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     IE_DESC = '爱奇艺'
  24
  25     _VALID_URL = r'http://(?:www\.)iqiyi.com/v_.+?\.html'
  26
  27     _TESTS = [{
  28         'url': 'http://www.iqiyi.com/v_19rrojlavg.html',
  29         'md5': '2cb594dc2781e6c941a110d8f358118b',
  30         'info_dict': {
  31             'id': '9c1fb1b99d192b21c559e5a1a2cb3c73',
  32             'title': '美国德州空中惊现奇异云团 酷似UFO',
  33             'ext': 'f4v',
  34         }
  35     }, {
  36         'url': 'http://www.iqiyi.com/v_19rrhnnclk.html',
  37         'info_dict': {
  38             'id': 'e3f585b550a280af23c98b6cb2be19fb',
  39             'title': '名侦探柯南第752集',
  40         },
  41         'playlist': [{
  42             'md5': '7e49376fecaffa115d951634917fe105',
  43             'info_dict': {
  44                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part1',
  45                 'ext': 'f4v',
  46                 'title': '名侦探柯南第752集',
  47             },
  48         }, {
  49             'md5': '41b75ba13bb7ac0e411131f92bc4f6ca',
  50             'info_dict': {
  51                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part2',
  52                 'ext': 'f4v',
  53                 'title': '名侦探柯南第752集',
  54             },
  55         }, {
  56             'md5': '0cee1dd0a3d46a83e71e2badeae2aab0',
  57             'info_dict': {
  58                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part3',
  59                 'ext': 'f4v',
  60                 'title': '名侦探柯南第752集',
  61             },
  62         }, {
  63             'md5': '4f8ad72373b0c491b582e7c196b0b1f9',
  64             'info_dict': {
  65                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part4',
  66                 'ext': 'f4v',
  67                 'title': '名侦探柯南第752集',
  68             },
  69         }, {
  70             'md5': 'd89ad028bcfad282918e8098e811711d',
  71             'info_dict': {
  72                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part5',
  73                 'ext': 'f4v',
  74                 'title': '名侦探柯南第752集',
  75             },
  76         }, {
  77             'md5': '9cb1e5c95da25dff0660c32ae50903b7',
  78             'info_dict': {
  79                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part6',
  80                 'ext': 'f4v',
  81                 'title': '名侦探柯南第752集',
  82             },
  83         }, {
  84             'md5': '155116e0ff1867bbc9b98df294faabc9',
  85             'info_dict': {
  86                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part7',
  87                 'ext': 'f4v',
  88                 'title': '名侦探柯南第752集',
  89             },
  90         }, {
  91             'md5': '53f5db77622ae14fa493ed2a278a082b',
  92             'info_dict': {
  93                 'id': 'e3f585b550a280af23c98b6cb2be19fb_part8',
  94                 'ext': 'f4v',
  95                 'title': '名侦探柯南第752集',
  96             },
  97         }],
  98     }]
  99
 100     _FORMATS_MAP = [
 101         ('1', 'h6'),
 102         ('2', 'h5'),
 103         ('3', 'h4'),
 104         ('4', 'h3'),
 105         ('5', 'h2'),
 106         ('10', 'h1'),
 107     ]
 108
 109     def construct_video_urls(self, data, video_id, _uuid):
 110         def do_xor(x, y):
 111             a = y % 3
 112             if a == 1:
 113                 return x ^ 121
 114             if a == 2:
 115                 return x ^ 72
 116             return x ^ 103
 117
 118         def get_encode_code(l):
 119             a = 0
 120             b = l.split('-')
 121             c = len(b)
 122             s = ''
 123             for i in range(c - 1, -1, -1):
 124                 a = do_xor(int(b[c - i - 1], 16), i)
 125                 s += chr(a)
 126             return s[::-1]
 127
 128         def get_path_key(x, format_id, segment_index):
 129             mg = ')(*&^flash@#$%a'
 130             tm = self._download_json(
 131                 'http://data.video.qiyi.com/t?tn=' + str(random.random()), video_id,
 132                 note='Download path key of segment %d for format %s' % (segment_index + 1, format_id)
 133             )['t']
 134             t = str(int(math.floor(int(tm) / (600.0))))
 135             return hashlib.md5((t + mg + x).encode('utf8')).hexdigest()
 136
 137         video_urls_dict = {}
 138         for format_item in data['vp']['tkl'][0]['vs']:
 139             if 0 < int(format_item['bid']) <= 10:
 140                 format_id = self.get_format(format_item['bid'])
 141             else:
 142                 continue
 143
 144             video_urls = []
 145
 146             video_urls_info = format_item['fs']
 147             if not format_item['fs'][0]['l'].startswith('/'):
 148                 t = get_encode_code(format_item['fs'][0]['l'])
 149                 if t.endswith('mp4'):
 150                     video_urls_info = format_item['flvs']
 151
 152             for segment_index, segment in enumerate(video_urls_info):
 153                 vl = segment['l']
 154                 if not vl.startswith('/'):
 155                     vl = get_encode_code(vl)
 156                 key = get_path_key(
 157                     vl.split('/')[-1].split('.')[0], format_id, segment_index)
 158                 filesize = segment['b']
 159                 base_url = data['vp']['du'].split('/')
 160                 base_url.insert(-1, key)
 161                 base_url = '/'.join(base_url)
 162                 param = {
 163                     'su': _uuid,
 164                     'qyid': uuid.uuid4().hex,
 165                     'client': '',
 166                     'z': '',
 167                     'bt': '',
 168                     'ct': '',
 169                     'tn': str(int(time.time()))
 170                 }
 171                 api_video_url = base_url + vl + '?' + \
 172                     compat_urllib_parse.urlencode(param)
 173                 js = self._download_json(
 174                     api_video_url, video_id,
 175                     note='Download video info of segment %d for format %s' % (segment_index + 1, format_id))
 176                 video_url = js['l']
 177                 video_urls.append(
 178                     (video_url, filesize))
 179
 180             video_urls_dict[format_id] = video_urls
 181         return video_urls_dict
 182
 183     def get_format(self, bid):
 184         matched_format_ids = [_format_id for _bid, _format_id in self._FORMATS_MAP if _bid == str(bid)]
 185         return matched_format_ids[0] if len(matched_format_ids) else None
 186
 187     def get_bid(self, format_id):
 188         matched_bids = [_bid for _bid, _format_id in self._FORMATS_MAP if _format_id == format_id]
 189         return matched_bids[0] if len(matched_bids) else None
 190
 191     def get_raw_data(self, tvid, video_id, enc_key, _uuid):
 192         tm = str(int(time.time()))
 193         param = {
 194             'key': 'fvip',
 195             'src': hashlib.md5(b'youtube-dl').hexdigest(),
 196             'tvId': tvid,
 197             'vid': video_id,
 198             'vinfo': 1,
 199             'tm': tm,
 200             'enc': hashlib.md5(
 201                 (enc_key + tm + tvid).encode('utf8')).hexdigest(),
 202             'qyid': _uuid,
 203             'tn': random.random(),
 204             'um': 0,
 205             'authkey': hashlib.md5(
 206                 (tm + tvid).encode('utf8')).hexdigest()
 207         }
 208
 209         api_url = 'http://cache.video.qiyi.com/vms' + '?' + \
 210             compat_urllib_parse.urlencode(param)
 211         raw_data = self._download_json(api_url, video_id)
 212         return raw_data
 213
 214     def get_enc_key(self, swf_url, video_id):
 215         enc_key = '8e29ab5666d041c3a1ea76e06dabdffb'
 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