Whenever we issue a COMMIT statement, all the open cursors will get closed. This is a very common case when we have to frequently use the commit statement after a UPDATE while working with a cursor. In this case we can use the “WITH HOLD” clause during the cursor declaration.The “WITH HOLD” clause will keep the cursor open even after firing the COMMIT statement. We can give the “WITH HOLD” clause in the following way.EXEC SQL       DECLARE ORDER_CUR CURSOR WITH HOLD FOR          SELECT ORDER_ID, TRANSACTION_ID FROM ORDERS             WHERE ORDER_DATE = ‘2020-07-28’ END-EXEC

The “WHERE CURRENT OF” clause will place the exclusive lock on the row once the UPDATE statement is executed. The “WHERE CURRENT OF” clause will point to the most recently fetched row of the cursor.We can update the rows in cursor using “WHERE CURRENT OF” in the following way.CURSOR definition.EXEC SQL DECLARE ORDER_CUR CURSOR FOR SELECT ORDER_ID, TRANSACTION_ID FROM ORDERS WHERE ORDER_DATE = ‘2020-07-28’ END-EXECOPEN cursorEXEC SQL OPEN ORDER_CUR END-EXECFETCH cursor and Update rowSET WF-END-CURSOR-N TO TRUE    PERFORM UNTIL WF-END-CURSOR-Y    EXEC SQL    FETCH ORDER_CUR INTO :ORDER-ID, :TRANSACTION-ID END-EXEC IF TRANSACTION-ID NOT = SPACES    EXEC SQL   ... Read More

The “FOR UPDATE OF” clause is given in the cursor declaration and it is used when we want to update the table. All the columns which need to be updated should be given in the cursor declaration.The “FOR UPDATE OF” clause will place the exclusive lock on all the qualifying rows once the cursor is open. We can also update the table without using “FOR UPDATE CLAUSE” but in that case the exclusive lock will be placed on the row only when the UPDATE query will be executed.

The cursor can be used to fetch multiple rows from a DB2 table. However, we have to fetch this cursor in a loop so that values corresponding to a single row are assigned to host variables at a time. Based on this logic we have to process our loop till the cursor reaches the last row result.The SQLCODE field takes the value as 100 when there are no more rows left in the cursor to be fetched. Practically, we can implement this in the following way.SET WF-END-CURSOR-N TO TRUE PERFORM UNTIL WF-END-CURSOR-Y    EXEC SQL       FETCH ORDER_CUR ... Read More

The CURSOR is used when we have to fetch multiple rows from a table. There are 4 steps involved in order to use a cursor in a COBOL-DB2 program.DECLARE cursor− In this step we will define the layout of the cursor. We will give the query we want to use. For example−EXEC SQL DECLARE ORDER_CUR CURSOR FOR SELECT ORDER_ID FROM ORDERS WHERE ORDER_DATE = ‘2020-07-28’ END-EXECOPEN cursor− Next we will open our cursor. This statement readies the cursor for data retrieval. For example−EXEC SQL OPEN ORDER_CUR END-EXECFETCH cursor− In this statement, we start fetching the data from DB2 and the ... Read More

The standalone SELECT statement is generally used when we use a primary or an alternate key in the WHERE clause. Therefore, In this case we are certain that the standalone select statement will return only one row since the primary key cannot have a duplicate value (multiple rows).If we want to query the database using a non-unique key which could return multiple rows from a DB2 table, we have to use a cursor to handle the multiple rows returned. We can access the cursor in a loop to read each row data one by one.For example, if we want to ... Read More

A NULL indicator is a 2 bytes field which serves the multiple purpose. This indicator takes the value as -1 when any DB2 column has NULL value and it takes the value as 0 when DB2 column has a non NULL value.Although the main purpose of the NULL indicator is to check whether a column has a NULL value or not, this indicator can take a value of -2 as well. A -2 value in this indicator states that a NULL value was assigned to the host variable as a result of evaluating an expression with an arithmetic error, or ... Read More

When we define the DB2 table, we can declare any column as “NOT NULL” which means that in any case this column cannot store NULL value.Now if we try to store a NULL value in this column in our COBOL-DB2 program using -1 value in the NULL indicator then our query will fail. In this case the SQLCODE field of SQLCA will give the error code as -407. As per the IBM documentation -407 error code states that−“AN UPDATE, INSERT, OR SET VALUE IS NULL, BUT THE OBJECT COLUMN CANNOT CONTAIN NULL VALUES”.

We will make use of NULL indicator in order to store a NULL value in any column of a DB2 table. Firstly, we should move a value -1 in the NULL indicator in our COBOL-DB2 program. After that we execute UPDATE or INSERT query to store the NULL value.For example, if we have to update a NULL value in the ORDER_DESCRIPTION column of the ORDER table where ORDER_ID is 3345612.A020-UPDATE-ORDERS.    MOVE -1 TO ORDER-DESCRIPTION-N    MOVE SPACES TO ORDER-DESCRIPTION-DATA    EXEC SQL       UPDATE ORDERS          SET ORDER_DESCRIPTION =           ... Read More

The COUNT function in DB2 is used to return the number of rows which satisfies the given criteria. The GROUP BY is used to divide the rows in the groups based on the criteria given in the query.If we perform the GROUP BY on INVOICE_ID and there are few rows having NULL value in INVOICE_ID then the null values form a separate group. For example, if we have below table.ORDER_IDINVOICE_IDA112343214A556113214A99867NULLA556713214A88907NULLA560126701On executing the query which performs GROUP BY on INVOICE_ID and counts the number of rows, we will get the result below.SELECT INVOICE_ID, COUNT(*) AS INVOICE COUNT FROM ORDERS GROUP BY ... Read More