Features

  • Strong type safety for Firestore - Automatically provide type information to nested documents without unsafe type assertions, from the simple schema. Also support data decoding.
  • Security rules generation - Generate firestore.rules file including data type validation and access control from the schema.
  • React Hooks - Get realtime updates with React Hooks.
  • Type safety for Cloud Functions
    • Automatically provide type information to snapshot data on Firestore Trigger Function based on the path string.
    • Guard HTTPS callable function's request/response data type both on compile time and runtime.


Requirement

  • TypeScript (>= 4.2)


Install

yarn add fireschema
yarn add -D typescript ts-node


Setup

Custom Transformer

To generate Firestore security rules or embedding type validation code into Callable Function, you need to compile codes via Fireschema's custom transformer that retrives type information from TypeScript AST.

Currently official TypeScript package doesn't support custom transformers, and you need to use ttypescript that wraps TypeScript compiler.

Add following options to your tsconfig.json,

{
  "compilerOptions": {
    "plugins": [
      {
        "transform": "fireschema/transformer"
      }
    ]
  }
}

and replace the commands.

before after
typescript tsc ttsc (ttypescript's compile command)
ts-node ts-node ts-node --compiler ttypescript

ttsc and ts-node supports specifying tsconfig.json by using environment variable TS_NODE_PROJECT.

If you use ts-jest, add following options to jest config.

module.exports = {
  globals: {
    'ts-jest': {
      tsconfig: 'tsconfig.json',
      compiler: 'ttypescript',
    },
  },
}




Example - Firestore

Do not use following variable names except importing from fireschema.

  • $collectionSchema
  • __$__

Data structure of examples

  • users/{uid} - User
  • users/{uid}/posts/{postId} - PostA | PostB


1. Define schema

The schema definition must be named exported as firestoreSchema.

import { Merge } from 'type-fest'
import {
  $allow,
  $collectionGroups,
  $collectionSchema,
  $docLabel,
  $functions,
  $or,
  $schema,
  createFirestoreSchema,
  FTypes,
} from 'fireschema'

// user
export type User = {
  name: string
  displayName: string | null
  age: number
  timestamp: FTypes.Timestamp
  options: { a: boolean } | undefined
}
export type UserDecoded = Merge<User, { timestamp: Date }>

const UserSchema = $collectionSchema<User, UserDecoded>()({
  decoder: (data) => ({
    ...data,
    timestamp: data.timestamp.toDate(),
  }),
})

// post
type PostA = {
  type: 'a'
  tags: { id: number; name: string }[]
  text: string
}
type PostB = {
  type: 'b'
  tags: { id: number; name: string }[]
  texts: string[]
}
const PostSchema = $collectionSchema<PostA | PostB>()({
  selectors: (q) => ({
    byTag: (tag: string) => q.where('tags', 'array-contains', tag),
  }),
})

export const firestoreSchema = createFirestoreSchema({
  [$functions]: {
    // whether /admins/<uid> exists
    ['isAdmin()']: `
      return exists(/databases/$(database)/documents/admins/$(request.auth.uid));
    `,

    // whether uid matches
    ['matchesUser(uid)']: `
      return request.auth.uid == uid;
    `,
  },

  [$collectionGroups]: {
    posts: {
      [$docLabel]: 'postId',
      [$schema]: PostSchema,
      [$allow]: {
        read: true,
      },
    },
  },

  // /users/{uid}
  users: {
    [$docLabel]: 'uid', // {uid}
    [$schema]: UserSchema, // collectionSchema
    [$allow]: {
      // access control
      read: true, // all user
      write: $or(['matchesUser(uid)', 'isAdmin()']), // only users matching {uid} or admins
    },

    // /users/{uid}/posts/{postId}
    posts: {
      [$docLabel]: 'postId',
      [$schema]: PostSchema,
      [$allow]: {
        read: true,
        write: 'matchesUser(uid)',
      },
    },
  },
})


2. Generate firestore.rules

yarn fireschema <path-to-schema>.ts

Environment variable TS_NODE_PROJECT is supported similarly to ttsc and ts-node.

Example of generated firestore.rules rules rules_version = '2'; service cloud.firestore { match /databases/{database}/documents { function isAdmin() { return exists(/databases/$(database)/documents/admins/$(request.auth.uid)); } function matchesUser(uid) { return request.auth.uid == uid; } match /{path=**}/posts/{postId} { allow read: if true; } match /users/{uid} { function __validator_0__(data) { return ( (!("_createdAt" in data) || data._createdAt == request.time) && (!("_updatedAt" in data) || data._updatedAt == request.time) && data.name is string && (data.displayName == null || data.displayName is string) && (data.age is int || data.age is float) && data.timestamp is timestamp && (!("options" in data) || data.options.a is bool) ); } allow read: if true; allow write: if ((matchesUser(uid) || isAdmin()) && __validator_0__(request.resource.data)); match /posts/{postId} { function __validator_1__(data) { return (( (!("_createdAt" in data) || data._createdAt == request.time) && (!("_updatedAt" in data) || data._updatedAt == request.time) && data.type == "a" && (data.tags.size() == 0 || ((data.tags[0].id is int || data.tags[0].id is float) && data.tags[0].name is string)) && data.text is string ) || ( (!("_createdAt" in data) || data._createdAt == request.time) && (!("_updatedAt" in data) || data._updatedAt == request.time) && data.type == "b" && (data.tags.size() == 0 || ((data.tags[0].id is int || data.tags[0].id is float) && data.tags[0].name is string)) && (data.texts.size() == 0 || data.texts[0] is string) )); } allow read: if true; allow write: if (matchesUser(uid) && __validator_1__(request.resource.data)); } } } }


3. Read/write collections and documents

The Firestore interface of Fireschema supports both Web SDK and Admin SDK.

import firebase from 'firebase/app' // or firebase-admin
import { TypedFirestore } from 'fireschema'
import { firestoreSchema } from './1-1-schema'

const app: firebase.app.App = firebase.initializeApp({
  // ...
})
const firestoreApp = app.firestore()
firestoreApp.settings({ ignoreUndefinedProperties: true })

/**
 * Initialize TypedFirestore
 */
export const typedFirestore: TypedFirestore<
  typeof firestoreSchema,
  firebase.firestore.Firestore
> = new TypedFirestore(firestoreSchema, firebase.firestore, firestoreApp)

/**
 * Reference collections/documents and get snapshot
 */
const users = typedFirestore.collection('users') // TypedCollectionRef instance
const user = users.doc('userId') // TypedDocumentRef instance

const posts = user.collection('posts')
const post = posts.doc('123')
const techPosts = user.collectionQuery(
  'posts',
  (q) => q.byTag('tech'), // selector defined in schema
)

!(async () => {
  await user.get() // DocumentSnapshot<User>

  await post.get() // DocumentSnapshot<PostA | PostB>
  await posts.get() // QuerySnapshot<PostA | PostB>
  await techPosts.get() // QuerySnapshot<PostA | PostB>
})

/**
 * Get child collection of retrived document snapshot
 */
!(async () => {
  const snap = await users.get()
  const firstUserRef = snap.docs[0]!.ref

  await typedFirestore.wrapDocument(firstUserRef).collection('posts').get()
})

/**
 * Reference parent collection/document
 */
const _posts = post.parentCollection()
const _user = posts.parentDocument()

/**
 * Reference collections groups and get snapshot
 */
const postsGroup = typedFirestore.collectionGroup(
  'posts', // collection name: passed to original collectionGroup method
  'users.posts', // to get schema options
)
const techPostsGroup = typedFirestore.collectionGroupQuery(
  'posts',
  'users.posts',
  (q) => q.byTag('tech'),
)

!(async () => {
  await postsGroup.get() // QuerySnapshot<PostA | PostB>
  await techPostsGroup.get() // QuerySnapshot<PostA | PostB>
})

/**
 * Write data
 */
!(async () => {
  await user.create({
    name: 'test',
    displayName: 'Test',
    age: 20,
    timestamp: typedFirestore.firestoreStatic.FieldValue.serverTimestamp(),
    options: { a: true },
  })
  await user.setMerge({
    age: 21,
  })
  await user.update({
    age: 21,
  })
  await user.delete()
})

/**
 * Transaction
 */
!(async () => {
  await typedFirestore.runTransaction(async (tt) => {
    const snap = await tt.get(user)
    tt.update(user, {
      age: snap.data()!.age + 1,
    })
  })
})


4. React Hooks

import React from 'react'
import { useTypedDocument, useTypedQuery } from 'fireschema/hooks'
import { typedFirestore } from './1-3-typed-firestore'

/**
 * Get realtime updates of collection/query
 */
export const UsersComponent = () => {
  const users = useTypedQuery(typedFirestore.collection('users'))
  if (!users.data) {
    return <span>{'Loading...'}</span>
  }

  return (
    <ul>
      {users.data.map((user, i) => (
        <li key={i}>{user.displayName}</li>
      ))}
    </ul>
  )
}

/**
 * Get realtime updates of document
 */
export const UserComponent = ({ id }: { id: string }) => {
  const user = useTypedDocument(typedFirestore.collection('users').doc(id))
  if (!user.data) {
    return <span>{'Loading...'}</span>
  }

  return <span>{user.data.displayName}</span>
}




Example - Cloud Functions

1. Create functions

import { firestore } from 'firebase-admin'
import * as functions from 'firebase-functions'
import { Merge } from 'type-fest'
import { $jsonSchema, TypedFunctions } from 'fireschema'
import { firestoreSchema, User } from './1-1-schema'

/**
 * Initialize TypedFunctions
 */
const timezone = 'Asia/Tokyo'
const typedFunctions = new TypedFunctions(
  firestoreSchema,
  firestore,
  functions,
  timezone,
)
const builder = functions.region('asia-northeast1')

/**
 * functions/index.ts file
 */
export type UserJson = Merge<User, { timestamp: string }>
export const callable = {
  createUser: typedFunctions.callable({
    schema: [
      $jsonSchema<UserJson>(), // schema of request data (automatically validate on request)
      $jsonSchema<{ result: boolean }>(), // schema of response data
    ],
    builder,
    handler: async (data, context) => {
      console.log(data) // UserJson

      return { result: true }
    },
  }),
}

export const firestoreTrigger = {
  onUserCreate: typedFunctions.firestoreTrigger.onCreate({
    builder,
    path: 'users/{uid}',
    handler: async (decodedData, snap, context) => {
      console.log(decodedData) // UserDecoded (provided based on path string)
      console.log(snap) // QueryDocumentSnapshot<User>
    },
  }),
}

export const http = {
  getKeys: typedFunctions.http({
    builder,
    handler: (req, resp) => {
      if (req.method !== 'POST') {
        resp.status(400).send()
        return
      }
      resp.json(Object.keys(req.body))
    },
  }),
}

export const topic = {
  publishMessage: typedFunctions.topic('publish_message', {
    schema: $jsonSchema<{ text: string }>(),
    builder,
    handler: async (data) => {
      data // { text: string }
    },
  }),
}

export const schedule = {
  cron: typedFunctions.schedule({
    builder,
    schedule: '0 0 * * *',
    handler: async (context) => {
      console.log(context.timestamp)
    },
  }),
}


2. Call HTTPS callable function

Automatically provide types to request/response data based on passed functions module type.

import firebase from 'firebase/app'
import React from 'react'
import { TypedCaller } from 'fireschema'

type FunctionsModule = typeof import('./2-1-typed-functions')

const app: firebase.app.App = firebase.initializeApp({
  // ...
})
const functionsApp = app.functions('asia-northeast1')

export const typedCaller = new TypedCaller<FunctionsModule>(functionsApp)

const Component = () => {
  const createUser = async () => {
    const result = await typedCaller.call('createUser', {
      name: 'test',
      displayName: 'Test',
      age: 20,
      timestamp: new Date().toISOString(),
      options: { a: true },
    })

    if (result.error) {
      console.error(result.error)
      return
    }
    console.log(result.data)
  }

  return <button onClick={createUser}></button>
}