You can, you have to write it slightly differently. But defined that way it can only ever have one element because NList only takes one.

mylist = NList (NList (NList (Elem 1)))

You could also change it to

data NestedList a = Elem a | NList a (NestedList a)

But at that point you have a nonempty list type (always at least one element) which already exists.

On Tue, Jan 26, 2021, 17:40 Lawrence Bottorff <borgauf@gmail.com> wrote:
Why can I not do this?

data NestedList1 a = Elem a | NList (NestedList1 a)

that is, with parens rather than square brackets, then

myList2 = (NList (Elem 1, NList (Elem 2, NList (Elem 3, Elem 4), Elem 5)))

It gives the error 

<interactive>:383:18-73: error:
    ,* Couldn't match expected type `NestedList1 a'
                  with actual type `(NestedList1 Integer, NestedList1 a0)'
    ,* In the first argument of `NList', namely
        `(Elem 1, NList (Elem 2, NList (Elem 3, Elem 4), Elem 5))'
      In the expression:
        (NList (Elem 1, NList (Elem 2, NList (Elem 3, Elem 4), Elem 5)))
      In an equation for `myList2':
          myList2
            = (NList (Elem 1, NList (Elem 2, NList (Elem 3, Elem 4), Elem 5)))
    ,* Relevant bindings include
        myList2 :: NestedList1 a (bound at <interactive>:383:1)

etc., etc.

On Tue, Jan 26, 2021 at 4:05 PM David McBride <toad3k@gmail.com> wrote:
Right that is a plain list of NestedLists.  So if you were to rewrite [a] as (Regularlist a) so to speak (not a real type), the definition of NestedList would be List (RegularList (NestedList a)).

Keep in mind that List is a constructor, everything after it is types.

On Tue, Jan 26, 2021 at 4:50 PM Lawrence Bottorff <borgauf@gmail.com> wrote:
So NestedList is using regular List? So in

data NestedList a = Elem a | List [NestedList a]

the second data constructor List [NestedList a] we see a "regular" list because of the square brackets?

On Tue, Jan 26, 2021 at 3:42 PM David McBride <toad3k@gmail.com> wrote:
In NestedList, the List constructor takes a regular list of NestedLists.  Therefore when pattern matching on it you can get access to those nested lists.  In your code, x is the first NestedList, and xs is the rest of the NestedLists.

On Tue, Jan 26, 2021 at 4:32 PM Lawrence Bottorff <borgauf@gmail.com> wrote:
I'm following this and yet I see this solution

data NestedList a = Elem a | List [NestedList a] deriving (Show)

flatten1 :: NestedList a -> [a]
flatten1 (Elem a   )   = [a]
flatten1 (List (x:xs)) = flatten1 x ++ flatten1 (List xs)
flatten1 (List [])     = []

What I find puzzling is this line

flatten1 (List (x:xs)) = flatten1 x ++ flatten1 (List xs)

where I see 

(List (x:xs)) as an argument. How is the NestedList type also able to be expressed as a normal consed list with x:xs argument? How is (:) interacting with NestedList?

LB
_______________________________________________
Beginners mailing list
Beginners@haskell.org
http://mail.haskell.org/cgi-bin/mailman/listinfo/beginners
_______________________________________________
Beginners mailing list
Beginners@haskell.org
http://mail.haskell.org/cgi-bin/mailman/listinfo/beginners
_______________________________________________
Beginners mailing list
Beginners@haskell.org
http://mail.haskell.org/cgi-bin/mailman/listinfo/beginners
_______________________________________________
Beginners mailing list
Beginners@haskell.org
http://mail.haskell.org/cgi-bin/mailman/listinfo/beginners
_______________________________________________
Beginners mailing list
Beginners@haskell.org
http://mail.haskell.org/cgi-bin/mailman/listinfo/beginners