Programatically use Solr on Drupal 8

README.md

🔍 Solr (6.1.0+) search Engine & Tika (1.13+) Extractor

We are using solr for search index.

Solr need to be configured for drupal. Follow the INSTALL.txt found in the search_api_solr module.

As a pre-requisite for running your own Solr server, you'll need Java 6 or higher.

Installation

Install all prerequisites and configuration from web/modules/contrib/search_api_solr/INSTALL.txt then

Start Solr with: $ bin/solr start

Create the new instance index-my-project $ bin/solr create -c index-my-project

The creation will generate a folder under $SOLR_PATH/6.1.0/server/solr/index-my-project

Copy all the configs files from web/modules/contrib/search_api_solr/solr-conf/6.x/* into $SOLR_PATH/6.1.0/server/solr/index-my-project/conf

$ cp web/modules/contrib/search_api_solr/solr-conf/6.x/* $SOLR_PATH/6.1.0/server/solr/index-my-project/conf

Restart Solr $ bin/solr restart

Check your Solr status with solr status and with http://localhost:8983/solr/#/

Usage

Start Solr with:

$ bin/solr start

Index all your content with:

$ drush cron

Clean the index:

$ drush sapi-c [index]

Index everything at once:

$ drush sapi-i [index]

Drupal Configuration

Check the Solr configuration on admin/config/search/search-api/server/solr/edit and the Tika under admin/config/search/search_api_attachments

Add partial search

By default Solr don't match partial search, you should edit the schema.xml file and add the following filters on the fieldType name="text".

<fieldType name="text" class="solr.TextField" positionIncrementGap="100">

<filter class="solr.LowerCaseFilterFactory"/>
<filter class="solr.EdgeNGramFilterFactory" minGramSize="2" maxGramSize="25" />

Improved natural language search

You should improved the natural search by editing the stopwords.txt file. Check stopwords on stopwords-iso

Replace the default SnowballPorterFilterFactory from language="English" to language="French",

You should improved the natural search by editing the synonyms.txt file.

• Check synonyms for english (~1500 words) from Roget's Thesaurus
• Check synonyms (~143405 words) for french from termsuite
• Check synonyms for french (~326 words) from kevinbouge

Highlights

http://www.codewrecks.com/blog/index.php/2013/05/27/hilight-matched-text-inside-documents-indexed-with-solr-plus-tika/

SearchController.php

<?php
/**
* @file
* Contains \Drupal\custom_search\Controller\SearchController.
*/

namespace Drupal\custom_search\Controller;

use Drupal\Core\Controller\ControllerBase;
use Symfony\Component\DependencyInjection\ContainerInterface;
use Drupal\search_api\ParseMode\ParseModePluginManager;
use Symfony\Component\HttpFoundation\RequestStack;
use Drupal\search_api\Entity\Index;
use Drupal\Core\Entity\EntityTypeManagerInterface;
use Drupal\Core\Url;

/**
* SearchController.
*/
class SearchController extends ControllerBase {
    /**
    * Number of result by page
    * @var Integer
    */
    private $limit = 12;

    /**
    * Request stack that controls the lifecycle of requests
    * @var RequestStack
    */
    private $request;

    /**
     * The node Storage.
     *
     * @var \Drupal\node\NodeStorageInterface
     */
    protected $nodeStorage;

    /**
     * The parse mode manager.
     *
     * @var \Drupal\search_api\ParseMode\ParseModePluginManager
     */
    private $parseModeManager;

    /**
    * Class constructor.
    */
    public function __construct(EntityTypeManagerInterface $entity, RequestStack $request, ParseModePluginManager $parse_mode_manager) {
      $this->nodeStorage      = $entity->getStorage('node');
      $this->request          = $request->getMasterRequest();
      $this->parseModeManager = $parse_mode_manager;
    }

    /**
    * {@inheritdoc}
    */
    public static function create(ContainerInterface $container) {
        // Instantiates this form class.
        return new static(
          $container->get('entity_type.manager'),
          $container->get('request_stack'),
          $container->get('plugin.manager.search_api.parse_mode')
        );
    }

    public function search() {
        $variables = array('results' => array(), 'search' => array());

        // Retrieve routes with parameters
        $route = $this->request->attributes->get('_route');
        $params = $this->request->query->all();

        $search = $this->request->query->get('q');
        $type = $this->request->query->get('type');

        if (!empty($search)) {

            /* @var $search_api_index \Drupal\search_api\IndexInterface */
            $search_api_index = Index::load('full_website');

            // Create the query.
            $query = $search_api_index->query([
                'limit'  => $this->limit,
                'offset' => !is_null($this->request->get('page')) ? $this->request->get('page') * $this->limit : 0,
            ]);

            $query->setFulltextFields(['title', 'body', 'filename', 'saa_field_file_document', 'saa_field_file_news', 'saa_field_file_page']);

            $parse_mode = $this->parseModeManager->createInstance('terms');
            $query->setParseMode($parse_mode);
            // $parse_mode->setConjunction('OR');
            $query->keys($search);

            $query->sort('search_api_relevance', 'DESC');

            // Facets
            $server = $search_api_index->getServerInstance();
            if ($server->supportsFeature('search_api_facets')) {
              $query->setOption('search_api_facets', [
                'type' => [
                  'field' => 'type',
                  'limit' => 20,
                  'operator' => 'AND',
                  'min_count' => 1,
                  'missing' => TRUE,
                ],
              ]);
            }

            // Retrieve facets before.
            $query_facets = clone $query;
            $results_facets = $query_facets->execute();
            $facets = $results_facets->getExtraData('search_api_facets', []);

            if (!empty($type)) {
              $query = $query->addCondition('type', $type);
            }
            $results = $query->execute();

            if (isset($facets['type']) && !empty($facets['type'])) {
              foreach ($facets['type'] as $key => $facet) {
                $facets['type'][$key]['filter'] = trim($facet['filter'], '"');
                $facets['type'][$key]['filter_name'] = $this->typeMachineNameToHuman(trim($facet['filter'], '"'));

                $params['type'] = $facets['type'][$key]['filter'];
                $facets['type'][$key]['url'] = Url::fromRoute($route, $params);
              }
            }

            $params = $this->request->query->all();

            $variables['search'] = array(
                'search'    => $search,
                'count'     => $results->getResultCount(),
                'facets'    => $facets,
                'type'      => $type,
                'type_name' => $this->typeMachineNameToHuman($type),
            );

            $solr_response = $results->getExtraData('search_api_solr_response');
            foreach ($results as $key => $result) {
              // $highlights = $this->getHighlights($result->getId(), $solr_response['highlighting']);
              // $excerpt = $this->getExcerpt($highlights);
                $excerpt = $result->getExcerpt();
                $variables['results'][$key] = array(
                    'nid'           => isset($result->getField('nid')->getValues()[0]) ? $result->getField('nid')->getValues()[0] : NULL,
                    'title'         => isset($result->getField('title')->getValues()[0]) ? $result->getField('title')->getValues()[0] : NULL,
                    'body'          => $excerpt,
                    'original_body' => isset($result->getField('body')->getValues()[0]) ? $result->getField('body')->getValues()[0] : NULL
                );

                if (isset($result->getField('type')->getValues()[0])) {
                  $label = $this->typeMachineNameToHuman($result->getField('type')->getValues()[0]);

                  $variables['results'][$key]['type'] = $label;
                }
            }

            // Pager
            pager_default_initialize($results->getResultCount(), $this->limit);
            $variables['pager'] = array(
                '#type' => 'pager',
                '#quantity' => '5',
            );

        }

        return [
          '#theme'     => 'custom_search_search_page',
          '#variables' => $variables,
          // Set cache for 0 seconds.
          '#cache' => ['max-age' => 0],
        ];
    }

    private function typeMachineNameToHuman($machine_name) {
      $label = ucfirst($machine_name);

      switch ($machine_name) {
          case 'news':
          $label = t('Actualité');
          break;

          case 'advisor':
          $label = t('Conseiller');
          break;

          case 'product':
          $label = t('Produit');
          break;

          case 'faq':
          $label = t('Question-réponse');
          break;

          case 'partnership':
          $label = t('Partenaire');
          break;

          case 'offer':
          $label = t('Coupons Bella Vita');
          break;

          case 'building':
          $label = t('Construction');
          break;

          case 'managementcontracts':
          case 'management_contracts':
          $label = t('Mandats de gestion');
          break;
      }

      return $label;
    }


    private function getHighlights($item_id, $solr_highlighting) {
      foreach ($solr_highlighting as $id => $highlights) {
        if (strpos($id, $item_id) !== false) {
          return $highlights;
        }
      }
    }

    private function getExcerpt(array $highlighting) {
      $highlights = [];
      foreach ($highlighting as $highlight) {
        foreach ($highlight as $subitem) {
          $highlights[] = $subitem;
        }
      }
      $excerpt = implode('... ', $highlights);
      $excerpt = str_replace('[HIGHLIGHT]', '<mark>', $excerpt);
      $excerpt = str_replace('[/HIGHLIGHT]', '</mark>', $excerpt);

      return $excerpt;
    }
}