Git hook pre-commit script for stripping output and irrelevant metadata from Jupyter notebooks

pre-commit

#!/usr/bin/env python3

"""
  Git pre-commit hook script

  Strips output from Jupyter Notebooks
"""

import os
import sys
from subprocess import check_output, check_call

files = check_output(['git', 'diff', '--name-only', '--cached'], encoding='ascii').splitlines()

def strip_output(file_path):
    """ Strip all outputs, to reduce git noise """
    d, file_name = os.path.split(file_path)
....
    jq_format = '(.cells[] | select(has("outputs")) | .outputs) = []| (.cells[] | select(has("execution_count")) | .execution_count) = null

    try:
        stripped = check_output(['jq', '--indent', '1', jq_format, file_name], cwd=d)
        #print(stripped)
    except FileNotFoundError:
        print(file_name)
        print("Please install jq for pre-commit clean-up!")
        print("Aborting commit...")
        sys.exit(1)

    with open(file_path, "wb") as f:
        f.write(stripped)

for f in files:
    if f[-5:] == b'ipynb':
        print(f"Stripping output from {f}...")
        strip_output(f)
        check_call(['git', 'add', f])

files = check_output(['git', 'diff', '--name-only', '--cached']).splitlines()

if not files:
    print("No changes after stripping output, aborting commit!")
    sys.exit(1)