add filter_bbox

[clausmichele]

Adding the filter_bbox process

[codecov]

Codecov Report

Merging #118 (39d82ee) into main (1b6e369) will decrease coverage by 1.19%.
The diff coverage is 55.55%.

@@            Coverage Diff             @@
##             main     #118      +/-   ##
==========================================
- Coverage   77.84%   76.65%   -1.19%     
==========================================
  Files          25       25              
  Lines        1020     1071      +51     
==========================================
+ Hits          794      821      +27     
- Misses        226      250      +24     

Impacted Files	Coverage Δ
...sses_dask/process_implementations/cubes/_filter.py	59.59% <54.71%> (-7.08%)	⬇️
..._dask/process_implementations/cubes/_xr_interop.py	92.42% <100.00%> (ø)

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[LukeWeidenwalker]

Thanks for this change, couple of comments!

[clausmichele]

@LukeWeidenwalker I updated the code using the .openeo accessor. However I noticed that calling data.openeo.spatial_dims throws an exception if the data has a single spatial dimension (you can get to this state using reduce_dimension(dimension='x',reducer='mean') for example). So, I updated the accessor to handle this scenario.

Additionally, there's also another edge case I'm not sure how to solve, when there's a single value on x_dim or y_dim: we need to check if this coordinate is actually inside the bbox or not. If not, should we raise an exception?

[LukeWeidenwalker]

@clausmichele I just merged #125, please let me know if this doesn't resolve the issue that was blocking this PR!

[clausmichele]

@LukeWeidenwalker now it should be fine and ready to be merged, thanks!
There's still an edge case which I didn't have time to decide how to handle, but the process is already functional like this.
I will try to fix that later on in another PR.

[LukeWeidenwalker]

Thanks for the changes, this lgtm now!