Github user ravipesala commented on the issue:
https://github.com/apache/carbondata/pull/3019
@chenliang613 if we are planning to contribute to Presto then it would be
under presto-hive connector, it is just like how ORC, parquet and other formats
supported under this connector. It would be much simpler and lesser code than
the current implementation. Now we are overriding HiveConnector to add our
carbon format to it, if we change presto-hive connector to add carbon format
then it would be adding of read/write factories to it would be enough.
CarbonData cannot be another individual connector on its own as it lacks
features like metadata , authentication etc which are readily available in
presto-hive connector. That is the reason we are overriding HiveConnector.
---
