0
0
mirror of https://github.com/ezyang/htmlpurifier.git synced 2024-09-20 19:25:19 +00:00
htmlpurifier/library/HTMLPurifier/ChildDef/Required.php
Edward Z. Yang 0767bbc12d Rewrite FixNesting implementation to be tree-based.
This mega-patch rips out the FixNesting implementation and the related
ChildDef components.  The primary algorithmic change is to convert from
use of tokens to tree nodes, which are far more amenable to the style
of processing that FixNesting uses.  Additionally, FixNesting has been
changed to go bottom-up rather than top-down, in order to avoid needing
to implement backtracking.

This patch simplifies a good deal of the relevant logic, since we no
longer need to continually recalculate the nesting structure when
processing things.  However, the conversion to the alternate format
incurs some overhead, so for small inputs these changes are not a win.
One possibility to greatly reduce the constant factors here is to switch
to entirely using libxml's representation, and never serializing tokens;
this would require one to rewrite injectors, however.

The iterative post-order traversal in FixNesting is a bit subtle, but
we have essentially reified the stack and continuations.

We've removed support for %Core.EscapeInvalidChildren.

Signed-off-by: Edward Z. Yang <ezyang@mit.edu>
2013-10-20 22:37:01 -07:00

119 lines
3.3 KiB
PHP

<?php
/**
* Definition that allows a set of elements, but disallows empty children.
*/
class HTMLPurifier_ChildDef_Required extends HTMLPurifier_ChildDef
{
/**
* Lookup table of allowed elements.
* @type array
*/
public $elements = array();
/**
* Whether or not the last passed node was all whitespace.
* @type bool
*/
protected $whitespace = false;
/**
* @param array|string $elements List of allowed element names (lowercase).
*/
public function __construct($elements)
{
if (is_string($elements)) {
$elements = str_replace(' ', '', $elements);
$elements = explode('|', $elements);
}
$keys = array_keys($elements);
if ($keys == array_keys($keys)) {
$elements = array_flip($elements);
foreach ($elements as $i => $x) {
$elements[$i] = true;
if (empty($i)) {
unset($elements[$i]);
} // remove blank
}
}
$this->elements = $elements;
}
/**
* @type bool
*/
public $allow_empty = false;
/**
* @type string
*/
public $type = 'required';
/**
* @param array $children
* @param HTMLPurifier_Config $config
* @param HTMLPurifier_Context $context
* @return array
*/
public function validateChildren($children, $config, $context)
{
// Flag for subclasses
$this->whitespace = false;
// if there are no tokens, delete parent node
if (empty($children)) {
return false;
}
// the new set of children
$result = array();
// 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;
$stack = array_reverse($children);
while (!empty($stack)) {
$node = array_pop($stack);
if (!empty($node->is_whitespace)) {
$result[] = $node;
continue;
}
$all_whitespace = false; // phew, we're not talking about whitespace
if (!isset($this->elements[$node->name])) {
// special case text
// XXX One of these ought to be redundant or something
if ($pcdata_allowed && $node instanceof HTMLPurifier_Node_Text) {
$result[] = $node;
continue;
}
// spill the child contents in
// ToDo: Make configurable
if ($node instanceof HTMLPurifier_Node_Element) {
for ($i = count($node->children) - 1; $i >= 0; $i--) {
$stack[] = $node->children[$i];
}
continue;
}
continue;
}
$result[] = $node;
}
if (empty($result)) {
return false;
}
if ($all_whitespace) {
$this->whitespace = true;
return false;
}
return $result;
}
}
// vim: et sw=4 sts=4