yt_dlp/extractor/xiaohongshu.py

   1
   2 from .common import InfoExtractor
   3 from ..utils import (
   4     float_or_none,
   5     int_or_none,
   6     js_to_json,
   7     url_or_none,
   8     urlhandle_detect_ext,
   9 )
  10 from ..utils.traversal import traverse_obj
  11
  12
  13 class XiaoHongShuIE(InfoExtractor):
  14     _VALID_URL = r'https?://www\.xiaohongshu\.com/(?:explore|discovery/item)/(?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         'url': 'https://www.xiaohongshu.com/discovery/item/674051740000000007027a15?xsec_token=CBgeL8Dxd1ZWBhwqRd568gAZ_iwG-9JIf9tnApNmteU2E=',
  31         'info_dict': {
  32             'id': '674051740000000007027a15',
  33             'ext': 'mp4',
  34             'title': '相互喜欢就可以了',
  35             'uploader_id': '63439913000000001901f49a',
  36             'duration': 28.073,
  37             'description': '#广州[话题]# #深圳[话题]# #香港[话题]# #街头采访[话题]# #是你喜欢的类型[话题]#',
  38             'thumbnail': r're:https?://sns-webpic-qc\.xhscdn\.com/\d+/[\da-f]+/[^/]+',
  39             'tags': ['广州', '深圳', '香港', '街头采访', '是你喜欢的类型'],
  40         },
  41     }]
  42
  43     def _real_extract(self, url):
  44         display_id = self._match_id(url)
  45         webpage = self._download_webpage(url, display_id)
  46         initial_state = self._search_json(
  47             r'window\.__INITIAL_STATE__\s*=', webpage, 'initial state', display_id, transform_source=js_to_json)
  48
  49         note_info = traverse_obj(initial_state, ('note', 'noteDetailMap', display_id, 'note'))
  50         video_info = traverse_obj(note_info, ('video', 'media', 'stream', ..., ...))
  51
  52         formats = []
  53         for info in video_info:
  54             format_info = traverse_obj(info, {
  55                 'fps': ('fps', {int_or_none}),
  56                 'width': ('width', {int_or_none}),
  57                 'height': ('height', {int_or_none}),
  58                 'vcodec': ('videoCodec', {str}),
  59                 'acodec': ('audioCodec', {str}),
  60                 'abr': ('audioBitrate', {int_or_none(scale=1000)}),
  61                 'vbr': ('videoBitrate', {int_or_none(scale=1000)}),
  62                 'audio_channels': ('audioChannels', {int_or_none}),
  63                 'tbr': ('avgBitrate', {int_or_none(scale=1000)}),
  64                 'format': ('qualityType', {str}),
  65                 'filesize': ('size', {int_or_none}),
  66                 'duration': ('duration', {float_or_none(scale=1000)}),
  67             })
  68
  69             formats.extend(traverse_obj(info, (('masterUrl', ('backupUrls', ...)), {
  70                 lambda u: url_or_none(u) and {'url': u, **format_info}})))
  71
  72         if origin_key := traverse_obj(note_info, ('video', 'consumer', 'originVideoKey', {str})):
  73             # Not using a head request because of false negatives
  74             urlh = self._request_webpage(
  75                 f'https://sns-video-bd.xhscdn.com/{origin_key}', display_id,
  76                 'Checking original video availability', 'Original video is not available', fatal=False)
  77             if urlh:
  78                 formats.append({
  79                     'format_id': 'direct',
  80                     'ext': urlhandle_detect_ext(urlh, default='mp4'),
  81                     'filesize': int_or_none(urlh.get_header('Content-Length')),
  82                     'url': urlh.url,
  83                     'quality': 1,
  84                 })
  85
  86         thumbnails = []
  87         for image_info in traverse_obj(note_info, ('imageList', ...)):
  88             thumbnail_info = traverse_obj(image_info, {
  89                 'height': ('height', {int_or_none}),
  90                 'width': ('width', {int_or_none}),
  91             })
  92             for thumb_url in traverse_obj(image_info, (('urlDefault', 'urlPre'), {url_or_none})):
  93                 thumbnails.append({
  94                     'url': thumb_url,
  95                     **thumbnail_info,
  96                 })
  97
  98         return {
  99             'id': display_id,
 100             'formats': formats,
 101             'thumbnails': thumbnails,
 102             'title': self._html_search_meta(['og:title'], webpage, default=None),
 103             **traverse_obj(note_info, {
 104                 'title': ('title', {str}),
 105                 'description': ('desc', {str}),
 106                 'tags': ('tagList', ..., 'name', {str}),
 107                 'uploader_id': ('user', 'userId', {str}),
 108             }),
 109         }