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     }]
  85
  86     _FORMATS_MAP = [
  87         ('1', 'h6'),
  88         ('2', 'h5'),
  89         ('3', 'h4'),
  90         ('4', 'h3'),
  91         ('5', 'h2'),
  92         ('10', 'h1'),
  93     ]
  94
  95     def construct_video_urls(self, data, video_id, _uuid):
  96         def do_xor(x, y):
  97             a = y % 3
  98             if a == 1:
  99                 return x ^ 121
 100             if a == 2:
 101                 return x ^ 72
 102             return x ^ 103
 103
 104         def get_encode_code(l):
 105             a = 0
 106             b = l.split('-')
 107             c = len(b)
 108             s = ''
 109             for i in range(c - 1, -1, -1):
 110                 a = do_xor(int(b[c - i - 1], 16), i)
 111                 s += chr(a)
 112             return s[::-1]
 113
 114         def get_path_key(x, format_id, segment_index):
 115             mg = ')(*&^flash@#$%a'
 116             tm = self._download_json(
 117                 'http://data.video.qiyi.com/t?tn=' + str(random.random()), video_id,
 118                 note='Download path key of segment %d for format %s' % (segment_index + 1, format_id)
 119             )['t']
 120             t = str(int(math.floor(int(tm) / (600.0))))
 121             return hashlib.md5((t + mg + x).encode('utf8')).hexdigest()
 122
 123         video_urls_dict = {}
 124         for format_item in data['vp']['tkl'][0]['vs']:
 125             if 0 < int(format_item['bid']) <= 10:
 126                 format_id = self.get_format(format_item['bid'])
 127             else:
 128                 continue
 129
 130             video_urls = []
 131
 132             video_urls_info = format_item['fs']
 133             if not format_item['fs'][0]['l'].startswith('/'):
 134                 t = get_encode_code(format_item['fs'][0]['l'])
 135                 if t.endswith('mp4'):
 136                     video_urls_info = format_item['flvs']
 137
 138             for segment_index, segment in enumerate(video_urls_info):
 139                 vl = segment['l']
 140                 if not vl.startswith('/'):
 141                     vl = get_encode_code(vl)
 142                 key = get_path_key(
 143                     vl.split('/')[-1].split('.')[0], format_id, segment_index)
 144                 filesize = segment['b']
 145                 base_url = data['vp']['du'].split('/')
 146                 base_url.insert(-1, key)
 147                 base_url = '/'.join(base_url)
 148                 param = {
 149                     'su': _uuid,
 150                     'qyid': uuid.uuid4().hex,
 151                     'client': '',
 152                     'z': '',
 153                     'bt': '',
 154                     'ct': '',
 155                     'tn': str(int(time.time()))
 156                 }
 157                 api_video_url = base_url + vl + '?' + \
 158                     compat_urllib_parse.urlencode(param)
 159                 js = self._download_json(
 160                     api_video_url, video_id,
 161                     note='Download video info of segment %d for format %s' % (segment_index + 1, format_id))
 162                 video_url = js['l']
 163                 video_urls.append(
 164                     (video_url, filesize))
 165
 166             video_urls_dict[format_id] = video_urls
 167         return video_urls_dict
 168
 169     def get_format(self, bid):
 170         matched_format_ids = [_format_id for _bid, _format_id in self._FORMATS_MAP if _bid == str(bid)]
 171         return matched_format_ids[0] if len(matched_format_ids) else None
 172
 173     def get_bid(self, format_id):
 174         matched_bids = [_bid for _bid, _format_id in self._FORMATS_MAP if _format_id == format_id]
 175         return matched_bids[0] if len(matched_bids) else None
 176
 177     def get_raw_data(self, tvid, video_id, enc_key, _uuid):
 178         tm = str(int(time.time()))
 179         param = {
 180             'key': 'fvip',
 181             'src': hashlib.md5(b'youtube-dl').hexdigest(),
 182             'tvId': tvid,
 183             'vid': video_id,
 184             'vinfo': 1,
 185             'tm': tm,
 186             'enc': hashlib.md5(
 187                 (enc_key + tm + tvid).encode('utf8')).hexdigest(),
 188             'qyid': _uuid,
 189             'tn': random.random(),
 190             'um': 0,
 191             'authkey': hashlib.md5(
 192                 (tm + tvid).encode('utf8')).hexdigest()
 193         }
 194
 195         api_url = 'http://cache.video.qiyi.com/vms' + '?' + \
 196             compat_urllib_parse.urlencode(param)
 197         raw_data = self._download_json(api_url, video_id)
 198         return raw_data
 199
 200     def get_enc_key(self, swf_url, video_id):
 201         enc_key = '8e29ab5666d041c3a1ea76e06dabdffb'
 202         return enc_key
 203
 204     def _real_extract(self, url):
 205         webpage = self._download_webpage(
 206             url, 'temp_id', note='download video page')
 207         tvid = self._search_regex(
 208             r'data-player-tvid\s*=\s*[\'"](\d+)', webpage, 'tvid')
 209         video_id = self._search_regex(
 210             r'data-player-videoid\s*=\s*[\'"]([a-f\d]+)', webpage, 'video_id')
 211         swf_url = self._search_regex(
 212             r'(http://[^\'"]+MainPlayer[^.]+\.swf)', webpage, 'swf player URL')
 213         _uuid = uuid.uuid4().hex
 214
 215         enc_key = self.get_enc_key(swf_url, video_id)
 216
 217         raw_data = self.get_raw_data(tvid, video_id, enc_key, _uuid)
 218
 219         if raw_data['code'] != 'A000000':
 220             raise ExtractorError('Unable to load data. Error code: ' + raw_data['code'])
 221
 222         if not raw_data['data']['vp']['tkl']:
 223             raise ExtractorError('No support iQiqy VIP video')
 224
 225         data = raw_data['data']
 226
 227         title = data['vi']['vn']
 228
 229         # generate video_urls_dict
 230         video_urls_dict = self.construct_video_urls(
 231             data, video_id, _uuid)
 232
 233         # construct info
 234         entries = []
 235         for format_id in video_urls_dict:
 236             video_urls = video_urls_dict[format_id]
 237             for i, video_url_info in enumerate(video_urls):
 238                 if len(entries) < i + 1:
 239                     entries.append({'formats': []})
 240                 entries[i]['formats'].append(
 241                     {
 242                         'url': video_url_info[0],
 243                         'filesize': video_url_info[-1],
 244                         'format_id': format_id,
 245                         'preference': int(self.get_bid(format_id))
 246                     }
 247                 )
 248
 249         for i in range(len(entries)):
 250             self._sort_formats(entries[i]['formats'])
 251             entries[i].update(
 252                 {
 253                     'id': '%s_part%d' % (video_id, i + 1),
 254                     'title': title,
 255                 }
 256             )
 257
 258         if len(entries) > 1:
 259             info = {
 260                 '_type': 'multi_video',
 261                 'id': video_id,
 262                 'title': title,
 263                 'entries': entries,
 264             }
 265         else:
 266             info = entries[0]
 267             info['id'] = video_id
 268             info['title'] = title
 269
 270         return info