usm_ndarray.real and usm_ndarray.imag now set flags correctly

[ndgrigorian]

The real and imag properties of usm_ndarray were setting the flags of the returned view incorrectly.

In most cases, the output was not contiguous, but the input was, which caused most functions to behave incorrectly on views into the real or imaginary parts of complex arrays.

This PR corrects that, and also allows real and imag to work on 0-dimensional arrays.

• 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?
• If this PR is a work in progress, are you opening the PR as a draft?

[github-actions]

View rendered docs @ https://intelpython.github.io/dpctl/pulls/1355/index.html

[coveralls]

coverage: 85.584%. remained the same when pulling 72775fc on flags-real-imag-views into 96293fd on master.

[github-actions]

Array API standard conformance tests for dpctl=0.14.6dev4=py310ha25a700_2 ran successfully.
Passed: 913
Failed: 87
Skipped: 119

[oleksandr-pavlyk]

Good catch, @ndgrigorian !

It is reminds me that it is possible to set up real/imag values of complex arrays:

import dpctl.tensor as dpt

sp_vals = dpt.asarray([dpt.nan, -dpt.nan, dpt.inf, -dpt.inf, -1.0, -0.0, 0.0, 1.0], dtype="f4")
x = dpt.empty((sp_vals.shape[0], sp_vals.shape[0]), dtype="c8")
x.real[...] = sp_vals[:, dpt.newaxis]
x.imag[...] = sp_vals[dpt.newaxis, :]

[github-actions]

Deleted rendered PR docs from intelpython.github.com/dpctl, latest should be updated shortly. 🤞

[github-actions]

Array API standard conformance tests for dpctl=0.14.6dev4=py310ha25a700_2 ran successfully.
Passed: 916
Failed: 84
Skipped: 119