Various boilerplates used in almost all of my Python packages.
This package includes boilerplates for various common tasks in Python packages, such as building the package, testing the packaging process, storing the package config, logging for the package or creating a CLI.
Contents
Python version 3.8 or later.
Python libraries as specified in requirements.txt.
Building and running tests additionally requires packages listed in requirements_test.txt.
Tested on Linux, macOS and Windows.
Module boilerplates.setup
provides a class Package
that abstracts out
many of the common tasks needed to set up the package. The class has a
setup()
class method that can be called from the if __name__ == '__main__'
block
in your setup.py
file.
To avoid setup script boilerplate, create setup.py
file with the minimal contents as given
below and modify it according to the specifics of your package.
See the implementation of boilerplates.setup.Package
for all other available options.
Some fields don't need to be entered and will be automatically initialised using various detectors.
Also, some fields have default values.
See DEFAULT_*
constants in the boilerplates.setup
for those values.
Example setup.py
:
"""Setup script."""
import boilerplates.setup
class Package(boilerplates.setup.Package):
"""Package metadata."""
name = ''
description = ''
url = 'https://github.com/mbdevpl/...'
author = '...'
author_email = '...'
classifiers = [
'Development Status :: 1 - Planning',
'Programming Language :: Python :: 3.8',
'Programming Language :: Python :: 3.9',
'Programming Language :: Python :: 3.10',
'Programming Language :: Python :: 3.11',
'Programming Language :: Python :: 3 :: Only']
keywords = []
if __name__ == '__main__':
Package.setup()
You will also need the following in your pyproject.toml
file:
[build-system]
requires = ['boilerplates[setup] ~= <version>']
As an extension of setup boilerplate, there's an extra boilerplate for testing the packaging process,
in a way that enables 100% code coverage, including the setup.py
script.
In order to use it, all you need to do is create a file as follows in your tests directory.
Example test/test_packaging.py
:
"""Tests for packaging."""
import boilerplates.packaging_tests
class Tests(boilerplates.packaging_tests.PackagingTests):
pass
And, you will need to add the following to your requirements_test.txt
file (or equivalent):
boilerplates[packaging_tests] ~= <version>
Module boilerplates.config
provides few utility functions useful to handle local configuration.
Example usage:
import boilerplates.config
...
boilerplates.config.initialize_config_directory('app_name')
And, you will need to add the following to your requirements.txt
file (or equivalent):
boilerplates[config] ~= <version>
Assumptions for this boilerplate are that you want to use the standard built-in Python
logging module (logging
), and that your application probably has a CLI entry point
or some executable script, as opposed to only being a library.
To reduce boilerplate necessary to setup logging for such application,
add the following in your __main__.py
:
"""Entry point of the command-line interface."""
import boilerplates.logging
class Logging(boilerplates.logging.Logging):
"""Logging configuration."""
packages = ['package_name']
...
if __name__ == '__main__':
Logging.configure()
...
More advanced usage could be (just changing the Logging
class definition):
class Logging(boilerplates.logging.Logging):
"""Logging configuration."""
packages = ['package_name']
level_global = logging.INFO
enable_file = True
directory = 'package_name'
You can and should adjust the class fields to your needs, please take a look
at the boilerplates.logging.Logging
class implementation for details.
You may also use this boilerplate in tests even if your code is just a library. In such case,
add the following to your test/__init__.py
:
"""Initialization of tests."""
import logging
import boilerplates.logging
class TestsLogging(boilerplates.logging.Logging):
"""Logging configuration for tests."""
packages = ['package_name']
level_global = logging.INFO
TestsLogging.configure()
If you wish, you can make your test logging config be a variant of your application logging config, like so:
"""Initialization of tests."""
from my_package.__main__ import Logging
class TestsLogging(Logging):
"""Logging configuration for tests."""
level_global = logging.DEBUG # relevant if level_global is set to e.g. INFO in parent class
enable_file = False # relevant if enable_file is set to True in parent class
As for using the logging in your code, you can use it as usual, for example:
# in a standalone script:
_LOG = logging.getLogger(pathlib.Path(__file__).stem)
# in a standalone script that can also be imported:
_LOG = logging.getLogger(pathlib.Path(__file__).stem if __name__ == '__main__' else __name__)
# in __main__.py:
_LOG = logging.getLogger(pathlib.Path(__file__).parent.name)
# in usual module files:
_LOG = logging.getLogger(__name__)
And, you will need to add the following to your requirements.txt
file (or equivalent):
boilerplates[logging] ~= <version>
This boilerplate aims at simplifying the process of setting up Sentry integration for your Python application.
Assumptions for this boilerplate are similar to logging boilerplate, in that
you want to use the standard built-in Python
logging module (logging
), and that your application probably has a CLI entry point
or some executable script, as opposed to only being a library.
Then, the example __main__.py
file may look like:
"""Entry point of the command-line interface."""
import boilerplates.sentry
from ._version import VERSION
class Sentry(boilerplates.sentry.Sentry):
"""Sentry configuration."""
release = VERSION
...
if __name__ == '__main__':
Sentry.init()
...
You can and should adjust the class fields to your needs, please take a look
at the boilerplates.sentry.Sentry
class implementation for details.
And, you will need to add the following to your requirements.txt
file (or equivalent):
boilerplates[sentry] ~= <version>
This boilerplate aims at making CLIs easier to write, by providing a few utility functions.
Your example cli.py
file which defines your command-line interface may look like:
"""Command-line interface definition."""
import argparse
import boilerplates.cli
def main(args=None):
"""Entry point of the command-line interface."""
parser = argparse.ArgumentParser(
prog='my-cli',
description='''My command-line interface.''',
epilog=boilerplates.cli.make_copyright_notice(
2019, 2023, author='The Author', license_name='Apache License 2.0',
url='https://github.com/...'))
boilerplates.cli.add_version_option(parser, '1.0.1')
boilerplates.cli.add_verbosity_group(parser)
parsed_args = parser.parse_args(args)
verbosity = boilerplates.cli.get_verbosity_level(parsed_args)
...
You can see the above example in action in the examples.ipynb notebook.
Please see the boilerplates.cli
module for details of the available features.
And then, an example __main__.py
file may look like:
"""Entry point of the command-line interface."""
# PYTHON_ARGCOMPLETE_OK
from my_package import cli
if __name__ == '__main__':
cli.main()
And, you will need to add the following to your requirements.txt
file (or equivalent):
boilerplates[cli] ~= <version>
Then, the output of running python -m my_package -h
will look like:
usage: my-cli [-h] [--version] [--verbose | --quiet | --verbosity LEVEL]
My command-line interface.
options:
-h, --help show this help message and exit
--version show program's version number and exit
--verbose, -v be more verbose than by default (repeat up to 3 times for
stronger effect)
--quiet, -q be more quiet than by default (repeat up to 2 times for
stronger effect)
--verbosity LEVEL set verbosity level explicitly (normally from 0 to 5)
Copyright 2019-2023 by The Author. Apache License 2.0. https://github.com/...
And the output of running python -m my_package --version
will look like:
my-cli 1.0.1, Python 3.11.0 (main, Feb 13 2023, 00:02:15) [GCC 12.1.0]
This boilerplate aims at making it easier to test your package in a context of a git repository.
It's only useful if you create a Python package that operates on git repositories, and helps to create and modify synthetic git repositories for testing purposes.
To start using boilerplates.git_repo_tests
, you can start with a file like this
in your test folder, for example test/test_with_git_repo.py
:
"""Perform tests on and in synthetic git repositories."""
import pathlib
import boilerplates.git_repo_tests
class Tests(boilerplates.git_repo_tests.GitRepoTests):
...
However, you will need to check the boilerplates.git_repo_tests.GitRepoTests
class
for details of available features.
And, you will need to add the following to your requirements_test.txt
file (or equivalent):
boilerplates[git_repo_tests] ~= <version>