Repository to capture best practice tooling for Python based developments.
Getting my initial thoughts down here about a manual process for setting up an environment using best practice approach. The idea being that it can then be migrated into a best practice pipeline for Azure DevOps.
Install Python, the latest version can be downloaded from python.org
Note - during the installation process choose the "Add Python X.X to PATH" option.
Check the installation by opening a terminal and running the following command:
python --version
Download and install Git from the official Git web site.
Check the installation of Git by running the following command:
git --version
There are some steps you need to take to get started with Git
The core ones being setting your git user and email address, this is required for pushing changes to a remote repository:
git config --global user.name "John Doe"
git config --global user.email [email protected]
Clone the template project from GitHub. From a local directory where we you want to host your repos, such as %HOMEPATH%\repos run the following command:
git clone <URL path to the repository>
Change directory into root of the cloned repo.
Use 'pip' (package installer for Python) to upgrade to the latest version of pip and pipenv:
python -m pip install --upgrade pip pipenv
Now lauch 'pipenv' shell - this will auotmatically create virtual environment:
pipenv shell
Now install dependencies using the --dev extensions so that development packages such as pytest, flake8 and the packaing tools are included. This installs the packages defined in the Pipfile and Pipfile.lock files.
pipenv install --dev
The following commands are useful to visualise the packages that are installed and the dependencies between them:
pipenv graph
pipend graph --reverse
For packages required in production use:
pipenv install <package name>
For packages only required for development use:
pipenv install <package name> --dev
The following tools are pre-packaged with this template project:
Install setuptools which is used to prepare a Python package:
pipenv install --dev setuptools
Install pipenv-setup which enables the the creation and management of setup.py file that governs how the Python code is packaged using setuptools:
pipenv install --dev pipenv-setup
To ensure the setup.py file is synchronized with the pipfile, issue the following command:
pipenv-setup sync
To create a packaged installation, run the following command:
pipenv run python setup.py sdist bdist_wheel
Install wheel which is used to zip up the package ready for deployment:
pipenv install --dev wheel keyring artifacts-keyring
There are some pre-requisites to the following command, such as setting up environemnt variables containing credentials - see https://twine.readthedocs.io/en/latest/.
Publish the wheel artefact:
pipenv run twine upload dist/* --non-interactive
Use flake8 to check the code meets consistent standards. Also recommend installing this as a Python linting plug-in for VS Code.
pipenv install flake8 --dev
Note : see entry in setup.cfg for linting configuration.
To run linting tests:
pipenv run flake8
Use pytest for automated testing:
pipenv install pytest --dev
Note : see entry in setup.cfg for test configuration. Current assumption is that tests will reside in .\test folder.
To run the tests:
pipenv run pytest
Add the following tooling to measure test coverage:
pipenv install pytest-cov --dev
The pytest-cov tool is governed by the following file .coveragerc file.
The following command will package up the Python package so that it can be published:
pipenv run python setup.py sdist bdist_wheel
The following command uses Twine to publish the package to the Azure Devops Artefacts repository so that it can be consumed by Python package installer pip or pipenv.
pipenv run twine upload dist/* --repository-url <Azure DevOps Artefect Repository URL>
See the configuration file .pre-commit-config.yaml. This will run a bunch of tests and checks before allowing code to be committed.