Dotan> One question, though, is that code unicode-safe in the event that Dotan> there are unicode characters in there?
Off the top of my head I don't know (and it may well vary by Python version). Just leave out the dangerous parts of the loop to check: for f in glob.glob('*.vcf'): # corrupt data uncooked = open(f, 'rb').read() # fix it cooked = uncooked.replace('\n ', '') # and write it to the bit bucket open(os.path.devnull, 'wb').write(cooked) Skip -- http://mail.python.org/mailman/listinfo/python-list