scrape.yml

name: scrape

on:
  push:
    branches:
      - main
  schedule:
    - cron: "0 4 * * *"

jobs:
  scrape:
    runs-on: ubuntu-latest
    steps:
      - name: Checkout
        uses: actions/checkout@v4

      - name: Setup Python
        uses: actions/setup-python@v5
        with:
          python-version: "3.11"
          cache: "pip"

      - name: Install Python dependencies
        run: python -m pip install -r requirements.txt

      # For some reason the 'overwrite' option for the Scrapy feed exporters
      # doesn't work reliably for me ¯\_(ツ)_/¯
      - name: Clear data file
        run: rm items.json

      - name: Scrape
        run: scrapy crawl czech-political-parties

      - name: Check that data got exported
        run: "[ -s items.json ] || exit 1"

      - name: Save to Git
        uses: EndBug/add-and-commit@v9
        with:
          add: items.json
          author_name: "scraper"
          author_email: "scraper@honzajavorek.cz"
          message: "update items 📥"
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}