Replication package of the paper titled "Is PSO a valid option for search-based test case generation in the context of dynamically-typed languages?" published at TU Delft 2023.
The aforementioned paper introduces two Particle Swarm Optimization adaptations for the context of search-based test case generation for dynamically-typed languages. This repository contains all the necessary tools needed to replicate the experiment, including tools to visualize the results.
To select the configuration of the experiment and select the related parameters, the .syntest.json file should be accessed:
target-root-directoryspecifies the benchmark directory of the files, the available directories can be found hereincludespecifies the files for which the algorithm should try to generate tests, the available files can be found herepresetspecifies the algorithm, valid options areMOSAPSO,DynaMOSAPSO,DynaMOSA,MOSA, andNSGAII.total-time: represents the maximum search time of the algorithm.
In order to run the experiment, two bash scripts are provided build.sh and clean-build.sh the first one builds and runs the experiment, while the latter deltes and reinstall all the node_modules, then builds and runs the experiment. We suggest to always use the first file, unless new changes were pulled from any of the repositories.
To see the coverage over time of the different files tested, run the extract.py file insidethe plot/ directory, followed by the name of the folder containing the results. e.g, "python extract.py base"