adr-005-rich-domain-models.md

ADR 005: Rich Domain Models vs Anemic CRUD Entities

Status

Accepted

Context

We need to decide how to structure our domain models and where to place business logic within our application architecture.

Decision

We will use rich domain models that encapsulate business logic rather than anemic CRUD entities.

Consequences

Positive

• Business logic is centralized in the domain layer
• Improved encapsulation and data integrity
• Better alignment with Domain-Driven Design principles
• Easier to maintain and extend as the application grows

Negative

• Steeper learning curve for developers used to anemic models
• Potential for increased complexity in simple CRUD operations

Example

public class Book : AggregateRoot<BookId>
{
    public string Title { get; private set; }
    public Money Price { get; private set; }
    public List<AuthorId> AuthorIds { get; private set; }

    public void UpdatePrice(Money newPrice)
    {
        if (newPrice.Amount < 0)
            throw new InvalidOperationException("Price cannot be negative");

        Price = newPrice;
        AddDomainEvent(new BookPriceUpdatedEvent(Id, newPrice));
    }

    public void AddAuthor(AuthorId authorId)
    {
        if (AuthorIds.Count >= 5)
            throw new InvalidOperationException("A book cannot have more than 5 authors");

        AuthorIds.Add(authorId);
    }
}

This rich domain model encapsulates business rules (e.g., price validation, author limit) and raises domain events, unlike a simple CRUD entity.