diff options
| author | Neil <nyamatongwe@gmail.com> | 2025-02-04 11:47:48 +1100 |
|---|---|---|
| committer | Neil <nyamatongwe@gmail.com> | 2025-02-04 11:47:48 +1100 |
| commit | 4c9ddc3121d0488914858ee511028520b96fd0e9 (patch) | |
| tree | 03989eae1ce94f479749ef74e1e6c76c10f3e332 /src/UniConversion.h | |
| parent | ef961772c3ced424f034c2055263d7231eccee01 (diff) | |
| download | scintilla-mirror-4c9ddc3121d0488914858ee511028520b96fd0e9.tar.gz | |
Fix segmentation of long lexemes to avoid breaking before modifiers like accents
that must be drawn with their base letters.
This is only a subset of implementing grapheme cluster boundaries but it
improves behaviour with some Asian scripts like Thai and Javanese.
Javanese is mostly written with (ASCII) Roman characters so issues will be rare
but Thai uses Thai script.
Also slightly improves placement of combining accents in European texts.
https://github.com/notepad-plus-plus/notepad-plus-plus/issues/14822
https://github.com/notepad-plus-plus/notepad-plus-plus/issues/16115
Diffstat (limited to 'src/UniConversion.h')
| -rw-r--r-- | src/UniConversion.h | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/src/UniConversion.h b/src/UniConversion.h index 7a51b2d08..5990cca8c 100644 --- a/src/UniConversion.h +++ b/src/UniConversion.h @@ -49,6 +49,10 @@ constexpr bool UTF8IsTrailByte(unsigned char ch) noexcept { return (ch >= 0x80) && (ch < 0xc0); } +constexpr bool UTF8IsFirstByte(unsigned char ch) noexcept { + return (ch >= 0xc2) && (ch <= 0xf4); +} + constexpr bool UTF8IsAscii(unsigned char ch) noexcept { return ch < 0x80; } |