From 033ea0e7fb0f72338ae95aa0413da838206440bb Mon Sep 17 00:00:00 2001
From: Shannon Booth <shannon@serenityos.org>
Date: Sat, 10 Aug 2024 17:16:01 +1200
Subject: [PATCH] AK: Add String::from_utf8_with_replacement_character

This takes a byte sequence and converts it to a UTF-8 string with the
replacement character.
---
 AK/String.cpp           | 10 ++++++++++
 AK/String.h             |  3 +++
 Tests/AK/TestString.cpp | 13 +++++++++++++
 3 files changed, 26 insertions(+)

diff --git a/AK/String.cpp b/AK/String.cpp
index dd6502740f6..fd4c50c33e5 100644
--- a/AK/String.cpp
+++ b/AK/String.cpp
@@ -20,6 +20,16 @@
 
 namespace AK {
 
+String String::from_utf8_with_replacement_character(StringView view)
+{
+    StringBuilder builder;
+
+    for (auto c : Utf8View { view })
+        builder.append_code_point(c);
+
+    return builder.to_string_without_validation();
+}
+
 String String::from_utf8_without_validation(ReadonlyBytes bytes)
 {
     String result;
diff --git a/AK/String.h b/AK/String.h
index 130087eaae8..14f8e2101c3 100644
--- a/AK/String.h
+++ b/AK/String.h
@@ -51,6 +51,9 @@ public:
     // Creates a new String from a sequence of UTF-8 encoded code points.
     static ErrorOr<String> from_utf8(StringView);
 
+    // Creates a new String using the replacement character for invalid bytes
+    [[nodiscard]] static String from_utf8_with_replacement_character(StringView);
+
     template<typename T>
     requires(IsOneOf<RemoveCVReference<T>, ByteString, DeprecatedFlyString, FlyString, String>)
     static ErrorOr<String> from_utf8(T&&) = delete;
diff --git a/Tests/AK/TestString.cpp b/Tests/AK/TestString.cpp
index f803e799529..ddb012f4efc 100644
--- a/Tests/AK/TestString.cpp
+++ b/Tests/AK/TestString.cpp
@@ -173,6 +173,19 @@ TEST_CASE(invalid_utf8)
     EXPECT(string3.error().string_literal().contains("Input was not valid UTF-8"sv));
 }
 
+TEST_CASE(with_replacement_character)
+{
+    auto string1 = String::from_utf8_with_replacement_character("long string \xf4\x8f\xbf\xc0"sv); // U+110000
+    Array<u8, 24> string1_expected { 0x6c, 0x6f, 0x6e, 0x67, 0x20, 0x73, 0x74, 0x72, 0x69, 0x6e, 0x67, 0x20, 0xef, 0xbf, 0xbd, 0xef, 0xbf, 0xbd, 0xef, 0xbf, 0xbd, 0xef, 0xbf, 0xbd };
+    EXPECT_EQ(string1.bytes(), string1_expected);
+
+    auto string3 = String::from_utf8_with_replacement_character("A valid string!"sv);
+    EXPECT_EQ(string3, "A valid string!"sv);
+
+    auto string4 = String::from_utf8_with_replacement_character(""sv);
+    EXPECT_EQ(string4, ""sv);
+}
+
 TEST_CASE(from_code_points)
 {
     for (u32 code_point = 0; code_point < 0x80; ++code_point) {