#!/usr/bin/env perl
use warnings;
use strict;
use Data::Dumper;
$Data::Dumper::Useqq = 1;
$Data::Dumper::Quotekeys = 0;
$Data::Dumper::Sortkeys = 1;

for my $folder ( grep {-d} glob('UNITAS_*') ) {
	print Data::Dumper->Dump([$folder], [qw/folder/]);
	opendir my $dh, $folder or die "$folder: $!";
	while ( my $file = readdir($dh) ) {
		next if $file !~ /\.mapped_sequences$/;
		print Data::Dumper->Dump([$file], [qw/file/]);
		
		my $reads = 0;
		open my $fileone, '<', "$folder/$file"
			or die "$folder/$file: $!";
		while ( my $tocount = <$fileone> ) {
			chomp $tocount;
			$tocount =~ s/>//g;
			next if $tocount =~ /[A-Za-z]/;
			if ( $tocount =~ /[0-9]/ ) {
				print Data::Dumper->Dump([$tocount], [qw/tocount/]);
				$reads += $tocount;
			}
		}
		close $fileone;
		print Data::Dumper->Dump([$reads], [qw/reads/]);
		
		my %hash;
		my $trftable = 'unitas.tRF-table.txt';
		open my $trf, '<', "$folder/$trftable"
			or die "$folder/$trftable: $!";
		<$trf> for 1 .. 4;
		while ( my $line = <$trf> ) {
			chomp $line;
			my @line = split /\t/, $line;
			#print Data::Dumper->Dump([\@line], [qw/*line/]);
			my $tRNAname;
			if ( $line[0] =~ s/tRNA-[^-]+-...// )
				{ $tRNAname = $& }
			else
				{ ( $tRNAname = $line[0] ) =~ s/-ENS.+$// }
			print Data::Dumper->Dump([$tRNAname], [qw/tRNAname/]);
			my $h = ( $hash{$tRNAname} //= {} );
			$h->{"5p-tR-halves"} += $line[ 1] / $reads * 1000000;
			$h->{"5p-tRFs"}      += $line[ 3] / $reads * 1000000;
			$h->{"3p-tR-halves"} += $line[ 5] / $reads * 1000000;
			$h->{"3p-CCA-tRFs"}  += $line[ 7] / $reads * 1000000;
			$h->{"3p-tRFs"}      += $line[ 9] / $reads * 1000000;
			$h->{"tRF-1"}        += $line[11] / $reads * 1000000;
			$h->{"tRNA-leader"}  += $line[13] / $reads * 1000000;
			$h->{"misc-tRFs"}    += $line[15] / $reads * 1000000;
		}
		close $trf;
		print Data::Dumper->Dump([\%hash], [qw/*hash/]);
		
		open my $merge, '>>', "$folder/merge"
			or die "$folder/merge: $!";
		my @tRF_types = ("5p-tR-halves", "5p-tRFs", "3p-tR-halves",
			"3p-CCA-tRFs", "3p-tRFs", "tRF-1", "tRNA-leader",
			"misc-tRFs");
		for my $tRNAname ( sort keys %hash ) {
			print $merge $tRNAname;
			for my $tRF_type (@tRF_types) {
				print $merge "\t$hash{$tRNAname}{$tRF_type}";
			}
			print $merge "\n";
		}
		close $merge;
	}
	close $dh;
}