Fix in case 'dz' was actually 0-dimensional

[johnomotani]

Fixes f49585e "Changes required to load output from a grid-free BOUT.0.nc file, necessary for MMS tests of the differential operators." in case dz was actually zero dimensional, rather than 2-dimensional.

Sorry, I missed that the tests failed on #309, because the final commit by the linter didn't run all the tests, so they weren't listed in the box, and I forgot for a minute that we have lots of pytest tests that should have been there. This PR should fix the test failures.

[johnomotani]

There seem to be test failures unrelated to #309. Is anyone working on these? For example:

>       n_highres = n.bout.interpolate_parallel("core", n=res_factor)

xbout/tests/test_boutdataarray.py:668: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
xbout/boutdataarray.py:313: in interpolate_parallel
    da = da.bout.from_region(region.name, with_guards={xcoord: 0, ycoord: 2})
xbout/boutdataarray.py:218: in from_region
    return _from_region(self.data, name, with_guards)
xbout/region.py:1744: in _from_region
    result = _concat_lower_guards(result, ds_or_da, mxg, myg)
xbout/region.py:1566: in _concat_lower_guards
    da_lower = da_lower.reset_index(xcoord).set_xindex(xcoord)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <xarray.DataArray 'n' (t: 2, x: 3, theta: 2, zeta: 3)>
array([[[[-2.03570987, -2.03570987, -2.03570987],
         [-1....     None
    regions:        {'core': <xbout.region.Region>\n\tname\tcore\n\txinner_in...
    geometry:       toroidal
name = 'set_xindex'

    def __getattr__(self, name: str) -> Any:
        if name not in {"__dict__", "__setstate__"}:
            # this avoids an infinite loop when pickle looks for the
            # __setstate__ attribute before the xarray object is initialized
            for source in self._attr_sources:
                with suppress(KeyError):
                    return source[name]
>       raise AttributeError(
            f"{type(self).__name__!r} object has no attribute {name!r}"
        )
E       AttributeError: 'DataArray' object has no attribute 'set_xindex'

../../../miniforge3/envs/gen/lib/python3.10/site-packages/xarray/core/common.py:256: AttributeError

[mrhardman]

I also have noticed new errors testing Hermes-3, it seems like my attempt to slice dz.data was too simple a fix.

$ ctest --rerun-failed --output-on-failure
Test project /home/mrhardman/hermes-3-work/hermes-3/build-debug-nonorthogonal
    Start  9: drift-wave
1/2 Test  #9: drift-wave .......................***Failed   21.95 sec
ln: failed to create symbolic link 'hermes-3': File exists
Running B = 0.1
getmpirun: using the default mpirun -np
Traceback (most recent call last):
  File "/home/mrhardman/hermes-3-work/hermes-3/build-debug-nonorthogonal/tests/integrated/drift-wave/runtest", line 42, in <module>
    result = analysis.extract_data("data")
  File "/home/mrhardman/hermes-3-work/hermes-3/build-debug-nonorthogonal/tests/integrated/drift-wave/analysis.py", line 6, in extract_data
    bd = xhermes.open(path)
  File "/home/mrhardman/hermes-3-work/xhermes/xhermes/load.py", line 421, in open
    return open_hermesdataset(
        datapath=os.path.join(path, "BOUT.dmp.*.nc"),
        inputfilepath=os.path.join(path, "BOUT.settings"),
        **kwargs,
    )
  File "/home/mrhardman/hermes-3-work/xhermes/xhermes/load.py", line 31, in open_hermesdataset
    ds = open_boutdataset(
        datapath=datapath,
    ...<4 lines>...
        **kwargs,
    )
  File "/home/mrhardman/hermes-3-work/xBOUT/xbout/load.py", line 361, in open_boutdataset
    ds = geometries.apply_geometry(ds, geometry, grid=grid)
  File "/home/mrhardman/hermes-3-work/xBOUT/xbout/geometries.py", line 217, in apply_geometry
    z1 = z0 + nz * dz.data[0,0]
                   ~~~~~~~^^^^^
  File "/home/mrhardman/hermes-3-work/python3.13.0-hermes-venv/lib/python3.13/site-packages/dask/array/core.py", line 1991, in __getitem__
    index2 = normalize_index(index, self.shape)
  File "/home/mrhardman/hermes-3-work/python3.13.0-hermes-venv/lib/python3.13/site-packages/dask/array/slicing.py", line 826, in normalize_index
    raise IndexError("Too many indices for array")
IndexError: Too many indices for array

[johnomotani]

@mrhardman this PR should fix that error.

[mrhardman]

I confirm that this fix works for me for the Hermes ctest and for the python post processing.