Hashing dictionary

dict_hasher.py

import reprlib
import hashlib
import base64
# Idea taken from http://stackoverflow.com/questions/5884066/hashing-a-dictionary
class DictHasher(object):

    def __init__(self):
        self._repr_funcs = [
            ([int, str, float, bool], lambda data: data),
            ([tuple, list], lambda data: tuple([self._make_repr(e) for e in data])),
            ([dict, ], lambda data: tuple(sorted((k, self._make_repr(v)) for k, v in data.items()))),
            ([set, frozenset], lambda data: tuple(sorted(self._make_repr(e) for e in o))),
        ]
        self.encode_func = base64.urlsafe_b64encode
        self.hash_func = lambda x :hashlib.sha1(x).digest()

    def _make_repr(self, o):
        if o is None:
            return o
        for types, func in self._repr_funcs:
            if isinstance(o, tuple(types)):
                return func(o)
        raise ValueError("type {type} is not representable. Please provide a function".format(type=type(o).name))

    def add_repr_func(self, types, func):
        self._repr_funcs.append((types, func))

    def hash(self, obj):
        rep = reprlib.repr(self._make_repr(obj)).encode("utf-8")
        hashed = self.hash_func(rep)
        encoded = self.encode_func(hashed).decode("utf-8")
        return encoded


hasher = DictHasher()
hash_obj = hasher.hash