yt_dlp/extractor/huya.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import hashlib
   5 import random
   6
   7 from ..compat import compat_urlparse, compat_b64decode
   8
   9 from ..utils import (
  10     ExtractorError,
  11     int_or_none,
  12     js_to_json,
  13     str_or_none,
  14     try_get,
  15     unescapeHTML,
  16     update_url_query,
  17 )
  18
  19 from .common import InfoExtractor
  20
  21
  22 class HuyaLiveIE(InfoExtractor):
  23     _VALID_URL = r'https?://(?:www\.|m\.)?huya\.com/(?P<id>[^/#?&]+)(?:\D|$)'
  24     IE_NAME = 'huya:live'
  25     IE_DESC = 'huya.com'
  26     TESTS = [{
  27         'url': 'https://www.huya.com/572329',
  28         'info_dict': {
  29             'id': '572329',
  30             'title': str,
  31             'description': str,
  32             'is_live': True,
  33             'view_count': int,
  34         },
  35         'params': {
  36             'skip_download': True,
  37         },
  38     }, {
  39         'url': 'https://www.huya.com/xiaoyugame',
  40         'only_matching': True
  41     }]
  42
  43     _RESOLUTION = {
  44         '蓝光4M': {
  45             'width': 1920,
  46             'height': 1080,
  47         },
  48         '超清': {
  49             'width': 1280,
  50             'height': 720,
  51         },
  52         '流畅': {
  53             'width': 800,
  54             'height': 480
  55         }
  56     }
  57
  58     def _real_extract(self, url):
  59         video_id = self._match_id(url)
  60         webpage = self._download_webpage(url, video_id=video_id)
  61         json_stream = self._search_regex(r'"stream":\s+"([a-zA-Z0-9+=/]+)"', webpage, 'stream', default=None)
  62         if not json_stream:
  63             raise ExtractorError('Video is offline', expected=True)
  64         stream_data = self._parse_json(compat_b64decode(json_stream).decode(), video_id=video_id,
  65                                        transform_source=js_to_json)
  66         room_info = try_get(stream_data, lambda x: x['data'][0]['gameLiveInfo'])
  67         if not room_info:
  68             raise ExtractorError('Can not extract the room info', expected=True)
  69         title = room_info.get('roomName') or room_info.get('introduction') or self._html_extract_title(webpage)
  70         screen_type = room_info.get('screenType')
  71         live_source_type = room_info.get('liveSourceType')
  72         stream_info_list = stream_data['data'][0]['gameStreamInfoList']
  73         formats = []
  74         for stream_info in stream_info_list:
  75             stream_url = stream_info.get('sFlvUrl')
  76             if not stream_url:
  77                 continue
  78             stream_name = stream_info.get('sStreamName')
  79             re_secret = not screen_type and live_source_type in (0, 8, 13)
  80             params = dict(compat_urlparse.parse_qsl(unescapeHTML(stream_info['sFlvAntiCode'])))
  81             fm, ss = '', ''
  82             if re_secret:
  83                 fm, ss = self.encrypt(params, stream_info, stream_name)
  84             for si in stream_data.get('vMultiStreamInfo'):
  85                 rate = si.get('iBitRate')
  86                 if rate:
  87                     params['ratio'] = rate
  88                 else:
  89                     params.pop('ratio', None)
  90                 if re_secret:
  91                     params['wsSecret'] = hashlib.md5(
  92                         '_'.join([fm, params['u'], stream_name, ss, params['wsTime']]))
  93                 formats.append({
  94                     'ext': stream_info.get('sFlvUrlSuffix'),
  95                     'format_id': str_or_none(stream_info.get('iLineIndex')),
  96                     'tbr': rate,
  97                     'url': update_url_query(f'{stream_url}/{stream_name}.{stream_info.get("sFlvUrlSuffix")}',
  98                                             query=params),
  99                     **self._RESOLUTION.get(si.get('sDisplayName'), {}),
 100                 })
 101
 102         self._sort_formats(formats)
 103
 104         return {
 105             'id': video_id,
 106             'title': title,
 107             'formats': formats,
 108             'view_count': room_info.get('totalCount'),
 109             'thumbnail': room_info.get('screenshot'),
 110             'description': room_info.get('contentIntro'),
 111             'http_headers': {
 112                 'Origin': 'https://www.huya.com',
 113                 'Referer': 'https://www.huya.com/',
 114             },
 115         }
 116
 117     def encrypt(self, params, stream_info, stream_name):
 118         ct = int_or_none(params.get('wsTime'), 16) + random.random()
 119         presenter_uid = stream_info['lPresenterUid']
 120         if not stream_name.startswith(str(presenter_uid)):
 121             uid = presenter_uid
 122         else:
 123             uid = int_or_none(ct % 1e7 * 1e6 % 0xffffffff)
 124         u1 = uid & 0xffffffff00000000
 125         u2 = uid & 0xffffffff
 126         u3 = uid & 0xffffff
 127         u = u1 | u2 >> 24 | u3 << 8
 128         params.update({
 129             'u': str_or_none(u),
 130             'seqid': str_or_none(int_or_none(ct * 1000) + uid),
 131             'ver': '1',
 132             'uuid': int_or_none(ct % 1e7 * 1e6 % 0xffffffff),
 133             't': '100',
 134         })
 135         fm = compat_b64decode(params['fm']).decode().split('_', 1)[0]
 136         ss = hashlib.md5('|'.join([params['seqid'], params['ctype'], params['t']]))
 137         return fm, ss