[jdom-interest] Feature Request

[Chris B.]

Elliotte Rusty Harold wrote: At 1:40 PM +1100 2/20/04, Chris B. wrote:

> Correct me if I'm wrong, but wouldn't it have been better if NekoHTML 
> wasn't written to conform to the non-standard Xerces class, but rather 
> if it implemented an XMLReader?



 >Probably, but that would have required writing it from scratch without 
making use of the existing Xerces infrastructre.

I wrote to the Neko author and this is what he said....




-------- Original Message --------
Subject: 	Re: NekoHTML
Date: 	Thu, 19 Feb 2004 19:28:00 -0800
From: 	Andy Clark <andyc at apache.org>
To: 	Chris B. <chris at tech.com.au>
References: 	<403557E6.4090703 at tech.com.au>



Chris B. wrote:
> I'm not a true expert of the internals of SAX, but from my 
> understanding, wouldn't the Right Thing have been for NekoHTML to have 
> provided an implementation of XMLReader, which could then be plugged 
> into other scenarios? (e.g. JDOM)??

It already does. The org.cyberneko.html.parsers.SAXParser
implements XMLReader so you can use it to drive anything
that is built on SAX. But by implementing it on XNI, it
can be used to drive existing and future Xerces parsers,
as well as borrowing Xerces components for use within its
parsing pipeline.

-- 
Andy Clark * andyc at apache.org