Modified KffDataObjectHandler.hash(..) method so it explicitly replaces IBDO params it generates.

[drivenflywheel]

A given object should have only one value for each type of hash we compute. With this change, the last one computed wins. Need to make sure that's the behavior we want.

Honestly, we should probably instead modify the behavior so that the hash method hashes a given payload at-most one time (so the computed hashes represent the data in its original form), with any subsequent calls being no-op'ed.

[fbruton]

Rebased and observed the following errors:

[ERROR]   IBaseDataObjectDiffHelperTest.testDiffCurrentForm:137 » ExceptionInInitializer
[ERROR]   UnixFilePlaceTest>IdentificationTest.testIdentificationPlace:61 »  java.lang.reflect.InvocationTargetException
[ERROR]   KffDataObjectHandlerTest.setUp:60 »  java.lang.reflect.InvocationTargetException
[ERROR]   KffHashPlaceTest.testProcess:24 » IllegalState No config directory specified

[drivenflywheel]

Rebased and observed the following errors:

[ERROR]   IBaseDataObjectDiffHelperTest.testDiffCurrentForm:137 » ExceptionInInitializer
[ERROR]   UnixFilePlaceTest>IdentificationTest.testIdentificationPlace:61 »  java.lang.reflect.InvocationTargetException
[ERROR]   KffDataObjectHandlerTest.setUp:60 »  java.lang.reflect.InvocationTargetException
[ERROR]   KffHashPlaceTest.testProcess:24 » IllegalState No config directory specified

Run it again. That's an intermittent-but-recurring issue unrelated to this change. Since the order of unit test execution is randomized, it might indicate that the codebase has one or more unit test classes that don't extend Emissary's UnitTest class. I don't recall the details exactly, but I believe there's a order-of-operations issue that interferes with the initialization of tests executed after one of these "non-confirming" tests. Symptoms suggest contamination of system properties or the PROJECT_BASE environment variable

[dev-mlb]

Rebased and observed the following errors:

[ERROR]   IBaseDataObjectDiffHelperTest.testDiffCurrentForm:137 » ExceptionInInitializer
[ERROR]   UnixFilePlaceTest>IdentificationTest.testIdentificationPlace:61 »  java.lang.reflect.InvocationTargetException
[ERROR]   KffDataObjectHandlerTest.setUp:60 »  java.lang.reflect.InvocationTargetException
[ERROR]   KffHashPlaceTest.testProcess:24 » IllegalState No config directory specified

Run it again. That's an intermittent-but-recurring issue unrelated to this change. Since the order of unit test execution is randomized, it might indicate that the codebase has one or more unit test classes that don't extend Emissary's UnitTest class. I don't recall the details exactly, but I believe there's a order-of-operations issue that interferes with the initialization of tests executed after one of these "non-confirming" tests. Symptoms suggest contamination of system properties or the PROJECT_BASE environment variable

Ugh...I thought I had gotten all of those at one point. This should fix it: #597

addressed comment