caches_cron_job.yml

name: Create caches for ephy_testing_data and conda env

on:
  workflow_dispatch: # Workflow can be triggered manually via GH actions webinterface
  push:  # When something is pushed into master this checks if caches need to re-created
    branches:
      - master
  schedule:
    - cron: "0 12 * * *"  # Daily at noon UTC

jobs:

  create-conda-env-cache-if-missing:
    name: Caching conda env
    runs-on: "ubuntu-latest"
    strategy:
      fail-fast: true
    defaults:
      # by default run in bash mode (required for conda usage)
      run:
        shell: bash -l {0}
    steps:
      - uses: actions/checkout@v3

      - name: Get current year-month
        id: date
        run: |
          echo "date=$(date +'%Y-%m')" >> $GITHUB_OUTPUT

      - name: Get current dependencies hash
        id: dependencies
        run: |
          echo "hash=${{hashFiles('**/pyproject.toml', '**/environment_testing.yml')}}" >> $GITHUB_OUTPUT

      - uses: actions/cache@v3
        # the cache for python package is reset:
        #   * every month
        #   * when package dependencies change
        id: cache-conda-env
        with:
          path: /usr/share/miniconda/envs/neo-test-env
          key: ${{ runner.os }}-conda-env-${{ steps.dependencies.outputs.hash }}-${{ steps.date.outputs.date }}

      - name: Cache found?
        run: echo "Cache-hit == ${{steps.cache-conda-env.outputs.cache-hit == 'true'}}"

      # activate environment if not restored from cache
      - uses: conda-incubator/setup-miniconda@v2.2.0
        if: steps.cache-conda-env.outputs.cache-hit != 'true'
        with:
          activate-environment: neo-test-env
          environment-file: environment_testing.yml
          python-version: 3.9

      - name: Create the conda environment to be cached
        if: steps.cache-conda-env.outputs.cache-hit != 'true'
        # create conda env, configure git and install pip, neo and test dependencies from master
        # for PRs that change dependencies, this environment will be updated in the test workflow
        run: |
          git config --global user.email "neo_ci@fake_mail.com"
          git config --global user.name "neo CI"
          python -m pip install -U pip  # Official recommended way
          pip install --upgrade -e .[test]

  create-data-cache-if-missing:
    name: Caching data env
    runs-on: "ubuntu-latest"
    steps:

      - name: Get current hash (SHA) of the ephy_testing_data repo
        id: ephy_testing_data
        run: |
          echo "dataset_hash=$(git ls-remote https://gin.g-node.org/NeuralEnsemble/ephy_testing_data.git HEAD | cut -f1)" >> $GITHUB_OUTPUT

      - uses: actions/cache@v3
        # Loading cache of ephys_testing_dataset
        id: cache-datasets
        with:
          path: ~/ephy_testing_data
          key: ${{ runner.os }}-datasets-${{ steps.ephy_testing_data.outputs.dataset_hash }}

      - name: Cache found?
        run: echo "Cache-hit == ${{steps.cache-datasets.outputs.cache-hit == 'true'}}"

      - name: Installing datalad and git-annex
        if: steps.cache-datasets.outputs.cache-hit != 'true'
        run: |
          git config --global user.email "neo_ci@fake_mail.com"
          git config --global user.name "neo CI"
          python -m pip install -U pip  # Official recommended way
          pip install datalad-installer
          datalad-installer --sudo ok git-annex --method datalad/packages
          pip install datalad
          git config --global filter.annex.process "git-annex filter-process"  # recommended for efficiency

      - name: Download dataset
        if: steps.cache-datasets.outputs.cache-hit != 'true'
        # Download repository and also fetch data
        run: |
          cd ~
          datalad install --recursive --get-data https://gin.g-node.org/NeuralEnsemble/ephy_testing_data

      - name: Show size of the cache to assert data is downloaded
        run: |
          cd ~
          pwd
          du -hs ephy_testing_data
          cd ephy_testing_data
          pwd