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

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import hashlib
   5 import math
   6 import random
   7 import time
   8 import uuid
   9
  10 from .common import InfoExtractor
  11 from ..compat import compat_urllib_parse
  12 from ..utils import ExtractorError
  13
  14
  15 class IqiyiIE(InfoExtractor):
  16     IE_NAME = 'iqiyi'
  17     IE_DESC = '爱奇艺'
  18
  19     _VALID_URL = r'http://(?:www\.)iqiyi.com/v_.+?\.html'
  20
  21     _TESTS = [{
  22         'url': 'http://www.iqiyi.com/v_19rrojlavg.html',
  23         'md5': '2cb594dc2781e6c941a110d8f358118b',
  24         'info_dict': {
  25             'id': '9c1fb1b99d192b21c559e5a1a2cb3c73',
  26             'title': '美国德州空中惊现奇异云团 酷似UFO',
  27             'ext': 'f4v',
  28         }
  29     }, {
  30         'url': 'http://www.iqiyi.com/v_19rrhnnclk.html',
  31         'info_dict': {
  32             'id': 'e3f585b550a280af23c98b6cb2be19fb',
  33             'title': '名侦探柯南第752集',
  34         },
  35         'playlist': [{
  36             'info_dict': {
  37                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part1',
  38                 'ext': 'f4v',
  39                 'title': '名侦探柯南第752集',
  40             },
  41         }, {
  42             'info_dict': {
  43                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part2',
  44                 'ext': 'f4v',
  45                 'title': '名侦探柯南第752集',
  46             },
  47         }, {
  48             'info_dict': {
  49                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part3',
  50                 'ext': 'f4v',
  51                 'title': '名侦探柯南第752集',
  52             },
  53         }, {
  54             'info_dict': {
  55                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part4',
  56                 'ext': 'f4v',
  57                 'title': '名侦探柯南第752集',
  58             },
  59         }, {
  60             'info_dict': {
  61                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part5',
  62                 'ext': 'f4v',
  63                 'title': '名侦探柯南第752集',
  64             },
  65         }, {
  66             'info_dict': {
  67                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part6',
  68                 'ext': 'f4v',
  69                 'title': '名侦探柯南第752集',
  70             },
  71         }, {
  72             'info_dict': {
  73                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part7',
  74                 'ext': 'f4v',
  75                 'title': '名侦探柯南第752集',
  76             },
  77         }, {
  78             'info_dict': {
  79                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part8',
  80                 'ext': 'f4v',
  81                 'title': '名侦探柯南第752集',
  82             },
  83         }],
  84         'params': {
  85             'skip_download': True,
  86         },
  87     }]
  88
  89     _FORMATS_MAP = [
  90         ('1', 'h6'),
  91         ('2', 'h5'),
  92         ('3', 'h4'),
  93         ('4', 'h3'),
  94         ('5', 'h2'),
  95         ('10', 'h1'),
  96     ]
  97
  98     @staticmethod
  99     def md5_text(text):
 100         return hashlib.md5(text.encode('utf-8')).hexdigest()
 101
 102     def construct_video_urls(self, data, video_id, _uuid):
 103         def do_xor(x, y):
 104             a = y % 3
 105             if a == 1:
 106                 return x ^ 121
 107             if a == 2:
 108                 return x ^ 72
 109             return x ^ 103
 110
 111         def get_encode_code(l):
 112             a = 0
 113             b = l.split('-')
 114             c = len(b)
 115             s = ''
 116             for i in range(c - 1, -1, -1):
 117                 a = do_xor(int(b[c - i - 1], 16), i)
 118                 s += chr(a)
 119             return s[::-1]
 120
 121         def get_path_key(x, format_id, segment_index):
 122             mg = ')(*&^flash@#$%a'
 123             tm = self._download_json(
 124                 'http://data.video.qiyi.com/t?tn=' + str(random.random()), video_id,
 125                 note='Download path key of segment %d for format %s' % (segment_index + 1, format_id)
 126             )['t']
 127             t = str(int(math.floor(int(tm) / (600.0))))
 128             return self.md5_text(t + mg + x)
 129
 130         video_urls_dict = {}
 131         for format_item in data['vp']['tkl'][0]['vs']:
 132             if 0 < int(format_item['bid']) <= 10:
 133                 format_id = self.get_format(format_item['bid'])
 134             else:
 135                 continue
 136
 137             video_urls = []
 138
 139             video_urls_info = format_item['fs']
 140             if not format_item['fs'][0]['l'].startswith('/'):
 141                 t = get_encode_code(format_item['fs'][0]['l'])
 142                 if t.endswith('mp4'):
 143                     video_urls_info = format_item['flvs']
 144
 145             for segment_index, segment in enumerate(video_urls_info):
 146                 vl = segment['l']
 147                 if not vl.startswith('/'):
 148                     vl = get_encode_code(vl)
 149                 key = get_path_key(
 150                     vl.split('/')[-1].split('.')[0], format_id, segment_index)
 151                 filesize = segment['b']
 152                 base_url = data['vp']['du'].split('/')
 153                 base_url.insert(-1, key)
 154                 base_url = '/'.join(base_url)
 155                 param = {
 156                     'su': _uuid,
 157                     'qyid': uuid.uuid4().hex,
 158                     'client': '',
 159                     'z': '',
 160                     'bt': '',
 161                     'ct': '',
 162                     'tn': str(int(time.time()))
 163                 }
 164                 api_video_url = base_url + vl + '?' + \
 165                     compat_urllib_parse.urlencode(param)
 166                 js = self._download_json(
 167                     api_video_url, video_id,
 168                     note='Download video info of segment %d for format %s' % (segment_index + 1, format_id))
 169                 video_url = js['l']
 170                 video_urls.append(
 171                     (video_url, filesize))
 172
 173             video_urls_dict[format_id] = video_urls
 174         return video_urls_dict
 175
 176     def get_format(self, bid):
 177         matched_format_ids = [_format_id for _bid, _format_id in self._FORMATS_MAP if _bid == str(bid)]
 178         return matched_format_ids[0] if len(matched_format_ids) else None
 179
 180     def get_bid(self, format_id):
 181         matched_bids = [_bid for _bid, _format_id in self._FORMATS_MAP if _format_id == format_id]
 182         return matched_bids[0] if len(matched_bids) else None
 183
 184     def get_raw_data(self, tvid, video_id, enc_key, _uuid):
 185         tm = str(int(time.time()))
 186         tail = tm + tvid
 187         param = {
 188             'key': 'fvip',
 189             'src': self.md5_text('youtube-dl'),
 190             'tvId': tvid,
 191             'vid': video_id,
 192             'vinfo': 1,
 193             'tm': tm,
 194             'enc': self.md5_text(enc_key + tail),
 195             'qyid': _uuid,
 196             'tn': random.random(),
 197             'um': 0,
 198             'authkey': self.md5_text(self.md5_text('') + tail),
 199         }
 200
 201         api_url = 'http://cache.video.qiyi.com/vms' + '?' + \
 202             compat_urllib_parse.urlencode(param)
 203         raw_data = self._download_json(api_url, video_id)
 204         return raw_data
 205
 206     def get_enc_key(self, swf_url, video_id):
 207         # TODO: automatic key extraction
 208         # last update at 2015-10-22 for Zombie::bite
 209         # '7223c67061dbea1259d0ceb44f44b6d62288f4f80c972170de5201d2321060270e05'[2:66][0::2]
 210         enc_key = '2c76de15dcb44bd28ff0927d50d31620'
 211         return enc_key
 212
 213     def _real_extract(self, url):
 214         webpage = self._download_webpage(
 215             url, 'temp_id', note='download video page')
 216         tvid = self._search_regex(
 217             r'data-player-tvid\s*=\s*[\'"](\d+)', webpage, 'tvid')
 218         video_id = self._search_regex(
 219             r'data-player-videoid\s*=\s*[\'"]([a-f\d]+)', webpage, 'video_id')
 220         swf_url = self._search_regex(
 221             r'(http://[^\'"]+MainPlayer[^.]+\.swf)', webpage, 'swf player URL')
 222         _uuid = uuid.uuid4().hex
 223
 224         enc_key = self.get_enc_key(swf_url, video_id)
 225
 226         raw_data = self.get_raw_data(tvid, video_id, enc_key, _uuid)
 227
 228         if raw_data['code'] != 'A000000':
 229             raise ExtractorError('Unable to load data. Error code: ' + raw_data['code'])
 230
 231         if not raw_data['data']['vp']['tkl']:
 232             raise ExtractorError('No support iQiqy VIP video')
 233
 234         data = raw_data['data']
 235
 236         title = data['vi']['vn']
 237
 238         # generate video_urls_dict
 239         video_urls_dict = self.construct_video_urls(
 240             data, video_id, _uuid)
 241
 242         # construct info
 243         entries = []
 244         for format_id in video_urls_dict:
 245             video_urls = video_urls_dict[format_id]
 246             for i, video_url_info in enumerate(video_urls):
 247                 if len(entries) < i + 1:
 248                     entries.append({'formats': []})
 249                 entries[i]['formats'].append(
 250                     {
 251                         'url': video_url_info[0],
 252                         'filesize': video_url_info[-1],
 253                         'format_id': format_id,
 254                         'preference': int(self.get_bid(format_id))
 255                     }
 256                 )
 257
 258         for i in range(len(entries)):
 259             self._sort_formats(entries[i]['formats'])
 260             entries[i].update(
 261                 {
 262                     'id': '%s_part%d' % (video_id, i + 1),
 263                     'title': title,
 264                 }
 265             )
 266
 267         if len(entries) > 1:
 268             info = {
 269                 '_type': 'multi_video',
 270                 'id': video_id,
 271                 'title': title,
 272                 'entries': entries,
 273             }
 274         else:
 275             info = entries[0]
 276             info['id'] = video_id
 277             info['title'] = title
 278
 279         return info