yt_dlp/extractor/iqiyi.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import hashlib
   5 import itertools
   6 import re
   7 import time
   8
   9 from .common import InfoExtractor
  10 from ..compat import (
  11     compat_str,
  12     compat_urllib_parse_urlencode,
  13 )
  14 from .openload import PhantomJSwrapper
  15 from ..utils import (
  16     clean_html,
  17     decode_packed_codes,
  18     ExtractorError,
  19     float_or_none,
  20     get_element_by_id,
  21     get_element_by_attribute,
  22     int_or_none,
  23     js_to_json,
  24     ohdave_rsa_encrypt,
  25     parse_age_limit,
  26     parse_duration,
  27     parse_iso8601,
  28     parse_resolution,
  29     qualities,
  30     remove_start,
  31     str_or_none,
  32     traverse_obj,
  33     urljoin,
  34 )
  35
  36
  37 def md5_text(text):
  38     return hashlib.md5(text.encode('utf-8')).hexdigest()
  39
  40
  41 class IqiyiSDK(object):
  42     def __init__(self, target, ip, timestamp):
  43         self.target = target
  44         self.ip = ip
  45         self.timestamp = timestamp
  46
  47     @staticmethod
  48     def split_sum(data):
  49         return compat_str(sum(map(lambda p: int(p, 16), list(data))))
  50
  51     @staticmethod
  52     def digit_sum(num):
  53         if isinstance(num, int):
  54             num = compat_str(num)
  55         return compat_str(sum(map(int, num)))
  56
  57     def even_odd(self):
  58         even = self.digit_sum(compat_str(self.timestamp)[::2])
  59         odd = self.digit_sum(compat_str(self.timestamp)[1::2])
  60         return even, odd
  61
  62     def preprocess(self, chunksize):
  63         self.target = md5_text(self.target)
  64         chunks = []
  65         for i in range(32 // chunksize):
  66             chunks.append(self.target[chunksize * i:chunksize * (i + 1)])
  67         if 32 % chunksize:
  68             chunks.append(self.target[32 - 32 % chunksize:])
  69         return chunks, list(map(int, self.ip.split('.')))
  70
  71     def mod(self, modulus):
  72         chunks, ip = self.preprocess(32)
  73         self.target = chunks[0] + ''.join(map(lambda p: compat_str(p % modulus), ip))
  74
  75     def split(self, chunksize):
  76         modulus_map = {
  77             4: 256,
  78             5: 10,
  79             8: 100,
  80         }
  81
  82         chunks, ip = self.preprocess(chunksize)
  83         ret = ''
  84         for i in range(len(chunks)):
  85             ip_part = compat_str(ip[i] % modulus_map[chunksize]) if i < 4 else ''
  86             if chunksize == 8:
  87                 ret += ip_part + chunks[i]
  88             else:
  89                 ret += chunks[i] + ip_part
  90         self.target = ret
  91
  92     def handle_input16(self):
  93         self.target = md5_text(self.target)
  94         self.target = self.split_sum(self.target[:16]) + self.target + self.split_sum(self.target[16:])
  95
  96     def handle_input8(self):
  97         self.target = md5_text(self.target)
  98         ret = ''
  99         for i in range(4):
 100             part = self.target[8 * i:8 * (i + 1)]
 101             ret += self.split_sum(part) + part
 102         self.target = ret
 103
 104     def handleSum(self):
 105         self.target = md5_text(self.target)
 106         self.target = self.split_sum(self.target) + self.target
 107
 108     def date(self, scheme):
 109         self.target = md5_text(self.target)
 110         d = time.localtime(self.timestamp)
 111         strings = {
 112             'y': compat_str(d.tm_year),
 113             'm': '%02d' % d.tm_mon,
 114             'd': '%02d' % d.tm_mday,
 115         }
 116         self.target += ''.join(map(lambda c: strings[c], list(scheme)))
 117
 118     def split_time_even_odd(self):
 119         even, odd = self.even_odd()
 120         self.target = odd + md5_text(self.target) + even
 121
 122     def split_time_odd_even(self):
 123         even, odd = self.even_odd()
 124         self.target = even + md5_text(self.target) + odd
 125
 126     def split_ip_time_sum(self):
 127         chunks, ip = self.preprocess(32)
 128         self.target = compat_str(sum(ip)) + chunks[0] + self.digit_sum(self.timestamp)
 129
 130     def split_time_ip_sum(self):
 131         chunks, ip = self.preprocess(32)
 132         self.target = self.digit_sum(self.timestamp) + chunks[0] + compat_str(sum(ip))
 133
 134
 135 class IqiyiSDKInterpreter(object):
 136     def __init__(self, sdk_code):
 137         self.sdk_code = sdk_code
 138
 139     def run(self, target, ip, timestamp):
 140         self.sdk_code = decode_packed_codes(self.sdk_code)
 141
 142         functions = re.findall(r'input=([a-zA-Z0-9]+)\(input', self.sdk_code)
 143
 144         sdk = IqiyiSDK(target, ip, timestamp)
 145
 146         other_functions = {
 147             'handleSum': sdk.handleSum,
 148             'handleInput8': sdk.handle_input8,
 149             'handleInput16': sdk.handle_input16,
 150             'splitTimeEvenOdd': sdk.split_time_even_odd,
 151             'splitTimeOddEven': sdk.split_time_odd_even,
 152             'splitIpTimeSum': sdk.split_ip_time_sum,
 153             'splitTimeIpSum': sdk.split_time_ip_sum,
 154         }
 155         for function in functions:
 156             if re.match(r'mod\d+', function):
 157                 sdk.mod(int(function[3:]))
 158             elif re.match(r'date[ymd]{3}', function):
 159                 sdk.date(function[4:])
 160             elif re.match(r'split\d+', function):
 161                 sdk.split(int(function[5:]))
 162             elif function in other_functions:
 163                 other_functions[function]()
 164             else:
 165                 raise ExtractorError('Unknown function %s' % function)
 166
 167         return sdk.target
 168
 169
 170 class IqiyiIE(InfoExtractor):
 171     IE_NAME = 'iqiyi'
 172     IE_DESC = '爱奇艺'
 173
 174     _VALID_URL = r'https?://(?:(?:[^.]+\.)?iqiyi\.com|www\.pps\.tv)/.+\.html'
 175
 176     _NETRC_MACHINE = 'iqiyi'
 177
 178     _TESTS = [{
 179         'url': 'http://www.iqiyi.com/v_19rrojlavg.html',
 180         # MD5 checksum differs on my machine and Travis CI
 181         'info_dict': {
 182             'id': '9c1fb1b99d192b21c559e5a1a2cb3c73',
 183             'ext': 'mp4',
 184             'title': '美国德州空中惊现奇异云团 酷似UFO',
 185         }
 186     }, {
 187         'url': 'http://www.iqiyi.com/v_19rrhnnclk.html',
 188         'md5': 'b7dc800a4004b1b57749d9abae0472da',
 189         'info_dict': {
 190             'id': 'e3f585b550a280af23c98b6cb2be19fb',
 191             'ext': 'mp4',
 192             # This can be either Simplified Chinese or Traditional Chinese
 193             'title': r're:^(?:名侦探柯南 国语版：第752集 迫近灰原秘密的黑影 下篇|名偵探柯南 國語版：第752集 迫近灰原秘密的黑影 下篇)$',
 194         },
 195         'skip': 'Geo-restricted to China',
 196     }, {
 197         'url': 'http://www.iqiyi.com/w_19rt6o8t9p.html',
 198         'only_matching': True,
 199     }, {
 200         'url': 'http://www.iqiyi.com/a_19rrhbc6kt.html',
 201         'only_matching': True,
 202     }, {
 203         'url': 'http://yule.iqiyi.com/pcb.html',
 204         'info_dict': {
 205             'id': '4a0af228fddb55ec96398a364248ed7f',
 206             'ext': 'mp4',
 207             'title': '第2017-04-21期 女艺人频遭极端粉丝骚扰',
 208         },
 209     }, {
 210         # VIP-only video. The first 2 parts (6 minutes) are available without login
 211         # MD5 sums omitted as values are different on Travis CI and my machine
 212         'url': 'http://www.iqiyi.com/v_19rrny4w8w.html',
 213         'info_dict': {
 214             'id': 'f3cf468b39dddb30d676f89a91200dc1',
 215             'ext': 'mp4',
 216             'title': '泰坦尼克号',
 217         },
 218         'skip': 'Geo-restricted to China',
 219     }, {
 220         'url': 'http://www.iqiyi.com/a_19rrhb8ce1.html',
 221         'info_dict': {
 222             'id': '202918101',
 223             'title': '灌篮高手 国语版',
 224         },
 225         'playlist_count': 101,
 226     }, {
 227         'url': 'http://www.pps.tv/w_19rrbav0ph.html',
 228         'only_matching': True,
 229     }]
 230
 231     _FORMATS_MAP = {
 232         '96': 1,    # 216p, 240p
 233         '1': 2,     # 336p, 360p
 234         '2': 3,     # 480p, 504p
 235         '21': 4,    # 504p
 236         '4': 5,     # 720p
 237         '17': 5,    # 720p
 238         '5': 6,     # 1072p, 1080p
 239         '18': 7,    # 1080p
 240     }
 241
 242     def _real_initialize(self):
 243         self._login()
 244
 245     @staticmethod
 246     def _rsa_fun(data):
 247         # public key extracted from http://static.iqiyi.com/js/qiyiV2/20160129180840/jobs/i18n/i18nIndex.js
 248         N = 0xab86b6371b5318aaa1d3c9e612a9f1264f372323c8c0f19875b5fc3b3fd3afcc1e5bec527aa94bfa85bffc157e4245aebda05389a5357b75115ac94f074aefcd
 249         e = 65537
 250
 251         return ohdave_rsa_encrypt(data, e, N)
 252
 253     def _login(self):
 254         username, password = self._get_login_info()
 255
 256         # No authentication to be performed
 257         if not username:
 258             return True
 259
 260         data = self._download_json(
 261             'http://kylin.iqiyi.com/get_token', None,
 262             note='Get token for logging', errnote='Unable to get token for logging')
 263         sdk = data['sdk']
 264         timestamp = int(time.time())
 265         target = '/apis/reglogin/login.action?lang=zh_TW&area_code=null&email=%s&passwd=%s&agenttype=1&from=undefined&keeplogin=0&piccode=&fromurl=&_pos=1' % (
 266             username, self._rsa_fun(password.encode('utf-8')))
 267
 268         interp = IqiyiSDKInterpreter(sdk)
 269         sign = interp.run(target, data['ip'], timestamp)
 270
 271         validation_params = {
 272             'target': target,
 273             'server': 'BEA3AA1908656AABCCFF76582C4C6660',
 274             'token': data['token'],
 275             'bird_src': 'f8d91d57af224da7893dd397d52d811a',
 276             'sign': sign,
 277             'bird_t': timestamp,
 278         }
 279         validation_result = self._download_json(
 280             'http://kylin.iqiyi.com/validate?' + compat_urllib_parse_urlencode(validation_params), None,
 281             note='Validate credentials', errnote='Unable to validate credentials')
 282
 283         MSG_MAP = {
 284             'P00107': 'please login via the web interface and enter the CAPTCHA code',
 285             'P00117': 'bad username or password',
 286         }
 287
 288         code = validation_result['code']
 289         if code != 'A00000':
 290             msg = MSG_MAP.get(code)
 291             if not msg:
 292                 msg = 'error %s' % code
 293                 if validation_result.get('msg'):
 294                     msg += ': ' + validation_result['msg']
 295             self.report_warning('unable to log in: ' + msg)
 296             return False
 297
 298         return True
 299
 300     def get_raw_data(self, tvid, video_id):
 301         tm = int(time.time() * 1000)
 302
 303         key = 'd5fb4bd9d50c4be6948c97edd7254b0e'
 304         sc = md5_text(compat_str(tm) + key + tvid)
 305         params = {
 306             'tvid': tvid,
 307             'vid': video_id,
 308             'src': '76f90cbd92f94a2e925d83e8ccd22cb7',
 309             'sc': sc,
 310             't': tm,
 311         }
 312
 313         return self._download_json(
 314             'http://cache.m.iqiyi.com/jp/tmts/%s/%s/' % (tvid, video_id),
 315             video_id, transform_source=lambda s: remove_start(s, 'var tvInfoJs='),
 316             query=params, headers=self.geo_verification_headers())
 317
 318     def _extract_playlist(self, webpage):
 319         PAGE_SIZE = 50
 320
 321         links = re.findall(
 322             r'<a[^>]+class="site-piclist_pic_link"[^>]+href="(http://www\.iqiyi\.com/.+\.html)"',
 323             webpage)
 324         if not links:
 325             return
 326
 327         album_id = self._search_regex(
 328             r'albumId\s*:\s*(\d+),', webpage, 'album ID')
 329         album_title = self._search_regex(
 330             r'data-share-title="([^"]+)"', webpage, 'album title', fatal=False)
 331
 332         entries = list(map(self.url_result, links))
 333
 334         # Start from 2 because links in the first page are already on webpage
 335         for page_num in itertools.count(2):
 336             pagelist_page = self._download_webpage(
 337                 'http://cache.video.qiyi.com/jp/avlist/%s/%d/%d/' % (album_id, page_num, PAGE_SIZE),
 338                 album_id,
 339                 note='Download playlist page %d' % page_num,
 340                 errnote='Failed to download playlist page %d' % page_num)
 341             pagelist = self._parse_json(
 342                 remove_start(pagelist_page, 'var tvInfoJs='), album_id)
 343             vlist = pagelist['data']['vlist']
 344             for item in vlist:
 345                 entries.append(self.url_result(item['vurl']))
 346             if len(vlist) < PAGE_SIZE:
 347                 break
 348
 349         return self.playlist_result(entries, album_id, album_title)
 350
 351     def _real_extract(self, url):
 352         webpage = self._download_webpage(
 353             url, 'temp_id', note='download video page')
 354
 355         # There's no simple way to determine whether an URL is a playlist or not
 356         # Sometimes there are playlist links in individual videos, so treat it
 357         # as a single video first
 358         tvid = self._search_regex(
 359             r'data-(?:player|shareplattrigger)-tvid\s*=\s*[\'"](\d+)', webpage, 'tvid', default=None)
 360         if tvid is None:
 361             playlist_result = self._extract_playlist(webpage)
 362             if playlist_result:
 363                 return playlist_result
 364             raise ExtractorError('Can\'t find any video')
 365
 366         video_id = self._search_regex(
 367             r'data-(?:player|shareplattrigger)-videoid\s*=\s*[\'"]([a-f\d]+)', webpage, 'video_id')
 368
 369         formats = []
 370         for _ in range(5):
 371             raw_data = self.get_raw_data(tvid, video_id)
 372
 373             if raw_data['code'] != 'A00000':
 374                 if raw_data['code'] == 'A00111':
 375                     self.raise_geo_restricted()
 376                 raise ExtractorError('Unable to load data. Error code: ' + raw_data['code'])
 377
 378             data = raw_data['data']
 379
 380             for stream in data['vidl']:
 381                 if 'm3utx' not in stream:
 382                     continue
 383                 vd = compat_str(stream['vd'])
 384                 formats.append({
 385                     'url': stream['m3utx'],
 386                     'format_id': vd,
 387                     'ext': 'mp4',
 388                     'quality': self._FORMATS_MAP.get(vd, -1),
 389                     'protocol': 'm3u8_native',
 390                 })
 391
 392             if formats:
 393                 break
 394
 395             self._sleep(5, video_id)
 396
 397         self._sort_formats(formats)
 398         title = (get_element_by_id('widget-videotitle', webpage)
 399                  or clean_html(get_element_by_attribute('class', 'mod-play-tit', webpage))
 400                  or self._html_search_regex(r'<span[^>]+data-videochanged-title="word"[^>]*>([^<]+)</span>', webpage, 'title'))
 401
 402         return {
 403             'id': video_id,
 404             'title': title,
 405             'formats': formats,
 406         }
 407
 408
 409 class IqIE(InfoExtractor):
 410     IE_NAME = 'iq.com'
 411     IE_DESC = 'International version of iQiyi'
 412     _VALID_URL = r'https?://(?:www\.)?iq\.com/play/(?:[\w%-]*-)?(?P<id>\w+)'
 413     _TESTS = [{
 414         'url': 'https://www.iq.com/play/one-piece-episode-1000-1ma1i6ferf4',
 415         'md5': '2d7caf6eeca8a32b407094b33b757d39',
 416         'info_dict': {
 417             'ext': 'mp4',
 418             'id': '1ma1i6ferf4',
 419             'title': '航海王 第1000集',
 420             'description': 'Subtitle available on Sunday 4PM（GMT+8）.',
 421             'duration': 1430,
 422             'timestamp': 1637488203,
 423             'upload_date': '20211121',
 424             'episode_number': 1000,
 425             'episode': 'Episode 1000',
 426             'series': 'One Piece',
 427             'age_limit': 13,
 428             'average_rating': float,
 429         },
 430         'params': {
 431             'format': '500',
 432         },
 433         'expected_warnings': ['format is restricted']
 434     }]
 435     _BID_TAGS = {
 436         '100': '240P',
 437         '200': '360P',
 438         '300': '480P',
 439         '500': '720P',
 440         '600': '1080P',
 441         '610': '1080P50',
 442         '700': '2K',
 443         '800': '4K',
 444     }
 445     _LID_TAGS = {
 446         '1': 'zh_CN',
 447         '2': 'zh_TW',
 448         '3': 'en',
 449         '18': 'th',
 450         '21': 'my',
 451         '23': 'vi',
 452         '24': 'id',
 453         '26': 'es',
 454         '28': 'ar',
 455     }
 456
 457     _DASH_JS = '''
 458         console.log(page.evaluate(function() {
 459             var tvid = "%(tvid)s"; var vid = "%(vid)s"; var src = "%(src)s";
 460             var dfp = "%(dfp)s"; var mode = "%(mode)s"; var lang = "%(lang)s"; var bid_list = %(bid_list)s;
 461             var tm = new Date().getTime();
 462             var cmd5x_func = %(cmd5x_func)s; var cmd5x_exporter = {}; cmd5x_func({}, cmd5x_exporter, {}); var cmd5x = cmd5x_exporter.cmd5x;
 463             var authKey = cmd5x(cmd5x('') + tm + '' + tvid);
 464             var k_uid = Array.apply(null, Array(32)).map(function() {return Math.floor(Math.random() * 15).toString(16)}).join('');
 465             var dash_paths = {};
 466             bid_list.forEach(function(bid) {
 467                 var query = {
 468                     'tvid': tvid,
 469                     'bid': bid,
 470                     'ds': 1,
 471                     'vid': vid,
 472                     'src': src,
 473                     'vt': 0,
 474                     'rs': 1,
 475                     'uid': 0,
 476                     'ori': 'pcw',
 477                     'ps': 1,
 478                     'k_uid': k_uid,
 479                     'pt': 0,
 480                     'd': 0,
 481                     's': '',
 482                     'lid': '',
 483                     'slid': 0,
 484                     'cf': '',
 485                     'ct': '',
 486                     'authKey': authKey,
 487                     'k_tag': 1,
 488                     'ost': 0,
 489                     'ppt': 0,
 490                     'dfp': dfp,
 491                     'prio': JSON.stringify({
 492                         'ff': 'f4v',
 493                         'code': 2
 494                     }),
 495                     'k_err_retries': 0,
 496                     'up': '',
 497                     'su': 2,
 498                     'applang': lang,
 499                     'sver': 2,
 500                     'X-USER-MODE': mode,
 501                     'qd_v': 2,
 502                     'tm': tm,
 503                     'qdy': 'a',
 504                     'qds': 0,
 505                     'k_ft1': 141287244169348,
 506                     'k_ft4': 34359746564,
 507                     'k_ft5': 1,
 508                     'bop': JSON.stringify({
 509                         'version': '10.0',
 510                         'dfp': dfp
 511                     }),
 512                     'ut': 0, // TODO: Set ut param for VIP members
 513                 };
 514                 var enc_params = [];
 515                 for (var prop in query) {
 516                     enc_params.push(encodeURIComponent(prop) + '=' + encodeURIComponent(query[prop]));
 517                 }
 518                 var dash_path = '/dash?' + enc_params.join('&'); dash_path += '&vf=' + cmd5x(dash_path);
 519                 dash_paths[bid] = dash_path;
 520             });
 521             return JSON.stringify(dash_paths);
 522         }));
 523         saveAndExit();
 524     '''
 525
 526     def _extract_vms_player_js(self, webpage, video_id):
 527         player_js_cache = self._downloader.cache.load('iq', 'player_js')
 528         if player_js_cache:
 529             return player_js_cache
 530         webpack_js_url = self._proto_relative_url(self._search_regex(
 531             r'<script src="((?:https?)?//stc.iqiyipic.com/_next/static/chunks/webpack-\w+\.js)"', webpage, 'webpack URL'))
 532         webpack_js = self._download_webpage(webpack_js_url, video_id, note='Downloading webpack JS', errnote='Unable to download webpack JS')
 533         webpack_map1, webpack_map2 = [self._parse_json(js_map, video_id, transform_source=js_to_json) for js_map in self._search_regex(
 534             r'\(({[^}]*})\[\w+\][^\)]*\)\s*\+\s*["\']\.["\']\s*\+\s*({[^}]*})\[\w+\]\+["\']\.js', webpack_js, 'JS locations', group=(1, 2))]
 535         for module_index in reversed(webpack_map2.keys()):
 536             module_js = self._download_webpage(
 537                 f'https://stc.iqiyipic.com/_next/static/chunks/{webpack_map1.get(module_index, module_index)}.{webpack_map2[module_index]}.js',
 538                 video_id, note=f'Downloading #{module_index} module JS', errnote='Unable to download module JS', fatal=False) or ''
 539             if 'vms request' in module_js:
 540                 self._downloader.cache.store('iq', 'player_js', module_js)
 541                 return module_js
 542         raise ExtractorError('Unable to extract player JS')
 543
 544     def _extract_cmd5x_function(self, webpage, video_id):
 545         return self._search_regex(r',\s*(function\s*\([^\)]*\)\s*{\s*var _qda.+_qdc\(\)\s*})\s*,',
 546                                   self._extract_vms_player_js(webpage, video_id), 'signature function')
 547
 548     def _update_bid_tags(self, webpage, video_id):
 549         extracted_bid_tags = self._parse_json(
 550             self._search_regex(
 551                 r'arguments\[1\][^,]*,\s*function\s*\([^\)]*\)\s*{\s*"use strict";?\s*var \w=({.+}})\s*,\s*\w\s*=\s*{\s*getNewVd',
 552                 self._extract_vms_player_js(webpage, video_id), 'video tags', default=''),
 553             video_id, transform_source=js_to_json, fatal=False)
 554         if not extracted_bid_tags:
 555             return
 556         self._BID_TAGS = {
 557             bid: traverse_obj(extracted_bid_tags, (bid, 'value'), expected_type=str, default=self._BID_TAGS.get(bid))
 558             for bid in extracted_bid_tags.keys()
 559         }
 560
 561     def _get_cookie(self, name, default=None):
 562         cookie = self._get_cookies('https://iq.com/').get(name)
 563         return cookie.value if cookie else default
 564
 565     def _real_extract(self, url):
 566         video_id = self._match_id(url)
 567         webpage = self._download_webpage(url, video_id)
 568         self._update_bid_tags(webpage, video_id)
 569
 570         next_props = self._search_nextjs_data(webpage, video_id)['props']
 571         page_data = next_props['initialState']['play']
 572         video_info = page_data['curVideoInfo']
 573
 574         # bid 0 as an initial format checker
 575         dash_paths = self._parse_json(PhantomJSwrapper(self).get(
 576             url, html='<!DOCTYPE html>', video_id=video_id, note2='Executing signature code', jscode=self._DASH_JS % {
 577                 'tvid': video_info['tvId'],
 578                 'vid': video_info['vid'],
 579                 'src': traverse_obj(next_props, ('initialProps', 'pageProps', 'ptid'),
 580                                     expected_type=str, default='01010031010018000000'),
 581                 'dfp': self._get_cookie('dfp', ''),
 582                 'mode': self._get_cookie('mod', 'intl'),
 583                 'lang': self._get_cookie('lang', 'en_us'),
 584                 'bid_list': '[' + ','.join(['0', *self._BID_TAGS.keys()]) + ']',
 585                 'cmd5x_func': self._extract_cmd5x_function(webpage, video_id),
 586             })[1].strip(), video_id)
 587
 588         formats, subtitles = [], {}
 589         initial_format_data = self._download_json(
 590             urljoin('https://cache-video.iq.com', dash_paths['0']), video_id,
 591             note='Downloading initial video format info', errnote='Unable to download initial video format info')['data']
 592
 593         preview_time = traverse_obj(initial_format_data, ('boss_ts', 'data', 'previewTime'), expected_type=float_or_none)
 594         if preview_time:
 595             self.report_warning(f'This preview video is limited to {preview_time} seconds')
 596
 597         # TODO: Extract audio-only formats
 598         for bid in set(traverse_obj(initial_format_data, ('program', 'video', ..., 'bid'), expected_type=str_or_none, default=[])):
 599             dash_path = dash_paths.get(bid)
 600             if not dash_path:
 601                 self.report_warning(f'Unknown format id: {bid}. It is currently not being extracted')
 602                 continue
 603             format_data = traverse_obj(self._download_json(
 604                 urljoin('https://cache-video.iq.com', dash_path), video_id,
 605                 note=f'Downloading format data for {self._BID_TAGS[bid]}', errnote='Unable to download format data',
 606                 fatal=False), 'data', expected_type=dict)
 607
 608             video_format = next((video_format for video_format in traverse_obj(
 609                 format_data, ('program', 'video', ...), expected_type=dict, default=[]) if str(video_format['bid']) == bid), {})
 610             extracted_formats = []
 611             if video_format.get('m3u8Url'):
 612                 extracted_formats.extend(self._extract_m3u8_formats(
 613                     urljoin(format_data.get('dm3u8', 'https://cache-m.iq.com/dc/dt/'), video_format['m3u8Url']),
 614                     'mp4', m3u8_id=bid, fatal=False))
 615             if video_format.get('mpdUrl'):
 616                 # TODO: Properly extract mpd hostname
 617                 extracted_formats.extend(self._extract_mpd_formats(
 618                     urljoin(format_data.get('dm3u8', 'https://cache-m.iq.com/dc/dt/'), video_format['mpdUrl']),
 619                     mpd_id=bid, fatal=False))
 620             if video_format.get('m3u8'):
 621                 ff = video_format.get('ff', 'ts')
 622                 if ff == 'ts':
 623                     m3u8_formats, _ = self._parse_m3u8_formats_and_subtitles(
 624                         video_format['m3u8'], ext='mp4', m3u8_id=bid, fatal=False)
 625                     extracted_formats.extend(m3u8_formats)
 626                 elif ff == 'm4s':
 627                     mpd_data = traverse_obj(
 628                         self._parse_json(video_format['m3u8'], video_id, fatal=False), ('payload', ..., 'data'), expected_type=str)
 629                     if not mpd_data:
 630                         continue
 631                     mpd_formats, _ = self._parse_mpd_formats_and_subtitles(
 632                         mpd_data, bid, format_data.get('dm3u8', 'https://cache-m.iq.com/dc/dt/'))
 633                     extracted_formats.extend(mpd_formats)
 634                 else:
 635                     self.report_warning(f'{ff} formats are currently not supported')
 636
 637             if not extracted_formats:
 638                 if video_format.get('s'):
 639                     self.report_warning(f'{self._BID_TAGS[bid]} format is restricted')
 640                 else:
 641                     self.report_warning(f'Unable to extract {self._BID_TAGS[bid]} format')
 642             for f in extracted_formats:
 643                 f.update({
 644                     'quality': qualities(list(self._BID_TAGS.keys()))(bid),
 645                     'format_note': self._BID_TAGS[bid],
 646                     **parse_resolution(video_format.get('scrsz'))
 647                 })
 648             formats.extend(extracted_formats)
 649
 650         self._sort_formats(formats)
 651
 652         for sub_format in traverse_obj(initial_format_data, ('program', 'stl', ...), expected_type=dict, default=[]):
 653             lang = self._LID_TAGS.get(str_or_none(sub_format.get('lid')), sub_format.get('_name'))
 654             subtitles.setdefault(lang, []).extend([{
 655                 'ext': format_ext,
 656                 'url': urljoin(initial_format_data.get('dstl', 'http://meta.video.iqiyi.com'), sub_format[format_key])
 657             } for format_key, format_ext in [('srt', 'srt'), ('webvtt', 'vtt')] if sub_format.get(format_key)])
 658
 659         extra_metadata = page_data.get('albumInfo') if video_info.get('albumId') and page_data.get('albumInfo') else video_info
 660         return {
 661             'id': video_id,
 662             'title': video_info['name'],
 663             'formats': formats,
 664             'subtitles': subtitles,
 665             'description': video_info.get('mergeDesc'),
 666             'duration': parse_duration(video_info.get('len')),
 667             'age_limit': parse_age_limit(video_info.get('rating')),
 668             'average_rating': traverse_obj(page_data, ('playScoreInfo', 'score'), expected_type=float_or_none),
 669             'timestamp': parse_iso8601(video_info.get('isoUploadDate')),
 670             'categories': traverse_obj(extra_metadata, ('videoTagMap', ..., ..., 'name'), expected_type=str),
 671             'cast': traverse_obj(extra_metadata, ('actorArr', ..., 'name'), expected_type=str),
 672             'episode_number': int_or_none(video_info.get('order')) or None,
 673             'series': video_info.get('albumName'),
 674         }
 675
 676
 677 class IqAlbumIE(InfoExtractor):
 678     IE_NAME = 'iq.com:album'
 679     _VALID_URL = r'https?://(?:www\.)?iq\.com/album/(?:[\w%-]*-)?(?P<id>\w+)'
 680     _TESTS = [{
 681         'url': 'https://www.iq.com/album/one-piece-1999-1bk9icvr331',
 682         'info_dict': {
 683             'id': '1bk9icvr331',
 684             'title': 'One Piece',
 685             'description': 'Subtitle available on Sunday 4PM（GMT+8）.'
 686         },
 687         'playlist_mincount': 238
 688     }, {
 689         # Movie/single video
 690         'url': 'https://www.iq.com/album/九龙城寨-2021-22yjnij099k',
 691         'info_dict': {
 692             'ext': 'mp4',
 693             'id': '22yjnij099k',
 694             'title': '九龙城寨',
 695             'description': 'md5:8a09f50b8ba0db4dc69bc7c844228044',
 696             'duration': 5000,
 697             'timestamp': 1641911371,
 698             'upload_date': '20220111',
 699             'series': '九龙城寨',
 700             'cast': ['Shi Yan Neng', 'Yu Lang', 'Peter  lv', 'Sun Zi Jun', 'Yang Xiao Bo'],
 701             'age_limit': 13,
 702             'average_rating': float,
 703         },
 704         'expected_warnings': ['format is restricted']
 705     }]
 706
 707     def _entries(self, album_id_num, page_ranges, album_id=None, mode_code='intl', lang_code='en_us'):
 708         for page_range in page_ranges:
 709             page = self._download_json(
 710                 f'https://pcw-api.iq.com/api/episodeListSource/{album_id_num}', album_id,
 711                 note=f'Downloading video list episodes {page_range.get("msg", "")}',
 712                 errnote='Unable to download video list', query={
 713                     'platformId': 3,
 714                     'modeCode': mode_code,
 715                     'langCode': lang_code,
 716                     'endOrder': page_range['to'],
 717                     'startOrder': page_range['from']
 718                 })
 719             for video in page['data']['epg']:
 720                 yield self.url_result('https://www.iq.com/play/%s' % (video.get('playLocSuffix') or video['qipuIdStr']),
 721                                       IqIE.ie_key(), video.get('qipuIdStr'), video.get('name'))
 722
 723     def _real_extract(self, url):
 724         album_id = self._match_id(url)
 725         webpage = self._download_webpage(url, album_id)
 726         next_data = self._search_nextjs_data(webpage, album_id)
 727         album_data = next_data['props']['initialState']['album']['videoAlbumInfo']
 728
 729         if album_data.get('videoType') == 'singleVideo':
 730             return self.url_result('https://www.iq.com/play/%s' % album_id, IqIE.ie_key())
 731         return self.playlist_result(
 732             self._entries(album_data['albumId'], album_data['totalPageRange'], album_id,
 733                           traverse_obj(next_data, ('props', 'initialProps', 'pageProps', 'modeCode')),
 734                           traverse_obj(next_data, ('props', 'initialProps', 'pageProps', 'langCode'))),
 735             album_id, album_data.get('name'), album_data.get('desc'))