#! /usr/bin/perl
#
#  scrubkeds (by Justin Appleby)
#  [justin.appleby@gmail.com]
#
#  This perl script finds occurences of O##, l##, and
#  other bad entries in KEDS data files.  Scrubbed files
#  are written to input_filename.clean
#
#  Usage: ./scrubkeds [filenames]

foreach $arg (@ARGV){
  # read MAC style endlines
  $/ = "\r";

  print "File to parse: $arg \n";
  open (IN, "<$arg") || die "cannot open: $!";
  my @lines = <IN>;
  close(IN);

  my $outfile = $arg . ".clean";
  open(OUT, ">$outfile");

  my $i=0;
  my $rm=0;
  my $errors=0;

  foreach $line (@lines)
  {
      $i++;
      # print back bad lines when found 
      if ( $line =~ /O\d\d/ ){ 
	      print "Entry $i changed ('O##' found)\n"; 
	      $errors++;
	  } 
      if ( $line =~ /l\d\d/ ){ 
	      print "Entry $i changed ('l##' found)\n"; 
	      $errors++;
	  } 
      if ( $line =~ /-*]/ ){ 
              print "Line $i removed (bad data found)\n"; 
	      $rm++;
      } 

      # substitute proper values in bad lines
      $line =~ s/O(\d\d)/0\1/g;
      $line =~ s/l(\d\d)/1\1/g;
      $line =~ s/^.*---].*$//g;

      # write to new file
      print OUT "$line";
  }
  close(OUT);
  print "\n$errors lines fixed in the data set\n";
  print "$rm bad lines removed\n";
  print "Scrubbed file written to: $outfile\n";
}