yt_dlp/extractor/huya.py

   1 import base64
   2 import hashlib
   3 import random
   4 import re
   5 import urllib.parse
   6
   7 from .common import InfoExtractor
   8 from ..utils import (
   9     ExtractorError,
  10     int_or_none,
  11     parse_duration,
  12     str_or_none,
  13     try_get,
  14     unescapeHTML,
  15     unified_strdate,
  16     update_url_query,
  17     url_or_none,
  18 )
  19 from ..utils.traversal import traverse_obj
  20
  21
  22 class HuyaLiveIE(InfoExtractor):
  23     _VALID_URL = r'https?://(?:www\.|m\.)?huya\.com/(?!(?:video/play/))(?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             'ext': 'flv',
  32             'description': str,
  33             'is_live': True,
  34             'view_count': int,
  35         },
  36         'params': {
  37             'skip_download': True,
  38         },
  39     }, {
  40         'url': 'https://www.huya.com/xiaoyugame',
  41         'only_matching': True,
  42     }]
  43
  44     _RESOLUTION = {
  45         '蓝光': {
  46             'width': 1920,
  47             'height': 1080,
  48         },
  49         '超清': {
  50             'width': 1280,
  51             'height': 720,
  52         },
  53         '流畅': {
  54             'width': 800,
  55             'height': 480,
  56         },
  57     }
  58
  59     def _real_extract(self, url):
  60         video_id = self._match_id(url)
  61         webpage = self._download_webpage(url, video_id=video_id)
  62         stream_data = self._search_json(r'stream:\s', webpage, 'stream', video_id=video_id, default=None)
  63         room_info = try_get(stream_data, lambda x: x['data'][0]['gameLiveInfo'])
  64         if not room_info:
  65             raise ExtractorError('Can not extract the room info', expected=True)
  66         title = room_info.get('roomName') or room_info.get('introduction') or self._html_extract_title(webpage)
  67         screen_type = room_info.get('screenType')
  68         live_source_type = room_info.get('liveSourceType')
  69         stream_info_list = stream_data['data'][0]['gameStreamInfoList']
  70         if not stream_info_list:
  71             raise ExtractorError('Video is offline', expected=True)
  72         formats = []
  73         for stream_info in stream_info_list:
  74             stream_url = stream_info.get('sFlvUrl')
  75             if not stream_url:
  76                 continue
  77             stream_name = stream_info.get('sStreamName')
  78             re_secret = not screen_type and live_source_type in (0, 8, 13)
  79             params = dict(urllib.parse.parse_qsl(unescapeHTML(stream_info['sFlvAntiCode'])))
  80             fm, ss = '', ''
  81             if re_secret:
  82                 fm, ss = self.encrypt(params, stream_info, stream_name)
  83             for si in stream_data.get('vMultiStreamInfo'):
  84                 display_name, bitrate = re.fullmatch(
  85                     r'(.+?)(?:(\d+)M)?', si.get('sDisplayName')).groups()
  86                 rate = si.get('iBitRate')
  87                 if rate:
  88                     params['ratio'] = rate
  89                 else:
  90                     params.pop('ratio', None)
  91                     if bitrate:
  92                         rate = int(bitrate) * 1000
  93                 if re_secret:
  94                     params['wsSecret'] = hashlib.md5(
  95                         '_'.join([fm, params['u'], stream_name, ss, params['wsTime']]))
  96                 formats.append({
  97                     'ext': stream_info.get('sFlvUrlSuffix'),
  98                     'format_id': str_or_none(stream_info.get('iLineIndex')),
  99                     'tbr': rate,
 100                     'url': update_url_query(f'{stream_url}/{stream_name}.{stream_info.get("sFlvUrlSuffix")}',
 101                                             query=params),
 102                     **self._RESOLUTION.get(display_name, {}),
 103                 })
 104
 105         return {
 106             'id': video_id,
 107             'title': title,
 108             'formats': formats,
 109             'view_count': room_info.get('totalCount'),
 110             'thumbnail': room_info.get('screenshot'),
 111             'description': room_info.get('contentIntro'),
 112             'http_headers': {
 113                 'Origin': 'https://www.huya.com',
 114                 'Referer': 'https://www.huya.com/',
 115             },
 116         }
 117
 118     def encrypt(self, params, stream_info, stream_name):
 119         ct = int_or_none(params.get('wsTime'), 16) + random.random()
 120         presenter_uid = stream_info['lPresenterUid']
 121         if not stream_name.startswith(str(presenter_uid)):
 122             uid = presenter_uid
 123         else:
 124             uid = int_or_none(ct % 1e7 * 1e6 % 0xffffffff)
 125         u1 = uid & 0xffffffff00000000
 126         u2 = uid & 0xffffffff
 127         u3 = uid & 0xffffff
 128         u = u1 | u2 >> 24 | u3 << 8
 129         params.update({
 130             'u': str_or_none(u),
 131             'seqid': str_or_none(int_or_none(ct * 1000) + uid),
 132             'ver': '1',
 133             'uuid': int_or_none(ct % 1e7 * 1e6 % 0xffffffff),
 134             't': '100',
 135         })
 136         fm = base64.b64decode(params['fm']).decode().split('_', 1)[0]
 137         ss = hashlib.md5('|'.join([params['seqid'], params['ctype'], params['t']]))
 138         return fm, ss
 139
 140
 141 class HuyaVideoIE(InfoExtractor):
 142     _VALID_URL = r'https?://(?:www\.)?huya\.com/video/play/(?P<id>\d+)\.html'
 143     IE_NAME = 'huya:video'
 144     IE_DESC = '虎牙视频'
 145
 146     _TESTS = [{
 147         'url': 'https://www.huya.com/video/play/1002412640.html',
 148         'info_dict': {
 149             'id': '1002412640',
 150             'ext': 'mp4',
 151             'title': '8月3日',
 152             'thumbnail': r're:https?://.*\.jpg',
 153             'duration': 14,
 154             'uploader': '虎牙-ATS欧卡车队青木',
 155             'uploader_id': '1564376151',
 156             'upload_date': '20240803',
 157             'view_count': int,
 158             'comment_count': int,
 159             'like_count': int,
 160         },
 161     },
 162         {
 163         'url': 'https://www.huya.com/video/play/556054543.html',
 164         'info_dict': {
 165             'id': '556054543',
 166             'ext': 'mp4',
 167             'title': '我不挑事 也不怕事',
 168             'thumbnail': r're:https?://.*\.jpg',
 169             'duration': 1864,
 170             'uploader': '卡尔',
 171             'uploader_id': '367138632',
 172             'upload_date': '20210811',
 173             'view_count': int,
 174             'comment_count': int,
 175             'like_count': int,
 176         },
 177     }]
 178
 179     def _real_extract(self, url: str):
 180         video_id = self._match_id(url)
 181         video_data = self._download_json(
 182             'https://liveapi.huya.com/moment/getMomentContent', video_id,
 183             query={'videoId': video_id})['data']['moment']['videoInfo']
 184
 185         formats = []
 186         for definition in traverse_obj(video_data, ('definitions', lambda _, v: url_or_none(v['url']))):
 187             formats.append({
 188                 'url': definition['url'],
 189                 **traverse_obj(definition, {
 190                     'format_id': ('defName', {str}),
 191                     'width': ('width', {int_or_none}),
 192                     'height': ('height', {int_or_none}),
 193                     'filesize': ('size', {int_or_none}),
 194                 }),
 195             })
 196
 197         return {
 198             'id': video_id,
 199             'formats': formats,
 200             **traverse_obj(video_data, {
 201                 'title': ('videoTitle', {str}),
 202                 'thumbnail': ('videoCover', {url_or_none}),
 203                 'duration': ('videoDuration', {parse_duration}),
 204                 'uploader': ('nickName', {str}),
 205                 'uploader_id': ('uid', {str_or_none}),
 206                 'upload_date': ('videoUploadTime', {unified_strdate}),
 207                 'view_count': ('videoPlayNum', {int_or_none}),
 208                 'comment_count': ('videoCommentNum', {int_or_none}),
 209                 'like_count': ('favorCount', {int_or_none}),
 210             }),
 211         }