#!/usr/bin/perl

# Task: Extract GeneID-Number and gene information

use strict;
use warnings;

my $in;
my $data;
my @array;
my $GeneID;
my @BMB;
my $flag = 0;
my %hash;


# 1) open the .gff Inputfile and while reading line by line split $data at each tab and put them in the @array


open $in, '<', "Genomteil.gff" or die $!;

while ($data = <$in>) {
    @array = split (/\t/, $data);

    
        if ($array[2] =~ /gene/) {    	#if you find the word 'gene' a textbloxk follows which contains some information I want to extract and put in an array)
            $flag = 1; 		     	 # Set the flag, start of important data
	    @BMB = ($array[3], $array[4], $array[6]);    #the array will be used as values for my hash later
        }						 
        if ($array[8] =~ /.*;db_xref=GeneID:(\d+)\n/) {    #if you find the word 'GeneID' extract the following number and put it in my hash (as key), then put the 								    array in my hash
	 $GeneID = $1;
        } 						 
   
        if ($array[2] =~ /CDS/) {
            push (@BMB, $array[2]);
            						#put more data in my array
        } elsif ($array[2] =~ /exon/) {
            push (@BMB, $array[2]);
        }
        $hash{$GeneID} = [ @BMB ];

	if ($array [8]=~ /.*;exon_number=1/){
	  $flag = 0;	
	}						# Reset the flag. I am expecting a 'gene'-line next
    							# end while ($data = <$in>)
}
close $in;



foreach my $key (keys %hash) {
    foreach my $val (@{$hash{$key}}) {
        print "$key --> $val\n";
    }
}