Layered Architecture

Layer Rules

@RestController        ← HTTP only. No business logic. No JPA entities in responses.
      ↓ DTOs
@Service               ← All business logic lives here. Orchestrates repositories.
      ↓ Domain objects / Entities
@Repository            ← Data access only. No business logic. Returns entities or projections.
      ↓ JPA / JDBC
Database

Controller Layer

• Handles HTTP: parsing requests, validating input (@Valid), returning responses
• Calls ONE service method per endpoint — no orchestration in controllers
• Never returns @Entity classes directly — always map to response DTOs
• Never injects @Repository — always goes through a @Service
• Exception handling via @ControllerAdvice, never try/catch in controllers

// ✅ GOOD
@PostMapping("/orders")
public ResponseEntity<OrderResponse> createOrder(@Valid @RequestBody CreateOrderRequest request) {
    Order order = orderService.createOrder(request);
    return ResponseEntity.status(HttpStatus.CREATED).body(OrderResponse.from(order));
}

// ❌ BAD — business logic in controller
@PostMapping("/orders")
public ResponseEntity<Order> createOrder(@RequestBody CreateOrderRequest request) {
    if (request.getItems().isEmpty()) throw new RuntimeException("No items");
    Order order = orderRepository.save(new Order(request)); // direct repo access
    return ResponseEntity.ok(order); // returning entity
}

Service Layer

• Contains all business logic, validation rules, and orchestration
• @Transactional lives here, not in controllers or repositories
• Constructor injection only — never @Autowired field injection
• One service per aggregate root (OrderService, not OrderAndPaymentService)
• Returns domain objects or DTOs — never HttpServletRequest / HttpServletResponse

// ✅ GOOD
@Service
@RequiredArgsConstructor
public class OrderService {
    private final OrderRepository orderRepository;
    private final InventoryService inventoryService;

    @Transactional
    public Order createOrder(CreateOrderRequest request) {
        inventoryService.reserve(request.items()); // records use accessor names, not getX()
        Order order = Order.from(request);
        return orderRepository.save(order);
    }
}

// ❌ BAD — field injection, HTTP concern in service
@Service
public class OrderService {
    @Autowired private OrderRepository orderRepository;

    public ResponseEntity<Order> createOrder(...) { ... } // HTTP type in service
}

Repository Layer

• Extends JpaRepository<Entity, ID> or CrudRepository
• Custom queries via @Query or query derivation — no raw SQL unless unavoidable
• Returns entities or Spring Data Projections — never raw Object[]
• No business logic — pure data access

DTOs

• Separate Request / Response DTOs — never use the same class for both
• Validation annotations (@NotNull, @Size, etc.) on Request DTOs only
• Static factory method ResponseDto.from(Entity entity) for mapping
• Use records for immutable DTOs (Java 16+)

// ✅ GOOD
public record OrderResponse(UUID id, String status, List<LineItemResponse> items, Instant createdAt) {
    public static OrderResponse from(Order order) {
        return new OrderResponse(order.getId(), order.getStatus().name(),
            order.getItems().stream().map(LineItemResponse::from).toList(),
            order.getCreatedAt());
    }
}

Mapper Pattern

• Keep mapping logic out of controllers and services — use dedicated mapper classes or static factory methods
• Mapper is a plain class or utility — not a Spring bean unless it needs injected dependencies
• Entity → Response DTO: static method on the response DTO (OrderResponse.from(order))
• Request DTO → Entity: static factory on the entity (Order.from(request)) or a mapper class
• Collection mapping: use .stream().map(OrderResponse::from).toList() — never manual loops

// ✅ GOOD — dedicated mapper for complex mappings
public class OrderMapper {

    public static OrderResponse toResponse(Order order) {
        return new OrderResponse(
            order.getId(),
            order.getStatus().name(),
            order.getItems().stream().map(OrderMapper::toLineItem).toList(),
            order.getCreatedAt()
        );
    }

    public static Order toEntity(CreateOrderRequest request, User user) {
        Order order = Order.create(request.customerEmail(), user);
        request.items().forEach(item ->
            order.addItem(item.productId(), item.quantity()));
        return order;
    }

    private static LineItemResponse toLineItem(OrderItem item) {
        return new LineItemResponse(item.getProductId(), item.getQuantity(), item.getPrice());
    }
}

Configuration Layer

• @Configuration classes live in a config/ package — never in service/ or controller/
• Configuration never imports service or controller classes
• Use @ConfigurationProperties for type-safe config — never raw @Value for groups of related settings
• Bean definitions for infrastructure concerns only (RestTemplate, ObjectMapper, SecurityFilterChain)

Cross-Cutting Concerns

• Logging: use @Slf4j — never System.out.println
• Validation: @Valid on controller parameters, custom validators as @Component
• Exception handling: single @RestControllerAdvice class, never try/catch in controllers
• Auditing: @CreatedDate / @LastModifiedDate with @EnableJpaAuditing

Gotchas

• Agent tends to put @Transactional on controllers — move it to services
• Agent uses @Autowired field injection — always use constructor injection (@RequiredArgsConstructor)
• Agent returns List<Entity> from controllers — always map to List<ResponseDto>
• Agent creates OrderAndInventoryService god classes — split by aggregate
• Agent puts mapping logic inside controllers — extract to mapper class or DTO factory method
• Agent creates @Configuration classes that depend on @Service beans — configuration should only wire infrastructure