How to Integrate IBM Watson with Drupal

Prerequisites

• Ensure you have a functioning Drupal installation and administrative access.

 

• Sign up for IBM Cloud and create an IBM Watson service instance (e.g., Watson Assistant, Watson Language Translator).

 

• Install the necessary modules in Drupal for integration, such as the JSON:API module for web service interaction.

Install Required Modules

• Navigate to the Drupal admin interface and go to Extend

 

• Search for the JSON:API module and enable it.

 

• Use Drush or Composer to manage modules if preferred. You can use a command like:

composer require drupal/jsonapi

Obtain IBM Watson Credentials

• In your IBM Cloud Dashboard, locate your Watson service instance and retrieve the API Key and Endpoint URL.

 

• Store these credentials securely, as they'll be needed for API requests.

Configure Drupal for IBM Watson Integration

• Create a custom module in Drupal if one doesn't exist. Navigate to the modules/custom directory and create a folder named ibm_watson_integration.

 

• Create an .info.yml file inside your module directory:

name: 'IBM Watson Integration'
type: module
description: 'Custom module to integrate IBM Watson services.'
package: Custom
core_version_requirement: ^8 || ^9
dependencies:
  - drupal:jsonapi

• Create a .module file in the same directory to write custom functions and interaction logic.

Create a Service for Watson API Interaction

• Inside the src directory of your custom module, create a directory named Service and add a PHP file WatsonClient.php.

namespace Drupal\ibm_watson_integration\Service;

use GuzzleHttp\ClientInterface;
use Drupal\Component\Serialization\Json;

class WatsonClient {
  protected $httpClient;
  protected $apiKey;
  protected $endpoint;

  public function __construct(ClientInterface $http_client) {
    $this->httpClient = $http_client;
    $this->apiKey = 'YOUR_WATSON_API_KEY';
    $this->endpoint = 'YOUR_WATSON_ENDPOINT_URL';
  }

  public function queryWatson($query) {
    $response = $this->httpClient->request('POST', $this->endpoint, [
      'headers' => [
        'Content-Type' => 'application/json',
        'Authorization' => 'Bearer ' . $this->apiKey,
      ],
      'body' => Json::encode(['text' => $query]),
    ]);

    return Json::decode($response->getBody());
  }
}

Register the Service

• Create a services.yml file in your module directory and register the WatsonClient service.

services:
  ibm_watson_integration.watson_client:
    class: 'Drupal\ibm_watson_integration\Service\WatsonClient'
    arguments: ['@http_client']

Use the Service in a Custom Block

• Create a Plugin/Block directory within your module’s src directory and add a PHP file WatsonBlock.php.

namespace Drupal\ibm_watson_integration\Plugin\Block;

use Drupal\Core\Block\BlockBase;
use Drupal\Core\Form\FormStateInterface;
use Drupal\ibm_watson_integration\Service\WatsonClient;
use Symfony\Component\DependencyInjection\ContainerInterface;
use Drupal\Core\Plugin\ContainerFactoryPluginInterface;

/**
 * Provides a 'Watson Query Block' Block.
 *
 * @Block(
 *   id = "watson_query_block",
 *   admin_label = @Translation("Watson Query Block"),
 * )
 */
class WatsonBlock extends BlockBase implements ContainerFactoryPluginInterface {
  protected $watsonClient;

  public function __construct(array $configuration, $plugin_id, $plugin_definition, WatsonClient $watson_client) {
    parent::__construct($configuration, $plugin_id, $plugin_definition);
    $this->watsonClient = $watson_client;
  }

  public static function create(ContainerInterface $container, array $configuration, $plugin_id, $plugin_definition) {
    return new static(
      $configuration,
      $plugin_id,
      $plugin_definition,
      $container->get('ibm_watson_integration.watson_client')
    );
  }

  public function build() {
    $result = $this->watsonClient->queryWatson('Hello World!');
    return [
      '#markup' => $this->t('Response from IBM Watson: @response', ['@response' => $result['response']]),
    ];
  }
}

Enable the Module and Clear Cache

• Enable your custom module via the command-line or the Drupal admin UI.

 

• Clear the Drupal cache to ensure all configurations are reloaded.

drush en ibm_watson_integration
drush cache:rebuild

Test the Integration

• Add the custom block (Watson Query Block) to a page through the block layout UI.

 

• Verify that the block displays the correct response from IBM Watson to ensure successful integration.

By following these steps, you achieve a successful integration of IBM Watson services with Drupal, allowing for enhanced AI capabilities directly within your content management system.