Mailing-List: contact java-dev-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: java-dev@lucene.apache.org
Received-SPF: pass (herse.apache.org: local policy)
Message-ID: <4564BBAD.6050601@mikemccandless.com>
Date: Wed, 22 Nov 2006 16:05:49 -0500
From: Michael McCandless <lucene@mikemccandless.com>
User-Agent: Thunderbird 1.5.0.8 (Windows/20061025)
MIME-Version: 1.0
To: java-dev@lucene.apache.org
Subject: Making RAMDirectory non final?
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Hi,

I'm working on a unit test for:

     http://issues.apache.org/jira/browse/LUCENE-702

which is the "disk full during addIndexes() can corrupt index" issue.

I think the simplest way to do this is to subclass RAMDirectory to
create a MockDiskFullRAMDirectory class (and a corresponding
MockDiskFullRAMOutputStream) that lets you set the max size and
then throws IOException to simulate disk full.

The recent commit for LUCENE-709 is wonderful because it incrementally
tracks total size so it's real simple to check :)

But to do this I'd need to make RAMDirectory non-final.  Any
objections to this?

If we do this we can also put other "test only" methods (eg the
"getRecomputedSizeInBytes()") into the same (or different) subclass,
over time.

Mike

---------------------------------------------------------------------
To unsubscribe, e-mail: java-dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-dev-help@lucene.apache.org