#!/usr/local/bin/perl

#######################
#  @rrays and %ashes  #
#######################
############ INPUT same as in cass3.pl##########

unless(@ARGV==1) {die "please provide name of the file in the command line!!\n";}

$filename=$ARGV[0];

#the file to be read contains a  list of GI numbers
#could need a check for contents

open(IN, "< $filename") || die "cannot open $filename:$!";

#again, the program will exit, if it cannot open the file 
#(good against typos in the filename)
#this (the ||) works because the program firast executes the left side. If this works 
#the left side is true and Perl doesn't even try to evaluate the right hand side

@gi=(); #initialized an array where I will store GI numbers

while(defined($line=<IN>)){   #populate array wih GI numbers

	#chomp($line);
	$line =~ s/\s//g;# substitutes all white spaces \s with nothing globally in $line
	#chomp only removes the \n at the end
	if ($line ne '') {push(@gi,$line)};
}

close(IN);
# Done with input 
##########Something was wrong - empty line treated as GI number 
# debugging screening for empty gi numbers
#$dummy=0;
#foreach (@gi) {
#print " $dummy  $_ \n";
#$dummy ++;
#}
#exit; added empty line test above and remove white spaces 

#how many gi numbers did I read in?
# $number_gis=@gi;
# print "I read in $number_gis GI numbers\n";

#$number_gis=scalar (@gi);
$number_gis=(@gi); #would work too
print "I read in $number_gis GI numbers\n\n";

#with list of GI numbers I could now send a request to NCBI to return me sequences
#corresponding to these GI numbers

#Now, suppose that not all GI numbers in my file are unique, i.e. the same GIs 
#occur more than once in the list. This often happens when GIs from several
#BLAST searches are combined.
#So, we want to get rid of duplicates.

#%gi_hash=(); 
#foreach $i (@gi){
#
#	if(!(exists $gi_hash{$i})){
#	 	$gi_hash{$i}=1;
#	}
#} 
######################################
#       task 4.2 :                   #
# HOW MANY DUPLICATES PER GI NUMBER? #
######################################

%gi_hash=(); #initialize hash

foreach (@gi)  #this loop populate gi_hash with key=gi-number and value how often encountered
	{
    $gi_hash{$_} += 1 ; #Hash contains Key: gi number  Values: "how often did it occur?"
    }

@gi_names = sort(keys(%gi_hash)); #assigns keys to an array

$number_gis=@gi_names; # determines number of different keys (the array in scalar context thing)

print "I read $number_gis different gi numbers\n"; 
# above answers task "4.4. Write a script that determines the number of elements in a %ash"

foreach (@gi_names){
        print "$_  occurred $gi_hash{$_} times\n";
        }
# above answers task 4.5 Write a script (or subroutine) that prints out a hash sorted on the keys in alphabetical order.


#########################################################
# But usually you want to SORT BY VALUES not by keys    #
# The sort command allows you to define a subroutine as follows 
# see old handout  

#@sorted_by_value = sort { $gi_hash{$a} <=> $gi_hash{$b}} keys %gi_hash;  

#this also works: 

#@sorted_by_value = sort by_value keys %gi_hash; # in contrast to the comment in the subroutine chapter this does not work with & 
#sub by_value { $gi_hash{$a} <=> $gi_hash{$b}} # defines the order smaller befor larger (a before b)

#even better 

@sorted_by_value = sort by_value (keys (%gi_hash));  
sub by_value { 
	$gi_hash{$a} <=> $gi_hash{$b}
	or
	$a <=> $b  #if the values are the same, then sort ascibethically (cmp) or numerically (<=>) on the keys
	} # defines the order smaller befor larger (a before b)

foreach (@sorted_by_value)
	{ 
 	print "$_ => \t$gi_hash{$_}\n";
	}
# exit;



################################
#          task 4.3:
#count how many unique GI numbers are in the list, i.e. count number of elements 
#in our %gi_hash hash.
###############################
$unique_count=0;
%gi_unique=();
foreach  (@gi_names) # could have used @gi
	{
	if ($gi_hash{$_}==1) #if the value is exactly 1
		{
		$unique_count++;    #increase the uncique counter
		$gi_unique{$_}=1 #
		};
	
	};

print "\n$unique_count gi numbers occurred only once.\n\n";

@gi_unique_array = sort(keys(%gi_unique));
foreach (@gi_unique_array){
        print "$_  occurred $gi_hash{$_} times\n";
        }