AstroCluster is a framework-agnostic clustering tool specializing in semantic clustering of projects and technical debt, with a vision of supporting various languages and clustering paradigms.
The user can view the clustered file suggestions based on the MVC (Model View Controller) clustering paradigm in a graph.
They can also download the technical debt calculation either as a PDF or a CSV file, alongside with optional
parameters that contribute to the calculation (Average Lines of Code, Lines of Code per hour and Developer Salary per hour).
The user can also view a bar chart that depicts the percentages of each cluster in the project. This can prove useful for refactoring
larger projects by identifying large groups of similar files.
- Docker Compose (version ≥ 3.8)
- Python (for evaluating and fine-tuning models only, ignore for build process)
- In order to full build and deploy the project as it is, run the following script from the root of the project:
./build.shThis will build and deploy all services with the default configuration (skips no service). To customize the build process, the user can enter any combination of the following flags:
--skip-ui: Skips the build process of the client UI.--skip-server: Skips the build process of the backend server.--skip-grpc: Skips the automatic gRPC generation from proto files.--skip-ac: Skips the build process of the AstroCluster clustering service.--skip-db: Skips the build process of the database.
Feel free to contribute to the project (either creating new models, improving existing models or bugs) by opening issues here.
The project follows the Eclipse Public License v2.0 (EPL-2.0 license).
This tool was developed in the context of my Bachelor Thesis "AstroCluster: A Framework Agnostic Clustering Tool for Semantic Analysis of Projects and Technical Debt" under the supervision of Professor A. Ampatzoglou and N. Nikolaidis.