Skip to content

apollo-server-integrations/apollo-server-integration-aws-lambda

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

@as-integrations/aws-lambda

Getting started

Apollo Server runs as a part of your Lambda handler, processing GraphQL requests. This package allows you to easily integrate Apollo Server with AWS Lambda. This integration comes with built-in request handling functionality for ProxyV1, ProxyV2, and ALB events with extensible typing. You can also create your own integrations via a Custom Handler and submitted as a PR if others might find them valuable.

First, install Apollo Server, graphql-js, and the Lambda handler package:

npm install @apollo/server graphql @as-integrations/aws-lambda

Then, write the following to server.mjs. (By using the .mjs extension, Node treats the file as a module, allowing us to use ESM import syntax.)

import { ApolloServer } from '@apollo/server';
import {
  startServerAndCreateLambdaHandler,
  handlers,
} from '@as-integrations/aws-lambda';

// The GraphQL schema
const typeDefs = `#graphql
  type Query {
    hello: String!
  }
`;

// A map of functions which return data for the schema.
const resolvers = {
  Query: {
    hello: () => 'world',
  },
};

// Set up Apollo Server
const server = new ApolloServer({
  typeDefs,
  resolvers,
});

export default startServerAndCreateLambdaHandler(
  server,
  handlers.createAPIGatewayProxyEventV2RequestHandler(),
);

Context

As with all Apollo Server 4 integrations, the context resolution is done in the integration. For the Lambda integration, it will look like the following:

import { ApolloServer } from '@apollo/server';
import {
  startServerAndCreateLambdaHandler,
  handlers,
} from '@as-integrations/aws-lambda';

type ContextValue = {
  isAuthenticated: boolean;
};

// The GraphQL schema
const typeDefs = `#graphql
  type Query {
    hello: String!
    isAuthenticated: Boolean!
  }
`;

// Set up Apollo Server
const server = new ApolloServer<ContextValue>({
  typeDefs,
  resolvers: {
    Query: {
      hello: () => 'world',
      isAuthenticated: (root, args, context) => {
        // For context typing to be valid one of the following must be implemented
        //   1. `resolvers` defined inline in the server config (not particularly scalable, but works)
        //   2. Add the type in the resolver function. ex. `(root, args, context: ContextValue)`
        //   3. Propagate the type from an outside definition like GraphQL Codegen
        return context.isAuthenticated;
      },
    },
  },
});

export default startServerAndCreateLambdaHandler(
  server,
  handlers.createAPIGatewayProxyEventV2RequestHandler(),
  {
    context: async ({ event }) => {
      // Do some parsing on the event (parse JWT, cookie, auth header, etc.)
      return {
        isAuthenticated: true,
      };
    },
  },
);

Middleware

For mutating the event before passing off to @apollo/server or mutating the result right before returning, middleware can be utilized.

Note, this middleware is strictly for event and result mutations and should not be used for any GraphQL modification. For that, plugins from @apollo/server would be much better suited.

For example, if you need to set cookie headers with a V2 Proxy Result, see the following code example:

import {
  startServerAndCreateLambdaHandler,
  handlers,
} from '@as-integrations/aws-lambda';
import type { APIGatewayProxyEventV2 } from 'aws-lambda';
import { server } from './server';

async function regenerateCookie(event: APIGatewayProxyEventV2) {
  // ...
  return 'NEW_COOKIE';
}

export default startServerAndCreateLambdaHandler(
  server,
  handlers.createAPIGatewayProxyEventV2RequestHandler(),
  {
    middleware: [
      // Both event and result are intended to be mutable
      async (event) => {
        const cookie = await regenerateCookie(event);
        return (result) => {
          result.cookies.push(cookie);
        };
      },
    ],
  },
);

Middleware Typing

If you want to define strictly typed middleware outside of the middleware array, the easiest way would be to extract your request handler into a variable and utilize the typeof keyword from Typescript. You could also manually use the RequestHandler type and fill in the event and result values yourself.

import {
  startServerAndCreateLambdaHandler,
  middleware,
  handlers,
} from '@as-integrations/aws-lambda';
import type {
  APIGatewayProxyEventV2,
  APIGatewayProxyStructuredResultV2,
} from 'aws-lambda';
import { server } from './server';

const requestHandler = handlers.createAPIGatewayProxyEventV2RequestHandler();

// Utilizing typeof
const cookieMiddleware: middleware.MiddlewareFn<typeof requestHandler> = (
  event,
) => {
  // ...
  return (result) => {
    // ...
  };
};

// Manual event filling
const otherMiddleware: middleware.MiddlewareFn<
  RequestHandler<APIGatewayProxyEventV2, APIGatewayProxyStructuredResultV2>
> = (event) => {
  // ...
  return (result) => {
    // ...
  };
};

export default startServerAndCreateLambdaHandler(server, requestHandler, {
  middleware: [
    // cookieMiddleware will always work here as its signature is
    // tied to the `requestHandler` above
    cookieMiddleware,

    // otherMiddleware will error if the event and result types do
    // not sufficiently overlap, meaning it is your responsibility
    // to keep the event types in sync, but the compiler may help
    otherMiddleware,
  ],
});

Middleware Short Circuit

In some situations, a middleware function might require the execution end before reaching Apollo Server. This might be a global auth guard or session token lookup.

To achieve this, the request middleware function accepts ResultType or Promise<ResultType> as a return type. Should middleware resolve to such a value, that result is returned and no further execution occurs.

import {
  startServerAndCreateLambdaHandler,
  middleware,
  handlers,
} from '@as-integrations/aws-lambda';
import type {
  APIGatewayProxyEventV2,
  APIGatewayProxyStructuredResultV2,
} from 'aws-lambda';
import { server } from './server';

const requestHandler = handlers.createAPIGatewayProxyEventV2RequestHandler();

// Utilizing typeof
const sessionMiddleware: middleware.MiddlewareFn<typeof requestHandler> = (
  event,
) => {
  // ... check session
  if (!event.headers['X-Session-Key']) {
    // If header doesn't exist, return early
    return {
      statusCode: 401
      body: 'Unauthorized'
    }
  }
};

export default startServerAndCreateLambdaHandler(server, requestHandler, {
  middleware: [
    sessionMiddleware,
  ],
});

Event Extensions

Each of the provided request handler factories has a generic for you to pass a manually extended event type if you have custom authorizers, or if the event type you need has a generic you must pass yourself. For example, here is a request that allows access to the lambda authorizer:

import {
  startServerAndCreateLambdaHandler,
  middleware,
  handlers,
} from '@as-integrations/aws-lambda';
import type { APIGatewayProxyEventV2WithLambdaAuthorizer } from 'aws-lambda';
import { server } from './server';

export default startServerAndCreateLambdaHandler(
  server,
  handlers.createAPIGatewayProxyEventV2RequestHandler<
    APIGatewayProxyEventV2WithLambdaAuthorizer<{
      myAuthorizerContext: string;
    }>
  >(), // This event will also be applied to the MiddlewareFn type
);

Custom Request Handlers

When invoking a lambda manually, or when using an event source we don't currently support (feel free to create a PR), a custom request handler might be necessary. A request handler is created using the handlers.createHandler function which takes two function arguments eventParser and resultGenerator, and two type arguments EventType and ResultType.

eventParser Argument

There are two type signatures available for parsing events:

Method A: Helper Object

This helper object has 4 properties that will complete a full parsing chain, and abstracts some of the work required to coerce the incoming event into a HTTPGraphQLRequest. This is the recommended way of parsing events.

parseHttpMethod(event: EventType): string

Returns the HTTP verb from the request.

Example return value: GET

parseQueryParams(event: EventType): string

Returns the raw query param string from the request. If the request comes in as a pre-mapped type, you may need to use URLSearchParams to re-stringify it.

Example return value: foo=1&bar=2

parseHeaders(event: EventType): HeaderMap

Import from here: import {HeaderMap} from "@apollo/server";

Return an Apollo Server header map from the event. HeaderMap automatically normalizes casing for you.

parseBody(event: EventType, headers: HeaderMap): string

Return a plaintext body. Be sure to parse out any base64 or charset encoding. Headers are provided here for convenience as some body parsing might be dependent on content-type

Method B: Parser Function

If the helper object is too restrictive for your use-case, the other option is to create a function with (event: EventType): HTTPGraphQLRequest as the signature. Here you can do any parsing and it is your responsibility to create a valid HTTPGraphQLRequest.

resultGenerator Argument

There are two possible result types, success and error, and they are to be defined as function properties on an object. Middleware will always run, regardless if the generated result was from a success or error. The properties have the following signatures:

success(response: HTTPGraphQLResponse): ResultType

Given a complete response, generate the desired result type.

error(e: unknown): ResultType

Given an unknown type error, generate a result. If you want to create a basic parser that captures everything, utilize the instanceof type guard from Typescript.

error(e) {
  if(e instanceof Error) {
    return {
      ...
    }
  }
  // If error cannot be determined, panic and use lambda's default error handler
  // Might be advantageous to add extra logging here so unexpected errors can be properly handled later
  throw e;
}

Custom Handler Example

import {
  startServerAndCreateLambdaHandler,
  handlers,
} from '@as-integrations/aws-lambda';
import type { APIGatewayProxyEventV2 } from 'aws-lambda';
import { HeaderMap } from '@apollo/server';
import { server } from './server';

type CustomInvokeEvent = {
  httpMethod: string;
  queryParams: string;
  headers: Record<string, string>;
  body: string;
};

type CustomInvokeResult =
  | {
      success: true;
      body: string;
    }
  | {
      success: false;
      error: string;
    };

const requestHandler = handlers.createRequestHandler<
  CustomInvokeEvent,
  CustomInvokeResult
>(
  {
    parseHttpMethod(event) {
      return event.httpMethod;
    },
    parseHeaders(event) {
      const headerMap = new HeaderMap();
      for (const [key, value] of Object.entries(event.headers)) {
        headerMap.set(key, value);
      }
      return headerMap;
    },
    parseQueryParams(event) {
      return event.queryParams;
    },
    parseBody(event) {
      return event.body;
    },
  },
  {
    success({ body }) {
      return {
        success: true,
        body: body.string,
      };
    },
    error(e) {
      if (e instanceof Error) {
        return {
          success: false,
          error: e.toString(),
        };
      }
      console.error('Unknown error type encountered!', e);
      throw e;
    },
  },
);

export default startServerAndCreateLambdaHandler(server, requestHandler);