Would it be possible to implement a second type of custom transformer (CT) where the creator can state that all attributes created inside the custom transformer are local and therefore will not be added to the feature entering the CT. In the output port(s), those attributes that are the result of the analysis will be visible. In the parent workspace, the user that uses this CT can than be allowed to catch the values returned by the custom transformer ('the function') and store it in attributes with names that can be chosen by the user (or by typing or by value of an attribute).
Implementing a custom transformer in this way, will allow us to completely differentiate between the feature that initiates the CT and the analysis done within the CT AND it will allow us to catch results in different attributenames without overwriting a possible existing attribute.
Is it clear what I'm looking for?
What do you think of this idea?
