Fix [RT#66098] -- stricter checking on SvIVX exposed a lack of SvIOK check

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

diff --git a/t/op/split.t b/t/op/split.t
index 55b2839..b3a9741 100755 (executable)
--- a/t/op/split.t
+++ b/t/op/split.t
@@ -6,7 +6,7 @@ BEGIN {
     require './test.pl';
 }
 
-plan tests => 52;
+plan tests => 136;
 
 $FS = ':';
 
@@ -50,12 +50,8 @@ $_ = join(':', split(/:/,'1:2:3:4:5:6:::', 999));
 is($_ , '1:2:3:4:5:6:::');
 
 # Does assignment to a list imply split to one more field than that?
-if ($^O eq 'MSWin32') { $foo = `.\\perl -D1024 -e "(\$a,\$b) = split;" 2>&1` }
-elsif ($^O eq 'NetWare') { $foo = `perl -D1024 -e "(\$a,\$b) = split;" 2>&1` }
-elsif ($^O eq 'VMS')  { $foo = `./perl "-D1024" -e "(\$a,\$b) = split;" 2>&1` }
-elsif ($^O eq 'MacOS'){ $foo = `$^X "-D1024" -e "(\$a,\$b) = split;"` }
-else                  { $foo = `./perl -D1024 -e '(\$a,\$b) = split;' 2>&1` }
-ok($foo =~ /DEBUGGING/ || $foo =~ /SV = (VOID|IV\(3\))/);
+$foo = runperl( switches => ['-Dt'], stderr => 1, prog => '($a,$b)=split;' );
+ok($foo =~ /DEBUGGING/ || $foo =~ /const\n?\Q(IV(3))\E/);
 
 # Can we say how many fields to split to when assigning to a list?
 ($a,$b) = split(' ','1 2 3 4 5 6', 2);
@@ -279,6 +275,92 @@ ok(@ary == 3 &&
 {
     $p="a,b";
     utf8::upgrade $p;
-    @a=split(/[, ]+/,$p);
+    eval { @a=split(/[, ]+/,$p) };
     is ("$@-@a-", '-a b-', '#20912 - split() to array with /[]+/ and utf8');
 }
+
+{
+    is (\@a, \@{"a"}, '@a must be global for following test');
+    $p="";
+    $n = @a = split /,/,$p;
+    is ($n, 0, '#21765 - pmreplroot hack used to return undef for 0 iters');
+}
+
+{
+    # [perl #28938]
+    # assigning off the end of the array after a split could leave garbage
+    # in the inner elements
+
+    my $x;
+    @a = split /,/, ',,,,,';
+    $a[3]=1;
+    $x = \$a[2];
+    is (ref $x, 'SCALAR', '#28938 - garbage after extend');
+}
+{
+    # check the special casing of split /\s/ and unicode
+    use charnames qw(:full);
+    # below test data is extracted from
+    # PropList-5.0.0.txt
+    # Date: 2006-06-07, 23:22:52 GMT [MD]
+    #
+    # Unicode Character Database
+    # Copyright (c) 1991-2006 Unicode, Inc.
+    # For terms of use, see http://www.unicode.org/terms_of_use.html
+    # For documentation, see UCD.html
+    my @spaces=(
+       ord("\t"),      # Cc       <control-0009>
+       ord("\n"),      # Cc       <control-000A>
+       # not PerlSpace # Cc       <control-000B>
+       ord("\f"),      # Cc       <control-000C>
+       ord("\r"),      # Cc       <control-000D>
+       ord(" "),       # Zs       SPACE
+       ord("\N{NEL}"), # Cc       <control-0085>
+       ord("\N{NO-BREAK SPACE}"),
+                       # Zs       NO-BREAK SPACE
+        0x1680,         # Zs       OGHAM SPACE MARK
+        0x180E,         # Zs       MONGOLIAN VOWEL SEPARATOR
+        0x2000..0x200A, # Zs  [11] EN QUAD..HAIR SPACE
+        0x2028,         # Zl       LINE SEPARATOR
+        0x2029,         # Zp       PARAGRAPH SEPARATOR
+        0x202F,         # Zs       NARROW NO-BREAK SPACE
+        0x205F,         # Zs       MEDIUM MATHEMATICAL SPACE
+        0x3000          # Zs       IDEOGRAPHIC SPACE
+    );
+    #diag "Have @{[0+@spaces]} to test\n";
+    foreach my $cp (@spaces) {
+       my $msg = sprintf "Space: U+%04x", $cp;
+        my $space = chr($cp);
+        my $str="A:$space:B\x{FFFD}";
+        chop $str;
+
+        my @res=split(/\s+/,$str);
+        ok(@res == 2 && join('-',@res) eq "A:-:B", "$msg - /\\s+/");
+
+        my $s2 = "$space$space:A:$space$space:B\x{FFFD}";
+        chop $s2;
+
+        my @r2 = split(' ',$s2);
+        ok(@r2 == 2 && join('-', @r2) eq ":A:-:B",  "$msg - ' '");
+
+        my @r3 = split(/\s+/, $s2);
+        ok(@r3 == 3 && join('-', @r3) eq "-:A:-:B", "$msg - /\\s+/ No.2");
+    }
+}
+
+{
+    my $src = "ABC \0 FOO \0  XYZ";
+    my @s = split(" \0 ", $src);
+    my @r = split(/ \0 /, $src);
+    is(scalar(@s), 3);
+    is($s[0], "ABC");
+    is($s[1], "FOO");
+    is($s[2]," XYZ");
+    is(join(':',@s), join(':',@r));
+}
+
+{
+    use constant BANG => {};
+    () = split m/,/, "", BANG;
+    ok(1);
+}