Re: variable analysis

Hi Jan,

Jan Wielemaker <[email protected]> writes:

> I guess we either need an library that can be called to optimize the
> generated code or we need some way for term/goal expansion to tell the
> compiler not to worry about this code. Ideas?

One thing I noticed is that compiling:

   :- use_module(library(clpfd)).

   n_factorial(0, 1).
   n_factorial(N, F) :- N #> 0, N1 #= N - 1, F #= N * F1,
                        n_factorial(N1, F1).

yields:

   Warning: ... Test is always true: var(N1)

However, compiling the expanded version (generated by listing/2) yields
no warnings (6.5.0-105-g28da4ed). At least this should be consistent.

All the best,
Markus