stemmers

git-svn-id: https://ws10smt.googlecode.com/svn/trunk@737 ec762483-ff6d-05da-a07a-a48fb63a330f

1 files changed, 27 insertions, 0 deletions

diff --git a/word-aligner/stemmers/zh.pl b/word-aligner/stemmers/zh.pl
new file mode 100755
index 00000000..14ba8aab
--- /dev/null
+++ b/word-aligner/stemmers/zh.pl
@@ -0,0 +1,27 @@
+#!/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);
+  $vocab = 1;
+}
+
+my %dict;
+while(<STDIN>) {
+  chomp;
+  my @words = split /\s+/;
+  my @out = @words;
+  if ($vocab) {
+    die "Expected exactly one word per line with --vocab: $_" unless scalar @out == 1;
+    print "$_ @out\n";
+  } else {
+    print "@out\n";
+  }
+}
+