get size of parquet file in HDFS for repartition with Spark in Scala

I have many parquet file directories on HDFS that contain a few thousands of small(most < 100kb) parquet files each. They slow down my Spark job, so I want to combine them. With the following c...