What have you tried? Please post your code and see "How do I post a question effectively?". If you don't have any code yet, start with perlintro; we'll be glad to help you learn but this isn't a code-writing service. You may also find the "flip-flop" operator useful for operating only on lines between #ERRORS and cellpattern.

hi monk, okok here is the code i wrote. i didnt put here at first because i just want to know the algorithm on the flow . btw, my code is pretty crude so hope do not mind on it.

if (exists $args{dir})
{
     if(-e $args{dir})
     {
        $directory =$args{dir};
        {
         print "Path: $directory\n\n";
        
         
        }
     }
     else

     {
        print "-E- the dir does not exist $args{dir}\n"; exit;
     }
}

opendir (DIR , "$directory") || die "Cannot open $directory\n";

chdir($directory);
$cmd = "ls -t";
chomp ( @files = grep { /log.sum/} `$cmd` );

#print "@files\n";
      

$flag_flow  = 0 ;
$flag_tool  = 0 ;

print "Script version : v1\n\n";
print "Timestamp        Cell             Status\t  Flow\t\t\t         
+      # Errors\t       # Warnings\n";
print "=========       ======            ======          ======       
+                        =========       ==========\n";
foreach (@files)
{
  
  if(/(\S+)\.(\S+).iss.log.sum/)
  {
     $cellname = $1;
  }
  
  @full = "$directory/"."$_";
  foreach  my $files (@full)
  {
     
     open INFILE , "<$files" ; 
     open OUTPUT , ">>result.txt" ; 
     my @lines = <INFILE>;
     

    
     my $timestamp = ctime(stat($files)->mtime);
     my $time = substr($timestamp , 4 , -8);
       



    foreach (@lines) 
    { 
        
       if (/Status\s+Flow/)
       { 
        $flag_flow = 1;
       }   
       if($flag_flow)
       {
          next if (/====/);
          if (/^(clean|DIRTY)\s+(\S+)\s+(\S+)\s+(\S+)/)
          {
           
           
               $data{$a}{Status}   = $1;
               $data{$a}{Flow}     = $2;
               $data{$a}{Errors}   = $3;
               $data{$a}{Warnings} = $4;
              push @array , $_;
                

                #print "$time  $cellname  $_";
                            
               foreach $array (keys %data)
               {   
                  
                  
                                   
                 


                   #@summary ="$time\t $cellname $data{$array}{Status}
+\t $data{$array}{Flow}\t             $data{$array}{Errors}\t\t $data{
+$array}{Warnings}\n";
                                     
 
                   print "$time\t $cellname $data{$array}{Status}\t $d
+ata{$array}{Flow}\t             $data{$array}{Errors}\t\t $data{$arra
+y}{Warnings}\n";
              
              
               }
               

          }    
          @array = ();
                    
                    

          
           
          

       }
       
       if(/^\s*$/) 
       {
          $flag_flow = 0 ;
                     
           
       } 
       
   
             

       if(/See\s+\$PDSLOGS\/(\S+)\.(\S+)\.icv/)
       {
             $error_cellname = $1; 
             $error_flow = $2;
             #print"$_\n";             
             #push @match , "cell is $error_cellname , flow is $error_
+flow\n"  ;                        
             
                          
             





             print "CELL = $error_cellname , FLOW = $error_flow\n";
             print "===========================================\n";
             $error_count = 1;
            
       }
       elsif(/PLAYERR Files Created/)
       {  
            $error_count = 0;
              
                      
       }
       elsif($error_count)
       {
            
           
           #push (@err , $_);
           print "$_";
          
       }
       
              
    }
    close INFILE ;
  }
  
  
} 


exit;
[download]

as you can see , after i print out $error_cellname or $error_flow , all the element will be treated as single element (due to the way i code my main code). i would like to print out the element individually so that i can produce the output i desired(as in my first post)

I'm sorry to say there are a lot of issues with your code that make it really hard to read and debug. If you want help with the code, the first thing you need to do is Use strict and warnings, and second you should run it through perltidy and get rid of all those extra blank lines. You've got some unused handles (e.g. DIR and OUTPUT) and you'll need to decide on the variables' scopes. Plus a few more things, like missing error handling (... or die) on opens, ...

I'm still a little unclear on the problem statement, but I can give a general answer to what was posted below:

use warnings;
use strict;
my @errors;
for my $input (1..3) {       # main loop, just a demo
    # do something with $input here
    print "pattern$input\n";
    if ($input%2) {          # whatever your error condition is
       push @errors, "error$input";
    }
}
print "\n#errors\n";           # main loop over, now output @errors
for my $error (@errors) {
   print "$error\n";
}
print "cellpattern\n";
[download]

As you operate over the input to produce the output, you accumulate the errors in @errors, and then, after the main loop, you loop over @errors to output the contents of that array. The output of the above snippet is:

pattern1
pattern2
pattern3

#errors
error1
error3
cellpattern
[download]

hi Laurent ok , let me further clarify want i want to produce as my output is display <pattern*> top of file and <#errors to cellpattern> bottom of file. right now whenever a file has #errors, it will parse and display together with pattern* .this is because i can't print them out of the main code.

desired output

pattern1
pattern2
pattern3


#errors
.
..
cellpattern
[download]

current output i get
pattern1
#errors
.
..
cellpattern

pattern2
#errors
.
..
cellpattern

pattern3
#errors
.
..
cellpattern
[download]

Disclaimer: Not tested.

Maybe something like this:

my $getPatterns;
while (<>) {
    if (/^#ERRORS/ .. /^CELLS/) {
        push(@errors, $_);
        $getPatterns = 1;
        next;
    } elsif (defined $getPatterns) {
        push(@errors, $_);
        $getPatterns = undef if (/^\s*$/);
        next;
    }
    ...; # other processing
}
print "Errors:\n";
for (@errors) {
    print "    $_";
}
[download]

Update: I think this could be written as:

Update 2; No, it won't work.

while (<>) {
    if (/^#ERRORS/ .. /^CELLS/) {
        push(@errors, $_);
    } elsif (/^CELLS/ .. /^\s*$/) {
        push(@errors, $_);
    } else {
        ...; # other processing
    }
}
print "Errors:\n";
for (@errors) {
    print "    $_";
}
[download]