library\search\CharacterFilter

Complexity

Total Complexity

6

Size/Duplication

Total Lines	73
Duplicated Lines	0 %

Coupling/Cohesion

Components	0
Dependencies	0

Importance

Changes

0

5 Methods

Rating	Name	Duplication	Size	Complexity
A	__construct()	0	8	1
A	__toString()	0	4	1
A	filterSpecialCharacters()	0	11	1
A	convertToUTF8()	0	11	2
A	getFilteredString()	0	4	1

<?php
/**
 * User: jensk
 * Date: 21-2-2017
 * Time: 10:44
 */

namespace library\search;

class CharacterFilter
{
	protected $originalString;
	protected $filteredString = '';

	/**
	 * CharacterFilter constructor.
	 *
	 * @param $string
	 */
	public function __construct($string)
	{
		$this->originalString = $string;
		$string = $this->convertToUTF8($string);
		$string = mb_strtolower($string);
		$string = $this->filterSpecialCharacters($string);
		$this->filteredString = $string;
	}

	/**
	 * Returns the filtered string
	 * @return string|void
	 */
	function __toString()

It is generally recommended to explicitly declare the visibility for methods.

	{
		return $this->filteredString;
	}

	/**
	 * Filter out all special characters, like punctuation and characters with accents
	 *
	 * @param $string
	 *
	 * @return mixed|string
	 */
	private function filterSpecialCharacters($string)
	{
		$string = strip_tags($string);
		$string = iconv('UTF-8', 'ASCII//TRANSLIT//IGNORE', $string); // Remove special alphanumeric characters
		$string = str_replace(array('+', '=', '!', ',', '.',';', ':', '?'), ' ', $string); // Replace sentence breaking charaters with spaces
		$string = preg_replace("/[\r\n]+/", " ", $string); // Replace multiple newlines with a single space.
		$string = preg_replace("/[\t]+/", " ", $string); // Replace multiple tabs with a single space.
		$string = preg_replace("/[^a-zA-Z0-9 ]/", '', $string); // Filter out everything that is not alphanumeric or a space
		$string = preg_replace('!\s+!', ' ', $string); // Replace multiple spaces with a single space
		return $string;
	}

	/**
	 * Convert the string to UTF-8 encoding
	 * @param $string
	 *
	 * @return string
	 */
	private function convertToUTF8($string)
	{
		$encoding = mb_detect_encoding($string, mb_detect_order(), false);

		if($encoding == "UTF-8") {
			$string = mb_convert_encoding($string, 'UTF-8', 'UTF-8');
		}

		$out = iconv(mb_detect_encoding($string, mb_detect_order(), false), "UTF-8//IGNORE", $string);
		return $out;
	}

	/**
	 * @return mixed|string
	 */
	public function getFilteredString()
	{
		return $this->filteredString;
	}
}