Beginner’s Question: • _I have a number of flows which ingest one or more jsons and transform them into a single dataframe_: • The flows are currently wired to write their output in a target path, that is parameterized (via a block, but this is still TBD). • But I also like to unit-test and in some use cases do not require to land the data at all. In effect I have reusable flows. What is the best pattern for this? Optionally persist as an input parameter, optionally return the DF or None? Control behavior via blocks?

Maybe have a parameter (let's say it's called

target_path

) to the flow that controls output destination, but the param is optional so output writing logic is behind

if target_path:

And then you could have multiple deployments of that flow, with

target_path

set/not set for convenience in the UI

ok, that makes sense, thanks!