yt_dlp/extractor/xiaohongshu.py

   1 import functools
   2
   3 from .common import InfoExtractor
   4 from ..utils import (
   5     float_or_none,
   6     int_or_none,
   7     js_to_json,
   8     url_or_none,
   9 )
  10 from ..utils.traversal import traverse_obj
  11
  12
  13 class XiaoHongShuIE(InfoExtractor):
  14     _VALID_URL = r'https?://www\.xiaohongshu\.com/explore/(?P<id>[\da-f]+)'
  15     IE_DESC = '小红书'
  16     _TESTS = [{
  17         'url': 'https://www.xiaohongshu.com/explore/6411cf99000000001300b6d9',
  18         'md5': '2a87a77ddbedcaeeda8d7eae61b61228',
  19         'info_dict': {
  20             'id': '6411cf99000000001300b6d9',
  21             'ext': 'mp4',
  22             'uploader_id': '5c31698d0000000007018a31',
  23             'description': '#今日快乐今日发[话题]# #吃货薯看这里[话题]# #香妃蛋糕[话题]# #小五卷蛋糕[话题]# #新手蛋糕卷[话题]#',
  24             'title': '香妃蛋糕也太香了吧🔥不需要卷❗️绝对的友好',
  25             'tags': ['今日快乐今日发', '吃货薯看这里', '香妃蛋糕', '小五卷蛋糕', '新手蛋糕卷'],
  26             'duration': 101.726,
  27             'thumbnail': r're:https?://sns-webpic-qc\.xhscdn\.com/\d+/[a-z0-9]+/[\w]+',
  28         },
  29     }]
  30
  31     def _real_extract(self, url):
  32         display_id = self._match_id(url)
  33         webpage = self._download_webpage(url, display_id)
  34         initial_state = self._search_json(
  35             r'window\.__INITIAL_STATE__\s*=', webpage, 'initial state', display_id, transform_source=js_to_json)
  36
  37         note_info = traverse_obj(initial_state, ('note', 'noteDetailMap', display_id, 'note'))
  38         video_info = traverse_obj(note_info, ('video', 'media', 'stream', ('h264', 'av1', 'h265'), ...))
  39
  40         formats = []
  41         for info in video_info:
  42             format_info = traverse_obj(info, {
  43                 'fps': ('fps', {int_or_none}),
  44                 'width': ('width', {int_or_none}),
  45                 'height': ('height', {int_or_none}),
  46                 'vcodec': ('videoCodec', {str}),
  47                 'acodec': ('audioCodec', {str}),
  48                 'abr': ('audioBitrate', {int_or_none}),
  49                 'vbr': ('videoBitrate', {int_or_none}),
  50                 'audio_channels': ('audioChannels', {int_or_none}),
  51                 'tbr': ('avgBitrate', {int_or_none}),
  52                 'format': ('qualityType', {str}),
  53                 'filesize': ('size', {int_or_none}),
  54                 'duration': ('duration', {functools.partial(float_or_none, scale=1000)}),
  55             })
  56
  57             formats.extend(traverse_obj(info, (('mediaUrl', ('backupUrls', ...)), {
  58                 lambda u: url_or_none(u) and {'url': u, **format_info}})))
  59
  60         thumbnails = []
  61         for image_info in traverse_obj(note_info, ('imageList', ...)):
  62             thumbnail_info = traverse_obj(image_info, {
  63                 'height': ('height', {int_or_none}),
  64                 'width': ('width', {int_or_none}),
  65             })
  66             for thumb_url in traverse_obj(image_info, (('urlDefault', 'urlPre'), {url_or_none})):
  67                 thumbnails.append({
  68                     'url': thumb_url,
  69                     **thumbnail_info,
  70                 })
  71
  72         return {
  73             'id': display_id,
  74             'formats': formats,
  75             'thumbnails': thumbnails,
  76             'title': self._html_search_meta(['og:title'], webpage, default=None),
  77             **traverse_obj(note_info, {
  78                 'title': ('title', {str}),
  79                 'description': ('desc', {str}),
  80                 'tags': ('tagList', ..., 'name', {str}),
  81                 'uploader_id': ('user', 'userId', {str}),
  82             }),
  83         }