Affiliations: [a] Computer Science Department, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel | [b] Sauder School of Business, The University of British Columbia, Vancouver, BC V6T 1Z2, Canada
Note: [1] Partially supported by the Lynn and William Frankel center for Computer Sciences, and by the Paul Ivanir Center for
Robotics and Production Management at Ben-Gurion University of the Negev.
Abstract: Enterprise information systems distinguish the Domain layer that handles the major business logic of an application, from the Data (Persistent) layer that handles storage concerns alone. The integration of these layers is not straightforward since usually the requirement is for partial persistency, i.e., persistency is required only for a subset of the Domain layer classes. Industry tools provide partial help by supporting convenient abstractions on top of concrete database systems. Nevertheless, the developer still has to design the concrete ties between the layers. Full automation can be a great improvement in complex system development, and goes along with the Model Driven Engineering approach. In this paper we introduce a set of independent Data Access Patterns that provide the missing link towards full automation of the Domain-Data layer interaction. Each pattern is a simple refactoring that yields some Data Access layer constructs, and relies on local information in the Domain layer. All patterns are Domain layer transparent, i.e., their application leaves the Domain layer intact. We provide an algorithm for combined pattern application and show its correctness. An implementation of our method is on the way.
