diff options
| author | nyamatongwe <unknown> | 2010-03-23 05:22:21 +0000 |
|---|---|---|
| committer | nyamatongwe <unknown> | 2010-03-23 05:22:21 +0000 |
| commit | a9235ffc9976527a6147ca87ca371ea70582404f (patch) | |
| tree | 31b1e6454ad4f4e191bcffa08aef846825b4b2c2 /src | |
| parent | 187818f68f76f51a38c83a212ebce1fc8072496c (diff) | |
| download | scintilla-mirror-a9235ffc9976527a6147ca87ca371ea70582404f.tar.gz | |
Added function for finding how many bytes are in a UTF-8 character.
Diffstat (limited to 'src')
| -rw-r--r-- | src/UniConversion.cxx | 12 | ||||
| -rw-r--r-- | src/UniConversion.h | 1 |
2 files changed, 13 insertions, 0 deletions
diff --git a/src/UniConversion.cxx b/src/UniConversion.cxx index 482a0a581..0064e3cba 100644 --- a/src/UniConversion.cxx +++ b/src/UniConversion.cxx @@ -61,6 +61,18 @@ void UTF8FromUTF16(const wchar_t *uptr, unsigned int tlen, char *putf, unsigned putf[len] = '\0'; } +unsigned int UTF8CharLength(unsigned char ch) { + if (ch < 0x80) { + return 1; + } else if (ch < 0x80 + 0x40 + 0x20) { + return 2; + } else if (ch < 0x80 + 0x40 + 0x20 + 0x10) { + return 3; + } else { + return 4; + } +} + unsigned int UTF16Length(const char *s, unsigned int len) { unsigned int ulen = 0; unsigned int charLen; diff --git a/src/UniConversion.h b/src/UniConversion.h index fd420a688..8cc3d0a18 100644 --- a/src/UniConversion.h +++ b/src/UniConversion.h @@ -7,6 +7,7 @@ unsigned int UTF8Length(const wchar_t *uptr, unsigned int tlen); void UTF8FromUTF16(const wchar_t *uptr, unsigned int tlen, char *putf, unsigned int len); +unsigned int UTF8CharLength(unsigned char ch); unsigned int UTF16Length(const char *s, unsigned int len); unsigned int UTF16FromUTF8(const char *s, unsigned int len, wchar_t *tbuf, unsigned int tlen); |