support to configure multiple partitioner and encode sequentially

[daehokimm]

Hi All :)

TL;DR

When the s3 partition is blunt, athena's query take long time.

---

In my environment, the s3 sink connector encode partition by FieldPartitioner.
As a result, one of s3 partition has so many objects.
and it cause long time query when using athena.
if s3 sink connector can configure more than one partitioner(like FieldPartitioner + DailyPartitioner), the s3 objects are encoded in detail.

So more than one partitioning is needed. plz support this 🙏

[raphaelauv]

You can write your own partitionner and use it with the confluent connector

Look to the current partitionners present in the repo