view release on metacpan or  search on metacpan

lib/AI/Prolog/Article.pod  view on Meta::CPAN

 father(Person) :-
   parent(Person, _),
   male(Person).

In the above rule, the underscore is called an "anonymous vairable" and means
"I don't care what this value is."  Prolog may still bind the variable
internally (though this behavior is not guaranteed), but its value will not be
taken into account when trying to determine if terms unify.

Taking the first term in the rule, the logic engine might try to unify this
with the first fact in the knowledge base, C<parent(sally, tom)>. C<Person>
unifies with I<sally>.  The underscore, C<_>, unifies with I<tom> but since
we stated this unification is unimportant, we can ignore that.

We now have a fact which unifies with the first term in the rule, so we push
this information onto a stack.  Since there are still additional facts we can
try, we set a "choice point" in the stack telling us which fact we last tried.
If we have to backtrack to see a choice point, we move on to the next fact and
try again.

Moving on to the next term in the rule, C<male(Person)>, we know that "sally"
is unified to C<Person>, so we now try to unify C<male(sally)> with all of the
corresponding rules in the knowledge base. Since we can't, the logic engine
backs up to the last item where we could make a new choice and sees
C<parent(bill, tom)>. C<Person> gets unified with I<bill>. Then in moving to
the next rule we see that we unify with C<male(bill)>. Now, we check the first
item in the rule and see that it's C<father(Person)>. and the logic engine
reports that I<bill> is a father.

Note that we can then force the engine to backtrack and by continuously
following this procedure, we can determine who all the fathers are.

And that's how logic programming works. Simple, eh? (Well, individual items can
be lists or other rules, but you get the idea).

=head2 Executing Prolog in Perl

Getting back to Perl, how would we implement that in a Perl program?

The basic process for using C<AI::Prolog> looks something like this:

 use AI::Prolog;
 my $prolog = AI::Prolog->new($prolog_code);

Create a new C<AI::Prolog> object, passing Prolog code as the argument.  If you
prefer, you can wrap the constructor in a C<BEGIN> block:

 my $prolog;
 BEGIN {
   $prolog = AI::Prolog->new(<<'  END_PROLOG');
     % some Prolog code goes here
   END_PROLOG
 }

This is not strictly necessary, but if your Prolog code has a syntax error, it
will be a compile-time error, not a run-time error, and you'll get an error
message similar to:

 Unexpected character: (Expecting: ')'.  Got (.)) at line number 12.
 BEGIN failed--compilation aborted at test.pl line 7.

Note that the line number for "Unexpected character" is relative to the Prolog
code, not the Perl code.

After the contructor, issue your query:

 $prolog->query($some_query);

And do something with the results:

 while ( my $results = $prolog->results ) {
     print "@$results\n";
 }

Results are usually each returned as an array reference with the first argument
being the functor and subsequent arguments being the values.  If any value is a
list, it will be represented as an array reference.  We'll see more on that
later as we cover lists.

Now let's see the full program:

 #!/usr/bin/perl
 use strict;
 use warnings;
 use AI::Prolog;
 
 my $prolog;
 
 # If reading from DATA, we need a CHECK block to ensure that
 # DATA is available by the time the constructor is called
 CHECK {
   $prolog = AI::Prolog->new( do { local $/; <DATA> } );
 }
 
 $prolog->query( 'father(WHO).' );
 while ( my $results = $prolog->results ) {
     print "@$results\n";
 }
 
 __DATA__
 parent(sally, tom).
 parent(bill, tom).
 parent(tom, sue).
 parent(alice, sue).
 parent(sarah, tim).
 
 male(bill).
 male(tom).
 male(tim).
 
 father(Person) :-
     parent(Person, _),
     male(Person).

If you run this program, it will quite happily print out "father bill" and
"father tom."  In fact, if you really want to see what's going on internally,
after you issue the query you can "trace" the execution:

 $prolog->query('father(Who)');
 $prolog->trace(1); # after the query, before the results

lib/AI/Prolog/Article.pod  view on Meta::CPAN

 X = [], Y = [1,2,3,4,5] 
 X = [1],  Y = [1,2,3,4]
 X = [1,2],  Y = [1,2,3]
 X = [1,2,3],  Y = [1,2]
 X = [1,2,3,4],  Y = [1]
 X = [1,2,3,4,5], Y = []

And the Perl equivalent:

  use AI::Prolog;
  
  my $prolog = AI::Prolog->new(<<"END_PROLOG");
      append([], X, X).
      append([W|X], Y, [W|Z]) :- append(X, Y, Z).
  END_PROLOG
  
  my $list = $prolog->list( qw/1 2 3 4 5/ );
  
  $prolog->query("append(X,Y,[$list]).");
  while ( my $results = $prolog->results ) {
      my ( $x, $y, $z ) = @{ $results }[ 1, 2, 3 ];
      $" = ', '; # Array separator
      print "[@$x],     [@$y],     [@$z]\n";
  }

As you can see, Prolog lists will be returned to Perl as array references.
C<< $results->[0] >> is the name of the predicate, C<append>, and the next three
elements are the successive values generated by Prolog.

=head1 Problems with Prolog

This article wouldn't be complete without listing some of the issues that have
hampered Prolog.

Perhaps the most significant is the depth-first exhaustive search algorithm
used for deduction.  This is slow and due to the dynamically typed nature of
Prolog, many of the optimizations that have been applied to databases cannot be
applied to Prolog.  Many Prolog programmers, understanding how the language
works internally, use the "cut" operator, C<!> (an exclamation point), to prune
the search trees.  This leads to our next problem.

The "cut" operator is what is known as an "extra-logical" predicate.  This,
along with I/O and predicates that assert and retract facts in the database
are a subject of much controversy.  They cause issues because they frequently
cannot be backtracked over and the order of the clauses sometimes becomes
important when using them.  They can be very useful and simplify many problems,
but they are more imperative in nature than logical and can introduce subtle
bugs.

Math is also handled poorly in Prolog.  Consider the following program:

  convert(Celsius, Fahrenheit) :-
       Celsius is (Fahrenheit - 32) * 5 / 9.

You can then issue the query C<convert(Celsuis, 32)> and it will dutifully
report that the celsius value is zero.  However, C<convert(0, 32)> fails.  This
is because Prolog is not able to solve the right hand side of the equation
unless C<Fahrenheit> has a value at the time that Prolog starts examining the 
equation.  One simplistic way of getting around this limitation is with multiple
predicates and testing which argument is an unbound variable:

 convert(Celsius, Fahrenheit) :-
     var(Celsius),
     not(var(Fahrenheit)),
     Celsius is (Fahrenheit - 32) * 5 / 9.

 convert(Celsius, Fahrenheit) :-
     var(Fahrenheit),
     not(var(Celsius)),
     Fahrenheit is (Celsius * (9/5)) + 32.

This has a variety of problems, not the least of which is that it offers no
advantages over imperative programming.

ISO-Prolog does not define it, but many Prolog implementations support
constraints and these, though beyond the scope of this article, can get around
this and other issues.  They can allow "logical" math and ensure that
"impossible" search branches are never explored.  This can help to alleviate
many of the aforementioned concerns.

=head1 Conclusion

We've barely scratched the surface of what the Prolog language can do.  The
language has many detractors and some criticisms are quite valid.  However, the
language's simplicity and ease-of-use has kept it around.  It's an excellent
starting point for understanding logic programming and exploration of AI.  In
fact, many common uses for Prolog are heavily used in the AI arena:

=over 4

=item * Agent-based programming

=item * Expert Systems

=item * Fraud prevention (via inductive logic programming)

=item * Natural language processing

=item * Rules-based systems

=item * Scheduling systems

=item * And much more (it was even embedded in Windows NT) 

=back

At the present time, I would not recommend C<AI::Prolog> for production work.
Attempts to bring logic programming to Perl are still relatively recent and no
module (to this author's knowledge) is currently ready for widespread use.  The
C<AI::Prolog> distribution has a number of sample programs in the C<examples/>
directory and two complete, albeit small, games in the C<data/> directory.

=head1 References

=head2 Online Resources

=over 4

=item Adventure in Prolog