drift.yml

name: Scheduled drift check

on:
  schedule:
    # 06:00 UTC daily.
    - cron: "0 6 * * *"
  workflow_dispatch:
    inputs:
      dry_run:
        description: "If true, run the helper in --dry-run mode (no git/gh side effects)."
        required: false
        default: "false"
        type: choice
        options:
          - "false"
          - "true"

permissions:
  # Need write access to push branches and open PRs with the default
  # GITHUB_TOKEN. No id-token / packages permissions are needed.
  contents: write
  pull-requests: write

env:
  # Base branch the bot branches off of and targets PRs against. Override here
  # once the refactor lands on dev/main.
  BASE_BRANCH: dev

jobs:
  drift:
    runs-on: ubuntu-latest
    # Schedule events on forks run with the upstream token but we still want to
    # be defensive: only run on the canonical repo to avoid surprise pushes
    # from fork-triggered manual dispatches.
    if: github.repository == 'bigbio/hvantk' || github.event_name == 'workflow_dispatch'

    steps:
      - name: Checkout base branch
        uses: actions/checkout@v4
        with:
          # Need history + branches for the helper to create new branches and
          # push them. fetch-depth: 0 keeps things simple at the cost of size.
          fetch-depth: 0
          ref: ${{ env.BASE_BRANCH }}

      - name: Install OpenBLAS
        run: |
          sudo apt-get update
          sudo apt-get install -y libopenblas-dev

      - name: Set up Python 3.10
        uses: actions/setup-python@v3
        with:
          python-version: "3.10"

      - name: Add conda to system path
        run: |
          echo $CONDA/bin >> $GITHUB_PATH

      - name: Install dependencies via conda env
        run: |
          conda env update --file environment.yml --name base

      - name: Install hvantk in development mode
        run: |
          pip install -e .

      - name: Configure git author for bot commits
        run: |
          git config user.name "hvantk-drift-bot"
          git config user.email "bot@hvantk.dev"

      - name: Run drift check (capture JSON)
        # Tolerate non-zero exits: drift_cli returns 1 on drifted, 2 on
        # probe_failed. Those are signals, not failures.
        run: |
          set +e
          python -m hvantk.hvantk drift --all --json > /tmp/drift_report.json
          rc=$?
          echo "drift exit code: $rc"
          if [ $rc -gt 2 ]; then
            echo "drift CLI exited with $rc (unexpected; treating as failure)" >&2
            exit $rc
          fi
          exit 0

      - name: Show drift report
        run: |
          echo "--- /tmp/drift_report.json ---"
          cat /tmp/drift_report.json || true

      - name: Open / update draft PRs per drifted dataset
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          BASE_BRANCH: ${{ env.BASE_BRANCH }}
          GITHUB_REPOSITORY: ${{ github.repository }}
          DRY_RUN: ${{ github.event.inputs.dry_run }}
        run: |
          ARGS="--report /tmp/drift_report.json --base-branch ${BASE_BRANCH}"
          if [ "$DRY_RUN" = "true" ]; then
            ARGS="$ARGS --dry-run"
          fi
          python .github/scripts/drift_to_pr.py $ARGS