improve package management and some cleaning #17

Workflow file for this run

.github/workflows/install_on_colab.yml at 99343fb

	# This workflow tests if SpectrochemPy can be installed and run in a Google Colab environment
	# It uses Docker to simulate the Colab environment locally on GitHub's servers

	name: Test Colab Notebook

	# Define when this workflow should run
	on:
	push:
	branches:
	- develop
	- feature/*
	- fix/*
	pull_request:
	workflow_dispatch: # Allows manual trigger from GitHub UI

	# Prevent multiple copies of this workflow from running simultaneously
	# This saves resources and prevents conflicts
	concurrency:
	group: ${{ github.workflow }}-${{ github.ref }}
	cancel-in-progress: true

	jobs:
	test:
	name: Test notebook in Colab environment
	runs-on: ubuntu-latest # Use Ubuntu as our base system

	steps:
	# Step 1: Free up space on the GitHub runner
	- name: Maximize disc space
	uses: AdityaGarg8/remove-unwanted-software@v4.1
	with:
	remove-dotnet: "true"
	remove-android: "true"
	remove-haskell: "true"
	remove-codeql: "true"

	# Step 2: Get our code
	- name: Checkout code
	uses: actions/checkout@v4

	# Step 3: Set up Docker build system with caching support
	- name: Set up Docker Buildx
	uses: docker/setup-buildx-action@v3

	# Step 4: Cache Docker images to speed up subsequent runs
	- name: Cache Docker images
	uses: actions/cache@v3
	with:
	path: /tmp/.buildx-cache
	key: ${{ runner.os }}-buildx-${{ github.sha }}
	restore-keys: \|
	${{ runner.os }}-buildx-

	# Step 5: Cache pip packages to speed up installation
	- name: Cache pip packages
	uses: actions/cache@v3
	with:
	path: ~/.cache/pip
	key: ${{ runner.os }}-pip-${{ hashFiles('requirements/requirements.txt') }}
	restore-keys: \|
	${{ runner.os }}-pip-

	# Step 6: Download the official Google Colab Docker image
	# This image contains the same environment as Google Colab
	- name: Pull Colab image
	run: docker pull europe-docker.pkg.dev/colab-images/public/runtime:latest

	# Step 7: Create a test notebook
	# This creates a Jupyter notebook file that will:
	# 1. Install our package and its dependencies
	# 2. Import and test basic functionality
	# The notebook format is JSON with specific cells that will be executed
	- name: Create test notebook
	run: \|
	cat << EOF > test_install.ipynb
	{
	"cells": [
	{
	"cell_type": "code",
	"execution_count": null,
	"metadata": {
	"tags": ["parameters"]
	},
	"source": [
	"# This cell is for papermill parameters\n",
	"workspace_dir = '/workspace'"
	],
	"outputs": []
	},
	{
	"cell_type": "code",
	"execution_count": null,
	"metadata": {},
	"source": [
	"import sys\n",
	"import subprocess\n",
	"import importlib\n",
	"\n",
	"def install_package(path):\n",
	" # Install requirements\n",
	" requirements = path + '/requirements/requirements_colab.txt'\n",
	" subprocess.check_call([sys.executable, '-m', 'pip', 'install', '-q', '-r', requirements])\n",
	" # Install spectrochempy\n",
	" subprocess.check_call([sys.executable, '-m', 'pip', 'install', '-q', '--no-deps', path])\n",
	" print('Installation completed')\n",
	"\n",
	"install_package(workspace_dir)"
	],
	"outputs": []
	},
	{
	"cell_type": "code",
	"execution_count": null,
	"metadata": {},
	"source": [
	"# Force reload sys.path\n",
	"import site\n",
	"import sys\n",
	"sys.path.insert(0, '')\n",
	"importlib.reload(site)\n",
	"\n",
	"# Now import and test\n",
	"from zoneinfo import ZoneInfo # For datetime handling\n",
	"import spectrochempy as scp\n",
	"print('Development version:', scp.version)\n",
	"\n",
	"# Test basic functionality\n",
	"ds = scp.NDDataset([1,2,3])\n",
	"print('Created dataset:', ds)"
	],
	"outputs": []
	}
	],
	"metadata": {
	"kernelspec": {
	"display_name": "Python 3",
	"language": "python",
	"name": "python3"
	}
	},
	"nbformat": 4,
	"nbformat_minor": 4
	}
	EOF

	# Step 8: Run the test notebook in the Colab environment
	# This step:
	# 1. Creates a directory for pip cache
	# 2. Starts the Colab container with our code mounted
	# 3. Installs required tools (papermill)
	# 4. Runs our test notebook
	#
	# Docker run flags explained:
	# --rm: Remove container when it stops
	# -d: Run in detached mode (background)
	# -p 8888:8888: Map container's port 8888 to host
	# -v: Mount directories (workspace and pip cache)
	- name: Run notebook in Colab environment
	run: \|
	# Create pip cache directory
	mkdir -p ~/.cache/pip

	# Start the Colab container with pip cache mounted
	docker run --rm -d \
	-p 8888:8888 \
	-v ${{ github.workspace }}:/workspace \
	-v ~/.cache/pip:/root/.cache/pip \
	--name colab_container \
	europe-docker.pkg.dev/colab-images/public/runtime:latest

	# Wait for Jupyter to start
	sleep 10

	# Install papermill and ipykernel in the container
	docker exec colab_container \
	pip install -q papermill ipykernel

	# Execute the notebook
	docker exec colab_container \
	papermill /workspace/test_install.ipynb /workspace/output.ipynb \
	-p workspace_dir /workspace \
	--no-progress-bar

	# Stop container
	docker stop colab_container

	# Step 9: Save the results
	# This saves the executed notebook as an artifact
	# Artifacts can be downloaded from the GitHub Actions interface
	# This runs even if previous steps fail (if: always)
	- name: Upload notebook results
	if: always()
	uses: actions/upload-artifact@v4
	with:
	name: notebook-output
	path: output.ipynb

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

improve package management and some cleaning #17

Workflow file

improve package management and some cleaning #17

Jobs

Run details

Workflow file for this run