Spark Dataframe writing to google pubsub

Question

I am trying to write the Parquet files to Pubsub through Spark on a Dataproc cluster.

I have used below pseudo code

dataFrame
      .as[MyCaseClass]