Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 25 May 2016 15:42:39 +0000 (UTC)
From:      "Pedro F. Giffuni" <pfg@FreeBSD.org>
To:        src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org
Subject:   svn commit: r300684 - head/usr.bin/sed
Message-ID:  <201605251542.u4PFgdRG076996@repo.freebsd.org>

next in thread | raw e-mail | index | archive | help
Author: pfg
Date: Wed May 25 15:42:39 2016
New Revision: 300684
URL: https://svnweb.freebsd.org/changeset/base/300684

Log:
  sed: convert sed to use REG_STARTEND more explicitly.
  
  Summarizing the findings in the OpenBSD list:
  
  This solves a reproduceable issue with very recent Mesa where REG_NOTBOL
  combined with a match at the begin of the string causes our regex library
  to treat the word as not begin of word.
  
  Thanks to Martijn van Duren and Ingo Schwarze for taking the time to
  solve this in the least invasive way.
  
  PR:		209352, 209387
  Taken from:     openbsd-tech (Martijn van Duren)
  MFC after:	1 month

Modified:
  head/usr.bin/sed/process.c

Modified: head/usr.bin/sed/process.c
==============================================================================
--- head/usr.bin/sed/process.c	Wed May 25 15:35:23 2016	(r300683)
+++ head/usr.bin/sed/process.c	Wed May 25 15:42:39 2016	(r300684)
@@ -71,7 +71,8 @@ static inline int	 applies(struct s_comm
 static void		 do_tr(struct s_tr *);
 static void		 flush_appends(void);
 static void		 lputs(char *, size_t);
-static int		 regexec_e(regex_t *, const char *, int, int, size_t);
+static int		 regexec_e(regex_t *, const char *, int, int, size_t,
+			     size_t);
 static void		 regsub(SPACE *, char *, char *);
 static int		 substitute(struct s_command *);
 
@@ -281,7 +282,7 @@ new:		if (!nflag && !pd)
  * (lastline, linenumber, ps).
  */
 #define	MATCH(a)							\
-	((a)->type == AT_RE ? regexec_e((a)->u.r, ps, 0, 1, psl) :	\
+	((a)->type == AT_RE ? regexec_e((a)->u.r, ps, 0, 1, 0, psl) :	\
 	    (a)->type == AT_LINE ? linenum == (a)->u.l : lastline())
 
 /*
@@ -381,6 +382,7 @@ substitute(struct s_command *cp)
 	regex_t *re;
 	regoff_t slen;
 	int lastempty, n;
+	size_t le = 0;
 	char *s;
 
 	s = ps;
@@ -392,7 +394,7 @@ substitute(struct s_command *cp)
 					linenum, fname, cp->u.s->maxbref);
 		}
 	}
-	if (!regexec_e(re, s, 0, 0, psl))
+	if (!regexec_e(re, s, 0, 0, 0, psl))
 		return (0);
 
 	SS.len = 0;				/* Clean substitute space. */
@@ -402,28 +404,30 @@ substitute(struct s_command *cp)
 
 	do {
 		/* Copy the leading retained string. */
-		if (n <= 1 && match[0].rm_so)
-			cspace(&SS, s, match[0].rm_so, APPEND);
+		if (n <= 1 && match[0].rm_so - le)
+			cspace(&SS, s, match[0].rm_so - le, APPEND);
 
 		/* Skip zero-length matches right after other matches. */
-		if (lastempty || match[0].rm_so ||
+		if (lastempty || (match[0].rm_so - le) ||
 		    match[0].rm_so != match[0].rm_eo) {
 			if (n <= 1) {
 				/* Want this match: append replacement. */
-				regsub(&SS, s, cp->u.s->new);
+				regsub(&SS, ps, cp->u.s->new);
 				if (n == 1)
 					n = -1;
 			} else {
 				/* Want a later match: append original. */
-				if (match[0].rm_eo)
-					cspace(&SS, s, match[0].rm_eo, APPEND);
+				if (match[0].rm_eo - le)
+					cspace(&SS, s, match[0].rm_eo - le,
+					    APPEND);
 				n--;
 			}
 		}
 
 		/* Move past this match. */
-		s += match[0].rm_eo;
-		slen -= match[0].rm_eo;
+		s += (match[0].rm_eo - le);
+		slen -= (match[0].rm_eo - le);
+		le = match[0].rm_eo;
 
 		/*
 		 * After a zero-length match, advance one byte,
@@ -434,13 +438,15 @@ substitute(struct s_command *cp)
 				slen = -1;
 			else
 				slen--;
-			if (*s != '\0')
+			if (*s != '\0') {
 			 	cspace(&SS, s++, 1, APPEND);
+				le++;
+			}
 			lastempty = 1;
 		} else
 			lastempty = 0;
 
-	} while (n >= 0 && slen >= 0 && regexec_e(re, s, REG_NOTBOL, 0, slen));
+	} while (n >= 0 && slen >= 0 && regexec_e(re, ps, 0, 0, le, psl));
 
 	/* Did not find the requested number of matches. */
 	if (n > 1)
@@ -652,7 +658,7 @@ lputs(char *s, size_t len)
 
 static int
 regexec_e(regex_t *preg, const char *string, int eflags, int nomatch,
-	size_t slen)
+	size_t start, size_t stop)
 {
 	int eval;
 
@@ -663,8 +669,8 @@ regexec_e(regex_t *preg, const char *str
 		defpreg = preg;
 
 	/* Set anchors */
-	match[0].rm_so = 0;
-	match[0].rm_eo = slen;
+	match[0].rm_so = start;
+	match[0].rm_eo = stop;
 
 	eval = regexec(defpreg, string,
 	    nomatch ? 0 : maxnsub + 1, match, eflags | REG_STARTEND);



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201605251542.u4PFgdRG076996>