Efficient Path Selection Strategy Based on Static Analysis for Regression Testing

Abstract

Abstract-Regression testing is used to revalidate modified program and provide confidence that changes does not harm the behavior of the existing code. Test suites grows in size as software evolve, a simple approach of regression testing is re-test all approach in which all the pre-existing test suites are executed on the code but it is too expensive and increase the cost of testing activity. Different problems have been involved with regression testing, e.g. test suites minimization problem, test selection problem, coverage identification problem, test case execution problem, test case maintenance problem etc. Another problem may occur, when tester has to select the changed paths from the set of modified paths for test case execution. This paper presents a new path selection strategy based on static analysis for regression testing which enables the tester to execute the test cases in an order that increases their effectiveness to find faults taking minimum efforts. With the proposed approach, tester can select the paths among the set of paths in an order to achieve the testing objective. Infeasible paths are also identified by the proposed approach which can reduce the effort, time, and cost.

An approach for reduced test path set selection

“…For this (i) they determined the similarity based test case pairs, which is suited for merging (ii) they identified the number of identical path sections based on local sequence alignment algorithm. Debbarma et al, [11] have proposed an efficient path selection strategy based on static analysis. In 2010, Gotlieb et al, [12] have introduced a pathoriented random testing to perform random testing at the path level.…”