iMath writes: > which package to use ?
Read the text in as a "bytes object" (bytes), then it has a .decode method that you can experiment with. Strings (str) are Unicode and have an .encode method. These methods allow you to specify a desired encoding and and what to do when there are errors. help(bytes.decode) help(str.encode) help(open) <http://docs.python.org/3.3/library/stdtypes.html> In Python 2.7 and before, strings seem to do double duty and have both the .encode and .decode methods, so Python version matters here. -- http://mail.python.org/mailman/listinfo/python-list