This is a simple Java program to illustrate the use of User Defined Functions (UDFs) in Apache SparkSql.
This POC will show you how to register, define, and call a UDF on a dataset. There are examples of how to pass parameters to the UDF as well.
As of now, there's only one input as a command line argument to the program. You need to pass the input file path to the program.
The input file should be a .csv with the columns name and number. A sample is included in the root of the project as an example. The file content is as follows:
name,number
name1,1
name2,2
name3,3
name4,4
The first row of the CSV file will be ignored as it is assumed to be the header.
You need to build the project first before running it. You can build it by running the following command from the root of the project directory:
mvn clean installAfter this, run the following command to run the project from the same directory:
java -jar target/SparkSqlUDF-POC-1.0-SNAPSHOT.jar inputFile.csv