> Am 24.03.2016 um 16:30 schrieb David Waite via swift-evolution > <[email protected]>: > > >>> On Mar 23, 2016, at 3:36 PM, Andrey Tarantsov via swift-evolution >>> <[email protected]> wrote: >>> >>> * Are the current stdlib names for optional map and flatMap misleading? >> >> Not so much misleading, as being unfortunate when optionals and collections >> end up as neighbors in the code. >> >> >>> * Are the current stdlib functions for optional closure application >>> appropriate and sufficient? >>> >>> public func f1<U>(@noescape f: (Wrapped) throws -> U) rethrows -> U? >>> public func f2<U>(@noescape f: (Wrapped) throws -> U!) rethrows -> U! >>> public func f3<U>(@noescape f: (Wrapped) throws -> U) rethrows -> Void >> >> I don't see why flatMap needs to be separate from map. It just does not make >> sense. I would only have >> >> public func f2<U>(@noescape f: (Wrapped) throws -> U?) rethrows -> U? >> >> and would call it "ifPresent", "then", "unwrap", "transform" or something >> like that. > > Optionals of Optionals are valid, and it would be ambiguous which would be > returned from a map overloaded to allow both optional and non-optional > transform functions. So, map always assumes the transform returns a > non-optional, and if it returns an optional value (and that was not what you > wanted) you have to flatten. Hence flatMap to combine the two steps.
Exactly. And opposed to those the suggested ifPresent would have result type Void which is the whole point of it, i.e. public func ifPresent(@noescape f: (Wrapped) throws -> Void) rethrows -> Void Just to get the discussion back on track ;-) -Thorsten
_______________________________________________ swift-evolution mailing list [email protected] https://lists.swift.org/mailman/listinfo/swift-evolution
