Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/youku.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import base64
   5 import itertools
   6 import random
   7 import re
   8 import string
   9 import time
  10
  11 from .common import InfoExtractor
  12 from ..compat import (
  13     compat_urllib_parse_urlencode,
  14     compat_ord,
  15 )
  16 from ..utils import (
  17     ExtractorError,
  18     get_element_by_attribute,
  19     sanitized_Request,
  20 )
  21
  22
  23 class YoukuIE(InfoExtractor):
  24     IE_NAME = 'youku'
  25     IE_DESC = '优酷'
  26     _VALID_URL = r'''(?x)
  27         (?:
  28             http://(?:v|player)\.youku\.com/(?:v_show/id_|player\.php/sid/)|
  29             youku:)
  30         (?P<id>[A-Za-z0-9]+)(?:\.html|/v\.swf|)
  31     '''
  32
  33     _TESTS = [{
  34         # MD5 is unstable
  35         'url': 'http://v.youku.com/v_show/id_XMTc1ODE5Njcy.html',
  36         'info_dict': {
  37             'id': 'XMTc1ODE5Njcy_part1',
  38             'title': '★Smile﹗♡ Git Fresh -Booty Music舞蹈.',
  39             'ext': 'flv'
  40         }
  41     }, {
  42         'url': 'http://player.youku.com/player.php/sid/XNDgyMDQ2NTQw/v.swf',
  43         'only_matching': True,
  44     }, {
  45         'url': 'http://v.youku.com/v_show/id_XODgxNjg1Mzk2_ev_1.html',
  46         'info_dict': {
  47             'id': 'XODgxNjg1Mzk2',
  48             'title': '武媚娘传奇 85',
  49         },
  50         'playlist_count': 11,
  51         'skip': 'Available in China only',
  52     }, {
  53         'url': 'http://v.youku.com/v_show/id_XMTI1OTczNDM5Mg==.html',
  54         'info_dict': {
  55             'id': 'XMTI1OTczNDM5Mg',
  56             'title': '花千骨 04',
  57         },
  58         'playlist_count': 13,
  59     }, {
  60         'url': 'http://v.youku.com/v_show/id_XNjA1NzA2Njgw.html',
  61         'note': 'Video protected with password',
  62         'info_dict': {
  63             'id': 'XNjA1NzA2Njgw',
  64             'title': '邢義田复旦讲座之想象中的胡人—从“左衽孔子”说起',
  65         },
  66         'playlist_count': 19,
  67         'params': {
  68             'videopassword': '100600',
  69         },
  70     }, {
  71         # /play/get.json contains streams with "channel_type":"tail"
  72         'url': 'http://v.youku.com/v_show/id_XOTUxMzg4NDMy.html',
  73         'info_dict': {
  74             'id': 'XOTUxMzg4NDMy',
  75             'title': '我的世界☆明月庄主☆车震猎杀☆杀人艺术Minecraft',
  76         },
  77         'playlist_count': 6,
  78     }]
  79
  80     def construct_video_urls(self, data):
  81         # get sid, token
  82         def yk_t(s1, s2):
  83             ls = list(range(256))
  84             t = 0
  85             for i in range(256):
  86                 t = (t + ls[i] + compat_ord(s1[i % len(s1)])) % 256
  87                 ls[i], ls[t] = ls[t], ls[i]
  88             s = bytearray()
  89             x, y = 0, 0
  90             for i in range(len(s2)):
  91                 y = (y + 1) % 256
  92                 x = (x + ls[y]) % 256
  93                 ls[x], ls[y] = ls[y], ls[x]
  94                 s.append(compat_ord(s2[i]) ^ ls[(ls[x] + ls[y]) % 256])
  95             return bytes(s)
  96
  97         sid, token = yk_t(
  98             b'becaf9be', base64.b64decode(data['security']['encrypt_string'].encode('ascii'))
  99         ).decode('ascii').split('_')
 100
 101         # get oip
 102         oip = data['security']['ip']
 103
 104         fileid_dict = {}
 105         for stream in data['stream']:
 106             if stream.get('channel_type') == 'tail':
 107                 continue
 108             format = stream.get('stream_type')
 109             fileid = stream['stream_fileid']
 110             fileid_dict[format] = fileid
 111
 112         def get_fileid(format, n):
 113             number = hex(int(str(n), 10))[2:].upper()
 114             if len(number) == 1:
 115                 number = '0' + number
 116             streamfileids = fileid_dict[format]
 117             fileid = streamfileids[0:8] + number + streamfileids[10:]
 118             return fileid
 119
 120         # get ep
 121         def generate_ep(format, n):
 122             fileid = get_fileid(format, n)
 123             ep_t = yk_t(
 124                 b'bf7e5f01',
 125                 ('%s_%s_%s' % (sid, fileid, token)).encode('ascii')
 126             )
 127             ep = base64.b64encode(ep_t).decode('ascii')
 128             return ep
 129
 130         # generate video_urls
 131         video_urls_dict = {}
 132         for stream in data['stream']:
 133             if stream.get('channel_type') == 'tail':
 134                 continue
 135             format = stream.get('stream_type')
 136             video_urls = []
 137             for dt in stream['segs']:
 138                 n = str(stream['segs'].index(dt))
 139                 param = {
 140                     'K': dt['key'],
 141                     'hd': self.get_hd(format),
 142                     'myp': 0,
 143                     'ypp': 0,
 144                     'ctype': 12,
 145                     'ev': 1,
 146                     'token': token,
 147                     'oip': oip,
 148                     'ep': generate_ep(format, n)
 149                 }
 150                 video_url = \
 151                     'http://k.youku.com/player/getFlvPath/' + \
 152                     'sid/' + sid + \
 153                     '_00' + \
 154                     '/st/' + self.parse_ext_l(format) + \
 155                     '/fileid/' + get_fileid(format, n) + '?' + \
 156                     compat_urllib_parse_urlencode(param)
 157                 video_urls.append(video_url)
 158             video_urls_dict[format] = video_urls
 159
 160         return video_urls_dict
 161
 162     @staticmethod
 163     def get_ysuid():
 164         return '%d%s' % (int(time.time()), ''.join([
 165             random.choice(string.ascii_letters) for i in range(3)]))
 166
 167     def get_hd(self, fm):
 168         hd_id_dict = {
 169             '3gp': '0',
 170             '3gphd': '1',
 171             'flv': '0',
 172             'flvhd': '0',
 173             'mp4': '1',
 174             'mp4hd': '1',
 175             'mp4hd2': '1',
 176             'mp4hd3': '1',
 177             'hd2': '2',
 178             'hd3': '3',
 179         }
 180         return hd_id_dict[fm]
 181
 182     def parse_ext_l(self, fm):
 183         ext_dict = {
 184             '3gp': 'flv',
 185             '3gphd': 'mp4',
 186             'flv': 'flv',
 187             'flvhd': 'flv',
 188             'mp4': 'mp4',
 189             'mp4hd': 'mp4',
 190             'mp4hd2': 'flv',
 191             'mp4hd3': 'flv',
 192             'hd2': 'flv',
 193             'hd3': 'flv',
 194         }
 195         return ext_dict[fm]
 196
 197     def get_format_name(self, fm):
 198         _dict = {
 199             '3gp': 'h6',
 200             '3gphd': 'h5',
 201             'flv': 'h4',
 202             'flvhd': 'h4',
 203             'mp4': 'h3',
 204             'mp4hd': 'h3',
 205             'mp4hd2': 'h4',
 206             'mp4hd3': 'h4',
 207             'hd2': 'h2',
 208             'hd3': 'h1',
 209         }
 210         return _dict[fm]
 211
 212     def _real_extract(self, url):
 213         video_id = self._match_id(url)
 214
 215         self._set_cookie('youku.com', '__ysuid', self.get_ysuid())
 216
 217         def retrieve_data(req_url, note):
 218             headers = {
 219                 'Referer': req_url,
 220             }
 221             self._set_cookie('youku.com', 'xreferrer', 'http://www.youku.com')
 222             req = sanitized_Request(req_url, headers=headers)
 223
 224             cn_verification_proxy = self._downloader.params.get('cn_verification_proxy')
 225             if cn_verification_proxy:
 226                 req.add_header('Ytdl-request-proxy', cn_verification_proxy)
 227
 228             raw_data = self._download_json(req, video_id, note=note)
 229
 230             return raw_data['data']
 231
 232         video_password = self._downloader.params.get('videopassword')
 233
 234         # request basic data
 235         basic_data_url = 'http://play.youku.com/play/get.json?vid=%s&ct=12' % video_id
 236         if video_password:
 237             basic_data_url += '&pwd=%s' % video_password
 238
 239         data = retrieve_data(basic_data_url, 'Downloading JSON metadata')
 240
 241         error = data.get('error')
 242         if error:
 243             error_note = error.get('note')
 244             if error_note is not None and '因版权原因无法观看此视频' in error_note:
 245                 raise ExtractorError(
 246                     'Youku said: Sorry, this video is available in China only', expected=True)
 247             elif error_note and '该视频被设为私密' in error_note:
 248                 raise ExtractorError(
 249                     'Youku said: Sorry, this video is private', expected=True)
 250             else:
 251                 msg = 'Youku server reported error %i' % error.get('code')
 252                 if error_note is not None:
 253                     msg += ': ' + error_note
 254                 raise ExtractorError(msg)
 255
 256         # get video title
 257         title = data['video']['title']
 258
 259         # generate video_urls_dict
 260         video_urls_dict = self.construct_video_urls(data)
 261
 262         # construct info
 263         entries = [{
 264             'id': '%s_part%d' % (video_id, i + 1),
 265             'title': title,
 266             'formats': [],
 267             # some formats are not available for all parts, we have to detect
 268             # which one has all
 269         } for i in range(max(len(v.get('segs')) for v in data['stream']))]
 270         for stream in data['stream']:
 271             if stream.get('channel_type') == 'tail':
 272                 continue
 273             fm = stream.get('stream_type')
 274             video_urls = video_urls_dict[fm]
 275             for video_url, seg, entry in zip(video_urls, stream['segs'], entries):
 276                 entry['formats'].append({
 277                     'url': video_url,
 278                     'format_id': self.get_format_name(fm),
 279                     'ext': self.parse_ext_l(fm),
 280                     'filesize': int(seg['size']),
 281                     'width': stream.get('width'),
 282                     'height': stream.get('height'),
 283                 })
 284
 285         return {
 286             '_type': 'multi_video',
 287             'id': video_id,
 288             'title': title,
 289             'entries': entries,
 290         }
 291
 292
 293 class YoukuShowIE(InfoExtractor):
 294     _VALID_URL = r'https?://(?:www\.)?youku\.com/show_page/id_(?P<id>[0-9a-z]+)\.html'
 295     IE_NAME = 'youku:show'
 296
 297     _TEST = {
 298         'url': 'http://www.youku.com/show_page/id_zc7c670be07ff11e48b3f.html',
 299         'info_dict': {
 300             'id': 'zc7c670be07ff11e48b3f',
 301             'title': '花千骨 未删减版',
 302             'description': 'md5:578d4f2145ae3f9128d9d4d863312910',
 303         },
 304         'playlist_count': 50,
 305     }
 306
 307     _PAGE_SIZE = 40
 308
 309     def _find_videos_in_page(self, webpage):
 310         videos = re.findall(
 311             r'<li><a[^>]+href="(?P<url>https?://v\.youku\.com/[^"]+)"[^>]+title="(?P<title>[^"]+)"', webpage)
 312         return [
 313             self.url_result(video_url, YoukuIE.ie_key(), title)
 314             for video_url, title in videos]
 315
 316     def _real_extract(self, url):
 317         show_id = self._match_id(url)
 318         webpage = self._download_webpage(url, show_id)
 319
 320         entries = self._find_videos_in_page(webpage)
 321
 322         playlist_title = self._html_search_regex(
 323             r'<span[^>]+class="name">([^<]+)</span>', webpage, 'playlist title', fatal=False)
 324         detail_div = get_element_by_attribute('class', 'detail', webpage) or ''
 325         playlist_description = self._html_search_regex(
 326             r'<span[^>]+style="display:none"[^>]*>([^<]+)</span>',
 327             detail_div, 'playlist description', fatal=False)
 328
 329         for idx in itertools.count(1):
 330             episodes_page = self._download_webpage(
 331                 'http://www.youku.com/show_episode/id_%s.html' % show_id,
 332                 show_id, query={'divid': 'reload_%d' % (idx * self._PAGE_SIZE + 1)},
 333                 note='Downloading episodes page %d' % idx)
 334             new_entries = self._find_videos_in_page(episodes_page)
 335             entries.extend(new_entries)
 336             if len(new_entries) < self._PAGE_SIZE:
 337                 break
 338
 339         return self.playlist_result(entries, show_id, playlist_title, playlist_description)