Extract Surface, Exchange and Agitation for SEO keywords

gistfile1.pl

#!/bin/perl
# use warnings;
use Data::Dumper;

$num_args = $#ARGV + 1;
if ($num_args != 2) {
  print "\nUsage: ./seokw.pl corpusA corpusB\n";
  exit;
}



my %Ca, my %Cb;

open (CSV, "<", $ARGV[0]) or die $!;
while(my $line = <CSV>) {
    my @cells = csvsplit($line); # or csvsplit($line, $my_custom_seperator)
    if($cells[4] <= 0){
      print "OUPS: $line";
    }
    else{
    	$Ca{$cells[0]} = $cells[4];
    }
}
close CSV;

open (CSV, "<", $ARGV[1]) or die $!;
while(my $line = <CSV>) {
    my @cells = csvsplit($line); # or csvsplit($line, $my_custom_seperator)
    if($cells[4] <= 0){
    	print "OUPS: $line";
    }
    else{
    	$Cb{$cells[0]} = $cells[4];
    }
}
close CSV;

# Compute Exchange
# Compute Agitation
my $Agitation = 0, my $Exchange = 0;
while( ($kw, $p) = each %Ca){
	if($Cb{$kw} != undef and $Cb{$kw} != $p){
		$Agitation++;
	}
	if($Cb{$kw} == undef){
		$Exchange++;
	}
}

# Compute Resulting Surface
my $Surface = 0;
while( ($kw, $p) = each %Cb){
	$Surface += 1 / $p;
}
my $CardCb = keys (%Cb);
$Surface = $Surface / $CardCb;

print "seo.keywords.all.agitation $Agitation\n";
print "seo.keywords.all.exchange $Exchange\n";
print "seo.keywords.all.surface $Surface\n";

# http://stackoverflow.com/questions/3065095/how-do-i-efficiently-parse-a-csv-file-in-perl
sub csvsplit {
        my $line = shift;
        my $sep = (shift or ',');

        return () unless $line;

        my @cells;
        $line =~ s/\r?\n$//;

        my $re = qr/(?:^|$sep)(?:"([^"]*)"|([^$sep]*))/;

        while($line =~ /$re/g) {
                my $value = defined $1 ? $1 : $2;
                push @cells, (defined $value ? $value : '');
        }

        return @cells;
}