The problem with this in its current state is that the set-model word
does not have a static stack effect, so if you define a model slot
named 'foo' then *any* code which accesses 'foo' accessors (on other
tuples, potentially) won't have a stack effect either, which is not
acceptable.
Once we move to 'full static stack safety', ideas like this will be
worth exploring further. At this stage, it would be too fragile.
Slava
On Wed, Dec 3, 2008 at 1:43 AM, Eduardo Cavazos <[EMAIL PROTECTED]> wrote:
> Hello,
>
> This is a much more experimental suggestion than the 'eval:' slot attribute.
>
> This:
>
> TUPLE: <abc> { a model-slot } b c ;
>
> Indicates that the 'a' slot will be a model and the reader and writer will act
> accordingly. I.e. given:
>
> 10 <model> 20 30 <abc> boa
>
> The 'a>>' will return '10' and '>>a' will update the model via 'set-model' so
> that observers are notified.
>
> It's somewhat controversial because the jury is still out on whether or not
> custom reader and writer methods should be used. This idiom is worth
> exploring. If custom methods enjoy wide usage, then words which access the
> slot directly by name would be good to have (can be achived via mirrors...).
>
> Code attached.
>
> Ed
>
> ----------------------------------------------------------------------
>
> IN: syntax
>
> SYMBOL: model-slot
>
> ! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
>
> IN: slots
>
> TUPLE: slot-spec name offset class initial read-only model-slot ;
>
> ! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
>
> : peel-off-attributes ( slot-spec array -- slot-spec array )
> dup empty?
> [
> unclip
> {
> { initial: [ [ first >>initial ] [ rest ] bi ] }
> { model-slot [ [ t >>model-slot ] dip ] }
> { read-only [ [ t >>read-only ] dip ] }
> [ bad-slot-attribute ]
> }
> case
> ]
> unless ;
>
> ! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
>
> :: model-reader-quot ( SLOT-SPEC -- quot )
>
> [let | OFFSET [ SLOT-SPEC offset>> ] |
>
> [ OFFSET slot value>> ] ] ;
>
> ! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
>
> :: define-model-reader ( CLASS SLOT-SPEC -- )
> CLASS
> SLOT-SPEC name>> reader-word
> SLOT-SPEC model-reader-quot
> SLOT-SPEC reader-props
> define-typecheck ;
>
> ! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
>
> :: model-writer-quot ( SLOT-SPEC -- quot )
>
> [let | OFFSET [ SLOT-SPEC offset>> ] |
>
> ! ( val obj -- )
>
> [ OFFSET slot set-model ] ] ;
>
> :: define-model-writer ( CLASS SLOT-SPEC -- )
> CLASS
> SLOT-SPEC name>> writer-word
> SLOT-SPEC model-writer-quot
> SLOT-SPEC writer-props
> define-typecheck ;
>
> ! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
>
> :: define-slot-methods ( CLASS SLOT-SPEC -- )
> SLOT-SPEC model-slot>>
> [
> CLASS SLOT-SPEC define-model-reader
> CLASS SLOT-SPEC define-model-writer
> ]
> [
> CLASS SLOT-SPEC
> [ define-reader ]
> [
> dup read-only>> [ 2drop ] [
> [ name>> define-setter drop ]
> [ name>> define-changer drop ]
> [ define-writer ]
> 2tri
> ] if
> ] 2bi
> ]
> if ;
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> Factor-talk mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/factor-talk
>
-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Factor-talk mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/factor-talk
