gnu-arch-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gnu-arch-users] ANNOUNCE: Spaces in filenames, finished


From: John Meacham
Subject: Re: [Gnu-arch-users] ANNOUNCE: Spaces in filenames, finished
Date: Thu, 18 Mar 2004 15:04:43 -0800
User-agent: Mutt/1.4.2i

On Thu, Mar 18, 2004 at 12:23:07PM +0100, address@hidden wrote:
> > Would any other rare character be better for this purpose? Some
> > candidates are ':', ';', '?', '%', '#', '^', '*'. All characters
> > probably have one or another problem, but IMHO the problem with
> > backslash is inherently higher. One more thing to think about is this
> > proprietary OS uses backslash as the directory delimiter.
> 
> If '\' is the escape character and the tla uses it on some interface
> seen by other tools, the tools need to be aware of escaping.
> 
> If '%' is the escape character and the tla uses it on some interface
> seen by other tools, the tools need to be aware of escaping.
> 
> If '#' is the escape character and the tla uses it on some interface
> seen by other tools, the tools need to be aware of escaping.
> 
> ....
> 
> Do u see any difference?
> 
> I favoured shell/C escaping because shell can handle it natively, Tom
> wanted pika escaping. But it really makes no difference even if tla
> would use shell compatible escaping, tools would need to be aware of it
> to do the transformations at the right place. There is really no sense
> in discussing about, because any sane escaping scheme will work and
> tools need to be aware of escaping no matter which scheme is used.

All I would like to see is null terminated. like -print0 in find. it is
completly unambiguous and most tools can be made to support it nativly.
(or xargs can be used for such a purpose).

as for quoting, I like rc-style. single quotes quote anything and
doubled up single quotes stands for itself. A variation is to use '\''
to stand for a single quote in a quoted as that is backwards compatable
with 'sh' style shells.

In any case, all this is mainly a presentation issue as any quoting
mechanism will have problems, the correct answer is to allow null
termination for the writing of robust tools.

        John

-- 
---------------------------------------------------------------------------
John Meacham - California Institute of Technology, Alum. - address@hidden
---------------------------------------------------------------------------




reply via email to

[Prev in Thread] Current Thread [Next in Thread]