The sort-then-uniq strategy will handle data of any quantity and will do so with linear performance.

When will you get it through your thick skull that sorting is not linear; it is O(N log N). And disk sorting slower still.

Hi All, I used wind's method but with slight modifications as below

##########################prints unique and sorted salary
 use strict;
use warnings;

my %hash;
open my $fh, '<', 'abc.txt' or die $!;
while (<$fh>) {
    chomp;
    my ($key, $val , $val1) = split '\|';
    $hash{$val1}++;
}

print "$_\n" for  sort keys  %hash;




----abc.txt------
P|PERL|50000
Paa|JAVA|20000
Poo|VB|10000
AB|SAN|20000
ABCD|CSS|500
PQRS|HTML|200

--The output is--

10000
200
20000
500
50000
[download]

Though i sorted and printed the unique values,the issue is maybe we may want to print it like

50000
20000
10000
500
200
[download]

But according to "Perl in 21 days" sort only compares the first digit/letter and sorts accordingly. For printing the above specified output we may want to write a subroutine which compares two values and then swaps them and prints the output. And one more thing, we are not allowed to install CPAN modules in our firm,as we have restricted access.So whatever we do we have to use existing functions and code.

Got the desired output:

# use strict;
#use warnings;

my %hash;
open my $fh, '<','abc.txt' or die $!;
while (<$fh>) {
    chomp;
    my ($key, $val , $val1) = split '\|';
    $hash{$val1}++;
}

@sorted= reverse sort keys  %hash;
#print "After reverse sorting @sorted\n";

################sorts and prints the array salaries in descending orde
+r

for($i=0;$i<=20;$i++)
  {
      for($j=$i+1;$j<=20;$j++)
       {
          if ($sorted[$i] < $sorted[$j])
            {
              $temp = $sorted[$i];
              $sorted[$i] = $sorted[$j];
              $sorted[$j] = $temp;
             }
             
       }
   }
              
print "The salaries in unique and sorted descending order are\n";
print "@sorted\n";
[download]

Thanks All.