#!/usr/bin/perl -w

# merge_ts1_ts2	SAM-T08-server-domain_evaluate.rdb SAM-T08-server_TS2-domain_evaluate.rdb ... > SAM-T08-server_TS1_TS2.rdb
#	creates a new rdb file with all the cost columns of the input files.
# The "target" and "name" columns of the inputs are used to determine
# which data rows are merged.  The names in the output will have
# "_TS[12345]" stripped off the ends, and the stripped off identifier
# added to the column names.
#
# That means that we would have a line for T0388/D1 SAM-T08-server, with
#	columns n_ca_c_TS1 n_ca_c_TS2 nn1000_TS1 nn1000_TS2 and so forth.

    use English;    
    use File::Basename;
    use lib dirname($PROGRAM_NAME);
    use READ_RDB;


my %value;
# $value{$target. "\t"	.$server_name}{$colname."_TS1"}
#	has the value for column $colname for $server_name."_TS1" on $target

my %all_cols;
#	keys are all column names (except "name" and "target") for output
#	value is irrelevant

foreach my $file (@ARGV)
{
    open (RDBIN, "<$file") || die "Error: can't open $file for input";
    # clear the column name and number arrays
    
    # set up the column name and number arrays
    read_rdb_header(*RDBIN);
#    print STDERR "DEBUG: col_names=" .join(",", @col_names) . "\n";
    my $model_col = $col_num{"name"};
    my $target_col = $col_num{"target"};
#    print STDERR "DEBUG: model_col=$model_col target-col=$target_col\n";
    while(my $line=<RDBIN>)
    {   chomp($line);
        my @fields = split(/\t/, $line);
#	print STDERR "DEBUG: there are ". scalar(@fields) . " fields\n";
	my $model = $fields[$model_col];
	$model =~ /^(.+)(_TS[12345])$/;
	my $name=$1;	my $extension=$2;
	my $target_model = $fields[$target_col] . "\t" . $name;
	for ($i =0; $i< scalar(@col_names); $i++)
	{   my $col = $col_names[$i];
	    next if ($col eq "name" || $col eq "target" || $col eq "length");
	    my $name = $col . $extension;
	    $value{$target_model}{$name} = $fields[$i];
	    $all_cols{$name} = 1;
	}
    }
    close RDBIN;
    
}

@sorted_cols = sort(keys(%all_cols));
print "target\tname\t" . join("\t", @sorted_cols) . "\n";
print "10S\t20S" . ("\t10N" x scalar(@sorted_cols)) . "\n";

foreach my $row_name (sort(keys(%value)))
{    print $row_name . "\t" 
	    . join("\t", map { $value{$row_name}{$_} } @sorted_cols)
	    . "\n"
}