This is a Firestore DataSource for Apollo GraphQL Servers. It was adapted from the CosmosDB DataSource
Warning
Version 6.x and up supports apollo server v4, and, through the nature of apollo server v4 it works well with other graphql servers or even standalone.
If you need apollo server v3 support, use version 5.x of this package.
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.
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'
// the createdAt and updatedAt timestamps stored by firestore natively are
// available as properties as well
readonly createdAt: Timestamp
readonly updatedAt: Timestamp
name: string
groupId: number
}
export interface PostsDoc {
readonly id: string
readonly collection: 'posts'
// the createdAt and updatedAt timestamps stored by firestore natively are
// available as properties as well
readonly createdAt: Timestamp
readonly updatedAt: Timestamp
title: string
}
export class UserDataSource extends FirestoreDataStore<UserDoc> {}
export class PostsDataSource extends FirestoreDataStore<PostsDoc> {}
const usersCollection: CollectionReference<UserDoc> = firestore.collection('users')
const postsCollection: CollectionReference<PostsDoc> = firestore.collection('posts')
const server = new ApolloServer({
typeDefs,
resolvers
})
const { url } = await startStandaloneServer(server, {
async context () {
const { cache } = server
return {
// We create new instances of our data sources with each request.
// We can optionally pass in our server's cache, or logger
dataSources: {
users: new UserDataSource(usersCollection, { cache }),
posts: new PostsDataSource(postsCollection, { cache })
}
}
}
})
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 UserDataSource extends FirestoreDataStore<UserDoc> {
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
}
}
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 is provided on all id-based queries by DataLoader.
Caching is available on an opt-in basis by passing a ttl
option on queries.