USAGE Clause in COBOL

DISPLAY Usage (Default)

DISPLAY is the default USAGE in COBOL when no explicit USAGE is specified. It stores data in a human-readable character format, with each character or digit occupying one byte of storage.

1
2
3
4
5
6
7
8
9
10
11
12
       01  EMPLOYEE-RECORD.
           05  EMPLOYEE-ID           PIC 9(5) USAGE DISPLAY.
           05  EMPLOYEE-NAME         PIC X(30) USAGE DISPLAY.
           05  DEPARTMENT-CODE       PIC 99 USAGE IS DISPLAY.
           05  SALARY                PIC 9(7)V99 USAGE DISPLAY.
           
       *> Since DISPLAY is the default, these are equivalent to the above:
       01  EMPLOYEE-RECORD-EQUIV.
           05  EMPLOYEE-ID           PIC 9(5).
           05  EMPLOYEE-NAME         PIC X(30).
           05  DEPARTMENT-CODE       PIC 99.
           05  SALARY                PIC 9(7)V99.

Key characteristics of DISPLAY usage:

• Each digit or character occupies one byte of storage
• Numeric data is stored in character format (EBCDIC on mainframes, ASCII or Unicode on other platforms)
• Human-readable without conversion
• Sign information for numeric data is typically stored in the rightmost byte (overpunched)
• Less efficient for arithmetic operations compared to binary formats
• Works well for data that is primarily used for input/output and display purposes
• Compatible with most external data formats

COMPUTATIONAL and COMP Variants

COMPUTATIONAL (or COMP for short) usage types store numeric data in binary format, which is more efficient for arithmetic operations. COBOL provides several variants of the COMPUTATIONAL usage with different characteristics.

1
2
3
4
5
6
7
       01  CALCULATION-FIELDS.
           05  TOTAL-AMOUNT         PIC 9(9)V99 USAGE COMPUTATIONAL.
           05  COUNTER              PIC 9(5) USAGE COMP.
           05  RATE-PERCENT         PIC 9(3)V9(2) USAGE IS COMP.
           05  FLOAT-VALUE          PIC 9(5)V9(2) USAGE COMP-1.
           05  DOUBLE-VALUE         PIC 9(10)V9(5) USAGE COMP-2.
           05  SMALL-INTEGER        PIC S9(4) USAGE COMP-5.

Common COMPUTATIONAL variants:

• COMP or COMPUTATIONAL - Binary format, usually equivalent to COMP-4
• COMP-1 - Single-precision floating-point (typically 4 bytes)
• COMP-2 - Double-precision floating-point (typically 8 bytes)
• COMP-3 - Packed-decimal format (see next section)
• COMP-4 - Usually the same as COMP (binary)
• COMP-5 - Binary format with native byte order (useful for interoperability with other languages)

Storage Allocation for COMP/COMPUTATIONAL

Basic COMP (or COMP-4) allocates binary storage based on the number of digits in the PICTURE clause:

• 1-4 digits: 2 bytes (halfword)
• 5-9 digits: 4 bytes (fullword)
• 10-18 digits: 8 bytes (doubleword)

Advantages of COMPUTATIONAL usage:

• More efficient arithmetic operations
• Can use less storage than DISPLAY for larger numbers
• Better performance for intensive calculations
• Direct compatibility with binary arithmetic hardware

Disadvantages of COMPUTATIONAL usage:

• Not human-readable without conversion
• May require conversion for external data exchange
• Potential for platform-dependent behavior
• Potential precision issues for very large numbers

PACKED-DECIMAL

PACKED-DECIMAL (often equivalent to COMP-3) is a hybrid storage format that balances storage efficiency and computational performance. It's especially popular on mainframe systems with hardware support for packed decimal operations.

1
2
3
4
5
       01  FINANCE-FIELDS.
           05  ACCOUNT-BALANCE      PIC S9(13)V99 USAGE PACKED-DECIMAL.
           05  TRANSACTION-AMOUNT   PIC S9(7)V99 USAGE COMP-3.
           05  INTEREST-RATE        PIC 9(3)V9(5) USAGE IS PACKED-DECIMAL.
           05  CUSTOMER-BALANCE     PIC S9(11)V99 COMP-3.

Key characteristics of PACKED-DECIMAL:

• Each byte stores two decimal digits (except the rightmost byte)
• The rightmost byte contains the last digit and the sign
• Requires approximately (n/2) + 1 bytes for n digits
• More storage-efficient than DISPLAY for numeric data
• Often preferred for financial calculations on mainframes
• Provides exact decimal precision (no binary conversion issues)
• Typically offers good performance on mainframe systems

BINARY, BINARY-LONG, BINARY-SHORT

The BINARY usage and its variants store numeric data in native binary format. These usage types are often used for efficient arithmetic operations and interfacing with other programming languages.

1
2
3
4
5
       01  BINARY-DATA-ITEMS.
           05  ITEM-COUNT           PIC 9(4) USAGE BINARY.
           05  RECORD-SIZE          PIC 9(8) USAGE BINARY-LONG.
           05  FLAG-VALUE           PIC 9(2) USAGE BINARY-SHORT.
           05  LARGE-COUNTER        PIC 9(16) USAGE BINARY SYNC.

BINARY usage variants:

• BINARY - Standard binary format (often equivalent to COMP)
• BINARY-SHORT - Binary in short (halfword) format, typically 2 bytes
• BINARY-LONG - Binary in long (fullword) format, typically 4 bytes
• BINARY-DOUBLE - Binary in double (doubleword) format, typically 8 bytes

Key characteristics of BINARY usage:

• Stores data in the computer's native binary format
• Most efficient for arithmetic operations and loops
• Often used for index values, counters, and array subscripts
• Fixed storage size based on the variant used (not dependent on PICTURE digit count)
• Can be used with the SYNCHRONIZED clause for aligned memory access
• Directly compatible with system APIs and calls to other languages

POINTER and Other Special USAGEs

Modern COBOL provides several specialized USAGE types for specific purposes, such as memory addressing, function references, and object-oriented programming.

1
2
3
4
5
6
7
       01  SPECIAL-USAGES.
           05  MEMORY-ADDRESS        USAGE POINTER.
           05  PROCEDURE-ADDRESS     USAGE PROCEDURE-POINTER.
           05  FUNCTION-ADDRESS      USAGE FUNCTION-POINTER.
           05  OBJECT-REFERENCE      USAGE OBJECT REFERENCE.
           05  STRING-HANDLE         USAGE DISPLAY-1.
           05  NATIONAL-STRING       USAGE NATIONAL.

Common special USAGE types:

• POINTER - Holds memory addresses, typically for dynamic memory allocation
• PROCEDURE-POINTER - References a procedure or paragraph
• FUNCTION-POINTER - References a function entry point
• OBJECT REFERENCE - References an object in object-oriented COBOL
• DISPLAY-1 - For DBCS (Double-Byte Character Set) data
• NATIONAL - For Unicode data, typically using UTF-16

1
2
3
4
5
6
7
8
9
10
11
12
13
       WORKING-STORAGE SECTION.
       01  BLOCK-SIZE               PIC 9(8) COMP.
       01  MEMORY-BLOCK             USAGE POINTER.
       
       PROCEDURE DIVISION.
           SET BLOCK-SIZE TO 1024
           CALL "GETMAIN" USING BLOCK-SIZE
                                MEMORY-BLOCK
           
           *> Use the allocated memory
           
           *> Free the memory when done
           CALL "FREEMAIN" USING MEMORY-BLOCK

Key points about special USAGE types:

• Do not require a PICTURE clause (except for specific implementations)
• Size is determined by the implementation and architecture
• Provide advanced capabilities for modern programming techniques
• Support for these types varies significantly between COBOL compilers
• Typically used with specific statements like SET, CALL, or INVOKE
• Essential for certain system programming tasks and interface with other languages

Choosing the Right USAGE

Selecting the appropriate USAGE clause for your data items is an important design decision that affects performance, storage requirements, and compatibility. Here are some guidelines for choosing the right USAGE.

Considerations for USAGE selection:

• Performance - Choose COMP or BINARY for calculation-intensive fields
• Storage efficiency - PACKED-DECIMAL (COMP-3) often provides the best balance
• Precision requirements - PACKED-DECIMAL for exact decimal precision
• Compatibility - DISPLAY for external interfaces and human-readable data
• Platform considerations - Use PACKED-DECIMAL on mainframes with hardware support
• Data size - For large numbers, PACKED-DECIMAL or COMP may be more efficient
• Compiler support - Check which USAGE types are supported by your compiler

FAQ Section

Related Pages

• PICTURE Clause
• Data Description Entries
• Data Types and Representation
• VALUE Clause
• Arithmetic Operations