Converting Spark UDF written in SCALA to JAVA - scala

Question

Below is my spark UD, can anyone help me to convert this into java?
val customUDF = udf((array: Seq[String]) => {
val newts = array.filter(_.nonEmpty)
if (newts.size == 0) null
else newts.head
})

Answer 1

Something like:
UDF2 my_udf = new UDF2<WrappedArray<String>, String>() {
public String call(WrappedArray<String> arr) throws Exception {
String[] newts = arr.filter(_.nonEmpty)
if (newts.length == 0) {
return null
} else { newts[0] }
}
};
spark.udf().register("my_udf", my_udf, DataTypes.StringType);

Answer 2

You can do it in two ways
Inline using Lambda i.e. Scala style
Or you can define a method and register it.
import org.apache.spark.sql.SparkSession;
import org.apache.spark.sql.api.java.UDF0;
import org.apache.spark.sql.api.java.UDF1;
import org.apache.spark.sql.types.DataTypes;
import java.util.List;
public class SimpleUDF {
public static void main(String[] args) {
SparkSession spark = SparkSession.builder().master("local[*]").getOrCreate();
spark.sqlContext()
.udf()
.register("sampleUDFLambda", (List<String> array) -> array.stream().filter(element ->
!element.isEmpty()).findFirst().orElse(null), DataTypes.StringType);
}
//Or you can define a function
private UDF1< List<String>,String> sampleUdf()
{
return ( array ) -> array.stream().filter(element ->
!element.isEmpty()).findFirst().orElse(null);
}
}