• Aucun résultat trouvé

The Study of the Sequential Inclusion of Paths in the Analysis of Program Code for the Task of Selecting Input Test Data

N/A
N/A
Protected

Academic year: 2022

Partager "The Study of the Sequential Inclusion of Paths in the Analysis of Program Code for the Task of Selecting Input Test Data"

Copied!
10
0
0

Texte intégral

(1)

The Study of the Sequential Inclusion of Paths in the Analysis of Program Code f or the Task of Selecting Input

Test Data

Konstantin E. Serdyukov[оооо-оооз-41б2-129sJ and Tatiana V. Avdeenko[оооо-0002-sб14-59З4J NovosiЬirsk State Technical University, К. Marks avenue 20, 630073 NovosiЬirsk, Russia

zores@live.ru

Abstract. The article proposes the results of а study evaluating one and many paths of program code, the work is not finished yet. То solve the proЫem of generating data sets, it is proposed to use а genetic algorithm for determining the complexity of program code for many paths. The proposed method allows to generate data for multiple code paths Ьу sequentially eliminating operations to achieve maximum possiЫe code coverage. For each generated data set the oper­

ations on the path are excluded from following generation, which allows gener­

ating new sets along other paths. The article describes the results of this method with а description of the data generation method and analysis of the generated data sets for many paths.

Keywords: genetic algorithщ test data, control flow graph, fitness function, testing process, program code, optimization method, test data generation, code coverage

1 Review of Existing Researches of Testing Automation and Selection of Test Data

Software engineering is а comprehensive, systematic approach to the development and maintenance of software. When developing programs, the following stages are most often distinguished - analysis, design, programming and testing. At the analysis stage, software requirements are determined and documentation is performed. At the design stage, the appearance of the program is detailed, its intemal functionality is determined, the product structure is developed, and requirements for subsequent testing are intro­

duced. Writing the source code of а program in one of the programming languages is done at the programming stage.

One of the most important steps in developing software products is testing. The im­

portant goals of testing are the compliance of the developed program with the specified requirements, adherence to logic in the data processing processes and obtaining сопесt final results. Therefore, for testing it is very important to generate input data, on the basis of which the program will Ье checked for eпors and compliance with specified requirements. То assess the quality of the input data, the code coverage indicator is used, that is, what percentage of the entire program can the test suites "cover". It is

Copyright © 2020 for this paper Ьу its authors. Use permitted under Creative Commons License Attribution 4.0 Intemational (СС ВУ 4.0).

79

(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)

Références

Documents relatifs

Based on the solution of this initial problem, we can further build an algorithm for finding test data set that provides coverage of the most complex branches (as many as

As NGS data were mainly used for the assembly of the whole nuclear genome, most assembly algorithms were designed to assemble all reads as quickly and accurately as

We present the fault finding results in Tables 2 and 3, listing for each case example, coverage criterion, test generation method, and oracle: the average fault finding for test

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des

However, this strategy would force us to define metamorphic relations for each operation meanwhile our approach can be used generically to generate test data for any analysis

We tested the signal observed by three different models of the mass structure of the deep earth mantle, and showed that they can be distinguished with a SNR of 30 dB with respect

Contextualization of on farm ex-ante evaluation of the sustainability of innovative cropping systems in viticulture, using a multiple criteria assessment

Table 3 summarizes our present knowledge of the two structure functions measured in unpolarized VCS: the MAMI result, and the preliminary JLab results obtained so far, both below