All Unicode planes supported, not just the Basic Multilingual Plane.

1 files changed, 55 insertions, 12 deletions

diff --git a/src/UniConversion.cxx b/src/UniConversion.cxx
index 363db90f4..863eb82cd 100644
--- a/src/UniConversion.cxx
+++ b/src/UniConversion.cxx
@@ -9,49 +9,80 @@
 
 #include "UniConversion.h"
 
+enum { SURROGATE_LEAD_FIRST = 0xD800 };
+enum { SURROGATE_TRAIL_FIRST = 0xDC00 };
+enum { SURROGATE_TRAIL_LAST = 0xDFFF };
+
 unsigned int UTF8Length(const wchar_t *uptr, unsigned int tlen) {
 	unsigned int len = 0;
-	for (unsigned int i = 0; i < tlen && uptr[i]; i++) {
+	for (unsigned int i = 0; i < tlen && uptr[i];) {
 		unsigned int uch = uptr[i];
-		if (uch < 0x80)
+		if (uch < 0x80) {
 			len++;
-		else if (uch < 0x800)
+		} else if (uch < 0x800) {
 			len += 2;
-		else
-			len +=3;
+		} else if ((uch >= SURROGATE_LEAD_FIRST) &&
+			(uch <= SURROGATE_TRAIL_LAST)) {
+			len += 4;
+			i++;
+		} else {
+			len += 3;
+		}
+		i++;
 	}
 	return len;
 }
 
-void UTF8FromUCS2(const wchar_t *uptr, unsigned int tlen, char *putf, unsigned int len) {
+void UTF8FromUTF16(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++) {
+	for (unsigned int i = 0; i < tlen && uptr[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 if ((uch >= SURROGATE_LEAD_FIRST) &&
+			(uch <= SURROGATE_TRAIL_LAST)) {
+			// Half a surrogate pair
+			i++;
+			unsigned int xch = 0x10000 + ((uch & 0x3ff) << 10) + (uptr[i] & 0x3ff);
+			putf[k++] = static_cast<char>(0xF0 | (xch >> 18));
+			putf[k++] = static_cast<char>(0x80 | (xch >> 12) & 0x3f);
+			putf[k++] = static_cast<char>(0x80 | ((xch >> 6) & 0x3f));
+			putf[k++] = static_cast<char>(0x80 | (xch & 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));
 		}
+		i++;
 	}
 	putf[len] = '\0';
 }
 
-unsigned int UCS2Length(const char *s, unsigned int len) {
+unsigned int UTF16Length(const char *s, unsigned int len) {
 	unsigned int ulen = 0;
-	for (unsigned int i=0;i<len;i++) {
+	unsigned int charLen;
+	for (unsigned int i=0;i<len;) {
 		unsigned char ch = static_cast<unsigned char>(s[i]);
-		if ((ch < 0x80) || (ch > (0x80 + 0x40)))
+		if (ch < 0x80) {
+			charLen = 1;
+		} else if (ch < 0x80 + 0x40 + 0x20) {
+			charLen = 2;
+		} else if (ch < 0x80 + 0x40 + 0x20 + 0x10) {
+			charLen = 3;
+		} else {
+			charLen = 4;
 			ulen++;
+		}
+		i += charLen;
+		ulen++;
 	}
 	return ulen;
 }
 
-unsigned int UCS2FromUTF8(const char *s, unsigned int len, wchar_t *tbuf, unsigned int tlen) {
+unsigned int UTF16FromUTF8(const char *s, unsigned int len, wchar_t *tbuf, unsigned int tlen) {
 	unsigned int ui=0;
 	const unsigned char *us = reinterpret_cast<const unsigned char *>(s);
 	unsigned int i=0;
@@ -63,12 +94,24 @@ unsigned int UCS2FromUTF8(const char *s, unsigned int len, wchar_t *tbuf, unsign
 			tbuf[ui] = static_cast<wchar_t>((ch & 0x1F) << 6);
 			ch = us[i++];
 			tbuf[ui] = static_cast<wchar_t>(tbuf[ui] + (ch & 0x7F));
-		} else {
+		} else if (ch < 0x80 + 0x40 + 0x20 + 0x10) {
 			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));
+		} else {
+			// Outside the BMP so need two surrogates
+			int val = (ch & 0x7) << 18;
+			ch = us[i++];
+			val += (ch & 0x3F) << 12;
+			ch = us[i++];
+			val += (ch & 0x3F) << 6;
+			ch = us[i++];
+			val += (ch & 0x3F);
+			tbuf[ui] = static_cast<wchar_t>(((val - 0x10000) >> 10) + SURROGATE_LEAD_FIRST);
+			ui++;
+			tbuf[ui] = static_cast<wchar_t>((val & 0x3ff) + SURROGATE_TRAIL_FIRST);
 		}
 		ui++;
 	}