emilford has asked for the wisdom of the Perl Monks concerning the following question:

I have data returned from a sql query that comes as: 
Row 1
   label1 : item1
   label2 : item1
Row 2
   label1 : item2
   label2 : item2

[download]
I need all of the information after the colons. The odd numbered values (label1) I want to use as a hash key and the even numbered values (label2) I want to use as the hash value. I've gotten that far with the following: my %hash = map { m!(?:\S+) : (\S+)! } @stuff_from_query;. The problem I've encountered is that the data after label1 could be duplicate, but with different values for label2. The current code stores only one value, overwritting the previous. I need to somehow incorporate a push so that the value of the hash is an anonymous array, containing all found label2 values.

Is there a nifty way I can still incorporate this into a map line?

Replies are listed 'Best First'.
Re: map to hash of arrays
by ccn (Vicar) on Jul 22, 2004 at 21:03 UTC

    Is it what you want? 

    use Data::Dumper;
my $data = <<EOT;
Row 1
   label1 : item1
   label2 : item1
Row 2
   label1 : item2
   label2 : item2
EOT

push @{$hash{$1}}, $2 while $data =~ /(\S+)\s*:\s*(\S+)/g;

print Dumper(\%hash);

    [download]

    $VAR1 = {
          'label1' => [
                        'item1',
                        'item2'
                      ],
          'label2' => [
                        'item1',
                        'item2'
                      ]
        };

    [download]
[reply]
[d/l]
[select]
      Not exactly. Let me try my explanation again. Here is a chunk of sample data to illustrate what I'd like to do. 
      Row 17
    object_label : yang
    object_id : 123456
Row 18
    object_label : yang
    object_id : 789101112
Row 19
    object_label : yang
    object_id : 13141516171819

      [download]
      The data comes in exactly like that, line by line. I need to ignore the Row \d+ lines and grab the rest of the information, so that I end up with a hash of arrays where 'object_label' are used as keys and 'object_id' with the same 'object_label' are items in an array. 
      %hash = {
          'yang' => [
                      '123456',
                      '789101112',
                      '13141516171819'
                    ]
         }

      [download]
[reply]
[d/l]
[select]

        please try this: push @{$hash{$1}}, $2 while $data =~ /object_label\s+:\s+(\S+)[^:]+:\s+(\S+)/g;

[reply]
[d/l]
Re: map to hash of arrays
by rnahi (Curate) on Jul 22, 2004 at 21:26 UTC

    Perhaps DBI Recipes has something useful for you.

    Check the sections about lists of hashes and hashes of hashes.

[reply]
Re: map to hash of arrays
by dpuu (Chaplain) on Jul 22, 2004 at 21:13 UTC
    It is obviously possible, but it would not be "nifty". In fact, I think it'd be worse solution than just using a C<for> loop in the first place. You just have to build a temportary hash off to the side. But if you're going to build a tmp hash, then you might as well just build the real one.

    --Dave
    Opinions my own; statements of fact may be in error.

[reply]
      That's my issue. I want to build the real hash of arrays without the need of a temporary one. I'm just not sure how to go about doing that.
[reply]
        Use a while loop, not a map: 
          my %hash;
  my $key;
  while (<>) {
    /\S+ : (\S+)/ or die "unexpected line format: $_";
    my $value = $1;
    if (defined $key) {
      push @{ $hash{$key} }, $value;
      undef $key;
    else {
      $key = $value;
    }
  }

        [download]
        --Dave
        Opinions my own; statements of fact may be in error.
[reply]
[d/l]

Back to Seekers of Perl Wisdom