Key Concepts

This article champions the Domain Model as a crucial, persistence-agnostic layer in software architecture. It emphasizes the clear definition of interactions between entities, focusing on data and behavior independent of the underlying database or storage mechanism. The article contrasts this with Database Models, which often intertwine domain logic with database access, hindering scalability and testability.

Building a rich Domain Model presents challenges. It demands meticulous definition of domain objects and their interactions, plus the implementation of a mapping layer to manage data flow between the model and persistence layers. However, the payoff is significant: portability and adaptability across diverse infrastructures.

The article advocates using Plain Old PHP Objects (POPOs) to encapsulate rich business logic within clean APIs, providing a practical PHP implementation example. It demonstrates the creation and interaction of entities (posts, comments, users) without relying on a specific persistence solution.

The proliferation of MVC implementations often leads to a Database Model, blurring the lines between domain logic and database access. While seemingly convenient for client code due to its simplified API (e.g., $user->save()), this approach compromises object-oriented design principles and introduces scalability and testability issues. Active Record and Table Data Gateway patterns, when coupled directly with domain logic, can exacerbate these problems.

A Domain Model, conversely, is an independent, persistence-agnostic layer clearly defining entity interactions through data and behavior. Creating a rich Domain Model with interacting objects and constraints is complex, requiring both model definition and a mapping layer for data transfer between persistence and model. Despite this added complexity, the resulting model's portability across different infrastructures is a major advantage.

The article demonstrates how Domain Models work well with PHP, using a blog application example. It showcases the use of POPOs to encapsulate business logic (validation, strategy) within a clean API. Interfaces (PostInterface, CommentInterface, UserInterface) define contracts for domain objects, allowing for flexible implementation swapping. An abstract AbstractEntity class simplifies field access using PHP magic methods (__set, __get). Concrete classes (Post, Comment, User) implement these interfaces, incorporating validation within their methods.

The example shows how to create and link domain objects using dependency injection. The application layer (controllers) acts as a simple mediator between the model and the presentation layer, highlighting the "Fat Models/Skinny Controllers" approach. A basic HTML template demonstrates how to present the model data. The entire implementation remains independent of any persistence mechanism, showcasing the model's portability and scalability.

While a Domain Model offers advantages over a Database Model, it requires a mapping layer to handle data transfer between the model and persistence. The choice between a custom or third-party ORM (like Doctrine or RedBeanPHP) depends on project needs.

The article concludes by highlighting the need for more emphasis on rich Domain Models in modern PHP frameworks and promises a future article detailing custom mapping layer implementation for MySQL integration.

Frequently Asked Questions (FAQs)

The FAQs section addresses key differences between Anemic and Rich Domain Models, the benefits of Domain Models in software development, their key components, and their relationship to SAFe and DDD. It also clarifies the distinction between Domain and Data Models, offers guidance on building effective Domain Models, discusses their evolution and challenges, and explains their role in improving communication between stakeholders.

The above is the detailed content of Building a Domain Model - An Introduction to Persistence Agnosticism. For more information, please follow other related articles on the PHP Chinese website!