Coding interoperable data types in COBOL and Java

Your COBOL program can use only certain data types when communicating with Java™.

Table 1. **Interoperable data types in COBOL and Java**
Primitive Java data type	Corresponding COBOL data type
boolean¹	`PIC X` followed by exactly two condition-names of this form: `level-number data-name PIC X. 88 data-name-false value X'00'. 88 data-name-true value X'01' through X'FF'.`
byte¹	Single-byte alphanumeric: `PIC X` or `PIC A`
short	`USAGE BINARY`, `COMP`, `COMP-4`, or `COMP-5`, with `PICTURE` clause of the form `S9(n)`, where 1<=`n`<=4
int	`USAGE BINARY`, `COMP`, `COMP-4`, or `COMP-5`, with `PICTURE` clause of the form `S9(n)`, where 5<=`n`<=9
long	`USAGE BINARY`, `COMP`, `COMP-4`, or `COMP-5`, with `PICTURE` clause of the form `S9(n)`, where 10<=`n`<=18
float²	`USAGE COMP-1`
double²	`USAGE COMP-2`
char	Single-character elementary national: `PIC N USAGE NATIONAL`. (Cannot be a national group.)
class types (object references)	`USAGE OBJECT REFERENCE` `class-name`
You must distinguish boolean from byte, because they each correspond to `PIC X`. `PIC X` is interpreted as boolean only if you define an argument or a parameter with the two condition-names as shown. Otherwise, a `PIC X` data item is interpreted as the Java byte type. Java floating-point data is formatted according to the IEEE Standard for Binary Floating Point Arithmetic. Enterprise COBOL, however, uses hexadecimal floating-point representation. When you pass floating-point arguments by using an `INVOKE` statement, or you receive floating-point data from a Java method, the arguments and data are automatically converted as needed.