summaryrefslogtreecommitdiff
path: root/word-aligner/support/generate-per-sentence-grammars.pl
blob: 695cfc17e68565f739bca34b4a248aef66274375 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
#!/usr/bin/perl -w
use strict;
use utf8;

die "Usage: $0 f.voc corpus.f-e grammar.f-e.gz\n" unless scalar @ARGV == 3;

my $MAX_INMEM = 1000;

open FV,"<$ARGV[0]" or die "Can't read $ARGV[0]: $!";
open C,"<$ARGV[1]" or die "Can't read $ARGV[1]: $!";
open G,"gunzip -c $ARGV[2]|" or die "Can't read $ARGV[2]: $!";

binmode FV, ":utf8";
binmode C, ":utf8";
binmode G, ":utf8";

my $vc = 0;
my %most_freq;
$most_freq{"<eps>"} = 1;
while(my $f = <FV>) {
  chomp $f;
  %most_freq{$f}=1;
  $vc++;
  last if $vc == $MAX_INMEM;
}
close FV;

print STDERR "Loaded $vc vocabulary items for permanent translation cache\n";

my $memrc = 0;
my $loadrc = 0;
while(<G>) {
  chomp;
  my ($f, $e, $feats) = split / \|\|\| /;
  if ($most_freq{$f}) {
    #print "$_\n";
    $memrc++;
  } else {
    $loadrc++;
    $grammar{$f}="$e ||| $feats";
  }
}

print STDERR "  mem rc: $memrc\n";
print STDERR " load rc: $loadrc\n";