C++ Unit Test Generator using Ollama 🚀

This project is part of the Keploy API Fellowship - Day 5 Assignment.

It demonstrates how to use Ollama (running open-source Large Language Models locally) to automatically generate unit tests for a C++ application and optionally refine them based on feedback.

---

📝 Project Overview

• ✅ Takes existing C++ source code (main.cpp)
• ✅ Uses AI (via Ollama and Codellama model) to generate unit tests
• ✅ Stores generated tests in tests/test_main.cpp
• ✅ Compiles and runs the tests using g++

---

📂 Project Structure

cpp-unit-test-generator/ ├── main.cpp ├── generator.py ├── yaml_prompts/ │ └── generate_tests.yaml ├── tests/ │ ├── functions.cpp │ ├── functions.h │ └── test_main.cpp └── README.md

⚙️ How to Run

1. Install Ollama and pull the model:

2. llama pull codellama:7b

3. Generate Unit Tests: python generator.py

4. Compile and Run: g++ main.cpp tests/functions.cpp tests/test_main.cpp -o test_program.exe test_program.exe

📊 Coverage & Improvements

• Test cases are auto-generated covering positive, negative, and edge cases.
• Currently no code coverage tool integrated (can be added using gcov later).

🙌 Acknowledgements

• Ollama: For running open-source LLMs locally.
• Keploy API Fellowship: For providing this hands-on learning opportunity. GitHub: https://github.com/toughcodersir