Add element to dict but don't overwrite value when key exists

Question

I want to add elements (key-value-pairs) to a dict. But I want to prevent overwriting existing values when the key exists.

But I don't want to do an if check. I would prefer an exception.

d = {}

d['a'] = 1
d['b'] = 2
d['a'] = 3  # <-- raise Exception! "KeyExistsException'

What I don't want

if not 'a' in d:
    d['a'] = 3

You could implement your own custom dict class to do this. You'll still need the `if key in dict` check, but it can be written once, inside the class. — John Gordon, Jul 27 '22 at 19:25
One option if you are on Python 3.9 or higher: `d = {'a': 3} | d`. This will update the dictionary, but conflicting keys will always use the existing values from `d`. — 0x5453, Jul 27 '22 at 19:25
Ahh, right. There is no built-in way to do that - you'll either need a manual `if` check or a custom `dict` class that has your desired behavior. — 0x5453, Jul 27 '22 at 19:29
This may help. [How to raise error if duplicates keys in dictionary](https://stackoverflow.com/questions/4999233/how-to-raise-error-if-duplicates-keys-in-dictionary) — Nesi, Jul 27 '22 at 19:36
It doesn't raise an exception, but if you use the `setdefault` dict method to add key-value pairs, it won't overwrite existing key values, e.g., `d.setdefault("a", 3)` — Matt Pitkin, Jul 27 '22 at 19:38
@0x5453 since you are already using a literal, I think `d = {"a":3, **d}` would be nicer. But both of these approache would be inefficient (linear time) so they shouldn't be used if this needs to happen repeatedly. But more idiomatically, i would just use `d.setdefault("a", 3)` — juanpa.arrivillaga, Jul 27 '22 at 19:40

JacobIRR · Answer 1 · 2022-07-27T19:30:57.510

You can subclass dict and in particular, override the __setitem__ method.

This sounds like what you want:

class SpecialDict(dict):
    def __setitem__(self, key, value):
        if not key in self:
            super(SpecialDict, self).__setitem__(key, value)
        else:
            raise Exception("Cannot overwrite key!") # You can make your own type here if needed


x = SpecialDict()


x['a'] = 1
x['b'] = 2
x['a'] = 3  #raises Exception

score 1 · Answer 2 · edited Jul 27 '22 at 19:37

1

Instead of subclassing dict as suggested by JacobIRR, you could also define a helper function for storing a key-value pair in a dict that throws an exception when the key already exists:

class KeyExistsException(Exception):
    pass

def my_add(the_dict, the_key, the_value):
    if the_key in the_dict:
        raise KeyExistsException("value already exists")
    the_dict[the_key] = the_value


d = {}

my_add(d, 'a', 1)
my_add(d, 'b', 2)
my_add(d, 'a', 3)  # <-- raise Exception! "KeyExistsException'

edited Jul 27 '22 at 19:37

juanpa.arrivillaga

88,713
10
131
172

answered Jul 27 '22 at 19:35

jboockmann

1,011
2
11
27

Also a nice solution. Keep in mind that today it is not recommended anymore to derive create your own Exceptions. – buhtz Jul 27 '22 at 20:08
1

Oh, I did not knew that. Do you have any sources for the rationale behind that? – jboockmann Jul 27 '22 at 20:32
pylint seems fine with it, if you replace pass with a docstring at least – Kenny Ostrom Jul 27 '22 at 21:02

score 1 · Answer 3 · answered Jul 27 '22 at 19:48

This might work for your use. I am a noob so I may be overlooking something.

d = {}
 
try:
    d['new_key']  # check if key exists.
except KeyError:  # if not exist create.
    d['new_key'] = 'value'
else:  # else raise exception KeyExists.
    raise Exception ('KeyExists')

Add element to dict but don't overwrite value when key exists

3 Answers3