Implement RainerScript field() function

8 files changed, 143 insertions, 8 deletions

diff --git a/ChangeLog b/ChangeLog
index 0d1e0e61..cdf4c046 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,4 +1,7 @@
 ---------------------------------------------------------------------------
+Version 7.1.5  [devel] 2012-09-??
+- implemented RainerScript field() function
+---------------------------------------------------------------------------
 Version 7.1.4  [devel] 2012-09-19
 - implemented ability for CEE-based properties to be stored in disk queues
 - implemented string concatenation in expressions via &-operator
diff --git a/doc/rainerscript.html b/doc/rainerscript.html
index fcc2674d..df31db2a 100644
--- a/doc/rainerscript.html
+++ b/doc/rainerscript.html
@@ -61,6 +61,11 @@ variable, if it exists. Returns an empty string if it does not exist.
 <li>cstr(expr) - converts expr to a string value
 <li>cnum(expr) - converts expr to a number (integer)
 <li>re_match(expr, re) - returns 1, if expr matches re, 0 otherwise
+<li>field(str, delim, matchnbr) - returns a field-based substring. str is the string
+to search, delim is the numerical ascii value of the field delimiter (so that
+non-printable characters can by specified) and matchnbr is the match to search
+for (the first match starts at 1). This works similar as the field based
+property-replacer option.
 </ul>
 <p>The following example can be used to build a dynamic filter based on some environment
 variable:
diff --git a/grammar/rainerscript.c b/grammar/rainerscript.c
index e7299e62..9eeda64e 100644
--- a/grammar/rainerscript.c
+++ b/grammar/rainerscript.c
@@ -744,11 +744,13 @@ var2Number(struct var *r, int *bSuccess)
 	long long n;
 	if(r->datatype == 'S') {
 		n = es_str2num(r->d.estr, bSuccess);
-	} else if(r->datatype == 'J') {
-		n = (r->d.json == NULL) ? 0 : json_object_get_int(r->d.json);
 	} else {
-		n = r->d.n;
-		if(bSuccess)
+		if(r->datatype == 'J') {
+			n = (r->d.json == NULL) ? 0 : json_object_get_int(r->d.json);
+		} else {
+			n = r->d.n;
+		}
+		if(bSuccess != NULL)
 			*bSuccess = 1;
 	}
 	return n;
@@ -782,6 +784,67 @@ var2String(struct var *r, int *bMustFree)
 	return estr;
 }
 
+static uchar*
+var2CString(struct var *r, int *bMustFree)
+{
+	uchar *cstr;
+	es_str_t *estr;
+	estr = var2String(r, bMustFree);
+	cstr = (uchar*) es_str2cstr(estr, NULL);
+	if(*bMustFree)
+		es_deleteStr(estr);
+	*bMustFree = 1;
+	return cstr;
+}
+
+rsRetVal
+doExtractField(uchar *str, uchar delim, int matchnbr, uchar **resstr)
+{
+	int iCurrFld;
+	int iLen;
+	uchar *pBuf;
+	uchar *pFld;
+	uchar *pFldEnd;
+	DEFiRet;
+
+	/* first, skip to the field in question */
+	iCurrFld = 1;
+	pFld = str;
+	while(*pFld && iCurrFld < matchnbr) {
+		/* skip fields until the requested field or end of string is found */
+		while(*pFld && (uchar) *pFld != delim)
+			++pFld; /* skip to field terminator */
+		if(*pFld == delim) {
+			++pFld; /* eat it */
+			++iCurrFld;
+		}
+	}
+	dbgprintf("field() field requested %d, field found %d\n", matchnbr, iCurrFld);
+	
+	if(iCurrFld == matchnbr) {
+		/* field found, now extract it */
+		/* first of all, we need to find the end */
+		pFldEnd = pFld;
+		while(*pFldEnd && *pFldEnd != delim)
+			++pFldEnd;
+		--pFldEnd; /* we are already at the delimiter - so we need to
+			    * step back a little not to copy it as part of the field. */
+		/* we got our end pointer, now do the copy */
+		iLen = pFldEnd - pFld + 1; /* the +1 is for an actual char, NOT \0! */
+		CHKmalloc(pBuf = MALLOC((iLen + 1) * sizeof(char)));
+		/* now copy */
+		memcpy(pBuf, pFld, iLen);
+		pBuf[iLen] = '\0'; /* terminate it */
+		if(*(pFldEnd+1) != '\0')
+			++pFldEnd; /* OK, skip again over delimiter char */
+		*resstr = pBuf;
+	} else {
+		ABORT_FINALIZE(RS_RET_FIELD_NOT_FOUND);
+	}
+finalize_it:
+	RETiRet;
+}
+
 /* Perform a function call. This has been moved out of cnfExprEval in order
  * to keep the code small and easier to maintain.
  */
@@ -793,8 +856,12 @@ doFuncCall(struct cnffunc *func, struct var *ret, void* usrptr)
 	int bMustFree;
 	es_str_t *estr;
 	char *str;
+	uchar *resStr;
 	int retval;
 	struct var r[CNFFUNC_MAX_ARGS];
+	int delim;
+	int matchnbr;
+	rsRetVal localRet;
 
 	dbgprintf("rainerscript: executing function id %d\n", func->fID);
 	switch(func->fID) {
@@ -862,8 +929,7 @@ doFuncCall(struct cnffunc *func, struct var *ret, void* usrptr)
 		break;
 	case CNFFUNC_RE_MATCH:
 		cnfexprEval(func->expr[0], &r[0], usrptr);
-		estr = var2String(&r[0], &bMustFree);
-		str = es_str2cstr(estr, NULL);
+		str = (char*) var2CString(&r[0], &bMustFree);
 		retval = regexp.regexec(func->funcdata, str, 0, NULL, 0);
 		if(retval == 0)
 			ret->d.n = 1;
@@ -874,10 +940,35 @@ doFuncCall(struct cnffunc *func, struct var *ret, void* usrptr)
 			}
 		}
 		ret->datatype = 'N';
-		if(bMustFree) es_deleteStr(estr);
+		if(bMustFree) free(str);
 		free(str);
 		if(r[0].datatype == 'S') es_deleteStr(r[0].d.estr);
 		break;
+	case CNFFUNC_FIELD:
+		cnfexprEval(func->expr[0], &r[0], usrptr);
+		cnfexprEval(func->expr[1], &r[1], usrptr);
+		cnfexprEval(func->expr[2], &r[2], usrptr);
+		str = (char*) var2CString(&r[0], &bMustFree);
+		delim = var2Number(&r[1], NULL);
+		matchnbr = var2Number(&r[2], NULL);
+		localRet = doExtractField((uchar*)str, (char) delim, matchnbr, &resStr);
+dbgprintf("RRRR: field() returns %d, str: '%s'\n", localRet, resStr);
+		if(localRet == RS_RET_OK) {
+			ret->d.estr = es_newStrFromCStr((char*)resStr, strlen((char*)resStr));
+			free(resStr);
+		} else if(localRet == RS_RET_OK) {
+			ret->d.estr = es_newStrFromCStr("***FIELD NOT FOUND***",
+					sizeof("***FIELD NOT FOUND***")-1);
+		} else {
+			ret->d.estr = es_newStrFromCStr("***ERROR in field() FUNCTION***",
+					sizeof("***ERROR in field() FUNCTION***")-1);
+		}
+		ret->datatype = 'S';
+		if(bMustFree) free(str);
+		if(r[0].datatype == 'S') es_deleteStr(r[0].d.estr);
+		if(r[1].datatype == 'S') es_deleteStr(r[1].d.estr);
+		if(r[2].datatype == 'S') es_deleteStr(r[2].d.estr);
+		break;
 	default:
 		if(Debug) {
 			fname = es_str2cstr(func->fname, NULL);
@@ -1805,6 +1896,13 @@ funcName2ID(es_str_t *fname, unsigned short nParams)
 			return CNFFUNC_INVALID;
 		}
 		return CNFFUNC_RE_MATCH;
+	} else if(!es_strbufcmp(fname, (unsigned char*)"field", sizeof("field") - 1)) {
+		if(nParams != 3) {
+			parser_errmsg("number of parameters for field() must be three "
+				      "but is %d.", nParams);
+			return CNFFUNC_INVALID;
+		}
+		return CNFFUNC_FIELD;
 	} else {
 		return CNFFUNC_INVALID;
 	}
diff --git a/grammar/rainerscript.h b/grammar/rainerscript.h
index f01b1d79..a6693ef0 100644
--- a/grammar/rainerscript.h
+++ b/grammar/rainerscript.h
@@ -207,7 +207,8 @@ enum cnffuncid {
 	CNFFUNC_TOLOWER,
 	CNFFUNC_CSTR,
 	CNFFUNC_CNUM,
-	CNFFUNC_RE_MATCH
+	CNFFUNC_RE_MATCH,
+	CNFFUNC_FIELD
 };
 
 struct cnffunc {
diff --git a/runtime/rsyslog.h b/runtime/rsyslog.h
index ceb277d0..fe9f856a 100644
--- a/runtime/rsyslog.h
+++ b/runtime/rsyslog.h
@@ -392,6 +392,7 @@ enum rsRetVal_				/** return value. All methods return this if not specified oth
 
 	/* RainerScript error messages (range 1000.. 1999) */
 	RS_RET_SYSVAR_NOT_FOUND = 1001, /**< system variable could not be found (maybe misspelled) */
+	RS_RET_FIELD_NOT_FOUND = 1002, /**< field() function did not find requested field */
 
 	/* some generic error/status codes */
 	RS_RET_OK = 0,			/**< operation successful */
diff --git a/tests/Makefile.am b/tests/Makefile.am
index f107a87e..fde9666c 100644
--- a/tests/Makefile.am
+++ b/tests/Makefile.am
@@ -65,6 +65,7 @@ TESTS +=  \
 	failover-no-basic.sh \
 	rcvr_fail_restore.sh \
 	rscript_contains.sh \
+	rscript_field.sh \
 	cee_simple.sh \
 	cee_diskqueue.sh \
 	linkedlistqueue.sh
@@ -269,6 +270,8 @@ EXTRA_DIST= 1.rstest 2.rstest 3.rstest err1.rstest \
 	   testsuites/arrayqueue.conf \
 	   rscript_contains.sh \
 	   testsuites/rscript_contains.conf \
+	   rscript_field.sh \
+	   testsuites/rscript_field.conf \
 	   cee_simple.sh \
 	   testsuites/cee_simple.conf \
 	   cee_diskqueue.sh \
diff --git a/tests/rscript_field.sh b/tests/rscript_field.sh
new file mode 100755
index 00000000..e989e666
--- /dev/null
+++ b/tests/rscript_field.sh
@@ -0,0 +1,13 @@
+# added 2012-09-20 by rgerhards
+# This file is part of the rsyslog project, released under ASL 2.0
+echo ===============================================================================
+echo \[rscript_field.sh\]: testing rainerscript field\(\) function
+source $srcdir/diag.sh init
+source $srcdir/diag.sh startup rscript_field.conf
+source $srcdir/diag.sh injectmsg  0 5000
+echo doing shutdown
+source $srcdir/diag.sh shutdown-when-empty
+echo wait on shutdown
+source $srcdir/diag.sh wait-shutdown 
+source $srcdir/diag.sh seq-check  0 4999
+source $srcdir/diag.sh exit
diff --git a/tests/testsuites/rscript_field.conf b/tests/testsuites/rscript_field.conf
new file mode 100644
index 00000000..d7eb9066
--- /dev/null
+++ b/tests/testsuites/rscript_field.conf
@@ -0,0 +1,11 @@
+$IncludeConfig diag-common.conf
+
+template(name="outfmt" type="list") {
+	property(name="$!usr!msgnum")
+	constant(value="\n")
+}
+
+if $msg contains 'msgnum' then {
+	set $!usr!msgnum = field($msg, 58, 2);
+	action(type="omfile" file="./rsyslog.out.log" template="outfmt")
+}