AK+Everywhere: Change StringView case conversions to return String

There's a bit of a UTF-8 assumption with this change. But nearly every
caller of these methods were immediately creating a String from the
resulting ByteString anyways.

AK/StringBase.h

@@ -73,6 +73,12 @@ public:
     [[nodiscard]] ALWAYS_INLINE FlatPtr raw(Badge<FlyString>) const { return bit_cast<FlatPtr>(m_impl); }
     [[nodiscard]] ALWAYS_INLINE FlatPtr raw(Badge<String>) const { return bit_cast<FlatPtr>(m_impl); }
 
+    template<typename Func>
+    ALWAYS_INLINE ErrorOr<void> replace_with_new_string(Badge<StringView>, size_t byte_count, Func&& callback)
+    {
+        return replace_with_new_string(byte_count, forward<Func>(callback));
+    }
+
 protected:
     template<typename Func>
     ErrorOr<void> replace_with_new_string(size_t byte_count, Func&& callback)

AK/StringImpl.cpp

@@ -75,28 +75,6 @@ NonnullRefPtr<StringImpl const> StringImpl::create(ReadonlyBytes bytes, ShouldCh
     return StringImpl::create(reinterpret_cast<char const*>(bytes.data()), bytes.size(), shouldChomp);
 }
 
-NonnullRefPtr<StringImpl const> StringImpl::create_lowercased(char const* cstring, size_t length)
-{
-    if (!length)
-        return the_empty_stringimpl();
-    char* buffer;
-    auto impl = create_uninitialized(length, buffer);
-    for (size_t i = 0; i < length; ++i)
-        buffer[i] = (char)to_ascii_lowercase(cstring[i]);
-    return impl;
-}
-
-NonnullRefPtr<StringImpl const> StringImpl::create_uppercased(char const* cstring, size_t length)
-{
-    if (!length)
-        return the_empty_stringimpl();
-    char* buffer;
-    auto impl = create_uninitialized(length, buffer);
-    for (size_t i = 0; i < length; ++i)
-        buffer[i] = (char)to_ascii_uppercase(cstring[i]);
-    return impl;
-}
-
 unsigned StringImpl::case_insensitive_hash() const
 {
     return case_insensitive_string_hash(characters(), length());

AK/StringImpl.h

@@ -28,8 +28,6 @@ public:
     static NonnullRefPtr<StringImpl const> create(char const* cstring, ShouldChomp = NoChomp);
     static NonnullRefPtr<StringImpl const> create(char const* cstring, size_t length, ShouldChomp = NoChomp);
     static NonnullRefPtr<StringImpl const> create(ReadonlyBytes, ShouldChomp = NoChomp);
-    static NonnullRefPtr<StringImpl const> create_lowercased(char const* cstring, size_t length);
-    static NonnullRefPtr<StringImpl const> create_uppercased(char const* cstring, size_t length);
 
     void operator delete(void* ptr)
     {

AK/StringUtils.cpp

@@ -488,7 +488,7 @@ ByteString to_snakecase(StringView str)
     return builder.to_byte_string();
 }
 
-ByteString to_titlecase(StringView str)
+String to_titlecase(StringView str)
 {
     StringBuilder builder;
     bool next_is_upper = true;
@@ -501,7 +501,7 @@ ByteString to_titlecase(StringView str)
         next_is_upper = ch == ' ';
     }
 
-    return builder.to_byte_string();
+    return MUST(builder.to_string());
 }
 
 // Finishes the replacing algorithm once it is known that ita least one

AK/StringUtils.h

@@ -105,7 +105,7 @@ enum class SearchDirection {
 Optional<size_t> find_any_of(StringView haystack, StringView needles, SearchDirection);
 
 ByteString to_snakecase(StringView);
-ByteString to_titlecase(StringView);
+String to_titlecase(StringView);
 
 ByteString replace(StringView, StringView needle, StringView replacement, ReplaceMode);
 ErrorOr<String> replace(String const&, StringView needle, StringView replacement, ReplaceMode);

AK/StringView.cpp

@@ -7,12 +7,14 @@
 #include <AK/AnyOf.h>
 #include <AK/ByteBuffer.h>
 #include <AK/ByteString.h>
+#include <AK/Enumerate.h>
 #include <AK/Find.h>
 #include <AK/FlyString.h>
 #include <AK/Function.h>
 #include <AK/String.h>
 #include <AK/StringBuilder.h>
 #include <AK/StringView.h>
+#include <AK/Utf8View.h>
 #include <AK/Vector.h>
 
 #include <simdutf.h>
@@ -204,17 +206,37 @@ bool StringView::is_ascii() const
     return simdutf::validate_ascii(characters_without_null_termination(), length());
 }
 
-ByteString StringView::to_lowercase_string() const
+String StringView::to_ascii_lowercase_string() const
 {
-    return StringImpl::create_lowercased(characters_without_null_termination(), length()).release_nonnull();
+    VERIFY(Utf8View { *this }.validate());
+
+    String result;
+
+    MUST(result.replace_with_new_string({}, length(), [&](Bytes buffer) -> ErrorOr<void> {
+        for (auto [i, character] : enumerate(bytes()))
+            buffer[i] = static_cast<u8>(AK::to_ascii_lowercase(character));
+        return {};
+    }));
+
+    return result;
 }
 
-ByteString StringView::to_uppercase_string() const
+String StringView::to_ascii_uppercase_string() const
 {
-    return StringImpl::create_uppercased(characters_without_null_termination(), length()).release_nonnull();
+    VERIFY(Utf8View { *this }.validate());
+
+    String result;
+
+    MUST(result.replace_with_new_string({}, length(), [&](Bytes buffer) -> ErrorOr<void> {
+        for (auto [i, character] : enumerate(bytes()))
+            buffer[i] = static_cast<u8>(AK::to_ascii_uppercase(character));
+        return {};
+    }));
+
+    return result;
 }
 
-ByteString StringView::to_titlecase_string() const
+String StringView::to_ascii_titlecase_string() const
 {
     return StringUtils::to_titlecase(*this);
 }

AK/StringView.h

@@ -105,9 +105,9 @@ public:
     [[nodiscard]] StringView trim(StringView characters, TrimMode mode = TrimMode::Both) const { return StringUtils::trim(*this, characters, mode); }
     [[nodiscard]] StringView trim_whitespace(TrimMode mode = TrimMode::Both) const { return StringUtils::trim_whitespace(*this, mode); }
 
-    [[nodiscard]] ByteString to_lowercase_string() const;
-    [[nodiscard]] ByteString to_uppercase_string() const;
-    [[nodiscard]] ByteString to_titlecase_string() const;
+    [[nodiscard]] String to_ascii_lowercase_string() const;
+    [[nodiscard]] String to_ascii_uppercase_string() const;
+    [[nodiscard]] String to_ascii_titlecase_string() const;
 
     [[nodiscard]] Optional<size_t> find(char needle, size_t start = 0) const
     {

Libraries/LibIDL/IDLParser.cpp

@@ -89,7 +89,7 @@ static ByteString convert_enumeration_value_to_cpp_enum_member(ByteString const&
 
         auto word = lexer.consume_while([](auto c) { return is_ascii_alphanumeric(c); });
         if (!word.is_empty()) {
-            builder.append(word.to_titlecase_string());
+            builder.append(word.to_ascii_titlecase_string());
         } else {
             auto non_alnum_string = lexer.consume_while([](auto c) { return !is_ascii_alphanumeric(c); });
             if (!non_alnum_string.is_empty())

Libraries/LibJS/Runtime/Intl/DisplayNames.cpp

@@ -103,7 +103,7 @@ ThrowCompletionOr<Value> canonical_code_for_display_names(VM& vm, DisplayNames::
             return vm.throw_completion<RangeError>(ErrorType::OptionIsNotValidValue, code, "region"sv);
 
         // b. Return the ASCII-uppercase of code.
-        return PrimitiveString::create(vm, code.to_uppercase_string());
+        return PrimitiveString::create(vm, code.to_ascii_uppercase_string());
     }
 
     // 3. If type is "script", then
@@ -119,7 +119,7 @@ ThrowCompletionOr<Value> canonical_code_for_display_names(VM& vm, DisplayNames::
         // c. Let first be the ASCII-uppercase of the substring of code from 0 to 1.
         // d. Let rest be the ASCII-lowercase of the substring of code from 1.
         // e. Return the string-concatenation of first and rest.
-        return PrimitiveString::create(vm, code.to_titlecase_string());
+        return PrimitiveString::create(vm, code.to_ascii_titlecase_string());
     }
 
     // 4. If type is "calendar", then
@@ -133,7 +133,7 @@ ThrowCompletionOr<Value> canonical_code_for_display_names(VM& vm, DisplayNames::
             return vm.throw_completion<RangeError>(ErrorType::OptionIsNotValidValue, code, "calendar"sv);
 
         // c. Return the ASCII-lowercase of code.
-        return PrimitiveString::create(vm, code.to_lowercase_string());
+        return PrimitiveString::create(vm, code.to_ascii_lowercase_string());
     }
 
     // 5. If type is "dateTimeField", then
@@ -154,7 +154,7 @@ ThrowCompletionOr<Value> canonical_code_for_display_names(VM& vm, DisplayNames::
         return vm.throw_completion<RangeError>(ErrorType::OptionIsNotValidValue, code, "currency"sv);
 
     // 8. Return the ASCII-uppercase of code.
-    return PrimitiveString::create(vm, code.to_uppercase_string());
+    return PrimitiveString::create(vm, code.to_ascii_uppercase_string());
 }
 
 // 12.5.2 IsValidDateTimeFieldCode ( field ), https://tc39.es/ecma402/#sec-isvaliddatetimefieldcode

Libraries/LibURL/Parser.cpp

@@ -527,10 +527,8 @@ static ErrorOr<String> domain_to_ascii(StringView domain, bool be_strict)
             }
         }
 
-        if (!slow_path) {
-            auto lowercase_domain = domain.to_lowercase_string();
-            return String::from_utf8_without_validation(lowercase_domain.bytes());
-        }
+        if (!slow_path)
+            return domain.to_ascii_lowercase_string();
     }
 
     Unicode::IDNA::ToAsciiOptions const options {

Libraries/LibWeb/DOM/DOMTokenList.cpp

@@ -227,7 +227,7 @@ WebIDL::ExceptionOr<bool> DOMTokenList::supports(StringView token)
         return WebIDL::SimpleException { WebIDL::SimpleExceptionType::TypeError, MUST(String::formatted("Attribute {} does not define any supported tokens", m_associated_attribute)) };
 
     // 2. Let lowercase token be a copy of token, in ASCII lowercase.
-    auto lowercase_token = token.to_lowercase_string();
+    auto lowercase_token = token.to_ascii_lowercase_string();
 
     // 3. If lowercase token is present in supported tokens, return true.
     if (supported_tokens->contains_slow(lowercase_token))

Libraries/LibWeb/DOM/Document.cpp

@@ -2072,17 +2072,16 @@ HTML::EnvironmentSettingsObject& Document::relevant_settings_object() const
 }
 
 // https://dom.spec.whatwg.org/#dom-document-createelement
-WebIDL::ExceptionOr<GC::Ref<Element>> Document::create_element(String const& a_local_name, Variant<String, ElementCreationOptions> const& options)
+WebIDL::ExceptionOr<GC::Ref<Element>> Document::create_element(String const& local_name, Variant<String, ElementCreationOptions> const& options)
 {
-    auto local_name = a_local_name.to_byte_string();
-
     // 1. If localName does not match the Name production, then throw an "InvalidCharacterError" DOMException.
-    if (!is_valid_name(a_local_name))
+    if (!is_valid_name(local_name))
         return WebIDL::InvalidCharacterError::create(realm(), "Invalid character in tag name."_string);
 
     // 2. If this is an HTML document, then set localName to localName in ASCII lowercase.
-    if (document_type() == Type::HTML)
-        local_name = local_name.to_lowercase();
+    auto local_name_lower = document_type() == Type::HTML
+        ? local_name.to_ascii_lowercase()
+        : local_name;
 
     // 3. Let is be null.
     Optional<String> is_value;
@@ -2100,7 +2099,7 @@ WebIDL::ExceptionOr<GC::Ref<Element>> Document::create_element(String const& a_l
         namespace_ = Namespace::HTML;
 
     // 6. Return the result of creating an element given this, localName, namespace, null, is, and with the synchronous custom elements flag set.
-    return TRY(DOM::create_element(*this, FlyString::from_utf8_without_validation(local_name.bytes()), move(namespace_), {}, move(is_value), true));
+    return TRY(DOM::create_element(*this, FlyString::from_utf8_without_validation(local_name_lower.bytes()), move(namespace_), {}, move(is_value), true));
 }
 
 // https://dom.spec.whatwg.org/#dom-document-createelementns

Libraries/LibWeb/Encoding/TextDecoder.cpp

@@ -32,7 +32,7 @@ WebIDL::ExceptionOr<GC::Ref<TextDecoder>> TextDecoder::construct_impl(JS::Realm&
     // 3. Set this’s encoding to encoding.
     // https://encoding.spec.whatwg.org/#dom-textdecoder-encoding
     // The encoding getter steps are to return this’s encoding’s name, ASCII lowercased.
-    auto lowercase_encoding_name = MUST(String::from_byte_string(encoding.value().to_lowercase_string()));
+    auto lowercase_encoding_name = encoding.value().to_ascii_lowercase_string();
 
     // 4. If options["fatal"] is true, then set this’s error mode to "fatal".
     auto fatal = options.value_or({}).fatal;

Libraries/LibWeb/HTML/TokenizedFeatures.cpp

@@ -68,7 +68,7 @@ TokenizedFeature::Map tokenize_open_features(StringView features)
         lexer.ignore_while(is_feature_separator);
 
         // 4. Collect a sequence of code points that are not feature separators from features given position. Set name to the collected characters, converted to ASCII lowercase.
-        name = MUST(String::from_byte_string(lexer.consume_until(is_feature_separator).to_lowercase_string()));
+        name = lexer.consume_until(is_feature_separator).to_ascii_lowercase_string();
 
         // 5. Set name to the result of normalizing the feature name name.
         name = normalize_feature_name(name);
@@ -85,7 +85,7 @@ TokenizedFeature::Map tokenize_open_features(StringView features)
         lexer.ignore_while([](auto character) { return Infra::is_ascii_whitespace(character) || character == '='; });
 
         // 2. Collect a sequence of code points that are not feature separators code points from features given position. Set value to the collected code points, converted to ASCII lowercase.
-        value = MUST(String::from_byte_string(lexer.consume_until(is_feature_separator).to_lowercase_string()));
+        value = lexer.consume_until(is_feature_separator).to_ascii_lowercase_string();
 
         // 8. If name is not the empty string, then set tokenizedFeatures[name] to value.
         if (!name.is_empty())

Meta/Lagom/Tools/CodeGenerators/LibWeb/GenerateWebGLRenderingContext.cpp

@@ -288,7 +288,7 @@ static void generate_get_parameter(SourceGenerator& generator, int webgl_version
         return JS::@type_name@::create(m_realm, @element_count@, array_buffer);
 )~~~");
         } else if (type_name == "WebGLProgram"sv || type_name == "WebGLBuffer"sv || type_name == "WebGLTexture"sv || type_name == "WebGLFramebuffer"sv || type_name == "WebGLRenderbuffer"sv) {
-            impl_generator.set("stored_name", name_and_type.name.to_lowercase_string());
+            impl_generator.set("stored_name", name_and_type.name.to_ascii_lowercase_string());
             impl_generator.append(R"~~~(
         if (!m_@stored_name@)
             return JS::js_null();