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