File Coverage

blib/lib/Lingua/TermWeight/WordSegmenter/LetterNgram.pm

Criterion	Covered	Total	%
statement	29	29	100.0
branch	6	6	100.0
condition			n/a
subroutine	7	7	100.0
pod	0	1	0.0
total	42	43	97.6

# SPDX-FileCopyrightText: 2014 Koichi SATOH
# SPDX-FileCopyrightText: 2026 Wesley Schwengle
#
# SPDX-License-Identifier: MIT

package Lingua::TermWeight::WordSegmenter::LetterNgram;
our $VERSION = '0.01';
# ABSTRACT: Word segmenter

use v5.20;
use utf8;
use warnings;
use Object::Pad;
use Carp qw(croak);

class Lingua::TermWeight::WordSegmenter::LetterNgram {

  field $n : param;

  ADJUST {
    croak "Word length must be 1+" if !defined($n) || $n <= 0;
  }

  method n { $n }

  method segment ($document) {
    $document = \"$document" unless ref $document;

    my $length = length $$document;
    my $index  = -1;
    my $n      = $self->n;

    return sub {
      GET_NEXT_NGRAM: {
        ++$index;
        return if $index + $n > $length;

        my $ngram = substr $$document, $index, $n;
        redo GET_NEXT_NGRAM if $ngram =~ /\s/;
        return $ngram;
      }
    };
  }
}

1;

__END__