Designing an n-tier architecture, part 2 – repositories
The next stage of the n-tier architecture we are going to discuss in this recipe is the bottom-most layer: the repositories, or the data layer. The purpose of a repository is to store and retrieve data. This layer can be implemented as a database, web service, or in the case of the Master Plan project, a simple in-memory cache. Unlike the controller layer, which is business logic-aware, the repository layer is only concerned with getting and storing data in its most abstract form. These classes should not even know about the model files that we created earlier.
The reason why repositories are so purposefully ignorant is to keep them focused entirely on their task – persistence. Communicating with a database or a web service can become complicated if you have many small requirements. These concerns are typically beneath business logic and are easier to resolve when you're only focused on abstract objects. Remember...
