From a8485b3c26308b4ebd260fc40b9de81e56e8b5eb Mon Sep 17 00:00:00 2001
From: Tim Waugh <twaugh@fedoraproject.org>
Date: Mon, 20 Feb 2006 14:36:53 +0000
Subject: [PATCH] - Applied Tim Robbins' patch for 'grep -w' (bug #179698).

---
 grep-2.5.1-w.patch | 72 ++++++++++++++++++++++++----------------------
 grep.spec          |  5 +++-
 2 files changed, 42 insertions(+), 35 deletions(-)

diff --git a/grep-2.5.1-w.patch b/grep-2.5.1-w.patch
index e4d6ea2..79ae2ae 100644
--- a/grep-2.5.1-w.patch
+++ b/grep-2.5.1-w.patch
@@ -1,32 +1,6 @@
---- grep-2.5.1a/src/search.c.w	2005-01-07 15:04:18.766280754 +0000
-+++ grep-2.5.1a/src/search.c	2005-01-07 16:59:19.287275172 +0000
-@@ -330,6 +330,7 @@
-   static int use_dfa;
-   static int use_dfa_checked = 0;
- #ifdef MBS_SUPPORT
-+  const char *last_char = NULL;
-   int mb_cur_max = MB_CUR_MAX;
-   mbstate_t mbs;
-   memset (&mbs, '\0', sizeof (mbstate_t));
-@@ -385,6 +386,8 @@
- 		  while (bytes_left)
- 		    {
- 		      size_t mlen = mbrlen (beg, bytes_left, &mbs);
-+
-+		      last_char = beg;
- 		      if (mlen == (size_t) -1 || mlen == 0)
- 			{
- 			  /* Incomplete character: treat as single-byte. */
-@@ -445,6 +448,8 @@
- 		  while (bytes_left)
- 		    {
- 		      size_t mlen = mbrlen (beg, bytes_left, &mbs);
-+
-+		      last_char = beg;
- 		      if (mlen == (size_t) -1 || mlen == 0)
- 			{
- 			  /* Incomplete character: treat as single-byte. */
-@@ -507,10 +512,84 @@
+--- grep-2.5.1a/src/search.c.w	2006-02-20 14:27:27.000000000 +0000
++++ grep-2.5.1a/src/search.c	2006-02-20 14:32:07.000000000 +0000
+@@ -507,10 +507,114 @@
  	      if (match_words)
  		while (start >= 0)
  		  {
@@ -44,11 +18,16 @@
 +			if (mb_cur_max > 1)
 +			  {
 +			    const char *s;
-+			    int mr;
++			    size_t mr;
 +			    wchar_t pwc;
 +
++			    /* Locate the start of the multibyte character
++			       before the match position (== beg + start). */
 +			    if (using_utf8)
 +			      {
++				/* UTF-8 is a special case: scan backwards
++				   until we find a 7-bit character or a
++				   lead byte. */
 +				s = beg + start - 1;
 +				while (s > buf
 +				       && (unsigned char) *s >= 0x80
@@ -56,15 +35,40 @@
 +				  --s;
 +			      }
 +			    else
-+			      s = last_char;
-+			    mr = mbtowc (&pwc, s, beg + start - s);
-+			    if (mr <= 0)
++			      {
++				/* Scan forwards to find the start of the
++				   last complete character before the
++				   match position.  */
++				size_t bytes_left = start - 1;
++				s = beg;
++				while (bytes_left > 0)
++				  {
++				    mr = mbrlen (s, bytes_left, &mbs);
++				    if (mr == (size_t) -1 || mr == 0)
++				      {
++					memset (&mbs, '\0', sizeof (mbs));
++					s++;
++					bytes_left--;
++					continue;
++				      }
++				    if (mr == (size_t) -2)
++				      {
++					memset (&mbs, '\0', sizeof (mbs));
++					break;
++				      }
++				    s += mr;
++				    bytes_left -= mr;
++				  }
++			      }
++			    mr = mbrtowc (&pwc, s, beg + start - s, &mbs);
++			    if (mr == (size_t) -2 || mr == (size_t) -1 ||
++				mr == 0)
 +			      {
 +				memset (&mbs, '\0', sizeof (mbstate_t));
 +				lword_match = 1;
 +			      }
 +			    else if (!(iswalnum (pwc) || pwc == L'_')
-+				     && mr == (int) (beg + start - s))
++				     && mr == beg + start - s)
 +			      lword_match = 1;
 +			  }
 +			else
diff --git a/grep.spec b/grep.spec
index 972cfde..8925184 100644
--- a/grep.spec
+++ b/grep.spec
@@ -1,7 +1,7 @@
 Summary: The GNU versions of grep pattern matching utilities.
 Name: grep
 Version: 2.5.1
-Release: 52.2
+Release: 53
 License: GPL
 Group: Applications/Text
 Source: ftp://ftp.gnu.org/pub/gnu/grep/grep-%{version}a.tar.bz2
@@ -94,6 +94,9 @@ fi
 %{_mandir}/*/*
 
 %changelog
+* Mon Feb 20 2006 Tim Waugh <twaugh@redhat.com> 2.5.1-53
+- Applied Tim Robbins' patch for 'grep -w' (bug #179698).
+
 * Fri Feb 10 2006 Jesse Keating <jkeating@redhat.com> - 2.5.1-52.2
 - bump again for double-long bug on ppc(64)