Looking at this thing from a slight distance, I suggest that you might want to do what you want to do using several statements instead of just one, and that you especially want to be sure that the resulting data structure does not depend on precisely what the function does return.   For example, you might want to put the function-results (whatever they turn out to be ...) say, into an arrayref, the element(s) of that arrayref being “whatever the function returned... no matter what it did return.”

While “one-liners” are sometimes convenient, quite frankly, I sometimes opine that they are more trouble than they’re worth.