GNU bug report logs - #29959
cc-mode: wrong indentation in absence of semicolon

Previous Next

Packages: cc-mode, emacs;

Reported by: Konstantin Kharlamov <hi-angel <at> yandex.ru>

Date: Wed, 3 Jan 2018 06:03:02 UTC

Severity: minor

Tags: fixed

Done: Noam Postavsky <npostavs <at> users.sourceforge.net>

Bug is archived. No further changes may be made.

Full log


Message #20 received at 29959 <at> debbugs.gnu.org (full text, mbox):

From: Konstantin Kharlamov <hi-angel <at> yandex.ru>
To: Alan Mackenzie <acm <at> muc.de>
Cc: 29959 <at> debbugs.gnu.org
Subject: Re: bug#29959: cc-mode: wrong indentation in absence of semicolon
Date: Thu, 18 Jan 2018 22:19:35 +0300
Sorry, I mean, not cool that it doesn't work, but cool there's a workaround.

On side note — though I am not experienced in elisp — I think it'd be 
nice to add a comment in the removed code that it's looking for C99 
compound literals. Otherwise I personally struggle to understand what's 
being looked up there. In fact, whole file doesn't mention sentence "C99 
compound literals" anywhere.

On 18.01.2018 22:08, Konstantin Kharlamov wrote:
> Cool, thanks, I applied it locally, works for me!
> 
> On 18.01.2018 21:41, Alan Mackenzie wrote:
>> Hello again, Konstantin.
>>
>> On Sat, Jan 13, 2018 at 20:37:52 +0000, Alan Mackenzie wrote:
>>> On Sat, Jan 13, 2018 at 14:11:31 +0300, Konstantin Kharlamov wrote:
>>>> Just tested with current git, the problem is still present.
>>
>>>> I am almost sure this is a regression, I don't remember having that
>>>> problem before. The real-world consequences are just too annoying 
>>>> for it
>>>> to go unnoticed — basically, every time I'm typing variables 
>>>> indentation
>>>> just doesn't work. And I also seem to remember a few irrelevant
>>>> situations where I was typing conditions with the like results.
>>
>>> The problem originates in a new feature, C99 compound literals, which 
>>> was
>>> committed on 2017-11-10.  This detects brace blocks (meaning something
>>> like the initialisation of a struct) in the middle of executable code,
>>> but it is clear the test for such a brace block isn't rigorous enough,
>>> since even function blocks are being mistaken for them.
>>
>> The new feature, C99 compound literals, has been removed from the
>> emacs-26 branch, so as to avoid delaying the relase process for that
>> branch.  The commit has hash 36edb6cb97ce3d53543c87643077d270bb5bdfd1,
>> and it should apply without problems to the master branch.
>>
>> It is to be hoped that a more elaborate and more correct solution will
>> be found soon for the master branch.
>>
>>>> On 03.01.2018 09:02, Konstantin Kharlamov wrote:
>>>>> # Steps to reproduce:
>>>>> 1. Open emacs -Q test.c
>>>>> 2. Write the following:
>>
>>>>> int main() {
>>>>>       int a,
>>>>> }
>>
>>>>> 3. put the caret after the comma symbol, and press enter, e.g. to
>>>>> continue writing variables on the new line.
>>
>>>>> # Expected:
>>>>> the new line alignment stands out with regard to the prev. line.
>>
>>>>> # Actual:
>>>>> the new line aligned to the beginning of the prev. line.
>>
>>>>> # Workarounds:
>>>>> Typing the following text:
>>
>>>>> int main() {
>>>>>       int a,;
>>>>> }
>>
>>>>> …then pressing enter right after the comma works as expected.
>>
>>>>> ------------
>>
>>>>> In GNU Emacs 27.0.50 (build 1, x86_64-pc-linux-gnu, GTK+ Version 
>>>>> 3.22.26)
>>>>>    of 2017-12-15 built on constantine-N61Ja
>>>>> Repository revision: 6c301afa70f6eac32ad1ce92412ea3cf6fcdeeca
>>>>> Windowing system distributor 'The X.Org Foundation', version 
>>>>> 11.0.11906000
>>>>> System Description: Arch Linux
>>
>>> [ .... ]
>>




This bug report was last modified 7 years and 109 days ago.

Previous Next


GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.