DB2 UDB V8.1 Family Application Development Certification:: User-Defined Routines

DB2 UDB V8.
1 Family Application Development

Certification: User-defined routines
Get ready for the exam
Skill Level: Introductory
Drew Bradstock (drewkb@ca.ibm.com)

Engagement Manager
08 Apr 2004
In this tutorial, you'll learn about using user-defined routines with the DB2 Universal
Database. You'll see how you can utilize programming logic within DB2 to simplify
and speed up your applications. This is the last tutorial in a series of seven that you
can use to help prepare for the DB2 UDB V8.1 Family Application Development
Certification exam (Exam 703).
Section 1. Before you start
What is this tutorial about?

In this tutorial, you'll learn about using user-defined routines with the DB2 Universal
Database. You'll see how you can utilize programming logic within DB2 to simplify
and speed up your applications. We'll discuss the following topics:
• User-defined functions and how to use them

• Stored procedures
• The DB2 Development Center tool
This is the seventh in a series of seven tutorials that you can use to help prepare for
the DB2 UDB V8.1 Family Application Development Certification exam (Exam 703).
The material in this tutorial primarily covers the objectives in Section 7 of the exam,
entitled "User-defined routines." You can view these objectives at:
http://www.ibm.com/certify/tests/obj703.shtml.
User-defined routines
© Copyright IBM Corporation 1994, 2007. All rights reserved. Page 1 of 47
developerWorks® ibm.com/developerWorks
You do not need a copy of DB2 Universal Database to complete this tutorial.
However, you can download a free trial version of IBM DB2 Universal Database from
the developerWorks downloads site for reference.
Who should take this tutorial?

To take the DB2 UDB V8.1 Family Application Development exam, you must have
already passed the DB2 UDB V8.1 Family Fundamentals exam (Exam 700). You
can use the DB2 Family Fundamentals tutorial series (see Resources) to prepare for
that test. It is a very popular tutorial series that has helped many people understand
the fundamentals of the DB2 family of products.
This tutorial is one of the tools that can help you prepare for Exam 703. You should
also review Resources at the end of this tutorial for more information.
This tutorial is designed to help you study for the user-defined routines section of the
DB2 Family Application Development exam. For this tutorial to be useful, you should
have a solid understanding of how SQL is used and how it works with a database.
An understanding of programming methodologies and flow of control usage would
also help. You should also have a base knowledge of:
• Databases
• Database objects
• Application compilation
Terminology review
Before you begin this tutorial, you need to be familiar with the concept of an SQL
access plan. This is a set of steps the DB2 engine uses to execute an SQL
statement. It includes the indexes that are used, the times at which fields are
retrieved from the database tables, and the order in which the steps of the query are
executed. The access plan is created by a database engine based on an SQL
statement sent to that engine.
We'll be using the term throughout this tutorial, so it's important to make sure you
understand it up front.
Notices and trademarks

Copyright, 2004 International Business Machines Corporation. All rights reserved.
IBM, DB2, DB2 Universal Database, DB2 Information Integrator, WebSphere and
WebSphere MQ are trademarks or registered trademarks of IBM Corporation in the
United States, other countries, or both.
Page 2 of 47 © Copyright IBM Corporation 1994, 2007. All rights reserved.
ibm.com/developerWorks developerWorks®
Other company, product, and service names may be trademarks or service marks of
others.
Section 2. Using user-defined functions
What are user-defined functions?

User-defined functions, or UDFs as they are commonly known, are functions defined
by the user aaqqthat run in the database engine. The functions are used to simplify
the logic within SQL statements. Instead of repeatedly including common or complex
logic in multiple statements, the developer can isolate that logic in a single function.
Traditionally, people think of a function as a program that takes in some inputs and
then returns a result. Within a database, you have a great deal more flexibility and
possibilities in what you can do with a function. There are a number of different types
of functions, and also different languages you can write them in.
Here are the types of available functions:
• SQL Scalar: Returns a single data type value using SQL

• SQL Row: Returns a row of data using SQL
• SQL Table: Returns an SQL table based on SQL
• External Scalar: Returns a single data type using an external language
• Sourced or Template: Are based on another function or template
And here are the programming languages that you can write them in:
• SQL (for functions written using one or more SQL statements)

• C/C++
• The Java language
• SQLJ
• Visual Basic
Function types
There are two distinct types of functions. The more common type is an internal
function, which is a function written using SQL statements. Functions can also be
written as references to external components of code, be they in the Java language,
C, or other languages. These functions are referred to as external functions. The

code for internal functions is stored in the database catalog tables, while external
functions are just references to locations where the compiled code is stored.
One important difference between the two types of functions is the location at which
you compile them. An SQL function can be compiled either remotely or locally, since
it will be stored in the database server's catalog tables. An external function,
however, is not as flexible. It must be compiled on the server, since the location of
the compiled code is then stored by the database server. If you compile it on a client,
then the code is stored on the client. The local client would be able to execute the
function, since the DB2 client would be able to find the program. However, the
server or any other clients would return an error, since the compiled code would not
be accessible to them. You could compile the function on the client and copy it to the
server at the exact same location to avoid this problem. But for that to work, the
operating system and underlying hardware need to be the same on both systems,
since the code would have been compiled by a specific compiler for the client
system and may not work with another computer/OS combination.
Creating a function
Internal functions are easy to create and use in your applications. A simplified
version of the syntax diagram is displayed below and is followed by an explanation
of its major parts.
>>-CREATE FUNCTION--function-name------------------------------->
>--(--+--------------------------------+--)--*------------------>
| .-,--------------------------. |
| V | |
'---parameter-name--data-type1-+-'
>--RETURNS--+-data-type2-----------------+--*------------------->
'-+-ROW---+--| column-list |-'
'-TABLE-'
.-LANGUAGE SQL-.
>--+-------------------------+--*--+--------------+--*---------->
'-SPECIFIC--specific-name-'
.-NOT DETERMINISTIC-. .-EXTERNAL ACTION----.
>--+-------------------+--*--+--------------------+--*---------->
'-DETERMINISTIC-----' '-NO EXTERNAL ACTION-'
.-READS SQL DATA---------.
>--+------------------------+--*--+-----------------+--*-------->
+-CONTAINS SQL-----------+
| |
'-MODIFIES SQL DATA------'
>--| SQL-function-body |--------------------------------------->
column-list:
.-,-----------------------.
V |
|--(----column-name--data-type3-+--)----------------------------|
SQL-function-body:
|--+-RETURN Statement-----------+-------------------------------|
'-dynamic-compound-statement-'
The major components of an SQL function are listed below. We'll look at some
examples on the next panel.
• Function name: Name of the function.

• RETURNS type : The type of function that is called. The types available
are scalar, row, or table . We'll look at each type in detail in Scalar
functions , Row functions , and Table functions. The scalar value is
actually a data type, not the word "SCALAR".
• Specific: You can assign the compiled function a specific name rather
than having DB2 assign a system-generated unique name for it.
• Deterministic: This specifies whether the function will return the same
result each time it is run. Deterministic functions include math functions
and functions that are not dependent on data in tables or changing data
sources.
• External Action: Specifies if the function has any impact on external
programs.
• Has SQL: Specifies how the function interacts with the database.
• Called on Null Input: Specifies if the function should be called if any of
the input values are null.
• SQL function body: This is the meat of the function, where the logic is
contained.
CREATE FUNCTION sections

This panel shows a number of examples to help you better understand what each of
the terms in the CREATE FUNCTION command mean.
RETURNS
The RETURNS specification controls the type of function that will be created. The
three major types are scalar, row, and table. A scalar function, like the following
example, only returns a single data type value:
CREATE FUNCTION TAN (X DOUBLE)

RETURNS DOUBLE
LANGUAGE SQL
CONTAINS SQL
NO EXTERNAL ACTION
DETERMINISTIC
RETURN SIN(X)/COS(X)
For more on scalar functions, see Scalar functions.
A row function, like the following example, separates a user-defined type into its
different components:
CREATE FUNCTION FROMPERSON (P PERSON)

RETURNS ROW (NAME VARCHAR(10), FIRSTNAME VARCHAR(10))
LANGUAGE SQL
CONTAINS SQL
NO EXTERNAL ACTION
DETERMINISTIC
RETURN VALUES (P..NAME, P..FIRSTNAME)
For more on row functions, see Row functions.
A table function, like the following example, will return zero or more rows of a table.
The table can be created from SQL or from application logic.
CREATE FUNCTION DEPTEMPLOYEES (DEPTNO CHAR(3))

RETURNS TABLE (
EMPNO CHAR(6),
LASTNAME VARCHAR(15),
FIRSTNAME VARCHAR(12))
LANGUAGE SQL
READS SQL DATA
NO EXTERNAL ACTION
DETERMINISTIC
RETURN
SELECT EMPNO, LASTNAME, FIRSTNME
FROM EMPLOYEE
WHERE EMPLOYEE.WORKDEPT = DEPTEMPLOYEES.DEPTNO
For more on table functions, see Table functions.
SPECIFIC
The SPECIFIC command is used to ensure that you have a specific name for the
function that allows you to remember what the compiled function is. This specific
name can then be referenced when you are sourcing, dropping, or commenting the
function. This is useful when you have overloaded the function and have more than
one version of the function with different data inputs.
The following two examples illustrate such overloading. The first function joins two
integer values by adding them together. The second joins the string new_ to the
input string.
CREATE FUNCTION joinData (x INT, y INT)

RETURNS DOUBLE
SPECIFIC join_int2
LANGUAGE SQL
CONTAINS SQL
NO EXTERNAL ACTION
DETERMINISTIC
RETURN x + y;
*******
CREATE FUNCTION joinData (x VARCHAR(10)
RETURNS VARCHAR(15)
SPECIFIC join_str
LANGUAGE SQL
CONTAINS SQL
NO EXTERNAL ACTION
DETERMINISTIC
RETURN 'new_' || x;
DETERMINISTIC
The DETERMINISTIC option is used to specify whether the function always returns
the same value. This information can then be used by DB2 to optimize how the
function is called, since DB2 would already know the value of the function if it had
been run once before and is deterministic. If a function uses any special registers, or
it calls any nondeterministic functions, it must be nondeterministic.
Here's an example of a deterministic function:
CREATE FUNCTION joinData (x INT, y INT)

RETURNS DOUBLE
LANGUAGE SQL
CONTAINS SQL
NO EXTERNAL ACTION
DETERMINISTIC
RETURN x + y;
And here's an example of a nondeterministic function:
CREATE FUNCTION futureDate (x INT)

RETURNS DATE
LANGUAGE SQL
CONTAINS SQL
NO EXTERNAL ACTION
NOT DETERMINISTIC
RETURN CURRENT DATE + x MONTHS;
EXTERNAL ACTION
This option specifies whether the function will change any objects outside the
database system. This option must be set to EXTERNAL ACTION if the function calls
any functions that have an external impact. An example would be a function that
modifies a flat file or alters data in an external source, such as a genome data file.
HAS SQL
This option lets DB2 know how the function is going to interact with the database.
You have a number of options here:
• CONTAINS SQL: Indicates that SQL statements that neither read nor
modify SQL data can be executed by the function.
• READS SQL DATA: Indicates that SQL statements that do not modify
SQL data can be executed by the function.
• MODIFIES SQL DATA: Indicates that all SQL statements supported in

the dynamic-compound statement can be executed by the function. This
option was added in DB2 V8.1 Fixpak 4. Previously, SQL functions could
not modify the data in the database.
Function rules
There are a number of restrictions on functions that you should be aware of when
dealing with DB2 SQL functions:
• If an SQL function contains multiple references to any of the date or time

special registers, all references return the same value. This will be the
same value returned by the register invocation in the statement that called
the function.
• The body of an SQL function cannot contain a recursive call to itself or to
another function or method that calls it, since such a function could not
exist to be called.
Privileges
When a function is created, the EXECUTE privilege is automatically given to the

definer of the function. The definer is also given the right to drop the function, even if
it does not have DBADM authority. Along with these authorities, the definer of the
function is given the WITH GRANT OPTION, which allows it to pass on the
authorities it has on the function to others. The function definer must have the
required privileges explicitly and cannot have them as result of being part of a
security GROUP.
Using compound statements in functions

What is a compound statement?
You are no longer limited to a single line of code for your SQL function logic. A
compound SQL statement is a group of statements contained within a BEGIN-END
block. The SQL statements are then treated as one atomic unit. This block of
statements can be dynamically prepared.
Here's the syntax diagram for a compound SQL block:
>>-+-------------+--BEGIN ATOMIC-------------------------------->
>--+-----------------------------------------+------------------>
| .-------------------------------------. |
| V | |
'---+-| SQL-variable-declaration |-+--;-+-'
'-| condition-declaration |----'
.-,--------------------------.
V |
>----SQL-procedure-statement--;-+--END--+-------+-------------->
SQL-variable-declaration:
.-,-----------------.
V |
|--DECLARE----SQL-variable-name-+--data-type-------------------->
.-DEFAULT NULL------------.
>--+-------------------------+----------------------------------|
'-DEFAULT--default-values-'
condition-declaration:
|--DECLARE--condition-name--CONDITION--FOR---------------------->
.-VALUE-.
.-SQLSTATE--+-------+-.
>--+---------------------+--string-constant---------------------|
The major parts of the compound statement are:
DECLARE: This allows you to declare variables within an SQL block. The data type
can be any user-defined type or standard SQL data type. If a default value is not
given for a data type, it is automatically set to null when it is declared. Here are
some examples:
DECLARE drewInt INTEGER;

DECLARE drewChar CHAR(6);
DECLARE drewI2 INTEGER DEFAULT 0;
DECLARE booBoo VARCHAR(100) DEFAULT NULL;
CONDITIONAL HANDLING: This option is currently not available for functions.
SQL control statements: The following statements can be used within a compound
statement:
• FOR
• GET DIAGNOSTICS
• IF
• ITERATE
• LEAVE
• SIGNAL
• WHILE
More information can be found on these statements in the DB2 online help.
The following SQL statements can be issued:
• fullselect (a common-table-expression can precede the fullselect)
• Searched UPDATE
• Searched DELETE
• INSERT
• SET variable statement
Working with stored procedures covers the different parts of SQL compound
statements and provides some more detailed examples. There are a number of
differences between the use of procedural SQL in functions and in procedures.
Scalar functions
The different function types offer you different ways to return data to your SQL
statements. It is important to understand how the function types can be used, and
also how easily complex functions can be written.
An SQL scalar function is a traditional function that is probably the first type you
think of when you hear the word function. It returns a single value in any of the data
types within DB2. This function is normally constructed using an SQL statement.
The simple example below illustrates how you can hide the logic of your programs in
a function. The function changeSal is created with a single line of real code:
RETURN sal * 2. The remainder of the lines are the function definition. The
function takes in as input the salary figure of an employee, or any other double field.
It also works with other number values, since DB2 will implicitly cast them for you.
Here's how you'd create your function:
CREATE FUNCTION changeSal (v_sal double)

RETURNS DOUBLE
LANGUAGE SQL
CONTAINS SQL
NO EXTERNAL ACTION
DETERMINISTIC
RETURN v_sal * 2;
And here's how you'd use it:
SELECT empno,
changeSal(salary) as newSalary
FROM employee
WHERE edlevel > 12;
Result from the DB2 sample database:
EMPNO NEWSAL
------ ----------------------
000030 87975.00
Scalar functions are normally more complicated than this, though, and can involve
full SQL statements. Let's create a more complex function:
CREATE FUNCTION edCount (v_edLevel double)

RETURNS INT
LANGUAGE SQL
READS SQL DATA
NO EXTERNAL ACTION
RETURN SELECT count(*)
FROM employee
WHERE edLevel = v_edLevel;
Now, let's use it in an SQL statement:
SELECT edLevel,
edCount(edLevel)
FROM employee
GROUP BY edlevel;
DB2 takes the SQL function and in-line it into the SQL statement in which the
function is called. When you in-line a function, the function call in the calling SQL
statement is actually replaced by the function text. As a result, the DB2 optimizer
can then create an optimized access plan for the entire statement and not just a
section of it. This leads to better overall access plans.
Let's look again at the first sample SQL statement:
SELECT empno,
changeSal(salary) as newSalary
FROM employee
WHERE edlevel > 12;
DB2 rewrites it so it looks like this:
SELECT empno,
sal * 2 as newSalary
FROM employee
WHERE edlevel > 12;
The second example's in-lining is much more complicated despite how simple the
initial SQL statement appears. This example clearly illustrates the benefit of moving
logic into the function instead of your own SQL statements. Here's the original SQL
statement again:
SELECT edLevel,
edCount(edLevel)
FROM employee
GROUP BY edLevel;
And here's what it looks like after DB2 rewrites it:
SELECT Q3.$C0 AS "EDLEVEL", Q6.$C0

FROM
(SELECT Q2.$C0
FROM
(SELECT Q1.EDLEVEL
FROM DB2ADMIN.EMPLOYEE AS Q1) AS Q2
GROUP BY Q2.$C0) AS Q3,
(SELECT COUNT(* )
FROM
(SELECT $RID$
FROM DB2ADMIN.EMPLOYEE AS Q4
WHERE (Q4.EDLEVEL = DOUBLE(Q3.$C0))) AS Q5)AS Q6
Row functions
With functions, you can also return an entire row of data, instead of just a single data
type. The row function does not return a row of data, as you might expect from its
name. It is instead used to transform structured data types into their individual
components. User-defined types, or UDTs, are data types defined by the user that
contain references to one or more DB2 data types. (UDTs are covered in detail in
the sixth tutorial in this series, see Resources.) You would therefore only use row
functions if you have UDTs in your database. Row functions can only be defined as
SQL functions, and you cannot use the other programming languages that are
normally allowed for DB2 functions.
The following example illustrates how you can use a row function to determine the
last name and first name of a person. The person object in the example is a UDT
that contains both the lastName and firstName fields. Our function can be used
to break apart the information you need from the person type.
Here's the command creating the person type:
CREATE TYPE person AS (

lastname varchar(15),
firstname varchar(12))
MODE DB2SQL
);
Next, create a table with the person information:
CREATE TABLE empInfo (

personInfo person,
empno char(6)
);
Then create the row function:
CREATE FUNCTION FROMPERSON (P PERSON)

RETURNS ROW (LNAME VARCHAR(15), FNAME VARCHAR(12))
LANGUAGE SQL
CONTAINS SQL
NO EXTERNAL ACTION
DETERMINISTIC
RETURN VALUES (P..LASTNAME, P..FIRSTNAME)
Table functions
One of the most powerful aspects of DB2 functions is their ability to return an entire
table of data instead of a single value. This opens up many sources of information
that you can use in your SQL statements. Instead of referring to a database table,
you could instead write a C function that referred to a live data stream such as stock
market data.
Table functions are actually quite easy to write and to reference. Instead of
referencing a single data value, as in scalar functions, you reference multiple
variables that refer to the different columns that will be returned in the table data, as
in the following example:
CREATE FUNCTION DEPTEMPLOYEES (DEPTNO CHAR(3))

RETURNS TABLE (EMPNO CHAR(6),
LASTNAME VARCHAR(15),
FIRSTNAME VARCHAR(12))
LANGUAGE SQL
READS SQL DATA
NO EXTERNAL ACTION
RETURN
SELECT EMPNO, LASTNAME, FIRSTNME
FROM EMPLOYEE
WHERE EMPLOYEE.WORKDEPT = DEPTEMPLOYEES.DEPTNO
There are restrictions to this: You must know the data types and number of
parameters that your SQL table will return. If you have a function that refers to the
entire table's columns and you add an additional column, then your function may not
work properly. For instance, suppose we create the following table:
CREATE TABLE testTab (

varOne int,
varTwo int
);
This function will work properly with that table:
CREATE FUNCTION returnAllTest (v_v1 int)

RETURNS TABLE (v_varOne INT,
v_varTwo INT)
LANGUAGE SQL
READS SQL DATA
NO EXTERNAL ACTION
RETURN
SELECT *
FROM testTab
WHERE varOne = v_v1;
But now we add a new column to the table:
ALTER TABLE testTab ADD varThree int;
Our function call will no longer return the expected value of all three columns, but will
instead return just the columns that were defined when it was created. This is
because the '*' used in the function definition is expanded out at creation time and
not at execution. Here's the SQL output for the modified table:
select * from testtab

VARONE VARTWO VARTHREE
----------- ----------- -----------
0 record(s) selected.
And here's the function output:
select * from table ( returnAllTest(1)) as test

V_VARONE V_VARTWO
----------- -----------
0 record(s) selected.
Section 3. Working with external functions
Using external functions

The ability to reference external programs as functions gives you more capabilities in
your SQL. By building functions in C or the Java language, you can perform very
complex string manipulation that would have required multiple (and confusing) SQL
statements to duplicate. You can also process outside data sources or perform
external actions outside of the database with these programming languages. For a
list of supported languages, see What are user-defined functions?.
In order to be able to create an external function, you must have at least one of the
following privileges:
• SYSADM or DBADM authority

• CREATE_EXTERNAL_ROUTINE authority on the database and at least one
of:
• IMPLICIT_SCHEMA authority on the database, if the schema name of
the function does not refer to an existing schema.
• CREATEIN privilege on the schema, if the schema name of the
function refers to an existing schema.
The functions can run either unfenced -- that is, inside the database engine -- or
fenced -- that is, in memory outside of the database engine. The advantage to
having an unfenced function is that it shares the memory area with the engine and
can communicate it with it more quickly. However, if an unfenced function has not
been written properly, a memory leak may occur. If this happens, the memory leak
could start overwriting the memory associated with the DB2 engine, which could
have horrible repercussions. If you are writing C or C++ functions it is highly
recommended that you write them as fenced.
To create an unfenced procedure, you will need one of the following privileges:
• CREATE_NOT_FENCED_ROUTINE authority on the database

• SYSADM or DBADM authority
Creating external functions

Here's a simplified syntax diagram for creating an external function:
>>-CREATE FUNCTION--function-name------------------------------->
>--(--+----------------------------------------------------+--)-->
| .-,----------------------------------------------. |
| V | |
'---+----------------+--data-type1--+------------+-+-'
'-parameter-name-' '-AS LOCATOR-'
>--*------------------------------------------------------------>
>--RETURNS--+-data-type2--+------------+------------------------+-->
| '-AS LOCATOR-' |
'-data-type3--CAST FROM--data-type4--+------------+-'
'-AS LOCATOR-'
>--*--+-------------------------+--*---------------------------->
>--EXTERNAL--+----------------------+--*------------------------>
'-NAME--+-'string'---+-'
'-identifier-'
>--LANGUAGE--+-C----+-------*----------------------------------->
+-JAVA-+
'-OLE--'
.-NOT DETERMINISTIC-.
>--PARAMETER STYLE--+-DB2GENERAL-+--*--+-------------------+---->
+-JAVA-------+ '-DETERMINISTIC-----'
'-SQL--------'
.-FENCED------------------------.
>--*--+-------------------------------+--*---------------------->
+-FENCED--*--+-THREADSAFE-----+-+
| '-NOT THREADSAFE-' |
| .-THREADSAFE-. |
'-NOT FENCED--*--+------------+-'
.-RETURNS NULL ON NULL INPUT-. .-READS SQL DATA-.
>--+----------------------------+--*--+----------------+--*----->
'-CALLED ON NULL INPUT-------' +-NO SQL---------+
'-CONTAINS SQL---'
.-EXTERNAL ACTION----.
>--+-----------------+--*--+--------------------+--*------------>
'-NO EXTERNAL ACTION-'
.-NO SCRATCHPAD----------. .-NO FINAL CALL-.
>--+------------------------+--*--+---------------+--*---------->
| .-100----. | '-FINAL CALL----'
'-SCRATCHPAD--+--------+-'
'-length-'
.-NO DBINFO-.
>--+-------------------+--*--+-----------+--*------------------->
+-ALLOW PARALLEL----+ '-DBINFO----'
'-DISALLOW PARALLEL-'
We explored the majority of the components of the FUNCTION command in Creating

a function. There are a few new components that are particular to external functions,
and we'll discuss them here.
• Language: This identifies the language that the function is written in.
• Parameter style: This clause is used to specify the conventions used for
passing parameters to and returning the value from functions.
• External action: This determines if external actions can be performed by
the function.
• Scratchpad: The scratchpad is used as a memory area where data can
be stored by the function.
• Final call: This is used to deallocate memory that is being utilized by the
function.
Major external function options

One of the most important mandatory parameters in an external function is the
EXTERNAL NAME option. This controls where DB2 will look for the compiled function
code.
C functions
In C, the portion of the command is:
>>-'--+-library_id-------+--+------------+--'------------------>|
'-absolute_path_id-' '-!--func_id-'
The different components of it are as follows:
• library_id: This is the location of the library containing the function. The
database manager will look for the library in the .../sqllib/function
directory (on UNIX-based systems) or the
...\sqllib\instance_name\function directory (on Windows
operating systems), as specified by the DB2INSTPROF registry variable.

The library_id will then be appended onto the function directory location.
For example, if drewFunc were the library ID, then the database
manager would look for the function in library
/u/production/sqllib/function/drewFunc, assuming the
instance was installed in /u/production.
• absolute_path_id: Identifies the full path name of the file containing the
function.
• ! func_id: Identifies the entry point name of the function to be invoked.
The ! serves as a delimiter between the library ID and the function ID. In
a UNIX-based system, for example, mymod!funcDrew directs the
database manager to look for the library
$inst_home_dir/sqllib/function/mymod and to use entry point
funcDrew() within that library. Under Windows operating systems,
mymod!funcDrew directs the database manager to load the mymod.dll
file and call the funcDrew() function in the dynamic link library (DLL).
Java functions
The syntax in the Java language for this command is:
>>-'--+----------+--class_id--+-.-+--method_id--'-------------->|
'-jar_id :-' '-!-'
The different components are:
• jar_id: Identifies the JAR identifier given to the JAR collection when it was
installed in the database. It can be either a simple identifier, such as
myJar, or a schema-qualified identifier, such as mySchema.myJar.
• class_id: Identifies the class identifier of the Java object. If the class is
part of a package, the class identifier part must include the complete
package prefix. For example, if you use myPacks.UserFuncs, the Java
Virtual Machine will look in the directory .../myPacks/UserFuncs/ (or
...\myPacks\UserFuncs\ on Windows) for the classes.
• method_id: Identifies the method name of the Java object to be invoked.
OLE functions
The OLE syntax for this command is:
>>-'--+-progid-+--!--method_id--'------------------------------>|
'-clsid--'
The different components are:
• progid: Identifies the programmatic identifier of the OLE object.

• clsid: Identifies the class identifier of the OLE object to create. It can be
used as an alternative for specifying a progid in case an OLE object is
not registered with a progid. The clsid has the form
{nnnnnnnn-nnnn-nnnn-nnnn-nnnnnnnnnnnn}, where n is an
alphanumeric character.
• method_id: Identifies the method name of the OLE object to be invoked.
Function definition examples

Let's start with an example in C. The example below calls as input a single
SMALLINT parameter and then returns as output a SMALLINT field as well. No SQL
is called during the execution of the function.
CREATE FUNCTION ntest1 (SMALLINT)

RETURNS SMALLINT
EXTERNAL NAME 'ntest1!nudft1'
LANGUAGE C PARAMETER STYLE SQL
DETERMINISTIC NOT FENCED NULL CALL
NO SQL NO EXTERNAL ACTION
Now let's try a Java language example. The following example calls the findc
function. No SQL is called in the function.
CREATE FUNCTION findc ( CLOB(100K))

RETURNS INTEGER
FENCED
LANGUAGE JAVA
PARAMETER STYLE JAVA
EXTERNAL NAME 'javaUDFs.findCnst'
NO EXTERNAL ACTION
CALLED ON NULL INPUT
DETERMINISTIC
NO SQL
Binding a function
If you create an external function that executes SQL, you will have to PREPARE and
BIND the application that will be connected to the database.
The PREPARE command checks that the embedded static SQL in your application is
properly formed. For more details on this command, check the DB2 online help
under "PRECOMPILE Command."
The BIND command invokes the bind utility, which prepares SQL statements stored
in the bind file generated by the precompiler and creates a package that is stored in
the database. Further information on this step can be found in the DB2 online help
under "BIND Command."
Let's look at an example. The following C program retrieves the application ID for the
application that calls it. The program simply retrieves the appl_id field from the
dbinfo data structure. This information cannot be extracted using SQL, so the
function can be useful. Here's the C code
#include <string.h>
#include <sqludf.h>
void SQL_API_FN getApplicationId(
SQLUDF_CHAR *applId, SQLUDF_NULLIND *applId_ind,
SQLUDF_TRAIL_ARGS, SQLUDF_DBINFO *dbinfo)
{
strncpy(applId, dbinfo->appl_id, 128);
*applId_ind = 0;
}
To transform this function from C code to a state where it can be called in SQL, you
need to walk through the following steps:
1. Connect to the database
2. PREP the .sqc function file
3. BIND the .bnd function file
4. Compile the .c function file
5. Create the external function in DB2
The next panel covers these steps in detail.
Binding a function: Example

This example script will prepare and bind our sample function from the previous
panel. The script will run on a Windows system, but a C compiler must first be
installed and configured. (See Creating a stored procedure and Resources for more
on C compilers.)
@ echo off
rem ------------------------------------------
rem Script to create the C function
rem crFunc.bat databaseName [user [pwd]]
rem ------------------------------------------
rem Default compiler is set to Microsoft Visual C++
rem Microsoft C/C++ Compiler
set BLDCOMP=cl
set PROG=application_Id
rem call sdkvars
rem Precompile and bind the program.
rem Connect to a database.
if "%1" == "" goto error
if "%2" == "" goto case1

goto error
:case1
db2 connect to %1
if NOT %ERRORLEVEL% == 0 GOTO connectError
goto prep
:case2
db2 connect to %1 user %2
goto prep
:case3
db2 connect to %1 user %2 using %3
goto prep
:prep
db2 prep %PROG%.sqc bindfile
:bind
rem bind step
db2 bind %PROG%.bnd
if NOT %ERRORLEVEL% == 0 GOTO errorBind
db2 connect reset
rem Compile the program.
%BLDCOMP% -Zi -Od -c -W2 -DWIN32 -MD %PROG%.c
if NOT %ERRORLEVEL% == 0 GOTO errorCompile
rem Link the program.
link -debug -out:%PROG%.dll -dll %PROG%.obj db2api.lib -def:%PROG%.def
if NOT %ERRORLEVEL% == 0 GOTO errorLink
rem Copy the routine DLL to the 'function' directory
copy %PROG%.dll "%DB2PATH%\function"
if NOT %ERRORLEVEL% == 0 GOTO errorCopy
goto exit
:connectError
echo The connection to the database failed. Please check the database,
echo username and password.
goto :error
:errorBind
db2 connect reset
echo The application failed to prep and bind the program to the database.
goto :error
:errorCompile
echo The program failed to compile. Please check that the sdkvars program has
echo been run.
goto :error
:errorLink
echo The program failed to link the files together.
goto :error
:errorCopy
echo The dll file could not be copied to the %DB2PATH%\FUNCTION directory.
goto :error
:error
echo Usage crFunc.bat databasename [user [pwd]]
:exit
@echo on
Once you've run this script, the function can be created in the database:
CREATE FUNCTION application_id()

RETURNS VARCHAR(128)
SPECIFIC applId EXTERNAL NAME 'application_id!getApplicationId'
NOT FENCED LANGUAGE C PARAMETER STYLE SQL DETERMINISTIC
NO SQL NO EXTERNAL ACTION ALLOW PARALLEL DBINFO
The function can now be called in any SQL statement, like so:
VALUES(application_id())
Section 4. Working with stored procedures
What are stored procedures?

Originally, stored procedures were complex C programs used to process and
execute multiple SQL commands on the database server. The creation of these
programs was quite complex and required in-depth knowledge of C/C++
programming and database interface languages such as the DB2 Call Level
Interface (CLI).
As databases evolved, the stored procedures languages grew into procedural

languages using SQL. DB2 uses a language called SQL PL, Oracle uses PL/SQL,
and SQL Server and Sybase use Transact-SQL (or T-SQL for short). These
languages are quite different, but they all follow a common framework of keywords,
control statements and behaviour. There is a section in the SQL standard that
specifies the accepted keywords and behaviour for a stored procedure language.
Each RDBMS system has implemented this standard differently, so it is important to
understand that moving between them will not necessarily be an easy task.
The stored procedures covered in this section are all written using the DB2 SQL
Procedure Language, known as SQL PL. Stored procedures can be written in a
variety of languages, including C, the Java language, and OLE, however, it is easier
to illustrate how they can be used if simple SQL language examples are used
instead. This is the language we'll use in this section.
Why use stored procedures?

You may want to use stored procedures in your applications for a number of
reasons:
• Reduced network traffic: Because the stored procedures send only the
procedure call and the final result across the network, not all of the interim
data sets need to be sent back to the client. For applications where the
client and the server are geographically distant, this can dramatically
improve performance.
• Improved performance: The stored procedures will be executed on the
server instead of on the client. In the vast majority of architectures, the
server is a powerful machine that can process the SQL much quicker than
the client can.
• Reduced complexity: By placing all of the database application code in
modular units, you have isolated the application's interaction with the
database to a single layer. This allows for easier testing and debugging
as well as maintenance.
• Division of responsibilities: Having all of the data interaction isolated to
the stored procedures allows the interactions to be controlled by the DBA.
The DBA is the one who knows the system best, and he or she will be
able to respond the quickest to any performance problems.
• Ease of use: Programming applications that process data using stored
procedures is quite simple and allows business logic to be easily
translated into application logic. By leaving the application in the data
format that accesses the database, less application programming skill is
needed.
When to use -- and not to use -- stored procedures

The previous panel noted a number of benefits of using stored procedures. There
are, however, limitations on using stored procedures.
• Stored procedures can return any DB2 system data type. The return
value of the stored procedure, however, can only be an integer.
• A stored procedure cannot be called within an SQL statement.
• Stored procedures cannot be called within triggers or functions.
• The length of a procedure is limited to 64 KB of text.
Unfortunately, the workaround for using stored procedure logic in functions and
triggers is to embed the entire stored procedure logic in the function or trigger.
Stored procedures are most efficient when they are used to encapsulate a large
section of related logic. You should treat your stored procedures just like application
programs. If a section of code is repeated often throughout your application or
procedures, then it probably makes sense to encapsulate it into a stored procedure.
However, if the logic is only called once, then it makes sense to leave it as code.
You should also avoid writing very large stored procedures that perform multiple
tasks, unless the stored procedure is clearly defined as doing so. You may want to
put all of your database logic into a stored procedure. If you have a simple call to the
database to return a result set, then creating a stored procedure for this task would
be overkill. It would be better to open a cursor directly within your application.
Creating a stored procedure

Creating a stored procedure is not a difficult process. First, you need to have a C
compiler installed on your system. Stored procedures are currently translated from
DB2 SQL PL into C code and then compiled as binaries. Without the compiler, you
will not be able to create any new procedures, though you will be able to take
existing stored procedures from other systems and install them on your current
server. (You can do this using the GET ROUTINE and PUT ROUTINE commands.
More detail about the commands and the restrictions on moving procedures is
provided in the DB2 online documentation.)
You can use existing free or open source compilers for compiling the stored
procedures. The two most common compilers used are the GCC compiler and the
trial edition of the Microsoft Visual C++ compiler (see Resources). Microsoft
currently has a trial version of its compiler available that can be used for compiling
stored procedures. This is for development purposes only. A full license must be
purchased if you are going to use it in production.
Creating a procedure
Once you have a compiler installed, you are ready to create a procedure. The syntax
is similar to that for creating a function.
>>-CREATE PROCEDURE--procedure-name----------------------------->
>--+----------------------------------------------------+--*---->
'-(--+------------------------------------------+--)-'
| .-,------------------------------------. |
| V .-IN----. | |
'---+-------+--parameter-name--data-type-+-'
+-OUT---+
'-INOUT-'
>--+-------------------------+--*------------------------------->
.-DYNAMIC RESULT SETS 0--------. .-MODIFIES SQL DATA-.
>--+------------------------------+--*--+-------------------+--->
'-DYNAMIC RESULT SETS--integer-' +-CONTAINS SQL------+
'-READS SQL DATA----'
.-NOT DETERMINISTIC-.
>--*--+-------------------+--*----------------------------*----->
'-DETERMINISTIC-----'
.-LANGUAGE SQL-.
>--+----*--+--------------+--*-------->
>--| SQL-procedure-body |-------------------------------------->|
SQL-procedure-body:
|--SQL-procedure-statement--------------------------------------|
The major components of the command are:
• Parameter type: There are three parameter types. IN is used for input
parameters. Changes to these parameters are not passed back to the
calling application. OUT is used for output parameters. Changes to these
parameters are passed back out to the calling application. INOUT is used
for input and output. Changes to these parameters and their input values
affect the stored procedure and the calling application.
• Specific: A specific name can be assigned to label a stored procedure
internally. DB2 assigns a system-generated value if a name is not
specified.
• Dynamic Result Sets: Specifies the number of results sets that will be
left open and passed back to the calling program or application.
• Deterministic: Specifies if the stored procedure will return the same
result set if the same input is given.
• Language: Currently, only the DB2 SQL PL language is supported for
SQL procedures.
SQL statements in routines

There are four settings for stored procedures that specify how they interact with the
database and what SQL they use.
• NO SQL: Indicates that no SQL is used in the procedure.

• CONTAINS SQL: Indicates that SQL statements that neither read nor
modify SQL data can be executed by the stored procedure.
• READS SQL: Indicates that some SQL statements that do not modify SQL
data can be included in the stored procedure.
• MODIFIES SQL: Indicates that the stored procedure can execute any
SQL statement except statements that are not supported in stored
procedures.
If a statement invokes a routine, the effective SQL data access indication for the
statement will be the greater of:
• The SQL data access indication of the statement from the following table.
• The SQL data access indication of the routine specified when the routine
was created.
The table below is taken from the DB2 documentation:
SQL Statement NO SQL CONTAINS SQL READS SQL MODIFIES SQL

DATA DATA
COMPOUND SQL N Y Y Y
CONNECT N N N N
CREATE N N N Y
DECLARE Y Y Y Y
CURSOR
DECLARE N N N Y
GLOBAL
TEMPORARY
TABLE
DELETE N N N Y
DESCRIBE Y Y Y Y
DISCONNECT N N N N
DROP ... N N N Y
END DECLARE Y Y Y Y
SECTION
EXECUTE N Y Y Y
EXECUTE N Y Y Y
IMMEDIATE
EXPLAIN N N N Y
FETCH N N Y Y
FREE LOCATOR N Y Y Y
FLUSH EVENT N N N Y
MONITOR
GRANT ... N N N Y
INCLUDE Y Y Y Y
INSERT N N N Y
LOCK TABLE N Y Y Y
OPEN N N Y Y
PREPARE N Y Y Y
REFRESH TABLE N N N Y
RELEASE N N N N
CONNECTION
RELEASE N N N Y
SAVEPOINT
RENAME TABLE N N N Y
REVOKE ... N N N Y
ROLLBACK N N N N
ROLLBACK TO N N N Y
SAVEPOINT
SAVEPOINT N N N Y
SELECT INTO N N Y Y
SET N N N N
CONNECTION
SET INTEGRITY N N N Y
SET special N Y Y Y
register
UPDATE N N N Y
VALUES INTO N N Y Y
WHENEVER Y Y Y Y
Creating stored procedures

This tutorial does not cover all of the semantics of creating a stored procedure.
What's most important to understand is how a stored procedure can be created from
the command line. Normally a script can be executed using the following syntax:
db2 -tvf fileName
This syntax will not necessarily work for your stored procedure, depending on how
line endings are handled. The different options for the db2 command that are used
with running scripts are:
• t: Indicates that ; is to be used to indicate the end of a line. This allows

SQL statements to be spread over more than one line in the script. If a
character token is put after the -t option, it will be the token used for line
delimitation.
• v: Turns on verbose mode. All of the script commands will be repeated in
the script output. This is useful if you are trying to identify a failed
command. When a script file becomes large, debugging can be a quite
time consuming and difficult process without verbose mode on.
• f: Specifies the name of input file.
• d: Specifies that the token for the end of command is separate from the
token for end of line. This is essential for stored procedures, as you need
to be able to differentiate between the end of a line and the end of the
procedure.
Let's look at an example procedure:
CREATE PROCEDURE drewProc1 (IN p_varOne INT,

OUT p_varTwo INT)
LANGUAGE SQL
BEGIN
DECLARE v_varUno INT DEFAULT 0; -- (1)
SET v_varUno = 1 + p_varOne;

SET p_varTwo - v_varUno * p_varOne;
END
@ -- (2)
If the script is compiled using the command db2 -tvf fileName, DB2 will stop at
(1) and treat it as one command, resulting in an error. To avoid this problem, you
have to compile it with the command db2 -td@ -f fileName. DB2 will then treat
the ; character at (1) as an end-of-line character and continue processing the
procedure. Only when the @ character is reached at (2) will DB2 stop processing it
as one entire procedure.
Stored procedure examples

Below are a couple of simple examples of stored procedures. The first example
could have been written as a function as well, since it only returns the count of the
number of employees of a certain education level.
CREATE PROCEDURE empCount (IN p_edLevel,

OUT p_edCount)
LANGUAGE SQL
BEGIN
DECLARE c_cnt CURSOR FOR
SELECT count(*)
FROM employee
WHERE edlevel = p_edLevel;
OPEN c_cnt;
FETCH FROM c_cnt INTO p_edCount;
END
The second example is more complex and expands on the first. It checks how many
employees are in a work department and gives them a larger budget if they have
less than four people.
CREATE PROCEDURE increaseSize (IN p_size DECIMAL)

LANGUAGE SQL
BEGIN
-- Declare returncode
DECLARE SQLSTATE CHAR(5);
-- Declare variables
DECLARE v_deptCount INT DEFAULT 0;
DECLARE v_dept CHAR(3) DEFAULT '';
-- Declare the cursor
DECLARE c_cnt CURSOR FOR
SELECT workdept,count(*)
FROM employee
group by workdept;
-- Open the cursor
OPEN c_cnt;
-- Fetch from the cursor until it is empty
FETCH FROM c_cnt INTO v_dept, v_deptCount;
WHILE ( SQLSTATE = '00000' ) DO

-- Check if the department has four or less people
IF v_deptCount < 5 THEN
-- update the salary if they do
UPDATE employee
SET salary = salary * (1 + p_size)
WHERE workdept = v_dept;
END IF;
FETCH FROM c_cnt INTO v_dept, v_deptCount;
END WHILE;
END
@
Section 5. Using the DB2 Development Center
What is the DB2 Development Center?

Creating and debugging stored procedures can be a difficult task, as most common
IDEs don't support SQL. To overcome this problem, IBM created the DB2
Development Center, known in DB2 UDB V7 as the DB2 Stored Procedure Builder.
A great deal of functionality has been added to the Development Center (commonly
referred to as the DC) since its initial release.
The Development Center is included with DB2 as a free tool. You can run it by
selecting it from the IBM DB2 options in the Windows Start Menu:
You can also run it from within any of the other DB2 GUI tools, such as the DB2
Control Center.
If you are going to be using the DC from a UNIX or Linux client, then you must have
the X Window System or equivalent software installed to ensure that the GUI will be
displayed.
Setting up a project
When you first open the Development Center, you will be asked to create a project.
The project stores all the information about the databases you have connections to,
the procedures you are working on, and the status of your project. Previously, the
Stored Procedure Builder would display all of the procedures that you had in the
database. By being able to pick which ones you want to work with, projects can be
handled much more clearly.
The project needs an initial connection to a database. The list of databases provided
to you is taken from the databases that you have catalogued on your local client. If
the required database is not in the list, you can add it by selecting the Add option.
This catalogues the new database on the client.
Creating an object
An SQL PL procedure isn't the only kind that you can build in the Development
Center. You can also work with Java procedures, though only in a project working
with user-defined functions. We'll cover the use of Java procedures in Advanced use
of the Development Center .
If you want to build a new procedure, the DC walks you through all of the options
step by step. The Object Building wizard is simple enough that even novices to SQL
procedures can build a procedure, but also complex enough that experts can build
fully functional stored procedures. The figure below shows the different options for
the header of the stored procedure.
You should also give the procedure a specific name in case any overloading is used.
(The use of specific names with stored procedures is discussed in Creating a
procedure.)
Using the Development Center

There are four main views in the Development Center:
• Project view: Lists all of projects that are currently open, the databases
they have connections to, and the stored procedures that are part of a
particular project.
• Server view: Summarizes the items based on database connections and
includes information on stored procedures, triggers, functions, tables, and
views.
• Output view: Lists the status of the actions you have just completed or
are working on.
• Editor view: Displays objects for editing.
The following figure shows you an overall picture of the Development Center
interface.
Changing the DC interface

The DC is designed to be very flexible and adaptable. None of the locations of the
menus or their sizes is fixed. If you hold your mouse over the dark blue menu line for
any of the views, a hand-shaped icon appears. You can use this icon to drag the
menu option around the screen.
The DC can display a view in a number of ways:
• Outside of the DC: The view can be dragged outside the DC to become
a stand-alone object. This is useful for viewing a list of objects in a
separate frame, which reduces the clutter of the Development Center.
• Tab: The views can also be lined up as tabs. All of the information can be
quickly viewed but still remain hidden when you don't need to see it.
Selecting this option can be a little tricky. You have to select the view with
the hand icon and move it over to the edge of another view. Move the
object around the edge until you see an object appear that resembles a
set of file tabs. You can then let go and the views will be combined into a
set of tabs. Two or more views can be combined this way.
The views can also be reset to the default settings by selecting

View=>Reset Views from the menu.
Working with stored procedures

The stored procedure editing window contains all the information that you need to
edit, build, and debug your procedures. The main editor window is a graphical IDE
that highlights the code based on the syntax. This makes writing the procedures
easier.
The individual icons used for building and debugging the procedures are covered in
the following panels.
Building a procedure
There are two main modes that the stored procedures can be built in: normal build
and build for debug. The build for debug option allows you to debug the procedure. It
automatically inserts all the required breakpoints and step-through information into
the procedure. It is therefore recommended that for a production environment you
ensure that your stored procedures have been built in normal mode. The debug
mode significantly slows the performance of the procedures.
• Build (1): This command builds the stored procedure in regular mode.
• Run (2): This command executes the stored procedure.
• Build in Debug (3): This command builds the stored procedure in debug
mode.
Debugging a stored procedure

One of the most important functions of the DC is the ability to step through your
procedures in debug mode. This allows you to find your errors and keep track of how
variables are changing. Once you have built your stored procedure in debug mode,
you can add breakpoints to the code.
• Add Breakpoint (1): This adds a breakpoint to a line of code.

• Toggle Breakpoint (2): This changes how the breakpoints are handled,
changing it from enabled to disable or vice versa.
• Remove breakpoint (3): This removes the breakpoint.
• Remove all breakpoints (4): This removes all breakpoints.
• Run in Debug (5): This runs the stored procedure in debug mode.
Not all of the commands support breakpoints. If you try to run a program in debug
mode and do not have any breakpoints, or have breakpoints on incorrect
commands, the program will run normally and you will not be able to step through it.
Category of statements Statement

Statements that accept breakpoints ALLOCATE CURSOR, ASSOCIATE
LOCATORS, CASE (EXPRESSION), COMMIT,
CREATE PROCEDURE, CREATE {table,view,
index}, DEFAULT (VALUE), DROP {table,view,
index}, ELSEIF (EXPRESSION), EXECUTE,
EXECUTE IMMEDIATE, FETCH {..}, INTO, FOR
v1 AS {SQLstmt}, GET DIAGNOSTICS,
GOTO(LABEL), IF (EXPRESSION),
RETURN(value), SELECT {..} INTO, SET
(EXPRESSION), UNTIL (EXPRESSION), WHEN
(VALUE), WHILE (EXPRESSION)
Statements that change variables CALL FETCH {..}, INTO, GET DIAGNOSTICS,
SELECT {..} INTO, SET
Statements that do nothing BEGIN, BEGIN NOT ATOMIC, BEGIN ATOMIC,
CLOSE CURSOR, DECLARE cursor, WITH,
RETURN, FOR {sql statement}, DECLARE , var
without default, DECLARE CONDITION
(CONDITION) FOR SQLSTATE (VALUE) "..." ,
DECLARE CONTINUE HANDLER, DECLARE
CURSOR, DECLARE EXIT HANDLER,
DECLARE RESULT_SET_LOCATOR
[VARYING], DECLARE SQLSTATE, DECLARE
SQLCODE (unless there is a default), DECLARE
UNDO HANDLER (unless they are entered), DO,
ELSE END, END, CASE END, IF END, FOR
END, REPEAT END, WHILE, ITERATE, LEAVE,
LOOP, OPEN CURSOR, REPEAT (as a keyword
alone), RESIGNAL, SIGNAL, THEN labels, i.e.
P1:
Section 6. Advanced use of the Development Center
Using the Java language with stored procedures

The development center allows you to work with stored procedures written in the
Java language. There is even a wizard that follows very similar steps to those that
you used to create an SQL procedure. The main differences become apparent when
you reach the options page. At this point in the wizard, you can select what the JAR
ID for the stored procedure will be and also the package name.
Stored procedures can be developed using both SQLJ (static JDBC) or standard
Java code using JDBC.
Editing Java procedures

The editor window for Java stored procedures is slightly different from that for SQL
PL. It does, however, offer a full editor that allows you to add and work with Java
programs utilizing the JDBC interface for working with the database.
A Java procedure can also be run normally as if it were an SQL PL procedure. The
dual run and debug capabilities of the DC allow you to use it for all your stored
procedure testing needs.
Creating user-defined functions

The Development Center is not limited to just stored procedures; it can also be used
for user-defined functions. When you select the wizard to create an object, you can
choose to build a user-defined function rather than a stored procedure. From there,
you can choose a variety of programming languages that you can work with.
The list of languages that the Development Center supports is limited to a subset of
the total available languages for UDFs. If you want to develop UDFs in other
languages, you have to obtain the appropriate IDE and follow the steps laid out in
Using user-defined functions .
Creating an OLE function

If you would like to create an OLE DB-based function, the wizard walks you through
the required steps. Initially, you have to select the OLE provider you'll be using.
Once the provider is selected, you'll need to provide your user ID and password for
the connection. This information is used to build the connection string for the OLE
function.
Creating an SQL function

SQL functions can be created quickly by selecting the User Defined Function folder
in the Project view. You can then create the function by using the wizard, or by
having the DC quickly create a base template for you to fill in.
The UDF editing window is similar to the procedure editor, but you no longer have
the debugging option.
Section 7. Conclusion
Summary
This tutorial covered how functions and stored procedures can be used in your
application. User-defined functions can be used in a variety of environments and
with multiple languages. Both SQL and external programs can be used to build
complex functions that can be integrated into your SQL statements.
Stored procedures are becoming an increasingly important component of any

application that interacts with a database. By understanding why and when stored
procedures should be used, you can improve the performance and manageability of
your code. Like functions, stored procedures can be written in multiple languages,
including the Java language, C, and SQL PL. This tutorial focused on using SQL PL,
but many companies use Java procedures to ensure portability between database
platforms.
The discussion of stored procedures and functions was all tied together in the
overview of the DB2 Development Center. The Development Center is a leap
forward from the original stored procedure IDE, the Stored Procedure Builder. The
DC allows you to build and debug stored procedures written in both SQL PL and the
Java language. It now has the capacity to build user-defined functions written in SQL
PL and a variety of external languages as well. You can use projects to contain all of
your database objects, procedures, and functions into one modular unit. This new
packaging in the DC will help you build your applications with DB2.
Resources
Learn
• This tutorial will help you study for the application development exam, but there
is only so much space and time to explain the many concepts. Each of the
sections has information at the end that refers to you to keywords that you can
look up in the DB2 documentation for further information. The best source of
information for DB2 is the new Information Center. This Information Center was
added in DB2 V8.1 Fixpak 4 and has an entirely new interface written using the
Eclipse programming framework. It is quick and easy to use, and the search
utility built into it is far better than any of the previous incarnations. It also has
links to other sources of DB2 information on the Web, such as Google and the
DB2 Technotes. >All of my references at the end of the sections refer to this
source and not to the DB2 documentation included with the product itself. The
new Information Center is written by the DB2 Information Development team
and is the new official reference for DB2 help. Go out and try it! The DB2
Information Center is also available on the Web.
• There is an excellent and in-depth reference on stored procedures available. I
am slightly biased, since I am one of the authors but I still recommend it to
anyone who has to work extensively with procedures: DB2 SQL Procedural
Language for Linux, Unix and Windows, Yip, Bradstock, Curtis, Gao,
Janmohamed, Liu, and McArthur (Prentice Hall, 2002).
• In Creating a stored procedure, we discussed the fact that you need a C
compiler on your system in order to compile stored procedures. There are
several options available to you:
• GCC compiler
• Microsoft is offering a free .Net compiler (and other .Net development
tools) on its Web site. There are two products you need in order to use
these compilers. (They also must be installed in the order presented here.)
The first is the .Net Framework Version 1.1 Redistributable Package; the
second is the .Net Framework SDK Version 1.1.
• For more information on the DB2 UDB V8.1 Family Application Development
Certification (Exam 703), see IBM DB2 Information Management -- Training and
certification for information on classes, certifications available and additional
resources.
• As mentioned earlier, this tutorial is just one tutorial in a series of seven to help
you prepare for the DB2 UDB V8.1 Family Application Development
Certification exam (Exam 703). The complete list of all tutorials in this series is
provided below:
1. Database objects and Programming Methods
2. Data Manipulation
3. Embedded SQL Programming
4. ODBC/CLI Programming
5. Java Programming
6. Advanced Programming
7. User-Defined Routines
• Before you take the certification exam (DB2 UDB V8.1 Application
Development, Exam 703) for which this tutorial was created to help you
prepare, you should have already taken and passed the DB2 V8.1 Family
Fundamentals certification exam (Exam 700). Use the DB2 V8.1 Family
Fundamentals certification prep tutorial seriesto prepare for that exam. A set of
six tutorials covers the following topics:
• DB2 planning
• DB2 security
• Accessing DB2 UDB data
• Working with DB2 UDB data
• Working with DB2 UDB objects
• Data concurrency
• Use the DB2 V8.1 Database Administration certification prep tutorial seriesto
prepare for the DB2 UDB V8.1 for Linux, UNIX and Windows Database
Administration certification exam (Exam 701). A set of six tutorials covers the
following topics:
• Server management
• Data placement
• Database access
• Monitoring DB2 activity
• DB2 utilities
• Backup and recovery
• Stay current with developerWorks technical events and webcasts.

Get products and technologies
• Build your next development project with IBM trial software, available for
download directly from developerWorks.
Discuss
• Participate in the discussion forum for this content.
About the author

Drew Bradstock
Drew Bradstock is an engagement manager for IBM Competitive Migration Services.
He was previously part of the IBM Data Management Business Partner Enablement
team, where he worked with IBM's partners to migrate their applications to DB2. As
part of this team, he was a coauthor of the DB2 textbook DB2 SQL Procedural
Language for Linux, UNIX, and Windows. He was also a coauthor of the IBM
Redbook Scaling DB2 UDB on Windows Server 2003. Drew has specialized in
performance-tuning databases and migrating application code and architectures.
Drew is an IBM-certified DB2 V8.1 Advanced Administrator, Business Intelligence
expert and Application Developer. He was also a member of the team that developed
the DB2 V8 for Linux, UNIX, and Windows (LUW) certification exams. You can
contact Drew at drewkb@ca.ibm.com.

DB2 UDB V8.1 Family Application Development Certification:: User-Defined Routines

Uploaded by

Copyright:

Available Formats

DB2 UDB V8.1 Family Application Development Certification:: User-Defined Routines

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

DB2 UDB V8.1 Family Application Development Certification:: User-Defined Routines

Uploaded by

Copyright:

Available Formats

DB2 UDB V8.

1 Family Application Development

Skill Level: Introductory

Drew Bradstock (drewkb@ca.ibm.com)

Section 1. Before you start

What is this tutorial about?

• User-defined functions and how to use them

Who should take this tutorial?

Notices and trademarks

Section 2. Using user-defined functions

What are user-defined functions?

Here are the types of available functions:

• SQL Scalar: Returns a single data type value using SQL

• SQL (for functions written using one or more SQL statements)

C, or other languages. These functions are referred to as external functions. The

• Function name: Name of the function.

CREATE FUNCTION sections

CREATE FUNCTION TAN (X DOUBLE)

For more on scalar functions, see Scalar functions.

CREATE FUNCTION FROMPERSON (P PERSON)

For more on row functions, see Row functions.

CREATE FUNCTION DEPTEMPLOYEES (DEPTNO CHAR(3))

For more on table functions, see Table functions.

CREATE FUNCTION joinData (x INT, y INT)

Here's an example of a deterministic function:

CREATE FUNCTION joinData (x INT, y INT)

And here's an example of a nondeterministic function:

CREATE FUNCTION futureDate (x INT)

• MODIFIES SQL DATA: Indicates that all SQL statements supported in

• If an SQL function contains multiple references to any of the date or time

When a function is created, the EXECUTE privilege is automatically given to the

Using compound statements in functions

Here's the syntax diagram for a compound SQL block:

The major parts of the compound statement are:

DECLARE drewInt INTEGER;

CONDITIONAL HANDLING: This option is currently not available for functions.

The following SQL statements can be issued:

• fullselect (a common-table-expression can precede the fullselect)

Here's how you'd create your function:

CREATE FUNCTION changeSal (v_sal double)

And here's how you'd use it:

CREATE FUNCTION edCount (v_edLevel double)

Now, let's use it in an SQL statement:

Let's look again at the first sample SQL statement:

DB2 rewrites it so it looks like this:

And here's what it looks like after DB2 rewrites it:

SELECT Q3.$C0 AS "EDLEVEL", Q6.$C0

Here's the command creating the person type:

CREATE TYPE person AS (

Next, create a table with the person information:

CREATE TABLE empInfo (

Then create the row function:

CREATE FUNCTION FROMPERSON (P PERSON)

CREATE FUNCTION DEPTEMPLOYEES (DEPTNO CHAR(3))

CREATE TABLE testTab (

This function will work properly with that table:

CREATE FUNCTION returnAllTest (v_v1 int)

But now we add a new column to the table:

ALTER TABLE testTab ADD varThree int;