From a9235ffc9976527a6147ca87ca371ea70582404f Mon Sep 17 00:00:00 2001
From: nyamatongwe <unknown>
Date: Tue, 23 Mar 2010 05:22:21 +0000
Subject: Added function for finding how many bytes are in a UTF-8 character.

---
 src/UniConversion.cxx | 12 ++++++++++++
 src/UniConversion.h   |  1 +
 2 files changed, 13 insertions(+)

(limited to 'src')

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);
 
-- 
cgit v1.2.3