Allow numeric bounds when clipping dimensionless quantities - #282

Clipping a dimensionless Quantity with numeric limits currently raises AttributeError:

import quantities as pq

q = pq.Quantity([-2., 0.5, 3.])
print(q.clip(0, 1))

This should return [0., 0.5, 1.] dimensionless. Convert the bounds to quantities before rescaling them, after the existing check that requires quantity-valued bounds for dimensional inputs.

Fixes #108. The regression covers scalar, array and one-sided bounds, np.clip, output buffers and incompatible units. Both the full suite and README doctests pass on Python 3.10/NumPy 1.24 and Python 3.12/NumPy 2.4 and 2.5. Mypy passes on NumPy 1.24 and 2.4; NumPy 2.5 has the same quantity.pyi error on unchanged master, already addressed by #277. The advisory lint results are unchanged.

I also checked clipping normalized channels from PhysioNet's S001R01 EEG recording against NumPy magnitude calculations: all 24 previously failing configurations now complete, and the eight configurations with explicit Quantity bounds retain identical results.