Using Doctrine's Lexer for own parsing purpose

Code
<?php

use Doctrine\Common\Lexer\AbstractLexer;

class 
MathLexer extends AbstractLexer
{
    
/**
     * better: a token for every operator
     * instead of T_OPERATOR
     */
    
const T_OPERATOR 1;
    
    const 
T_FORMULA 2;
    
     
/**
     * Lexical catchable patterns.
     *
     * @return array
     */
    
protected function getCatchablePatterns()
    {
        return array(
            
'[\+\-\/\*]+',
            
'[a-zA-Z]+',
        );
    }

    
/**
     * Lexical non-catchable patterns.
     *
     * @return array
     */
    
protected function getNonCatchablePatterns()
    {
        return array(
            
'\s+',
        );
    }

    
/**
     * Retrieve token type.
     *
     * @param string $value
     *
     * @return integer
     */
    
protected function getType(&$value)
    {
        if (
preg_match('/[\+\-\/\*]+/'$value)) {
            return 
self::T_OPERATOR;
        }
        
        return 
self::T_FORMULA;
    }
}

$lexer = new MathLexer();
$lexer->setInput('X + Y - Z');

$lexer->moveNext();
while(
null !== $lexer->lookahead) {
    
// lookahead is an array with token elements
    // value: value of token
    // position: position in initial string
    // type: integer (see constants)
    
echo '"' $lexer->lookahead['value'] . '"';
    echo 
' is of type ' $lexer->lookahead['type'] . PHP_EOL;
    
$lexer->moveNext();
}
Result
"X" is of type 2
"+" is of type 1
"Y" is of type 2
"-" is of type 1
"Z" is of type 2
Used Versions
PHP 7.1, Zend 3.0, Symfony 3.2, Laravel 5.2, PHPUnit 5.6, Doctrine 2.6