[swift-evolution] Proposal: Split extensions into implementing methods and adding static functions Was: [swift-evolution-announce] [Review] SE-0164: Remove final support in protocol extensions
Howard Lovatt
howard.lovatt at gmail.com
Thu Apr 13 17:10:40 CDT 2017
I don't see that retroactive conformance needs to be exportable. If it is exported then you cannot prevent clashes from two modules, this is a known problem in C#. Because of this and other problems with C# extensions, this style of extension were rejected by other language communities (notably Java and Scala).
A better alternative for export is a new class that encapsulates the standard type but with added methods for the protocol to be added. This way there is no clash between modules. EG:
public protocol P {
func m() -> String
}
public class PInt: P {
var value = 0
func m() -> String { return "PI.m" }
}
-- Howard.
> On 13 Apr 2017, at 10:31 pm, Xiaodi Wu <xiaodi.wu at gmail.com> wrote:
>
> The retroactive conformance needs to be exportable. If one cannot vend a library that conforms standard library types to new protocols, then that is a non-starter.
>
>
>> On Thu, Apr 13, 2017 at 06:07 Howard Lovatt <howard.lovatt at gmail.com> wrote:
>> @Xiaodi,
>>
>> You can safely post-hoc add protocols and methods provided that they are final, do not override, and are not exported from the module. See version 2 of the proposal below.
>>
>> -- Howard.
>>
>> # Proposal: Split extension usage up into implementing methods and adding methods and protocols post-hoc
>>
>> Draft 2 (Added support for post-hoc conformance to a protocol - replaced static final extensions with final extensions)
>>
>> ## Introduction
>>
>> Currently extension methods are confusing because they have different dispatch rules for the same calling syntax. EG:
>>
>> public protocol P {
>> func mP() -> String
>> }
>> extension P {
>> func mP() -> String { return "P.mP" }
>> func mE() -> String { return "P.mE" }
>> }
>> struct S: P {
>> func mP() -> String { return "S.mP" }
>> func mE() -> String { return "S.mE" }
>> }
>> let s = S()
>> s.mP() // S.mP as expected
>> s.mE() // S.mE as expected
>> let p: P = s // Note: s now typed as P
>> p.mP() // S.mP as expected
>> p.mE() // P.mE unexpected!
>>
>> Extension methods can also cause compatibility problems between modules, consider:
>>
>> In Module A
>> extension Int: P {
>> func m() -> String { print("A.m") }
>> }
>>
>> In Module B
>> extension Int: P {
>> func m() -> String { print("B.m") }
>> }
>>
>> In Module C
>> import A
>> import B // Should this be an error
>> let i = 0
>> i.m() // Should it return A.m or B.m?
>>
>> This proposal cures the above two problems by separating extension methods into two seperate use cases: implementations for methods and adding methods and protocols post-hoc.
>>
>> ## Implementing methods
>>
>> If the extension is in the same file as the protocol/struct/class declaration then it implements the methods and is dispatched using a Vtable. EG:
>>
>> File P.swift
>> protocol/struct/class P {
>> // func m() not declared in type since it is added by the extension, under this proposal it is an error to include a declaration in a type *and* in an extension
>> }
>> extension P {
>> func m() { print("P.m") } // m is added to the protocol/struct/class declaration
>> }
>>
>> Same or other file
>> struct S: P {
>> override func m() { print("S.m") } // Note override required because m already has an implementation from the extension
>> }
>> let p: P = S() // Note typed as P
>> p.m() // Now prints S.m as expected
>>
>> Extensions in the same file as the declaration can have any access, can be final, and can have where clauses and provide inheritable implementations.
>>
>> In a protocol at present there is a difference in behaviour between a protocol that declares a method that is then implemented in an extension and a protocol that just has the method implemented in an extension and no declaration. This situation only applies to protocols, for structs and classes you cannot declare in type and implement in extensions. The proposal unifies the behaviour of protocol/struct/class with extensions and prevents the error of a minor typo between the protocol and extension adding two methods instead of generating an error.
>>
>> The implementation needed to achieve this is that a value instance typed as a protocol is copied onto the heap, a pointer to its Vtable added, and it is passed as a pointer. IE it becomes a class instance. No change needed for a class instance typed as a protocol.
>>
>> ## Post-hoc adding protocols and methods
>>
>> A new type of extension is proposed, a "final extension", which can be either in or outside the file in which the protocol/struct/class declaration is in. EG:
>>
>> protocol P2 {
>> func m2P()
>> }
>> final extension S: P2 { // Note extension marked final
>> func m2P() { print("SP2.m2P") } // Implicitly final, completely implements P2
>> func m2E() { print("SP2.m2E") } // Implicitly final, not an existing method
>> }
>>
>> Which are called as any other method would be called:
>>
>> let s = S()
>> s.m2P() // Prints SP2.m2P
>> s.m2E() // Prints SP2.m2E
>>
>> A method added by a final extension is implicitly final, as the name would suggest, and cannot be overridden.
>>
>> If the final extension:
>>
>> 1. Adds a method, e.g. m2E, that method cannot already exist. IE a final extension cannot override an existing method or implement a protocol declared method that lacks an implementation unless it also post-hoc adds the protocol.
>>
>> 2. Adds a protocol then it must implement all the methods in that protocol that are not currently implemented.
>>
>> 3. Is outside of the file in which the protocol/struct/class declaration is in then the extension and the methods can only have fileprivate or internal access. This prevents post-hoc extensions from numerous modules clashing, since they are not exported outside of the module.
>>
>> ## Possible future work (not part of this proposal)
>>
>> This proposal will naturally allow bodies to be added to protocols directly rather than via an extension, since under the proposal the extension adds the declaration to the type so it is a small step to allow the protocol methods to have an implementation.
>>
>> In an opposite sense to the above adding bodies to protocols, extensions could be allowed to add method declarations without bodies to protocols.
>>
>> The two above future work proposals, if both added, would add symmetry to where declarations and bodies may appear for protocols.
>>
>> ## In summary.
>>
>> The proposal formalises the split use of extensions into their two uses: implementing methods and post-hoc adding protocols and methods. Syntax is added that clarifies the two use cases, the former are termed extensions and must be in the same file as the type is declared, and the latter are termed final extensions and can be in any file, however if they are not in the type's file the they can only have fileprivate or internal access.
>>
>> Note the distinction between an extension in the same file and in a separate file is consistent with the philosophy that there is special status to the same file as proposed for private in https://github.com/apple/swift-evolution/blob/master/proposals/0169-improve-interaction-between-private-declarations-and-extensions.md.
>>
>>> On 11 Apr 2017, at 1:26 pm, Xiaodi Wu <xiaodi.wu at gmail.com> wrote:
>>>
>>> As far as I'm aware, eliminating retroactive conformances is a non-starter.
>>>
>>>
>>>> On Mon, Apr 10, 2017 at 21:44 Howard Lovatt <howard.lovatt at gmail.com> wrote:
>>>> @Xiaodi,
>>>>
>>>> You make two drugs.
>>>>
>>>> 1. Deliberately making retroactive conformance outside of the file in which the type is declared illegal because of the problems it causes. See all the questions on Swift Users and watch people learning Swift get caught out.
>>>>
>>>> 2. Outside of the file in which the type is declared the static final extension is restricted to internal or fileprivate so that multiple modules can add static final extensions without clashes.
>>>>
>>>>
>>>> -- Howard.
>>>>
>>>>> On 11 Apr 2017, at 8:51 am, Xiaodi Wu <xiaodi.wu at gmail.com> wrote:
>>>>>
>>>>> On Mon, Apr 10, 2017 at 5:35 PM, Howard Lovatt via swift-evolution <swift-evolution at swift.org> wrote:
>>>>> In response to Jordan Rose's comment I suggest the following change:
>>>>>
>>>>> Proposal: Split extension usage up into implementing methods and adding static functions
>>>>>
>>>>> Currently extension methods are confusing because they have different dispatch rules for the same syntax. EG:
>>>>>
>>>>> protocol P {
>>>>> func m()
>>>>> }
>>>>> extension P {
>>>>> func m() { print("P.m") }
>>>>> }
>>>>> struct S: P {
>>>>> func m() { print("S.m") }
>>>>> }
>>>>> val p: P = S() // Note typed as P
>>>>> p.m() // Surprisingly prints P.m even though S implements its own m
>>>>>
>>>>> This is incorrect. This prints "S.m", not "P.m".
>>>>>
>>>>> val s = S() // Note typed as S
>>>>> s.m() // Prints S.m as expected
>>>>>
>>>>> This proposal cures the above problem by separating extension methods into two seperate use cases: implementations for methods and adding static functions.
>>>>>
>>>>> First implementing methods.
>>>>>
>>>>> If the extension is in the same file as the protocol/struct/class declaration then it implements the methods and is dispatched using a Vtable. EG:
>>>>>
>>>>> File P.swift
>>>>> protocol/struct/class P {
>>>>> func m()
>>>>> }
>>>>> extension P {
>>>>> func m() { print("P.m") }
>>>>> }
>>>>>
>>>>> Same or other file
>>>>> struct S: P {
>>>>> override func m() { print("S.m") } // Note override required because m already has an implementation from the extension
>>>>>
>>>>> Requiring `override` breaks retroactive conformance of types to protocols. This idea has been brought up over half a dozen times. Each time it fails in not being able to accommodate retroactive conformance.
>>>>>
>>>>> }
>>>>> val p: P = S() // Note typed as P
>>>>> p.m() // Now prints S.m as expected
>>>>>
>>>>> Extensions in the same file as the declaration can have any access, can be final, and can have where clauses and provide inheritable implementations.
>>>>>
>>>>> The implementation needed to achieve this is that a value instance typed as a protocol is copied onto the heap, a pointer to its Vtable added, and it is passed as a pointer. IE it becomes a class instance. No change needed for a class instance typed as a protocol.
>>>>>
>>>>> The second use case is adding static functions.
>>>>>
>>>>> A new type of extension is proposed, a static final extension, which can be either in or outside the file in which the protocol/struct/class declaration is in. EG:
>>>>>
>>>>> static final extension P { // Note extension marked static final
>>>>> func m() { print("P.m") }
>>>>> }
>>>>>
>>>>> Which is called as any other static function would be called:
>>>>>
>>>>> val s = S()
>>>>> P.m(s) // Prints P.m as expected
>>>>>
>>>>> The new static final extension is shorthand, particularly in the case of multiple functions, for:
>>>>>
>>>>> extension P {
>>>>> static final func m(_ this: P) { print("P.m") }
>>>>> }
>>>>>
>>>>> If the static final extension is outside of the file in which the protocol/struct/class declaration is in then the extension and the methods can only have fileprivate and internal access.
>>>>>
>>>>> What is the use case for having this restriction? What is the problem you are trying to solve?
>>>>>
>>>>>
>>>>> As at present protocol/struct/class can have both a static and instance method of the same name, m in the case of the example, because the usage syntax is distinct. As at present, static final extensions, both the extension and the individual functions, can have where clauses.
>>>>>
>>>>> In summary.
>>>>>
>>>>> The proposal formalises the split use of extensions into their two uses: implementing methods and adding static functions. Syntax is added that clarifies both for declarations and usage which type of extension is provided/in use.
>>>>>
>>>>> Note the distinction between an extension in the same file and in a separate file is consistent with the proposed use of private in https://github.com/apple/swift-evolution/blob/master/proposals/0169-improve-interaction-between-private-declarations-and-extensions.md.
>>>>>
>>>>> Comments?
>>>>>
>>>>> -- Howard.
>>>>>
>>>>>> On 7 Apr 2017, at 4:49 am, Jordan Rose <jordan_rose at apple.com> wrote:
>>>>>>
>>>>>> [Proposal: https://github.com/apple/swift-evolution/blob/master/proposals/0164-remove-final-support-in-protocol-extensions.md]
>>>>>>
>>>>>>>> On Apr 5, 2017, at 16:15, Howard Lovatt via swift-evolution <swift-evolution at swift.org> wrote:
>>>>>>>>
>>>>>>>> The review of SE-0164 "Remove final support in protocol extensions"
>>>>>>>>
>>>>>>>
>>>>>>>> What is your evaluation of the proposal?
>>>>>>> The present situation isn't great. People get confused about which method will called with protocol extensions. Seems like every week there is a variation on this confusion on Swift Users mailing list. Therefore something needs to be done.
>>>>>>>
>>>>>>> However I am not keen on this proposal since it makes behaviour inconsistent between methods in protocol extensions, classes, and structs.
>>>>>>>
>>>>>>> I think a better solution would be one of the following alternatives:
>>>>>>>
>>>>>>> 1. Must use final and final means it cannot be overridden; or
>>>>>>> 2. If not final dispatches using a table like a class and if marked final cannot be overridden and if marked dynamic uses obj-c dispatching; or
>>>>>>> 3. Must be marked dynamic and uses obj-c dispatching.
>>>>>>>
>>>>>>> My preference would be option 2 but I think any of the three is superior to the present situation or the proposal.
>>>>>>
>>>>>> People have suggested all of these before, but none of them are obviously correct. It's true that we have a difference between extension members that satisfy requirements and those that don't, and that that confuses people. However, an extension-only member of one protocol can be used to satisfy the requirements of another protocol today, which is a tool for code reuse.
>>>>>>
>>>>>> (I think we managed to convince everyone that it's just a bug that a protocol extension method that satisfies a requirement cannot be overridden in a subclass, so at least that isn't an issue on top of the rest of this.)
>>>>>>
>>>>>> Oh, and we can't retroactively add members of a protocol extension to existing adopters, which is why protocol extension members cannot be @objc. There are limited circumstances where that would be safe, but that would be a separate proposal.
>>>>>>
>>>>>> Jordan
>>>>>
>>>>> _______________________________________________
>>>>> 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/20170414/fca81007/attachment.html>
More information about the swift-evolution
mailing list