For more about motivation & design of TMap and its relationship to Map, see Denotational design with type class morphisms.  -- Conal

On Wed, Nov 14, 2012 at 1:07 PM, Jake McArthur <jake.mcarthur@gmail.com> wrote:
Now I'm just spamming, but I feel I should elaborate more.

A slight superset of the original semantics of Map can be recovered like this:

    type Map k v = TMap k (Maybe v)

So, nothing is really lost, apart from the specific library I linked not being very rich. Your example having intersection semantics would be written like this:

    (liftA3.liftA3) f amap bmap cmap

In fact, even this restricted form of TMap is still a monad (having the same semantics as ReaderT k Maybe v). Come to think of it, there should be a monad transformer version of TMap.

- Jake


On Wed, Nov 14, 2012 at 3:50 PM, Jake McArthur <jake.mcarthur@gmail.com> wrote:
Sorry for the double send, Henning. I forgot the list.



On Wed, Nov 14, 2012 at 2:46 PM, Henning Thielemann <lemming@henning-thielemann.de> wrote:

An ZipList-like Applicative instance for Data.Map would be nice. I have an application where I like to write
   liftA3 f amap bmap cmap
 meaning
   Map.intersectionWith ($) (Map.intersectionWith f amap bmap) cmap

But I cannot complete the instance implementation because there is no sensible definition for 'pure' for Data.Map. :-(

_______________________________________________
Libraries mailing list
Libraries@haskell.org
http://www.haskell.org/mailman/listinfo/libraries



_______________________________________________
Libraries mailing list
Libraries@haskell.org
http://www.haskell.org/mailman/listinfo/libraries