[swift-evolution] [Review] SE-0161: Smart KeyPaths: Better Key-Value Coding for Swift

[Karl Wagner]

> On 2 Apr 2017, at 02:56, Karl Wagner <karl.swift at springsup.com> wrote:
> 
> struct UnaryPredicate<Parameter> {
>     let evaluate: (Parameter) -> Bool
> }
> struct BinaryPredicate<Left, Right> {
>     let evaluate: (Left, Right) -> Bool
> }
> 
> extension KeyPath where Value: Equatable {
>     func equals(_ value: Value) -> UnaryPredicate<Root> {
>         return UnaryPredicate { $0[keyPath: self] == value }
>     }
>     func equals<KP: KeyPath>(_ other: KP) -> BinaryPredicate<Root, KP.Root> where KP.Value == Value {
>         return BinaryPredicate { $0[keyPath: self] == $1[keyPath: other] }
>     }
> }
> 
> let isDog = #keypath(Pet, .type).equals(.dog) // UnaryPredicate<Pet>
> if isDog.evaluate(somePet) {
>     print(“It’s a dog”)
> }
> 
> let areSameLength = #keypath(Array<Int>, .count).equals(#keypath(Array<String>, .count))
> // BinaryPredicate<Array<Int>, Array<String>>
> if areSameLength.evaluate([1,2,3], [“a”, “b”, “c”]) {
>     print(“same lengths”)
> }


Apologies, been working with lots of generic closures today and the ‘Result’ just came out instinctively. Of course the Result of a predicate is a Boolean…

Still, it’s useful to think about what some more advanced use-cases of keypaths might look like with this proposal. As far as syntax is concerned, operators may ultimately be the best way to make long predicate combinations cleaner.

extension UnaryPredicate {
    static func &&<P>(left: UnaryPredicate<P>, right: UnaryPredicate<P>) -> UnaryPredicate<P> {
        return UnaryPredicate { self($0) && other($0) }
    }
}

let isHealthyPuppy = #keypath(Pet, .type).equals(.dog) && #keypath(Pet, .age).isLessThan(12) && #keypath(Pet, .health).equals(.good)

vs

let isHealthyPuppy = #keypath(Pet, .type).equals(.dog).and(#keypath(Pet, .age).isLessThan(12)).and(#keypath(Pet, .health).equals(.good))

- Karl
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.swift.org/pipermail/swift-evolution/attachments/20170402/575ac7ef/attachment.html>