Chris@0: <?php
Chris@0: 
Chris@0: /**
Chris@0:  * @file
Chris@0:  * Provides Unicode-related conversions and operations.
Chris@0:  */
Chris@0: 
Chris@0: use Drupal\Component\Utility\Unicode;
Chris@0: 
Chris@0: /**
Chris@0:  * Returns Unicode library status and errors.
Chris@0:  */
Chris@17: 
Chris@0: /**
Chris@0:  * Moves unicode_requirements() logic to system_requirements().
Chris@0:  *
Chris@0:  * @deprecated in Drupal 8.4.0 and will be removed before Drupal 9.0.0.
Chris@0:  *
Chris@0:  * @see https://www.drupal.org/node/2884698
Chris@0:  */
Chris@0: function unicode_requirements() {
Chris@0:   @trigger_error('unicode_requirements() is deprecated in Drupal 8.4.0 and will be removed before Drupal 9.0.0. There is no replacement; system_requirements() now includes the logic instead. See https://www.drupal.org/node/2884698', E_USER_DEPRECATED);
Chris@0: 
Chris@0:   $libraries = [
Chris@0:     Unicode::STATUS_SINGLEBYTE => t('Standard PHP'),
Chris@0:     Unicode::STATUS_MULTIBYTE => t('PHP Mbstring Extension'),
Chris@0:     Unicode::STATUS_ERROR => t('Error'),
Chris@0:   ];
Chris@0:   $severities = [
Chris@0:     Unicode::STATUS_SINGLEBYTE => REQUIREMENT_WARNING,
Chris@0:     Unicode::STATUS_MULTIBYTE => NULL,
Chris@0:     Unicode::STATUS_ERROR => REQUIREMENT_ERROR,
Chris@0:   ];
Chris@0:   $failed_check = Unicode::check();
Chris@0:   $library = Unicode::getStatus();
Chris@0: 
Chris@0:   $requirements['unicode'] = [
Chris@0:     'title' => t('Unicode library'),
Chris@0:     'value' => $libraries[$library],
Chris@0:     'severity' => $severities[$library],
Chris@0:   ];
Chris@0:   switch ($failed_check) {
Chris@0:     case 'mb_strlen':
Chris@0:       $requirements['unicode']['description'] = t('Operations on Unicode strings are emulated on a best-effort basis. Install the <a href="http://php.net/mbstring">PHP mbstring extension</a> for improved Unicode support.');
Chris@0:       break;
Chris@0: 
Chris@0:     case 'mbstring.func_overload':
Chris@0:       $requirements['unicode']['description'] = t('Multibyte string function overloading in PHP is active and must be disabled. Check the php.ini <em>mbstring.func_overload</em> setting. Please refer to the <a href="http://php.net/mbstring">PHP mbstring documentation</a> for more information.');
Chris@0:       break;
Chris@0: 
Chris@0:     case 'mbstring.encoding_translation':
Chris@0:       $requirements['unicode']['description'] = t('Multibyte string input conversion in PHP is active and must be disabled. Check the php.ini <em>mbstring.encoding_translation</em> setting. Please refer to the <a href="http://php.net/mbstring">PHP mbstring documentation</a> for more information.');
Chris@0:       break;
Chris@0: 
Chris@0:     case 'mbstring.http_input':
Chris@0:       $requirements['unicode']['description'] = t('Multibyte string input conversion in PHP is active and must be disabled. Check the php.ini <em>mbstring.http_input</em> setting. Please refer to the <a href="http://php.net/mbstring">PHP mbstring documentation</a> for more information.');
Chris@0:       break;
Chris@0: 
Chris@0:     case 'mbstring.http_output':
Chris@0:       $requirements['unicode']['description'] = t('Multibyte string output conversion in PHP is active and must be disabled. Check the php.ini <em>mbstring.http_output</em> setting. Please refer to the <a href="http://php.net/mbstring">PHP mbstring documentation</a> for more information.');
Chris@0:       break;
Chris@0:   }
Chris@0: 
Chris@0:   return $requirements;
Chris@0: }
Chris@0: 
Chris@0: /**
Chris@0:  * Prepares a new XML parser.
Chris@0:  *
Chris@0:  * This is a wrapper around xml_parser_create() which extracts the encoding
Chris@0:  * from the XML data first and sets the output encoding to UTF-8. This function
Chris@0:  * should be used instead of xml_parser_create(), because PHP 4's XML parser
Chris@0:  * doesn't check the input encoding itself. "Starting from PHP 5, the input
Chris@0:  * encoding is automatically detected, so that the encoding parameter specifies
Chris@0:  * only the output encoding."
Chris@0:  *
Chris@0:  * This is also where unsupported encodings will be converted. Callers should
Chris@0:  * take this into account: $data might have been changed after the call.
Chris@0:  *
Chris@0:  * @param $data
Chris@0:  *   The XML data which will be parsed later.
Chris@0:  *
Chris@0:  * @return
Chris@0:  *   An XML parser object or FALSE on error.
Chris@0:  *
Chris@0:  * @ingroup php_wrappers
Chris@0:  *
Chris@0:  * @deprecated in Drupal 8.3.0 and will bre removed in Drupal 9.0.0. Use
Chris@0:  *   xml_parser_create() and
Chris@0:  *   xml_parser_set_option($xml_parser, XML_OPTION_TARGET_ENCODING, 'utf-8')
Chris@0:  *   instead.
Chris@0:  */
Chris@0: function drupal_xml_parser_create(&$data) {
Chris@0:   // Default XML encoding is UTF-8
Chris@0:   $encoding = 'utf-8';
Chris@0:   $bom = FALSE;
Chris@0: 
Chris@0:   // Check for UTF-8 byte order mark (PHP5's XML parser doesn't handle it).
Chris@0:   if (!strncmp($data, "\xEF\xBB\xBF", 3)) {
Chris@0:     $bom = TRUE;
Chris@0:     $data = substr($data, 3);
Chris@0:   }
Chris@0: 
Chris@0:   // Check for an encoding declaration in the XML prolog if no BOM was found.
Chris@0:   if (!$bom && preg_match('/^<\?xml[^>]+encoding="(.+?)"/', $data, $match)) {
Chris@0:     $encoding = $match[1];
Chris@0:   }
Chris@0: 
Chris@0:   // Unsupported encodings are converted here into UTF-8.
Chris@0:   $php_supported = ['utf-8', 'iso-8859-1', 'us-ascii'];
Chris@0:   if (!in_array(strtolower($encoding), $php_supported)) {
Chris@0:     $out = Unicode::convertToUtf8($data, $encoding);
Chris@0:     if ($out !== FALSE) {
Chris@0:       $encoding = 'utf-8';
Chris@0:       $data = preg_replace('/^(<\?xml[^>]+encoding)="(.+?)"/', '\\1="utf-8"', $out);
Chris@0:     }
Chris@0:     else {
Chris@0:       \Drupal::logger('php')->warning('Could not convert XML encoding %s to UTF-8.', ['%s' => $encoding]);
Chris@0:       return FALSE;
Chris@0:     }
Chris@0:   }
Chris@0: 
Chris@0:   $xml_parser = xml_parser_create($encoding);
Chris@0:   xml_parser_set_option($xml_parser, XML_OPTION_TARGET_ENCODING, 'utf-8');
Chris@0:   return $xml_parser;
Chris@0: }