Consolidate CompressionOption enums

[jeffreyAaron]

Is your feature request related to a problem? Please describe.
There are currently two CompressionOption enums providing similar compression functionality. One of these CompressionOptions is used for S3 pipelines and the other is being used for KafkaBuffer. When additional compression engines are added to one CompressionOption, these changes are not reflected in the other CompressionOption.

S3 CompressionOption: org.opensearch.dataprepper.plugins.sink.s3.compression
Codec/KafkaBuffer CompressionOption: org.opensearch.dataprepper.plugins.codec

Adding ZSTD compression functionality to KafkaBuffer did not provide additional ZSTD functionality in the S3 pipeline.

Ideally, the two enums should be consolidated to reduce duplicate code.

[KarstenSchnitter]

Thanks for providing this issue. I am not sure, that each data store supports all available compression codecs. Kafka supports lz4, gzip, snappy, and zstd. S3 might support more. If both options are joined, how will unsupported codecs be handled?

[kkondaka]

@KarstenSchnitter I think we can add check for supported/unsupported options for each of the sources as applicable.

[dlvenable]

I do think we should try to consolidate these.

With Kafka specifically we have also done some recent work to perform compression before encryption (#5778) since this is generally better for performance. So in that case, we use the custom compression algorithms.

I think from the perspective of a user, they just want to compress. They probably don't care much if Data Prepper is compression using our own compression enum or if the Kafka client is doing it.

And yes, we could have different plugins support only subsets of the compression options.