internal-links-check.yml

name: Check internal links

on:
    schedule:
        - cron: '0 11 * * 1' # Every Monday at 11am UTC
    workflow_dispatch:
        inputs:
            save_results:
                description: 'Save results to file'
                required: false
                default: 'true'
                type: boolean
            send_slack_notification:
                description: 'Send Slack notification'
                required: false
                default: 'true'
                type: boolean

jobs:
    check-internal-links:
        runs-on: ubuntu-latest

        env:
            NODE_OPTIONS: '--max-old-space-size=8192 --expose-gc'
            GATSBY_CPU_COUNT: 1

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

            - name: Setup Node.js
              uses: actions/setup-node@v4
              with:
                  node-version: '22'
                  cache: 'pnpm'

            - name: Install dependencies
              run: pnpm install --frozen-lockfile

            - name: Build site
              run: pnpm build

            - name: Check links (console only)
              if: ${{ github.event_name == 'workflow_dispatch' && inputs.save_results == false }}
              run: |
                  OUTPUT=$(node scripts/check-links-post-build.js 2>&1)
                  echo "$OUTPUT"
                  echo "LINK_CHECK_OUTPUT<<EOF" >> $GITHUB_ENV
                  echo "$OUTPUT" >> $GITHUB_ENV
                  echo "EOF" >> $GITHUB_ENV

            - name: Check links (with file output)
              if: ${{ github.event_name == 'schedule' || (github.event_name == 'workflow_dispatch' && inputs.save_results != false) }}
              run: |
                  OUTPUT=$(node scripts/check-links-post-build.js link-check-results 2>&1)
                  echo "$OUTPUT"
                  echo "LINK_CHECK_OUTPUT<<EOF" >> $GITHUB_ENV
                  echo "$OUTPUT" >> $GITHUB_ENV
                  echo "EOF" >> $GITHUB_ENV

            - name: Upload link check results
              if: ${{ always() && (github.event_name == 'schedule' || (github.event_name == 'workflow_dispatch' && inputs.save_results != false)) }}
              id: upload-results
              uses: actions/upload-artifact@v4
              with:
                  name: link-check-results
                  path: link-check-results/*.json
                  if-no-files-found: warn
                  retention-days: 30

            - name: Send Slack notification
              if: always()
              run: |
                  # Default to true for scheduled runs, use input value for manual runs
                  SEND_SLACK="${{ inputs.send_slack_notification }}"
                  if [ "$SEND_SLACK" = "" ]; then SEND_SLACK="true"; fi

                  if [ "$SEND_SLACK" = "true" ] && [ -n "${{ secrets.SLACK_LINKS_CHECK_WEBHOOK }}" ]; then
                    WORKFLOW_URL="https://github.com/${{ github.repository }}/actions/runs/${{ github.run_id }}"
                    
                    if [ "${{ job.status }}" = "success" ]; then
                      COLOR="good"
                      EMOJI=":white_check_mark:"
                      STATUS="completed successfully"
                    else
                      COLOR="danger"
                      EMOJI=":x:"
                      STATUS="failed"
                    fi
                    
                    # Set triggered_by based on event type
                    if [ "${{ github.event_name }}" = "workflow_dispatch" ]; then
                      TRIGGERED_BY="manually run by ${{ github.actor }}"
                    elif [ "${{ github.event_name }}" = "schedule" ]; then
                      TRIGGERED_BY="Triggered via schedule"
                    else
                      TRIGGERED_BY="${{ github.actor }}"
                    fi
                    
                    # Extract statistics from the environment variable (suppress extra output)
                    if [ -n "$LINK_CHECK_OUTPUT" ]; then
                      MARKDOWN_FILES=$(echo "$LINK_CHECK_OUTPUT" | grep -o "Scanned [0-9]* markdown files" | grep -o "[0-9]*" | head -1 || echo "0")
                      TOTAL_INTERNAL_LINKS=$(echo "$LINK_CHECK_OUTPUT" | grep -o "Processed [0-9]* internal links" | grep -o "[0-9]*" | head -1 || echo "0")
                      EXCLUDED_LINKS=$(echo "$LINK_CHECK_OUTPUT" | grep -o "Found [0-9]* excluded links (skipped)" | grep -o "[0-9]*" | head -1 || echo "0")
                      REDIRECTED_LINKS=$(echo "$LINK_CHECK_OUTPUT" | grep -o "Found [0-9]* redirected links (skipped)" | grep -o "[0-9]*" | head -1 || echo "0")
                      BROKEN_LINKS=$(echo "$LINK_CHECK_OUTPUT" | grep -o "Found [0-9]* broken links" | grep -o "[0-9]*" | head -1 || echo "0")
                      BROKEN_ANCHORS=$(echo "$LINK_CHECK_OUTPUT" | grep -o "Found [0-9]* broken anchor links" | grep -o "[0-9]*" | head -1 || echo "0")
                    else
                      MARKDOWN_FILES="0"
                      TOTAL_INTERNAL_LINKS="0"
                      EXCLUDED_LINKS="0"
                      REDIRECTED_LINKS="0"
                      BROKEN_LINKS="0"
                      BROKEN_ANCHORS="0"
                    fi
                    
                    # Prepare variables for Slack payload
                    # Default to true for scheduled runs, use input value for manual runs
                    SAVE_RESULTS="${{ inputs.save_results }}"
                    if [ "$SAVE_RESULTS" = "" ]; then SAVE_RESULTS="true"; fi
                    
                    RESULTS_DOWNLOAD=""
                    if [ "$SAVE_RESULTS" = "true" ]; then
                      ARTIFACT_ID="${{ steps.upload-results.outputs.artifact-id }}"
                      if [ -n "$ARTIFACT_ID" ]; then
                        RESULTS_DOWNLOAD="https://github.com/${{ github.repository }}/actions/runs/${{ github.run_id }}/artifacts/$ARTIFACT_ID"
                      else
                        RESULTS_DOWNLOAD="$WORKFLOW_URL"
                      fi
                    fi
                    
                    curl -s -X POST "${{ secrets.SLACK_LINKS_CHECK_WEBHOOK }}" \
                      -H "Content-Type: application/json" \
                      -d "{
                        \"status\": \"$STATUS\",
                        \"emoji\": \"$EMOJI\",
                        \"repository\": \"${{ github.repository }}\",
                        \"triggered_by\": \"$TRIGGERED_BY\",
                        \"workflow_url\": \"$WORKFLOW_URL\",
                        \"markdown_files\": \"$MARKDOWN_FILES\",
                        \"links_checked\": \"$TOTAL_INTERNAL_LINKS\",
                        \"excluded_links\": \"$EXCLUDED_LINKS\",
                        \"redirected_links\": \"$REDIRECTED_LINKS\",
                        \"broken_links\": \"$BROKEN_LINKS\",
                        \"broken_anchors\": \"$BROKEN_ANCHORS\",
                        \"results_download\": \"$RESULTS_DOWNLOAD\",
                        \"save_results\": \"$SAVE_RESULTS\"
                      }" > /dev/null 2>&1
                    
                    echo "✅ Slack notification sent successfully"
                  else
                    if [ "$SEND_SLACK" = "false" ]; then
                      echo "Slack notifications disabled by user input"
                    else
                      echo "SLACK_LINKS_CHECK_WEBHOOK not configured, skipping Slack notification"
                    fi
                  fi

            - name: Report results
              if: always()
              run: |
                  # Default to true for scheduled runs, use input value for manual runs
                  SAVE_RESULTS="${{ inputs.save_results }}"
                  if [ "$SAVE_RESULTS" = "" ]; then SAVE_RESULTS="true"; fi

                  echo "Link check completed!"
                  echo "Check the job output above for detailed results."
                  if [ "$SAVE_RESULTS" = "true" ]; then
                    echo "Results files have been uploaded as artifacts."
                  fi