I'm trying to translate the Haskell core library's Arrows into F# (I think it's a good exercise to understanding Arrows and F# better, and I might be able to use them in a project I'm working on.) However, a direct translation isn't possible due to the difference in paradigms. Haskell uses type-classes to express this stuff, but I'm not sure what F# constructs best map the functionality of type-classes with the idioms of F#. I have a few thoughts, but figured it best to bring it up here and see what was considered to be the closest in functionality.
For the tl;dr crowd: How do I translate type-classes (a Haskell idiom) into F# idiomatic code?
For those accepting of my long explanation:
This code from the Haskell standard lib is an example of what I'm trying to translate:
class Category cat where
id :: cat a a
comp :: cat a b -> cat b c -> cat a c
class Category a => Arrow a where
arr :: (b -> c) -> a b c
first :: a b c -> a (b,d) (c,d)
instance Category (->) where
id f = f
instance Arrow (->) where
arr f = f
first f = f *** id
Attempt 1: Modules, Simple Types, Let Bindings
My first shot at this was to simply map things over directly using Modules for organization, like:
type Arrow<'a,'b> = Arrow of ('a -> 'b)
let arr f = Arrow f
let first f = //some code that does the first op
That works, but it loses out on polymorphism, since I don't implement Categories and can't easily implement more specialized Arrows.
Attempt 1a: Refining using Signatures and types
One way to correct some issues with Attempt 1 is to use a .fsi file to define the methods (so the types enforce easier) and to use some simple type tweaks to specialize.
type ListArrow<'a,'b> = Arrow<['a],['b]>
//or
type ListArrow<'a,'b> = LA of Arrow<['a],['b]>
But the fsi file can't be reused (to enforce the types of the let bound functions) for other implementations, and the type renaming/encapsulating stuff is tricky.
Attempt 2: Object models and interfaces
Rationalizing that F# is built to be OO also, maybe a type hierarchy is the right way to do this.
type IArrow<'a,'b> =
abstract member comp : IArrow<'b,'c> -> IArrow<'a,'c>
type Arrow<'a,'b>(func:'a->'b) =
interface IArrow<'a,'b> with
member this.comp = //fun code involving "Arrow (fun x-> workOn x) :> IArrow"
Aside from how much of a pain it can be to get what should be static methods (like comp and other operators) to act like instance methods, there's also the need to explicitly upcast the results. I'm also not sure that this methodology is still capturing the full expressiveness of type-class polymorphism. It also makes it hard to use things that MUST be static methods.
Attempt 2a: Refining using type extensions
So one more potential refinement is to declare the interfaces as bare as possible, then use extension methods to add functionality to all implementing types.
type IArrow<'a,'b> with
static member (&&&) f = //code to do the fanout operation
Ah, but this locks me into using one method for all types of IArrow. If I wanted a slightly different (&&&) for ListArrows, what can I do? I haven't tried this method yet, but I would guess I can shadow the (&&&), or at least provide a more specialized version, but I feel like I can't enforce the use of the correct variant.
Help me
So what am I supposed to do here? I feel like OO should be powerful enough to replace type-classes, but I can't seem to figure out how to make that happen in F#. Were any of my attempts close? Are any of them "as good as it gets" and that'll have to be good enough?
My brief answer is:
OO is not powerful enough to replace type classes.
The most straightforward translation is to pass a dictionary of operations, as in one typical typeclass implementation. That is if typeclass Foo defines three methods, then define a class/record type named Foo, and then change functions of
Foo a => yadda -> yadda -> yadda
to functions like
Foo -> yadda -> yadda -> yadda
and at each call site you know the concrete 'instance' to pass based on the type at the call-site.
Here's a short example of what I mean:
// typeclass
type Showable<'a> = { show : 'a -> unit; showPretty : 'a -> unit } //'
// instances
let IntShowable =
{ show = printfn "%d"; showPretty = (fun i -> printfn "pretty %d" i) }
let StringShowable =
{ show = printfn "%s"; showPretty = (fun s -> printfn "<<%s>>" s) }
// function using typeclass constraint
// Showable a => [a] -> ()
let ShowAllPretty (s:Showable<'a>) l = //'
l |> List.iter s.showPretty
// callsites
ShowAllPretty IntShowable [1;2;3]
ShowAllPretty StringShowable ["foo";"bar"]
See also
https://web.archive.org/web/20081017141728/http://blog.matthewdoig.com/?p=112
Here's the approach I use to simulate Typeclasses (from http://code.google.com/p/fsharp-typeclasses/ ).
In your case, for Arrows could be something like this:
let inline i2 (a:^a,b:^b ) =
((^a or ^b ) : (static member instance: ^a* ^b -> _) (a,b ))
let inline i3 (a:^a,b:^b,c:^c) =
((^a or ^b or ^c) : (static member instance: ^a* ^b* ^c -> _) (a,b,c))
type T = T with
static member inline instance (a:'a ) =
fun x -> i2(a , Unchecked.defaultof<'r>) x :'r
static member inline instance (a:'a, b:'b) =
fun x -> i3(a, b, Unchecked.defaultof<'r>) x :'r
type Return = Return with
static member instance (_Monad:Return, _:option<'a>) = fun x -> Some x
static member instance (_Monad:Return, _:list<'a> ) = fun x -> [x]
static member instance (_Monad:Return, _: 'r -> 'a ) = fun x _ -> x
let inline return' x = T.instance Return x
type Bind = Bind with
static member instance (_Monad:Bind, x:option<_>, _:option<'b>) = fun f ->
Option.bind f x
static member instance (_Monad:Bind, x:list<_> , _:list<'b> ) = fun f ->
List.collect f x
static member instance (_Monad:Bind, f:'r->'a, _:'r->'b) = fun k r -> k (f r) r
let inline (>>=) x (f:_->'R) : 'R = T.instance (Bind, x) f
let inline (>=>) f g x = f x >>= g
type Kleisli<'a, 'm> = Kleisli of ('a -> 'm)
let runKleisli (Kleisli f) = f
type Id = Id with
static member instance (_Category:Id, _: 'r -> 'r ) = fun () -> id
static member inline instance (_Category:Id, _:Kleisli<'a,'b>) = fun () ->
Kleisli return'
let inline id'() = T.instance Id ()
type Comp = Comp with
static member instance (_Category:Comp, f, _) = (<<) f
static member inline instance (_Category:Comp, Kleisli f, _) =
fun (Kleisli g) -> Kleisli (g >=> f)
let inline (<<<) f g = T.instance (Comp, f) g
let inline (>>>) g f = T.instance (Comp, f) g
type Arr = Arr with
static member instance (_Arrow:Arr, _: _ -> _) = fun (f:_->_) -> f
static member inline instance (_Arrow:Arr, _:Kleisli<_,_>) =
fun f -> Kleisli (return' <<< f)
let inline arr f = T.instance Arr f
type First = First with
static member instance (_Arrow:First, f, _: 'a -> 'b) =
fun () (x,y) -> (f x, y)
static member inline instance (_Arrow:First, Kleisli f, _:Kleisli<_,_>) =
fun () -> Kleisli (fun (b,d) -> f b >>= fun c -> return' (c,d))
let inline first f = T.instance (First, f) ()
let inline second f = let swap (x,y) = (y,x) in arr swap >>> first f >>> arr swap
let inline ( *** ) f g = first f >>> second g
let inline ( &&& ) f g = arr (fun b -> (b,b)) >>> f *** g
Usage:
> let f = Kleisli (fun y -> [y;y*2;y*3]) <<< Kleisli ( fun x -> [ x + 3 ; x * 2 ] ) ;;
val f : Kleisli<int,int list> = Kleisli <fun:f#4-14>
> runKleisli f <| 5 ;;
val it : int list = [8; 16; 24; 10; 20; 30]
> (arr (fun y -> [y;y*2;y*3])) 3 ;;
val it : int list = [3; 6; 9]
> let (x:option<_>) = runKleisli (arr (fun y -> [y;y*2;y*3])) 2 ;;
val x : int list option = Some [2; 4; 6]
> ( (*) 100) *** ((+) 9) <| (5,10) ;;
val it : int * int = (500, 19)
> ( (*) 100) &&& ((+) 9) <| 5 ;;
val it : int * int = (500, 14)
> let x:List<_> = (runKleisli (id'())) 5 ;;
val x : List<int> = [5]
Note: use id'() instead of id
Update: you need F# 3.0 to compile this code, otherwise here's the F# 2.0 version.
And here's a detailed explanation of this technique which is type-safe, extensible and as you can see works even with some Higher Kind Typeclasses.
Related
I was reading Idris tutorial. And I can't understand the following code.
disjoint : (n : Nat) -> Z = S n -> Void
disjoint n p = replace {P = disjointTy} p ()
where
disjointTy : Nat -> Type
disjointTy Z = ()
disjointTy (S k) = Void
So far, what I figure out is ...
Void is the empty type which is used to prove something is impossible.
replace : (x = y) -> P x -> P y
replace uses an equality proof to transform a predicate.
My questions are:
which one is an equality proof? (Z = S n)?
which one is a predicate? the disjointTy function?
What's the purpose of disjointTy? Does disjointTy Z = () means Z is in one Type "land" () and (S k) is in another land Void?
In what way can an Void output represent contradiction?
Ps. What I know about proving is "all things are no matched then it is false." or "find one thing that is contradictory"...
which one is an equality proof? (Z = S n)?
The p parameter is the equality proof here. p has type Z = S n.
which one is a predicate? the disjointTy function?
Yes, you are right.
What's the purpose of disjointTy?
Let me repeat the definition of disjointTy here:
disjointTy : Nat -> Type
disjointTy Z = ()
disjointTy (S k) = Void
The purpose of disjointTy is to be that predicate replace function needs. This consideration determines the type of disjointTy, viz. [domain] -> Type. Since we have equality between naturals numbers, [domain] is Nat.
To understand how the body has been constructed we need to take a look at replace one more time:
replace : (x = y) -> P x -> P y
Recall that we have p of Z = S n, so x from the above type is Z and y is S n. To call replace we need to construct a term of type P x, i.e. P Z in our case. This means the type P Z returns must be easily constructible, e.g. the unit type is the ideal candidate for this. We have justified disjointTy Z = () clause of the definition of disjointTy. Of course it's not the only option, we could have used any other inhabited (non-empty) type, like Bool or Nat, etc.
The return value in the second clause of disjointTy is obvious now -- we want replace to return a value of Void type, so P (S n) has to be Void.
Next, we use disjointTy like so:
replace {P = disjointTy} p ()
^ ^ ^
| | |
| | the value of `()` type
| |
| proof term of Z = S n
|
we are saying "this is the predicate"
As a bonus, here is an alternative proof:
disjoint : (n : Nat) -> Z = S n -> Void
disjoint n p = replace {P = disjointTy} p False
where
disjointTy : Nat -> Type
disjointTy Z = Bool
disjointTy (S k) = Void
I have used False, but could have used True -- it doesn't matter. What matters is our ability to construct a term of type disjointTy Z.
In what way can an Void output represent contradiction?
Void is defined like so:
data Void : Type where
It has no constructors! There is no way to create a term of this type whatsoever (under some conditions: like Idris' implementation is correct and the underlying logic of Idris is sane, etc.). So if some function claims it can return a term of type Void there must be something fishy going on. Our function says: if you give me a proof of Z = S n, I will return a term of the empty type. This means Z = S n cannot be constructed in the first place because it leads to a contradiction.
Yes, p : x = y is an equality proof. So p is a equality proof and Z = S k is a equality type.
Also yes, usually any P : a -> Type is called predicate, like IsSucc : Nat -> Type. In boolean logic, a predicate would map Nat to true or false. Here, a predicate holds, if we can construct a proof for it. And it is true, if we can construct it (prf : ItIsSucc 4). And it is false, if we cannot construct it (there is no member of ItIsSucc Z).
At the end, we want Void. Read the replace call as Z = S k -> disjointTy Z -> disjointTy (S k), that is Z = S K -> () -> Void. So replace needs two arguments: the proof p : Z = S k and the unit () : (), and voilà, we have a void. By the way, instead of () you could use any type that you can construct, e.g. disjointTy Z = Nat and then use Z instead of ().
In dependent type theory we construct proofs like prf : IsSucc 4. We would say, we have a proof prf that IsSucc 4 is true. prf is also called a witness for IsSucc 4. But with this alone we could only proove things to be true. This is the definiton for Void:
data Void : Type where
There is no constructor. So we cannot construct a witness that Void holds. If you somehow ended up with a prf : Void, something is wrong and you have a contradiction.
I am trying to create module/interface (i dont exactly know how its called, i am new to the language) for basic operations on BST in OCaml. My goal is to have an implementation that lets me doing something like this:
T.create();;
T.push(2);;
T.push(3);;
T.push(5);;
in order to get a bst tree consisting of 2,3,5.
But at the moment to achieve this i have to write something like this:
let teeBst = T.push(2)(T.push(3)(T.push(5)(T.create())));;
So when I am checking/using my code I have to do it like this:
let tee2 = T.push(2)(T.push(3)(T.push(5)(T.create())));;
T.postorder(tee2);;
The output is fine:
# val tee2 : T.bt = <abstr>
# - : int list = [2; 3; 5]
But, as I said before, I would like to achieve this doing as below:
T.push(2);;
T.push(3);;
T.push(5);;
T.postorder();;
(I realise this requires some changes to my postorder function but the one I am currently using is a temporary one so I can check the tree I have atm )
Below is my implementation. If you see the solution, please let me know ;)
module type Tree =
sig
type bt
val create: unit -> bt
val push: int -> bt -> bt
val find: int -> bt -> bool
val preorder: bt -> int list
val postorder: bt -> int list
val inorder: bt -> int list
end;;
module T : Tree =
struct
type bt = E | B of bt * int * bt
let create () = E
let rec push x = function
| E -> B(E, x, E)
| B (l, y, r) when x<y -> B(push x l, y, r)
| B (l, y, r) when x>y -> B(l, y, push x r)
| xs -> xs;;
let rec find x = function
| E -> false
| B(l, y,_) when x< y -> find x l
| B(_,y,r) when x>y -> find x r
| _ -> true;;
let rec preorder = function
| B(l,v,r) -> v::(preorder r) # (preorder l)
| E -> [];;
let rec inorder = function
| B(l,v,r) ->(inorder r) # v::(inorder l)
| E -> []
let rec postorder = function
| B(l,v,r) -> (postorder r) # (postorder l) # [v]
| E -> []
end;;
It seems like you want modules to be classes, but I'd advise you to consider more idiomatic solutions. Have you considered using the pipe operator?
T.create()
|> T.push(2)
|> T.push(3)
|> T.push(5)
|> T.postorder;;
Or with local open (which makes more sense if you have a module with a longer name than just T of course) you can even do
T.(
create()
|> push(2)
|> push(3)
|> push(5)
|> postorder
);
What you're asking for would require introducing global mutable state, which isn't just "some changes" but an entirely different paradigm. And one that is generally frowned upon because it makes your code unpredictable and hard to debug since it relies on state that might change at any moment from anywhere.
Another possibility is to actually use classes, since OCaml has those too. Then you'd still have mutable state, but it would at least be contained.
I'm trying to write a functor that takes a pair of ordered things and produces another ordered thing (with ordering defined lexicographically).
However, I want the resulting "ordered type" to be abstract, rather than an OCaml tuple.
This is easy enough to do with an inline/anonymous signature.
(* orderedPairSetInlineSig.ml *)
module type ORDERED_TYPE = sig
type t
val compare : t -> t -> int
end
module MakeOrderedPairSet (X : ORDERED_TYPE) :
sig
type t
val get_fst : t -> X.t
val get_snd : t -> X.t
val make : X.t -> X.t -> t
val compare : t -> t -> int
end = struct
type t = X.t * X.t
let combine_comparisons fst snd =
if fst = 0 then snd else fst
let compare (x, y) (a, b) =
let cmp = X.compare x a in
let cmp' = X.compare y b in
combine_comparisons cmp cmp'
let get_fst ((x, y) : t) = x
let get_snd ((x, y) : t) = y
let make x y = (x, y)
end
I want to give my anonymous signature a name like ORDERED_PAIR_SET_TYPE and move it outside the definition of MakeOrderedPairSet, like so (warning: not syntactically valid) :
(* orderedPairSet.ml *)
module type ORDERED_TYPE = sig
type t
val compare : t -> t -> int
end
module type ORDERED_PAIR_SET_TYPE = sig
type t
type el
val get_fst : t -> el
val get_snd : t -> el
val make : el -> el -> t
val compare : t -> t -> int
end
module MakeOrderedPairSet (X : ORDERED_TYPE) :
(ORDERED_PAIR_SET_TYPE with type el = X.t) = struct
type t = X.t * X.t
let combine_comparisons fst snd =
if fst = 0 then snd else fst
let compare (x, y) (a, b) =
let cmp = X.compare x a in
let cmp' = X.compare y b in
combine_comparisons cmp cmp'
let get_fst ((x, y) : t) = x
let get_snd ((x, y) : t) = y
let make x y = (x, y)
end
with el being an abstract type in the signature that I'm trying to bind to X.t inside the body of MakeOrderedPairSet.
However, I can't figure out how to fit everything together.
(ORDERED_PAIR_SET_TYPE with type el = X.t) is the most obvious way I can think of to say "give me a signature that's just like this one, but with an abstract type replaced with a concrete one (or differently-abstract in this case)". However, it isn't syntactically valid in this case (because of the parentheses). Taking the parentheses off does not result in a valid "module-language-level expression" either; I left it on because I think it makes my intent more obvious.
So ... how do you use a named signature to restrict the visibility into a [module produced by a functor]/[parameterized module]?
If you don't want to add el to the exports of the module then there are two ways:
Use a substitution constraint:
ORDERED_PAIR_SET_TYPE with type el := X.t
That will remove the specification of el from the signature.
Use a parameterised signature. Unfortunately, that is not expressible directly in OCaml, but requires a bit of extra functor gymnastics around the definition of your signature:
module SET_TYPE (X : ORDERED_TYPE) =
struct
module type S =
sig
type t
val get_fst : t -> X.el
val get_snd : t -> X.el
val make : X.el -> X.el -> t
val compare : t -> t -> int
end
end
With that you can write:
module MakeOrderedPairSet (X : ORDERED_TYPE) : SET_TYPE(X).S = ...
I need a function that produces primes in F#. I found this:
let primesSeq =
let rec nextPrime n p primes =
if primes |> Map.containsKey n then
nextPrime (n + p) p primes
else
primes.Add(n, p)
let rec prime n primes =
seq {
if primes |> Map.containsKey n then
let p = primes.Item n
yield! prime (n + 1) (nextPrime (n + p) p (primes.Remove n))
else
yield n
yield! prime (n + 1) (primes.Add(n * n, n))
}
prime 2 Map.empty
This works very well, but sometimes I need to work with int64/BigInts as well. Is there a more clever way of reusing this code than providing another sequences like these:
let primesSeq64 = Seq.map int64 primesSeq
let primesBigInts = Seq.map (fun (x : int) -> BigInteger(x)) primesSeq
I've heard about modifying a code using "inline" and "LanguagePrimitives", but all I've found was connected with function while my problem is related to a value.
Moreover - I'd like to have a function that works with integer types and computes a floor of a square root.
let inline sqRoot arg = double >> Math.Sqrt >> ... ?
but I can't see a way of returning the same type as "arg" is, as Math.Sqrt returns a double. Again - is there anything better than reimplementing the logic that computes a square root by myself ?
So the general way to do this requires a function and languageprimitives - in your case everywhere you have 1 you write LanguagePrimitives.GenericOne which will produce 1 or 1.0 etc depending on what is required.
To get this to work, you need to create a function value - you can avoid this by doing something like:
let inline primesSeq() = ...
let primesintSeq = primesSeq() //if you use this as an int seq later the compiler will figure it out, otherwise you use
let specified : int seq = primesSeq()
I am not so sure about the sqrt case though - it probably depends on how hacky you are willing to make the solution.
A naïve implementation of generic sqRoot may go along these lines:
let sqRoot arg =
let inline sqrtd a = (double >> sqrt) a
let result = match box(arg) with
| :? int64 as i -> (sqrtd i) |> int64 |> box
| :? int as i -> (sqrtd i) |> int |> box
// cases for other relevant integral types
| _ -> failwith "Unsupported type"
unbox result
and then, checking in FSI:
> let result: int = sqRoot 4;;
val result : int = 2
> let result: int64 = sqRoot 9L;;
val result : int64 = 3L
I find the following C# extension method very useful:
public static bool In<T>(this T x, params T[] xs)
{
return xs.Contains(x);
}
allowing for C# calls such as
var s = "something else";
var rslt = s.In("this","that","other") ? "Yay" : "Boo";
and
var i = 1;
var rslt = i.In(1,2,3) ? "Yay" : "Boo";
I have been trying to come up with an F# (near-)equivalent, allowing e.g.:
let s = "something else"
let rslt = if s.In("this","that","other") then "Yay" else "Boo"
It seems like I would need something like:
type 'T with
static member this.In([ParamArray] xs : 'T )
{
return xs.Contains(x);
}
but that is not legal F# syntax. I can't see how to declare a extension method on a generic class in F#. Is it possible? Or is there a better way to achieve similar results? (I imagine I could just link in the C# project and call it from F#, but that would be cheating! :-)
The best I could come up with was:
let inline In (x : 'a, [<ParamArray>] xs : 'a[]) = Array.Exists( xs, (fun y -> x = y) )
which I expected to allow for calls like (which are not really acceptable anyway imho):
if In(ch, '?', '/') then "Yay" else "Boo"
but in fact required:
if In(ch, [| '?'; '/' |]) then "Yay" else "Boo"
implying that the ParamArray attribute is being ignored (for reasons I've yet to fathom).
Fwiw, the latest version of F# (3.1) contains exactly what I was after (yay!):
[<Extension>]
type ExtraCSharpStyleExtensionMethodsInFSharp () =
[<Extension>]
static member inline In(x: 'T, xs: seq<'T>) = xs |> Seq.exists (fun o -> o = x)
[<Extension>]
static member inline Contains(xs: seq<'T>, x: 'T) = xs |> Seq.exists (fun o -> o = x)
[<Extension>]
static member inline NotIn(x: 'T, xs: seq<'T>) = xs |> Seq.forall (fun o -> o <> x)
providing usages as
if s.In(["this","that","other"]) then ....
if (["this","that","other"]).Contains(s) then ...
etc.