Presentation is loading. Please wait.

Presentation is loading. Please wait.

Type Variables in ML Until we know the type of a value (perhaps never!), we use a variable for its type Book uses, e.g., t1, tx, tf PL literature uses.

Similar presentations


Presentation on theme: "Type Variables in ML Until we know the type of a value (perhaps never!), we use a variable for its type Book uses, e.g., t1, tx, tf PL literature uses."— Presentation transcript:

1 Type Variables in ML Until we know the type of a value (perhaps never!), we use a variable for its type Book uses, e.g., t1, tx, tf PL literature uses Greek letters:  x.x :  →  (lambda (a b) b) :  →  ML uses 'a, 'b, 'c,... > fun foo(a,b) = b; val foo = fn : 'a * 'b -> 'b

2 Recall: fun reverse(nil) = nil | reverse(x::t) = [x]; ML: Let's Write... sum: int list -> int member: 'a * 'a list -> bool

3 Recall (?) mergesort: Divide-and-conquer for sorting a list Split list into two sorted sublists, then merge them: merge([1,3,5], [2,4,6]) = [1,2,3,4,5,6] So first write merge : ML: Local environments using let

4 > funmerge(nil,M) = M |merge(L,nil) = L |merge(x::xs, y::ys) = if (x:int) < y then x::merge(xs,y::ys) else y::merge(x::xs,ys); val merge = fn : int list * int list -> int list > merge([1,3,5], [2,4,6]); val it = [1,2,3,4,5,6] : int list ML: Local environments using let Now write split :

5 > funsplit(nil) = (nil,nil) | split([a]) = ([a], nil) | split(a::b::cs) = let val (M,N) = split(cs) in (a::M, b::N) end; val split = fn : 'a list -> 'a list * 'a list > split([1,2,3,4,5]); val it = ([1,3,5],[2,4]) : int list * int list Finally write mergeSort :

6 > funmergeSort(nil) = nil | mergeSort([a]) = [a] | mergeSort(L) = let val (M,N) = split(L); val M = mergeSort(M); val N = mergeSort(N) in merge(M,N) end; val mergeSort = fn : int list -> int list > mergeSort([5,4,3,2,1]); val it = [1,2,3,4,5] : int list

7 ML: Polymorphism Note that merge is the only function here explicitly declaring a type ( int ): funmerge(nil,M) = M |merge(L,nil) = L |merge(x::xs, y::ys) = if (x:int) < y then x::merge(xs,y::ys) else y::merge(x::xs,ys); We can “factor out” the int comparison, and make merge and mergeSort polymorphic:

8 ML: Polymorphism > fun merge(nil,M,C) = M (* C is comparator fun *) |merge(L,nil,C) = L |merge(x::xs, y::ys,C) = if C(x,y)then x::merge(xs,y::ys,C) else y::merge(x::xs,ys,C); val merge = fn : 'a list * 'a list * ('a * 'a -> bool) -> 'a list

9 ML: Polymorphism > funmerge Sort(nil,C) = nil | mergeSort([a],C) = [a] | mergeSort(L,C) = let val (M,N) = split(L); val M = mergeSort(M,C); val N = mergeSort(N,C) in merge(M,N,C) end; val mergeSort = fn : 'a list * ('a * 'a -> bool) -> 'a list

10 ML: Polymorphism > fun intcmp(a:int, b) = a < b; val intcmp = fn : int * int -> bool > fun realcmp(a:real, b) = a < b; val realcmp = fn : real * real -> bool > mergeSort([5,4,3,2,1], intcmp); val it = [1,2,3,4,5] : int list > mergeSort([5.5, 4.4, 3.3, 2.2, 1.1], realcmp); val it = [1.1,2.2,3.3,4.4,5.5] : real list

11 ML: Polymorphism Q.: What's the point? A.: ML supports functions as first -class objects (as in Scheme), while also providing strong type- checking (as in Java) Java supports this ability (abstract comparison) through the Comparable interface.


Download ppt "Type Variables in ML Until we know the type of a value (perhaps never!), we use a variable for its type Book uses, e.g., t1, tx, tf PL literature uses."

Similar presentations


Ads by Google