220 29836 <572a5592-ec0f-47f2-8253-aedb23ae29ce@isocpp.org> article Path: news.gmane.org!.POSTED!not-for-mail From: alexandre.isoard@gmail.com Newsgroups: gmane.comp.lang.c++.isocpp.proposals Subject: Looping construct with in-the-middle test: do { ... } while(...) { ... } Date: Fri, 16 Dec 2016 06:49:47 -0800 (PST) Lines: 366 Approved: news@gmane.org Message-ID: <572a5592-ec0f-47f2-8253-aedb23ae29ce@isocpp.org> Reply-To: std-proposals@isocpp.org NNTP-Posting-Host: blaine.gmane.org Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="----=_Part_501_553955834.1481899787455" X-Trace: blaine.gmane.org 1481899794 1112 195.159.176.226 (16 Dec 2016 14:49:54 GMT) X-Complaints-To: usenet@blaine.gmane.org NNTP-Posting-Date: Fri, 16 Dec 2016 14:49:54 +0000 (UTC) To: ISO C++ Standard - Future Proposals Original-X-From: std-proposals+bncBDXMXE5S7EMRBDH6Z7BAKGQEGOPVNVY@isocpp.org Fri Dec 16 15:49:50 2016 Return-path: Envelope-to: gclcip-std-proposals@m.gmane.org Original-Received: from mail-yb0-f197.google.com ([209.85.213.197]) by blaine.gmane.org with esmtp (Exim 4.84_2) (envelope-from ) id 1cHtpQ-0007ME-F2 for gclcip-std-proposals@m.gmane.org; Fri, 16 Dec 2016 15:49:44 +0100 Original-Received: by mail-yb0-f197.google.com with SMTP id h184sf107843699ybb.7 for ; Fri, 16 Dec 2016 06:49:49 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=isocpp-org.20150623.gappssmtp.com; s=20150623; h=date:from:to:message-id:subject:mime-version: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=ubvxV4TZaDYv+NfBjJCMBfbENa7r5vhbwDDx/wtOmG4=; b=mU5orLnem5UgLtdt3AxCFVgLXabF+3GAqS4wKo4QUjPbTBK1YvoLBleP8yJ5s9IcyB 4//04JVdP1hAJ5Huyvlq5opXjg04nVvV3fAjXoZwHuJVQDL04d0XiCfHbv2oKLqN4X5A g8S73e/f/ujB5kH2aD36PmMpcpZyKiACwjt5d2C1vTN5Ul5A/VboKmutkYFkS9470RW4 Wo7jNjmesHjvGaOIv7OdCJ746J+3xdb9AAEW6DLnz7Cd2ItOzGw9EsuT/qrR7qcYrgW8 nnXueRTpNK/exZgmUG5MJEJLIxfi/Egl68i5CbPOp9c80CSkShmm0fsR2TO40qgTJqei BH6w== DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=date:from:to:message-id:subject:mime-version: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=ubvxV4TZaDYv+NfBjJCMBfbENa7r5vhbwDDx/wtOmG4=; b=n7Ol1Vt8aCPTqC6ReFBmvvxk8oBPG1hyXHC1wM730dPcGBfGqUTcYe0OWdeS1jJixD L3ZoSyh8O05+SqXdXXUeU13BxuW0JRNhhAQtvBBTShPNwe8ijGZTWqlV73kPn/rzVJOQ ynKt3mI6JEghapCSPSVr44SaF6jjBqPNruna5B9fsEekZQOGMe+syhV0UODry86/6u2w QjC39J0R+K1I1KcCcjgTrjLiDvNpuQ3qQn+WOQFcjI5VDYvU1xRE5cdOor4//wMXb6e2 V/x9cj6slvgu8QKoHvHUGQ00mtlK70JbF8Fo7W+MZDJGK+YbBIXzBUSz3aqZxzHA75jz U79Q== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:date:from:to:message-id:subject:mime-version :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=ubvxV4TZaDYv+NfBjJCMBfbENa7r5vhbwDDx/wtOmG4=; b=aVT2VznJrZwkRyZiRqBwts+hZr5Tz4zDB3u+jyIDlUyiFEwdZLij+/ChRCO0opwPqs HxbqdhTcUZq9bDTJx0Nvb8qC2gwmOci2lKoBDSw6sFX+oZbVaEM4m+gwLSa7vLsLSPAh 8rnmRj10r0VW96u6Gy24sHwvr7CSN2YuHtz8P6dqJWHcR1QxHAkfwohyT9lVoO0U9RH3 WUANhf9Bn5YxDhmCBGMgzKuVf6Ssomql2f7wNquSAC+Qu5+FwyvuYV7LmsQGg8SBLyP+ uK24rPS07UwtpQ3aU/vOtW6GGzxIMlfqD37kWL8MnWaYi6oKNkCgekzXhnMmdsHSfbch LMOw== X-Gm-Message-State: AIkVDXJyhfDXGRyoRQtYNHWz09MiwBcY12JX5aPg/0xUJyCRgxHkQkTA4uPkXdXUZUuKVw== X-Received: by 10.129.26.213 with SMTP id a204mr853252ywa.111.1481899788652; Fri, 16 Dec 2016 06:49:48 -0800 (PST) X-BeenThere: std-proposals@isocpp.org Original-Received: by 10.157.1.164 with SMTP id e33ls8391281ote.43.gmail; Fri, 16 Dec 2016 06:49:47 -0800 (PST) X-Received: by 10.157.18.211 with SMTP id g77mr150252otg.14.1481899787834; Fri, 16 Dec 2016 06:49:47 -0800 (PST) X-Original-Sender: alexandre.isoard@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:29836 Archived-At: ------=_Part_501_553955834.1481899787455 Content-Type: multipart/alternative; boundary="----=_Part_502_973598154.1481899787456" ------=_Part_502_973598154.1481899787456 Content-Type: text/plain; charset=UTF-8 C++ provides two kind of while loop structures: while loop with at-beginning test: bool condition = test(); while (condition) { action(); condition = test(); } Useful when we want to perform the content of the loop as long as the test is true and we don't want to perform the content if the test is false at the beginning. while loop with at-end test: do { action(); bool condition = test(); } while (condition); Useful when we want to perform the content of the loop as long as the test is true but we still want to perform the content of the loop at least once even when the test is false at the beginning. Frequent when looking for fixpoint in an algorithm. We perform one iteration, see if it is sufficient, else continue. I propose an additional while loop structures that fit the syntax of the second form: do { bool condition = test(); } while (condition) { action(); } The semantic is as follow: 1. The do part is always executed when enter the loop (even if the test would be false) 2. We interpret the condition 3. If the test is false, we exit the loop structure (and therefore skip the while part) 4. If the test is true, we execute the while part *and *the do part and go back to point 2 As you may see, this third form is a generalization of the second form. We simply allow the mandatory ';' to be replaced with code. You may also remark that the third form has exactly the same behavior as the first form, except that the condition variable has a live-range restricted to the loop structure. An other use case: std::ostream &join(std::ostream &os, iterator it, iterator end, const std:: string &sep) { if (it != end) { os << *it; while(it != end) { os << sep; os << *it; } } return os; } Could be written: std::ostream &join(std::ostream &os, iterator it, iterator end, const std:: string &sep) { if (it != end) do { os << *it; } while(it != end) { os << sep; } return os; } There are many weird stuff with such a construct: - Potential syntax problems with LALR parsers. - We may want live-ranges of the variables declared in the do section to extend to the condition and the while section. - We may not want the live-ranges of the variables declared in the while section to extend to the do section and the condition. What do you think? -- 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. To view this discussion on the web visit https://groups.google.com/a/isocpp.org/d/msgid/std-proposals/572a5592-ec0f-47f2-8253-aedb23ae29ce%40isocpp.org. ------=_Part_502_973598154.1481899787456 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
C++ provides two kind of while loop structures:
=

while loop with at-beginning test:
bool con= dition =3D= test();
while (condition) {=C2=A0 action();=C2=A0 condition =3D test();<= /span>
= }
Useful when we want to perform the content of the loo= p as long as the test is true and we don't want to perform the content = if the test is false at the beginning.

while loop with = at-end test:
do {
=C2=A0 action();
=C2=A0 bool condition =3D t= est();
} while (condition);
Useful when we want t= o perform the content of the loop as long as the test is true but we still = want to perform the content of the loop at least once even when the test is= false at the beginning.
Frequent when looking for fixpoint in an= algorithm. We perform one iteration, see if it is sufficient, else continu= e.

I propose an additional while loop structures that fit the synt= ax of the second form:
do {
= =C2=A0 boo= l conditio= n =3D test();
} while (condition= ) <= span style=3D"color: #660;" class=3D"styled-by-prettify">{
=C2=A0 action();
}
The se= mantic is as follow:
  1. The do part is always executed when enter the loop (even if the te= st would be false)
  2. We interpret the condition
  3. If the test i= s false, we exit the loop structure (and therefore skip the while part)
  4. If the test is = true, we execute the while part and the do part and go back to point 2
As = you may see, this third form is a generalization of the second form. We sim= ply allow the mandatory ';' to be replaced with code.
You= may also remark that the third form has exactly the same behavior as the f= irst form, except that the condition = variable has a live-range restricted to the loop structure.

An other use case:
std::ostream &join= (std::ostream &os, iterator it, iterator end, const std::string &sep) {=
=C2=A0= if= (it !=3D end) {
=C2=A0 =C2= =A0 os <= ;< *it;
=C2=A0 =C2=A0 while(it !=3D end) {
=C2=A0 =C2=A0 =C2=A0 = os <<= ; <= font color=3D"#008800">sep;=C2=A0 =C2=A0 =C2=A0 os << *it;
=C2=A0 =C2=A0 }
=C2=A0 }
=C2=A0 return os;
}

Could b= e written:
std::ostrea= m &join(std::ostream &os, iterator it, iterator end, const std::string &sep) = {
=C2=A0 <= span style=3D"color: #008;" class=3D"styled-by-prettify">if (it !=3D end)
=C2= =A0 =C2=A0 do {
=C2=A0 =C2=A0 =C2=A0 = os <<= ; <= span style=3D"color: #660;" class=3D"styled-by-prettify">*it;
=C2=A0 =C2=A0 } while(it = !=3D end) {
=C2=A0 =C2=A0 =C2=A0 os << <= span style=3D"color: #000;" class=3D"styled-by-prettify">sep<= span style=3D"color: #660;" class=3D"styled-by-prettify">;
=C2=A0 =C2=A0 }
=C2=A0 return os;
}

There are many weird stuff= with such a construct:
  • Potential syntax problems with LA= LR parsers.
  • We may want live-ranges of the variables declared in th= e do section to extend to the condition and the while section.
  • We m= ay not want the live-ranges of the variables declared in the while section = to extend to the do section and the condition.
What do you th= ink?

--
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.
To view this discussion on the web visit https://groups.google.= com/a/isocpp.org/d/msgid/std-proposals/572a5592-ec0f-47f2-8253-aedb23ae29ce= %40isocpp.org.
------=_Part_502_973598154.1481899787456-- ------=_Part_501_553955834.1481899787455-- .