Mailing-List: contact user-help@excalibur.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Excalibur Users List" <user@excalibur.apache.org>
Received-SPF: pass (hermes.apache.org: domain of pchristmann@gmail.com
 designates 64.233.170.199 as permitted sender)
DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
        s=beta; d=gmail.com;
        h=received:message-id:date:from:reply-to:to:subject:mime-version:content-type:content-transfer-encoding;
        b=cEDr5auV8380RuDibcACMTYHvbt1y0sD/D3tKfETH401jgOzfcRFblIXriCJY6WCmgmiw6n9x5obcwBnwFPJRvl8pAd4DqF7ijwY8RTF5FoDyK8ZareN/3FCALhy7miHkl02pJE4kblxlomQYxPRok+JHkgGS/+ARBIufIdgSQA=
Message-ID: <db3371b20501041303691d73d0@mail.gmail.com>
Date: Tue, 4 Jan 2005 15:03:18 -0600
From: Paul Christmann <pchristmann@gmail.com>
Reply-To: paul@priorartisans.com
To: Excalibur Users List <user@excalibur.apache.org>
Subject: URLSource bug?
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

I've been digging through code in Cocoon and Excalibur, trying to
decide if there is a bug in either that is keeping me from deploying a
cocoon web application as a WAR.

(I've posted details of my problem on the cocoon users mailing list. 
See http://marc.theaimsgroup.com/?l=xml-cocoon-users&m=110418663626644&w=2
if you're interested.  But in short: deploying as a WAR fails,
deploying as an expanded folder works).

It seems that the problem is with the
org.apache.excalibur.source.impl.URLSource: exists() returns true even
if the URL provided to the init method does not exist.  In cocoon, the
URL looks something like "jndi:/localhost/webapp/com.java".  (Why such
a weird URL?  For the life of me, I can't figure it out: something
with how rhino is interpreting a javascript file with an "importClass"
entry).

But - the URLSource constructed with that URL happily says it exists,
which causes a cocoon error when it tries to access it.

I've written two tests (I didn't see a URLSourceTestCase anywhere?):
one with an invalid http URL (http://localhost/garbage) and one with a
valid http url (http://www.example.com), and the exists method returns
true for both.

I've got a simple one line fix for URLSource that seems to do the
trick: set m_exists based on the returned contentLength.

But - before I submit to bugzilla (the patch and the test case), I
thought I'd check first: am I misunderstanding how the URLSource
should work, and will I be breaking other things if I make exists()
return false?
-- 
PC

Paul Christmann
Prior Artisans, LLC
mailto:paul@priorartisans.com
504-587-9072

---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@excalibur.apache.org
For additional commands, e-mail: user-help@excalibur.apache.org