From a378c619485cb900c2de8f1aa768ffaf981d2320 Mon Sep 17 00:00:00 2001
From: Emilia Bopp <contact@ebopp.de>
Date: Sun, 10 Nov 2024 14:39:21 +0100
Subject: [PATCH] libutil: fix non-ASCII chars in URL encoding

Due to the cast to (unsigned int), the encoding appended broken
bytes padding. This is fixed here with a bitmask.

Fixes: https://git.lix.systems/lix-project/lix/issues/562
Change-Id: I0c93bd2b8c2f82df208d4693b7254544e3121dc3
---
 src/libutil/url.cc        | 2 +-
 tests/unit/libutil/url.cc | 7 ++++++-
 2 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/src/libutil/url.cc b/src/libutil/url.cc
index 2de50dd4d..f0586189b 100644
--- a/src/libutil/url.cc
+++ b/src/libutil/url.cc
@@ -104,7 +104,7 @@ std::string percentEncode(std::string_view s, std::string_view keep)
             || keep.find(c) != std::string::npos)
             res += c;
         else
-            res += fmt("%%%02X", (unsigned int) c);
+            res += fmt("%%%02X", 0xff & (unsigned int) c);
     return res;
 }
 
diff --git a/tests/unit/libutil/url.cc b/tests/unit/libutil/url.cc
index bfd9a228a..ecf544618 100644
--- a/tests/unit/libutil/url.cc
+++ b/tests/unit/libutil/url.cc
@@ -303,7 +303,6 @@ namespace nix {
         ASSERT_EQ(d, s);
     }
 
-
     /* ----------------------------------------------------------------------------
      * percentEncode
      * --------------------------------------------------------------------------*/
@@ -336,4 +335,10 @@ namespace nix {
         ASSERT_EQ(d, s);
     }
 
+    TEST(percentEncode, utf8Input) {
+        std::string s = percentEncode("ä");
+        std::string d = "%C3%A4";
+
+        ASSERT_EQ(d, s);
+    }
 }