Alexander Burger wrote:
 Hi Cle,

Hi Alex,

(...)

 But please note that there is one difference (or bug?) that Jon
 Kleiser found out about two years ago, and which I could not
 resolve:

 Anonymous variables ('@' in Pilog, '_' in Prolog) may not work as
 expected, when there is more than one of them in a complex rule,
 with each in a different context.

 I could not find a solution for that problem, and finally decided to
 live with it. I hope this is acceptable.

urghs! This is a severe difference from Prolog. I would call it a bug ;-) But I do not know, if it is acceptable or if it comes into my way and make me stumble.

But anyway, the most important thing is, to *know* about that restriction! Then -- perhaps -- one can solve or work-around it ... so, thank you for this hint!

Now coming to another problem, I am already fiddle for several hours now. There is no Pilog 'retract'. So I decided to implement one. But I would like it to have the same semantic as in Prolog. In Prolog, 'retract' will first try to match and bind a clause, and then remove that clause from the database. Something like this:

   foo(1).
   foo(2).
   foo(3).

   retract(foo(N)), print(N), nl.

Would try to match the clause foo(N). Success with foo(1). Therefore N would be bound to 1 and the clause foo(1) would be retracted.

For this, however, I would have to be able to pass a predicate to 'retract' execute it and retract afterwards.

If I try it interactively I can perform the operation like this:

   (be foo (1))
   (be foo (2))
   (be foo (3))

(? (@CL list 'foo '(@N)) (equal (@P . (@A)) @CL) (call @P . @A) (@ retract (-> @CL))

Here the predicate I want to retract is the first one unifying with '(foo @N)'. This should be the first one '(foo 1)'. So '@N' is bound to '1' and the corresponding clause is retracted.

This works as I expect!

But I seem not to be able to get the code into another clause. I tried it like this:

   (be retract (@CL)
     (equal (@P . (@A)) @CL)
     (call @P . @A)
     (@ retract (-> @CL)) )

But this does not work. If I try

   (? (retract (foo (@N))))

'@N' got not bound to anything, an no fact was retracted. I have tried several other implementations of 'retract' but I cannot get it to work.

So perhaps you can shed any light over me, what I do miss?


Thanks and ciao,
Cle.



--------------000509060708050706030409
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: 8bit

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
 <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
 <title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
<br>
<br>
Alexander Burger wrote:<br>
<span style="white-space: pre;">&gt; Hi Cle,<br>
</span><br>
Hi Alex,<br>
<span style="white-space: pre;"><br>
(...)</span><br>
<br>
<span style="white-space: pre;">&gt; But please note that there is one
difference (or bug?) that Jon<br>
&gt; Kleiser found out about two years ago, and which I could not<br>
&gt; resolve:<br>
&gt; <br>
&gt; Anonymous variables ('@' in Pilog, '_' in Prolog) may not work as <br>
&gt; expected, when there is more than one of them in a complex rule,<br>
&gt; with each in a different context.<br>
&gt; <br>
&gt; I could not find a solution for that problem, and finally decided
to <br>
&gt; live with it. I hope this is acceptable.<br>
</span><br>
urghs! This is a severe difference from Prolog. I would call it a bug
;-) But I do not know, if it is acceptable or if it comes into my way
and make me stumble.<br>
<br>
But anyway, the most important thing is, to *know* about that
restriction! Then -- perhaps -- one can solve or work-around it ... so,
thank you for this hint!<br>
<br>
Now coming to another problem, I am already fiddle for several hours
now. There is no Pilog 'retract'. So I decided to implement one. But I
would like it to have the same semantic as in Prolog. In Prolog,
'retract' will first try to match and bind a clause, and then remove
that clause from the database. Something like this:<br>
<br>
   foo(1).<br>
   foo(2).<br>
   foo(3).<br>
   <br>
   retract(foo(N)), print(N), nl.<br>
<br>
Would try to match the clause foo(N). Success with foo(1). Therefore N
would be bound to 1 and the clause foo(1) would be retracted.<br>
<br>
For this, however, I would have to be able to pass a predicate to
'retract' execute it and retract afterwards.<br>
<br>
If I try it interactively I can perform the operation like this:<br>
<br>
   (be foo (1))<br>
   (be foo (2))<br>
   (be foo (3))<br>
<br>
   (? (@CL list 'foo '(@N)) (equal (@P . (@A)) @CL) (call @P . @A) (@
retract (-&gt; @CL))<br>
<br>
Here the predicate I want to retract is the first one unifying with
'(foo @N)'. This should be the first one '(foo 1)'. So '@N' is bound to
'1' and the corresponding clause is retracted.<br>
<br>
This works as I expect!<br>
<br>
But I seem not to be able to get the code into another clause. I tried
it like this:<br>
<br>
   (be retract (@CL)<br>
     (equal (@P . (@A)) @CL)<br>
     (call @P . @A)<br>
     (@ retract (-&gt; @CL)) )<br>
<br>
But this does not work. If I try<br>
<br>
   (? (retract (foo (@N))))<br>
<br>
'@N' got not bound to anything, an no fact was retracted. I have tried
several other implementations of 'retract' but I cannot get it to work.<br>
<br>
So perhaps you can shed any light over me, what I do miss?<br>
<br>
<br>
Thanks and ciao,<br>
Cle.<br>
<br>
<br>
</body>
</html>

--------------000509060708050706030409--
--
UNSUBSCRIBE: mailto:picol...@software-lab.de?subject=unsubscribe

Reply via email to