[mercury-users] fact tables (was Re: &&/if-then-else semantics (was Re: Request for comments on this code

[Peter Moulder]

On Fri, Jul 14, 2006 at 07:16:16AM -0500, doug.auclair at logicaltypes.com wrote:

> >The mercury compiler itself has memory difficulties when 
> >compiling computer-generated source files.
> 
> Would this file be a fact-table equivalent?

Maria's Hal work would sometimes involve the compiler using hundreds of
megabytes, though I'm not sure what exactly she was compiling.

I've certainly had problems with the fact-table stuff -- even when I
actually used the fact_table declaration.

I think the 0.13 NEWS mentions doing better for this sort of code ("We
have significantly improved the compiler's performance on predicates
with many clauses."), though I haven't compared.

In my case, I first tried using libsqlite (after trying fact tables and
trying hlc grade), but found sqlite's runtime performance somewhat
lacking: it didn't make good use of the fact that the input arguments
were all smallish non-negative ints (corresponding to enumerated union
types).  So I ended up doing my own generated C code to do the look-up.
I wonder if the code generation code might be useful for the Mercury
compiler.

pjrm.
--------------------------------------------------------------------------
mercury-users mailing list
post:  mercury-users at csse.unimelb.edu.au
administrative address: owner-mercury-users at csse.unimelb.edu.au
unsubscribe: Address: mercury-users-request at csse.unimelb.edu.au Message: unsubscribe
subscribe:   Address: mercury-users-request at csse.unimelb.edu.au Message: subscribe
--------------------------------------------------------------------------