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

hi Monks,
Trying to make a little csv-parsing script more efficient. Everything works fine. I realize my current set-up is going to look really ugly if I add more "selection-rules" for a change in value of column 2. I chose current setup as I'm using regexes for the real data but I'm losing track of all the curly braces at the end of my script.

Is there a different way of setting the selections up? Via subs? Eg. if I'd like to add "boat" after "Sailing".
Thanks,
Gert 
#!/usr/bin/perl -w
use strict;
use diagnostics;
use Text::CSV_XS qw( );

my $qfn_in = 'data.txt';

open( my $fh_in, '<', $qfn_in )
  or die("Can't open file \"$qfn_in\": $!\n");

my $fh_out = \*STDOUT;

my $csv_in = Text::CSV_XS->new( { sep_char => ',', eol => $/ } );
my $csv_out = Text::CSV_XS->new( { eol => $/ } );

my $selection = "Apple";
my $code      = " ####";
my $filter    = "CX89";

while ( my $row = $csv_in->getline($fh_in) ) {

    if ( $row->[1] eq $selection ) {
        $code = "tree";
        if ( $row->[0] eq $filter ) {
            $fh_out->print("$row->[0] $row->[1] $code \n");
        }
    }
}

[download]
data.txt 
"CX89",Apple
"CX89",Sailing
"CX89",Apple
"AC75",Apple
"CX89",Sailing

[download]

Replies are listed 'Best First'.
Re: parsing CSV with TEXT::CSV
by Anonymous Monk on Jul 22, 2009 at 13:12 UTC 
    CX89 Apple tree
CX89 Sailing boat
CX89 Apple tree
AC75 Apple candy
CX89 Sailing boat
    #!/usr/bin/perl --

use strict;
use warnings;

use diagnostics;

use Text::CSV;
use Text::CSV_XS qw( );

#main
{
  use autodie 2.06;
  open( my $fh_in, '<', 'data.txt' );

  my $col_rules = [
    [
      [ 1 => eq => 'Apple'],
      [ 0 => eq => 'CX89'],
      [ "%s %s tree\n" => [ 0, 1 ] ],
    ],
    [
      [ 1 => eq => 'Sailing'],
      [ 0 => eq => 'CX89'],
      [ "%s %s boat\n" => [ 0, 1 ] ],
    ],
    [
      [ 0 => ne => 'CX89'],
      [ 1 => eq => 'Apple'],
      [ 1 => ne => 'Sailing'],
      [ "%s %s candy\n" => [ 0, 1 ] ],
    ],
  ];
  process_rules( $col_rules, $fh_in, \*STDOUT );
}

sub process_rules {
  my ( $rules, $in, $out ) = @_;
  my $csv_in = Text::CSV->new( { sep_char => ',', eol => $/ } )
    or die Text::CSV->error_diag();
  my $csv_out = Text::CSV->new( { eol => $/ } )
    or die Text::CSV->error_diag();

  while ( my $row = $csv_in->getline($in) ) {
    my $rulematch = 0;
    for my $ruleset (@$rules) {
      last if $rulematch;
      my $ruleset_match = 0;
      for my $rule ( @$ruleset[0 .. $#$ruleset ] ) {
        my ( $ix, $op, $val ) = @$rule;
        $ruleset_match++ if $op eq 'eq' and $val eq $row->[$ix];
        $ruleset_match++ if $op eq 'ne' and $val ne $row->[$ix];
      }

#~ warn '$rulematch ',$rulematch;
#~ warn '$ruleset_match ', $ruleset_match ;
#~ warn '$#$ruleset ', $#$ruleset ;

      if ( $ruleset_match == $#$ruleset ) {
        $rulematch++;
        my ( $printf, $indices ) = @{ $ruleset->[-1] };
        $out->printf( $printf, (@$row)[@$indices] );
      }
    }
  }
}


__DATA__
"CX89",Apple
"CX89",Sailing
"CX89",Apple
"AC75",Apple
"CX89",Sailing

    [download]
[reply]
[d/l]
      This code is for me definitely worth to study in detail. It brings me a step further in my goal and learning process. Right now trying to sort out what this autodie module is all about.
      Many thanks!
      Gert
[reply]
        autodie stops you from having to write 
        open(...) or die;
close ... or die;

        [download]
        and allows you to use eval{} to trap fatal errors 
        eval { open ...; open ...; 1 } or print "ERROR $@\n";

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

Back to Seekers of Perl Wisdom