This is the reproduction package for Trident. It contains scripts to reproduce two examples chosen to demonstrate the difference in the evaluated configurations.
- Codeflaws 676-A-bug-18247191-18247237
- Coreutils c160afe
This package also path data for Codeflaws dataset.
eval- data and scriptscodeflaws_test- one example program from Codeflaws 676-A-bug-18247191-18247237coreutils_test- one example program from Coreutils c160afedata- path count data for Codeflawsscripts- scripts to generate plots
Dockerfile- Dockerfile to build a container for executing examplesTrident- Trident's source code
This section describes how to reproduce experiments with the two examples from Codeflaws and Coreutils
Building the Dockerfile
docker build . -t reproduction_package
Going into the docker container
docker run -v $(pwd):/home/Trident/ --rm -ti reproduction_package /bin/bash
The package helps in running the following tools
- Trident --> Mentioned in the paper
- Semfix --> semfix with support for l-value patches
- klee-merge --> Runs semfix with support for l-value patches and state merge with klee's
use-merge.
Running the sample codeflaws test with semfix
cd eval/codeflaws_test
python3 script.py 676-A-bug-18247191-18247237 --semfix
with klee
python3 script.py 676-A-bug-18247191-18247237 --klee-merge
running with Trident, specifying no option defaults it to Trident
python3 script.py 676-A-bug-18247191-18247237
You can run the patch prioritisation by using --priority.
For help run
python3 script.py -h
The <tool_name> can be replaced with trident, semfix or klee-merge
> cd eval/coreutils_test/test_c160afe
> ./run.sh <tool_name>
For example we can run klee-merge method with
> ./run.sh klee-merge
This command saves the plot in the file codeflaws_paths.png in the current directory
>python3 eval/scripts/plot_codeflaws_paths.py