GitHub dataset

The GitHub dataset consists of real-world PHP code samples collected from GitHub repositories, aimed at uncovering actual vulnerabilities through analysis.

The GitHub Blog has published a writeup about Bioclipse’s method for importing scripts from GitHub. Although there are obvious security issues with allowing arbitrary code to be run directly from the Web without a sandbox, this approach has some potential. GitHub has many other potential uses for chemistry.

The dataset contains info from 169 Open Source Software development GitHub Repositories. The dataset includes:Commits (along with commit messages)Issues (and issue discussions)Pull requests (and pull request discussions)Project detailsCollaboration characteristics and SonarQube violations

A Curated Dataset of GitHub Repository Metrics for Software Analytics Research

This dataset is extracted from GitHub and contains 172,919 java source codes written by 3,128 authors. It can be used for authorship attribution.