[Task]: Clean up Java Example PreCommit/PostCommit Dataflow

[Abacn]

What needs to happen?

Currently we have multiple Example PreCommit/PostCommit on Dataflow:

• PreCommit Example V1 (8, 11, 17) - run wordcount
• PostCommit Example V1 (11, 17) - run wordcount and impersonate wordcount
• PostCommit Example V2(8, 11, 17) - run full example (except wordcount)
• PostCommit V1 - run full example (except wordcount)
• PostCommit V2 - run full example (except wordcount)

There are a few issues

• No Dataflow Pipeline run on runner v2 at PreCommit - this has caused regression escaped e.g. Remove TableSchema to JSON conversion. #28274
• Wordcount never runs on runner v2 - either on Example PostCommit or the "big" PostCommit
• Duplicate test run, e.g. both PostCommit V2 and PostCommit Example V2 runs Java example
• Example tests are defined in two different files (runner/google-cloud-dataflow-java/build.gradle (for runner v1 tests) and runner/google-cloud-dataflow-java/example/build.gradle (for runner v2 tests))

We should put all example tests to example/build.gradle also

Issue Priority

Priority: 2 (default / most normal work should be filed as P2)

Issue Components

• Component: Python SDK
• Component: Java SDK
• Component: Go SDK
• Component: Typescript SDK
• Component: IO connector
• Component: Beam examples
• Component: Beam playground
• Component: Beam katas
• Component: Website
• Component: Spark Runner
• Component: Flink Runner
• Component: Samza Runner
• Component: Twister2 Runner
• Component: Hazelcast Jet Runner
• Component: Google Cloud Dataflow Runner