caml-list - the Caml user's mailing list
 help / color / mirror / Atom feed
From: Andreas Rossberg <rossberg@ps.uni-sb.de>
To: caml-list@inria.fr
Subject: Re: Compiler translation of array indexing
Date: Mon, 24 Jan 2000 18:06:10 +0100	[thread overview]
Message-ID: <388C8682.6E3A8E4@ps.uni-sb.de> (raw)
In-Reply-To: <200001192250.XAA25766@pauillac.inria.fr>

Pierre Weis wrote:
> 
> To summerize, I really would like to write
>  v.(1) instead of Array.get v 1
>  s.(1) instead of String.get s 1
> 
> I would consider as an additional benefit of the new type system, if I
> could use:
>  t.("ok") instead of Hashtbl.find t "ok",
>  l.("ok") instead of List.assoc "ok" l.

Recent work by Mark Jones tries to make this work in Haskell/Hugs by
extending multi parameter type classes with dependency annotations:

	http://www.cse.ogi.edu/~mpj/fds.html

Using dependency annotations, one can declare a class like (in some
mixed Haskell/Caml pseudo syntax):

        class Index 'c 'i 'e | 'c -> 'i 'e where
                .() : 'c -> 'i -> 'e

(note that all parameters have kind *) and instances

        instance Index ('a array) int 'a
        instance Index string int char
        instance Index (('a,'b) Hashtbl.t) 'a 'b
        instance Index (('a*'b) list) 'a 'b

These should provide the desired type-driven behaviour for indexing. The
dependency annotation ('c -> 'i 'e) in the class ensures that index and
element type are determined by the collection type. For example, the
compiler could infer the following:

	[|1;2;3|].(0)         : int
	"hello".(3)           : char
	[(2,39);(4,0)].(2)    : int
	fun i -> "hello".(i)  : int -> char

Note however, that overloading can quickly become ambiguous, even with
dependencies. While

	let f x = x.(0)

would be possible and have qualified type like

	(Index 'c int 'e) => 'c -> 'e,

the only slightly more involved

	let g x = x.(0).(0)

would not be allowed without additional type annotations since

	(Index 'c1 int 'c2, Index 'c2 int 'e) => 'c1 -> 'e

is an ambiguous type.

Best regards,

        - Andreas



  reply	other threads:[~2000-01-25  8:39 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2000-01-19 17:33 Simon Peyton-Jones
2000-01-19 22:50 ` Pierre Weis
2000-01-24 17:06   ` Andreas Rossberg [this message]
  -- strict thread matches above, loose matches on Subject: below --
2000-01-18 16:34 David McClain
2000-01-19 14:25 ` Pierre Weis
2000-01-19 17:08   ` Markus Mottl
2000-01-19 21:17     ` Pierre Weis

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=388C8682.6E3A8E4@ps.uni-sb.de \
    --to=rossberg@ps.uni-sb.de \
    --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).