[perl #23769] Unicode regex broken on simple example

[perl5.git] / t / op / pat.t

diff --git a/t/op/pat.t b/t/op/pat.t
index 360892b..54f67fc 100755 (executable)
--- a/t/op/pat.t
+++ b/t/op/pat.t
@@ -6,7 +6,7 @@
 
 $| = 1;
 
-print "1..986\n";
+print "1..1015\n";
 
 BEGIN {
     chdir 't' if -d 't';
@@ -20,9 +20,8 @@ $x = "abc\ndef\n";
 if ($x =~ /^abc/) {print "ok 1\n";} else {print "not ok 1\n";}
 if ($x !~ /^def/) {print "ok 2\n";} else {print "not ok 2\n";}
 
-$* = 1;
-if ($x =~ /^def/) {print "ok 3\n";} else {print "not ok 3\n";}
-$* = 0;
+# used to be a test for $*
+if ($x =~ /^def/m) {print "ok 3\n";} else {print "not ok 3\n";}
 
 $_ = '123';
 if (/^([0-9][0-9]*)/) {print "ok 4\n";} else {print "not ok 4\n";}
@@ -69,9 +68,8 @@ if (m|bc/*d|) {print "ok 22\n";} else {print "not ok 22\n";}
 
 if (/^$_$/) {print "ok 23\n";} else {print "not ok 23\n";}
 
-$* = 1;                # test 3 only tested the optimized version--this one is for real
-if ("ab\ncd\n" =~ /^cd/) {print "ok 24\n";} else {print "not ok 24\n";}
-$* = 0;
+# used to be a test for $*
+if ("ab\ncd\n" =~ /^cd/m) {print "ok 24\n";} else {print "not ok 24\n";}
 
 $XXX{123} = 123;
 $XXX{234} = 234;
@@ -1367,10 +1365,10 @@ print "ok 247\n";
     print "ok 263\n";
 }
 
-{
+SKIP: {
     my $test = 264; # till 575
 
-    use charnames ':full';
+    use charnames ":full";
 
     # This is far from complete testing, there are dozens of character
     # classes in Unicode.  The mixing of literals and \N{...} is
@@ -3090,6 +3088,140 @@ ok("bbbbac" =~ /$pattern/ && $1 eq 'a', "[perl #3547]");
     ok( $1 ne "\xC4\x80", '$1 is not non-utf-8' );
 }
 
-# last test 984
+{
+    use utf8;
+    my $attr = 'Name-1' ;
+
+    my $NormalChar          = qr/[\p{IsDigit}\p{IsLower}\p{IsUpper}]/;
+    my $NormalWord          = qr/${NormalChar}+?/;
+    my $PredNameHyphen      = qr/^${NormalWord}(\-${NormalWord})*?$/;
+
+    $attr =~ /^$/;
+    ok( $attr =~ $PredNameHyphen, "[perl #19767] original test" );
+}
+
+{
+    use utf8;
+    "a" =~ m/[b]/;
+    ok ( "0" =~ /\p{N}+\z/, "[perl #19767] variant test" );
+}
+
+{
+
+    $p = 1;
+    foreach (1,2,3,4) {
+           $p++ if /(??{ $p })/
+    }
+    ok ($p == 5, "[perl #20683] (??{ }) returns stale values");
+    { package P; $a=1; sub TIESCALAR { bless[] } sub FETCH { $a++ } }
+    tie $p, P;
+    foreach (1,2,3,4) {
+           /(??{ $p })/
+    }
+    ok ( $p == 5, "(??{ }) returns stale values");
+}
+
+{
+  # Subject: Odd regexp behavior
+  # From: Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>
+  # Date: Wed, 26 Feb 2003 16:53:12 +0000
+  # Message-Id: <E18o4nw-0008Ly-00@wisbech.cl.cam.ac.uk>
+  # To: perl-unicode@perl.org
+    
+  $x = "\x{2019}\nk"; $x =~ s/(\S)\n(\S)/$1 $2/sg;
+  ok($x eq "\x{2019} k", "Markus Kuhn 2003-02-26");
+
+  $x = "b\nk"; $x =~ s/(\S)\n(\S)/$1 $2/sg;
+  ok($x eq "b k", "Markus Kuhn 2003-02-26");
+
+  ok("\x{2019}" =~ /\S/, "Markus Kuhn 2003-02-26");
+}
+
+{
+    my $i;
+    ok('-1-3-5-' eq join('', split /((??{$i++}))/, '-1-3-5-'),
+       "[perl #21411] (??{ .. }) corrupts split's stack");
+    split /(?{'WOW'})/, 'abc';
+    ok('a|b|c' eq join ('|', @_),
+       "[perl #21411] (?{ .. }) version of the above");
+}
+
+{
+    split /(?{ split "" })/, "abc";
+    ok(1,'cache_re & "(?{": it dumps core in 5.6.1 & 5.8.0');
+}
+
+{
+    ok("\x{100}\n" =~ /\x{100}\n$/, "UTF8 length cache and fbm_compile");  
+}
+
+{
+    package Str;
+    use overload q/""/ => sub { ${$_[0]}; };
+    sub new { my ($c, $v) = @_; bless \$v, $c; }
+
+    package main;
+    $_ = Str->new("a\x{100}/\x{100}b");
+    ok(join(":", /\b(.)\x{100}/g) eq "a:/", "re_intuit_start and PL_bostr");
+}
+
+{
+    $_ = "code:   'x' { '...' }\n"; study;
+    my @x; push @x, $& while m/'[^\']*'/gx;
+    ok(join(":", @x) eq "'x':'...'",
+       "[perl #17757] Parse::RecDescent triggers infinite loop");
+}
+
+{
+    my $re = qq/^([^X]*)X/;
+    utf8::upgrade($re);
+    ok("\x{100}X" =~ /$re/, "S_cl_and ANYOF_UNICODE & ANYOF_INVERTED");
+}
+
+# bug #22354
+sub func ($) {
+    ok( "a\nb" !~ /^b/, $_[0] );
+    ok( "a\nb" =~ /^b/m, "$_[0] - with /m" );
+}
+func "standalone";
+$_ = "x"; s/x/func "in subst"/e;
+$_ = "x"; s/x/func "in multiline subst"/em;
+#$_ = "x"; /x(?{func "in regexp"})/;
+#$_ = "x"; /x(?{func "in multiline regexp"})/m;
+
+# bug #19049
+$_="abcdef\n";
+@x = m/./g;
+ok("abcde" eq "$`", '# TODO #19049 - global match not setting $`');
+
+ok("123\x{100}" =~ /^.*1.*23\x{100}$/, 'uft8 + multiple floating substr');
+
+# LATIN SMALL/CAPITAL LETTER A WITH MACRON
+ok("  \x{101}" =~ qr/\x{100}/i,
+   "<20030808193656.5109.1@llama.ni-s.u-net.com>");
+
+# LATIN SMALL/CAPITAL LETTER A WITH RING BELOW
+ok("  \x{1E01}" =~ qr/\x{1E00}/i,
+   "<20030808193656.5109.1@llama.ni-s.u-net.com>");
+
+# DESERET SMALL/CAPITAL LETTER LONG I
+ok("  \x{10428}" =~ qr/\x{10400}/i,
+   "<20030808193656.5109.1@llama.ni-s.u-net.com>");
+
+# LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X'
+ok("  \x{1E01}x" =~ qr/\x{1E00}X/i,
+   "<20030808193656.5109.1@llama.ni-s.u-net.com>");
+
+{
+    # [perl #23769] Unicode regex broken on simple example
+    # regrepeat() didn't handle UTF-8 EXACT case right.
+
+    my $s = "\x{a0}\x{a0}\x{a0}\x{100}"; chop $s;
+
+    ok($s =~ /\x{a0}/,       "[perl #23769]");
+    ok($s =~ /\x{a0}+/,      "[perl #23769]");
+    ok($s =~ /\x{a0}\x{a0}/, "[perl #23769]");
+}
 
+# last test 1015