If you look back at the thread, you'll see that Diane's solution simulates the use of a nested <mapper> in <delete>, because <mapper> is currently not supported for this task. You'll also see that the <present> selector selects the wrong files, and that it cannot be made to select the right ones (AFAIK), thus the need for the second <mapper> in <delete>.
The current turn around of using <PathConvert> with the caveat of pathname transformation looks like a hack to me. OTOH, I did just now realize that a <mapper> doesn't make sense for a task that takes a single pathname, as opposed to <copy> which takes two... So it appears it might be a deficiency of the selectors not being able to select the right files!?!? Did we miss something Bruce? Thanks, --DD -----Original Message----- From: Stefan Bodewig [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 9:01 AM To: [EMAIL PROTECTED] Subject: Re: reverse glob mapper pattern? On Mon, 23 Sep 2002, Dominique Devienne <[EMAIL PROTECTED]> wrote: > Shouldn't delete simply take a <mapper>, as does <copy> I don't think so, what would you use it for? Isn't the <present> selector (which accepts a <mapper>) all that is needed (sorry, I haven't followed the thread very closely)? Stefan -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]> -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
