pull-practice-data.yml

name: Pull Practice Session Data

on:
  # Run Saturday 22:00 UTC during F1 season — captures FP1/FP2/FP3 for all circuits,
  # including late-running Americas sessions (COTA ~18:30 UTC, Miami ~18:30 UTC).
  schedule:
    - cron: '0 22 * 3-11 sat'   # Mar–Nov (EDT/EST): Saturday 22:00 UTC
    - cron: '0 22 * 12,1,2 sat' # Dec–Feb (EST): Saturday 22:00 UTC
  workflow_dispatch:
    inputs:
      force_repull:
        description: 'Re-pull sessions already in all_practice_laps.csv (rewrites file)'
        type: boolean
        default: false

env:
  PYTHON_VERSION: '3.11'
  FORCE_JAVASCRIPT_ACTIONS_TO_NODE24: 'true'

jobs:
  pull-practice:
    runs-on: ubuntu-latest
    timeout-minutes: 90   # practice data is large; FastF1 downloads can be slow
    permissions:
      contents: write     # needed to push the updated CSV back to main

    steps:
      - uses: actions/checkout@v5
        with:
          fetch-depth: 0  # full history so we can pull/rebase before pushing

      - uses: actions/setup-python@v6
        with:
          python-version: ${{ env.PYTHON_VERSION }}
          cache: 'pip'

      - name: Install dependencies
        run: pip install -r requirements.txt

      - name: Show job context
        run: |
          echo "Job started at $(date -u)"
          echo "Python: $(python --version)"
          git log -1 --oneline
          git fetch origin main
          echo "origin/main is at $(git rev-parse origin/main)"

      - name: Pull practice session data (incremental)
        run: python fastF1-practices.py
        env:
          LOCAL_RUN: '1'    # enables FastF1 disk cache

      - name: Upload practice data as artifact
        uses: actions/upload-artifact@v6
        with:
          name: practice-laps-${{ github.run_id }}
          path: data_files/all_practice_laps.csv
          retention-days: 14

      - name: Commit updated practice data
        run: |
          git config --local user.email "github-actions[bot]@users.noreply.github.com"
          git config --local user.name "github-actions[bot]"
          # Bring our branch up to date in case of concurrent pushes
          git pull --ff-only origin main || true
          git add data_files/all_practice_laps.csv
          if git diff --staged --quiet; then
            echo "No new practice data — nothing to commit"
          else
            ROWS=$(python -c "import pandas as pd; df=pd.read_csv('data_files/all_practice_laps.csv', sep='\t'); print(len(df))" 2>/dev/null || echo "?")
            git commit -m "data: update all_practice_laps.csv (${ROWS} rows) [skip ci]"
            git push
            echo "Pushed updated practice data to main"
          fi

      - name: Show post-push status
        run: |
          echo "Job finished at $(date -u)"
          git log -1 --oneline