JarCastingNLP Client Spring Boot Starter
NLP Client project provides integration with various nlp clients (only dialog-flow for now). It is based on Bot Framework. It uses IntentContainer from Bot Framework. To read more about it, check the link below:
https://github.com/botscrew-projects/bot-framework-core
Getting Started
- Add
NLP Clientdependency
<dependency>
<groupId>com.botscrew</groupId>
<artifactId>bot-framework-nlp-spring-boot-starter</artifactId>
<version>1.1.2</version>
</dependency>
NLP Client already depends on the Bot Framework so you don't need to add it to project by yourself.
Usage
To send query to the nlp provider you need to autowire NlpClient and call NlpClient#query() method
public class TextHandler {
@Autowired
private NlpClient nlpClient;
@Text
public void handleText(User user, @Text String text) {
nlpClient.query(user, text);
}
}
You will be able to get an answer in your Intent handling method:
@Intent
public void handleIntent(User user, @Text String originalQuery) {
// ...
}
You can get entities from the nlp provider as parameters in your handling method:
@Intent("name")
public void handleNameIntent(User user, @Param("name") String name) {
// ...
}
If you need to get any complex objects you can define your own model and get them as parameters:
public class Age {
private String unit;
private Integer amount;
}
// ...
@Intent("age")
public void handleAgeIntent(@Param("age") Age age) {}
IntentContainer supports user states so, you will be able to define different intent handlers for different user states:
@Intent(value = "PHONE_NUMBER", states = {"DEFAULT"})
public void handlePhoneNumberIntent(User user, @Param("number") String number) {
// ...
}
@Intent(value = "PHONE_NUMBER", states = {"ONBOARDING"})
public void handlePhoneNumberInOnboarding(User user, @Param("number") String number) {
// ...
}
Providers
- DialogFlow APIv1
To work with DialogFlow API you need define the following properties: nlp.provider.dialog-flow.v1.client-token
If you need get some specific data from the original DialogFlow response, you can get it by using the @NlpResponse annotation:
@Intent
public void intent(@NlpResponse AiResponse aiResponse) {
// ...
}