Python-to-Executable-Or-Module

This GitHub Action builds Windows, Mac, and Linux programs and extension modules from Python using the amazingly compatible nuitka Python compiler.

Key Features

• Build Standalone Executables - Build an executable from your Python code (standalone *.exe or *.bin file executables and even .app bundles for Mac)
• Build Binary Python Modules - Build binary *.pyd modules that are importable by other Python scripts
• Mac, Linux, and Windows - Support for Windows, Mac (including .app bundles), and Linux
• GUI Support - Supports GUIs made, for example, with TkInter** and Qt PySide6), and (PyQt6
• Lots More! - All the features of [Nuitka Python Compiler, including support for Nuitka Commercial Features like obfuscation, embedding data files, and more (for those with a license).

Simple to Use

See Usage Details below for more info.

1) Create a Python script

## hello_world.py
print("hello world!")

Run it in python

C:\> python hello_world.py
hello world!

2) Build an executable

Use this action as a step in one of your project's CI workflow jobs (details below):

# Build python script into a stand-alone exe
- uses: Nuitka/Nuitka-Action@main
  with:
    nuitka-version: main
    script-name: hello_world.py

3) Run the executable

C:\> hello_world.exe
hello world!

Current Limitations

• Not enough examples yet that demonstrate how to use this action in practice. Please help proving them.

Common traps

• Uploading artifacts should make sure include-hidden-files is present or else incomplete folders will be copied in case of .libs folders.

• Onefile is actually the default right now and needs to be deactivated.

Some Example Projects

Project	Example Workflow (YAML)
Node Editor GUI using Qt/Pyside6
Kasa TP-Link CLI App

Usage Details

See action.yml for details on how this action works under the hood. It is actually extremely simple.

Build a python script into an exe

See jimkring/test-nuitka-action/ for examples of this workflow in action.

jobs:

  build:
    # Windows is currently the only platform this action supports
    runs-on: windows-latest

    steps:

      # Check-out repository
      - uses: actions/checkout@v4

      # Setup Python
      - uses: actions/setup-python@v5
        with:
          python-version: '3.x' # Version range or exact version of a Python version to use, using SemVer's version range syntax
          architecture: 'x64' # optional x64 or x86. Defaults to x64 if not specified

      # Build python script into a stand-alone exe
      - uses: Nuitka/Nuitka-Action@main
        with:
          nuitka-version: main
          script-name: hello_world.py
          onefile: true

      # Uploads artifact
      - name: Upload Artifact
        uses: actions/upload-artifact@v4
        with:
          name: exe
          path: build/hello_world.exe
          include-hidden-files: true

GUI Builds

Similar to the others, but with enable-plugins: pyside6 or enable-plugins:tk-inter to ensure that those libraries are included correctly.

- name: Qt GUI with PySide6
  uses: Nuitka/Nuitka-Action@main
  with:
    nuitka-version: main
    script-name: my_qt_gui_app.py
    standalone: true
    enable-plugins: pyside6

- name: Python GUI With TkInter
  uses: Nuitka/Nuitka-Action@main
  with:
    nuitka-version: main
    script-name: my_tkinter_gui_app.py
    standalone: true
    enable-plugins: tk-inter

Multi-Platform Builds

Configure a runner of the appropriate operating system to build for a given platform. You can even do multiple platforms in a single workflow using a matrix strategy, as shown below:

Here is a workflow from the jimkring/kasa-cli project.

jobs:
  build:
    strategy:
      matrix:
        os: [macos-latest, ubuntu-latest, windows-latest]

    runs-on: ${{ matrix.os }}

    steps:
      - name: Check-out repository
        uses: actions/checkout@v4

      - name: Setup Python
        uses: actions/setup-python@v5
        with:
          python-version: '3.10' # Version range or exact version of a Python version to use, using SemVer's version range syntax
          architecture: 'x64' # optional x64 or x86. Defaults to x64 if not specified
          cache: 'pip'
          cache-dependency-path: |
            **/requirements*.txt

      - name: Install Dependencies
        run: |
          pip install -r requirements.txt -r requirements-dev.txt

      - name: Build Executable
        uses: Nuitka/Nuitka-Action@main
        with:
          nuitka-version: main
          script-name: kasa_cli
          onefile: true

      - name: Upload Artifacts
        uses: actions/upload-artifact@v4
        with:
          name: ${{ runner.os }} Build
          path: |
            build/*.exe
            build/*.bin
            build/*.app/**/*
          include-hidden-files: true

And here's what a resulting job run looks like:

https://github.com/jimkring/kasa-cli/actions/runs/2682890462

You can see that it creates executable binaries for Mac, Linux, and Windows.

Python and Package Dependencies

This action installs the following Python packages specified by the requirements.txt of this action repo.

ordered-set==4.1.0
    # via -r requirements.in
wheel==0.38.4
    # via -r requirements.in
zstandard==0.20.0

Value syntax

Since Action workflows accept no list values, for options that in Nuitka can be given multiple times, there is support for splitting those arguments by newline, which allows you to specify multiple values like this.

include-data-dir: |
  source_path_dir1=dest_path_dir1
  source_path_dir2=dest_path_dir2
  source_path_dir3=dest_path_dir3

Additional Documentation

See Nuitka for full documentation on Nuitka. It's a really fantastic tool!

License

Nuitka Action scripts and documentation in this project are under the MIT License.

Nuitka has the Apache 2.0 License

Python has the Python Software Foundation (PSF) License.

You are Responsible for Complying with your Project's Dependencies' Licenses

This tool compiles and copies your project's package dependencies (and their dependencies) into the output executable, which will be considered a combined or derivative work of those packages.

Important: You are responsible for compliance with the licenses of your project's package dependencies. Please consult with an attorney about your individual/project's compliance needs and strategy.

How to Comply With Dependency Package Licenses

There are some license checker tools that you might consider integrating with your project. Generally speaking, they enable you to specify which licenses (or types) are approved or disapproved and alert you whenever your project has a package dependency that is not approved.

Here is a list of license checker tools:

• python-license-check - can be run as a GitHub pre-commit hook.
• dependencies-license-compliance-checker - a github action that you can run before your executable build.