[swift-evolution] [Proposal Draft] Literal Syntax Protocols

Dave Abrahams dabrahams at apple.com
Fri Jul 1 15:35:25 CDT 2016


on Fri Jul 01 2016, Erica Sadun <erica-AT-ericasadun.com> wrote:

> The best way to pass the Dave Test is to ask him directly, for example:
>
> Dave:
>
> Do you think the stdlib team would be okay with a naming scheme like 
> Syntax.Literal.ArrayProtocol,  Syntax.Literal.IntegerProtocol, etc. We think 
> this produces a clear description of the conformant role and one that is unlikely to be 
> misinterpreted. It may read less fluently but it's also less subject
> to confusing users. 

No.  IMO Putting “Protocol” in the name just adds redundant type
information that doesn't help readability.

> This naming scheme uses the Syntax namespacing, and creates a Literal
> subspace.  Each protocol is named as "XXXXProtocol". This introduction
> a distinction between "This is/can be used as an integer literal" and
> "Conforming to this protocol ensures that an instance of the type can
> be written as an integer literal".
>
> The problem with earlier approximations was that people saw
> "Syntax.IntegerLiteralXXX" and thought the typed could be substituted
> into expressions where an integer literal was used, and not that an
> integer literal could be be used to write an instance of the type.
>
> So what do (and your team) think of this idea?

I think if `Syntax.IntegerLiteral` is actually unclear then the best
cure is `ExpressibleAsIntegerLiteral` (no namespace needed).  None of
the other suggestions I've seen describe what the protocol means as well
as that.  I've asked Matthew to update the proposal accordingly.

>
> -- Adrian (and Erica)
>
>> On Jul 1, 2016, at 2:08 AM, Adrian Zubarev via swift-evolution <swift-evolution at swift.org> wrote:
>> 
>> We haven’t pass the dave test yet? :D 
>> 
>> Still curious what he’d say about Syntax.Literal.*Protocol
>> 
>> One more question:
>> 
>> What can the namespace Syntax could be used for except for literals, any idea? (I have no clue.)
>

-- 
Dave


More information about the swift-evolution mailing list