<?php
// $Id: apachesolr_search.module,v 1.1.2.6.2.63 2009-01-27 21:32:34 pwolanin Exp $

/**
 * @file
 *   Provides a content search implementation for node content for use with the
 *   Apache Solr search application.
 */

/**
 * Implementation of hook_help().
 */
function apachesolr_search_help($section) {
  switch ($section) {
    case 'admin/settings/apachesolr/index':
      $remaining = 0;
      $total = 0;
      // Collect the stats
      $status = module_invoke('apachesolr_search', 'search', 'status');
      $remaining += $status['remaining'];
      $total += $status['total'];

      return t('The Apache Solr search index is generated by !cron. %percentage of the site has been indexed. There @items left to index.', array(
        '!cron' => l(t('running cron'), 'admin/reports/status/run-cron', array('query' => array('destination' => 'admin/settings/apachesolr/index'))),
        '%percentage' => ((int)min(100, 100 * ($total - $remaining) / max(1, $total))) .'%',
        '@items' => format_plural($remaining, t('is 1 item'), t('are @count items')
      )));
  }
}

/**
 * Implementation of hook_menu().
 */
function apachesolr_search_menu() {
  $items['admin/settings/apachesolr/query-fields'] = array(
    'title'            => 'Query field settings',
    'page callback'    => 'apachesolr_search_settings_page',
    'access arguments' => array('administer site configuration'),
    'weight'           => 1,
    'type'             => MENU_LOCAL_TASK,
    'file'             => 'apachesolr_search.admin.inc',
  );

  return $items;
}

/**
 * Implementation of hook_update_index().
 */
function apachesolr_search_update_index() {
  $cron_limit = variable_get('apachesolr_cron_limit', 50);
  $result = apachesolr_get_nodes_to_index('apachesolr_search', $cron_limit);
  apachesolr_index_nodes($result, 'apachesolr_search');
}

/**
 * Implementation of hook_search().
 */
function apachesolr_search_search($op = 'search', $keys = NULL) {

  switch ($op) {
    case 'name':
      return t('Search');

    case 'reset':
      apachesolr_clear_last_index('apachesolr_search');
      return;

    case 'status':
      return apachesolr_index_status('apachesolr_search');

    case 'search':

      try {
        // This is the object that knows about the query coming from the user.
        $query = apachesolr_current_query($keys);
        if (is_null($query)) {
          throw new Exception(t('Could not construct a Solr query in function apachesolr_search_search()'));
        }

        $results = array();

        $params = array(
          'fl' => 'id,nid,title,comment_count,type,created,changed,score,url,uid,name',
          'rows' => variable_get('apachesolr_rows', 10),
          'facet' => 'true',
          'facet.mincount' => 1,
          'facet.sort' => 'true'
        );

        /**
         * Highlighting settings
         * These settings are set in solrconfig.xml.
         * See the defaults there.
         * If you wish to override them, you can via settings.php
         */

        $params['hl'] = variable_get('apachesolr_hl_active', NULL);
        $params['hl.fragsize']= variable_get('apachesolr_hl_textsnippetlength', NULL);
        $params['hl.simple.pre'] = variable_get('apachesolr_hl_pretag', NULL);
        $params['hl.simple.post'] = variable_get('apachesolr_hl_posttag', NULL);
        $params['hl.snippets'] = variable_get('apachesolr_hl_numsnippets', NULL);
        $params['hl.fl'] = variable_get('apachesolr_hl_fieldtohightlight', NULL);
        // We default to getting snippets from the body.
        $hl_fl = is_null($params['hl.fl']) ? 'body' : $params['hl.fl'];

        if (variable_get('apachesolr_search_spellcheck', FALSE)) {
          //Add new parameter to the search request
          $params['spellcheck.q'] = $query->get_query_basic();
          $params['spellcheck'] = 'true';
        }

        $facet_query_limits = variable_get('apachesolr_facet_query_limits', array());
        // Request all enabled facets.
        foreach (apachesolr_get_enabled_facets() as $module => $module_facets) {
          foreach($module_facets as $delta => $facet_field) {
            $params['facet.field'][] = $facet_field;
            // Facet limits
            if (isset($facet_query_limits[$module][$delta])) {
              $params['f.' . $facet_field . '.facet.limit'] = $facet_query_limits[$module][$delta];
            }
          }
        }
        if (!empty($params['facet.field'])) {
          // Add a default limit for fields where no limit was set.
          $params['facet.limit'] = variable_get('apachesolr_facet_query_limit_default', 20);
        }

        if (isset($_GET['solrsort'])) {
          $sort = check_plain($_GET['solrsort']);
        }

        // Validate sort parameter
        if (isset($sort) && preg_match('/^([a-z0-9_]+ (asc|desc)(,)?)+$/i', $sort)) {
          $params['sort'] = $sort;
        }

        $page = isset($_GET['page']) ? $_GET['page'] : 0;
        $params['start'] = $page * $params['rows'];
        // This is the object that does the communication with the solr server.
        $solr = apachesolr_get_solr();

        // Note - we have query fields set in solrconfig.xml, which will operate when
        // none are set.
        $qf = variable_get('apachesolr_search_query_fields', array());
        $fields = $solr->getFields();
        if ($qf && $fields) {
          foreach ($fields as $field_name => $field) {
            if (!empty($qf[$field_name])) {
              if ($field_name == 'body') {
                // Body is the only normed field.
                $qf[$field_name] *= 40.0;
              }
              $params['qf'][] = $field_name . '^'. $qf[$field_name];
            }
          }
        }
        // Note: we use 2 since 1 fails on Ubuntu Hardy.
        $data = $solr->getLuke(2);
        if (isset($data->index->numDocs)) {
          $total = $data->index->numDocs;
        }
        else {
          $total = db_result(db_query("SELECT COUNT(nid) FROM {node}"));
        }
        $date_settings = variable_get('apachesolr_search_date_boost', '4:200.0');
        list($date_steepness, $date_boost) = explode(':', $date_settings);
        // Default date-biasing function, as suggested (but steeper) at
        // http://wiki.apache.org/solr/DisMaxRequestHandler
        // rord() returns 1 for the newset doc, and the number in the index for
        // the oldest doc.  The function is thus: $total/(rord()*$steepness + $total).
        if ($date_boost) {
          $params['bf'][] = "recip(rord(created),$date_steepness,$total,$total)^$date_boost";
        }
        $comment_settings = variable_get('apachesolr_search_comment_boost', '4:200.0');
        list($comment_steepness, $comment_boost) = explode(':', $comment_settings);
        // Default date-biasing function, as suggested (but steeper) at
        // http://wiki.apache.org/solr/DisMaxRequestHandler
        // rord() returns 1 for the newset doc, and the number in the index for
        // the oldest doc.  The function is thus: $total/(rord()*$steepness + $total).
        if ($comment_boost) {
          $params['bf'][] = "recip(rord(comment_count),$comment_steepness,$total,$total)^$comment_boost";
        }
        // This hook allows modules to modify the query and params objects.
        apachesolr_modify_query($query, $params, 'apachesolr_search');
        if (!$query) {
          return array();
        }

        $response = $solr->search($query->get_query_basic(), $params['start'], $params['rows'], $params);
        // The response is cached so that it is accessible to the blocks and anything
        // else that needs it beyond the initial search.
        $total = $response->response->numFound;
        apachesolr_static_response_cache($response);
        apachesolr_has_searched(TRUE);
        pager_query("SELECT %d", $params['rows'], 0, NULL, $total);
        if ($total > 0) {
          foreach ($response->response->docs as $doc) {
            $extra = array();
            $snippet = isset($response->highlighting->{$doc->id}->$hl_fl) ? theme('apachesolr_search_snippets', $doc, $response->highlighting->{$doc->id}->$hl_fl) : '';
            if (!isset($doc->body)) {
              $doc->body = $snippet;
            }
            $doc->created = strtotime($doc->created);
            $doc->changed = strtotime($doc->changed);
            $extra += node_invoke_nodeapi($doc, 'search result');
            $extra['score'] = $doc->score;
            $results[] = array(
              'link' => $doc->url,
              'type' => node_get_types('name', $doc),
              'title' => $doc->title,
              'user' => theme('username', $doc),
              'date' => $doc->created,
              'node' => $doc,
              'extra' => $extra,
              'score' => $doc->score,
              'snippet' => $snippet,
            );
          }

          // Hook to allow modifications of the retrieved results
          foreach (module_implements('apachesolr_process_results') as $module) {
            $function = $module .'_apachesolr_process_results';
            call_user_func_array($function, array(&$results));
          }
        }
        // Set breadcrumb
        drupal_set_breadcrumb($query->get_breadcrumb());
        return $results;

      } // try
      catch (Exception $e) {
        watchdog('Apache Solr', $e->getMessage(), NULL, WATCHDOG_ERROR);
        apachesolr_failure(t('Solr search'), is_null($query) ? $keys : $query->get_query_basic());
      }
      break;
  } // switch
}

/**
 * Implementation of hook_apachesolr_facets().
 *
 * Returns an array keyed by block delta.
 */
function apachesolr_search_apachesolr_facets() {
  $facets = array();

  $facets['type'] = array(
    'info' => t('Apache Solr Search: Filter by content type'),
    'facet_field' => 'type',
  );
  $facets['uid'] = array(
    'info' => t('Apache Solr Search: Filter by author'),
    'facet_field' => 'uid',
  );
  $facets['language'] = array(
    'info' => t('Apache Solr Search: Filter by language'),
    'facet_field' => 'language',
  );

  // Get taxonomy vocabulary facets.
  if (module_exists('taxonomy')) {
    $vocabs = taxonomy_get_vocabularies();
    foreach ($vocabs as $vid => $vocab) {
      // In this case the delta and facet field are the same.
      $delta = 'imfield_vid_' . $vid;
      $facets[$delta] = array(
        'info' => t('Apache Solr Search: Filter by @name', array('@name' => $vocab->name)),
        'facet_field' => $delta,
      );
    }
  }

  // Get CCK field facets.
  $fields = apachesolr_cck_fields();
  if ($fields) {
    foreach ($fields as $name => $field) {
      // $delta can only be 32 chars, and the CCK field name may be this
      // long also, so we cannot add anything to it.
      $facets[$field['field_name']] = array(
        'info' => t('Apache Solr Search: Filter by @field', array('@field' => $field['label'])),
        'facet_field' => apachesolr_index_key($field),
      );
    }
  }
  return $facets;
}

/**
 * Implementation of hook_block().
 */
function apachesolr_search_block($op = 'list', $delta = 0, $edit = array()) {

  switch ($op) {
    case 'list':
      $enabled_facets = apachesolr_get_enabled_facets('apachesolr_search');
      $facets = apachesolr_search_apachesolr_facets();
      // Add the blocks
      $blocks = array();
      foreach ($enabled_facets as $delta => $facet_field) {
        $blocks[$delta] = $facets[$delta] + array('cache' => BLOCK_CACHE_PER_PAGE,);
      }
      $blocks['currentsearch'] = array(
        'info' => t('Apache Solr Search: Current search'),
        'cache' => BLOCK_CACHE_PER_PAGE,
      );
      return $blocks;

    case 'view':
      if (apachesolr_has_searched()) {
        // Get the query and response. Without these no blocks make sense.
        $response = apachesolr_static_response_cache();
        if (empty($response)) {
          return;
        }
        $query = apachesolr_current_query();

        $facets = apachesolr_get_enabled_facets('apachesolr_search');
        if (empty($facets[$delta]) && ($delta != 'currentsearch')) {
          return;
        }

        // Get information needed by the taxonomy blocks about limits.
        $initial_limits = variable_get('apachesolr_facet_query_initial_limits', array());
        $limit_default = variable_get('apachesolr_facet_query_initial_limit_default', 10);

        // Handle taxonomy vocabulary facets
        if ((strpos($delta, 'imfield_vid_') === 0) && module_exists('taxonomy')) {

          if (is_object($response->facet_counts->facet_fields->$delta)) {
            $contains_active = FALSE;
            $terms = array();

            foreach ($response->facet_counts->facet_fields->$delta as $tid => $count) {
              $unclick_link = '';
              unset($active);
              $term = taxonomy_get_term($tid);
              $new_query = clone $query;
              if ($active = $query->has_field('tid', $tid)) {
                $contains_active = TRUE;
                $new_query->remove_field('tid', $term->tid);
                $path = 'search/' . arg(1) . '/' . $new_query->get_query_basic();
                $querystring = $new_query->get_url_querystring();
                $unclick_link = theme('apachesolr_unclick_link', $path, $querystring);
              }
              else {
                $new_query->add_field('tid', $term->tid);
                $path = 'search/' . arg(1) . '/' . $new_query->get_query_basic();
                $querystring = $new_query->get_url_querystring();
              }
              $countsort = $count == 0 ? '' : 1 / $count;
              // if numdocs == 1 and !active, don't add.
              if ($response->numFound == 1 && !$active) {
                // skip
              }
              else {
                $terms[$term->vid][$active ? $countsort . $term->name : 1 + $countsort . $term->name] = theme('apachesolr_facet_item', $term->name, $count, $path, $querystring, $active, $unclick_link, $response->numFound);
              }
            }
          }
          $vid = substr($delta, 12);
          $vocab = taxonomy_vocabulary_load($vid);
          if (is_numeric($vid) && is_array($terms) && isset($terms[$vid]) && is_array($terms[$vid])) {
            ksort($terms[$vid]);
            $limit = isset($initial_limits['apachesolr_search'][$delta]) ? $initial_limits['apachesolr_search'][$delta] : $limit_default;
            return array(
              'subject' => t('Filter by @name', array('@name' => $vocab->name)),
              'content' => theme('apachesolr_facet_list', $terms[$vid], $limit),
            );
          }
          return;
        }

        switch ($delta) {
          case 'currentsearch':
            $fields = $query->get_fields();
            $search_keys = $query->get_query_basic();
            $path = 'search/' . arg(1) . '/' . $search_keys;
            $options = array();
            if (!$fields) {
              $options['attributes']['class'] = 'active';
            }
            $links[] = apachesolr_l($search_keys, $path, $options);
            foreach($fields as $field) {
              if ($field['#name']) {
                $new_query = clone $query;
                $new_query->remove_field($field['#name'], $field['#value']);
                $path = 'search/'. arg(1) .'/'. $new_query->get_query_basic();
                $querystring = $new_query->get_url_querystring();
                $unclick_link = theme('apachesolr_unclick_link', $path, $querystring);
                if (! $fielddisplay = theme("apachesolr_breadcrumb_". $field['#name'], $field['#value'])) {
                  $fielddisplay = $field['#value'];
                }
                $links[] = theme('apachesolr_facet_item', $fielddisplay, NULL, $path, $querystring, $active, $unclick_link, $response->numFound);
              }
            }
            $content = theme('apachesolr_currentsearch', $response->response->numFound, $links);
            return array('subject' => t('Current search'), 'content' => $content);

          case 'language':
            return apachesolr_facet_block($response, $query, 'apachesolr_search', $delta, $delta, t('Filter by language'), 'locale_language_name');
           case 'uid':
            return apachesolr_facet_block($response, $query, 'apachesolr_search', $delta, $delta, t('Filter by author'), 'apachesolr_search_get_username');
          case 'type':
            return apachesolr_facet_block($response, $query, 'apachesolr_search', $delta, $delta, t('Filter by type'), 'apachesolr_search_get_type');

          default:
           if ($fields = apachesolr_cck_fields()) {
            foreach ($fields as $name => $field) {
              if ($field['field_name'] == $delta) {
                $index_key = apachesolr_index_key($field);
                return apachesolr_facet_block($response, $query, 'apachesolr_search', $delta, $index_key, t('Filter by @field', array('@field' => $field['label'])));
              }
            }
          }
        }
        break;
      }
      break;

    case 'configure':
      if ($delta != 'currentsearch') {
        return apachesolr_facetcount_form('apachesolr_search', $delta);
      }
      break;

    case 'save':
      if ($delta != 'currentsearch') {
        apachesolr_facetcount_save($edit);
      }
      break;
  }
}

/**
 * Callback function for the 'Filter by name' facet block.
 */
function apachesolr_search_get_username($facet) {
  if ($facet == 0) {
    return variable_get('anonymous', t('Anonymous'));
  }
  else {
    return db_result(db_query("SELECT name FROM {users} WHERE uid = %d", $facet));
  }
}

/**
 * Callback function for the 'Filter by type' facet block.
 */
function apachesolr_search_get_type($facet) {
  return node_get_types('name', $facet);
}

/**
 * Implementation of hook_form_[form_id]_alter().
 *
 * This adds spelling suggestions to the search form.
 */
function apachesolr_search_form_search_form_alter(&$form, $form_state) {

  if (($form['module']['#value'] == 'apachesolr_search') && variable_get('apachesolr_search_spellcheck', FALSE) && apachesolr_has_searched() && ($response = apachesolr_static_response_cache())) {
    //Get spellchecker suggestions into an array.
    $suggestions = get_object_vars($response->spellcheck->suggestions);

    if ($suggestions) {
      //Get the original query and replace words.
      $query = apachesolr_current_query();

      foreach($suggestions as $word => $value) {
        $replacements[$word] = $value->suggestion[0];
      }
      $new_keywords = strtr($query->get_query_basic(), $replacements);

      $form['basic']['suggestion'] = array(
        '#prefix' => '<div class="spelling-suggestions">',
        '#suffix' => '</div>',
        '#type' => 'item',
        '#title' => t('Did you mean'),
        '#value' => l($new_keywords, 'search/'. arg(1) .'/'. $new_keywords),
      );
    }
  }
}

/**
 * Implementation of hook_form_[form_id]_alter().
 *
 * This adds spelling suggestions to the search form.
 */
function apachesolr_search_form_apachesolr_settings_alter(&$form, $form_state) {
  $form['apachesolr_search_spellcheck'] = array(
    '#type' => 'checkbox',
    '#title' => t('Enable spellchecker and suggestions'),
    '#default_value' => variable_get('apachesolr_search_spellcheck', FALSE),
    '#description' => t('Enable spellchecker and get word suggestions. Also known as the "Did you mean ... ?" feature.'),
  );

  $form['#submit'][] = 'apachesolr_search_build_spellcheck';
  // Move buttons to the bottom.
  $buttons = $form['buttons'];
  unset($form['buttons']);
  $form['buttons'] = $buttons;
}

function apachesolr_search_build_spellcheck() {
  try {
    $solr = apachesolr_get_solr();
    $params['spellcheck'] = 'true';
    $params['spellcheck.build'] = 'true';
    $response = $solr->search('solr', 0, 0, $params);
  }
  catch (Exception $e) {
    watchdog('Apache Solr', $e->getMessage(), NULL, WATCHDOG_ERROR);
  }
}

/**
 * Implementation of hook_theme().
 */
function apachesolr_search_theme() {
  return array(
    'apachesolr_breadcrumb_uid' => array(
      'arguments' => array('uid' => NULL),
    ),
    'apachesolr_breadcrumb_tid' => array(
      'arguments' => array('tid' => NULL),
    ),
    'apachesolr_breadcrumb_type' => array(
      'arguments' => array('type' => NULL),
    ),
    'apachesolr_currentsearch' => array(
      'arguments' => array('total_found' => NULL, 'links' => NULL),
    ),
    'apachesolr_search_snippets' => array(
      'arguments' => array('doc' => NULL, 'snippets' => NULL),
    ),
  );
}

/**
 * Return the username from $uid
 */
function theme_apachesolr_breadcrumb_uid($uid) {
  return apachesolr_search_get_username($uid);
}

/**
 * Return the term name from $tid.
 */
function theme_apachesolr_breadcrumb_tid($tid) {
  $term = taxonomy_get_term($tid);
  return $term->name;
}

/**
 * Return the human readable text for a content type.
 */
function theme_apachesolr_breadcrumb_type($type) {
  return node_get_types('name', $type);
}

/**
 * Return current search block contents
 */
function theme_apachesolr_currentsearch($total_found, $links) {
  return theme_item_list($links, t('Search found @count items', array('@count' => $total_found)));
}

/**
 * Returns the snipit text for a search entry
 *
 * @param object $doc
 * @param array $snippets
 *
 */
function theme_apachesolr_search_snippets($doc, $snippets) {
  return implode(' ... ', $snippets);
}