A Survey on Regression Test-Case Prioritization

Regression testing is crucial for ensuring the quality of modern software systems, but can be extremely costly in practice. Test-case prioritization has been proposed to improve the effectiveness of regression testing by scheduling the execution order of test cases to detect regression bugs faster. Since its first proposal, test-case prioritization has been intensively studied in the literature. In this chapter, we perform an extensive survey and analysis on existing test-case prioritization techniques, as well as pointing out future directions for test-case prioritization. More specifically, we collect 191 papers on test-case prioritization from 1997 to 2016 and conduct a detailed survey to systematically investigate these work from six aspects, i.e., algorithms, criteria, measurements, constraints, empirical studies, and scenarios. For each of the six aspects, we discuss the existing work and the trend during the evolution of test-case prioritization. Furthermore, we discuss the current limitations/issues in test-case prioritization research, as well as potential future directions on test-case prioritization. Our analyses provide the evidence that test-case prioritization topic is attracting increasing interests, while the need for practical test-case prioritization tools remains.