Towards Supporting Open Source Library Maintainers with Community-Based Analytics

Overview

This replication package contains data and scripts used in our study. The package is structured into four main components.

Folder Structure

1. Data

• ProcessedData: Contains refined datasets that guide our research questions.

• RawData: Contains raw scraped data about dependents from GitHub, selected for analysis.

2. RepoClonerDataAnalyser

• The starting repository for the study.

• Select the top 10 libraries and their dependents.

• Clones repositories and analyzes all research questions.

• Implemented in Python.

3. methodTypeResolutionJavaParser

• A Java project used for method resolution.

• This tool is used for parsing and resolving method types after cloning repositories and filtering potential Java files using the RepoClonerDataAnalyzer project.

4. JacocoCoverageReporter

• Converts raw JaCoCo HTML coverage reports into CSV format.

• Implemented in Python 

5. Survey Forms

• Contains pdf version of survey forms that was shared to maintainers.

Usage Instructions

Each project within this package has its own README file with detailed setup and execution instructions. Below is a high-level guide:

1. Data Collection: Use RepoClonerDataAnalyser to select, clone, and filter dependents.

2. Method Resolution: Run methodTypeResolutionJavaParser on filtered Java files.

3. Coverage Analysis: Use JacocoCoverageReporter to convert JaCoCo HTML reports into CSV format and then Use RepoClonerDataAnalyser for further analysis. 

4. Data Analysis: Utilize the processed data in the Data folder for research insights.

Requirements

• Python 3.x

• Java 8+

• Required dependencies (listed in individual project README files. In version 1, you might notice a random GitHub repository URL provided in individual README. It is intended solely for context and clarity. It does not lead to an accessible resource and results in a 404 error. We have removed it in version 2 to avoid any confusion.)