To ensure proper and accurate data matches between the data source and your COBOL program, you must match the SQL data type with the appropriate COBOL data type. Note that only fixed-length and variable-length character strings are supported.
The following table describes data type compatibility for DB2.
| SQL Type
| COBOL Type
| Description
|
| CHAR(n)
| 01 name PIC X(n).
| Fixed-length character string
|
| DATE
| 01 name PIC X(10).
| 10-byte character string
|
| DECIMAL(p,n)
| 01 name PIC S9(m) V9(n) COMP-3.
| Packed decimal
(where m = p - n) |
| DOUBLE
| 01 name USAGE IS DOUBLE.
| Double-precision floating point
|
| INTEGER
| 01 name PIC S9(9) COMP-5.
| 32-bit signed integer
|
| LONG VARCHAR(n)
| 01 name.
03 name-length PIC S9(4) COMP-5. | Variable length character string
|
| REAL
| 01 name PIC S9(4) USAGE IS COMP-1.
| Single-precision floating point
|
| SMALLINT
| 01 name PIC S9(4) COMP-5.
| 16-bit signed integer
|
| TIME
| 01 name PIC X(8).
| 8-byte character string
|
| TIMESTAMP
| 01 name PIC X(26).
| 26-byte character string
|
| VARCHAR(n)
| 01 name.
03 name-length PIC S9(4) COMP-5. | Variable-length character string (n <= 4000)
|
The following table describes data type compatibility for MSSQL.
| SQL Type
| COBOL Type
| Description
|
| CHAR(n)
| 01 name PIC X(n).
| Fixed-length character string
|
| DATETIME
| 01 name PIC X(25).
| 25-byte character string
|
| DECIMAL(p,n)
| 01 name PIC S9(m) V9(n) COMP-3.
| Packed decimal
(where m = p - n) |
| DOUBLE PRECISION
| 01 name USAGE IS DOUBLE.
| Double-precision floating point
|
| INTEGER
| 01 name PIC S9(9) COMP-5.
| 32-bit signed integer
|
| REAL
| 01 name PIC S9(4) USAGE IS COMP-1.
| Single-precision floating point
|
| SMALLINT
| 01 name PIC S9(4) COMP-5.
| 16-bit signed integer
|
| VARCHAR(n)
| 01 name.
03 name-length PIC S9(4) COMP-5. | Variable-length character string (n <= 4000)
|