Other components

This components perform various operations like data verification, command performing and so on.

Common attributes

Checks a defined foreign key against a table of primary keys to verify whether the foreign key is valid. If the foreign key is not found among primary keys, default foreign key is substituted. The resulting foreign record is broadcast to all connected outputs. The table containing the “primary” keys may contain duplicates but they will be ignored.

Input ports: * 0 - foreign records.

• 1 - primary records.

Output ports:

• [0] - foreign records with eventually new key
• [1] - (optional) foreign records with invalid key

Xml attributes:

Example:

  <Node id="CHECKFOREIGN" 
        type="CHECK_FOREIGN_KEY" 
        primaryKey="CustomerID"
        foreignKey="CustomerID"
        defaultForeignKey="-1"
  />
  <Node defaultForeignKey="-;-;" foreignKey="$city=$city;$country=$country;#" id="CHECK_FOREIGN_KEY0" type="CHECK_FOREIGN_KEY"/>

This component executes specified command (SQL/DML) against specified DB.

Input ports:

• one input port defined/connected (optional) - input records for stored procedure or (since 2.7 ver.) sql statements

Output ports:

• [0] (optional) - stored procedure output parameters and/or query result set
• [1] (optional) - logging error - fields with autofilling=“ErrCode” and/or autofilling=“ErrText” are filled by error code and/or error message of sql error respectively - since 2.6 version

Xml attributes:

Example:

  <Node id="DATABASE_RUN"
        type="DB_EXECUTE"
        dbConnection="NorthwindDB"
        sqlQuery="drop table employee_z"
  />
 
  <Node id="DATABASE_RUN" 
        type="DB_EXECUTE"
        dbConnection="NorthwindDB"
        errorActions="CONTINUE">
        <SQLCode>
            create table testTab (
                name varchar(20)
            );
            insert into testTab ('nobody');
            insert into testTab ('somebody'); 
        </SQLCode>
  </Node>
 
  <Node dbConnection="Connection1" id="DB_EXECUTE1" type="DB_EXECUTE">
    <attr name="sqlQuery">create table proc_table (
  	id INTEGER,
 	    string VARCHAR(80),
 	    date DATETIME
    );
    CREATE PROCEDURE SPDownload
      @last_dl_ts DATETIME
    AS
    BEGIN
      SELECT id, string, date
        FROM proc_table
           WHERE date >=  @last_dl_ts
    END;></attr>
 </Node>
 
<Node callStatement="true" dbConnection="Connection1" id="DB_EXECUTE2" inParameters="1:=$date" 
 			outputFields="id;string;date" type="DB_EXECUTE" sqlQuery="{call SPDownload(?)}" >
 
<Node dbConnection="Connection0" errorActions="MIN_INT=CONTINUE;" id="DB_EXECUTE0" printStatements="true" 
 			type="DB_EXECUTE" url="port:$0.field1:discrete"/>

This component is simple HTTP requestor. Connector is dedicated to generate HTTP requests and to receive responses from web servers.

Generally there are two ways how to specify a request content and how to handle a response:

1. single HTTP connect - If you need just one HTTP interaction, you can use 'Input file URL' for request content and 'Output file URL' as a storage for response. The request content can be also passed via 'Request content' attribute.
2. multiple HTTP interaction are possible through the use of input port and output port - 'Input field' attribute specifies field name of input record where is the request content. 'Output field' attribute specifies field name of outptu record where the reposonse will be stored. For each incoming record one HTTP interaction is perfomed and one output record sent. Both fields has to be string data field.
3. no combination of two previous approaches is possible

*) properties are accepted in key=value form i.e.

attribute1=value1
attribute2=value2
attribute3=value3

Example:

<Node guiName="HTTP connector" id="HTTP_CONNECTOR0" 
        outFileUrl="${DATAOUT_DIR}/result.txt" 
        requestMethod="POST" 
        type="HTTP_CONNECTOR" 
        url="http://almighty.pri.univie.ac.at/~mangler/helloService.php">
        <attr name="requestContent">
                <![CDATA[<?xml version="1.0" encoding="UTF-8"?>
                <soapenv:Envelope xmlns:soapenv="http://schemas.xmlsoap.org/soap/envelope/" xmlns:xsd="http://www.w3.org/2001/XMLSchema" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
                <soapenv:Body>
                <ns1:sayHelloTo soapenv:encodingStyle="http://schemas.xmlsoap.org/soap/encoding/" xmlns:ns1="urn:helloService">
                <name xsi:type="xsd:string">Agata</name>
                </ns1:sayHelloTo>
                </soapenv:Body>
                </soapenv:Envelope>]]>
        </attr>
</Node>

This component executes Java code. It dooesn't handle any input and output. This component is a wrapper around class implementing org.jetel.component.JavaRunnable interface, which contains Java code to be executed (method run()).

Input ports:

• none

Output ports:

• none

Example:

<Node id="JAVA_EXECUTE0"
      type="JAVA_EXECUTE"
      properties="testingProperty=Testing output string">
<attr name="runnable"><![CDATA[
import org.jetel.component.BasicJavaRunnable;
 
public class testJavaRunnable extends BasicJavaRunnable {
 
	public void run() {
		System.out.println("Basic example of Java Execute, value of testingProperty = " 
				+ parameters.getProperty("testingProperty"));
	}
}
]]></attr>
</Node>

Writes/reads records to/from lookup table. Depending on ports connected records are read put to lookup table and/or read from it.

Input ports:

• if no input port connected, read from lookup table and write to output ports, output port(s) must be defined/connected.
• if one input port connected, read from one input port and write to lookup table.

Output ports:

• if no output port connected, write to lookup table, input port must be defined/connected.
• if at least one output port connected, write to output port(s), read from lookup table.

Xml attributes:

Example:

  <Node id="READ" type="LOOKUP_TABLE_READER_WRITER" lookupTable="LookupTable1" />

since 2.4

This component runs specified graphs. Filenames of the graphs to be executed can be supplied by the input edge. The first field in input metadata must be String. In case the input is not connected a single filename must be specified as the attribute graphName.

Graph can be executed

1. in the same JVM: Graph is run by existing executor in current JVM instance. Log output is flushed together with current graph.
2. in separated JVM: Graph is run as external process. Log output is written into specified file. Executed JVM will have the same classpath as current JVM.
   In this mode, supplying the command line for clover is necessary (at least the -plugins argument).
   Command line arguments can be supplied by cloverCmdLineArgs attribute or by second field in input port (Supplying by field can be made only in in/out mode).
   If cloverCmdLineArgs attribute is defined and input port is connected then field from input port has higher priority. Thus when some graph defined in input port hasn't assigned command line arguments cloverCmdLineArgs attribute is used.
   When command line arguments is defined by input port then different command line arguments can be used for each graph.
   Note: using input port to defined command line arguments is implemented from version 2.5.

If you know that your graphs need more memory, set the alternativeJavaCmdLine attribute to “java -Xmx1g -cp” or equivalent. Remember that sameInstance must be set to false, otherwise this alternativeJavaCmdLine setting would be ignored.
If you set sameInstance to true, you should set the maximum needed memory in the Open Run Dialog before running the graph.
Remember that memory cannot be set as Command line argument.

Choosing of processing mode depend on attribute graphName. If attribute graphName is defined then pipeline mode is used else in-out mode is used.

In this mode graph to execute is specified by attribute graphName. If a input is connected then it's ignored. Each RunGraph node may be used as resolution node, because it chooses output port according to graph execution success/fail.

Input ports: no input

Output ports:

• port 0 - one record will be put on this port only if graph execution success
• port 1 - one record will be put on this port only if graph execution fails

In this mode input is connected and attribute graphName is not set. The first field in input metadata must be String containing path to graph file definition. Thus it executes graph for each incoming record.
If input metadata contains more then one filed then the second field must be String and it contains clover command line arguments.

Input ports: port 0 - first field of record must be String path to graph file definition.
Second field is optional - it must be String and contains clover command line arguments.

Output ports: port 0 - one record for each graph execution is generated; it may describe fail or success otherwise

Additional attributes for execution in separated instance of JVM

if sameInstance is true, all following attributes are ignored

Example:

<Node id="RUN1" type="RUN_GRAPH" graphName="${WORKSPACE}/graphXmlWriter.grf" />

<Metadata id="Metadata0">
<Record name="outdata" recordSize="-1" type="delimited">
<Field delimiter=";" name="graph" nullable="true" type="string"/>
<Field delimiter=";" name="result" nullable="true" type="string"/>
<Field delimiter=";" name="description" nullable="true" type="string"/>
<Field delimiter=";" name="message" nullable="true" type="string"/>
<Field delimiter="\n" name="duration" nullable="true" type="decimal"/>
</Record>
</Metadata>

Output metadata fields

since 2.6

Checks incoming records based on specified key whether all records are in the properly order. The key is a name (or combination of names) of field(s) from the input record. The sort order is either Ascending (default) or Descending.

Input ports: * one input port defined/connected.

Output ports:

• 0-n output ports defined/connected.

Xml attributes:

Example:

  <Node id="SEQUENCE_CHECKER0" 
        type="SEQUENCE_CHECKER"
        sortKey="LastName" />

The same as Simple Copy, but it waits specified amount of milliseconds before each record is sent to output. All records from input port are copied onto all connected output ports.

Input ports: * one input port defined/connected.

Output ports:

• at least one output port defined/connected.

Xml attributes:

Example:

  <Node id="SPEED_LIMITER0" type="SPEED_LIMITER" delay="2000" />

This component executes the specified command and arguments in a separate process and send output from process to output port. If there is specified command interpreter, command string is written to tmp file and there is called this interpreter to execute system script. If there is not connected output port, output from process (together with error) can be send to a file.

Input ports:

• one input port defined/connected (optional) - input records for system command (process' input).

Output ports:

• one input port defined/connected (optional) - output stream from system command (process' output).

Xml attributes:

Example:

  <Node id="SYS_EXECUTE0" type="SYS_EXECUTE">
        <attr name="capturedErrorLines">3</attr>
        <attr name="command">rm test.txt</attr>
  </Node>
 
  <Node id="SYS_EXECUTE0" type="SYS_EXECUTE">
        <attr name="interpreter">sh ${}</attr>
        <attr name="outputFile">data/data.txt</attr>
        <attr name="command">touch data/data.txt
                             cat
                             dir /home/username
        </attr>
  </Node>
 
  <Node id="SYS_EXECUTE0" type="SYS_EXECUTE">
        <attr name="interpreter">sh ${} /tmp</attr>
        <attr name="command">cat
                             ls -l $1
        </attr>
 </Node>
 
 <Node append="false" id="SYS_EXECUTE0" interpreter="sh ${}" outputFile="data-tmp/data.txt" type="SYS_EXECUTE">
       <attr name="environment">PATH=/mybin;MY_VARIABLE=my_value</attr>
       <attr name="deleteBatch">false</attr>
       <attr name="workingDirectory">/home/user/home/tmp</attr>
       <attr name="command">echo $PATH
                            echo $MY_VARIABLE
                            touch data/data.txt
                            cat
                            dir .
       </attr>
</Node>
 
When you want to use input records as arguments of your command:
<Node id="SYS_EXECUTE0" interpreter="xargs sh ${}" type="SYS_EXECUTE">
       <attr name="command">mv -t processed-in $@</attr>
</Node>

In following example, the record with file name is sent to System Execute component, that process' the file, creating a xml file as output. This resulting file is sent to XMLExtract component. You can send whole file (if it is small) as one record with following metadata:

<Metadata id="Metadata1" >
  <Record name="oneRecord" type="delimited">
    <Field eofAsDelimiter="true" name="data" nullable="true" type="string"/>
  </Record>
</Metadata>

<Metadata id="Metadata1" >
  <Record name="oneRecord" type="delimited" recordDelimiter="\n">
    <Field eofAsDelimiter="true" name="data" nullable="true" type="string"/>
  </Record>
</Metadata>