homepage

# pickle.dumps is not able to process an instance of
# a class that inherits from 'dict' and
# overrides the built-in __getattribute__ method
# but can successfully process one that 
# overrides the__getattr__ method
>>> class Examp1(dict):
... def __getattr__(self,name):
... return self[name]
... 
>>> class Examp2(dict):
... def __getattribute__(self,name):
... return self[name]
... 
>>> ex1 = Examp1()
>>> ex2 = Examp2()
>>> ex1['aKey'] = (3,4)
>>> ex2['aKey2'] = (4,5)
>>> ex1
{'aKey': (3, 4)}
>>> ex1.aKey
(3, 4)
>>> ex2
{'aKey2': (4, 5)}
>>> ex2.aKey2
(4, 5)
>>> import pickle
>>> pickle.dumps(ex1)
b'\x80\x03c__main__\nexamp1\nq\x00)\x81q\x01X\x04\x00\x00\x00aKeyq\x02K\x03K\x04\x86q\x03s}q\x04b.'
>>> pickle.dumps(ex2)
Traceback (most recent call last):
 File "<stdin>", line 1, in <module>
 File "[hidden]/python3/3.0b2/common/lib/python3.0/pickle.py", line
1319, in dumps
 Pickler(f, protocol).dump(obj)
 File "<stdin>", line 3, in __getattribute__
KeyError: '__reduce_ex__'

Is this related to #1730480?

Should this be closed for the same reason #1730480 was?
If not, would this effectively be a feature request and have to wait for a new version (3.3)?

I don't think it is a bug.
The posted code completely breaks the expected behavior of __getattribute__. With a such implementation, there is nothing we can do with this object as we cannot introspect it.
Use the following if you really need this kind of behaviour:
class E(dict):
 def __getattribute__(self,name):
 try:
 return self[name]
 except KeyError:
 return dict.__getattribute__(self, name)