Deductive database technology represents an important step towards the goal of developing highly-declarative database programming languages. In order to make deductive databases a practical technology, deductive rules have to be extended to provide a dynamic behavior. In particular, current applications require not only a support for updates and transactions but also the ability to automatically react to the occurrence of particular events. This is possible by integrating typical deductive rules, whose execution is user-dependent, with active rules, whose execution is event-dependent. Current solutions to this problem are not completely satisfactory. In particular, they often lack a clear semantics, guaranteeing termination, confluence and efficient evaluation. The aim of this paper is to propose a new language for integrating active rules, deductive rules and updates in a uniform logical context. The language we propose is based on the U-Datalog language [9], and extends it with support for active rules, modeled according to the PARK semantics [23]. The resulting language allows the representation of several dynamic aspects, such as transaction execution, reactive behavior and update propagation, in a uniform logical framework, admitting a clear and flexible semantics.
Active-U-Datalog: Integrating active rules in a logical update language.
GERVASI, VINCENZO;
1998-01-01
Abstract
Deductive database technology represents an important step towards the goal of developing highly-declarative database programming languages. In order to make deductive databases a practical technology, deductive rules have to be extended to provide a dynamic behavior. In particular, current applications require not only a support for updates and transactions but also the ability to automatically react to the occurrence of particular events. This is possible by integrating typical deductive rules, whose execution is user-dependent, with active rules, whose execution is event-dependent. Current solutions to this problem are not completely satisfactory. In particular, they often lack a clear semantics, guaranteeing termination, confluence and efficient evaluation. The aim of this paper is to propose a new language for integrating active rules, deductive rules and updates in a uniform logical context. The language we propose is based on the U-Datalog language [9], and extends it with support for active rules, modeled according to the PARK semantics [23]. The resulting language allows the representation of several dynamic aspects, such as transaction execution, reactive behavior and update propagation, in a uniform logical framework, admitting a clear and flexible semantics.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.