UNIX for Advanced & Expert Users

IMHO the definitive answer is installing gawk, at least that's the routine advice given over at the HP forums to people facing the same situation, that really helps resolving this issue, and it'll also come very handy in the future. From my experience, sed is able to manipulate huge files in HP-UX ( 11.00 ).

If that's not an option then perl will be your friend. In the meantime you can use awk 2 perl converter ( a2p code.awk ) that comes with perl install package, for an equivalent perl solution from an awk code.The following one might be helpful for your problem, and I'm sure that it can be optimized even more, but it'll get the job done:

Code:

#!/usr/bin/perl

$[ = 1;
$\ = "\n";
$FS = "|";
while (<>) {
    chomp;
     @Fld = split($FS, $_, -1);
     if ($a{$Fld[1] . $FS . $Fld[2] . $FS . $Fld[3] . $FS .

      $Fld[4] . $FS . $Fld[5] . $FS . $Fld[6] . $FS . $Fld[7]}++ == 0) {
        print $_;
    }
}

[ Tested on Cygwin, not near a HP box now. ]
______________________________________

EDIT: On HP-UX worked fine, it took a while though. It ran about 4 min on a file with ~ 200,000 lines, ( ~ 13 MB ).

Hi, I know this is old, but couldn't you do this?

#!/usr/bin/ksh

while read name

do

something_to $name

done < $data_to_read

This is line by line, from "mastering shell scripting".