Merge pull request #240 from Masterminds/self-closing-table-elements

add support for optional end tags for table elements
Masterminds · Oct 14, 2023 · b80cdb3 · b80cdb3
2 parents f47dcf3 + de52ce1
commit b80cdb3
Show file tree

Hide file tree

Showing 3 changed files with 82 additions and 0 deletions.
diff --git a/src/HTML5/Elements.php b/src/HTML5/Elements.php
@@ -71,6 +71,24 @@ class Elements
      */
     const BLOCK_ONLY_INLINE = 128;
 
+    /**
+     * Elements with optional end tags that cause auto-closing of previous and parent tags,
+     * as example most of the table related tags, see https://www.w3.org/TR/html401/struct/tables.html
+     * Structure is as follows:
+     * TAG-NAME => [PARENT-TAG-NAME-TO-CLOSE1, PARENT-TAG-NAME-TO-CLOSE2, ...].
+     *
+     * Order is important, after auto-closing one parent with might have to close also their parent.
+     *
+     * @var array<string, string[]>
+     */
+    public static $optionalEndElementsParentsToClose = array(
+        'tr' => array('td', 'tr'),
+        'td' => array('td', 'th'),
+        'th' => array('td', 'th'),
+        'tfoot' => array('td', 'th', 'tr', 'tbody', 'thead'),
+        'tbody' => array('td', 'th', 'tr', 'thead'),
+    );
+
     /**
      * The HTML5 elements as defined in http://dev.w3.org/html5/markup/elements.html.
      *

diff --git a/src/HTML5/Parser/DOMTreeBuilder.php b/src/HTML5/Parser/DOMTreeBuilder.php
@@ -359,6 +359,16 @@ public function startTag($name, $attributes = array(), $selfClosing = false)
             $this->onlyInline = null;
         }
 
+        // some elements as table related tags might have optional end tags that force us to auto close multiple tags
+        // https://www.w3.org/TR/html401/struct/tables.html
+        if ($this->current instanceof \DOMElement && isset(Elements::$optionalEndElementsParentsToClose[$lname])) {
+            foreach (Elements::$optionalEndElementsParentsToClose[$lname] as $parentElName) {
+                if ($this->current instanceof \DOMElement && $this->current->tagName === $parentElName) {
+                    $this->autoclose($parentElName);
+                }
+            }
+        }
+
         try {
             $prefix = ($pos = strpos($lname, ':')) ? substr($lname, 0, $pos) : '';
 

diff --git a/test/HTML5/Html5Test.php b/test/HTML5/Html5Test.php
@@ -56,6 +56,60 @@ public function testImageTagsInSvg()
         $this->assertEmpty($this->html5->getErrors());
     }
 
+    public function testSelfClosingTableHierarchyElements()
+    {
+        $html = '
+                <table>
+                    <thead>
+                        <tr>
+                            <th>0
+                    <tbody>
+                        <tr>
+                            <td>A
+                        <tr>
+                            <td>B1
+                            <td>B2
+                        <tr>
+                            <td>C
+                    <tfoot>
+                        <tr>
+                            <th>1
+                            <td>2
+                    </table>';
+
+        $expected = '<table>
+                        <thead>
+                            <tr>
+                                <th>0</th>
+                            </tr>
+                        </thead>
+                        <tbody>
+                            <tr>
+                                <td>A</td>
+                            </tr>
+                            <tr>
+                                <td>B1</td>
+                                <td>B2</td>
+                            </tr>
+                            <tr>
+                                <td>C</td>
+                            </tr>
+                        </tbody>
+                        <tfoot>
+                            <tr>
+                                <th>1</th>
+                                <td>2</td>
+                            </tr>
+                        </tfoot>
+                    </table>';
+
+        $doc = $this->html5->loadHTMLFragment($html);
+        $this->assertSame(
+            preg_replace('/\s+/', '', $expected),
+            preg_replace('/\s+/', '', $this->html5->saveHTML($doc))
+        );
+    }
+
     public function testLoadOptions()
     {
         // doc