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