This is (a) one of the big advantages of records and (b) why it's a mistake to have the record discussion without taking into account some notion of lens. Haskell field names are a red herring because they're selector functions - what we need is a record system which generates lenses and solves the namespacing issues.

Jules

On 19 Sep 2011, at 19:58, Edward Kmett wrote:

If you use lenses you can do this today with no real need to adulterate the parser.

setL (l2 . l1) x rec

This goes one step further as it can be written point free so you don't even have to give rec a name if you don't want to. ;)

-Edward

On Mon, Sep 19, 2011 at 9:09 AM, Barney Hilken <b.hilken@ntlworld.com> wrote:
All this talk about records got me thinking. I don't really like the current syntax for record update (because it looks too much like function application) but here is an extension to it which might be useful.

If you use nested records, many languages allow you to update the inner records directly, for example rect.bottomLeft.xCoord += 3. We could allow a similar thing in Haskell by extending the syntax like this:

fbind   ->      qvar = exp
               |       qvar { fbind1 , ... , fbindn }

then

       rec { l1 { l2 = x }}

would mean

       rec { l1 = (l1 rec) { l2 = x }}

The advantage (apart from convenience) is that we don't have to repeat rec, which could be a complex expression.

Is this any use?

Barney.


_______________________________________________
Glasgow-haskell-users mailing list
Glasgow-haskell-users@haskell.org
http://www.haskell.org/mailman/listinfo/glasgow-haskell-users

_______________________________________________
Glasgow-haskell-users mailing list
Glasgow-haskell-users@haskell.org
http://www.haskell.org/mailman/listinfo/glasgow-haskell-users