From 62c754fe23f3382036a38f72c4bd947d9501a5f1 Mon Sep 17 00:00:00 2001
From: MDLeom <2809763-curben@users.noreply.gitlab.com>
Date: Wed, 19 Mar 2025 10:22:34 +0000
Subject: [PATCH] fix(clean_url): encode ^
this is unrelated to address separator of adblock filter syntax
https://adblockplus.org/filter-cheatsheet#blocking1
which does not need to be escaped
uBO can handle this pattern just fine
||example.com/so%5Emany?ca^r=e^t^$all
query string does not need to percent-encode ^,
only pathname
---
src/clean_url.js | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/src/clean_url.js b/src/clean_url.js
index b9613710..9581aeb9 100644
--- a/src/clean_url.js
+++ b/src/clean_url.js
@@ -30,7 +30,11 @@ for await (const line of createInterface({ input: process.stdin, terminal: false
url = new URL(url.searchParams.get('url'))
}
- const outUrl = `${url.host.replace(/^www\./, '')}${url.pathname}${url.search}`
+ url.host = url.host.replace(/^www\./, '')
+ // nodejs does not percent-encode ^ yet
+ // https://github.com/nodejs/node/issues/57313
+ url.pathname = url.pathname.replaceAll('^', encodeURI('^'))
+ const outUrl = `${url.host}${url.pathname}${url.search}`
// remove trailing slash from domain except path #43
.replace(/(^[^/]*)\/+$/, '$1')