Skip to content

Latest commit

 

History

History
62 lines (41 loc) · 3.21 KB

README.md

File metadata and controls

62 lines (41 loc) · 3.21 KB
Raw

pip-clean

This is AI generated, may not be 100% accurate.

pip-clean is a Python tool for identifying unused dependencies in your project. It analyzes your project's manifest file (e.g., Pipfile, requirements.txt) and compares it with the actual imports used in your Python code. The tool helps keep your dependency list lean and secure by highlighting dependencies that can be removed.

Features

  • Support for Multiple Manifest Types: Works with Pipfile (beta), requirements.txt (experimental), and pyproject.toml (coming soon).
  • Accurate Analysis: Cross-references the dependencies in your manifest with actual imports in your Python files.
  • Simple CLI Interface: Easy-to-use command-line tool for efficient dependency management.

Usage

  1. Clone the repository:
    git clone https://github.com/yourusername/pip-clean.git
    cd pip-clean

Run the tool using the following format:
uv run python pip_clean/clean.py <manifest_type> -m <manifest_file> -t <target_directory>

Arguments

  • <manifest_type>: The type of manifest file to analyze. Supported values are:
    • Pipfile (beta)
    • requirements.txt (experimental)
    • pyproject.toml (coming soon)
  • -m, --manifest: Path to the manifest file (e.g., Pipfile or requirements.txt).
  • -t, --target: Path to the root directory of your Python project containing the codebase to analyze.

Example

To analyze a project using a Pipfile located at /tmp/projects/example:
uv run python pip_clean/clean.py pipfile -m /tmp/projects/example/Pipfile -t /tmp/projects/example/

Output

  • Unused Dependencies: The tool will list any dependencies in the manifest file that are not used in the codebase. For example:
    Unused dependencies:
    - flask
    - requests

  • No Unused Dependencies: If all dependencies are in use, it will output:
    No unused dependencies found.

Known Limitations

While pip-clean is a helpful tool, there are some limitations to be aware of:

  1. Mismatch Between Package and Import Names: Some Python packages are installed with a name that differs from the one used to import them in code. For example:

    • Package name: python-dateutil
    • Import name: dateutil This can result in false positives where a dependency is marked as unused because its import name doesn’t directly match the package name.

  2. Dynamic Imports: The tool does not detect dynamically generated imports, such as those created using importlib or __import__. These imports might not be visible in the static analysis.

  3. Indirect Dependencies: Dependencies that are imported indirectly (e.g., through another library) are not analyzed. The tool focuses only on imports explicitly used in your code. Pip-clean expects requirements.txt and Pipfile files to include direct dependencies only. Dev dependencies and other non-production dependencies are also ignored.

  4. Manifest Type Limitations: Currently, Pipfile is the most stable format supported. Analysis for requirements.txt is experimental, and support for pyproject.toml is planned but not yet implemented.

Contributing

Contributions are welcome! Feel free to open an issue or submit a pull request if you have suggestions or improvements.