Skip to content

Latest commit

 

History

History
78 lines (58 loc) · 2.96 KB

DEVELOPMENT.md

File metadata and controls

78 lines (58 loc) · 2.96 KB
Raw

Development

Project tooling and choices:

  • poetry - packaging and dependency management
  • poetry-dynamic-versioning - dynamic versioning based on tags and commits in git (poetry plugin)
  • poethepoet - project tasks runner that works well with poetry (poetry plugin)
  • tox - testing automation in different Python environments
  • black - code formatter
  • isort - imports sorter that works well with black
  • mypy - static typing checker
  • flake8 - linter for code and docstrings
  • docstrings style - Google (example)

Getting Started

Install main development dependencies (poetry+plugins, tox) to your system at once with

python3 -m pip install -U -r requirements-dev.txt

Then use poetry to manage all other dependencies in the underlying virtual environment:

# this step is optional and only required if you want to specify an interpeter explicitly
poetry env use 3.8

# install all dependencies and a project to a virtual environment
poetry install

poetry-dynamic-versioning temporary substitute current version in sources when the most of poetry commands are used, but with poetry poe changes are stay on the disk. This leads to unwanted "diff noise", so consider adding this variable to your environment to restrict allowed commands for dynamic versioning:

export POETRY_DYNAMIC_VERSIONING_COMMANDS="publish,build,version"

Code quality control

Common development tasks (lint, format, etc.) are collected in pyproject.toml file and can be run via poetry poe <task>. Run poetry poe -h to get a list with available tasks:

...
CONFIGURED TASKS
  clean          Clean project directory from temporary files and folders
  test           Run all tests
  test-cov       Run all tests with coverage collection
  lint           Invoke linting checks
  type           Invoke typing checks
  format         Format all Python files
  check-format   Check Python formatting
  pre-commit     Pre-commit routine to ensure code quality

Testing automation

Testing automation in different Python environments can be accomplished with tox. Usually, you don't run tox locally, because CI does this for you. However, some useful commands are described below:

# install python versions for testing
pyenv install 3.8.6 3.9 3.10 3.11

# run everything (tests on all interpeters, lint check, format check, etc.)
tox run

# list all jobs (environments)
tox l

# run specific environment
# here, tests on python3.11 + lint
tox -e py3.11 lint

# run specific environments under label `aux`: format, lint, type
tox run -m aux