SavePip

A tool to intelligently install and save clean package dependencies for both pip and conda environments.

Installation

pip install savepip

Usage

# Install packages with pip
savepip install pandas numpy

# Install packages with conda
savepip -m conda numpy pandas

# Save current environment
savepip save

# Upgrade packages
savepip -u requests pandas

# Save to custom file
savepip -o custom_requirements.txt requests pandas

Features

• Smart Dependency Management: Tracks installed packages in a memory file for future reference
• Requirements.txt Preservation: Maintains packages in requirements.txt even if not in memory
• Selective Package Tracking: Only includes packages explicitly installed with savepip or already in requirements.txt
• Synchronization: Automatically syncs memory with requirements.txt content
• Clean Output: Removes build hashes, development versions, and unnecessary information
• Package Manager Support: Works with both pip and conda environments
• Alphabetical Sorting: Dependencies are sorted alphabetically for readability

How It Works

SavePip maintains a memory of installed packages in .savepip/memory.json and uses this to:

1. Track packages explicitly installed with savepip
2. Preserve packages already in requirements.txt
3. Update package versions when they change
4. Generate clean requirements files

Unlike pip freeze, savepip only includes packages you've specifically chosen to track, avoiding bloated requirements files with unnecessary dependencies.

Category Management

SavePip supports organizing packages into categories, allowing you to manage different sets of dependencies for different purposes:

# Create a new category
savepip mk-category dev

# Switch to a category
savepip use-category dev

# Show current category
savepip cur-category

# List all categories
savepip ls-category

# Install packages to the current category
savepip install pytest coverage

# Save dependencies from specific categories
savepip save --categories dev,prod

This feature is useful for managing different types of dependencies (e.g., development, production, testing) within the same project.

License

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

Contribution

We welcome contributions! Feel free to submit issues or pull requests.