COBOL Table Declaration: How to Use OCCURS for Data Arrays ????

1?? What is OCCURS in COBOL?

The OCCURS clause defines a repeating group of fields within a structure, allowing you to store multiple values under a single table name. Instead of creating separate variables (ITEM-1, ITEM-2, ITEM-3...), you define a single item with OCCURS.

?? Basic Syntax:

01  DATA-TABLE.
    05 ITEM  PIC X(10) OCCURS 5 TIMES.        

Here, ITEM is an array of 5 elements, each holding 10 characters.

2?? Example: Storing 5 Customer Names in an Array

IDENTIFICATION DIVISION.
PROGRAM-ID. CUSTOMER-TABLE.
DATA DIVISION.
WORKING-STORAGE SECTION.
01  CUSTOMER-NAMES.
    05 CUSTOMER  PIC X(20) OCCURS 5 TIMES.

PROCEDURE DIVISION.
    MOVE 'John Doe' TO CUSTOMER(1).
    MOVE 'Alice Smith' TO CUSTOMER(2).
    MOVE 'Bob Johnson' TO CUSTOMER(3).
    
    DISPLAY "First Customer: " CUSTOMER(1).
    DISPLAY "Second Customer: " CUSTOMER(2).
    
    STOP RUN.        

?? Output:

First Customer: John Doe  
Second Customer: Alice Smith        

? Pros: Cleaner and more maintainable than declaring separate variables.

3?? Accessing OCCURS with PERFORM (Loops)

To process all elements in a table, use a PERFORM loop with an index:

WORKING-STORAGE SECTION.
       01  CUSTOMER-TABLE.
           05  CUSTOMER  PIC X(20) OCCURS 5 TIMES INDEXED BY CUSTOMER-IDX.

       PROCEDURE DIVISION.
           PERFORM VARYING CUSTOMER-IDX FROM 1 BY 1 UNTIL CUSTOMER-IDX > 5
               DISPLAY "Customer " CUSTOMER-IDX ": " CUSTOMER(CUSTOMER-IDX)
           END-PERFORM.        

? Best Use Case: When you need to iterate over an unknown number of records.

4?? OCCURS with DEPENDING ON – Dynamic Arrays

Want to define a flexible table size at runtime? Use OCCURS DEPENDING ON:

01  MAX-CUSTOMERS  PIC 9(2) VALUE 3.
01  CUSTOMER-NAMES.
    05 CUSTOMER  PIC X(20) OCCURS 1 TO 10 TIMES DEPENDING ON MAX-CUSTOMERS.        

?? Here, the array size can be changed dynamically based on MAX-CUSTOMERS.

5?? OCCURS with MULTIPLE LEVELS – 2D Tables

You can nest OCCURS inside OCCURS to create multi-dimensional tables.

?? Example: Storing 5 customers, each with 3 transaction amounts

01  CUSTOMER-DATA.
    05 CUSTOMER OCCURS 5 TIMES.  *> 5 Customers (1st dimension)
        10 NAME        PIC X(20).
        10 TRANSACTIONS PIC 9(6)V99 OCCURS 3 TIMES.  *> 3 Transactions per Customer (2nd dimension)
        

To access a specific transaction of a customer:

       MOVE "John Doe" TO NAME(1).  *> Customer 1
       MOVE 1250.75 TO TRANSACTIONS(1,1).  *> First transaction of first customer
       MOVE 890.50 TO TRANSACTIONS(1,2).  *> Second transaction of first customer

       MOVE "Alice Smith" TO NAME(2).  *> Customer 2
       MOVE 3000.00 TO TRANSACTIONS(2,1).  *> First transaction of second customer        

? Best Use Case: When you need to store structured data like customer orders, sales, or matrix-based calculations.

6?? OCCURS vs. Indexed Files – When to Use?

?? Rule of Thumb:

• Use OCCURS when working within program memory.
• Use indexed files for large datasets requiring search operations.

7?? Key Takeaways

?? OCCURS simplifies handling multiple data items in a structured way.

?? PERFORM VARYING helps iterate over table elements efficiently.

?? OCCURS DEPENDING ON allows dynamic sizing.

?? Nested OCCURS enables multi-dimensional data storage.