Dimension names are lost in mode nczarr,noxarray

[malmans2]

>> nc-config --version
netCDF 4.9.1

When I write a file in mode "nczarr,noxarray" the attributes describing the dimension names are lost.

>> ncdump -h file://test-nczarr-noxarray#mode=nczarr

netcdf test-nczarr-noxarray {
dimensions:
	_zdim_9 = 9 ;
	_zdim_10 = 10 ;
	_zdim_20 = 20 ;
	_zdim_8 = 8 ;
variables:
	double dim2(_zdim_9) ;
		dim2:_FillValue = NaN ;
	int64 numbers(_zdim_10) ;
	int64 time(_zdim_20) ;
		time:units = "days since 2000-01-01 00:00:00" ;
		time:calendar = "proleptic_gregorian" ;
	double var1(_zdim_8, _zdim_9) ;
		var1:_FillValue = NaN ;
		var1:foo = "variable" ;
	double var2(_zdim_8, _zdim_9) ;
		var2:_FillValue = NaN ;
		var2:foo = "variable" ;
	double var3(_zdim_10, _zdim_8) ;
		var3:_FillValue = NaN ;
		var3:foo = "variable" ;
		var3:coordinates = "numbers" ;
}

This is the same file written in mode `"nczarr"`

>> ncdump -h file://test-nczarr#mode=nczarr

netcdf test-nczarr {
dimensions:
	dim2 = 9 ;
	time = 20 ;
	dim1 = 8 ;
	dim3 = 10 ;
variables:
	double dim2(dim2) ;
		dim2:_FillValue = NaN ;
	int64 time(time) ;
		time:units = "days since 2000-01-01 00:00:00" ;
		time:calendar = "proleptic_gregorian" ;
	double var1(dim1, dim2) ;
		var1:_FillValue = NaN ;
		var1:foo = "variable" ;
	double var2(dim1, dim2) ;
		var2:_FillValue = NaN ;
		var2:foo = "variable" ;
	double var3(dim3, dim1) ;
		var3:_FillValue = NaN ;
		var3:foo = "variable" ;
		var3:coordinates = "numbers" ;
	int64 numbers(dim3) ;
}

Here is the code and environment to reproduce the MRE

from xarray.tests.test_dataset import create_test_data
ds = create_test_data()
ds = ds.drop_vars("dim3")
for mode in ("nczarr", "nczarr,noxarray"):
    ds.to_netcdf(f"file://test-{mode.replace(',', '-')}#mode={mode}")

>> conda list
# packages in environment at /Users/mattia/mambaforge/envs/xarray-new:
#
# Name                    Version                   Build  Channel
appnope                   0.1.3              pyhd8ed1ab_0    conda-forge
asciitree                 0.3.3                      py_2    conda-forge
asttokens                 2.2.1              pyhd8ed1ab_0    conda-forge
attrs                     22.2.0             pyh71513ae_0    conda-forge
backcall                  0.2.0              pyh9f0ad1d_0    conda-forge
backports                 1.0                pyhd8ed1ab_3    conda-forge
backports.functools_lru_cache 1.6.4              pyhd8ed1ab_0    conda-forge
bzip2                     1.0.8                h0d85af4_4    conda-forge
c-ares                    1.18.1               h0d85af4_0    conda-forge
ca-certificates           2022.12.7            h033912b_0    conda-forge
cftime                    1.6.2           py311hd5badaa_1    conda-forge
colorama                  0.4.6              pyhd8ed1ab_0    conda-forge
curl                      7.88.1               h6df9250_0    conda-forge
decorator                 5.1.1              pyhd8ed1ab_0    conda-forge
entrypoints               0.4                pyhd8ed1ab_0    conda-forge
exceptiongroup            1.1.0              pyhd8ed1ab_0    conda-forge
executing                 1.2.0              pyhd8ed1ab_0    conda-forge
fasteners                 0.17.3             pyhd8ed1ab_0    conda-forge
hdf4                      4.2.15               h7aa5921_5    conda-forge
hdf5                      1.12.2          nompi_h48135f9_101    conda-forge
icu                       70.1                 h96cf925_0    conda-forge
iniconfig                 2.0.0              pyhd8ed1ab_0    conda-forge
ipython                   8.11.0             pyhd1c38e8_0    conda-forge
jedi                      0.18.2             pyhd8ed1ab_0    conda-forge
jpeg                      9e                   hb7f2c08_3    conda-forge
krb5                      1.20.1               h049b76e_0    conda-forge
libaec                    1.0.6                hf0c8a7f_1    conda-forge
libblas                   3.9.0           16_osx64_openblas    conda-forge
libcblas                  3.9.0           16_osx64_openblas    conda-forge
libcurl                   7.88.1               h6df9250_0    conda-forge
libcxx                    15.0.7               h71dddab_0    conda-forge
libedit                   3.1.20191231         h0678c8f_2    conda-forge
libev                     4.33                 haf1e3a3_1    conda-forge
libffi                    3.4.2                h0d85af4_5    conda-forge
libgfortran               5.0.0           10_4_0_h97931a8_28    conda-forge
libgfortran5              11.3.0              h082f757_28    conda-forge
libiconv                  1.17                 hac89ed1_0    conda-forge
liblapack                 3.9.0           16_osx64_openblas    conda-forge
libnetcdf                 4.9.1           nompi_h56690a0_101    conda-forge
libnghttp2                1.51.0               he2ab024_0    conda-forge
libopenblas               0.3.21          openmp_h429af6e_3    conda-forge
libsqlite                 3.40.0               ha978bb4_0    conda-forge
libssh2                   1.10.0               h47af595_3    conda-forge
libxml2                   2.10.3               hb9e07b5_0    conda-forge
libzip                    1.9.2                h6db710c_1    conda-forge
libzlib                   1.2.13               hfd90126_4    conda-forge
llvm-openmp               15.0.7               h61d9ccf_0    conda-forge
matplotlib-inline         0.1.6              pyhd8ed1ab_0    conda-forge
msgpack-python            1.0.4           py311hd2070f0_1    conda-forge
ncurses                   6.3                  h96cf925_1    conda-forge
netcdf4                   1.6.2           nompi_py311he6f34b2_101    conda-forge
numcodecs                 0.11.0          py311h814d153_1    conda-forge
numpy                     1.24.2          py311ha9d2c9f_0    conda-forge
openssl                   3.0.8                hfd90126_0    conda-forge
packaging                 23.0               pyhd8ed1ab_0    conda-forge
pandas                    1.5.3           py311hd84f3f5_0    conda-forge
parso                     0.8.3              pyhd8ed1ab_0    conda-forge
pexpect                   4.8.0              pyh1a96a4e_2    conda-forge
pickleshare               0.7.5                   py_1003    conda-forge
pip                       23.0.1             pyhd8ed1ab_0    conda-forge
pluggy                    1.0.0              pyhd8ed1ab_5    conda-forge
prompt-toolkit            3.0.38             pyha770c72_0    conda-forge
prompt_toolkit            3.0.38               hd8ed1ab_0    conda-forge
ptyprocess                0.7.0              pyhd3deb0d_0    conda-forge
pure_eval                 0.2.2              pyhd8ed1ab_0    conda-forge
pygments                  2.14.0             pyhd8ed1ab_0    conda-forge
pytest                    7.2.1              pyhd8ed1ab_0    conda-forge
python                    3.11.0          he7542f4_1_cpython    conda-forge
python-dateutil           2.8.2              pyhd8ed1ab_0    conda-forge
python_abi                3.11                    3_cp311    conda-forge
pytz                      2022.7.1           pyhd8ed1ab_0    conda-forge
readline                  8.1.2                h3899abd_0    conda-forge
setuptools                67.4.0             pyhd8ed1ab_0    conda-forge
six                       1.16.0             pyh6c4a22f_0    conda-forge
stack_data                0.6.2              pyhd8ed1ab_0    conda-forge
tk                        8.6.12               h5dbffcc_0    conda-forge
tomli                     2.0.1              pyhd8ed1ab_0    conda-forge
traitlets                 5.9.0              pyhd8ed1ab_0    conda-forge
tzdata                    2022g                h191b570_0    conda-forge
wcwidth                   0.2.6              pyhd8ed1ab_0    conda-forge
wheel                     0.38.4             pyhd8ed1ab_0    conda-forge
xarray                    2023.2.0           pyhd8ed1ab_0    conda-forge
xz                        5.2.6                h775f41a_0    conda-forge
zarr                      2.13.6             pyhd8ed1ab_0    conda-forge
zlib                      1.2.13               hfd90126_4    conda-forge
zstd                      1.5.2                hbc0c0cd_6    conda-forge

[DennisHeimbigner]

Currently, it is the case that (no)xarray implies pure Zarr.
So it is the case that one cannot have NCZarr and XARRAY at the same time.

[malmans2]

Thanks, I think I misunderstood the usage of xarray/noxarray.

To recap, there are 3 scenarios:

• (mode=nczarr): write both _ARRAY_DIMENSIONS and _nczarr_attr
• (mode=zarr; mode=zarr,xarray; mode=nczarr,xarray): write _ARRAY_DIMENSIONS only
• (mode=zarr,noxarray; mode=nczarr,noxarray): do NOT write _ARRAY_DIMENSIONS nor _nczarr_attr

Therefore it is not possible to write _nczarr_attr only?

[DennisHeimbigner]

I think that is correct.
I suppose there is no harm in writing _ARRAY_DIMENSIONS in all cases (unless noxarray), so in effect
that would separate any connection between the (nczarr,zarr) choice and the (xarray,noxarray) choice.
Any opinion on that?b

[malmans2]

Yes, I think that would make it more clear.

Maybe it would be even better to make nczarr and noxarray mutually exclusive, meaning that mode=nczarr,noxarray would raise an error and nczarr is always compatible with xarray by definition.