Automation of test scripts in software product line using Model driven architecture

Author(s):  
Abdelgaffar Hamed Ahmed ◽  
Abeer Abd Allah SidAhmed ◽  
Rehab Bakheet Eltoum
Author(s):  
Maria Eugenia Cabello ◽  
Isidro Ramos ◽  
Oscar Alberto Santana ◽  
Saúl Iván Beristain

This paper presents a process, a method and a framework for developing families of software systems in a domain. The process is generic (domain-independent) and produces skeleton software architectures as Software Product Lines. The genericity is supported by the metamodels (abstract languages) that are defined in order to describe the Reference Architecture (structure view, behavior view and variability view) of the system domain. A standardized Production Plan takes the Reference Architecture as input and produces the equivalent Skeleton Software Architecture (component-connector view) using a Feature Model configuration (describing the system to be) as output. This Skeleton Software Architecture includes the structure and behavior of the target software product. A framework has been implemented to support the approach. The process is applied, as an example, to the Diagnostic Expert Systems domain. Our approach is based on Model-Driven Engineering techniques and the Software Product Line paradigm. A domain analysis must be done in order to build the Reference Architecture.


2009 ◽  
pp. 1280-1312 ◽  
Author(s):  
Gan Deng ◽  
Jeff Gray ◽  
Douglas C. Schmidt ◽  
Yuehua Lin ◽  
Aniruddha Gokhale ◽  
...  

This chapter describes our approach to modeldriven engineering (MDE)-based product line architectures (PLAs) and presents a solution to address the domain evolution problem. We use a case study of a representative software-intensive system from the distributed real-time embedded (DRE) systems domain to describe key challenges when facing domain evolution and how we can evolve PLAs systematically and minimize human intervention. The approach uses a mature metamodeling tool to define a modeling language in the representative DRE domain, and applies a model transformation tool to specify model-tomodel transformation rules that precisely define metamodel and domain model changes. Our approach automates many tedious, time consuming, and error-prone tasks of model-to-model transformation, thus significantly reducing the complexity of PLA evolution.


Sign in / Sign up

Export Citation Format

Share Document