X-Git-Url: http://git.tdb.fi/?p=libs%2Fcore.git;a=blobdiff_plain;f=source%2Futf8.cpp;h=c7e1705f814e2d099d3b0180730c57e3aa210d85;hp=0217790277a2a1b14112668552cc1675125028d6;hb=5b1368cb791cab043f0435628cacbaff36e39b7b;hpb=58384e355b4a78730d69243f1092e47591f2f384

diff --git a/source/utf8.cpp b/source/utf8.cpp
index 0217790..c7e1705 100644
--- a/source/utf8.cpp
+++ b/source/utf8.cpp
@@ -1,92 +1,92 @@
+/* $Id$
+
+This file is part of libmspstrings
+Copyright © 2006-2007 Mikko Rasa
+Distributed under the LGPL
+*/
+
 #include "utf8.h"
 
 using namespace std;
 
 namespace Msp {
+namespace Codecs {
 
-void Utf8::Encoder::encode_char(wchar_t c)
+void Utf8::Encoder::encode_char(UnicodeChar ch, string &buf)
 {
-	unsigned code=c;
-	if(code>0x10FFFF)
-		throw CodecError("Can't express character in UTF-8");
-
-	unsigned bytes=1;
-	if(code>0xFFFF)
-		bytes=4;
-	else if(code>0x7FF)
-		bytes=3;
-	else if(code>0x7F)
-		bytes=2;
+	if(ch<0 || ch>0x10FFFF)
+		return error(ch, buf, "Can't express character in UTF-8");
+
+	unsigned bytes = 1;
+	if(ch>0xFFFF)
+		bytes = 4;
+	else if(ch>0x7FF)
+		bytes = 3;
+	else if(ch>0x7F)
+		bytes = 2;
 
 	if(bytes==1)
-		append(code);
+		buf += ch;
 	else
 	{
-		char buf[4];
-		
-		buf[0]=0xFF<<(8-bytes) | code>>(bytes*6-6);
+		char utf[4];
+
+		utf[0] = 0xFF<<(8-bytes) | ch>>(bytes*6-6);
 		for(unsigned j=bytes-1; j>0; --j)
 		{
-			buf[j]=0x80 | code&0x3F;
-			code>>=6;
+			utf[j] = 0x80 | (ch&0x3F);
+			ch >>= 6;
 		}
 
-		append(buf, bytes);
+		buf.append(utf, bytes);
 	}
 }
 
+void Utf8::Encoder::transliterate(UnicodeChar, string &buf)
+{
+	buf.append("\357\277\275", 3);  // � U+FFFE Replacement Character
+}
 
-void Utf8::Decoder::decode_char(const string &str, string::const_iterator &i)
+
+UnicodeChar Utf8::Decoder::decode_char(const string &str, string::const_iterator &i)
 {
-	while(i!=str.end())
+	if(i==str.end())
+		return error("No input");
+
+	if((*i&0xC0)==0x80)
 	{
-		if(bytes==0)
-		{
-			if((*i&0xC0)==0x80)
-				throw CodecError("Invalid UTF-8 string (tail byte when expecting head)");
-
-			else if(*i&0x80)
-			{
-				unsigned mask=0x40;
-				for(; *i&mask; mask>>=1)
-					++bytes;
-
-				if(bytes>3)
-					throw CodecError("Invalid UTF-8 string (overlong multibyte sequence)");
-
-				code=(*i++)&(mask-1);
-				if(!code)
-					throw CodecError("Invalid UTF-8 string (denormalized multibyte sequence)");
-			}
-			else
-			{
-				append(*i++);
-				break;
-			}
-		}
-		else
-		{
-			if((*i&0xC0)!=0x80)
-				throw CodecError("Invalid UTF-8 string (head byte when expecting tail)");
-
-			code=code<<6 | (*i++)&0x3F;
-			--bytes;
-
-			if(!bytes)
-			{
-				if(code>0x10FFFF)
-					throw CodecError("Invalid UTF-8 string (character out of range)");
-				append(code);
-				break;
-			}
-		}
+		UnicodeChar result = error("UTF-8 tail byte found when expecting head");
+		++i;
+		return result;
 	}
-}
+	else if(*i&0x80)
+	{
+		unsigned bytes = 2;
+		unsigned mask = 0x20;
+		for(; *i&mask; mask>>=1)
+			++bytes;
 
-void Utf8::Decoder::sync()
-{
-	if(bytes)
-		throw CodecError("Sync in the middle of multibyte UTF-8 sequence");
+		string::const_iterator j = i;
+
+		UnicodeChar result = (*j++)&(mask-1);
+
+		unsigned k;
+		for(k=1; (k<bytes && j!=str.end() && (*j&0xC0)==0x80); ++k)
+			result = (result<<6) | ((*j++)&0x3F);
+
+		if(k<bytes)
+			result = error("Incomplete UTF-8 character");
+		else if(!(result>>(bytes*5-4)) || !(result>>7))
+			result = error("Denormalized UTF-8 multibyte sequence");
+		else if(result>0x10FFFF || (result>=0xD800 && result<=0xDFFF))
+			result = error("Invalid Unicode code point");
+
+		i = j;
+		return result;
+	}
+	else
+		return *i++;
 }
 
+} // namespace Codecs
 } // namespace Msp