FOL

FOL

[Stefan Israelsson Tampe]

Hi,

I got the prolog parser reporting syntax errors for prolog programs. 
It's soo nice to have an small analysis and a position where the error 
is located. With this I can take on writing more complex prolog programs. An
 example that is in the pipe is a FOL (first order logic) 
theorem solver, quite fun and a nice testbed to try out the postpone logic.

For setting up the error reporting I used a pdf
describing the syntax-parse macro

  http://www.ccs.neu.edu/scheme/pubs/icfp10-cf.pdf

I was wondering if you would like to see it ported to guile? I could try
to do that!

Regards
Stefan

Re: FOL

[Andy Wingo]

Hi,

On Thu 16 Sep 2010 21:00, Stefan Israelsson Tampe <stefan.tampe@spray.se> writes:

>   http://www.ccs.neu.edu/scheme/pubs/icfp10-cf.pdf
>
> I was wondering if you would like to see it ported to guile? I could try
> to do that!

I think we could use an implementation in any case. Is there a free
implementation already, based on syntax-case? (Is syntax-parse
implementable in syntax-case?)

I haven't read the paper yet, but it's on my list. It sounds
interesting, and a little daunting as well ;)

Andy
-- 
http://wingolog.org/

Re: FOL

[Ludovic Courtès]

Hi!

Stefan Israelsson Tampe <stefan.tampe@spray.se> writes:

> I got the prolog parser reporting syntax errors for prolog programs. 
> It's soo nice to have an small analysis and a position where the error 
> is located. With this I can take on writing more complex prolog programs. An
>  example that is in the pipe is a FOL (first order logic) 
> theorem solver, quite fun and a nice testbed to try out the postpone logic.

Woow, cool!

What are your plans for your Prolog compiler, in terms of releases,
merges, and whatnot?  :-)

> For setting up the error reporting I used a pdf
> describing the syntax-parse macro
>
>   http://www.ccs.neu.edu/scheme/pubs/icfp10-cf.pdf

Yeah, that seems like an improvement in terms of error detection and
error reporting (better than “source expression failed to match any
pattern”...).

However, I don’t like the aesthetics: keywords, and, worse, type
annotations embedded in identifiers.

Thanks,
Ludo’.

Re: FOL

[Stefan Israelsson Tampe]

> 
> What are your plans for your Prolog compiler, in terms of releases,
> merges, and whatnot?  :-)

This is on my list before I think that it will be merged/released a beta
1.
I'm in the process to replace the matchers with modified Shinn. The main
problem using prompts is that I don't get tail call's, I do have it with 
the fragile umatch and Shinn! Also I will not need to patch guile sources.
I'll use a pure Scheme backend and a c-code version.

2. Write test test code - which now is a good time to start when there are
decent syntax error reports.


> > For setting up the error reporting I used a pdf
> > describing the syntax-parse macro
> > 
> >   http://www.ccs.neu.edu/scheme/pubs/icfp10-cf.pdf
> 
> Yeah, that seems like an improvement in terms of error detection and
> error reporting (better than “source expression failed to match any
> pattern”...).
> 
> However, I don’t like the aesthetics: keywords, and, worse, type
> annotations embedded in identifiers.

The syntax can be fixed, any ideas?

/Stefan

Re: FOL

[Ludovic Courtès]

Hi,

Stefan Israelsson Tampe <stefan.tampe@spray.se> writes:

>> > For setting up the error reporting I used a pdf
>> > describing the syntax-parse macro
>> > 
>> >   http://www.ccs.neu.edu/scheme/pubs/icfp10-cf.pdf
>> 
>> Yeah, that seems like an improvement in terms of error detection and
>> error reporting (better than “source expression failed to match any
>> pattern”...).
>> 
>> However, I don’t like the aesthetics: keywords, and, worse, type
>> annotations embedded in identifiers.
>
> The syntax can be fixed, any ideas?

No, but I think that having a consensual syntax among Schemers may be
more sustainable than rolling our own macro system.  :-)

Thanks,
Ludo’.