Files
apskel-pos-backend/internal
efrilmandClaude Opus 5 ebf666c004 feat(product): require a unit for weight-based products
A product sold by weight with no unit produces order lines with nothing to
print: the receipt would read "4,2" with no idea of what. Until now nothing
stopped that — the mistake only surfaced at the cashier.

Enforce it in two places, because neither alone sees the whole picture. On
create, the validator has everything it needs. On update, the request may
omit unit_id for a product that already has one, so the check runs in the
processor against the merged product: what is rejected is the end state, a
product sold by weight with no unit.

Also fixes two things this uncovered:

The struct tags on the product contracts are decorative — this validator is
hand-written and never calls validator.Struct — so `oneof=unit weight` was
never enforced, and an unknown sell_by was silently rewritten to "unit" by
the mapper. It is now rejected with a message that names the valid values.

The update validator's "at least one field" guard did not list unit_id,
sell_by or print_to_checker, so an update carrying only one of those was
turned away as an empty request.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-09-06 17:29:34 +07:00
..
2026-08-13 14:38:28 +07:00
2026-05-09 12:24:44 +07:00
2026-06-19 13:58:36 +07:00
2026-05-09 01:01:25 +07:00
2026-08-13 14:38:28 +07:00
…
…

Internal Architecture

This document describes the clean architecture implementation for the POS backend with complete separation of concerns between database entities, business models, and constants.

📁 Package Structure

/constants - Business Constants

  • Purpose: All business logic constants, enums, and validation helpers
  • Usage: Used by models, services, and validation layers
  • Features:
    • Type-safe enums (UserRole, OrderStatus, PaymentStatus, etc.)
    • Business validation functions (IsValidUserRole, etc.)
    • Default values and limits
    • No dependencies on database or frameworks

/entities - Database Models

  • Purpose: Database-specific models with GORM tags and hooks
  • Usage: ONLY used by repository layer for database operations
  • Features:
    • GORM annotations (gorm: tags)
    • Database relationships and constraints
    • BeforeCreate/AfterCreate hooks
    • Table name specifications
    • SQL-specific data types
    • Never used in business logic

/models - Business Models

  • Purpose: Pure business domain models without any framework dependencies
  • Usage: Used by services, handlers, and business logic
  • Features:
    • Clean JSON serialization (json: tags)
    • Validation rules (validate: tags)
    • Request/Response DTOs
    • Zero GORM dependencies
    • Zero database annotations
    • Uses constants package for type safety
    • Pure business logic methods

/mappers - Data Transformation

  • Purpose: Convert between entities and business models
  • Usage: Bridge between repository and service layers
  • Features:
    • Entity ↔ Model conversion functions
    • Request DTO → Entity conversion
    • Entity → Response DTO conversion
    • Null-safe conversions
    • Slice/collection conversions
    • Type conversions between constants and entities

/repository - Data Access Layer

  • Purpose: Database operations using entities exclusively
  • Usage: Only works with database entities
  • Features:
    • CRUD operations with entities
    • Query methods with entities
    • Private repository implementations
    • Interface-based contracts
    • Never references business models

🔄 Data Flow

API Request (JSON)
       ↓
Request DTO (models)
       ↓
Business Logic (services with models + constants)
       ↓
Entity (via mapper)
       ↓
Repository Layer (entities only)
       ↓
Database
       ↓
Entity (from database)
       ↓
Business Model (via mapper)
       ↓
Response DTO (models)
       ↓
API Response (JSON)

🎯 Key Design Principles

✅ Clean Business Models


type User struct {
    ID    uuid.UUID           `json:"id"`
    Role  constants.UserRole  `json:"role"`
    }

type User struct {
    ID   uuid.UUID `gorm:"primaryKey" json:"id"`
    Role string    `gorm:"size:50" json:"role"`
    }

✅ Type-Safe Constants


type UserRole string
const (
    RoleAdmin UserRole = "admin"
    )
func IsValidUserRole(role UserRole) bool { /* ... */ }

const AdminRole = "admin"  ```

### ✅ **Repository Isolation**
```go

func (r *userRepository) Create(ctx context.Context, user *entities.User) error {
    return r.db.Create(user).Error
}

func (r *userRepository) Create(ctx context.Context, user *models.User) error {
    }

📊 Example Usage

Service Layer (Business Logic)

func (s *userService) CreateUser(req *models.UserCreateRequest) (*models.UserResponse, error) {
        if !constants.IsValidUserRole(req.Role) {
        return nil, errors.New("invalid role")
    }
    
        entity := mappers.UserCreateRequestToEntity(req, hashedPassword)
    
        err := s.userRepo.Create(ctx, entity)
    if err != nil {
        return nil, err
    }
    
        return mappers.UserEntityToResponse(entity), nil
}

Repository Layer (Data Access)

func (r *userRepository) Create(ctx context.Context, user *entities.User) error {
        return r.db.WithContext(ctx).Create(user).Error
}

Handler Layer (API)

func (h *userHandler) CreateUser(c *gin.Context) {
    var req models.UserCreateRequest
    
        if err := c.ShouldBindJSON(&req); err != nil {
        c.JSON(400, gin.H{"error": err.Error()})
        return
    }
    
        resp, err := h.userService.CreateUser(&req)
    if err != nil {
        c.JSON(500, gin.H{"error": err.Error()})
        return
    }
    
        c.JSON(201, resp)
}

🏗️ Architecture Benefits

  1. 🎯 Single Responsibility: Each package has one clear purpose
  2. 🔒 Zero Database Leakage: Business logic never sees database concerns
  3. 🧪 Testability: Easy to mock interfaces and test business logic
  4. 🔧 Maintainability: Changes to database don't affect business models
  5. 🚀 Flexibility: Can change ORM without touching business logic
  6. 📜 API Stability: Business models provide stable contracts
  7. 🛡️ Type Safety: Constants package prevents invalid states
  8. 🧹 Clean Code: No mixed concerns anywhere in the codebase

📋 Development Guidelines

Constants Package (/constants)

  • ✅ Define all business enums and constants
  • ✅ Provide validation helper functions
  • ✅ Include default values and limits
  • ❌ Never import database or framework packages
  • ❌ No business logic, only constants and validation

Models Package (/models)

  • ✅ Pure business structs with JSON tags only
  • ✅ Use constants package for type safety
  • ✅ Include validation tags for input validation
  • ✅ Separate Request/Response DTOs
  • ✅ Add business logic methods (validation, calculations)
  • ❌ NEVER include GORM tags or database annotations
  • ❌ NEVER import database packages
  • ❌ No database relationships or foreign keys

Entities Package (/entities)

  • ✅ Include GORM tags and database constraints
  • ✅ Define relationships and foreign keys
  • ✅ Add database hooks (BeforeCreate, etc.)
  • ✅ Use database-specific types
  • ❌ NEVER use in business logic or handlers
  • ❌ NEVER add business validation rules

Mappers Package (/mappers)

  • ✅ Always check for nil inputs
  • ✅ Handle type conversions between constants and strings
  • ✅ Provide slice conversion helpers
  • ✅ Keep conversions simple and direct
  • ❌ No business logic in mappers
  • ❌ No database operations

Repository Package (/repository)

  • ✅ Work exclusively with entities
  • ✅ Use private repository implementations
  • ✅ Provide clean interface contracts
  • ❌ NEVER reference business models
  • ❌ NEVER import models package

🚀 Migration Complete

All packages have been successfully reorganized:

  • ✅ 4 Constants files - All business constants moved to type-safe enums
  • ✅ 10 Clean Model files - Zero GORM dependencies, pure business logic
  • ✅ 11 Entity files - Database-only models with GORM tags
  • ✅ 11 Repository files - Updated to use entities exclusively
  • ✅ 2 Mapper files - Handle conversions between layers
  • ✅ Complete separation - No cross-layer dependencies

The codebase now follows strict clean architecture principles with complete separation of database concerns from business logic! 🎉