Yes. It creates an empty table for a failing variant call.
Its difficult to follow what you are saying if you don’t post the
code. Is it again somewhere on GitHub by any chance?
If the tabling solution can be made fast, I am happy to change
this solution here, which also shortens the code itself, so
that the code is only 621 bytes. Maybe its shorter with a tabling
directive, shorter than 621 bytes, the manual memoization takes
quite some space in the solution:
Prolog, 621 + 2036 - 100 = 2557
answered Oct 7 at 2:02
https://codegolf.stackexchange.com/a/265839/119893