Make use of Dask-Arrays in xcube.core.geom

[forman]

In this PR:

• The functions mask_dataset_by_geometry() and rasterize_features() of module xcube.core.geom have been reimplemented to generate lazy dask arrays. Both should now be applicable to datasets that have arbitrarily large spatial dimensions. The spatial chunk sizes to be used can be specified using keyword argument tile_size.

• Deprecated following functions of module xcube.core.geom:

  • is_dataset_y_axis_inverted() is no longer used;
  • get_geometry_mask() is no longer used;
  • convert_geometry() has been renamed to normalize_geometry().

Closes #593.
Closes #666.

Checklist:

• Add unit tests and/or doctests in docstrings
• Add docstrings and API docs for any new/modified user-facing classes and functions
• New/modified features documented in docs/source/*
• Changes documented in CHANGES.md
• AppVeyor CI passes
• Test coverage remains or increases (target 100%)

[codecov-commenter]

Codecov Report

Merging #672 (1b1b1e7) into master (bcf0fb0) will increase coverage by 0.06%.
The diff coverage is 99.55%.

@@            Coverage Diff             @@
##           master     #672      +/-   ##
==========================================
+ Coverage   92.33%   92.39%   +0.06%     
==========================================
  Files         307      307              
  Lines       28844    28973     +129     
==========================================
+ Hits        26632    26769     +137     
+ Misses       2212     2204       -8     

Impacted Files	Coverage Δ
xcube/core/geom.py	97.32% <99.15%> (+2.96%)	⬆️
test/core/test_geom.py	100.00% <100.00%> (ø)
xcube/core/timeseries.py	96.70% <100.00%> (ø)
xcube/core/schema.py	96.04% <0.00%> (+0.79%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update bcf0fb0...1b1b1e7. Read the comment docs.

[pont-us]

Looks good! Demo notebook works nicely, and I successfully ran a number of stress tests with larger data volumes.