X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=blobdiff_plain;f=regcomp.h;h=2ac1be11e995394266a2fc276249e4d1a0e8a601;hb=1bc4b319ba6d50bfdf5332d4378c85af1205184b;hp=8f14a200c157d45d741f800ed69309e3b355b225;hpb=f7819f85d6af0dcf9114284a4fe5ef21855e4e5a;p=p5sagit%2Fp5-mst-13.2.git

diff --git a/regcomp.h b/regcomp.h
index 8f14a20..2ac1be1 100644
--- a/regcomp.h
+++ b/regcomp.h
@@ -1,7 +1,7 @@
 /*    regcomp.h
  *
  *    Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
- *    2000, 2001, 2002, 2003, 2005, 2006 by Larry Wall and others
+ *    2000, 2001, 2002, 2003, 2005, 2006, 2007, by Larry Wall and others
  *
  *    You may distribute under the terms of either the GNU General Public
  *    License or the Artistic License, as specified in the README file.
@@ -18,6 +18,24 @@ typedef OP OP_4tree;			/* Will be redefined later. */
 /* Be really agressive about optimising patterns with trie sequences? */
 #define PERL_ENABLE_EXTENDED_TRIE_OPTIMISATION 1
 
+/* Use old style unicode mappings for perl and posix character classes
+ *
+ * NOTE: Enabling this essentially breaks character class matching against unicode 
+ * strings, so that POSIX char classes match when they shouldn't, and \d matches 
+ * way more than 10 characters, and sometimes a charclass and its complement either
+ * both match or neither match.
+ * NOTE: Disabling this will cause various backwards compatibility issues to rear 
+ * their head, and tests to fail. However it will make the charclass behaviour 
+ * consistant regardless of internal string type, and make character class inversions
+ * consistant. The tests that fail in the regex engine are basically broken tests.
+ *
+ * Personally I think 5.12 should disable this for sure. Its a bit more debatable for
+ * 5.10, so for now im leaving it enabled.
+ *
+ * -demerphq
+ */
+#define PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS 1
+
 /* Should the optimiser take positive assertions into account? */
 #define PERL_ENABLE_POSITIVE_ASSERTION_STUDY 0
 
@@ -317,9 +335,9 @@ struct regnode_charclass_class {	/* has [[:blah:]] classes */
 #define ANYOF_NALNUM	 1
 #define ANYOF_SPACE	 2	/* \s */
 #define ANYOF_NSPACE	 3
-#define ANYOF_DIGIT	 4
+#define ANYOF_DIGIT	 4	/* \d */
 #define ANYOF_NDIGIT	 5
-#define ANYOF_ALNUMC	 6	/* isalnum(3), utf8::IsAlnum, ALNUMC */
+#define ANYOF_ALNUMC	 6	/* [[:alnum:]] isalnum(3), utf8::IsAlnum, ALNUMC */
 #define ANYOF_NALNUMC	 7
 #define ANYOF_ALPHA	 8
 #define ANYOF_NALPHA	 9
@@ -528,7 +546,10 @@ struct reg_data {
 #define check_offset_max substrs->data[2].max_offset
 #define check_end_shift substrs->data[2].end_shift
 
-
+#define RX_ANCHORED_SUBSTR(rx)	(((struct regexp *)SvANY(rx))->anchored_substr)
+#define RX_ANCHORED_UTF8(rx)	(((struct regexp *)SvANY(rx))->anchored_utf8)
+#define RX_FLOAT_SUBSTR(rx)	(((struct regexp *)SvANY(rx))->float_substr)
+#define RX_FLOAT_UTF8(rx)	(((struct regexp *)SvANY(rx))->float_utf8)
 
 /* trie related stuff */
 
@@ -796,7 +817,7 @@ re.pm, especially to the documentation.
     const char * const rpv =                                       \
         pv_pretty((dsv), (pv), (l), (m), \
             PL_colors[0], PL_colors[1], \
-            ( PERL_PV_PRETTY_QUOTE | PERL_PV_ESCAPE_RE | PERL_PV_PRETTY_ELIPSES |      \
+            ( PERL_PV_PRETTY_QUOTE | PERL_PV_ESCAPE_RE | PERL_PV_PRETTY_ELLIPSES | \
               ((isuni) ? PERL_PV_ESCAPE_UNI : 0))                  \
         )
 
@@ -814,5 +835,12 @@ re.pm, especially to the documentation.
 
 #endif /* DEBUG RELATED DEFINES */
 
-
-
+/*
+ * Local variables:
+ * c-indentation-style: bsd
+ * c-basic-offset: 4
+ * indent-tabs-mode: t
+ * End:
+ *
+ * ex: set ts=8 sts=4 sw=4 noet:
+ */