├── README.md
└── MyExplode
    └── src
        └── com
            └── example
                └── hive
                    └── udf
                        └── MyExplode.java


/README.md:
--------------------------------------------------------------------------------
 1 | HDInsight
 2 | =========
 3 | 
 4 | HDInsight : Hadoop - Hive - Microsoft Business Intelligence
 5 | 
 6 | -------------------------------------------------------------------------------------------------------------
 7 | 
 8 | The UDF Explode() take an array and explode it : SELECT EXPLODE(COL) AS MyCOL FROM MyTable;
 9 | 
10 |      Col       	 MyCOL
11 |   | [1,2] |	->	|  1  |
12 |   | [3,4] |		|  2  |
13 |   				|  3  |
14 |   				|  4  |
15 | 
16 | With LATERAL VIEW : SELECT ID, MyCol FROM MyTable LATERAL VIEW EXPLODE(Col) MyVirtualTable AS MyCol;
17 | 
18 |     ID    Col          ID  MyCol
19 |    | 1 | [1,2] |  ->  | 1 |  1  |
20 |    | 2 | [3,4] |      | 1 |  2  |
21 |                       | 2 |  3  |
22 |                       | 2 |  4  |
23 | 
24 | With Multiple LATERAL VIEW : 
25 |   SELECT MyCol1, MyCol2 FROM MaTable 
26 |   LATERAL VIEW explode(Col1) MyVirtualTable1 AS MyCol1
27 |   LATERAL VIEW explode(Col2) MyVirtualTable2 AS MyCol2;
28 | 
29 |       Col1    Col2          MyCol1  MyCol2
30 |    | [1,2] | [5,6] |  ->  |   1   |   5   |
31 |    | [3,4] | [7,8] |      |   1   |   6   |
32 |                           |   2   |   5   |
33 |                           |   2   |   6   |
34 |                           |   3   |   7   |
35 |                           |   3   |   8   |
36 |                           |   4   |   7   |
37 |                           |   4   |   8   |
38 | 
39 | !!! With MyExplode() UDF :
40 |   SELECT MyCol1, MyCol2 FROM MyTable 
41 |   LATERAL VIEW MyExplode(Array(1,2,3), Array(4,5,6)) MyVirtualTable As MyCol1, MyCol2;
42 | 
43 |        Col1      Col2           MyCol1  MyCol2
44 |    | [1,2,3] | [4,5,6] |  ->  |   1   |   4   |
45 |                               |   2   |   5   |
46 |                               |   3   |   6   |
47 | 
48 | -------------------------------------------------------------------------------------------------------------
49 | Instalation : 
50 | 
51 | hive>  ADD JAR /#PATH#/MyExplode.jar; 
52 | hive>  CREATE TEMPORARY FUNCTION MyExplode AS 'com.example.hive.udf.MyExplode';
53 | 
54 | 


--------------------------------------------------------------------------------
/MyExplode/src/com/example/hive/udf/MyExplode.java:
--------------------------------------------------------------------------------
 1 | package com.example.hive.udf;
 2 | 
 3 | import java.util.ArrayList;
 4 | import java.util.List;
 5 | 
 6 | import org.apache.hadoop.hive.ql.exec.Description;
 7 | import org.apache.hadoop.hive.ql.exec.UDFArgumentException;
 8 | import org.apache.hadoop.hive.ql.metadata.HiveException;
 9 | import org.apache.hadoop.hive.ql.udf.generic.GenericUDTF;
10 | import org.apache.hadoop.hive.serde2.objectinspector.ListObjectInspector;
11 | import org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector;
12 | import org.apache.hadoop.hive.serde2.objectinspector.ObjectInspectorFactory;
13 | import org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector;
14 | import org.apache.hadoop.hive.serde2.objectinspector.PrimitiveObjectInspector.PrimitiveCategory;
15 | import org.apache.hadoop.hive.serde2.objectinspector.primitive.PrimitiveObjectInspectorFactory;
16 | 
17 | @Description(name = "MyExplode",
18 | 	value = "_FUNC_(a, b) - separates the elements of arrays a and b into multiple rows")
19 | 
20 | public class MyExplode extends GenericUDTF {
21 | 
22 |   private ListObjectInspector listOI0 = null;
23 |   private ListObjectInspector listOI1 = null;
24 | 
25 |   @Override
26 |   	public void close() throws HiveException {
27 |   }
28 | 
29 |   @Override
30 |   public StructObjectInspector initialize(ObjectInspector[] args) throws UDFArgumentException {
31 |    
32 |     if (args[0].getCategory() != ObjectInspector.Category.LIST) {
33 |       throw new UDFArgumentException("explode() takes an array as a parameter");
34 |     }
35 |     listOI0 = (ListObjectInspector) args[0];
36 |     listOI1 = (ListObjectInspector) args[1];
37 | 
38 |     ArrayList<String> fieldNames = new ArrayList<String>();
39 |     ArrayList<ObjectInspector> fieldOIs = new ArrayList<ObjectInspector>();
40 |     fieldNames.add("col0");
41 |     fieldNames.add("col1");
42 |     fieldOIs.add(PrimitiveObjectInspectorFactory.getPrimitiveJavaObjectInspector(PrimitiveCategory.STRING));
43 |     fieldOIs.add(PrimitiveObjectInspectorFactory.getPrimitiveJavaObjectInspector(PrimitiveCategory.STRING));
44 |     return ObjectInspectorFactory.getStandardStructObjectInspector(fieldNames, fieldOIs);
45 |   }
46 | 
47 |   private final Object[] forwardObj = new Object[2];
48 | 
49 |   @Override
50 |   public void process(Object[] o) throws HiveException {
51 |     List<?> list0 = listOI0.getList(o[0]);
52 |     List<?> list1 = listOI1.getList(o[1]);
53 |     if(list0 == null) {
54 |       return;
55 |     }
56 |     for(int i=0; i< list0.size(); i++)
57 |     {
58 |      Object r0 = list0.get(i);
59 |      Object r1 = list1.get(i);
60 |      forwardObj[0] = r0.toString();
61 |      forwardObj[1] = r1.toString();
62 |        forward(forwardObj);
63 |     }
64 |   }
65 | 
66 |   @Override
67 |   public String toString() {
68 |     return "MyExplode";
69 |   }
70 | }


--------------------------------------------------------------------------------