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

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