system/serenity
system/serenity
1
0
Fork 0
mirror of https://github.com/SerenityOS/serenity synced 2024-07-09 09:57:28 +00:00
Code Issues Projects Releases Wiki Activity

AK: Make StringBuilder::try_append_code_point actually fallible

Browse Source 
It currently uses the non-fallible `append` method to append each UTF-8
encoded byte of the code point.
This commit is contained in:
Timothy Flynn 2023-01-07 14:48:20 -05:00 committed by Linus Groh
parent aee7c44064
commit 39bda0073e
2 changed files with 31 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
AK/StringBuilder.cpp
View File

@ -141,7 +141,7 @@ void StringBuilder::clear()
ErrorOr<void> StringBuilder::try_append_code_point(u32 code_point)
{
auto nwritten = AK::UnicodeUtils::code_point_to_utf8(code_point, [this](char c) { append(c); });
auto nwritten = TRY(AK::UnicodeUtils::try_code_point_to_utf8(code_point, [this](char c) { return try_append(c); }));
if (nwritten < 0) {
TRY(try_append(0xef));
TRY(try_append(0xbf));

30
AK/UnicodeUtils.h
View File

@ -6,6 +6,8 @@
#pragma once
#include <AK/Concepts.h>
#include <AK/Error.h>
#include <AK/Forward.h>
namespace AK::UnicodeUtils {
@ -35,4 +37,32 @@ template<typename Callback>
return -1;
}
template<FallibleFunction<char> Callback>
[[nodiscard]] ErrorOr<int> try_code_point_to_utf8(u32 code_point, Callback&& callback)
{
if (code_point <= 0x7f) {
TRY(callback(static_cast<char>(code_point)));
return 1;
}
if (code_point <= 0x07ff) {
TRY(callback(static_cast<char>((((code_point >> 6) & 0x1f) | 0xc0))));
TRY(callback(static_cast<char>((((code_point >> 0) & 0x3f) | 0x80))));
return 2;
}
if (code_point <= 0xffff) {
TRY(callback(static_cast<char>((((code_point >> 12) & 0x0f) | 0xe0))));
TRY(callback(static_cast<char>((((code_point >> 6) & 0x3f) | 0x80))));
TRY(callback(static_cast<char>((((code_point >> 0) & 0x3f) | 0x80))));
return 3;
}
if (code_point <= 0x10ffff) {
TRY(callback(static_cast<char>((((code_point >> 18) & 0x07) | 0xf0))));
TRY(callback(static_cast<char>((((code_point >> 12) & 0x3f) | 0x80))));
TRY(callback(static_cast<char>((((code_point >> 6) & 0x3f) | 0x80))));
TRY(callback(static_cast<char>((((code_point >> 0) & 0x3f) | 0x80))));
return 4;
}
return -1;
}
}