[antonwolfy]

NumPy 2.6 defines numpy.positive and the unary + operator for boolean arrays: they return a boolean copy of the input instead of raising TypeError. This PR aligns dpnp.positive and + on dpnp.ndarray with that behavior.

>>> import dpnp as np
>>> x = np.array([True, False])
>>> +x
array([ True, False])

dpnp.tensor.positive has no boolean loop, so for boolean input dpnp.positive runs the dpnp.tensor abs kernel, which returns boolean values unchanged. NumPy reuses its absolute loop for this case the same way. Other dtypes still use the positive kernel. dpnp.tensor.positive and + on usm_ndarray are unchanged.

The acceptance function that raised TypeError for boolean input is removed, since boolean input now has its own loop and never goes through type promotion.

• Have you provided a meaningful PR description?
• Have you added a test, reproducer or referred to an issue with a reproducer?
• Have you tested your changes locally for CPU and GPU devices?
• Have you made sure that new changes do not introduce compiler warnings?
• Have you checked performance impact of proposed changes?
• Have you added documentation for your changes, if necessary?
• Have you added your changes to the changelog?

[github-actions]

View rendered docs @ https://intelpython.github.io/dpnp/pull/3128/index.html

[coveralls]

coverage: 82.644% (+0.004%) from 82.64% — support-boolean-positive into master