220 12903 article Path: news.gmane.org!not-for-mail From: Richard Smith Newsgroups: gmane.comp.lang.c++.isocpp.proposals Subject: Re: unevaluated this Date: Mon, 15 Sep 2014 15:43:11 -0700 Lines: 139 Approved: news@gmane.org Message-ID: References: <009e6218-1e2b-402d-ab1a-e8ebf29f8f77@isocpp.org> Reply-To: std-proposals@isocpp.org NNTP-Posting-Host: plane.gmane.org Mime-Version: 1.0 Content-Type: multipart/alternative; boundary=001a1133375c11cb350503225c92 X-Trace: ger.gmane.org 1410821005 13311 80.91.229.3 (15 Sep 2014 22:43:25 GMT) X-Complaints-To: usenet@ger.gmane.org NNTP-Posting-Date: Mon, 15 Sep 2014 22:43:25 +0000 (UTC) To: std-proposals@isocpp.org Original-X-From: std-proposals+bncBDVNBJG4YAIBBAGX3WQAKGQEDMLMH6Y@isocpp.org Tue Sep 16 00:43:18 2014 Return-path: Envelope-to: gclcip-std-proposals@m.gmane.org Original-Received: from mail-qa0-f71.google.com ([209.85.216.71]) by plane.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1XTezK-0004Vq-5w for gclcip-std-proposals@m.gmane.org; Tue, 16 Sep 2014 00:43:14 +0200 Original-Received: by mail-qa0-f71.google.com with SMTP id k15sf12290903qaq.2 for ; Mon, 15 Sep 2014 15:43:13 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:sender:in-reply-to:references:date :message-id:subject:from:to:x-original-sender :x-original-authentication-results:reply-to:precedence:mailing-list :list-id:list-post:list-help:list-archive:list-subscribe :list-unsubscribe:content-type; bh=LohwjooK+Jf3u8GapFwI6ShuZ2dVpr/yCMnLJJ2WdXQ=; b=gsmw3+VYfye2WvN4rzLIRVBTa1T49eV05hf+GJDgi+qmWMP6NPSctjvXNE72hSpGhA ks4CTkOffYGyoe85Cnh91pnVDW0D2hyUKDBRzzcKcmQaEhj8L9ytVgJX9IEyxM27yzgR Vj/DuR2OJNXsXh1SPxmLt2yQst5EJjG+FYx2F+NNH8dJ0tElV1ebeyoSxfmmo5CNC9Vb lp4f1+s4S4X4EQ+NJM0/2Q4dYRh5tS/cN33cE320I5NdZH5Mwrn5H3LfNAbbkgZ5WW1Z Zl3Q49x/cITQZS7+8Ff6qqy2TJzbZq7enIzcEs3tdQod4155RhNiQroHV9I66bvTRhzU /PKQ== X-Gm-Message-State: ALoCoQnJ/4lUwhZfR6hYiVXjGg+p2BFLficKHottZ+B04MX4MQHUTsMr9I+G0oKwh53w3Qf9roHY X-Received: by 10.236.8.74 with SMTP id 50mr16875700yhq.42.1410820993348; Mon, 15 Sep 2014 15:43:13 -0700 (PDT) X-BeenThere: std-proposals@isocpp.org Original-Received: by 10.140.47.87 with SMTP id l81ls1729241qga.51.gmail; Mon, 15 Sep 2014 15:43:12 -0700 (PDT) X-Received: by 10.221.45.132 with SMTP id uk4mr4596500vcb.57.1410820992458; Mon, 15 Sep 2014 15:43:12 -0700 (PDT) Original-Received: from mail-vc0-x22c.google.com (mail-vc0-x22c.google.com [2607:f8b0:400c:c03::22c]) by mx.google.com with ESMTPS id f2si6896575vcy.28.2014.09.15.15.43.12 for (version=TLSv1 cipher=ECDHE-RSA-RC4-SHA bits=128/128); Mon, 15 Sep 2014 15:43:12 -0700 (PDT) Received-SPF: pass (google.com: domain of metafoo@gmail.com designates 2607:f8b0:400c:c03::22c as permitted sender) client-ip=2607:f8b0:400c:c03::22c; Original-Received: by mail-vc0-f172.google.com with SMTP id hy10so4127293vcb.31 for ; Mon, 15 Sep 2014 15:43:12 -0700 (PDT) X-Received: by 10.221.23.66 with SMTP id qz2mr26688760vcb.46.1410820991992; Mon, 15 Sep 2014 15:43:11 -0700 (PDT) Original-Sender: metafoo@gmail.com Original-Received: by 10.220.76.200 with HTTP; Mon, 15 Sep 2014 15:43:11 -0700 (PDT) In-Reply-To: X-Original-Sender: richard@metafoo.co.uk X-Original-Authentication-Results: mx.google.com; spf=pass (google.com: domain of metafoo@gmail.com designates 2607:f8b0:400c:c03::22c as permitted sender) smtp.mail=metafoo@gmail.com; dkim=pass header.i=@gmail.com Precedence: list Mailing-list: list std-proposals@isocpp.org; contact std-proposals+owners@isocpp.org List-ID: X-Google-Group-Id: 399137483710 List-Post: , List-Help: , List-Archive: List-Subscribe: , List-Unsubscribe: , Xref: news.gmane.org gmane.comp.lang.c++.isocpp.proposals:12903 Archived-At: --001a1133375c11cb350503225c92 Content-Type: text/plain; charset=UTF-8 On Mon, Sep 15, 2014 at 3:09 PM, Zhihao Yuan wrote: > On Mon, Sep 15, 2014 at 5:24 PM, Richard Smith > wrote: > >> >> Oops, sorry, I meant to refer to the previous proposal of >> 'decltype(class)'. =) >> > > Err, I misparsed the paper, too. > > Copy-pasting Ville's comment on decltype(class): > > "The use of the type and value category of 'this' is already allowed in > static member > functions as per [expr.prim.general]/3. > This is only permitted as a hack: compilers don't know for sure whether they have an out-of-line definition of a static or non-static member function until they get to the end of the function declarator. 'this' is permitted before then. To support this, the type and value category of 'this' are defined for a static member function, but your program is ill-formed if it actually uses 'this' in a static member function. Example: struct S { template auto f() -> S*; template static auto f() -> const S*; }; template auto S::f() -> decltype(this) {} Here, we cannot tell whether the out-of-line definition is static or not unless we know the type of 'this'. So we define that 'this' has type 'S*' for a static member function, and we're able to resolve the redefinition to being the non-static function, and all is OK. Allowing similar uses in the class definition > has its problems, but decltype(class) likely doesn't solve those problems > (the type may > be incomplete because the class definition is not yet known sufficiently, > all > sorts of chicken-and-egg problems arise). " > > // and he referred you to provide more details about type completeness here > Class completeness should not be a problem here any more than it would be if you named the class by its name; I don't anticipate any new problems in that direction, but we would have a new syntax for writing the existing problems. (I'd expect decltype(class) to be identical to naming the class in almost all ways -- except where the class is anonymous, or its name is hidden, or a typedef-name is disallowed etc.) -- --- You received this message because you are subscribed to the Google Groups "ISO C++ Standard - Future Proposals" group. To unsubscribe from this group and stop receiving emails from it, send an email to std-proposals+unsubscribe@isocpp.org. To post to this group, send email to std-proposals@isocpp.org. Visit this group at http://groups.google.com/a/isocpp.org/group/std-proposals/. --001a1133375c11cb350503225c92 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
On M= on, Sep 15, 2014 at 3:09 PM, Zhihao Yuan <zy@miator.net> wrote:<= br>
On Mon, Sep 15, 2014 = at 5:24 PM, Richard Smith <richard@metafoo.co.uk> wrote:=
=

=
Oops, sorry, I meant to refer to the previous proposal of= 'decltype(class)'. =3D)=C2=A0
=

Err, I misparsed the pape= r, too.

Copy-pasting Ville's co= mment on decltype(class):

=C2=A0 "The use of the type and value= category of 'this' is already allowed in static member
function= s as per [expr.prim.general]/3.

This is only permitted as a hack: compilers don't know for sure whethe= r they have an out-of-line definition of a static or non-static member func= tion until they get to the end of the function declarator. 'this' i= s permitted before then. To support this, the type and value category of &#= 39;this' are defined for a static member function, but your program is = ill-formed if it actually uses 'this' in a static member function.<= /div>

Example:

struct S {<= /div>
=C2=A0 template<typename T> auto f() -> S*; =C2=A0 =C2= =A0 =C2=A0
=C2=A0 template<typename T> static auto f() ->= ; const S*;
};
template<typename T> auto S::f() -= > decltype(this) {}

Here, we cannot tell = whether the out-of-line definition is static or not unless we know the type= of 'this'. So we define that 'this' has type 'S*' = for a static member function, and we're able to resolve the redefinitio= n to being the non-static function, and all is OK.

=
Allowing simil= ar uses in the class definition
has its problems, but decltype(class) li= kely doesn't solve those problems (the type may
be incomplete becaus= e the class definition is not yet known sufficiently, all
sorts of chick= en-and-egg problems arise). "

= // and he referred you to provide more details about type completeness here=

Class completeness should not = be a problem here any more than it would be if you named the class by its n= ame; I don't anticipate any new problems in that direction, but we woul= d have a new syntax for writing the existing problems. (I'd expect decl= type(class) to be identical to naming the class in almost all ways -- excep= t where the class is anonymous, or its name is hidden, or a typedef-name is= disallowed etc.)

--

---
You received this message because you are subscribed to the Google Groups &= quot;ISO C++ Standard - Future Proposals" group.
To unsubscribe from this group and stop receiving emails from it, send an e= mail to std-proposa= ls+unsubscribe@isocpp.org.
To post to this group, send email to std-proposals@isocpp.org.
Visit this group at http://groups.google.com/a/isocpp.org/group/std-proposals/<= /a>.
--001a1133375c11cb350503225c92-- .