Skip to content

Latest commit

 

History

History
101 lines (76 loc) · 6.09 KB

CONTRIBUTING.md

File metadata and controls

101 lines (76 loc) · 6.09 KB
Raw

Tan-chan contributing guidelines

First let me thank you in advanced for any contributions you may make to Tan-chan; contributions are always welcome and any amount of time spent on this project is greatly appreciated.

But before you get started contributing, it's recommended that you read through the following guide in-order to ensure that any pull-requests you open can be at their best from the start.

Pipelines

The most important thing to consider while contributing towards Tan-chan is the checks the library's run against. While these are run against all PRs by Github Actions, you can run these locally before any PR is opened using Nox.

To run the tests and checks locally you'll have to go through the following steps.

  1. Ensure your current working directory is Tan-chan's top-level directory.
  2. pip install -r nox-requirements.txt to install Nox.
  3. Use nox -s to run the default tasks.

A list of all the available tasks can be found by running nox -l with blue names being the tasks which are run by default when nox -s is called alone. To call specific tasks you just call nox -s name1 name2 where any number of tasks can be called at once by specifying their name after -s.

It's worth noting that the reformat nox task (which is run by default) will reformat additions to the project in-order to make them match the expected style and that nox will generate virtual environments for each task instead of pollution the environment it was installed into.

Documentation style

This project's documentation is written in numpydoc style and should also use styles which are specific to pdoc.

A few examples of pdoc style would be:

  • Links: Unlike sphinx, regardless of whether you're linking to a module, class, function or variable the link will always be in the style of [link.to.thing][] with no type information included.
  • Documenting fluent methods: The return type for fluent methods should be given as Self with the description for it following the lines of something like "the {x} instance to enable chained calls".
  • Union types should be annotated in code and documentation using | style and T | None/T | hikari.UndefinedType are preferred over T | None/hikari.UndefinedOr[T]

CHANGELOG.md

While you aren't required to update the changelog as a contributor, a reference on the schema CHANGELOG.md follows can be found here.

It should be noted that not all changes will be included in the changelog (since some are just not significant enough) and it comes down to a maintainer's discretion as to what is included.

Tests

All changes contributed to this project should be tested. This repository uses pytest and nox -s test for an easier and less likely to be problematic way to run the tests.

Type checking

All contributions to this project will have to be "type-complete" and, while the nox tasks let you check that the type hints you've added/changed are type safe, pyright's type-completness guidelines and standard typing library's type-completness guidelines are good references for how projects should be type-hinted to be type-complete.

NOTES

  • This project deviates from the common convention of importing types from the typing module and instead imports the typing module itself to use generics and types in it like typing.Annotated.
  • Since this project supports python 3.11+, the typing types which were deprecated by PEP 585 should be avoided in favour of their collections.abc, builtin, re and contextlib equivalents.
  • The standard approach for using collections.abc types within this project is to import collections.abc as collections.
  • All exported symbols should have docstrings.

Versioning

This project follows semantic versioning 2.0.0 and PEP 440.

General enforced style

  • All modules present in Tan-chan should start with the commented out licence (including the source encoding and cython language level declarations), a relevant component documentation string, from __future__ import annotations, an __all__ declaration and then imports. For an example see any of Tan-chan's current components.
  • Only callback type variables (e.g. CommandCallbackSig = collections.Callable[..., collections.Awaitable[None]]) should public, meaning that they should also be included in the __all__ of the module they're declared in and should also be documented but should not included in the __all__ of any parent module(s). Any other type variables should be protected (prefixed with _).
  • pep8 should be followed as much as possible with notable cases where its ignored being that black style may override this.
  • The maximum character count for a line is 120 characters and this may only ever be ignored for docstrings where types go over this count, in which case a # noqa: E501 should be added after the doc-string (on the same line as its trailing """.
  • All public top-level modules should be explicitly imported into tanchan.__init__ and included in tanchan.__init__.__all__.
  • for type-completness with only the most important of their contents needing to be included in tanchan.__init__.__all__. Note here that star imports should not be used; import each entry you want to re-export explicitly and let isort reformat the imports.
  • When an abstract class is defined outside of tanchan.abc (meaning that its not a part of the standard interface), its name should be prefixed with Abstract.
  • collections.abc.Awaitable is generally preferred over collections.abc.Coroutine for interfaces and types.
  • Explicit | unions are preferred over using generic union shorthands (e.g. UndefinedOr, Optional, SnowflakeishOr) in documentation.