Chunking works only with dask installed

I am not sure if this is a bug and, if so, of what.

>>> path = "tests/data/S1B_IW_SLC__1SDV_20210401T052622_20210401T052650_026269_032297_EFA4.SAFE/manifest.safe"
>>> group = 'IW1/R168-N471-E0118'
>>> xs.sentinel1.open_dataset(path, group=group)
<xarray.Dataset>
Dimensions:           (azimuth_time: 1501, slant_range_time: 21632)
Coordinates:
    line              (azimuth_time) float64 0.5 1.5 2.5 ... 1.5e+03 1.5e+03
    pixel             (slant_range_time) float64 0.5 1.5 ... 2.163e+04 2.163e+04
  * azimuth_time      (azimuth_time) datetime64[ns] 2021-04-01T05:26:24.20999...
  * slant_range_time  (slant_range_time) float64 0.005343 0.005343 ... 0.005679
Data variables:
    VH                (azimuth_time, slant_range_time) complex128 ...
    VV                (azimuth_time, slant_range_time) complex128 ...
Attributes: (12/14)
    constellation:              sentinel-1
    platform:                   sentinel-1b
    instrument:                 ['c-sar']
    sat:orbit_state:            descending
    sat:absolute_orbit:         26269
    sat:relative_orbit:         168
    ...                         ...
    sar:instrument_mode:        IW
    sar:polarizations:          ['VV', 'VH']
    sar:product_type:           SLC
    xs:instrument_mode_swaths:  ['IW1', 'IW2', 'IW3']
    Conventions:                CF-1.7
    history:                    created by xarray_sentinel-999

>>> xs.sentinel1.open_dataset(path, group=group, chunks=1000)
---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
<ipython-input-5-79175c680e53> in <module>
---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
<ipython-input-7-90f5963f6ea0> in <module>
----> 1 xs.sentinel1.open_dataset(path, group=group, chunks=1000)

~\xarray-sentinel\xarray_sentinel\sentinel1.py in open_dataset(product_urlpath, drop_variables, group, chunks, fs)
    384         else:
    385             annotation_path = list(groups[group]["annotation_path"].values())[0]
--> 386             ds = open_burst_dataset(
    387                 manifest_path,
    388                 measurement_paths=groups[group]["measurement_path"],

~\xarray-sentinel\xarray_sentinel\sentinel1.py in open_burst_dataset(manifest_path, burst_position, measurement_paths, annotation_path, chunks)
    297     data_vars = {}
    298     for pol, data_path in measurement_paths.items():
--> 299         arr = rioxarray.open_rasterio(data_path, chunks=chunks)
    300
    301         arr = arr.squeeze("band").drop_vars(["band", "spatial_ref"])

~\miniconda3\envs\xr-sentinel\lib\site-packages\rioxarray\_io.py in open_rasterio(filename, parse_coordinates, chunks, cache, lock, masked, mask_and_scale, variable, group, default_name, **open_kwargs)
    863
    864     if chunks is not None:
--> 865         result = _prepare_dask(result, riods, filename, chunks)
    866
    867     # Make the file closeable

~\miniconda3\envs\xr-sentinel\lib\site-packages\rioxarray\_io.py in _prepare_dask(result, riods, filename, chunks)
    573     """
    574     # pylint: disable=import-outside-toplevel
--> 575     from dask.base import tokenize
    576
    577     # augment the token with the file modification time

ModuleNotFoundError: No module named 'dask'

[aurghs]

1. we could add an xfails on the tests with dask
2. I agree
3. I think @corrado9999 was preparing the PR.

[alexamici]

1. we do install dask for testing already, no need to test without dask, we don’t have any dask related logic to test. (I propose no action)
2. dask is not a needed dependency so it must not be listed in the install_requires (no action)
3. we agree, @aurghs would you please take care of it?