BTW there is also a setup() method you can override. This is called during pipeline setup - before any components are executed.
Regards Jorg Heymans
Bent Andr� Solheim wrote:
First of all, thank you for your answer Jorg!
I'm having a bit trouble with your suggestion. It appears the transformer is configured after MyGenerator.compose(ComponentManager), but before MyGenerator.generate(). It is in generate() I look up the transformation id and put it in the Request, so when the transformer is being configured, the parameter is not set, and I get a FileNotFoundException (Unable to get transformer handler for file...).
I cannot do the lookup in compose, because this is only called on generator creation. Have I misunderstood something here?
Best regards, Bent.
On Thu, 11 Nov 2004 15:30:09 +0100, Jorg Heymans <[EMAIL PROTECTED]> wrote:
you can make your generator set a request attribute, and retrieve it in the sitemap via the request-attr module .
Note that you cannot make the transformer *type* dynamic.
If you want to execute a different pipeline depending on parameters then i think you could selectors for this (i've never used them so i could be wrong)
does this make sense?
Jorg
Bent Andr� Solheim wrote:
Hi Cocoon users.
I'm developing an application where we keep thousands of gzipped xml documents in some 20 different formats in a legacy relational database. These documents can be accessed using a document id. I have written a custom generator that retrieves a gzipped document from the database based on a request parameter, unzips it, and passes sax events on to a transformer. Which transformation the transformer should use was earlier passed as a request parameter along with the document id, but new requirements dictates that only the document id can be passed as parameter. Now I have to look the transformation id up in a database (in the same table as the gzipped xml document).
The application will be under heavy load constantly, so I would like to minimize the number of database accesses pr. request. Hence, I would like to look up the gzipped xml data and the transformation id in one query. My question is; is there a way for my custom generator to pass the transformation id to the next transformer in the pipeline (file based)? Or do I need to write my own transformer that looks up the transformation id based on the document id request parameter and figures out what transformation to apply by it self?
To summarize; is there a way for my generator to tell the next transformer which transformation to use?
Any suggestion would be recieved with great appreciation! :)
Best regards Bent Andr� Solheim
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
