Question

Purely academic question, and I don’t see instructions banning them here (although there is no ‘academic’-like tag I could find).

If I have an existing hash like the following, I can take a slice(?) of it as shown:

my %hash = (one=>1, two=>2, three=>3, four=>4);
my ($two, $four) = @hash{'two','four'};

Is there a way to do this if the hash is returned from an example function like this?

sub get_number_text
{
    my %hash = (one=>1, two=>2, three=>3, four=>4);
    return %hash;
}

One way that works is:

my ($two, $four) = @{ { get_number_text() } }{'two', 'four'};

As I understand it, function returns a list of hash keys/values, the inner {} creates an anonymous hash/ref, and @{} uses the reference to “cast” it to a list aka a hash slice since Perl knows the ref is a hash. (I was a little surprised that the last bit worked, but more power to Perl, etc.)

But is that the clearest way to write that admittedly strange access in one expression?

Answer 1

In general, avoid returning a flattened hash (return %foo) from a subroutine; it makes it harder to work with without copying it into another hash. Better to return a hash reference (return \%foo).

But yes, that is the clearest way. Though often lists of hardcoded keys are given using qw:

my ($two, $four) = @{ { returnit() } }{ qw/two four/ };