#!/usr/local/bin/perl -w

##############################################
##############################################
# SCRIPT NAME:  QUICKMINE_script
# FUNCTION:    Do lots of command line processing quickly!
##############################################

# This script is intended to be as simple as possible.
# Its purpose it to quickly population a MINE database
# with a variety of analysis files (e.g. blast reports)
# by automating the process of doing batch jobs
# on the command line.

# In summary, this script accepts a group of files
# and a list of commands.  It performs all the commands
# on each file and outputs a result file that shares
# the same 'basename' as the input file.

# Files must be in fasta format.


# This script accepts a single argument - the
# name of your config file (must
# end in .config)  

# Usage: scriptname your.config

############################################################
# Default Usage:  ./QUICKMINE_script QUICKMINE.config      #
############################################################


############################################################
# print a reminder about local applications required	   #
############################################################
print <<MESSAGE;
The following applications must be locally installed 
to run all the commands in the QUICKMINE.config file.

1) NCBI Blast
2) The EMBOSS package of applications 

If you don't have a local copy the relevant commands are simple skipped - no worries.

You can learn more about EMBOSS and the command line interfaces at
http://www.hgmp.mrc.ac.uk/Software/EMBOSS/

MESSAGE



############################################################
# initialize all variables local to this script            #        
###########################################################

my $config_file = undef;	# config file to read
my $cmd		= undef;	# each command as it's executed
undef (my @cmds);		# user set commands to use with system calls
undef (my @options);		# split each $cmd into words to get $output_file
undef my (@input_files);		# list of input files in fasta format
$file		= undef;	# each file from @input_files as it is processed
my $output_file = undef;        # output file name made from ext added to $file.all
undef (my @output_files);       # a list of $output_files

#################################################
# set the files to glob!
#################################################
@input_files = <junk*.fasta>;

###########################################################
# Parse the command line and die if it doesn't look right #
###########################################################

unless (@ARGV ==1) {
        die "\n\nProper Command Line Usage: QUICKMINE_script your.config\nPlease try again.\n\n\n";}

# shift the first command line arg passed through @ARGV
$config_file = shift;

unless ($config_file =~ /([a-zA-Z0-9]+)\.config/) {
        die "\n\n Your config file must contain only alphanumeric characters and be the first argument and must end with .config\nPlease try again.\n\n";}

print "You have chosen to process ".@input_files." input files with 
	the commands in config file $config_file\n";

########################################################
# make a blast database using formatdb #
########################################################
# the most basic command for formatdb
# formatdb -i SELF_blast_database -p F -o F
# type "formatdb" on the command line to see all options for preparing blast databases
# path to the formatdb application (comes with blastall from NCBI) -i path to the database (add flags you want)

#system ("/usr/software/blast/blast2/formatdb -i /usr/users/paed/dfield/public_html/cgi-bin/BIOPERL/$contig -p F -o F");
#$blast_database = $contig;



#################################################
# LOOP over each file ...                       #
#################################################

foreach $file (@input_files) {

        ##########################################################
	# check if if the config file exists and get commands  	 #
        ##########################################################

        # user set commands
        # all values to be interpolated inside the commands must be set at this point
        # or they go into @cmd as blanks

        # Here's where the array of commands should go - we'll get them
        # from the $config_file with a do statement -
        # This is just like having them in the code, but since they
        # are separate, we can change them independently to create
        # lots of config files

	# if it does, we should now have an array called @cmd
	unless ($return = do $config_file) {
        	warn "couldn't parse $config_file: $@" 	if $@;
        	warn "couldn't do $config_file: $!"    	unless defined $return;
        	warn "couldn't run $config_file"       	unless $return;
		die "Sorry, couldn't open your config file.  Does it exist? Is the permission 644?\nError: $!";
	}

	print "Total commands read from config file $config_file: ". @cmd." \n";


	#################################################
	# ...AND DO ALL THE WORK - loop over commands   #
	#################################################
	
	for $i (0 .. $#cmd) {
		$ext = $cmd[$i][0];	# get the extension
		$cmd = $cmd[$i][1];	# get the command

		# append the $ext onto the end of the output file
                #if ($file eq $contig) {
                #        # append the $ext onto the end of the output file
                #        $cmd .= ".all.".$ext;
                #} else {
			$cmd .= ".".$ext;
		#}

		# get the output file name
		@options = split (/\s+/,$cmd);
		$output_file = pop @options;

		
		# make a system call
		print "System call should create output file $output_file.\n";
		print "Running command: $cmd ...\n";
		system($cmd);
		warn "Warning: Couldn't run $cmd! :$?" if $?;
                push (@output_files, $output_file) if !$?;

		# make it web readable
		system ("chmod 644 $output_file");
	} # end foreach $cmd

} # end foreach file

#################################################
# Print all the new files we made and exit  ... #
#################################################


print "\n\nCreated the following output files:\n\n";
for (@output_files) {
	print "$_\n";
}


print "\n\n\nDone!\n\n\n";