Comparing two text files

[perlnoobster]

Hi perl monks, I am fairly new to perl and have given an attempt to compare two text files, text file A contains several columns containing numbers, text file B contains a column of numbers that I need to match. the 3rd column from text file A contains a unique value ($textpac3) and that unique value may be in text file B ($text1[0]). If both unique values appear in both files and a number from column 11 ($textpac11) matches the $text11 value from text file B then the results will print, however it doesnt seem to print correctly to the matches.txt file, please can someone help me?

use warnings;
use strict;

my $A_file = 'A.txt';
my $B = 'B.txt';
my $matches ='matches.txt';


open (INPUT, $A_file) or die "ERROR: cannot find file $A_file\n";
open (OUT2, ">$matches");

while (<INPUT>) 

{        
    my @text1=();
    @text1 = split /\t/,$_;
    chomp @text1;
    next if ($text1[0]=~ /L1/gi);

            open (OUT1, "$B");
        
                
                while(<OUT1>) {
                my @textpac=();
                @textpac = split /\t/,$_;
                chomp @textpac;
                
                next if ($textpac[0]=~ /OID/gi);
                        
            if($textpac[3] eq $text1[0] && $textpac[11] eq $text1[11])
+ 
            {
                print OUT2 join( "\t", @text1[ 1, 2, 3 ]  ), "\n";
            }
      }
       
}

close ($A_file);
 close OUT1;

[davido]

Two things to look at: First, "open (OUT2, ">$matches");" should have an "or die $!" appended to it so that you know for certain whether or not you're even successfully opening the file for output. Second, temporarily change "print OUT2 join(...." to just "print join(..." and visually inspect that the output to your screen is what you think it should be. That in and of itself is usually enlightening enough to help you to isolate the issue.

[CountZero]

Perl arrays are zero-based. When you split to @arraythe third column will actually end up in $array[2]. So you may want to review your indexes in the if test to see if you are matching the right columns.

Also, contrary to your explanation, the fields of A.txt actually end up in @text and the fields of B.txt end up in @textpac.

[space_monk]

Try something like this, which includes the suggestions by davido and Anonymous_Monk. As CountZero has suggested, you may also need to check some or all of the index values as arrays start at element zero in Perl ... :-)

use strict; use warnings;

open my $bh, "<", $B_file or die "$0: open $B_file: $!";

my %bHash;

# slurp text file B
# could generate a unique key consisting of field 0 and 11, 
# but never mind.
foreach (<$bh>) {
   my @text=split(/\t/);
   $bHash{$text[0]} = $text[11];
} 

close $bh;

open my $ah, "<", $A_file or die "$0: open $A_file: $!";
# read in text file A
foreach (<$ah>) {
    my @text = split(/\t/);
    if ($bHash{$text[3]} eq $text[11]) {
       # we have a match, print it
       print join( "\t", @text[1..3]),"\n";
    }
}
close $ah;

[perlnoobster]

Thank you all for your advice! i've taken it aboard and it works great! I have one last question though, the value that could match the $text11 of file A can be found in several other columns such as 12,13,14,15 How can I adjust the original hash: <code>$bHash{$text[0]} = $text11; so that it includes text 12....15?<\code> is it possible to implement that? if so how? Thank you ! :) This doesnt work: $bHash{$text[0]} = $text11|$text12|$text13;

[Anonymous Monk]

Hi,

I would start by reading the contents of file B into a hash to use as a look up for the values you want to check in file A.

That should simplify things a bit.

J.C.