Writing Efficient Code

The ECLiPSe compiler tries its best, however there are some constructs which can be compiled more efficiently than others. On the other hand, many Prolog programmers overemphasise the importance of efficient code and write completely unreadable programs which can be only hardly maintained and which are only marginally faster than simple, straightforward and readable programs. The advice is therefore Try the simple and straightforward solution first! The second rule is to keep this original program even if you try to optimise it. You may find out that the optimisation was not worth the effort.

To achieve the maximum speed of your programs, you must produce the optimised code with the flag debug_compile being off, e.g. by calling nodbgcomp/0 or set_flag(debug_compile, off), or using the pragma nodebug. Setting the flag variable_names can also cause slight performance degradations and it is thus better to have it off, unless variable names have to be kept. Unlike in the previous releases, the flag coroutine has now no influence on the execution speed. Some programs spend a lot of time in the garbage collection, collecting the stacks and/or the dictionary. If the space is known to be deallocated anyway, e.g. on failure, the programs can be often speeded up considerably by switching the garbage collector off or by increasing the gc_interval flag. As the global stack expands automatically, this does not cause any stack overflow, but it may of course exhaust the machine memory.

When the program is running and its speed is still not satisfactory, use the profiling tools. The profiler can tell you which predicates are the most expensive ones, and the statistics tool tells you why. A program may spend its time in a predicate because the predicate itself is very time consuming, or because it was frequently executed. The statistics tool gives you this information. It can also tell whether the predicate was slow because it has created a choice point or because there was too much backtracking due to bad indexing.

One of the very important points is the selection of the clause that matches the current call. If there is only one clause that can potentially match, the compiler is expected to recognise this and generate code that will directly execute the right clause instead of trying several subsequent clauses until the matching one is found. Unlike most of the current Prolog compilers, the ECLiPSe

compiler tries to base this selection ( indexing) on the most suitable argument of the predicate. It is therefore not necessary to reorder the predicate arguments so that the first one is the crucial argument for indexing. However, the decision is still based only on one argument. If it is necessary to look at two arguments in order to select the matching clause, e.g. in

p(a, a) :- a.
p(b, a) :- b.
p(a, b) :- c.
p(d, b) :- d.
p(b, c) :- e.

p(X, a) :- pa(X).
p(X, b) :- pb(X).
p(b, c) :- e.

pa(a) :- a. pa(b) :- b.

pb(a) :- c. pb(d) :- d.

The compiler also tries to use for indexing all type-testing information that appears at the beginning of the clause body:

• Type testing predicates free/1, var/1, meta/1, atom/1, integer/1, real/1, number/1, string/1, atomic/1, compound/1, nonvar/1 and nonground/1.

• Explicit unification and value testing =/2, ==/2, /2 and /2.

• Combinations of tests with ,/2, ;/2, not/1, ->/2.

• Arithmetic testing predicates </2, =</2, >/2, >=/2 if one argument is an integer constant and the other one known to be of the integer type.

• A cut after the type tests.

If the compiler can decide about the clause selection at compile time, the type tests are never executed and thus they incur no overhead. When the clauses are not disjoint because of the type tests, either a cut after the test or more tests into the other clauses can be added. For example, the following procedure will be recognised as deterministic and all tests are optimised away:

    % a procedure without cuts
    p(X) :- var(X), ...
    p(X) :- (atom(X); integer(X)), X \= [], ...
    p(X) :- nonvar(X), X = [_|_], ...
    p(X) :- nonvar(X), X = [], ...

Another example:

    % A procedure with cuts
    p(X{_}) ?- !, ...
    p(X) :- var(X), !, ...
    p(X) :- integer(X), ...
    p(X) :- real(X), ...
    p([H|T]) :- ...
    p([]) :- ...

Integers less than or greater than a constant can also be recognised by the compiler:

    p(X) :- integer(X), X < 5, ...
    p(7) :- ...
    p(9) :- ...
    p(X) :- integer(X), X >= 15, ...

If the clause contains tests of several head arguments, only the first one is taken into account for indexing.

Here are some more hints for efficient coding with ECLiPSe :

• Arguments which are repeated in the clause head and in the first regular goal in the body do not require any data moving and thus they do not cost anything. For example,

  p(X, Y, Z, T, U) :- q(X, Y, Z, T, U).
  

  is as expensive as

  p :- q.
  

  On the other hand, switching arguments requires data moves and so

  p(A, B, C) :- q(B, C, A).
  

  is significantly more expensive.

• When accessing an argument of a structure whose functor is known, unification is better than arg/3. Note, however, that for better maintainability the library structures should be used to define the structures.  

• Tests are generally rather slow unless they can be compiled away (see indexing).
• When processing all arguments of a structure, using =../2 and list predicates is always faster, more readable and easier analyzable by automated tools than using functor/3 and arg/3 loops.

• Similarly, when adding one new element to a structure, using =../2 and append/3 is faster than functor/arg.

• Waking is less expensive than metacalling and more expensive than direct calling. Metacalls, although generally slow, are still a lot faster than in some other Prolog systems.

• Sorting using sort/2 is very efficient and it does not use much space. Using setof/3, findall/3 etc. is also efficient enough to be used every time a list of all solutions is needed.

• using not not Goal is optimised in the compiler to use only one choice point.

• =/2, when expanded by the compiler, is faster than ==/2 or =:=/2.

• call_explicit/2 is optimised away by the compiler if both argument are known.

• Using several clauses is much more efficient than using a disjunction if the clause heads contain nonvariables which can be used for indexing. If no indexing can be made anyway, using a disjunction is slightly faster.

• Conditionals with are compiled more efficiently if the condition is a simple built-in test. However, using several clauses can be faster if the compiler optimises the test away.