GraphQL Java

Table of Contents

• GraphQL Best practices
  • DDOS
  • Playground
• GraphQL Java
  • GraphQLResolver vs GraphQLQueryResolver<DataClassName>
  • Exception Handling
  • DataFetcherResult
  • Mutation
• DataFetchingEnviroment
• Scalars
  • Steps to use custom Scalar
• GraphQLServletListener
• Pagination
• Custom Context
• N+1 Problem
• Instrumentation and Tracing
  • Tracing
• Data Loader

GraphQL Best practices

DDOS

1. Attack

2. Solution

Set max-query-depth in application.yml

graphql:
    max-query-depth: 3

Playground

DO NOT ENABLE Playground for Prod.

GraphQL Java

GraphQLResolver vs GraphQLQueryResolver<DataClassName>

Warning

I still can’t grasp the difference and there is not official doc for these Classes!

Exception Handling

Default GraphQL Exception Handler

DefaultGraphQLErrorHandler.processErrors is called for any Runtime.class or GraphQLException.class is called (thrown by ThrowableGraphQLError)

This project has common GraphQLExceptionHandler with 2 methods

1. ThrowableGraphQLError handle(GraphQLException graphQLException)

   • For Client side errors (Giving MORE info about Client error)

2. ThrowableGraphQLError handle(RuntimeException graphQLException)

   • For Server side errors (Giving NO info about Server error)

Custom GraphQL ErrorHandler

Can extend GraphQLErrorHandler for custom exception handling.

• Defined in exception.CustomGraphQLErrorHandler

• CustomGraphQLErrorHandler.processErrors is called instead of DefaultGraphQLErrorHandler.processErrors

DataFetcherResult

Contains both data, errors and local context to be sent as final result to graphql server.

Mutation

Write operations

Upload file

Use DataFetchingEnvironment and DefaultGraphQLServletContext from the GraphQL Servers

curl --location --request POST 'http://localhost:8080/graphql' \
--form 'operations="{\"query\":\"mutation {uploadFile}\" , \"variables\":{}}"' \
--form 'file=@"/D:/Coding/learngraphql-java/README.adoc"' // Remem  ber this is of type File. CURL import on postman will fail this command

// "{"query":"mutation {uploadFile}" , "variables":{}}"

DataFetchingEnviroment

Documentation

Powerful object with useful methods for GraphQL interceptors

1. getSelectionSet Useful for DB layer

2. getContext: Useful for Auth etc

Scalars

GraphQL spec includes default scalar types Int, Float, String, Boolean, and ID. Some applications need to support other atomic data types (such as Date) or add validation to an existing type(NonNegativeInteger)

To enable this

1. Define custom scalar types. Apollo Doc

2. Use library. Java GraphQL Scalar

Steps to use custom Scalar

1. Add scalar <ScalarName> in .graphqls or any GrpahQL file.

2. For Java, create a Bean. Example

   @Bean
   public GraphQLScalarType nonNegativeInt() {
       return ExtendedScalars.NonNegativeInt;
   }

   • TODO: Typescript

3.

GraphQLServletListener

Interfaces for request states.

1. onRequest

2. RequestCallback

   • onRequest

   • onSuccess

   • onFinally

Pagination

GraphQL docs

// GET Bank Accounts Pagination
query GET_BANK_ACCOUNTS{
    bankAccounts(first: 5, after: "<uuid>") {
        edges {
            cursor // uuid
            node {
                id
            }
        }
        pageInfo {
            hasPreviousPage
            hasNextPage
            startCursor
            endCursor
        }
    }
}

Response:

{
  "data": {
    "bankAccounts": {
      "edges": [
        {
          "cursor": "NDEwZjU5MTktZTUwYi00NzkwLWFhZTMtNjVkMmQ0YjIxYzc3",
          "node": {
            "id": "410f5919-e50b-4790-aae3-65d2d4b21c77"
          }
        },
        {
          "cursor": "NDhlNGE0ODQtYWYyYy00MzY2LThjZDQtMjUzMzA1OTc0NzNm",
          "node": {
            "id": "48e4a484-af2c-4366-8cd4-25330597473f"
          }
        }
      ],
      "pageInfo": {
        "hasPreviousPage": true,
        "hasNextPage": false,
        "startCursor": "NDEwZjU5MTktZTUwYi00NzkwLWFhZTMtNjVkMmQ0YjIxYzc3",
        "endCursor": "NDhlNGE0ODQtYWYyYy00MzY2LThjZDQtMjUzMzA1OTc0NzNm"
      }
    }
  }
}

Usually pagination is offset-based. ID-based based pagination uses unique identifier which is more reliable than offset-based.

Custom Context

We have custom context so that we can intercept network request and get info such as header from HttpServletRequest from GraphQLServletContext

We used GraphQLServletContext in junction with GraphQLServletContextBuilder to get Context info.

Note	user_id is the header name given in HTTP Headers in GraphQL playground.

N+1 Problem

Instrumentation and Tracing

Ability to GraphQL query and when it is executing in server, provide instrument data (log data; tracing) during execution of query.

Tracing

application.properties

• graphql.servolet.tracing-enabled=true Get tracing info on RH corner

Data Loader

Very important.

Warning

Please read this Github Java Dataloader doc