clean_notebooks.py

#! $HOME/anaconda/bin/python
# swap $HOME with your absolute path if this doesn't work

import sys
import io
import os
from nbformat import read, write, current_nbformat
from argparse import ArgumentParser
from glob import glob
from shutil import rmtree

def remove_outputs(nb):
    for cell in nb.cells:
        if cell.cell_type == 'code':
            cell.execution_count = None
            cell.outputs = []

def process_file(fname, overwrite=False):
    with io.open(fname, 'r') as f:
        nb = read(f, current_nbformat)
    
    remove_outputs(nb)

    if overwrite:
        new_ipynb = fname
        print("Overwriting %s" % new_ipynb)
    else:
        base, ext = os.path.splitext(fname)
        new_ipynb = "%s_removed%s" % (base, ext)
        print("New file %s" % new_ipynb)

    with io.open(new_ipynb, 'w', encoding='utf8') as f:
        write(nb, f, current_nbformat)

if __name__ == '__main__':
    parser = ArgumentParser()
    parser.add_argument("-p", "--path", help="Start directory", type=str, default='**/*.ipynb')
    parser.add_argument("-d", "--delete_path", help="Start directory", type=str, default='**/.ipynb_checkpoints')
    parser.add_argument("-o", "--overwrite", help="overwrite original file", action="store_true")
    args = parser.parse_args()

    to_delete = glob(args.delete_path, recursive=True)
    for fname in to_delete:
        rmtree(fname)
    
    to_clean = glob(args.path, recursive=True)
    for fname in to_clean:
        process_file(fname, args.overwrite)