[swift-users] is this a defect in equatable for swift tuples?
Martin R
martinr448 at gmail.com
Sun Jul 9 14:30:53 CDT 2017
> On 9. Jul 2017, at 21:20, David Baraff <davidbaraff at gmail.com> wrote:
>
> Nice: i hadn’t seen elementsEqual.
>
> (1) Why do you have to pass in “by: ==“ ? is not that the default
There are two versions: One taking an explicit predicate, and another which requires sequences of Equatable elements. As observed earlier in this thread, an array of tuples is not Equatable.
>
> (2) not a big deal, but if the sequence type’s length can be determined a priori (e.g. in the case of an Array, or perhaps a Collection if that has a count member, haven’t checked) does the elementsEqual function short circuit by first checking that the lengths are equal before beginning the loop?
As far as I can tell from the implementation at https://github.com/apple/swift/blob/master/stdlib/public/core/SequenceAlgorithms.swift.gyb#L292 <https://github.com/apple/swift/blob/master/stdlib/public/core/SequenceAlgorithms.swift.gyb#L292> , the length is not checked a-priori, so you might want to add that to your code:
class Tree : Equatable {
let rootData:Int = 0
let children:[(String, Tree)] = []
static public func ==(_ lhs:Tree, _ rhs:Tree) -> Bool {
return lhs.rootData == rhs.rootData &&
lhs.children.count == rhs.children.count &&
lhs.children.elementsEqual(rhs.children, by: ==)
}
}
>
> But again, that’s a great one to know.
>
>> On Jul 9, 2017, at 12:14 PM, Martin R via swift-users <swift-users at swift.org <mailto:swift-users at swift.org>> wrote:
>>
>>
>>> On 9. Jul 2017, at 21:00, Jens Persson via swift-users <swift-users at swift.org <mailto:swift-users at swift.org>> wrote:
>>>
>>> Since Array has .elementsEqual, another workaround (until conditional conformance) is:
>>>
>>> class Tree : Equatable {
>>> let rootData:Int
>>> let children:[(String, Tree)]
>>>
>>> init(rootData: Int, children: [(String, Tree)]) {
>>> self.rootData = rootData
>>> self.children = children
>>> }
>>> static public func ==(_ lhs:Tree, _ rhs:Tree) -> Bool {
>>> return lhs.rootData == rhs.rootData &&
>>> lhs.children.elementsEqual(rhs.children, by: { (a: (String, Tree), b: (String, Tree)) -> Bool in
>>> return a.0 == b.0 && a.1 == b.1
>>> })
>>> }
>>> }
>>
>>
>> Slightly simpler (since == is already defined for the tuples):
>>
>> class Tree : Equatable {
>> let rootData:Int = 0
>> let children:[(String, Tree)] = []
>>
>> static public func ==(_ lhs:Tree, _ rhs:Tree) -> Bool {
>> return lhs.rootData == rhs.rootData &&
>> lhs.children.elementsEqual(rhs.children, by: ==)
>> }
>> }
>>
>>
>>>
>>>
>>> On Sun, Jul 9, 2017 at 8:44 PM, David Sweeris <davesweeris at mac.com <mailto:davesweeris at mac.com>> wrote:
>>>
>>> On Jul 9, 2017, at 10:06, David Baraff via swift-users <swift-users at swift.org <mailto:swift-users at swift.org>> wrote:
>>>
>>>>
>>>>> On Jul 9, 2017, at 8:27 AM, Jens Persson <jens at bitcycle.com <mailto:jens at bitcycle.com>> wrote:
>>>>>
>>>>> (Also, note that your implementation of == uses lhs === rhs thus will only return true when lhs and rhs are the same instance of SomeClass.)
>>>> Of course — i threw that in just to make a simple example.
>>>>
>>>> Followup question: what I really wanted to write was an == operator for a tree:
>>>>
>>>> // silly tree, useful for nothing
>>>> class Tree : Equatable {
>>>> let rootData:Int
>>>> let children:[(String, Tree)]
>>>>
>>>> static public func ==(_ lhs:Tree, _ rhs:Tree) {
>>>> return lhs.rootData == rhs.rootData &&
>>>> lhs.children == rhs.children // sadly, this doesn’t compile
>>>> }
>>>> }
>>>
>>> Right, the `==` func is *defined* for 2-element tuples where both elements conform to `Equatable`, but that tuple type doesn't itself *conform* to `Equatable`. So the`==` func that's defined on "Array where Element: Equatable" can't see it.
>>>
>>> We'd need both "conditional conformance" and "tuple conformance" in order for that to Just Work.
>>>
>>> - Dave Sweeris
>>>
>>> _______________________________________________
>>> swift-users mailing list
>>> swift-users at swift.org <mailto:swift-users at swift.org>
>>> https://lists.swift.org/mailman/listinfo/swift-users <https://lists.swift.org/mailman/listinfo/swift-users>
>> _______________________________________________
>> swift-users mailing list
>> swift-users at swift.org <mailto:swift-users at swift.org>
>> https://lists.swift.org/mailman/listinfo/swift-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.swift.org/pipermail/swift-users/attachments/20170709/1b1a9ce3/attachment.html>
More information about the swift-users
mailing list