<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 12pt;
font-family:Calibri
}
--></style></head>
<body class='hmmessage'><div dir='ltr'><pre style="line-height: 21px; white-space: normal; color: rgb(68, 68, 68); font-size: 15px;">Hi David,<br>I need to add another one boolean parameter for OpenMP pragmas parsing <br>to skip counting braces/brackets/parens and Hal Finkel proposed to <br>convert all these booleans into a bit-based single parameter.<br><br>Best regards,<br>Alexey Bataev<br>=============<br>Software Engineer<br>Intel Compiler Team<br>Intel Corp.</pre><br><div>> Date: Wed, 6 Nov 2013 14:24:26 +0400<br>> From: a.bataev@hotmail.com<br>> To: reviews+D2108+public+7ef5c38906feb203@llvm-reviews.chandlerc.com; dgregor@apple.com; cbergstrom@pathscale.com; hfinkel@anl.gov<br>> CC: cfe-commits@cs.uiuc.edu; david.majnemer@gmail.com<br>> Subject: Re: [PATCH] Replaced bool parameters in SkipUntil function with single bit-based parameter.<br>> <br>> Hi David,<br>> I need to add another one boolean parameter for OpenMP pragmas parsing <br>> to skip counting braces/brackets/parens and Hal Finkel proposed to <br>> convert all these booleans into a bit-based single parameter.<br>> <br>> Best regards,<br>> Alexey Bataev<br>> =============<br>> Software Engineer<br>> Intel Compiler Team<br>> Intel Corp.<br>> <br>> 06.11.2013 14:10, David Majnemer пишет:<br>> > From my experience,<br>> > `SkipUntil(EndKind, /*StopAtSemi=*/true, /*DontConsume=*/true);`<br>> > is more typical in clang than<br>> > `SkipUntil(EndKind, StopAtSemi | DontConsume);`<br>> ><br>> > It seems that some of the calls that you changed were previously nasty (i.e. `SkipUntil(EndKind, true, true)`) which justifies a cleanup.<br>> > However I'm not sure we want a bitfield here.<br>> ><br>> > What is your justification?<br>> ><br>> ><br>> > ================<br>> > Comment at: include/clang/Parse/Parser.h:751<br>> > @@ -741,3 +750,3 @@<br>> > /// token will ever occur, this skips to the next token, or to some likely<br>> > - /// good stopping point. If StopAtSemi is true, skipping will stop at a ';'<br>> > - /// character.<br>> > + /// good stopping point. If Flags has bit set at StopAtSemi, skipping will<br>> > + /// stop at a ';' character.<br>> > ----------------<br>> > The wording "has bit set" seems strange.<br>> ><br>> ><br>> > <a href="http://llvm-reviews.chandlerc.com/D2108" target="_blank" rel="noreferrer" style="cursor:help;display:inline !important;">http://llvm-reviews.chandlerc.com/D2108</a><br>> <br></div> </div></body>
</html>