name: 02.1. Intelligent Link Cleaner & Dedup

on:
  schedule:
    - cron: '0 0 1 */3 *' # Quarterly: 1st of Jan, Apr, Jul, Oct
  workflow_dispatch:
    inputs:
      force_full_check:
        description: 'Force full re-validation (bypasses 21-day cache)'
        type: boolean
        default: false

permissions:
  contents: write
  pull-requests: write
  issues: write

concurrency:
  group: link-cleaner-${{ github.ref }}
  cancel-in-progress: true

jobs:
  intelligent-clean-sequential:
    runs-on: ubuntu-latest
    env:
      FORCE_JAVASCRIPT_ACTIONS_TO_NODE24: true
    steps:
      - name: Repository Synchronization
        uses: actions/checkout@v6
        with:
          ref: develop

      - name: Python 3.11 Environment Provisioning
        uses: actions/setup-python@v6
        with:
          python-version: '3.11'
          cache: 'pip'

      - name: Dependencies Installation
        run: |
          python -m pip install --upgrade pip
          pip install --no-cache-dir pydantic PyGithub aiohttp beautifulsoup4 httpx fake-useragent pytz python-dotenv playwright PyYAML tenacity

      - name: Cache Playwright Binaries
        uses: actions/cache@v5
        id: playwright-cache
        with:
          path: ~/.cache/ms-playwright
          key: ${{ runner.os }}-playwright-cleaner

      - name: Install Playwright Browsers
        if: steps.playwright-cache.outputs.cache-hit != 'true'
        run: playwright install chromium --with-deps

      - name: Restore Link Checker Cache
        uses: actions/cache/restore@v4
        id: link-checker-cache
        with:
          path: |
            shard_result.json
            triage_report.md
            health_check_checkpoint.json
          key: link-cleaner-results-${{ github.run_id }}-${{ github.run_attempt }}
          restore-keys: |
            link-cleaner-results-${{ github.run_id }}-
            link-cleaner-results-

      - name: Global Intelligent Cleaning Execution (Sequential Fast-Track)
        env:
          GEMINI_API_KEY_1: ${{ secrets.GEMINI_API_KEY_1 }}
          GEMINI_API_KEY_2: ${{ secrets.GEMINI_API_KEY_2 }}
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          FORCE_FULL_CHECK: ${{ github.event.inputs.force_full_check || 'false' }}
          PYTHONPATH: ${{ github.workspace }}
          PYTHONUNBUFFERED: 1
        run: |
          if [ -s shard_result.json ]; then
            echo "shard_result.json found in cache/restored files and has content. Skipping health checker execution."
          else
            python src/intelligent_health_checker.py
          fi

      - name: Run Inventory Reducer & Apply Changes
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          PYTHONPATH: ${{ github.workspace }}
        run: |
          # Since it's sequential, the script produces shard_result.json (default name for no shard)
          python src/inventory_reducer.py

      - name: Open Triage Issue for High-Value Links
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: |
          if [ -f triage_report.md ] && [ -s triage_report.md ]; then
            ISSUE_TITLE="🚨 Manual Triage Required: High-Value Links Failing ($(date +'%Y-%m-%d'))"
            EXISTING_ISSUE=$(gh issue list --search "$ISSUE_TITLE" --state open --json number --jq '.[0] | select(. != null) | .number')
            
            if [ -z "$EXISTING_ISSUE" ] || [ "$EXISTING_ISSUE" = "null" ]; then
              gh issue create --title "$ISSUE_TITLE" --body-file triage_report.md --label "triage, automated-report" || gh issue create --title "$ISSUE_TITLE" --body-file triage_report.md
            else
              echo "A triage issue already exists (#$EXISTING_ISSUE). Updating with new report..."
              gh issue comment $EXISTING_ISSUE --body-file triage_report.md
            fi
          fi

      - name: Ensure Cache Files Exist
        if: always()
        run: |
          touch shard_result.json triage_report.md health_check_checkpoint.json

      - name: Save Link Checker Cache
        if: always()
        uses: actions/cache/save@v4
        with:
          path: |
            shard_result.json
            triage_report.md
            health_check_checkpoint.json
          key: link-cleaner-results-${{ github.run_id }}-${{ github.run_attempt }}