RGBCube/serenity
1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-05-31 07:48:11 +00:00
Code Issues Activity Actions

LibRegex: Support property escapes of Unicode General Categories

Browse source 
This changes LibRegex to parse the property escape as a Variant of
Unicode Property & General Category values. A byte code instruction is
added to perform matching based on General Category values.
This commit is contained in:
Timothy Flynn 2021-07-31 17:46:05 -04:00 • committed by Ali Mohammad Pur
parent 5de6d3dd90
commit 1e10d6d7ce
5 changed files with 77 additions and 19 deletions
Download patch file Download diff file Expand all files Collapse all files

20
Userland/Libraries/LibRegex/RegexByteCode.cpp
View file

@ -537,6 +537,10 @@ ALWAYS_INLINE ExecutionResult OpCode_Compare::execute(MatchInput const& input, M
auto property = static_cast<Unicode::Property>(m_bytecode->at(offset++));
compare_property(input, state, property, current_inversion_state(), inverse_matched);
} else if (compare_type == CharacterCompareType::GeneralCategory) {
auto general_category = static_cast<Unicode::GeneralCategory>(m_bytecode->at(offset++));
compare_general_category(input, state, general_category, current_inversion_state(), inverse_matched);
} else {
warnln("Undefined comparison: {}", (int)compare_type);
VERIFY_NOT_REACHED();
@ -742,6 +746,22 @@ ALWAYS_INLINE void OpCode_Compare::compare_property(MatchInput const& input, Mat
}
}
ALWAYS_INLINE void OpCode_Compare::compare_general_category(MatchInput const& input, MatchState& state, Unicode::GeneralCategory general_category, 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_general_category(code_point, general_category);
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());