hashprob.py

import collections

def nest_variables(variables):
    nested = dict()
    for key, value in variables.items():
        segments = key.split('.')
        location = nested
        for segment in segments[:-1]:
            if segment not in location:
                location[segment] = dict()
            location = location[segment]
            print("location: %s" % location)
            print("nested: %s" % nested)
            if not isinstance(location, dict):
                print('good')
                return

        location[segments[-1]] = value

    print('bad')

# use ordereddict, just to be sure order is irrelevant
nest_variables(collections.OrderedDict({'foo.bar': '1', 'foo': '2'}))

Run about 5 times in terminal (should return different output):

• $ PYTHONHASHSEED="random" python2.7 hashprob.py

and then about 5 times as (works as expected):

• $ python2.7 hashprob.py

Same problem with python3.3 that turns random hashseed by default

I an reproduce it. It misses about half of the time.

python 2.7.3 (default, Nov 17 2012, 19:54:34)
[GCC 4.2.1 Compatible Apple Clang 4.1 ((tags/Apple/clang-421.11.66))] on darwin

Thanks!

Seems to be the case of two bugs. One being introduction of hash randomization (or better say dict really not keeping the order) and the other of test case not covering a bug, which I was hunting.

Bottom of line, this fails due to careless passing of dict instead of ordereddict.