Apollo DataSource for Firestore

This is a Firestore DataSource for Apollo GraphQL Servers. It was adapted from the CosmosDB DataSource

Install

npm i apollo-datasource-firestore

Usage

Use by creating a new class extending the FirestoreDataSource, with the desired document type. Use separate DataSources for each data type, and preferably different collections in Firestore too. Initialise the class by passing a collection reference created by the Firestore library.

import { FirestoreDataSource } from 'apollo-datasource-firestore';

export interface UserDoc {
  /**
   * A string id value is required for entities using this library.
   * It will be used for the firestore document ID but not stored in the document in firestore.
   */
  readonly id: string
  readonly collection: 'users'
  name: string
  groupId: number
}

export interface PostsDoc {
  readonly id: string
  readonly collection: 'posts'
  title: string
}

export class UsersDataSource extends FirestoreDataSource<UsersDoc, ApolloContext> {}
export class PostsDataSource extends FirestoreDataSource<PostsDoc, ApolloContext> {}

const usersCollection: CollectionReference<UsersDoc> = firestore.collection('users')
const postsCollection: CollectionReference<PostsDoc> = firestore.collection('posts')

const server = new ApolloServer({
  typeDefs,
  resolvers,
  dataSources: () => ({
    users: new UsersDataSource(usersCollection),
    posts: new PostsDataSource(postsCollection)
  })
})

Custom queries

FirestoreDataSource has a findByQuery method that accepts a function taking the collection as its only argument, which you can then create a query based on. Can be used in resolvers or to create wrappers.

Example of derived class with custom query methods:

export class UsersDataSource extends FirestoreDataSource<UsersDoc, ApolloContext> {
  async findManyByGroupId (groupId: number) {
    return this.findManyByQuery(c => c.where('groupId', '==', groupId).limit(2))
  }

  async findOneByEmail (email: string) {
    return this.findManyByQuery(c => c.where('email', '==', email).limit(1))?.[0] ?? undefined
  }
}

Write Operations

This DataSource has some built-in mutations that can be used to create, update and delete documents.

await context.dataSources.users.createOne(userDoc)

await context.dataSources.users.updateOne(userDoc)

await context.dataSources.users.updateOnePartial(userId, { name: 'Bob' })

await context.dataSources.users.deleteOne(userId)

Batching

Batching is provided on all id-based queries by DataLoader.

Caching

Caching is available on an opt-in basis by passing a ttl option on queries.

alexieyizhe / apollo-datasource-firestore