"alesj" wrote :
| Does this somehow fit into the picture you have/had?
|
No not really.
I wanted you to seperate the annotation scanning from the component creation.
i.e the annotation @Bean or a "collection" of them @Beans?
becomes an attachment in its own right.
The component deployer should not care where the annotation(s) come from.
One possible source is the annoation scanning.
Schematically:
BeanAnnotationScanner: class scan -> annotation attachments
Component deployer: metadata or annotations -> BeanMetaData components
That way if I dont want annotation scanning for beans or I want to use
some other logic I can remove, replace or insert a new deployer to do what I want.
e.g. Hypothetically: suppose I want all beans scanned from classes to be
"on-demand"
I can add a deployer inbetween that modifies the annotations but doesn't touch
those created using normal metadata.
Like I said before, I don't really think anybody will do this kind of thing,
but we should do it properly just in case.
Also it keeps the processing to "TO DO ONE THING WELL"
i.e. one deployer knows how to scan for bean annotations
one knows how to create components from a set of such annotations
(mod those that already have metadata).
View the original post :
http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4154690#...
Reply to the post :
http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...