Creating a hash in Perl

How do I create a hash in Perl which uses the directory name as the key, and then stores both the count of files in the director开发者_如何学编程y as well as the names each of the files? Is it possible using hash of hashes or hash of arrays?

I would appreciate any pointers.

---

Hash values must be scalars, so the real question is how to get two values into one scalar. References are scalars, so a reference to a hash would work.

$data{$dir} = {
   file_count => 0+@files,
   files      => \@files,
};

Note that the file count is redundant. 0+@{ $data{$dir}{files} } could be used for the file count. If you choose to get rid of this redundancy, you could use

$files{$dir} = \@files;

The file count is available as

0+@{ $files{$dir} }

and the files are available as

@{ $files{$dir} }

(The 0+ can be omitted in scalar context.)

---

If I understand you correctly, this seems to do the trick (the printing of the hash using Dumper() at the end is just to show you what the hashref contains):

#!/usr/bin/perl -w

use strict;
use Data::Dumper;

my $dir = $ENV{PWD};
opendir( DIR, $dir ) or die $!;
my @files = grep { -f "$dir/$_" } readdir( DIR );
my $hash = {
    $dir => {
        count => scalar( @files ),
        files => \@files,
    }
};

print Dumper( $hash ), "\n";

---

Personally almost always I use hash references instead perl hashes (and arrayrefs instead perl arrays, too). Example:

my $dirs = {
     '/home/user' => [ '.profile', '.bashrc', 'My_pic.png' ],
     '/root'      => [ '.profile', '.zshrc' ]
};

my $var = { (...) } makes hash reference, => is just a synonym of comma , but allows distinguishing between hash keys and values. [ (...) ] makes annonymous array reference which is assigned as hash value.

You don't have to store redundant information like number of files, you can just evaluate array in scalar context:

my $root_files = $dirs->{'/root'};
$size = scalar @{$root_files};

You can read more about hashes here and here.