Allow to use the spreadsheet iterator as a migrate_plus data parser

Added missing return types for Migrate Plus 6.x compatibility

Thank you @duaelfr and @Matthijs for this plugin. We've used for a couple of years. It worked great.
But recently migrated to Drupal 10 and also changed some data import logic.
We needed "id-less" data import. Well, actually migrate can't do that. But we decided to use
"row_index_column" field, provided by this module, and xls file_name as ids.

Here is our migration config:

source:
  plugin: custom_source_extends_url_source
  data_fetcher_plugin: file
  data_parser_plugin: custom_data_parser_extends_migrate_spreadsheet
  track_changes: true
  worksheet: 'Sheet1'
  origin: A2
  header_row: 1
  item_selector: null
  // Filled up on runtime
  urls: []
  columns: []
  fields: []

  # Actual ids for migrate source plugin will be next:
  # - file_name
  # - spreadsheet_row_index
  ids:
    spreadsheet_row_index:
      type: string
  row_index_column: 'spreadsheet_row_index'

And then we faced a number of issues with data parser plugin.
I'm not sure the roots of them, maybe updated migrate/migrate_plus/drupal apis.

But here are how we ended up using data parser plugin and fixed issues.

@claudiu.cristea was right about that file in openSourceUrl() can come not only from remote, but from local fs.

Here how delt with it:

protected function openSourceUrl($url): bool {
    $fetcher = $this->getDataFetcherPlugin();

    if ($fetcher->getPluginId() === 'file') {
      $this->configuration['file'] = $url;
    }
    else {
      // Why we want to read file two times?
      /** @var \GuzzleHttp\Psr7\Response $response */
      $response = $this->getDataFetcherPlugin()->getResponse($url);

      preg_match('/filename="(.*?)"/', $response->getHeader('Content-Disposition')[0] ?? '', $matches);
      $this->configuration['file'] = $this->fileSystem->saveData($response->getBody(), 'temporary://' . $matches[1], FileSystemInterface::EXISTS_RENAME);
    }

    $this->initializeIterator();
    return TRUE;
  }

Next thing. In our case checking for an empty current item in valid() was not enought. We got infinite loop.
So ended up with checking row_index_column and rows count:

public function valid(): bool {
    $currentItem = $this->currentItem;
    if (!$currentItem) {
      return FALSE;
    }
    if (!parent::valid()) {
      return FALSE;
    }

    // Check iterator for current/max index.
    // Fix issue with lost last record.
    $rowIndexCol = $this->configuration['row_index_column'] ?? NULL;
    if ($rowIndexCol) {
      $index = $currentItem[$rowIndexCol];
      $max = $this->spreadsheetIterator->getRowsCount();

      if ($index <= $max) {
        return TRUE;
      }
    }
    // Fallback to default check, which doesn't work for us,
    // but should work for others if they use "row_index_column" feature.
    else {
      return $this->spreadsheetIterator->valid();
    }

    return FALSE;
  }

BTW: there is also php notice for undefined row_index_column array key in SpreadsheetIterator class but I'll report this issue separately and link it here.