An archival version of this repository is available at https://zenodo.org/record/3262370#.XRgAoS2ZP4M
This repository provides the implementation of the final three steps of the automated framework presented in our FSE'2019 paper (included as ./PDF). The first step, Sample Generation, relies on the work done by Jeho et al. [31], so we do not include it here. so we do not include it here. Instead, we provide the results of that step in ./cases, and provide the rest of the framework, which runs off-the-shelf bug detectors, deduplicates their results, automatically extracts relevant features, and outputs those results in a unified format.
See INSTALL.md for instructions on running the experiments.
The scripts directory contains the artifacts that run the experiments. Namely, we provide two artifacts which encompass the final three steps of the automated framework presented in the FSE'19 paper. Namely, the run_tools artifact runs analysis tools on configurations of target programs, and the deduplicate_project artifact performs deduplication, formatting, and automatic feature identification of the warnings produced by the run_tools project. See INSTALL.md for instructions on how to run the two artifacts. See the READMEs listed in both tools' root directories for more specific information about their organization.
The cases directory contains both the source code and the sampled configurations of the three target programs we used in our study. Under each program's root directory, bugs/configs contains the 1000 sampled configurations. Each folder also includes a tarball containing the source code of the version of the code we are testing.