Recursion depth limit when hashing numpy.datetime64

[jcipar]

Describe the bug
When hashing a numpy.datetime64 deepdiff.DeepHash fails with RecursionError: maximum recursion depth exceeded.

This can be fixed by adding numpy.datetime64 to deepdiff.helper.times.

To Reproduce
This is a test case that demonstrates the problem:

from deepdiff import DeepHash
import numpy as np

def test_hash_datetime():
    DeepHash(np.datetime64("2025-03-25T11:15:04Z"))

This is a monkeypatch that I'm using to work around the problem:

import deepdiff.deephash
import deepdiff.helper
import numpy as np

deepdiff.helper.times = (*deepdiff.helper.times, np.datetime64,)
deepdiff.deephash.times = deepdiff.helper.times

Expected behavior
It produces a hash value for the datetime64.

OS, DeepDiff version and Python version (please complete the following information):

• OS: MacOS
• Version: Darwin Darwin Kernel Version 23.6.0: Fri Jul 5 17:55:37 PDT 2024; root:xnu-10063.141.1~2/RELEASE_ARM64_T6030 arm64
• Python Version: 3.12.8
• DeepDiff Version: 8.4.2
• Numpy Version: 2.2.4

Additional context
It looks like there's some other stuff in deepdiff.helper that might be relevant. For instance, the datetimes tuple also doesn't have numpy.datetime64, but adding it there seems unnecessary to fix this issue. It looks like the only use of datetimes is to add to the numbers tuple, and I'm not sure whether this should be considered a number.

[seperman]

Hi @jcipar
Thanks for the detailed explanation. Do you want to make a PR? You can then add yourself to the contributors list too. And please include unit tests.
We do check for isinstance(obj, datetimes) btw and convert datetimes to proper timezone for comparison.

[jcipar]

Sounds good, I can have something ready shortly. I'm not sure where you'd prefer the tests for diffing. It seems like both test_diff_numpy and test_diff_datetime would make sense.