Skip to content

Latest commit

 

History

History
119 lines (84 loc) · 4.79 KB

README.md

File metadata and controls

119 lines (84 loc) · 4.79 KB
Raw

gg-java

Java wrapper for the smash.gg GraphQL API

Current status

This is only somewhat maintained at the moment. It should work, but it might not be completely up-to-date with the API in the future (and there are some library changes I'd want to make). I'm open to work on this a bit again if you want to use this. In that case please open an issue.

Download

This repo is published on Maven Central. You can download it here.

If you are using Maven, add the following to your dependencies in your pom.xml:

<dependency>
    <groupId>com.github.gpluscb</groupId>
    <artifactId>gg-java</artifactId>
    <version>1.0.2</version>
</dependency>

If you are using Gradle, add the following to your repositories in your build.gradle:

mavenCentral()

And this to your dependencies:

implementation group: 'com.github.gpluscb', name: 'gg-java', version: '1.0.2'

Usage

String testQuery = "query TournamentQuery{tournament(slug:\"evo2018\"){events{name standings(query:{page:1,perPage:3}){nodes{standing entrant{name}}}}}}";

GGClient client = GGClient.builder("your-token-here").build();

client.query(testQuery)
        .thenAccept(r -> {
            r.onError(e -> System.out.println("Oh no, an error!"));
            r.onSuccess(data -> System.out.println("Success! First event name: " + data.getTournament().getEvents().get(0).getName().getValue()));
        }).exceptionally(t -> {t.printStackTrace(); return null;});
client.shutdown();

The smash.gg API docs will be an useful resource, especially the GraphQL reference. For a tutorial on how to construct the GraphQL queries, visit https://graphql.org/learn/.

Use the GGClient#query overloads for queries, GGClient#mutation for mutations. If you only want to get the raw JsonObject of the response, you can skip the deserialization and use GGClient#request.

These methods return a CompletableFuture<GGResponse<QueryResponse>>, CompletableFuture<GGResponse<MutationResponse>> and CompletableFuture<JsonObject> respectively for handling stuff asynchronously.

For reference, here's the gson repo used for all the JSON stuff. The responses will have the format {"data": {...}}.

XResponse classes

The XResponse classes (not including GGResponse) provide getters for all the fields of the GraphQL types they are representing.

However, depending on your query, some of these fields are not requested by you and therefore not provided in the response. In these cases, the getter for the field will not return null, but rather an XResponse instance of which the isProvided() method returns false. All other getters will throw an IllegalStateException. The getter for the field will return null if it was provided in the response as null.

This system has the consequence that Scalars and Enums cannot be given directly in their native java representation. Instead they are wrapped into a XResponse class, of which the getValue() method will return the underlying java type or throw if it is not provided.

For deserialization of unions to work, you will need to include the __typename field.

Variables

String testQuery = "query TournamentQuery($slug:String){tournament(slug:$slug){events{name standings(query:{page:1,perPage:3}){nodes{standing entrant{name}}}}}}";
JsonObject testVariables = new JsonObject();
testVariables.addProperty("slug", "evo2018");

client.query(testQuery, testVariables)
    // ...
    ;

Rate limiting

The rate limiting is handled automatically and you should never run into one. If you still do, a message will be printed to System.err and the request will retry after a back-off.

You can still specify custom rate limits if you want to. For example if you want to adhere to a 80/40s bucket:

GGClient client = GGClient.builder("your-token-here").limiter(RateLimiter.bucketBuilder().tasksPerPeriod(80).period(40000L).build()).build();

I have done some testing with the rate limit system. It seems like, for the default rate limits, clients have to respect an 80/61s bucket instead of an 80/60s one as the docs would suggest, so that's the default.

Docs

The docs can be found on the github-pages site here.

TODO

  • More efficient way to add variables
  • Better Error/Exception handling
  • More/better documentation
  • Proper configurable logging / allow ratelimit handler
  • Generally better code structure (whatever that means)
  • Builder for GraphQL queries, later specific to the project for more tight syntax and less room for illegal queries