#! perl -slw 
use strict; 

our $L ||= 200; 
our $SEED ||= 1; 
our $RANGE ||= "'a'..'c'"; 

#srand $SEED; 

my @chars = eval $RANGE; 
my $s = join'', map{ ( @chars )[ rand @chars ] } 1 .. $L; 
my @rls; use constant { BASESEQ => 0, SAVED => 1, OFFSET => 2, LENGTH => 3, REPEATS => 4 }; 

## Find runs an record them 
push @rls, [ 
    $1, 
    length $2, 
    pos($s) - length( $1 ), 
    length( $1 ) + length( $2 ), 
    length( $2 ) / length( $1 ), 
] while $s =~ m[(.+)(?=(\1{1,254}))]g; 

## Eliminate overlaps deleting the 
## lesser saving in each case. 

for my $p ( 0 .. $#rls ) { 
    next unless defined $rls[ $p ]; 
    for my $q ( $p+1 .. $#rls ) { 
        next unless defined $rls[ $p ] 
            and defined $rls[ $q ]; 
        last if $rls[ $q ][ OFFSET ] 
              > $rls[ $p ][ OFFSET ] + $rls[ $p ][ LENGTH ]; 
              
           $rls[ $q ][ OFFSET ] >= $rls[ $p ][ OFFSET ] 
        && $rls[ $q ][ OFFSET ] <  $rls[ $p ][ OFFSET ] + $rls[ $p ][ LENGTH ] 
        and delete $rls[ 
            $rls[ $p ][ SAVED ] < $rls[ $q ][ SAVED ] 
            ? $p 
            : $q 
        ]; 
    }
} 

## Remove the overlaps 
@rls = grep defined, @rls; 

## Display the results 
print $s; 
print ' ' x $_->[ OFFSET ], uc( $_->[ BASESEQ ] ), $_->[ BASESEQ ] x $_->[ REPEATS ] for @rls; 

__END__