From:	sandmann AT clio DOT rice DOT edu (Charles Sandmann)
Message-Id:	<9606061324.AA17918@clio.rice.edu>
Subject:	Re: Downcasing pathnames under LFN
To:	eliz AT is DOT elta DOT co DOT il (Eli Zaretskii)
Date:	Thu, 6 Jun 1996 08:24:28 -0600 (CDT)
Cc:	djgpp-workers AT delorie DOT com
In-Reply-To:	<Pine.SUN.3.91.960606130954.28496C-100000@is> from "Eli Zaretskii" at Jun 6, 96 01:19:25 pm

> I think that under LFN, the default should be to not automatically 
> downcase pathnames we get from the OS, as people seem to want the case to 
> be preserved.  If I'm correct, then the following library functions 
> should be fixed:
> 
> 	getcwd, _file_tree_walk, srchpath, readdir, 
> 	__crt0_setup_arguments

Then you see real uglyless with the all upper case 8.3 names.  What I was
going to do before I got sidetracked is to suppress the conversion if 
the original name contained any lower case chars.  I planned to have this
be a single routine called by all of the above locations.

> Maybe the handling of _CRT0_FLAG_PRESERVE_UPPER_CASE should be also 
> different under LFN?
> 
> `glob' looks at the pattern and, if all the characters there are 
> lower-case, decides to match case-insensitively.  I wonder if that is 
> correct for case-preserving OSes.

You would need to ask for volume information for each device to see if case
is significant.  Right now, it's not on anything I have tested.  It makes
it a bunch more complicated, so I would leave this one until someone with
a test system wants to work it and sees a problem.

- Raw text -