Merge branch 'v4-beta' into v4-devel

3 files changed, 35 insertions, 4 deletions

diff --git a/ChangeLog b/ChangeLog
index cb6ba7a1..df0b69bf 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,11 @@
 ---------------------------------------------------------------------------
 Version 4.5.3  [v4-beta] (rgerhards), 2009-08-??
+- bugfix: message sanitation had some issues:
+  - control character DEL was not properly escaped
+  - NUL and LF characters were not properly stripped if no control
+    character replacement was to be done
+  - NUL characters in the message body were silently dropped (this was
+    a regeression introduced by some of the recent optimizations)
 - bugfix: strings improperly reused, resulting in some message properties
   be populated with strings from previous messages. This was caused by
   an improper predicate check. [backported from v5]
diff --git a/runtime/msg.h b/runtime/msg.h
index 0b346f7b..3a02365b 100644
--- a/runtime/msg.h
+++ b/runtime/msg.h
@@ -192,6 +192,23 @@ uchar *propIDToName(propid_t propID);
 extern void (*funcMsgPrepareEnqueue)(msg_t *pMsg);
 #define MsgPrepareEnqueue(pMsg) funcMsgPrepareEnqueue(pMsg)
 
+
+/* ------------------------------ some inline functions ------------------------------ */
+
+/* set raw message size. This is needed in some cases where a trunctation is necessary
+ * but the raw message must not be newly set. The most important (and currently only)
+ * use case is if we remove trailing LF or NUL characters. Note that the size can NOT
+ * be extended, only shrunk!
+ * rgerhards, 2009-08-26
+ */
+static inline void
+MsgSetRawMsgSize(msg_t *pMsg, size_t newLen)
+{
+	assert(newLen <= (size_t) pMsg->iLenRawMsg);
+	pMsg->iLenRawMsg = newLen;
+}
+
+
 #endif /* #ifndef MSG_H_INCLUDED */
 /* vim:set ai:
  */
diff --git a/runtime/parser.c b/runtime/parser.c
index a2538fa3..db11ac5b 100644
--- a/runtime/parser.c
+++ b/runtime/parser.c
@@ -163,6 +163,7 @@ sanitizeMessage(msg_t *pMsg)
 	size_t iDst;
 	size_t iMaxLine;
 	size_t maxDest;
+	bool bUpdatedLen = FALSE;
 	uchar szSanBuf[32*1024]; /* buffer used for sanitizing a string */
 
 	assert(pMsg != NULL);
@@ -177,6 +178,7 @@ sanitizeMessage(msg_t *pMsg)
 	/* remove NUL character at end of message (see comment in function header) */
 	if(pszMsg[lenMsg-1] == '\0') {
 		DBGPRINTF("dropped NUL at very end of message\n");
+		bUpdatedLen = TRUE;
 		lenMsg--;
 	}
 
@@ -187,6 +189,7 @@ sanitizeMessage(msg_t *pMsg)
 	 */
 	if(bDropTrailingLF && pszMsg[lenMsg-1] == '\n') {
 		DBGPRINTF("dropped LF at very end of message (DropTrailingLF is set)\n");
+		bUpdatedLen = TRUE;
 		lenMsg--;
 	}
 
@@ -197,7 +200,7 @@ sanitizeMessage(msg_t *pMsg)
 	 */
 	int bNeedSanitize = 0;
 	for(iSrc = 0 ; iSrc < lenMsg ; iSrc++) {
-		if(pszMsg[iSrc] < 32) {
+		if(iscntrl(pszMsg[iSrc])) {
 			if(pszMsg[iSrc] == '\0' || bEscapeCCOnRcv) {
 				bNeedSanitize = 1;
 				break;
@@ -205,8 +208,11 @@ sanitizeMessage(msg_t *pMsg)
 		}
 	}
 
-	if(!bNeedSanitize)
+	if(!bNeedSanitize) {
+		if(bUpdatedLen == TRUE)
+			MsgSetRawMsgSize(pMsg, lenMsg);
 		FINALIZE;
+	}
 
 	/* now copy over the message and sanitize it */
 	iMaxLine = glbl.GetMaxLine();
@@ -219,8 +225,10 @@ sanitizeMessage(msg_t *pMsg)
 		CHKmalloc(pDst = malloc(sizeof(uchar) * (iMaxLine + 1)));
 	iSrc = iDst = 0;
 	while(iSrc < lenMsg && iDst < maxDest - 3) { /* leave some space if last char must be escaped */
-		if(pszMsg[iSrc] == '\0') { /* guard against \0 characters... */
-		} else if(iscntrl((int) pszMsg[iSrc])) {
+		if(iscntrl((int) pszMsg[iSrc])) {
+			/* note: \0 must always be escaped, the rest of the code currently
+			 * can not handle it! -- rgerhards, 2009-08-26
+			 */
 			if(pszMsg[iSrc] == '\0' || bEscapeCCOnRcv) {
 			/* we are configured to escape control characters. Please note
 			 * that this most probably break non-western character sets like