1 files changed, 77 insertions, 0 deletions
diff --git a/src/UniConversion.cxx b/src/UniConversion.cxx
new file mode 100644
index 000000000..9306f307c
--- /dev/null
+++ b/src/UniConversion.cxx
@@ -0,0 +1,77 @@
+// UniConversion.h - functions to handle UFT-8 and UCS-2 strings
+// Copyright 1998-2000 by Neil Hodgson <neilh@scintilla.org>
+// The License.txt file describes the conditions under which this software may be distributed.
+
+#include <stdlib.h>
+
+#include "UniConversion.h"
+
+unsigned int UTF8Length(const wchar_t *uptr, unsigned int tlen) {
+	unsigned int len = 0;
+	for (unsigned int i = 0; i < tlen && uptr[i]; i++) {
+		unsigned int uch = uptr[i];
+		if (uch < 0x80)
+			len++;
+		else if (uch < 0x800)
+			len+=2;
+		else
+			len +=3;
+	}
+	return len;
+}
+
+void UTF8FromUCS2(const wchar_t *uptr, unsigned int tlen, char *putf, unsigned int len) {
+	int k = 0;
+	for (unsigned int i = 0; i < tlen && uptr[i]; i++) {
+		unsigned int uch = uptr[i];
+		if (uch < 0x80) {
+			putf[k++] = static_cast<char>(uch);
+		} else if (uch < 0x800) {
+			putf[k++] = static_cast<char>(0xC0 | (uch >> 6));
+			putf[k++] = static_cast<char>(0x80 | (uch & 0x3f));
+		} else {
+			putf[k++] = static_cast<char>(0xE0 | (uch >> 12));
+			putf[k++] = static_cast<char>(0x80 | ((uch >> 6) & 0x3f));
+			putf[k++] = static_cast<char>(0x80 | (uch & 0x3f));
+		}
+	}
+	putf[len] = '\0';
+}
+
+unsigned int UCS2Length(const char *s, unsigned int len) {
+	unsigned int ulen = 0;
+	for (unsigned int i=0;i<len;i++) {
+		unsigned char ch = static_cast<unsigned char>(s[i]);
+		if ((ch < 0x80) || (ch > (0x80 + 0x40)))
+			ulen++;
+	}
+	return ulen;
+}
+
+unsigned int UCS2FromUTF8(const char *s, unsigned int len, wchar_t *tbuf, unsigned int tlen) {
+#ifdef USE_API
+	return ::MultiByteToWideChar(CP_UTF8, 0, s, len, tbuf, tlen);
+#else 
+	unsigned int ui=0;
+	const unsigned char *us = reinterpret_cast<const unsigned char *>(s);
+	unsigned int i=0;
+	while ((i<len) && (ui<tlen)) {
+		unsigned char ch = us[i++];
+		if (ch < 0x80) {
+			tbuf[ui] = ch;
+		} else if (ch < 0x80 + 0x40 + 0x20) {
+			tbuf[ui] = static_cast<wchar_t>((ch & 0x1F) << 6);
+			ch = us[i++];
+			tbuf[ui] = static_cast<wchar_t>(tbuf[ui] + (ch & 0x7F));
+		} else {
+			tbuf[ui] = static_cast<wchar_t>((ch & 0xF) << 12);
+			ch = us[i++];
+			tbuf[ui] = static_cast<wchar_t>(tbuf[ui] + ((ch & 0x7F) << 6));
+			ch = us[i++];
+			tbuf[ui] = static_cast<wchar_t>(tbuf[ui] + (ch & 0x7F));
+		}
+		ui++;
+	}
+	return ui;
+#endif
+}