From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 995 invoked from network); 21 Nov 1997 19:32:05 -0000 Received: from math.gatech.edu (list@130.207.146.50) by ns1.primenet.com.au with SMTP; 21 Nov 1997 19:32:05 -0000 Received: (from list@localhost) by math.gatech.edu (8.8.5/8.8.5) id OAA23265; Fri, 21 Nov 1997 14:25:00 -0500 (EST) Resent-Date: Fri, 21 Nov 1997 14:23:27 -0500 (EST) Message-ID: <19971121142356.16832@astaroth.nit.gwu.edu> Date: Fri, 21 Nov 1997 14:23:56 -0500 From: Sweth Chandramouli To: zsh-users@math.gatech.edu Subject: Re: path/PATH function References: <19971119103756.34168@astaroth.nit.gwu.edu> <971119082751.ZM574@candle.brasslantern.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.88e In-Reply-To: <971119082751.ZM574@candle.brasslantern.com>; from Bart Schaefer on Wed, Nov 19, 1997 at 08:27:51AM -0800 Resent-Message-ID: <"S_K-k2.0.ug5.k-TTq"@math> Resent-From: zsh-users@math.gatech.edu X-Mailing-List: archive/latest/1147 X-Loop: zsh-users@math.gatech.edu X-Loop: zsh-workers@math.gatech.edu Precedence: list Resent-Sender: zsh-workers-request@math.gatech.edu On Wed, Nov 19, 1997 at 08:27:51AM -0800, Bart Schaefer wrote: > This "feature" is restricted at present to a few pairs of variables that are > hardwired at compile time. They include: > > CDPATH cdpath > FPATH fpath > MAILPATH mailpath > MANPATH manpath > PATH path maybe the documentation i have isn't that complete, but all i could find in the manpage for zshparam is the fact that the variables like cdpath and path exist and are "special" varaibles; there isn't any mention of how they are mapped on to their capitalized counterparts. my next question, then, is if these do, in fact, work the way i surmised (that is, they remove redundancies and map their colon-delimited versions onto the capitalized equivalent variable). also, what exactly is the format for defining them? is it just paren-enclosed, space-delimited lists? > You might also wish to look into `typeset -U` which causes an array to > remove all non-unique elements from itself. A function to test whether > any given directory is already in your path shouldn't be necessary. the version of typeset on my system doesn't seem to support this option (or at least, the man page doesn't mention it). typeset doesn't return an error if i try to use the -U switch, however; could someone summarize the syntax so i could see if it is, in fact, implemented? tia, sweth.