Keeper Integration API

[osopardo1]

This PR is changing ONLY the API to connect to the Keeper. It does not introduce any logic.
PR #77 is the one solving #41. Here I only take care of matching calls between the different libraries we want to implement.

In order to delegate the maintenance of the index to a Keeper instance, this is the configuration you need to use:

$SPARK_HOME/bin/spark-shell --master local[*]\
 --packages io.delta:delta-core_2.12:1.0.0,io.qbeast:qbeast-spark-keeper_2.12:0.1.0-a2\
 --jars ./target/scala-2.12/qbeast-spark-assembly-0.2.0.jar\
 --conf spark.sql.extensions=io.qbeast.spark.internal.QbeastSparkSessionExtension\
 --conf spark.qbeast.keeper.host=localhost\
 --conf spark.qbeast.keeper.port=50051

As you can see, there's three new parameters:

1. --packages io.qbeast:qbeast-spark-keeper_2.12:0.1.0-a2 -> Package of the qbeast-spark-keeper driver. It maps calls between the qbeast-spark and qbeast-keeper. It also shades grpc/netty libraries to the Spark.
2. --conf spark.qbeast.keeper.host=localhost -> The ip/host of the keeper
3. --conf spark.qbeast.keeper.port=50051 -> The port of the keeper

[eavilaes]

Looks good to me!

Should we add some tests to check that --conf spark.qbeast.keeper.XXX is used in the Keeper?

[cugni]

I have some doubts about how the keeper service is defined. Also, I can't find documentation explaining how (at least not in this PR 😅)

[cugni]

My bad, I misread the cood. It seems all good 👍🏻