I'd suggest:

• Providing code to call your subroutine (with data to drive it),
• Simplifying your code removing things unrelated to the problem (such as the regexes reformatting the data) if your sample program is large,
• Describing the problem completely--IOW, what does "not able to print the contents" actually mean? What's the symptom?

For more information, see Ask questions the smart way, I know what I mean. Why don't you? and How (Not) To Ask A Question.

...roboticus

In this case,not able to print the contents.

What happens?

It would be easier if you minimized your code to a few lines which demonstrate your problem, and without the other code that is of no interest to us.

You should also read perlreftut, which introduces you to references in Perl.

Is argument 6 @ff? If so it looks as though it is getting clobbered when you "determine the field names".

  # determine the field names
  if ($colNamesHdl) {
    @ff=();
    for ($i=0; $i <= $#$colNamesHdl; $i++) {
      push(@ff, $$colNamesHdl[$i]);
    }

  } else {
    $k  = $$rowsHdl[0];
    @ff = keys(%$k);
  }
[download]

update: changed shift to unshift

As per the given details, Its not clear what really is your requirement. If getting the contents of 6th argument($rateType ) means, You have to access the array as @$rateType.

It would be lot helpful if you mention the requirement with sample input and details about how this subroutine is called etc

The above function being called via the following

&pkoutFile($OUTFP, ';:=~',"301","WPSPD", $valDate1, "SPREAD", \@retArr
+ay);
[download]

where \@retArray is the array reference which contains the result set. We are able to access the contents of the result set in the above function, but unable to copy the contents to the hash variable @ff. Code snippet for this is given below. $rowsHdl is the array reference

# determine the field names
  if ($colNamesHdl) {
    @ff=();
    for ($i=0; $i <= $#$colNamesHdl; $i++) {
      push(@ff, $$colNamesHdl[$i]);
    }
  } else {
    $k  = $$rowsHdl[0];
    @ff = keys(%$k);
  }

  $j = $#$rowsHdl;
  for ($i=0; $i <= $j; $i++) {
    @paramVals = ();
    foreach $k (@ff) {
      $k1= $k;
      $k1 =~ s/(^[\n\r\s\t]+)|([\n\r\s\t]+$)//g;
      $val = $$rowsHdl[$i]->{$k};
      push(@paramVals, $k . $delim2 . $val);
    }
[download]

I would suggest adding some print Dumper statements in to make sure that you have the kind of thing that you think that you have...

Running some simple test should provide some light on what data this pkoutFile sub is really getting. I simplified your statement to input the sub's input args - avoid using indicies unless you have to - Perl is designed so they aren't needed nearly as often as in other languages.

#!/usr/bin/perl -w
use warnings;
use Data::Dumper;

pkoutFile($OUTFP, ';:=~',"301","WPSPD", $valDate1, "SPREAD", 
         \@retArray);
         
sub pkoutFile
{
  my( $outFH,       $delims,   $branchID,  $sourceSystem,  
      $valDateY4MD, $rateType, $rowsHdl,   $colNamesHdl)= @_;
      
  print Dumper \$rowsHdl;
  
  #$colNamesHdl is undefined in this case
  #so, it appears that this following code is executed
  
  $k  = $$rowsHdl[0];  #  $rowsHdl->[0]
  
  print Dumper \$k;
  
  @ff = keys(%$k);
  
  print Dumper \@ff;

}
[download]

update: I noticed that for example the k1 value has a calculation done on it, but is never printed. With further looking, if the intent of the regex you want is to remove leading and trailing spaces, I would suggest this:

foreach ($delims, $branchID, $valDateY4MD,
         $sourceSystem, $rateType)
{
    s/^\s+//;  #remove leading spaces
    s/\s+$//;  #remove trailing spaces
}
[download]

This sets $_ to each of the scalar variables in turn and then runs 2 simple and fast regex'es that do not require the /g switch. In this case $_ is an alias for the variable (like $delims) and changes in the loop are reflected in changes to the actual variable.