Implementing Set Operators over Class Hierarchies

this paper we introduce the concept of an entity database model. This model is not intended to be a complete working database model in the sense that various semantic database models, (e.g. DAPLEX [Shi81], SDM, Semantic Data Model [HaM81], Galileo [ACO85], IFO [AbH87]) or object oriented database models, (e.g. Smalltalk [CoM84, MSe86], Orient84/K [IsT86], and [SSE87]) are. These latter models are "complete" in the sense that entire computational languages with syntax and semantics have, or could be, implemented with respect to them. Instead, we will characterize our generic entity database model by only two fundamental properties which are common to all of them. Then we will derive several necessary consequences arising from the implementation of set operators in any such language belonging to the entity model and thereby contrast such languages with the many languages adhering to the relational database model, (e.g. [CGY81,RoS87, Sho79, SWK76]). We hasten to point out that our entity database model is quite distinct from the entity-relationship model introduced by Peter Chen [Che76] and extended by several others (e.g. [CaS84, TYF86]). Their's can be viewed as a design model which is extremely useful for organizing the structure of real world data prior to representing it with respect to some particular language, which is often relational. Our entity database model is strictly an implementational model. The key presupposition of the entity database model is that the database is implemented by creating uniquely identifiable entities, or objects. These entities may be transient or persistent. These representational entities may, or may not, correspond directly with what we would choose to regard as "entities" in the real world situation we are modeling. It is customary t..