http://www.perlmonks.org?node_id=512534

thealienz1 has asked for the wisdom of the Perl Monks concerning the following question:

Hello monks

I am currently working on a data mining algorithm program in Perl, which requires the need to read the arff file format. Currently, I have gotten as far as I can in a regular expression I am working on.

The regexp is being designed to load from the arff file the @attribute directive. Currently I am trying to implement the different possible formats of this into a single regexp.

possible values

With my regexp I am trying to pull the name and type from directive line. So I would get the name=Sex and type=Male, Female or name=Feature 0 and type=real, etc... Currently this the code I have written for testing.

my @lines = (
    "@attribute 'Sex' { Male, Female}",
    "@attribute 'Feature 0' real",
    "@attribute 'Feature 1' real",
    "@attribute 'Feature 2' real",
    "@attribute Malic_acid REAL",
    "@attribute Ash REAL"
    );
foreach my $line (@lines) {
    $line =~ /^@attribute\s+[']?(.*?)[']?\s+[{]?(.*?)[}]?$/;
    print "relation name: $1\n";
    print "type: $2\n";
}

[download]

With the current output of the program being

perl test.pl
relation name: Sex
type:  Male, Female
relation name: Feature
type: 0' real
relation name: Feature
type: 1' real
relation name: Feature
type: 2' real
relation name: Malic_acid
type: REAL
relation name: Ash
type: REAL

[download]

As you can see the relation name for the Feature 0,1,2 are not being correctly read. I no I am so close, but it is getting late, and I have no idea where I am going with this regexp. I though I'd leave it open for the monks to tinker with.

Night

Back to Seekers of Perl Wisdom