RGBCube/serenity
1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-05-20 13:55:08 +00:00
Code Issues Activity Actions

LibRegex+LibUnicode: Begin implementing Unicode property escapes

Browse source 
This supports some binary property matching. It does not support any
properties not yet parsed by LibUnicode, nor does it support value
matching (such as Script_Extensions=Latin).
This commit is contained in:
Timothy Flynn 2021-07-29 14:18:51 -04:00 • committed by Linus Groh
parent f1dd770a8a
commit d485cf29d7
11 changed files with 230 additions and 33 deletions
Download patch file Download diff file Expand all files Collapse all files

21
Userland/Libraries/LibRegex/RegexByteCode.cpp
View file

@ -9,6 +9,7 @@
#include "RegexDebug.h"
#include <AK/CharacterTypes.h>
#include <AK/Debug.h>
#include <LibUnicode/CharacterTypes.h>
namespace regex {
@ -532,6 +533,10 @@ ALWAYS_INLINE ExecutionResult OpCode_Compare::execute(MatchInput const& input, M
if (!compare_string(input, state, str, had_zero_length_match))
return ExecutionResult::Failed_ExecuteLowPrioForks;
} else if (compare_type == CharacterCompareType::Property) {
auto property = static_cast<Unicode::Property>(m_bytecode->at(offset++));
compare_property(input, state, property, current_inversion_state(), inverse_matched);
} else {
warnln("Undefined comparison: {}", (int)compare_type);
VERIFY_NOT_REACHED();
@ -721,6 +726,22 @@ ALWAYS_INLINE void OpCode_Compare::compare_character_range(MatchInput const& inp
}
}
ALWAYS_INLINE void OpCode_Compare::compare_property(MatchInput const& input, MatchState& state, Unicode::Property property, bool inverse, bool& inverse_matched)
{
if (state.string_position == input.view.length())
return;
u32 code_point = input.view[state.string_position];
bool equal = Unicode::code_point_has_property(code_point, property);
if (equal) {
if (inverse)
inverse_matched = true;
else
++state.string_position;
}
}
String const OpCode_Compare::arguments_string() const
{
return String::formatted("argc={}, args={} ", arguments_count(), arguments_size());