--- parser3/src/main/pa_charset.C	2008/07/16 17:07:16	1.61
+++ parser3/src/main/pa_charset.C	2009/04/17 09:04:49	1.70
@@ -1,11 +1,11 @@
 /** @file
 	Parser: Charset connection implementation.
 
-	Copyright(c) 2001-2005 ArtLebedev Group (http://www.artlebedev.com)
+	Copyright(c) 2001-2009 ArtLebedev Group (http://www.artlebedev.com)
 	Author: Alexander Petrosyan<paf@design.ru>(http://paf.design.ru)
 */
 
-static const char * const IDENT_CHARSET_C="$Date: 2008/07/16 17:07:16 $";
+static const char * const IDENT_CHARSET_C="$Date: 2009/04/17 09:04:49 $";
 
 #include "pa_charset.h"
 #include "pa_charsets.h"
@@ -15,6 +15,8 @@ static const char * const IDENT_CHARSET_
 #endif
 
 //#define PA_PATCHED_LIBXML_BACKWARD
+
+// reduce memory usage by pre-calculation utf-8 string length
 #define PRECALCULATE_DEST_LENGTH
 
 // globals
@@ -86,7 +88,6 @@ static void element2case(unsigned char f
 
 // methods
 
-extern "C" unsigned char pcre_default_tables[]; // pcre/chartables.c
 Charset::Charset(Request_charsets* charsets, const String::Body ANAME, const String* afile_spec): 
 	FNAME(ANAME),
 	FNAME_CSTR(ANAME.cstrm()) {
@@ -100,7 +101,7 @@ Charset::Charset(Request_charsets* chars
 	} else {
 		fisUTF8=true;
 		// grab default onces [for UTF-8 so to be able to make a-z =>A-Z
-		memcpy(pcre_tables, pcre_default_tables, sizeof(pcre_tables));
+		memcpy(pcre_tables, _pcre_default_tables, sizeof(pcre_tables));
 	}
 
 #ifdef XML
@@ -432,19 +433,21 @@ of ocetes consumed.
 	return 0;
 }
 
-static bool is_escaped(char c){
+static bool isEscaped(XMLByte c){
 	return
-		!(c<=127
-		&& (
-			((c>='0') && (c<='9'))
-			|| ((c>='A') && (c<='Z'))
-			|| ((c>='a') && (c<='z'))
-			|| strchr("*@-_+./", c)!=0
-		));
+		!(
+			(c<=127)
+			&& (
+				((c>='0') && (c<='9'))
+				|| ((c>='A') && (c<='Z'))
+				|| ((c>='a') && (c<='z'))
+				|| strchr("*@-_+./", c)!=0
+			)
+		);
 }
 
-// read one utf8 character, return number of bytes needed for store it
-static unsigned int readChar(const XMLByte*& srcPtr, const XMLByte* srcEnd, XMLByte& firstByte, XMLCh& UTF8Char){
+// read one UTF8 char and return length of this char (in bytes)
+static unsigned int readUTF8Char(const XMLByte*& srcPtr, const XMLByte* srcEnd, XMLByte& firstByte, XMLCh& UTF8Char){
 	if(!srcPtr || !*srcPtr || srcPtr>=srcEnd)
 		return 0;
 
@@ -478,13 +481,18 @@ static unsigned int readChar(const XMLBy
 	return trailingBytes+1;
 }
 
-static unsigned int readChar(const XMLByte*& srcPtr, const XMLByte* srcEnd){
-	XMLByte firstByte;
-	XMLCh UTF8Char;
-	return readChar(srcPtr, srcEnd, firstByte, UTF8Char);
+// skip UTF8 char and return length of this char (in bytes)
+static unsigned int skipUTF8Char(const XMLByte*& srcPtr, const XMLByte* srcEnd){
+	if(!srcPtr || !*srcPtr || srcPtr>=srcEnd)
+		return 0;
+
+	unsigned int trailingBytes=gUTFBytes[*srcPtr]+1;
+	srcPtr+=trailingBytes;
+
+	return trailingBytes;
 }
 
-// read char, return number of bytes needed for store it as UTF8
+// read non-UTF8 char, and return number of bytes needed for store this char in UTF8
 static unsigned int readChar(const XMLByte*& srcPtr, const XMLByte* srcEnd, XMLByte& firstByte, XMLCh& UTF8Char, const Charset::Tables& tables){
 	if(!srcPtr || !*srcPtr || srcPtr>=srcEnd)
 		return 0;
@@ -510,7 +518,7 @@ static unsigned int readChar(const XMLBy
 	return 1;
 }
 
-static int escape(const XMLByte* srcData, size_t& srcLen,
+static int escapeUTF8(const XMLByte* srcData, size_t& srcLen,
 			     XMLByte* toFill, size_t& toFillLen) {
 	const XMLByte* srcPtr=srcData;
 	const XMLByte* srcEnd=srcData+srcLen;
@@ -521,9 +529,9 @@ static int escape(const XMLByte* srcData
 	uint charSize;
 
 	// loop until we either run out of input data, or room to store
-	while((outPtr < outEnd) && (charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char))){
+	while((outPtr < outEnd) && (charSize=readUTF8Char(srcPtr, srcEnd, firstByte, UTF8Char))){
 		if(charSize==1){
-			if(is_escaped(firstByte)) // %XX
+			if(isEscaped(firstByte)) // %XX
 				outPtr+=sprintf((char*)outPtr, "%%%02X", firstByte);
 			else
 				*outPtr++=firstByte;
@@ -546,7 +554,6 @@ static int escape(const XMLByte* srcData
 	const XMLByte* srcPtr=srcData;
 	const XMLByte* srcEnd=srcData+srcLen;
 	XMLByte* outPtr=toFill;
-	//XMLByte* outEnd=toFill+toFillLen;
 	XMLByte firstByte;
 	XMLCh UTF8Char;
 	uint charSize;
@@ -554,7 +561,7 @@ static int escape(const XMLByte* srcData
 	while(charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char, tables)){
 		if(charSize==1){
 			if(firstByte){
-				if(is_escaped(firstByte)) // %XX
+				if(isEscaped(firstByte)) // %XX
 					outPtr+=sprintf((char*)outPtr, "%%%02X", firstByte);
 				else
 					*outPtr++=firstByte;
@@ -587,16 +594,16 @@ String::C Charset::escape(const String::
 	XMLCh UTF8Char;
 
 	if(source_charset.isUTF8()){
-		while(uint charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char)){
+		while(uint charSize=readUTF8Char(srcPtr, srcEnd, firstByte, UTF8Char)){
 			if(charSize==1)
-				dest_length+=!is_escaped(firstByte)?1:3/*%XX*/;
+				dest_length+=!isEscaped(firstByte)?1:3/*%XX*/;
 			else
 				dest_length+=6; // '%uXXXX'
 		}
 	} else {
 		while(uint charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char, source_charset.tables)){
 			if(charSize==1)
-				dest_length+=(!firstByte/*replacement char '?'*/ || !is_escaped(firstByte))?1:3/*'%XX'*/;
+				dest_length+=(!firstByte/*replacement char '?'*/ || !isEscaped(firstByte))?1:3/*'%XX'*/;
 			else
 				dest_length+=6; // '%uXXXX'
 		}
@@ -614,7 +621,7 @@ String::C Charset::escape(const String::
 
 	int status;
 	if(source_charset.isUTF8()){
-		status=::escape((XMLByte *)src.str, src_length, dest_body, dest_length);
+		status=::escapeUTF8((XMLByte *)src.str, src_length, dest_body, dest_length);
 	} else {
 		status=::escape((XMLByte *)src.str, src_length, dest_body, dest_length, source_charset.tables);
 	}
@@ -622,13 +629,28 @@ String::C Charset::escape(const String::
 	if(status<0)
 		throw Exception(0,
 			0,
-			"Charset::escapeString buffer overflow");
+			"Charset::escape buffer overflow");
 
 	assert(dest_length<=saved_dest_length);
 	dest_body[dest_length]=0; // terminator
 	return String::C((char*)dest_body, dest_length);
 }
 	
+String::Body Charset::escape(const String::Body src, const Charset& source_charset) {
+	const char *src_ptr=src.cstr();
+	size_t src_size=strlen(src_ptr);
+
+	String::C dest=Charset::escape(String::C(src_ptr, src_size), source_charset);
+
+	return String::Body(dest.str, dest.length);
+}
+
+String& Charset::escape(const String& src, const Charset& source_charset) {
+	if(!src.length())
+		return *new String("", 0, false);
+
+	return *new String(escape((String::Body)src, source_charset), String::L_CLEAN);
+}
 
 const String::C Charset::transcodeToUTF8(const String::C src) const {
 	size_t src_length=src.length;
@@ -637,8 +659,8 @@ const String::C Charset::transcodeToUTF8
 	size_t dest_length=0;
 	const XMLByte* srcPtr=(XMLByte*)src.str;
 	const XMLByte* srcEnd=srcPtr+src_length;
-	XMLByte firstByte;
-	XMLCh UTF8Char;
+ 	XMLByte firstByte;
+ 	XMLCh UTF8Char;
 	while(uint charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char, tables))
 		dest_length+=charSize;
 #else
@@ -814,7 +836,7 @@ const String::C Charset::transcodeFromUT
 	const XMLByte* srcEnd=srcPtr+src_length;
 	XMLByte firstByte;
 	XMLCh UTF8Char;
-	while(uint charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char)){
+	while(uint charSize=readUTF8Char(srcPtr, srcEnd, firstByte, UTF8Char)){
 		if(charSize==1)
 			dest_length++;
 		else
@@ -1122,7 +1144,7 @@ void Charset::transcode(HashStringString
 
 size_t getUTF8BytePos(const XMLByte* srcBegin, const XMLByte* srcEnd, size_t charPos){
 	const XMLByte* ptr=srcBegin;
-	while(charPos-- && readChar(ptr, srcEnd));
+	while(charPos-- && skipUTF8Char(ptr, srcEnd));
 
 	return ptr-srcBegin;
 }
@@ -1131,7 +1153,7 @@ size_t getUTF8CharPos(const XMLByte* src
 	size_t charPos=0;
 	const XMLByte* ptr=srcBegin;
 	const XMLByte* ptrEnd=srcBegin+bytePos;
-	while(readChar(ptr, srcEnd)){
+	while(skipUTF8Char(ptr, srcEnd)){
 		if(ptr>ptrEnd)
 			return charPos;
 		charPos++;
@@ -1145,7 +1167,7 @@ size_t getUTF8CharPos(const XMLByte* src
 
 size_t lengthUTF8(const XMLByte* srcBegin, const XMLByte* srcEnd){
 	size_t size=0;
-	while(readChar(srcBegin, srcEnd))
+	while(skipUTF8Char(srcBegin, srcEnd))
 		size++;
 
 	return size;