Hi Bob, as your subroutine is only called when the user command indicates that the execute button has been pressed if the editing is successful then the editing is completely invisible to the end user. So in this case it is exactly equivalent to having the code editing inside the start-of-selection. The only way there could be a difference in end user experience is if an error message is thrown inside this subroutine which is why I was asking if it was possible to separate out the editing from the checks.
It seems strange to me to have a routine that edits what the user enters on screen without them knowing about it before starting the start-of-selection event. The only use that comes to mind is to convert an external to internal format but then a conversion routine might be more appropriate. Anyway, I think Klaus's suggestion below should make the program work as you want.
Don't forget to award points if helpful.