From 0bf565b97fc3b4d63d8cded2d17737f7190ca5fb Mon Sep 17 00:00:00 2001
From: Timothy Flynn <trflynn89@pm.me>
Date: Sat, 2 Aug 2025 19:12:16 -0400
Subject: [PATCH] AK: Allow comparing UTF-16 strings to UTF-8 strings

Before now, you could compare a Utf16View to a StringView, but it would
only be valid if the StringView were ASCII. When porting code to UTF-16,
it will be handy to have a code point-aware implementation for non-ASCII
StringViews.
---
 AK/Utf16View.h               | 17 ++++++++++++++++-
 Tests/AK/TestUtf16String.cpp |  2 +-
 Tests/AK/TestUtf16View.cpp   | 18 ++++++++++++++++++
 3 files changed, 35 insertions(+), 2 deletions(-)

diff --git a/AK/Utf16View.h b/AK/Utf16View.h
index 715a66582e3..1b78520c644 100644
--- a/AK/Utf16View.h
+++ b/AK/Utf16View.h
@@ -20,6 +20,7 @@
 #include <AK/Traits.h>
 #include <AK/Types.h>
 #include <AK/UnicodeUtils.h>
+#include <AK/Utf8View.h>
 #include <AK/Vector.h>
 
 namespace AK {
@@ -232,7 +233,21 @@ public:
     {
         if (has_ascii_storage())
             return StringView { m_string.ascii, length_in_code_units() } == other;
-        return *this == Utf16View { other.characters_without_null_termination(), other.length() };
+
+        if (other.is_ascii())
+            return *this == Utf16View { other.characters_without_null_termination(), other.length() };
+
+        Utf8View other_utf8 { other };
+
+        auto this_it = begin();
+        auto other_it = other_utf8.begin();
+
+        for (; this_it != end() && other_it != other_utf8.end(); ++this_it, ++other_it) {
+            if (*this_it != *other_it)
+                return false;
+        }
+
+        return this_it == end() && other_it == other_utf8.end();
     }
 
     [[nodiscard]] constexpr bool equals_ignoring_case(Utf16View const& other) const
diff --git a/Tests/AK/TestUtf16String.cpp b/Tests/AK/TestUtf16String.cpp
index c5b5ba0e69b..1fee98ae332 100644
--- a/Tests/AK/TestUtf16String.cpp
+++ b/Tests/AK/TestUtf16String.cpp
@@ -1070,7 +1070,7 @@ TEST_CASE(equals_ascii)
     test("ababababab"sv, "😀😀😀😀😀"_utf16);
 
     // Non-ASCII string comparison.
-    EXPECT_NE("😀"sv, "😀"_utf16);
+    EXPECT_EQ("😀"sv, "😀"_utf16);
 }
 
 TEST_CASE(equals_ignoring_ascii_case)
diff --git a/Tests/AK/TestUtf16View.cpp b/Tests/AK/TestUtf16View.cpp
index 00137a8ea49..fe1803259f8 100644
--- a/Tests/AK/TestUtf16View.cpp
+++ b/Tests/AK/TestUtf16View.cpp
@@ -442,6 +442,24 @@ TEST_CASE(to_ascii_titlecase)
     EXPECT_EQ(u"foo 😀 bar"sv.to_ascii_titlecase(), u"Foo 😀 Bar"sv);
 }
 
+TEST_CASE(equals_utf8)
+{
+    EXPECT_EQ(u""sv, ""sv);
+
+    EXPECT_EQ(u"foo bar"sv, "foo bar"sv);
+    EXPECT_NE(u"foo bar"sv, "foo ba"sv);
+    EXPECT_NE(u"foo bar"sv, "foo"sv);
+    EXPECT_NE(u"foo bar"sv, ""sv);
+
+    EXPECT_EQ(u"foo 😀 bar"sv, "foo 😀 bar"sv);
+    EXPECT_NE(u"foo 😀 bar"sv, "foo 😀"sv);
+    EXPECT_NE(u"foo 😀 bar"sv, "foo"sv);
+    EXPECT_NE(u"foo 😀 bar"sv, ""sv);
+
+    EXPECT_NE(u"foo 😀 bar"sv, "foo 😂 bar"sv);
+    EXPECT_NE(u"foo 😂 bar"sv, "foo 😀 bar"sv);
+}
+
 TEST_CASE(equals_ignoring_case)
 {
     auto string1 = Utf16String::from_utf8("foobar"sv);