To me, this sounds pretty domain specific, so it is probably not a good fit for Pivot. However, maybe others will have a different opinion. G
On Thursday, November 12, 2009, at 06:38PM, "Sandro Martini" <[email protected]> wrote: >Hi to all, >some time ago I wrote some utility classes to handle in a transparent >way the (dynamic) loading of Classes, but in a way like ResourceBundke >does. My (real, business) problem was to handle standard code that in >some cases had to be customized (for example only for some Customers, >and any of these with its own version) overriding the standard >behavior. > >The logical flow is: >- set the full classname of the base (standard) class to load, >- set the code of the variant (for example the Customer Code), >- set other parameters, >- then calculate the full class name of the variant (depending on the >type of style, like subpackage, prefix, postfix, etc ... by default >postfix like in ResourceBundles, ex. MyBeanName_Customer), >- then try to load the customized version of the class, otherwise the >standard version. > > >I can do a rewrite of this for Pivot (1.4.1 or 1.5), but some time is >required (it was for a Web Enterprise Application) to drop some part, >and simplify others. > >Probably this is not best suited for Pivot core, but what do you think ? >Could help to have a support for this type of common problems. > >Bye, >Sandro > >
