Ability to pass in Spatial Index precision as a hint during the join

[halfabrane]

There are use cases where within the same Spark run (ie with the same spark context) we want to join multiple different spatial datasets at different resolutions.
This PR allows you to pass in the resolution as a hint as follows:

import org.apache.spark.sql.magellan.dsl.expressions._
point.join(polygons index 15).where($"point" within $"polygon")

allows you to pass a hint to the optimizer that the index resolution should be set to 15.

Future enhancements:

• Ability to pass in a min/ max precision to greedily cover the geometries?
• Ability to index geometries at the right precision on the fly?

[codecov-io]

Codecov Report

Merging #144 into master will decrease coverage by 0.02%.
The diff coverage is 84.61%.

@@            Coverage Diff             @@
##           master     #144      +/-   ##
==========================================
- Coverage   85.47%   85.44%   -0.03%     
==========================================
  Files          47       48       +1     
  Lines        1363     1374      +11     
  Branches       96       95       -1     
==========================================
+ Hits         1165     1174       +9     
- Misses        198      200       +2

Impacted Files	Coverage Δ
src/main/scala/magellan/Utils.scala	100% <100%> (ø)	⬆️
src/main/scala/magellan/catalyst/SpatialJoin.scala	100% <100%> (ø)	⬆️
src/main/scala/magellan/dsl/package.scala	84.61% <100%> (+4.61%)	⬆️
...main/scala/magellan/catalyst/SpatialJoinHint.scala	50% <50%> (ø)

---

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 5b8c595...938ace8. Read the comment docs.