Trigger point framework
A trigger point is the location in a method of an object where
externalized business rules are invoked.
The proper placement of trigger points can add substantially to the flexibility
and speed with which a business application adapts to new business practices.
Wherever a trigger point is placed in user-written code, the Business Rule
Beans (BRBeans) trigger point framework needs to do the following:
- Assemble the parameter list to send to the rules
- Find the potential rules that apply
- Filter out any rules which do not apply (optional)
- Fire the rules in the filtered rule set
- Combine the results of the rule firings is some meaningful way
The application code that contains the trigger point needs to perform the
following functions:
- Establish a value for the target object. Usually the target object is
the object in which the trigger point is encountered. The target object is
one of the parameters passed to the fire method of the RuleImplementor.
However, this is an optional parameter. If the rule implementor does require
a target object, null can be passed instead.
- Build the array of objects containing the run-time parameters needed to
satisfy the trigger point's business purpose. This array is normally passed
as one of the parameters of the fire method of
the RuleImplementor. If firing parameters are specified
on the rule itself, then those firing parameters are passed instead of the
ones passed by the caller.
- Invoke the trigger(), triggerClassifier(), or triggerSituational() method of the TriggerPoint class.
- Catch and handle any exceptions that might occur as a result of firing
the rules. Otherwise, take action based upon the rule firing results.
The two simple trigger methods, trigger and triggerClassifier, perform their function in four steps:
- Find the rules
- Filter out those rules which are not desired
- Fire the remaining rules
- Combine the results and return to the caller
The complex trigger method, triggerSituational does
this sequence of steps twice. In the first phase, the method performs the
four steps once to find a rule that returns a classification. This classification
is fed into the second phase. The second phase triggers rules that have the
name specified in the triggerSituational method and have a classification
equal to the value returned by the first phase.
How each of these steps is performed can be modified through various methods
on the TriggerPoint object. The implementation of each step is defined by
a strategy object. For more information on strategies, see Administering strategy objects to control
triggers.

Overview of Business Rule Beans
Simple trigger point
Classifier trigger point
Situational trigger point

Administering strategy objects to control triggers
Searchable topic ID:
cbrb_trgpnt
Last updated: Jun 21, 2007 8:07:48 PM CDT
WebSphere Business Integration Server Foundation, Version 5.0.2
http://publib.boulder.ibm.com/infocenter/wasinfo/index.jsp?topic=/com.ibm.wasee.doc/info/ee/brb/concepts/cbrb_trgpnt.html