Test Cases Reduction Based On Program Slicing Algorithm

Information industry now is more and more widely used, significant events happened because of software quality are not uncommon, for example millions of the national interbank transactions cannot be achieved because China Union Pay business interrupted for6hours. And the key of informationization is software, so the primary job is to test the software errors in order to ensure it’s smoothly implementation. So test case execution strategy has a crucial influence on test results, it can save a lot of manpower and material resources, in the same time it also reduces maintenance costs. It is necessary to select a set of effective test cases from the original test cases to complete the testing process; the selected new test cases can also test every test requirement of the test requirements set and ensure error detection capability. Test cases selection directly affects the quality of the software, so the study of methods of test cases minimization has a very vital significance.Program slicing technique deletes the points which are besides the interest points to have a convenient to understand and analysis program. In this paper, program slicing technique is applied to the optimization of the test set. Program slicing techniques only select procedures which are related to the interested test points in the program, or procedures affecting the interested test points. This can reduce the number of test cases, and use the least test cases to test the most bugs.This research work of this paper mainly includes the following several aspects:(1) According to the reverse program flow, functional dependency set and arguments parameter table, then give a algorithm of program slicing;(2) Function calls program slicing algorithm is also applicable to inter-program slicing, the later only depends on the reverse process stream, the execution of the algorithm is more simple;(3) Test cases minimization algorithm use improved GRE algorithm. According to the interest points of test requirements of section to generate program slicing, and then generate the corresponding test cases for the points in slicing. First remove redundant test cases, then selected essential test cases, finally, use double selection mechanism algorithm to instead greedy algorithm.In double selection mechanism, if there isn’t a test case can cover all test requirements, then to combine every two test cases. If the combination covers more test requirements than the test case which covers the most test requirements, then choose the combination, or choose the test case which covers the most test requirements; in the same time, delete test requirements which are covered. Continue to cycle the above steps until all test requirements are met.Finally, the paper verifies the feasibility and practicality of the algorithm through the giving example, and compares the proposed algorithm with other algorithms, and then provides the direction and content of the next step and content.