C++ Logo

STD-DISCUSSION

Advanced search

Subject: Re: A contradiction between [class.base.init#9] and [class.copy.ctor#14]
From: jim x (xmh970252187_at_[hidden])
Date: 2020-09-26 20:51:48


I don't agree. AFAIK, the rules in [class.base.init]/9 still conform to
implicitly-defined default constructors. And it also applies to
user-provided copy/move constructors which have no corresponding
member-initializer, such as this example <https://godbolt.org/z/YeW668>. I
think that `as if` is equivalent to `looks like` but does not do these
things actually, the similar wordings are also used in
[expr.prim.lambda.capture#6]
<https://eel.is/c++draft/expr.prim.lambda.capture#6>, however it definitely
does not mean that do these things would result in a well-formed code. This
is my opinion.

Andrew Schepler <aschepler_at_[hidden]> 于2020年9月26日周六 下午9:31写道:

> An implicitly-defined copy/move constructor can't be specified as if by a
> definition with mem-initializer-list, because of array members. There's no
> simple way to write the equivalent of the element-wise copy for arrays
> specified in [class.copy.ctor]/(14.1)
> <https://eel.is/c++draft/class.copy.ctor#14.1>.
>
> If a clarification for this is needed, it would be simpler to change
> [class.base.init]/9. For example, "In a non-defaulted, non-delegating
> constructor, ..."
>
> On Wed, Sep 23, 2020 at 4:53 AM jim x via Std-Discussion <
> std-discussion_at_[hidden]> wrote:
>
>> Hi, everyone, wish you a nice day.
>>
>> I have an issue about two rules in [class.base.init] and
>> [class.copy.ctor], respectively. They are:
>>
>> [class.base.init#9] <https://eel.is/c++draft/class.base.init#9>
>> In a non-delegating constructor, if a given potentially constructed
>> subobject *is not designated by a mem-initializer-id* (including the
>> case where there is no mem->initializer-list because the constructor has no
>> ctor-initializer), then
>>
>>
>> [class.copy.ctor#14] <https://eel.is/c++draft/class.copy.ctor#14>
>> The implicitly-defined copy/move constructor for a non-union class X
>> performs a memberwise copy/move of its bases and members. [Note 7: Default
>> member initializers of non-static data members are ignored. See also the
>> example in [class.base.init]. — end note] The order of initialization is
>> the same as the order of initialization of bases and members in a *user-defined
>> constructor *(see [class.base.init]). Let x be either the parameter of
>> the constructor or, for the move constructor, an xvalue referring to the
>> parameter. Each base or non-static data member is copied/moved in the
>> manner appropriate to its type
>>
>>
>> Except for the tips mentioned in `Note`, there's no any wording in the
>> standard that says the `implicitly-defined copy/move constructor` can
>> suppress the rules in [class.base.init#9] to apply to it. Due
>> to `implicitly-defined copy/move constructor` has no ctor-initializer, it
>> should obey [class.base.init#9]. So, it appears to be vague.
>>
>> So, Is it necessary to reword the rule in [class.copy.ctor#14]
>> <https://eel.is/c++draft/class.copy.ctor#14> to suppress the rule
>> [class.base.init#9] explicitly? Such as:
>>
>> The implicitly-defined copy/move constructor for a non-union class X
>> performs a memberwise copy/move of its bases and members *as if it
>> exists a corresponding member-initializer for each potentially constructed
>> subobject*.
>>
>> After this modification, this rule "Virtual base class subobjects shall
>> be initialized only once by the implicitly-defined copy/move constructor"
>> will also be clear due to it will conform to this rule, namely "A
>> mem-initializer where the mem-initializer-id denotes a virtual base class
>> is ignored during execution of a constructor of any class that is not the
>> most derived class." in [class.base.init#7]
>> <https://eel.is/c++draft/class.base.init#7>
>> Although, there's a rule in the standard to suppress that the constructor
>> of non most derived class to initialize the virtual base subobject, that
>> is:
>> [class.base.init#13.1] <https://eel.is/c++draft/class.base.init#13.1>
>> First, and only for the constructor of the most derived class
>> ([intro.object]), virtual base classes are initialized in the order they
>> appear on a depth-first left-to-right traversal of the directed acyclic
>> graph of base classes, where “left-to-right” is the order of appearance of
>> the base classes in the derived class base-specifier-list.
>>
>> However, it's a bit obscure, not obvious. Well, this modification is not
>> completely for why the virtual base subobject only initialized once, but
>> explicitly reject the rule in [class.base.init#9]
>> <https://eel.is/c++draft/class.base.init#9> which may apply
>> to `implicitly-defined copy/move constructor`. BTW, the sentence "The order
>> of initialization is the same as the order of initialization of bases and
>> members in a *user-defined constructor *(see [class.base.init])." sounds
>> like that these rules in [class.base.init] is applied to `user-defined
>> constructor`.
>>
>> Anyhow, It's only my reading. If you have any other opinions, please tell
>> me, thanks.
>>
>>
>>
>>
>> --
>> Std-Discussion mailing list
>> Std-Discussion_at_[hidden]
>> https://lists.isocpp.org/mailman/listinfo.cgi/std-discussion
>>
>



STD-DISCUSSION list run by std-discussion-owner@lists.isocpp.org

Older Archives on Google Groups