yt_dlp/extractor/douyutv.py

   1 import hashlib
   2 import time
   3 import urllib
   4 import uuid
   5
   6 from .common import InfoExtractor
   7 from .openload import PhantomJSwrapper
   8 from ..utils import (
   9     ExtractorError,
  10     UserNotLive,
  11     determine_ext,
  12     int_or_none,
  13     js_to_json,
  14     parse_resolution,
  15     str_or_none,
  16     traverse_obj,
  17     unescapeHTML,
  18     url_or_none,
  19     urlencode_postdata,
  20     urljoin,
  21 )
  22
  23
  24 class DouyuBaseIE(InfoExtractor):
  25     def _download_cryptojs_md5(self, video_id):
  26         for url in [
  27             'https://cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/rollups/md5.js',
  28             'https://cdn.bootcdn.net/ajax/libs/crypto-js/3.1.2/rollups/md5.js',
  29         ]:
  30             js_code = self._download_webpage(
  31                 url, video_id, note='Downloading signing dependency', fatal=False)
  32             if js_code:
  33                 self.cache.store('douyu', 'crypto-js-md5', js_code)
  34                 return js_code
  35         raise ExtractorError('Unable to download JS dependency (crypto-js/md5)')
  36
  37     def _get_cryptojs_md5(self, video_id):
  38         return self.cache.load('douyu', 'crypto-js-md5') or self._download_cryptojs_md5(video_id)
  39
  40     def _calc_sign(self, sign_func, video_id, a):
  41         b = uuid.uuid4().hex
  42         c = round(time.time())
  43         js_script = f'{self._get_cryptojs_md5(video_id)};{sign_func};console.log(ub98484234("{a}","{b}","{c}"))'
  44         phantom = PhantomJSwrapper(self)
  45         result = phantom.execute(js_script, video_id,
  46                                  note='Executing JS signing script').strip()
  47         return {i: v[0] for i, v in urllib.parse.parse_qs(result).items()}
  48
  49     def _search_js_sign_func(self, webpage, fatal=True):
  50         # The greedy look-behind ensures last possible script tag is matched
  51         return self._search_regex(
  52             r'(?:<script.*)?<script[^>]*>(.*?ub98484234.*?)</script>', webpage, 'JS sign func', fatal=fatal)
  53
  54
  55 class DouyuTVIE(DouyuBaseIE):
  56     IE_DESC = '斗鱼直播'
  57     _VALID_URL = r'https?://(?:www\.)?douyu(?:tv)?\.com/(topic/\w+\?rid=|(?:[^/]+/))*(?P<id>[A-Za-z0-9]+)'
  58     _TESTS = [{
  59         'url': 'https://www.douyu.com/pigff',
  60         'info_dict': {
  61             'id': '24422',
  62             'display_id': 'pigff',
  63             'ext': 'mp4',
  64             'title': 're:^【PIGFF】.* [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
  65             'description': r'≥15级牌子看鱼吧置顶帖进粉丝vx群',
  66             'thumbnail': str,
  67             'uploader': 'pigff',
  68             'is_live': True,
  69             'live_status': 'is_live',
  70         },
  71         'params': {
  72             'skip_download': True,
  73         },
  74     }, {
  75         'url': 'http://www.douyutv.com/85982',
  76         'info_dict': {
  77             'id': '85982',
  78             'display_id': '85982',
  79             'ext': 'flv',
  80             'title': 're:^小漠从零单排记！——CSOL2躲猫猫 [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
  81             'description': 'md5:746a2f7a253966a06755a912f0acc0d2',
  82             'thumbnail': r're:^https?://.*\.png',
  83             'uploader': 'douyu小漠',
  84             'is_live': True,
  85         },
  86         'params': {
  87             'skip_download': True,
  88         },
  89         'skip': 'Room not found',
  90     }, {
  91         'url': 'http://www.douyutv.com/17732',
  92         'info_dict': {
  93             'id': '17732',
  94             'display_id': '17732',
  95             'ext': 'flv',
  96             'title': 're:^清晨醒脑！根本停不下来！ [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
  97             'description': r're:.*m7show@163\.com.*',
  98             'thumbnail': r're:^https?://.*\.png',
  99             'uploader': '7师傅',
 100             'is_live': True,
 101         },
 102         'params': {
 103             'skip_download': True,
 104         },
 105     }, {
 106         'url': 'https://www.douyu.com/topic/ydxc?rid=6560603',
 107         'info_dict': {
 108             'id': '6560603',
 109             'display_id': '6560603',
 110             'ext': 'flv',
 111             'title': 're:^阿余：新年快乐恭喜发财！ [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
 112             'description': 're:.*直播时间.*',
 113             'thumbnail': r're:^https?://.*\.png',
 114             'uploader': '阿涛皎月Carry',
 115             'live_status': 'is_live',
 116         },
 117         'params': {
 118             'skip_download': True,
 119         },
 120     }, {
 121         'url': 'http://www.douyu.com/xiaocang',
 122         'only_matching': True,
 123     }, {
 124         # \"room_id\"
 125         'url': 'http://www.douyu.com/t/lpl',
 126         'only_matching': True,
 127     }]
 128
 129     def _get_sign_func(self, room_id, video_id):
 130         return self._download_json(
 131             f'https://www.douyu.com/swf_api/homeH5Enc?rids={room_id}', video_id,
 132             note='Getting signing script')['data'][f'room{room_id}']
 133
 134     def _extract_stream_formats(self, stream_formats):
 135         formats = []
 136         for stream_info in traverse_obj(stream_formats, (..., 'data')):
 137             stream_url = urljoin(
 138                 traverse_obj(stream_info, 'rtmp_url'), traverse_obj(stream_info, 'rtmp_live'))
 139             if stream_url:
 140                 rate_id = traverse_obj(stream_info, ('rate', {int_or_none}))
 141                 rate_info = traverse_obj(stream_info, ('multirates', lambda _, v: v['rate'] == rate_id), get_all=False)
 142                 ext = determine_ext(stream_url)
 143                 formats.append({
 144                     'url': stream_url,
 145                     'format_id': str_or_none(rate_id),
 146                     'ext': 'mp4' if ext == 'm3u8' else ext,
 147                     'protocol': 'm3u8_native' if ext == 'm3u8' else 'https',
 148                     'quality': rate_id % -10000 if rate_id is not None else None,
 149                     **traverse_obj(rate_info, {
 150                         'format': ('name', {str_or_none}),
 151                         'tbr': ('bit', {int_or_none}),
 152                     }),
 153                 })
 154         return formats
 155
 156     def _real_extract(self, url):
 157         video_id = self._match_id(url)
 158
 159         webpage = self._download_webpage(url, video_id)
 160         room_id = self._search_regex(r'\$ROOM\.room_id\s*=\s*(\d+)', webpage, 'room id')
 161
 162         if self._search_regex(r'"videoLoop"\s*:\s*(\d+)', webpage, 'loop', default='') == '1':
 163             raise UserNotLive('The channel is auto-playing VODs', video_id=video_id)
 164         if self._search_regex(r'\$ROOM\.show_status\s*=\s*(\d+)', webpage, 'status', default='') == '2':
 165             raise UserNotLive(video_id=video_id)
 166
 167         # Grab metadata from API
 168         params = {
 169             'aid': 'wp',
 170             'client_sys': 'wp',
 171             'time': int(time.time()),
 172         }
 173         params['auth'] = hashlib.md5(
 174             f'room/{room_id}?{urllib.parse.urlencode(params)}zNzMV1y4EMxOHS6I5WKm'.encode()).hexdigest()
 175         room = traverse_obj(self._download_json(
 176             f'http://www.douyutv.com/api/v1/room/{room_id}', video_id,
 177             note='Downloading room info', query=params, fatal=False), 'data')
 178
 179         # 1 = live, 2 = offline
 180         if traverse_obj(room, 'show_status') == '2':
 181             raise UserNotLive(video_id=video_id)
 182
 183         js_sign_func = self._search_js_sign_func(webpage, fatal=False) or self._get_sign_func(room_id, video_id)
 184         form_data = {
 185             'rate': 0,
 186             **self._calc_sign(js_sign_func, video_id, room_id),
 187         }
 188         stream_formats = [self._download_json(
 189             f'https://www.douyu.com/lapi/live/getH5Play/{room_id}',
 190             video_id, note="Downloading livestream format",
 191             data=urlencode_postdata(form_data))]
 192
 193         for rate_id in traverse_obj(stream_formats[0], ('data', 'multirates', ..., 'rate')):
 194             if rate_id != traverse_obj(stream_formats[0], ('data', 'rate')):
 195                 form_data['rate'] = rate_id
 196                 stream_formats.append(self._download_json(
 197                     f'https://www.douyu.com/lapi/live/getH5Play/{room_id}',
 198                     video_id, note=f'Downloading livestream format {rate_id}',
 199                     data=urlencode_postdata(form_data)))
 200
 201         return {
 202             'id': room_id,
 203             'formats': self._extract_stream_formats(stream_formats),
 204             'is_live': True,
 205             **traverse_obj(room, {
 206                 'display_id': ('url', {str}, {lambda i: i[1:]}),
 207                 'title': ('room_name', {unescapeHTML}),
 208                 'description': ('show_details', {str}),
 209                 'uploader': ('nickname', {str}),
 210                 'thumbnail': ('room_src', {url_or_none}),
 211             })
 212         }
 213
 214
 215 class DouyuShowIE(DouyuBaseIE):
 216     _VALID_URL = r'https?://v(?:mobile)?\.douyu\.com/show/(?P<id>[0-9a-zA-Z]+)'
 217
 218     _TESTS = [{
 219         'url': 'https://v.douyu.com/show/mPyq7oVNe5Yv1gLY',
 220         'info_dict': {
 221             'id': 'mPyq7oVNe5Yv1gLY',
 222             'ext': 'mp4',
 223             'title': '四川人小时候的味道“蒜苗回锅肉”，传统菜不能丢，要常做来吃',
 224             'duration': 633,
 225             'thumbnail': str,
 226             'uploader': '美食作家王刚V',
 227             'uploader_id': 'OVAO4NVx1m7Q',
 228             'timestamp': 1661850002,
 229             'upload_date': '20220830',
 230             'view_count': int,
 231             'tags': ['美食', '美食综合'],
 232         },
 233     }, {
 234         'url': 'https://vmobile.douyu.com/show/rjNBdvnVXNzvE2yw',
 235         'only_matching': True,
 236     }]
 237
 238     _FORMATS = {
 239         'super': '原画',
 240         'high': '超清',
 241         'normal': '高清',
 242     }
 243
 244     _QUALITIES = {
 245         'super': -1,
 246         'high': -2,
 247         'normal': -3,
 248     }
 249
 250     _RESOLUTIONS = {
 251         'super': '1920x1080',
 252         'high': '1280x720',
 253         'normal': '852x480',
 254     }
 255
 256     def _real_extract(self, url):
 257         url = url.replace('vmobile.', 'v.')
 258         video_id = self._match_id(url)
 259
 260         webpage = self._download_webpage(url, video_id)
 261
 262         video_info = self._search_json(
 263             r'<script>\s*window\.\$DATA\s*=', webpage,
 264             'video info', video_id, transform_source=js_to_json)
 265
 266         js_sign_func = self._search_js_sign_func(webpage)
 267         form_data = {
 268             'vid': video_id,
 269             **self._calc_sign(js_sign_func, video_id, video_info['ROOM']['point_id']),
 270         }
 271         url_info = self._download_json(
 272             'https://v.douyu.com/api/stream/getStreamUrl', video_id,
 273             data=urlencode_postdata(form_data), note="Downloading video formats")
 274
 275         formats = []
 276         for name, url in traverse_obj(url_info, ('data', 'thumb_video', {dict.items}, ...)):
 277             video_url = traverse_obj(url, ('url', {url_or_none}))
 278             if video_url:
 279                 ext = determine_ext(video_url)
 280                 formats.append({
 281                     'format': self._FORMATS.get(name),
 282                     'format_id': name,
 283                     'url': video_url,
 284                     'quality': self._QUALITIES.get(name),
 285                     'ext': 'mp4' if ext == 'm3u8' else ext,
 286                     'protocol': 'm3u8_native' if ext == 'm3u8' else 'https',
 287                     **parse_resolution(self._RESOLUTIONS.get(name))
 288                 })
 289             else:
 290                 self.to_screen(
 291                     f'"{self._FORMATS.get(name, name)}" format may require logging in. {self._login_hint()}')
 292
 293         return {
 294             'id': video_id,
 295             'formats': formats,
 296             **traverse_obj(video_info, ('DATA', {
 297                 'title': ('content', 'title', {str}),
 298                 'uploader': ('content', 'author', {str}),
 299                 'uploader_id': ('content', 'up_id', {str_or_none}),
 300                 'duration': ('content', 'video_duration', {int_or_none}),
 301                 'thumbnail': ('content', 'video_pic', {url_or_none}),
 302                 'timestamp': ('content', 'create_time', {int_or_none}),
 303                 'view_count': ('content', 'view_num', {int_or_none}),
 304                 'tags': ('videoTag', ..., 'tagName', {str}),
 305             }))
 306         }