function asset_search_parse_feed in Asset 6

Same name and namespace in other branches

1. 5.2 contrib/asset_search/asset_search.parser.inc \asset_search_parse_feed()

Parse an rss feed and return an array of items Taken from aggregator_parse_feed

1 call to asset_search_parse_feed()

asset_search_fetch in contrib/asset_search/asset_search.module

Fetch a RSS feed and return its items as pseudo-assets

File

contrib/asset_search/asset_search.parser.inc, line 13

Include file to parse RSS feeds into an array of items that will then be used as pseudo-assets.

Code

function asset_search_parse_feed(&$data, $type, $value) {
  global $items, $image, $channel;
  $tmp_assets = array();
  $type = asset_search_types($type);

  // Unset the global variables before we use them:
  unset($GLOBALS['element'], $GLOBALS['item'], $GLOBALS['tag']);
  $items = array();
  $image = array();
  $channel = array();

  // parse the data:
  $xml_parser = drupal_xml_parser_create($data);
  xml_set_element_handler($xml_parser, 'asset_search_element_start', 'asset_search_element_end');
  xml_set_character_data_handler($xml_parser, 'asset_search_element_data');
  if (!xml_parse($xml_parser, $data, 1)) {
    watchdog('asset_search', t('The feed seems to be broken, due to an error "%error" on line %line.', array(
      '%error' => xml_error_string(xml_get_error_code($xml_parser)),
      '%line' => xml_get_current_line_number($xml_parser),
    )), WATCHDOG_WARNING);
    drupal_set_message(t('The feed seems to be broken, because of error "%error" on line %line.', array(
      '%error' => xml_error_string(xml_get_error_code($xml_parser)),
      '%line' => xml_get_current_line_number($xml_parser),
    )), 'error');
    return 0;
  }
  xml_parser_free($xml_parser);

  // Initialize variables
  $title = $link = $author = $description = $guid = NULL;
  foreach ($items as $item) {
    unset($title, $link, $author, $description, $guid);

    // Prepare the item:
    foreach ($item as $key => $value) {
      $item[$key] = trim($value);
    }

    /*
     ** Resolve the item's title. If no title is found, we use
     ** up to 40 characters of the description ending at a word
     ** boundary but not splitting potential entities.
     */
    if ($item['TITLE']) {
      $title = $item['TITLE'];
    }
    else {
      $title = preg_replace('/^(.*)[^\\w;&].*?$/', "\\1", truncate_utf8($item['DESCRIPTION'], 40));
    }

    /*
     ** Resolve the items link.
     */
    if ($item['LINK']) {
      $link = $item['LINK'];
    }
    if ($item['GUID']) {
      $guid = $item['GUID'];
    }

    /**
     * Atom feeds have a CONTENT and/or SUMMARY tag instead of a DESCRIPTION tag
     */
    if ($item['CONTENT:ENCODED']) {
      $item['DESCRIPTION'] = $item['CONTENT:ENCODED'];
    }
    else {
      if ($item['SUMMARY']) {
        $item['DESCRIPTION'] = $item['SUMMARY'];
      }
      else {
        if ($item['CONTENT']) {
          $item['DESCRIPTION'] = $item['CONTENT'];
        }
      }
    }

    /*
     ** Try to resolve and parse the item's publication date. If no
     ** date is found, we use the current date instead.
     */
    if ($item['PUBDATE']) {
      $date = $item['PUBDATE'];
    }
    else {
      if ($item['DC:DATE']) {
        $date = $item['DC:DATE'];
      }
      else {
        if ($item['DCTERMS:ISSUED']) {
          $date = $item['DCTERMS:ISSUED'];
        }
        else {
          if ($item['DCTERMS:CREATED']) {
            $date = $item['DCTERMS:CREATED'];
          }
          else {
            if ($item['DCTERMS:MODIFIED']) {
              $date = $item['DCTERMS:MODIFIED'];
            }
            else {
              if ($item['ISSUED']) {
                $date = $item['ISSUED'];
              }
              else {
                if ($item['CREATED']) {
                  $date = $item['CREATED'];
                }
                else {
                  if ($item['MODIFIED']) {
                    $date = $item['MODIFIED'];
                  }
                  else {
                    if ($item['PUBLISHED']) {
                      $date = $item['PUBLISHED'];
                    }
                    else {
                      if ($item['UPDATED']) {
                        $date = $item['UPDATED'];
                      }
                      else {
                        $date = 'now';
                      }
                    }
                  }
                }
              }
            }
          }
        }
      }
    }
    $timestamp = strtotime($date);

    // As of PHP 5.1.0, strtotime returns FALSE on failure instead of -1.
    if ($timestamp <= 0) {
      $timestamp = asset_search_parse_w3cdtf($date);

      // Returns FALSE on failure
      if (!$timestamp) {
        $timestamp = time();

        // better than nothing
      }
    }
    $tmp = new stdClass();
    $tmp->aid = -1;
    $tmp->created = $timestamp;
    $tmp->title = $title;
    $tmp->link = $link;
    $tmp->guid = $guid;
    $tmp->cid = md5($guid);
    $tmp->search_type = $type;
    $tmp = module_invoke($type['module'], 'asset_search', 'feed item', $type, $tmp);
    cache_set($tmp->cid, 'cache_asset_search', serialize($tmp));
    $tmp_assets[$tmp->cid] = $tmp;
  }
  $channel['items'] = $tmp_assets;
  return $channel;
}