Software test case generation method with every two cases forming one combination

Abstract

The invention provides a software test case generation method with every two cases forming one combination. According to the method, a boundary value algorithm, an equivalent partitioning algorithm, arandom algorithm and other algorithms are applied to extraction of software input factors, and the generation cycle of a value set is lowered; a combination coverage method is adopted to combine theinput factors in sequence according to the order of products of every two weights to generate a test case set with a priority order; in the testing process, if the test cases with high priorities areused first, a problem can be discovered earlier, and test resources can be saved; and by performing position improvement on limited combinations in the test cases, the problem that a constraint relation exists in the test case set is solved. Through the method, redundancy of the test cases is remarkably reduced, influences of all the factors on a software system are better checked with less test cases, and system efficiency is improved.

Description

technical field [0001] The invention relates to software testing technology. Background technique [0002] With the vigorous development of computer software technology, software testing technology has gradually become one of the popular science and technology. Software testing plays an extremely important part in software development and is one of the main means to ensure the quality of software development. The workload of software testing is generally very large. According to the statistics of a large number of testers, in the development cycle of a piece of software, the testing time will account for about 40%. In aerospace or other fields that require high software reliability , testing time will account for 60% of the development time. Normally, the software testing process can be divided into six stages: test planning, test design, test execution, test recording, test analysis, and test summary. The most important thing for a better test is the design and generatio...

AI Technical Summary

Problems solved by technology

However, according to previous scholars' proofs, N-dimensional combination tests and variable-strength combination tests have NP-Complete properties, and it is difficult to find the optimal solution. Many algorithms are still being studied, so most of the attention in engineering practice is pairwise combination tests Use case generation technology, in this regard, people use the orthogonal test design method of orthogonal matrix to design test cases, and use it in engineering, but because there are some special parameters for the input parameters when constructing the orthogonal matrix Requirements, this will cause more redundancy in the generated test case set; some scholars use heuristic algorithms to generate approximately optimal combined test case sets, such as the DDA algorithm, this method is in the process of constructing test case sets , introduces global density and local density, and uses a greedy algorithm in the assignment order and value selection of parameters. This method can generate a smaller test case set, so it also has its limitations. A smaller test case set will cause coverage. rate won't be too good

In addition, in the process of software testing, the test case set is relatively large. If each test case is executed in sequence, it will waste a lot of time and greatly affect the test efficiency.

Images