Revision: 1411
http://vcs.pcre.org/viewvc?view=rev&revision=1411
Author: ph10
Date: 2013-12-06 17:11:44 +0000 (Fri, 06 Dec 2013)
Log Message:
-----------
Fix internal error for XCLASS in 16/32-bit non-UCP versions with heap
recursion, and get rid of some compiler warnings when UCP is not configured.
Modified Paths:
--------------
code/trunk/ChangeLog
code/trunk/pcre_compile.c
code/trunk/pcre_exec.c
code/trunk/pcre_jit_compile.c
code/trunk/testdata/testinput17
code/trunk/testdata/testoutput17
Modified: code/trunk/ChangeLog
===================================================================
--- code/trunk/ChangeLog 2013-12-06 16:21:06 UTC (rev 1410)
+++ code/trunk/ChangeLog 2013-12-06 17:11:44 UTC (rev 1411)
@@ -200,6 +200,13 @@
42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
mean "start of word" and "end of word", respectively, as a transition aid.
+
+43. A minimizing repeat of a class containing codepoints greater than 255 in
+ non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
+ compiled to use the heap for recursion.
+
+44. Got rid of some compiler warnings for unused variables when UTF but not UCP
+ is configured.
Version 8.33 28-May-2013
Modified: code/trunk/pcre_compile.c
===================================================================
--- code/trunk/pcre_compile.c 2013-12-06 16:21:06 UTC (rev 1410)
+++ code/trunk/pcre_compile.c 2013-12-06 17:11:44 UTC (rev 1411)
@@ -2860,12 +2860,17 @@
const pcre_uint8 *fcc, pcre_uint32 *list)
{
pcre_uchar c = *code;
+pcre_uchar base;
const pcre_uchar *end;
-const pcre_uint32 *clist_src;
-pcre_uint32 *clist_dest;
pcre_uint32 chr;
-pcre_uchar base;
+#ifdef SUPPORT_UCP
+pcre_uint32 *clist_dest;
+const pcre_uint32 *clist_src;
+#else
+utf = utf; /* Suppress "unused parameter" compiler warning */
+#endif
+
list[0] = c;
list[1] = FALSE;
code++;
Modified: code/trunk/pcre_exec.c
===================================================================
--- code/trunk/pcre_exec.c 2013-12-06 16:21:06 UTC (rev 1410)
+++ code/trunk/pcre_exec.c 2013-12-06 17:11:44 UTC (rev 1411)
@@ -167,7 +167,7 @@
{
PCRE_PUCHAR eptr_start = eptr;
register PCRE_PUCHAR p = md->start_subject + md->offset_vector[offset];
-#ifdef SUPPORT_UTF
+#if defined SUPPORT_UTF && defined SUPPORT_UCP
BOOL utf = md->utf;
#endif
@@ -195,8 +195,7 @@
if (caseless)
{
-#ifdef SUPPORT_UTF
-#ifdef SUPPORT_UCP
+#if defined SUPPORT_UTF && defined SUPPORT_UCP
if (utf)
{
/* Match characters up to the end of the reference. NOTE: the number of
@@ -230,7 +229,6 @@
}
else
#endif
-#endif
/* The same code works when not in UTF-8 mode and in UTF-8 mode when there
is no UCP support. */
@@ -3075,9 +3073,10 @@
/* Control never gets here */
- /* Match an extended character class. This opcode is encountered only
- when UTF-8 mode mode is supported. Nevertheless, we may not be in UTF-8
- mode, because Unicode properties are supported in non-UTF-8 mode. */
+ /* Match an extended character class. In the 8-bit library, this opcode is
+ encountered only when UTF-8 mode mode is supported. In the 16-bit and
+ 32-bit libraries, codepoints greater than 255 may be encountered even when
+ UTF is not supported. */
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
case OP_XCLASS:
@@ -6219,10 +6218,10 @@
LBL(53) LBL(54) LBL(55) LBL(56) LBL(57) LBL(58) LBL(63) LBL(64)
LBL(65) LBL(66)
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
- LBL(21)
+ LBL(20) LBL(21)
#endif
#ifdef SUPPORT_UTF
- LBL(16) LBL(18) LBL(20)
+ LBL(16) LBL(18)
LBL(22) LBL(23) LBL(28) LBL(30)
LBL(32) LBL(34) LBL(42) LBL(46)
#ifdef SUPPORT_UCP
Modified: code/trunk/pcre_jit_compile.c
===================================================================
--- code/trunk/pcre_jit_compile.c 2013-12-06 16:21:06 UTC (rev 1410)
+++ code/trunk/pcre_jit_compile.c 2013-12-06 17:11:44 UTC (rev 1411)
@@ -4023,7 +4023,6 @@
jump_list *found = NULL;
jump_list **list = (*cc & XCL_NOT) == 0 ? &found : backtracks;
pcre_int32 c, charoffset;
-const pcre_uint32 *other_cases;
struct sljit_jump *jump = NULL;
pcre_uchar *ccbegin;
int compares, invertcmp, numberofcmps;
@@ -4032,6 +4031,7 @@
BOOL needstype = FALSE, needsscript = FALSE, needschar = FALSE;
BOOL charsaved = FALSE;
int typereg = TMP1, scriptreg = TMP1;
+const pcre_uint32 *other_cases;
pcre_int32 typeoffset;
#endif
Modified: code/trunk/testdata/testinput17
===================================================================
--- code/trunk/testdata/testinput17 2013-12-06 16:21:06 UTC (rev 1410)
+++ code/trunk/testdata/testinput17 2013-12-06 17:11:44 UTC (rev 1411)
@@ -301,4 +301,7 @@
/a*[b-\x{200}]?a#a*[b-\x{200}]?b#[a-f]*[g-\x{200}]*#[g-\x{200}]*[a-c]*#[g-\x{200}]*[a-h]*/BZ
+/^[\x{1234}\x{4321}]{2,4}?/
+ \x{1234}\x{1234}\x{1234}
+
/-- End of testinput17 --/
Modified: code/trunk/testdata/testoutput17
===================================================================
--- code/trunk/testdata/testoutput17 2013-12-06 16:21:06 UTC (rev 1410)
+++ code/trunk/testdata/testoutput17 2013-12-06 17:11:44 UTC (rev 1411)
@@ -551,4 +551,8 @@
End
------------------------------------------------------------------
+/^[\x{1234}\x{4321}]{2,4}?/
+ \x{1234}\x{1234}\x{1234}
+ 0: \x{1234}\x{1234}
+
/-- End of testinput17 --/
