Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1N2W9q-0004Pi-7X for qemu-devel@nongnu.org; Mon, 26 Oct 2009 16:27:14 -0400 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1N2W9p-0004PI-Bx for qemu-devel@nongnu.org; Mon, 26 Oct 2009 16:27:13 -0400 Received: from [199.232.76.173] (port=36625 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1N2W9p-0004PD-6h for qemu-devel@nongnu.org; Mon, 26 Oct 2009 16:27:13 -0400 Received: from hall.aurel32.net ([88.191.82.174]:51025) by monty-python.gnu.org with esmtps (TLS-1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.60) (envelope-from ) id 1N2W9o-0003r5-GJ for qemu-devel@nongnu.org; Mon, 26 Oct 2009 16:27:13 -0400 Date: Mon, 26 Oct 2009 21:27:07 +0100 From: Aurelien Jarno Subject: Re: [Qemu-devel] [PATCH] CODING_STYLE: don't allow non-indented statements after if/else blocks Message-ID: <20091026202707.GR26129@hall.aurel32.net> References: <20091026062637.GA21562@volta.aurel32.net> <20091026200326.GA9669@volta.aurel32.net> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-15 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: Sender: Aurelien Jarno List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Blue Swirl Cc: qemu-devel@nongnu.org On Mon, Oct 26, 2009 at 10:20:34PM +0200, Blue Swirl wrote: > On Mon, Oct 26, 2009 at 10:03 PM, Aurelien Jarno wrote: > > On Mon, Oct 26, 2009 at 06:02:52PM +0200, Blue Swirl wrote: > >> On Mon, Oct 26, 2009 at 8:26 AM, Aurelien Jarno wrote: > >> > Rationale: The following code is difficult to read, but allowed by the > >> > current coding style. > >> > >> Fully agree. > >> > >> > +Every control flow statement is followed by a new indented and braced > >> > +block; even if the block contains just one statement.  The opening brace > >> > +is on the line that contains the control flow statement that introduces > >> > +the new block; the closing brace is on the same line as the else keyword, > >> > +or on a line by itself if there is no else keyword.  Example: > >> > >> I think an exception should be granted for "else if" case, otherwise > >> the style would require braces around "if", like: > >>     if (a == 5) { > >>         printf("a was 5.\n"); > >>     } else { > >>         if (a == 6) { > >>             printf("a was 6.\n"); > >>         } > >>     } else { > >>         printf("a was something else entirely.\n"); > >>     } > >> > >> Picking nits: "while" is a control flow statement, even in "do {} > >> while" statement and then it would illegal to require a braced block > >> after the "while" statement. > > > > Good point. Please find another try below: > > > > From: Aurelien Jarno > > > > Rationale: The following code is difficult to read: > > > >    if (a == 5) printf("a was 5.\n"); > >    else if (a == 6) printf("a was 6.\n"); > >    else printf("a was something else entirely.\n"); > > > > Signed-off-by: Aurelien Jarno > > --- > >  CODING_STYLE |   12 +++++++----- > >  1 files changed, 7 insertions(+), 5 deletions(-) > > > > diff --git a/CODING_STYLE b/CODING_STYLE > > index a579cb1..c17c3f3 100644 > > --- a/CODING_STYLE > > +++ b/CODING_STYLE > > @@ -51,11 +51,13 @@ QEMU coding style. > > > >  4. Block structure > > > > -Every indented statement is braced; even if the block contains just one > > -statement.  The opening brace is on the line that contains the control > > -flow statement that introduces the new block; the closing brace is on the > > -same line as the else keyword, or on a line by itself if there is no else > > -keyword.  Example: > > +Every control flow statement is followed by a new indented and braced > > +block, except if it is followed by another control flow statement (else > > +if) or by a condition (do {} while ()); even if the block contains just > > +one statement.  The opening brace is on the line that contains the > > +control flow statement that introduces the new block; the closing > > +brace is on the same line as the else keyword, or on a line by itself > > +if there is no else keyword.  Example: > > Nice try, but does it prevent this: > if (x) for (;;) do { > } while (0); > ? We should find a way to define "for (;;)" as a single statement. > Maybe also "break" and "goto" can be considered control flow statements. Correct, they should be excluded from there as they don't need to be followed by braces. > How about something like "wherever C syntax allows potentially > ambiguous sequence of statements, braces must be used, with the > exception of 'else' followed by 'if'"? Now the problem becomes > defining ambiguous sequences of statements. That's the problem. We have seen that people already take advantage of the ambiguities, as I would have never have imagined someone writing the code in the rationale of this patch to avoid putting braces. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurelien@aurel32.net http://www.aurel32.net