Mailing-List: contact ant-dev-help@jakarta.apache.org; run by ezmlm
To: ant-dev@jakarta.apache.org
Subject: Re: Path.java update
References: <393EA22B.CFD7BCDA@haas-wyss.ch> <m3ya4gh1jn.fsf@sbodewig.bost.de>
	<393F8289.D01993C@softwired-inc.com> <m3pupsgzpb.fsf@sbodewig.bost.de>
	<393F8EE0.91F523F2@softwired-inc.com>
From: Stefan Bodewig <bodewig@bost.de>
Date: 08 Jun 2000 14:49:59 +0200
In-Reply-To: Thomas Haas's message of "Thu, 08 Jun 2000 14:17:37 +0200"
Message-ID: <m3d7lsfibs.fsf@sbodewig.bost.de>
Lines: 18
User-Agent: Gnus/5.0807 (Gnus v5.8.7) XEmacs/21.1 (Canyonlands)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii

>>>>> "TH" == Thomas Haas <thomas.haas@softwired-inc.com> writes:

 TH> However I may be pleased by doing it the other way round: let
 TH> translatePath first split the path into single elements and then
 TH> call translateFile for each element. What do you think?

Of course, this is the cleanest solution.

But this would mean you'd need to scan the full path twice, first to
split the path (and do the DOS check to detect ':') and then once
again. I feel the primary intention of Project.translatePath was to
avoid this - else you could do the / \ replacement with two
String.replace calls instead of the loop.

Personally I don't think you would notice the difference - how many
Path's do we need to split?

Stefan