Path: ...!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: Julio Di Egidio <julio@diegidio.name>
Newsgroups: comp.lang.prolog
Subject: Re: Can we do this with ancestral cuts or something?
Date: Mon, 9 Dec 2024 17:19:28 +0100
Organization: A noiseless patient Spider
Lines: 44
Message-ID: <vj75ag$clr4$4@dont-email.me>
References: <vj6rmq$clr4$1@dont-email.me> <vj6u0o$l219$1@solani.org>
 <vj73f8$clr4$2@dont-email.me> <vj752t$clr4$3@dont-email.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Injection-Date: Mon, 09 Dec 2024 17:19:28 +0100 (CET)
Injection-Info: dont-email.me; posting-host="7790aa2e18bcf5e7788978594a704952";
	logging-data="415588"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1+FQnskfKBYhPYbdQCVPc9qYC1H6vS7alo="
User-Agent: Mozilla Thunderbird
Cancel-Lock: sha1:/HamzBQFvxjGEOmT5jL9lkSb0k4=
In-Reply-To: <vj752t$clr4$3@dont-email.me>
Content-Language: en-GB
Bytes: 2515

On 09/12/2024 17:15, Julio Di Egidio wrote:
> On 09/12/2024 16:47, Julio Di Egidio wrote:
>> On 09/12/2024 15:14, Mild Shock wrote:
>>> Julio Di Egidio schrieb:
>>
>>>> Given a list of goals representing a conjunction, I would like it to 
>>>> fail as soon as any goal fails, but I would like *not* to cut on the 
>>>> search space otherwise.
>>>>
>>>> Here is my actual test case, only considering ground goals 
>>>> (otherwise too many ways to play with it that seem immaterial to the 
>>>> problem):
>>>
>>> First create a conjunction:
>>>     A1,...,An
>>> Wrap each goal Aj into a soft cut:
>>>    (Aj *-> true; !).
>>> The cut will abort all previous goals.
>>
>> Indeed soft-cuts do it, thank you very much!
>>
>> Though with this scheme: `(Aj *-> true; !, fail)`.
>>
>> I have also managed to abstract the recursive scheme away to a 
>> meta-predicate, by combining soft with ancestral cuts.  But I am not 
>> sure how good it is, especially since term rewriting also can do the 
>> trick and quite more simply so...
> 
> Actually, this one works (no need for a soft cut in `scut`), and it at 
> least starts making sense:
> 
> ```
> scut(G, ChF) :-
>      (    call(G, ChF), !
>      ;    prolog_cut_to(ChF),
>          fail
>      ).
> ```

No no, forget that: that is cutting on the search space.  The previous 
solution is it.

-Julio