|
This article is cited in 2 scientific papers (total in 2 papers)
Technologies for automatic testing of a software package for realistic computer graphics
E. Y. Denisov, A. G. Voloboy, E. D. Birukov, M. S. Kopylov, I. A. Kalugina Keldysh Institute of Applied Mathematics of RAS
Abstract:
The article describes the technology of automatic software testing in relation to industrial systems of computer graphics and optical simulation. Test automation becomes vital in the face of limited resources with the frequent release of product versions, which often occur among software product manufacturers. There are presented both methods of regression testing the computational kernel of such systems, and methods of testing the user interface. Scripting mechanism based on Python is used for regression testing, its multithreading capabilities which allow significant decreasing of testing time are also described. Python allows two ways of parallelization - multithreading and multiprocessing, both of them are considered. Due to the stochastic methods used in optical simulation calculation results may differ from time to time, which complicates regression testing. In this case, it is proposed to apply some (in each case - your own) threshold when comparing the simulation results. Separately automated testing of user interface which was elaborated basing on the AutoIt tool is described. The approach for testing the user interface of systems implemented in the form of plugins to existing CAD/PDM complexes, the source code of which is closed and not available to the authors of automatic tests, are described as well.
Keywords:
software testing, automatic testing, computer graphics, lighting simulation, software product robustness.
Citation:
E. Y. Denisov, A. G. Voloboy, E. D. Birukov, M. S. Kopylov, I. A. Kalugina, “Technologies for automatic testing of a software package for realistic computer graphics”, Proceedings of ISP RAS, 32:1 (2020), 71–88
Linking options:
https://www.mathnet.ru/eng/tisp486 https://www.mathnet.ru/eng/tisp/v32/i1/p71
|
Statistics & downloads: |
Abstract page: | 112 | Full-text PDF : | 52 | References: | 18 |
|