moses-smt/mosesdecoder
1
1
Fork 0
mirror of https://github.com/moses-smt/mosesdecoder.git synced 2024-12-29 06:52:34 +03:00
Code Issues Packages Projects Releases Wiki Activity

option to output factors

Browse Source
This commit is contained in:
Hieu Hoang 2015-04-27 17:35:19 +04:00
parent da648fd65b
commit a47fc00635
1 changed files with 53 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

57
scripts/training/wrappers/madamira-wrapper.perl
View File

@ -8,21 +8,32 @@ use File::Basename;
use FindBin qw($RealBin);
use Cwd 'abs_path';
sub GetFactors;
my $TMPDIR = "tmp";
my $SCHEME = "D2";
my $KEEP_TMP = 0;
my $MADA_DIR;
my $FACTORS_STR;
my @FACTORS;
GetOptions(
"scheme=s" => \$SCHEME,
"tmpdir=s" => \$TMPDIR,
"keep-tmp" => \$KEEP_TMP,
"mada-dir=s" => \$MADA_DIR
"mada-dir=s" => \$MADA_DIR,
"factors=s" => \$FACTORS_STR
) or die("ERROR: unknown options");
$TMPDIR = abs_path($TMPDIR);
print STDERR "TMPDIR=$TMPDIR \n";
if (defined($FACTORS_STR)) {
@FACTORS = split(",", $FACTORS_STR);
}
#binmode(STDIN, ":utf8");
#binmode(STDOUT, ":utf8");
@ -75,13 +86,21 @@ while(my $line = <MADA_OUT>) {
print "\n";
}
elsif (index($line, ";;WORD") == 0) {
# word
# word
my $word = substr($line, 7, length($line) - 8);
#print STDERR "FOund $word\n";
#print STDERR "FOund $word\n";
for (my $i = 0; $i < 4; ++$i) {
$line = <MADA_OUT>;
}
my $factors = GetFactors($line, \@FACTORS);
$word .= $factors;
print "$word ";
}
else {
#print STDERR "NADA\n";
#print STDERR "NADA\n";
}
}
close (MADA_OUT);
@ -91,3 +110,33 @@ if ($KEEP_TMP == 0) {
# `rm -rf $TMPDIR`;
}
###########################
sub GetFactors
{
my $line = shift;
my $factorsRef = shift;
my @factors = @{$factorsRef};
# all factors
my %allFactors;
my @toks = split(" ", $line);
for (my $i = 1; $i < scalar(@toks); ++$i) {
#print " tok=" .$toks[$i];
my ($key, $value) = split(":", $toks[$i]);
$allFactors{$key} = $value;
}
my $ret = "";
my $factorType;
foreach $factorType(@factors) {
#print "factorType=$factorType ";
my $value = $allFactors{$factorType};
$ret .= "|$value";
}
return $ret;
}