Configure the "namespace" and "name" attributes generated by the CDC replication engine for Kafka

Description of the issue: We want to generate Java classes from the schema registry subjects associated to target topics in order to deserialize the keys and values of Avro messages written to this topics by the CDC replication engine for Kafka of Infosphere Data Replication. There are standard tools that generate the classes from the Avro record stored in the schema registry under a specific subject version (subject = topic-key or topic-value). These tools use the "namespace" and "name" attributes of the record as package name and class name respectively. The "namespace" and "name" attributes generated by the CDC replication engine for Kafka do not suite our needs for the following reasons:

1. The "namespace"/"name" do not conform to our organizational naming conventions for Java classes.
2. The "namespace" contains references to the environment at several places and is therefore not usable generically in our Java code.

We tried the following obvious workaround: read the records from the schema registry, modify the name und namespace in a new avro schema file and generate the classes from the new files. That unfortunately did not work (apparently the package and class names are matched against the namespace and name attributes at some point) and the messages could not be deserialized. Please let us know if you arrive to a different conclusion here.

IBM Workaround: The workaround supplied  did work, as far as we could tell. However, we would like a solution that is both officially supported by IBM and is directly integrated to the distribution. If possible, for basic functionality, we would like to avoid having to write and maintain our own KCOPs and rely instead on a solution provided and supported by IBM.