Merge pull request #118 from spaceml-org/cleanup-pages-deploy

Clean up legacy ghp-import deploy path and restore FloodMapper logo

Author: will-fawcett-trillium

Makefile

@@ -1,4 +1,4 @@
-.PHONY: help install install-dev sync lock format lint type test check clean build publish build-jupyterbook clean-jupyterbook publish-docs
+.PHONY: help install install-dev sync lock format lint type test check clean build publish build-jupyterbook clean-jupyterbook
 .DEFAULT_GOAL = help
 
 PKGROOT = ml4floods
@@ -49,9 +49,6 @@ build-jupyterbook:  ## Build jupyter book
 clean-jupyterbook:  ## Clean jupyter book html
 	uv run jupyter-book clean jupyterbook
 
-publish-docs:  ## Publish docs to gh-pages
-	uv run ghp-import -n -p -f jupyterbook/_build/html
-
 ##@ Build & Publish
 
 clean:  ## Remove build artifacts

jupyterbook/README.md

@@ -1,34 +1,27 @@
 # Create docs
 
-Install `jupyter-book` and `ghp-import`. From the main directory run
+The documentation site is built with [Jupyter Book](https://jupyterbook.org/).
 
-```bask
-make build-jupyterbook
-# Check doc in local looks good
+Build it locally from the repository root:
 
-# Upload the doc to GitHub
-ghp-import -n -p -f jupyterbook/_build/html
+```bash
+make build-jupyterbook
 ```
-The commit should appear in branch `gh-pages` and the page will be live at [spaceml-org.github.io/ml4floods](https://spaceml-org.github.io/ml4floods)
 
-# Publish package in pip
+Then open `jupyterbook/_build/html/index.html` to review it.
 
-Install `twine`.
+Deployment is automatic: on every merge to `main`, the `deploy` GitHub Actions
+workflow builds the book and publishes it to GitHub Pages at
+[spaceml-org.github.io/ml4floods](https://spaceml-org.github.io/ml4floods).
 
-First update the version number in `ml4floods/__init__.py`
+# Publish package to PyPI
 
-```
-rm -rf build/
-rm -rf dist/
-python setup.py sdist bdist_wheel
+First bump the version number in `ml4floods/__init__.py`, then from the
+repository root:
 
-# upload to testpypi
-python -m twine upload --repository testpypi dist/*
-
-# Upload to real pypi
-python -m twine upload dist/*
+```bash
+make build      # build sdist + wheel (uv build)
+make publish    # upload to PyPI (uv publish)
 ```
-The package in PyPi will be shown at [pypi.org/project/ml4floods/](https://pypi.org/project/ml4floods/). The test version can be checked at [test.pypi.org/project/ml4floods/](https://test.pypi.org/project/ml4floods/)
-
 
-Follow [this tutorial](https://towardsdatascience.com/how-to-publish-a-python-package-to-pypi-7be9dd5d6dcd) to understand how pip works
+The package is listed at [pypi.org/project/ml4floods](https://pypi.org/project/ml4floods/).

jupyterbook/content/config.md

@@ -1,7 +1,7 @@
 # Introduction
 
 ---
-ML4Floods is python package to do *end-to-end* flood extent estimation from optical images using deep learning models.
+FloodMapper (also known as ML4Floods) is python package to do *end-to-end* flood extent estimation from optical images using deep learning models.
 
 ```{figure} ./ml4ops/ts_albania.gif
 ---

jupyterbook/ml4cc_logo.png

@@ -56,7 +56,7 @@ Homepage = "https://github.com/spaceml-org/ml4floods"
 [project.optional-dependencies]
 dev = ["ruff", "mypy", "pre-commit", "pyprojroot"]
 tests = ["pytest"]
-docs = ["jupyter-book<2", "ghp-import"]
+docs = ["jupyter-book<2"]
 serve = ["flask", "gunicorn"]
 cloudmasks = ["segmentation-models-pytorch", "s2cloudless"]
 kappazeta = ["tensorflow"]