It's easier to split to do that. When you want to handle uniqueness or duplication, think of using a hash.

my %hash;
while (<>) {
    for (split) {
        print "dup word '$_' occurs at line $.\n"
            if $hash{$_}++;
    }
}
[download]

You can make that $hash{+lc}++ if you want to ignore case.

First of all, wrapping your code in "codetags" means that you need to start with "<code>", NOT with "<here is my code>".  You'll get more people who are willing to take a shot at helping you if they can read the code easily from the beginnning.

Secondly, what is the "expected result"?  We have no way for sure of knowing unless you tell us that up front!

Hi i feel we can do it in this way i have used my own example which is in the cookbook to explian it.

To generalise it better to find the N th match in a string, not just the first one. For example, you'd like to find the word preceding the third occurrence of "fish":

One fish two fish red fish blue fish
[download]

Solution

Use the /g modifier in a while loop, keeping count of matches:

   $WANT = 3;
   $count = 0;
   while (/(\w+)\s+fish\b/gi) 
    {
    if (++$count == $WANT) 
      {
        print "The third fish is a $1 one.\n";
        # Warning: don't `last' out of this loop
      }
    }
[download]

Or use a repetition count and repeated pattern like this:

  /(?:\w+\s+fish\s+){2}(\w+)\s+fish/i;
[download]

Code tags added by GrandFather

#!/usr/bin/perl -w
use strict;

my @para = qw|this is my paragraph. this is not all that interesting. 
+hey ho|;
my %hash = ();
print "non-unique words: \n";
## the interesting bit ##
map {$hash{$_}++} @para;
for (keys %hash){print "$_ " if($hash{$_}>1)};
#########################
print "\n";
[download]

$ perl tmp.pl
non-unique words:
is, this,
$
[download]