Marks start of statement block in a CASE construct.
Syntax
CASE ( case-selector ) [ construct-name ] |
- case-selector
is a comma-separated list of ranges of values that
are candidates for matching against the case index specified by
the SELECT CASE statement. Each item in the list can take one
of the following forms:
where:
case-value, low, and high
are scalar initialization expressions of type integer, character,
or logical
DEFAULT
indicates the statement block to execute if none of the other CASE statements in the CASE construct produces a match.
- construct-name
is the name given to the CASE construct.
Description
The CASE statement is used in a CASE construct to mark the start of a statement block.
The CASE construct can consist of multiple blocks; at most,
one is selected for execution. Selection is determined by comparing
the case index produced by the SELECT CASE statement to the case-selector in each CASE statement. If a match is found, the statement
block under the matching case-selector executes. A match between the case index (c) and case-selector is determined for each form of case-selector, as follows:
- case-value
For integer and character types, a match occurs
if c .EQ. case-value.
For logical types, a match occurs if c .EQV. case-value.
- low:
For integer and character types, a match occurs
if c .GE. low.
- :high
For integer and character types, a match occurs
if c .LE. high.
- low : high
For integer and character types, a match occurs
if c .GE. low .AND. c .LE. high.
- DEFAULT
For integer, character, and logical types, a match
occurs if no match is found with any other case-selector and DEFAULT is specified as a case-selector.
If CASE DEFAULT is not present and no match is found with any
of the other CASE statements, none of the statement blocks within
the CASE construct executes and execution resumes with the
first executable statement following the END SELECT statement.
At most only one DEFAULT selector can appear within a CASE construct.
Each CASE statement must specify a unique value or range
of values within a particular CASE construct. Only one match can occur, and only
one statement block can execute.
All case-selectors and the case index within a particular CASE construct must be of the same type: integer, character,
or logical. However, the lengths of character types can differ.
The colon forms—low:, :high, or low:high—are not permitted for a logical type.
Although putting the CASE statements in order according to range may improve
readability, it is not necessary for correct or optimal execution
of the CASE construct. In particular, DEFAULT can appear anywhere among the CASE statements and need not be the last.
CASE statements inside a named CASE construct need not specify construct-name; but if they do, the name they specify must match
that of the SELECT CASE.
A CASE statement can have an empty statement block.
Examples
The following example considers a person’s credits
and debits and prints a message indicating whether a resulting account
balance will be overdrawn, empty, uncomfortably small, or sufficient:
INTEGER :: credits, debits
SELECT CASE (credits - debits)
CASE (:-1)
PRINT *, 'OVERDRAWN'
CALL TRANSFERFUNDS
CASE (0)
PRINT *, 'NO MONEY LEFT'
CASE (1:50)
PRINT *, 'BALANCE LOW'
CASE (51:)
PRINT *, 'BALANCE OKAY'
END SELECT |
Related statements
SELECT CASE and END (construct)
Related concepts
The CASE construct is described in “CASE
construct”.
Printable version
