seek, or close and re-open the file. ...or make your filehandle lexically scoped to a block just outside your while loop (which will handle closing your file for you). In other words:

{
    open my $fh, '<', $filename or die $!
    while ( <$fh> ) {
        # do your stuff.
    }
    # Note, your $fh is about to pass out of scope, which
    # will close the filehandle in cleanup.
}
{
    $open my $fh, '<', $filename or die $!
    while ( <$fh> ) {
        # do your other stuff.
    }
}
# Now the second filehandle fell out of scope and closed too.
[download]

Hi Dave,

I check the perl doc for SEEK. after the first while but before closing the file. I Added,

seek , 0, 1;

while <FH> ...

it did not work, Is this the right way of using SEEk?

No: you need to specify a filehandle.

seek FH, 0, 1;
[download]

Take another look at the seek docs.

It's also nice to check the return status of the file operation for success:

seek FH, 0, 1 or die "seek failed: $!";
[download]

Do you really need to read the TSV file twice in order to process your data? For large TSV files it might be faster to read the file only once line-by-line and perform the two operations - currently distributed over two different while-loops - within the same loop.

On the other hand, if processing time doesn't matter and code keeps cleaner, seek might be the better option.

Yeah, that was basically my first thought, too. If the file is small enough to easily fit in memory, read it all into an array and then walk the array for each operation. If it's too big for that, then read through it once doing both operations as you go. Either way, you'll get better performance out of it, since disk access is almost always the slowest operation.