of algorithm implementations, such as the correct number of rules
for a given support and confidence level on a given dataset. Further,
it does not test negative cases, that is, those involving error conditions.
For example, if a vendor claims not to support a given capability, the
DME implementation should return an UnsupportedFeatureException
if a feature associated with this capability is invoked by the caller.
To run the TCK tests, you will need the following jar files:
• jdm.jar: the jar containing interfaces defined in JDM.
• junit.jar: the TCK is based on the JUnit unitary test frame-
work [JUNIT 2006].
• jdmTCK1_1_0.jar: the jar containing the test scenarios to be
run on the implementation.
• jdmTCKFramework1_1_0.jar: a jar containing utilities to
start the test process.
The four jar files must be in the CLASSPATH of your runtime
environment to be able to run the TCK. Then, assuming that the jar
file containing your implementation is called MyJDMImpl.jar, you
will have to perform the following operations:
• Create a test directory.
• Under the test directory, create (1) a libs directory in which
the five jar files will be stored, (2) a config directory in
which all eight configuration files provided with the down-
load will be copied, and (3) a reports directory in which the
TCK will produce the reports.
• Edit the configuration files. Most of them contain specifica-
tions for datasets that are expressed using URIs; thus, they
must be edited because the URI syntax is vendor dependent
and must point to valid datasets in the test environment.
The master configuration file, called tckconfig.pro , points to
six others, and a connection.pro that contains the class path
used by the test framework to point to the JDM implemen-
tation connection factory.
Then, go to the test directory and launch the following command:
java.exe -classpath libs\jdmTCKFramework1_1_0.jar;libs\jdmTCK1_1_0.jar;