I already reported that as a Bug to the official Spark bug tracker: https://issues.apache.org/jira/browse/SPARK-46876A short summary: When reading a tab separated file, that has lines that only contain of tabs, then this line will not show up in the ...