gianm commented on issue #13923:
URL: https://github.com/apache/druid/issues/13923#issuecomment-1477330662

   Ah, I hadn't thought about the cross-extension dependency. I had imagined 
the `iceberg` source literally delegating to the `hdfs` one, but yeah, that 
wouldn't be available as a Java object.
   
   I think we can fix this though. Even though InputSources from other 
extensions aren't available as Java objects, they're still available from class 
registries that may be created in core, such as the ObjectMapper deserializers. 
So, you could create the JSON in Java code, then use Jackson to turn it into an 
InputSource and use it that way. This is what allows, for example, the 
`druid-multi-stage-query` extension to use any kind of InputSource from any 
extension, without depending on those extensions.
   
   Wonder if this is a good path or if it's too weird?
   
   If too weird, it'd also be an option to create a different kind of registry 
instead of using ObjectMapper. `ComplexMetricSerde` could be a model here.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to