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