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