pre-commit hook stops .ipynb outputs

pre-commit.py

#! ./python
import json
import sys
from pathlib import Path
import subprocess

def get_staged_files():
    """Get list of files that are staged for commit."""
    try:
        result = subprocess.run(
            ['git', 'diff', '--cached', '--name-only'], 
            capture_output=True, 
            text=True
        )
        if result.returncode != 0:
            print(f"Error getting staged files: {result.stderr}")
            return []
        
        staged_files = result.stdout.strip().split('\n')
        # Filter for only .ipynb files
        return [file for file in staged_files if file.endswith('.ipynb')]
    except Exception as e:
        print(f"Error getting staged files: {e}")
        return []

def check_notebook_outputs(notebook_path):
    """Check if a Jupyter notebook has any outputs or execution counts."""
    try:
        # Read the notebook file
        notebook_content = Path(notebook_path).read_text()
        notebook = json.loads(notebook_content)
        
        # Check if it's a valid notebook format
        if 'cells' not in notebook:
            return False
        
        # Process each cell to check for outputs
        for cell in notebook.get('cells', []):
            if cell.get('cell_type') == 'code':
                # Check execution count
                if 'execution_count' in cell and cell['execution_count'] is not None:
                    return True
                
                # Check outputs
                if 'outputs' in cell and cell['outputs']:
                    return True
        
        # No outputs found
        return False
    except Exception as e:
        print(f"Error processing {notebook_path}: {e}")
        return False

def main():
    """Find staged .ipynb files in the repository and check for outputs."""
    # Get staged files first
    staged_notebooks = get_staged_files()
    
    if not staged_notebooks:
        print("No staged notebooks found for commit.")
        return 0
    
    notebooks_with_outputs = []
    
    # Check each staged notebook
    for notebook_path in staged_notebooks:
        if check_notebook_outputs(notebook_path):
            notebooks_with_outputs.append(notebook_path)
    
    # If any notebooks have outputs, abort the commit
    if notebooks_with_outputs:
        # Output error as one line with files separated by semicolons
        print(f"ERROR: Please clear notebook(s) with outputs: {'; '.join(str(path) for path in notebooks_with_outputs)}")
        return 1
    
    print("All notebooks are clean (no outputs).")
    return 0

if __name__ == "__main__":
    sys.exit(main())