Add binding so that inspect will work correctly

closes numpy#21908

The array_api take() doesn't flatten the array by default, so the axis
argument must be provided for multidimensional arrays. However, it should be
optional when the input array is 1-D, which the signature previously did not
allow.

c.f. data-apis/array-api#644

BUG: Fix the signature for np.array_api.take

[skip ci]

BLD: update OpenBLAS to an intermediate commit

Bumps [pypa/cibuildwheel](https://github.com/pypa/cibuildwheel) from 2.14.0 to 2.14.1.
- [Release notes](https://github.com/pypa/cibuildwheel/releases)
- [Changelog](https://github.com/pypa/cibuildwheel/blob/main/docs/changelog.md)
- [Commits](pypa/cibuildwheel@66b46d0...f21bb83)

---
updated-dependencies:
- dependency-name: pypa/cibuildwheel
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>

…ta()

Closes numpygh-24203

…ypa/cibuildwheel-2.14.1

MAINT: Bump pypa/cibuildwheel from 2.14.0 to 2.14.1

`f2py_init_func` is defined in `fortranobject.h` as:

    typedef void (*f2py_init_func)(int *, npy_intp *,
                  f2py_set_data_func, int *);

`f90mod_rules.py` attempts to assign functions with the following
prototype to `func` of `struct FortranDataDef`:

    void(*)(int*, int*, void(*)(char*, int*), int*)

This causes the `test_module_doc.py` test to fail when `npy_intp` is
`long` and clang 16 is used as the compiler because the assignment is
identified as an invalid function pointer conversion, which clang 16
treats as an error by default.

Using `npy_intp` instead of `int` fixes the error.

BUG: random: Fix check for both uniform variates being 0 in random_beta()

MAINT: do not use copyswap in flatiter internals

The implementation of Johnk's algorithm for beta(a, b) could generate nan
values if both a and b were extremely small (i.e. subnormal or a small
multiple of the smallest normal double precision float).

The fix is to handle variate generation in this case by noting that when
both a and b are extremely small, the probability of generating a double
precision value that is not either 0 or 1 is also extremely small.  In
particular, if a and b are less than 3e-103, the probability of generating
a (double precision) value that is not 0 or 1 is less than approximately
1e-100.  So instead of using Johnk's algorithm in this extreme case, we
can generate the values 0 or 1 as Bernoulli trials, with the probability
of 1 being a/(a + b).

Closes numpygh-24266.

BUG: random: Fix generation of nan by beta.

CI: Enable running intel_spr_sde_test with Intel SDE

The check needs to be in the python path, because also the printing
for `str()` subtly changed.

To be the same as tuples, we should actually print `(np.float32(3.), np.int8(1))`
for `str()` (tuples include the repr), while for `repr()` we include the dtype
so we would print similar (but ideally with full precision) to arrays.

This isn't quite ideal, I would be happy to print the full repr in the `str()`
but I guess that might be a bit annoying in practice, so maybe our Numeric types
are special enough for now.

Remove legacy-inner-loop-selector and clean up the masked one
(This was effectivly removed a while ago, but a deprecation was added for
ABI compatibility if compiling against a newer NumPy that still had
the slot).

Removed the `reservedX` struct member docs, seems clear enough to not
have them explicitly...

API: Remove several niche objects for numpy 2.0 python API cleanup

BUG: Move legacy check for void printing

BUG: do not modify the input to ufunc_at

API: Remove legacy-inner-loop-selector

``np.source`` and ``np.lookfor`` was removed (numpygh-24144).

DOC: Remove ``np.source`` and ``np.lookfor``

DOC: fix a couple typos and rst formatting errors in NEP 0053

This is a follow-up to numpygh-24272 which missed a few files.

Disable using BLAS in the PyPy job on Azure because it was broken.
Before this PR, it uses to silently not find OpenBLAS and continue,
now we have to be explicit about it.

BUG: Further fixes to indexing loop and added tests

CI: correct URL in cirrus.star

Fix rst formatting to correct rendered html

CI/BLD: fail by default if no BLAS/LAPACK, add 32-bit Python on Windows CI job

…ble (numpy#24289)

As suggested by Chuck in a review comment on a previous PR. Sending as a
separate follow-up also because it helps to check Cirrus CI triggering
which we are looking at in issue 24280.

[skip azp] [skip circle] [skip actions] [skip travis]