Setting Up Spring for GraphQL (spring-graphql)

Setting up Spring for GraphQL is three pieces: dependency, schema files, and resolver beans. The official Spring project is simpler to integrate than older libraries and provides native testing support.

Dependency and auto-configuration:

• Add spring-boot-starter-graphql and a web starter to pom.xml; Boot auto-configures the GraphQL engine and a /graphql HTTP endpoint with zero additional wiring or factory beans
• Drop schema files into src/main/resources/graphql/ with a .graphqls extension; Boot discovers and merges them automatically, so you can split schemas by domain — book.graphqls, author.graphqls — and they compose into one schema

Resolver conventions:

• Resolvers are plain @Controller beans: @QueryMapping handles reads, @MutationMapping handles writes, @SchemaMapping resolves nested fields like Book.author
• Method names match schema field names by convention, so there's almost no boilerplate; use @SchemaMapping(typeName, field) explicitly only when names diverge

What to wire from day one:

• Enable GraphiQL in dev profiles so the team can explore the schema before any client code exists — it's a built-in IDE for your API
• Wire DataLoader for any @SchemaMapping that resolves to a related entity immediately, not after the N+1 problem surfaces in production; retrofitting DataLoader into existing resolvers is significantly harder

Why Spring for GraphQL over alternatives:

• Integrates natively with @GraphQlTest slice tests, Spring Security filters, and Actuator; older libraries like DGS work but lack this cohesion with the Spring ecosystem

I recommend enabling GraphiQL in dev and setting up DataLoader for nested fields on the first day of development — both are much easier to start with than to add retroactively.