SFML 2.6.2.0

Python bindings for SFML

PySFML

PySFML provides Python bindings for SFML.

The current package targets SFML 2.6.2 with modern Python packaging, pytest-based validation, and wheel builds.

PySFML is distributed under the terms of the zlib license.

Installation

Install the published package with pip:

python -m pip install SFML

Binary wheels are the preferred installation path.

Current Baseline

• Supported Python versions: 3.10+
• Bound SFML release: 2.6.2
• Packaging baseline: pyproject.toml, setuptools, setuptools-scm, Cython 3.2.x
• Validation baseline: pytest smoke coverage for all compiled modules plus curated window and graphics coverage under Xvfb on Linux

Quick Example

The following PySFML program is the direct Python-style equivalent of the classic SFML C++ example that opens a window, displays a sprite and some text, and plays music:

from sfml import audio, graphics, window


render_window = graphics.RenderWindow(
	window.VideoMode(800, 600),
	"PySFML window",
)

texture = graphics.Texture.from_file("cute_image.jpg")
sprite = graphics.Sprite(texture)

font = graphics.Font.from_file("arial.ttf")
text = graphics.Text("Hello PySFML", font, 50)

music = audio.Music.from_file("nice_music.ogg")
music.play()

while render_window.is_open:
	for event in render_window.events:
		if event.type == window.EventType.CLOSED:
			render_window.close()

	render_window.clear()
	render_window.draw(sprite)
	render_window.draw(text)
	render_window.display()

The main differences from the C++ form are:

• resource loading is constructor-style via from_file(...)
• event kinds use window.EventType
• the event loop is exposed as render_window.events
• module imports are the preferred style instead of a single monolithic namespace

See docs/getting-started.md for source-build prerequisites and the local build flow.

Package Layout

The bindings are organized by SFML module:

• sfml.system
• sfml.window
• sfml.graphics
• sfml.audio
• sfml.network

The sfml.sf module remains available as a convenience import that re-exports the full binding surface, but the module-specific imports are the primary structure of the project.

See docs/modules.md for a module-by-module orientation. See docs/compatibility.md for Python-specific behavior differences, unsupported features, and known parity gaps.

Examples

The examples tree is curated around maintained upstream-aligned examples only.

See examples/README.md for the current examples policy and classification.

Development

Local source-build prerequisites and test commands are documented in docs/getting-started.md.

Packaging metadata and test configuration live in pyproject.toml.

The CI and release workflows referenced below live in .github/workflows/.

CI And Releases

GitHub Actions is centered on:

• editable-install validation on Linux against a pinned SFML 2.6.2 build
• packaging validation for both wheel and sdist artifacts
• smoke-install checks on Linux, macOS, and Windows
• curated typing validation using mypy
• tag-driven wheel and sdist builds via cibuildwheel and python -m build

Release publishing is restricted to v* tags:

• prerelease tags containing a, b, or rc publish to TestPyPI
• other v* tags publish to PyPI

Release artifacts are validated before publish with .github/scripts/validate-dist.py and twine check, and the wheel builds run the deterministic wheel-safe pytest subset configured in pyproject.toml.

There is also a manual TestPyPI Validation GitHub Actions workflow for installing an already-uploaded TestPyPI version on fresh Linux, macOS, and Windows runners.

Documentation

• docs/getting-started.md: installation and source-build guidance
• docs/recipes.md: minimal task-oriented snippets for common PySFML patterns
• docs/modules.md: module overview and import strategy
• docs/compatibility.md: support boundaries, known differences, and unsupported features
• examples/README.md: maintained examples policy and classification
• pyproject.toml: packaging metadata and test configuration
• .github/workflows/: CI, release, and TestPyPI validation workflows

The repository source of truth is GitHub:

• Repository: https://github.com/intjelic/python-sfml
• Issues: https://github.com/intjelic/python-sfml/issues

Author

Jonathan De Wachter [e-mail]