1. Javaflow or flowscript gets the initial call
2. All the request parameters, etc are all dealt with in flow
3. Hibernate is used from within flow to deal with the business request
4. When the request is processed, java objects representing the result are passed to the jxtemplate generator
5. Normal cocoon pipeline processing (jxtg, xsl, etc) are used to send final result
Irv
On 1/9/06, David Kavanagh <[EMAIL PROTECTED]> wrote:
FWIW, You can probably get more mileage out of a transformer. That way, you can pass config information into it via the pipeline, which would allow you to configure it via request params more easily. I've bene this route with Excel input and ended up with a transformer because it was a lot more flexible. Imagine, you can also throw it anyplace in your pipeline and have it operate on just the namespace it uses. (I'm sure others have said it better...)
However, in my case (and after looking at the cocoon-hibernate tutorial), we've decided to bypass cocoon in favor of our own SQL Transformer (which is quite a bit different and more advanced than the stanard cocoon one). We just didn't need the caching and lazy loading that hibernate would have provided. We need to drive an AJAX tree view of a tree data structure.
David
Thus Spoke Beat De Martin:
David Kavanagh wrote:
Has anyone interfaced with Hibernate in a pipeline? Did you write a
transformer, call JavaFlow or something else?
I'm thinking of using the Hibernate XML export feature to load objects
into my pipeline (inside a transformer). I'd certainly appreciate
hearing about any existing work in this area.
I'd like to do the same. I may do it with a generator.
Bye
Beat De Martin
