####
#  This subroutine takes a file containing any number of FASTA sequences, and
#  by reading the sequences in a line at a time, searches all the sequences for
#  matches to binding sites entered.  The header line of the FASTA sequence is
#  set to a key in a hash, and each site searched for is set to a key in a sub hash
#  for each Fasta ID.  The position of the 1st base of the matches is set to the 
#  values of each subhash.  Only the position of the matches is
#  saved, no actual text.  Any formatting or output of the matches needs to be 
#  done elsewhere.
####

sub SEARCHFASTA {
  my $site = 0; 
  my $fasta_id = shift;
  my $position = 0; 
  my $buffer = ''; 
  my %matches = ();
 
  open(DNA,"<", "$filename") or croak("Cannot open $filename:$!\n");
  seek(DNA,$index_hash{$fasta_id},0);
    
  GETSLINE: while($newline = <DNA> ) {
   # commented out debugging statement: print "\$newline is: $newline";
 
  # If new fasta header or blank line,
    if($newline =~ /(^>)|(^$)/) {  
      chomp $newline;
      if ($newline eq $fasta_id) {
        print "Currently searching sequence:\n     $newline\n";
        unless ($buffer eq '') {
          $i = 0;
          # commented out debugging statement: print "in unless buffer loop\n";
          while ($i < $num) {
            while($buffer =~ /$pattern[$i]/gxsi) {
              # commented out debugging statement: print "  match is $1\n";
              push @{$matches{$i}}, $position + $-[0]; 
              }
            $i++;
            } 
          $buffer = '';
          $position = 0;
          }
        }
      # if the newline is not the fasta_id, then we finished searching it, and
      # need to exit the loop structure, and close the file.
      if ($newline ne $fasta_id) {
        last GETSLINE;
        }
      }       
    
    chomp $newline;
    $buffer .= $newline;
    
    if (length($buffer) < $span ) {
      next GETSLINE;
      }
    
    # Search for the DNA pattern
    $i = 0;
    while ($i < $num) {
      while($buffer =~ /$pattern[$i]/gsxi) {
        push @{$matches{$i}}, $position + $-[0]; 
        }
      $i++;
      } 
      
    # Reset the position counter
    # (will be accurate after you reset the buffer, next)
    $position += (length($buffer) - $lg[0]+1);
  
    # Reset the buffer
    # Discard the first $max worth of data in the buffer
    $buffer = substr($buffer, (length($buffer) - $lg[0] + 1));     
    } 
      
  # This erases some of the larger variables from the search routine   
  $buffer = '';
  $position = 0; 
   
  for my $tfbs (keys %matches) {
    foreach (@{$matches{$tfbs}}) {
      $_ -= length($fasta_id);
      }
    } 
  close(DNA) or croak("Cannot close $filename:$!\n");
  return %matches; 
}