220 22141 <73b5fa06-b0e2-4f64-96a5-703fbdde4aa2@isocpp.org> article Path: news.gmane.org!not-for-mail From: Nicol Bolas Newsgroups: gmane.comp.lang.c++.isocpp.proposals Subject: Re: N4542: Should we be able to "widen" variant to variant, or "narrow" it to variant? Date: Thu, 29 Oct 2015 16:10:23 -0700 (PDT) Lines: 198 Approved: news@gmane.org Message-ID: <73b5fa06-b0e2-4f64-96a5-703fbdde4aa2@isocpp.org> References: Reply-To: std-proposals@isocpp.org NNTP-Posting-Host: plane.gmane.org Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="----=_Part_14_1132156555.1446160224012" X-Trace: ger.gmane.org 1446160229 373 80.91.229.3 (29 Oct 2015 23:10:29 GMT) X-Complaints-To: usenet@ger.gmane.org NNTP-Posting-Date: Thu, 29 Oct 2015 23:10:29 +0000 (UTC) To: ISO C++ Standard - Future Proposals Original-X-From: std-proposals+bncBCEKFTV6ZUMBBYOOZKYQKGQEYYNXD5Q@isocpp.org Fri Oct 30 00:10:27 2015 Return-path: Envelope-to: gclcip-std-proposals@m.gmane.org Original-Received: from mail-ig0-f198.google.com ([209.85.213.198]) by plane.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1ZrwKx-0005Mc-4S for gclcip-std-proposals@m.gmane.org; Fri, 30 Oct 2015 00:10:27 +0100 Original-Received: by igbxf8 with SMTP id xf8sf119319255igb.1 for ; Thu, 29 Oct 2015 16:10:26 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=date:from:to:message-id:in-reply-to:references:subject:mime-version :content-type:x-original-sender:reply-to:precedence:mailing-list :list-id:x-spam-checked-in-group:list-post:list-help:list-archive :list-subscribe:list-unsubscribe; bh=7DQh4eOlq6Gzj7yfp5BJ3uCFl96vt7HSY2/HsWSAHEk=; b=QsK+H3qq3wBAMrnfZwZV/p2TG/mpMu2rJyKYx+iw2rGxIISi2j6QKxaxD8AIfSG94v TauQRqXpr6aslT9M8mfduJW9qD6XFGwGNFrBRH5vjGiT4QW41IT/CjL+nZUJ+HU7M7JG D/XNmEEB3gZbdfDWLNiFPYdxKRWSXl0YxnZ/U9BjtCXO33/CMFJKO1LvMf5qIAe1B17h tiJ+mHgBsQj7sDCTDdHsdne5MN+mtKFFw/JDL3M6sjE9hjic1epQo0X3efx+9F7jCcU/ Pdhh9Jro5dsw0FaSpc6cA3SXI1DBvcOUM1Zl+Rj7N1AkZ9VEa4xN3gJqTZPNN7M6NCGv tE6g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:date:from:to:message-id:in-reply-to:references :subject:mime-version:content-type:x-original-sender:reply-to :precedence:mailing-list:list-id:x-spam-checked-in-group:list-post :list-help:list-archive:list-subscribe:list-unsubscribe; bh=7DQh4eOlq6Gzj7yfp5BJ3uCFl96vt7HSY2/HsWSAHEk=; b=jhpNsA2cWDe4m8Ovn3uhE8Vx8ygpdNYbq/G87nr78aKrPapDWy4QHHQ1puDhwE7PbE wiM2nw8qULzI44KnpOdP+LLHwo8Dsc+FrTz8y6NvJTNN5RZ9s3eaERQpGBLS0pQ6Yj5I YOAxr/PaMdzGhoDKn2imQSkIHik8S9FFasyGt9syhQjVek1VaojPO/wIPLYtSwegXFLw HmY8/hbZs3PF40GqjSLxBFuMsHBYBU9hJjzIkeCrQQnsPy23Eup7dQddFvAN2Rtfm0MU 1syw2wD5hteoeYAUwtoVvEsGb554376SPPgqVziC2LcWOlaIHaYh53pVrMRv1TOoBqGZ +CyA== X-Gm-Message-State: ALoCoQmnS25haytVbGQjv+7ukdWgYAXhOXZzf6nwOdlcMGV6xeQGp3pRegO4vTrhk+Q7HrzX45m+ X-Received: by 10.107.47.219 with SMTP id v88mr4776426iov.25.1446160226062; Thu, 29 Oct 2015 16:10:26 -0700 (PDT) X-BeenThere: std-proposals@isocpp.org Original-Received: by 10.107.27.83 with SMTP id b80ls449429iob.20.gmail; Thu, 29 Oct 2015 16:10:25 -0700 (PDT) X-Received: by 10.50.77.70 with SMTP id q6mr359577igw.4.1446160225067; Thu, 29 Oct 2015 16:10:25 -0700 (PDT) In-Reply-To: X-Original-Sender: jmckesson@gmail.com Precedence: list Mailing-list: list std-proposals@isocpp.org; contact std-proposals+owners@isocpp.org List-ID: X-Spam-Checked-In-Group: std-proposals@isocpp.org 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:22141 Archived-At: ------=_Part_14_1132156555.1446160224012 Content-Type: multipart/alternative; boundary="----=_Part_15_2090865752.1446160224012" ------=_Part_15_2090865752.1446160224012 Content-Type: text/plain; charset=UTF-8 On Thursday, October 29, 2015 at 6:09:38 PM UTC-4, Arthur O'Dwyer wrote: > > This question came up at the Mountain View meetup last night. > > I can imagine wanting to do this: > > using Value = std::variant; > using Simple = std::variant; > using Numeric = std::variant; > > void func(Simple x); > > Numeric n; > func(n); // Should this work, somehow, since every Numeric "is-a" > Simple in some sense? > Could it work? Yes. Should it work? Not implicitly; if it should be allowed, there ought to be a function for it. When should it work? Well, that's the question. > > Value v; > func(v); // Should this compile, somehow, and then throw at runtime > if v holds a List or Object? > No, it should not. Not implicitly. The "narrowing" case is sort of like dynamic_cast; > In no sense is it like a dynamic cast. `dynamic_cast` is about casting within a class hierarchy. Specifically, from child to parent. The only way it's similar is that it can conceptually fail. Or, is there a clever trick to get this kind of behavior already, without > having to write an arcane and expensive helper function that does a bunch > of get()s and assignments? > There is nothing arcane about a simple visitor: template variant_conv { template variant operator()(const T &t) const { //If T not in Args, throw. return t; } }; auto v = visit(variant_conv{}, variant{ 3.4f}); You can even wrap that up in a template function. -- --- 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/. ------=_Part_15_2090865752.1446160224012 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
On Thursday, October 29, 2015 at 6:09:38 PM UTC-4, Arthur = O'Dwyer wrote:
This question came up at the Mountain View meetup last night.

=
I can imagine wanting to do this:

=C2= =A0 =C2=A0 using Value =3D std::variant<int,double,std::string,List= ,Object>;
=C2=A0 =C2=A0 using Simple =3D std::variant<int,d= ouble,std::string>;
=C2=A0 =C2=A0 using Numeric =3D std::= variant<int,double>;

=C2=A0 =C2=A0 void func= (Simple x);

=C2=A0 =C2=A0 Numeric n;
=C2= =A0 =C2=A0 func(n); =C2=A0// Should this work, somehow, since every Numeric= "is-a" Simple in some sense?

Co= uld it work? Yes. Should it work? Not implicitly; if it should be allowed, = there ought to be a function for it. When should it work? Well, that's = the question.
=C2=A0
<= div dir=3D"ltr">

=C2=A0 =C2=A0 Value v;
=C2=A0= =C2=A0 func(v); =C2=A0// Should this compile, somehow, and then throw at r= untime if v holds a List or Object?

No, it= should not. Not implicitly.

The "narrowing" case= is sort of like dynamic_cast;

In no sense= is it like a dynamic cast. `dynamic_cast` is about casting within a class = hierarchy. Specifically, from child to parent. The only way it's simila= r is that it can conceptually fail.

Or, is there a clever = trick to get this kind of behavior already, without having to write an arca= ne and expensive helper function that does a bunch of get()s and assignment= s?

There is nothing arcane about a simple = visitor:

template<type= name ...Args>
variant_conv
{
=C2=A0 template<typename T>
=C2=A0 variant<Args...> operator()(const<= /span> T <= span style=3D"color: #660;" class=3D"styled-by-prettify">&t) const
=C2=A0 {
=C2=A0 =C2=A0 //If T not in Args, throw.
=C2=A0 =C2=A0 return t;
=C2=A0 }
};

= auto v =3D visit(variant_conv<int, float, std::string= >{}, variant<int, float>{3.4f});

You can even wrap that up in a tem= plate function.

--

---
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>.
------=_Part_15_2090865752.1446160224012-- ------=_Part_14_1132156555.1446160224012-- .