Joerg Heinicke wrote:
On 10.10.2007 4:45 Uhr, Leszek Gawron wrote:
1. Can somebody tell me why do we need all these:
protected static final ContentHandler EMPTY_CONTENT_HANDLER = new
DefaultHandler();
/** The <code>XMLConsumer</code> receiving SAX events. */
protected XMLConsumer xmlConsumer;
/** The <code>ContentHandler</code> receiving SAX events. */
protected ContentHandler contentHandler = EMPTY_CONTENT_HANDLER;
/** The <code>LexicalHandler</code> receiving SAX events. */
protected LexicalHandler lexicalHandler =
DefaultLexicalHandler.NULL_HANDLER;
in AbstractXMLProducer?
I might understand why the need for separate lexicalHandler and
contentHandler. Still why the need for the default value? Why aren't
they simply null?
I'd guess exactly for NOT being null and avoiding NPEs.
/**
* Set the <code>XMLConsumer</code> that will receive XML data.
* <br>
* This method will simply call <code>setContentHandler(consumer)</code>
* and <code>setLexicalHandler(consumer)</code>.
*/
public void setConsumer(XMLConsumer consumer) {
this.xmlConsumer = consumer;
setContentHandler(consumer);
setLexicalHandler(consumer);
}
It looks like we are fulfilling some really old dependencies here...
What's your point? Are you concerned that consumer is used 3-times?
That's probably because some components don't want to have it set
directly but implement the setters and wrap the consumer with some filter.
I am just asking what is the point of such implementation and the answer
clears all. Thanks.
2. Look at SitemapModelComponent:
void setup(SourceResolver resolver, Map objectModel, String src,
Parameters par) throws ProcessingException, SAXException,
IOException;
}
It looks like we could drop the source resolver and objectModel. Both
can be injected and AFAIU both are properly scoped with some custom
scope of ours.
But that's an interface and the method is probably called by the
pipeline. You still have to support both Avalon and Spring and need at
least a migration path in the pipeline.
Of course not now but that is just a pointer that spring invalidates
somehow this interface.
--
Leszek Gawron http://www.mobilebox.pl/krs.html
CTO at MobileBox Ltd.
