Hi Lionel:

>
>Yes actually I think you're right. I just hope that the QName java class 
>won't be 'final'.

 I agree with you that this can't be final. And all the indications from the 
JAXP 1.3 EG are that it won't be final.
 
 Regards
 -Ramesh
 
 
 
 
 I just have to do is to rely on the expression factory 
>to creates/gets a qname.
>Thanks!
>
>Lionel Villard
>IBM T.J Watson
>
>
>
>
>
>Ramesh Mandava <[EMAIL PROTECTED]>
>07/25/2003 02:25 PM
>Please respond to xalan-dev
> 
>        To:     [EMAIL PROTECTED]
>        cc: 
>        Subject:        Re: XPath API issues and plans
>
>
>Hi Lionel:
>
>
>>Hi all,
>>        I want to give you an overall of the remaining issues that need 
>to 
>>
>>be addressed to fulfill the XPath API requirements (in xstl20 branch 
>under 
>>
>>the xpath_rwapi directory). One of these requirements is to allow several 
>
>>applications to implement and/or reuse efficiently the XPath API and many 
>
>>problems need to be solved in order to reach it. Here the list of these 
>>issues and also more general tasks to be done:
>> 
>>XPath public API:
>>
>>- namespace/qname resolver : most of applications use a qname manager for 
>
>>example to share the same qname pool throughout their application. 
>>Currently the public XPath API relies on the JAXP QName definition. I 
>want 
>>
>>to remove this dependency and to define a QName interface that 
>>applications can implement within their QName manager. 
>
> Can we get away with defining another QName interface. If needed we can 
>have, 
>JAXP QName ( supposed to be the standard )   wrapped inside the QName 
>Manager.
>  As the main aim of JAXP QName is to standardize QName, our adaption 
>would help 
>that goal. 
> 
> Regards
> -Ramesh 
> 
>
>

Reply via email to