Html::transformRootRelativeUrlsToAbsolute

public static Html::transformRootRelativeUrlsToAbsolute($html, $scheme_and_host)

Converts all root-relative URLs to absolute URLs.

Does not change any existing protocol-relative or absolute URLs. Does not change other relative URLs because they would result in different absolute URLs depending on the current path. For example: when the same content containing such a relative URL (for example 'image.png'), is served from its canonical URL (for example 'http://example.com/some-article') or from a listing or feed (for example 'http://example.com/all-articles') their "current path" differs, resulting in different absolute URLs: 'http://example.com/some-article/image.png' versus 'http://example.com/all-articles/image.png'. Only one can be correct. Therefore relative URLs that are not root-relative cannot be safely transformed and should generally be avoided.

Necessary for HTML that is served outside of a website, for example, RSS and e-mail.

Parameters

string $html: The partial (X)HTML snippet to load. Invalid markup will be corrected on import.

string $scheme_and_host: The root URL, which has a URI scheme, host and optional port.

Return value

string The updated (X)HTML snippet.

File

core/lib/Drupal/Component/Utility/Html.php, line 451

Class

Html
Provides DOMDocument helpers for parsing and serializing HTML strings.

Namespace

Drupal\Component\Utility

Code

public static function transformRootRelativeUrlsToAbsolute($html, $scheme_and_host) {
  assert('empty(array_diff(array_keys(parse_url($scheme_and_host)), ["scheme", "host", "port"]))', '$scheme_and_host contains scheme, host and port at most.');
  assert('isset(parse_url($scheme_and_host)["scheme"])', '$scheme_and_host is absolute and hence has a scheme.');
  assert('isset(parse_url($scheme_and_host)["host"])', '$base_url is absolute and hence has a host.');

  $html_dom = Html::load($html);
  $xpath = new \DOMXpath($html_dom);

  // Update all root-relative URLs to absolute URLs in the given HTML.
  foreach (static::$uriAttributes as $attr) {
    foreach ($xpath->query("//*[starts-with(@$attr, '/') and not(starts-with(@$attr, '//'))]") as $node) {
      $node->setAttribute($attr, $scheme_and_host . $node->getAttribute($attr));
    }
    foreach ($xpath->query("//*[@srcset]") as $node) {
      // @see https://html.spec.whatwg.org/multipage/embedded-content.html#attr-img-srcset
      // @see https://html.spec.whatwg.org/multipage/embedded-content.html#image-candidate-string
      $image_candidate_strings = explode(',', $node->getAttribute('srcset'));
      $image_candidate_strings = array_map('trim', $image_candidate_strings);
      for ($i = 0; $i < count($image_candidate_strings); $i++) {
        $image_candidate_string = $image_candidate_strings[$i];
        if ($image_candidate_string[0] === '/' && $image_candidate_string[1] !== '/') {
          $image_candidate_strings[$i] = $scheme_and_host . $image_candidate_string;
        }
      }
      $node->setAttribute('srcset', implode(', ', $image_candidate_strings));
    }
  }
  return Html::serialize($html_dom);
}
doc_Drupal
2016-10-29 09:18:45
Comments
Leave a Comment

Please login to continue.