[swift-evolution] Proposal: An assignment operator := that ensures nil before assignment.
Howard Lovatt
howard.lovatt at gmail.com
Sat Feb 27 20:58:52 CST 2016
Sorry -1 from me. Not worth adding. Too few use cases.
On Sunday, 28 February 2016, Brent Royal-Gordon via swift-evolution <
swift-evolution at swift.org> wrote:
> > Here’s a use case. Suppose you are filling out a dictionary in some
> non-trivial way. However, it should be the case that the value for each key
> is computed and assigned only once. And so you would use the := operator to
> ensure this: dictionary[k] := value
>
> This is actually a great use case, and it's exactly the kind of thing you
> should mention when you suggest a change to Swift.
>
> However, I think that specific use case is better served by a method on
> Dictionary. Perhaps something equivalent to:
>
> mutating func initializeValue(value: Value, forKey key: Key) {
> let oldValue = updateValue(value, forKey: key)
> precondition(oldValue == nil, "Initialized \(key) when it
> was already set")
> }
>
> Why do I think this approach is better?
>
> * `:=` doesn't really explain what it does—if you've never seen the
> operator before, the most you can be sure of is that it's probably some
> kind of assignment. `initializeValue(_:forKey:)` *does* explain the
> intended semantic—this should set a value for the first time—which suggests
> that it shouldn't be used with a value that's already set.
>
> * I would need to see compelling examples of `:=` used outside of
> Dictionaries before I thought it would be useful as something applied to
> any lvalue. Dictionary is an unusual case because they have a
> dynamically-determined set of keys which are created simply by assigning
> them, so it's reasonable to not know whether a particular key is `nil` or
> not. Swift variables and properties, and most Swift collections, separate
> adding a value from updating it, so they may not need an analogous
> operation.
>
> * By putting the precondition inside `initializeValue(_:forKey:)` itself,
> the error message can actually include the key, which may aid in debugging.
>
> --
> Brent Royal-Gordon
> Architechies
>
> _______________________________________________
> swift-evolution mailing list
> swift-evolution at swift.org <javascript:;>
> https://lists.swift.org/mailman/listinfo/swift-evolution
>
--
-- Howard.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.swift.org/pipermail/swift-evolution/attachments/20160228/42850bff/attachment.html>
More information about the swift-evolution
mailing list