When I gave a Python Packaging Course, I advised my students to always use Python 3. After all, Python 2 is dead. But one student said she had to use Python 2 as there is a package she definitely needs to use. And it's Python 2 only. What can one do in such a scenario?
I see the following options:
- Ask the maintainer to move the package
- Create a fork and move the package yourself
- Use an interface
Don't forget to be polite when you ask. After all, you want something. And you're most likely asking somebody who does this in their free-time.
When you have a program which has an interface (e.g. a web API like REST or a
command line interface), then you can execute it with Python 2 from Python 3.
Basically doing a system-level call, e.g. with
os.system. That is
rarely the case and pretty straight forward. So let's come to the interesting
part: Forking a package.
Naming
The package propy was
not maintained anymore and the original authors didn't reply. So I chose to
create a fork which I wanted to work with Python 3.
I thought about just keeping the name, but propy on PyPI
was already taken by an unrelated project. And, after all, making sure there
is some difference might not be the worst idea.
So I chose the name propy3 - a Python 3 compatible version of propy.
Structure
I want all my projects to have this structure:
propy3
├── docs
│ ├── Makefile
│ ├── requirements.txt
│ └── source
├── Makefile
├── MANIFEST.in
├── propy
│ └── __init__.py
├── README.md
├── setup.cfg
├── setup.py
├── tests
└── tox.ini
with a setup.py like this:
from setuptools import setup
setup()
and a setup.cfg like this:
[metadata]
# https://setuptools.readthedocs.io/en/latest/setuptools.html#configuring-setup-using-setup-cfg-files
name = propy3
version = 1.0.0a2
author = Dongsheng Cao
author_email = [email protected]
maintainer = Martin Thoma
maintainer_email = [email protected]
license = GPL
url = http://cbdd.csu.edu.cn/index
description = Compute protein descriptors
long_description = file: README.md
long_description_content_type = text/markdown
keywords = Bio-Informatics, Protein
classifiers =
License :: OSI Approved :: GNU General Public License (GPL)
Programming Language :: Python :: 3
Programming Language :: Python :: 3.6
Programming Language :: Python :: 3.7
Programming Language :: Python :: 3.8
Topic :: Scientific/Engineering :: Bio-Informatics
[options]
include_package_data = True
packages = find:
[tool:pytest]
addopts = --doctest-modules --mccabe --cov=./propy --cov-append --cov-report html:tests/reports/coverage-html --cov-report xml:tests/reports/coverage.xml --cov-report term --flake8 --ignore=docs/ --ignore=propy/__main__.py --durations=3 --black
doctest_encoding = utf-8
# Just temporarily: Increase from 10 to 25
mccabe-complexity=25
[pydocstyle]
convention=numpy
match_dir = propy
[flake8]
match_dir = propy
max-line-length = 80
select = C,E,F,W,B,B950
# Just temporarily: E731
ignore = E501,W503,E203, E731
[mypy]
ignore_missing_imports=true
check_untyped_defs=true
disallow_untyped_defs=false
warn_redundant_casts=true
warn_unused_configs=true
disallow_untyped_calls=false
follow_imports=skip
mypy_path=typeshed/pyi:typeshed/imports
General Cleanup
- Move documentation to
/docs - Remove generated files and add a
.gitignorefile - Apply
black - Apply
isort - Remove trailing spaces
I do this by putting the following .pre-commit-config.yaml file in the root
directory of the project:
# pre-commit run --all-files
repos:
- repo: https://github.com/pre-commit/pre-commit-hooks
rev: v2.3.0
hooks:
- id: end-of-file-fixer
- id: trailing-whitespace
- repo: https://github.com/asottile/seed-isort-config
rev: v1.9.3
hooks:
- id: seed-isort-config
- repo: https://github.com/pre-commit/mirrors-isort
rev: v4.3.21
hooks:
- id: isort
- repo: https://github.com/psf/black
rev: stable
hooks:
- id: black
language_version: python3.6
And then executing pre-commit run --all-files.
2to3
The tool 2to3 can help to make the move.
However, it is not optimal:
- It adds too many paranthesis for print sometimes
- It sometimes wraps zip / map / filter unnecessarily in a
list
Unit Tests
Setting up pytest and continuous integration (e.g. Travis) is pretty helpful
Sphinx and ReadTheDocs
Might help to make the tool discoverable.