Fix (u)int8 upcasting as per docs and numpy

[s-ol]

Close #649.

For testing with CircuitPython, see the branch https://github.com/s-ol/micropython-ulab/tree/fix-int8-upcasting.

Apologies if this breaks any tests that should have been updated, I couldn't get the test build to work on my system.

[v923z]

The upcasting rules are given here:

micropython-ulab/code/ndarray_operators.c

Line 32 in 4bde4ef

uint8 + int8 => int16

, and here https://github.com/v923z/micropython-ulab/blob/master/docs/manual/source/ulab-ndarray.rst#upcasting

[v923z]

Same here.

[v923z]

I think this is OK now, thanks! Could you, please, increment the sub-minor version here

micropython-ulab/code/ulab.c

Line 36 in 4bde4ef

#define ULAB_VERSION 6.4.1

, and add an entry here https://github.com/v923z/micropython-ulab/blob/master/docs/ulab-change-log.md? You'll also have to fix the test https://github.com/v923z/micropython-ulab/actions/runs/7153101407/job/19492990488?pr=650, but I think that's just a single line in https://github.com/v923z/micropython-ulab/blob/master/tests/2d/numpy/operators.py.exp, here

micropython-ulab/tests/2d/numpy/operators.py.exp

Line 97 in 4bde4ef

array([5, 7, 9], dtype=uint16)

.

[v923z]

OK, thanks!