mirror of
https://github.com/ezyang/htmlpurifier.git
synced 2025-10-17 06:56:06 +02:00
Split out code in Definition.php .
git-svn-id: http://htmlpurifier.org/svnroot/htmlpurifier/trunk@73 48356398-32a2-884e-a903-53898d9a118a
This commit is contained in:
Edward Z. Yang
194
Definition.php
194
Definition.php
@@ -85,7 +85,7 @@ class HTMLPurifier_Definition
|
||||
$this->info['blockquote'] =
|
||||
$this->info['dd'] =
|
||||
$this->info['li'] =
|
||||
$this->info['div'] = new HTMLDTD_Element($e_Flow);
|
||||
$this->info['div'] = new HTMLPurifier_ElementDef($e_Flow);
|
||||
|
||||
$this->info['em'] =
|
||||
$this->info['strong'] =
|
||||
@@ -121,31 +121,31 @@ class HTMLPurifier_Definition
|
||||
$this->info['h3'] =
|
||||
$this->info['h4'] =
|
||||
$this->info['h5'] =
|
||||
$this->info['h6'] = new HTMLDTD_Element($e_Inline);
|
||||
$this->info['h6'] = new HTMLPurifier_ElementDef($e_Inline);
|
||||
|
||||
$this->info['ol'] =
|
||||
$this->info['ul'] =
|
||||
new HTMLDTD_Element(
|
||||
new HTMLPurifier_ElementDef(
|
||||
new HTMLPurifier_ChildDef_Required('li')
|
||||
);
|
||||
|
||||
$this->info['dl'] =
|
||||
new HTMLDTD_Element(
|
||||
new HTMLPurifier_ElementDef(
|
||||
new HTMLPurifier_ChildDef_Required('dt|dd')
|
||||
);
|
||||
$this->info['address'] =
|
||||
new HTMLDTD_Element(
|
||||
new HTMLPurifier_ElementDef(
|
||||
new HTMLPurifier_ChildDef_Optional("#PCDATA | p | $e_inline".
|
||||
" | $e_misc_inline")
|
||||
);
|
||||
|
||||
$this->info['img'] =
|
||||
$this->info['br'] =
|
||||
$this->info['hr'] = new HTMLDTD_Element(new HTMLPurifier_ChildDef_Empty());
|
||||
$this->info['hr'] = new HTMLPurifier_ElementDef(new HTMLPurifier_ChildDef_Empty());
|
||||
|
||||
$this->info['pre'] = new HTMLDTD_Element($e_pre_content);
|
||||
$this->info['pre'] = new HTMLPurifier_ElementDef($e_pre_content);
|
||||
|
||||
$this->info['a'] = new HTMLDTD_Element($e_a_content);
|
||||
$this->info['a'] = new HTMLPurifier_ElementDef($e_a_content);
|
||||
|
||||
}
|
||||
|
||||
@@ -424,191 +424,17 @@ class HTMLPurifier_Definition
|
||||
|
||||
}
|
||||
|
||||
class HTMLDTD_Element
|
||||
class HTMLPurifier_ElementDef
|
||||
{
|
||||
|
||||
var $child_def;
|
||||
var $attr_def = array();
|
||||
|
||||
function HTMLDTD_Element($child_def, $attr_def = array()) {
|
||||
function HTMLPurifier_ElementDef($child_def, $attr_def = array()) {
|
||||
$this->child_def = $child_def;
|
||||
$this->attr_def = $attr_def;
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
// HTMLPurifier_ChildDef and inheritance have three types of output:
|
||||
// true = leave nodes as is
|
||||
// false = delete parent node and all children
|
||||
// array(...) = replace children nodes with these
|
||||
|
||||
// this is the hardest one to implement. We'll use fancy regexp tricks
|
||||
// right now, we only expect it to return TRUE or FALSE (it won't attempt
|
||||
// to fix the tree)
|
||||
|
||||
// we may end up writing custom code for each HTML case
|
||||
// in order to make it self correcting
|
||||
class HTMLPurifier_ChildDef
|
||||
{
|
||||
var $type = 'custom';
|
||||
var $dtd_regex;
|
||||
var $_pcre_regex;
|
||||
function HTMLPurifier_ChildDef($dtd_regex) {
|
||||
$this->dtd_regex = $dtd_regex;
|
||||
$this->_compileRegex();
|
||||
}
|
||||
function _compileRegex() {
|
||||
$raw = str_replace(' ', '', $this->dtd_regex);
|
||||
if ($raw{0} != '(') {
|
||||
$raw = "($raw)";
|
||||
}
|
||||
$reg = str_replace(',', ',?', $raw);
|
||||
$reg = preg_replace('/([#a-zA-Z0-9_.-]+)/', '(,?\\0)', $reg);
|
||||
$this->_pcre_regex = $reg;
|
||||
}
|
||||
function validateChildren($tokens_of_children) {
|
||||
$list_of_children = '';
|
||||
$nesting = 0; // depth into the nest
|
||||
foreach ($tokens_of_children as $token) {
|
||||
if (!empty($token->is_whitespace)) continue;
|
||||
|
||||
$is_child = ($nesting == 0); // direct
|
||||
|
||||
if ($token->type == 'start') {
|
||||
$nesting++;
|
||||
} elseif ($token->type == 'end') {
|
||||
$nesting--;
|
||||
}
|
||||
|
||||
if ($is_child) {
|
||||
$list_of_children .= $token->name . ',';
|
||||
}
|
||||
}
|
||||
$list_of_children = rtrim($list_of_children, ',');
|
||||
|
||||
$okay =
|
||||
preg_match(
|
||||
'/^'.$this->_pcre_regex.'$/',
|
||||
$list_of_children
|
||||
);
|
||||
|
||||
return (bool) $okay;
|
||||
}
|
||||
}
|
||||
class HTMLPurifier_ChildDef_Simple extends HTMLPurifier_ChildDef
|
||||
{
|
||||
var $elements = array();
|
||||
function HTMLPurifier_ChildDef_Simple($elements) {
|
||||
if (is_string($elements)) {
|
||||
$elements = str_replace(' ', '', $elements);
|
||||
$elements = explode('|', $elements);
|
||||
}
|
||||
$elements = array_flip($elements);
|
||||
foreach ($elements as $i => $x) $elements[$i] = true;
|
||||
$this->elements = $elements;
|
||||
$this->gen = new HTMLPurifier_Generator();
|
||||
}
|
||||
function validateChildren() {
|
||||
trigger_error('Cannot call abstract function!', E_USER_ERROR);
|
||||
}
|
||||
}
|
||||
class HTMLPurifier_ChildDef_Required extends HTMLPurifier_ChildDef_Simple
|
||||
{
|
||||
var $type = 'required';
|
||||
function validateChildren($tokens_of_children) {
|
||||
// if there are no tokens, delete parent node
|
||||
if (empty($tokens_of_children)) return false;
|
||||
|
||||
// the new set of children
|
||||
$result = array();
|
||||
|
||||
// current depth into the nest
|
||||
$nesting = 0;
|
||||
|
||||
// whether or not we're deleting a node
|
||||
$is_deleting = false;
|
||||
|
||||
// whether or not parsed character data is allowed
|
||||
// this controls whether or not we silently drop a tag
|
||||
// or generate escaped HTML from it
|
||||
$pcdata_allowed = isset($this->elements['#PCDATA']);
|
||||
|
||||
// a little sanity check to make sure it's not ALL whitespace
|
||||
$all_whitespace = true;
|
||||
|
||||
foreach ($tokens_of_children as $token) {
|
||||
if (!empty($token->is_whitespace)) {
|
||||
$result[] = $token;
|
||||
continue;
|
||||
}
|
||||
$all_whitespace = false; // phew, we're not talking about whitespace
|
||||
|
||||
$is_child = ($nesting == 0);
|
||||
|
||||
if ($token->type == 'start') {
|
||||
$nesting++;
|
||||
} elseif ($token->type == 'end') {
|
||||
$nesting--;
|
||||
}
|
||||
|
||||
if ($is_child) {
|
||||
$is_deleting = false;
|
||||
if (!isset($this->elements[$token->name])) {
|
||||
$is_deleting = true;
|
||||
if ($pcdata_allowed) {
|
||||
$result[] = new HTMLPurifier_Token_Text(
|
||||
$this->gen->generateFromToken($token)
|
||||
);
|
||||
}
|
||||
continue;
|
||||
}
|
||||
}
|
||||
if (!$is_deleting) {
|
||||
$result[] = $token;
|
||||
} elseif ($pcdata_allowed) {
|
||||
$result[] =
|
||||
new HTMLPurifier_Token_Text(
|
||||
$this->gen->generateFromToken( $token )
|
||||
);
|
||||
} else {
|
||||
// drop silently
|
||||
}
|
||||
}
|
||||
if (empty($result)) return false;
|
||||
if ($all_whitespace) return false;
|
||||
if ($tokens_of_children == $result) return true;
|
||||
return $result;
|
||||
}
|
||||
}
|
||||
|
||||
// only altered behavior is that it returns an empty array
|
||||
// instead of a false (to delete the node)
|
||||
class HTMLPurifier_ChildDef_Optional extends HTMLPurifier_ChildDef_Required
|
||||
{
|
||||
var $type = 'optional';
|
||||
function validateChildren($tokens_of_children) {
|
||||
$result = parent::validateChildren($tokens_of_children);
|
||||
if ($result === false) return array();
|
||||
return $result;
|
||||
}
|
||||
}
|
||||
|
||||
// placeholder
|
||||
class HTMLPurifier_ChildDef_Empty extends HTMLPurifier_ChildDef
|
||||
{
|
||||
var $type = 'empty';
|
||||
function HTMLPurifier_ChildDef_Empty() {}
|
||||
function validateChildren() {
|
||||
return false;
|
||||
}
|
||||
}
|
||||
|
||||
class HTMLPurifier_AttrDef
|
||||
{
|
||||
var $def;
|
||||
function HTMLPurifier_AttrDef($def) {
|
||||
$this->def = $def;
|
||||
}
|
||||
}
|
||||
|
||||
?>
|
||||
Reference in New Issue
Block a user