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      * Cached contents of %AutoFormat.RemoveEmpty.Predicate
  32      * @type array
  33      */
  34     private $exclude;
  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->exclude = $config->get('AutoFormat.RemoveEmpty.Predicate');
  49         foreach ($this->exclude as $key => $attrs) {
  50             if (!is_array($attrs)) {
  51                 // HACK, see HTMLPurifier/Printer/ConfigForm.php
  52                 $this->exclude[$key] = explode(';', $attrs);
  53             }
  54         }
  55         $this->attrValidator = new HTMLPurifier_AttrValidator();
  56     }
  57
  58     /**
  59      * @param HTMLPurifier_Token $token
  60      */
  61     public function handleElement(&$token)
  62     {
  63         if (!$token instanceof HTMLPurifier_Token_Start) {
  64             return;
  65         }
  66         $next = false;
  67         $deleted = 1; // the current tag
  68         for ($i = count($this->inputZipper->back) - 1; $i >= 0; $i--, $deleted++) {
  69             $next = $this->inputZipper->back[$i];
  70             if ($next instanceof HTMLPurifier_Token_Text) {
  71                 if ($next->is_whitespace) {
  72                     continue;
  73                 }
  74                 if ($this->removeNbsp && !isset($this->removeNbspExceptions[$token->name])) {
  75                     $plain = str_replace("\xC2\xA0", "", $next->data);
  76                     $isWsOrNbsp = $plain === '' || ctype_space($plain);
  77                     if ($isWsOrNbsp) {
  78                         continue;
  79                     }
  80                 }
  81             }
  82             break;
  83         }
  84         if (!$next || ($next instanceof HTMLPurifier_Token_End && $next->name == $token->name)) {
  85             $this->attrValidator->validateToken($token, $this->config, $this->context);
  86             $token->armor['ValidateAttributes'] = true;
  87             if (isset($this->exclude[$token->name])) {
  88                 $r = true;
  89                 foreach ($this->exclude[$token->name] as $elem) {
  90                     if (!isset($token->attr[$elem])) $r = false;
  91                 }
  92                 if ($r) return;
  93             }
  94             if (isset($token->attr['id']) || isset($token->attr['name'])) {
  95                 return;
  96             }
  97             $token = $deleted + 1;
  98             for ($b = 0, $c = count($this->inputZipper->front); $b < $c; $b++) {
  99                 $prev = $this->inputZipper->front[$b];
 100                 if ($prev instanceof HTMLPurifier_Token_Text && $prev->is_whitespace) {
 101                     continue;
 102                 }
 103                 break;
 104             }
 105             // This is safe because we removed the token that triggered this.
 106             $this->rewindOffset($b+$deleted);
 107             return;
 108         }
 109     }
 110 }
 111
 112 // vim: et sw=4 sts=4