Fix hasSingleTagInsideElement method

It would fail for e.g. `<div> <p>foo</p> </div>`.

mozilla/readability uses children for the tag lookup, which return only elements.
PHP does not have children property so b580cf2
mistakenly used `childNodes` instead, but that can return any node type.

Let’s filter the children ourselves.

Also add comments from mozilla/readability’s `_hasSingleTagInsideElement`.

Author: jtojnar

src/Readability.php

@@ -1536,16 +1536,23 @@ private function isPhrasingContent($node): bool
             );
     }
 
+    /**
+     * Checks if `$node` has only whitespace and a single element with `$tag` for the tag name.
+     * Returns false if `$node` contains non-empty text nodes
+     * or if it contains no element with given tag or more than 1 element.
+     */
     private function hasSingleTagInsideElement(JSLikeHTMLElement $node, string $tag): bool
     {
-        if (1 !== $node->childNodes->length || $node->childNodes->item(0)->nodeName !== $tag) {
-            return false;
+        $childNodes = iterator_to_array($node->childNodes);
+        $children = array_filter($childNodes, fn ($childNode) => $childNode instanceof \DOMElement);
+
+        // There should be exactly 1 element child with given tag
+        if (1 !== \count($children) || $children[0]->nodeName !== $tag) {
+            return null;
         }
 
-        $a = array_filter(
-            iterator_to_array($node->childNodes),
-            fn ($childNode) => $childNode instanceof \DOMText && preg_match($this->regexps['hasContent'], $this->getInnerText($childNode))
-        );
+        // And there should be no text nodes with real content
+        $a = array_filter($childNodes, fn ($childNode) => $childNode instanceof \DOMText && preg_match($this->regexps['hasContent'], $this->getInnerText($childNode)));
 
         return 0 === \count($a);
     }