#!/usr/bin/perl

use strict;
use warnings;

use Fcntl qw(O_CREAT O_WRONLY O_TRUNC);
use Getopt::Long;
use Pod::Usage;

=head1 NAME

fasta-most - finds the most frequently occuring sequence length

=head1 SYNOPSIS

fasta-most [options]

 Options:
  -min        <length>  minimum length to accept
  -max        <length>  maximum length to accept

 Reads fasta sequences on standard input.
 Writes the most frequently occuring sequence length and it's count to 
  standard output.

=cut

my $min = 1;
my $max = undef;
my $help = 0; #FALSE

GetOptions(
  "min=i" => \$min, 
  "max=i" => \$max,
  "help|?" => \$help
) or pod2usage(2);

pod2usage(1) if $help;
pod2usage("Min must be larger or equal to zero.") if $min < 0;

my $in_sequence = 0; # FALSE
my $len = 0;
my $line;
my %lengths = ();

while ($line = <STDIN>) {
  if ($line =~ m/^>/) {
    if ($in_sequence && $len >= $min && (!defined($max) || $len <= $max)) {
      if ($lengths{$len}) {
        $lengths{$len} += 1;
      } else {
        $lengths{$len} = 1;
      }
    }
    $in_sequence = 1; # TRUE
    $len = 0;
  } elsif ($in_sequence) {
    if ($line =~ m/^\s*((?:\S+\s*?)+)\s*$/) {
      my @chunks = split(/\s+/, $1);
      foreach my $chunk (@chunks) {
        $len += length($chunk); 
      }
      $in_sequence = 0 if defined($max) && $len > $max;
    }
  }
}
if ($in_sequence && $len >= $min && (!defined($max) || $len <= $max)) {
  if ($lengths{$len}) {
    $lengths{$len} += 1;
  } else {
    $lengths{$len} = 1;
  }
}

unless (%lengths) {
  print "$min 0\n";
} else {
  my ($length, $count);
  my ($best_length, $best_count);
  keys %lengths; # reset iterator
  foreach $length (sort {$b <=> $a} keys %lengths) {
    $count = $lengths{$length};
    if (!defined($best_count) || $count > $best_count) {
      $best_length = $length;
      $best_count = $count;
    }
  }
  print "$best_length $best_count\n";
}