library/HTMLPurifier/Injector/RemoveEmpty.php

   1 <?php
   2
   3 class HTMLPurifier_Injector_RemoveEmpty extends HTMLPurifier_Injector
   4 {
   5     /**
   6      * @type HTMLPurifier_Context
   7      */
   8     private $context;
   9
  10     /**
  11      * @type HTMLPurifier_Config
  12      */
  13     private $config;
  14
  15     /**
  16      * @type HTMLPurifier_AttrValidator
  17      */
  18     private $attrValidator;
  19
  20     /**
  21      * @type bool
  22      */
  23     private $removeNbsp;
  24
  25     /**
  26      * @type bool
  27      */
  28     private $removeNbspExceptions;
  29
  30     /**
  31      * @type array
  32      * TODO: make me configurable
  33      */
  34     private $_exclude = array('colgroup' => 1, 'th' => 1, 'td' => 1, 'iframe' => 1);
  35
  36     /**
  37      * @param HTMLPurifier_Config $config
  38      * @param HTMLPurifier_Context $context
  39      * @return void
  40      */
  41     public function prepare($config, $context)
  42     {
  43         parent::prepare($config, $context);
  44         $this->config = $config;
  45         $this->context = $context;
  46         $this->removeNbsp = $config->get('AutoFormat.RemoveEmpty.RemoveNbsp');
  47         $this->removeNbspExceptions = $config->get('AutoFormat.RemoveEmpty.RemoveNbsp.Exceptions');
  48         $this->attrValidator = new HTMLPurifier_AttrValidator();
  49     }
  50
  51     /**
  52      * @param HTMLPurifier_Token $token
  53      */
  54     public function handleElement(&$token)
  55     {
  56         if (!$token instanceof HTMLPurifier_Token_Start) {
  57             return;
  58         }
  59         $next = false;
  60         for ($i = $this->inputIndex + 1, $c = count($this->inputTokens); $i < $c; $i++) {
  61             $next = $this->inputTokens[$i];
  62             if ($next instanceof HTMLPurifier_Token_Text) {
  63                 if ($next->is_whitespace) {
  64                     continue;
  65                 }
  66                 if ($this->removeNbsp && !isset($this->removeNbspExceptions[$token->name])) {
  67                     $plain = str_replace("\xC2\xA0", "", $next->data);
  68                     $isWsOrNbsp = $plain === '' || ctype_space($plain);
  69                     if ($isWsOrNbsp) {
  70                         continue;
  71                     }
  72                 }
  73             }
  74             break;
  75         }
  76         if (!$next || ($next instanceof HTMLPurifier_Token_End && $next->name == $token->name)) {
  77             if (isset($this->_exclude[$token->name])) {
  78                 return;
  79             }
  80             $this->attrValidator->validateToken($token, $this->config, $this->context);
  81             $token->armor['ValidateAttributes'] = true;
  82             if (isset($token->attr['id']) || isset($token->attr['name'])) {
  83                 return;
  84             }
  85             $token = $i - $this->inputIndex + 1;
  86             for ($b = $this->inputIndex - 1; $b > 0; $b--) {
  87                 $prev = $this->inputTokens[$b];
  88                 if ($prev instanceof HTMLPurifier_Token_Text && $prev->is_whitespace) {
  89                     continue;
  90                 }
  91                 break;
  92             }
  93             // This is safe because we removed the token that triggered this.
  94             $this->rewind($b - 1);
  95             return;
  96         }
  97     }
  98 }
  99
 100 // vim: et sw=4 sts=4