Date: Tue, 11 Feb 2003 07:45:30 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: Richard Dawe <rich AT phekda DOT freeserve DOT co DOT uk>
cc: djgpp-workers AT delorie DOT com
Subject: Re: Take on __solve_symlinks()
In-Reply-To: <3E4830C4.6AD557DB@phekda.freeserve.co.uk>
Message-ID: <Pine.SUN.3.91.1030211074233.25449A-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
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
Precedence: bulk


On Mon, 10 Feb 2003, Richard Dawe wrote:

> It still doesn't work for the cases where ".."s go above the
> root directory. So either findfirst (& findnext, I guess) or __solve_symlinks
> needs to be fixed. But does the fix belong in findfirst or __solve_symlinks?
> 
> We could "fix" findfirst & findnext to cope with excess ".."s on DOS.

IMHO, the fix should be in __solve_symlinks.  findfirst/findnext are 
more-or-less direct interfaces to DOS system calls, with no simple 
equivalent on Posix systems, and DOS barfs when there are excess ".." in 
the file name.  The feature whereby "/.." is equivalent to "/" is a Posix 
feature, so we should emulate that in the Posix layer.  Pure DOS 
functions such as findfirst should behave like DOS does.