[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [xmlblaster] XML newby question


On Mon, Feb 17, 2003 at 09:52:17AM +0100, Marcel Ruff wrote:
> Michael Atighetchi wrote:
> >The documentation in XmlKey gives this as a typical publish example:
> >
> >    <key oid='4711' contentMime='text/xml'>
> >        <AGENT id='' subId='1' type='generic'>
> >           <DRIVER id='FileProof' pollingFreq='10'>
> >           </DRIVER>
> >        </AGENT>
> >    </key>
> >
> >
> >What is the common way for an application to parse this XML chunck in
> >the subscribing client to get out id, subId and type in <AGENT> ?
> This message meta information is put into the big internal DOM tree
> and is immutable for the lifetime of a topic.
> You can NOT extract parts of it, you can only query on it.
> Only the first (administrative) message of a topic (a specific key oid)
> is used to store the key DOM.
> Sending further messages with the same key oid but different
> key meta informations have not effect (the key meta information
> is ignored).
> The key meta information (here AGENT and DRIVER) are only used
> to describe a topic and to find it with XPath.
> The message content delivers your problem domain data.
> If the content contains XML syntax you can consider using
> Peter Antmans XPath mime plugin to further filter
> messages depending on the content (full text search), see
> http://www.xmlblaster.org/xmlBlaster/doc/requirements/mime.plugin.access.xpath.html
thanks for the explanation (see email to Cyrille in this thread).

> Where should we improve the documentation to make this clear?
> (where did you expect to find this information?)
You could add someting in the JAVA DOC API for MsgKeyData, saying that
application level tags are passed to the subscribing application and
have to be parsed via DOMs or java regexpressions. Now that I
understand XML a little better, it seems obvious though.

> regards
> Marcel
> >
> >How does this relate to the implicit DOM tree that xmlblaster creates
> >(I say a lot of references to this in the documentation, but could not
> >figure out how to extract the attribute values out of the DOM).
> >
> >Michael
> >
> >

matighet at bbn.com   BBN Technologies