Replication Package for "Beyond the YAML File: Understanding Real-World GitHub Actions Workflow Adoption"

This repository contains the replication package for the study "Beyond the YAML File: Understanding Real-World GitHub Actions Workflow Adoption" . It provides the code, data artifacts, and instructions necessary to reproduce and access the quantitative and qualitative analyses from the paper.

Repository structure

- data-pipelines-and-analysis/
  - data_pipeline/: End-to-end data collection, storage, and analysis pipeline.
    - collect_data.py: Entry point to collect data from the GitHub API.
    - config/: Configuration (API tokens, runtime options).
    - crawlers/: Modular crawlers for repositories, commits, pull requests, workflow runs, and jobs.
    - database/: DB connection and ORM-like models used during collection.
    - persistence/: Storage interfaces and stores.
    - data/: Lightweight helpers and logs for local data handling.
    - analysis/: Reproduction scripts for figures, tables, and statistics used in the paper. Includes figures/ and intermediate data/ CSVs used by scripts.
    - services/ and scripts/: Utilities and checks (e.g., data quality scripts).
    - README.md: Detailed setup, configuration, and execution steps for the pipeline and analyses.
  - manual_and_qualitative/: CSVs and notes for manual and qualitative analysis. See its README.md for details and suggested usage.

Quick start

Reproducing the pipeline and analysis:

1. Navigate to data-pipelines-and-analysis/data_pipeline/.
2. Create and activate a virtual environment.
3. Install requirements: pip install -r requirements.txt
4. Follow the instructions in data-pipelines-and-analysis/data_pipeline/README.md to configure credentials, run data collection, and execute analysis scripts in analysis/.

For qualitative artifacts and manual coding resources, see data-pipelines-and-analysis/manual_and_qualitative/README.md.

Reproducing figures and tables

The analysis/ folder inside data_pipeline/ contains scripts that generate the figures and summary tables used in the paper. Many scripts read from analysis/data/ and write outputs to analysis/figures/. Refer to each script's docstring and the data_pipeline/README.md.