From owner-svn-src-user@FreeBSD.ORG  Mon Aug 22 23:20:13 2011
Return-Path: <owner-svn-src-user@FreeBSD.ORG>
Delivered-To: svn-src-user@freebsd.org
Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34])
	by hub.freebsd.org (Postfix) with ESMTP id D22BB106566C;
	Mon, 22 Aug 2011 23:20:13 +0000 (UTC)
	(envelope-from gabor@FreeBSD.org)
Received: from svn.freebsd.org (svn.freebsd.org [IPv6:2001:4f8:fff6::2c])
	by mx1.freebsd.org (Postfix) with ESMTP id B755A8FC0C;
	Mon, 22 Aug 2011 23:20:13 +0000 (UTC)
Received: from svn.freebsd.org (localhost [127.0.0.1])
	by svn.freebsd.org (8.14.4/8.14.4) with ESMTP id p7MNKDIQ072566;
	Mon, 22 Aug 2011 23:20:13 GMT (envelope-from gabor@svn.freebsd.org)
Received: (from gabor@localhost)
	by svn.freebsd.org (8.14.4/8.14.4/Submit) id p7MNKDNg072564;
	Mon, 22 Aug 2011 23:20:13 GMT (envelope-from gabor@svn.freebsd.org)
Message-Id: <201108222320.p7MNKDNg072564@svn.freebsd.org>
From: Gabor Kovesdan <gabor@FreeBSD.org>
Date: Mon, 22 Aug 2011 23:20:13 +0000 (UTC)
To: src-committers@freebsd.org, svn-src-user@freebsd.org
X-SVN-Group: user
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Cc: 
Subject: svn commit: r225094 - user/gabor/tre-integration/contrib/tre/lib
X-BeenThere: svn-src-user@freebsd.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "SVN commit messages for the experimental &quot; user&quot;
	src tree" <svn-src-user.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/svn-src-user>,
	<mailto:svn-src-user-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/svn-src-user>
List-Post: <mailto:svn-src-user@freebsd.org>
List-Help: <mailto:svn-src-user-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/svn-src-user>,
	<mailto:svn-src-user-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Mon, 22 Aug 2011 23:20:14 -0000

Author: gabor
Date: Mon Aug 22 23:20:13 2011
New Revision: 225094
URL: http://svn.freebsd.org/changeset/base/225094

Log:
  - With the previous change we can allow more accurate heuristics

Modified:
  user/gabor/tre-integration/contrib/tre/lib/tre-heuristic.c

Modified: user/gabor/tre-integration/contrib/tre/lib/tre-heuristic.c
==============================================================================
--- user/gabor/tre-integration/contrib/tre/lib/tre-heuristic.c	Mon Aug 22 23:19:20 2011	(r225093)
+++ user/gabor/tre-integration/contrib/tre/lib/tre-heuristic.c	Mon Aug 22 23:20:13 2011	(r225094)
@@ -135,57 +135,63 @@ tre_compile_heur(heur_t *h, const tre_ch
 
 	      /*
 	       * If a repetition marker, erases the repeting character
-	       * and terminates the segment.
-	       * Otherwise just terminates the segment (XXX).
+	       * and terminates the segment, otherwise treated as a normal
+	       * character.
 	       */
 	      case TRE_CHAR('{'):
 		PARSE_UNIT('{', '}');
 		if (escaped)
-		  pos--;
-		END_SEGMENT;
+		  {
+		    pos--;
+		    END_SEGMENT;
+		  }
+		heur[pos++] = regex[i];
 		break;
 
 	      /*
-	       * Terminates the current segment whether a subexpression
-	       * marker or not. (XXX)
+	       * Terminates the current segment when escaped,
+	       * otherwise treated as a normal character.
 	       */
 	      case TRE_CHAR('('):
 		PARSE_UNIT('(', ')');
-		END_SEGMENT;
+		if (escaped)
+		  END_SEGMENT;
+		heur[pos++] = regex[i];
 		break;
 
 	      /*
 	       * Sets escaped flag.
-	       * Escaped escape terminates current segment. (XXX)
+	       * Escaped escape is treated as a normal character.
+	       * (This is also the GNU behaviour.)
 	       */
 	      case TRE_CHAR('\\'):
 		if (escaped)
-		  END_SEGMENT;
+		  heur[pos++] = regex[i];
 		escaped = !escaped;
 		continue;
 
 	      /*
-	       * If not the first character, erases the last character
-	       * and terminates the segment.
-	       * Otherwise heuristic construction fails. (XXX)
+	       * If not the first character and not escaped, erases the
+	       * last character and terminates the segment.
+	       * Otherwise treated as a normal character.
 	       */
 	      case TRE_CHAR('*'):
-		if (i != 0)
+		if ((i != 0) && !escaped)
 		  {
 		    pos--;
 		    END_SEGMENT;
 		  }
 		else
-		  goto badpat1;
+		  heur[pos++] = regex[i];
 		break;
 
 	      /*
-	       * If a backreference (escaped digit), terminates segment.
+	       * If escaped, terminates segment.
 	       * Otherwise adds current character to the current segment
 	       * by copying it to the temporary space.
 	       */
 	      default:
-		if (escaped && tre_isdigit(regex[i]))
+		if (escaped)
 		  END_SEGMENT;
 		heur[pos++] = regex[i];
 		continue;
@@ -218,7 +224,7 @@ end_segment:
 	      goto space1;
 	    }
 
-	  ret = tre_compile_fast(h->start, heur, pos, 0);
+	  ret = tre_compile_fast(h->start, heur, pos, _REG_HEUR);
 	  if (ret != REG_OK)
 	    {
 	      errcode = REG_BADPAT;
@@ -248,7 +254,7 @@ end_segment:
 	      goto space2;
 	    }
 	    
-	  ret = tre_compile_fast(h->end, heur, pos, 0);
+	  ret = tre_compile_fast(h->end, heur, pos, _REG_HEUR);
 	  if (ret != REG_OK)
 	    {
 	      xfree(h->end);