[swift-evolution] Warn about unused Optional.some(())

[Jordan Rose]

I think I see Alex's point here. Optional chaining is still intended to be a substitute for Objective-C's nil-swallowing, and therefore foo?.bar() should not warn if 'bar' has a discardable result, even though there is semantic information about whether the method was actually called. I think that of the three things under consideration here:

1. foo?.bar() should not warn
2. foo.map(baz) should warn
3. Ternaries should be consistent with non-ternaries

#1 is the most important, at least to me. The Swift 3 change was to sacrifice #2 in favor of #3, which I'm not sure I would have done, but I wouldn't want to sacrifice #1 in favor of #2.

I wouldn't mind the model of the type being '@discardableResult Optional<Void>' or whatever, but I think that's probably more work than anyone wants to sign up for.

Jordan


> On Jan 31, 2017, at 08:16, Daniel Duan via swift-evolution <swift-evolution at swift.org> wrote:
> 
> Good to know the history. If I were to fix the inconsistency, I'd add the warning to optional chaining instead. 
> 
> Deliberately make the compiler give us *less* information for esthetic reasons feels wrong to me. As I mentioned in the original email, this has cost us a few unnoticed bad patterns slipping into our production. That's the opposite of what this type of warning is supposed to achieve.
> 
>> On Jan 31, 2017, at 1:07 AM, Alex Hoppen <alex at ateamer.de> wrote:
>> 
>> This was a deliberate change between Swift 3 beta 1 and beta 2 after a friend of mine pointed the following inconsistency out to me:
>> 
>> struct Foo {
>> func bar() {}
>> }
>> let foo: Foo? = Foo()
>> foo?.bar() // Does not create a warning
>> true ? foo?.bar() : foo?.bar()  // expression of type '()?' is unused
>> 
>> After some offline discussion at WWDC with the Swift team we decided to move to a consistent model where ()?, ()??, … is always discardable since we didn't want to take the convenience of foo?.bar() away (something that regularly occurs with weak variables, e.g. captures in closures).
>> 
>> So much for the history of this feature.
>> 
>> – Alex
>> 
>> 
>>> On 30 Jan 2017, at 22:58, Daniel Duan via swift-evolution <swift-evolution at swift.org> wrote:
>>> 
>>> Hi all,
>>> 
>>> Right now, expressions that evaluates to Optional<()>, Optional<Optional<()>>… gets special treatment when it’s unused. For example:
>>> 
>>> func f(s: String) {}
>>> let s: String = “”
>>> s.map(f) // no warning here, even tho the resulting type is `Optional<()>` and unused.
>>> 
>>> func g() throws {}
>>> try? g() // no warnings here neither.
>>> 
>>> This is convenient, but encourages composing map/filter/reduce, etc with side-effect-ful functions, which we have found a few cases of in our production code recently. Granted, these cases could’ve been caught with more careful code reviews. But we wouldn’t have missed them if this “feature” didn’t exist.
>>> 
>>> I think we should remove the special treatment so that code in the example above would generate a warning about `()?` being unused. Users can silence it manually by assigning the result to `_`. 
>>> 
>>> OTOH, this would undermine the convenience of `try?` when the throwing function don’t return anything.
>>> 
>>> What do y’all think?
>>> 
>>> Daniel Duan
>>> _______________________________________________
>>> swift-evolution mailing list
>>> swift-evolution at swift.org
>>> https://lists.swift.org/mailman/listinfo/swift-evolution
>> 
> 
> _______________________________________________
> swift-evolution mailing list
> swift-evolution at swift.org
> https://lists.swift.org/mailman/listinfo/swift-evolution

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.swift.org/pipermail/swift-evolution/attachments/20170206/06870d4a/attachment.html>