The id bbb is null. 

By my understanding, the reason is:

1. the findRevision method return a null RevisionBean for PropertyTag. 

2. in the PropertyTagSupport class, the method :
        getProperty(RevisionBean revision, String name, String namespace) 
    will therefore return a null PropertyBean, since revision is null

3. the null object is put into pageContext with id: bbb

4. the tag <bean:write name='bbb' property='value'/> will throw a null point
exception.


As to objectnode type definition, I would suggest put it in the domain.xml
just like definition of the roles. Or use the full class name as the
type?????

Haiying Qiao




-----Original Message-----
From: Christopher Lenz [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, March 19, 2002 10:43 AM
To: Slide Users Mailing List
Subject: Re: RE: tag for nodebean.


19.03.2002 16:16:38, Haiying Qiao <[EMAIL PROTECTED]> wrote:
>I found a way to get around it ( not using property tag).
>
>The problem I mentioned before is:
>
>                <slide:node uri='<%= uri %>' id='nodebean' >    // seems I
>can struts tag for uri!!!

well, if I understand your comment here right, you'd like attributes like 
uriName/uriProperty to retrieve the uri-value from a bean-property, right ?
if so, I've been thinking about adding that, but haven't come around to it
yet. 
that's a problem with most of the slide tags (luckily, such flexibility is
much 
easier to support in the JSTL-based tag library).

>                  <slide:property id='bbb' namespace='DAV:'
>name='resourcetype' />
>                    <bean:write name='bbb' property='value'/>
>
>it return a null exception.

that's weird. can you tell exactly where the NullPointerException is being 
thrown? Actually, if the revision or property couldn't be found, the body of
the 
property-tag should be skipped.

>I check the source code:
>in StutsUtilsTagUtils.java the method :findRevision(Tag from, PageContext
>context)
>if the tag type is PropertyTagSupport ( since it extends the TagSupport).
[snip]

I don't understand what you're getting at, is there something wrong with the

code ?

>Maybe the propertyBean can still be retrieved without hte RevisonBean, in
>that case, this is a not problem.

You need a RevisionBean to retrieve the property, but if the revision tag
hadn't 
been explicitly used, the poperty-tag should use the latest-revision of the 
node.

>Here is one suggestion. In the iterate tag, the type for includetypes and
>excludetypes attribute is hard coded. I think it will be good to specify it
>in the configuration file just as role's definition. It will be very
>flexible since we may define our own ObjectNode.

right, but I'm not sure where such configuration could be added.

-chris
_______________________________________________
 /=/ cmlenz at gmx.de





--
To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>

Reply via email to