Hi All, I am a relative struts newbie, using the following: Struts 1.0.2, Sun JDK1.3.1, Tomcat3.3, MySQL, mm.mysql jdbc driver.
In my application, the only components that need JDBC are model components, which I am trying to keep decoupled from any one particular front-end for the sake of reusability. I would like to use the connection pooling available in Struts by passing a javax.sql.DataSource into model-land at initialization time. Unfortunately I have not found a good way to do that. I have tried the obvious method of writing a simple servlet (subclass of HttpServlet) and set it up in web.xml to load AFTER Struts ActionServlet when the webapp is started. I hoped I could somehow get the DataSource via ActionServlet.findDataSource(), the only widely-documented way to get it. I would then pass this DataSource to a configuration singleton class in model-land for use from there. I was hoping that ActionServlet would leave a reference to itself in the ServletContext, but alas. It appears that the only way to have access to the ActionServlet, and hence the DataSource, is to be an Action invoked by it. Would the 'right way' be to subclass ActionServlet and initialize my model from init() after calling the superclass? What it the right way to (generically) initialize model components with things like DataSource, log sink, parameters & resources from the struts / webapp framework? --Jeff P.S. as an OT aside, could anyone recomend a decent OSS java connection-pooling implementation? (I have already ruled out PoolMan) -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
