# Problem in spliting record

**URL:** <https://forum.cloverdx.com/t/problem-in-spliting-record/20>\
**Category:** CloverDX Platform\
**Created:** [July 16, 2007, 12:00am UTC](https://forum.cloverdx.com/t/problem-in-spliting-record/20 "2007-07-16T00:00:00Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![oldforum](https://avatars.discourse-cdn.com/v4/letter/o/13edae/32.png) [@oldforum](https://forum.cloverdx.com/u/oldforum)\
**Post date:** [July 16, 2007, 12:00am UTC](https://forum.cloverdx.com/t/problem-in-spliting-record/20/1 "2007-07-16T00:00:00Z")

</div>

hello  
i want to develope a component which split record into two different record

for example input is sales detail then one output is date and other is remaining data.

how to do this?  
how can i create new empy dataRecord and asign new data value to each of its field in my run methode?

---

<div class="post-metadata">

**Author:** ![oldforum](https://avatars.discourse-cdn.com/v4/letter/o/13edae/32.png) [@oldforum](https://forum.cloverdx.com/u/oldforum)\
**Post date:** [July 16, 2007, 8:23am UTC](https://forum.cloverdx.com/t/problem-in-spliting-record/20/2 "2007-07-16T08:23:44Z")

</div>

hello

i have created my SplitRecord class and modified component factory

then when i execute my testSplit graph i get exception that INPUT port has no out put port but there is in my graph.

i am providing full resources so

this is my graph  
[testSplit.grf]

\<?xml version="1.0" encoding="UTF-8"?\> 0 and sv.current\_version and sv.sales\_representativ=em.id and em.person\_master\_id=pm.id and sd.sv\_id=sv.id"/\> 

here what i get on console

[console dump]  
\*\*\* CloverETL framework/transformation graph runner ver1.7, (c) 2002-05 D.Pavlis, released under GNU Lesser General Public License \*\*\*

Graph definition file: E:/CloverETL/Examples/testSplit.txt  
phase No. 0  
phase No. 0  
java.lang.NullPointerException  
at org.jetel.graph.TransformationGraphXMLReaderWriter.instantiateNodes(TransformationGraphXMLReaderWriter.java:381)  
at org.jetel.graph.TransformationGraphXMLReaderWriter.instantiatePhases(TransformationGraphXMLReaderWriter.java:348)  
at org.jetel.graph.TransformationGraphXMLReaderWriter.read(TransformationGraphXMLReaderWriter.java:242)  
at org.jetel.main.runGraph.main(runGraph.java:124)  
Initializing DB connection: jdbc:postgresql://192.168.0.3/vtech0506 … OK  
Initializing DB connection: jdbc:postgresql://192.168.0.3/olap … OK  
[Clover] starting WatchDog thread …  
[WatchDog] Thread started.  
[WatchDog] Running on 1 CPU(s) max available memory for JVM 1639 KB  
[Clover] Initializing phase: 0  
initializing edges: all edges initialized successfully…  
initializing nodes:  
INPUT …FAILED !  
Nov 12, 2005 5:12:02 PM org.jetel.graph.Phase init  
SEVERE: At least one output port has to be defined for node INPUT  
[WatchDog] !!! Phase finished with error - stopping graph run !!!  
Failed starting graph !  
[Clover] WatchDog thread finished - total execution time: 0 (sec)  
[Clover] !!! Graph execution finished with errors !!!

here is my SplitRecord class

/\*  
\* Created on Nov 10, 2005  
\*  
\* TODO To change the template for this generated file go to  
\* Window - Preferences - Java - Code Style - Code Templates  
\*/  
package org.jetel.component;

/\*\*  
\* @author Deepak  
\*  
\* TODO To change the template for this generated type comment go to  
\* Window - Preferences - Java - Code Style - Code Templates  
\*/  
import java.io.IOException;

import org.jetel.data.DataRecord;  
import org.jetel.exception.ComponentNotReadyException;  
import org.jetel.graph.InputPort;  
import org.jetel.graph.Node;  
import org.jetel.graph.OutputPort;

/\*\*  
\* @author Deepak  
\*  
\* TODO To change the template for this generated type comment go to  
\* Window - Preferences - Java - Code Style - Code Templates  
\*/  
public class SplitRecord extends Node{

public static final String COMPONENT\_TYPE=“SPLIT\_RECORD”;

public SplitRecord(String id)  
{  
super(id);  
}

public String getType() {  
return COMPONENT\_TYPE;  
}

public void init() throws ComponentNotReadyException {  
// test that we have at least one input port and one output  
if (inPorts.size()\<1){  
throw new ComponentNotReadyException("At least one input port has to be defined for node " + id);  
}else if (outPorts.size()\<1){  
throw new ComponentNotReadyException("At least one output port has to be defined for node " + id);  
}  
}

public void run()  
{  
boolean isDriverDifferent;

// get all ports involved  
InputPort inPort = getInputPort(0);  
DataRecord driverRecordread = new DataRecord(inPort.getMetadata());  
driverRecordread.init();

int index=0;

// get array of all ports defined/connected - use collection Collection - getInPorts();  
OutputPort outPorts;  
outPorts = (OutputPort) getOutPorts().toArray(new OutputPort[0]);

System.out.println("the no. of out put pin in split is " + outPorts.length);

int cou = new int[5];  
Integer timeid=new Integer (1);

while (runIt)  
{  
try  
{  
driverRecordread = inPort.readRecord(driverRecordread);  
if(driverRecordread != null)  
{  
DataRecord record0 = new DataRecord(outPorts[0].getMetadata());  
record0.init();  
record0.getField(0).setValue(driverRecordread.getField(1).getValue());

DataRecord record1=new DataRecord(outPorts[1].getMetadata());  
record1.init();  
record1.getField(0).setValue(new Integer(1));  
record1.getField(1).setValue(driverRecordread.getField(0).getValue());  
record1.getField(2).setValue(timeid );  
record1.getField(3).setValue(driverRecordread.getField(2).getValue());  
record1.getField(4).setValue(driverRecordread.getField(3).getValue());  
record1.getField(5).setValue(driverRecordread.getField(4).getValue());  
record1.getField(6).setValue(driverRecordread.getField(5).getValue());  
record1.getField(7).setValue(“KG”);  
record1.getField(8).setValue(driverRecordread.getField(6).getValue());  
record1.getField(9).setValue(driverRecordread.getField(7).getValue());  
record1.getField(10).setValue(driverRecordread.getField(8).getValue());  
record1.getField(11).setValue(driverRecordread.getField(9).getValue());

outPorts[0].writeRecord(record0 );  
outPorts[1].writeRecord(record1 );

}else  
{  
break;  
}  
} catch (IOException ex)  
{  
resultMsg = ex.getMessage();  
resultCode = Node.RESULT\_ERROR;  
closeAllOutputPorts();  
System.out.println(“IO error in split:” + ex.getMessage());  
ex.printStackTrace();  
return;  
} catch (Exception ex)  
{  
resultMsg = ex.getMessage();  
resultCode = Node.RESULT\_FATAL\_ERROR;  
System.out.println(“error in split:” + ex.getMessage());  
ex.printStackTrace();  
return;  
}  
}

broadcastEOF();

if (runIt) {  
resultMsg = “OK”;  
} else  
{  
resultMsg = “STOPPED”;  
}  
resultCode = Node.RESULT\_OK;  
}

public boolean checkConfig() {  
return true;  
}

}

i made this entry in component factory class

registerComponent(SplitRecord.COMPONENT\_TYPE,“org.jetel.component.SplitRecord”);

so  
there is two thing wrong here

1. though there is output port in INPUT node  
the input node can’t initialise due to it not get any output port

2. graphNode.setPhase(phaseNum);  
instruction in org.jetel.graph.TransformationGraphXMLReaderWriter.instantiateNodes() function throws nullpontexception

if any body figure out problem then inform me

thanks in advance

---

<div class="post-metadata">

**Author:** ![oldforum](https://avatars.discourse-cdn.com/v4/letter/o/13edae/32.png) [@oldforum](https://forum.cloverdx.com/u/oldforum)\
**Post date:** [July 16, 2007, 8:23am UTC](https://forum.cloverdx.com/t/problem-in-spliting-record/20/3 "2007-07-16T08:23:45Z")

</div>

Creating a new data record is easy. Following example shows how to create metadata object first and then how to create data record based on this:

-----------------  
DataRecordMetadata metadata=new DataRecordMetadata(“TestInput”,DataRecordMetadata.DELIMITED\_RECORD);

metadata.addField(new DataFieldMetadata(“Name”,DataFieldMetadata.STRING\_FIELD, “;”));  
metadata.addField(new DataFieldMetadata(“Age”,DataFieldMetadata.NUMERIC\_FIELD, “|”));  
metadata.addField(new DataFieldMetadata(“City”,DataFieldMetadata.STRING\_FIELD, “\n”));

DataRecord record = new DataRecord(metadata);  
record.init();

// assign value to Name field:  
record.getField(0).setValue(“John Brown”);

// assign value to Age field:  
record.getField(“Age”).setValue(31);

-----------------  
Note: Unless you want to use the record  
to be output to text file, it does not matter  
whether you create DELIMITED\_RECORD  
and which delimiters you assign to fields - any  
character(s) will work.

Internally, the data (record) is stored as  
variable-length data structure. You may  
later redefine the (text) formatting by  
creating record with the same structure but  
different delimiters.

For info about custom components building, see "Step-by-step component building " on  
Clover’s web site. Documentation section.