drupol\phpngrams\NGramsTrait

Complexity

Total Complexity

11

Size/Duplication

Total Lines	66
Duplicated Lines	0 %

Coupling/Cohesion

Dependencies

0

Test Coverage

Coverage

100%

Importance

Changes

0

3 Methods

Rating	Name	Duplication	Size	Complexity
A	frequency()	0	8	1
B	ngramsFactory()	0	18	5
B	doNgrams()	0	14	5

<?php

namespace drupol\phpngrams;

trait NGramsTrait
{

    /**
     * @param \Generator $ngrams
     * @param string $substring
     *
     * @return float|int
     */
    public function frequency(\Generator $ngrams, $substring)
    {
        $ngrams = iterator_to_array($ngrams);

Consider using a different name than the parameter $ngrams. This often makes code more readable.

        return count(array_filter($ngrams, function ($n) use ($substring) {

Avoid variables with short names like $n. Configured minimum length is 3.

            return $n === $substring;
        }))/count($ngrams);
    }

    /**
     * @param $data
     * @param int $n
     * @param bool $cyclic
     *
     * @return bool|\Generator

Consider making the return type a bit more specific; maybe use \Generator.

     */
    public function ngramsFactory($data, $n = 1, $cyclic = true)

Avoid variables with short names like $n. Configured minimum length is 3.

    {
        $ngrams = [];

        if (is_string($data)) {
            foreach ($this->doNgrams(str_split($data), $n, $cyclic) as $item) {
                yield implode('', $item);
            }
        }

        if (is_array($data)) {
            foreach ($this->doNgrams($data, $n, $cyclic) as $item) {
                yield $item;
            }
        }

        return $ngrams;
    }

    /**
     * @param $data
     * @param $n
     * @param $cyclic
     *
     * @return \Generator
     */
    private function doNgrams($data, $n = 1, $cyclic = true)

Avoid variables with short names like $n. Configured minimum length is 3.

    {
        $dataLength = count($data);
        $n = $n > $dataLength ? $dataLength : $n;

Consider using a different name than the parameter $n. This often makes code more readable.

        $length = (false === $cyclic ? $dataLength - $n + 1 : $dataLength);

        for ($j = 0; $j < $length; $j++) {
            $ngrams = [];
            for ($i = $j; $i < $n + $j; $i++) {
                $ngrams[] = $data[$i%$dataLength];
            }
            yield $ngrams;
        }
    }
}