Craig R. McClanahan wrote:
>
> On Sat, 12 May 2001, Scott Sanders wrote:
>
>
>> Hello all,
>>
>> I have updated Digester to use the SAX2 interfaces, but I have a few
>> questions.
>>
>> Basically, when we have a start element, should Digester match on
>> localName (just the element name, no namespace) or match on qName (ns
>> prefix and element name)?
>>
>
>
> If you turn namespace awareness off, you're matching on
> "prefix:element". It would be consistent to make the same policy for
> namespace-aware matches, although that means the Digester author has to
> know what prefix the document developer used.
>
> Maybe we need another boolean configuration variable to let the Digester
> author choose.
That is something that we have to think about. For now it just matches
on the element name.
>
>
>> Or, should we match on the element name only for now, as that is what
>> Digester is doing, and find some way to better include namespace
>> processing into the digester framework, using the namespaceURI?
>>
>> Anyway, I am committing the SAX2 changes that I have, that do exactly
>> what Digester is doing now, but eliminate the deprecation warnings. I
>> think that we should go beyond this later, though.
>>
>
>
> After these changes, does Digester remain compatible with JAXP/1.0 APIs?
> That is important for Struts users who will want to be able to plug in
> whatever parser they want.
No, it does not. JAXP 1.0 uses SAX 1, and JAXP uses SAX2. The
PROPOSAL.html states that Digester will be JDK1.2, JAXP 1.1, and SAX2
compliant. I did not know there was a need to stay back at JAXP 1.0.
Is there anything stopping Struts people from using JAXP 1.1?
>
>
>> Scott Sanders
>>
>>
>
>
> Craig
