rasa-java-client-library

A Java SDK for Rasa Open Source

The rasa-java-client-library wraps the Rasa OpenSource HTTP API and enables you to build your Rasa assistant application with Java.

Use the comfortable RasaClient to connect to the Rasa Server and talk to or train your Rasa assistant
Use the flexible ModelApi, DomainApi and TrackerApi to implement additional advanced features, e.g. interactive learning

Version

current version: 0.2.0 (unstable)

Requirements

java 11 and higher

Installation

Attention!

This is a GitHub package. You need to authenticate to GitHub Packages first.
Currently available version 0.2.0 is unstable!

Maven

<dependency>
  <groupId>io.github.arturkorb</groupId>
  <artifactId>rasa-java-client-library</artifactId>
  <version>0.2.0</version>
</dependency>

Gradle

dependencies {
  implementation 'io.github.arturkorb:rasa-java-client-library:0.2.0'
}

Compatibility with Rasa

rasa-java-client-library version	compatible Rasa version
`0.2.0`	`>=2.2.0`

Use Case

Use the rasa-java-client-library if you consider building your Rasa assistant application with Java. A typical architecture of such a Rasa assistant application could look like this:

Rasa assistant application architecture

The picture shows in which components the rasa-java-client-library can be used.

Used in the communication component for talking with the bot and retrieving context (NLU parse results, etc.)
Used in the maintenance/development component for training and configuration
Used in Action Service for building response and event objects when executing action

Usage

Examples:

        /*RasaClient usage example*/
        RasaClient rasaClient = new RasaClient().withBasePath("http://<rasa server url>:5005");
        
        //build your request from your training data
        YAMLTrainingRequest yamlTrainingRequest = new YAMLTrainingRequest();
        yamlTrainingRequest
                .config("<content of 'config.yml' as String>")
                .domain("<content of 'domain.yml' as String>")
                .nlu("<content of 'nlu.yml' as String>")
                .stories("<content of 'stories.yml' as String>")
                .rules("<content of 'rules.yml' as String>");
        
        //training
        rasaClient.trainModel(yamlTrainingRequest);
                
        //talking
        BotAnswer botAnswer = rasaClient.sendMessage("hi", "<user id>");
                
        //context
        Context context = rasaClient.sendMessageWithContextRetrieval("hi", "<user id");
        ParseResult result = context.getParseResult();
        Intent intent = result.getIntent();        

        /*ServerInformationApi usage example*/
        ServerInformationApi serverInformationApi = new ServerInformationApi();
        InlineResponse200 version = serverInformationApi.getVersion();

        /*DomainApi usage example*/
        DomainApi domainApi = new DomainApi();
        Domain domain = domainApi.getDomain();
        DomainConfig config = domain.getConfig();

        /*TrackerApi usage example*/
        TrackerApi trackerApi = new TrackerApi();
        Tracker conversationTracker = trackerApi.getConversationTracker("<conversation_ID>", "ALL", null);

        /*ModelApi usage example*/
        ModelApi modelApi = new ModelApi();
        ModelRequest modelRequest = new ModelRequest();
        modelRequest = modelRequest.modelFile("/app/models/" + "<your model file name>");
        modelApi.replaceModel(modelRequest);