Negative Complement of a Set of Vulnerability-Fixing Commits: Supplementary Material

MSR 2024 Data and Tool Showcase Track (Submission)

This archive contains the supporting documents for the paper:
Rocío Cabrera Lozoya, Antonino Sabetta, Tommaso Aiello. "Negative Complement of a Set of Vulnerability-Fixing Commits: Method and Dataset"
submitted to the 2024 MSR Data and Tools Showcase Track.
Its structure is as follows:
- data
   - commit_pairs_final.csv : Dataset of 534 commit pairs corresponding to a positiive (security-relevant) commit and a negative sample obtained by the approach described in the paper.
   - single_positive.csv : Contains a single positive instance (taken from the MSR2019 dataset) which can be used to test the generate_negative_complement.py script.
- scripts
   - generate_negative_complement.py : Takes in a single .java file and obfuscates its developer-defined identifiers.
   - obfuscate_java_file.py : Generates a negative complement for a security-relevant dataset. The ouput is written to couple_dataset.csv.
   - requirements.txt : Requirements needed in the virtual environment to successfully run the previous scripts.