hf-sync.yml

name: HF ↔ GitHub Sync

# Trigger on every push to main (GitHub → HF direction)
# and hourly to pull any HF-side changes back (HF → GitHub direction)
on:
  push:
    branches: [main]
  schedule:
    - cron: '0 * * * *'   # hourly HF pull-check
  workflow_dispatch:
    inputs:
      force_direction:
        description: 'Force sync direction (hf-wins | gh-wins | auto)'
        required: false
        default: 'auto'

env:
  HF_REPO_TYPE: space          # model | dataset | space
  HF_REPO: ${{ vars.HF_REPO }} # e.g. your-org/codecraftlab

jobs:
  sync:
    name: Sync HuggingFace ↔ GitHub
    runs-on: ubuntu-latest
    permissions:
      contents: write

    steps:
      - name: Checkout (full history)
        uses: actions/checkout@v4
        with:
          fetch-depth: 0
          token: ${{ secrets.GITHUB_TOKEN }}

      - name: Configure git identity
        run: |
          git config user.email "sync-bot@codecraftlab.noreply"
          git config user.name "CodeCraftLab Sync Bot"

      - name: Install git-lfs
        run: |
          sudo apt-get install -y git-lfs
          git lfs install

      - name: Add HuggingFace remote
        run: |
          git remote add hf \
            "https://user:${{ secrets.HF_TOKEN }}@huggingface.co/${HF_REPO_TYPE}s/${HF_REPO}"
          git fetch hf --prune

      - name: Detect divergence and resolve
        id: sync
        env:
          FORCE_DIRECTION: ${{ github.event.inputs.force_direction || 'auto' }}
        run: |
          set -euo pipefail

          HF_HEAD=$(git rev-parse hf/main 2>/dev/null || echo "NONE")
          GH_HEAD=$(git rev-parse HEAD)

          if [ "$HF_HEAD" = "NONE" ]; then
            echo "action=push-to-hf" >> "$GITHUB_OUTPUT"
            echo "reason=HF remote has no main branch — initial push"
            exit 0
          fi

          BASE=$(git merge-base HEAD hf/main 2>/dev/null || echo "NONE")

          if [ "$FORCE_DIRECTION" = "hf-wins" ]; then
            echo "action=hf-wins" >> "$GITHUB_OUTPUT"
            echo "reason=Forced HF-wins override"
          elif [ "$FORCE_DIRECTION" = "gh-wins" ]; then
            echo "action=push-to-hf" >> "$GITHUB_OUTPUT"
            echo "reason=Forced GH-wins override"
          elif [ "$HF_HEAD" = "$GH_HEAD" ]; then
            echo "action=in-sync" >> "$GITHUB_OUTPUT"
            echo "reason=Already in sync"
          elif [ "$BASE" = "$GH_HEAD" ]; then
            # HF is ahead — pull HF → GitHub
            echo "action=hf-wins" >> "$GITHUB_OUTPUT"
            echo "reason=GitHub is behind HF — fast-forward"
          elif [ "$BASE" = "$HF_HEAD" ]; then
            # GitHub is ahead — push GitHub → HF
            echo "action=push-to-hf" >> "$GITHUB_OUTPUT"
            echo "reason=HF is behind GitHub — pushing"
          else
            # Both diverged — HF is source of truth
            echo "action=hf-wins" >> "$GITHUB_OUTPUT"
            echo "reason=CONFLICT: both diverged — HF wins (source of truth)"
          fi

      - name: "[In-sync] Nothing to do"
        if: steps.sync.outputs.action == 'in-sync'
        run: echo "✅ HF and GitHub are in sync — no action required."

      - name: "[Push] GitHub → HuggingFace"
        if: steps.sync.outputs.action == 'push-to-hf'
        run: |
          echo "📤 Pushing GitHub → HuggingFace"
          git push hf main

      - name: "[HF Wins] HuggingFace → GitHub"
        if: steps.sync.outputs.action == 'hf-wins'
        run: |
          echo "📥 HuggingFace wins — overwriting GitHub main"
          git reset --hard hf/main
          git push origin main --force-with-lease || git push origin main --force

      - name: Summary
        if: always()
        run: |
          echo "### Sync Result" >> "$GITHUB_STEP_SUMMARY"
          echo "- **Action:** ${{ steps.sync.outputs.action }}" >> "$GITHUB_STEP_SUMMARY"
          echo "- **Trigger:** ${{ github.event_name }}" >> "$GITHUB_STEP_SUMMARY"
          echo "- **Branch:** main" >> "$GITHUB_STEP_SUMMARY"

  # ------------------------------------------------------------------
  # Validate HF Space config on every push
  # ------------------------------------------------------------------
  validate-space-config:
    name: Validate HF Space README config
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - name: Check README frontmatter
        run: |
          python3 - <<'EOF'
          import re, sys

          with open("README.md") as f:
              content = f.read()

          match = re.match(r"^---\n(.*?)\n---", content, re.DOTALL)
          if not match:
              print("❌ README is missing HF Space YAML frontmatter")
              sys.exit(1)

          frontmatter = match.group(1)
          required_keys = ["title", "sdk", "app_port", "license"]
          missing = [k for k in required_keys if k + ":" not in frontmatter]
          if missing:
              print(f"❌ Missing frontmatter keys: {missing}")
              sys.exit(1)

          if "sdk: streamlit" in frontmatter:
              print("❌ sdk is still 'streamlit' — should be 'docker' for FastAPI")
              sys.exit(1)

          print("✅ HF Space frontmatter is valid")
          EOF