caml-list - the Caml user's mailing list
 help / color / mirror / Atom feed
From: Tom <tom.primozic@gmail.com>
To: caml-list <caml-list@inria.fr>
Subject: Type system infering 'a and '_a
Date: Wed, 13 Sep 2006 18:10:32 +0200	[thread overview]
Message-ID: <c1490a380609130910s37d76ed6k4c73f2c6df9e1cee@mail.gmail.com> (raw)

[-- Attachment #1: Type: text/plain, Size: 1594 bytes --]

I was playing a bit with Ocaml type system the other day...

  type 'a t1 = Empty1 | Full1 of 'a

  # let a1 = Empty1;;
  val a1 : 'a t1 = Empty1

Smart.

  type 'a t2 = Empty2 of 'a option ref | Full2 of 'a

  # let a2 = Empty2 (ref None);;
  val a2 : '_a t2 = Empty2 {contents = None}


The type system won't infer a2 as being polymorphic, as it is mutable.
Smart.

  type 'a t3 = Empty3 of string | Full3 of 'a

  # let a3 = Empty3 "foo";;
  val a3 : 'a t3 = Empty3 "foo"

Even thou a3 is mutable, mutating it can't bind 'a, so a3 is polymorphic.
Smart.

  type ('a, 'b) t4 = Empty4 of 'a option ref | Full4 of 'a * 'b

  # let a4 = Empty4 (ref None);;
  val a4 : ('_a, 'b) t4 = Empty4 {contents = None}

As only 'a is "contained" in a mutable type, a4 is polymorphic only in 'b.
Smart

But now...

  type ('a, 'b) t5 =
      Empty_a5 of 'a option ref
    | Empty_b5 of 'b option ref
    | Full5 of 'a * 'b

  # let a5 = Empty_a5 (ref None);;
  val a5 : ('_a, '_b) t5 = Empty_a5 {contents = None}

Stupid. Value a5 should be polymorphic in 'b!

(And, with a bit of hacking:

  type 'a t1 = Empty1 | Full1 of 'a

  # let b1 = Full1 (Obj.magic 0);;
  val b1 : 'a t1 = Full1 <poly>

Smart.

  type ('a, 'b) t4 = Empty4 of 'a option ref | Full4 of 'a * 'b

  # let b4 = Full4 (Obj.magic 0, Obj.magic 0);;
  val b4 : ('_a, 'b) t4 = Full4 (<poly>, <poly>)

Stupid.

)

So we see that "mutability" of each type variable is defined for the whole
type, not for the individual value constructors only. Is there a theoretical
reason (of course, one reason is probably an easier implementation)?

- Tom

[-- Attachment #2: Type: text/html, Size: 2053 bytes --]

             reply	other threads:[~2006-09-13 16:10 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-09-13 16:10 Tom [this message]
2006-09-14  0:24 ` [Caml-list] " Jacques Garrigue
2006-09-14  2:13   ` skaller

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=c1490a380609130910s37d76ed6k4c73f2c6df9e1cee@mail.gmail.com \
    --to=tom.primozic@gmail.com \
    --cc=caml-list@inria.fr \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).