COBOL Manual for TNS/E Programs (H06.08+, J06.03+)

The descriptions of the action of the PERFORM statement in this section are expressed in terms
of the out-of-line PERFORM statement. The execution of an in-line PERFORM statement is exactly
equivalent to that of an out-of-line PERFORM statement, with the exception that the statements
contained in imperative-statement in the in-line PERFORM statement are executed in
place of the statements within the range of procedure-group. Unless specially qualified
by the term in-line or out-of-line, all the considerations that apply to the out-of-line PERFORM
statement also apply to the in-line PERFORM statement.
• Execution and Transfers of Control
Chapter 8: Procedure Division (page 227) explains the execution of sections and paragraphs
as well as explicit and implicit transfers of control.
• Range of the PERFORM Statement
The range of an out-of-line PERFORM statement is the statements contained within the range
of procedure-group.
The range of an in-line PERFORM statement is the statements contained within the PERFORM
statement.
• Logical Range of the PERFORM Statement
The logical range of any PERFORM statement is all statements that are executed as a result
of the PERFORM statement, including the transfer of control to the statement following the
PERFORM statement (or to the statement following the END-PERFORM). The logical range
includes all statements executed as a result of a GO TO, PERFORM, or CALL statement in the
range of the original PERFORM statement, as well as all statements in the Declaratives Portion
that might be executed. There is no requirement for statements in the range of a PERFORM
statement to appear consecutively.
• Nested PERFORM Statements
The logical range of a PERFORM cycle can include another PERFORM statement (called a
nested PERFORM statement), but these restrictions apply:
◦ The logical range of the nested PERFORM statement must be either totally included in or
totally excluded from the logical range of the outer PERFORM statement; therefore, an
active PERFORM statement whose execution point begins within the range of another
active PERFORM statement must not allow control to pass to the exit of the other active
PERFORM statement.
◦ The final paragraph in the range of one active PERFORM statement cannot be the same
as the final paragraph in the range of any other active PERFORM statement, because the
program terminates only one PERFORM cycle for each such final paragraph. The final
paragraph would have to be executed twice to terminate both cycles.
◦ The sequence of statements executed in any cycle of one PERFORM statement must not
allow control to reach the return point for any other active PERFORM statement in which
it is nested. If this does happen, neither PERFORM cycle terminates, and control falls
through to succeeding statements.
◦ The sequence of statements executed in any cycle of one PERFORM statement must not
allow control to reach any active PERFORM statement, including itself.
◦ As a consequence of the preceding rules, a cycle of an active PERFORM statement cannot
end until after any PERFORM statements nested within it complete their execution.
PERFORM 391