# .github/workflows/train_and_push.yml
# Three triggers:
#   1. train        — daily 7am EST (start_year=2016) + manual single-year
#   2. sweep_YYYY   — manual consensus sweep (parallel jobs, missing years only)
#   3. cleanup      — daily midnight EST, wipes single-year outputs only

name: Daily TFT Training

on:
  schedule:
    - cron: "0 12 * * 1-5"   # 12:00 UTC = 7am EST Mon-Fri (single-year default)
    - cron: "0 5 * * *"       # 05:00 UTC = midnight EST daily cleanup
    - cron: "0 1 * * 2-6"    # 01:00 UTC = 8pm EST Mon-Fri (daily consensus sweep)
    - cron: "0 0 * * 2-6"    # 00:00 UTC = 7pm EST Mon-Fri (wipe previous day's sweep)
  workflow_dispatch:
    inputs:
      start_year:
        description: "Start year for single-year run"
        required: false
        default: "2016"
        type: string
      sweep_mode:
        description: "Comma-separated sweep years e.g. 2008,2014,2016,2019,2021 (leave blank for single-year)"
        required: false
        default: ""
        type: string
      force_refresh:
        description: "Force full dataset rebuild"
        required: false
        default: "false"
        type: boolean

jobs:

  # ── 7pm EST: wipe previous day sweep cache from HF dataset ─────────────────
  wipe_sweep_cache:
    if: >
      github.event_name == 'schedule' &&
      github.event.schedule == '0 0 * * 2-6'
    runs-on: ubuntu-latest
    timeout-minutes: 10
    env:
      HF_TOKEN: ${{ secrets.HF_TOKEN }}
    steps:
      - name: Delete stale dated sweep files from HF dataset
        run: |
          pip install huggingface_hub -q
          python3 - <<'PYEOF'
          import os
          from datetime import datetime, timezone, timedelta
          from huggingface_hub import HfApi
          api   = HfApi()
          token = os.environ["HF_TOKEN"]
          repo  = "P2SAMAPA/p2-etf-tft-outputs"
          today = (datetime.now(timezone.utc) - timedelta(hours=5)).strftime("%Y%m%d")
          try:
              files = api.list_repo_files(repo_id=repo, repo_type="dataset", token=token)
          except Exception as e:
              print(f"Could not list files: {e}"); files = []
          deleted = 0
          for f in files:
              if f.startswith("signals_") and f.endswith(".json") and "_" in f[9:]:
                  parts = f.replace(".json","").split("_")
                  if len(parts) == 3 and parts[2] != today:
                      try:
                          api.delete_file(path_in_repo=f, repo_id=repo,
                              repo_type="dataset", token=token,
                              commit_message=f"Wipe stale sweep: {f}")
                          print(f"Deleted {f}"); deleted += 1
                      except Exception as e:
                          print(f"Could not delete {f}: {e}")
          print(f"Wiped {deleted} stale sweep files")
          PYEOF

  # ── Midnight cleanup — wipes single-year outputs only ────────────────────────
  cleanup:
    if: >
      github.event_name == 'schedule' &&
      github.event.schedule == '0 5 * * *'
    runs-on: ubuntu-latest
    timeout-minutes: 10
    env:
      HF_TOKEN: ${{ secrets.HF_TOKEN }}
    steps:
      - name: Wipe single-year outputs (preserve sweep cache)
        run: |
          pip install huggingface_hub -q
          python3 - <<'PYEOF'
          import os
          from huggingface_hub import HfApi
          api = HfApi()
          token = os.environ["HF_TOKEN"]
          repo_id = "P2SAMAPA/p2-etf-tft-outputs"
          for filename in ["model_outputs.npz", "signals.json", "training_meta.json"]:
              try:
                  api.delete_file(path_in_repo=filename, repo_id=repo_id,
                      repo_type="dataset", token=token,
                      commit_message=f"Midnight cleanup: {filename}")
                  print(f"Deleted {filename}")
              except Exception as e:
                  print(f"{filename} not found: {e}")
          print("Cleanup done")
          PYEOF

  # ── Single-year train — daily default + manual (no sweep_mode) ──────────────
  train:
    if: >
      (github.event_name == 'schedule' &&
       github.event.schedule == '0 12 * * 1-5') ||
      (github.event_name == 'workflow_dispatch' &&
       (github.event.inputs.sweep_mode == '' || github.event.inputs.sweep_mode == null))
    runs-on: ubuntu-latest
    timeout-minutes: 180
    env:
      HF_TOKEN:     ${{ secrets.HF_TOKEN }}
      FRED_API_KEY: ${{ secrets.FRED_API_KEY }}
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"
          cache: "pip"
      - name: Install dependencies
        run: pip install --no-cache-dir -r requirements.txt
      - name: Train
        run: |
          START_YEAR="${{ github.event.inputs.start_year }}"
          START_YEAR="${START_YEAR:-2016}"
          ARGS="--start-year $START_YEAR"
          if [ "${{ github.event.inputs.force_refresh }}" = "true" ]; then
            ARGS="$ARGS --force-refresh"
          fi
          python train_pipeline.py $ARGS
      - if: always()
        uses: actions/upload-artifact@v4
        with:
          name: logs-single-${{ github.run_id }}
          path: "*.log"
          retention-days: 7

  # ── Sweep 2008 ───────────────────────────────────────────────────────────────
  sweep_2008:
    if: >
      (github.event_name == 'schedule' && github.event.schedule == '0 1 * * 2-6') ||
      (github.event_name == 'workflow_dispatch' && contains(github.event.inputs.sweep_mode, '2008'))
    runs-on: ubuntu-latest
    timeout-minutes: 180
    env:
      HF_TOKEN:     ${{ secrets.HF_TOKEN }}
      FRED_API_KEY: ${{ secrets.FRED_API_KEY }}
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"
          cache: "pip"
      - run: pip install --no-cache-dir -r requirements.txt
      - run: python train_pipeline.py --start-year 2008 --sweep-date $(date -u -d "-5 hours" +%Y%m%d 2>/dev/null || date -u +%Y%m%d)
      - if: always()
        uses: actions/upload-artifact@v4
        with:
          name: logs-sweep-2008-${{ github.run_id }}
          path: "*.log"
          retention-days: 7

  # ── Sweep 2014 ───────────────────────────────────────────────────────────────
  sweep_2014:
    if: >
      (github.event_name == 'schedule' && github.event.schedule == '0 1 * * 2-6') ||
      (github.event_name == 'workflow_dispatch' && contains(github.event.inputs.sweep_mode, '2014'))
    runs-on: ubuntu-latest
    timeout-minutes: 180
    env:
      HF_TOKEN:     ${{ secrets.HF_TOKEN }}
      FRED_API_KEY: ${{ secrets.FRED_API_KEY }}
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"
          cache: "pip"
      - run: pip install --no-cache-dir -r requirements.txt
      - run: python train_pipeline.py --start-year 2014 --sweep-date $(date -u -d "-5 hours" +%Y%m%d 2>/dev/null || date -u +%Y%m%d)
      - if: always()
        uses: actions/upload-artifact@v4
        with:
          name: logs-sweep-2014-${{ github.run_id }}
          path: "*.log"
          retention-days: 7

  # ── Sweep 2016 ───────────────────────────────────────────────────────────────
  sweep_2016:
    if: >
      (github.event_name == 'schedule' && github.event.schedule == '0 1 * * 2-6') ||
      (github.event_name == 'workflow_dispatch' && contains(github.event.inputs.sweep_mode, '2016'))
    runs-on: ubuntu-latest
    timeout-minutes: 180
    env:
      HF_TOKEN:     ${{ secrets.HF_TOKEN }}
      FRED_API_KEY: ${{ secrets.FRED_API_KEY }}
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"
          cache: "pip"
      - run: pip install --no-cache-dir -r requirements.txt
      - run: python train_pipeline.py --start-year 2016 --sweep-date $(date -u -d "-5 hours" +%Y%m%d 2>/dev/null || date -u +%Y%m%d)
      - if: always()
        uses: actions/upload-artifact@v4
        with:
          name: logs-sweep-2016-${{ github.run_id }}
          path: "*.log"
          retention-days: 7

  # ── Sweep 2019 ───────────────────────────────────────────────────────────────
  sweep_2019:
    if: >
      (github.event_name == 'schedule' && github.event.schedule == '0 1 * * 2-6') ||
      (github.event_name == 'workflow_dispatch' && contains(github.event.inputs.sweep_mode, '2019'))
    runs-on: ubuntu-latest
    timeout-minutes: 180
    env:
      HF_TOKEN:     ${{ secrets.HF_TOKEN }}
      FRED_API_KEY: ${{ secrets.FRED_API_KEY }}
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"
          cache: "pip"
      - run: pip install --no-cache-dir -r requirements.txt
      - run: python train_pipeline.py --start-year 2019 --sweep-date $(date -u -d "-5 hours" +%Y%m%d 2>/dev/null || date -u +%Y%m%d)
      - if: always()
        uses: actions/upload-artifact@v4
        with:
          name: logs-sweep-2019-${{ github.run_id }}
          path: "*.log"
          retention-days: 7

  # ── Sweep 2021 ───────────────────────────────────────────────────────────────
  sweep_2021:
    if: >
      (github.event_name == 'schedule' && github.event.schedule == '0 1 * * 2-6') ||
      (github.event_name == 'workflow_dispatch' && contains(github.event.inputs.sweep_mode, '2021'))
    runs-on: ubuntu-latest
    timeout-minutes: 180
    env:
      HF_TOKEN:     ${{ secrets.HF_TOKEN }}
      FRED_API_KEY: ${{ secrets.FRED_API_KEY }}
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"
          cache: "pip"
      - run: pip install --no-cache-dir -r requirements.txt
      - run: python train_pipeline.py --start-year 2021 --sweep-date $(date -u -d "-5 hours" +%Y%m%d 2>/dev/null || date -u +%Y%m%d)
      - if: always()
        uses: actions/upload-artifact@v4
        with:
          name: logs-sweep-2021-${{ github.run_id }}
          path: "*.log"
          retention-days: 7