test-external-links.yml

name: Test external links

on:
  # Ideally this would be a scheduled action to catch if a website *did* work, but subsequently went offline
  # But then it would have to notify *someone*, whereas a PR author notification is easier
  pull_request:
     
jobs:
  test-external-links:
    runs-on: ubuntu-latest

    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with:
          node-version: 20
          cache: 'npm'

      - name: Install Lynx
        run: |
          sudo apt-get update
          sudo apt-get install -y lynx

      - name: Build documentation
        run: |
          npm i
          npm run-script build-local

      - run: |
          echo "temp_file=$(mktemp)" >> $GITHUB_ENV

      - name: Extract links
        run: |
          # Extract all unique URLs
          # Faster than potentially checking the same link on multiple pages
          find . -name "*.html" | while read -r file; do
              lynx -dump -listonly -nonumbers "${file}" | { grep -E "^http" || test $? = 1; } >> "${temp_file}"
          done

      - name: Check links
        run: |
          distinct_urls=$(sort -u "${temp_file}")

          while read -r url; do
              if [[ -n "${url}" ]]; then
                  echo "::debug::Checking URL '${url}'..."

                  status=$(curl --globoff --no-progress-meter --output /dev/null --location --head --write-out "%{http_code}" "${url}" || true)

                  if [[ "${status}" -eq 404 ]]; then
                      locations=$(grep -rl "${url}")
                      echo "::error::❌ URL '${url}' had status ${status} (found in ${locations})" 1>&2
                      found_error=1
                  else
                      echo "::debug::✅ URL '${url}' had status ${status}"
                  fi
              fi
          done <<< "${distinct_urls}"
          
          if [[ "${found_error}" -eq 1 ]]; then
              exit 1
          else
              exit 0
          fi