#!/usr/bin/perl -s

use strict;
use warnings;

use NAT::Client;
use Data::Dumper;
use Lingua::StarDict::Gen;

our($name);
my $id=shift || "EuroParl-PT-EN";
our $client=NAT::Client->new(Local=>$id);
$name ||= $id ;

my $dic;
our $i = 0;
$client->iterate( \&proc );

Lingua::StarDict::Gen::writeDict($dic,"$name");

sub proc {
	my %param=@_;
	my $word = $param{word};
	return unless $word;
	return unless length($word) > 2;
	return unless $param{count} > 1;
	my $string = "$param{count}\n";
	my %trads = %{$param{trans}};

	# return if $i > 2000;

##	$string = "Palavra: $word\n\n";

	for my $t (sort {$trads{$b}<=>$trads{$a}} keys %trads) {
		next if $trads{$t} < 0.03;

		my $p = sprintf("%5.2f",$trads{$t}*100);
##		$string.= " - P(Trad( $word ) = $t) = $p\n\n";
  		$string.= "==> $t ($p):\n";
##		$string.= "   - Exemplos de uso:\n\n";

		my $concs = $client->conc({count=>4},$word,$t);
		for my $c (@$concs) {
				$string.= "  - ".trunca($word,$c->[0])."\n";
				$string.= "  - ".trunca($t,$c->[1])."\n\n";
		}

		# $string.= "   - Entradas Terminologicas:\n\n";
		# my @terms = `grep -E '\\<$word\\>' EuroParl-PT-EN/___|grep '\\<$t\\>'`;
		# for my $term (@terms) {
			# chomp($term);
			# my ($count,$l,$r) = ($term =~ m/\s*(\d+)\s+(.*?)\s=!.+?!=\s(.*)/);
			# $string .="     - $l | $r ($count)\n";
		# }
		$string .="\n\n";
	}
	$dic->{$word} = $string;

	$i++;
	print STDERR "." unless $i%1000;

}

sub trunca {
    my ($word, $sentence) = @_;

	$sentence =~ s/^.*(.{30}\Q$word\E)/...$1/is;
	$sentence =~ s/(\Q$word\E.{30}).*$/$1.../is;

    return $sentence;
}