gistfile1.py

import unicodedata

from pymongo.son import SON
from pymongo.son_manipulator import SONManipulator


class NormalizeUnicode(SONManipulator):
    def __init__(self, form="NFKD"):
        self.__form = form

    def transform_incoming(self, son, collection):
        def transform_value(value):
            if isinstance(value, unicode):
                return unicodedata.normalize(self.__form, value)
            elif isinstance(value, list):
                return [transfom_value(v) for v in value]
            elif isinstance(value, dict):
                return transform_dict(SON(value))
            return value

        def transform_dict(object):
            for (key, value) in object.items():
                object[key] = transform_value(value)
            return object

        return transform_dict(SON(son))