Your first problem is definitely the easiest. To loop over each line of the file, use while (<>) like so:

open my $outfile, '>', 'out.txt';
while (<>)
{
  print $outfile $_ if /^(?:A id|M X)/;
}
close $outfile;
[download]

Obviously you may need a more rigorous regular expression, but this works for your example data. Your second problem is more difficult, so it'll have to wait for now.

Update: On second glance, your second probem is not well stated. What kind of comparison do you want to do? What should the output look like? Afraid I can't help with a poor problem statement.

This code will read your input file 1 line at a time, and filter out just the lines you require, without reading the whole file into memory at once. You would need to redirect the output to a file.

use strict;
use warnings;

my $flag = 0;
while (<DATA>) {
    if (/^A id/) {
        print;
        $flag = 1;
    }
    print if $flag and /^M/
}

__DATA__
A class 1 1 1 1 1 1
A id 12 12 15 15 16 16 
B class 0 0 0 0 0 0 
B id 0 0 0 0 0 0 0 
C id 0 0 0 0 0 0 0 
M X1 1 2 2 2 2 2 
M X2 2 1 1 1 1 1 
M X3 1 2 2 2 2 2 
M X4 2 2 2 1 1 2 
M X5 1 1 1 1 1 1 
M X6 1 1 1 2 2 1 
M X7 1 1 1 1 1 1 
M X8 1 1 1 1 1 1 
M X9 1 1 1 1 1 1 
M X10 1 1 1 1 1 1 
M X11 1 2 1 1 1 1 
M X12 2 2 2 2 2 2 
M X13 2 1 2 2 2 2 
M X14 2 1 2 2 2 2 
M X15 1 2 1 1 1 1 
M X16 1 1 2 2 2 2 
M X17 1 2 2 2 2 2
[download]

Problem #2

I'm trying to understand your comparison requirements. Could you elaborate?

Update: Here's a guess...

use strict;
use warnings;

while (<DATA>) {
    next if /^A id/;
    my ($x, $c1, $c2, @cols) = (split)[1..7];

    print "$x: col 12, 1st: $c1\n";
    for my $col (@cols) {
        if ($c1 == $col) {
            print "    matches\n";
        }
        else {
            print "    does not match\n";
        }
    }

    print "$x: col 12, 2nd: $c2\n";
    for my $col (@cols) {
        if ($c2 == $col) {
            print "    matches\n";
        }
        else {
            print "    does not match\n";
        }
    }
}

__DATA__
A id 12 12 15 15 16 16
M X1 1 2 2 2 2 2
M X2 2 1 1 1 1 1
M X3 1 2 2 2 2 2
M X4 2 2 2 1 1 2
[download]

X1: col 12, 1st: 1
    does not match
    does not match
    does not match
    does not match
X1: col 12, 2nd: 2
    matches
    matches
    matches
    matches
X2: col 12, 1st: 2
    does not match
    does not match
    does not match
    does not match
X2: col 12, 2nd: 1
    matches
    matches
    matches
    matches
X3: col 12, 1st: 1
    does not match
    does not match
    does not match
    does not match
X3: col 12, 2nd: 2
    matches
    matches
    matches
    matches
X4: col 12, 1st: 2
    matches
    does not match
    does not match
    matches
X4: col 12, 2nd: 2
    matches
    does not match
    does not match
    matches
[download]

Thanks a lot fopr the reply and sorry for bad explainations. Since, I have only two values i.e. 1 or 2 in both the columns, I would like to compare the values which are identical in the same position of the row but are different in position of the columns. for eg:

A id 12(F.C.) 12 (S.C.) 15(F.C.) 15(S.C.) 16(F.C.)16 (S.C.)
M X1   1       2         2        2        2       2 
M X2   2       1         1        1        1       1 
M X3   1       2         2        2        2       2
[download]

It is the subset of the above data, where F.C. represents First Column and S.C. represents Second Column (included for being more descriptive). Here, we see that the second column of 12 is identical to first and second column of 15 and 16. Therefore, I would like to know the longest stretch of the two similar/identical columns. Similarly, I would like to do it for all the other columns. Remember: that I can't compare the first column of 12 with second column of 12.