#!/usr/bin/perl -w

use strict;

my $numArgs = $#ARGV + 1;  # calculate number of arguments passed to program from command line
our $inputfile;
our $outputfile;  # initialise filename variables

if ($numArgs >= 2) {
	$inputfile = "$ARGV[0]";
	print "Input filename received from command line.  Opening '$inputfile'...\n";
	open INPUTFILE, "<$inputfile" || die "An error occured whilst opening your input file '$inputfile': $!";
	
	$outputfile = "$ARGV[1]";
	print "Output filename received from command line.  Opening '$outputfile'...\n";
	open OUTPUTFILE, ">$outputfile" || die "An error occured whilst creating your output file '$outputfile': $!";
}

elsif ($numArgs == 1) {
	$inputfile = "$ARGV[0]";
	print "Input filename received from command line.  Opening '$inputfile'...\n";
	open INPUTFILE, "<$inputfile" || die "An error occured whilst opening your input file '$inputfile': $!";
	
	print "Please enter the path to your output file: ";
	chomp($outputfile = <STDIN>);
	open OUTPUTFILE, ">$outputfile" || die "An error occured whilst creating your output file '$outputfile': $!";
}

else {
	print "Please enter the path to your input file: ";
	chomp($inputfile = <STDIN>);
	open INPUTFILE, "<$inputfile" || die "An error occured whilst opening your input file '$inputfile': $!";
	
	print "Please enter the path to your output file: ";
	chomp($outputfile = <STDIN>);
	open OUTPUTFILE, ">$outputfile" || die "An error occured whilst creating your output file '$outputfile': $!";
}	

print "INPUT: $inputfile\n";
print "OUTPUT: $outputfile\n";  #display input and output file names

while ((&chooseprogram != 1) && (&chooseprogram != 2) && (&chooseprogram != 3)) {}


sub chooseprogram {
	
	print "Please enter 1 to parse a csv file and extract a certain field from each line...\n";
	print "Enter 2 to parse a csv file and bin the data according to a specific column...\n";
	print "Enter 3 to parse a csv file and remove a series of invalid data rows...\n";
	print "Enter 4 to parse a csv file and replace invalid characters with blanks...\n";
	print "Or enter 5 to select a string (using regex) from a repeating sequence of lines in a text file: ";
	
	chomp(my $programchoice = <STDIN>);

	eval {
		if ($programchoice == 1) {

			&csvextract;
		}

		elsif ($programchoice == 2) {

			&csvbin;
		}
		
		elsif ($programchoice == 3) {
		
			&csvstrip;
		}
		
		elsif ($programchoice == 4) {
		
			&csvclean;
		}
		
		elsif ($programchoice == 5) {
			
			&compoundselect;
		}
	};
	
	return $programchoice;
}


sub csvextract {

	print "CSVextract\n";
	
	my @requiredcolumn;
	my $requiredcolumn;
	my $firstline = 0;
	my @fields;
	my $fields;
	my $lineindex = 0;
	my $finished = "y";
	my $selections = 0;  # initialise variables
	
	while ($finished =~ /^[y]$/) {  # while the user still wants to enter more columns for selection
	
		print "Please enter the number of the column that you wish to select (first column is zero): ";
		chomp($requiredcolumn[$selections] = <STDIN>);  # read in column number for selection from keyboard
		
		print "Would you like to select another column?  Please enter 'y' if yes: ";
		chomp($finished = <STDIN>);  # read in choice of user
		
		$selections++;  # increment number of column selections counter
	}
	
	print "Please enter the line number of the first line to be parsed (line number of the first line in the file is zero): ";
	chomp($firstline = <STDIN>);  # read in line number of the first line from keyboard
	
	my $requiredcolumns = @requiredcolumn;  # count the number of required columns
	
	
	while(defined(my $line = <INPUTFILE>)) {  # read through each line in turn while they still exist 
		
		if (($lineindex) >= $firstline) {  # if the line is required to be read
			$line =~ s/["]+//g;  # remove quotation marks from SMILES strings
			$line =~ s/[\n]+//g;  # remove new line characters
			@fields = split(/,\t*/, $line);  # split line into fields according to comma delimiter
			for my $i (0 .. $requiredcolumns - 1) {  # for each required column
			
				print OUTPUTFILE "$fields[$requiredcolumn[$i]] ";  # print the correct field with a space delimiter
			}
			
			print OUTPUTFILE "\n";  # add new line character as all the fields in this row have been printed
		}
		
		$lineindex++;  # increment current line

	}
	
	close INPUTFILE;
	close OUTPUTFILE;  # close input and output files
	
	return;
}


sub compoundselect {

	print "StringSelect\n";
	
	print "Please enter the number of compounds to select: ";
	chomp(my $selectcmpds = <STDIN>);  # read in number of compounds to select from keyboard

	print "Please enter the line number of the first compound line (first line is zero): ";
	chomp(my $firstline = <STDIN>);  # read in first compound line number from keyboard

	print "Please enter the number of lines between each compound line: ";
	chomp(my $increment = <STDIN>);  # read in increment from keyboard

	print "Please enter the regex for selection (compound will be taken from $1): "; # e.g. <(.+)>
	chomp(my $regex = <STDIN>);  # read in compound selection regex from keyboard

	my @compounds;
	my $compounds;  # initialise compounds array
	my $lineindex = 0;  # initialise current line variable
	$increment++;  # add 1 to the increment variable to account for the compound line itself
	my $offset = $increment - $firstline; # calculate offset


	while(defined(my $line = <INPUTFILE>)) {  # read through each line in turn while they still exist 
		
		if (($lineindex+$offset)%$increment == 0) {  # if the line is a compound line
			$line =~ /$regex/;  # match with the regex
			$compounds[($lineindex+$offset)/$increment] = "$1";  # Read each compound into the compounds array
		}
		$lineindex++;  # increment current line
	}


	for my $i (0 .. $selectcmpds) { # for each compound to be selected
		my $totalcmpds = @compounds;  # total number of compounds is the length of the compounds array
		my $rnd = int(rand($totalcmpds));  # create a random number corresponding to a compound in the array
		print OUTPUTFILE "$compounds[$rnd] $i\n";  # write the random compound into the output file
		splice(@compounds,$rnd,1);  #remove the written compound from the array so that it can't be selected again
	}

	close INPUTFILE;
	close OUTPUTFILE;  # close input and output files
	
	return;
}
	
sub csvstrip {
	
	print "CSVstrip\n";
	
	print "Please enter the line number of the first compound line (first line is zero): ";
	chomp(my $firstline = <STDIN>);  # read in first compound line number from keyboard

	print "Please enter the column to use for validation (first column is zero): ";
	chomp(my $valcolumn = <STDIN>);  # read in number of compounds to select from keyboard

	print "Please enter the regex for validation: "; # e.g. ^"[\.]"$
	chomp(my $regex = <STDIN>);  # read in compound selection regex from keyboard

	my @fields;
	my $fields;
	my $lineindex = 0;
	
	while(defined(my $line = <INPUTFILE>)) {  # read through each line in turn while they still exist 
		
		if ($lineindex >= $firstline) {
			
			$line =~ s/[\n]+//g;  # remove new line characters
			@fields = split(/,\t*/, $line);  # split line into fields according to comma delimiter
	
			if ($fields[$valcolumn] !~ /$regex/) {	
				print "field: $fields[$valcolumn]...printing \$lineindex: $lineindex\n";
				
				print OUTPUTFILE "$line\n";
			}				
		}
		
		$lineindex++;
	}
	
	close INPUTFILE;
	close OUTPUTFILE;  # close input and output files
	
	return;
}

sub csvclean {
	
	print "CSVclean\n";
	
	print "Please enter the line number of the first compound line (first line is zero): ";
	chomp(my $firstline = <STDIN>);  # read in first compound line number from keyboard

	print "Please enter the regex matching the characters to be removed: "; # e.g. ^"[\.]"$
	chomp(my $regex = <STDIN>);  # read in compound selection regex from keyboard

	my @fields;
	my $fields;
	my $lineindex = 0;
	
	while(defined(my $line = <INPUTFILE>)) {  # read through each line in turn while they still exist 
		
		if ($lineindex >= $firstline) {
			$line =~ s/[\n]+//g;  # remove new line characters
			@fields = split(/,\t*/, $line);  # split line into fields according to comma delimiter
			my $numfields = @fields;
			for my $i (0 .. $numfields) {
				#print "field: $fields[$valcolumn]...printing \$lineindex: $lineindex\n";
				$fields[$i] =~ s/^"[\.]"$//g;
				$fields[$i] =~ s/^"[?]"$//g;
				if ($i == $numfields) {
					print OUTPUTFILE "$fields[$i]\n";
				}
				else {
					print OUTPUTFILE "$fields[$i],";
				}
			}
							
		}
		
		$lineindex++;
	}
	
	close INPUTFILE;
	close OUTPUTFILE;  # close input and output files
	
	return;
}


sub csvbin {

	print "CSVBin\n";
	
	print "Please enter the line number of the first compound line (first line is zero): ";
	chomp(my $firstline = <STDIN>);  # read in first compound line number from keyboard
	
	print "Please enter the number of bins required: ";
	chomp(my $bins = <STDIN>);  # read in number of compounds to select from keyboard
	
	print "Please enter the maximum bin value: ";
	chomp(my $binmax = <STDIN>);  # read in number of compounds to select from keyboard

	print "Please enter the column to use for binning (first column is zero): ";
	chomp(my $bincolumn = <STDIN>);  # read in number of compounds to select from keyboard

	my @fields;
	my $fields;
	my @bin;
	my $bin;
	my $lineindex = 0;
	
	my $binsize = $binmax / $bins;
	
	for my $i (0 .. $bins) {  # for each bin 
		$lineindex = 0;
		$bin[$i] = 0;
		open OUTPUTFILE, ">bin" . $i . "_output.txt" || die "An error occured whilst creating your output file 'bin . $i . _output': $!";
		open INPUTFILE, "<$inputfile" || die "An error occured whilst opening your input file '$inputfile': $!";

		while(defined(my $line = <INPUTFILE>)) {  # read through each line in turn while they still exist 

			if ($lineindex >= $firstline) {
				@fields = split(/,\t*/, $line);  # split line into fields according to comma delimiter
		
				if ($fields[$bincolumn] !~ /^[\.]$/) {	

					if ($i == $bins) {

						if ($fields[$bincolumn] >= ($i * $binsize)) {  # if the current value fits into this particular bin

							$bin[$i]++;
							$line =~ s/[\n]+//g;  # remove new line characters
						
							print OUTPUTFILE "$line\n";
							
							}
						
				
					}
						
						
					elsif (($fields[$bincolumn] >= ($i * $binsize)) && ($fields[$bincolumn] < (($i + 1) * $binsize))) {  # if the current value fits into this particular bin
					
							$bin[$i]++;
							$line =~ s/[\n]+//g;  # remove new line characters
						
							print OUTPUTFILE "$line\n";
						
					}
					
				
			
				}
		
			}

			$lineindex++;  # increment current line
		}
		
		close INPUTFILE;
		close OUTPUTFILE;
	}
	
	open OUTPUTFILE, ">$outputfile" || die "An error occured whilst creating your output file '$outputfile': $!";
		
	for my $i (0 .. $bins) {  # for each bin 
		
		my $binminval = $i * $binsize;
		my $binmaxval = ($i + 1) * $binsize;
		
		if ($i == $bins) {	
			
			print OUTPUTFILE "$binminval" . "+," . "$bin[$i]\n";
		}
		
		else {
			print OUTPUTFILE "$binminval" . "-" . "$binmaxval" . "," . "$bin[$i]\n";
		}

	}
	
	close INPUTFILE;
	close OUTPUTFILE;
	
	return;
}