Test case selection method combining static dependence and dynamic execution rules

Abstract

The invention provides a test case selection method combining static dependence and dynamic execution rules. In order to select the test cases more accurately to further reduce the test scale, the invention provides a test case selection method of the method level, and the method collects more accurate dependencies by combining static dependency analysis and dynamic execution rules so as to selecta test set with a smaller accurate scale. The method comprises the following steps: firstly, positioning a change method in a submitted code, and collecting a method-level accurate dependency relationship through static analysis and dynamic execution rules; and designing a BFS-based incremental iterative search algorithm to search all affected methods of the change method set, and identifying anaffected test case set from the affected methods. According to the method, the problem that some necessary tests are omitted or some unnecessary tests are selected due to unreasonable dependence analysis of an existing static test case selection method can be solved, and the method has higher application value for frequent and continuous integration tests of a large complex system.

Description

technical field [0001] The invention belongs to the field of computer software testing, relates to continuous integration test case selection technology, studies how to collect accurate dependencies, and designs a search algorithm based on a breadth-first strategy (BFS), incrementally and iteratively searches out tests affected by code changes Use example set. Background technique [0002] Internet-based software development requires rapid iteration of products while maintaining product quality. Continuous Integration (CI) is a software development practice that solves this problem. Continuous integration requires developers to frequently integrate changed code into the mainline code base, and each integration verifies code quality through automated builds (including compilation and testing), thereby discovering integration errors as soon as possible. This development practice has been widely used in well-known software organizations and open source communities such as Goog...

AI Technical Summary

Problems solved by technology

[0007] 1. Omission of dependencies leads to omission of necessary tests, such as inability to handle method call relationships in reflection;

[0008] 2. Inaccurate dependencies, resulting in unnecessary tests, such as the dynamic binding call relationship in inheritance cannot be accurately constructed

[0009] The above two problems are mainly due to the fact that some dependencies are dynamically determined at runtime and cannot be completely obtained through existing static analysis techniques.

Images