From d56da8cf9a0298d3e6e3ee8984f60d4a3217183b Mon Sep 17 00:00:00 2001
From: Shannon Booth <shannon@serenityos.org>
Date: Thu, 15 Aug 2024 20:40:10 +1200
Subject: [PATCH] LibWeb: Sort URLSearchParams using UTF-16 code units

We were previously sorting using code points which could give the wrong
result for certain inputs.

Fixes the last two failing tests on:

https://wpt.live/url/urlsearchparams-sort.any.html
---
 ...rl-search-params-sort-utf16-code-units.txt |  4 ++++
 ...l-search-params-sort-utf16-code-units.html | 17 ++++++++++++++
 .../LibWeb/DOMURL/URLSearchParams.cpp         | 23 ++++++++-----------
 3 files changed, 31 insertions(+), 13 deletions(-)
 create mode 100644 Tests/LibWeb/Text/expected/URL/url-search-params-sort-utf16-code-units.txt
 create mode 100644 Tests/LibWeb/Text/input/URL/url-search-params-sort-utf16-code-units.html

diff --git a/Tests/LibWeb/Text/expected/URL/url-search-params-sort-utf16-code-units.txt b/Tests/LibWeb/Text/expected/URL/url-search-params-sort-utf16-code-units.txt
new file mode 100644
index 00000000000..26e63107b37
--- /dev/null
+++ b/Tests/LibWeb/Text/expected/URL/url-search-params-sort-utf16-code-units.txt
@@ -0,0 +1,4 @@
+%EF%BF%BC=&%EF%BF%BD=x&%EF%BF%BD=a
+'\ufffc' => ''
+'\ufffd' => '\u0078'
+'\ufffd' => '\u0061'
diff --git a/Tests/LibWeb/Text/input/URL/url-search-params-sort-utf16-code-units.html b/Tests/LibWeb/Text/input/URL/url-search-params-sort-utf16-code-units.html
new file mode 100644
index 00000000000..1101024bcfa
--- /dev/null
+++ b/Tests/LibWeb/Text/input/URL/url-search-params-sort-utf16-code-units.html
@@ -0,0 +1,17 @@
+<script src="../include.js"></script>
+<script>
+    function escapeUnicode(str) {
+        return str.replace(/[\s\S]/g, function(c) {
+            return '\\u' + ('0000' + c.charCodeAt(0).toString(16)).slice(-4);
+        });
+    }
+
+    test(() => {
+        let params = new URLSearchParams("\uFFFD=x&\uFFFC&\uFFFD=a");
+        params.sort();
+        println(params.toString())
+        for (const [key, value] of params) {
+            println(`'${escapeUnicode(key)}' => '${escapeUnicode(value)}'`);
+        }
+    })
+</script>
diff --git a/Userland/Libraries/LibWeb/DOMURL/URLSearchParams.cpp b/Userland/Libraries/LibWeb/DOMURL/URLSearchParams.cpp
index 5a386928d0f..3a78274ae66 100644
--- a/Userland/Libraries/LibWeb/DOMURL/URLSearchParams.cpp
+++ b/Userland/Libraries/LibWeb/DOMURL/URLSearchParams.cpp
@@ -325,26 +325,23 @@ void URLSearchParams::set(String const& name, String const& value)
     update();
 }
 
+// https://url.spec.whatwg.org/#dom-urlsearchparams-sort
 void URLSearchParams::sort()
 {
     // 1. Sort all name-value pairs, if any, by their names. Sorting must be done by comparison of code units. The relative order between name-value pairs with equal names must be preserved.
     insertion_sort(m_list, [](auto& a, auto& b) {
-        Utf8View a_code_points { a.name };
-        Utf8View b_code_points { b.name };
+        // FIXME: There should be a way to do this without converting to utf16
+        auto a_utf16 = MUST(utf8_to_utf16(a.name));
+        auto b_utf16 = MUST(utf8_to_utf16(b.name));
 
-        if (a_code_points.starts_with(b_code_points))
-            return false;
-        if (b_code_points.starts_with(a_code_points))
-            return true;
+        auto common_length = min(a_utf16.size(), b_utf16.size());
 
-        for (auto k = a_code_points.begin(), l = b_code_points.begin();
-             k != a_code_points.end() && l != b_code_points.end();
-             ++k, ++l) {
-            if (*k != *l) {
-                return *k < *l;
-            }
+        for (size_t position = 0; position < common_length; ++position) {
+            if (a_utf16[position] != b_utf16[position])
+                return a_utf16[position] < b_utf16[position];
         }
-        VERIFY_NOT_REACHED();
+
+        return a_utf16.size() < b_utf16.size();
     });
 
     // 2. Update this.