GNU bug report logs - #34479
Incorrect classification of messages in compilation-mode buffer

Previous Next

Full log

View this message in rfc822 format

From: help-debbugs <at> gnu.org (GNU bug Tracking System)
To: Daniel Lopez <daniel.lopez999 <at> gmail.com>
Subject: bug#34479: closed (Re: bug#34479: Incorrect classification of
 messages in compilation-mode buffer)
Date: Fri, 15 Feb 2019 08:03:02 +0000

[Message part 1 (text/plain, inline)]

Your bug report

#34479: Incorrect classification of messages in compilation-mode buffer

which was filed against the emacs package, has been closed.

The explanation is attached below, along with your original report.
If you require more details, please reply to 34479 <at> debbugs.gnu.org.

-- 
34479: http://debbugs.gnu.org/cgi/bugreport.cgi?bug=34479
GNU Bug Tracking System
Contact help-debbugs <at> gnu.org with problems

[Message part 2 (message/rfc822, inline)]

From: Eli Zaretskii <eliz <at> gnu.org>
To: Daniel Lopez <daniel.lopez999 <at> gmail.com>
Cc: 34479-done <at> debbugs.gnu.org
Subject: Re: bug#34479: Incorrect classification of messages in
 compilation-mode buffer
Date: Fri, 15 Feb 2019 10:02:19 +0200

> From: Daniel Lopez <daniel.lopez999 <at> gmail.com>
> Date: Thu, 14 Feb 2019 16:05:10 +0000
> 
> Argh, I forgot to attach the patch. Here it is!

Thanks.  This is a relatively recent regression, so I pushed it to the
emacs-26 branch.

Please in the future include a ChangeLog-style commit log message with
your patches.

[Message part 3 (message/rfc822, inline)]

From: Daniel Lopez <daniel.lopez999 <at> gmail.com>
To: bug-gnu-emacs <at> gnu.org
Subject: Incorrect classification of messages in compilation-mode buffer
Date: Thu, 14 Feb 2019 16:00:08 +0000

[Message part 4 (text/plain, inline)]

Reproduce:

- Open the attached test program compilation_test.c in Emacs
- M-x compile
- Enter compile command "gcc -c compilation_test.c"

In the resulting *compilation* buffer, there should be a mixture of 
"error", "warning" and "note" messages. Trouble is they're all parsed as 
errors - hence they're all coloured in red, and querying any one's type 
by moving point onto it and evaluating "(compilation--message->type 
(get-text-property (point) 'compilation-message))" always returns 2, 
though warnings should return 1 and notes should return 0. Commands that 
jump between previous/next error also behave correspondingly wrongly.

Fix:

The attached fix_compilation_message_type.patch, applied to Git master, 
fixes it for me.

It seems like the old code was accidentally using the same variable for 
the TYPE field of the current rule from compilation-error-regexp-alist, 
and in an inner loop, the actual type that the current message from the 
compilation buffer was resolved to be, so the resolved type of the first 
message tends to get carried down into the following ones.

Daniel

[compilation_test.c (text/x-csrc, attachment)]

Previous Next

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