Automated testing assistant to enforce method coverage in testing OOP
Ihssan Alkadi, Ghassan Alkadi, H. Etheridge · 2004
(337) 482-6206 Abstract - Re-implementation testing is performed to capture software errors, reduce unnecessary post- implementation testing, and prevent error recurrence. Rigorous pre-implementation testing is never sufficient to eliminate all probability of error occurrence and generally is cost-prohibitive; however, a minimum level of testing is necessary to reduce the probability of serious errors. Regardless of the importance that this minimum level of testing be performed prior to implementation, no defined framework designed to ensure sufficient testing of objects currently exists. The purpose of this paper is to describe such a framework. Testing all methods in an object generally is sufficiently similar to the testing of an object for pre-implementation testing purposes. In this paper, we present a remodeled testing assistant, the Object-Oriented Testing Assistant 11 (OOTA III) (1,14,15,16) that facilitates the testing of object-oriented code by enforcing .method coverage for each class. Method coverage involves triggering each method in every object in order to test all methods in that object. Consequently, testing at the object level is completed before proceeding to the class-level of testing. OOTA' was updated to generate test drivers for C , C#, and Java objects. Testing is a critical phase of Object Oriented (00) software development, and