Hi

I give that as a starting point. Nr. 1 should be solved. the rest is an execise for tsk1979.

#!/usr/bin/perl
use warnings;
use strict;

use Text::CSV;
use Data::Dumper;

die "ERROR: You have to provide at least one csv file."
    unless(@ARGV);

my %RESULT;

foreach my $filename (@ARGV) {
    if(grep { $_ eq $filename } keys %RESULT) {
        warn "WARN: Why dou you want to compare a file '$filename' whi
+ch you have read already. Skipped.";
        next;
    }
    my $csv = Text::CSV->new(
         { binary => 1 }
    ) or die "Cannot use CSV: " . Text::CSV->error_diag ();

    open my $fh, "<", $filename or die "ERROR: Can't open '$filename':
+ $!";
    my $counter = 0;
    while (my $row = $csv->getline($fh)) {
        $counter++;
        next if($counter < 2);
        unless(@$row <= 11) {
            die "ERROR: Line '$counter' of file '$filename' has wrong 
+format. Check your data crap.";
        }
        my $index = $row->[0];
        my $index_id = $row->[1];
        $RESULT{$filename}->{$index}->{$index_id} = { 'record' => $row
+ };
    }
    $csv->eof or die "ERROR: Couldn't read csv file '$filename': " . $
+csv->error_diag();
    close $fh;
}

# Missing keys
my %FOUND_KEYS;
my $files_loaded = scalar keys %RESULT;

foreach my $filename (keys %RESULT) {
    foreach my $index (keys %{$RESULT{$filename}}) {
        foreach my $index_id (keys %{$RESULT{$filename}->{$index}}) {
            my $key = $index . '@' . $index_id;
            my $hash_ref = $FOUND_KEYS{$key} || {};
            $hash_ref->{$filename} = 1;
            $FOUND_KEYS{$key} = $hash_ref;
        }
    }
}
foreach my $key (keys %FOUND_KEYS) {
    if(scalar keys %{$FOUND_KEYS{$key}} != $files_loaded) {
        print "INFO: Key combination '$key' only found in the followin
+g files: " . join(', ', sort keys %{$FOUND_KEYS{$key}}) . "\n";
    }
}
[download]

Best regards
McA

I have a script ...

I cant help with a script I cant see - please update with script?

Hi,

IMHO the csv file format is not correct. The first line of the file lets the user assume you allways get 11 fields. Afterwards there are not enough commas to define 11 fields even if they are empty.

Best regards
McA

1. Use Text::CSV_XS to parse your CSV
2. Read your files into two separate arrays of arrays, populating the arrays with only the fields you need
3. Use Set::Scalar to do a difference operation to get your diff

use strict;
use warnings;
if ($#ARGV < 1) { die "usage csvout_compare.pl <csvoutout1> <csvout2> 
+<csvoutout3>.....\n"};
my $f=0;

my %allhash;
my @indexes_to_print;
my %data;
my @line;
my $index;
foreach my $files (@ARGV) {
    $f++;
    open FILE,$files or die "Cannot open $files\n";
    print "Opened $files \n";
    while (<FILE>) {
        next if (/csvout_ID/);
        chomp;
        @line = split(/,/,$_);
        $index = $line[0].".".$line[1];
        $allhash{$index} = 1;

        $data{$f}{$index} = "$_";
            }
    close FILE;
}

foreach $index (keys %allhash) {
    my @temparray;
    my $flag = 0;
    #First lets take all those indexes which are not in all files
    foreach my $dataset (keys %data) {
        if (!(exists $data{$dataset}{$index})) {
            push (@indexes_to_print,$index);
            #delete $allhash{$index};
            $flag = 1;
        }
    }
    next if ($flag eq 1);

    #Now we take those indexes where the Final result is different
    for (my $i=1;$i<=$f;$i++) {
        my @temp = split(/,/,$data{$i}{$index});
        $temparray[$i] = $temp[2];
        if ($i>1) {
            if ($temparray[$i] !~ $temparray[$i-1]) {
                push (@indexes_to_print,$index);
            }
        }
    }
}
#uniquify @indexes_to_print
#
 my %seen   = map { $_, 1 } @indexes_to_print;
 my @unique_indexes = keys %seen;

 foreach my $index (@unique_indexes) {
     print "$index";
    for (my $i=1;$i<=$f;$i++) {
        if (exists $data{$i}{$index}) {
            my @temp = split(/,/,$data{$i}{$index});
            #print ",$i,$temp[2]";
            print ",$i";
            foreach (my $j=2;$j<=10;$j++) {
                if (exists ($temp[$j])) {
                    print ",$temp[$j]";
                } else {
                    print ",";
                }
            }
        } else {
            print ",$i,NA,,,,,,,,";
        }
    }
    print "\n";
}
[download]