1: <?php
  2: 
  3: /**
  4:  * This file is part of the Nette Framework (https://nette.org)
  5:  * Copyright (c) 2004 David Grudl (https://davidgrudl.com)
  6:  */
  7: 
  8: namespace Nette\Utils;
  9: 
 10: use Nette;
 11: 
 12: 
 13: /**
 14:  * Simple lexical analyser. Internal class.
 15:  *
 16:  * @author     David Grudl
 17:  * @internal
 18:  */
 19: class Tokenizer extends Nette\Object
 20: {
 21:     const VALUE = 0,
 22:         OFFSET = 1,
 23:         TYPE = 2;
 24: 
 25:     /** @var string */
 26:     private $re;
 27: 
 28:     /** @var array */
 29:     private $types;
 30: 
 31: 
 32:     /**
 33:      * @param  array of [(int) symbol type => pattern]
 34:      * @param  string  regular expression flag
 35:      */
 36:     public function __construct(array $patterns, $flags = '')
 37:     {
 38:         $this->re = '~(' . implode(')|(', $patterns) . ')~A' . $flags;
 39:         $keys = array_keys($patterns);
 40:         $this->types = $keys === range(0, count($patterns) - 1) ? FALSE : $keys;
 41:     }
 42: 
 43: 
 44:     /**
 45:      * Tokenize string.
 46:      * @param  string
 47:      * @return array
 48:      */
 49:     public function tokenize($input)
 50:     {
 51:         if ($this->types) {
 52:             $tokens = Strings::matchAll($input, $this->re);
 53:             $len = 0;
 54:             $count = count($this->types);
 55:             foreach ($tokens as & $match) {
 56:                 $type = NULL;
 57:                 for ($i = 1; $i <= $count; $i++) {
 58:                     if (!isset($match[$i])) {
 59:                         break;
 60:                     } elseif ($match[$i] != NULL) {
 61:                         $type = $this->types[$i - 1]; break;
 62:                     }
 63:                 }
 64:                 $match = array(self::VALUE => $match[0], self::OFFSET => $len, self::TYPE => $type);
 65:                 $len += strlen($match[self::VALUE]);
 66:             }
 67:             if ($len !== strlen($input)) {
 68:                 $errorOffset = $len;
 69:             }
 70: 
 71:         } else {
 72:             $tokens = Strings::split($input, $this->re, PREG_SPLIT_NO_EMPTY | PREG_SPLIT_OFFSET_CAPTURE);
 73:             $last = end($tokens);
 74:             if ($tokens && !Strings::match($last[0], $this->re)) {
 75:                 $errorOffset = $last[1];
 76:             }
 77:         }
 78: 
 79:         if (isset($errorOffset)) {
 80:             list($line, $col) = $this->getCoordinates($input, $errorOffset);
 81:             $token = str_replace("\n", '\n', substr($input, $errorOffset, 10));
 82:             throw new TokenizerException("Unexpected '$token' on line $line, column $col.");
 83:         }
 84:         return $tokens;
 85:     }
 86: 
 87: 
 88:     /**
 89:      * Returns position of token in input string.
 90:      * @param  int token number
 91:      * @return array [line, column]
 92:      */
 93:     public static function getCoordinates($text, $offset)
 94:     {
 95:         $text = substr($text, 0, $offset);
 96:         return array(substr_count($text, "\n") + 1, $offset - strrpos("\n" . $text, "\n") + 1);
 97:     }
 98: 
 99: }
100: 
101: 
102: /**
103:  * The exception that indicates tokenizer error.
104:  */
105: class TokenizerException extends \Exception
106: {
107: }
108: