Source code level program variation generation method and device

Abstract

The invention discloses a source code level program variation generation method and device. According to the method, a to-be-varied function is obtained, an abstract syntax tree is constructed, and literal constants and object classes capable of being called by variation are extracted through analysis of the abstract syntax tree. Then the abstract syntax sub-trees of the function to be varied is traversed, if the nodes are variables, whether the variables have variable conditions are judged, and if the variables have variable conditions, corresponding variation information is constructed to form a variation set; the variation information comprises variation statements. The variation statements are divided into literal constant assignment statements of the basic class variables and statements of the object class methods capable of being called for variation calling of the object classes. And finally, a corresponding variation function is constructed for each piece of variation information, and a variation response code is inserted into the original function to be varied. The variation response code calls a corresponding variation function through an externally acquired variation ID. In this way, during testing, which variation function is called is determined through external configuration.

Description

technical field [0001] The invention relates to software testing, in particular to variation testing at the level of software source codes. Background technique [0002] It is an important indicator of software testing to evaluate whether the testing process of software program is perfect. In software testing, a good testing process should be able to perceive subtle changes in the program under test. If the testing process cannot perceive subtle changes in the program under test, it indicates that the testing process is flawed, or the testing is not perfect. [0003] Mutation testing is a testing process evaluation technique based on the above principles. The principle of this technology is to produce slight variations to the original program, and then test through the testing process to determine whether the testing process can perceive such subtle variations. [0004] Existing program mutation tools, such as PIT and Major, inject mutations in the intermediate code stage...

AI Technical Summary

Problems solved by technology

[0005] The problem to be solved by the present invention: the variation injected in the intermediate code stage cannot be read and understood by developers, and it is not helpful for the positioning of defect problems

Images