Ngram tokenizer in php

NgramTokenizer.php

<?php

class NgramTokenizer
{
    public static function tokenize($word, $n)
    {
        if ($n === 1) {
            return str_split($word);
        }

        $grams = array();

        for ($i = 0; $i <= strlen($word) - $n; ++$i) {
            $grams[] = substr($word, $i, $n);
        }

        return $grams;
    }
}

NgramTokenizerTest.php

<?php

class NgramTokenizerTest extends \PHPUnit_Framework_TestCase
{
    public function testUnigram()
    {
        $tokens = NgramTokenizer::tokenize("abcd", 1);

        $this->assertEquals(range('a', 'd'), $tokens);
    }

    public function testBigram()
    {
        $tokens = NgramTokenizer::tokenize("abcd", 2);

        $this->assertEquals(array('ab', 'bc', 'cd'), $tokens);
    }

    public function testTrigram()
    {
        $tokens = NgramTokenizer::tokenize("abcdef", 3);

        $this->assertEquals(array('abc', 'bcd', 'cde', 'def'), $tokens);
    }
}

ASCII string only NGram tokenizer. No support for strings with unicode or breaking things up by certain characters like spaces.