function feeds_crawler_find_next in Feeds Crawler 6

Same name and namespace in other branches

1. 6.2 feeds_crawler.module \feeds_crawler_find_next()

1 call to feeds_crawler_find_next()

feeds_crawler_batch in ./feeds_crawler.module

Batch callback.

File

./feeds_crawler.module, line 71

Code

function feeds_crawler_find_next($autodetect, $xpath, $raw, $base_url, $html) {
  if ($html == 'html') {
    $dom = new DOMDocument();
    $success = @$dom
      ->loadHTML($raw);
    if (!$success) {
      drupal_set_message(t('There was an error parsing the HTML document at %url.', array(
        '%url' => $base_url,
      )), 'error');
      return FALSE;
    }
    $xml = simplexml_import_dom($dom);
    unset($dom);
  }
  else {
    $xml = @new SimpleXMLElement($raw);
  }
  if ($autodetect && $html == 'xml') {
    feeds_include_library('common_syndication_parser.inc', 'common_syndication_parser');
    $format = _parser_common_syndication_feed_format_detect($xml);
    if ($format) {
      $xml
        ->registerXpathNamespace('atom', 'http://www.w3.org/2005/Atom');
      $xpath = 'atom:link[@rel="next"]/@href';
    }
    else {
      $xpath = 'link[@rel="next"]/@href';
    }
  }
  else {
    if ($autodetect && $html == 'html') {
      $xpath = 'link[@rel="next"]';
    }
  }
  $href = $xml
    ->xpath($xpath);
  unset($xml);
  if ($href === FALSE || empty($href)) {
    return FALSE;
  }
  $href = (string) $href[0];
  if (strpos($href, 'http') !== 0) {
    $href = $base_url . $href;
  }
  return $href;
}