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://(?:[^.]+\.)?iqiyi\.com/.+\.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         'url': 'http://www.iqiyi.com/w_19rt6o8t9p.html',
  89         'only_matching': True,
  90     }, {
  91         'url': 'http://www.iqiyi.com/a_19rrhbc6kt.html',
  92         'only_matching': True,
  93     }, {
  94         'url': 'http://yule.iqiyi.com/pcb.html',
  95         'only_matching': True,
  96     }]
  97
  98     _FORMATS_MAP = [
  99         ('1', 'h6'),
 100         ('2', 'h5'),
 101         ('3', 'h4'),
 102         ('4', 'h3'),
 103         ('5', 'h2'),
 104         ('10', 'h1'),
 105     ]
 106
 107     @staticmethod
 108     def md5_text(text):
 109         return hashlib.md5(text.encode('utf-8')).hexdigest()
 110
 111     def construct_video_urls(self, data, video_id, _uuid):
 112         def do_xor(x, y):
 113             a = y % 3
 114             if a == 1:
 115                 return x ^ 121
 116             if a == 2:
 117                 return x ^ 72
 118             return x ^ 103
 119
 120         def get_encode_code(l):
 121             a = 0
 122             b = l.split('-')
 123             c = len(b)
 124             s = ''
 125             for i in range(c - 1, -1, -1):
 126                 a = do_xor(int(b[c - i - 1], 16), i)
 127                 s += chr(a)
 128             return s[::-1]
 129
 130         def get_path_key(x, format_id, segment_index):
 131             mg = ')(*&^flash@#$%a'
 132             tm = self._download_json(
 133                 'http://data.video.qiyi.com/t?tn=' + str(random.random()), video_id,
 134                 note='Download path key of segment %d for format %s' % (segment_index + 1, format_id)
 135             )['t']
 136             t = str(int(math.floor(int(tm) / (600.0))))
 137             return self.md5_text(t + mg + x)
 138
 139         video_urls_dict = {}
 140         for format_item in data['vp']['tkl'][0]['vs']:
 141             if 0 < int(format_item['bid']) <= 10:
 142                 format_id = self.get_format(format_item['bid'])
 143             else:
 144                 continue
 145
 146             video_urls = []
 147
 148             video_urls_info = format_item['fs']
 149             if not format_item['fs'][0]['l'].startswith('/'):
 150                 t = get_encode_code(format_item['fs'][0]['l'])
 151                 if t.endswith('mp4'):
 152                     video_urls_info = format_item['flvs']
 153
 154             for segment_index, segment in enumerate(video_urls_info):
 155                 vl = segment['l']
 156                 if not vl.startswith('/'):
 157                     vl = get_encode_code(vl)
 158                 key = get_path_key(
 159                     vl.split('/')[-1].split('.')[0], format_id, segment_index)
 160                 filesize = segment['b']
 161                 base_url = data['vp']['du'].split('/')
 162                 base_url.insert(-1, key)
 163                 base_url = '/'.join(base_url)
 164                 param = {
 165                     'su': _uuid,
 166                     'qyid': uuid.uuid4().hex,
 167                     'client': '',
 168                     'z': '',
 169                     'bt': '',
 170                     'ct': '',
 171                     'tn': str(int(time.time()))
 172                 }
 173                 api_video_url = base_url + vl + '?' + \
 174                     compat_urllib_parse.urlencode(param)
 175                 js = self._download_json(
 176                     api_video_url, video_id,
 177                     note='Download video info of segment %d for format %s' % (segment_index + 1, format_id))
 178                 video_url = js['l']
 179                 video_urls.append(
 180                     (video_url, filesize))
 181
 182             video_urls_dict[format_id] = video_urls
 183         return video_urls_dict
 184
 185     def get_format(self, bid):
 186         matched_format_ids = [_format_id for _bid, _format_id in self._FORMATS_MAP if _bid == str(bid)]
 187         return matched_format_ids[0] if len(matched_format_ids) else None
 188
 189     def get_bid(self, format_id):
 190         matched_bids = [_bid for _bid, _format_id in self._FORMATS_MAP if _format_id == format_id]
 191         return matched_bids[0] if len(matched_bids) else None
 192
 193     def get_raw_data(self, tvid, video_id, enc_key, _uuid):
 194         tm = str(int(time.time()))
 195         tail = tm + tvid
 196         param = {
 197             'key': 'fvip',
 198             'src': self.md5_text('youtube-dl'),
 199             'tvId': tvid,
 200             'vid': video_id,
 201             'vinfo': 1,
 202             'tm': tm,
 203             'enc': self.md5_text(enc_key + tail),
 204             'qyid': _uuid,
 205             'tn': random.random(),
 206             'um': 0,
 207             'authkey': self.md5_text(self.md5_text('') + tail),
 208         }
 209
 210         api_url = 'http://cache.video.qiyi.com/vms' + '?' + \
 211             compat_urllib_parse.urlencode(param)
 212         raw_data = self._download_json(api_url, video_id)
 213         return raw_data
 214
 215     def get_enc_key(self, swf_url, video_id):
 216         # TODO: automatic key extraction
 217         # last update at 2015-12-18 for Zombie::bite
 218         enc_key = '8b6b683780897eb8d9a48a02ccc4817d'[::-1]
 219         return enc_key
 220
 221     def _real_extract(self, url):
 222         webpage = self._download_webpage(
 223             url, 'temp_id', note='download video page')
 224         tvid = self._search_regex(
 225             r'data-player-tvid\s*=\s*[\'"](\d+)', webpage, 'tvid')
 226         video_id = self._search_regex(
 227             r'data-player-videoid\s*=\s*[\'"]([a-f\d]+)', webpage, 'video_id')
 228         swf_url = self._search_regex(
 229             r'(http://[^\'"]+MainPlayer[^.]+\.swf)', webpage, 'swf player URL')
 230         _uuid = uuid.uuid4().hex
 231
 232         enc_key = self.get_enc_key(swf_url, video_id)
 233
 234         raw_data = self.get_raw_data(tvid, video_id, enc_key, _uuid)
 235
 236         if raw_data['code'] != 'A000000':
 237             raise ExtractorError('Unable to load data. Error code: ' + raw_data['code'])
 238
 239         if not raw_data['data']['vp']['tkl']:
 240             raise ExtractorError('No support iQiqy VIP video')
 241
 242         data = raw_data['data']
 243
 244         title = data['vi']['vn']
 245
 246         # generate video_urls_dict
 247         video_urls_dict = self.construct_video_urls(
 248             data, video_id, _uuid)
 249
 250         # construct info
 251         entries = []
 252         for format_id in video_urls_dict:
 253             video_urls = video_urls_dict[format_id]
 254             for i, video_url_info in enumerate(video_urls):
 255                 if len(entries) < i + 1:
 256                     entries.append({'formats': []})
 257                 entries[i]['formats'].append(
 258                     {
 259                         'url': video_url_info[0],
 260                         'filesize': video_url_info[-1],
 261                         'format_id': format_id,
 262                         'preference': int(self.get_bid(format_id))
 263                     }
 264                 )
 265
 266         for i in range(len(entries)):
 267             self._sort_formats(entries[i]['formats'])
 268             entries[i].update(
 269                 {
 270                     'id': '%s_part%d' % (video_id, i + 1),
 271                     'title': title,
 272                 }
 273             )
 274
 275         if len(entries) > 1:
 276             info = {
 277                 '_type': 'multi_video',
 278                 'id': video_id,
 279                 'title': title,
 280                 'entries': entries,
 281             }
 282         else:
 283             info = entries[0]
 284             info['id'] = video_id
 285             info['title'] = title
 286
 287         return info