The fields of the input record that should be populated.
The fields of the input record that should be populated. If a projection should not be applied, this should be null. Code upstream will check for a null value, so this is safe.
A schema describing the projection to apply to all records.
Run method to be implemented by plugin implementor.
Run method to be implemented by plugin implementor.
Java Spark Context that this plugin is running inside of.
Input RDD of records.
Command line arguments passed to this plugin.
Output RDD to be saved to disk.
Plugins that do not want to apply a predicate should extend this class.