hashirama/goldendict-ng
hashirama/goldendict-ng
0
0
Fork 0
mirror of https://github.com/xiaoyifang/goldendict-ng.git synced 2024-11-23 16:04:06 +00:00
Code Issues Actions 1 Projects Releases Wiki Activity

opt: simplify Folding::apply
Some checks are pending
SonarCloud / Build and analyze (push) Waiting to run
Details

Browse source
This commit is contained in:
shenleban tongying 2024-11-12 22:06:27 -05:00
parent 20fcea33e1
commit fbec70e41a
1 changed files with 15 additions and 33 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

42
src/common/folding.cc
View file

@ -20,41 +20,23 @@ bool isCombiningMark( wchar ch )
wstring apply( wstring const & in, bool preserveWildcards )
{
//remove space and accent;
auto withPunc = QString::fromStdU32String( in )
// remove diacritics (normalization), white space, punt,
auto temp = QString::fromStdU32String( in )
.normalized( QString::NormalizationForm_KD )
.remove( RX::markSpace )
.removeIf( [ preserveWildcards ]( const QChar & ch ) -> bool {
return ch.isPunct()
&& !( preserveWildcards && ( ch == '\\' || ch == '?' || ch == '*' || ch == '[' || ch == ']' ) );
} )
.toStdU32String();
//First, strip diacritics and apply ws/punctuation removal
wstring withoutDiacritics;
withoutDiacritics.reserve( withPunc.size() );
for ( auto const & ch : withPunc ) {
if ( !isPunct( ch )
|| ( preserveWildcards && ( ch == '\\' || ch == '?' || ch == '*' || ch == '[' || ch == ']' ) ) ) {
withoutDiacritics.push_back( ch );
}
}
// Now, fold the case
wstring caseFolded;
caseFolded.reserve( withoutDiacritics.size() * foldCaseMaxOut );
wchar const * nextChar = withoutDiacritics.data();
// case folding
std::u32string caseFolded;
caseFolded.reserve( temp.size() );
wchar buf[ foldCaseMaxOut ];
for ( size_t left = withoutDiacritics.size(); left--; ) {
caseFolded.append( buf, foldCase( *nextChar++, buf ) );
for ( const char32_t ch : temp ) {
auto n = foldCase( ch, buf );
caseFolded.append( buf, n );
}
return caseFolded;
}