IBM Mainframe Data Types

Double Prec float (Mainframe)

DOUBLE

8

8-byte big-endian mainframe double precision:  USAGE COMP-2.

Longword (BE)

INTEGER

4

Signed big-endian 4-byte INTEGER PIC S9(n) COMP, where 5 <= n <= 9.

PACKED (F) Decimal -> Double

DOUBLE

1-10

Unsigned Packed Decimal PIC 9(precision,scale) USAGE COMP-3, where scale > 0 or 10 <= precision <= 18, or both.

PACKED (F) Decimal -> Integer

LONG

1-5

Unsigned Packed Decimal PIC 9(n) USAGE COMP-3, where 1 <= n <= 9.

PACKED Decimal

-> Decimal

DECIMAL

1-10

Signed Packed Decimal PIC S9(precision,scale) USAGE COMP-3, where scale > 0 or 10 <= precision <= 18, or both.

PACKED Decimal
-> Integer

LONG

1-5

Signed Packed Decimal PIC S9(n) USAGE COMP-3, where 1 <= n <= 9.

Signed Overpunch
-> Decimal

DECIMAL

1-18

Signed Numeric Display PIC S9(precision,scale), where scale > 0 or 10 <= precision <= 18, or both.

Signed Overpunch
-> Integer

LONG

1-9

Signed Numeric Display PIC S9(n), where 1 <= n <= 9.

Single Prec float (Mainframe)

REAL

4

4-byte big-endian mainframe float:  USAGE COMP-1.

Text -> Decimal

DECIMAL

1-18

Unsigned Numeric Display PIC 9(precision,scale), where scale > 0 or 10 <= precision <= 18, or both.

Text -> Integer

INTEGER

1-9

Unsigned Numeric Display PIC 9(n), where 1 <= n <= 9

Text (Right Space Padded)

CHAR

1-32767

Character display PIC X(nnnnn), where

1 <= n <= 32767

Unsigned Long 4 bytes (BE)

INTEGER

4

Unsigned big-endian 4-byte INTEGER PIC S9(n) COMP, where 5 <= n <= 9.

Unsigned Word 2 Bytes (BE)

SMALLINT

2

Unsigned big-endian 2-byte INTEGER PIC 9(n) COMP, where 1 <= n <= 4.

Word (BE)

SMALLINT

2

Signed big-endian 2-byte INTEGER PIC S9(n) COMP, where 1 <= n <= 4.