The following code is just a Q&D proof of concept of Re^2: sorting a table columns using hashes (weaving output).

The idea is to interweave the output of a pp just after the code position where pp was called.

Actually the following code IS the output of itself (well stripping the POD of former calls is still needed)

Might help producing code examples for the monastery...

use Data::Dump;
use Data::Dumper;

sub LINENUMBERS(){0}


my %output;
my $order=0;
sub pp {
  my($package,$file,$line)=caller;
  $output{$file}{$line}=Data::Dump::pp(@_);
}

END {
  #print Dumper( \%output),"\n";
  for my $file (keys %output){
    my $h_lines=$output{$file};
    #print Dumper($h_lines),"\n";
    
    open my $f_code,"<",$file;
    while (my $line=<$f_code>){
      
      print "$.: " if LINENUMBERS;
      print "$line";
      if (exists $h_lines->{$.}) {
        print <<"__POD";
=output_pp
$h_lines->{$.}
=cut
__POD
      }
    }
  }
}

  
$a++;
pp($a);
=output_pp
1
=cut

$a++;
pp($a);
=output_pp
2
=cut

$a++;
pp($a);
=output_pp
3
=cut

[download]

Cheers Rolf

UPDATE: Hmm ...could be used for regression tests...

Back to Meditations