jambazid-datar 0.0.58

Minimal working example of a packaged Python project. Key integrations include:

• Build System (Poetry).
• Test Framework (PyTest).
• Auto-docs with Markdown and ReST support (pdoc).
• CI/CD (GitHub Actions).

Contents

• Contents
• Structure
• Markdown and ReST support
• Data models
• Advanced tools

Structure

Documentation is generated from Python docstrings and takes a structure mirroring that of the source code (API documentation).

Initial focus could be on high-level (public) APIs intended for direct use by end-users.

Developers can thus grow the documentation organically within a single task-based workflow. As soon as a new feature/function is added - the developer need only include a docstring in their commit and it will reflect in the hosted page.

Markdown and ReST support

The pdoc framework renders Markdown in docstrings to HTML:

• Lists
• Are
• Supported

As are code blocks:

# Example invocation

result: Type = function(param="value", *variadic_args, **variadic_kwargs)

Data models

The pydantic framework has quickly become the leading data validation library for Python.

It's not a stretch to say that there's a whole ecosystem being built around it with everything from web frameworks to data science toolkits using it as the base for their data models.

From a documentation perspective, if you can describe your data source as a pydantic derived data model and then group them accordingly such links could be fronted as data schemas. Standard representations such as "JSON Schema" can be generated directly from pydantic models.

This also has advantages for testing.

Advanced tools

The pdoc framework is arguably the simplest auto-documentation tool for Python.

Below are some more advanced tools that are popular with large open-source projects:

• mkdocstrings
• Sphinx