[Summary view] [Print] [Text view]

   1  <?php
   2  /*
   3   * Limb PHP Framework
   4   *
   5   * @link http://limb-project.com 
   6   * @copyright  Copyright &copy; 2004-2007 BIT(http://bit-creative.com)
   7   * @license    LGPL http://www.gnu.org/copyleft/lesser.html 
   8   */
   9  
  10  /**

  11   * The scanner tokenizes input content ({@link input()})

  12   * and provides a "scanner" interface which can go forward

  13   * ({@link next()}) and backward (({@link back()}) )

  14   * along the tokens. It also correlates the toekens with line

  15   * number ({@link line()}).

  16   *

  17   * Part of the credits for this scanner is due to phpDocumentor,

  18   * from which I borrowed some ideas to make it work.

  19   *

  20   * @author Oak Nauhygon <ezpdo4php@gmail.com>

  21   * @package core
  22   * @version $Id$
  23   */
  24  class lmbPHPTokenizer
  25  {
  26    /**

  27     * List of tokens that can contain a newline

  28     * @var array

  29     */
  30    static public $newline_tokens = array(
  31        T_WHITESPACE,
  32        T_ENCAPSED_AND_WHITESPACE,
  33        T_COMMENT,
  34        T_DOC_COMMENT,
  35        T_OPEN_TAG,
  36        T_CLOSE_TAG,
  37        T_INLINE_HTML
  38        );
  39  
  40    /**

  41     * The input content

  42     * @var string

  43     */
  44    protected $input;
  45  
  46    /**

  47     * tokenized array from {@link token_get_all()}

  48     * @var array

  49     */
  50    protected $tokens;
  51  
  52    /**

  53     * current token position

  54     * @var integer

  55     */
  56    protected $pos = 0;
  57  
  58    /**

  59     * current source line number

  60     * @var integer

  61     */
  62    protected $line = 0;
  63  
  64    /**

  65     * Constructor

  66     * @param string input content

  67     */
  68    public function __construct($input = '') {
  69        if (!empty($input)) {
  70            $this->input($input);
  71        }
  72    }
  73  
  74    /**

  75     * get input if no param supplied or set input if param is a non-empty string

  76     * @param string input content

  77     * @return string|bool

  78     */
  79    public function input($input = false) {
  80  
  81        // if input is false, return

  82        if ($input === false) {
  83            return $this->input;
  84        }
  85  
  86        // trim the \r\n input content

  87        $input = rtrim(ltrim($input, "\r\n"));
  88        if (empty($input)) {
  89            return false;
  90        }
  91  
  92        // use reference to save memory

  93        $this->input = & $input;
  94  
  95        // unset the tokens so when next() is called the frist

  96        // time, it will call reset()

  97        unset($this->tokens);
  98  
  99        return true;
 100    }
 101  
 102    /**

 103     * Tokenize input content and reset current

 104     * token position and line number

 105     * @return void

 106     */
 107    public function reset() {
 108        $this->tokens = @token_get_all($this->input);
 109        $this->pos = 0;
 110        $this->line = 0;
 111    }
 112  
 113    /**

 114     * Fetch the next token

 115     * @return string|array token from tokenizer

 116     */
 117    function next() {
 118  
 119        // check if we need to reset (tokenize input)

 120        if (empty($this->tokens)) {
 121            $this->reset();
 122        }
 123  
 124        // check if token at the cur position set

 125        if ( !isset($this->tokens[$this->pos]) ) {
 126            return false;
 127        }
 128  
 129        // keep track of the old line

 130        $oldline = $this->line;
 131  
 132        // now get the current token

 133        $word = $this->tokens[$this->pos++];
 134  
 135        // correlate line and token

 136        if ( is_array($word) ) {
 137  
 138            // count line num for special tokens ({@link $newline_tokens})

 139            if ( in_array($word[0], lmbPHPTokenizer::$newline_tokens) ) {
 140                $this->line += substr_count($word[1], "\n");
 141            }
 142  
 143            // always skip whitespace

 144            if ( $word[0] == T_WHITESPACE )    {
 145                return $this->next();
 146            }
 147        }
 148  
 149        return $word;
 150    }
 151  
 152    /**

 153     * Go back one token (reverse of {@link next()})

 154     * @return false|string|array

 155     */
 156    public function back() {
 157  
 158        $this->pos --;
 159  
 160        // check if it's the beginning

 161        if ($this->pos < 0) {
 162            $this->pos = 0;
 163            return false;
 164        }
 165  
 166        $word = $this->tokens[$this->pos];
 167  
 168        if ( is_array($word) ) {
 169  
 170            if ( $word[0] == T_WHITESPACE )
 171                return $this->next();
 172  
 173            if ( in_array($word[0], lmbPHPTokenizer::$newline_tokens) ) {
 174                $this->line -= substr_count($word[1], "\n");
 175            }
 176        }
 177    }
 178  
 179    /**

 180     * Get the current line number

 181     * @return integer

 182     */
 183    public function line() {
 184        return $this->line;
 185    }
 186  }
 187  
 188  ?>