From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <zsh-workers-return-8573-mason-zsh=primenet.com.au@sunsite.auc.dk>
Received: (qmail 22614 invoked from network); 5 Nov 1999 15:48:37 -0000
Received: from sunsite.auc.dk (130.225.51.30)
  by ns1.primenet.com.au with SMTP; 5 Nov 1999 15:48:37 -0000
Received: (qmail 26751 invoked by alias); 5 Nov 1999 15:48:29 -0000
Mailing-List: contact zsh-workers-help@sunsite.auc.dk; run by ezmlm
Precedence: bulk
X-No-Archive: yes
X-Seq: 8573
Received: (qmail 26744 invoked from network); 5 Nov 1999 15:48:26 -0000
Date: Fri, 5 Nov 1999 16:48:21 +0100 (MET)
Message-Id: <199911051548.QAA11447@beta.informatik.hu-berlin.de>
From: Sven Wischnowsky <wischnow@informatik.hu-berlin.de>
To: zsh-workers@sunsite.auc.dk
In-reply-to: Zefram's message of Fri, 5 Nov 1999 15:20:20 +0000 (GMT)
Subject: Re: Bug with parameter expansion


Zefram wrote:

> Sven Wischnowsky wrote:
> >Tanaka Akira wrote:
> >> Maybe, `~' should be treated as `='?
> >
> >Yep.
> 
> Nope.  ~ in the middle of a word is special if EXTENDED_GLOB is set.

I was about to say:

I had exactly the same idea, had written the mail and the patch this
way and then checked. The `~' is only special if there are other glob
special characters. So I didn't change it.


... but of course, this isn't correct because bslashquote() doesn't
test if there are other special characters. Oh well...

Bye
 Sven

diff -u oldsrc/utils.c Src/utils.c
--- oldsrc/utils.c	Fri Nov  5 16:45:51 1999
+++ Src/utils.c	Fri Nov  5 16:46:19 1999
@@ -3075,8 +3075,9 @@
 	}
 	else if (ispecial(*u) &&
 		 ((*u != '=' && *u != '~') ||
-		  (u == s || (isset(MAGICEQUALSUBST) &&
-			      (u[-1] == '=' || u[-1] == ':')))) &&
+		  u == s ||
+		  (isset(MAGICEQUALSUBST) && (u[-1] == '=' || u[-1] == ':')) ||
+		  (*u == '~' && isset(EXTENDEDGLOB))) &&
 	    (!instring ||
 	     (isset(BANGHIST) && *u == (char)bangchar) ||
 	     (instring == 2 &&

--
Sven Wischnowsky                         wischnow@informatik.hu-berlin.de