'pseudo model 2' that strictly favors a diagonal, with tunable parameters for p(null) and how sharp/flat the alignment distribution is around the diagonal

1 files changed, 39 insertions, 0 deletions

diff --git a/word-aligner/stemmers/ar.pl b/word-aligner/stemmers/ar.pl
new file mode 100755
index 00000000..c85e883a
--- /dev/null
+++ b/word-aligner/stemmers/ar.pl
@@ -0,0 +1,39 @@
+#!/usr/bin/perl -w
+
+use strict;
+use utf8;
+
+binmode(STDIN, ":utf8");
+binmode(STDOUT,":utf8");
+
+my $vocab = undef;
+if (scalar @ARGV > 0) {
+  die "Only allow --vocab" unless ($ARGV[0] eq '--vocab' && scalar @ARGV == 1);
+  $vocab = 1;
+}
+
+my %dict;
+while(<STDIN>) {
+  chomp;
+  my @words = split /\s+/;
+  my @out = ();
+  for my $w (@words) {
+    my $tw = $dict{$w};
+    if (!defined $tw) {
+      my $el = 4;
+      if ($w =~ /^(.st|.n|Al)/) { $el+=2; }
+      if ($w =~ /^(y|t|n)/) { $el++; }
+      if ($el > length($w)) { $el = length($w); }
+      $tw = substr $w, 0, $el;
+      $dict{$w} = $tw;
+    }
+    push @out, $tw;
+  }
+  if ($vocab) {
+    die "Expected exactly one word per line with --vocab: $_" unless scalar @out == 1;
+    print "$_ @out\n";
+  } else {
+    print "@out\n";
+  }
+}
+