From: Alex Riesen <raa.lkml@gmail.com>
Date: Sun, 13 Jan 2008 23:33:23 GMT
Subject: Re: [PATCH] index: be careful when handling long names
Message-ID: <20080113233323.GB19970@steel.home>
In-Reply-To: <7v4pdhpbmw.fsf@gitster.siamese.dyndns.org>

Junio C Hamano, Mon, Jan 14, 2008 00:08:07 +0100:
> Alex Riesen <raa.lkml@gmail.com> writes:
> 
> > Junio C Hamano, Sun, Jan 13, 2008 23:36:34 +0100:
> >> +test_expect_success 'very long name in the index handled sanely' '
> >> +
> >> +	a=a && # 1
> >> +	a=$a$a$a$a$a$a$a$a$a$a$a$a$a$a$a$a && # 16
> >> +	a=$a$a$a$a$a$a$a$a$a$a$a$a$a$a$a$a && # 256
> >> +	a=$a$a$a$a$a$a$a$a$a$a$a$a$a$a$a$a && # 4096
> >
> > I'd expect it to fail on some systems (everywindowsthing up to w2k,
> > maybe some commercial unices).
> 
> My understanding is that Everywindowsthing do not come with any
> (POSIX compliant) shell that we support by default, so if you
> are talking about a limit of shell variable value, I do not
> think it is an issue to begin with.  It is just the matter of

Oh, right. The file system wont even see it, it is passed directly to
update-index.

> picking a sensible shell (I understand both Cygwin and msys
> ports use a shell that supports more than 4k bytes in value
> given to a variable).

can't check right now, but I believe it is so

> I would agree that it might overflow the argument limit when
> this is given to "echo", though.  We cannot do much about it,
> but you may have cleverer ideas.

I thought about conditionally disabling the test, like it was done
when the tabs in filenames had to be tested. Wont be needed for this
particular case.