docs: migrate documentation from Sphinx/Read the Docs to VitePress with GitHub Pages deployment

[nixel2007]

Summary

Migrate the documentation system from Sphinx/Read the Docs to VitePress with GitHub Pages deployment.

Changes

• Replace Sphinx/Read the Docs with VitePress for documentation
• Add GitHub Pages deployment workflow (triggered on main branch)
• Auto-generate API docs from Python source code using pydoc-markdown
• Add documentation deployment badge to README.md
• Configure VitePress with search, base path, and proper styling

[Copilot]

Pull request overview

Migrates the project documentation from Sphinx/Read the Docs to a VitePress-based site published via GitHub Pages, with API reference pages generated from the Python source using pydoc-markdown.

Changes:

• Add a docs generation script (pydoc-markdown → Markdown) and wire it into VitePress build/dev commands.
• Introduce a GitHub Pages deployment workflow for the VitePress site and update README badge/links.
• Remove the previous Sphinx/Read the Docs configuration and source files.

Reviewed changes

Copilot reviewed 20 out of 22 changed files in this pull request and generated 6 comments.

Show a summary per file

File	Description
scripts/generate_docs.py	New script to generate docs/api.md, docs/models.md, docs/exceptions.md via pydoc-markdown.
pyproject.toml	Ruff per-file ignore tweak for scripts/*.
package.json	Add docs scripts (generate/dev/build/preview) and VitePress dependency.
package-lock.json	Lockfile updates for VitePress and its transitive dependencies.
docs/usage.md	New Markdown usage page for VitePress docs.
docs/index.md	New VitePress home page frontmatter/content.
docs/.vitepress/config.mts	VitePress site configuration (nav/sidebar/search/base).
README.md	Switch documentation badge/link from Read the Docs to GitHub Pages.
.github/workflows/deploy-docs.yaml	New workflow to build and deploy docs to GitHub Pages.
.prettierignore	Ignore VitePress build artifacts and generated API docs.
.gitignore	Ignore VitePress build artifacts and generated API docs.
.readthedocs.yaml	Remove Read the Docs build configuration.
docs/requirements.txt	Remove Sphinx requirements.
docs/Makefile	Remove Sphinx Makefile.
docs/make.bat	Remove Sphinx Windows build script.
docs/__init__.py	Remove docs package marker (Sphinx-era).
docs/source/conf.py	Remove Sphinx configuration.
docs/source/index.rst	Remove Sphinx index page.
docs/source/usage.rst	Remove Sphinx usage page.
docs/source/api.rst	Remove Sphinx API module page.
docs/source/models.rst	Remove Sphinx models reference page.
docs/source/__init__.py	Remove docs source package marker (Sphinx-era).

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The usage example calls Tado("username", "password"), but Tado.__init__ does not accept username/password (it takes an optional refresh_token, etc.). This example won’t run and will mislead users; update it to match the actual public API (e.g., device-flow usage with await tado.async_init() / async with Tado() and/or passing refresh_token=...).

Suggested change

	async with Tado("username", "password") as tado:
	async with Tado(refresh_token="your_refresh_token") as tado:
	await tado.async_init()

[Copilot]

This paragraph mentions calling tado.close() when not using async with, but it doesn’t mention that callers also need to run the async initialization (await tado.async_init()) before making API calls. Add that guidance so the non-context-manager usage path is complete.

Suggested change

	Keep in mind that if you use `async with`, the `__aenter__` and `__aexit__` methods are called automatically. It's recommended you use session control; if you don't use `async with`, you need to call `tado.close()` to close the session.
	Keep in mind that if you use `async with`, the `__aenter__` and `__aexit__` methods are called automatically. It's recommended you use session control; if you don't use `async with`, you need to run `await tado.async_init()` before making any API calls and call `tado.close()` to close the session when you're done.

[Copilot]

subprocess.run passes -I src as a relative path, so the script will fail if invoked from a working directory other than the repo root. Consider deriving an absolute src path from __file__ (or setting cwd to the repo root) so doc generation works regardless of where it’s executed from.

[Copilot]

Path.write_text() is called without an explicit encoding. For consistent output across environments/CI, specify encoding="utf-8" (and optionally ensure the docs output directory exists before writing).

Suggested change

	output.write_text(content)
	print(f"Generated {output}") # noqa: T201
	def main() -> None:
	"""Generate all documentation pages."""
	pydoc_md = _find_pydoc_markdown()
	output.write_text(content, encoding="utf-8")
	print(f"Generated {output}") # noqa: T201
	def main() -> None:
	"""Generate all documentation pages."""
	pydoc_md = _find_pydoc_markdown()
	DOCS_DIR.mkdir(parents=True, exist_ok=True)

[Copilot]

The docs:preview script runs vitepress preview docs without ensuring the site has been built first. Since vitepress preview serves the build output, this should typically depend on docs:build (and/or docs:generate) to avoid preview failing on a clean checkout.

Suggested change

	"docs:preview": "vitepress preview docs"
	"docs:preview": "npm run docs:build && vitepress preview docs"

[erwindouna]

Is Copilot right here?

[Copilot]

This workflow installs pydoc-markdown via pip install directly, which bypasses the repo’s existing uv-based, locked dependency approach used in other workflows. To make docs builds reproducible, consider adding pydoc-markdown to the project’s dev dependencies and using uv sync/uv run here as well (or at least pin the installed version).

Suggested change

	run: pip install pydoc-markdown
	run: pip install pydoc-markdown==4.8.2

[erwindouna]

We don't use PIP, bu UV instead. :)

[erwindouna]

There are some merge conflicts now, can you please have a look? :)