Update propagation in chimera, an active DOOD language

Abstract

Propagating updates is an important task to be performed within many database services such as integrity checking, maintenance of materialized views, and condition monitoring. This paper is concerned with the propagation of updates in an active DOOD language. The approach proposed is to make use of Chimera triggers for computing induced updates. It will be shown how a subset of Chimera's deductive rules can be compiled to update propagation triggers. In its expressiveness the rule set considered corresponds to that of Datalog with sets and negation. Using triggers for implementing update propgation has the advantage that no special component has to be implemented as a trigger mechanism has. to exist anyway. In this paper we will not propose new techniques for computing induced updates but will transfer the techniques - well-known for the relational model - to the object-oriented case.