🚀 go-pugleaf

In article <7141@skye.ed.ac.uk> ken@aiai.ed.ac.uk (Ken Johnson) writes:
>
>   In article <189@syllogi.syllogic.nl> fdb@syllogic.nl (Frank de Bruijn)
>   writes:
>
>   #  ...a meta-interpreter could handle the symmetry, transitivity, etc.  of
>   #  all relations declared as such.  One implementation technique that can
>   #  be used in such a meta-interpreter is to examine the goal stack for an
>   #  earlier call-frame that matches the current call frame.  This works, but
>   #  it's costly......... 
>
>   I hadn't thought of that.  It is a fascinating possibility.  Have you
>   actually implemented anything this way?

This is something the metalogic programming language Reflective Prolog
is really good at. In fact, this very problem is one of the first
examples the authors take up in their papers about the language. There
is a paper by Costantini and Lanzarone about Reflective Prolog in the
proceedings of the Lisbon ICLP. There is a slow prototype
implementation available from University of Milan and an efficient
compiling version (of a somewhat modified language) is under
development through a collaboration between University of Uppsala and
University of Milan.

-- Jonas B.
--
Jonas Barklund, Computing Science Department, Uppsala University
Email jonas@csd.uu.se, phone +46-18-181050, fax +46-18-521270