mirror of
https://github.com/ezyang/htmlpurifier.git
synced 2025-08-03 20:58:11 +02:00
Release 1.6.0, merged in r875-930.
git-svn-id: http://htmlpurifier.org/svnroot/htmlpurifier/branches/strict@931 48356398-32a2-884e-a903-53898d9a118a
This commit is contained in:
@@ -43,6 +43,14 @@ HTMLPurifier_ConfigSchema::define(
|
||||
'is set to a non-empty value! This directive was available since 1.2.0.'
|
||||
);
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'Attr', 'IDBlacklistRegexp', null, 'string/null',
|
||||
'PCRE regular expression to be matched against all IDs. If the expression '.
|
||||
'is matches, the ID is rejected. Use this with care: may cause '.
|
||||
'significant degradation. ID matching is done after all other '.
|
||||
'validation. This directive was available since 1.6.0.'
|
||||
);
|
||||
|
||||
/**
|
||||
* Validates the HTML attribute ID.
|
||||
* @warning Even though this is the id processor, it
|
||||
@@ -94,6 +102,11 @@ class HTMLPurifier_AttrDef_HTML_ID extends HTMLPurifier_AttrDef
|
||||
$result = ($trim === '');
|
||||
}
|
||||
|
||||
$regexp = $config->get('Attr', 'IDBlacklistRegexp');
|
||||
if ($regexp && preg_match($regexp, $id)) {
|
||||
return false;
|
||||
}
|
||||
|
||||
if (/*!$this->ref && */$result) $id_accumulator->add($id);
|
||||
|
||||
// if no change was made to the ID, return the result
|
||||
|
75
library/HTMLPurifier/AttrDef/HTML/LinkTypes.php
Normal file
75
library/HTMLPurifier/AttrDef/HTML/LinkTypes.php
Normal file
@@ -0,0 +1,75 @@
|
||||
<?php
|
||||
|
||||
require_once 'HTMLPurifier/AttrDef.php';
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'Attr', 'AllowedRel', array(), 'lookup',
|
||||
'List of allowed forward document relationships in the rel attribute. '.
|
||||
'Common values may be nofollow or print. By default, this is empty, '.
|
||||
'meaning that no document relationships are allowed. This directive '.
|
||||
'was available since 1.6.0.'
|
||||
);
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'Attr', 'AllowedRev', array(), 'lookup',
|
||||
'List of allowed reverse document relationships in the rev attribute. '.
|
||||
'This attribute is a bit of an edge-case; if you don\'t know what it '.
|
||||
'is for, stay away. This directive was available since 1.6.0.'
|
||||
);
|
||||
|
||||
/**
|
||||
* Validates a rel/rev link attribute against a directive of allowed values
|
||||
* @note We cannot use Enum because link types allow multiple
|
||||
* values.
|
||||
* @note Assumes link types are ASCII text
|
||||
*/
|
||||
class HTMLPurifier_AttrDef_HTML_LinkTypes extends HTMLPurifier_AttrDef
|
||||
{
|
||||
|
||||
/** Lookup array of attribute names to configuration name */
|
||||
var $configLookup = array(
|
||||
'rel' => 'AllowedRel',
|
||||
'rev' => 'AllowedRev'
|
||||
);
|
||||
|
||||
/** Name config attribute to pull. */
|
||||
var $name;
|
||||
|
||||
function HTMLPurifier_AttrDef_HTML_LinkTypes($name) {
|
||||
if (!isset($this->configLookup[$name])) {
|
||||
trigger_error('Unrecognized attribute name for link '.
|
||||
'relationship.', E_USER_ERROR);
|
||||
return;
|
||||
}
|
||||
$this->name = $this->configLookup[$name];
|
||||
}
|
||||
|
||||
function validate($string, $config, &$context) {
|
||||
|
||||
$allowed = $config->get('Attr', $this->name);
|
||||
if (empty($allowed)) return false;
|
||||
|
||||
$string = $this->parseCDATA($string);
|
||||
$parts = explode(' ', $string);
|
||||
|
||||
// lookup to prevent duplicates
|
||||
$ret_lookup = array();
|
||||
foreach ($parts as $part) {
|
||||
$part = strtolower(trim($part));
|
||||
if (!isset($allowed[$part])) continue;
|
||||
$ret_lookup[$part] = true;
|
||||
}
|
||||
|
||||
if (empty($ret_lookup)) return false;
|
||||
|
||||
$ret_array = array();
|
||||
foreach ($ret_lookup as $part => $bool) $ret_array[] = $part;
|
||||
$string = implode(' ', $ret_array);
|
||||
|
||||
return $string;
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
?>
|
28
library/HTMLPurifier/AttrTransform/BgColor.php
Normal file
28
library/HTMLPurifier/AttrTransform/BgColor.php
Normal file
@@ -0,0 +1,28 @@
|
||||
<?php
|
||||
|
||||
require_once 'HTMLPurifier/AttrTransform.php';
|
||||
|
||||
/**
|
||||
* Pre-transform that changes deprecated bgcolor attribute to CSS.
|
||||
*/
|
||||
class HTMLPurifier_AttrTransform_BgColor
|
||||
extends HTMLPurifier_AttrTransform {
|
||||
|
||||
function transform($attr, $config, &$context) {
|
||||
|
||||
if (!isset($attr['bgcolor'])) return $attr;
|
||||
|
||||
$bgcolor = $attr['bgcolor'];
|
||||
unset($attr['bgcolor']);
|
||||
// some validation should happen here
|
||||
|
||||
$attr['style'] = isset($attr['style']) ? $attr['style'] : '';
|
||||
$attr['style'] = "background-color:$bgcolor;" . $attr['style'];
|
||||
|
||||
return $attr;
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
?>
|
28
library/HTMLPurifier/AttrTransform/Border.php
Normal file
28
library/HTMLPurifier/AttrTransform/Border.php
Normal file
@@ -0,0 +1,28 @@
|
||||
<?php
|
||||
|
||||
require_once 'HTMLPurifier/AttrTransform.php';
|
||||
|
||||
/**
|
||||
* Pre-transform that changes deprecated border attribute to CSS.
|
||||
*/
|
||||
class HTMLPurifier_AttrTransform_Border
|
||||
extends HTMLPurifier_AttrTransform {
|
||||
|
||||
function transform($attr, $config, &$context) {
|
||||
|
||||
if (!isset($attr['border'])) return $attr;
|
||||
|
||||
$border_width = $attr['border'];
|
||||
unset($attr['border']);
|
||||
// some validation should happen here
|
||||
|
||||
$attr['style'] = isset($attr['style']) ? $attr['style'] : '';
|
||||
$attr['style'] = "border:{$border_width}px solid;" . $attr['style'];
|
||||
|
||||
return $attr;
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
?>
|
33
library/HTMLPurifier/AttrTransform/Length.php
Normal file
33
library/HTMLPurifier/AttrTransform/Length.php
Normal file
@@ -0,0 +1,33 @@
|
||||
<?php
|
||||
|
||||
require_once 'HTMLPurifier/AttrTransform.php';
|
||||
|
||||
/**
|
||||
* Class for handling width/height length attribute transformations to CSS
|
||||
*/
|
||||
class HTMLPurifier_AttrTransform_Length extends HTMLPurifier_AttrTransform
|
||||
{
|
||||
|
||||
var $name;
|
||||
var $cssName;
|
||||
|
||||
function HTMLPurifier_AttrTransform_Length($name, $css_name = null) {
|
||||
$this->name = $name;
|
||||
$this->cssName = $css_name ? $css_name : $name;
|
||||
}
|
||||
|
||||
function transform($attr, $config, &$context) {
|
||||
if (!isset($attr[$this->name])) return $attr;
|
||||
$length = $attr[$this->name];
|
||||
unset($attr[$this->name]);
|
||||
if(ctype_digit($length)) $length .= 'px';
|
||||
|
||||
$attr['style'] = isset($attr['style']) ? $attr['style'] : '';
|
||||
$attr['style'] = $this->cssName . ":$length;" . $attr['style'];
|
||||
|
||||
return $attr;
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
?>
|
31
library/HTMLPurifier/AttrTransform/Name.php
Normal file
31
library/HTMLPurifier/AttrTransform/Name.php
Normal file
@@ -0,0 +1,31 @@
|
||||
<?php
|
||||
|
||||
require_once 'HTMLPurifier/AttrTransform.php';
|
||||
|
||||
/**
|
||||
* Pre-transform that changes deprecated name attribute to ID if necessary
|
||||
*/
|
||||
class HTMLPurifier_AttrTransform_Name extends HTMLPurifier_AttrTransform
|
||||
{
|
||||
|
||||
function transform($attr, $config, &$context) {
|
||||
|
||||
if (!isset($attr['name'])) return $attr;
|
||||
|
||||
$name = $attr['name'];
|
||||
unset($attr['name']);
|
||||
|
||||
if (isset($attr['id'])) {
|
||||
// ID already set, discard name
|
||||
return $attr;
|
||||
}
|
||||
|
||||
$attr['id'] = $name;
|
||||
|
||||
return $attr;
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
?>
|
@@ -6,7 +6,7 @@ require_once 'HTMLPurifier/AttrTransform.php';
|
||||
* Pre-transform that changes deprecated align attribute to text-align.
|
||||
*/
|
||||
class HTMLPurifier_AttrTransform_TextAlign
|
||||
extends HTMLPurifier_AttrTransform {
|
||||
extends HTMLPurifier_AttrTransform {
|
||||
|
||||
function transform($attr, $config, &$context) {
|
||||
|
||||
|
@@ -162,7 +162,9 @@ class HTMLPurifier_CSSDefinition
|
||||
new HTMLPurifier_AttrDef_CSS_Percentage()
|
||||
));
|
||||
|
||||
$this->info['width'] = new HTMLPurifier_AttrDef_CSS_Composite(array(
|
||||
$this->info['width'] =
|
||||
$this->info['height'] =
|
||||
new HTMLPurifier_AttrDef_CSS_Composite(array(
|
||||
new HTMLPurifier_AttrDef_CSS_Length(true),
|
||||
new HTMLPurifier_AttrDef_CSS_Percentage(true),
|
||||
new HTMLPurifier_AttrDef_Enum(array('auto'))
|
||||
|
@@ -1,250 +1,281 @@
|
||||
<?php
|
||||
|
||||
// components
|
||||
require_once 'HTMLPurifier/HTMLModuleManager.php';
|
||||
|
||||
// this definition and its modules MUST NOT define configuration directives
|
||||
// outside of the HTML or Attr namespaces
|
||||
|
||||
// will be superceded by more accurate doctype declaration schemes
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'HTML', 'Strict', false, 'bool',
|
||||
'Determines whether or not to use Transitional (loose) or Strict rulesets. '.
|
||||
'This directive has been available since 1.3.0.'
|
||||
);
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'HTML', 'BlockWrapper', 'p', 'string',
|
||||
'String name of element to wrap inline elements that are inside a block '.
|
||||
'context. This only occurs in the children of blockquote in strict mode. '.
|
||||
'Example: by default value, <code><blockquote>Foo</blockquote></code> '.
|
||||
'would become <code><blockquote><p>Foo</p></blockquote></code>. The '.
|
||||
'<code><p></code> tags can be replaced '.
|
||||
'with whatever you desire, as long as it is a block level element. '.
|
||||
'This directive has been available since 1.3.0.'
|
||||
);
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'HTML', 'Parent', 'div', 'string',
|
||||
'String name of element that HTML fragment passed to library will be '.
|
||||
'inserted in. An interesting variation would be using span as the '.
|
||||
'parent element, meaning that only inline tags would be allowed. '.
|
||||
'This directive has been available since 1.3.0.'
|
||||
);
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'HTML', 'AllowedElements', null, 'lookup/null',
|
||||
'If HTML Purifier\'s tag set is unsatisfactory for your needs, you '.
|
||||
'can overload it with your own list of tags to allow. Note that this '.
|
||||
'method is subtractive: it does its job by taking away from HTML Purifier '.
|
||||
'usual feature set, so you cannot add a tag that HTML Purifier never '.
|
||||
'supported in the first place (like embed, form or head). If you change this, you '.
|
||||
'probably also want to change %HTML.AllowedAttributes. '.
|
||||
'<strong>Warning:</strong> If another directive conflicts with the '.
|
||||
'elements here, <em>that</em> directive will win and override. '.
|
||||
'This directive has been available since 1.3.0.'
|
||||
);
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'HTML', 'AllowedAttributes', null, 'lookup/null',
|
||||
'IF HTML Purifier\'s attribute set is unsatisfactory, overload it! '.
|
||||
'The syntax is \'tag.attr\' or \'*.attr\' for the global attributes '.
|
||||
'(style, id, class, dir, lang, xml:lang).'.
|
||||
'<strong>Warning:</strong> If another directive conflicts with the '.
|
||||
'elements here, <em>that</em> directive will win and override. For '.
|
||||
'example, %HTML.EnableAttrID will take precedence over *.id in this '.
|
||||
'directive. You must set that directive to true before you can use '.
|
||||
'IDs at all. This directive has been available since 1.3.0.'
|
||||
);
|
||||
|
||||
/**
|
||||
* Definition of the purified HTML that describes allowed children,
|
||||
* attributes, and many other things.
|
||||
*
|
||||
* Conventions:
|
||||
*
|
||||
* All member variables that are prefixed with info
|
||||
* (including the main $info array) are used by HTML Purifier internals
|
||||
* and should not be directly edited when customizing the HTMLDefinition.
|
||||
* They can usually be set via configuration directives or custom
|
||||
* modules.
|
||||
*
|
||||
* On the other hand, member variables without the info prefix are used
|
||||
* internally by the HTMLDefinition and MUST NOT be used by other HTML
|
||||
* Purifier internals. Many of them, however, are public, and may be
|
||||
* edited by userspace code to tweak the behavior of HTMLDefinition.
|
||||
*
|
||||
* HTMLPurifier_Printer_HTMLDefinition is a notable exception to this
|
||||
* rule: in the interest of comprehensiveness, it will sniff everything.
|
||||
*/
|
||||
class HTMLPurifier_HTMLDefinition
|
||||
{
|
||||
|
||||
/** FULLY-PUBLIC VARIABLES */
|
||||
|
||||
/**
|
||||
* Associative array of element names to HTMLPurifier_ElementDef
|
||||
* @public
|
||||
*/
|
||||
var $info = array();
|
||||
|
||||
/**
|
||||
* Associative array of global attribute name to attribute definition.
|
||||
* @public
|
||||
*/
|
||||
var $info_global_attr = array();
|
||||
|
||||
/**
|
||||
* String name of parent element HTML will be going into.
|
||||
* @public
|
||||
*/
|
||||
var $info_parent = 'div';
|
||||
|
||||
/**
|
||||
* Definition for parent element, allows parent element to be a
|
||||
* tag that's not allowed inside the HTML fragment.
|
||||
* @public
|
||||
*/
|
||||
var $info_parent_def;
|
||||
|
||||
/**
|
||||
* String name of element used to wrap inline elements in block context
|
||||
* @note This is rarely used except for BLOCKQUOTEs in strict mode
|
||||
* @public
|
||||
*/
|
||||
var $info_block_wrapper = 'p';
|
||||
|
||||
/**
|
||||
* Associative array of deprecated tag name to HTMLPurifier_TagTransform
|
||||
* @public
|
||||
*/
|
||||
var $info_tag_transform = array();
|
||||
|
||||
/**
|
||||
* Indexed list of HTMLPurifier_AttrTransform to be performed before validation.
|
||||
* @public
|
||||
*/
|
||||
var $info_attr_transform_pre = array();
|
||||
|
||||
/**
|
||||
* Indexed list of HTMLPurifier_AttrTransform to be performed after validation.
|
||||
* @public
|
||||
*/
|
||||
var $info_attr_transform_post = array();
|
||||
|
||||
/**
|
||||
* Nested lookup array of content set name (Block, Inline) to
|
||||
* element name to whether or not it belongs in that content set.
|
||||
* @public
|
||||
*/
|
||||
var $info_content_sets = array();
|
||||
|
||||
|
||||
|
||||
/** PUBLIC BUT INTERNAL VARIABLES */
|
||||
|
||||
var $setup = false; /**< Has setup() been called yet? */
|
||||
var $config; /**< Temporary instance of HTMLPurifier_Config */
|
||||
|
||||
var $manager; /**< Instance of HTMLPurifier_HTMLModuleManager */
|
||||
|
||||
/**
|
||||
* Performs low-cost, preliminary initialization.
|
||||
* @param $config Instance of HTMLPurifier_Config
|
||||
*/
|
||||
function HTMLPurifier_HTMLDefinition(&$config) {
|
||||
$this->config =& $config;
|
||||
$this->manager = new HTMLPurifier_HTMLModuleManager();
|
||||
}
|
||||
|
||||
/**
|
||||
* Processes internals into form usable by HTMLPurifier internals.
|
||||
* Modifying the definition after calling this function should not
|
||||
* be done.
|
||||
*/
|
||||
function setup() {
|
||||
|
||||
// multiple call guard
|
||||
if ($this->setup) {return;} else {$this->setup = true;}
|
||||
|
||||
$this->processModules();
|
||||
$this->setupConfigStuff();
|
||||
|
||||
unset($this->config);
|
||||
unset($this->manager);
|
||||
|
||||
}
|
||||
|
||||
/**
|
||||
* Extract out the information from the manager
|
||||
*/
|
||||
function processModules() {
|
||||
|
||||
$this->manager->setup($this->config);
|
||||
|
||||
foreach ($this->manager->activeModules as $module) {
|
||||
foreach($module->info_tag_transform as $k => $v) $this->info_tag_transform[$k] = $v;
|
||||
foreach($module->info_attr_transform_pre as $k => $v) $this->info_attr_transform_pre[$k] = $v;
|
||||
foreach($module->info_attr_transform_post as $k => $v) $this->info_attr_transform_post[$k]= $v;
|
||||
}
|
||||
|
||||
$this->info = $this->manager->getElements($this->config);
|
||||
$this->info_content_sets = $this->manager->contentSets->lookup;
|
||||
|
||||
}
|
||||
|
||||
/**
|
||||
* Sets up stuff based on config. We need a better way of doing this.
|
||||
*/
|
||||
function setupConfigStuff() {
|
||||
|
||||
$block_wrapper = $this->config->get('HTML', 'BlockWrapper');
|
||||
if (isset($this->info_content_sets['Block'][$block_wrapper])) {
|
||||
$this->info_block_wrapper = $block_wrapper;
|
||||
} else {
|
||||
trigger_error('Cannot use non-block element as block wrapper.',
|
||||
E_USER_ERROR);
|
||||
}
|
||||
|
||||
$parent = $this->config->get('HTML', 'Parent');
|
||||
$def = $this->manager->getElement($parent, $this->config);
|
||||
if ($def) {
|
||||
$this->info_parent = $parent;
|
||||
$this->info_parent_def = $def;
|
||||
} else {
|
||||
trigger_error('Cannot use unrecognized element as parent.',
|
||||
E_USER_ERROR);
|
||||
$this->info_parent_def = $this->manager->getElement(
|
||||
$this->info_parent, $this->config);
|
||||
}
|
||||
|
||||
// setup allowed elements, SubtractiveWhitelist module
|
||||
$allowed_elements = $this->config->get('HTML', 'AllowedElements');
|
||||
if (is_array($allowed_elements)) {
|
||||
foreach ($this->info as $name => $d) {
|
||||
if(!isset($allowed_elements[$name])) unset($this->info[$name]);
|
||||
}
|
||||
}
|
||||
$allowed_attributes = $this->config->get('HTML', 'AllowedAttributes');
|
||||
if (is_array($allowed_attributes)) {
|
||||
foreach ($this->info_global_attr as $attr_key => $info) {
|
||||
if (!isset($allowed_attributes["*.$attr_key"])) {
|
||||
unset($this->info_global_attr[$attr_key]);
|
||||
}
|
||||
}
|
||||
foreach ($this->info as $tag => $info) {
|
||||
foreach ($info->attr as $attr => $attr_info) {
|
||||
if (!isset($allowed_attributes["$tag.$attr"]) &&
|
||||
!isset($allowed_attributes["*.$attr"])) {
|
||||
unset($this->info[$tag]->attr[$attr]);
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
|
||||
}
|
||||
|
||||
?>
|
||||
<?php
|
||||
|
||||
// components
|
||||
require_once 'HTMLPurifier/HTMLModuleManager.php';
|
||||
|
||||
// this definition and its modules MUST NOT define configuration directives
|
||||
// outside of the HTML or Attr namespaces
|
||||
|
||||
// will be superceded by more accurate doctype declaration schemes
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'HTML', 'Strict', false, 'bool',
|
||||
'Determines whether or not to use Transitional (loose) or Strict rulesets. '.
|
||||
'This directive has been available since 1.3.0.'
|
||||
);
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'HTML', 'BlockWrapper', 'p', 'string',
|
||||
'String name of element to wrap inline elements that are inside a block '.
|
||||
'context. This only occurs in the children of blockquote in strict mode. '.
|
||||
'Example: by default value, <code><blockquote>Foo</blockquote></code> '.
|
||||
'would become <code><blockquote><p>Foo</p></blockquote></code>. The '.
|
||||
'<code><p></code> tags can be replaced '.
|
||||
'with whatever you desire, as long as it is a block level element. '.
|
||||
'This directive has been available since 1.3.0.'
|
||||
);
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'HTML', 'Parent', 'div', 'string',
|
||||
'String name of element that HTML fragment passed to library will be '.
|
||||
'inserted in. An interesting variation would be using span as the '.
|
||||
'parent element, meaning that only inline tags would be allowed. '.
|
||||
'This directive has been available since 1.3.0.'
|
||||
);
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'HTML', 'AllowedElements', null, 'lookup/null',
|
||||
'If HTML Purifier\'s tag set is unsatisfactory for your needs, you '.
|
||||
'can overload it with your own list of tags to allow. Note that this '.
|
||||
'method is subtractive: it does its job by taking away from HTML Purifier '.
|
||||
'usual feature set, so you cannot add a tag that HTML Purifier never '.
|
||||
'supported in the first place (like embed, form or head). If you change this, you '.
|
||||
'probably also want to change %HTML.AllowedAttributes. '.
|
||||
'<strong>Warning:</strong> If another directive conflicts with the '.
|
||||
'elements here, <em>that</em> directive will win and override. '.
|
||||
'This directive has been available since 1.3.0.'
|
||||
);
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'HTML', 'AllowedAttributes', null, 'lookup/null',
|
||||
'IF HTML Purifier\'s attribute set is unsatisfactory, overload it! '.
|
||||
'The syntax is \'tag.attr\' or \'*.attr\' for the global attributes '.
|
||||
'(style, id, class, dir, lang, xml:lang).'.
|
||||
'<strong>Warning:</strong> If another directive conflicts with the '.
|
||||
'elements here, <em>that</em> directive will win and override. For '.
|
||||
'example, %HTML.EnableAttrID will take precedence over *.id in this '.
|
||||
'directive. You must set that directive to true before you can use '.
|
||||
'IDs at all. This directive has been available since 1.3.0.'
|
||||
);
|
||||
|
||||
/**
|
||||
* Definition of the purified HTML that describes allowed children,
|
||||
* attributes, and many other things.
|
||||
*
|
||||
* Conventions:
|
||||
*
|
||||
* All member variables that are prefixed with info
|
||||
* (including the main $info array) are used by HTML Purifier internals
|
||||
* and should not be directly edited when customizing the HTMLDefinition.
|
||||
* They can usually be set via configuration directives or custom
|
||||
* modules.
|
||||
*
|
||||
* On the other hand, member variables without the info prefix are used
|
||||
* internally by the HTMLDefinition and MUST NOT be used by other HTML
|
||||
* Purifier internals. Many of them, however, are public, and may be
|
||||
* edited by userspace code to tweak the behavior of HTMLDefinition.
|
||||
*
|
||||
* HTMLPurifier_Printer_HTMLDefinition is a notable exception to this
|
||||
* rule: in the interest of comprehensiveness, it will sniff everything.
|
||||
*/
|
||||
class HTMLPurifier_HTMLDefinition
|
||||
{
|
||||
|
||||
/** FULLY-PUBLIC VARIABLES */
|
||||
|
||||
/**
|
||||
* Associative array of element names to HTMLPurifier_ElementDef
|
||||
* @public
|
||||
*/
|
||||
var $info = array();
|
||||
|
||||
/**
|
||||
* Associative array of global attribute name to attribute definition.
|
||||
* @public
|
||||
*/
|
||||
var $info_global_attr = array();
|
||||
|
||||
/**
|
||||
* String name of parent element HTML will be going into.
|
||||
* @public
|
||||
*/
|
||||
var $info_parent = 'div';
|
||||
|
||||
/**
|
||||
* Definition for parent element, allows parent element to be a
|
||||
* tag that's not allowed inside the HTML fragment.
|
||||
* @public
|
||||
*/
|
||||
var $info_parent_def;
|
||||
|
||||
/**
|
||||
* String name of element used to wrap inline elements in block context
|
||||
* @note This is rarely used except for BLOCKQUOTEs in strict mode
|
||||
* @public
|
||||
*/
|
||||
var $info_block_wrapper = 'p';
|
||||
|
||||
/**
|
||||
* Associative array of deprecated tag name to HTMLPurifier_TagTransform
|
||||
* @public
|
||||
*/
|
||||
var $info_tag_transform = array();
|
||||
|
||||
/**
|
||||
* Indexed list of HTMLPurifier_AttrTransform to be performed before validation.
|
||||
* @public
|
||||
*/
|
||||
var $info_attr_transform_pre = array();
|
||||
|
||||
/**
|
||||
* Indexed list of HTMLPurifier_AttrTransform to be performed after validation.
|
||||
* @public
|
||||
*/
|
||||
var $info_attr_transform_post = array();
|
||||
|
||||
/**
|
||||
* Nested lookup array of content set name (Block, Inline) to
|
||||
* element name to whether or not it belongs in that content set.
|
||||
* @public
|
||||
*/
|
||||
var $info_content_sets = array();
|
||||
|
||||
|
||||
|
||||
/** PUBLIC BUT INTERNAL VARIABLES */
|
||||
|
||||
var $setup = false; /**< Has setup() been called yet? */
|
||||
var $config; /**< Temporary instance of HTMLPurifier_Config */
|
||||
|
||||
var $manager; /**< Instance of HTMLPurifier_HTMLModuleManager */
|
||||
|
||||
/**
|
||||
* Performs low-cost, preliminary initialization.
|
||||
* @param $config Instance of HTMLPurifier_Config
|
||||
*/
|
||||
function HTMLPurifier_HTMLDefinition(&$config) {
|
||||
$this->config =& $config;
|
||||
$this->manager = new HTMLPurifier_HTMLModuleManager();
|
||||
}
|
||||
|
||||
/**
|
||||
* Processes internals into form usable by HTMLPurifier internals.
|
||||
* Modifying the definition after calling this function should not
|
||||
* be done.
|
||||
*/
|
||||
function setup() {
|
||||
|
||||
// multiple call guard
|
||||
if ($this->setup) {return;} else {$this->setup = true;}
|
||||
|
||||
$this->processModules();
|
||||
$this->setupConfigStuff();
|
||||
|
||||
unset($this->config);
|
||||
unset($this->manager);
|
||||
|
||||
}
|
||||
|
||||
/**
|
||||
* Extract out the information from the manager
|
||||
*/
|
||||
function processModules() {
|
||||
|
||||
$this->manager->setup($this->config);
|
||||
|
||||
foreach ($this->manager->activeModules as $module) {
|
||||
foreach($module->info_tag_transform as $k => $v) $this->info_tag_transform[$k] = $v;
|
||||
foreach($module->info_attr_transform_pre as $k => $v) $this->info_attr_transform_pre[$k] = $v;
|
||||
foreach($module->info_attr_transform_post as $k => $v) $this->info_attr_transform_post[$k]= $v;
|
||||
}
|
||||
|
||||
$this->info = $this->manager->getElements($this->config);
|
||||
$this->info_content_sets = $this->manager->contentSets->lookup;
|
||||
|
||||
}
|
||||
|
||||
/**
|
||||
* Sets up stuff based on config. We need a better way of doing this.
|
||||
*/
|
||||
function setupConfigStuff() {
|
||||
|
||||
$block_wrapper = $this->config->get('HTML', 'BlockWrapper');
|
||||
if (isset($this->info_content_sets['Block'][$block_wrapper])) {
|
||||
$this->info_block_wrapper = $block_wrapper;
|
||||
} else {
|
||||
trigger_error('Cannot use non-block element as block wrapper.',
|
||||
E_USER_ERROR);
|
||||
}
|
||||
|
||||
$parent = $this->config->get('HTML', 'Parent');
|
||||
$def = $this->manager->getElement($parent, $this->config);
|
||||
if ($def) {
|
||||
$this->info_parent = $parent;
|
||||
$this->info_parent_def = $def;
|
||||
} else {
|
||||
trigger_error('Cannot use unrecognized element as parent.',
|
||||
E_USER_ERROR);
|
||||
$this->info_parent_def = $this->manager->getElement(
|
||||
$this->info_parent, $this->config);
|
||||
}
|
||||
|
||||
// support template text
|
||||
$support = "(for information on implementing this, see the ".
|
||||
"support forums) ";
|
||||
|
||||
// setup allowed elements, SubtractiveWhitelist module
|
||||
$allowed_elements = $this->config->get('HTML', 'AllowedElements');
|
||||
if (is_array($allowed_elements)) {
|
||||
foreach ($this->info as $name => $d) {
|
||||
if(!isset($allowed_elements[$name])) unset($this->info[$name]);
|
||||
unset($allowed_elements[$name]);
|
||||
}
|
||||
// emit errors
|
||||
foreach ($allowed_elements as $element => $d) {
|
||||
trigger_error("Element '$element' is not supported $support", E_USER_WARNING);
|
||||
}
|
||||
}
|
||||
|
||||
$allowed_attributes = $this->config->get('HTML', 'AllowedAttributes');
|
||||
$allowed_attributes_mutable = $allowed_attributes; // by copy!
|
||||
if (is_array($allowed_attributes)) {
|
||||
foreach ($this->info_global_attr as $attr_key => $info) {
|
||||
if (!isset($allowed_attributes["*.$attr_key"])) {
|
||||
unset($this->info_global_attr[$attr_key]);
|
||||
} elseif (isset($allowed_attributes_mutable["*.$attr_key"])) {
|
||||
unset($allowed_attributes_mutable["*.$attr_key"]);
|
||||
}
|
||||
}
|
||||
foreach ($this->info as $tag => $info) {
|
||||
foreach ($info->attr as $attr => $attr_info) {
|
||||
if (!isset($allowed_attributes["$tag.$attr"]) &&
|
||||
!isset($allowed_attributes["*.$attr"])) {
|
||||
unset($this->info[$tag]->attr[$attr]);
|
||||
} else {
|
||||
if (isset($allowed_attributes_mutable["$tag.$attr"])) {
|
||||
unset($allowed_attributes_mutable["$tag.$attr"]);
|
||||
} elseif (isset($allowed_attributes_mutable["*.$attr"])) {
|
||||
unset($allowed_attributes_mutable["*.$attr"]);
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
// emit errors
|
||||
foreach ($allowed_attributes_mutable as $elattr => $d) {
|
||||
list($element, $attribute) = explode('.', $elattr);
|
||||
if ($element == '*') {
|
||||
trigger_error("Global attribute '$attribute' is not ".
|
||||
"supported in any elements $support",
|
||||
E_USER_WARNING);
|
||||
} else {
|
||||
trigger_error("Attribute '$attribute' in element '$element' not supported $support",
|
||||
E_USER_WARNING);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
|
||||
}
|
||||
|
||||
?>
|
@@ -1,6 +1,7 @@
|
||||
<?php
|
||||
|
||||
require_once 'HTMLPurifier/HTMLModule.php';
|
||||
require_once 'HTMLPurifier/AttrDef/HTML/LinkTypes.php';
|
||||
|
||||
/**
|
||||
* XHTML 1.1 Hypertext Module, defines hypertext links. Core Module.
|
||||
@@ -21,8 +22,8 @@ class HTMLPurifier_HTMLModule_Hypertext extends HTMLPurifier_HTMLModule
|
||||
// 'charset' => 'Charset',
|
||||
'href' => 'URI',
|
||||
//'hreflang' => 'LanguageCode',
|
||||
//'rel' => 'LinkTypes',
|
||||
//'rev' => 'LinkTypes',
|
||||
'rel' => new HTMLPurifier_AttrDef_HTML_LinkTypes('rel'),
|
||||
'rev' => new HTMLPurifier_AttrDef_HTML_LinkTypes('rev'),
|
||||
//'tabindex' => 'Number',
|
||||
//'type' => 'ContentType',
|
||||
);
|
||||
|
@@ -8,6 +8,10 @@ require_once 'HTMLPurifier/TagTransform/Font.php';
|
||||
|
||||
require_once 'HTMLPurifier/AttrTransform/Lang.php';
|
||||
require_once 'HTMLPurifier/AttrTransform/TextAlign.php';
|
||||
require_once 'HTMLPurifier/AttrTransform/BgColor.php';
|
||||
require_once 'HTMLPurifier/AttrTransform/Border.php';
|
||||
require_once 'HTMLPurifier/AttrTransform/Name.php';
|
||||
require_once 'HTMLPurifier/AttrTransform/Length.php';
|
||||
|
||||
/**
|
||||
* Proprietary module that transforms deprecated elements into Strict
|
||||
@@ -20,7 +24,8 @@ class HTMLPurifier_HTMLModule_TransformToStrict extends HTMLPurifier_HTMLModule
|
||||
var $name = 'TransformToStrict';
|
||||
|
||||
// we're actually modifying these elements, not defining them
|
||||
var $elements = array('h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'p', 'blockquote');
|
||||
var $elements = array('h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'p',
|
||||
'blockquote', 'table', 'td', 'th', 'tr', 'img', 'a', 'hr');
|
||||
|
||||
var $info_tag_transform = array(
|
||||
// placeholders, see constructor for definitions
|
||||
@@ -73,6 +78,23 @@ class HTMLPurifier_HTMLModule_TransformToStrict extends HTMLPurifier_HTMLModule
|
||||
$this->info['blockquote']->content_model_type = 'strictblockquote';
|
||||
$this->info['blockquote']->child = false; // recalculate please!
|
||||
|
||||
$this->info['table']->attr_transform_pre['bgcolor'] =
|
||||
$this->info['tr']->attr_transform_pre['bgcolor'] =
|
||||
$this->info['td']->attr_transform_pre['bgcolor'] =
|
||||
$this->info['th']->attr_transform_pre['bgcolor'] = new HTMLPurifier_AttrTransform_BgColor();
|
||||
|
||||
$this->info['img']->attr_transform_pre['border'] = new HTMLPurifier_AttrTransform_Border();
|
||||
|
||||
$this->info['img']->attr_transform_pre['name'] =
|
||||
$this->info['a']->attr_transform_pre['name'] = new HTMLPurifier_AttrTransform_Name();
|
||||
|
||||
$this->info['td']->attr_transform_pre['width'] =
|
||||
$this->info['th']->attr_transform_pre['width'] =
|
||||
$this->info['hr']->attr_transform_pre['width'] = new HTMLPurifier_AttrTransform_Length('width');
|
||||
|
||||
$this->info['td']->attr_transform_pre['height'] =
|
||||
$this->info['th']->attr_transform_pre['height'] = new HTMLPurifier_AttrTransform_Length('height');
|
||||
|
||||
}
|
||||
|
||||
var $defines_child_def = true;
|
||||
|
Reference in New Issue
Block a user