Question

I am reading f# code and I am confused by the syntax. A Parser type is introduced as follows:

type Parser<'r> = Parser of (char list -> ('r*char list) list)

This is evaluated by the interpreter as:

type Parser<'r> = | Parser of (char list -> ('r * char list) list)

which makes sense to me. Then, a new line of code is introduced: ‘A parser function also needs to be applied so we define a partial function for that:’, and the code that follows:

let parse (Parser p) = p

and the interpreter output is:

Parser<'a> -> (char list -> ('a * char list) list)

I am surprised this is even valid syntax. What is it and why is it needed?

Many thanks

Answer 1

In general

let fpattern=body

is equivalent to

let f = function
|pattern ->body

or, even more verbosely,

let f x =
    match x with
    |pattern->body

This allows you to avoid introducing a new identifier which is immediately destructured and then never used again.

In this particular example, that means that parse is equivalent to:

let parse x = 
    match x with
    | Parser p -> p

Since there is only a single case in the Parser type, this destructuring will always succeed.