If I have a string containing surrogate pairs like this in Python 3.4: "\udb40\udd9d" How do I convert it into the proper form: "\U000E019D" ? The answer appears not to be "unicodedata.normalize".