data1_url = 'http://v.youku.com/player/getPlayList/VideoIDS/%s' % video_id
 157         data2_url = 'http://v.youku.com/player/getPlayList/VideoIDS/%s/Pf/4/ctype/12/ev/1' % video_id
 158
 159         raw_data1 = self._download_json(data1_url, video_id)
 160         raw_data2 = self._download_json(data2_url, video_id)
 161         data1 = raw_data1['data'][0]
 162         data2 = raw_data2['data'][0]
 163
 164         error_code = data1.get('error_code')
 165         if error_code:
 166             # -8 means blocked outside China.
 167             # Chinese and English, separated by newline.
 168             error = data1.get('error')
 169             raise ExtractorError(
 170                 error or 'Server reported error %i' %
 171                 error_code,
 172                 expected=True)
 173
 174         title = data1['title']
 175
 176         # generate video_urls_dict
 177         video_urls_dict = self.construct_video_urls(data1, data2)
 178
 179         # construct info
 180         entries = []
 181         for fm in data1['streamtypes']:
 182             #formats = []
 183             video_urls = video_urls_dict[fm]
 184             for i in range(len(video_urls)):
 185                 if len(entries) < i+1:
 186                     entries.append({'formats': []})
 187                 entries[i]['formats'].append(
 188                     {
 189                         'url': video_urls[i],
 190                         'format_id': fm,
 191                         'ext': self.parse_ext_l(fm),
 192                         'filesize': int(data1['segs'][fm][i]['size'])
 193                     }
 194                 )
 195
 196         for i in range(len(entries)):
 197             entries[i].update(
 198                 {
 199                     'id': '_part%d' % (i+1),
 200                     'title': title,
 201                 }
 202             )
 203
 204         if len(entries) > 1:
 205             info = {
 206                 '_type': 'multi_video',
 207                 'id': video_id,
 208                 'title': title,
 209                 'entries': entries,
 210             }
 211         else:
 212             info = entries[0]
 213             info['id'] = video_id
 214
 215         return info