From 42d7a87578e23be199b16b60a223c261df2f1668 Mon Sep 17 00:00:00 2001
From: enured <enis.nuredini@uzh.ch>
Date: Mon, 31 Oct 2022 16:21:36 +0100
Subject: [PATCH] F14320: Implemented use for library html purifier and
finished implementation for new feature htmlAllow. Added documentation for
htmlAllow.
---
Documentation/Form.rst | 17 ++++
extension/Classes/Core/AbstractBuildForm.php | 8 +-
extension/Classes/Core/Save.php | 99 ++++++++++----------
3 files changed, 72 insertions(+), 52 deletions(-)
diff --git a/Documentation/Form.rst b/Documentation/Form.rst
index 0f6609708..a9a331fb2 100644
--- a/Documentation/Form.rst
+++ b/Documentation/Form.rst
@@ -1378,6 +1378,12 @@ General input for any text.
* *step* = Step size of the up/down buttons which increase/decrease the number of in the input field. Optional.
Default 1. Only useful with `inputType=number` (defined explicit via `inputType` or detected automatically).
* *textareaResize* = 0|1 (optional). Be default = 1 (=on). A textarea element is resizable by the user.
+ * *htmlAllow* = p,br,img,table,u,ol,b,h2,h3,h5,sup (optional). By default every html tag is allowed. Allow only specific html tags. If following tags are given in list then associated tags will be added automatically:
+
+ * table: td, tr, th, tbody, thead
+ * ol,ul: li
+ * b: strong
+ * u,ins,del,s: span
.. _`input-typeahead`:
@@ -1615,6 +1621,17 @@ Type: editor
* *FormElement.size* = <min_height>,<max_height>: in pixels, including top and bottom bars. E.g.: 300,600
+ Define allowed html tags. TinyMce Settings will be overwritten if this parameter is set.
+* Following tags are not used from tinyMce: u,del,ins,s. In this case use textDecoration to get comparable function and correct configuration. Example: ::
+
+ htmlAllow = p,br,h1,h3,table,b,textDecoration,ul,img
+
+* By default every html tag is allowed. If following tags are given in list then associated tags will be added automatically:
+
+ * table: td, tr, th, tbody, thead
+ * ol,ul: li
+ * b: strong
+ * textDecoration: span
Type: annotate
^^^^^^^^^^^^^^
diff --git a/extension/Classes/Core/AbstractBuildForm.php b/extension/Classes/Core/AbstractBuildForm.php
index da767926e..d5afe0755 100644
--- a/extension/Classes/Core/AbstractBuildForm.php
+++ b/extension/Classes/Core/AbstractBuildForm.php
@@ -3393,8 +3393,8 @@ abstract class AbstractBuildForm {
// Set defaults for tinyMce
$imgToken = 'img[longdesc|usemap|src|border|alt=|title|hspace|vspace|width|height|align]';
$textDecoration = 'span[style]';
- $table = 'table[style|align],td[style],th[style],tr[style],tbody[style],thead[style]';
- $url = 'a[href|target]';
+ $table = 'table[style|align|border],td[style],th[style],tr[style],tbody[style],thead[style]';
+ $url = 'a[href|target|title]';
$paragraphToken = 'p[align]';
$strong = 'strong';
@@ -3437,6 +3437,8 @@ abstract class AbstractBuildForm {
if(!$listFlag) {
$htmlAllowArray[$htmlToken] = 'ul,li';
$listFlag = true;
+ } else {
+ $htmlAllowArray[$htmlToken] = 'ul';
}
self::setTinymceEditorToolbarAttributes($customEditorToolbarFlags, $customEditorToolbar, 'bullist');
break;
@@ -3444,6 +3446,8 @@ abstract class AbstractBuildForm {
if(!$listFlag) {
$htmlAllowArray[$htmlToken] = 'ol,li';
$listFlag = true;
+ } else {
+ $htmlAllowArray[$htmlToken] = 'ol';
}
self::setTinymceEditorToolbarAttributes($customEditorToolbarFlags, $customEditorToolbar, 'numlist');
break;
diff --git a/extension/Classes/Core/Save.php b/extension/Classes/Core/Save.php
index 2430f35f5..688946540 100644
--- a/extension/Classes/Core/Save.php
+++ b/extension/Classes/Core/Save.php
@@ -9,6 +9,7 @@
namespace IMATHUZH\Qfq\Core;
use DOMDocument;
+use HTMLPurifier;
use IMATHUZH\Qfq\Core\Database\Database;
use IMATHUZH\Qfq\Core\Form\FormAction;
use IMATHUZH\Qfq\Core\Helper\EncryptDecrypt;
@@ -410,10 +411,10 @@ class Save {
foreach($this->feSpecNative as $fe) {
$feColumnTypes[$fe['name']] = $fe['type'];
}
+
// Get htmlAllow parameters of all formValues and store in $feSpecsTags
$feSpecsTags = $this->getHtmlAllowTags($this->feSpecNative, $formValues);
-
- // For htmlAllow: Define with flags if '<li>' or '<span>' is needed for tinyMce and add them
+ // For htmlAllow: Define with flags if '<li>' or '<span>' is needed for tinyMce (used for underline and other text decoration) and add them
$feSpecsTags = $this->setTinyMceSpecificTags($feSpecsTags);
@@ -470,10 +471,10 @@ class Save {
Support::setIfNotSet($formValues, $column);
}
- // Check for existing htmlAllow and strip tags
+ // Check for existing htmlAllow and strip tags, purify html result to prevent XSS
if(isset($feSpecsTags[$column]) && $feSpecsTags[$column] !== '') {
$formValues[$column] = $this->custom_strip_tags($formValues[$column], $feSpecsTags[$column]);
- $formValues[$column] = $this->strip_attributes($formValues[$column], 'style,name,align');
+ $formValues[$column] = $this->purifierHtml($formValues[$column]);
}
$newValues[$column] = $formValues[$column];
@@ -535,17 +536,32 @@ class Save {
private function setTinyMceSpecificTags($feSpecsTags): array {
$listFlag = false;
$decorationFlag = false;
+ $tableFlag = false;
+ $strongFlag = false;
foreach ($feSpecsTags as $key => $value) {
$feSpecsTagArray[$key] = explode(',',$value);
foreach ($feSpecsTagArray[$key] as $key2 => $tag) {
- if(($tag === 'ul' || $tag === 'ol')) {
- $listFlag = true;
- } elseif ($tag === 'textDecoration' || $tag === 'u' || $tag === 'ins' || $tag === 'del' || $tag === 's') {
- $decorationFlag = true;
- }
-
- if($tag !== 'textDecoration') {
- $feSpecsTagArray[$key][$key2] = $tag;
+ switch ($tag) {
+ case 'ul':
+ case 'ol':
+ $listFlag = true;
+ break;
+ case 'textDecoration':
+ case 'u':
+ case 'ins':
+ case 'del':
+ case 's':
+ $decorationFlag = true;
+ break;
+ case 'table':
+ $tableFlag = true;
+ break;
+ case 'b':
+ $strongFlag = true;
+ break;
+ default:
+ $feSpecsTagArray[$key][$key2] = $tag;
+ break;
}
}
@@ -559,6 +575,18 @@ class Save {
$feSpecsTagArray[$key][] = "span";
$decorationFlag = false;
}
+
+
+ if($strongFlag) {
+ $feSpecsTagArray[$key][] = "strong";
+ $strongFlag = false;
+ }
+
+ if($tableFlag) {
+ array_push($feSpecsTagArray[$key],"th","td","tr","tbody","thead");
+ $tableFlag = false;
+ }
+
$feSpecsTags[$key] = implode(',', $feSpecsTagArray[$key]);
}
@@ -1750,48 +1778,19 @@ class Save {
}
/**
- * Remove not allowed attributes
+ * Remove not allowed attributes and content which is not in whitelist
+ * Used in combination with htmlAllow.
+ * Author: Edward Z. Yang
+ * Website: http://htmlpurifier.org/
*
* @param $html
- * @param string $allowedAttributes
* @return array|string|string[]|null
*/
- function strip_attributes($html, string $allowedAttributes) {
- $regex_attributes = '/(<\s*\w+\s+)(\w+)\s*=\s*".*"\s*\/*(>)/U';
- $regex_attributes2 = '/(<\s*\w+\s+)(\w+)\s*\W*(>)/U';
- $regex_attributes3 = '/(<\s*\w+\s+)(\w+)\s*=\s*.+\W*(>)/U';
- $allowed_attributes = explode(',',$allowedAttributes);
- $allowed_attributes = array_map(strtolower,$allowed_attributes);
-
- //
-
- $regexHtmlBlock = '/<\s*\w+\s*\w*=*.*>.*<\/.*>/U';
- $matchAllBlocks = array();
- preg_match_all($regexHtmlBlock, $html, $matchAllBlocks);
-
- foreach ($matchAllBlocks[0] as &$block) {
- $matchesType1 = array();
- preg_match_all($regex_attributes, $block, $matchesType1);
- $matchesType2 = array();
- preg_match_all($regex_attributes2, $block, $matchesType2);
- $matchesType3 = array();
- preg_match_all($regex_attributes3, $block, $matchesType3);
-
- if(empty($matchesType1[0])) {
- if(empty($matchesType2[0])) {
- $regex_attributes = $regex_attributes3;
- } else {
- $regex_attributes = $regex_attributes2;
- }
- }
- $allMatches = array_replace_recursive($matchesType1, $matchesType2, $matchesType3);
- $block = preg_replace_callback($regex_attributes, function ($allMatches) use (&$allowed_attributes) {
- return in_array(strtolower($allMatches[2]),$allowed_attributes)?$allMatches[0]:$allMatches[1].$allMatches[3];
- },$block);
- }
-
-
- $rhtml = implode('',$matchAllBlocks[0]);
+ function purifierHtml($html) {
+ $libraryPath = Path::absoluteExt('vendor/htmlpurifier-4.15.0-lite/library/HTMLPurifier.auto.php');
+ require_once $libraryPath;
+ $purifier = new HTMLPurifier();
+ $rhtml = $purifier->purify($html);
return $rhtml;
}
--
GitLab