Revision: 976
http://vcs.pcre.org/viewvc?view=rev&revision=976
Author: ph10
Date: 2012-06-16 18:53:17 +0100 (Sat, 16 Jun 2012)
Log Message:
-----------
Fix capture problem with repeated, empty-string-matching groups.
Modified Paths:
--------------
code/trunk/ChangeLog
code/trunk/pcre_exec.c
code/trunk/testdata/testinput1
code/trunk/testdata/testoutput1
Modified: code/trunk/ChangeLog
===================================================================
--- code/trunk/ChangeLog 2012-06-02 11:03:06 UTC (rev 975)
+++ code/trunk/ChangeLog 2012-06-16 17:53:17 UTC (rev 976)
@@ -126,7 +126,11 @@
35. Improve JIT code generation for greedy plus quantifier.
+36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
+ 1 to "aa" instead of to an empty string. The bug affected repeated groups
+ that could potentially match an empty string.
+
Version 8.30 04-February-2012
-----------------------------
Modified: code/trunk/pcre_exec.c
===================================================================
--- code/trunk/pcre_exec.c 2012-06-02 11:03:06 UTC (rev 975)
+++ code/trunk/pcre_exec.c 2012-06-16 17:53:17 UTC (rev 976)
@@ -37,7 +37,6 @@
-----------------------------------------------------------------------------
*/
-
/* This module contains pcre_exec(), the externally visible function that does
pattern matching using an NFA algorithm, trying to mimic Perl as closely as
possible. There are also some static supporting functions. */
@@ -907,7 +906,6 @@
}
else /* OP_KETRMAX */
{
- md->match_function_type = MATCH_CBEGROUP;
RMATCH(eptr, prev, offset_top, md, eptrb, RM66);
if (rrc != MATCH_NOMATCH) RRETURN(rrc);
ecode += 1 + LINK_SIZE;
@@ -1036,7 +1034,8 @@
for (;;)
{
- if (op >= OP_SBRA || op == OP_ONCE) md->match_function_type = MATCH_CBEGROUP;
+ if (op >= OP_SBRA || op == OP_ONCE)
+ md->match_function_type = MATCH_CBEGROUP;
/* If this is not a possibly empty group, and there are no (*THEN)s in
the pattern, and this is the final alternative, optimize as described
@@ -2009,7 +2008,6 @@
}
if (*prev >= OP_SBRA) /* Could match an empty string */
{
- md->match_function_type = MATCH_CBEGROUP;
RMATCH(eptr, prev, offset_top, md, eptrb, RM50);
RRETURN(rrc);
}
@@ -2018,7 +2016,6 @@
}
else /* OP_KETRMAX */
{
- if (*prev >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
RMATCH(eptr, prev, offset_top, md, eptrb, RM13);
if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;
if (rrc != MATCH_NOMATCH) RRETURN(rrc);
@@ -3431,7 +3428,7 @@
maximizing, find the maximum number of characters and work backwards. */
DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,
- max, eptr));
+ max, (char *)eptr));
if (op >= OP_STARI) /* Caseless */
{
@@ -3700,7 +3697,7 @@
characters and work backwards. */
DPRINTF(("negative matching %c{%d,%d} against subject %.*s\n", fc, min, max,
- max, eptr));
+ max, (char *)eptr));
if (op >= OP_NOTSTARI) /* Caseless */
{
Modified: code/trunk/testdata/testinput1
===================================================================
--- code/trunk/testdata/testinput1 2012-06-02 11:03:06 UTC (rev 975)
+++ code/trunk/testdata/testinput1 2012-06-16 17:53:17 UTC (rev 976)
@@ -5256,4 +5256,10 @@
/(?!a(*COMMIT)b)ac|cd/
ac
+/((?:a?)*)*c/
+ aac
+
+/((?>a?)*)*c/
+ aac
+
/-- End of testinput1 --/
Modified: code/trunk/testdata/testoutput1
===================================================================
--- code/trunk/testdata/testoutput1 2012-06-02 11:03:06 UTC (rev 975)
+++ code/trunk/testdata/testoutput1 2012-06-16 17:53:17 UTC (rev 976)
@@ -8723,4 +8723,14 @@
ac
0: ac
+/((?:a?)*)*c/
+ aac
+ 0: aac
+ 1:
+
+/((?>a?)*)*c/
+ aac
+ 0: aac
+ 1:
+
/-- End of testinput1 --/
