From: Ilya Zakharevich <ilya@math.berkeley.edu>
Date: Wed, 8 Jul 1998 01:30:15 +0000 (-0400)
Subject: Switch modifiers in RE off
X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=commitdiff_plain;h=ca9dfc883b7dc3204e375cea5d38c48cb1715f83;p=p5sagit%2Fp5-mst-13.2.git

Switch modifiers in RE off

	Message-Id: <199807080530.BAA14072@monk.mps.ohio-state.edu>

p4raw-id: //depot/perl@1375
---

diff --git a/pod/perlre.pod b/pod/perlre.pod
index f6fdc29..c72a71c 100644
--- a/pod/perlre.pod
+++ b/pod/perlre.pod
@@ -277,6 +277,8 @@ C<)> in the comment.
 
 =item C<(?:pattern)>
 
+=item C<(?imsx-imsx:pattern)>
+
 This is for clustering, not capturing; it groups subexpressions like
 "()", but doesn't make backreferences as "()" does.  So
 
@@ -288,6 +290,15 @@ is like
 
 but doesn't spit out extra fields.
 
+The letters between C<?> and C<:> act as flags modifiers, see
+L<C<(?imsx-imsx)>>.  In particular,
+
+    /(?s-i:more.*than).*million/i
+
+is equivalent to more verbose
+
+    /(?:(?s-i)more.*than).*million/i
+
 =item C<(?=pattern)>
 
 A zero-width positive lookahead assertion.  For example, C</\w+(?=\t)/>
@@ -448,7 +459,7 @@ Say,
 matches a chunk of non-parentheses, possibly included in parentheses
 themselves.
 
-=item C<(?imsx)>
+=item C<(?imsx-imsx)>
 
 One or more embedded pattern-match modifiers.  This is particularly
 useful for patterns that are specified in a table somewhere, some of
@@ -464,6 +475,8 @@ pattern.  For example:
     $pattern = "(?i)foobar";
     if ( /$pattern/ ) { } 
 
+Letters after C<-> switch modifiers off.
+
 These modifiers are localized inside an enclosing group (if any).  Say,
 
     ( (?i) blah ) \s+ \1
diff --git a/regcomp.c b/regcomp.c
index 6292466..663933d 100644
--- a/regcomp.c
+++ b/regcomp.c
@@ -982,6 +982,9 @@ reg(I32 paren, I32 *flagp)
     /* Make an OPEN node, if parenthesized. */
     if (paren) {
 	if (*regcomp_parse == '?') {
+	    U16 posflags = 0, negflags = 0;
+	    U16 *flagsp = &posflags;
+
 	    regcomp_parse++;
 	    paren = *regcomp_parse++;
 	    ret = NULL;			/* For look-ahead/behind. */
@@ -1117,11 +1120,24 @@ reg(I32 paren, I32 *flagp)
                 break;
 	    default:
 		--regcomp_parse;
+	      parse_flags:
 		while (*regcomp_parse && strchr("iogcmsx", *regcomp_parse)) {
 		    if (*regcomp_parse != 'o')
-			pmflag(&regflags, *regcomp_parse);
+			pmflag(flagsp, *regcomp_parse);
+		    ++regcomp_parse;
+		}
+		if (*regcomp_parse == '-') {
+		    flagsp = &negflags;
 		    ++regcomp_parse;
+		    goto parse_flags;
 		}
+		regflags |= posflags;
+		regflags &= ~negflags;
+		if (*regcomp_parse == ':') {
+		    regcomp_parse++;
+		    paren = ':';
+		    break;
+		}		
 	      unknown:
 		if (*regcomp_parse != ')')
 		    FAIL2("Sequence (?%c...) not recognized", *regcomp_parse);
diff --git a/t/op/re_tests b/t/op/re_tests
index 78d89be..7ac20c3 100644
--- a/t/op/re_tests
+++ b/t/op/re_tests
@@ -359,6 +359,35 @@ a(?:b|(c|e){1,2}?|d)+?(.)	ace	y	$1$2	ce
 ((?i)a)b	Ab	y	$&:$1	Ab:A
 (?:(?i)a)b	aB	n	-	-
 ((?i)a)b	aB	n	-	-
+(?i:a)b	ab	y	$&	ab
+((?i:a))b	ab	y	$&:$1	ab:a
+(?i:a)b	Ab	y	$&	Ab
+((?i:a))b	Ab	y	$&:$1	Ab:A
+(?i:a)b	aB	n	-	-
+((?i:a))b	aB	n	-	-
+'(?:(?-i)a)b'i	ab	y	$&	ab
+'((?-i)a)b'i	ab	y	$&:$1	ab:a
+'(?:(?-i)a)b'i	aB	y	$&	aB
+'((?-i)a)b'i	aB	y	$&:$1	aB:a
+'(?:(?-i)a)b'i	Ab	n	-	-
+'((?-i)a)b'i	Ab	n	-	-
+'(?:(?-i)a)b'i	aB	y	$&	aB
+'((?-i)a)b'i	aB	y	$1	a
+'(?:(?-i)a)b'i	AB	n	-	-
+'((?-i)a)b'i	AB	n	-	-
+'(?-i:a)b'i	ab	y	$&	ab
+'((?-i:a))b'i	ab	y	$&:$1	ab:a
+'(?-i:a)b'i	aB	y	$&	aB
+'((?-i:a))b'i	aB	y	$&:$1	aB:a
+'(?-i:a)b'i	Ab	n	-	-
+'((?-i:a))b'i	Ab	n	-	-
+'(?-i:a)b'i	aB	y	$&	aB
+'((?-i:a))b'i	aB	y	$1	a
+'(?-i:a)b'i	AB	n	-	-
+'((?-i:a))b'i	AB	n	-	-
+'((?-i:a.))b'i	a\nB	n	-	-
+'((?s-i:a.))b'i	a\nB	y	$1	a\n
+'((?s-i:a.))b'i	B\nB	n	-	-
 (?:c|d)(?:)(?:a(?:)(?:b)(?:b(?:))(?:b(?:)(?:b)))	cabbbb	y	$&	cabbbb
 (?:c|d)(?:)(?:aaaaaaaa(?:)(?:bbbbbbbb)(?:bbbbbbbb(?:))(?:bbbbbbbb(?:)(?:bbbbbbbb)))	caaaaaaaabbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb	y	$&	caaaaaaaabbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb
 '(ab)\d\1'i	Ab4ab	y	$1	Ab