Paired Code Smells and Test Smells: A Fine-Grained Longitudinal Empirical Study

This artifact accompanies the paper "Paired Code Smells and Test Smells: A Fine-Grained Longitudinal Empirical Study". Contents: ├── README.md ├── Dockerfile ├── run.sh # Shell script for batch execution ├── paired_smell_annotation.csv # Annotations for paired smells ├── unpaired_smell_annotation.csv # Annotations for unpaired smells ├── rules_summary.csv # Summary of rules ├── scripts/ │ ├── filtering.py # Processing the input scv by filtering for resolved smells │ ├── gitlog_plus_smell.py # Enriches smell data with Git commit logs │ ├── rq1_analysis.py # RQ1: Association Rule Mining │ ├── rq2_analysis.py # RQ2: Survival Analysis (KM Curves, Cox PH, Log-rank) │ ├── single_smell.py # Post-processing: Adds Git stats (LOC, files) to survival data │ └── export_git_log.py # Generate git log record └── merge_test_code_file.py # Merge test code and linking production code └── stat/ ├── counts_rq1.py # Aggregates RQ1 association rules from multiple files ├── counts_rq2.py # Aggregates RQ2 survival statistics ├── eval_direct_root_agreement.py # Evaluates agreement on manual vs. AI labeling ├── gpt-as-judge.py # Uses LLM to evaluate/judge removal reasons └── gpt_eval.py # Main script to query OpenAI for causal analysis