The Bean Machine is a simple tool for ordering a list of beans in your application. Beans are annotated
in such a way as to define a dependency graph between them, using @Precedes / @Succeeds. Additional granularity
is available by using the @First and @Last annotations, which provide ordering hints at each level of the tree.
BeanManagerabstraction, with out-of-the-box Spring and Pojo managers- Dependency free, with Spring being completely optional now!
- Annotate your beans and register them with Spring through component scanning, or manually delcare them.
- Scan or declare the
SpringBeanManagercomponent. - Inject the live
SpringBeanManagerinto your application and create a newBeanMachinewith it. - request a list of beans by type.
#getOrderedList(Class<T> type)
- Create a new
PojoBeanManager - Create a new
BeanMachinewith it. - Load it up with beans.
- Request an ordered list of beans.
When a list is requested, it is retreived either from the cache or recomputed. The cache is cleared automatically when the context is refreshed. When the list of beans is first computed the graph is checked for cycles, and if one is detected an exception is thrown.
See the tests for full usage examples.
The tool was initially written for several projects where pluggable processing steps were being used. Developers could add their own steps at loadtime or runtime simply by declaring their position relative to other steps (after 'A' but before 'D'). The beans are retrieved programmatically from Spring, which means that prototype beans without constructor arguments are also possible.
Care to share how you are using it differently? Send an email!
- sequenced processing of beans
- ordered event notification
- ordering guarantees without having to keep track of
@Ordervalues