Key Principle: All dependencies point inward. The Application layer defines interfaces; Infrastructure implements them.

Folder Structure

src/
├── domain/                    # Business logic (zero dependencies)
│   ├── entities/             # User
│   ├── value-objects/        # Email, UserId
│   └── repositories/         # IUserRepository (interface)
│
├── application/               # Business use cases
│   ├── dto/                  # Data Transfer Objects
│   ├── interfaces/           # IRepositoryFactory
│   ├── services/             # UserManagementService, UserProfileService
│   └── use-cases/            # CreateUserUseCase, GetUserByIdUseCase, etc.
│
├── infrastructure/            # Technical implementations
│   ├── config/
│   │   ├── repositories.ts   # Singleton factory functions
│   │   ├── RepositoryFactory.ts  # Implements IRepositoryFactory
│   │   └── serviceFactory.ts     # Service composition functions
│   ├── database/             # MySQL and MongoDB connections
│   └── persistence/
│       ├── mysql/           # MySQLUserRepository
│       └── mongodb/         # MongoDBUserRepository
│
└── presentation/              # HTTP interface
    ├── controllers/          # UserController
    └── routes/              # Composition root

How It Works

Dependency Injection

Dependencies are wired together in the composition root (route files):

// src/presentation/routes/userRoutes.ts

import { createUserProfileService, createUserManagementService }
  from '../../infrastructure/config/serviceFactory';

// 1. Create services using factory functions
const profileService = createUserProfileService();
const managementService = createUserManagementService();

// 2. Inject into controller
const userController = new UserController(profileService, managementService);

// 3. Define routes
router.post('/', (req, res) => userController.createUser(req, res));

Abstract Factory Pattern

The Application layer defines the factory interface:

// src/application/interfaces/IRepositoryFactory.ts
export interface IRepositoryFactory {
  createMySQLUserRepository(): IUserRepository;
  createMongoDBUserRepository(): IUserRepository;
}

Infrastructure implements the interface:

// src/infrastructure/config/RepositoryFactory.ts
export class RepositoryFactory implements IRepositoryFactory {
  createMySQLUserRepository(): IUserRepository {
    return getMySQLUserRepository(); // Singleton
  }

  createMongoDBUserRepository(): IUserRepository {
    return getMongoDBUserRepository(); // Singleton
  }
}

Services receive the factory via constructor:

// src/application/services/UserManagementService.ts
export class UserManagementService {
  constructor(private repositoryFactory: IRepositoryFactory) {}

  async createUser(data: CreateUserDTO) {
    // Service creates use case with repositories from factory
    const mysqlRepo = this.repositoryFactory.createMySQLUserRepository();
    const mongoRepo = this.repositoryFactory.createMongoDBUserRepository();

    const useCase = new CreateUserUseCase(mysqlRepo, mongoRepo);
    return useCase.execute(data);
  }
}

Benefits:

• Application layer never imports from Infrastructure
• Easy to test with mock factories
• True dependency inversion
• All dependencies point inward

Bounded Contexts

Services represent bounded contexts:

• UserProfileService - Read operations (profile viewing)
• UserManagementService - Write operations (CRUD lifecycle)

Each service groups related use cases by business capability.

Use Cases

Use cases implement single business operations:

export class CreateUserUseCase {
  constructor(
    private mysqlRepository: IUserRepository,
    private mongoRepository: IUserRepository
  ) {}

  async execute(data: CreateUserDTO): Promise<UserDTO> {
    const user = User.create(data);

    // Dual-write pattern
    await this.mongoRepository.save(user);
    await this.mysqlRepository.save(user);

    return this.mapToDTO(user);
  }
}

Multi-Database Architecture

Dual-Write Pattern

Data is written to both MySQL and MongoDB simultaneously:

• Create → Writes to both databases
• Update → Updates both databases
• Delete → Removes from both databases
• Read → Reads from MongoDB (document-based queries)

Configuration

Both databases are required:

# MySQL
MYSQL_HOST=localhost
MYSQL_PORT=3306
MYSQL_USER=root
MYSQL_PASSWORD=password
MYSQL_DATABASE=hexacore

# MongoDB
MONGODB_URI=mongodb://localhost:27017/hexacore

API Endpoints

User Operations

# Create user
POST /api/users
{
  "name": "John Doe",
  "email": "john@example.com",
  "age": 30
}

# Get all users
GET /api/users

# Get user by ID
GET /api/users/:id

# Update user
PUT /api/users/:id
{
  "name": "Jane Doe",
  "email": "jane@example.com"
}

# Delete user
DELETE /api/users/:id

# Health check
GET /health

Domain Model

Entities

Entities encapsulate business logic:

export class User {
  private constructor(
    private readonly id: UserId,
    private name: string,
    private email: Email,
    private readonly age?: number
  ) {}

  static create(props: CreateUserProps): User {
    // Validation and business rules
    return new User(UserId.generate(), props.name, new Email(props.email), props.age);
  }

  updateEmail(newEmail: string): void {
    this.email = new Email(newEmail); // Validates
  }
}

Value Objects

Value objects ensure domain invariants:

export class Email {
  constructor(private readonly value: string) {
    if (!this.isValid(value)) {
      throw new Error('Invalid email format');
    }
  }

  private isValid(email: string): boolean {
    return /^[^\s@]+@[^\s@]+\.[^\s@]+$/.test(email);
  }
}

Adding New Features

Add a New Use Case

1. Create use case in src/application/use-cases/
2. Add method to appropriate service
3. Service creates and executes the use case

Add a New Service (Bounded Context)

1. Create service in src/application/services/
2. Service constructor accepts IRepositoryFactory
3. Add factory function in src/infrastructure/config/serviceFactory.ts
4. Inject service into controller

Add a New Entity

1. Create entity in src/domain/entities/
2. Create repository interface in src/domain/repositories/
3. Implement repositories in src/infrastructure/persistence/
4. Update IRepositoryFactory and RepositoryFactory
5. Create use cases and services
6. Create controller and routes

Testing

Unit Test Use Cases

const mockMySQLRepo = createMock<IUserRepository>();
const mockMongoRepo = createMock<IUserRepository>();

const useCase = new CreateUserUseCase(mockMySQLRepo, mockMongoRepo);
await useCase.execute(userData);

expect(mockMySQLRepo.save).toHaveBeenCalled();
expect(mockMongoRepo.save).toHaveBeenCalled();

Unit Test Services

const mockFactory: IRepositoryFactory = {
  createMySQLUserRepository: () => mockMySQLRepo,
  createMongoDBUserRepository: () => mockMongoRepo,
};

const service = new UserManagementService(mockFactory);
await service.createUser(userData);

Unit Test Domain

test('Email validates format', () => {
  expect(() => new Email('invalid')).toThrow('Invalid email format');
  expect(() => new Email('valid@example.com')).not.toThrow();
});

Technologies

• TypeScript 5.3 - Type safety with strict mode
• Node.js 16+ - JavaScript runtime
• Express 4.18 - Web framework
• MySQL2 3.6 - MySQL database driver
• Mongoose 8.0 - MongoDB ODM
• class-validator - DTO validation
• class-transformer - Object transformation

Scripts

yarn dev        # Development with hot reload
yarn build      # Build TypeScript to JavaScript
yarn start      # Production: build + run
yarn lint       # Run ESLint
yarn format     # Format with Prettier

Architecture Benefits

✅ Testability - Easy to mock dependencies ✅ Maintainability - Clear separation of concerns ✅ Flexibility - Swap implementations without changing business logic ✅ Independence - Business logic independent of frameworks and databases ✅ Scalability - Add features without affecting existing code ✅ Type Safety - Full TypeScript support

Design Principles

• Dependency Inversion - High-level modules define interfaces
• Single Responsibility - Each class has one reason to change
• Open/Closed - Open for extension, closed for modification
• Interface Segregation - Small, focused interfaces
• Clean Architecture - Dependencies point inward only

Documentation

• ARCHITECTURE.md - Detailed architecture documentation
• CLAUDE.md - Guide for Claude Code AI assistant

License

MIT

Author

Created by Shawon Kanji

Built with Clean Architecture and Domain-Driven Design principles for maximum flexibility and maintainability.