linkcheck.yml

name: Link Checker [Anaconda, Linux]
on:
  schedule:
    # UTC 23:00 is early morning in Australia (9am)
    - cron:  '0 23 * * 1'
  workflow_dispatch:
jobs:
  link-checking:
    name: QuantEcon AI link checking
    runs-on: "ubuntu-latest"
    permissions:
      issues: write   # required for QuantEcon link-checker
    steps:
      # Download the latest release HTML archive (permanent, not subject to artifact expiry)
      - name: Get latest release asset URL
        id: release
        env:
          GH_TOKEN: ${{ github.token }}
        run: |
          ASSET_URL=$(gh api repos/${{ github.repository }}/releases/latest \
            --jq '.assets[] | select(.name | endswith(".tar.gz")) | .browser_download_url')
          echo "asset-url=$ASSET_URL" >> $GITHUB_OUTPUT
      - name: Download and extract release HTML
        run: |
          mkdir -p _site
          curl -sL "${{ steps.release.outputs.asset-url }}" | tar -xz -C _site
      - name: Pre-process HTML to exclude known bot-blocking domains
        env:
          # Domains that block automated crawlers (space-separated).
          # These are also listed in linkcheck_ignore in lectures/_config.yml.
          SKIP_DOMAINS: "fred.stlouisfed.org"
        run: |
          python3 - <<'PYEOF'
          import os
          import re
          from pathlib import Path
          skip_domains = os.environ.get('SKIP_DOMAINS', '').split()
          count = 0
          for html_file in Path('_site').rglob('*.html'):
              try:
                  content = html_file.read_text(encoding='utf-8')
              except UnicodeDecodeError as exc:
                  print(f"Skipping {html_file}: encoding error - {exc}")
                  continue
              for domain in skip_domains:
                  pattern = r'href="https?://' + re.escape(domain) + r'[^"]*"'
                  content = re.sub(pattern, 'href="#"', content)
              html_file.write_text(content, encoding='utf-8')
              count += 1
          print(f"Pre-processing complete ({count} files processed)")
          PYEOF
      - name: AI-Powered Link Checker
        uses: QuantEcon/action-link-checker@main
        with:
          html-path: '_site'
          fail-on-broken: 'false'
          silent-codes: '403,503'
          ai-suggestions: 'true'
          create-issue: 'true'