bug in `xcube.core.resampling.resample_in_space

[konstntokas]

Describe the bug
When trying to use resample_in_space to convert Sentinel-2 data from EPSG:4326 to EPGS:3035 a ValueError: new_grid_mapping_from_dataset occurs when trying to get a GridMappingfrom dataset in `xcube.core.resampling.resample_in_space()

To Reproduce
The following script, produces the error:

from xcube.core.store import new_data_store
from xcube.core.resampling import resample_in_space
import pyproj


# create sentinelhub datastore
store = new_data_store("sentinelhub")

# selection parameters
data_id = "S2L2A"
bbox = [9.6, 47.4, 9.8, 47.6]
tile_size = [1024, 1024]
time_range = ("2020-06-01", "2020-07-01")
variable_names = ["B03", "B04"]

# get data in EPSG:4326
spatial_res = 0.00018
ds_4326 = store.open_data(
    data_id,
    variable_names=variable_names,
    bbox=bbox,
    spatial_res=spatial_res,
    time_range=time_range,
    tile_size=tile_size,
)
print(ds_4326)

# get data in EPSG:3035
transformer = pyproj.Transformer.from_crs("EPSG:4326", "EPSG:3035", always_xy=True)
pmin = transformer.transform(bbox[0], bbox[1])
pmax = transformer.transform(bbox[2], bbox[3])
bbox_3035 = [pmin[0], pmin[1], pmax[0], pmax[1]]
spatial_res_3035 = 20
ds_3035 = store.open_data(
    data_id,
    variable_names=variable_names,
    bbox=bbox_3035,
    spatial_res=spatial_res_3035,
    time_range=time_range,
    tile_size=tile_size,
    crs="EPSG:3035",
)
print(ds_3035)

# resample the data
ds_resample = resample_in_space(ds_4326.isel(time=4), ref_ds=ds_3035.isel(time=4))
print(ds_resample)

Expected behavior
The function should convert the projection and resample to the grid mapping of the reference dataset.

[konstntokas]

This now works since #1082