Hash or hash's of hash?

[armbsu]

Hello,

I am a beginner in Perl. I am trying to read and then process a pipe delimited file. My file looks something like this-

size|3
directory|c:\un icorn
sub-directories|sub |new.....

The examples I have seen online show code where it is known how many columns the file consists. However, this is going to be a configuration file and number of columns in sub-directories is unknown.

My question is what do I need has or hash's of hash or something else in this case?

Thanks!

[RonB]

You can use a hash, or a hash of arrays, or a hash of hashes. You haven't provided enough info for use to be able to make a clear recommendation.

Is the file format set in stone i.e., is it being dictated by your instructor/boss? If you can, I'd recommend changing it to a more standard config file format and possibly use one of the config modules on cpan to parse it.

[RonB]

Here's one option.

#!/usr/bin/env perl

use strict;
use warnings;
use Data::Dumper;

my %cfg;

while (my $line = <DATA>) {
    chomp $line;
    my ($key, @values) = split /\|/, $line;
    
    if (@values > 1) {
        $cfg{$key} = \@values;
    }
    else {
        $cfg{$key} = $values[0];
    }
}
print Dumper \%cfg;

__DATA__
size|3
directory|c:\unicorn
sub-directories|sub|new

which outputs:

$VAR1 = {
          'sub-directories' => [
                                 'sub',
                                 'new'
                               ],
          'directory' => 'c:\\unicorn',
          'size' => '3'
        };

[numberwhun]

If I may add to this, take a look at this link. I have used it as a reference for hashes in Perl and it has been extremely handy. Hopefully it helps.

Regards,

Jeff