youtube_dl/extractor/youku.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import base64
   5
   6 from .common import InfoExtractor
   7 from ..compat import (
   8     compat_urllib_parse,
   9     compat_ord,
  10 )
  11 from ..utils import (
  12     ExtractorError,
  13     sanitized_Request,
  14 )
  15
  16
  17 class YoukuIE(InfoExtractor):
  18     IE_NAME = 'youku'
  19     IE_DESC = '优酷'
  20     _VALID_URL = r'''(?x)
  21         (?:
  22             http://(?:v|player)\.youku\.com/(?:v_show/id_|player\.php/sid/)|
  23             youku:)
  24         (?P<id>[A-Za-z0-9]+)(?:\.html|/v\.swf|)
  25     '''
  26
  27     _TESTS = [{
  28         'url': 'http://v.youku.com/v_show/id_XMTc1ODE5Njcy.html',
  29         'md5': '5f3af4192eabacc4501508d54a8cabd7',
  30         'info_dict': {
  31             'id': 'XMTc1ODE5Njcy_part1',
  32             'title': '★Smile﹗♡ Git Fresh -Booty Music舞蹈.',
  33             'ext': 'flv'
  34         }
  35     }, {
  36         'url': 'http://player.youku.com/player.php/sid/XNDgyMDQ2NTQw/v.swf',
  37         'only_matching': True,
  38     }, {
  39         'url': 'http://v.youku.com/v_show/id_XODgxNjg1Mzk2_ev_1.html',
  40         'info_dict': {
  41             'id': 'XODgxNjg1Mzk2',
  42             'title': '武媚娘传奇 85',
  43         },
  44         'playlist_count': 11,
  45     }, {
  46         'url': 'http://v.youku.com/v_show/id_XMTI1OTczNDM5Mg==.html',
  47         'info_dict': {
  48             'id': 'XMTI1OTczNDM5Mg',
  49             'title': '花千骨 04',
  50         },
  51         'playlist_count': 13,
  52         'skip': 'Available in China only',
  53     }, {
  54         'url': 'http://v.youku.com/v_show/id_XNjA1NzA2Njgw.html',
  55         'note': 'Video protected with password',
  56         'info_dict': {
  57             'id': 'XNjA1NzA2Njgw',
  58             'title': '邢義田复旦讲座之想象中的胡人—从“左衽孔子”说起',
  59         },
  60         'playlist_count': 19,
  61         'params': {
  62             'videopassword': '100600',
  63         },
  64     }]
  65
  66     def construct_video_urls(self, data):
  67         # get sid, token
  68         def yk_t(s1, s2):
  69             ls = list(range(256))
  70             t = 0
  71             for i in range(256):
  72                 t = (t + ls[i] + compat_ord(s1[i % len(s1)])) % 256
  73                 ls[i], ls[t] = ls[t], ls[i]
  74             s = bytearray()
  75             x, y = 0, 0
  76             for i in range(len(s2)):
  77                 y = (y + 1) % 256
  78                 x = (x + ls[y]) % 256
  79                 ls[x], ls[y] = ls[y], ls[x]
  80                 s.append(compat_ord(s2[i]) ^ ls[(ls[x] + ls[y]) % 256])
  81             return bytes(s)
  82
  83         sid, token = yk_t(
  84             b'becaf9be', base64.b64decode(data['security']['encrypt_string'].encode('ascii'))
  85         ).decode('ascii').split('_')
  86
  87         # get oip
  88         oip = data['security']['ip']
  89
  90         fileid_dict = {}
  91         for stream in data['stream']:
  92             format = stream.get('stream_type')
  93             fileid = stream['stream_fileid']
  94             fileid_dict[format] = fileid
  95
  96         def get_fileid(format, n):
  97             number = hex(int(str(n), 10))[2:].upper()
  98             if len(number) == 1:
  99                 number = '0' + number
 100             streamfileids = fileid_dict[format]
 101             fileid = streamfileids[0:8] + number + streamfileids[10:]
 102             return fileid
 103
 104         # get ep
 105         def generate_ep(format, n):
 106             fileid = get_fileid(format, n)
 107             ep_t = yk_t(
 108                 b'bf7e5f01',
 109                 ('%s_%s_%s' % (sid, fileid, token)).encode('ascii')
 110             )
 111             ep = base64.b64encode(ep_t).decode('ascii')
 112             return ep
 113
 114         # generate video_urls
 115         video_urls_dict = {}
 116         for stream in data['stream']:
 117             format = stream.get('stream_type')
 118             video_urls = []
 119             for dt in stream['segs']:
 120                 n = str(stream['segs'].index(dt))
 121                 param = {
 122                     'K': dt['key'],
 123                     'hd': self.get_hd(format),
 124                     'myp': 0,
 125                     'ypp': 0,
 126                     'ctype': 12,
 127                     'ev': 1,
 128                     'token': token,
 129                     'oip': oip,
 130                     'ep': generate_ep(format, n)
 131                 }
 132                 video_url = \
 133                     'http://k.youku.com/player/getFlvPath/' + \
 134                     'sid/' + sid + \
 135                     '_00' + \
 136                     '/st/' + self.parse_ext_l(format) + \
 137                     '/fileid/' + get_fileid(format, n) + '?' + \
 138                     compat_urllib_parse.urlencode(param)
 139                 video_urls.append(video_url)
 140             video_urls_dict[format] = video_urls
 141
 142         return video_urls_dict
 143
 144     def get_hd(self, fm):
 145         hd_id_dict = {
 146             'flv': '0',
 147             'mp4': '1',
 148             'hd2': '2',
 149             'hd3': '3',
 150             '3gp': '0',
 151             '3gphd': '1',
 152             'flvhd': '0',
 153             'mp4hd': '1',
 154             'mp4hd2': '1'
 155         }
 156         return hd_id_dict[fm]
 157
 158     def parse_ext_l(self, fm):
 159         ext_dict = {
 160             'flv': 'flv',
 161             'mp4': 'mp4',
 162             'mp4hd': 'mp4',
 163             'mp4hd2': 'flv',
 164             'mp4hd3': 'flv',
 165             'hd2': 'flv',
 166             'hd3': 'flv',
 167             '3gp': 'flv',
 168             '3gphd': 'mp4',
 169             'flvhd': 'flv'
 170         }
 171         return ext_dict[fm]
 172
 173     def get_format_name(self, fm):
 174         _dict = {
 175             '3gp': 'h6',
 176             '3gphd': 'h5',
 177             'flvhd': 'h4',
 178             'flv': 'h4',
 179             'mp4': 'h3',
 180             'hd2': 'h2',
 181             'hd3': 'h1',
 182             'mp4hd': 'h3',
 183             'mp4hd3': 'h4',
 184             'mp4hd2': 'h4'
 185         }
 186         return _dict[fm]
 187
 188     def _real_extract(self, url):
 189         video_id = self._match_id(url)
 190
 191         def retrieve_data(req_url, note):
 192             headers = {
 193                 'Referer': req_url,
 194             }
 195             self._set_cookie('youku.com', 'xreferrer', 'http://www.youku.com')
 196             req = sanitized_Request(req_url, headers=headers)
 197
 198             cn_verification_proxy = self._downloader.params.get('cn_verification_proxy')
 199             if cn_verification_proxy:
 200                 req.add_header('Ytdl-request-proxy', cn_verification_proxy)
 201
 202             raw_data = self._download_json(req, video_id, note=note)
 203
 204             return raw_data['data']
 205
 206         video_password = self._downloader.params.get('videopassword', None)
 207
 208         # request basic data
 209         basic_data_url = "http://play.youku.com/play/get.json?vid=%s&ct=12" % video_id
 210         if video_password:
 211             basic_data_url += '&pwd=%s' % video_password
 212
 213         data = retrieve_data(
 214             basic_data_url,
 215             'Downloading JSON metadata 1')
 216
 217         error = data.get('error')
 218         if error:
 219             error_note = error.get('note')
 220             if error_note is not None and '因版权原因无法观看此视频' in error_note:
 221                 raise ExtractorError(
 222                     'Youku said: Sorry, this video is available in China only', expected=True)
 223             else:
 224                 msg = 'Youku server reported error %i' % error.get('code')
 225                 if error is not None:
 226                     msg += ': ' + error_note
 227                 raise ExtractorError(msg)
 228
 229         # get video title
 230         title = data['video']['title']
 231
 232         # generate video_urls_dict
 233         video_urls_dict = self.construct_video_urls(data)
 234
 235         # construct info
 236         entries = [{
 237             'id': '%s_part%d' % (video_id, i + 1),
 238             'title': title,
 239             'formats': [],
 240             # some formats are not available for all parts, we have to detect
 241             # which one has all
 242         } for i in range(max(len(v.get('segs')) for v in data['stream']))]
 243         for stream in data['stream']:
 244             fm = stream.get('stream_type')
 245             video_urls = video_urls_dict[fm]
 246             for video_url, seg, entry in zip(video_urls, stream['segs'], entries):
 247                 entry['formats'].append({
 248                     'url': video_url,
 249                     'format_id': self.get_format_name(fm),
 250                     'ext': self.parse_ext_l(fm),
 251                     'filesize': int(seg['size']),
 252                 })
 253
 254         return {
 255             '_type': 'multi_video',
 256             'id': video_id,
 257             'title': title,
 258             'entries': entries,
 259         }