view release on metacpan or  search on metacpan

lib/Lingua/FreeLing3/Splitter.pm  view on Meta::CPAN

use 5.010;

use warnings;
use strict;

use Carp;
use Lingua::FreeLing3;
use Lingua::FreeLing3::Config;
use File::Spec::Functions 'catfile';
use Lingua::FreeLing3::Bindings;
use Lingua::FreeLing3::Sentence;

use parent -norequire, 'Lingua::FreeLing3::Bindings::splitter';

our $VERSION = "0.02";

=encoding UTF-8

=head1 NAME

Lingua::FreeLing3::Splitter - Interface to FreeLing3 Splitter

=head1 SYNOPSIS

   use Lingua::FreeLing3::Splitter;
   use Lingua::FreeLing3::Tokenizer;

   my $pt_tok = Lingua::FreeLing3::Tokenizer->new("pt");
   my $pt_split = Lingua::FreeLing3::Splitter->new("pt");

   # compute list of Lingua::FreeLing3::Words
   my $list_of_words = $pt_tok->tokenize( $text );
   my $list_of_sentences = $pt_split->split($list_of_words);

=head1 DESCRIPTION

Interface to the FreeLing3 splitter library.

=head2 C<new>

Object constructor. One argument is required: the languge code
(C<Lingua::FreeLing3> will search for the splitter data file).

Returns the splitter object for that language, or undef in case of
failure.

=cut

sub new {
    my ($class, $lang) = @_;

    my $config = Lingua::FreeLing3::Config->new($lang);
    my $file = $config->config('SplitterFile');

    unless (-f $file) {
        carp "Cannot find splitter data file. Tried [$file]\n";
        return undef;
    }

    my $self = $class->SUPER::new($file);
    return bless $self => $class
}

=head2 C<split>

This is the only available method for the splitter object. It receives
a list of L<Lingua::FreeLing3::Word> objects (you can obtain one using
the L<Lingua::FreeLing3::Tokenizer>), and splits the text to a list of
sentences.

Without any further configuration option, it will return a reference
to a list of L<Lingua::FreeLing3::Sentence>. The option C<to_text> can
be set, and it will return a reference to a list of strings, where the
words/tokens will be separated by a simple space.

 $list_of_sentences = $pt_split->split($list_of_words, to_text => 1 )

The C<buffered> option can also be set to the value C<0> if the
function should not buffer tokens while processing. The default is to
buffer.

 $list_of_sentences = $pt_split->split($list_of_words, buffered => 0 )

B<NOTE:> Before exiting, your application you B<should> run the split
method without the buffered feature, so that all the text is really
processed!

=cut

sub split {
    my ($self, $tokens, %opts) = @_;

    unless (Lingua::FreeLing3::_is_word_list($tokens)) {
        carp "Error: split argument should be a list of words";
        return undef;
    }

    my $buffered = $opts{buffered} // 1;
    my $result = $self->SUPER::split($tokens, $buffered);

    for my $s (@$result) {
        $s->ACQUIRE();
        $s = Lingua::FreeLing3::Sentence->_new_from_binding($s);
        $s = $s->to_text if $opts{to_text};
    }

    return $result;
}

1;

__END__

=head1 SEE ALSO

Lingua::FreeLing3(3) for the documentation table of contents. The
freeling library for extra information, or perl(1) itself.

=head1 AUTHOR

Alberto Manuel Brandão Simões, E<lt>ambs@cpan.orgE<gt>