Date:	Tue, 8 May 2001 10:47:31 +0300 (IDT)
From:	Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender:	eliz AT is
To:	"Mark E." <snowball3 AT bigfoot DOT com>
cc:	djgpp-workers AT delorie DOT com
Subject:	Re: snprintf?
In-Reply-To:	<3AF7358B.19349.38937B@localhost>
Message-ID:	<Pine.SUN.3.91.1010508104711.3965T-100000@is>
MIME-Version:	1.0
Reply-To:	djgpp-workers AT delorie DOT com
Errors-To:	nobody AT delorie DOT com
X-Mailing-List:	djgpp-workers AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

On Mon, 7 May 2001, Mark E. wrote:

> The last snprintf diff is at:
> http://www.delorie.com/djgpp/mail-archives/browse.cgi?p=djgpp-workers/2000/11/18/19:19:56
> 
> it sounds like it takes care of most of the objections raised:
> 
> > * Correct texinfo, as directed by Eli.
> > 
> > * Extended tests, as suggested by Eli. I'm not sure how to induce a
> > encoding error in printf() though. Formats like '%!' result in '!' as
> > output. I've renamed the test t-snprtf.c from snprintf.c, for debugging
> > with gdb.
> > 
> > * snprintf() is now just a wrapper for vsnprintf().
> > 
> > * n == 0 handled properly now. The buffer pointer in FILE is set to NULL
> > in this case, to catch bogus buffer accesses.

This last issue is what bothered me.  It seemed to me that the code
didn't cope well with this (setting a pointer to NULL is looking for
SIGSEGV's ;-), and IIRC the test harness didn't test all of the
possible cases (I forget the details).  The case of n == 0 is very
important for snprintf, so it must be rock-solid.

- Raw text -