xml-dev - Re: XML and LDAP: Common APIs?

[Jeff Turner <jeff@socialchange.net.au>]

mark hu wrote:
> 
> Yes, I agree - and the stuff I am working on today is bring a collection of
> XML objects ( in which every object has the same XML structure ), and the
> number of objects could be changed very often, and futher the relation-ship
> between these objects are parent and child, and one root level parent could
> any number child and again these child could have multi-level sub child ,
> and so on ..... all these objects are linked together baed on 2 of the XML
> tags ( e.g. nodeIDTag and parentIDTag ) - if it's NULL value means root
> level parent, if it's not NULL follows this rule: all the parentIDTag in a
> child note would has the value of its nearest parentIDTag value, so all the
> objects would eventually construct a OBJECT TREE, and on the browser side -
> I'll construct the whole TREE as a TREE VIEW list .... is there any good
> idea to do this ????? right now, I am counting the max level of the TREE
> level first ( both vertical and horizontal ... ), then go through a loop
> based on these 2 params to creat vertical based objects collection ( which
> has its nearest parent on the left ), then I have a JavaScript code to
> generate client side TREE VIEW which takes these collections to populate all
> the nodes value ..... thanks.

Well.. if you've got all these objects which you can represent as XML,
why not store them as a DOM Document? If you're using Java, JDOM
(http://www.jdom.org) would be easiest.

Otherwise, if you want simple object->XML mapping, there's various APIs
to do that:

http://www.jxml.com/quick/index.html
http://castor.exolab.org

--Jeff

> 
> FYI:
> 
> <ROOTCOLLECTION>
>     <ROOT1>
>         <nodeIDTag>1</nodeIDTag>
>         <parentIDTag></parentIDTag>
>     </ROOT1>
>     <CHILD1.1>
>         <nodeIDTag>1.1</nodeIDTag>
>         <parentIDTag>1</parentIDTag>
>     </CHILD1.1>
>     <CHILD1.1.1>
>         <nodeIDTag>1.1.1</nodeIDTag>
>         <parentIDTag>1.1</parentIDTag>
>     </CHILD1.1.1>
>     <CHILD1.1.1.1>
>         <nodeIDTag>1.1.1.1</nodeIDTag>
>         <parentIDTag>1.1.1</parentIDTag>
>     </CHILD1.1.1.1>
>     <CHILD1.2>
>         <nodeIDTag>1.2</nodeIDTag>
>         <parentIDTag>1</parentIDTag>
>     </CHILD1.2>
>     <CHILD1.3>
>         <nodeIDTag>1.3</nodeIDTag>
>         <parentIDTag>1</parentIDTag>
>     </CHILD1.3>
>     <ROOT2>
>         <nodeIDTag>2</nodeIDTag>
>         <parentIDTag></parentIDTag>
>     </ROOT2>
>     <CHILD2.1>
>         <nodeIDTag>2.1</nodeIDTag>
>         <parentIDTag>2</parentIDTag>
>     </CHILD1.1>
>     <ROOT3>
>         <nodeIDTag>3</nodeIDTag>
>         <parentIDTag></parentIDTag>
>     </ROOT3>
>     <CHILD3.1>
>         <nodeIDTag>3.1</nodeIDTag>
>         <parentIDTag>3</parentIDTag>
>     </CHILD3.1>
> </ROOTCOLLECTION>
> 
> mark
> 
> "Life Is a Such a B......"
> 
> ----- Original Message -----
> From: Lauren Wood <lauren@sqwest.bc.ca>
> To: <xml-dev@lists.xml.org>
> Sent: Thursday, August 24, 2000 12:27 PM
> Subject: Re: XML and LDAP: Common APIs?
> 
> > On 24 Aug 2000, at 13:58, Jeff Turner wrote:
> >
> > > Hi,
> > >
> > > I was just thinking how similar the XML and LDAP data models are. LDAP
> > > gives you a tree, where each node can contain data -and- can act as a
> > > namespace for other nodes. This corresponds pretty closely to XML (or
> > > maybe SML).
> > >
> > > Has anyone explored the possibility of crossover APIs for accessing all
> > > these 'tree-like' data structures in a standardised way? I'd like to be
> > > able to traverse an LDAP directory with DOM or SAX.
> >
> > I can't see any reason why this shouldn't work; as long as you have
> > some sort of tree structure, why not use getParentNode and have it
> > return as close to the same thing as possible? Not everything from
> > DOM or SAX will be applicable, but it would be good to have those
> > methods which do the same thing being called the same as well,
> > so the meaning is recognisable.
> >
> > On a side note, I've even had people talk to me about using DOM
> > methods to walk a text document, where you map the lines into
> > the equivalents of elements, with the line breaks being the
> > demarcation between them. Then you can apply (some) DOM
> > methods to the resulting tree. I'm still not sure why anyone would
> > *want* to do this, but it's possible.
> >
> >
> >
> >
> > Lauren
> >