File Coverage

blib/lib/String/Normal/Type/Title.pm

Criterion	Covered	Total	%
statement	35	35	100.0
branch	5	6	83.3
condition	4	5	80.0
subroutine	7	7	100.0
pod	2	2	100.0
total	53	55	96.3

package String::Normal::Type::Title;
use strict;
use warnings;
use String::Normal::Type;
use String::Normal::Config;

use Lingua::Stem;
our $STEM;
our $title_stem;
our $title_stop;

sub transform {
    my ($self,$value) = @_;

    $value =~ s/\([^)]*\)/ /g if $value =~ /^[^(]|[^)]$/;

    $value = String::Normal::Type::_scrub_value( $value );

    # tokenize and stem
    my @tokens = ();
    for my $token (split ' ', $value) {
        #$token = defined( $title_stem->{$token} ) ? $title_stem->{$token} : $token;
        push @tokens, $token;
    }

    # Remove all middle stop words that are safe to remove, based on the number of
    # tokens, of course.
    my @filtered = map {
        my $count = $title_stop->{middle}{$_} || '';
        (length $count and @tokens >= $count) ? () : $_;
    } @tokens;

    # stem, but override if Stemmer "blanks out" token
    my @copy = @filtered;
    $STEM->stem_in_place( @copy );
    for my $i (0 .. $#copy) {
        $filtered[$i] = $copy[$i] unless $filtered[$i] =~ /\d/;
    }

    return join ' ', @filtered;
}

sub new {
    my $self = shift;
    $title_stem = String::Normal::Config::TitleStem::_data( @_ );
    $title_stop = String::Normal::Config::TitleStop::_data( @_ );
    $STEM = Lingua::Stem->new;
    $STEM->add_exceptions( $title_stem );
    return bless {@_}, $self;
}


1;

__END__