join of 2 files on the bais of 2 columns of by perl

amar_lahiri · 06-13-2012, 07:04 AM

I want to write script in perl for below logic-

I have two files: file1 file2, and I wish to extract lines in file1 that have no match with second file. script should match 1st and 2nd cloumn of first file with 1st and 2nd column of second file ,then give the output on the basis of 1st file (values of file1 whioch are not present in second file)

FILE1-
6781 9547253220
55555 9777676804
3334 8016307705
8990 9777676804
121 9679234253
123 9938365782
6781 9679234253
55555 9938365782

FILE2-
55555 8018555766
55555 9938365782
6781 9679234253
6781 8018555766

so the output should be on the basis of first file or output extracts the line from first file -Menas value of FILE1 which are not present in FILE2

OUTPUT-
6781 9547253220
55555 9777676804
3334 8016307705
8990 9777676804
121 9679234253
123 9938365782

Tinkster · 06-14-2012, 04:13 PM

Not perl, but does what you asked for ...

Code:

 grep -v -f file2 file1|sort -u

121 9679234253
123 9938365782
3334 8016307705
55555 9777676804
6781 9547253220
8990 9777676804

David the H. · 06-14-2012, 04:40 PM

Please use ***[code][/code] tags*** around your code and data, to preserve formatting and to improve readability. Please do not use quote tags, colors, or other fancy formatting.

I'm not experienced in perl, but assuming the input files aren't too large to keep in memory, here's how I'd do it in awk. The concept should be transferable to other languages.

Code:

awk 'NR==FNR { arr[$1$2]=1 ; next } ( ! arr[$1$2] ){ print }' File2.txt File1.txt

1) Read file2 in a hash (assoc. array), with the index equal to field1+field2. The value only needs to be a positive string of some kind, as all we need to do is keep track of existing fields.

2) Iterate over file1, and if the first two fields fail to match an existing index, print the line. Otherwise ignore it.

Incidentally, if you aren't wedded to perl, and if the format is unique enough that a simple whole-line comparison can't result in false positives, then a simple grep could also do the job:

Code:

grep -v -f File2.txt File1.txt

Valdis Grinbergs · 06-15-2012, 08:51 AM

The other posters are right, tools other than perl are probably a better fit for this problem. A tool box should have more than one tool and a skilled worker uses the right tool for the right job. That being said, if you really need to use perl, here is one solution:

Code:

use strict;
use warnings;

my $record;
my %set1;
my %set2;

my ($filename1, $filename2) = @ARGV;

open my $f1, "<", $filename1 or die "Could not open '$filename1' - $!";
while ($record = <$f1>) {
    $set1{$record} = 1;
}
close $f1;

open my $f2, "<", $filename2 or die "Could not open '$filename2' - $!";
while ($record = <$f2>) {
    $set2{$record} = 1;
}
close $f2;

foreach my $item (sort keys %set1) {
    print $item if not exists $set2{$item};
}

Use it with the command:
perl scriptname.pl FILE1 FILE2

Tinkster · 06-16-2012, 05:39 PM

Very nice solution valdis!