Add tvb_get_token_len

This is intended to be a replacement for get_token_len (from strutil.h) when its used on a tvb.  It should be a little safer and remove the need for a dissector to use tvb_get_ptr.

Change-Id: Ib2d4a79718b6fba4eb9acc0129b13be6c8199a43
Reviewed-on: https://code.wireshark.org/review/30892
Petri-Dish: Michael Mann <mmann78@netscape.net>
Tested-by: Petri Dish Buildbot
Reviewed-by: Anders Broman <a.broman58@gmail.com>

1 files changed, 67 insertions, 0 deletions

diff --git a/epan/tvbuff.c b/epan/tvbuff.c
index 9d3ca4b82f..110c437437 100644
--- a/epan/tvbuff.c
+++ b/epan/tvbuff.c
@@ -3705,6 +3705,73 @@ tvb_skip_guint8(tvbuff_t *tvb, int offset, const int maxlength, const guint8 ch)
 	return offset;
 }
 
+static ws_mempbrk_pattern pbrk_whitespace;
+
+int tvb_get_token_len(tvbuff_t *tvb, const gint offset, int len, gint *next_offset, const gboolean desegment)
+{
+	gint   eob_offset;
+	gint   eot_offset;
+	int    tokenlen;
+	guchar found_needle = 0;
+	static gboolean compiled = FALSE;
+
+	DISSECTOR_ASSERT(tvb && tvb->initialized);
+
+	if (len == -1) {
+		len = _tvb_captured_length_remaining(tvb, offset);
+		/* if offset is past the end of the tvbuff, len is now 0 */
+	}
+
+	eob_offset = offset + len;
+
+	if (!compiled) {
+		ws_mempbrk_compile(&pbrk_whitespace, " \r\n");
+		compiled = TRUE;
+	}
+
+	/*
+	* Look either for a space, CR, or LF.
+	*/
+	eot_offset = tvb_ws_mempbrk_pattern_guint8(tvb, offset, len, &pbrk_whitespace, &found_needle);
+	if (eot_offset == -1) {
+		/*
+		* No space, CR or LF - token is presumably continued in next packet.
+		*/
+		if (desegment) {
+			/*
+			* Tell our caller we saw no whitespace, so they can
+			* try to desegment and get the entire line
+			* into one tvbuff.
+			*/
+			return -1;
+		}
+		else {
+			/*
+			* Pretend the token runs to the end of the tvbuff.
+			*/
+			tokenlen = eob_offset - offset;
+			if (next_offset)
+				*next_offset = eob_offset;
+		}
+	}
+	else {
+		/*
+		* Find the number of bytes between the starting offset
+		* and the space, CR or LF.
+		*/
+		tokenlen = eot_offset - offset;
+
+		/*
+		* Return the offset of the character after the last
+		* character in the line, skipping over the last character
+		* in the line terminator.
+		*/
+		if (next_offset)
+			*next_offset = eot_offset + 1;
+	}
+	return tokenlen;
+}
+
 /*
  * Format a bunch of data from a tvbuff as bytes, returning a pointer
  * to the string with the formatted data, with "punct" as a byte