| Abstract [eng] |
Currently in software development more and more attention is paid to its quality. One of the most frequently used methods for quality assurance and improvement of software is testing. Since it is very time and money consuming process (might take over 50% of the total project time), everyone tries to automate it. One of the biggest problems of test automation is test data generation. To solve it, there is created a lot of different data generation methods, which are being constantly improved. Not all methods are equally suitable for generating test data for various applications. Because of this it was decided to investigate the suitability of several data generation methods for various specifics programs. For this work there was chosen and examined 4 dynamic methods: Random, Hill Climbing, Simulated Annealing and Genetic. It was also decided to compare the performance of the methods with different code coverage criteria (lines of code, instructions and branches). For this research was designed and developed Eclipse IDE plugin that allows generation of test data for selected Java class methods using data generation methods being researched. This plugin is also able to compare them with each other. In the conclusions presented experimental study results argues that the appropriate data generation method can save up to 43% of generation time. |
