Modify GCS Bucket Name for Nightly Build Jar

[prashastia]

Note: Before merging:

• Change he cloud build variable _GCS_JAR_LOCATION to gs://flink-bq-connector-nightly-job/flink-bq-connector-app-jars
• Post modification, run the e2e testing pipeline for verification.

Details:
Currently, the jar is copied to a common named GCS bucket and is then used by Dataproc Clusters to execute the job.

When multiple cloudbuild runs are running simultaneously, this can create a problem as we might loose control over the jar version being tested.
To prevent this, The Jar Name should be modified with a timestamp of execution (like all other aspects in the test). This jar name should then be saved to a file in the VM and used by all the steps(Tests) in the Run.
So now instead of gs://.../BigQueryIntegrationTest.jar The jar would be something like - gs://.../[TIMESTAMP]/BigQueryIntegrationTest.jar

• Copy maven jar to gs://.../[TIMESTAMP]/BigQueryIntegrationTest.jar
• Save the name of this file to _GCS_JAR_LOCATION_FILE
• Replace _GCS_JAR_LOCATION variable with _GCS_JAR_LOCATION_FILE in every step of yaml and read the value of GCS_JAR_LOCATION from this file (in the run_test() function)

/gcbrun