Merge branch 'v6-devel'

Conflicts:
	ChangeLog

6 files changed, 275 insertions, 80 deletions

diff --git a/ChangeLog b/ChangeLog
index 64d29d3d..ef3eb344 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,4 +1,15 @@
 ---------------------------------------------------------------------------
+Version 7.1.3  [devel] 2012-09-??
+- bugfix: missing support for escape sequences in RainerScript
+  only \' was supported. Now the usual set is supported. Note that v5
+  used \x as escape where x was any character (e.g. "\n" meant "n" and NOT
+  LF). This also means there is some incompatibility to v5 for well-know
+  sequences. Better break it now than later.
+- bugfix: invalid property name in property-filter could cause abort
+  if action chaining (& operator) was used
+  http://bugzilla.adiscon.com/show_bug.cgi?id=355
+  Thanks to pilou@gmx.com for the bug report
+---------------------------------------------------------------------------
 Version 7.1.2  [devel] 2012-09-12
 - bugfix: messages were duplicated, sometimes massively
   regression from new code in 7.1.1 and reason for early release
@@ -28,6 +39,18 @@ Version 7.1.0  [devel] 2012-09-06
 - bugfix: imtcp could abort on exit due to invalid free()
 - imported bugfixes from 6.4.1
 ---------------------------------------------------------------------------
+Version 6.5.1  [devel] 2012-08-??
+- added pure JSON output plugin parameter passing mode
+- ommongodb now supports templates
+- bugfix: imtcp could abort on exit due to invalid free()
+- bugfix: remove invalid socket option call from imuxsock
+  Thanks to Cristian Ionescu-Idbohrn and Jonny Törnbom 
+- bugfix: missing support for escape sequences in RainerScript
+  only \' was supported. Now the usual set is supported. Note that v5
+  used \x as escape where x was any character (e.g. "\n" meant "n" and NOT
+  LF). This also means there is some incompatibility to v5 for well-know
+  sequences. Better break it now than later.
+---------------------------------------------------------------------------
 Version 6.5.0  [devel] 2012-08-28
 - imrelp now supports non-cancel thread termination
   (but now requires at least librelp 1.0.1)
@@ -53,6 +76,15 @@ Version 6.5.0  [devel] 2012-08-28
   Note: patches were released under ASL 2.0, see
   http://bugzilla.adiscon.com/show_bug.cgi?id=353
 ---------------------------------------------------------------------------
+Version 6.4.2  [V6-STABLE] 2012-09-??
+- bugfix: remove invalid socket option call from imuxsock
+  Thanks to Cristian Ionescu-Idbohrn and Jonny Törnbom 
+- bugfix: missing support for escape sequences in RainerScript
+  only \' was supported. Now the usual set is supported. Note that v5
+  used \x as escape where x was any character (e.g. "\n" meant "n" and NOT
+  LF). This also means there is some incompatibility to v5 for well-know
+  sequences. Better break it now than later.
+---------------------------------------------------------------------------
 Version 6.4.1  [V6-STABLE] 2012-09-06
 - bugfix: multiple main queues with same queue file name were not detected
   This lead to queue file corruption. While the root cause is a config
@@ -607,6 +639,10 @@ expected that interfaces, even new ones, break during the initial
   [ported from v4]
 ---------------------------------------------------------------------------
 Version 5.10.1  [V5-STABLE], 2012-0?-??
+- bugfix: invalid property name in property-filter could cause abort
+  if action chaining (& operator) was used
+  http://bugzilla.adiscon.com/show_bug.cgi?id=355
+  Thanks to pilou@gmx.com for the bug report
 - bugfix: remove invalid socket option call from imuxsock
   Thanks to Cristian Ionescu-Idbohrn and Jonny Törnbom 
 ---------------------------------------------------------------------------
diff --git a/doc/v6compatibility.html b/doc/v6compatibility.html
index eec0784b..edb6dedf 100644
--- a/doc/v6compatibility.html
+++ b/doc/v6compatibility.html
@@ -181,6 +181,15 @@ As you see, here you may include spaces between user names.
 so it is a wise decision to change config files at least to the legacy
 format (with ":omusrmsg:" in front of the name).
 
+<h2>Escape Sequences in Script-Based Filters<h2>
+<p>In v5, escape sequences were very simplistic. Inside a string, "\x" meant
+"x" with x being any character. This has been changed so that the usual set of
+escapes is supported, must importantly "\n", "\t", "\xhh" (with hh being hex digits)
+and "\ooo" with (o being octal digits). So if one of these sequences was used
+previously, results are obviously different. However, that should not create any
+real problems, because it is hard to envision why someone should have done that
+(why write "\n" when you can also write "n"?).
+---------------------------------------------------------------------------
 <p>[<a href="manual.html">manual index</a>] [<a href="http://www.rsyslog.com/">rsyslog site</a>]</p>
 <p><font size="2">This documentation is part of the
 <a href="http://www.rsyslog.com/">rsyslog</a> project.<br>
diff --git a/grammar/lexer.l b/grammar/lexer.l
index db875386..8b930713 100644
--- a/grammar/lexer.l
+++ b/grammar/lexer.l
@@ -122,9 +122,15 @@ int fileno(FILE *stream);
 <EXPR>0x[0-7a-f] |		/* hex number, following rule is dec; strtoll handles all! */
 <EXPR>([1-9][0-9]*|0)		{ yylval.n = strtoll(yytext, NULL, 0); return NUMBER; }
 <EXPR>\$[$!]{0,1}[a-z][a-z0-9\-_\.]*	{ yylval.s = strdup(yytext); return VAR; }
-<EXPR>\'([^'\\]|\\['])*\'	 { yylval.estr = es_newStrFromBuf(yytext+1, yyleng-2);
+<EXPR>\'([^'\\]|\\['"\\$bntr]|\\x[0-9a-f][0-9a-f]|\\[0-7][0-7][0-7])*\'	 {
+				   yytext[yyleng-1] = '\0';
+				   unescapeStr((uchar*)yytext+1, yyleng-2);
+				   yylval.estr = es_newStrFromBuf(yytext+1, strlen(yytext));
 				   return STRING; }
-<EXPR>\"([^"\\]|\\["])*\"	 { yylval.estr = es_newStrFromBuf(yytext+1, yyleng-2);
+<EXPR>\"([^"\\$]|\\["'\\$bntr]|\\x[0-9a-f][0-9a-f]|\\[0-7][0-7][0-7])*\" {
+				   yytext[yyleng-1] = '\0';
+				   unescapeStr((uchar*)yytext+1, yyleng-2);
+				   yylval.estr = es_newStrFromBuf(yytext+1, yyleng-2);
 				   return STRING; }
 <EXPR>[ \t\n]
 <EXPR>[a-z][a-z0-9_]*		{ yylval.estr = es_newStrFromCStr(yytext, yyleng);
diff --git a/grammar/rainerscript.c b/grammar/rainerscript.c
index d588ccc9..b0a79458 100644
--- a/grammar/rainerscript.c
+++ b/grammar/rainerscript.c
@@ -1910,3 +1910,141 @@ initRainerscript(void)
 finalize_it:
 	RETiRet;
 }
+
+/* we need a function to check for octal digits */
+static inline int
+isodigit(uchar c)
+{
+	return(c >= '0' && c <= '7');
+}
+
+/**
+ * Get numerical value of a hex digit. This is a helper function.
+ * @param[in] c a character containing 0..9, A..Z, a..z anything else
+ * is an (undetected) error.
+ */
+static inline int
+hexDigitVal(char c)
+{
+	int r;
+	if(c < 'A')
+		r = c - '0';
+	else if(c < 'a')
+		r = c - 'A' + 10;
+	else
+		r = c - 'a' + 10;
+	return r;
+}
+
+/* Handle the actual unescaping.
+ * a helper to unescapeStr(), to help make the function easier to read.
+ */
+static inline void
+doUnescape(unsigned char *c, int len, int *iSrc, int iDst)
+{
+	if(c[*iSrc] == '\\') {
+		if(++(*iSrc) == len) {
+			/* error, incomplete escape, treat as single char */
+			c[iDst] = '\\';
+		}
+		/* regular case, unescape */
+		switch(c[*iSrc]) {
+		case 'a':
+			c[iDst] = '\007';
+			break;
+		case 'b':
+			c[iDst] = '\b';
+			break;
+		case 'f':
+			c[iDst] = '\014';
+			break;
+		case 'n':
+			c[iDst] = '\n';
+			break;
+		case 'r':
+			c[iDst] = '\r';
+			break;
+		case 't':
+			c[iDst] = '\t';
+			break;
+		case '\'':
+			c[iDst] = '\'';
+			break;
+		case '"':
+			c[iDst] = '"';
+			break;
+		case '?':
+			c[iDst] = '?';
+			break;
+		case '$':
+			c[iDst] = '$';
+			break;
+		case '\\':
+			c[iDst] = '\\';
+			break;
+		case 'x':
+			if(    (*iSrc)+2 >= len
+			   || !isxdigit(c[(*iSrc)+1])
+			   || !isxdigit(c[(*iSrc)+2])) {
+				/* error, incomplete escape, use as is */
+				c[iDst] = '\\';
+				--(*iSrc);
+			}
+			c[iDst] = (hexDigitVal(c[(*iSrc)+1]) << 4) +
+				  hexDigitVal(c[(*iSrc)+2]);
+			*iSrc += 2;
+			break;
+		case '0': /* octal escape */
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+			if(    (*iSrc)+2 >= len
+			   || !isodigit(c[(*iSrc)+1])
+			   || !isodigit(c[(*iSrc)+2])) {
+				/* error, incomplete escape, use as is */
+				c[iDst] = '\\';
+				--(*iSrc);
+			}
+			c[iDst] = ((c[(*iSrc)  ] - '0') << 6) +
+			          ((c[(*iSrc)+1] - '0') << 3) +
+			          ( c[(*iSrc)+2] - '0');
+			*iSrc += 2;
+			break;
+		default:
+			/* error, incomplete escape, indicate by '?' */
+			c[iDst] = '?';
+			break;
+		}
+	} else {
+		/* regular character */
+		c[iDst] = c[*iSrc];
+	}
+}
+
+void
+unescapeStr(uchar *s, int len)
+{
+	int iSrc, iDst;
+	assert(s != NULL);
+
+	/* scan for first escape sequence (if we are luky, there is none!) */
+	iSrc = 0;
+	while(iSrc < len && s[iSrc] != '\\')
+		++iSrc;
+	/* now we have a sequence or end of string. In any case, we process
+	 * all remaining characters (maybe 0!) and unescape.
+	 */
+	if(iSrc != len) {
+		iDst = iSrc;
+		while(iSrc < len) {
+			doUnescape(s, len, &iSrc, iDst);
+			++iSrc;
+			++iDst;
+		}
+		s[iDst] = '\0';
+	}
+}
diff --git a/grammar/rainerscript.h b/grammar/rainerscript.h
index 5306188d..7637ae48 100644
--- a/grammar/rainerscript.h
+++ b/grammar/rainerscript.h
@@ -291,6 +291,7 @@ struct cnfstmt * cnfstmtNewLegaAct(char *actline);
 void cnfstmtDestruct(struct cnfstmt *root);
 char* getFIOPName(unsigned iFIOP);
 rsRetVal initRainerscript(void);
+void unescapeStr(uchar *s, int len);
 
 /* debug helper */
 void cstrPrint(char *text, es_str_t *estr);
diff --git a/runtime/rule.c b/runtime/rule.c
index 67ecd077..f6160bdc 100644
--- a/runtime/rule.c
+++ b/runtime/rule.c
@@ -184,78 +184,83 @@ shouldProcessThisMessage(rule_t *pRule, msg_t *pMsg, sbool *bProcessMsg)
 		DBGPRINTF("result of rainerscript filter evaluation: %d\n", bRet);
 	} else {
 		assert(pRule->f_filter_type == FILTER_PROP); /* assert() just in case... */
-		pszPropVal = MsgGetProp(pMsg, NULL, pRule->f_filterData.prop.propID,
-					pRule->f_filterData.prop.propName, &propLen, &pbMustBeFreed);
-
-		/* Now do the compares (short list currently ;)) */
-		switch(pRule->f_filterData.prop.operation ) {
-		case FIOP_CONTAINS:
-			if(rsCStrLocateInSzStr(pRule->f_filterData.prop.pCSCompValue, (uchar*) pszPropVal) != -1)
-				bRet = 1;
-			break;
-		case FIOP_ISEMPTY:
-			if(propLen == 0)
-				bRet = 1; /* process message! */
-			break;
-		case FIOP_ISEQUAL:
-			if(rsCStrSzStrCmp(pRule->f_filterData.prop.pCSCompValue,
-					  pszPropVal, ustrlen(pszPropVal)) == 0)
-				bRet = 1; /* process message! */
-			break;
-		case FIOP_STARTSWITH:
-			if(rsCStrSzStrStartsWithCStr(pRule->f_filterData.prop.pCSCompValue,
-					  pszPropVal, ustrlen(pszPropVal)) == 0)
-				bRet = 1; /* process message! */
-			break;
-		case FIOP_REGEX:
-			if(rsCStrSzStrMatchRegex(pRule->f_filterData.prop.pCSCompValue,
-					(unsigned char*) pszPropVal, 0, &pRule->f_filterData.prop.regex_cache) == RS_RET_OK)
-				bRet = 1;
-			break;
-		case FIOP_EREREGEX:
-			if(rsCStrSzStrMatchRegex(pRule->f_filterData.prop.pCSCompValue,
-					  (unsigned char*) pszPropVal, 1, &pRule->f_filterData.prop.regex_cache) == RS_RET_OK)
-				bRet = 1;
-			break;
-		default:
-			/* here, it handles NOP (for performance reasons) */
-			assert(pRule->f_filterData.prop.operation == FIOP_NOP);
-			bRet = 1; /* as good as any other default ;) */
-			break;
-		}
-
-		/* now check if the value must be negated */
-		if(pRule->f_filterData.prop.isNegated)
-			bRet = (bRet == 1) ?  0 : 1;
-
-		if(Debug) {
-			char *cstr;
-			if(pRule->f_filterData.prop.propID == PROP_CEE) {
-				cstr = es_str2cstr(pRule->f_filterData.prop.propName, NULL);
-				DBGPRINTF("Filter: check for CEE property '%s' (value '%s') ",
-					cstr, pszPropVal);
-				free(cstr);
-			} else {
-				DBGPRINTF("Filter: check for property '%s' (value '%s') ",
-					propIDToName(pRule->f_filterData.prop.propID), pszPropVal);
+		if(pRule->f_filterData.prop.propID == PROP_INVALID) {
+			DBGPRINTF("invalid property ID, filter always returns 0\n");
+			bRet = 0;
+		} else {
+			pszPropVal = MsgGetProp(pMsg, NULL, pRule->f_filterData.prop.propID,
+						pRule->f_filterData.prop.propName, &propLen, &pbMustBeFreed);
+
+			/* Now do the compares (short list currently ;)) */
+			switch(pRule->f_filterData.prop.operation ) {
+			case FIOP_CONTAINS:
+				if(rsCStrLocateInSzStr(pRule->f_filterData.prop.pCSCompValue, (uchar*) pszPropVal) != -1)
+					bRet = 1;
+				break;
+			case FIOP_ISEMPTY:
+				if(propLen == 0)
+					bRet = 1; /* process message! */
+				break;
+			case FIOP_ISEQUAL:
+				if(rsCStrSzStrCmp(pRule->f_filterData.prop.pCSCompValue,
+						  pszPropVal, ustrlen(pszPropVal)) == 0)
+					bRet = 1; /* process message! */
+				break;
+			case FIOP_STARTSWITH:
+				if(rsCStrSzStrStartsWithCStr(pRule->f_filterData.prop.pCSCompValue,
+						  pszPropVal, ustrlen(pszPropVal)) == 0)
+					bRet = 1; /* process message! */
+				break;
+			case FIOP_REGEX:
+				if(rsCStrSzStrMatchRegex(pRule->f_filterData.prop.pCSCompValue,
+						(unsigned char*) pszPropVal, 0, &pRule->f_filterData.prop.regex_cache) == RS_RET_OK)
+					bRet = 1;
+				break;
+			case FIOP_EREREGEX:
+				if(rsCStrSzStrMatchRegex(pRule->f_filterData.prop.pCSCompValue,
+						  (unsigned char*) pszPropVal, 1, &pRule->f_filterData.prop.regex_cache) == RS_RET_OK)
+					bRet = 1;
+				break;
+			default:
+				/* here, it handles NOP (for performance reasons) */
+				assert(pRule->f_filterData.prop.operation == FIOP_NOP);
+				bRet = 1; /* as good as any other default ;) */
+				break;
 			}
+
+			/* now check if the value must be negated */
 			if(pRule->f_filterData.prop.isNegated)
-				DBGPRINTF("NOT ");
-			if(pRule->f_filterData.prop.operation == FIOP_ISEMPTY) {
-				DBGPRINTF("%s : %s\n",
-				       getFIOPName(pRule->f_filterData.prop.operation),
-				       bRet ? "TRUE" : "FALSE");
-			} else {
-				DBGPRINTF("%s '%s': %s\n",
-				       getFIOPName(pRule->f_filterData.prop.operation),
-				       rsCStrGetSzStrNoNULL(pRule->f_filterData.prop.pCSCompValue),
-				       bRet ? "TRUE" : "FALSE");
+				bRet = (bRet == 1) ?  0 : 1;
+
+			if(Debug) {
+				char *cstr;
+				if(pRule->f_filterData.prop.propID == PROP_CEE) {
+					cstr = es_str2cstr(pRule->f_filterData.prop.propName, NULL);
+					dbgprintf("Filter: check for CEE property '%s' (value '%s') ",
+						cstr, pszPropVal);
+					free(cstr);
+				} else {
+					dbgprintf("Filter: check for property '%s' (value '%s') ",
+						propIDToName(pRule->f_filterData.prop.propID), pszPropVal);
+				}
+				if(pRule->f_filterData.prop.isNegated)
+					dbgprintf("NOT ");
+				if(pRule->f_filterData.prop.operation == FIOP_ISEMPTY) {
+					dbgprintf("%s : %s\n",
+					       getFIOPName(pRule->f_filterData.prop.operation),
+					       bRet ? "TRUE" : "FALSE");
+				} else {
+					dbgprintf("%s '%s': %s\n",
+					       getFIOPName(pRule->f_filterData.prop.operation),
+					       rsCStrGetSzStrNoNULL(pRule->f_filterData.prop.pCSCompValue),
+					       bRet ? "TRUE" : "FALSE");
+				}
 			}
-		}
 
-		/* cleanup */
-		if(pbMustBeFreed)
-			free(pszPropVal);
+			/* cleanup */
+			if(pbMustBeFreed)
+				free(pszPropVal);
+		}
 	}
 
 finalize_it:
@@ -401,16 +406,16 @@ CODESTARTobjDebugPrint(rule)
 	} else {
 		dbgprintf("PROPERTY-BASED Filter:\n");
 		dbgprintf("\tProperty.: '%s'\n", propIDToName(pThis->f_filterData.prop.propID));
-		if(pThis->f_filterData.prop.propName != NULL) {
-			cstr = es_str2cstr(pThis->f_filterData.prop.propName, NULL);
-			dbgprintf("\tCEE-Prop.: '%s'\n", cstr);
-			free(cstr);
-		}
-		dbgprintf("\tOperation: ");
-		if(pThis->f_filterData.prop.isNegated)
-			dbgprintf("NOT ");
-		dbgprintf("'%s'\n", getFIOPName(pThis->f_filterData.prop.operation));
-		if(pThis->f_filterData.prop.pCSCompValue != NULL) {
+		if(pThis->f_filterData.prop.propID != PROP_INVALID) {
+			if(pThis->f_filterData.prop.propName != NULL) {
+				cstr = es_str2cstr(pThis->f_filterData.prop.propName, NULL);
+				dbgprintf("\tCEE-Prop.: '%s'\n", cstr);
+				free(cstr);
+			}
+			dbgprintf("\tOperation: ");
+			if(pThis->f_filterData.prop.isNegated)
+				dbgprintf("NOT ");
+			dbgprintf("'%s'\n", getFIOPName(pThis->f_filterData.prop.operation));
 			dbgprintf("\tValue....: '%s'\n",
 			       rsCStrGetSzStrNoNULL(pThis->f_filterData.prop.pCSCompValue));
 		}