trait SpamspanDomTrait in SpamSpan filter 8

trait SpamspanDomTrait {

  /**
   * Replaces email addresses using DOM and regex.
   *
   * @param string $text
   *   Input text.
   * @param bool $altered
   *   Set to true if any replacements happen.
   *
   * @return string
   *   Output text.
   */
  protected function processAsDom($text, &$altered) {
    $document = $this
      ->loadHtmlDocument($text);

    // Process mailto <a> tags.
    foreach ($document
      ->getElementsByTagName('a') as $atag) {
      $href = trim($atag
        ->getAttribute('href'));
      if (strpos($href, 'mailto:') === 0) {
        $atag
          ->setAttribute('href', $href);
        $text = $this
          ->replaceMailtoLinks($document
          ->saveHTML($atag), $altered);
        $this
          ->replaceDomNode($atag, $text);
      }
    }

    // Parse all text nodes.
    $xpath = new \DomXPath($document);
    foreach ($xpath
      ->query('//text()') as $text_node) {
      $text = $text_node->nodeValue;
      $node_altered = FALSE;
      if (!empty($this->settings['spamspan_use_form'])) {
        $text = $this
          ->replaceEmailAddressesWithOptions($text, $node_altered);
      }
      $text = $this
        ->replaceBareEmailAddresses($text, $node_altered);
      if ($node_altered) {
        $this
          ->replaceDomNode($text_node, $text);
        $altered = TRUE;
      }
    }
    return $this
      ->toStringHtmlDocument($document);
  }

  /**
   * Replace DOM node with another one created from text.
   *
   * @param \DOMNode $old_node
   *   Node to be replaced.
   * @param string $new_text
   *   Html for new node.
   */
  protected function replaceDomNode(\DOMNode $old_node, $new_text) {
    $fragment = $this
      ->loadHtmlDocument($new_text);
    $div = $fragment
      ->getElementsByTagName('div')
      ->item(0);
    foreach ($div->childNodes as $child) {
      $new_node = $old_node->ownerDocument
        ->importNode($child, TRUE);
      $old_node->parentNode
        ->insertBefore($new_node, $old_node);
    }
    $old_node->parentNode
      ->removeChild($old_node);
  }

  /**
   * Load text as DOM Document.
   *
   * @param string $text
   *   Text to load.
   *
   * @return \DOMDocument
   *   DOM Document.
   */
  protected function loadHtmlDocument($text) {
    $document = new \DOMDocument();

    // Replace CRLF with LF, because CR will be encoded as &#13 otherwise.
    $text = str_replace("\r\n", "\n", $text);

    // Ignore warnings with '@' due to unknown HTML5 tags (section, aside, etc).
    @$document
      ->loadHTML('<?xml encoding="UTF-8"><div>' . $text . '</div>', LIBXML_HTML_NOIMPLIED | LIBXML_HTML_NODEFDTD);
    return $document;
  }

  /**
   * Convert DOM Document back to html.
   *
   * @param \DOMDocument $document
   *   DOM Document.
   *
   * @return string
   *   Html.
   */
  protected function toStringHtmlDocument(\DOMDocument $document) {
    foreach ($document->childNodes as $item) {
      if ($item->nodeType == XML_PI_NODE) {
        $document
          ->removeChild($item);
        break;
      }
    }

    // Use saveHTML(documentElement) instead of simply saveHTML() to prevent
    // utf-8 characters (e.g. accented letters) from being turned into entities
    // At the same time, this will strip the DOCTYPE.
    $div = $document
      ->getElementsByTagName("div")
      ->item(0);
    return implode('', array_map([
      $document,
      'saveHTML',
    ], iterator_to_array($div->childNodes)));
  }

}