HP OpenVMS Systems Documentation

HP COBOL
Reference Manual

When both operands are numeric, their algebraic values are compared. The program performs the necessary conversion if the data descriptions of the operands specify different USAGE. When you use operands that are literals or arithmetic expressions, their length (in terms of the number of digits represented) is not significant.

Unsigned numeric operands are assumed to be positive for comparison. A zero value is always treated the same way, whether or not the operand contains a sign.

6.5.1.2 Comparison of Nonnumeric Operands

When one (or both) of the operands is nonnumeric, each operand is considered a string of alphanumeric characters. Therefore, the operands are compared according to the program's collating sequence. (See the Section 4.1.2 paragraph in Chapter 4.)

If one of the operands is numeric, it must be either an integer literal or a data item described as an integer. The data item must be implicitly or explicitly described with USAGE DISPLAY. The treatment of the numeric data item is further affected by the following:

• If the nonnumeric operand is an elementary data item or a nonnumeric literal, the numeric data item is treated as though it were moved to an elementary alphanumeric data item of the same size. The content of this alphanumeric data item is then compared to the nonnumeric operand.
• If the nonnumeric operand is a group item, the numeric operand is treated as though it were moved to a group item of the same size. The content of this group item is then compared to the nonnumeric operand.
• When a numeric operand contains a sign, its sign is part of the string only if the other operand is a group item. Otherwise, the sign is removed and is not part of the comparison.

The two operands are compared character by character, beginning at the left end of each string. When the operation finds an unequal character pair, it uses that pair to evaluate the comparison. The greater operand is the one that contains the character with the higher collating sequence position. If the operands are of unequal size, the shorter operand is treated as if it were extended on the right with spaces to make it the same size as the other. Therefore, ABCD is greater than ABC (unless the program's collating sequence dictates otherwise).

Comparisons of Index-Names or Index Data Items

A program can compare the following:

• Two index-names
• One index-name and one literal or data item (other than an index data item)
• One index-name and one index data item
• Two index data items

The class condition tests whether the contents of an operand are numeric or alphabetic. It also determines if an alphabetic operand contains only uppercase characters, only lowercase characters, or if an operand is in conformance with class-name. The general format is as follows:

The identifier must reference a data item whose usage is explicitly or implicitly DISPLAY or COMP-3. If the identifier is a function-identifier, it must reference an alphanumeric function.

The following rules apply to the NUMERIC test:

1. The test is true when the operand contains only the characters 0 to 9 and the operational sign (subject to the next rule); otherwise, it is false.
2. The operand must contain an operational sign if its PICTURE clause specifies a sign. If the PICTURE clause does not specify a sign, the operand must not contain one. If the operand contains a sign that is not specified, or if a sign is specified and the operand does not contain one, the NUMERIC test is false.
3. You cannot use the test for an operand described as alphabetic or a group item containing signed elementary items.

The following rules apply to the ALPHABETIC test:

1. The test is true when the operand contains only the characters A to Z, a to z, and the space; otherwise, it is false.
2. You cannot use the ALPHABETIC test for an operand described as numeric.

The ALPHABETIC-LOWERCASE test is true when the operand contains only the characters a to z, and the space; otherwise, it is false.

The ALPHABETIC-UPPERCASE test is true when the operand contains only the characters A to Z, and the space; otherwise, it is false.

The class-name test is true when the operand consists entirely of the characters listed in the definition of class-name in the SPECIAL-NAMES paragraph. The class-name test must not be used with an item whose data description describes the item as numeric.

NOT and the key word following it are treated as a unit. For example, NOT NUMERIC is a test for determining that the operand is nonnumeric.

6.5.3 Condition-Name Condition

The condition-name condition determines if a data item contains a value assigned to one of that item's condition-names. The term conditional variable refers to the data item. condition-name refers to a level 88 entry associated with that item.

The general format for this condition is:

The condition is true if one of the values corresponding to condition-name equals the value of the associated conditional variable. The data description for a variable can associate condition-name with one or more ranges of values. In this case, the condition tests to determine if the value of the variable falls in the specified range (end values included).

The following example illustrates testing condition-names associated with both one value and a range of values:

    WORKING-STORAGE SECTION.
    01  STUDENT-REC.
        05  YEAR-ID           PIC 99.
        88  FRESHMAN                    VALUE IS 1.
        88  SOPHOMORE                   VALUE IS 2.
        88  JUNIOR                      VALUE IS 3.
        88  SENIOR                      VALUE IS 4.
        88  GRADUATE                    VALUE IS 5 THRU 10.
        .
        .
        .
PROCEDURE DIVISION.
        .
        .
        .
    IF FRESHMAN ...
    IF SOPHOMORE ...
    IF JUNIOR ...
    IF SENIOR ...
    IF GRADUATE ...

When your program evaluates a conditional variable and its condition-name, the procedure is the same as the one used with the relation condition. (See Section 6.5.1.)

6.5.4 Switch-Status Condition

The switch-status condition tests the on or off setting of an external logical program switch. Its general format is as follows:

You use the SWITCH clause of the SPECIAL-NAMES paragraph to associate condition-name with a logical switch setting. (See the Section 4.1.3 paragraph in Chapter 4.) The condition is true if the switch setting in effect during program execution is the same one assigned to condition-name.

6.5.5 Sign Condition

Its general format is as follows:

An operand is defined as:

• POSITIVE, if its value is greater than zero
• NEGATIVE, if its value is less than zero
• ZERO, if its value equals zero

arithmetic-expression must contain at least one reference to a variable.

NOT and the key word following it are treated as a unit. For example, NOT ZERO tests for a nonzero condition.