etl.yaml

name: Auto-update

on:
  schedule:
    - cron: "0 * * * *"
  workflow_dispatch:

jobs:
  update:
    runs-on: ubuntu-latest

    steps:
    - name: Checkout code
      uses: actions/checkout@v4
      with:
        ref: main

    - name: Install dependencies
      uses: ./.github/actions/setup-python-poetry
      with:
        python-version: "3.12"
        poetry-version: "1.8.3"
        poetry-install-args: --without dev

    - name: Run ETL Process
      id: etl
      run: |
        source .venv/bin/activate
        python webapp/update/etl.py

    - name: Check for changes in data/*.csv
      id: check_changes
      run: |
        changes=$(git diff --name-only | grep '^data/.*\.csv$') || true
        untracked=$(git ls-files --others --exclude-standard | grep '^data/.*\.csv$') || true

        if [ -n "$changes" ] || [ -n "$untracked" ]; then
          echo "has_changes=true" >> $GITHUB_ENV
        else
          echo "has_changes=false" >> $GITHUB_ENV
        fi

    - name: Commit and push changes
      if: env.has_changes == 'true'
      run: |
        git config user.name 'GitHub Actions'
        git config user.email 'actions@github.com'

        echo $(date +%s) > data/metadata
        git add data
        git commit -m "chore: update data"
        git status
        git push