[swift-evolution] Smart KeyPaths
Matthew Johnson
matthew at anandabits.com
Mon Mar 20 09:52:53 CDT 2017
> On Mar 20, 2017, at 9:23 AM, Christopher Kornher via swift-evolution <swift-evolution at swift.org> wrote:
>
>
>
>> On Mar 20, 2017, at 5:12 AM, David Hart via swift-evolution <swift-evolution at swift.org> wrote:
>>
>>
>>
>>> On 20 Mar 2017, at 10:39, Jonathan Hull via swift-evolution <swift-evolution at swift.org> wrote:
>>>
>>> +1. This is my favorite solution so far.
>>>
>>> With ‘Person.keypath.name' it is obvious that we are creating a key path. There is no ambiguity for the reader. With autocomplete it will be very little extra typing anyway…
>>
>> But that adds a lot of verbosity. They disregarded #keyPath because it was too verbose.
>
> The syntax in the original proposal is terse and elegant, and will probably be fine when a developer who is experienced with Swift and a particular codebase is first writing the code. Using “key path” or “keypaths” of perhaps a shorter term or even a single leading character (`#` ?) will make this feature more discoverable, tool-friendly and its usages more maintainable.
It also makes it inconsistent with how unbound methods behave. I have yet to hear a convincing argument about why key paths should be treated different syntactically.
>
> An extra term or character does add verbosity. How much is subjective, but I would not call it “a lot”. It does not add any nesting or code complexity. KVO code is usually a small fraction of most Objective-C projects (in my experience, at least) and it is probably safe to assume that the usage of this feature in Swift will be similar.
>
> Verbosity vs clarity is often a tradeoff and I think that on balance, for a feature like this a little extra verbosity is worth it. Swift does not have the most terse syntax possible. `++` was removed, for example.
>
> Just because an assignment is already implicitly typed in Swift does not mean that the ambiguity has to keep increasing without end for implementation of all new features, especially for ones that are not used very frequently.
>
>
>>
>>> Thanks,
>>> Jon
>>>
>>>> On Mar 19, 2017, at 9:20 PM, Dietmar Planitzer via swift-evolution <swift-evolution at swift.org> wrote:
>>>>
>>>> Key paths of this form:
>>>>
>>>> Person.name
>>>>
>>>> will always make it harder than necessary to:
>>>>
>>>> * search for all places where we are using key paths
>>>>
>>>> * do efficient code completion. Eg you’ll get a mix of static properties and key paths
>>>>
>>>>
>>>> We’ve been using this kind of setup in our projects for some time now:
>>>>
>>>> class Person {
>>>>
>>>> struct keypath {
>>>>
>>>> static let name = #keyPath(Person.name)
>>>> …
>>>> }
>>>>
>>>> …
>>>> }
>>>>
>>>> where a keypath is then used like this:
>>>>
>>>> Person.keypath.name
>>>>
>>>> and this has worked very well. It makes it easy to see where we are using a keypath rather than accessing some static property, it works very nicely with code completion and it makes it very easy to search for all places where we are using key paths from the Person type.
>>>>
>>>> I would prefer that the proposed keypath model would automatically organize key paths like this.
>>>>
>>>>
>>>> Regards,
>>>>
>>>> Dietmar Planitzer
>>>>
>>>>
>>>>> On Mar 19, 2017, at 20:49, Matthew Johnson via swift-evolution <swift-evolution at swift.org> wrote:
>>>>>
>>>>>
>>>>>
>>>>> Sent from my iPad
>>>>>
>>>>>> On Mar 19, 2017, at 10:31 PM, Jaden Geller via swift-evolution <swift-evolution at swift.org> wrote:
>>>>>>
>>>>>> I think the clarity desired is more similar to that obtained by the `try` keyword. Ya, the compiler knows that this function throws already, but Swift aims for clarity in the source code. Clarity is often achieved by providing potentially redundant information for the programmer.
>>>>>>
>>>>>> As proposed, it is difficult to distinguish a key path from a static variable. Maybe that's not problematic? Well, it's up to the community to decide.
>>>>>
>>>>> Why don't we just say all instance properties are shadowed by a static constant property of the same name with the appropriate key path type. This makes it not mysterious at all but instead very straightforward. We could even say that static and class properties are shadowed by a key path property on the meta type.
>>>>>
>>>>>
>>>>>> I do think it is a bit worrisome that static variable access might cause side effects (or at least, might take a while to compute) but creating key paths should not, but that's a fringe case probably.
>>>>>>
>>>>>> On Mar 19, 2017, at 6:32 PM, Brent Royal-Gordon via swift-evolution <swift-evolution at swift.org> wrote:
>>>>>>
>>>>>>>> On Mar 19, 2017, at 4:47 PM, Charles Srstka <cocoadev at charlessoft.com> wrote:
>>>>>>>>
>>>>>>>>> This is true of many things. It is why IDEs make type information readily available.
>>>>>>>>
>>>>>>>> Is clarity not a thing to be desired?
>>>>>>>
>>>>>>> Clarity is in the eye of the beholder. Here's one notion of clarity:
>>>>>>>
>>>>>>> sum :: (Num a, Foldable t) => t a -> a
>>>>>>> sum = foldl (+) 0
>>>>>>>
>>>>>>> Here's another:
>>>>>>>
>>>>>>> int sum(int array[], size_t len) {
>>>>>>> int total = 0;
>>>>>>> for(size_t i = 0; i < len; i++) {
>>>>>>> total += array[i];
>>>>>>> }
>>>>>>> return total;
>>>>>>> }
>>>>>>>
>>>>>>> And another:
>>>>>>>
>>>>>>> SUM PROC
>>>>>>> ; this procedure will calculate the sum of an array
>>>>>>> ; input : SI=offset address of the array
>>>>>>> ; : BX=size of the array
>>>>>>> ; output : AX=sum of the array
>>>>>>>
>>>>>>> PUSH CX ; push CX onto the STACK
>>>>>>> PUSH DX ; push DX onto the STACK
>>>>>>>
>>>>>>> XOR AX, AX ; clear AX
>>>>>>> XOR DX, DX ; clear DX
>>>>>>> MOV CX, BX ; set CX=BX
>>>>>>>
>>>>>>> @SUM: ; loop label
>>>>>>> MOV DL, [SI] ; set DL=[SI]
>>>>>>> ADD AX, DX ; set AX=AX+DX
>>>>>>> INC SI ; set SI=SI+1
>>>>>>> LOOP @SUM ; jump to label @SUM while CX!=0
>>>>>>>
>>>>>>> POP DX ; pop a value from STACK into DX
>>>>>>> POP CX ; pop a value from STACK into CX
>>>>>>>
>>>>>>> RET ; return control to the calling procedure
>>>>>>> SUM ENDP
>>>>>>>
>>>>>>>
>>>>>>> And one more:
>>>>>>>
>>>>>>> extension Sequence where Element: Arithmetic {
>>>>>>> func sum() {
>>>>>>> return reduce(0, +)
>>>>>>> }
>>>>>>> }
>>>>>>>
>>>>>>> Clarity is not achieved by explicitly stating every detail of your code. It's achieved by explicitly stating what needs to be said, and *not* explicitly stating what *doesn't* need to be said.
>>>>>>>
>>>>>>> The people who oppose using a special syntax for this feature think that, by and large, clarity is best served by *not* explicitly stating when you're using a key path. They believe that you are unlikely to run into ambiguity and, when you do, it will be easy to work around it. This is an opinion, so it's open to disagreement, but that's where they stand on it.
>>>>>>>
>>>>>>> --
>>>>>>> Brent Royal-Gordon
>>>>>>> Architechies
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> 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
>>>>> _______________________________________________
>>>>> 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
>>>
>>> _______________________________________________
>>> 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
>
> _______________________________________________
> swift-evolution mailing list
> swift-evolution at swift.org
> https://lists.swift.org/mailman/listinfo/swift-evolution
More information about the swift-evolution
mailing list