Test case generating method based on relationship

Abstract

The present invention relates a test example generating method based on relationship, comprising the following steps of: selecting the test point of each control of a graphical user interface to be tested, according to types; taking the value of each control according to a principle that all levels of each factor are taken and assembling the related controls completely, and then combining; generating a test example according to the reference relation of the function to be tested and the example template; combining the test examples to generate all test examples of the function to be tested. With boundary value testing and equivalence testing, the method carries out strict, normative and comprehensive test point design according to the common controls in an interface test. The test designing points of each type of controls are solidified into a group of test example generating criterion for citing and expanding. The example is designed with a control relation graph and a function relation graph according to the analysis of control relativity and function relativity, so that the selection of the input value of the test example has more pertinences, thereby being capable of greatly improving the efficiency and the quality of GUI test designing.

Description

technical field [0001] The invention relates to a software test case, in particular to a method for generating a test case based on a control relationship and a function relationship in a graphical user interface GUI. Background technique [0002] (1) Software testing [0003] Software testing in a broad sense includes the testing and verification of software to meet requirements at each stage of the software life cycle, such as whether the product meets the needs of users, product performance, ease of use, etc. [0004] Software testing in a narrow sense can be summarized as a test on the model: Y=F(X1, X2, ..., Xn), where X1, X2, ..., Xn are input variables, and F is the pair of variables X1, X2, ..., a transformation of Xn, Y is the desired result. Various test case design methods are actually different designs of the function F. [0005] In software, many variables may have infinitely many values, and due to resource constraints and engineering needs, software testing...

AI Technical Summary

Problems solved by technology

[0007] (2) Existing software test case design method

[0028] The test cases designed by the uniform test case design method take values ​​for each level of each factor, but the interaction between factors is not considered enough. For example, X1 and X2 are a pair of keywords, then X1 and X2 should be X2 takes repeated values ​​for testing, and the uniform use case design method cannot do this

[0029] Both the orthogonal test case design method and the uniform test case design method are derived from the experimental design. They are experimental methods designed to determine the influence of the relationship between various factors on the results. In the software test case design, the relationship between the various factors The interrelationship among them has been determined, and there are many factors that are independent of each other and do not affect each other. At this time, the above-mentioned method is used for test case design, and multiple cross-tests are still performed on these mutually independent factors, resulting in a large number of redundant test cases. Excessive or missing tests increase the cost and risk of testing

[0030] Each function in the software is often interrelated, and the orthogonal test case design method and the uniform test case design method cannot design use cases for the reference relationship between functions.

Images