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     def construct_video_urls(self, data, video_id, _uuid):
  99         def do_xor(x, y):
 100             a = y % 3
 101             if a == 1:
 102                 return x ^ 121
 103             if a == 2:
 104                 return x ^ 72
 105             return x ^ 103
 106
 107         def get_encode_code(l):
 108             a = 0
 109             b = l.split('-')
 110             c = len(b)
 111             s = ''
 112             for i in range(c - 1, -1, -1):
 113                 a = do_xor(int(b[c - i - 1], 16), i)
 114                 s += chr(a)
 115             return s[::-1]
 116
 117         def get_path_key(x, format_id, segment_index):
 118             mg = ')(*&^flash@#$%a'
 119             tm = self._download_json(
 120                 'http://data.video.qiyi.com/t?tn=' + str(random.random()), video_id,
 121                 note='Download path key of segment %d for format %s' % (segment_index + 1, format_id)
 122             )['t']
 123             t = str(int(math.floor(int(tm) / (600.0))))
 124             return hashlib.md5((t + mg + x).encode('utf8')).hexdigest()
 125
 126         video_urls_dict = {}
 127         for format_item in data['vp']['tkl'][0]['vs']:
 128             if 0 < int(format_item['bid']) <= 10:
 129                 format_id = self.get_format(format_item['bid'])
 130             else:
 131                 continue
 132
 133             video_urls = []
 134
 135             video_urls_info = format_item['fs']
 136             if not format_item['fs'][0]['l'].startswith('/'):
 137                 t = get_encode_code(format_item['fs'][0]['l'])
 138                 if t.endswith('mp4'):
 139                     video_urls_info = format_item['flvs']
 140
 141             for segment_index, segment in enumerate(video_urls_info):
 142                 vl = segment['l']
 143                 if not vl.startswith('/'):
 144                     vl = get_encode_code(vl)
 145                 key = get_path_key(
 146                     vl.split('/')[-1].split('.')[0], format_id, segment_index)
 147                 filesize = segment['b']
 148                 base_url = data['vp']['du'].split('/')
 149                 base_url.insert(-1, key)
 150                 base_url = '/'.join(base_url)
 151                 param = {
 152                     'su': _uuid,
 153                     'qyid': uuid.uuid4().hex,
 154                     'client': '',
 155                     'z': '',
 156                     'bt': '',
 157                     'ct': '',
 158                     'tn': str(int(time.time()))
 159                 }
 160                 api_video_url = base_url + vl + '?' + \
 161                     compat_urllib_parse.urlencode(param)
 162                 js = self._download_json(
 163                     api_video_url, video_id,
 164                     note='Download video info of segment %d for format %s' % (segment_index + 1, format_id))
 165                 video_url = js['l']
 166                 video_urls.append(
 167                     (video_url, filesize))
 168
 169             video_urls_dict[format_id] = video_urls
 170         return video_urls_dict
 171
 172     def get_format(self, bid):
 173         matched_format_ids = [_format_id for _bid, _format_id in self._FORMATS_MAP if _bid == str(bid)]
 174         return matched_format_ids[0] if len(matched_format_ids) else None
 175
 176     def get_bid(self, format_id):
 177         matched_bids = [_bid for _bid, _format_id in self._FORMATS_MAP if _format_id == format_id]
 178         return matched_bids[0] if len(matched_bids) else None
 179
 180     def get_raw_data(self, tvid, video_id, enc_key, _uuid):
 181         tm = str(int(time.time()))
 182         param = {
 183             'key': 'fvip',
 184             'src': hashlib.md5(b'youtube-dl').hexdigest(),
 185             'tvId': tvid,
 186             'vid': video_id,
 187             'vinfo': 1,
 188             'tm': tm,
 189             'enc': hashlib.md5(
 190                 (enc_key + tm + tvid).encode('utf8')).hexdigest(),
 191             'qyid': _uuid,
 192             'tn': random.random(),
 193             'um': 0,
 194             'authkey': hashlib.md5(
 195                 (tm + tvid).encode('utf8')).hexdigest()
 196         }
 197
 198         api_url = 'http://cache.video.qiyi.com/vms' + '?' + \
 199             compat_urllib_parse.urlencode(param)
 200         raw_data = self._download_json(api_url, video_id)
 201         return raw_data
 202
 203     def get_enc_key(self, swf_url, video_id):
 204         enc_key = '8e29ab5666d041c3a1ea76e06dabdffb'
 205         return enc_key
 206
 207     def _real_extract(self, url):
 208         webpage = self._download_webpage(
 209             url, 'temp_id', note='download video page')
 210         tvid = self._search_regex(
 211             r'data-player-tvid\s*=\s*[\'"](\d+)', webpage, 'tvid')
 212         video_id = self._search_regex(
 213             r'data-player-videoid\s*=\s*[\'"]([a-f\d]+)', webpage, 'video_id')
 214         swf_url = self._search_regex(
 215             r'(http://[^\'"]+MainPlayer[^.]+\.swf)', webpage, 'swf player URL')
 216         _uuid = uuid.uuid4().hex
 217
 218         enc_key = self.get_enc_key(swf_url, video_id)
 219
 220         raw_data = self.get_raw_data(tvid, video_id, enc_key, _uuid)
 221
 222         if raw_data['code'] != 'A000000':
 223             raise ExtractorError('Unable to load data. Error code: ' + raw_data['code'])
 224
 225         if not raw_data['data']['vp']['tkl']:
 226             raise ExtractorError('No support iQiqy VIP video')
 227
 228         data = raw_data['data']
 229
 230         title = data['vi']['vn']
 231
 232         # generate video_urls_dict
 233         video_urls_dict = self.construct_video_urls(
 234             data, video_id, _uuid)
 235
 236         # construct info
 237         entries = []
 238         for format_id in video_urls_dict:
 239             video_urls = video_urls_dict[format_id]
 240             for i, video_url_info in enumerate(video_urls):
 241                 if len(entries) < i + 1:
 242                     entries.append({'formats': []})
 243                 entries[i]['formats'].append(
 244                     {
 245                         'url': video_url_info[0],
 246                         'filesize': video_url_info[-1],
 247                         'format_id': format_id,
 248                         'preference': int(self.get_bid(format_id))
 249                     }
 250                 )
 251
 252         for i in range(len(entries)):
 253             self._sort_formats(entries[i]['formats'])
 254             entries[i].update(
 255                 {
 256                     'id': '%s_part%d' % (video_id, i + 1),
 257                     'title': title,
 258                 }
 259             )
 260
 261         if len(entries) > 1:
 262             info = {
 263                 '_type': 'multi_video',
 264                 'id': video_id,
 265                 'title': title,
 266                 'entries': entries,
 267             }
 268         else:
 269             info = entries[0]
 270             info['id'] = video_id
 271             info['title'] = title
 272
 273         return info