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 re
   8 import time
   9 import uuid
  10 import zlib
  11
  12 from .common import InfoExtractor
  13 from ..compat import compat_urllib_parse
  14 from ..utils import ExtractorError
  15
  16
  17 class IqiyiIE(InfoExtractor):
  18     IE_NAME = 'iqiyi'
  19
  20     _VALID_URL = r'http://(?:www\.)iqiyi.com/v_.+?\.html'
  21
  22     _TESTS = [{
  23         'url': 'http://www.iqiyi.com/v_19rrojlavg.html',
  24         'md5': '2cb594dc2781e6c941a110d8f358118b',
  25         'info_dict': {
  26             'id': '9c1fb1b99d192b21c559e5a1a2cb3c73',
  27             'title': '美国德州空中惊现奇异云团 酷似UFO',
  28             'ext': 'f4v',
  29         }
  30     }, {
  31         'url': 'http://www.iqiyi.com/v_19rrhnnclk.html',
  32         'info_dict': {
  33             'id': 'e3f585b550a280af23c98b6cb2be19fb',
  34             'title': '名侦探柯南第752集',
  35         },
  36         'playlist': [{
  37             'md5': '7e49376fecaffa115d951634917fe105',
  38             'info_dict': {
  39                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part1',
  40                 'ext': 'f4v',
  41                 'title': '名侦探柯南第752集',
  42             },
  43         }, {
  44             'md5': '41b75ba13bb7ac0e411131f92bc4f6ca',
  45             'info_dict': {
  46                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part2',
  47                 'ext': 'f4v',
  48                 'title': '名侦探柯南第752集',
  49             },
  50         }, {
  51             'md5': '0cee1dd0a3d46a83e71e2badeae2aab0',
  52             'info_dict': {
  53                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part3',
  54                 'ext': 'f4v',
  55                 'title': '名侦探柯南第752集',
  56             },
  57         }, {
  58             'md5': '4f8ad72373b0c491b582e7c196b0b1f9',
  59             'info_dict': {
  60                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part4',
  61                 'ext': 'f4v',
  62                 'title': '名侦探柯南第752集',
  63             },
  64         }, {
  65             'md5': 'd89ad028bcfad282918e8098e811711d',
  66             'info_dict': {
  67                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part5',
  68                 'ext': 'f4v',
  69                 'title': '名侦探柯南第752集',
  70             },
  71         }, {
  72             'md5': '9cb1e5c95da25dff0660c32ae50903b7',
  73             'info_dict': {
  74                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part6',
  75                 'ext': 'f4v',
  76                 'title': '名侦探柯南第752集',
  77             },
  78         }, {
  79             'md5': '155116e0ff1867bbc9b98df294faabc9',
  80             'info_dict': {
  81                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part7',
  82                 'ext': 'f4v',
  83                 'title': '名侦探柯南第752集',
  84             },
  85         }, {
  86             'md5': '53f5db77622ae14fa493ed2a278a082b',
  87             'info_dict': {
  88                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part8',
  89                 'ext': 'f4v',
  90                 'title': '名侦探柯南第752集',
  91             },
  92         }],
  93     }]
  94
  95     _FORMATS_MAP = [
  96         ('1', 'h6'),
  97         ('2', 'h5'),
  98         ('3', 'h4'),
  99         ('4', 'h3'),
 100         ('5', 'h2'),
 101         ('10', 'h1'),
 102     ]
 103
 104     def construct_video_urls(self, data, video_id, _uuid):
 105         def do_xor(x, y):
 106             a = y % 3
 107             if a == 1:
 108                 return x ^ 121
 109             if a == 2:
 110                 return x ^ 72
 111             return x ^ 103
 112
 113         def get_encode_code(l):
 114             a = 0
 115             b = l.split('-')
 116             c = len(b)
 117             s = ''
 118             for i in range(c - 1, -1, -1):
 119                 a = do_xor(int(b[c - i - 1], 16), i)
 120                 s += chr(a)
 121             return s[::-1]
 122
 123         def get_path_key(x, format_id, segment_index):
 124             mg = ')(*&^flash@#$%a'
 125             tm = self._download_json(
 126                 'http://data.video.qiyi.com/t?tn=' + str(random.random()), video_id,
 127                 note='Download path key of segment %d for format %s' % (segment_index + 1, format_id)
 128             )['t']
 129             t = str(int(math.floor(int(tm) / (600.0))))
 130             return hashlib.md5((t + mg + x).encode('utf8')).hexdigest()
 131
 132         video_urls_dict = {}
 133         for format_item in data['vp']['tkl'][0]['vs']:
 134             if 0 < int(format_item['bid']) <= 10:
 135                 format_id = self.get_format(format_item['bid'])
 136             else:
 137                 continue
 138
 139             video_urls = []
 140
 141             video_urls_info = format_item['fs']
 142             if not format_item['fs'][0]['l'].startswith('/'):
 143                 t = get_encode_code(format_item['fs'][0]['l'])
 144                 if t.endswith('mp4'):
 145                     video_urls_info = format_item['flvs']
 146
 147             for segment_index, segment in enumerate(video_urls_info):
 148                 vl = segment['l']
 149                 if not vl.startswith('/'):
 150                     vl = get_encode_code(vl)
 151                 key = get_path_key(
 152                     vl.split('/')[-1].split('.')[0], format_id, segment_index)
 153                 filesize = segment['b']
 154                 base_url = data['vp']['du'].split('/')
 155                 base_url.insert(-1, key)
 156                 base_url = '/'.join(base_url)
 157                 param = {
 158                     'su': _uuid,
 159                     'qyid': uuid.uuid4().hex,
 160                     'client': '',
 161                     'z': '',
 162                     'bt': '',
 163                     'ct': '',
 164                     'tn': str(int(time.time()))
 165                 }
 166                 api_video_url = base_url + vl + '?' + \
 167                     compat_urllib_parse.urlencode(param)
 168                 js = self._download_json(
 169                     api_video_url, video_id,
 170                     note='Download video info of segment %d for format %s' % (segment_index + 1, format_id))
 171                 video_url = js['l']
 172                 video_urls.append(
 173                     (video_url, filesize))
 174
 175             video_urls_dict[format_id] = video_urls
 176         return video_urls_dict
 177
 178     def get_format(self, bid):
 179         matched_format_ids = [_format_id for _bid, _format_id in self._FORMATS_MAP if _bid == str(bid)]
 180         return matched_format_ids[0] if len(matched_format_ids) else None
 181
 182     def get_bid(self, format_id):
 183         matched_bids = [_bid for _bid, _format_id in self._FORMATS_MAP if _format_id == format_id]
 184         return matched_bids[0] if len(matched_bids) else None
 185
 186     def get_raw_data(self, tvid, video_id, enc_key, _uuid):
 187         tm = str(int(time.time()))
 188         param = {
 189             'key': 'fvip',
 190             'src': hashlib.md5(b'youtube-dl').hexdigest(),
 191             'tvId': tvid,
 192             'vid': video_id,
 193             'vinfo': 1,
 194             'tm': tm,
 195             'enc': hashlib.md5(
 196                 (enc_key + tm + tvid).encode('utf8')).hexdigest(),
 197             'qyid': _uuid,
 198             'tn': random.random(),
 199             'um': 0,
 200             'authkey': hashlib.md5(
 201                 (tm + tvid).encode('utf8')).hexdigest()
 202         }
 203
 204         api_url = 'http://cache.video.qiyi.com/vms' + '?' + \
 205             compat_urllib_parse.urlencode(param)
 206         raw_data = self._download_json(api_url, video_id)
 207         return raw_data
 208
 209     def get_enc_key(self, swf_url, video_id):
 210         req = self._request_webpage(
 211             swf_url, video_id, note='download swf content')
 212         cn = req.read()
 213         cn = zlib.decompress(cn[8:])
 214         pt = re.compile(b'MixerRemote\x08(?P<enc_key>.+?)\$&vv')
 215         enc_key = self._search_regex(pt, cn, 'enc_key').decode('utf8')
 216         return enc_key
 217
 218     def _real_extract(self, url):
 219         webpage = self._download_webpage(
 220             url, 'temp_id', note='download video page')
 221         tvid = self._search_regex(
 222             r'data-player-tvid\s*=\s*[\'"](\d+)', webpage, 'tvid')
 223         video_id = self._search_regex(
 224             r'data-player-videoid\s*=\s*[\'"]([a-f\d]+)', webpage, 'video_id')
 225         swf_url = self._search_regex(
 226             r'(http://[^\'"]+MainPlayer[^.]+\.swf)', webpage, 'swf player URL')
 227         _uuid = uuid.uuid4().hex
 228
 229         enc_key = self.get_enc_key(swf_url, video_id)
 230
 231         raw_data = self.get_raw_data(tvid, video_id, enc_key, _uuid)
 232
 233         if raw_data['code'] != 'A000000':
 234             raise ExtractorError('Unable to load data. Error code: ' + raw_data['code'])
 235
 236         if not raw_data['data']['vp']['tkl']:
 237             raise ExtractorError('No support iQiqy VIP video')
 238
 239         data = raw_data['data']
 240
 241         title = data['vi']['vn']
 242
 243         # generate video_urls_dict
 244         video_urls_dict = self.construct_video_urls(
 245             data, video_id, _uuid)
 246
 247         # construct info
 248         entries = []
 249         for format_id in video_urls_dict:
 250             video_urls = video_urls_dict[format_id]
 251             for i, video_url_info in enumerate(video_urls):
 252                 if len(entries) < i + 1:
 253                     entries.append({'formats': []})
 254                 entries[i]['formats'].append(
 255                     {
 256                         'url': video_url_info[0],
 257                         'filesize': video_url_info[-1],
 258                         'format_id': format_id,
 259                         'preference': int(self.get_bid(format_id))
 260                     }
 261                 )
 262
 263         for i in range(len(entries)):
 264             self._sort_formats(entries[i]['formats'])
 265             entries[i].update(
 266                 {
 267                     'id': '%s_part%d' % (video_id, i + 1),
 268                     'title': title,
 269                 }
 270             )
 271
 272         if len(entries) > 1:
 273             info = {
 274                 '_type': 'multi_video',
 275                 'id': video_id,
 276                 'title': title,
 277                 'entries': entries,
 278             }
 279         else:
 280             info = entries[0]
 281             info['id'] = video_id
 282             info['title'] = title
 283
 284         return info