You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 28 Next »

Introduction 

The JITL PLSQLJob provides a parameterizable interface for executing Oracle® PL/SQL statements.

JS7 offers out-of-the-box capabilities:

  • for executing PL/SQL,
  • for passing arguments to PL/SQL,
  • for collecting and passing on the results of PL/SQL execution to subsequent jobs,
  • for using a JS7 - Credential Store.

The PLSQLJob can be used to execute files that include PL/SQL statements.

The job supports encryption of database credentials if used with a Hibernate configuration file, see JS7 - How to encrypt and decrypt Database Credentials.

Changes

  • Starting from Release 2.0.0 the PLSQLJob
    • is implemented on top of the JDBC Interface of the Oracle® JDBC Driver.
    • The Oracle® JDBC Driver performs a commit operation when a connection is closed without an explicit commit or rollback.
  • Starting from Release 2.7.1 the PLSQLJob is implemented using Hibernate
    • no commit or rollback operation is performed when the connection is closed.
  • Starting from Release 2.7.2 the PLSQLJob
    • makes use of the Hikari connection pool that ships with Hibernate,
    • performs a rollback operation when a connection is closed,
    • closes the connection in the event that a cancel/kill or suspend/kill operation is performed on the order.
    • JITL-743 - Getting issue details... STATUS

Usage

When defining the job either:

  • invoke the Wizard that is available from the job properties tab in the Configuration view and select the JITL PLSQLJob and respective arguments from the Wizard

or

  • specify the JITL job class and com.sos.jitl.jobs.db.oracle.PLSQLJob Java class name and add arguments as explained in the documentation below.

Example

Download: dbPLSQLExecution.json

The following example explains the basic use of the PLSQLJob. The job executes PL/SQL anonymous code blocks - selecting the current system date and writing it to the stdout channel.

Documentation

The Job Documentation including the full list of arguments can be found under: https://www.sos-berlin.com/doc/JS7-JITL/PLSQLJob.xml

The PLSQLJob supports the following arguments:

Name

Purpose

Required

Default Value

Example

command

PL/SQL statements to be executed

One of these arguments has to be specified


DECLARE
  return_cursor SYS_REFCURSOR;
BEGIN
  OPEN return_cursor FOR 'SELECT FIRST_NAME, LAST_NAME, AGE FROM EMPLOYEES';
END;

command_script_filePL/SQL statements from a file
/home/sos/some_script.sql

db_url

JDBC connection string

Either a DB URL, location of a Hibernate configuration file or a Credential Store reference is used.

User and password for database access can be specified by arguments with the same name and can be omitted if an Oracle Wallet is used or can be specified with the Hibernate configuration file, from a Job Resource or from a Credential Store Reference.


jdbc:oracle:thin:@localhost:1521:XE

db_user

User name for database access


scott

db_password

Password for database access


tiger
hibernate_configuration_fileReplaces db_url, db_user, db_password arguments from a hibernate configuration file.
./config/private/hibernate.cfg.xml

variable_parser_reg_expr

Regular expression to parse output from the DBMS_OUTPUT package and to set order variables for next jobs

no^SETs+(\\s)\\s*ISs(.*)$



result_set_as_variables


Specifies how a result set is forwarded to order variables for subsequent jobs:


  • false: no result set is processed (default)
  • column_value: for the first record of the result set from each column name a variable is created that holds the column value.
  • name_value:  the first two columns of the result set will be added to order variables. The values of the first column of each record included with the result set will become the variable names, the values of the second column will become the variable values. 


  • csv: the column names will be written to the first line of the file specified with the result_file argument, any column values of the result set will be written to subsequent lines of this file. Column names and values will be enclosed with double quotes and will be separated by commas.
    • Example:
    • "FIRST_NAME","LAST_NAME","AGE"
      "John","Doe","30"
  • xml: rows are created as a sequence of <ROW> elements of the <RESULTSET> root element. The column names will be created as XML element names. XML output is written to the file specified with the result_file argument. 
    • Example:
    • <?xml version="1.0" encoding="UTF-8"?>
      <RESULTSET>
        <ROW>
          <FIRST_NAME><![CDATA[John]]></FIRST_NAME>
          <LAST_NAME><![CDATA[Doe]]></LAST_NAME>
          <AGE><![CDATA[30]]></AGE>
        </ROW>
      </RESULTSET>

    • json: the column names will be written as an array of JSON objects to the file specified with the result_file argument, any column values of the result set will be written as object properties to this file. 
      • Example:
      • [
          {
            "FIRST_NAME" : "John",
            "LAST_NAME" : "DOE",
            "AGE" : 30
          }
        ]

no

result_fileSpecifies the path to a result file if the resultset_as_variables argument is used with one of the values csv, xml, json. An absolute path can be specified and a relative path can be used that starts from the Agent's working directory.no

credential_store_fileLocation of a credential store database (*.kdbx)false
./config/private/jobs.kdbx
credential_store_keyLocation of a credential store key file (*.key)false
./config/private/jobs.key


The PLSQLJob can be used with a credential store to hold sensitive arguments. For use of the credential_store_* arguments see JS7 - Use of Credential Store with JITL Jobs.

Argument: command

PL/SQL code can be specified as the value of the command argument.

The following PL/SQL code creates output that can be parsed by use of the variable_parser_reg_expr  argument to create workflow variables:

PL/SQL procedure returning parsable output
DECLARE   
  v_order_date DATE := SYSDATE; 
BEGIN      		    
  SELECT SYSDATE INTO v_order_date FROM DUAL;     
  DBMS_OUTPUT.PUT_LINE(' +++              +++');   
  DBMS_OUTPUT.PUT_LINE('SET order_date IS '|| v_order_date);  
  DBMS_OUTPUT.PUT_LINE(' +++              +++');   
END;


The following PL/SQL code creates a cursor that can be used to store the result set to a file when specifying the result_set_as_variables argument with one of the values csv, xml or json. In addition, the result_file argument is used to specify the file which selected data is written to.

PL/SQL procedure returning result set cursor
DECLARE
  return_cursor SYS_REFCURSOR;
BEGIN
  OPEN return_cursor FOR 'SELECT FIRST_NAME, LAST_NAME, AGE FROM EMPLOYEES';
END;

Argument: command_script_file

PL/SQL code can be used from from a file that is assigned this argument, for example:

./config/get_order.sql

Argument: db_url

The job requires a standard JDBC database connection string such as jdbc:oracle:thin:@localhost:1521:XE

Argument: db_user

This argument specifies the user account for a database schema which has been assigned the required privileges to execute the PL/SQL code.

Argument: db_password

This argument specifies the password for the user account given with the db_user argument.

Argument: variable_parser_reg_expr

This argument specifies a regular expression for parsing the output created by the the DBMS_OUTPUT package from the PL/SQL statements. For each match an order variable is created that is available for subsequent jobs. For example, the statement:

EXEC DBMS_OUTPUT.PUT_LINE( 'SET order_date IS ' || TO_CHAR(CURRENT_DATE, 'YYYY-MM-DD') );

writes the following output to the console:

SET order_date is 2021-05-04

The output will be parsed by the regular expression:

^SETs+(\\s)\\s*ISs(.*)$

and will result in an order variable:

order_date = "2021-05-04"

Subsequent jobs can make use of the $order_date variable.

Manage Arguments

Arguments from Job Resources

For example, database connection settings can be added to JS7 - Job Resources in order to be shared by a number of jobs. This allows central management of connections settings and limits visibility of passwords with individual database jobs.


The Job Resource is then assigned the job or workflow. As the same argument names have been used as for the Pl/SQL job they are automatically mapped to the job.

Passing Arguments to PL/SQL Code

Arguments can be passed to PL/SQL statements. Such arguments can be added to the PL/SQL code and will be replaced at run-time. The syntax for arguments includes stating ${VARIABLE} - such variables are replaced from any location in the PL/SQL script code. For example, a job argument or order variable with the name $dateFormat specifying the output format of a date can be added to PL/SQL script code like this:

Passing variables to PL/SQL
DECLARE 
    v_order_date VARCHAR2(16) := SYSDATE;       
BEGIN 
    /* pass variables to PL/SQL by using the syntax: ${VARIABLE} */
    SELECT to_char(SYSDATE, '${dateFormat}' )    
      INTO v_order_date       
      FROM DUAL;  

      DBMS_OUTPUT.PUT_LINE(' +++              +++');
      DBMS_OUTPUT.PUT_LINE('SET order_date IS '|| v_order_date);
      DBMS_OUTPUT.PUT_LINE(' +++              +++');
END;


The above example can be used:

  • for PL/SQL code included as the value of the command argument.
  • for PL/SQL code included with a script file that is referenced by the command_script_file argument.

Manage Return Values

Passing PL/SQL results to subsequent instructions and jobs

JS7 jobs can create return values that are available for later instructions and jobs in a workflow. The PLSQLJob can pass on the result of PL/SQL execution, e.g. calculated dates, by use of the DBMS_OUTPUT package.

This package allows writing to the stdout channel that is checked by the PLSQLJob for patterns which indicate return values. The output is parsed by the PLSQLJob using a regular expression that is specified by the argument variable_parser_reg_expr. This defaults to ^SETs+(\\s)\\s*ISs(.*)$.

The following line of Pl/SQL code:

Example how to specify a return variable
EXEC DBMS_OUTPUT.PUT_LINE( 'SET order_date IS ' || TO_CHAR(CURRENT_DATE, 'YYYY-MM-DD') );

results in output such as:

Example for a return variable
SET order_date IS 2021-05-04

The output will be parsed by the regular expression:

^SETs+(\\s)\\s*ISs(.*)$

and will result in an order variable:

$order_date = "2021-05-04"

Subsequent instructions and jobs can make use of the $order_date variable.

Return Variables automatically available from the PLSQLJob

This job automatically creates the following order variables which are available to subsequent instructions and jobs. 

Return Variable: sql_error

  • The sql_error order variable contains the error messages generated during PL/SQL execution. This variable will be empty if no errors occur.

Return Variable: std_out_output

  • The std_out_output order variable contains the messages spooled to the stdout channel by PL/SQL. 

Cancellation

The following operations are available:

  • The cancel operation is applied to the order, not to the PLSQLJob. In case of cancellation the job will continue until completion and the order will be put in the failed state.
  • The cancel/kill operation is applied to the order and to the job. The PLSQLJob will be killed, the database connection will be closed and the DBMS session will be terminated. The order will be put in the failed state.
    • The DBMS decides about rolling back transactions when closing the client session.

Further Resources


  • No labels