Make List module HTML5 spec compliant

library/HTMLPurifier/ChildDef/HTML5/Div.php

@@ -0,0 +1,140 @@
+<?php
+
+/**
+ * HTML5 compliant div content model
+ *
+ * In HTML5 divs can be nested directly under dl elements. When they are,
+ * they are expected to have dt and dd children.
+ */
+class HTMLPurifier_ChildDef_HTML5_Div extends HTMLPurifier_ChildDef_Optional
+{
+    public $type = 'div';
+
+    public $allow_empty = true;
+
+    public $elements = array(
+        '#PCDATA' => true,
+        'Flow'    => true,
+        'dt'      => true,
+        'dd'      => true,
+    );
+
+    protected $init = false;
+
+    public function __construct()
+    {
+        parent::__construct($this->elements);
+    }
+
+    /**
+     * @param HTMLPurifier_Config $config
+     * @return array
+     */
+    public function getAllowedElements($config)
+    {
+        $this->init($config);
+        return $this->elements;
+    }
+
+    protected function init(HTMLPurifier_Config $config)
+    {
+        if ($this->init) {
+            return;
+        }
+        $def = $config->getHTMLDefinition();
+        $elements = array();
+        foreach ($this->elements as $name => $_) {
+            if (isset($def->info_content_sets[$name])) {
+                $elements = array_merge($elements, $def->info_content_sets[$name]);
+            } else {
+                $elements[$name] = true;
+            }
+        }
+        $this->elements = $elements;
+        $this->init = true;
+    }
+
+    /**
+     * @param HTMLPurifier_Node[] $children
+     * @param HTMLPurifier_Config $config
+     * @param HTMLPurifier_Context $context
+     * @return array|bool
+     */
+    public function validateChildren($children, $config, $context)
+    {
+        $this->init($config);
+
+        $currentNode = $context->get('CurrentNode', true);
+        if (empty($currentNode->attr['DlDiv'])) {
+            return $this->validateDivChildren($children, $config, $context);
+        }
+
+        return $this->validateDlDivChildren($children);
+    }
+
+    /**
+     * @param HTMLPurifier_Node[] $children
+     * @param HTMLPurifier_Config $config
+     * @param HTMLPurifier_Context $context
+     * @return array|bool
+     */
+    protected function validateDivChildren(array $children, HTMLPurifier_Config $config, HTMLPurifier_Context $context)
+    {
+        // Filter out dt and dd, but try to retain their contents (if any),
+        // just as MakeWellFormed strategy would do to unrecognized elements
+        $result = array();
+        foreach ($children as $child) {
+            if (isset($child->name) && ($child->name === 'dt' || $child->name === 'dd')) {
+                foreach ($child->children as $c) {
+                    $result[] = $c;
+                }
+            } else {
+                $result[] = $child;
+            }
+        }
+        return parent::validateChildren($result, $config, $context);
+    }
+
+    /**
+     * @param HTMLPurifier_Node[] $children
+     * @return array|bool
+     */
+    protected function validateDlDivChildren(array $children)
+    {
+        // div in dl is required to have (dt+, dd+) content
+        // https://html.spec.whatwg.org/multipage/grouping-content.html#the-dl-element
+        // Related discussion: https://github.com/whatwg/html/issues/1937
+
+        $dt = null;
+        $dd = null;
+        $result = array();
+
+        foreach ($children as $child) {
+            if (!empty($child->is_whitespace)) {
+                $result[] = $child;
+                continue;
+            }
+            if ($child->name === 'dt' && !$dd) {
+                $dt = $child;
+                $result[] = $child;
+            } elseif ($child->name === 'dd') {
+                $dd = $child;
+                $result[] = $dd;
+            }
+        }
+
+        if (!$dd && !$dt) {
+            return false;
+        }
+
+        if (!$dt) {
+            array_unshift($result, new HTMLPurifier_Node_Element('dt'));
+        }
+
+        if (!$dd) {
+            $result[] = new HTMLPurifier_Node_Element('dd');
+        }
+
+        return $result;
+    }
+}

library/HTMLPurifier/ChildDef/HTML5/Dl.php

@@ -0,0 +1,119 @@
+<?php
+
+/**
+ * HTML5 compliant dl content model
+ *
+ * @see https://html.spec.whatwg.org/multipage/grouping-content.html#the-dl-element
+ */
+class HTMLPurifier_ChildDef_HTML5_Dl extends HTMLPurifier_ChildDef
+{
+    public $type = 'dl';
+
+    public $elements = array(
+        'dt'  => true,
+        'dd'  => true,
+        'div' => true,
+    );
+
+    /**
+     * @param HTMLPurifier_Node[] $children
+     * @param HTMLPurifier_Config $config
+     * @param HTMLPurifier_Context $context
+     * @return array|bool|void
+     */
+    public function validateChildren($children, $config, $context)
+    {
+        // if dt or dd is not allowed, delete parent node
+        if (!isset($config->getHTMLDefinition()->info['dt'])) {
+            trigger_error("Cannot allow dl without allowing dt", E_USER_WARNING);
+            return false;
+        }
+        if (!isset($config->getHTMLDefinition()->info['dd'])) {
+            trigger_error("Cannot allow dl without allowing dd", E_USER_WARNING);
+            return false;
+        }
+
+        // Content model:
+        // The dl element represents a description list of zero or more term-description
+        // groups. Each term-description group consists of one or more terms (represented
+        // by dt elements) possibly as children of a div element child, and one or more
+        // descriptions (represented by dd elements possibly as children of a div element
+        // child), ignoring any nodes other than dt and dd element children, and dt and dd
+        // elements that are children of div element children within a single dl element.
+        // https://html.spec.whatwg.org/multipage/grouping-content.html#the-dl-element
+
+        // Related discussion: https://github.com/whatwg/html/issues/1937
+
+        // Detect if the first child element is a div, if yes, then all
+        // children are expected to be divs
+        foreach ($children as $node) {
+            if (!empty($node->is_whitespace)) {
+                continue;
+            }
+            if ($node->name === 'div') {
+                $result = $this->validateDivChildren($children);
+                return $result;
+            }
+            break;
+        }
+
+        return $this->validateDtDdChildren($children);
+    }
+
+    /**
+     * @param HTMLPurifier_Node[] $children
+     * @return HTMLPurifier_Node[]
+     */
+    protected function validateDivChildren(array $children)
+    {
+        $result = array();
+        foreach ($children as $node) {
+            if (!empty($node->is_whitespace)) {
+                $result[] = $node;
+                continue;
+            }
+            if ($node->name === 'div') {
+                $result[] = $node;
+            }
+        }
+        return $result;
+    }
+
+    /**
+     * @param HTMLPurifier_Node[] $children
+     * @return HTMLPurifier_Node[]
+     */
+    protected function validateDtDdChildren(array $children)
+    {
+        $result = array();
+
+        $dt = null;
+        $dd = null;
+
+        foreach ($children as $node) {
+            if (!empty($node->is_whitespace)) {
+                $result[] = $node;
+                continue;
+            }
+            if ($node->name === 'dt') {
+                $dt = $node;
+                $dd = null;
+                $result[] = $node;
+            } elseif ($node->name === 'dd') {
+                if ($dt === null) {
+                    $dt = new HTMLPurifier_Node_Element('dt');
+                    $result[] = $dt;
+                }
+                $dd = $node;
+                $result[] = $node;
+            }
+        }
+
+        // There must be at least one dd after dt
+        if ($dt && !$dd) {
+            $result[] = new HTMLPurifier_Node_Element('dd');
+        }
+
+        return $result;
+    }
+}

library/HTMLPurifier/ChildDef/HTML5/List.php

@@ -0,0 +1,57 @@
+<?php
+
+/**
+ * HTML5 compliant content definition for ul and ol elements.
+ *
+ * The only difference between this and {@link HTMLPurifier_ChildDef_List}
+ * is that HTML5 spec allows empty lists.
+ */
+class HTMLPurifier_ChildDef_HTML5_List extends HTMLPurifier_ChildDef
+{
+    public $type = 'list';
+
+    public $elements = array(
+        'li' => true,
+        'ul' => true,
+        'ol' => true,
+    );
+
+    /**
+     * @param HTMLPurifier_Node[] $children
+     * @param HTMLPurifier_Config $config
+     * @param HTMLPurifier_Context $context
+     * @return array|bool
+     */
+    public function validateChildren($children, $config, $context)
+    {
+        // if li is not allowed, delete parent node
+        if (!isset($config->getHTMLDefinition()->info['li'])) {
+            trigger_error("Cannot allow ul/ol without allowing li", E_USER_WARNING);
+            return false;
+        }
+
+        $result = array();
+        $li = null;
+
+        foreach ($children as $node) {
+            if (!empty($node->is_whitespace)) {
+                $result[] = $node;
+                continue;
+            }
+            if ($node->name === 'li') {
+                $li = $node;
+                $result[] = $node;
+            } else {
+                // tuck this element into the previous li
+                if ($li === null) {
+                    $li = new HTMLPurifier_Node_Element('li');
+                    $result[] = $li;
+                }
+                $li->children[] = $node;
+                $li->empty = false;
+            }
+        }
+
+        return $result;
+    }
+}

library/HTMLPurifier/HTML5Config.php

@@ -2,7 +2,7 @@
 
 class HTMLPurifier_HTML5Config extends HTMLPurifier_Config
 {
-    const REVISION = 2020053001;
+    const REVISION = 2020060101;
 
     /**
      * @param  string|array|HTMLPurifier_Config $config

library/HTMLPurifier/HTML5Definition.php

@@ -48,6 +48,8 @@ public static function setupHTMLDefinition(HTMLPurifier_HTMLDefinition $def, HTM
 
         $def->manager->attrTypes->set('Datetime', new HTMLPurifier_AttrDef_HTML5_Datetime());
 
+        $def->info_injector[] = new HTMLPurifier_Injector_HTML5_DlDiv();
+
         return $def;
     }
 }