[Caml-list] record declaration, SML

[Caml-list] record declaration, SML

[A. Ozmen]

Hi,

What's best translation of the following SML type?
datatype t = C of { f : int }

I have to define extra record types, I guess. Then, not including these
types in interface (.mli) files? Is this the right approach?

Thanks.
__________________________________________________________
Get your Private, Free Email from HTTP://www.DmailMan.Com
-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

Re: [Caml-list] record declaration, SML

[brogoff]

Hi,
    You've noticed that OCaml records aren't like SML records. Here are a few 
places where they differ. 

You can't have anonymous records in OCaml, so in the case of a datatype 
declaration like yours you need something like 

type t = C of t_record 
and t_record = { f : int } 

If you hide the record definition, then you can't match on it, but I imagine 
you know that already. 

In the case of anonymous records as function arguments, you could try labels. 

In the case of records in a module sharing field names, well, you can't do 
that either, so you'll need to use classes or find some way to disambiguate 
the fields. This is one of those annoyances of Caml (like the lack of 
overloading :) that you just need to get used to. SML is a bit better here 
IMO, but still not ideal; I think that there was an SML variant (SML#) that 
was appealing but it looks like that idea isn't going anywhere. 

There are a few other conveniences in SML record handling too, like the ... 
notation, that don't exist in OCaml. 

Oh yeah, also be careful initializing records with stateful functions, because 
you may be surprised at the order of evaluation. 

-- Brian

On Thu, 9 Jan 2003, A. Ozmen wrote:

> Hi,
> 
> What's best translation of the following SML type?
> datatype t = C of { f : int }
> 
> I have to define extra record types, I guess. Then, not including these
> types in interface (.mli) files? Is this the right approach?
> 
> Thanks.
> __________________________________________________________
> Get your Private, Free Email from HTTP://www.DmailMan.Com
> -------------------
> To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
> Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> 

-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

Re: [Caml-list] record declaration, SML

[A. Ozmen]

Thanks for the answers.

Another thing with the records is that, if I don't open the module the
record defined in, I also have to prefix field names with the module name,
even though I prefix the constructor (M.C{M.f=1}). 

ps:  Anybody having problems with yahoo?  I tried to subscribe to the
beginners list.  Gives some server error.  I also experienced the same
a few weeks ago with free email subscription.
__________________________________________________________
Get your Private, Free Email from HTTP://www.DmailMan.Com
-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

Re: [Caml-list] record declaration, SML

[Xavier Leroy]

> > What's best translation of the following SML type?
> > datatype t = C of { f : int }
> In the case of records in a module sharing field names, well, you can't do 
> that either, so you'll need to use classes or find some way to disambiguate 
> the fields.

Sometimes, it's acceptable to just omit the record type:

        type t = C of int

You lose the naming of the arguments of the constructor, but for small
numbers of arguments (e.g. 1 or 2), this is often tolerable.

> There are a few other conveniences in SML record handling too, like the ... 
> notation, that don't exist in OCaml. 

You mean, in pattern-matching over records?  Caml offers the same
functionality without the ... notation, e.g.

        type r = { x: int; y: int }

        match r with { x = 1 } -> ...

Because records are declared in advance, there is no requirement that
all record labels be mentioned in a pattern matching.

- Xavier Leroy
-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

Re: [Caml-list] record declaration, SML

[brogoff]

On Fri, 10 Jan 2003, Xavier Leroy wrote:
> > There are a few other conveniences in SML record handling too, like the ... 
> > notation, that don't exist in OCaml. 
> 
> You mean, in pattern-matching over records?  Caml offers the same
> functionality without the ... notation, e.g.
> 
>         type r = { x: int; y: int }
> 
>         match r with { x = 1 } -> ...
> 
> Because records are declared in advance, there is no requirement that
> all record labels be mentioned in a pattern matching.

Right, but if I'm not mistaken, SML forces you to use the ... in record 
pattern matching when you want to ignore some labels in the match. As you 
suggest, this is a lot more important in SML, but even in OCaml it would be 
a bit better IMO to explicitly distinguish between ignoring some labels 
and forgetting some, so that slovenly programmers (like yours truly) could 
depend on the type checker to slap them when they miss fields. Unfortunately, 
OCaml behaves as though every record pattern match had an implicit ..., so I 
don't see a good way out even if the developers agreed that this was worth 
fixing.

File that one under "petty complaints", or, if you're feeling generous in this 
new year, under the non-petty "Oh how I wish we had more polymorphism in 
records!" complaint :-)

-- Brian


-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

Re: [Caml-list] record declaration, SML

[Daniel de Rauglaudre]

Hi,

> What's best translation of the following SML type?
> datatype t = C of { f : int }

A solution is:
  type t = C of < f : int >

It is what does the converter SML->OCaml by Camlp4:
  camlp4 pa_sml.cmo pr_o.cmo -impl file.sml

-- 
Daniel de RAUGLAUDRE
http://cristal.inria.fr/~ddr/
-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

Re: [Caml-list] record declaration, SML

[Alessandro Baretta]

Daniel de Rauglaudre wrote:
> Hi,
> 
> 
>>What's best translation of the following SML type?
>>datatype t = C of { f : int }
> 
> 
> A solution is:
>   type t = C of < f : int >

Interesting solution. However, Ocaml does not support 
defining objects on the fly. The following
 > let c = new object method f = 1 end in C c
is not valid Ocaml.

So, in order to define just one value one needs to define an 
entire class as in the following example.

class f_one = object method f = 1 end
let res = C (new f_one)

BTW, why is something like the above not in the language? It 
seems only natural to have such a feature since an analogous 
construct exists for structs.

let module C = struct let f = 1 end in C.f

Alex



-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

Re: [Caml-list] record declaration, SML

[Daniel de Rauglaudre]

Hi,

On Fri, Jan 10, 2003 at 12:14:17AM +0100, Alessandro Baretta wrote:
> 
> Interesting solution. However, Ocaml does not support 
> defining objects on the fly. The following
>  > let c = new object method f = 1 end in C c
> is not valid Ocaml.

You can do this:

  let module M =
    struct
      class a = object val f = 1 method f = f end
     end
  in
  C (new M.a)

It is a little bit tricky, but useful in an automatic translation
from SML to OCaml.

-- 
Daniel de RAUGLAUDRE
http://cristal.inria.fr/~ddr/
-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

Re: [Caml-list] record declaration, SML

[brogoff]

On Fri, 10 Jan 2003, Alessandro Baretta wrote:
> Daniel de Rauglaudre wrote:
> > Hi,
> > 
> > 
> >>What's best translation of the following SML type?
> >>datatype t = C of { f : int }
> > 
> > 
> > A solution is:
> >   type t = C of < f : int >
> 
> Interesting solution. However, Ocaml does not support 
> defining objects on the fly. The following
>  > let c = new object method f = 1 end in C c
> is not valid Ocaml.
> 
> So, in order to define just one value one needs to define an 
> entire class as in the following example.
> 
> class f_one = object method f = 1 end
> let res = C (new f_one)

Is this such a big problem? I imagine we're more likely to provide a 
smart constructor for t, something like 

    let mkC o = C(o :> < f : int >)

and just stuff any object with method f : int in there. If we only want there 
to be one kind of thing, we're better off using a record. Of course, Daniel 
doesn't have that option since he is discussing a syntactic transformation 
from SML records into OCaml constructs, and hence using the class system makes 
sense on account of the differences between SML record typing and that of 
OCaml. 

> BTW, why is something like the above not in the language? It 
> seems only natural to have such a feature since an analogous 
> construct exists for structs.

There are no anonymous records (like SML has) either, so if you see classes as 
analogous to records rather than structs it doesn't seem so unnatural. No, I 
realize that's not a real explanation. 

-- Brian


-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

Re: [Caml-list] record declaration, SML

[Alessandro Baretta]

brogoff@speakeasy.net wrote:
> On Fri, 10 Jan 2003, Alessandro Baretta wrote:
>>So, in order to define just one value one needs to define an 
>>entire class as in the following example.
>>
>>class f_one = object method f = 1 end
>>let res = C (new f_one)
> 
> 
> Is this such a big problem? I imagine we're more likely to provide a 
> smart constructor for t, something like 
> 
>     let mkC o = C(o :> < f : int >)

No, it's not a bit problem. I was just wondering

> and just stuff any object with method f : int in there. If we only want there 
> to be one kind of thing, we're better off using a record. Of course, Daniel 
> doesn't have that option since he is discussing a syntactic transformation 
> from SML records into OCaml constructs, and hence using the class system makes 
> sense on account of the differences between SML record typing and that of 
> OCaml. 

Besides, Daniel showed us that you can actually define a 
record on the fly without cluttering the class namespace.

   let module M =
     struct
       class a = object val f = 1 method f = f end
      end
   in
   C (new M.a)

This raises one more question:
let module P =
	struct type t = { f : int } let x = { f = 1 } end
in P.x;;

The above is not valid Ocaml because P.x has a type which 
escapes its scope. This is quite evident. Now I would expect 
Daniel's code to be rejected by the same standards, but it 
is not. Why? Does class a not escape its scope?

> 
>>BTW, why is something like the above not in the language? It 
>>seems only natural to have such a feature since an analogous 
>>construct exists for structs.
> 
> 
> There are no anonymous records (like SML has) either, so if you see classes as 
> analogous to records rather than structs it doesn't seem so unnatural. No, I 
> realize that's not a real explanation. 

Right. Even unnamed records are missing. But it's different 
because the Ocaml type system does not support record 
subtyping--even if record type b in module B defines all 
fields of record type a in module A with the same types, B 
is not a subtype of A. On the other hand, object subtyping 
exists already in the language, so on-the-fly definitions of 
objects à la façon de Daniel are more meaningful and useful.

Alex

-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners