A robust tool for scraping, processing, and solving coding/math problems using AI-driven automation.
- Problem Crawling: Automatically scrape problems from online platforms using Selenium-based crawlers
- PDF Processing: Extract questions and metadata from PDF documents
- AI-Powered Solutions: Leverage Google's Generative AI to generate code/math solutions
- Smart Problem Selection: Filter and curate problems based on difficulty/subject using ML algorithms
- End-to-End Workflow: From data ingestion to solution generation in a single CLI interface
# Clone repository
git clone https://github.com/your-username/codemath-solver.git
cd codemath-solver
# Install dependencies
conda create --name codemath_solver python=3.10
pip install -r requirements.txt
# Configure settings
cp .env.example .env # Create config file# Run webUI with streamlit
streamlit run app.py --server.runOnSave true
# Run chrome debugging profile in port 9222
/Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome --remote-debugging-port=9222 --user-data-dir="/tmp/ChromeProfile"codemath-solver/
├── app.py # Main CLI entrypoint
├── codemath_solver/ # Core modules
│ ├── agent/ # AI interaction components
│ │ └── llm.py # Large Language Model interface
│ ├── crawl/ # Web scraping modules
│ │ └── core.py # Crawling engine
│ └── utils/ # Utility functions
│ ├── process_pdf.py
│ └── select_problems.py
├── conf/ # Configuration files
├── data/ # Input/output data storage
└── tests/ # Test suites
- Fork the repository
- Create feature branch:
git checkout -b feature/YourFeature - Implement changes with tests
- Lint code with
pre-commit run --all-files - Submit PR with clear description
Released under the MIT License. Feel free to modify and distribute.
Made with ❤️ using Python and AI technologies