--- parser3/src/main/untaint.C	2011/02/18 06:03:53	1.162
+++ parser3/src/main/untaint.C	2015/10/03 00:03:49	1.167
@@ -1,11 +1,11 @@
 /** @file
 	Parser: String class part: untaint mechanizm.
 
-	Copyright(c) 2001-2005 ArtLebedev Group (http://www.artlebedev.com)
+	Copyright (c) 2001-2012 Art. Lebedev Studio (http://www.artlebedev.com)
 	Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru)
 */
 
-static const char * const IDENT_UNTAINT_C="$Date: 2011/02/18 06:03:53 $";
+volatile const char * IDENT_UNTAINT_C="$Id: untaint.C,v 1.167 2015/10/03 00:03:49 moko Exp $";
 
 
 #include "pa_string.h"
@@ -52,29 +52,34 @@ extern "C" { // author forgot to do that
 	} \
     }
 
-
 #define escape_fragment(action) \
 	for(; fragment_length--; CORD_next(info->pos)) { \
 		char c=CORD_pos_fetch(info->pos); \
 		action \
 	}
-#define _default CORD_ec_append(info->result, c)
+
 #define encode(need_encode_func, prefix, otherwise)  \
 	if(need_encode_func(c)) { \
-		static const char* hex="0123456789ABCDEF"; \
 		CORD_ec_append(info->result, prefix); \
-		CORD_ec_append(info->result, hex[((unsigned char)c)/0x10]); \
-		CORD_ec_append(info->result, hex[((unsigned char)c)%0x10]); \
+		to_hex(c); \
 	} else \
 		CORD_ec_append(info->result, otherwise);
+
+#define to_hex(c) \
+	{ \
+		CORD_ec_append(info->result, hex_digits[((unsigned char)c) >> 4]); \
+		CORD_ec_append(info->result, hex_digits[((unsigned char)c) & 0x0F]); \
+	}
+
 #define to_char(c)  { CORD_ec_append(info->result, c); whitespace=false; }
 #define to_string(s)  { CORD_ec_append_cord(info->result, s); whitespace=false; }
+#define _default CORD_ec_append(info->result, c)
 
 inline bool need_file_encode(unsigned char c){
 	// russian letters and space ENABLED
 	// encoding only these...
 	return strchr(
-			"*?'\"<>|"
+			"*?\"<>|"
 #ifndef WIN32
 			":\\"
 #endif
@@ -393,8 +398,8 @@ int cstr_to_string_body_block(String::La
 				}
 				//RFC   + An 'encoded-word' MUST NOT appear in any portion of an 'addr-spec'.
 				if(!email && (
-					!to_quoted_printable && (c & 0x80)  // starting quote-printable-encoding on first 8bit char
-					|| to_quoted_printable && !mail_header_char_valid_within_Qencoded(c)
+					( !to_quoted_printable && (c & 0x80) )  // starting quote-printable-encoding on first 8bit char
+					|| ( to_quoted_printable && !mail_header_char_valid_within_Qencoded(c) )
 					)) {
 					if(!to_quoted_printable) {
 						to_string("=?");
@@ -495,7 +500,7 @@ int cstr_to_string_body_block(String::La
 		{
 			if(info->charsets==NULL || info->charsets->client().isUTF8()){
 				// escaping to \uXXXX is not needed
-				escape_fragment(switch(c) {
+				escape_fragment(switch((unsigned char)c) {
 					case '\n': to_string("\\n");  break;
 					case '"' : to_string("\\\""); break;
 					case '\\': to_string("\\\\"); break;
@@ -504,7 +509,33 @@ int cstr_to_string_body_block(String::La
 					case '\r': to_string("\\r");  break;
 					case '\b': to_string("\\b");  break;
 					case '\f': to_string("\\f");  break;
-					default  : _default; break;
+					case 0xE2: // \u2028 and \u2029 (line/paragraph separators), check bug #1023
+						if(info->charsets && info->charsets->source().isUTF8() && fragment_length>=2){
+							CORD_next(info->pos);
+							char c1=CORD_pos_fetch(info->pos);
+							CORD_next(info->pos);
+							char c2=CORD_pos_fetch(info->pos);
+							if((unsigned char)c1 == 0x80 && ((unsigned char)c2 >= 0xA8 && (unsigned char)c2 <= 0xAF)){
+								to_string("\\u20");
+								to_hex(((unsigned char)c2-0x80));
+							} else {
+								CORD_ec_append(info->result, c);
+								CORD_ec_append(info->result, c1);
+								CORD_ec_append(info->result, c2);
+							}
+							fragment_length-=2;
+						} else {
+							_default;
+						}
+						break;
+					default: 
+						if((unsigned char)c < 0x20){
+							to_string("\\u00");
+							to_hex(c);
+						} else {
+							_default;
+						}
+						break;
 				});
 			} else {
 				const char *fragment_str=info->body->mid(info->fragment_begin, fragment_length).cstr();