Modular Semantics for Object-Oriented Models

Hamie, A. and Howse, J. and Kent, S. (1998) Modular Semantics for Object-Oriented Models. In: Proceedings of Northern Formal Methods Workshop. (Full text available)

PDF
Download (125kB)
[img]
Preview

Abstract

This paper describes a formal, modular approach to the semantics of object-oriented models. These models are expressed in the UML notation augmented with the Object Constraint language (OCL), which provides a precise textual language for expressing assertions: invariant constraints and operation specifications. The approach to semantics we adopt could easily be applied to other, similar notation sets. An important aspect of our approach is to treat every component of a model, down to the level of individual diagram elements, as distinct semantic entities which, here, are theories in Larch. The semantics of a model or part of a model is then the composition of the semantic entities corresponding to the individual model elements. This leads to a highly modular approach, allow-ing reuse of generic semantic entities, resulting in an elegant, transparent semantics for individual models. The fine-grained modularity promises to support the extraction and manipulation of different ''views'' of a model. The composition of model elements supported by the semantics promises to underpin the development of systems from reusable components. Larch is used as it supports theory composition, ensures that the semantics is relatively accessible (essentially theories of FOPL), is well-defined, and is close to technologies most likely to provide auto-mated tool support (including a toolset to support it directly).

Item Type: Conference or workshop item (Paper)
Subjects: Q Science > QA Mathematics (inc Computing science) > QA 76 Software, computer programming,
Divisions: Faculties > Science Technology and Medical Studies > School of Computing > Systems Architecture Group
Depositing User: Mark Wheadon
Date Deposited: 22 Aug 2009 23:12
Last Modified: 06 Sep 2011 04:00
Resource URI: http://kar.kent.ac.uk/id/eprint/21617 (The current URI for this page, for reference purposes)
  • Depositors only (login required):

Downloads

Downloads per month over past year