Re: How do I write subs that take bare blocks as args?

[DamnDirtyApe]

in reply to How do I write subs that take bare blocks as args?

Thank you both. I've found the relevant sections now in perlsub and in the Camel book. For the record, it was as easy as:

sub each_item (&@)
{
    my $test = shift ;
    foreach ( @_ ) { return 0 unless ( eval &$test ) }
    return 1 ;
}

Thanks again.

Update: On reading kvale's node a little closer, I tried changing my code to:

sub each_item (&@)
{
    my $test = shift ;
    foreach ( @_ ) { return 0 unless &{$test} }
    return 1 ;
}

...which also worked. Is there any advantage/disadvantage to this over eval?

_______________
D a m n D i r t y A p e
Home Node | Email

[educated_foo]

..which also worked. Is there any advantage/disadvantage to this over eval?

The eval version probably isn't doing what you want, unless you are doing something deliciously perverse. I think your original code is doing this:

# "$x = eval &$foo" becomes:
my $tmp = &$foo(@_);
$x = eval $tmp;

So you're passing @_ to the user's block, then eval'ing the result. This ends up working because the block presumably accesses $_ directly, and returns an integer (which evals as itself). But it's an ... interesting way to get the job done. An optimist would say that the eval version is "very general".

/s

[Courage]

Among those two calls of &$test they are almost equivalent, and first one will trap any errors, so you can later check $@ variable to see if subroutine died. Like "try" in C++.

OTOH my personal preference is to call anonymous sub as

$test->();
$test->('arg1',$arg2);

Courage, the Cowardly Dog.
PS. Something fishy is going on there, or my name is Vadim Konovalov. And it's not.