pokeapi-reactor

A non-blocking, reactive API client for PokeAPI with caching for Spring Boot projects.

Features

• Simple, single entry point for all client operations.
• Non-blocking HTTP operations and non-blocking caching.
• Fully customizable caching. Supports Spring Boot's generic CacheManager for caching.
• Caching and non-caching configurations.

Getting started

Project Configuration

Properties

You can (must) configure the location of the PokeAPI instance you want to use. Add the following property to your application.properties:

skaro.pokeapi.base-uri=https://pokeapi.co/api/v2/ #or the url of your own instance

You may also configure the max buffer size for the WebClient, which is used to fetch resources from PokeAPI. Its default value is 565000 bytes, while the API request for "/pokemon/mew" can grow over the time, you may want to increase it yourself to a higher value. To achieve this, add the following property to your application.properties:

skaro.pokeapi.max-buffer-size=565000

Application Context

Import one of pokeapi-reactor's configurations as well as specify your own reactor.netty.http.client.HttpClient bean. Two configurations are available: caching and non-caching. Below is an example of a caching configuration which uses a flexible HttpClient tuned for high parallel throughput.

@Configuration
@Import(PokeApiReactorCachingConfiguration.class)
@EnableCaching
public class MyPokeApiReactorCachingConfiguration {
	@Bean
	public ConnectionProvider connectionProvider() {
	    return ConnectionProvider.builder("Auto refresh & no connection limit")
		    .maxIdleTime(Duration.ofSeconds(10))
		    .maxConnections(500)
		    .pendingAcquireMaxCount(-1)
		    .build();
	}

	@Bean
	public HttpClient httpClient(ConnectionProvider connectionProvider) {
		return HttpClient.create(connectionProvider)
                  .compress(true)
                  .resolver(DefaultAddressResolverGroup.INSTANCE);
	}
}

Or, if you'd rather not enable caching:

@Configuration
@Import(PokeApiReactorNonCachingConfiguration.class)
public class MyPokeApiReactorNonCachingConfiguration {
	@Bean
	public ConnectionProvider connectionProvider() { ... }
	
	@Bean
	public HttpClient httpClient(ConnectionProvider connectionProvider) { ... }
}

Both the PokeApiReactorCachingConfiguration and PokeApiReactorNonCachingConfiguration will register the appropriate PokeApiClient bean.

Fetching a resource

Inject the registered PokeApiClient into your class and request a resource.

@Autowired
private PokeApiClient pokeApiClient;

...

public void printPokemon() {
   pokeApiClient.getResource(Pokemon.class, "pikachu")
      .map(pokemon -> String.format("%s has %d forms", pokemon.getName(), pokemon.getForms().size()))
      .subscribe(System.out::println);
}

If you don't mind blocking, you can simply block for the resource.

public void printPokemon() {
   Pokemon pokemon = pokeApiClient.getResource(Pokemon.class, "pikachu").block();
   String pokemonInfo = String.format("%s has %d forms", pokemon.getName(), pokemon.getForms().size()));
   System.out.println(pokemonInfo);
}

Following a resource

Following links to other resources also has first-class support.

public void printPokemonForms() {
   pokeApiClient.getResource(Pokemon.class, "pikachu")
      .flatMapMany(pokemon -> pokeApiClient.followResources(pokemon::getForms, PokemonForm.class))
      .map(form -> String.format("Pikachu has a form called %s", form.getName()))
      .subscribe(System.out::println);
}

Again, with blocking:

public void printPokemonForms() {
   Pokemon pokemon = pokeApiClient.getResource(Pokemon.class, "pikachu").block();
   List<PokemonForm> forms = pokeApiClient.followResources(pokemon::getForms, PokemonForm.class)
      .collectList()
      .block();

   for(PokemonForm form : forms) {
      String formInfo = String.format("Pikachu has a form called %s", form.getName());
      System.out.println(formInfo);
   }
}

Maven Configuration

The latest release can be pulled from GitHub's Apache Maven repository. To pull from their repository, you must add your GitHub credentials to your settings.xml (located in ${user.home}/.m2/settings.xml). You can read GitHub's article about how to do that. But for impatient programmer's, below is yet another guide to configure Maven to pull from GitHub that holds your hand a little more closely - I always appreciate it when people do that.

Generate a token

Generate a PAT (Personal Access Token) with the read:packages privilege.

Add <server> and <repository> tags in settings.xml

Configure Maven to authenticate when trying to pull from GitHub's Apache Maven repository. If you don't have a settings.xml, create it under your .m2 directory. It should look something like this.

<settings xmlns="http://maven.apache.org/SETTINGS/1.0.0"
  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
  xsi:schemaLocation="http://maven.apache.org/SETTINGS/1.0.0
                      http://maven.apache.org/xsd/settings-1.0.0.xsd">

  <activeProfiles>
    <activeProfile>github</activeProfile>
  </activeProfiles>

  <profiles>
    <profile>
      <id>github</id>
      <repositories>
        <repository>
          <id>central</id>
          <url>https://repo1.maven.org/maven2</url>
        </repository>
        <repository>
          <id>github-skaro-pokeapi-reactor</id>
          <url>https://maven.pkg.github.com/SirSkaro/pokeapi-reactor</url>
        </repository>
      </repositories>
    </profile>
  </profiles>

  <servers>
    <server>
      <id>github-skaro-pokeapi-reactor</id>
      <username>YOUR USERNAME</username>
      <password>YOUR TOKEN YOU JUST CREATED</password>
    </server>
  </servers>
</settings>