library\search\Tokenizer

Complexity

Total Complexity

7

Size/Duplication

Total Lines	46
Duplicated Lines	0 %

Coupling/Cohesion

Components	1
Dependencies	0

Importance

Changes	3
Bugs	0	Features	0

4 Methods

Rating	Name	Duplication	Size	Complexity
A	__construct()	0	5	1
A	tokenize()	0	7	2
A	addTokenToVector()	0	10	3
A	getTokenVector()	0	4	1

<?php
/**
 * User: jensk
 * Date: 21-2-2017
 * Time: 16:23
 */

namespace library\search;

/**
 * Class Tokenizer
 * @package library\search
 */
class Tokenizer
{
	protected $inputString;
	protected $tokenVector = array();

	/**
	 * Tokenizer constructor.
	 *
	 * @param string $string Should preferably be parsed wit \library\search\CharacterFilter
	 * @see \library\search\CharacterFilter
	 */
	public function __construct($string)
	{
		$this->inputString = $string;
		$this->tokenize();
	}

	protected function tokenize()
	{
		$tokens = explode(' ', $this->inputString);
		foreach ($tokens as $token) {
			$this->addTokenToVector($token);
		}
	}

	protected function addTokenToVector($token)
	{
		if (!empty($token)) {
			if (isset($this->tokenVector[$token])) {
				$this->tokenVector[$token] += 1;
			} else {
				$this->tokenVector[$token] = 1;
			}
		}
	}

	/**
	 * @return array
	 */
	public function getTokenVector()
	{
		return $this->tokenVector;
	}


}