Missing linked coordinates of subgroup variable

[zxdawn]

What happened:

I have a NetCDF file that has groups like this:

--/PRODUCT/
--/PRODUCT/SUPPORT_DATA/DETAILED_RESULTS
.....

Screenshots

When I read the variable named air_mass_factor_clear inside /PRODUCT/SUPPORT_DATA/DETAILED_RESULTS, the coordinates are missing. But, it works for variable named nitrogendioxide_tropospheric_column in the /PRODUCT/ group.

What you expected to happen:

Coordinates of variables in subgroups are loaded.

Minimal Complete Verifiable Example:

xr.open_dataset('./201908/S5P_PAL__L2__NO2____20190811T022605_20190811T040734_09459_01_020301_20211110T023157.nc',
                group='PRODUCT/')['nitrogendioxide_tropospheric_column']

xr.open_dataset('./201908/S5P_PAL__L2__NO2____20190811T022605_20190811T040734_09459_01_020301_20211110T023157.nc',
                group='PRODUCT/SUPPORT_DATA/DETAILED_RESULTS/')['air_mass_factor_clear']

Environment:

Output of xr.show_versions()

INSTALLED VERSIONS

commit: None
python: 3.9.7 | packaged by conda-forge | (default, Sep 29 2021, 19:20:46)
[GCC 9.4.0]
python-bits: 64
OS: Linux
OS-release: 5.11.0-40-generic
machine: x86_64
processor: x86_64
byteorder: little
LC_ALL: None
LANG: en_US.UTF-8
LOCALE: ('en_US', 'UTF-8')
libhdf5: 1.12.1
libnetcdf: 4.8.1

xarray: 0.20.1
pandas: 1.3.4
numpy: 1.20.3
scipy: 1.7.3
netCDF4: 1.5.8
pydap: None
h5netcdf: None
h5py: 3.6.0
Nio: None
zarr: 2.10.3
cftime: 1.5.1.1
nc_time_axis: None
PseudoNetCDF: None
rasterio: 1.2.10
cfgrib: None
iris: None
bottleneck: None
dask: 2021.11.2
distributed: 2021.11.2
matplotlib: 3.5.0
cartopy: 0.20.1
seaborn: None
numbagg: None
fsspec: 2021.11.1
cupy: None
pint: 0.18
sparse: None
setuptools: 59.4.0
pip: 21.3.1
conda: 4.11.0
pytest: None
IPython: 7.30.0
sphinx: None

[TomNicholas]

Hi @zxdawn, thanks for the report. Would you say your issue is the same as the one in #1982 ?

[zxdawn]

Hi @TomNicholas, thanks and yes that's the same issue. Shall we close this duplicated one?

[TomNicholas]

Yes I think we should close this as a duplicate. By the way, this isn't a quick fix in xarray.Dataset as it is (see #1982 (comment) as for why), so if we do ever support this behaviour it's likely to be via the DataTree route described in #4118.