Hi all, with the following test file how do I get prove to report 6 (assertions) rather than 2 (subtests)?

use Test::More;

subtest foo => sub {
    ok 1;
    ok 2;
    ok 3;
};

subtest bar => sub {
    ok 4;
    ok 5;
    ok 6;
};

done_testing;
[download]

Thanks!

I'm getting an unexpected result from an attempt to use a module, Filesys::DfPortable, on CPAN. I'm seeing the same results in blocks, with different filesystems used as arguments to dfportable. The output of my code looks like this:

$ perl ~/Documents/perl-libdirs-installTime.pl
/usr/lib/perl5/5.40/x86_64-cygwin-threads  installarchlib
Total 1k blocks used in installarchlib:  175,087,368
/usr/share/perl5/5.40                      installprivlib
Total 1k blocks used in installprivlib:  175,087,368

So here's the code I'd like help with:

#!/usr/bin/env perl
#
use strict;
use warnings;
use Config;
use Number::Format qw(:subs);
use Filesys::DfPortable;

my $mlen = 0;
my %hop;
my @lodirs = qw(installarchlib installprivlib);
for (@lodirs) {
    my $p = $Config{ $_ };
    $mlen = length( $p ) > $mlen ? length( $p ) : $mlen;
    $hop{ $_ } = $p;
}

$mlen += 2;
for (sort keys( %hop )) {
    printf( "%-${mlen}s%s\n", $hop{ $_ }, $_ );
    my $mea = dfportable( $hop{ $_ }, 1024 );
    printf "Total 1k blocks used in %s:  %s\n",
           $_ , format_number( $mea->{bused} );
}
[download]

I started working on this little script in Linux and I see the same anomalous results on CygPerl. I don't think it's a PEBKAC but I need other eyes to check it out. Thanks very much, Monks / Nuns.

    – Soren
Apr 16, 2026 at 07:05 UTC

use A;
use B;

$A_obj = A->new(16);
$B_obj = B->new(6);

my $n1 = $A_obj - $B_obj; # $n1 is an A object with value 10
my $n2 = $B_obj - $A_obj; # Fatal error
[download]

I have the following Type defined.

package Types;

use v5.34;
use warnings;

use Type::Utils   qw( as coerce declare from via );
use Types::Common qw( Enum Str );
use Type::Library -base, -declare => qw(
    CreditType
    InvoiceType
);

declare InvoiceType, as Enum [qw/
    ACCPAY
    ACCREC
/];

coerce InvoiceType,
    from Str,
    via {
        #warn "Trying to coerce $_\n";
        my %types = (
            Invoice         => 'ACCREC',
            SupplierInvoice => 'ACCPAY',
        );
        #my $ret = $types{$_};
        #warn "Returning $ret\n";
        return $types{$_};
     }
;

declare CreditType, as Enum [qw/
    ACCPAYCREDIT
    ACCRECCREDIT
/];

coerce CreditType,
    from Str,
    via {
        warn "Trying to coerce $_\n";
        my %types = (
            Credit         => 'ACCRECCREDIT',
            SupplierCredit => 'ACCPAYCREDIT',
        );
        my $ret = $types{$_};
        warn "Returning $ret\n";
        return $types{$_};
     }
;

1;
[download]

I have a class that uses the types as a union, this or that

package MyApp;

use v5.34;
use warnings;

use Moo;
use Types qw/ InvoiceType CreditType /;

has 'type' => (
    is       => 'ro',
    isa      => InvoiceType | CreditType,
    required => 1,
    coerce   => 1,
);


sub run {
    my ($self) = @_;

    say "Running with ".$self->type;
}

1;
[download]

I have a test to make sure it's doing what I expect `t/type.t`

#!/opt/perl5/bin/perl

use v5.34;
use warnings;

use Test::More;

use Types qw( InvoiceType CreditType );

{
    subtest 'Type coercion' => sub {
        is InvoiceType->coerce('Invoice'), 'ACCREC',
            'Can coerce a sales invoice';
        is InvoiceType->coerce('SupplierInvoice'), 'ACCPAY',
            'Can coerce a supplier invoice';

        is CreditType->coerce('Credit'), 'ACCRECCREDIT',
            'Can coerce a credit type';
        is CreditType->coerce('SupplierCredit'), 'ACCPAYCREDIT',
            'Can coerce a supplier credit type';
    };
}

{
    my $class = 'MyApp';
    use_ok($class);

    subtest 'Class coercion' => sub {
        for my $t ( qw/Invoice SupplierInvoice Credit SupplierCredit /
+ ) {
            note "Type=$t";
            my $x = new_ok($class => [ type => $t ]);
            note "Now Type=".$x->type;
        }
    };

}

done_testing;
[download]

t/type.t ..
# Subtest: Type coercion
    ok 1 - Can coerce a sales invoice
    ok 2 - Can coerce a supplier invoice
    ok 3 - Can coerce a credit type
    ok 4 - Can coerce a supplier credit type
    1..4
ok 1 - Type coercion
ok 2 - use MyApp;
# Subtest: Class coercion
    # Type=Invoice
    ok 1 - An object of class 'MyApp' isa 'MyApp'
    # Type=SupplierInvoice
    ok 2 - An object of class 'MyApp' isa 'MyApp'
    # Type=Credit
    not ok 3 - MyApp->new() died

    #   Failed test 'MyApp->new() died'
    #   at t/type.t line 31.
    #     Error was:  Undef did not pass type constraint "InvoiceType|
+CreditType" (in $args->{"type"}) at /home/dpaikkos/spl/local/lib/perl
+5/Test/More.pm line 741
    #     "InvoiceType|CreditType" requires that the value pass "Credi
+tType" or "InvoiceType"
    #     Undef did not pass type constraint "InvoiceType"
    #         "InvoiceType" is a subtype of "Enum["ACCPAY","ACCREC"]"
    #         "Enum["ACCPAY","ACCREC"]" requires that the value is def
+ined
    #     Undef did not pass type constraint "CreditType"
[download]

I excel at leaving typos in my code but I am pretty sure there are none in the code so far. The coercions appear to work in a stand alone fashion but when used as a union, the 2nd Type does not appear to apply the coercion. If I swap the "CreditType" to be the first item, I find that the InvoiceType fails.

I suspect I could use some kind of named parameterized coercion and `plus_coercions` but I hit this snag and haven't been able to move forward.

Does anyone have any insights into what I'm doing wrong?

Is it possible to run WebPerl inside a PWA?

Hence effectively running Perl inside an app which can be installed on Android, Win, Linux?

Has it been attempted yet?

What are the results?

Does it reduce the startup time of Perl because it's running hot in the background?

When the program below reaches the start function, it fails with the seemingly nonsensical error "Can't locate auto/Tk/ROText/FILENO.al". Tieing only STDOUT to the same widget is no problem.

Does anyone have any insight as to what is happening here?

use warnings;
use strict;

use IPC::Run qw(start pump finish timeout);
use Tk;
require Tk::ROText;

my $mw = MainWindow->new(-title => "   NON BLOCKING");
my $outw = $mw->Scrolled('ROText',  
            -font => "{Courier New} 10 bold", 
                -background => 'DarkBlue', 
            -foreground => 'OldLace',
            -scrollbars => 'se',
            -wrap => 'none',
            -width => 100,
            -height => 10,
            )->pack(-fill => "both", -expand => 1);
                                                                      
+                                          
# tie *STDOUT, 'Tk::Text', $outw;
tie *STDERR, 'Tk::Text', $outw;

my ($in, $out, $err) = ('', '', '');    
my $h;

if (! defined(eval { $h = start ['cmd.exe', '/c', 'dir'], \$in, \$out,
+ \$err; })) 
{
  print "\nStart failed: $@\n";
}

while($h->pumpable) 
{
  $h->pump;
  print $out;
  $out = '';
}

$h->finish;

MainLoop;
[download]

Dear Monks, here monk Dominik, asking my siblings for wisdom.

I'm a fan of using pp to make standalone programs since v5.40.0, because it lets me use the new Perls and doesn't mess the target system by installing anything by sudo cpan somewhere nobody is finding things again or making updates at all.

Long story short: Delivering a pp archive is already pretty neat, but I'd like to go a step further. The application extracts itself to /tmp/par-<user name hash>/cache-<app hash or -T string> (or similar) by default. So, what I thought, to reduce the startup time, which is really poor for pp executables, compared to native Perl, I'd like to omit the extraction and verification step of the pp binary, but directly start the interpreter from the /tmp/par-.../cache-... directory, with the script as argument. Something I think the pp binary is doing. But the executables and libraries have very obscure names, so I don't know, what exact I need to execute to start up the interpreter from /tmp/par-.../cache-...

Maybe I'm completely wrong on this path anyway and this is any magic within the pp binary itself, not just from this folder.

So, what I actually try to achieve is: having a unpacked Perl interpreter, running my scripts, for example:

$ cat /usr/bin/myapp
#!/usr/bin/env bash

/usr/libexec/myapp/perl /usr/libexec/myapp/myapp.pl "$@"
[download]

I know, I could just compile a Perl residing entirely withing /usr/libexec/myapp, but maybe there is a more simple way to do so?

From time to time, I use https://metacpan.org/pod2html to check my POD. Recently, I noticed some strange behavior:

The following code renders as expected:

package FOO;
use strict;
use warnings;

sub foo {
  my $str = shift;
  return $str if $str =~ /^ssh\@github-/;
}
1;

__END__


=head1 NAME

Foo - The great Foo module
[download]

But if i change the "return" line like this, simply adding a \. to the regexp:

  return $str if $str =~ /^\.ssh\@github-/;
[download]

then i get "Error rendering POD - "

The error disappears when i use parens:

  return $str if ($str =~ /^\.ssh\@github-/);
[download]

Or, when i change the string "ssh" to something else, no error occurs:

  return $str if $str =~ /^\.sh\@github-/;
[download]

No error with the line above!

What's going on here? I would expect a POD renderer not to look at the active code at all. Is there perhaps a hidden feature that I'm conflicting with here?

perl -MIO::Prompter -e '$ret = prompt -num,"number:" ; prompt -num,-de
+f=>$ret, "another:"'
[download]

How can I explicitly extract the string/number from the object?

perl a 
(nothing happens)

perl b 
(nothing happens)

perl c
Unrecognized character \xCF; marked by <-- HERE after <-- HERE near co
+lumn 1 at c line 1.

perl v
Number found where operator expected (Do you need to predeclare "perl"
+?) at v line 2, near "perl 5"
syntax error at v line 2, near "perl 5"
Execution of v aborted due to compilation errors.
[download]