Description
Suggestion
I would like to propose something along the lines of $Call utility type in Flow so it is possible to represents the result of calling a generic function with a concrete parameters.
Alternatively could be an utility type that hoists parameters from generic function type into type itself e.g.
HoistTypeParams<<I, O>(inn:I[], f:(p:I) => O) => O[]> // => type DerivedType<I, O> = (inn:I[], f:(p:I) => O) => O[]
Or yet another alternative might be to add some way to parametrize through infer
e.g. I would have expected following to work:
type $Call<T extends (args: any[]) => any, Args extends unknown[]> =
T extends (...args:Args) => infer O ? O : never
🔍 Search Terms
List of keywords you searched for before creating this issue. Write them down here so that others can find this suggestion more easily and help provide feedback.
✅ Viability Checklist
My suggestion meets these guidelines:
- This wouldn't be a breaking change in existing TypeScript/JavaScript code
- This wouldn't change the runtime behavior of existing JavaScript code
- This could be implemented without emitting different JS based on the types of the expressions
- This isn't a runtime feature (e.g. library functionality, non-ECMAScript syntax with JavaScript output, new syntax sugar for JS, etc.)
- This feature would agree with the rest of TypeScript's Design Goals.
⭐ Suggestion
📃 Motivating Example
Today it seems impossible capture return type of the generic function, even when concrete type of the input is known. Both built-in ReturnType
and infer
based solutions turn generics into a unknowns.
Here is concrete example:
type Map = <I, O> (list:I[], f:(i:I) => O) => O[]
ReturnType<Map> // unknown[]
Map extends (...args:infer Args) => infer O ? O : never // unknown[]
Call<Map, [number[], (i:number) => string]> // unknown[]
type Call<F, Args extends unknown[]> = F extends (...args:Args) => infer O ? O : never
As per #33185 (comment) it is possible to accomplish this at the implementation level, however I do not believe it is possible to do it at the type level.
💻 Use Cases
Libraries that generate more complex interfaces from a simple ones, something like RPC client for a service with generics becomes impossible here to express:
type Service<T, Path extends PropertyKey[]=[]> = {
[K in keyof T]: T[K] extends (...args:infer I) => infer O ? Method<I, O, [...Path, K]> : Service<T[K], [...Path, K]>
}
interface Method<I extends unknown[], O, Path extends PropertyKey[]> {
<Args extends I> (...args:Args): Promise<O>
meta: { path: Path }
}
declare function service <T>(impl:T):Service<T>
const s = service({
map <I, O>(inn:I[], f:(inn:I) => O):O[] {
return inn.map(f)
}
})
s.map([1, 2, 3], (n:number):string => String(number))
// ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
// Argument of type '(n: number) => string' is not assignable to parameter of type '(inn: unknown) => unknown'.
// Types of parameters 'n' and 'inn' are incompatible.
// Type 'unknown' is not assignable to type 'number'.(2345)