sane_file_names.py

#!/usr/bin/env python

import os
import hashlib
import sys

def get_folders(folder):
  ret = []
  for root, subdirs, _files in os.walk(folder):
    for subdir in subdirs:
      ret.append(f"{root}/{subdir}")
  return ret


def get_files(folder):
  ret = []
  for root, _subdirs, files in os.walk(folder):
    for f in files:
      ret.append(f"{root}/{f}")
  return ret


def sort_folders(folders):
  return sorted(folders, key=lambda k: k.count('/'), reverse=True)


def get_safe_file_name(filename):
  keepcharacters = ('.','_', '/')
  return "".join(c for c in filename if c.isalnum() or c in keepcharacters).rstrip()


def rename_bad_folders(folders):
  for folder in folders:

    subfolders    = folder.split(os.path.sep)
    top           = subfolders[:-1]
    bottom        = subfolders[-1]
    safe_bottom   = get_safe_file_name(bottom)
    new_list      = top + [ safe_bottom ]
    new_name      = os.path.sep.join(new_list)

    if folder == new_name:
      print(f'{folder} is ok')
    else:
      print(f'renaming: {folder}')
      try:
        if not os.path.exists(new_name):
          os.rename(folder, new_name)
        else:
          os.rename(folder, f"{new_name}_._")
      except Exception as ex:
        print(f"Error while processing file: {folder}")
        print(ex)
        sys.exit(1)

  return 'ok'


def get_sha_sum(fname):
    hash_sha3 = hashlib.sha3_256()
    with open(fname, "rb") as f:
        for chunk in iter(lambda: f.read(4096), b""):
            hash_sha3.update(chunk)
    return hash_sha3.hexdigest()


def rename_files(files):
  try:
    for f in files:
      if f != sys.argv[0]:
        os.rename(f, f"sha3-{get_sha_sum(f)}{os.path.splitext(f)[1]}")
      else:
        print('ok')
  except Exception as ex:
    print(f"Error while processing file: {f}")
    print(ex)
    sys.exit(1)


def delete_empyt_folders(folders):
  try:
    for folder in folders:
      os.removedirs(folder)
  except Exception as ex:
    print(f"Error while processing file: {folder}")
    print(ex)

def main():
  old_folders = sort_folders(get_folders("."))

  for i in old_folders:
    print(i)

  rename_bad_folders(old_folders)
  new_folders = sort_folders(get_folders("."))

  for i in old_folders:
    print(i)

  files = get_files(".")
  rename_files(files)

  delete_empyt_folders(new_folders)

if __name__ == '__main__':
  exit(main())