youtube_dl/extractor/iqiyi.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import hashlib
   5 import math
   6 import os.path
   7 import random
   8 import re
   9 import time
  10 import uuid
  11 import zlib
  12
  13 from .common import InfoExtractor
  14 from ..compat import compat_urllib_parse
  15 from ..utils import (
  16     ExtractorError,
  17     url_basename,
  18 )
  19
  20
  21 class IqiyiIE(InfoExtractor):
  22     IE_NAME = 'iqiyi'
  23
  24     _VALID_URL = r'http://(?:www\.)iqiyi.com/v_.+?\.html'
  25
  26     _TESTS = [{
  27         'url': 'http://www.iqiyi.com/v_19rrojlavg.html',
  28         'md5': '2cb594dc2781e6c941a110d8f358118b',
  29         'info_dict': {
  30             'id': '9c1fb1b99d192b21c559e5a1a2cb3c73',
  31             'title': '美国德州空中惊现奇异云团 酷似UFO',
  32             'ext': 'f4v',
  33         }
  34     }, {
  35         'url': 'http://www.iqiyi.com/v_19rrhnnclk.html',
  36         'info_dict': {
  37             'id': 'e3f585b550a280af23c98b6cb2be19fb',
  38             'title': '名侦探柯南第752集',
  39         },
  40         'playlist': [{
  41             'md5': '7e49376fecaffa115d951634917fe105',
  42             'info_dict': {
  43                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part1',
  44                 'ext': 'f4v',
  45                 'title': '名侦探柯南第752集',
  46             },
  47         }, {
  48             'md5': '41b75ba13bb7ac0e411131f92bc4f6ca',
  49             'info_dict': {
  50                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part2',
  51                 'ext': 'f4v',
  52                 'title': '名侦探柯南第752集',
  53             },
  54         }, {
  55             'md5': '0cee1dd0a3d46a83e71e2badeae2aab0',
  56             'info_dict': {
  57                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part3',
  58                 'ext': 'f4v',
  59                 'title': '名侦探柯南第752集',
  60             },
  61         }, {
  62             'md5': '4f8ad72373b0c491b582e7c196b0b1f9',
  63             'info_dict': {
  64                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part4',
  65                 'ext': 'f4v',
  66                 'title': '名侦探柯南第752集',
  67             },
  68         }, {
  69             'md5': 'd89ad028bcfad282918e8098e811711d',
  70             'info_dict': {
  71                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part5',
  72                 'ext': 'f4v',
  73                 'title': '名侦探柯南第752集',
  74             },
  75         }, {
  76             'md5': '9cb1e5c95da25dff0660c32ae50903b7',
  77             'info_dict': {
  78                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part6',
  79                 'ext': 'f4v',
  80                 'title': '名侦探柯南第752集',
  81             },
  82         }, {
  83             'md5': '155116e0ff1867bbc9b98df294faabc9',
  84             'info_dict': {
  85                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part7',
  86                 'ext': 'f4v',
  87                 'title': '名侦探柯南第752集',
  88             },
  89         }, {
  90             'md5': '53f5db77622ae14fa493ed2a278a082b',
  91             'info_dict': {
  92                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part8',
  93                 'ext': 'f4v',
  94                 'title': '名侦探柯南第752集',
  95             },
  96         }],
  97     }]
  98
  99     _FORMATS_MAP = [
 100         ('1', 'h6'),
 101         ('2', 'h5'),
 102         ('3', 'h4'),
 103         ('4', 'h3'),
 104         ('5', 'h2'),
 105         ('10', 'h1'),
 106     ]
 107
 108     def construct_video_urls(self, data, video_id, _uuid):
 109         def do_xor(x, y):
 110             a = y % 3
 111             if a == 1:
 112                 return x ^ 121
 113             if a == 2:
 114                 return x ^ 72
 115             return x ^ 103
 116
 117         def get_encode_code(l):
 118             a = 0
 119             b = l.split('-')
 120             c = len(b)
 121             s = ''
 122             for i in range(c - 1, -1, -1):
 123                 a = do_xor(int(b[c - i - 1], 16), i)
 124                 s += chr(a)
 125             return s[::-1]
 126
 127         def get_path_key(x, format_id, segment_index):
 128             mg = ')(*&^flash@#$%a'
 129             tm = self._download_json(
 130                 'http://data.video.qiyi.com/t?tn=' + str(random.random()), video_id,
 131                 note='Download path key of segment %d for format %s' % (segment_index + 1, format_id)
 132             )['t']
 133             t = str(int(math.floor(int(tm) / (600.0))))
 134             return hashlib.md5((t + mg + x).encode('utf8')).hexdigest()
 135
 136         video_urls_dict = {}
 137         for format_item in data['vp']['tkl'][0]['vs']:
 138             if 0 < int(format_item['bid']) <= 10:
 139                 format_id = self.get_format(format_item['bid'])
 140             else:
 141                 continue
 142
 143             video_urls = []
 144
 145             video_urls_info = format_item['fs']
 146             if not format_item['fs'][0]['l'].startswith('/'):
 147                 t = get_encode_code(format_item['fs'][0]['l'])
 148                 if t.endswith('mp4'):
 149                     video_urls_info = format_item['flvs']
 150
 151             for segment_index, segment in enumerate(video_urls_info):
 152                 vl = segment['l']
 153                 if not vl.startswith('/'):
 154                     vl = get_encode_code(vl)
 155                 key = get_path_key(
 156                     vl.split('/')[-1].split('.')[0], format_id, segment_index)
 157                 filesize = segment['b']
 158                 base_url = data['vp']['du'].split('/')
 159                 base_url.insert(-1, key)
 160                 base_url = '/'.join(base_url)
 161                 param = {
 162                     'su': _uuid,
 163                     'qyid': uuid.uuid4().hex,
 164                     'client': '',
 165                     'z': '',
 166                     'bt': '',
 167                     'ct': '',
 168                     'tn': str(int(time.time()))
 169                 }
 170                 api_video_url = base_url + vl + '?' + \
 171                     compat_urllib_parse.urlencode(param)
 172                 js = self._download_json(
 173                     api_video_url, video_id,
 174                     note='Download video info of segment %d for format %s' % (segment_index + 1, format_id))
 175                 video_url = js['l']
 176                 video_urls.append(
 177                     (video_url, filesize))
 178
 179             video_urls_dict[format_id] = video_urls
 180         return video_urls_dict
 181
 182     def get_format(self, bid):
 183         matched_format_ids = [_format_id for _bid, _format_id in self._FORMATS_MAP if _bid == str(bid)]
 184         return matched_format_ids[0] if len(matched_format_ids) else None
 185
 186     def get_bid(self, format_id):
 187         matched_bids = [_bid for _bid, _format_id in self._FORMATS_MAP if _format_id == format_id]
 188         return matched_bids[0] if len(matched_bids) else None
 189
 190     def get_raw_data(self, tvid, video_id, enc_key, _uuid):
 191         tm = str(int(time.time()))
 192         param = {
 193             'key': 'fvip',
 194             'src': hashlib.md5(b'youtube-dl').hexdigest(),
 195             'tvId': tvid,
 196             'vid': video_id,
 197             'vinfo': 1,
 198             'tm': tm,
 199             'enc': hashlib.md5(
 200                 (enc_key + tm + tvid).encode('utf8')).hexdigest(),
 201             'qyid': _uuid,
 202             'tn': random.random(),
 203             'um': 0,
 204             'authkey': hashlib.md5(
 205                 (tm + tvid).encode('utf8')).hexdigest()
 206         }
 207
 208         api_url = 'http://cache.video.qiyi.com/vms' + '?' + \
 209             compat_urllib_parse.urlencode(param)
 210         raw_data = self._download_json(api_url, video_id)
 211         return raw_data
 212
 213     def get_enc_key(self, swf_url, video_id):
 214         filename, _ = os.path.splitext(url_basename(swf_url))
 215         enc_key_json = self._downloader.cache.load('iqiyi-enc-key', filename)
 216         if enc_key_json is not None:
 217             return enc_key_json[0]
 218
 219         req = self._request_webpage(
 220             swf_url, video_id, note='download swf content')
 221         cn = req.read()
 222         cn = zlib.decompress(cn[8:])
 223         pt = re.compile(b'MixerRemote\x08(?P<enc_key>.+?)\$&vv')
 224         enc_key = self._search_regex(pt, cn, 'enc_key').decode('utf8')
 225
 226         self._downloader.cache.store('iqiyi-enc-key', filename, [enc_key])
 227
 228         return enc_key
 229
 230     def _real_extract(self, url):
 231         webpage = self._download_webpage(
 232             url, 'temp_id', note='download video page')
 233         tvid = self._search_regex(
 234             r'data-player-tvid\s*=\s*[\'"](\d+)', webpage, 'tvid')
 235         video_id = self._search_regex(
 236             r'data-player-videoid\s*=\s*[\'"]([a-f\d]+)', webpage, 'video_id')
 237         swf_url = self._search_regex(
 238             r'(http://[^\'"]+MainPlayer[^.]+\.swf)', webpage, 'swf player URL')
 239         _uuid = uuid.uuid4().hex
 240
 241         enc_key = self.get_enc_key(swf_url, video_id)
 242
 243         raw_data = self.get_raw_data(tvid, video_id, enc_key, _uuid)
 244
 245         if raw_data['code'] != 'A000000':
 246             raise ExtractorError('Unable to load data. Error code: ' + raw_data['code'])
 247
 248         if not raw_data['data']['vp']['tkl']:
 249             raise ExtractorError('No support iQiqy VIP video')
 250
 251         data = raw_data['data']
 252
 253         title = data['vi']['vn']
 254
 255         # generate video_urls_dict
 256         video_urls_dict = self.construct_video_urls(
 257             data, video_id, _uuid)
 258
 259         # construct info
 260         entries = []
 261         for format_id in video_urls_dict:
 262             video_urls = video_urls_dict[format_id]
 263             for i, video_url_info in enumerate(video_urls):
 264                 if len(entries) < i + 1:
 265                     entries.append({'formats': []})
 266                 entries[i]['formats'].append(
 267                     {
 268                         'url': video_url_info[0],
 269                         'filesize': video_url_info[-1],
 270                         'format_id': format_id,
 271                         'preference': int(self.get_bid(format_id))
 272                     }
 273                 )
 274
 275         for i in range(len(entries)):
 276             self._sort_formats(entries[i]['formats'])
 277             entries[i].update(
 278                 {
 279                     'id': '%s_part%d' % (video_id, i + 1),
 280                     'title': title,
 281                 }
 282             )
 283
 284         if len(entries) > 1:
 285             info = {
 286                 '_type': 'multi_video',
 287                 'id': video_id,
 288                 'title': title,
 289                 'entries': entries,
 290             }
 291         else:
 292             info = entries[0]
 293             info['id'] = video_id
 294             info['title'] = title
 295
 296         return info