protected function Xliff::processForExport

Helper function to process the source text.

Parameters

string $source: Job data array.

array $key_array: The source item data key.

Return value

string

1 call to Xliff::processForExport()
Xliff::writeData in translators/tmgmt_file/src/Plugin/tmgmt_file/Format/Xliff.php
Writes text according to the XLIFF export settings.

File

translators/tmgmt_file/src/Plugin/tmgmt_file/Format/Xliff.php, line 464

Class

Xliff
Export to XLIFF format.

Namespace

Drupal\tmgmt_file\Plugin\tmgmt_file\Format

Code

protected function processForExport($source, array $key_array) {
  $tjiid = $key_array[0];
  $key_string = \Drupal::service('tmgmt.data')
    ->ensureStringKey($key_array);

  // The reason why we use DOMDocument object here and not just XMLReader
  // is the DOMDocument's ability to deal with broken HTML.
  $dom = new \DOMDocument();

  // We need to append the head with encoding so that special characters
  // are read correctly.
  $dom
    ->loadHTML("<html><head><meta http-equiv='Content-type' content='text/html; charset=UTF-8' /></head><body>" . $source . '</body></html>');
  $iterator = new \RecursiveIteratorIterator(new RecursiveDOMIterator($dom), \RecursiveIteratorIterator::SELF_FIRST);
  $writer = new \XMLWriter();
  $writer
    ->openMemory();
  $writer
    ->startDocument('1.0', 'UTF-8');
  $writer
    ->startElement('wrapper');
  $tray = array();
  $non_pair_tags = array(
    'br',
    'img',
  );
  $xliff_validation = $this->job
    ->getSetting('xliff_validation');

  /** @var \DOMElement $node */
  foreach ($iterator as $node) {
    if (in_array($node->nodeName, array(
      'html',
      'body',
      'head',
      'meta',
    ))) {
      continue;
    }
    if ($node->nodeType === XML_ELEMENT_NODE) {

      // Increment the elements count and compose element id.
      if (!isset($xliff_validation[$key_string])) {
        $xliff_validation[$key_string] = 0;
      }
      $xliff_validation[$key_string]++;
      $id = 'tjiid' . $tjiid . '-' . $xliff_validation[$key_string];
      $is_pair_tag = !in_array($node->nodeName, $non_pair_tags);
      if ($is_pair_tag) {
        $this
          ->writeBPT($writer, $node, $id);
      }
      elseif ($node->nodeName == 'img') {
        $this
          ->writeIMG($writer, $node, $id);
      }
      elseif ($node->nodeName == 'br') {
        $this
          ->writeBR($writer, $node, $id);
      }

      // Add to tray new element info.
      $tray[$id] = array(
        'name' => $node->nodeName,
        'id' => $id,
        'value' => $node->nodeValue,
        'built_text' => '',
        'is_pair_tag' => $is_pair_tag,
      );
    }
    elseif ($node->nodeName == '#text') {

      // Add the node value to the text output.
      $writer
        ->writeCdata($this
        ->toEntities($node->nodeValue));
      foreach ($tray as &$info) {
        $info['built_text'] .= $node->nodeValue;
      }
    }

    // Reverse so that pair tags are closed in the expected order.
    $reversed_tray = array_reverse($tray);
    foreach ($reversed_tray as $_info) {

      // If the build_text equals to the node value and it is not a pair tag
      // add the end pair tag markup.
      if ($_info['value'] == $_info['built_text'] && $_info['is_pair_tag']) {

        // Count also for the closing elements.
        $xliff_validation[$key_string]++;
        $this
          ->writeEPT($writer, $_info['name'], $_info['id']);

        // When the end pair tag has been written unset the element info
        // from the tray.
        unset($tray[$_info['id']]);
      }
    }
  }

  // Set the xliff_validation data and save the job.
  $this->job->settings->xliff_validation = $xliff_validation;
  $this->job
    ->save();
  $writer
    ->endElement();

  // Load the output with XMLReader so that we can easily get the inner xml.
  $reader = new \XMLReader();
  $reader
    ->XML($writer
    ->outputMemory());
  $reader
    ->read();
  return $reader
    ->readInnerXML();
}