Introduction
Functional programming (FP) has gained popularity for its composability, testability, and robustness. In the JavaScript ecosystem, libraries like fp-ts bring powerful FP concepts to TypeScript, allowing you to write cleaner and more reliable code.
This article explores fp-ts concepts like Option, Either, Task, Reader, and ReaderTaskEither. We’ll build a basic CRUD app using fp-ts, pg (PostgreSQL client), and Express.js to see how these abstractions shine in real-world applications.
Key Concepts
Before diving into the app, let’s briefly discuss the main concepts:
-
Option: Models the presence or absence of a value (
SomeorNone). -
Either: Represents computations that can succeed (
Right) or fail (Left). - Task: Represents lazy asynchronous computations.
- Reader: Injects dependencies into computations.
- ReaderTaskEither: Combines Reader, Task, and Either for async operations with dependencies and error handling.
Setting Up the Project
Initialize the Project
mkdir fp-ts-crud && cd fp-ts-crud
npm init -y
npm install express pg fp-ts io-ts
npm install --save-dev typescript @types/express ts-node-dev jest @types/jest ts-jest
Setup TypeScript
Create a tsconfig.json:
{
"compilerOptions": {
"target": "ES2020",
"module": "CommonJS",
"outDir": "dist",
"strict": true,
"esModuleInterop": true
},
"include": ["src/**/*"]
}
Project Structure
src/
index.ts # Entry point
db.ts # Database setup
models/ # Data models and validation
services/ # Business logic
controllers/ # CRUD operations
utils/ # fp-ts utilities
errors/ # Custom error classes
Implementing the CRUD App
Database Setup (db.ts)
import { Pool } from 'pg';
export const pool = new Pool({
user: 'postgres',
host: 'localhost',
database: 'fp_ts_crud',
password: 'password',
port: 5432,
});
Define Models and Validation (models/User.ts)
import * as t from 'io-ts';
import { isRight } from 'fp-ts/Either';
export const User = t.type({
id: t.number,
name: t.string,
email: t.string,
});
export const validateUser = (data: unknown): t.TypeOf<typeof User> | null => {
const result = User.decode(data);
return isRight(result) ? result.right : null;
};
Custom Error Handling (errors/AppError.ts)
export class AppError extends Error {
constructor(public statusCode: number, public code: string, public message: string) {
super(message);
this.name = 'AppError';
}
}
export const createAppError = (statusCode: number, code: string, message: string): AppError => {
return new AppError(statusCode, code, message);
};
Service Layer (services/UserService.ts)
import { pool } from '../db';
import { ReaderTaskEither, right, left } from 'fp-ts/ReaderTaskEither';
import { pipe } from 'fp-ts/function';
import { createAppError, AppError } from '../errors/AppError';
import { validateUser } from '../models/User';
type Dependencies = { db: typeof pool };
type User = { name: string; email: string };
export const createUser = (
user: User
): ReaderTaskEither<Dependencies, AppError, string> => (deps) => async () => {
// Validate the incoming user data
const validatedUser = validateUser(user);
if (!validatedUser) {
return left(createAppError(400, 'INVALID_USER_DATA', 'Invalid user data provided'));
}
try {
const result = await deps.db.query(
'INSERT INTO users (name, email) VALUES ($1, $2) RETURNING id',
[validatedUser.name, validatedUser.email]
);
return right(`User created with ID: ${result.rows[0].id}`);
} catch (error) {
return left(createAppError(500, 'USER_CREATION_FAILED', 'Failed to create user'));
}
};
export const getUser = (
id: number
): ReaderTaskEither<Dependencies, AppError, { id: number; name: string; email: string }> => (deps) => async () => {
try {
const result = await deps.db.query('SELECT * FROM users WHERE id = $1', [id]);
return result.rows[0]
? right(result.rows[0])
: left(createAppError(404, 'USER_NOT_FOUND', 'User not found'));
} catch {
return left(createAppError(500, 'USER_FETCH_FAILED', 'Failed to fetch user'));
}
};
CRUD Operations (controllers/UserController.ts)
import { pipe } from 'fp-ts/function';
import { createUser, getUser } from '../services/UserService';
import { pool } from '../db';
import { AppError } from '../errors/AppError';
const errorHandler = (err: unknown, res: express.Response): void => {
if (err instanceof AppError) {
res.status(err.statusCode).json({ error: { code: err.code, message: err.message } });
} else {
res.status(500).json({ error: { code: 'UNKNOWN_ERROR', message: 'An unexpected error occurred' } });
}
};
export const createUserHandler = (req: express.Request, res: express.Response): void => {
pipe(
createUser(req.body),
(task) => task({ db: pool }),
(promise) =>
promise.then((result) =>
result._tag === 'Left'
? errorHandler(result.left, res)
: res.json({ message: result.right })
)
);
};
export const getUserHandler = (req: express.Request, res: express.Response): void => {
pipe(
getUser(parseInt(req.params.id, 10)),
(task) => task({ db: pool }),
(promise) =>
promise.then((result) =>
result._tag === 'Left'
? errorHandler(result.left, res)
: res.json(result.right)
)
);
};
Express API (index.ts)
import express from 'express';
import { createUserHandler, getUserHandler } from './controllers/UserController';
const app = express();
app.use(express.json());
// Routes
app.post('/users', createUserHandler);
app.get('/users/:id', getUserHandler);
// Start Server
app.listen(3000, () => {
console.log('Server running on http://localhost:3000');
});
Running the App with Docker and Docker Compose
Dockerfile
# Stage 1: Build
FROM node:22 AS builder
WORKDIR /app
COPY package*.json .
RUN npm install
COPY . .
RUN npm run build
# Stage 2: Run
FROM node:22
WORKDIR /app
COPY --from=builder /app/dist ./dist
COPY package*.json ./
RUN npm install --production
CMD ["node", "dist/index.js"]
docker-compose.yml
version: '3.8'
services:
db:
image: postgres:15
environment:
POSTGRES_USER: postgres
POSTGRES_PASSWORD: password
POSTGRES_DB: fp_ts_crud
ports:
- "5432:5432"
volumes:
- db_data:/var/lib/postgresql/data
volumes:
db_data:
Run the App - Development Mode
# Start the database
docker-compose up -d
# Run the app
npx ts-node-dev src/index.ts
Run the App - Production Mode
# Build the docker image
docker build -t fp-ts-crud-app .
# Start the database
docker-compose up -d
# Run the container
docker run -p 3000:3000 fp-ts-crud-app
Writing Tests
Setup Jest
Update package.json scripts:
"scripts": {
"test": "jest",
"test:watch": "jest --watch"
}
Example Test (__tests__/UserService.test.ts)
import { createUser, getUser } from '../services/UserService';
import { pool } from '../db';
jest.mock('../db', () => ({
pool: {
query: jest.fn(),
},
}));
describe('UserService', () => {
afterEach(() => {
jest.clearAllMocks();
});
it('should create a user', async () => {
(pool.query as jest.Mock).mockResolvedValueOnce({ rows: [{ id: 1 }] });
const result = await createUser({ name: 'Alice', email: 'alice@example.com' })({ db: pool })();
expect(result._tag).toBe('Right');
if (result._tag === 'Right') {
expect(result.right).toBe('User created with ID: 1');
}
});
it('should return error if user not found', async () => {
(pool.query as jest.Mock).mockResolvedValueOnce({ rows: [] });
const result = await getUser(1)({ db: pool })();
expect(result._tag).toBe('Left');
if (result._tag === 'Left') {
expect(result.left.message).toBe('User not found');
}
});
it('should return error for invalid user data during creation', async () => {
const invalidData = { name: 123, email: 'invalid-email' };
const result = await createUser(invalidData)({ db: pool })();
expect(result._tag).toBe('Left');
if (result._tag === 'Left') {
expect(result.left.message).toBe('Invalid user data provided');
}
});
});
Conclusion
By leveraging fp-ts, Docker, and robust error handling, we built a functional, scalable, and maintainable Node.js CRUD application. Using functional programming patterns makes your code more predictable and reliable, especially when handling asynchronous workflows.
Top comments (0)