Re: [Yap-users] multiple instances of YAP from c

[Jean-Noel V. <jea...@un...>]

Hi Davide, hi Vitor,

Davide, I needed exactly what you are looking for so I modified Yap 6.2.2 to put all global variables of Yap into a C++ class, changed functions into methods and modified some code to be correct C++. The result is working well (note that signals are broken since there is now more than one Yap).

It might be an extension of the work done in the directory CXX in yap-6.3.4 in which I noticed that only one instance of YAPEngine can be created. The problem is I don’t see a clean way to keep together the C code and the C++ code together in order to make C++Yap only an extension of CYap.

Jean-Noël



Le 15 juin 2014 à 14:19, Vitor Santos Costa <vs...@gm...> a écrit :

> Dear Davide.
> 
> Sorry for the delay
> 
> 
>> is it possible to create multiple prolog interpreters in the same C/C++
>> program?
> 
> Yes, if you have -DTHREADS (or if you use --enable-threads).
> 
>> As far as I know, functions such as YAP_FastInit and YAP_RunGoalOnce are
>> global, so I cannot initialize multiple instances.
> 
> FastInit will create a single database, thaat will be shared by all
> threads. But threads can run separate, and RunGoal shsould be
> thread-local. Have a look at
> 
> YAP_ThreadCreateEngine, YAP_ThreadAttachEngine,
> YAP_ThreadDetachEngine, YAP_ThreadDestroyEngine. YAP also supports the
> thread creation ops in the SWi interface:
> 
> PL_thread_self()
> PL_unify_thread_id()
> PL_thread_attach_engine()
> PL_thread_destroy_engine()
> PL_thread_at_exit()
> 
> 
>> In particular, I'm creating a c++ class that calls yap, so I would like
>> to create a new instance of yap prolog every time I create a new object
>> of the class. Thus every object will call and use yap independently in
>> the same c++ program.
> 
> 
> You may want to have a look at the directory CXX in yap-6.3.4. It is
> designed to interface with the swig package, but it is growing
> quickly. Any feedback, extensions or patches will be much appreciated!
> 
> Vitor
> 
>> 
>> thanks!
>> 
>> regards,
>> Davide
>> 
>> Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
>> 
>> ------------------------------------------------------------------------------
>> HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
>> Find What Matters Most in Your Big Data with HPCC Systems
>> Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
>> Leverages Graph Analysis for Fast Processing & Easy Data Exploration
>> http://p.sf.net/sfu/hpccsystems
>> _______________________________________________
>> Yap-users mailing list
>> Yap...@li...
>> https://lists.sourceforge.net/lists/listinfo/yap-users
> 
> ------------------------------------------------------------------------------
> HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
> Find What Matters Most in Your Big Data with HPCC Systems
> Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
> Leverages Graph Analysis for Fast Processing & Easy Data Exploration
> http://p.sf.net/sfu/hpccsystems
> _______________________________________________
> Yap-users mailing list
> Yap...@li...
> https://lists.sourceforge.net/lists/listinfo/yap-users