220 4363 <4a382ec0-2050-4cfd-9ac6-e77e1445de31@isocpp.org> article Path: news.gmane.org!not-for-mail From: inkwizytoryankes@gmail.com Newsgroups: gmane.comp.lang.c++.isocpp.proposals Subject: Switch statmet support for variadic templates Date: Sat, 11 May 2013 07:40:13 -0700 (PDT) Lines: 423 Approved: news@gmane.org Message-ID: <4a382ec0-2050-4cfd-9ac6-e77e1445de31@isocpp.org> Reply-To: std-proposals@isocpp.org NNTP-Posting-Host: plane.gmane.org Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_289_14652115.1368283213408" X-Trace: ger.gmane.org 1368283216 3436 80.91.229.3 (11 May 2013 14:40:16 GMT) X-Complaints-To: usenet@ger.gmane.org NNTP-Posting-Date: Sat, 11 May 2013 14:40:16 +0000 (UTC) To: std-proposals@isocpp.org Original-X-From: std-proposals+bncBDDLTAGNTIBBBTVQXGGAKGQERBKL7WQ@isocpp.org Sat May 11 16:40:16 2013 Return-path: Envelope-to: gclcip-std-proposals@m.gmane.org Original-Received: from mail-vc0-f200.google.com ([209.85.220.200]) by plane.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1UbAy8-0002ZR-14 for gclcip-std-proposals@m.gmane.org; Sat, 11 May 2013 16:40:16 +0200 Original-Received: by mail-vc0-f200.google.com with SMTP id m16sf6171367vca.3 for ; Sat, 11 May 2013 07:40:15 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=x-received:x-beenthere:x-received:date:from:to:message-id:subject :mime-version:x-original-sender:reply-to:precedence:mailing-list :list-id:x-google-group-id:list-post:list-help:list-archive :list-subscribe:list-unsubscribe:content-type; bh=1yt8+16Xrr3MQKdahwaJDsdHXbCEDbtNGxmzWNYgOB4=; b=d36roFnKza69Q9YB8mv23Gi9VKk3PlQ0ZOGC0Fggbp4biSn7U81rggYxuHxwPKm6ul qPIq+/boU4sWOPXbeCWbK7xEi2Bqx5IPy841DOFCsnQC+6iJafy6f9ZbxZdPc+fta+UL YNVvS+ZNKf98sUK3EHu8MxcUyQJqJa5cbxt5rxy0nhMLg3iBohssbu4bDPFLD8do1S5J mQ9yKrkIb3xVjHnJaF2uW81rtx9oLLp8QOIOtlj4EI/bMuQPKjMVOxsfq9WndxSIkm1a Z/bSGKZHZ3MIrnYtXqXD2fCcStOGFn/pOtI/YHkXBSTDk/RyjnAVxh4WHfTs2PUaxkco RL0A== X-Received: by 10.224.10.6 with SMTP id n6mr14459899qan.4.1368283215161; Sat, 11 May 2013 07:40:15 -0700 (PDT) X-BeenThere: std-proposals@isocpp.org Original-Received: by 10.49.83.36 with SMTP id n4ls534364qey.76.gmail; Sat, 11 May 2013 07:40:13 -0700 (PDT) X-Received: by 10.49.1.197 with SMTP id 5mr1708347qeo.24.1368283213812; Sat, 11 May 2013 07:40:13 -0700 (PDT) X-Original-Sender: inkwizytoryankes@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:4363 Archived-At: ------=_Part_289_14652115.1368283213408 Content-Type: text/plain; charset=ISO-8859-1 In metaprograming use most of control structures using recursion. Thanks to inlinig this approach is effective, because it remove redundant calls leaving only program logic. But in case of switch its isn't true, because of flat nature of switch it can not be split to different function without degradation to equivalent if statement. This cause degradation of performance compare to hand written code. Right now C++11 have some workarounds but is have many limitations and is hard to use. template void func(int i) { constexpr Func f[] = {T::call...}; if(i < sizeof...(T)) f[i](); //cant be inlined } #define MACRO_CALL(x, f) case (x): get_param<(x), f...>::call(); break; //you need handle case where x >= szieof...(f) #define MACRO_CALL_2TIMES(x, f) MACRO_CALL((x+0), f) MACRO_CALL((x+1), f) #define MACRO_CALL_4TIMES(x, f) MACRO_CALL_2TIMES((x+0), f)MACRO_CALL_2TIMES ((x+2), f) /* etc. */ #define MACRO_CALL16TIMES(x, f) MACRO_CALL_8TIMES((x+0), f)MACRO_CALL_8TIMES ((x+8), f) template void func2(int i) { switch(i) //always 32 cases { MACRO_CALL16TIMES(0, T) MACRO_CALL16TIMES(15, T) //oops... } } I suggest to allow pack expansion in switch statement template void func3(int i) { switch(i) { case pos_param::value: { T::call(); break; }...; } } template void func4(int i) { switch(i) { case I: { select_func::call(); break; }...; } } template void func5() { switch(0) //from beging { case I: { select_func::call(); }...; //no break, call all function, not need to use recursion } } -- --- 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/?hl=en. ------=_Part_289_14652115.1368283213408 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable In metaprograming use most of control structures using recursion. Thanks to= inlinig this approach is effective, because it remove redundant calls leav= ing only program logic.
But in case of switch its isn't true, because of= flat nature of switch it can not be split to different function without de= gradation to equivalent if statement.
This cause degradation of performa= nce compare to hand written code.

Right now C++11 have some workarou= nds but is have many limitations and is hard to use.
template<typename... T>
void func(int i)
{
    constexpr Func f[] =3D {T::call...};
    if(i &= lt; sizeof...(T))
        f[i](); //cant be inlined
}

#define MACRO_CALL(x, f) case (x): get_pa= ram<(x), f...>::call(); break; //you n= eed handle case where x >=3D szieof...(f)
#define MACRO_CALL_2TIMES(x, f)<= /span> MACRO_CALL= ((<= span style=3D"color: #000;" class=3D"styled-by-prettify">x+0), f) MACRO_CALL((= x+<= span style=3D"color: #066;" class=3D"styled-by-prettify">1), f)
#define MACRO_CALL_4TIMES(x, f<= /span>) MACRO_CALL_2TIMES((x+0), f) MACRO_CALL_2TIMES((x+2), f)
/* etc. */
#define MACRO_CALL16TIMES(x, f) MACRO_CALL_8TIMES((x+0), f) = MACRO_CALL_8TIMES((x= +8), f)

template<typename... T>
<= /span>void= func2(int i)
{
    switch(i) //always 32 cases
   = ; {=
   =     MACRO_CALL16TIMES(0, T<= /span>)
    &n= bsp;   MACRO_CALL16TIMES(15, T) //oops...
    }
}

I suggest to allow <= /span>pack expansion in switch statement
<= div class=3D"prettyprint" style=3D"background-color: rgb(250, 250, 250); bo= rder-color: rgb(187, 187, 187); border-style: solid; border-width: 1px; wor= d-wrap: break-word;">
template<typename... T>
void func3(int i)
{
    switch(i)
    {
        case pos_param<T, T...>::value:<= /span> { T::call(); break; }...;
    <= /span>}
}
template<int... I>void func4<= span style=3D"color: #660;" class=3D"styled-by-prettify">(int i)
{
    switch(i)
  &nb= sp; {
  &nbs= p;     case I: { select_func<I>::call(); break; }...;
    <= span style=3D"color: #660;" class=3D"styled-by-prettify">}
}
template<int... I>= ;
void func5()
{
    switch(0) //f= rom beging=
    {<= br>        case I: <= /span>{ select_func<= span style=3D"color: #660;" class=3D"styled-by-prettify"><I>::call(); }...; //no break, call all function, not need to use recursion
    }
}

--
 
---
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-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-pro= posals/?hl=3Den.
 
 
------=_Part_289_14652115.1368283213408-- .