Terminology and concepts

The following sections define the terms and explain the concepts that are mentioned throughout this chapter.

Program units

A program consists of the following program units:

Main program unit
External procedure, which can be either a subroutine or a function
Module program unit
Block data program unit

A complete executable program contains one (and only one) main program unit and zero or more other program units, each of which is separately compilable. A program unit is an ordered set of constructs, statements, comments, and INCLUDE lines. The heading statement identifies the kind of program unit; it is optional in a main program unit only. An END statement marks the end of a program unit.

The only executable program units are the main program and external procedures. Program execution begins with the first executable statement in the main program and ends (typically) with the last. During execution, if the main program references an external procedure, control passes to the procedure, which executes and returns control to the main program. An executing procedure can also reference other procedures or even reference itself recursively.

The main program unit is described in “Main program”, and external procedures are described in “External procedures”.

The nonexecutable program units are:

The module program unit, which contains data declarations, user-defined type definitions, procedure interfaces, common block declarations, namelist group declarations, and subprogram definitions used by other program units. Modules are described in “Modules”.
The block data program unit, which specifies initial values for variables in named common blocks. Block data program units are described in “Block data program unit”.

Procedures

A procedure is a subroutine or function that contains a sequence of statements and that may be invoked during program execution. Depending on where and how it is used, a procedure can be one of the following:

Intrinsic procedures are defined by the language and are available for use without any declaration or definition. Intrinsic procedures implement common computations that are important to scientific and engineering applications. Intrinsic procedures are described in detail in Chapter 11 “Intrinsic procedures”.
An external procedure is a separately compilable program unit whose name and any additional entry points have global scope. External procedures are described in “External procedures”.
An internal procedure has more limited accessibility than an external procedure. It can appear only within a main program unit or an external procedure and cannot be accessed outside of its hosting program unit. Internal procedures are described in “Internal procedures”.
A module procedure can be defined only within a module program unit and can be accessed only by use association. Module procedures are described in “Modules”.

Scope

All defined Fortran entities have a scope within which their properties are known. For example, a label used within a subprogram cannot be referenced directly from outside the subprogram; the subprogram is the scoping unit of the label. A variable declared within a subprogram has a scope that is the subprogram. A common block name can be used in any program unit, and it refers to the same entity—that is, the name has global scope. At the other extreme, the index variable used within an implied-DO loop in a DATA statement or array constructor has a scope consisting only of the implied-DO loop construct itself.

Association

If the concept of scope limits the accessibility of entities, then the concept of association permits different entities to become accessible to each other in the same or different scope. The different types of association are:

Argument association is the association that is established between actual arguments and dummy arguments during a procedure reference. For more information, see “Argument association”.
Host association applies to nested scoping units, where the outer scoping unit (for example, an external procedure) plays host to the inner scoping unit (for example, an internal procedure). Host association allows the host and its nested scoping units to share data. For information about internal procedures, see “Internal procedures”.
Pointer association is the association between a pointer and its target that is established by a pointer assignment statement. For more information, see “Pointer association status” and “Pointer assignment”.
Sequence association is the association that is established between dummy and actual arguments when they are arrays of different rank. For more information, see “Array dummy argument”.
Storage association is the association of different objects with the same storage area and is established by the EQUIVALENCE and COMMON statements. For more information about storage association, refer to the descriptions of the EQUIVALENCE and COMMON statements in Chapter 10 “HP Fortran 90 statements”. Derived-type objects that include the SEQUENCE statement in their definition can also be storage associated; see “Sequence derived type”.
Use association allows different program units access to module entities by means of the USE statement. For more information about modules and the USE statement, see “Modules”.

Terminology and concepts

Technical documentation

» Table of Contents

» Glossary

Program units

Procedures

Scope

Association