On Thu, Feb 05, 2004 at 04:32:26PM +0100, Paul Bijnens wrote:
> stan wrote:
> 
> >On Thu, Feb 05, 2004 at 02:16:09PM +0100, Paul Bijnens wrote:
> >>stan wrote:
> >>
> >>>>>/foo/bar/[A-J]*
> >>>>>/foo/bar/[K-z]*
> >>
> >>Don't mix uppercase and lowercase in the character maps.
> >>It depends on the current LANG and/or LC_* settings what they
> >>expand to.  Probably not what you want.
> 
> >Is there a way to this that IS safe given the locale stuff?
> 
> 
> Stay in a logical range:
> 
> is ok:  [0-9]  [A-Z]  [a-z]  [!a-z]
>    (if you ever used a EBCDIC character set in unix, then even those
> last three were not safe, and included two extra ranges with braces
> etc; no sane person uses EBCDIC anymore today -- I did, but I'm not
> sane.)
> 
> is not ok:  [A-z]  [a-Z]
>   (it's not a syntax error, but it does not what you expect it to do!)
> 
> The "all except" is encoded by an exclamation mark.
> 
> Assume this directory structure:
> 
>    /d/
>    /d/l
>    /d/l/e
>    /d/d/d
> 
> then, this disklist entry does NOT backup /d/d/....:
> 
>    host /d /d {
>       include "./l"
>       }
> 
> so you don't need to an explicit exclude for it.

OK, I think I understand. If I do then the follwing should do what I
intended the original to do:

-- first set ---
include "./[A-K]*"
exclude "./[!A-K]*"

-- second set ---
include "./[!A-K]*"

Or, do I need an exclude on the 2nd set?

What I think this does is abckup _all_ files. Set on gets them if the satrt
with a capital letter in the range A _ >K, and the 2nd set gets all the _do
not_ fit this pattern. Hmm, writing that makes ne realize just how much
better regexs are at expressing these things that English :-)

Thanks for educating me here.

-- 
"They that would give up essential liberty for temporary safety deserve
neither liberty nor safety."
                                                -- Benjamin Franklin