#!/usr/local/bin/perl

#  Copyright @ 1993 MIT and University of Pennsylvania
#  Written by Eric Brill 

#THIS SOFTWARE IS PROVIDED "AS IS", AND M.I.T. MAKES NO REPRESENTATIONS 
#OR WARRANTIES, EXPRESS OR IMPLIED.  By way of example, but not 
#limitation, M.I.T. MAKES NO REPRESENTATIONS OR WARRANTIES OF 
#MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE OR THAT THE USE OF 
#THE LICENSED SOFTWARE OR DOCUMENTATION WILL NOT INFRINGE ANY THIRD PARTY 
#PATENTS, COPYRIGHTS, TRADEMARKS OR OTHER RIGHTS.  

# Add new bigrams from corpus being tagged

open(XX,$ARGV[0]); # lexical rule file
while(<XX>){
	if (/good/) {
	@temp = split;
	for ($count=0;$count<=$#temp;++$count) {
		$seen{$temp[$count]}++; }}}
open(YY,$ARGV[1]); # current bigram list
while(<YY>) {
	chop;
	@temp = split;
	if ($seen{$temp[0]} || $seen{$temp[1]}) {
		$store{$_}++;}}

open(ZZ,$ARGV[2]); # new corpus to be tagged
while (<ZZ>) {
  s/^\s+//;
  if (! /^$/) {			
      @words = split(/\s+/, $_); 
      if ($seen{$words[0]} && $words[0] !~ /[0-9]/ ) 
	{ ++$store{"S-T-A-R-T " . $words[0]};}
      for ($count=1;  $count <= $#words ; ++$count) {
	  if (($seen{$words[$count-1]} || $seen{$words[$count]}) &&
	     $words[$count-1 ] !~ /[0-9]/ && 
	     $words[$count] !~ /[0-9]/) {
	      ++$store{$words[$count-1] . " " . $words[$count]}; }
      }
  }
}

while(($key,$val) = each %store) {
    print "$key\n"; }