LibUnicode: Replace case transformations and comparison with ICUs

There are a couple of differences here due to using ICU:

1. Titlecasing behaves slightly differently. We previously transformed
   "123dollars" to "123Dollars", as we would use word segmentation to
   split a string into words, then transform the first cased character
   to titlecase. ICU doesn't go quite that far, and leaves the string
   as "123dollars". While this is a behavior change, the only user of
   this API is the `text-transform: capitalize;` CSS rule, and we now
   match the behavior of other browsers.

2. There isn't an API to compare strings with case insensitivity without
   allocating case-folded strings for both the left- and right-hand-side
   strings. Our implementation was previously allocation-free; however,
   in a benchmark, ICU is still ~1.4x faster.
This commit is contained in:
Timothy Flynn 2024-06-19 16:39:30 -04:00 committed by Andreas Kling
commit 5cf818e305
Notes: sideshowbarker 2024-07-16 23:17:55 +09:00
12 changed files with 111 additions and 1092 deletions

View file

@ -1,57 +1,141 @@
/*
* Copyright (c) 2023, Tim Flynn <trflynn89@serenityos.org>
* Copyright (c) 2023-2024, Tim Flynn <trflynn89@serenityos.org>
*
* SPDX-License-Identifier: BSD-2-Clause
*/
#define AK_DONT_REPLACE_STD
#include <AK/String.h>
#include <AK/StringBuilder.h>
#include <AK/Utf8View.h>
#include <LibUnicode/CharacterTypes.h>
#include <LibUnicode/UnicodeUtils.h>
#include <LibLocale/ICU.h>
#include <unicode/bytestream.h>
#include <unicode/casemap.h>
#include <unicode/stringoptions.h>
// This file contains definitions of AK::String methods which require UCD data.
namespace AK {
struct ResolvedLocale {
ByteString buffer;
char const* locale { nullptr };
};
static ResolvedLocale resolve_locale(Optional<StringView> const& locale)
{
if (!locale.has_value())
return {};
ResolvedLocale resolved_locale;
resolved_locale.buffer = *locale;
resolved_locale.locale = resolved_locale.buffer.characters();
return resolved_locale;
}
ErrorOr<String> String::to_lowercase(Optional<StringView> const& locale) const
{
StringBuilder builder;
TRY(Unicode::Detail::build_lowercase_string(code_points(), builder, locale));
UErrorCode status = U_ZERO_ERROR;
StringBuilder builder { bytes_as_string_view().length() };
icu::StringByteSink sink { &builder };
auto resolved_locale = resolve_locale(locale);
icu::CaseMap::utf8ToLower(resolved_locale.locale, 0, Locale::icu_string_piece(*this), sink, nullptr, status);
if (Locale::icu_failure(status))
return Error::from_string_literal("Unable to convert string to lowercase");
return builder.to_string_without_validation();
}
ErrorOr<String> String::to_uppercase(Optional<StringView> const& locale) const
{
StringBuilder builder;
TRY(Unicode::Detail::build_uppercase_string(code_points(), builder, locale));
UErrorCode status = U_ZERO_ERROR;
StringBuilder builder { bytes_as_string_view().length() };
icu::StringByteSink sink { &builder };
auto resolved_locale = resolve_locale(locale);
icu::CaseMap::utf8ToUpper(resolved_locale.locale, 0, Locale::icu_string_piece(*this), sink, nullptr, status);
if (Locale::icu_failure(status))
return Error::from_string_literal("Unable to convert string to uppercase");
return builder.to_string_without_validation();
}
ErrorOr<String> String::to_titlecase(Optional<StringView> const& locale, TrailingCodePointTransformation trailing_code_point_transformation) const
{
StringBuilder builder;
TRY(Unicode::Detail::build_titlecase_string(code_points(), builder, locale, trailing_code_point_transformation));
UErrorCode status = U_ZERO_ERROR;
StringBuilder builder { bytes_as_string_view().length() };
icu::StringByteSink sink { &builder };
auto resolved_locale = resolve_locale(locale);
u32 options = 0;
if (trailing_code_point_transformation == TrailingCodePointTransformation::PreserveExisting)
options |= U_TITLECASE_NO_LOWERCASE;
icu::CaseMap::utf8ToTitle(resolved_locale.locale, options, nullptr, Locale::icu_string_piece(*this), sink, nullptr, status);
if (Locale::icu_failure(status))
return Error::from_string_literal("Unable to convert string to titlecase");
return builder.to_string_without_validation();
}
static ErrorOr<void> build_casefold_string(StringView string, StringBuilder& builder)
{
UErrorCode status = U_ZERO_ERROR;
icu::StringByteSink sink { &builder };
icu::CaseMap::utf8Fold(0, Locale::icu_string_piece(string), sink, nullptr, status);
if (Locale::icu_failure(status))
return Error::from_string_literal("Unable to casefold string");
return {};
}
ErrorOr<String> String::to_casefold() const
{
StringBuilder builder;
TRY(Unicode::Detail::build_casefold_string(code_points(), builder));
StringBuilder builder { bytes_as_string_view().length() };
TRY(build_casefold_string(*this, builder));
return builder.to_string_without_validation();
}
bool String::equals_ignoring_case(String const& other) const
{
return Unicode::equals_ignoring_case(code_points(), other.code_points());
StringBuilder lhs_builder { bytes_as_string_view().length() };
if (build_casefold_string(*this, lhs_builder).is_error())
return false;
StringBuilder rhs_builder { other.bytes_as_string_view().length() };
if (build_casefold_string(other, rhs_builder).is_error())
return false;
return lhs_builder.string_view() == rhs_builder.string_view();
}
Optional<size_t> String::find_byte_offset_ignoring_case(StringView needle, size_t from_byte_offset) const
{
auto haystack = code_points().substring_view(from_byte_offset);
auto haystack = bytes_as_string_view().substring_view(from_byte_offset);
if (haystack.is_empty())
return {};
if (auto index = Unicode::find_ignoring_case(haystack, Utf8View { needle }); index.has_value())
StringBuilder lhs_builder { haystack.length() };
if (build_casefold_string(haystack, lhs_builder).is_error())
return {};
StringBuilder rhs_builder { needle.length() };
if (build_casefold_string(needle, rhs_builder).is_error())
return false;
if (auto index = lhs_builder.string_view().find(rhs_builder.string_view()); index.has_value())
return *index + from_byte_offset;
return {};