_ 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 from ..compat import (
  11     compat_urllib_parse,
  12     compat_ord,
  13 )
  14
  15
  16 class YoukuIE(InfoExtractor):
  17     IE_NAME = 'youku'
  18     _VALID_URL = r'''(?x)
  19         (?:
  20             http://(?:v|player)\.youku\.com/(?:v_show/id_|player\.php/sid/)|
  21             youku:)
  22         (?P<id>[A-Za-z0-9]+)(?:\.html|/v\.swf|)
  23     '''
  24
  25     _TEST = {
  26         'url': 'http://v.youku.com/v_show/id_XMTc1ODE5Njcy.html',
  27         'md5': '5f3af4192eabacc4501508d54a8cabd7',
  28         'info_dict': {
  29             'id': 'XMTc1ODE5Njcy',
  30             'title': '★Smile﹗♡ Git Fresh -Booty Music舞蹈.',
  31             'ext': 'flv'
  32         }
  33     }
  34
  35     def construct_video_urls(self, data1, data2):
  36         # get sid, token
  37         def yk_t(s1, s2):
  38             ls = list(range(256))
  39             t = 0
  40             for i in range(256):
  41                 t = (t + ls[i] + compat_ord(s1[i % len(s1)])) % 256
  42                 ls[i], ls[t] = ls[t], ls[i]
  43             s = bytearray()
  44             x, y = 0, 0
  45             for i in range(len(s2)):
  46                 y = (y + 1) % 256
  47                 x = (x + ls[y]) % 256
  48                 ls[x], ls[y] = ls[y], ls[x]
  49                 s.append(compat_ord(s2[i]) ^ ls[(ls[x] + ls[y]) % 256])
  50             return bytes(s)
  51
  52         sid, token = yk_t(
  53             b'becaf9be', base64.b64decode(data2['ep'].encode('ascii'))
  54         ).decode('ascii').split('_')
  55
  56         # get oip
  57         oip = data2['ip']
  58
  59         # get fileid
  60         string_ls = list(
  61             'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ/\:._-1234567890')
  62         shuffled_string_ls = []
  63         seed = data1['seed']
  64         N = len(string_ls)
  65         for ii in range(N):
  66             seed = (seed * 0xd3 + 0x754f) % 0x10000
  67             idx = seed * len(string_ls) // 0x10000
  68             shuffled_string_ls.append(string_ls[idx])
  69             del string_ls[idx]
  70
  71         fileid_dict = {}
  72         for format in data1['streamtypes']:
  73             streamfileid = [
  74                 int(i) for i in data1['streamfileids'][format].strip('*').split('*')]
  75             fileid = ''.join(
  76                 [shuffled_string_ls[i] for i in streamfileid])
  77             fileid_dict[format] = fileid[:8] + '%s' + fileid[10:]
  78
  79         def get_fileid(format, n):
  80             fileid = fileid_dict[format] % hex(int(n))[2:].upper().zfill(2)
  81             return fileid
  82
  83         # get ep
  84         def generate_ep(format, n):
  85             fileid = get_fileid(format, n)
  86             ep_t = yk_t(
  87                 b'bf7e5f01',
  88                 ('%s_%s_%s' % (sid, fileid, token)).encode('ascii')
  89             )
  90             ep = base64.b64encode(ep_t).decode('ascii')
  91             return ep
  92
  93         # generate video_urls
  94         video_urls_dict = {}
  95         for format in data1['streamtypes']:
  96             video_urls = []
  97             for dt in data1['segs'][format]:
  98                 n = str(int(dt['no']))
  99                 param = {
 100                     'K': dt['k'],
 101                     'hd': self.get_hd(format),
 102                     'myp': 0,
 103                     'ts': dt['seconds'],
 104                     'ypp': 0,
 105                     'ctype': 12,
 106                     'ev': 1,
 107                     'token': token,
 108                     'oip': oip,
 109                     'ep': generate_ep(format, n)
 110                 }
 111                 video_url = \
 112                     'http://k.youku.com/player/getFlvPath/' + \
 113                     'sid/' + sid + \
 114                     '_' + str(int(n) + 1).zfill(2) + \
 115                     '/st/' + self.parse_ext_l(format) + \
 116                     '/fileid/' + get_fileid(format, n) + '?' + \
 117                     compat_urllib_parse.urlencode(param)
 118                 video_urls.append(video_url)
 119             video_urls_dict[format] = video_urls
 120
 121         return video_urls_dict
 122
 123     def get_hd(self, fm):
 124         hd_id_dict = {
 125             'flv': '0',
 126             'mp4': '1',
 127             'hd2': '2',
 128             'hd3': '3',
 129             '3gp': '0',
 130             '3gphd': '1'
 131         }
 132         return hd_id_dict[fm]
 133
 134     def parse_ext_l(self, fm):
 135         ext_dict = {
 136             'flv': 'flv',
 137             'mp4': 'mp4',
 138             'hd2': 'flv',
 139             'hd3': 'flv',
 140             '3gp': 'flv',
 141             '3gphd': 'mp4'
 142         }
 143         return ext_dict[fm]
 144
 145     def get_format_name(self, fm):
 146         _dict = {
 147             '3gp': 'h6',
 148             '3gphd': 'h5',
 149             'flv': 'h4',
 150             'mp4': 'h3',
 151             'hd2': 'h2',
 152             'hd3': 'h1'
 153         }
 154         return _dict[fm]
 155
 156     def _real_extract(self, url):
 157         mobj = re.match(self._VALID_URL, url)
 158         video_id = mobj.group('id')
 159
 160         # request basic data
 161         data1_url = 'http://v.youku.com/player/getPlayList/VideoIDS/%s' % video_id
 162         data2_url = 'http://v.youku.com/player/getPlayList/VideoIDS/%s/Pf/4/ctype/12/ev/1' % video_id
 163
 164         raw_data1 = self._download_json(data1_url, video_id)
 165         raw_data2 = self._download_json(data2_url, video_id)
 166         data1 = raw_data1['data'][0]
 167         data2 = raw_data2['data'][0]
 168
 169         error_code = data1.get('error_code')
 170         if error_code:
 171             # -8 means blocked outside China.
 172             # Chinese and English, separated by newline.
 173             error = data1.get('error')
 174             raise ExtractorError(
 175                 error or 'Server reported error %i' %
 176                 error_code,
 177                 expected=True)
 178
 179         title = data1['title']
 180
 181         # generate video_urls_dict
 182         video_urls_dict = self.construct_video_urls(data1, data2)
 183
 184         # construct info
 185         entries = []
 186         for fm in data1['streamtypes']:
 187             video_urls = video_urls_dict[fm]
 188             for i in range(len(video_urls)):
 189                 if len(entries) < i + 1:
 190                     entries.append({'formats': []})
 191                 entries[i]['formats'].append(
 192                     {
 193                         'url': video_urls[i],
 194                         'format_id': self.get_format_name(fm),
 195                         'ext': self.parse_ext_l(fm),
 196                         'filesize': int(data1['segs'][fm][i]['size'])
 197                     }
 198                 )
 199
 200         for i in range(len(entries)):
 201             entries[i].update(
 202                 {
 203                     'id': '_part%d' % (i + 1),
 204                     'title': title,
 205                 }
 206             )
 207
 208         if len(entries) > 1:
 209             info = {
 210                 '_type': 'multi_video',
 211                 'id': video_id,
 212                 'title': title,
 213                 'entries': entries,
 214             }
 215         else:
 216             info = entries[0]
 217             info['id'] = video_id
 218
 219         return info