Glyph uses Machine Learning and Natural Language Processing to understand commit messages. This knowledge can be used for classifying commits into categories such as Bug-fixes, Feature additions, Improvements etc.
- Using Glyph with Kebechet, smart CHANGELOG entries out of commit messages can be generated.
- Glyph can also be used as a standalone library for analyzing commits from a locally stored repository (see usage below)
Currently Glyph ships with a model trained using Facebook’s fasttext library over a dataset of ~5000 commits collected from multiple large-scale open source projects (see referred publications for more details). The library can be easily extended to accommodate more models. Developers are welcome to contribute and improve the classification accuracy.
https://docs.google.com/presentation/d/1X3tnBngpIgN51jS7Rc0v9jwIlVfTUXpLJSHiKGmf1BY/edit#slide=id.p
Red Hat Intern: Tushar Sharma