Hive UDF Development and Application

Developing Hive UDFs

This sample implements one Hive UDF described in the following table.

• A common Hive UDF must be inherited from org.apache.hadoop.hive.ql.exec.UDF.
• A common Hive UDF must implement at least one evaluate(). The evaluate function supports overloading.
• Currently, only the following data types are supported:
  • boolean, byte, short, int, long, float, and double
  • Boolean, Byte, Short, Int, Long, Float, and Double
  • List and Map

  UDFs, UDAFs, and UDTFs currently do not support complex data types other than the preceding ones.

• Currently, Hive UDFs supports only less than or equal to five input parameters. UDFs with more than five input parameters will fail to be registered.
• If the input parameter of a Hive UDF is null, the call returns null directly without parsing the Hive UDF logic. As a result, the UDF execution result may be inconsistent with the Hive execution result.
• To add the hive-exec-3.1.1 dependency package to the Maven project, you can obtain the package from the Hive installation directory.
• (Optional) If the Hive UDF depends on a configuration file, you are advised to save the configuration file as a resource file in the resources directory so that it can be packed into the Hive UDF function package.

1. Create a Maven project. Set groupId to com.test.udf and artifactId to udf-test. The two values can be customized based on the site requirements.
2. Modify the pom.xml file as follows:

   <project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
          <modelVersion>4.0.0</modelVersion>
          <groupId>com.test.udf</groupId>
          <artifactId>udf-test</artifactId>
          <version>0.0.1-SNAPSHOT</version>
   
          <dependencies>
            <dependency>
               <groupId>org.apache.hive</groupId>
               <artifactId>hive-exec</artifactId>
               <version>3.1.1</version>
            </dependency>
          </dependencies>
   
          <build>
            <plugins>
               <plugin>
                   <artifactId>maven-shade-plugin</artifactId>
                   <executions>
                       <execution>
                           <phase>package</phase>
                           <goals>
                               <goal>shade</goal>
                           </goals>
                       </execution>
                   </executions>
               </plugin>
               <plugin>
                   <artifactId>maven-resources-plugin</artifactId>
                   <executions>
                       <execution>
                           <id>copy-resources</id>
                           <phase>package</phase>
                           <goals>
                               <goal>copy-resources</goal>
                           </goals>
                           <configuration>
                               <outputDirectory>${project.build.directory}/</outputDirectory>
                               <resources>
                                   <resource>
                                       <directory>src/main/resources/</directory>
                                       <filtering>false</filtering>
                                   </resource>
                               </resources>
                           </configuration>
                       </execution>
                   </executions>
               </plugin>
            </plugins>
          </build>
   </project>

3. Create the implementation class of the Hive UDF.

   import org.apache.hadoop.hive.ql.exec.UDF; 
   
   /** 
    * AutoAddOne 
    * 
    * @since 2020-08-24 
    */ 
   public class AutoAddOne extends UDF { 
       public int evaluate(int data) { 
           return data + 1; 
       } 
   }

4. Package the Maven project. The udf-test-0.0.1-SNAPSHOT.jar file in the target directory is the Hive UDF function package.
   

   You need to pack all dependencies into a JAR package.

Configuring Hive UDFs

In configuration file udf.properties, add registration information in the "Function_name Class_path" format to each line.

The following provides an example of registering four Hive UDFs in configuration file udf.properties:

booleanudf io.hetu.core.hive.dynamicfunctions.examples.udf.BooleanUDF
shortudf io.hetu.core.hive.dynamicfunctions.examples.udf.ShortUDF
byteudf io.hetu.core.hive.dynamicfunctions.examples.udf.ByteUDF
intudf io.hetu.core.hive.dynamicfunctions.examples.udf.IntUDF

• If the added Hive UDF registration information is incorrect, for example, the format is incorrect or the class path does not exist, the system ignores the incorrect registration information and prints the corresponding logs.
• If duplicate Hive UDFs are registered, the system will only register once and ignore the duplicate registrations.
• If the Hive UDF to be registered is the same as that already registered in the system, the system throws an exception and cannot be started properly. To solve this problem, you need to delete the Hive UDF registration information.

Deploying Hive UDFs

To use an existing Hive UDF in HetuEngine, you need to upload the UDF function package, udf.properties file, and configuration file on which the UDF depends to the specified HDFS directory, for example, /user/hetuserver/udf/, and restart the HetuEngine compute instance.

Using Hive UDFs

Use a client to access a Hive UDF:

1. Log in to the HetuEngine client. For details, see Using the HetuEngine Client.
2. Run the following command to use a Hive UDF:

   select AutoAddOne(1);

   select AutoAddOne(1);
   _col0
   -------
        2
   (1 row)