graphlib backport

Backport of the Python 3.9 graphlib module for older Python versions.

Github: https://github.com/mariushelf/graphlib_backport

Supported versions

This backport currently supports Python 3.6, 3.7, 3.8, 3.9, 3.10, 3.11 and pypy3 (tested with pypy3.6, 3.7, 3.8 and 3.9).

Support for Python 3.6 and 3.7 is somewhat experimental and due to be dropped because of the end of support for these old versions.

Installation

pip install graphlib_backport

It is recommended to limit installation of this package to version of Python <3.9 and use the original implementation for Python >=3.9.

For example in a poetry pyproject.toml this can be done by specifying the dependency as a Python restricted dependency:

[tool.poetry.dependencies]
graphlib_backport = {version="^1.0.0", python="<3.9"}

Usage

The package works the very same way as the original package. Here's the documentation.

Development

The sourcecode is hosted on github. To develop on this package, just clone it, work on it and submit a pull request.

Dev requirements

For testing against different Python versions, tox is required.

To download the latest original sourcecode into the repository, there is a make target:

make download_sourcecode

Warning: This overwrites the code in this repo. By default it uses the tag v3.9.0, but you can overwrite that with a tag environment variable.

Running tests

As simple as running tox on the command line.

The executables for all python versions must be in the path, e.g, python3.6, ..., python3.9, pypy3. You can install them with pyenv.

Publishing a new version

Create a new tag and create a new release on Github. This will automatically publish the package to PyPI.

Changelog

See CHANGELOG.md

Thanks

I did not create this code -- I only repackaged it so it can be pip-installed into older versions of Python.

So all thanks go to the original contributors of the original sourcecode.

License

PSF

Changes to the sourcecode and the tests: replace incompatible code (e.g., the := walrus operator) with constructions compatible to earlier versions of Python, and make the tests pass on all supported versions.

Oven