protected function FieldLink::canonicalizeUri in Drupal 10

Same name and namespace in other branches

1. 8 core/modules/link/src/Plugin/migrate/process/FieldLink.php \Drupal\link\Plugin\migrate\process\FieldLink::canonicalizeUri()
2. 9 core/modules/link/src/Plugin/migrate/process/FieldLink.php \Drupal\link\Plugin\migrate\process\FieldLink::canonicalizeUri()

Turn a Drupal 6/7 URI into a Drupal 8-compatible format.

Parameters

string $uri: The 'url' value from Drupal 6/7.

Return value

string The Drupal 8-compatible URI.

See also

\Drupal\link\Plugin\Field\FieldWidget\LinkWidget::getUserEnteredStringAsUri()

1 call to FieldLink::canonicalizeUri()

FieldLink::transform in core/modules/link/src/Plugin/migrate/process/FieldLink.php

Performs the associated process.

File

core/modules/link/src/Plugin/migrate/process/FieldLink.php, line 61

Class

FieldLink

Transform a pre-Drupal 8 formatted link for use in Drupal 8.

Namespace

Drupal\link\Plugin\migrate\process

Code

protected function canonicalizeUri($uri) {

  // If the path starts with 2 slashes then it is always considered an
  // external URL without an explicit protocol part.
  // @todo Remove this when https://www.drupal.org/node/2744729 lands.
  if (strpos($uri, '//') === 0) {
    return $this->configuration['uri_scheme'] . ltrim($uri, '/');
  }

  // If we already have a scheme, we're fine.
  if (parse_url($uri, PHP_URL_SCHEME)) {
    return $uri;
  }

  // Empty URI and non-links are allowed.
  if (empty($uri) || in_array($uri, [
    '<nolink>',
    '<none>',
  ])) {
    return 'route:<nolink>';
  }

  // Remove the <front> component of the URL.
  if (strpos($uri, '<front>') === 0) {
    $uri = substr($uri, strlen('<front>'));
  }
  else {

    // List of unicode-encoded characters that were allowed in URLs,
    // according to link module in Drupal 7. Every character between &#x00BF;
    // and &#x00FF; (except × &#x00D7; and ÷ &#x00F7;) with the addition of
    // &#x0152;, &#x0153; and &#x0178;.
    // @see https://git.drupalcode.org/project/link/blob/7.x-1.5-beta2/link.module#L1382
    // cSpell:disable-next-line
    $link_i_chars = '¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿŒœŸ';

    // Pattern specific to internal links.
    $internal_pattern = "/^(?:[a-z0-9" . $link_i_chars . "_\\-+\\[\\] ]+)";
    $directories = "(?:\\/[a-z0-9" . $link_i_chars . "_\\-\\.~+%=&,\$'#!():;*@\\[\\]]*)*";

    // Yes, four backslashes == a single backslash.
    $query = "(?:\\/?\\?([?a-z0-9" . $link_i_chars . "+_|\\-\\.~\\/\\\\%=&,\$'():;*@\\[\\]{} ]*))";
    $anchor = "(?:#[a-z0-9" . $link_i_chars . "_\\-\\.~+%=&,\$'():;*@\\[\\]\\/\\?]*)";

    // The rest of the path for a standard URL.
    $end = $directories . '?' . $query . '?' . $anchor . '?$/i';
    if (!preg_match($internal_pattern . $end, $uri)) {
      $link_domains = '[a-z][a-z0-9-]{1,62}';

      // Starting a parenthesis group with (?: means that it is grouped, but is not captured
      $authentication = "(?:(?:(?:[\\w\\.\\-\\+!\$&'\\(\\)*\\+,;=" . $link_i_chars . "]|%[0-9a-f]{2})+(?::(?:[\\w" . $link_i_chars . "\\.\\-\\+%!\$&'\\(\\)*\\+,;=]|%[0-9a-f]{2})*)?)?@)";
      $domain = '(?:(?:[a-z0-9' . $link_i_chars . ']([a-z0-9' . $link_i_chars . '\\-_\\[\\]])*)(\\.(([a-z0-9' . $link_i_chars . '\\-_\\[\\]])+\\.)*(' . $link_domains . '|[a-z]{2}))?)';
      $ipv4 = '(?:[0-9]{1,3}(\\.[0-9]{1,3}){3})';
      $ipv6 = '(?:[0-9a-fA-F]{1,4}(\\:[0-9a-fA-F]{1,4}){7})';
      $port = '(?::([0-9]{1,5}))';

      // Pattern specific to external links.
      $external_pattern = '/^' . $authentication . '?(' . $domain . '|' . $ipv4 . '|' . $ipv6 . ' |localhost)' . $port . '?';
      if (preg_match($external_pattern . $end, $uri)) {
        return $this->configuration['uri_scheme'] . $uri;
      }
    }
  }

  // Add the internal: scheme and ensure a leading slash.
  return 'internal:/' . ltrim($uri, '/');
}