[Ometa] True higher order rules

[Ondřej Bílka]

I do not know why they still do not have lambda. 
In ruby-ometa I added lambda without any trouble. 
Year ago I sent mail how js ometa can be improved including lambda
but it seem nothing was implemented. 

If you wish to port ometa compare my implementation https://github.com/neleai/ruby-ometa
vs javascript one. Mine offers better features with simpler
implementation.

I now work on amethyst which started on ometa and is much stronger. 

On Mon, Mar 19, 2012 at 12:36:35PM +0100, Loup Vaillant wrote:
> Higher order rules are currently used like this:
> 
>   ometa Foo {
>     until :rule = apply(rule)
>                 | anything until(rule),
>     close       = ']',
>     until_close = <until('close')>
>   }
>   Foo.matchAll('yada]yada', 'until_close') // prints 'yada]'
> 
> So, to pass the rule 'close' as a parameter, I have to refer to it by
> name (as a string), then invoke it with apply(). This causes 2 problems:
> 
>  - I can't pass around anonymous rules. (this is a limitation of
>    Ometa's syntax)
> 
>  - I can't do partial application:
>      foo('bar(baz)') // ankward attempt at partial application.
>                      // of course, it will fail miserably, as there is
>                      // no rule with parentheses in its name
> 
> In an ideal world, I'd like to be able to write:
> 
>   ometa Foo {
>     until :rule = rule           // apply() is now unnecessary
>                 | anything until(rule),
>     close       = ']',
>     until_close = <until(close)> // no quote around 'close', it is
>                                  // passed directly
>   }
> 
> Or even better:
> 
>   ometa Foo {
>     until :rule = rule
>                 | anything until(rule),
>     until_close = <until(\']')> // no, I don't mean the character ']'
>                                 // So I added a Haskell-like backslash
>                                 // to represent a "lambda" anonymous
>                                 // rule.
>   }
> 
> Now the question is, why Alessandro didn't do this in the first place?
> Are there implementation difficulties?  Which are those?  I'm currently
> trying to port OMeta to Lua, so it would be good to know.
> 
> Thanks,
> Loup.
> _______________________________________________
> OMeta mailing list
> OMeta at vpri.org
> http://vpri.org/mailman/listinfo/ometa

-- 

network down, IP packets delivered via UPS