I'm developing it for a book I'm writing on Ant, and I've posted one piece
of it here already - my HtmlDocument class that uses JTidy to DOM'ify the
HTML and rip out the title and body contents as two separate fields (without
HTML tags, of course).
I have every intention of giving all the code developed to Lucene or other
Jakarta projects where appropriate. I only haven't yet because its still
under development - its not top secret or anything. :) The Ant task
definitely deserves some additional Lucene expertise to make sure its doing
the right thing, but I have it checking dependencies by embedding a
non-indexed "last modified" field into the Lucene index too which it checks
before actually indexing a document again - so a second incremental run of
indexing is *much* faster since it skips files unless they are newer.
Erik
----- Original Message -----
From: "Andrew C. Oliver" <[EMAIL PROTECTED]>
To: "Lucene Developers List" <[EMAIL PROTECTED]>
Sent: Friday, February 08, 2002 8:19 AM
Subject: Re: Proposal for Lucene
> Is this open source? APL'd? Where can I look at it?
>
> On Thu, 2002-02-07 at 22:00, Erik Hatcher wrote:
> > I've developed something similar myself. I've created an Ant task
<index>
> > that uses DocumentHandler interface implementing classes - one that can
be
> > used (<index class="...">) is a FileExtensionDocumentHandler. At
build-time
> > I generate a Lucene index of static documents, and roll that into a web
> > application.
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
