Rename files in a directory by removing non-alphanumeric characters that shouldn't be in a filename.

rename_files.py

"""
Recurseively rename all files in a directory by replace whitespace with underscores
and erasing all non-('a-zA-Z0-9' or '_' or '-' or '.') characters.
"""
import re, os, shutil, argparse, sys

parser = argparse.ArgumentParser(description = "Rename all files in directory by replacing whitespace with underscores.")
parser.add_argument('directories', metavar="DIR", nargs='*', default = [os.getcwd()], help="directories to walk. Default: CWD")
err = sys.stderr.write

def new_names(filenames):
  """
  Generate new filenames for the given filenames
  """
  # Replace white space with underscores:
  filenames_new = (re.sub('\s+', '_', fn) for fn in filenames)
  # Erase non-alphanumeric-period-underscore characters:
  filenames_new = (re.sub('[^a-zA-Z0-9._-]', '',  fn) for fn in filenames_new)
  # Merge consecutive underscores for aesthetics
  filenames_new = [re.sub('_+', '_', fn) for fn in filenames_new]
  return filenames_new

def rename_files_in_dir(dir):
  """
  Walk a directory and rename all files in the path.
  """

  for (dirpath, dirnames, filenames) in os.walk(dir):

    # Rename the filenames
    filenames_new = new_names(filenames)
    for f_old, f_new in zip(filenames, filenames_new):

      f_old = os.path.join(dirpath, f_old)
      f_new = os.path.join(dirpath, f_new)

      if f_old == f_new:
        continue

      err("Renaming %s to %s\n"%(os.path.abspath(f_old), os.path.abspath(f_new)))
      shutil.move(f_old, f_new)

    # Rename the directories, and modify dirnames in place to aid os.walk
    dirnames_new = new_names(dirnames)
    for d_old, d_new in zip(dirnames, dirnames_new):

      d_old = os.path.join(dirpath, d_old)
      d_new = os.path.join(dirpath, d_new)

      if d_old == d_new:
        continue

      err("Renaming dir %s to %s\n"%(os.path.abspath(d_old), os.path.abspath(d_new)))
      shutil.move(d_old, d_new)

    del dirnames[:]
    dirnames.extend(dirnames_new)

    
if __name__ == '__main__':
  args = parser.parse_args()
  dirs = [os.path.abspath(d) for d in args.directories]
  for d in dirs:
    rename_files_in_dir(d)

Wow, works great! Exactly what I was looking for to remove weird UTF characters from my directory names, so I can work with them using web manager.