test_full.yaml

name: Full Testing Suite

on:
  push:
    paths:
    - src/**
    - scripts/**
    - tests/**
    - pyproject.toml
    - uv.lock
    - .github/workflows/test_full.yaml
  workflow_dispatch:
    inputs:
      no_cleanup:
        description: Keep intermediate files after running tests
        default: false
        type: boolean
      styxcache_dir:
        description: Override persistent styxcache directory on the self-hosted runner
        default: ""
        type: string

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

jobs:
  test:
    runs-on: arcana
    env:
      STYXCACHE_DIR_OVERRIDE: ${{ inputs.styxcache_dir }}
      STYXCACHE_MAX_GB: "50"
    steps:
    - uses: actions/checkout@v6
    - uses: astral-sh/setup-uv@v8.0.0
      with:
        version: "0.11.3"
        enable-cache: true # not automatic on self-hosted runners
    - run: uv sync

    - name: Prepare styx cache
      shell: bash
      run: |
        # Resolve cache dir: honor explicit override, else use the runner user's
        # home cache (persistent across runs, user-owned — no sudo required).
        STYXCACHE_DIR="${STYXCACHE_DIR_OVERRIDE:-$HOME/.cache/rbc-styxcache}"
        echo "STYXCACHE_DIR=$STYXCACHE_DIR" >> "$GITHUB_ENV"

        mkdir -p "$STYXCACHE_DIR"
        # Purge stale staging dirs from crashed previous runs (safe: atomic-rename commits)
        rm -rf "$STYXCACHE_DIR/.incoming" 2>/dev/null || true

        # Drop entries that contain any truncated .gz (styxcache's own
        # stdout/stderr streams or cached .nii.gz outputs). Atomic commits
        # should prevent these, but crashes before commit have been observed.
        corrupt=0
        while IFS= read -r entry; do
          bad=0
          for gz in $(find "$entry" -type f -name '*.gz'); do
            if ! gzip -t "$gz" 2>/dev/null; then
              bad=1
              break
            fi
          done
          if [ "$bad" = "1" ]; then
            rm -rf "$entry"
            corrupt=$((corrupt + 1))
          fi
        done < <(find "$STYXCACHE_DIR" -mindepth 2 -maxdepth 2 -type d 2>/dev/null)
        if [ "$corrupt" -gt 0 ]; then
          echo "evicted $corrupt entries with corrupt gzip streams"
        fi

        size_kb=$(du -sk "$STYXCACHE_DIR" 2>/dev/null | awk '{print $1}')
        size_kb=${size_kb:-0}
        cap_kb=$(( STYXCACHE_MAX_GB * 1024 * 1024 ))
        if [ "$size_kb" -gt "$cap_kb" ]; then
          echo "Cache size ${size_kb}KB exceeds cap ${cap_kb}KB; evicting oldest shards"
          find "$STYXCACHE_DIR" -mindepth 2 -maxdepth 2 -type d -printf '%T@ %p\n' \
            | sort -n \
            | while read -r _ path; do
                rm -rf "$path"
                cur=$(du -sk "$STYXCACHE_DIR" 2>/dev/null | awk '{print $1}')
                [ "${cur:-0}" -le "$cap_kb" ] && break
              done
        fi

        {
          echo "## styxcache"
          echo "- dir: \`$STYXCACHE_DIR\`"
          echo "- size before: $(du -sh "$STYXCACHE_DIR" 2>/dev/null | cut -f1)"
        } >> "$GITHUB_STEP_SUMMARY"

    - name: Run all tests
      shell: bash
      run: |
        # Clean up stale job tmp dirs from previous crashed runs
        find "$RUNNER_TEMP" -maxdepth 1 -name 'rbc_test_*' -mmin +120 -exec rm -rf {} + 2>/dev/null || true

        export JOB_TMP="$RUNNER_TEMP/rbc_test_$(date +%s%N)"
        mkdir -p $JOB_TMP
        echo "JOB_TMP=$JOB_TMP" >> $GITHUB_ENV
        echo "Job temp directory: $JOB_TMP"

        export PYTEST_CACHE_DIR="$JOB_TMP/.pytest_cache"
        export PYTHONPYCACHEPREFIX="$JOB_TMP/__pycache__"
        export COVERAGE_FILE="$JOB_TMP/.coverage"
        export RBC_STYXCACHE_DIR="$STYXCACHE_DIR"

        uv run pytest \
          -n 8 \
          --runner=podman \
          --basetemp=$JOB_TMP/pytest-tmp \
          --cov-report= \
          --cov=src \
          --durations=0 \
          --log-level=DEBUG \
          --verbose tests

    - name: Generate pipeline report
      if: always()
      continue-on-error: true
      shell: bash
      run: |
        uv run scripts/visualize_pipeline.py --output $JOB_TMP/pipeline_report.png

        if [ -f "$JOB_TMP/pipeline_report.png" ]; then
          echo "## Pipeline Report" >> $GITHUB_STEP_SUMMARY
          echo "Download the **pipeline-report** artifact for the full visualization." >> $GITHUB_STEP_SUMMARY
        fi

    - name: Upload pipeline report
      if: always()
      uses: actions/upload-artifact@v7
      with:
        name: pipeline-report
        path: ${{ env.JOB_TMP }}/pipeline_report.png

    - name: Coverage summary
      if: always()
      env:
        COVERAGE_FAIL: 75
      shell: bash
      run: |
        uv run coverage report \
          --data-file=$JOB_TMP/.coverage \
          --fail-under=$COVERAGE_FAIL \
          --show-missing \
          --format=markdown >> $GITHUB_STEP_SUMMARY

    - name: styxcache summary
      if: always()
      shell: bash
      run: |
        size=$(du -sh "$STYXCACHE_DIR" 2>/dev/null | cut -f1)
        entries=$(find "$STYXCACHE_DIR" -mindepth 2 -maxdepth 2 -type d 2>/dev/null | wc -l)
        {
          echo "- size after: ${size:-0}"
          echo "- entries: ${entries:-0}"
        } >> "$GITHUB_STEP_SUMMARY"

    - name: Cleanup
      if: ${{ always() && github.event.inputs.no_cleanup != true }}
      shell: bash
      run: |
        echo "Cleaning up job temp directory: $JOB_TMP"
        echo "Temp directory size: $(du -sh $JOB_TMP 2>/dev/null | cut -f1)"
        rm -rf $JOB_TMP