yt_dlp/extractor/ximalaya.py

   1 import base64
   2 import math
   3 import time
   4
   5 from .common import InfoExtractor
   6 from .videa import VideaIE
   7 from ..utils import (
   8     InAdvancePagedList,
   9     int_or_none,
  10     str_or_none,
  11     traverse_obj,
  12     try_call,
  13     update_url_query,
  14 )
  15
  16
  17 class XimalayaBaseIE(InfoExtractor):
  18     _GEO_COUNTRIES = ['CN']
  19
  20
  21 class XimalayaIE(XimalayaBaseIE):
  22     IE_NAME = 'ximalaya'
  23     IE_DESC = '喜马拉雅FM'
  24     _VALID_URL = r'https?://(?:www\.|m\.)?ximalaya\.com/(?:(?P<uid>\d+)/)?sound/(?P<id>[0-9]+)'
  25     _TESTS = [
  26         {
  27             'url': 'http://www.ximalaya.com/sound/47740352/',
  28             'info_dict': {
  29                 'id': '47740352',
  30                 'ext': 'm4a',
  31                 'uploader': '小彬彬爱听书',
  32                 'uploader_id': '61425525',
  33                 'uploader_url': 'http://www.ximalaya.com/zhubo/61425525/',
  34                 'title': '261.唐诗三百首.卷八.送孟浩然之广陵.李白',
  35                 'description': 'contains:《送孟浩然之广陵》\n作者：李白\n故人西辞黄鹤楼，烟花三月下扬州。\n孤帆远影碧空尽，惟见长江天际流。',
  36                 'thumbnail': r're:^https?://.*\.jpg',
  37                 'thumbnails': [
  38                     {
  39                         'name': 'cover_url',
  40                         'url': r're:^https?://.*\.jpg',
  41                     },
  42                     {
  43                         'name': 'cover_url_142',
  44                         'url': r're:^https?://.*\.jpg',
  45                         'width': 180,
  46                         'height': 180,
  47                     },
  48                 ],
  49                 'categories': ['其他'],
  50                 'duration': 93,
  51                 'view_count': int,
  52                 'like_count': int,
  53             },
  54         },
  55         {
  56             'url': 'http://m.ximalaya.com/61425525/sound/47740352/',
  57             'info_dict': {
  58                 'id': '47740352',
  59                 'ext': 'm4a',
  60                 'uploader': '小彬彬爱听书',
  61                 'uploader_id': '61425525',
  62                 'uploader_url': 'http://www.ximalaya.com/zhubo/61425525/',
  63                 'title': '261.唐诗三百首.卷八.送孟浩然之广陵.李白',
  64                 'description': 'contains:《送孟浩然之广陵》\n作者：李白\n故人西辞黄鹤楼，烟花三月下扬州。\n孤帆远影碧空尽，惟见长江天际流。',
  65                 'thumbnail': r're:^https?://.*\.jpg',
  66                 'thumbnails': [
  67                     {
  68                         'name': 'cover_url',
  69                         'url': r're:^https?://.*\.jpg',
  70                     },
  71                     {
  72                         'name': 'cover_url_142',
  73                         'url': r're:^https?://.*\.jpg',
  74                         'width': 180,
  75                         'height': 180,
  76                     },
  77                 ],
  78                 'categories': ['人文'],
  79                 'duration': 93,
  80                 'view_count': int,
  81                 'like_count': int,
  82             },
  83         },
  84         {
  85             # VIP-restricted audio
  86             'url': 'https://www.ximalaya.com/sound/562111701',
  87             'only_matching': True,
  88         },
  89     ]
  90
  91     @staticmethod
  92     def _decrypt_filename(file_id, seed):
  93         cgstr = ''
  94         key = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ/\\:._-1234567890'
  95         for _ in key:
  96             seed = float(int(211 * seed + 30031) % 65536)
  97             r = int(seed / 65536 * len(key))
  98             cgstr += key[r]
  99             key = key.replace(key[r], '')
 100         parts = file_id.split('*')
 101         filename = ''.join(cgstr[int(part)] for part in parts if part.isdecimal())
 102         if not filename.startswith('/'):
 103             filename = '/' + filename
 104         return filename
 105
 106     @staticmethod
 107     def _decrypt_url_params(encrypted_params):
 108         params = VideaIE.rc4(
 109             base64.b64decode(encrypted_params), 'xkt3a41psizxrh9l').split('-')
 110         # sign, token, timestamp
 111         return params[1], params[2], params[3]
 112
 113     def _real_extract(self, url):
 114         scheme = 'https' if url.startswith('https') else 'http'
 115
 116         audio_id = self._match_id(url)
 117         audio_info = self._download_json(
 118             f'{scheme}://m.ximalaya.com/tracks/{audio_id}.json', audio_id,
 119             'Downloading info json', 'Unable to download info file')
 120
 121         formats = []
 122         # NOTE: VIP-restricted audio
 123         if audio_info.get('is_paid'):
 124             ts = int(time.time())
 125             vip_info = self._download_json(
 126                 f'{scheme}://mpay.ximalaya.com/mobile/track/pay/{audio_id}/{ts}',
 127                 audio_id, 'Downloading VIP info json', 'Unable to download VIP info file',
 128                 query={'device': 'pc', 'isBackend': 'true', '_': ts})
 129             filename = self._decrypt_filename(vip_info['fileId'], vip_info['seed'])
 130             sign, token, timestamp = self._decrypt_url_params(vip_info['ep'])
 131             vip_url = update_url_query(
 132                 f'{vip_info["domain"]}/download/{vip_info["apiVersion"]}{filename}', {
 133                     'sign': sign,
 134                     'token': token,
 135                     'timestamp': timestamp,
 136                     'buy_key': vip_info['buyKey'],
 137                     'duration': vip_info['duration'],
 138                 })
 139             fmt = {
 140                 'format_id': 'vip',
 141                 'url': vip_url,
 142                 'vcodec': 'none',
 143             }
 144             if '_preview_' in vip_url:
 145                 self.report_warning(
 146                     f'This tracks requires a VIP account. Using a sample instead. {self._login_hint()}')
 147                 fmt.update({
 148                     'format_note': 'Sample',
 149                     'preference': -10,
 150                     **traverse_obj(vip_info, {
 151                         'filesize': ('sampleLength', {int_or_none}),
 152                         'duration': ('sampleDuration', {int_or_none}),
 153                     }),
 154                 })
 155             else:
 156                 fmt.update(traverse_obj(vip_info, {
 157                     'filesize': ('totalLength', {int_or_none}),
 158                     'duration': ('duration', {int_or_none}),
 159                 }))
 160
 161             fmt['abr'] = try_call(lambda: fmt['filesize'] * 8 / fmt['duration'] / 1024)
 162             formats.append(fmt)
 163
 164         formats.extend([{
 165             'format_id': f'{bps}k',
 166             'url': audio_info[k],
 167             'abr': bps,
 168             'vcodec': 'none',
 169         } for bps, k in ((24, 'play_path_32'), (64, 'play_path_64')) if audio_info.get(k)])
 170
 171         thumbnails = []
 172         for k in audio_info:
 173             # cover pics kyes like: cover_url', 'cover_url_142'
 174             if k.startswith('cover_url'):
 175                 thumbnail = {'name': k, 'url': audio_info[k]}
 176                 if k == 'cover_url_142':
 177                     thumbnail['width'] = 180
 178                     thumbnail['height'] = 180
 179                 thumbnails.append(thumbnail)
 180
 181         audio_uploader_id = audio_info.get('uid')
 182
 183         audio_description = try_call(
 184             lambda: audio_info['intro'].replace('\r\n\r\n\r\n ', '\n').replace('\r\n', '\n'))
 185
 186         return {
 187             'id': audio_id,
 188             'uploader': audio_info.get('nickname'),
 189             'uploader_id': str_or_none(audio_uploader_id),
 190             'uploader_url': f'{scheme}://www.ximalaya.com/zhubo/{audio_uploader_id}/' if audio_uploader_id else None,
 191             'title': audio_info['title'],
 192             'thumbnails': thumbnails,
 193             'description': audio_description,
 194             'categories': list(filter(None, [audio_info.get('category_name')])),
 195             'duration': audio_info.get('duration'),
 196             'view_count': audio_info.get('play_count'),
 197             'like_count': audio_info.get('favorites_count'),
 198             'formats': formats,
 199         }
 200
 201
 202 class XimalayaAlbumIE(XimalayaBaseIE):
 203     IE_NAME = 'ximalaya:album'
 204     IE_DESC = '喜马拉雅FM 专辑'
 205     _VALID_URL = r'https?://(?:www\.|m\.)?ximalaya\.com/(?:\d+/)?album/(?P<id>[0-9]+)'
 206     _TESTS = [{
 207         'url': 'http://www.ximalaya.com/61425525/album/5534601/',
 208         'info_dict': {
 209             'title': '唐诗三百首（含赏析）',
 210             'id': '5534601',
 211         },
 212         'playlist_mincount': 323,
 213     }, {
 214         'url': 'https://www.ximalaya.com/album/6912905',
 215         'info_dict': {
 216             'title': '埃克哈特《修炼当下的力量》',
 217             'id': '6912905',
 218         },
 219         'playlist_mincount': 41,
 220     }]
 221
 222     def _real_extract(self, url):
 223         playlist_id = self._match_id(url)
 224
 225         first_page = self._fetch_page(playlist_id, 1)
 226         page_count = math.ceil(first_page['trackTotalCount'] / first_page['pageSize'])
 227
 228         entries = InAdvancePagedList(
 229             lambda idx: self._get_entries(self._fetch_page(playlist_id, idx + 1) if idx else first_page),
 230             page_count, first_page['pageSize'])
 231
 232         title = traverse_obj(first_page, ('tracks', 0, 'albumTitle'), expected_type=str)
 233
 234         return self.playlist_result(entries, playlist_id, title)
 235
 236     def _fetch_page(self, playlist_id, page_idx):
 237         return self._download_json(
 238             'https://www.ximalaya.com/revision/album/v1/getTracksList',
 239             playlist_id, note=f'Downloading tracks list page {page_idx}',
 240             query={'albumId': playlist_id, 'pageNum': page_idx})['data']
 241
 242     def _get_entries(self, page_data):
 243         for e in page_data['tracks']:
 244             yield self.url_result(
 245                 self._proto_relative_url(f'//www.ximalaya.com{e["url"]}'),
 246                 XimalayaIE, e.get('trackId'), e.get('title'))