Object Attribute Accessor Syntax Extension

Object Attribute Accessor Syntax Extension

[Will M. Farr]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

I would like to learn more about camlp4, so I thought I would undertake 
a simple project, but I would like some comment on my ideas first (in 
the hope that my project would be useful to others, too).  It has 
always irked me that when defining a class which is essentially an 
extensible data-container, one has to specify the data accessor 
functions in addition to the data values:

class body (m : float) (qq : float array) (pp : float array) =
object
   val mutable m = m
   val q = qq
   val p = pp
   val mutable f = Array.make (Array.length pp) 0.0
   method q = q
   method p = p
   method m = m
   method set_m mm = m <- mm
   method set_f ff = f <- ff
   (* etc *)
end

I think it would be nice to have accessor functions defined 
automatically for simple datatypes (sort of a simplified version of the 
:accessor q properties of a data member in CLOS -- Ruby does something 
like this with its :attr_accessor a,b,c notes in the class definiton, 
too, and Python probably has something similar).  The sytax I was 
thinking of using was:

class body (m : float) (qq : float array) (pp : float array) =
object
   val mutable accessor m = m
   val reader q = qq
   val reader p = pp
   val writer f = Array.make (Array.length pp) 0.0
   (* etc *)
end

If anyone would like to comment on this idea (because you hate the 
syntax, or love the syntax, or think it's been done before, or think 
it's a bad idea, etc) before I go ahead and write it, let me know.

Will 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (Darwin)

iD8DBQFCQcr+jFCrhUweU3MRArHIAJoC68xlm5Plr6brtXhnmwkDZ5D6tgCaAy5d
BcHYs6KXDrBhbssPxGDoPmM=
=iy74
-----END PGP SIGNATURE-----

Re: [Caml-list] Object Attribute Accessor Syntax Extension

[Jacques Garrigue]

> I think it would be nice to have accessor functions defined 
> automatically for simple datatypes (sort of a simplified version of the 
> :accessor q properties of a data member in CLOS -- Ruby does something 
> like this with its :attr_accessor a,b,c notes in the class definiton, 
> too, and Python probably has something similar).  The sytax I was 
> thinking of using was:
> 
> class body (m : float) (qq : float array) (pp : float array) =
> object
>    val mutable accessor m = m
>    val reader q = qq
>    val reader p = pp
>    val writer f = Array.make (Array.length pp) 0.0
>    (* etc *)
> end

Looks nice. I actually wrote the same extension (with the same names)
experimentally in the yacc grammar last year. Having it in camlp4
would be more modular. 
Note that in a yacc grammar your syntax would require "reader" and
"writer" to be keywords, but it may be OK with camlp4.
It may also be nice to have another small extension, so that
        o#f <- expr
would be handled as
        o#set_f <- (expr)

Jacques Garrigue

Re: [Caml-list] Object Attribute Accessor Syntax Extension

[Alex Baretta]

Jacques Garrigue wrote:

> It may also be nice to have another small extension, so that
>         o#f <- expr
> would be handled as
>         o#set_f <- (expr)
> 
> Jacques Garrigue

Doesn't this look dangerously similar to C++? Let us not turn to the 
Dark Side of object orientation...

Alex

-- 
*********************************************************************
http://www.barettadeit.com/
Baretta DE&IT
A division of Baretta SRL

tel. +39 02 370 111 55
fax. +39 02 370 111 54

Our technology:

The Application System/Xcaml (AS/Xcaml)
<http://www.asxcaml.org/>

The FreerP Project
<http://www.freerp.org/>

Re: [Caml-list] Object Attribute Accessor Syntax Extension

[Jacques Garrigue]

From: Alex Baretta <alex@barettadeit.com>
> Jacques Garrigue wrote:
> 
> > It may also be nice to have another small extension, so that
> >         o#f <- expr
> > would be handled as
> >         o#set_f <- (expr)
> 
> Doesn't this look dangerously similar to C++? Let us not turn to the 
> Dark Side of object orientation...

I just took the syntax from ruby.
Note that both in ocaml and ruby, you cannot access directly object
fields, so the above syntax is not ambiguous: it can only be syntactic
sugar for a method call.
This is not to be mixed with overloading of assignment in C++, which
changes the semantics of an existing operator.

Jacques Garrigue