Python Dependency Scanner

A smart Python tool that automatically scans your Python projects, detects imports, generates requirements.txt, and sets up a virtual environment. It's especially useful for handling imports where package names differ from import names (like cv2 → opencv-python).

Key Features

• 🔍 Smart Import Detection: Recognizes common packages like cv2, mediapipe, tensorflow even if not installed
• 🗺️ Import Name Mapping: Handles cases where import names differ from package names
• 📝 Requirements Generation: Creates requirements.txt in your project directory
• 🔧 Virtual Environment: Automatically creates/reuses virtual environment
• 📦 Package Installation: Installs all required packages with confirmation
• 🔄 Encoding Support: Automatically handles different file encodings

Quick Start

# Install required package
pip install chardet

# Clone the repository
git clone https://github.com/science64/python-dependency-scanner.git
cd python-dependency-scanner

# Run the scanner (interactive mode)
python dependency_scanner.py

# Or specify directory directly
python dependency_scanner.py -i /path/to/your/project

Installation

1. Clone the repository:

git clone https://github.com/science64/python-dependency-scanner.git
cd python-dependency-scanner

2. Install required package:

pip install chardet

Usage Examples

Basic Usage

# Scan current directory
python dependency_scanner.py

# Scan specific project
python dependency_scanner.py -i /path/to/project

What It Does

When you run the scanner, it will:

1. Recursively scan all Python files in the directory
2. Detect imports like:
   • Standard imports (import numpy)
   • Aliased imports (import pandas as pd)
   • From imports (from PIL import Image)
3. Create requirements.txt
4. Set up virtual environment
5. Install packages after confirmation

Example Output

Scanning directory: /path/to/project

Found imports in main.py:
- opencv-python (from import cv2)
- mediapipe
- numpy
- python-dotenv

Creating requirements.txt...
Requirements file created at: /path/to/project/requirements.txt

Creating new virtual environment: .venv
The following modules will be installed:
- opencv-python
- mediapipe
- numpy
- python-dotenv

Do you want to proceed with the installation? (yes/no):

Smart Package Detection

The scanner recognizes packages in three ways:

1. 📗 Through the package mapping file
2. 🔍 By checking installed packages
3. 📦 Through a built-in list of common packages

Built-in Package Recognition

Common packages that are automatically recognized include:

• cv2 (opencv-python)
• mediapipe
• tensorflow
• torch
• pandas
• numpy
• PIL
• And more...

Package Name Mapping

The tool includes a package_mapping.json file that maps import names to package names. For example:

{
  "PIL": "Pillow",
  "cv2": "opencv-python",
  "sklearn": "scikit-learn",
  "bs4": "beautifulsoup4"
}

Adding Custom Mappings

You can add your own mappings:

1. Open package_mapping.json
2. Add new entries:

   {
     "your_import": "package-name"
   }

Project Structure

After running the scanner, your project will look like this:

your-project/
├── .venv/                  # Virtual environment
├── requirements.txt        # Generated requirements
├── your_python_files.py
└── ...

Debugging

If certain imports aren't being detected:

1. Run with -i flag to specify directory explicitly
2. Check if the Python file has proper .py extension
3. Verify file permissions
4. Ensure the file isn't in an excluded directory (like .venv)

Contributing

Contributions are welcome! Areas for improvement:

• Additional package mappings
• Support for more virtual environment configurations
• Enhanced import detection patterns
• Performance optimizations

License

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

Acknowledgments

• Thanks to the chardet library for encoding detection
• All contributors to the package mapping database
• Python community for package naming standards