Refactoring common operations into module

I have module that takes a variety of different tables (different column names) strips this down to the Key and Tags columns, does a split text to rows and then writes to out to a DB based on parameters sent to the module

I was able to do this once with a SANDBOX, but now I want to remove the sandbox before the SELECT-COLUMNS so that this module can be shared by many other module invokers.

the issue I have is it now complains about missing columns

I could try hacking around with the Input and Create Column transforms, and I bet your would appreciate this so that I was not bothering you so much in the forum. But I suspect this is a pattern that others might come across, so it might be useful to store as a forum question.

Here is the same module with an INPUT transform at the front:

As an idea - have an Input action, then have a conditionally derived table from it so that actions in the derived table are executed only when the input dataset is not empty. In this case there will be no error messages about missing columns even if there is no data.

True, but the derived table also looks like it is giving an error.

Feature suggestion: have some way of mocking up the Input schema with a dummy so that the steps further down don’t show errors. Right now, I can do this manually with a PASTE to INPUT, but it disappears each time the project is reloaded.

It’s not an error. It’s a message that informs that the actions in the derived table were skipped.

I’ve created a new topic:

1 Like