From 66f4c04e50d9213522095247666d3d90345ad5d1 Mon Sep 17 00:00:00 2001
From: The Hatsune Daishi <nao20010128@gmail.com>
Date: Wed, 8 Dec 2021 01:38:50 +0900
Subject: [extractor] Add `_search_nuxt_data` (#1921)

Authored by: nao20010128nao
---
 yt_dlp/extractor/common.py | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

(limited to 'yt_dlp/extractor/common.py')

diff --git a/yt_dlp/extractor/common.py b/yt_dlp/extractor/common.py
index 2180f879c..d8fc5272c 100644
--- a/yt_dlp/extractor/common.py
+++ b/yt_dlp/extractor/common.py
@@ -1513,6 +1513,24 @@ class InfoExtractor(object):
                 webpage, 'next.js data', **kw),
             video_id, **kw)
 
+    def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__'):
+        ''' Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function. '''
+        # not all website do this, but it can be changed
+        # https://stackoverflow.com/questions/67463109/how-to-change-or-hide-nuxt-and-nuxt-keyword-in-page-source
+        rectx = re.escape(context_name)
+        js, arg_keys, arg_vals = self._search_regex(
+            (r'<script>window\.%s=\(function\((?P<arg_keys>.*?)\)\{return\s(?P<js>\{.*?\})\}\((?P<arg_vals>.+?)\)\);?</script>' % rectx,
+             r'%s\(.*?\(function\((?P<arg_keys>.*?)\)\{return\s(?P<js>\{.*?\})\}\((?P<arg_vals>.*?)\)' % rectx),
+            webpage, context_name, group=['js', 'arg_keys', 'arg_vals'])
+
+        args = dict(zip(arg_keys.split(','), arg_vals.split(',')))
+
+        for key, val in args.items():
+            if val in ('undefined', 'void 0'):
+                args[key] = 'null'
+
+        return self._parse_json(js_to_json(js, args), video_id)['data'][0]
+
     @staticmethod
     def _hidden_inputs(html):
         html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html)
-- 
cgit v1.2.3