This content has been marked as final.
Show 2 replies
-
1. Re: Internal AbstractController State
adrian.brock Apr 3, 2007 9:14 AM (in response to adrian.brock)"adrian@jboss.org" wrote:
The reason I say it is dangerous is because there is more to updating a context's
state than allContexts. There are other "indexes" where the context could exist.
I'm actually currently changing it so you can have a name->context
in the allContexts which doesn't exist anywhere else.
These will be aliases.
So that will be even more work do in the scoped kernel,
i.e. move the aliases as well as the main name/context.
This says to me that we need a well defined api for these operations
(register/unregisterControllerContext) where this, future enhancements
and subclasses can plugin their new knowledge. -
2. Re: Internal AbstractController State
adrian.brock Apr 3, 2007 11:07 AM (in response to adrian.brock)I've added a protected [un]registerControllerContext(ControllerContext)
for the alias handling.
This works when invoked from the [un]installed methods,
but it does not handle maintinaing or clearing up the other internal state
for the context (there is none for aliases).