From d19b31529f28e88aa691a6a9fbd0215c54c6e81c Mon Sep 17 00:00:00 2001
From: Timothy Flynn <trflynn89@pm.me>
Date: Thu, 19 Sep 2024 14:25:38 -0400
Subject: [PATCH] AK+Meta: Update simdutf to version 5.5.0

Contains many fixes found upstream by fuzzers. Also includes fixes for
CPU-specific inconsistencies with null inputs.
---
 AK/Utf16View.cpp | 18 ------------------
 AK/Utf8View.cpp  | 10 ----------
 vcpkg.json       |  2 +-
 3 files changed, 1 insertion(+), 29 deletions(-)

diff --git a/AK/Utf16View.cpp b/AK/Utf16View.cpp
index 1276a37363c..449a1b0b88f 100644
--- a/AK/Utf16View.cpp
+++ b/AK/Utf16View.cpp
@@ -131,10 +131,6 @@ ErrorOr<void> code_point_to_utf16(Utf16Data& string, u32 code_point, Endianness
 
 size_t utf16_code_unit_length_from_utf8(StringView string)
 {
-    // FIXME: The CPU-specific implementations behave differently on null inputs. We treat null views as an empty string.
-    if (string.is_empty())
-        return 0;
-
     return simdutf::utf16_length_from_utf8(string.characters_without_null_termination(), string.length());
 }
 
@@ -300,10 +296,6 @@ bool Utf16View::starts_with(Utf16View const& needle) const
 
 bool Utf16View::validate() const
 {
-    // FIXME: The CPU-specific implementations behave differently on null inputs. We treat null views as an empty string.
-    if (is_empty())
-        return true;
-
     switch (m_endianness) {
     case Endianness::Host:
         return simdutf::validate_utf16(char_data(), length_in_code_units());
@@ -317,12 +309,6 @@ bool Utf16View::validate() const
 
 bool Utf16View::validate(size_t& valid_code_units) const
 {
-    // FIXME: The CPU-specific implementations behave differently on null inputs. We treat null views as an empty string.
-    if (is_empty()) {
-        valid_code_units = 0;
-        return true;
-    }
-
     auto result = [&]() {
         switch (m_endianness) {
         case Endianness::Host:
@@ -341,10 +327,6 @@ bool Utf16View::validate(size_t& valid_code_units) const
 
 size_t Utf16View::calculate_length_in_code_points() const
 {
-    // FIXME: The CPU-specific implementations behave differently on null inputs. We treat null views as an empty string.
-    if (is_empty())
-        return 0;
-
     // FIXME: simdutf's code point length method assumes valid UTF-16, whereas Utf16View uses U+FFFD as a replacement
     //        for invalid code points. If we change Utf16View to only accept valid encodings as an invariant, we can
     //        remove this branch.
diff --git a/AK/Utf8View.cpp b/AK/Utf8View.cpp
index 71f71618be2..76d2e3a04c8 100644
--- a/AK/Utf8View.cpp
+++ b/AK/Utf8View.cpp
@@ -74,10 +74,6 @@ Utf8View Utf8View::unicode_substring_view(size_t code_point_offset, size_t code_
 
 size_t Utf8View::calculate_length() const
 {
-    // FIXME: The CPU-specific implementations behave differently on null inputs. We treat null views as an empty string.
-    if (is_empty())
-        return 0;
-
     // FIXME: simdutf's code point length method assumes valid UTF-8, whereas Utf8View uses U+FFFD as a replacement
     //        for invalid code points. If we change Utf8View to only accept valid encodings as an invariant, we can
     //        remove this branch.
@@ -157,12 +153,6 @@ Utf8View Utf8View::trim(Utf8View const& characters, TrimMode mode) const
 
 bool Utf8View::validate(size_t& valid_bytes, AllowSurrogates allow_surrogates) const
 {
-    // FIXME: The CPU-specific implementations behave differently on null inputs. We treat null views as an empty string.
-    if (is_empty()) {
-        valid_bytes = 0;
-        return true;
-    }
-
     auto result = simdutf::validate_utf8_with_errors(m_string.characters_without_null_termination(), m_string.length());
     valid_bytes = result.count;
 
diff --git a/vcpkg.json b/vcpkg.json
index 243c92d80b3..849216a0020 100644
--- a/vcpkg.json
+++ b/vcpkg.json
@@ -102,7 +102,7 @@
     },
     {
       "name": "simdutf",
-      "version": "5.3.4#0"
+      "version": "5.5.0#0"
     },
     {
       "name": "skia",