Skip to content

README.md

Latest commit

 

History

History
263 lines (197 loc) · 7.12 KB

README.md

File metadata and controls

263 lines (197 loc) · 7.12 KB

GitProbe

License Python 3.8+ FastAPI

Turn any GitHub repository into comprehensive code analysis with interactive call graphs and multi-language support.

🚀 Features

  • Multi-language Analysis: Support for Python, JavaScript, TypeScript, Rust, Go, C, and C++
  • Tree-sitter Powered: Advanced syntax parsing with tree-sitter for accurate code analysis
  • Call Graph Generation: Interactive visualizations showing function relationships
  • Web API: RESTful API for integration with other tools and frontends
  • Real-time Analysis: Live progress tracking and results
  • Repository Insights: File structure, function counts, and relationship mapping
  • LLM-Ready Output: Structured JSON optimized for AI analysis

📸 Preview

GitProbe Preview

GitProbe's interactive call graph visualization showing function relationships and code structure analysis

📚 Requirements

  • Python 3.8+
  • Git (for repository cloning)
  • Internet access for GitHub repository analysis

📦 Installation

# Clone the repository
git clone https://github.com/your-org/gitprobe.git
cd gitprobe

# Create virtual environment
python -m venv env
source env/bin/activate  # On Windows: env\Scripts\activate

# Install dependencies
pip install -r requirements.txt

💡 Command line usage

Start the Web Server

# Start GitProbe server
./gitprobe server

# Server will be available at http://localhost:8000
# API documentation at http://localhost:8000/docs

CLI Analysis (Legacy)

# Analyze a GitHub repository
python -m gitprobe https://github.com/user/repository

# With custom output directory
python -m gitprobe https://github.com/user/repository --output ./analysis/

🌐 Web API Usage

Analyze Repository

# Start analysis
curl -X POST "http://localhost:8000/analyze" \
  -H "Content-Type: application/json" \
  -d '{"github_url": "https://github.com/psf/requests"}'

Python API Client

import requests

# Analyze repository
response = requests.post("http://localhost:8000/analyze", json={
    "github_url": "https://github.com/psf/requests",
    "include_patterns": ["*.py"],
    "exclude_patterns": ["*test*", "docs/"]
})

result = response.json()
print(f"Found {result['data']['summary']['total_functions']} functions")
print(f"Languages: {result['data']['summary']['languages_analyzed']}")

Example Response

{
  "status": "success",
  "data": {
    "summary": {
      "total_functions": 235,
      "total_calls": 657,
      "languages_analyzed": ["python"],
      "files_analyzed": 45
    },
    "functions": [...],
    "relationships": [...],
    "visualization": {
      "cytoscape": {...}
    }
  }
}

🧪 Testing

GitProbe includes a comprehensive integration test suite that validates all language analyzers:

# Install test dependencies
pip install rich

# Run quick tests (1 repo per language)
python tests/test_integration.py --quick

# Test all languages comprehensive
python tests/test_integration.py

# Test specific language
python tests/test_integration.py --language python

# Verbose output with detailed progress
python tests/test_integration.py --verbose

# JSON output for CI/CD
python tests/test_integration.py --json > results.json

Test Coverage

  • Python: rich, requests, flask, cpython
  • JavaScript: lodash, axios, express, node.js
  • TypeScript: vscode, typescript, angular
  • Rust: clap, ripgrep, rust compiler
  • Go: cobra, hugo, kubernetes
  • C: cJSON, libuv, curl
  • C++: fmt, catch2, protobuf

🏗️ Architecture

gitprobe/
├── src/gitprobe/
│   ├── analysis/           # Core analysis engine
│   │   ├── analysis_service.py
│   │   ├── call_graph_analyzer.py
│   │   └── repo_analyzer.py
│   ├── analyzers/          # Language-specific parsers
│   │   ├── python.py       # Python tree-sitter analyzer
│   │   ├── javascript.py   # JavaScript/TypeScript analyzer
│   │   ├── rust.py         # Rust analyzer
│   │   ├── go.py           # Go analyzer
│   │   ├── c_cpp.py        # C/C++ analyzer
│   │   └── ...
│   ├── web/               # FastAPI web server
│   │   └── server.py
│   └── models/            # Data models
│       └── ...
├── tests/                 # Integration test suite
│   ├── test_integration.py
│   └── README.md
└── requirements.txt

🎯 Language Support

Language Functions Calls Classes Imports Status
Python ✅ ✅ ✅ ✅ Stable
JavaScript ✅ ✅ ✅ ✅ Stable
TypeScript ✅ ✅ ✅ ✅ Stable
Rust ✅ ✅ ✅ ✅ Stable
Go ✅ ✅ ✅ ✅ Stable
C ✅ ✅ ❌ ✅ Stable
C++ ✅ ✅ ✅ ✅ Stable

🔧 Configuration

Environment Variables

# Optional: Custom server configuration
export GITPROBE_HOST=0.0.0.0
export GITPROBE_PORT=8000

Analysis Options

# Include/exclude patterns
{
  "github_url": "https://github.com/user/repo",
  "include_patterns": ["*.py", "*.js"],
  "exclude_patterns": ["*test*", "node_modules/", "__pycache__/"]
}

🤝 Contributing

Running Tests

# Start GitProbe server (in one terminal)
./gitprobe server

# Run integration tests (in another terminal)
python tests/test_integration.py --quick

Adding New Languages

  1. Create analyzer in src/gitprobe/analyzers/
  2. Add tree-sitter language dependency to requirements.txt
  3. Register analyzer in analysis service
  4. Add test repositories to tests/test_integration.py

Development Setup

# Install in development mode
pip install -e .

# Install development dependencies
pip install pytest black isort mypy

# Run code formatting
black .
isort .

🛠️ Stack

🐛 Known Issues

  • Large repositories (>1000 functions) are limited to 900 functions for performance
  • Some complex C++ template syntax may not parse correctly
  • Private repositories require local cloning

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

GitProbe - Comprehensive multi-language code analysis with interactive call graphs.