KALA: Kernel aspect language for advanced transactions

  • Authors:
  • Johan Fabry;Éric Tanter;Theo DHondt

  • Affiliations:
  • PLEIAD Lab, Computer Science Department (DCC), University of Chile, Blanco Encalada 2120, Santiago, Chile;PLEIAD Lab, Computer Science Department (DCC), University of Chile, Blanco Encalada 2120, Santiago, Chile;Vrije Universiteit Brussel, Programming Technology Lab, Pleinlaan 2, 1050 Brussel, Belgium

  • Venue:
  • Science of Computer Programming
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

Transaction management is a known crosscutting concern. Previous research has been conducted to express this concern as an aspect. However, such work has used general-purpose aspect languages which lack a formal foundation, and most importantly are unable to express advanced models of transaction management. In this paper, we propose a domain-specific aspect language for advanced transaction management, called KALA, that overcomes these limitations. First, KALA is based on a recognized formalism for the domain of advanced transaction management, called ACTA. Second, as a consequence of being based on the ACTA formalism, KALA covers a wide variety of models for transaction management. Finally, being a domain-specific aspect language, KALA allows programmers to express their needs at a higher level of abstraction than what is achieved with general-purpose aspect languages. This paper reports on the design of KALA and its implementation over Java, based on the Reflex AOP kernel for domain-specific aspect languages.