You are here

class Drupal_Apache_Solr_Service in Apache Solr Search 6.2

Same name and namespace in other branches
  1. 5.2 Drupal_Apache_Solr_Service.php \Drupal_Apache_Solr_Service
  2. 6 Drupal_Apache_Solr_Service.php \Drupal_Apache_Solr_Service

Hierarchy

Expanded class hierarchy of Drupal_Apache_Solr_Service

1 string reference to 'Drupal_Apache_Solr_Service'
apachesolr_get_solr in ./apachesolr.module
Factory method for solr singleton object. Structure allows for an arbitrary number of solr objects to be used based on the host, port, path combination. Get an instance like this: $solr = apachesolr_get_solr();

File

./Drupal_Apache_Solr_Service.php, line 5

View source
class Drupal_Apache_Solr_Service extends Apache_Solr_Service {
  protected $luke;
  protected $luke_cid;
  protected $stats;
  const LUKE_SERVLET = 'admin/luke';
  const STATS_SERVLET = 'admin/stats.jsp';

  /**
   * Whether {@link Apache_Solr_Response} objects should create {@link Apache_Solr_Document}s in
   * the returned parsed data
   *
   * @var boolean
   *
   * @override to FALSE by default.
   */
  protected $_createDocuments = FALSE;

  /**
   * Whether {@link Apache_Solr_Response} objects should have multivalue fields with only a single value
   * collapsed to appear as a single value would.
   *
   * @var boolean
   *
   * @override to FALSE by default
   */
  protected $_collapseSingleValueArrays = FALSE;

  /**
   * Call the /admin/ping servlet, to test the connection to the server.
   *
   * @param $timeout
   *   maximum time to wait for ping in seconds, -1 for unlimited (default 2).
   * @return
   *   (float) seconds taken to ping the server, FALSE if timeout occurs.
   */
  public function ping($timeout = 2) {
    $start = microtime(TRUE);
    if ($timeout <= 0.0) {
      $timeout = -1;
    }

    // Attempt a HEAD request to the solr ping url.
    list($data, $headers) = $this
      ->_makeHttpRequest($this->_pingUrl, 'HEAD', array(), NULL, $timeout);
    $response = new Apache_Solr_Response($data, $headers);
    if ($response
      ->getHttpStatus() == 200) {

      // Add 0.1 ms to the ping time so we never return 0.0.
      return microtime(TRUE) - $start + 0.0001;
    }
    else {
      return FALSE;
    }
  }

  /**
   * Sets $this->luke with the meta-data about the index from admin/luke.
   */
  protected function setLuke($num_terms = 0) {
    if (empty($this->luke[$num_terms])) {
      $url = $this
        ->_constructUrl(self::LUKE_SERVLET, array(
        'numTerms' => "{$num_terms}",
        'wt' => self::SOLR_WRITER,
      ));
      $this->luke[$num_terms] = $this
        ->_sendRawGet($url);
      cache_set($this->luke_cid, $this->luke, 'cache_apachesolr');
    }
  }

  /**
   * Get just the field meta-data about the index.
   */
  public function getFields($num_terms = 0) {
    return $this
      ->getLuke($num_terms)->fields;
  }

  /**
   * Get meta-data about the index.
   */
  public function getLuke($num_terms = 0) {
    if (!isset($this->luke[$num_terms])) {
      $this
        ->setLuke($num_terms);
    }
    return $this->luke[$num_terms];
  }

  /**
   * Sets $this->stats with the information about the Solr Core form /admin/stats.jsp
   */
  protected function setStats() {
    $data = $this
      ->getLuke();

    // Only try to get stats if we have connected to the index.
    if (empty($this->stats) && isset($data->index->numDocs)) {
      $url = $this
        ->_constructUrl(self::STATS_SERVLET);
      $this->stats_cid = "apachesolr:stats:" . md5($url);
      $cache = cache_get($this->stats_cid, 'cache_apachesolr');
      if (isset($cache->data)) {
        $this->stats = simplexml_load_string($cache->data);
      }
      else {
        $response = $this
          ->_sendRawGet($url);
        $this->stats = simplexml_load_string($response
          ->getRawResponse());
        cache_set($this->stats_cid, $response
          ->getRawResponse(), 'cache_apachesolr');
      }
    }
  }

  /**
   * Get information about the Solr Core.
   *
   * Returns a Simple XMl document
   */
  public function getStats() {
    if (!isset($this->stats)) {
      $this
        ->setStats();
    }
    return $this->stats;
  }

  /**
   * Get summary information about the Solr Core.
   */
  public function getStatsSummary() {
    $stats = $this
      ->getStats();
    $summary = array(
      '@pending_docs' => '',
      '@autocommit_time_seconds' => '',
      '@autocommit_time' => '',
      '@deletes_by_id' => '',
      '@deletes_by_query' => '',
      '@deletes_total' => '',
      '@schema_version' => '',
      '@core_name' => '',
    );
    if (!empty($stats)) {
      $docs_pending_xpath = $stats
        ->xpath('//stat[@name="docsPending"]');
      $summary['@pending_docs'] = (int) trim($docs_pending_xpath[0]);
      $max_time_xpath = $stats
        ->xpath('//stat[@name="autocommit maxTime"]');
      $max_time = (int) trim(current($max_time_xpath));

      // Convert to seconds.
      $summary['@autocommit_time_seconds'] = $max_time / 1000;
      $summary['@autocommit_time'] = format_interval($max_time / 1000);
      $deletes_id_xpath = $stats
        ->xpath('//stat[@name="deletesById"]');
      $summary['@deletes_by_id'] = (int) trim($deletes_id_xpath[0]);
      $deletes_query_xpath = $stats
        ->xpath('//stat[@name="deletesByQuery"]');
      $summary['@deletes_by_query'] = (int) trim($deletes_query_xpath[0]);
      $summary['@deletes_total'] = $summary['@deletes_by_id'] + $summary['@deletes_by_query'];
      $schema = $stats
        ->xpath('/solr/schema[1]');
      $summary['@schema_version'] = trim($schema[0]);
      $core = $stats
        ->xpath('/solr/core[1]');
      $summary['@core_name'] = trim($core[0]);
    }
    return $summary;
  }

  /**
   * Clear cached Solr data.
   */
  public function clearCache() {

    // Don't clear cached data if the server is unavailable.
    if (@$this
      ->ping()) {
      $this
        ->_clearCache();
    }
    else {
      throw new Exception('No Solr instance available when trying to clear the cache.');
    }
  }
  protected function _clearCache() {
    cache_clear_all("apachesolr:luke:", 'cache_apachesolr', TRUE);
    cache_clear_all("apachesolr:stats:", 'cache_apachesolr', TRUE);
    $this->luke = array();
    $this->stats = NULL;
  }

  /**
   * Clear the cache whenever we commit changes.
   *
   * @see Apache_Solr_Service::commit()
   */
  public function commit($optimize = TRUE, $waitFlush = TRUE, $waitSearcher = TRUE, $timeout = 3600) {
    parent::commit($optimize, $waitFlush, $waitSearcher, $timeout);
    $this
      ->_clearCache();
  }

  /**
   * Construct the Full URLs for the three servlets we reference
   *
   * @see Apache_Solr_Service::_initUrls()
   */
  protected function _initUrls() {
    parent::_initUrls();
    $this->_lukeUrl = $this
      ->_constructUrl(self::LUKE_SERVLET, array(
      'numTerms' => '0',
      'wt' => self::SOLR_WRITER,
    ));
  }

  /**
   * Make a request to a servlet (a path) that's not a standard path.
   *
   * @param string $servlet
   *   A path to be added to the base Solr path. e.g. 'extract/tika'
   *
   * @param array $params
   *   Any request parameters when constructing the URL.
   *
   * @param string $method
   *   'GET', 'POST', 'PUT', or 'HEAD'.
   *
   * @param array $request_headers
   *   Keyed array of header names and values.  Should include 'Content-Type'
   *   for POST or PUT.
   *
   * @param string $rawPost
   *   Must be an empty string unless method is POST or PUT.
   *
   * @param float $timeout
   *   Read timeout in seconds or FALSE.
   *
   * @return
   *  Apache_Solr_Response object
   */
  public function makeServletRequest($servlet, $params = array(), $method = 'GET', $request_headers = array(), $rawPost = '', $timeout = FALSE) {
    if ($method == 'GET' || $method == 'HEAD') {

      // Make sure we are not sending a request body.
      $rawPost = '';
    }

    // Add default params.
    $params += array(
      'wt' => self::SOLR_WRITER,
    );
    $url = $this
      ->_constructUrl($servlet, $params);
    list($data, $headers) = $this
      ->_makeHttpRequest($url, $method, $request_headers, $rawPost, $timeout);
    $response = new Apache_Solr_Response($data, $headers, $this->_createDocuments, $this->_collapseSingleValueArrays);
    $code = (int) $response
      ->getHttpStatus();
    if ($code != 200) {
      $message = $response
        ->getHttpStatusMessage();
      if ($code >= 400 && $code != 403 && $code != 404) {

        // Add details, like Solr's exception message.
        $message .= $response
          ->getRawResponse();
      }
      throw new Exception('"' . $code . '" Status: ' . $message);
    }
    return $response;
  }

  /**
   * Put Luke meta-data from the cache into $this->luke when we instantiate.
   *
   * @see Apache_Solr_Service::__construct()
   */
  public function __construct($host = 'localhost', $port = 8180, $path = '/solr/') {
    parent::__construct($host, $port, $path);
    $this->luke_cid = "apachesolr:luke:" . md5($this->_lukeUrl);
    $cache = cache_get($this->luke_cid, 'cache_apachesolr');
    if (isset($cache->data)) {
      $this->luke = $cache->data;
    }
  }

  /**
   * Switch to POST for search if resultant query string is too long.
   *
   * Default threshold is 4000 characters. Note that this is almost an
   * exact copy of Apache_Solr_Service::search().
   *
   * @see Apache_Solr_Service::search
   */
  public function search($query, $offset = 0, $limit = 10, $params = array(), $method = self::METHOD_GET) {
    if (!is_array($params)) {
      $params = array();
    }

    // Common parameters in this interface.
    $params['wt'] = self::SOLR_WRITER;
    $params['json.nl'] = $this->_namedListTreatment;
    $params['q'] = $query;
    $params['start'] = $offset;
    $params['rows'] = $limit;

    // Use http_build_query to encode our arguments because its faster
    // than urlencoding all the parts ourselves in a loop.
    $queryString = http_build_query($params, null, $this->_queryStringDelimiter);

    // Because http_build_query treats arrays differently than we want
    // to, correct the query string by changing foo[#]=bar (# being an actual
    // number) parameter strings to just multiple foo=bar strings. This regex
    // should always work since '=' will be urlencoded anywhere else the
    // regex isn't expecting it.
    $queryString = preg_replace('/%5B(?:[0-9]|[1-9][0-9]+)%5D=/', '=', $queryString);

    // Check string length of the query string, change method to POST
    // if longer than 4000 characters.
    if (strlen($queryString) > variable_get('apachesolr_search_post_threshold', 4000)) {
      $method = self::METHOD_POST;
    }
    if ($method == self::METHOD_GET) {
      return $this
        ->_sendRawGet($this->_searchUrl . $this->_queryDelimiter . $queryString);
    }
    else {
      if ($method == self::METHOD_POST) {
        return $this
          ->_sendRawPost($this->_searchUrl, $queryString, FALSE, 'application/x-www-form-urlencoded; charset=UTF-8');
      }
      else {
        throw new Exception("Unsupported method '{$method}', please use the Apache_Solr_Service::METHOD_* constants");
      }
    }
  }

  /**
   * Central method for making a get operation against this Solr Server
   *
   * @see Apache_Solr_Service::_sendRawGet()
   */
  protected function _sendRawGet($url, $timeout = FALSE) {
    list($data, $headers) = $this
      ->_makeHttpRequest($url, 'GET', array(), '', $timeout);
    $response = new Apache_Solr_Response($data, $headers, $this->_createDocuments, $this->_collapseSingleValueArrays);
    $code = (int) $response
      ->getHttpStatus();
    if ($code != 200) {
      $message = $response
        ->getHttpStatusMessage();
      if ($code >= 400 && $code != 403 && $code != 404) {

        // Add details, like Solr's exception message.
        $message .= $response
          ->getRawResponse();
      }
      throw new Exception('"' . $code . '" Status: ' . $message);
    }
    return $response;
  }

  /**
   * Central method for making a post operation against this Solr Server
   *
   * @see Apache_Solr_Service::_sendRawPost()
   */
  protected function _sendRawPost($url, $rawPost, $timeout = FALSE, $contentType = 'text/xml; charset=UTF-8') {
    $request_headers = array(
      'Content-Type' => $contentType,
    );
    list($data, $headers) = $this
      ->_makeHttpRequest($url, 'POST', $request_headers, $rawPost, $timeout);
    $response = new Apache_Solr_Response($data, $headers, $this->_createDocuments, $this->_collapseSingleValueArrays);
    $code = (int) $response
      ->getHttpStatus();
    if ($code != 200) {
      $message = $response
        ->getHttpStatusMessage();
      if ($code >= 400 && $code != 403 && $code != 404) {

        // Add details, like Solr's exception message.
        $message .= $response
          ->getRawResponse();
      }
      throw new Exception('"' . $code . '" Status: ' . $message);
    }
    return $response;
  }
  protected function _makeHttpRequest($url, $method = 'GET', $headers = array(), $content = '', $timeout = FALSE) {

    // Set a response timeout
    if ($timeout) {
      $default_socket_timeout = ini_set('default_socket_timeout', $timeout);
    }
    $result = drupal_http_request($url, $headers, $method, $content);

    // Restore the response timeout
    if ($timeout) {
      ini_set('default_socket_timeout', $default_socket_timeout);
    }
    if (!isset($result->code) || $result->code < 0) {
      $result->code = 0;
      $result->status_message = 'Request failed';
    }
    if (isset($result->error)) {
      $result->status_message .= ': ' . check_plain($result->error);
    }
    if (!isset($result->data)) {
      $result->data = '';
    }
    $headers[] = "{$result->protocol} {$result->code} {$result->status_message}";
    if (isset($result->headers)) {
      foreach ($result->headers as $name => $value) {
        $headers[] = "{$name}: {$value}";
      }
    }
    return array(
      $result->data,
      $headers,
    );
  }

}

Members

Namesort descending Modifiers Type Description Overrides
Drupal_Apache_Solr_Service::$luke protected property
Drupal_Apache_Solr_Service::$luke_cid protected property
Drupal_Apache_Solr_Service::$stats protected property
Drupal_Apache_Solr_Service::$_collapseSingleValueArrays protected property Whether {@link Apache_Solr_Response} objects should have multivalue fields with only a single value collapsed to appear as a single value would.
Drupal_Apache_Solr_Service::$_createDocuments protected property Whether {@link Apache_Solr_Response} objects should create {@link Apache_Solr_Document}s in the returned parsed data
Drupal_Apache_Solr_Service::clearCache public function Clear cached Solr data.
Drupal_Apache_Solr_Service::commit public function Clear the cache whenever we commit changes.
Drupal_Apache_Solr_Service::getFields public function Get just the field meta-data about the index.
Drupal_Apache_Solr_Service::getLuke public function Get meta-data about the index.
Drupal_Apache_Solr_Service::getStats public function Get information about the Solr Core.
Drupal_Apache_Solr_Service::getStatsSummary public function Get summary information about the Solr Core.
Drupal_Apache_Solr_Service::LUKE_SERVLET constant
Drupal_Apache_Solr_Service::makeServletRequest public function Make a request to a servlet (a path) that's not a standard path.
Drupal_Apache_Solr_Service::ping public function Call the /admin/ping servlet, to test the connection to the server.
Drupal_Apache_Solr_Service::search public function Switch to POST for search if resultant query string is too long.
Drupal_Apache_Solr_Service::setLuke protected function Sets $this->luke with the meta-data about the index from admin/luke.
Drupal_Apache_Solr_Service::setStats protected function Sets $this->stats with the information about the Solr Core form /admin/stats.jsp
Drupal_Apache_Solr_Service::STATS_SERVLET constant
Drupal_Apache_Solr_Service::_clearCache protected function
Drupal_Apache_Solr_Service::_initUrls protected function Construct the Full URLs for the three servlets we reference
Drupal_Apache_Solr_Service::_makeHttpRequest protected function
Drupal_Apache_Solr_Service::_sendRawGet protected function Central method for making a get operation against this Solr Server
Drupal_Apache_Solr_Service::_sendRawPost protected function Central method for making a post operation against this Solr Server
Drupal_Apache_Solr_Service::__construct public function Put Luke meta-data from the cache into $this->luke when we instantiate.