action.yml

name: 'NBStripout Check'
description: 'Verify that Jupyter notebooks have their output stripped'
inputs:
  python-version:
    description: 'Python version to use (supports versions, ranges, or "3.x"). Can also use a .python-version file by setting this to an empty string and ensuring the file exists.'
    required: false
    default: '3.x'
  paths:
    description: 'Newline-separated list of paths to check (supports wildcards)'
    required: false
    default: '**/*.ipynb'
  extra-keys:
    description: 'Extra metadata keys to strip (space-separated)'
    required: false
    default: ''
  keep-output:
    description: 'Keep output in notebooks'
    required: false
    default: 'false'
  keep-count:
    description: 'Keep execution counts'
    required: false
    default: 'false'
  strip-init-cells:
    description: 'Strip init cells'
    required: false
    default: 'false'
runs:
  using: 'composite'
  steps:
    - name: Setup Python
      uses: actions/setup-python@v6
      with:
        python-version: ${{ inputs.python-version }}
        pip-install: 'nbstripout'

    - name: Check notebooks are stripped
      shell: bash
      env:
        INPUT_PATHS: ${{ inputs.paths }}
        INPUT_EXTRA_KEYS: ${{ inputs.extra-keys }}
        INPUT_KEEP_OUTPUT: ${{ inputs.keep-output }}
        INPUT_KEEP_COUNT: ${{ inputs.keep-count }}
        INPUT_STRIP_INIT_CELLS: ${{ inputs.strip-init-cells }}
      run: |
        # Build nbstripout args as an array (bash 3.1+)
        args=(--verify)

        if [ "$INPUT_KEEP_OUTPUT" = "true" ]; then
          args+=(--keep-output)
        fi

        if [ "$INPUT_KEEP_COUNT" = "true" ]; then
          args+=(--keep-count)
        fi

        if [ "$INPUT_STRIP_INIT_CELLS" = "true" ]; then
          args+=(--strip-init-cells)
        fi

        if [ -n "$INPUT_EXTRA_KEYS" ]; then
          args+=(--extra-keys "$INPUT_EXTRA_KEYS")
        fi

        # Use Python for file discovery: correct ** glob handling on all platforms
        # Outputs null-delimited paths and deduplicates overlapping patterns.
        # INPUT_PATHS is quoted to prevent bash from expanding globs before Python sees them.
        notebook_list="${RUNNER_TEMP:-/tmp}/nbstripout_filelist.txt"

        python3 - "$INPUT_PATHS" << 'PYTHON' > "$notebook_list"
        import sys
        from pathlib import Path

        raw_input = sys.argv[1]
        patterns = [stripped for line in raw_input.splitlines() if (stripped := line.strip())]
        seen = set()
        for pat in patterns:
            for p in sorted(Path(".").glob(pat)):
                rp = str(p)
                if p.is_file() and rp not in seen:
                    seen.add(rp)
                    sys.stdout.write(rp + "\0")
        PYTHON

        if [ ! -s "$notebook_list" ]; then
          echo "No notebook files found matching: $INPUT_PATHS"
          exit 0
        fi

        # Count null-delimited entries (POSIX-compatible, works on macOS and Linux)
        notebook_count=$(tr -cd '\0' < "$notebook_list" | wc -c | tr -d ' ')
        echo "Found $notebook_count notebook(s) to check."

        # xargs -0 reads null-delimited paths (handles any filename characters)
        # and batches automatically for ARG_MAX safety on large repos
        if xargs -0 nbstripout "${args[@]}" < "$notebook_list"; then
          echo "All notebooks are properly stripped! ✅"
        else
          echo ""
          echo "================================================"
          echo "Some notebooks have output that should be stripped."
          echo "Please run: nbstripout <notebook-file>"
          echo "Or install nbstripout as a git filter with: nbstripout --install"
          echo "================================================"
          exit 1
        fi