From a3bd695cd4559416ce83194bf9ab6ea83dc8c961 Mon Sep 17 00:00:00 2001
From: Hieu Hoang <hieuhoang@gmail.com>
Date: Sun, 13 Jul 2014 02:54:58 +0100
Subject: [PATCH] factor for oov is 0, not <unk> - interferes with source
 input. Add extra argument to lowercase input words or not

---
 .../training/wrappers/make-factor-brown-cluster-mkcls.perl | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/scripts/training/wrappers/make-factor-brown-cluster-mkcls.perl b/scripts/training/wrappers/make-factor-brown-cluster-mkcls.perl
index 60f341de8..13aa7f912 100755
--- a/scripts/training/wrappers/make-factor-brown-cluster-mkcls.perl
+++ b/scripts/training/wrappers/make-factor-brown-cluster-mkcls.perl
@@ -2,7 +2,7 @@
 
 use strict;
 
-my ($cluster_file,$in,$out,$tmp) = @ARGV;
+my ($lowercase, $cluster_file,$in,$out,$tmp) = @ARGV;
 
 my $CLUSTER = &read_cluster_from_mkcls($cluster_file);
 
@@ -17,7 +17,10 @@ while(<IN>) {
   s/ $//;
   my $first = 1;
   foreach my $word (split) {
-    my $cluster = defined($$CLUSTER{$word}) ? $$CLUSTER{$word} : "<unk>";
+    if ($lowercase) {
+      $word = lc($word);
+    }
+    my $cluster = defined($$CLUSTER{$word}) ? $$CLUSTER{$word} : "0";
     print OUT " " unless $first;
     print OUT $cluster;
     $first = 0;