[swift-evolution] [Design Question] Unexpected forced optional unwrapping behavior

[John McCall]

> On Jan 8, 2016, at 6:14 PM, Charles Srstka via swift-evolution <swift-evolution at swift.org> wrote:
>> On Jan 8, 2016, at 7:56 PM, Michael Henson via swift-evolution <swift-evolution at swift.org <mailto:swift-evolution at swift.org>> wrote:
>> 
>> I've spent some time exploring different aspects of the language's type system and came across a case where I wanted to know the type of the thing an optional might hold, if it exists. The direct type of an optional variable is, unsurprisingly, an optional:
>> 
>> > var example: String? = "example"
>> example: String? = "example"
>> > example.dynamicType
>> $R1: String?.Type = String?
>> 
>> 
>> Ok, that didn't work. What if I dereference?
>> 
>> > example!.dynamicType
>> $R2: String.Type = String
>> 
>> Ok. What if the optional doesn't currently have a value? Does this still work?
>> 
>> > example = nil
>> $R3: String.Type = String
>> 
>> That's odd. I expected it not work, given how forced unwrapping of nil optionals usually turns out.
>> 
>> Is this explicitly intended behavior, something that really shouldn't work and shouldn't be relied on, or... other?
> 
> Well, there’s at least one case where it doesn’t work, since this only seems to work with ‘var’ variables, and not with ‘let’ constants.
> 
>  1> var variable: String? = nil
> variable: String? = nil
>   2> let constant: String? = nil
> constant: String? = nil
>   3> variable!.dynamicType
> $R0: String.Type = String
>   4> constant!.dynamicType
> fatal error: unexpectedly found nil while unwrapping an Optional value
> Execution interrupted. Enter Swift code to recover and continue.
> Enter LLDB commands to investigate (type :help for assistance.)
> 
> I, too, am confused as to why this ever works at all.

Please file a bug; we might be failing to evaluate the base when we can statically fold the .dynamicType expression.

John.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.swift.org/pipermail/swift-evolution/attachments/20160108/18e370d6/attachment.html>