From d0607dd9d43829284ebfe4909cc627bb1edf1ce8 Mon Sep 17 00:00:00 2001
From: John Hodge <tpg@mutabah.net>
Date: Sun, 23 Jun 2013 00:07:44 +0800
Subject: [PATCH] Kernel - Added UTF-16 functions

---
 KernelLand/Kernel/Makefile        |  2 +-
 KernelLand/Kernel/include/utf16.h | 16 +++++++
 KernelLand/Kernel/utf16.c         | 79 +++++++++++++++++++++++++++++++
 3 files changed, 96 insertions(+), 1 deletion(-)
 create mode 100644 KernelLand/Kernel/include/utf16.h
 create mode 100644 KernelLand/Kernel/utf16.c

diff --git a/KernelLand/Kernel/Makefile b/KernelLand/Kernel/Makefile
index f6820132..0aa55aba 100644
--- a/KernelLand/Kernel/Makefile
+++ b/KernelLand/Kernel/Makefile
@@ -54,7 +54,7 @@ BUILDINFO_SRC := $(OBJDIR)buildinfo.c$(OBJSUFFIX)
 
 OBJ := $(addprefix arch/$(ARCHDIR)/,$(A_OBJ))
 OBJ += pmemmap.o
-OBJ += heap.o logging.o debug.o lib.o libc.o adt.o time.o
+OBJ += heap.o logging.o debug.o lib.o libc.o adt.o time.o utf16.o
 OBJ += drvutil_video.o drvutil_disk.o
 OBJ += messages.o modules.o syscalls.o system.o
 OBJ += threads.o mutex.o semaphore.o workqueue.o events.o rwlock.o
diff --git a/KernelLand/Kernel/include/utf16.h b/KernelLand/Kernel/include/utf16.h
new file mode 100644
index 00000000..0a6666cf
--- /dev/null
+++ b/KernelLand/Kernel/include/utf16.h
@@ -0,0 +1,16 @@
+/*
+ * Acess2 Kernel
+ * - By John Hodge (thePowersGang)
+ *
+ * utf16.h
+ * - UTF-16 <-> UTF-8/UCS32 translation
+ */
+#ifndef _UTF16_H_
+#define _UTF16_H_
+
+extern int	ReadUTF16(const Uint16 *Str16, Uint32 *Codepoint);
+extern size_t	UTF16_ConvertToUTF8(size_t DestLen, char *Dest, size_t SrcLen, const Uint16 *Source);
+extern int	UTF16_CompareWithUTF8(size_t Str16Len, const Uint16 *Str16, const char *Str8);
+
+#endif
+
diff --git a/KernelLand/Kernel/utf16.c b/KernelLand/Kernel/utf16.c
new file mode 100644
index 00000000..514fc7cb
--- /dev/null
+++ b/KernelLand/Kernel/utf16.c
@@ -0,0 +1,79 @@
+/*
+ * Acess2 Kernel
+ * - By John Hodge (thePowersGang) 
+ *
+ * utf16.c
+ * - UTF-16 Translation/Manipulation
+ */
+#define DEBUG	1
+#include <acess.h>
+#include <utf16.h>
+
+int ReadUTF16(const Uint16 *Str16, Uint32 *Codepoint)
+{
+	if( 0xD800 < *Str16 && *Str16 <= 0xDFFF )
+	{
+		// UTF-16 surrogate pair
+		// > 0xDC00 is the second word
+		if( Str16[0] > 0xDC00 ) {
+			*Codepoint = 0;
+			return 1;
+		}
+		if( Str16[1] < 0xD800 || Str16[1] >= 0xDC00 ) {
+			*Codepoint = 0;
+			return 2;
+		}
+		// 2^16 + 20-bit
+		*Codepoint = 0x10000 + (((Str16[0] & 0x3FF) << 10) | (Str16[1] & 0x3FF));
+		return 2;
+	}
+	else {
+		*Codepoint = *Str16;
+		return 1;
+	}
+}
+
+size_t UTF16_ConvertToUTF8(size_t DestLen, char *Dest, size_t SrcLen, const Uint16 *Source)
+{
+	 int	len = 0;
+	for( ; *Source && SrcLen --; Source ++ )
+	{
+		// TODO: Decode/Reencode
+		if( Dest && len < DestLen )
+			Dest[len] = *Source;
+		len += 1;
+	}
+	if( Dest && len < DestLen )
+		Dest[len] = 0;
+	return len;
+}
+
+int UTF16_CompareWithUTF8(size_t Str16Len, const Uint16 *Str16, const char *Str8)
+{
+	 int	pos16 = 0, pos8 = 0;
+	const Uint8	*str8 = (const Uint8 *)Str8;
+	
+	while( pos16 < Str16Len && Str16[pos16] && str8[pos8] )
+	{
+		Uint32	cp8, cp16;
+		pos16 += ReadUTF16(Str16+pos16, &cp16);
+		pos8 += ReadUTF8(str8 + pos8, &cp8);
+	
+		LOG("cp16 = %x, cp8 = %x", cp16, cp8);
+		if(cp16 == cp8)	continue ;
+		
+		if(cp16 < cp8)
+			return -1;
+		else
+			return 1;
+	}
+	if( pos16 == Str16Len )
+		return 0;
+	if( Str16[pos16] && str8[pos8] )
+		return 0;
+	if( Str16[pos16] )
+		return 1;
+	else
+		return -1;
+}
+
-- 
2.20.1