Digital PDFs

EY-1233E-ID-001

April 1983

88 pages

Original

2.9MB

Document:	ANSI-77 FORTRAN Information Document
Order Number:	EY-1233E-ID
Revision:	001
Pages:	88
Original Filename:

OCR Text

EY-1233E-ID-OOO 1

ANSI-77 FORTRAN
Information Document

Prepared by Educational Services
Of
Digital Equipment Corporation

Is t

@ Di 9 ita I

Eq u i pm en t

Ed i t ion, Ap r i I

I 98 3

Co r PJ rat ion I 98 3 •

The
information in
this document is
subject
to
change
without notice.
Dig ital Equi pment Corporation assumes no
responsibiliity for any errors that may
a ppear in thi s man u31 •
The software described in this document
is furnished under a license and may not
be used or copied except
in accordance
with the terms of such license.
Di g i ta I Eq u i pm en t Co r po rat ion ass um e s no
responsibility
for
the
use
or
reliability of its software on equipment
that is not supplied by Digital.
The following are trademarks of
Equi pmen t Co r po ration:
DIBOL
~D~DD~D
DEC
MASSBUS
DECrnate
PDP
DECnet
P/OS
DECsystem-IO Pro fe ssio nal
DE C S YS T EM - 20 Rainbow
DECUS
RSTS
DECwr iter
RSX

Dig ital

TOPS-IO
TOPS-20
UNIB US
VAX
VMS

Wo r k Processo r

5/83-15

CONTENTS
PREFACE
CHAPTER 1

INTRODUCTION

1. 1
1.2
1.2. 1
1.2.2
1.2.3
1.2.4
1.2.5
1.2.6
1.2.7

ION'rRODUCTION •••••••••.•..•••••.••••••••••••••••••• 1-1
CHARACTER CONVENorIONS •••••••••••.•••••••••••.••.•. 1-1
Character Constants and Symbolic Constants .••••• 1-1
Character Variables and Arrays •••••••••••••••••• 1-2
Character Re1atiana1 Expressions •••••••••••••••• 1-3
Ch a r act erAs s i 9 nm en t S tat 9 men t s • • • • • • • • • • • . . • • • • 1-- 4
Sub s t r in 9 Re fer en c e and De fin i t ion • • • • • • • . • • • • • • 1-4
Character Expressions ••••••.•••••••••••••.•••••• 1-S
Character Var iab1es and Constants in DATA
Sota tern en ts •••••••••••••••••••••••••••••••••••••• 1-6
Character Variables, Array Elements, Arrays,
and Substrings in Input Lists ••••••••••••••••••• 1-5
Chara~ter Constants, Variables, Array
Elements, Arrays, Substrings, and
Ex pr e S5 ion sin Ou t pu t Li s t s ••••••••••••••••••••• 1-7
Character Functions ••••••••.••••••••••••••••••.• 1-7
Dummy and Actual Arguments of Type Character •••• 1-7
Intrinsic Functions that Operate on
Character Data •••••••••••••••••••••••••••••••••• 1-7
Character Variable Names, Array Element Names,
Array Names, and Substring Names in
EQUIVALENCE Statemen ts •••••••••••••••••••••••••• 1-7
Character Variable and Arrays in COMMON Blocks •• 1-8
Charactar Variables and Expressions as the
Values of the Specifiers in OPEN Stat-aments ••••• 1-8
Character Constants, Character Arrays, and
Character Variables as Format Specifiers •••••••• 1-8
C ONTR OL S 'rA TEJ'v1 ENTS •••••••••••••••••.••.••••••••••• 1-8
BLOCK IF, ELSE IF, and END IF Statements •.•••••• 1-8
DO Loop Semantics ••••••••••••••••••••••••••.••• 1-10
INPUT/OUTPUT ••••••••••••••••••••••••••••••••••••• 1-10
Format Edit Descriptors •••••••••••••••••••••••• l-10
Expressions on Output Lists . . . . . . . . . . . . . . . . . . . . . 1-12
Internal Fi1es ••••••••••••••••••••••••••••••••• 1-13
Unit Specifier and Identifier •••••••••••••••••• 1-13
Format Specifier and Identifier •••••••••••••••• 1-14
ASSUMED SIZE ARRAY DECLARATORS ••••••••••••••••••• 1--14
USE OF A FOR!'1AT S'rATE1'v1ENT LABEL IN AN ASSIGN
STATEMENT •••••••••••••••••••••••••.•••••••••••••• 1-14

1.2.8
1. 2.9
1.2.10
1.2. 11
1.2.12
1.2.13
1.2.14
1.2.15
1.2.16
1.3
1.3. 1
1.3.2
1.4
1.4.1
1.4.2
1.4.3
1.4.4
1.4.5
1.5
1.6

iii

CONTENTS

(Cont)

1.7
1.8
1.9
1.10

IN'rRINSIC AND EXTERNAL STATE!~ENTS ••..•.....••.•.• 1-14
SAVE STATENlENT ...•.••.•.••..•.....••.•.•.•..••..• 1-15
NULL ARGUMENr LISTS FOR FUNCTIO~S •..•.••....•.••• 1-15
CONSTANT EXPRESSIONS ••.•..•..•..•...••...•.••.... 1-15

CHAPTER 2

TOPS-IO/20 FORTRAN

2.1
2.2

MAJOR FEA'rURES OF VERSION 7 'rHAT WERE No'r IN

I·NTRODUCTION .•.•.•••••.•...••..•••...•••.•..••••.• 2-1
VE R S ION

2.2. 1
2.2.1.1
2.2.1.2
2.2.1.3
2.2.1.4
2.2.1.5
2.2.1.6
2.2.1.7
2.2.1.8
2.2.2
2.2.3
2.2.4
2.2.5
2.2.6
2.2.7
2.2.8
2.2.9
2.2.10
2.2.11
2.2.12
2.2.13

6......................................... 2 - 1

Features Supported for Character Data •...•.•••.. 2-1
Character Assignment Statements ..•.•••••.•.•.• 2-1
Character Expressions •••.•...••••..••..•..•.•• 2-2
Character Variables and Constants in DATA
Sta temen ts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2
Character Variables, Array Elements, Arrays,
and Substrin.~s in Input Lists •••.•.•.•...••••• 2-2
Dummy and Actual ArguInents of Type Character .• 2-2
EQUIVALENCE Statements ...•••••.•...•••.•...••. 2-~
Character Variables and Arrays in COMMON
Blocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2
Namel ists ••••••..•••.•••..•.•.••.•••••.••..•.. 2-2
IF THEN ELSE Statements •.•.••••••...••••..•••••. 2-3
Expressions on Output Lists ••.•.••••••.•..•••••• 2-3
In tr insic and Gener ic Func tions a t the
li'OR'rRAN-77 Full Lang uag e Level ••••••.•••.••••••• 2-3
Internal Files (.single-Record and Mu1tirecord) •. 2-3
FORTRAN-77 DO Loop Semantics •••..•••••••••.•.••• 2-4
Assumed-Size Array Declarators •••••••••••••••••• 2-4
Use of FORMAT Statement Numbecs in ASSIGN
Sta tamen ts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-4
·INTRINSIC Statement; EXTERNAL Statement •••.••••• 2-4
SAVE Statement ••••.••••••.••••••••••••••••.••••• 2-4
Null Arg umen t Li st s fo r Func tio ns ••••.•••...•••• 2-4
.Minor Syntax Extensions Required by the
FORTRAN-77 Standard •••••••••••••.•••••••.••.•••• 2-4
Compile Time Constant Expressions in
Declar; tions, as Array Bounds, and String
Bo und s ••••••••••••.•••••••.••••••••••••••••••••• 2-5

2.2.14
2.2.15
2.2.15.1
2.2.15.2
2.2.15.3
2.2.16
2.2.17
2.2.18
2.2.19

FORTRAN-77 PARAME'rER Statements •.•••••.••••••••• 2-5
DO WHILE and END DO Statements ••••••••••••••.••• 2-5
Optional Statement Label in the Indexed
(Standard) DO Statement •.••••••••••••••••••••• 2-5
DO WHILE Statement •••••••••.•••••••••••••.•••• 2-n
END DO Statement •••••••••..••••••••••••••••••• 2-6
LINKt ime Type-Chec king 0 f Subprog ram Arg umen ts •• 2-6
G-F1oating Double-precision Numbers •••.••••••••• 2-7
Native TOPS-20 Command Interface for the
Co mp i 1 e r • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 2 - 7
New Functionality in the ERRSET Subroutine •••••• 2-7
iv

- - - - - - - - - - - - - - _.. _ - - - - - . - - -

--_. ._... _-_ ..._--_._-- - - - - - -

CONTENTS

(Cant)

2.2.20
2.3
2.4

Utility Subroutine to Get a Free Unit Number •••• 2-8
FORTRAN-77 FEATURES THAT ARE NOT SUPPORTED •••••••• 2-8
SUMMARY OF EXTENSIONS TO THE FORTRAN-77 STANDARD •• 2-9

CHAPTER 3

VAX-II FORTRAN

3.1
3.2
3.2.1
3.2.2
3. 2. 3
3.2.4
3.2.5
3.2.6
3.2.7
3.2.8
3.2.9
3.2.10
3.2.11
3.2.12
3.2.13
3.2.14
3.2.15
3.2.16
3.2.17
3.2.18
3.2.19
3.2.20
3.2.21
3.3
3.4
3.5
3.6
3.6.1
3.6.2
3.6.3
3.6.4
3.6.5

INrrRODUCTION ••••.•••••••••••••••••••••••••••.••••• 3-1
NEW FEATURES SUPPOR'rED BY VERSION 3 ••••••••••••••• 3-2
Bit Functions ••••••••••••••••••••••••••••••••••• 3-2
Debugger Commands for Source Code Debugg ing ••••• 3-3
DEFAULTFILE Keyword, in the OPEN Statement ••••••• 3-3
IMPLICIT NONE Statement ••••••••••••••••••••••••• 3-3
FORTRAN Data Manipulation Language Preprocessor.3-4
Faster I/O Interface •••••••••••••••••••••••••••• 3-4
Zero-Extending Intrinsic Functions •••••••••••••• 3-4
Library-Based INCLUDE Statement ••••••••••••••••• 3-4
Improved Math Routines •••••••••••••••••••••••••• 3-5
Namelist-Directed I/O Statements •••••••••••••••• 3-5
Optimization of Generateg Code •••••••••••••••••• 3-5
Optimization of I/O Routines •••••••••••••••••••• 3-5
Checking for Extensions to the FORTRAN-77
Standard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-5
A Cross-Reference Listing ••••••••••••••••••••••• 3-5
Floating-Underflow Checking ••••••••••••••••••••• 3-6
Sub s t r in 9 - B 0 un d s Ch e c kin g • • • • • • • • • • • • • • • • • • • • • • • 3 - f)
OPTIONS Statement ••••••••••••••••••••••••••••••• 3-6
FORTRAN Definitions for System Symbols •••••••••• 3-6
Trigonometric Functions in Degrees •••••••••••••• 3-6
Run - Tim e I/O Err 0 r Me s sag e s • • • • • • • • • • • • • • • • • • • • • 3 - 6
/SHOW Qualifier ••••••••••••••••••••••••••••••••• 3-7
E XTENS IONS TO THE ANS I STANDARD ••••••••••••••••••• 3-7
ADDI'rIONAL FEATURES •••••••••••••••••••••••••••••• 3-10
VAX-II FOR'rRAN COMPILER OP'rr.r>4IZATIONS •••••••••••• 3-10
COMPATIB ILITY:
VAX-II FORTRAN AND FORTRAN-66 •••• 3-10
DO Loop Minimum Iteration Count •••••••••••••••• 3-11
E XT E RNA L S tat e men t • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 3 -11
OPEN Statement Keyword Defaults •••••••••••••••• 3-11
OPEN Statement' Status Keyword Default •••••••••• 3-11
X Format Edit Descriptor ••••••••••••••••••••••• 3-12

CHAPTER 4

PDP-II FORTRAN

4.1
4.2
4.2.1
4.2.1.1
4.2.1.2
4.2.1.3
4.2.1.4
4.2.1.5

INTRODUCTION •••••••••••••••••••••••••••••••••••••• 4-1
NEW FEATURES ••••••••••••••••••••••••••• ~ •••••••••• ~-2
Fea t ure s Suppo r ted fo r Cha r ac te r Da ta ••••••••••• 4-2
Character Constants ••••••••••••••••••••••••••• 4-2
Substr ing Reference and Definition •••••••••••• 4-2
Character Expressions ••••••••••••••••••••••••• 4-2
Character Variables and Arrays •••••••••••••••• 4-2
Charac ter Relational Expressions •••••••••••••• 4-2

CONTENTS
4.2.1.6
4.2.1.7

(Cont)

4.4
4.5
4.6
4.6.1
4. 6. 2
4.6.3
4.6.4
4.6.5
4.6.6

Character Assignment Statements ••••••••••••••• 4-2
Character Variables and Constants in DATA
Sta temen ts •••••••••••••••••••••••••••••••••••• 4-3
Character Functions ••••••••••••••••••••••••••• 4-3
IF rrHEN ELSE Sta temen ts ••••••••••••••••••••••••• 4-3
FEATURES OF THE FULL-LANGUAGE FORTRAN AS DEFINED
BY 'rHE ANS I STANDARD •••••••••••••••••••••••••••••• 4-3
Exponentiation Forms •••••••••••••••••••••••••••• 4-3
Format Edit Descriptors ••••••••••••••••••••••••• 4-3
IN'rR INS IC and EXTERN.~ L Sta temen ts ••••••••••••••• 4-4
Generic Function Selection Based on Argument
Da ta Type fo r FORTRAN-De fi ned Func tio ns ••••••••• 4-4
PARAMETER Statements •••••••••••••••••••••••••••• 4-5
General ized DO Loop Parameters •••••••••••••••••• 4-5
Lower and Upper Bound s Spec ification in Array
Declarators ••••••••••••••••••••••••••••••••••••• 4-5
Optional Syntax for I/O Statements (UNIT= and
F1VIT~) •••••••••••••••••••••••••••••••••••••••••• • 4-5
EXTENSIONS TO THE ANSI STANDARD ••••••••••••••••••• 4-5
ADDIT IONA L FEATURES ••••••••••••••••••••••••••••••• 4-7
COMPATIBILITY:
PDP-II FORTRAN AND FORTRAN-66 ••••• 4-7
DO Loop Minimum Iteration Count ••••••••••••••••• 4-7
E XT ERNA L S tat e men t • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 4 - 7
OPEN Statement Keyword Defaults ••••••••••••••••• 4-7
OPEN Sta temen t Sta tus Keywo rd De fa ul t ••••••••••• 4-8
Blank Common Block PSECT (.$$$$.) •••••••••••• ~ •• 4-8
X Format Edit Descriptor •••••••••••••••••••••••• 4-8

CHAPTER 5

A COMPARISON OF FORTRAN LANGUAGE FEATURES

4.2.1.8
4.2.2
4.3
4.3.1
4.3.2
4.3.3
4.3.4
4.3.5
4.3.6
4.3.7
4. 3. 8

TABLES
Table No.
3-1
4-1
5-1
5-2
5-3
5-4
5-5
5-6
5-7
5-8
5-9
5-10
5-11

Title

Tr ig onometr ic Func tions •••••••••••••••••••••••••• 3-6
Allowed Combinations of Types of Base and
Ex I?<> nen ts ••••••••••••••.•.••••••••••••••••••.•••• 4-3
Prog ramming Considerations ••••••••••••••••••••••• 5-1
Subprogram Statements •••••••••••••••••••••••••••• 5-3
Constant and Variable Types •••••••••••••••••••••• 5-4
Da ta Type s ....•....................•.•..••....•• . 5-5
Spec ification Statements ••••••••••••••••••••••••• 5-6
Data Initialization Statement •••••••••••••••••••• 5-8
Relational Operators ••••••••••••••••••••••••••••• 5-8
Logical Operators •••••••••••••••••••••••••••••••• 5-9
Ass ig nmen t Sta temen ts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-9
Control Statements •••••••••••••••••••••••••••••• 5-10
Sequential I/O Statements ••••••••••••••••••••••• 5-l2

vi
- - - - ---------------- -----

Page

TAB LE S

Table No.

5-12
5-13
5-14
5-15
5-16

( Co n t)

Title

Page

F i 1 e Co n t r 0 1 Stat erne n t s • • • • • • • • • • • • • • • • • • • • • • • • • 5-1 5
Direct Access I/O Statements •••••••••••••••••••• 5-15
Indexed I/O Statements •••••••••••••••••••••••••• 5-16
For mat and Typ e s 0 f Co n v e r s ion • • • • • • • • • • • • • • • • • • 5 -1 7
Library Functions ••••••••••••••••••••••••••••••• 5-19

vii

PREFACE
American National
Standard
programmin';J
Language
FORTRt\N,
X3.9-l978,
specifies
the
form
and
establishes the
ANSI
interpretation of programs expressed
in
the
FORTRAN language.
Its purpose is to promote portability of FORTRAN programs for use
on a var iety 0 f data processing system s.
FORTRAN 77 is a
reVISIon of American National
Standard
FORTRAN,
ANSI X3.9-l966.
It describes two levels of the FORTRAN
1 ang uag e, re ferred to as FORTRAN and Sub se t FORTRAN.
FORTRAN 77
includes the
subset,
American National Standard Basic FORTRAN,
ANSI X3.l0-l966.
This information document provides information about ANSI-77
standard FORTRAN and Dig ital' s FORTRAN features under TOPS-lO/20,
VAX, and PDP FORTRAN.
The primary functions of this document are
to
serve as a
reference point to and summary of more detailed
information, and to serve as a quick review of new features and
changes, such as the comparison and
contrast of different
versions or of different operating systems.
The
target audience
for
this
document
is
software
s pe cia lis t s wh 0 se res po n sib iIi tie sin c 1 ud e pr e sa 1 e, ins tall at ion,
and warranty support 0 f FORTRAN.
These spec ial ists should
have
the
following
prerequisite
skills and
training:
programming
experience; a knowledge of FORTRAN; and user-level knowledge of
the
TOPS-IO/20,
VMS, and the several operating systems that run
on the PDP-li family of computers.
This document is not a
tutorial in the application of FORTRAN.

-----------------------

CHAPTER 1
INTRODUCTION
1.1 INTRODUCTION
The
FORTRAN language was originally developed
for
solving
mathematically oriented problems.
The engineering and scientific
communities, as well as the business community use it extensively
for analytical
work.
FORTRAN appl ications have evolved
to
include more general appl ications involving
character and
file
manipulation.
In March 1976 a new standard was drafted to ensure
portability of FORTRAN programs that include the advanced
file
and character handling
features.
The
final version of this
standard was approved on 3 Apr il 1978 and is referred to in this
document as the FORTRAN-77 or ANSI-77 standard.
The following paragraphs summarize the appropriate
sections of
the
FORTRAN-77 full language standard (ANSI X3.9-1978).
(These
sections reflect the major areas of change involved
in the new
v e r s ion s 0 f
TO PS -1 0/2 a,
VM S ,
and
th e PD P fa mil y 0 f 0 pe rat in g
systems.) This is not meant to be a complete delineation of all
the changes that were added
to
the
previous standard (ANSI
X3. 9-1966).
Descriptions of these changes are
found
in the
Amer ic an Na tio nal
Stand ard
Prog r amm ing
Lang uag e FOR TRAN, ANS I
X3.9-1978 publ ished bOy-the ~errc an--Na tional Standard s -'InstItute.
'rhe following paragraphs refer to specific sections of this book.

1.2

CHARACTER CONVENTIONS

1.2.1 Character Constants and Symbolic Constants
A character datum has one character storage unit in a
storaSJe
sequence for each character in the datum.
A storage sequence 1S
a sequence of storage units (either numeric or character).
If a
datum
requires more than one storage unit in a storage sequence,
these storage units are consecutive.
(Section 2.13)
A character datum is a string of
characters consisting
of any
characters capable of being represented in the processor.
(Blank
characters are valid and significant.) The character position of
each character in the string is numbered consecutively from left
to rig ht •
(Sec tio n 4. 8)

A character constant is a string of printable ASCII characters
enclosed by apostrophes.
Blanks between delimiting apostrophes
are significant, but the delimiting apostrophes are not counted
1-1

INTRODUCTION
as part of the datum.
Within a character constant,
the
apostrophe
character
is
represented
by
two consecutive
apostrophes with no space or other character between them.
A character constant has the form:
'clc2c3 ••••• cn'
where cl,c2,c3, ••• ,cn are printable characters.
A character constant expression is a character expression in
which each operand is a character constant, the symbol ic name of
a char ac ter constan t, 0 r a c ha r ac ter constan t ex pr e ssio n enc losed
in parentheses.
(Variables, array elements, substrings, and
function references are not allowed.) (Section 6.2.3)
A parameter statement allows constants to be defined symbolically
during compilation.
Its form is:
PARA.I'V1E'fER (p=e (,p=e]

where p is a symbol ic name and e is a constant expression.
The constant acquires the same data type as the
symbol ic name.
In particular, if P is a character constant expression, e must
correspondingly be a character constant expression.
The scope of
a
parameter
is .the
program unit in which it is declared.
The
data type of a symbol ic name is specified by a type-statement or
IMPLICIT statement preceding
the defining PARAMETER statement.
The default length for the symbolic name is one;
other lengths
c an be spec i fi ed
in type- sta temen ts 0 r
IMPLIC 1'f sta temen ts
preceding the symbolic name.
(Refer to section 8.6.)
rrhe following

is an example of the parameter statement.

CHARACTER DE UfA, EPS ILON*3
PARAMETER (DELTA='DIFF',EPSILON='SUM' ,E=2.7)
1.2.2 Character Variables and Arrays
The form of a character type-statement is:
CHARACTER [*len (,]]

nam [,naml

where nam is of the form:
v (* lenl
a [(d)] [*len]
[v is a var iable name, symbol ic name 0 f a constan t, func tion
name, or dummy procedure name; a is an array name; a(d) is an
array declarator;
len is the number of characters 0 f a character
1-2

IN'rRODUC TION
variable, character array element, or character function.]
to se c t ion 8. 4. 2. )

(Refer

'rhe length specification Gan be an unsigned integer constant, a
positive-valued integer constant expression in parentheses, or an
asterisk in parentheses.
In the sta tem en t :
CHARACTER*4 ALPHA,BETA*1,GAMMA(lO)*8
the following character declarations are made.
ALPHA has a length specification of 4
BE'rA ha s a l eng th spec i fic ation 0 f 1
GAMM A has a 1 eng th s pe c i f i cat ion 0 f 8

1 • 2. 3 Ch a rae te r Re 1 at ion a I Ex pr e s s ion s
A character relational expression has the form a reI b where a
and b are character ·expressions and reI is a relational operator.
(Refer to section 0.3.4.)
This expression is interpreted as
logically true if the values of the operands satisfy the relation
specified by the operator;
it is interpreted as false
if the
relation is not satisfied.
(Section 6.3.5)
'rhe six relational operators are:
Less than
Less than or equal to
Equal to
Not equal to
Greater than or equal to
Greater than

• LT.
• LE.

• EQ.

.NE.
.GE.
.GT.

In character relational expressions, less than means "precedes in
the ASCII collating sequence." For example, the expression:
'ABCD'

.LT.

'CDEF'

states that
'ABCD'
is less
than
'CDEF'.
Because
this
relationship does exist, the value of the expression is true.
If
the relationship did not exist, the value of the
relationship
wo uld be fal se •
If the two character expressions in a relational
expression
not the same length, the
shorter of the two
is padded on

1-3

are
the

I N'rR ODUC TI ON

right with spaces until the lengths are equal.
the relational expressions:
'ABC'
'AB'

.EQ.
.LT.

For

example,

'l\BC'
'C'

the first has a value of true, even though the lengths of the
expressions are not equal;
and the second has a value of true
even though 'AB' is longer than 'C'.

1.2.4
A

Character Assignment Statements
character assignment statement has the form:

= e

where v is the name of a
element,
or
character
ex pr e s s ion.

character variable,
sub str ing ;
and
e

c ha r ac te r
a r ray
is a
character

Execution of this statement causes e to be evaluated and v to be
assigned
the value of e.
V and e may be of different lengths.
(Section 10.4) In the full language, only as much of the value of
e must be defined as is needed to define v. For example:
CHARACTER X*4,Y*5
X=Y

doe s not r e qu ire th a t

th e sub s t ring Y ( 4 : 5) be d e fin e d •

1.2.5

Substr ing Reference and Definition
character
is a contiguous segment of a
variable or character array element.
A character sub str ing is
identified by a substring name and can be assigned val ue sand
referenced.
(Section 5.7)
A character substring

character substring reference has one of the following
v( [el]: [e2])

a(s[,s] ••• )

([e1]:[e2])

where:
•

v is a character variable name

•

a is a character array name

1-4

forms:

INTRODUCTION
•

s i s a sub sc r i pt ex pr e s s ion

•

el is an optional numeric expression that specifies
leftmost character position of the substring

the

•

e2 is an optional nurner ic expression that spec ifies
rightmost character position of the substring.

the

Character positions within a character variable or array
are numbered from left to right, starting with 1.

element

The values of al and e2 must satisfy the inequalities:
1 • LE.

e 1 • LE.

e 2 • LE.

len

to be
If el is omitted, its value
is implied
om itt ed, its val ue
is
i mpI i e d
to be 1 en.
character substr ing is e2 - el + 1.

1;
if e2 is
The 1 eng th 0 f a

If el or e2 is not an integer, it
is converted
to
an
integer
val ue b y t r un cat i ng its f r act i 0 na 1 pa r t •
( Sec t ion s 5. 7 • 1, 5. 7 • 2)
For example, in the statement:
CHARACTER TLC*8,XIN(3,4)*6
'rLC=' START'
XIN(2,3)='OUTPUT'
the substring name TLC(2:5) has a character value of II TART"
and
the substring name XIN(2,3) (4:6) has a character value of "PUrr."
In the
CHARACTER type
statement, character variable TLC is
declared as having
a string
length of 8 and each element of
character array XIN has a string length of 6.

1. 2. 6

Char ac ter Ex pressions
Character expressions (including
the concatenation operator)
A
consist
of character operands and c ha r ~c te rope r a to r s •
character operand can be:
1.

A character constan t

A symbol ic

A character variable

A character a r ray e 1 em en t

A character sub str ing

name

0 f

a c harac ter constan t

1-5

INTRODUCTION
6.

A
char ac ter
parenthe se s

A character function reference.

ex pression,

optionall y

The character operator is the concatenation
character expression of the form:

enclosed

(II).

operator

character operand [Ilcharacter operand] •••
for example x21lxl, is interpreted 'as xl concatenated with x2.
The resulting character string has the value of the value of xl
on the right concatenated with the value of x2 on the left.
The
length of this resulting character string
is the sum of the
lehgths of xl and x2.
Parentheses do not affect the value of a character expression.
Variables, array elements, substrings, and function references are
not allowed. (Sections 6.2, 6.2.1, 6.2.2.1, 6.2.2.2, 6.2.3, 6.6.5)
For example, the value of the character expression
i s the c ha r a c te r s t ring 'X YZAB CD' •

'XYZ'II'ASCD'

1.2.7 Character Variables and COnstants in DATA Statements
The fo rm 0 f the DA'rA sta temen t i s :
DATA nl ist/cl ist/ [ [ ,1

nl ist/cl ist/l •••

where nlist is a list of variable names, array names, array
element names, substring names, and implied-DO lists; and clist
is a constant 1 ist of the form a[ ,al ••• where each a is either a
constant, a
parameter, or a repetition factor of the form n*
followed by a constant or parame.ter
(denoting
n successive
appearances of the constant or parameter separated by commas) •
Fo rex am pI e :
CHARACTER*4 TEXT
DA Tl\ TE XT / ' YY XX ' I
illustrates the use of a character constant in a DATA
(Re fer to sec tion 9.4.)

statement.

1.2.8 Character Variables, Array Elements, Arrays, and Substrings
in Input Li sts
These are discussed in Section 12.8.2.1.

1-6

INTRODUCTION
1.2.9 Character Constants, Variables, Array Elements, Arrays,
Substrings, and Expressions in Output Lists
Input/output lists are used in READ, WRITE, or
PRINT statements
to spec ify data to be transferred.
(Section 12.8.2.2)
1.2.10 Character Functions
These
include intrinsic
functions, external
statement
functions that return values of
(Sections 15.2, 15.2.2, 15.4)

func tions,
and
type character.

1.2.11 Dummy and Actual Arguments of Type Character
These are discussed in section 15.9.3.1.
1.2.12. Intr insic Functions that Operate on Character Data
These functions are LEN, INDEX, CHAR, ICBAR, LGE, LGT,
LLE,
and
LLT.
(Refer
to
section 15.10.)
LEN gives the length of the
character entity.
INDEX(al,a2) gives the location of substr ing
a2 in string
ale
CHAR(i)
returns the character in the i-th
posi tion of the processer collating sequence.
The val ue
is 0 f
type character of length one.
ICHAR gives the position of a
character in the processor collating sequence.
LGE(al,a2) returns the value true if al = a2 or if al follows a2
in the
ASCII standard collating sequence and returns the value
false otherwise.
LGT(al,a2) returns the value ·true if al follows
a2 in the standard collating sequence and otherwise returns the
value false.
LLE (al,a2) returns the value true if al = a2 or
if
al precedes a2 in the standard collating sequence, and otherwise
returns the value false.
LLT(al,a2) returns the value true if al
precedes a2 in the
standard collating sequence and otherwise
returns the value false.
1.2.13 Character Variable Names, Array Element Names, Array
Names, and Substr ing Names in EQUIVALENCE Statements
An EQUIVALENCE statement
is used
to
specify the
sharing
of
s to rag e
un its by two 0 r m0 r e en tit i e s i n a
pr og r am un it.
(Sections 8.2,
8.2.1,
8.2.3)
The
form
of the
EQUIVALENCE
s ta tern en t i s :
E QU I VA LE NC E ( nl i s t) [ , ( nl i s t) ] •••
where nlist is a list of variable names, array element names,
array names, and other character. substring names.
Character
strings may be equivalenced only with other character
strings.
The association is made between the first storage units occupied

1-7

INTRODUCTION
by entities appearing in
the equivalence 1 ist.
Any adjacent
characters may also have the same character storage unit and thus
may be associated.
For example:
CHARACTER A*6, C(4)*4
E QU I VA LE NC E ( A, C ( 2 ) )
causes the following associations:
Storage unit: 1 2 3 4 5 6 7 8
Variable A:
- - - - A - Array C:
- - C(2) -f- C(3)
1.2.14 Character Variable and Arrays in COMMON Blocks
The fo rm of a C01VlMON sta temen t i s :
COMMON [f[cb] f]

nlist [,] f[cb] f nlistl •••

where cb is a common block name and nlist is a list of var iable
names, array names, and
array decl arator s.
If a character
variable or array is placed in a common block, that block must
contain only character data.
(Section 8.3.1)
1.2.15 Character Variables and Expressions as the Values
Specifiers in OPEN Statements
These are discussed in section 12.10.1.

1.2.16 Character Constants, Character
Variables as Format Specifiers
The form of a FORMAT statement is:

Arrays,

and

the

Character

FORMAT fs
where fs is a format specification.
One of the options for
the
format
identifier is a character array name, character variable,
character array element, or character expression.
(Section
13.1.2)
1.3 CONTROL STATEMENTS

1.3.1 BLOCK IF, ELSE IF, and END IF Statements
FORTRAN-77 ha s add ed
BLOC K IF,
ELSE IF,
ELSE,
and
END
statements.
(Refer to sections 11.6, 11.7, 11.8, and 11.9.)

The IF-level of a statement s is nl - n2 where nl is the number
of block IF statements from the beg inning of the program unit up

1-8

- - _..

_ _. _ - - - - - - - - - - - - - - - - - - - -

....

_._-_.. _.. _-_._.._.- --_._-_ .. _..... _.__._ ......_._--_. -

INTRODUCTION
to and including s, and n2 is the number of END IF statements in
the
program unit
up to but not including s.
This facility
enables the processor to determine which IF-THEN-ELSE statements
correspond to eac h 0 the r •
The form of a BLOCK IF statement is:
IF (e) THEN
'it/here e is a log ical expression.
The processor executes a
BLOCK
IF statement by first evaluating e.
If the value of e is true,
and the IF block is empty, the program control is passed directly
to
the next END IF statement.
If the IF block is not empty, the
first statement of the IF block is executed.
If the value of e
is false, control
is transferred to the next ELSE IF, ELSE, or
END IF statement having
the
same
IF level
as the
BLOCK IF
statement.
(An
IF BLOCK consists of all
the executable
statements appearing between the BLOCK IF statement up to, but
no t inc 1 ud ing, the nex t ELSE IF, ELSE, 0 r END IF s ta temen t hav ing
the same IF level as the BLOCK IF statement) •
The fo rm of an ELSE IF sta temen t i s :
ELSE IF (e) 'rHEN
where e is a logical expression.
An ELSE IF
block consists of
all
the executable statements between the ELSE IF statement up
to, but no t
inc 1 ud ing, the nex t
ELSE
IF,
ELSE,
0 r
END IF
statement having the same IF level as the ELSE IF statement.
The processor executes the ELSE IF statement by first evaluating
the expression e.
If the value of e is true, and the ELSE IF
block is empty, control is transferred directly to the next END
IF statement with the same IF level.
If the value of e is true,
and the ELSE IF block is not empty, the first statement of the
ELSE IF block is executed.
If the value of e is false, control
is transferred to the next ELSE IF, ELSE, or
END IF statement
having the same IF level as the ELSE IF statement.
The form of the ELSE statement is:
ELSE
The form of the END IF statement is:
END IF
The END IF statement performs no executable function, but
a s a po in t 0 f r e fe r e nc e •

1-9

serves

INTRODuc'rION
1.3.2 DO Loop Seman tic 5
'rhe fo rm of a DO sta tem~n t i s :
DO s

[,1

i = el, e2 [,e31

where s is the statement label of an executable statement;
i
is
the n am e 0 f a n
in teg e r,
rea 1 , o r d 0 ub 1 e- pr e cis ion v a ria b 1 e
(called the DO variable in FORTRAN-77);
el, e2, e3 are each an
integer, real, or double-precision expression.
El is the initial
value of the DO variable, e2 is the
limiting
value of the
DO
variable, and e3 is the increment value for the DO variable.
Two new features of FORTRAN-77 DO loops are:
1.

The possibility of zero
trip loops
(The
FORTRAN-66
standard did
not specify how many iterations a DO loop
had.
Mo st
processo r s,
incl ud ing
Dig i tal
and
IBM,
executed loops at least once.)

The availability of the loop index
after
loop exit
(previously this was not available).
(Sections 11.10
t 11 r 0 ug h 11. 10 • 7)

1.4 INPUT/OUTPUT

1. 4. 1 Fa rm at Ed it De sc r i pta r s
item
size and format of a data
or items.
An edit descriptor specifies an editing function to be
performed on a data item or
items.
(Refer
to
sections 13. 2,
13.5, 13.5.3, 13.5.3.1, 13.5.6, 13.5.9.2.3)

A field descriptor describes the

The BN edit descriptor causes the processor
to
ignore all
the
embedded and trail ing blanks it encounters within a numeric input
field.
It ha s the fo rm:
BN
blanks
and
the
Th e
e f f e c t i s t ha t
0 f
act ua 11 y removing
fi eld • A field of all
ri'3ht-justifying
the
remainder
of the
blanks is treated as zero.
The BZ edit descriptor causes the
processor
to
treat all
the
embedded and trailing blanks it encounters within a numeric input
field as zeros.
It has the form:

1-10

INTR ODUC TION
The SP edit descriptor causes the processor
to
produce a
plus
character in any position where this character would otherwise be
optional.
It has the form:
SP
'rhe S8 edit descriptor causes the pro\;essor to suppress a leading
plus character
from
any position where this character would
normally be produced as an optional
character;
it has the
opposite effect of the SP field descriptor described below.
The
SS d esc r i pto r. ha s the fo rm :
SS
The S edit descriptor reinvokes optional plus characters
n umer ic 0 ut put fi eld s.
It ha s the fo rm :

(+)

The S descriptor counters the action of either
the
SP or
SS
descriptor by restoring
to
the
processor the decision-making
ability to produce plus characters on an optional basis.
the
The T ed it descr iptor spec ifies the
position, reI ative r.to
start of an external
record, of the nex t
character
to be
processed.
It has the form:
Tn

where the term n indicates the position in the external record of
the next character
to be processed.
The val ue of n must be
greater than or equal to 1, but not greater than
the number of
characters allowed in the record.
The TL edit descriptor is a
relative tabulation
tabb ing to the 1 eft.
It ha s the fo rm :

specifier

for

TLn

The term n specifies that the next character
to be transferred
from or
to
a record
is the nth character to the left of the
current character.
The value of n must be greater than or equal
to 1.
If the value of n is greater than or equal to the current
character
position, the
first character
in the
record
is
~pec i fied •
The TR edit descriptor is a
relative tabulation
tabb ing to the r ig ht •
It ha s the fo rm :
TRn

1-11

specifier

for

IN'rRODUC'rION
'fhe term n indicates that the next character
to be transferred
from or
to
a record
is the nth character to the right of the
current character.
The value of n must be greater than or equal
to 1.
The I field descriptor specifies decimal integer values.
the fo rrn :

has

Iw [ • en]

The corresponding I/O list element must be of integer data type.
On input, the I field descriptor specifies that w characters are
to be read
from
an external
file,
interpreted as a decimal
integer val ue, and
assigned
to
the correspond ing
I/O 1 ist
element.
On output, the I field descriptor specifies that the value of the
correspondin'j
I/O list element is to be transferred as a decimal
value, right-justified, to an external field w characters long.
If m is present, the external
field
consists of at least m
di'gits;
if necessary, zeros are added on the left to bring
the
total digits to ffi.
If the value exceeds the field width, the
en t ire fie 1 dis f i 11 ed wi t has t e r i s ks • 1ft h e val ue 0 f the 1 i s t
element is negative, the
field
will have a minus sign as its
leftmost, nonblank chatacter, provided
the
term w is large
enough.
Plus signs are optionally suppressed (at the discretion
of the processor) unless SP is specified.
rfh e G fie 1 d d esc r i pt 0 r s pe c i fie s rea lor d 0 ub 1 e pr e c 1 s 1 0 n val ue s ,
combin ing
Eor
F-type
formats accord ing
to the si ze 0 f the
number being output.
It has the form:
Gw.d [Ee]
The correspond ing
I/O 1 ist el emen t must be
0 f
real
or
double-precision data type, or it must be either the real or the
imag inary part 0 f a complex data type.
On input, the G field descr iptor does not differ from the
or D de sc r i pto r s .

On output, the G field descr iptor spec ifies that the val ue 0 f the
corresponding
I/O list element is to be transferred as a real or
double-precision value in either exponential or fixed-point
form
depending
on its magnitude, rounded to d decimal positions, and
right-justified, to an external field w characters long.

1.4. 2 Ex pressions on OUtput Li sts
An input/output list specifies the entities whose values
transferred by a data
transfer
input/output statement.
1-12

-------------~

~-~~--

-- - - - - - - - - ---- - - - - - - - - - - - - - - - - ---

-~----

are
An

INTRODUCTION
input/output 1 ist is a simple 1 ist, an impl ied-DO 1 ist, or
two
input/output lists separated by a comma. A simple input/output
1 ist i tern is one 0 f the following.
1.

A variable

2.
3.
4•

An array
A character substr ing
An e 1 em en t 0 fan a r ray

In addition, in the full language, a simple output list may be an
expression,
with
the exception of a character expression
containing a dummy argument with a length attr ibute specified
with
an
asterisk.
(Sections 12.8.2,
12.8.2.1,
12.8.2.2,
12.8.2.3)
Note that a constant, an expression involving
operators or
function references, or an expression enclosed in parentheses may
appear as an output list item but not as an input list item.
An

implied-DO list is of the form:
(dlist, i = el, e2 [,e3])

where i, el, e2, and e3 are as specified for the DO statement and
dl ist is an input/output 1 ist.

1.4.3 Internal Files
Internal files provide a means of transferring
and converting
data from internal storage to internal storage.
An internal file
is a character variable, character array element, character
array, or character
substring.
Reading and writing records is
accomplished by
sequential
access
formatted
input/output
statements.

1.4.4 Unit Specifier and Identifier
Th e fo rm

0 f

the un i t s pe c i fie r

is:

[UNIT=] u
vlhere u is an external
unit
identifier or an
internal
unit
identifier.
An external unit identifier is an integer expression
with a zero or positive value or is an asterisk,
identifying
a
particular
processor-determined
external
unit
that
is
preconnected for formatted sequential access.
(Section 12.3.3)

1-13

INTRODUCTION
1.4.5 Format Specifier and Identifier
Th e fo rm 0 f a fo rm at s pe c i fie r is:
[FMT=] f
where f is a statement label of a FORMAT statement, an
integer
variable that has been ASSIGNED the statement label of a FORMAT
statement, a character array name, any character expression other
than one involving concatenation of an operand whose length was
specified by an asterisk and that is not the symbolic name of a
constant, or an asterisk specifying list-directed formatting.
If
FMT= is omitted, the format spec ifier must be the second item
in
the control information list and the first item shall be the unit
specifier without UNIT=.
(Section 12.4)
1.5 ASSUMED SIZE ARRAY DECIARATORS
In an array declarator for a dummy array, the upper bound of the
last dimension may be spec ified by an *
(Sections 5.1.2,
5.1.2.2, 5.5)
The form of an array declarator is:

a (d [,d] ••• )
where a is the symbol ic name 0 f the array and d
decl arator •
A dummy array declarator is an array
array name is a dummy arg ument.

declarator

a
in

dimension
which

the

1.6 USE OF A FORMAT STATEMENT IABEL IN AN ASSIGN STATEMENT
'rhe form of a statement label assignment statement is:
ASSIGN s TO i
where s is a statement label of an executable statement or FORMAT
statement and
i
is an integer variable name.
(Sections 10.3,
12.4)
An integer variable assigned a statement number may be used only
as a
statement identifier in an assigned Go'ro statement or as a
format identifier in an input/output statement.

1.7

INTRINSIC AND EXTERNAL STATEMENTS

The INTRINSIC statement is used to identify a
symbolic name as
representing
an intrinsic
function.
(Refer to sections 8.7 and

1-14

I N'rR ODUC TION
8.8.) I t perm its the name to be used as an ac tual arg wnen t.
form of the INTRINSIC statement is:

The

INTRINSIC fun[ ,fun] •••
where fun is the symbolic name of
example:

intrinsic

function.

For

IN'rR INS IC SIN
CALL AIDS(A,B,C,SIN)
The form of the ExrrERNAL statement is:
EXTERNAL proc[,proc] •••
where proc is the symbol ic name 0 f an ex ternal proced ure, a dummy
procedure name, or a block data program unit.
The ExrrERNAL statement is used to identify a
symbol ic name as
an external
procedure or dummy procedure, and to
perm it suc h a name to be used as an ac tual arg umen t.

representin'~

1.8

SAVE STATEMENT

A SAVE statement is used to retain the definition of a variable,
an array, or a named common block after the execution of a RETURN
or END statement.
(Refer to Section 8.9.) The
form of a
SAVE
statement is:
SA VE

[a, [ ,a] ••• ]

where each a is a named common block name preceded
by a slash, a variable name, or an array name.

and

followed

1.9 NULL ARGUMENT LISTS FOR FUNCTIONS
The form:
fun ()
may be used to reference a function with a
(Section 15.2.1)

null

argument

list.

Statement functions with null argument lists may be declared.
A
null, parenthesized argument list appears in the declaration.
(Section 15.4.1)

1-15

I NTRODUC TION
The form of a statement function statement is:
f un

([ d , [ ,d • • • ]) = e

where fun is the symbol ic name 0 f the statement function, d is a
statement function dummy argument, and
e
is an expression.
Parentheses may optionally be used for null
argument 1 ists' in
FUNCTION statements.

1.10 CONSTANT EXPRESSIONS
Constant expressions (not just constants) may be used
in array
declarators,
PARAME'rER statements, and impl ied DO-lists in DATA
statements.
(Sections 6.1.3,
6.1.3.1,
6.2.3,
6.4.4,
6.7,
5.1.1.1, 8.6, 9.3)

1-16

CHAPTER 2
TOPS-10/20 FORTRAN
2.1 INTRODUCTION
The FORTRAN language as implemented on the
TOPS-IO and
TOPS-20
operating systems is compatible with and encompasses the standard
described in American National Standard FORTRAN, X3.9-1978 at the
subset level.
TOPS-IO/20 FOR"'rRAN--alSOsupportsmostOfeatures
from the FORTRAN-77 full language standard.
FORTRAN provides many extensions and additions to the
FORTRAN-77
standard
that greatly enhance
the
usefulness of FORTRAN and
increase its compatibility with FORrrRAN languages implemented by
other computer manufacturers.
The TOPS-IO/20 FORTR?\N Language Manual (AA-N383A-TK) reflects the
software as of Version-7ofthe FORTRAN-IO/20 object time system
(FORO'rS), and Version 7 of the
FORTRAN-I 0/20 debugging
program
(FORDDT).
This
manual describes the
FOR'rRAN language as
implemented for the TOPS-IO operating system (FORTRAN-IO)' and the
TOPS-20 operating system (FORTRAN-20) and notes any differences.
It supersedes AA-J127A-TK.
2.2 MAJOR FEATURES OF VERSION 7 THAT WERE NOT IN VERSION 6
2.2.1 Features Supported for Character Data
Version 7 supports character data as specified by the
full
language
FORTRAN-77
standard.
Supported
features include
character
assignments,
character
relationals,
substrings,
concatenation, and
char ac ter func tions and arg umen ts, incl ud ing
functions and dummy arguments of length *
Character data
is
supported
in DATA,
COMMON, and EQUIVALENCE statements; and in
formatted, binary, and image mode I/O.
2.2.1.1 Character Assignment Statements - Character constants may
be assigned
to
noncharacter variables.
The standard restricts
the assignment of character constants to character variables.
Version 7 extends the standard to support assignment statements
in which there is overlap between the left- and right-hand sides.
The results of such an assignment will be as if the expression on
the
right-hand
side were assigned to a temporary and then the
value of the temporary were assigned to the left-hand side.
2-1

TOPS-IO/20 FORTRAN
2.2.1.2 Character Expressions - Version 7 extends the standard to
allow concatenation of formal parameters that are length *
2.2.1.3 Character Variables and Constants in DATA Statements
For compatibility with previous versions, Version 7 supports the
use of character constants to initialize noncharacter variables.
2.2.1.4 Character Variables, Array
Elements,
Arrays,
and
Substrings in Input Lists - In addition to the A edit descriptor
for input/output I ist items of type character, Version 7 supports
the G edit descriptor.
The G edit descriptor functions as the A
edit desc:riptor
for
list items of type character.
R edit
descriptors are not supported for character data.
2.2.1.5 Dummy and Actual Arguments of Type Character - Version 7
extends the standard to provide support of character constants as
ac tual
arg uments correspond ing
to dummy arg uments tha t a r e
integer, real, double-precision, complex, or logical, as well as
character.
This feature does not work when the name of the
fun c t ion call ed i s i t s elf a d umm y a r g um en t •
I fan act ua I a r 9 urn en t i s 0 f type c ha r a c te rand i s not a con s tan t ,
the corresponding dummy must be of type character.
If a dummy
argument is of type character, the corresponding actual must be
of type character.
Actual
arguments may be longer
than
correspond ing
dummy
arguments.
Length * may be used for character dummy arguments.
2.2.1.6 EQUIVALENCE Statements - It is illegal to equivalence a
numeric variable to
a character variable.
Equivalencing
a
numeric variable to an unaligned character variable is fatal;
equivalencing
a numeric variable to
a word-aligned character
variable is nonfatal.
2.2.1.7 Character Variables and Arrays in COMMON Blocks - When a
character variable or character array is in a COMMON block, all
the entities in that COMMON block must be of type character.
If
both character and numer ic data are spec ified in the same COMMON
block, a nonfatal warning message is issued •
Variables other
than character var iables beg in on word boundar ies;
thus a C01~MON
block containing both character and numeric data would contain
unused character positions.
2.2.1.8 Name1ists - Version 7 supports
input, but not in namelist output.
2-2

-------.-.--

--~-------------.--.---~------

substrings

namelist

TOPS-lO/20 FORTRAN
Version 7 does not support global optimization of programs that
contain character data.
If the /OP'rIMIZE switch is specified for
suc h a prog ram, the warn ing d iag no st ic :
Global optimization not yet supported with
character data - 10FT ignored
is issued.

2.2.2 IF THEN ELSE Statements
Version 7 supports the block
statements.

IF,

ELSE

IF,

ELSE,

and

END

2. 2. 3 Ex pressions on Output Li sts
variable name, an array element
name, a character
substr ing
name, an array name, or any other
expression.
FORTRAN-20 extends the standard
to
support output
list expressions that include concatenation of operands of length
asterisk.

An output list item can be a

2.2.4 Intrinsic and Generic

runction.

the

rORTRAN-77

rull

Language Level
Version 7 supports all intrinsic and generic functions described
in section 15.10 of the
FORTRAN-77 standard.
The following
intrinsic functions are new in Version 7.
1.

DINT - Truncation for double-precision

ANINrr, DNINT - Nearest whole number

NINT,

DDIM - positive difference for double-prec ision

DPROD - Double-precision product of real arguments

ICHAR, CHAR, LEN, INDEX, LGE, LGT, LLE, LLT functions as described in the introduction

IDNINT - Nearest integer

Character

The following generic function names have been added:
ACOS,
AINT,
ANINT, ASIN, COSH, CMPLX, DBLE, DIM, LOG, LOGI0, MAX, MIN,
NINT, REAL, SINH, TAN, and TANH.
The second arg umen t to CMPLX is
now optional. The generic function name INT has been extended to
support arguments that are COMPLEX and IN'rEGER (as well
as REAL
and DOUBLE-PRECISION).

2.2.5 Internal Files (Single-Record and Multirecord)
Version 7 conforms to the F.ORTRAN-77 standard.

2-3

TOPS-lO/20 FORTRAN
2.2.6 FORTRAN-77 DO Loop Seman tic s
As an extension to the standard,
Version 7 supports "ex tend ed
range DO loops"
(transfer
into
the range of a
DO-loop is
permitted if a previous transfer out has occurred.)
2.2.7 Assumed-Size Array Dec1arators
Version 7 conforms to the FORTRAN-77 standard.
2.2.8 Use of FORMAT Statement Numbers in ASSIGN Statements
Version 7 conforms to the FORTRAN-77 standard.
2.2.9 INTRINSIC Statement; EXTERNAL Statement
In Version 7, if the name 0 f an intr insic function appears in an
EXTERNAL statement, that name is subsequently treated as the name
of a user-defined function.
(This is in accordance with the
FORTRAN-77 standard, but incompatible with previous versions of
FORTRAN 20.
In Version 6, an asterisk appearing in front of an
intrinsic name in an EXTERNAL statement is required to force that
name to become the name of some external procedure.)
2.2.10 SAVE Statement
In Version 7, if a FORTRAN overlay contains any local variables
that
are SAVEd, all writable storage in that overlay is
preserved.
If a named COI'1MON block is SAVEd, that common block
is preserved.
Blank COMMON is always preserved.
2.2.11 Null Argument Lists for Functions
Version 7 conforms to the FORTRAN-77 standard.
FORTRAN-77
2.2.12 Minor
syntax
Extensions Required by the
Standard
The comma is optional in the following:
DATA statements, COMMON
statements, assigned GOTO, and after the statement number in DO
sta tements.
Parentheses may optionally be used for
SUBROUTINE and CALL statements.

null

argument

I ists

Statement numbers are legal on nonexecutable statements.
Exponentiation to an integer power is allowed
expressions in DATA statements.

2-4

the

subscript

TOPS-IO/20 FORTRAN
2.2.13 Compile Time Constant Expressions in
Array Bounds, and String Bounds
Version 7 conforms to the FORTRAN-77 standard.

Declarations,

2.2.14 FORTRAN-77 PARAMETER Statements
Version 7 supports PARAIV1ETER statements in accordance
FORTRAN-77
stand ard •
Compi I e
time
ex pressions
mul ti pI ication, d iv ision, or ex ponentiation 0 f COMPLEX
not supported.

with the
involv ing
data are

In Version 6 the data type of a PARA.METER was determined by the
type of the constant;
in Version 7 the data type is determined
by its symbolic name.
In Version 6 the list of parameters is never enclosed
in
parentheses;
in Version 7
the list of parameters must be
enclosed in parentheses.
If the I ist of parameters is not
enclosed
in parentheses, the compiler assumes that it is not a
FORTRAN-77 PARAMETER statement and a warning message is issued.
(This warning message can be suppressed by compiling with the
/NOF77 (/F66) switch.)
In Version 7 (as in FORTRAN-77), PARAMETER statements may precede
type declaration statements except for
those statements that
specify the type of parameter.
In Version 6 the parameter may only be set to
simple constants;
in Version 7 the parameter may be set to a constant expression.
2.2.15 DO WHILE and END DO Statements
The DO WHILE/ END DO support involves the following
to the FORTRAN-77 standard.
2.2.15.1 Optional Statement Label in the

Indexed

enhancements

(Standard)

Statement - The syntax of the indexed DO statement is:

DO [ s [ ,]] v= e I , e 2 [ ,e 3 ]
where s is the label of the statement that terminates the I00p.
If s
is omitted, the loop must be terminated by an END DO
statement as discussed below.

2-5

TOPS-IO/20 FORTRAN

2.2.15.2 DO WHILE Statement - The
following syntax:
DO [s [ ,]] WHI LE

WHILE

statement

has

the

(e)

where s is the label of the statement that terminates the loop.
If s
is omitted, the loop must be terminated by an END DO
statement.
E is a log ical expression that is tested at the beg inning of each
execution of the loop, including the first.
If the value of the
expression is true, the statements in the body 0 f the loop are
executed;
if the value of the expression is false, control
transfers to the statement following the loop.

2.2.15.3 END DO Statement - The END DO statement has the syntax:
END DO
An END DO statement terminates the range of a
DO or DO WHILE
statement.
The
END DO statement must be used to terminate a DO
block if the
DO or
DO WHILE statement does not contain a
statement label.
It may also be used as a labeled terminal
statement if the DO or DO WHILE statement does contain a terminal
sta temen t I abel.

2.2.16 LINKtime Type-Checking of Subprogram Arguments
FOR'rRAN Version 7 and
LINK Version
5.1
provide
limited
type-checking
for character constants that are passed as actual
arguments that correspond to numer ic dummy arg uments. V'8rsion 7
has modified
the argument passing mechanism;
the argument
passing mechanism for quoted strings now involves passing
the
address of a descriptor
for
the
string rather than the word
address of the string (as is done
in Version 6).
These
two
methods of passing
arguments may b~ referred to as "passing by
descriptor" and "passing by address." If an actual
argument is
passed by descriptor and the corresponding formal is passed by
address,
LINK will
transform the actual
argument
into
a
passed-by-address
argument if the
following
conditions are
sa tisfied :
1.

rrhe arg umen t i s a constant.

The str ing is in the same section as the argument block.

The byte po inter word in the descr iptor
in the
user's
image is word-aligned.
(The object code generated by
FORTRAN Version 7 now includes descriptors for character
variables, primaries, and subprogram arguments) •

2-6

TOPS-IO/20 FORTRAN
No type-checking will be performed on calls involving
old REL
files since either the caller or the callee or both will not have
LINK argument descriptor blocks.
Version 7 also supports a new option to the DEBUG switch of the
form /DEBUG:PARAMETERS.
With this option specified, FORTRAN will
generate REL file blocks that specify that illegal argument type
mismatches should result in nonfatal error messages at load time.

2.2.17 G-Floating Double-Precision Numbers
FORTRAN-20 Version 7 provides support for
the
G-floating
double-precision number
format.
The exponent range for this
number format is 2.8D-309 to 8.9D+307.
G-floating
is
an
alternative
internal
format
for
double-precision, supported
only on KL model B processors.
The
user specifies G-floating format by specifying
the /GFLOATING
command line switch to the FORTRAN compiler. /NOGFLOATING (the
defaul t) spec ifies the old double-prec ision format.
REL files that use the two different double-prec ision formats ?re
not compatible.
If a
user attempts to LINK together programs
compiled with different values of the /GFLOATING switch, a
warning will be issued at LINK time.
2.2.18 Native TOPS-20 Command Inter face for the Compiler
The FORTRAN-20 Version 7 compiler's command line
interface now
provides
support
of
long
file names, .,?", and command
recognition.
COMPILE now works for any legal TOPS-20 file name.
However, a
user cannot do an EXECUTE or DEBUG of a long file
name.
(LINK does not yet support long
REL file names).
The
syntax for the EXEC commands, EXECUTE, DEBUG, and COMPILE, is not
affected by the new command scanner.

2.2.19 New Functionality in the ERRSET Subroutine
Version 7 provides ERRSET trapping
for
additional
classes of
errors. Also, the user can now write his own fix-up routines for
~rithmetic exceptions.
The calling sequence for ERRSET is:
CALL ERRSET (N)
CALL ERRSET (N,I)
or
CALL ERRSET (N, I, SUBR)
where N equals the maximum number of error messages to type and I
equals the error to which this call applies.
If
I equals -1 it

2-7

TOPS-10/20 FORTRAN
will trap to any of the following errors.
-1 is assumed.

o
1
4
5
6
8
9
10
11
12
21
22

If I is not specified,

Integer overflow
Integer divide check
Floating overflow
Floating divide check
Fl oa ting und er flow
Library routine error
Output field wid th too small
Input floating over flow
Input floating underflow
Input integer overflow
FORLIS warning s
Nonstandard usag e warn ing s

SUBR is the subroutine to calIon the
trap.
If
SUBR is not
specified, no routine is called on the arithmetic exception.
If
SUB R i s s pe c i fie d th e e f f e c t i s a s if:
CALL SUBR (I,

IPC, N2,

ITYPE, UNFIXED, FIXED)

were placed in the program just after the instruction causing the
trap.
I is the error number of the trap.
IPC is the program counter of
the trap instruction, or if the error number equals 9, IPC equals
the prog ram co un ter 0 f the FOROTS call.
N2 equal s
the
second
error number
(reserved for Digital).
ITYPE is the data type of
value.
UNFIXED is the val ue returned by the processor, and FIXED
is the value after the fix-up (can be changed by SUBR).
2.2.20 Utility Subroutine to Get a Free Unit Number
Version 7 provides an additional FORTRAN-supplied subroutine that
can be used
to get an unused unit number.
The routine FFUNIT
(first free unit) is called by:
CALL FFUNIT(IUNIT)
where IUNIT is an integer variable
available unit number by FFUNrr.

that

set

the

first

2.3 rORTRAN-77 FEATURES THAT ARE NOT SUPPORTED
1.

The INQUIRE statement (used
to
status of a file attribute) •

Comment lines and blank lines may not appear between an
initial line and
its first continuation line, nor may
the y a ppe arb e t we en two con tin ua t ion 1 in e s •
2-8

determine

the

current

TOPS-IO/20 FORTRAN
3.

The compile time expression in a
PARAMETER statement
cannot
contain
mul ti pI ication,
d iv ision,
or
exponentiation of COMPLEX data.

2.4 SUMMARY OF EXTENSIONS TO THE FORTRAN-77 STANDARD
I.

II.

III.

Programmer convenience
FOR TRAN-I i ke

FORDD'r:
command s

Optional
c hec king

LINKtime c hec king for subprog ram arguments

Sel ec tive suppression
warning s

compil e

time

and

run

time

User sel ec tion of d efa ul t
SWITCH. INI

swi tch

val ues

means

F•

Run tim e t r ac e b ac k 0 n
PAUSE

tr ac ebac k

INC LUDE s ta tt'~m en ts

On TOPS-20:
DIALOG mode

Interactive
array

debugger
c hac king

bound s

err 0 r s ;

with
and

opt ional

bounds

string

and recognition in compiler commands and

Structured programming
A.

DO WHILE statement

END DO statement

Syntax
A.

End-of-l ine commen ts

Multiple statements per line

2-9

TOPS-IO/20 FORTRAN
IV.

Very accurate single- and double-precision arithmetic
A.

36-bit integer (10.5 decimal digits)

36-bit floating-point (8 decimal digits)

72-bit floating-point (18 decimal digits)

8-bit floating-point exponent:

On 2060 systems, /GFL switch also provides for an II-bit
e x po n en t ;
1 0 * * ( - 3 0 8 ) to 1 0 * * ( + 3 0 8 )

10**(-38)

to 10**(+38)

I/O
A.

NAMELIST I/O

Logical device names provide run time device assignments

DIALOG mode
spec i fier s

IMAGE mode fil es

APPEND access to sequential files

DELETE, EXPUNGE,
CLOSE

Additional I/O statements

provides

run

SAVE,

time

PRINT,

selection

PUNCH

dispositions

REREAD read s prev io us record

TYPE outputs to the user's terminal

ACCEPT inputs from the user's terminal

ENCODE/DECODE formatted
rather than a file

reading/writing to an

2-10

--------------------------------

OPEN

array

TOPS-IO/20 FORTRAN
H.

DBMS-IO/20 support

SORT-IO/20 support

DIL-IO/20/VAX support

DI L ( data in te r c ha ng e l i bra r y) is a se t 0 f call ab I e sub r 0 uti n e s
that enables a COBOL or FORTRAN prog rammer to access and use data
that resides on another computer system.
The DIL allows
you to
pass data between programs on different systems or directly
,access records in files on other systems. You can use the DIL to
access a single record within a file and avoid having to transfer
the entire file to your system.
If the accessed data is of the
wrong
format or data type,
OIL prov ides the necessary data
conversion facilities.
To use the OIL in a multiple computer environment, the
must be interconnected by OECnet to form a network.
Refer to the Data
(AA-M58IA-'rK)-.-

Interch~~ge

Library

2-11

User IS

computers
Ver sion

CHAPTER 3
VAX-II FORTRAN

3.1 INTRODUCTION
VAX-II FOR'rRAN is an implementation of the
full
language ANSI
It
is a compatible superset of PDP-II
FORTRAN-77 standard.
FOR rrRAN-77.
(You can compile ex isting PDP-II FORTRAN-77 source
programs, as well
as new programs that incorporate features
available in VAX-II FORTRAN.) VAX-11 FORTRAN Version 3 is also
upwardly compatible from the previous versions of VAX-II FORTRAN
and VAX/VMS systems.
This means that object files produced by
VAX-II FORTRAN Version 2 can be correctly 1 inked and run on
VAX/VMS Version 3.
Executable images linked on previous versions
of VAX/VMS systems can be run without reI inking on VAX/VMS
Version 3 systems.
It is al so
po ssible
to
freel y mix
obj ec t
files from
prev io us ver sions 0 f VAX-II FORTRAN wi th Ver sion 3
obj ect files on Version 3 VAX/VMS systems.
However, there is no
backward
compatibility;
that is, it is not possible to link or
execute files from VAX-II FORTRAN Version 3 on previous versions
of VAX/VMS systems.
This chapter descr ibes the new features supported
in Version 3
and
also outlines the major features that are extensions to the
standard.
This information,
as
well
as
more
detailed
information, is fo und in the following documents.
Lang_~~g e

VAX-II FORTRAN
AA-D034C-TE

VAX-II FORTRAN User's Guide, (April 1982), AA-D035C-TE

VAX-II FORTRAN Installation Guide/Release
1982) AA-H953B-n-·---··_- ----.-----.--

The
following
prog rammer s.

documents

are

Re fe renc e Manual, (Apr il

interest

1982),

(Apr il

Notes
VAX-II

FORTRAN

VAX/VMS primer, AA-D030C-TE

VAX/VI'1S Command Language USer's Guide, AA-D023C-TE

VAX-II

-----

Symb~ic

Deb ugg er Re fe renc e Man ual, AA-D 02 6D-TE

'rhe VAX-II FORTRAN Lang uage Re ~e~~nc~ Manua!. descr ibes the
FORTRAN language elements supported by VAX-II FORTRAN.
It is
3-1

VAX-II FORTRAN
intended to be used as a reference manual
sour c e pr og ram s •

preparing

FORTRAN

The VAX-II FORTRAN User's Guide describes how to compile, link,
debug, and--eX"ecute
programs wr itten in the
VAX-II FORTRAN
language, using the facilities of the VAX/VMS operating
system.
It contains other information of interest to FORTRAN programmers,
such as FORTRAN input/output, error
processing, programming
efficiency, compatibility between VAX-II FORTRAN and VAX-II
FORTRAN 66, and compatibil ity between VAX-II FORTRAN and
PDP-II
FORTRAN.
The VAX-II FORTRAN Installation Guide/Release Notes contains
detailed
instructions for installing the"VAX"=-fl FORTRAN compiler
on the VAX/VMS operating system.
It also contains information
about the differences between VAX-II FORTRAN Version 3 and
previous versions of VAX-II
FORTRAN,
including
a
short
description of new features and a complete description of the
/SHOW qual ifier.

3.2 NEW FEATURES SUPPORTED BY VERSION 3

3.2.1 Bit Functions
The following
additional
have been added.
1.
2.
3•
4.
5•
6•

bit-manipulation

intrinsic

functions

IBSET sets a bit.
BTEST tests a bit.
IB C LR c I ear s a bit.
IB IT Sex tr ac ts a bit fie I d •
IS HF TC pe r fo r ms a c i r c ul a r sh i ft.
MVB I'r S mov e s a bit fie 1 d •

IBSET, BTEST, IBCLR, IBITS, and the subroutine MVBI'rS operate on
bit fields.
A bit field is a contiguous group of bits within a
binary pattern, specified by a
starting
bit position and
a
length.
IBSET,
IBCLR, and BTEST operate on I-bit fields and do
not require a length argument.
IBITS and
MVBI'rS operate on
general bit fields.
Both the starting position of a bit field
and its length are arguments to these intrinsics.
The MVBITS subroutine transfers a bit field
from
one storage
location
(source)
to
a
field
in a
second
storage location
(destination).
The call to MVBITS has the form:
CALL MVB I'rS (m ,i ,1 en,n ,j)
where m is an integer variable or array element that represents
the source location (that is, the location from which a bit field
is transferred);
i is an integer expression that identifies the
first bit position in the field transferred from m;
len is an
3-2

VAX-II FORTRAN
integer expression that
identifies the length of the
field
transferred
from m;
n is an integer variable or array element
that represents the destination location (that is, the location
to which a bit
field
is transferred);
and j is an integer
expression that identifies the bit in which the
transferr~d
bit
fie 1 d beg ins.
The MVBITS subroutine transfers len bits from positions i through
i + len - I of the source location (m) to positions j through j +
1 en - 1 0 f the des tin at ion 1 0 cat ion
( n) •
at her bit s 0 f
the
destination location and
all
the bits of the source location
remain unchanged.
The val ues 0 f i + len and j + len must be less
t han 3 2 •
IS HF TC s h i f t s a bin a r y pat t ern; a po sit i v e sh i ftc 0 un t
indicates a left circular shift, while a negative shift count
indicates a right circular shift.

3. 2. 2 Debugg er Command s for So urce Cod e Debugg ing
A new interface allowing FORTRAN users to
access
source code display fac il ity has been prov ided.
debugger commands can now be used:
1.
2.
3•
4.
5.

the debugger
The following

SET STEP SOURCE
TYPE rang e
EXAM INE/ SEARC H
SEARC H [rang e] II str ing"
SET SOURC E

Refer to the VAX-II Symbolic Debugge£

Refe~~nce

Manual.

3.2.3 DEFAULTFILE Keyword in the OPEN Statement
The DEFAULTFILE keyword has been added
to
the
OPEN statement.
DEFAULTFILE allows
you to
specify a default file name that
overrides the general FORTRAN defaults.

3.2.4 IMPLICIT NONE Statement
The I1VIPLICIT statement now has a second format.
IMPLIC IT NONE
The IMPLICIT NONE statement inval idates all impl ic it defaul ts and
causes error messages to be generated for names that are not
explicitly typed in a type declaration statement.
You must then
explicitly declare
the data types of all symbolic names in the
program unit.
If you specify IMPLIcrr NONE,
no other
IMPLICIT
statement can be included in the program unit.

3-3

VAX-II FORTRAN

3.2.5 FORTRAN Data Manipulation Language Preprocessor
An interface to an integrated FOR'rRAN data manipulation
language
preprocessor
has been provided
under
Version 3.
The FORTRAN
qualifier /DML controls the
use of
the
interface.
The
preprocessor produces an intermediate file of FORTRAN 'source code.
with FORTRAN DML commands expanded into FORTRAN statements.
The
compiler
then compiles the
intermed iate files and produces an
optional object and listing file.
The listing file includes the
DML statements and, optionally, their expansions into FORTRAN
source.
DML error messages appear in the correct place in
the
listing.
The
preprocessing and compilation are controlled by a
single qual ifier on the FORTRAN command so that the user does not
need to know that a preprocessor is present.
3.2.6 Faster I/O Interface
A faster I/O interface is now provided by the following changes.
1.

1m pI ied DO loops in I/O I ists have been optimized.

Sinqle-element, unformatted I/O lists now use a
in te r fac e.

A larger default record size is allowed for
I/O transfers.

special

unformatted

3.2.7 Zero-Extending Intrinsic Functions
The two new in tr insic
func tions IZ EXT and JZ EXT allow fo r
zero-extending
arguments.
FORTRAN normally converts a smaller
fixed-point data type
to
a larger
fixed-point data
type by
sign-extending the smaller value.
This means that the high-order
bits of the larger data type are set to the
same value as the
sign bit of the smaller data type.
The functions IZEXT and JZEXT
zero-extend a
value
to
either
INTEGER*2
or
INTEGER*4
respectively.
This means that the high-order bits of the larger
data type are set to zero rather than to
the
sign bit of the
smaller data type.
The generic function ZEx'r selects IZEXT or
JZEXT according to the setting of the /14 command qualifier.
3.2.8 Library-Based INCLUDE Statement
The INC LUDE sta tement ha s a second format.
The new format allows
you to include modules from VAX/VMS text libraries.
INCLUDE 'file-specification (module-name)

[/(NO]LIST]'

The INCLUDE statement specifies that the contents of a designated
file are to be incorporated into a compilation directly following
the sta tern en t •

3-4
- - - - - - ------------------

--------~------------------

VAX-Ii FORTRAN
3. 2.9 Improved Ma th Ro ut ine s
The accuracy of the following math
significantly.
1.
2.

routines

has

been

improved

SIN
COS

LOG

4.
5.
6.

ATAN
ASIN
ACOS

3.2.10 Name1ist-Directed I/O Statements
The following statements provide namelist-directed I/O.
1.

'rhe NAM E;LIS T s ta tem en t

The READ, WRI'rE, ACCEPT,
TYPE,
and
PRINT statements,
with a
namelist specifier
in place of format and I/O
lists

3.2.11 Opt.imization of Generated Code
The following improvements in compiler
im pI emen ted.

optimization

have

been

In-line expansion of statement functions

2•

El im ina t ion 0 fred un dan t s tor e

El imination of redundant argument (eduction for calls to
SIN and COS wi th the same arg umen ts

3.2.12 Optimization of I/O Routines
The speed of the
following
run-time
routines has been improved.
1.
2.

0 pe rat ion s

library

I/O

processing

Floating input conver sian
Fo rm at in te r pr e ta t ion

3.2.13 Checking for Extensions to the FORTRAN-77 Standard
The new FORTRAN qualifier /STANDARD provides optional
checking
for source code that does not conform to the FORTRAN-77 standard.
3.2.14 A Cross-Reference Listing
The new FORTRAN qualifier /CROSS-REFERENCE provides
cross-reference as part of the listing file.

3-5

optional

VAX-II FORTRAN

3.2.15 Floating-Underflow Checking
The
FORTRAN
qualifier
/CHECK=UNDERFLOW
floating-under flow chec king at r un time.

3.2.16 Substring-Bounds Checking
Th~
FORTRAN
qualifier
/CHECK=BOUNDS
substring-bounds checking at run time.

provides

optional

provides

opt ional

3.2.17 OPTIONS Statement
The OPTIONS statement overrides or confirms the
qua I i fie r sin a pr og r am un it.

FORTRAN

3.2.18 FORTRAN Definitions for System Symbols
A text library FORSYSDEF.TLB containing the
FORTRAN
fa r system symbol s i s now prov id ed •

command

definitions

3.2.19 Trigonometric Functions in Degrees
The trigonometric
functions in
Table
degrees or produce results in degrees.
Table 3-1

3-1

take

arguments

Tr igonometr ic Func tions

Func tion

Desc r iption

SIND, DSIND, QSIND

Sine functions \vith degree
arg uments

COSD, DCOSD, QCOSD

Cosine functions with degree
arg uments

TAND, DTAND, QTAND

Tangent functions with degree
arg uments

ASIND, DASIND,

QASIND

Ar'c sine wi th deg ree resu1 t

ACOSD, DACOSD,

QACOSD

Arc co sine wi th d eg ree resul t

ATAND, DATAND,

QATAND

Arc tang en t wi th d eg ree re suI t

ATAN2D, DATAN2D, QATAN2D

Two-argument arc tangent with
d eg r e ere s ul t

3.2.20 Run-Time I/O Error Messages
Additional diagnostic information is now provided with
I/O error messages.

3-6

run

time

VAX-II FORTRAN
3.2.21 /SHOW Qualifier
The /SHOW qualifier controls listing options for
INCLUDE
preprocessor-generated output, and the symbol table map.

files,

3.3 EXTENSIONS TO THE ANSI STANDARD
The following are some of the VAX-II FORTRAN-77 extensions to the
ANS I stand ard •
1.

Language elements for kayed
and
VAX-II RMS mul tikey ISAM fil es.

A set of data types
language FORTRAN-77

beyond

those

sequential
specified

access

for

full

units

for

LOGICAL*I, BYTE (synonymous)
LOGICAL*2
IN'rEGER*2
COMPLE X*16, DOUBLE COMPLEX
REAL*16
3.

Explicit specification of storage allocation
data types (e.g., REAL*8, INTEGER*4).

Data initialization in type declaration statements.

IMPLICIT NONE
sec tio n) •

DO, DO WHILE, END DO statements - As an extension to the
standard,
VAX-II
FORTRAN supports "extended
range
DO-loops (transfer
into
the
range of a
DO-loop is
permitted if a previous transfer out has occurred) •

statement

(described

The DO WHILE/ END DO support involves
enhancements to the FORTRAN-77 standard.

new

the

features

following

( stand ard)
Optional statement label in the
indexed
DO
statement.
The
syntax
of the
indexed
DO
statement is:
DO

[ s , ( ,]]

v= e 1 , e 2, ( e 3 ]

where
s
is the
label. of the
statement
that
terminates the loop.
If s is omitted, the loop must
be terminated by an END DO statement.

3-7

VAX-II FORTRAN
B.

The DO WHILE statement has the syntax:
DO [s[,1] WHILE

(e)

where s
is the label
of the
statement
that
terminates the loop.
If s is omitted, the loop must
be terminated by an END DO statement as discussed
below.
The
(e)
is a logical expression that is
tested at the beginning of each execution of the
1 00 p,
inc 1 ud ing
the
fi r st •
If the val ue 0 f the
expression is true, the statements in the body of
the
loop are exec uted;
if the val ue 0 f
the
ex pr e s s ion
is
fa 1 se, con t r 01
tr an sf e r s
to
th e
statement following the loop.
C.

The END DO statement has the syntax:
END DO
An END DO statement terminates the range of a DO or
DO WHILE statement.
The END DO statement must be
used to te rm ina te a DO bloc kif the DO 0 r
DO WHILE
statement does not contain a statement label.
It
may also be used as a labeled terminal statement if
the DO 0 r DO WHILE sta temen t does con ta in a te rm inal
sta temen t 1 abel.

Bit manipulation functions.

Hexadecimal and octal constants and Z and 0 format edit
descriptors applicable to all data types - The 0 field
descriptor specifies octal integer values; the Z field
descriptor specifies hexadecimal (base 16) values.

DEFINE FILE, FIND, ENCODE, DECODE, DELETE, REWRITE,
and
UNLOCK statements - The DEFINE FILE statement describes
direct-access sequential files that are associated
with
a logical unit number.
The ENCODE and DECODE statements transfe r d ate between
variables or arrays
in internal storage and translate
that data from
internal
to character
form, or
from
character
to
internal
form, accord ing
to
format
s pe c i fie r s •
The DELETE statement deletes records in relative files
and
in indexed files.
Specifically, it causes a record
to be mar ked as del eted;
record 5
so mar ked
are no t
acc essibl e to sub se quen tREAD 0 r REWRITE sta temen ts •

3-8

VAX-II FORTRAN
The
REWRITE statement transfers output
data
from
internal
storage to
the current record in an indexed
fil e.
The UNLOCK statement unlocks records in a relative or
indexed
file.
When a
record is locked, it cannot be
accessed by any other program or log ical unit.
10.

ACCEPT,
TYPE input/output statements
The
ACCEPT
statement transfers input data to internal storage from
external records accessed under the sequential mode of
access.
The
TYPE statement transfers output data from
internal storage to external records accessed under
the
sequential mode of access.

11.

USEROPEN subroutine invocation at file OPEN time
USEROPEN parameter has the form:

The

USEROPEN = P
where p is an external
function name.
The
USEROPEN
keyword
specifies a user-written external function that
con tr 01 s the 0 pe n i ng 0 f the f i 1 e •
12.

INCLUDE statement (described in new features).

13.

NAMELIST - Directed I/O (described in new features) •

14.

31-character identifiers that can
( $) and und e r 1 ine ( ).

15.

Comments allowed at the end of each so urce 1 ine.

16 •

De b ug

17.

Language elements that support the VAX-ll extended range
and
extended
precision floating-point architectural
features.

include

dollar

sign

s tat em en t sin so ur c e •

64- bit G-floating data type with an II-bit exponent
and 53-bit mantissa, which provides a
range of
0.56*10**-308 to 0.09*10**308 and a precision of 15
dec imal dig its

l28-bit H-floating data type with a 15-bit exponent
and
a 113-bit mantissa, which provides a range of
0.84*10**-4932 to 0.59*10**4932 and a precision of
33 decimal digits
(To execute G- and
H-floating data type extended
range instructions directly on the VAX-ll/780, both
the KU780 and the KE780 hardware options must be
3-9

VAX-II FORTRAN
present.
To execute these instructions directly on
the VAX-II/750, the KU750 hardware option must be
present.
The VAX-Il/730 does not require any
additional options for G- and H-floating data
type
instruction execution.)
18.

OIL support - 'rhe OIL (Data Interchange Library)
is a
set of callable subroutines that enables a COBOL or
FORTRAN programmer to
access and manipulate data on
another VAX or a
OECSYSTEM-20.
To use the OIL in a
multiple computer environment, the computers must be
interconnected by OECnet to form a network.

3.4 ADDITIONAL FEATURES
1.

Support for calls to VAX/VMS system service procedures

Generation of symbol
deb ugger

Generation 0 f c ro ss- referenc e 1 isting s

Generation of shareable code

Up to 255 actual arg uments in a CALL statement

Up to 250 named COMMON blocks per· subprog ram

tables

for

the

VAX-II

symbol ic

3.5 VAX-l! FORTRAN COMPILER OPTIMIZATIONS
1.

Constant fold ing

Optimizations of arithmetic IF, log ical
IF -'rHEN-E LSE

Common subexpression elimination

Removal of invariant expressions from DO loops

Allocation of general reg isters across DO loops

In line expansion of statement functions

IF,

and

block

3.6 CCMPATIBILITY: VAX-ll FORTRAN AND FORTRAN-66
The VAX-II FORTRAN
compiler
selects
FORTRAN-77
language
in te r pr e ta t ion s
by d e fa ul t • As are suI t ,
i t con ta ins c e r ta in

3-10

VAX-II FORTRAN
incompatibil ities wi th FOR'l'RAN implementations that are based
the previous standard, X3.9-1966.
The areas affected are:
I•
2.
3.
4.
5.

DO I 00 p min i In um i t era t ion co un t
EXTERNAL sta temen t
OPEN statement BLANK keyword default
OPEN statement STATUS keyword default
X format edit descriptor.

3.6.1 DO Loop Minimum Iteration Count
DO
The IF77 command qualifier controls the interpretation of the
loop minimum
iteration count.
In FOR'I'RAN-77, the body of a DO
loop is not executed if the' end condition of the loop is al read y
satisfied
when
the
DO
statement is executed.
In most
implementations of FOR'l'RAN-66, the body of a DO loop is al wa ys
exec uted at I east once.
3. 6. 2 EXTERNAL Statement
The IF77 command qualifier controls the
interpretation of the
EXTERNAL statement.
In FORTRAN-66 the EXTERNAL statement is used
to spec i fy tha t
a
symbol ic name is the
name 0 f
e i the r
a
user-defined
ex ternal
proced ure or a FORTRAN-suppl ied func tion.
In FORTRAN-77, the IN'fRINS Ie and EXTERNAL statements are used
to
accomplish this function.
3.6.3 OPEN Statement Keyword Defaults
In FOR'rRAN-77 the
OPEN statement BLANK keyword
controls the
interpretation of blanks in number input fields.
The FORTRAN-77
default is BLANK='NULL';
that is, blanks in numeric input fields
are
ignored.
The FOR'I'RAN-66 interpretation of blanks in numer ic
input fields is equivalent to BLANK='ZERO'.
If a logical unJt is opened without an explicit
OPEN statement,
VAX-II FORTRAN and FOR'I'RAN-66 both provide a default equivalent
to BLANK=' ZERO' •
The BLANK ke yword affec ts the
treatment 0 f bl anks
in n umer ic
input
fields read
with the
D,
E,
F,
G,
I, 0, and Z field
descriptors.
If BLANK='NULL' is in effect, embedded and trailing
blanks are
ignored;
the value is converted as if the nonblank
characters were right-justified in the field.
If BLANK='ZERO' is
in effect, embedded and trailing blanks are treated as zeros.

3.6.4 OPEN Statement Status Keyword Defaul t
In
FOR'rRAN-77,
the
OPEN statement STATUS keyword
specifies
ini tial
status
0 f
the
fil e
( 'OLD',
'NEW',
'SCRATCH', or
'UNKNOWN').
Tne FOR'I RAN-77 default
is STATUS='UNKNOWN';
that
is, an existing
file is opened or a new fil~ is created if the
1

3-11

VAX-II F ORrrRAN
file does not ex ist.
If you use the IF77 command
qual ifier and
you do not
specify STATUS (or TYPE) in an OPEN statement, the
compiler issues an informational message to warn you that it is
using
a default of STATUS='UNKNOWN'.
It is advisable to include
an explicit STATUS (or TYPE) keyword in every OPEN statement.

3.6.5 X Format Ed it Descr iptor
The nX edit descriptor causes transmission of the next character
to or from a record to occur at the position n characters to the
right of the current position.
In a FORTRAN-77 output statement,
character
positions that are skipped are not modified, and the
length of the output record is not affected.
However,
in many
FORTRAN-66
implementations, the X edit descriptor writes blanks
and may extend the output record.

3-12

---_._.._-_ ......_--_. __ ...-

. __ ._--------_.-

---

CHAPTER 4
PDP-II FORTRAN

4.1 INTRODUCTION
PDP-II FORTRAN is an extended implementation of the
ANSI subset
FORTRAN-77 standard.
Version 4 contains all the features 0 f the
ANSI FORTRAN-77 subset, many of the full-set language
features,
and extensions that are not included
in the ANSI FORTRAN-77
standard.
This chapter describes the new features supported
in Version 4
and also outlines the major features that are extensions to the
standards.
This information, as
well
as
more
detailed
information, is found in the following documents.
1.

PDP-II FORTRAN-77 Lang uage Reference Manual,
1981), AA-19791-TC

PDP-II FORTRAN-77
AA-1884D-TC

User's

PDP-II FORTRAN-77
AA-KS03B-TC-------

Install ation

The
PDP-II FORTRAN-77 Obj ect
(AA-1874C- rrC)"
isalso
of
programmers.

Guide

(September

1981) ,

Guid e/Rel ea se
-----------

Time System
interest-to

Re fe renc e
Man ual
PDP-flFoRTRAN-77

The PDP-II FORTRAN-77 Language
Reference Manual
describes the
syntax
and
semanticsOffhe-FORTRAN-77 implementation of PDP-II
FORTRAN.
It does not, however, present information spec ific
to
any operating system.
The PDP-II FORTRAN-77 User's Guide contains the
information
----_.- ------ ---necessary to create, link, and execute PDP-II FORTRAN-77 programs
on a PDP-II processor.
programming information is provided
for
the RSX-11M/M-PLUS, IAS, and RSTS/E operating systems.
The PDP-II FORTRAN-77 Installation Guide/Release Notes
the
procedures-fOr
install ing--PDP-ll
FORTRAN
RSX-IlM/M-PLUS, RSTS/E, and IAS operating systems.

4-1

describes
on
the

PDP-II FORTRAN

4. 2 NEW FEA TURES
Among the major features defined by the new ANSI subset lnnguage
FORTRAN standard and
not found
in either
the previous ANSI
standard or previous versions of Dig ital PDP-II FORTRAN are:
1.

CHARACTER data type

Block IF construct, including IF ••• THEN, ELSE IF,
ELSE,
and
END IF statements, for conditional execution of
bloc ks 0 f sta temen ts •

4.2.1 Features Supported for Character Data
Version 4 supports character data as specified
language FORTRAN-77 standard.

the

4.2.1.1 Character Constants - The length of a character
must be in the range 1 through 255.

subset

constant

4.2.1.2 Substring Reference and Definition - Version 4 supports
character substrings as outlined
in the full-language subset.
(Note that substrings are not included in the subset.)

4.2.1.3 Character Expressions
A character operand can be
character constant, character variable, character array element,
or character substring. A character expression has the form:
character operand
and can be enclosed in parentheses. Note that the concatenation
operator is not included in .the subset (nor in Version 4).

4.2.1.4 Character Variables and Arrays - The length specification
in the character
type-statement can be an unsigned integer
constant or
an
integer-constant
expression
enclosed
in
parentheses.
[An asterisk in parentheses (*) is not allowed].
When you specify CHARACTER*len, the length specification must be
in the range 1 to 255.

4.2.1.5 Character Relational Expressions - Version 4 conforms

the ANSI-77 full language standard.

4.2.1.6 Character Assigrunent Statements - Version 4 conforms to
the ANSI-77 subset language standard.
Note
that you cannot
assign a numeric value to a character variable, array element, or
substr ing •

4-2

PD P-ll FOR TR AN
4.2.1.7 Character Variables and Constants in DATA Statements
Version 4
(and the ANSI subset language) do not support impl ied
DO-lists in DATA statements.
Version 4 supports
LEN,
INDEX,
4.2.1.8 Character Functions No te tha t CHAR is no t suppo r ted.
IC HAR, LL'r, LLE, LGT, and LGE.
4.2.2 IF THEN ELSE Statements
Version 4 supports the block IF,
sta tements.

ELSE

IF,

ELSE,

and

END

4.3 FEATURES OF THE FULL-LANGUAGE FORTRAN AS DEFINED
BY THE ANSI STANDARD
4.3.1 Exponentiation Forms
(Th i s incl ud es doubl e-prec ision and compl ex
fo rms.)
Tabl e 4-1
summarizes the allowed combinations of data types of base and
exponent, and the data type of the result of exponentiation.
The
new features are underl ined.
Tab1 e 4-1 Allowed Combinations 0 f
Types of Base and Exponents
Base

Exponent
Integer

In teg er

Real

!buble

Complex

In teg er

Real
-

!b uble
---

Come~ex

Real

Do uble

Complex

Double

Complex

Compl ex

Com pI ex

4.3.2 Format Edit Descriptors
( Th i sin c I ud e s 5, 5 P , 5S , T, T L, TR , Iw. m, and Gw. d Ee .) Ve r s ion 4
conforms to the full FORTRAN-77 standard.

4-3

PDP-II FORTRAN
4.3.3 INTRINSIC and EXTERNAL Statements
Normally, the name of an intrinsic function refers to the FORTRAN
library function with that name.
However, the name can refer to
a user-defined function under any of the following conditions:
1.

The name is used in a function reference with arguments
of a different data type from that normally used.

The name appears in an EXTERNAL statement.

The EX'rERNAL and INTRINSIC statements enable the
programmer
to
use
subprogram names as actual arguments to other subprograms.
The semantics of the EXTERNAL statement are different in
FORTRAN
77 than
in
previous versions of PDP-II FORTRAN.
In previous
versions, the appearance of an
intrinsic
function name in an
EXTERNAL statement caused the processor to treat the name as the
name of an intrinsic function.
In FOR'rRAN 77, the appearance of
an intrinsic
function name in an EXTERNAL statement causes the
processor to treat the name as the name of an external
function.
In previous versions, an
intrinsic
function name had to be
preceded by an asterisk to be treated as an external
function.
The /NOF77 switch allows the programmer to select the previous
semantics, rather than FORTRAN-77 semantics.
The following shows
the equivalent statements.
/F77
EXTERNAL ex t
EXTERNAL in t
INTRINS IC in t

/NOF77
EXTERNAL ex t
EXTERNAL * in t
EXTERNAL in t

Except when they are used in an
EXTERNAL statement, intrinsic
function names are local to the program unit that refers to them.
Thus, they can be used for other purposes in other program units.
In addition, the data
type of an intrinsic function does not
change if you use an
IMPLICIT statement or an explicit type
declaration to change the impl ied data type rules.
You cannot have an intrinsic function and a user-defined function
wi th the same name in the same prog ram un it.

4.3.4 Generic Function Selection Based on Argument Data Type
FORTRAN-Defined Functions

for

Some intrinsic functions perform the same computation but handle
different data types.
These functions are references with the
same generic name. A generic-function reference refers to
the
category of the computation to be performed, not to a specific
function within the category.
The
selection of a
specific
function, that is, the actual computing procedure for a specific
data type, is left to the compiler, which chooses a
spec ific
fLlnc tion wi thin a category on the basi s 0 f the data type 0 f the
4-4

PDP-II FORTRAN
relevant
actual
ar~ument.
For
example,
if
D
is
a
double-precision variable, the generic function reference SIN (D)
refers to the double-precision sine function.
You need not write
DS IN (D) •
Generic function references are independent from one another.
Therefore, you could use both of the function references SIN (X)
where
X is a
real variable, and
SIN(D)
where
D
is
a
double-prec ision var iable, in the same prog ram unit.
The intrinsic and generic
functions
15.10 of the FORTRAN-77 standard.

are

described

section

4.3.5 PARAMETER Statements
Version 4 prov ides both the FORTRAN-77 and the earl ier
form of
the
PARAMETER statement.
The 1 ist in the earl ier form of the
PARAMETER statement is not bounded with parentheses, and the form
of the constant
(rather
than typing
of the
symbolic name)
determines the data type of the variable.

4.3.6 General ized DO Loop Parameters
Version 4 conforms to the full
FORTRAN-77
"extended range DO loops" are supported.

standard.

Moreover,

4.3.7 Lower and Upper Bounds Specifi,cation in Array Declarators
Version 4 conforms to the full FORTRAN-77 standard.
4.3.8 Optional Syntax for I/O Statements (UNIT= and FMT=)
Version 4 conforms to the full FORTRAN-77 standard.

4.4

EXTENSIONS TO THE ANSI STANDARD
1.

Language elements for keyed and sequential access to RMS
multikey ISAM files.

DEFINE FILE, FIND, ENCODE, DECODE, DELETE, REWRITE,
and
UNLOCK statements - The DEFINE FILE statement describes
direct-access sequential files that are associated
with
a log ical
unit number.
The OPEN statement, which can
also be used to describe direct-access sequential files,
is the preferred statement.
The FIND statement positions a direct-access file on a
specified unit to a particular record. No data transfer
ta ke s pl ac e •
The ENCODE and DECODE statements transfer data between
variables or arrays
in internal storage and translate

4-5

PDP-II FORTRAN
that data
from
internal
to character
form or
from
character
to
internal
form, according
to
format
specifiers. Similar results can be accomplished
using
internal
files with formatted sequential WRITE and READ
sta temen ts •
The DELETE statement deletes records in relative files
and
in indexed files.
Specifically, it causes a record
to be marked as deleted;
records so marked are not
accessible to subsequent READ or REWRITE statements.
The REWRITE statement repositions a
sequential
file
currently open for
sequential or append access to the
beg in n i ng 0 f the f i 1 e •
The UNLOCK statement unlocks records in a
relative or
indexed
file.
When a record is locked, it cannot be
accessed by any other program or log ical unit.
3.

TYPE and ACCEP'r input/output statements
The TYPE
statement transfers output data from internal storage to
external records accessed under the sequential mode of
access.
The ACCEPT statement transfers input data to
internal storage from external
records accessed
under
the sequential mode of access.

Comments permitted at the end of each source 1 ine.

INCLUDE statement - The INCLUDE statement specifies that
the contents of a designated file are to be incorporated
into a compilation directly following
the
statement.
INC LUDE ha s no e ffec t on prog ram ex ec ut io n.

BYTE data type - BYTE and LOGICAL*l are synonymous.

Explicit specification of storage allocation
data type s (e.g., INTEGER*4).

Hexadec imal and octal constants.

Virtual array support for systems with memory management
directives.
Virtual
arrays are memory-resident and
require enough main memory to contain all
elements of
all arrays.

10.

0 and Z format edit descriptors - The 0 field descriptor
specifies octal integer values; the Z field descriptor
spec ifies hexadec imal (base 16) val ues.

4-'5

units

for

- - - - - - - - - - - - -

--------------------------

PDP-II FORTRAN
4.5 ADDITIONAL FEATURES
The PDP-II FORTRAN-77 compiler produces direct PDP-II machine
code optimized
for execution-time efficiency on a PDP-II with a
floating-point
processor.
PDP-II
FORTRAN-77
compiler
o pt i m i za t ion sin c I ud e :
1.

Optimizations of arithmetic and logical IF statements

Common sUbexpression el imination

Removal of invariant expressions from DO loops

Allocation of processor
constr uc ts and DO loops.

registers

across

block

4.6 C<l-tPATIBILITY:
PDP-II FORTRAN AND FORTRAN-66
The
PDP-II FORTRAN
compiler
selects
FORTRAN-77
language
interpretations by defa ul t.
As a resul t, it contains certain
incompatibilities with FOR'rRAN implementations that are based
on
the prev io us stand ard, X3. 9-1966.
The area s a ffec ted are:
1•
2.
3.
4.
5.
6.

DO 1 00 p min im urn i t era t ion co un t
EXTERNAL statement
OPEN statement BLANK keyword default
OPEN statement STATUS keyword default
Blank common block PSECT
X format ed it descr iptor.

4.6.1 DO Loop Minimum Iteration Count
The IF77 command qualifier controls the interpretation of the
DO
loop mlnlmUffi
iteration count.
In FORTRAN-77, the body of a DO
loop is not executed if the end condition of the loop is al read y
satisfied
when
the
DO
statement is executed.
In most
implementations of FORTRAN-66, the body of a DO loop is al wa ys
executed at least once.
4.6.2 EXTERNAL Statement
The IF77 command qualifier controls the
interpretation of the
EXTERNAL statement.
In FORTRAN-66 the EXTERNAL statement is used
to specify that a
symbolic
name is the name of either a
user-defined
external
procedure or a FORTRAN-suppl ied function.
In FORTRAN-77 the INTRINSIC and EXTERNAL statements are
used
to
accomplish this function.
4.6.3 OPEN Statement Keyword Defaul ts
In FORTRAN-77 the
OPEN statement BI.J\NK keyword controls the
interpretation of blanks
in
the number
input
fields.
The
FORTRAN-77 default is BLANK='NULL'i
that is, blanks
in numeric
4-7

PDP-II FORTRAN
input
fields are
ignored.
The
FORTRAN-66
interpretation of
blanks in numeric input fields is equivalent to BLANK='ZERO'.
If a log ic al un it is 0 pened wi tho ut an ex pI ic it OPEN sta temen t,
PDP-II FORTRAN and FORTRAN-66 both prov ide a defaul t equivalent
to BLANK=' ZERO' •
The BLANK keyword affects the
treatment of blanks
in numeric
input
fields read
with the
0,
E,
F,
G,
1,0, and Z field
descriptors.
If BLANK='NULL' is in effect, embedded and trailing
blanks are
ignored;
the value is converted as if the nonblank
characters were right-justified in the field.
If BLANK='ZERO' is
in effect, embedded and trail ing blanks are treated as zeros.

4.6.4 OPEN Statement Status Keyword Default
In
FORTRAN-77,
the
OPEN statement STATUS keyword
spec ifies
initial
status
of the
file
('OLD',
'NEW',
'SCRATCH', or
'UNKNOWN').
The FORTRAN-77 default is STATUS='UNKNOWN';
that
lS,
an ex~isting
file is opened or a new file is created if the
file does not ex ist.
If you use the IF77 command
qual ifier and
you do not specifiy STATUS (or TYPE) in an OPEN statement, the
compiler issues an informational message to warn you that
it is
using
a default of STATUS='UNKNOWN'.
It is advisable to include
an ex pI ic it STATUS (or TYPE) ke ywo rd in ever y OPEN sta temen t.

4. 6. 5 B1 an k Common Bloc k PS EC T (. $ $ $ $ • )
Under PDP-II FORTRAN-77, the blank common block PSECT
(.$$$$.)
has the
SAV attribute.
The
SAV attribute on a PSECT has the
effect of pulling that PSECT into the root segment of an overlay.

4. 6. 6 X Fo rm a t Ed it

De sc r i pto r
The nX edit descriptor causes transmission of the next character
to or from a record to occur at the position n characters to the
right of the current position.
In a FORTRAN-77 output statement,
character
positions that are
skipped are not modified and the
length of the output record is not affected.
However,
in many
FORTRAN-66
implementation, the
X ed it descr iptor wr ites blanks
and may extend the output record.

4-8

--~-~-------

CHAPTER 5
A COMPARISON OF FORTRAN LANGUAGE FEATURES

Tabl es 5-1 thro ug h
5-16 prov ide
comparative
fea tur es.
Re fer al so to Lang uag e Fund amen tal s

Table 5-1
<-AMERlCAN->
NATlOOAL
STANDARDS
FORTRAN

X3.9 -77 (Full
1966 Language)

FORTRAN lang uag e
(AA-M460A-RK).

Programming Considerations

05/8 OS/78
FORT FORT
IV
IV

TOPS-l 0
<-PDP-11-> Vf>.:I../
RSX
RSX
VMS
TOPS-20
RSTS/E RSTS/E
RT-ll
lAS
Vf>.:I../11
lAS
(FOR)
(E77)
FORTRAN F-I0

<--PDP-8-->
05/8
FORT

Maximum number of dimensions
allowed for an array

Infinity

Mixed mode expressions

Double exponentiation
(e.g., A**B**C) permitted

Statement number size
(characters)

1-5

Maximum level of nesting
for 00 loops

-3

Infinity

Maximum number of characters
allowed in a PAUSE message

255

Infinity

Generalized subscripts
permitted

Adjustable dimensions
permitted in subprogram

Specification statement can
follow first executable
statement

Generic function selection

2 Right to left evaluation.
3 Undefined.
4 Produces warning.

5-1

A COl'-1.PARISON OF FORTRAN LANGUAGE FEATURES

Table 5-1

Programming Considerations (Cont)

X3.9 -77 (Full
1966 Language)

OS/8
FORT
II

TOPS-l 0
<-PDP-ll-> VAX/
RSX
VMS
RSX
RSTS/E RSTS/E
TOPS-20
!AS
OS/8 OS/78 RT-ll
FORT FORT lAS
VAX/II
(FOR)
(F77)
FORTRAN F-IO
IV
IV

Statement functions can
follow executable statements or precede related
specification statements

Maximum nunber of
continuation lines

Infinity

0-99

Infinity

Embedded blanks permitted
in key words

Key words reserved by the
compiler

Maximum characters in a
symbolic name

Maximum level of nesting
for implied 00 loops

-1

-2

Infinity

Comment line starts with

C,*

C,D,1

C,D,
* ,1

C,$,*,
/ ,D,1

Source code in EBCDIC

-1

Source code in BCD

Source code in ASCII

-1

End-o f-l ine corranents
delimited by "1"

<-AMERlCAN->
NATIONAL
STANDARDS
FORTRAN

Label on any statement

<--PDP-8-->

INCLUDE from source file

INCWDE from test library

1
2
3
4
5
6

Not specified.
Undefined.
Warning given for >6.
Warning given.
Warning given for over 31.
Not on a function statement.

5-2

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-2
<-AMERICAN->
NATIONAL
STANDARDS
FORTRAN

X3.9 -77 (Full
1966 Language)

Subprogram Statements

<-PDP-ll-> VN</
TOPS-l 0
RSX
RSX
VMS
RSTS/E RSTS/E
TOPS-20
lAS
as/8 OS/78 RT-ll
FORT FORT lAS
VN</ll
(F77)
FORTRAN F-IO
(FOR)
IV
IV

<--PDP-8-->
as/8
FORT

BLOCK DATA

ENTRY entry-point
[(arg[ ,arg] ••• )]

FUNCTION function-name
(arg [ ,arg] ••• )

function-name (arg[,arg] ••• )= X
arithmetic-expression

function-name (arg [ ,arg] ••• ) =
logical expression

function-name (arg[,arg] ••• )=
character expression

x 3 ,7

x 2 ,6

x 5 ,2,7

[nITffiffi
~~LE

]

PRECISION

FUNCTION x3

COMPLEX
LOGICAL
func-name
[*length] (arg [ ,arg] ••• )

SUBROUTINE subroutine-name
[(arg[ ,arg] ••• )]
1
2
3
4
5
6
7

Can have name.
Permit () and no args.
·*length not permitted.
INTEGER, REAL, LOGICAL only.
*length ignored in most cases; flagged with warning.
Also types DOUBLE COMPLEX, CHARACTER
Also type CHARACTER [*length].

5-3

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-3

Constant and Variable Types

05/8

OS/78

X3.9 -77 (Full
1966 Language)

FORT
II

FORT

<-PDP-ll-> VAX/
TOPS-IO
VMS
RSX
RSX
RSTS/E RSTS/E
TOPS-2O
RT-ll
lAS
lAS
VAX/ll
(FOR)
(F77)
FORTRAN F-10

Integer

Real

Double-precision

)(

Complex

Double complex

Logical

Literal (preceded by H)

Literal (enclosed in
single quotes)

Hexadecimal

Octal

Quadruple 'precision

Radix 50

<-AMERlCAN->
NATIONAL
STANDARDS
FORTRAN

<--PDP-8-->

CONSTANT FORMS

I
2
3
5
6
7

In DATA statements only.
["ddd] allowed anywhere [Oddd] is allowed in DATA statements.
In DATA, FORMAT, and subroutine arguments only.
Denotes Hollerith literal.
"ddd allowed anywhere.
Denotes CHARACTER constant.

5-4

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-4
<-AMERICAN->
NATIONAL

Data Types

<--PDP-8-->

STANDARDS

FORTRAN
X3.9 -77 (Full
1966 Language)

OS/8
FORT
II

05/8 05/78
FORT FORT
IV
IV

<-PDP-ll-> VAX/
TOPS-IO
RSX
RSX
VMS
RSTS/E RSTS/E
TOPS-20
RT-ll
lAS
!AS
VAX/ll
(FOR)
(F77)
FORTRAN F-I0

Floating point
REAL
REAL*4
REAL*8
REAL*16
DOUBLE PRECISION

-X

X
X
X

X
X

X
X
X
X
X

X
X
Xs
X
X

X
Xs
X
-

Complex
COMPLEX
COMPLEX*8
COMPLEX*16
DOUBLE COMPLEX

X
X

X
X
X
X

X
-

X
X

x2
x4

xl
x4
X
X

Xs
Xs
X
X

Character
CHARACTER*(length)
CHARACTER* (*)
Logical
LOGICAL
LOGlCAL*1
LOOICAL*2
LOGlCAL*4

Integer
INTEGER
INTEGER*1
INTEGER*2
INTEGER*4
I
2
3
4
5
7

Defaults to *2 or *4 at compile time.
Implemented as *4.
Four-byte allocation; two-byte precision.
BYTE is synonym for LOGlCAL*I.
Produces warning; * length ignored.
Only with FPP hardware.

5-5

X3
X

X
X

~s
X

A COMPARISON OF FOHTRAN LANGUAGE FEATURES

Table 5-5

Specification Statements

<-AMERICAN->
NATIONAL
STANDARDS
FORTRAN
X3.9 -77 (Full
1966 Language)

<-PDP-ll-> VlVl../
TOP5-1 0
RSX
RSX
VMS
RSTS/E RSTS/E
TOP5-2 0
lAS
as/8 OS/78 RT-ll
FORT FORT' lAS
VlVl../ll
(FOR)
IV
IV
(F77)
FORTRAN F-10

PRCGRAM name

NAMELIST /namelist-name/ name
[,name ••• ] [[,] /namelistname/ name [,name ••• ]] •••
1
PARAMETER var=constant, •••
2
PARAMETER (var=exp, ••• )

SAVE statement

IMPLICIT type (characters
[,type(characters)] •••

x 3 ,S,6

RFAL
1[INTEGER
~ DOUBLE PRECISION
COMPLEX
LOGICAL

<--PDP-8-->
as/8
FORT

{"arne
array:-declarator }
functlon-name

[,name

} ...J

,array~declarator

, functlon-name

INTEGER
RFAL [*8][*2]
COMPLEX [*16]
LOGICAL [*1]

fame
array-declo }
function-name
[*length] [/initial-value/]
1
2
3
4
5
6

Type of symbolic name determined by constant.

Type of symbolic name determined by first letter, IMPLICIT, and type declarations.
Warnings produced for INTEGER*2, COMPLEX*16, LOGICAL*l.
Also allows types DOUBLE COMPLEX, CHARACTER.
Also allows type CHARACTER.
Initial-value not allowed.

5-6

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-5

Specification Statements (Cont)

<-AMERlCAN->
NATIONAL

<--PDP-8-->

<-PDP-ll-> VAX/
RSX

RSX

TOP5-l0

VMS

X3.9
1966

-77 (Full

OS/8
FORT

Language)

RSTS/E RSTS/E
TOP5-20
OS/8 OS/78 RT-ll
!AS
FORT FORT lAS
VAX/ll
(F77)
IV
IV
(FOR)
FORTRAN F-IO

DIMENSION array-decl.
[,array-decl.] •••

EQUIVALENCE (name[,name] ••• )
[, (name [ ,name] ••• )] •••

~ X

STANDARDS

FORTRAN

COMMON~~:y_decl.~
Knarne
,array-decl.]

• ••

COMMON/block-name!~ame
array-d ec
Kname
,array-decl.~.

••• [/block-name/ ••• ] ••.

EXTERNAL~SUbPrOgram-name

external-proc-name

~SubprOgram-narne

]
,external-proc-name

•••

EXTERNAL
1[[*] subprogram-name
~
[*]external-proc-name •••
Array declarators allow
upper and lower dimension
bounds
INTRINSIC

1[subProgram-name~
1 & can be used in place of *.

5-7

A COMPARISON OF FORTRAN L.ANGUAGE FEATURES

Table 5-6

Data Initialization statement

<-AMERICAN->
NATIONAL

<-PDP-ll-> vAX!
TOPS-I 0
RSX
RSX
VMS
TOPS-20
RSTS/E RSTS/E
lAS
RT-ll
VAX/ll
lAS
(FOR)
(F77)
FORTRAN F-IO

<--PDP-8-->

STANDARDS

FORTRAN
X3.9 -77 (Full
1966 Language)

05/8
FORT
II

05/8 OSi78
FORT FORT
IV
IV

DATA name [,name] •••
/[number*] value
[,[number*]va1ue] •••
[ , / [ , name ••• ] ••• ] •••

Implied DO construct

Data values converted to
type of name

Data initialization allowed
in type declaration
statements

Name may be an array name implying all elements of array.

Table 5-7
<-AMERICAN->
NATIONAL

Relational Operators
<--PDP-8-->

STANDARDS

FORTRAN
X3.9 -77 (Full
1966 Language)

05/8
FORT
II

05/8 OSi78
FORT FORT
IV
IV

TOPS-I 0
<-PDP-ll-> VAX/
RSX
VMS
RSX
TOPS-20
RSTS/E RSTS/E
RT-ll
lAS
VAX/ll
lAS
(FOR)
FORTRAN F-IO
(F77)

.GT.

.GE.

.LT.

.LE.

.EQ.

.NE.

5-8

- - - - _..- - - - - _ . _ - -

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-8

Logical Operators

05/8 05/78
FORT FORT
IV
IV

TOPS-1O
<-PDP-ll-> VAX/
RSX
RSX
VMS
TOPS-20
RSTS/E RSTS/E
lAS
RT-ll
VAX/ll
!AS
(FOR)
(F77)
FORTRAN F-1O

.OR.

.XOR.

.E(]J.

.NEQV.

Table 5-9

Assignment Statements

<-AMERlCAN-->
NATlOOAL

<--PDP-8-->

STANDARDS

FORTRAN
X3.9 -77 (Full
1966 Language)

05/8
FORT

.Nor.

.AND.

05/8 05/78
FORT FORT
IV
IV

TOPS-1O
<-PDP-ll-> vAX!
RSX
RSX
VMS
TOPS-20
RSTS/E RSTS/E
lAS
RT-ll
VAX/ll
!AS
FORTRAN F-IO
(FOR)
(F77)

variable = logical expression X

ASSIGN statement-number TO
variable

variable = character
expression

<-MERleAN->
NATIONAL

<--PDP-8-->

STANDARDS

variable = arithmetic
expression

X3.9
1966

FORTRAN
-77 (Full
Language)

05/8
FORT

I Statement label may be label of a format statement.
2 Character expression must be variable, substring, or constant.

5-9

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-10
<-AMERICAN->
NATIONAL

Control Statements
<--PDP-8-->.

STANDARDS

<-PDP-ll-> VAX/
RSX
RSX
VMS
RSTS/E RSTS/E
RT-ll
lAS
lAS
VAX/ll
(FOR)
(F77)
FORTRAN

TOPS-I 0

F-IO

TOPS-20

FORTRAN
X3.9 -77 (Full
1966 Language)

OS/8
FORT
II

OS/8 OS/78
FORT FORT
IV
IV

CALL subroutine-name
[(argument [,arg[ ••• )]

CONTINUE

x 2 ,4,8

x 2 ,8,
9,10

x 2 ,8,9

X
x2,8,9

DO statement-number controlvariable = inJ tial-value,
test-value [ , increment]

END

GO TO statement-number

GO TO (statement-number,
statement-number
[,statement-number] ••• )
[ ,] variable

X
x 5 ,6

x 5 ,6

GO TO variable [[,]
(statement number,
statement-number
[,statement-number] ••• )]

IF (arithmetic-expression)
X
statement-number, statement
number, statement-number

IF (logical-expression)
executable-statement

1
2
3
4
5

Null argument permitted.
General expressions permitted.
Iteration count computer but minimum of one iteration.
Variable and expressions are of type INTEGER.
Index may be a general expression which will be converted automatically to
integer if expression of other type.
6 If index out of bounds, then acts as continue.
7 If list present and assigned label not in list, then acts as continue.
8 Optional comma after statement number.
9 Compiler switch to determine minimum iteration count (0 or 1).
10 Statement number optional.

5-10

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-10

Control Statements (Cont)

as/8 OS/78
FORT FORT
IV
IV

TOPS-10
<-PDP-ll-> VM../
RSX
VMS
RSX
RSTS/E RSTS/E
TOPS-20
RT-ll
!AS
lAS
VAX/II
(FOR)
FORTRAN F-IO
(F77)

DO-WHILE

END 00

PAUSE (one to six octal
digits)

PAUSE (one to five decimal
digits)

PAUSE 'message'

S'IDP (one to six octal digits) X

STOP (one to five decimal
digits)

STOP 'message'

Tracing after pause

RETURN

RETURN (expression)

<-AMERlCAN->
NATIONAL
STANDARDS
FORTRAN
X3.9 -77 (Full
1966 Language)
IF (logical-expression)
statement-number, statement
number

IF-THEN-ELSE-ENDIF

<--PDP-8-->
as/8
FORT

2 Up to 12 digits.
4 PAUSE enters command processor. User can STOP, CONTINUE, ASSIGN, SHOW, DEBUG, etc.

5-11

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-11
<-MERleAN->
NATIONAL

Sequential I/O Statements
<--PDP-8-->

STJ\NDARDS

TOPS-10
<-PDP-ll-> VAX/
RSX
RSX
VMS
TOPS-20
RSTS/E RSTS/E
RT-ll
lAS
VAX/l1
!AS
(F77)
FORTRAN F-10
(FOR)

FOR'l"RAN
X3.9 -77 (Full
1966 Language)

as/8
FORT
II

as/8 as/78
FORT FORT
IV
IV

ACCEPT N

BACKSPACE u

ENDFILE u

PRINT f,k

PRINT f

PRINT*,k

PRIm' N

PUNCH f,k

PUNCH f

PUNCH* ,K

N = namelist-name
u = uni t m.mber
f = label of a format
statement
k = an I/O list
* indicates listdirected I/O
ACCEPT f,k
ACCEPT f
ACCEPT*,k

5-12

~~~----~~--~~-.---~-----

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-11

Sequential I/O Statements (Cont)

<-MERleAN->
NATIONAL

<--PDP-8-->

STANDARDS

FORTRAN
X3.9 -77 (Full
1966 Language)

as/8 05/78
FORT FORT
IV
IV

05/8
FORT

<-PDP-ll-> VAX/
RSX
RSX
VMS
RSTS/E RSTS/E
lAS
RT-ll

TOPS-2O

VAXill

!AS

(FOR)

TOPS-I 0

(F77)

FORTRAN

F-IO

N = namelist-name
u = uni t nunber
f = label of a format
statement
k = an I/O list
* indicates listdirected I/O
RFAD f,k

READ (u,f) [k]

RFAD (u) [k]

RFAD (u [, f] [, END=SI]
[,ERR=S2] ) [k]

x l ,4

x l ,2

X1 ,2

RFAD* ,k

x l ,3

READ (u,*[,END=S]] [,ERR=S2])k

RFADN

x l ,3

x 3 ,s

BACKFILE u

SKIPFILE u

TYPE f[ ,k]

TYPE *[ ,k]

TYPE N

RERFAD f,k

REWIND u

RFAD (u,N[,END=SI] [,ERR=S2])

I
2
3
4
5

Keywords in either order.
Also UNIT = u, FMT = f.
Also UNIT = u.
f is mandatory (formatted only) •
Also NML=N.

5-13

A COMPARISON OF r'ORTRAN LANGUAGE FEATURES

Table 5-11

Sequential I/O Statements (Cont)

<-AMERlCAN->
NATIONAL
STANDARDS
FORTRAN
X3.9 -77 (Full
1966 Language)

OS/8 OS/78
FORT FORT
IV
IV

<-PDP-ll-.>
RSX
RSX
RSTS/E RSTS/E
RT-ll
lAS
lAS
(FOR)
(F77)

<--PDP-8-->
OS/8
FORT

VAX/
VMS

TOPS-l 0
TOPS-20

VAX/II
FORTRAN

F-I0

N = namelist-name
u = uni t nunber
f = label of a format
statement
k = an I/O list
* indicates listdirected I/O
ary = array, array element,
or variable
WRITE (u,f) [k]

3
X

WRITE (u) [k]

x l ,3

WRITE (u[,f] [,END=SI]
[ , ERR=S2] ) k

xl,S

x l ,3

WRITE (u,*[,END=SI]
[ , ERR=S2] ) k

xl ,4

x l ,4

WRITE (u,N[ ,END=SI] [,ERR=S2])

WRITE *,k

WRITE f,k

WRITE f

SKIPRECORD u

UNLOAD u

ENCODE/DECODE (cnt,fmt,ary)k

1
3
4
5
6

Keywords in either order.
Also UNIT = u, FMT = f.
Also UNIT = u.
f is mandatory (formatted only).
Also NML=N.

5-14

A COMPARISON OF FORTRAN LANGUAGE FEATURES
Table 5-12

File Control Statements

OS/8 OS/78
FORT FORT
IV
IV

<-PDP-ll-> VAX/
TOPS-I 0
RSX
RSX
VMS
TOPS-20
RSTS/E RSTS/E
RT-ll
lAS
VAX/ll
!AS
(FOR)
(F77)
FORTRAN F-I0

<-AMERlCAN->
NATlOOAL

<--PDP-8-->

STANDARDS

FORTRAN
X3.9 -77 (Full
1966 Language)

OS/8
FORT

OPEN

CLOSE
INQUIRE
1

Library subroutines.

Table 5-13

Direct Access I/O Statements

<-AMERleAN->
NATIONAL

<--PDP-8-->

OS/8 OS/78
FORT FORT
IV
IV

<-PDP-ll-> VAX!
TOPS-I 0
RSX
RSX
VMS
TOPS-20
RSTS/E RSTS/E
RT-ll
lAS
!AS
VAX/ll
(FOR)
(F77)
FORTRAN F-I0

STANDARDS

FORTRAN
X3.9 -77 (Full
1966 Language)

OS/8
FORT

u = unit number
n = number of records in file
rs = length in 16-bit words
of record
c = indicates unformatted
v = associated variable
DEFINE FILE u(n,rs,c,v)
[ ,u • •• ) •••
2
FIND (u'r)

x2,3

RFAD (u'r[,f)

xl!

READ (u'r[f) [ ,END=Sl)
[,ERR=S2) [K)

x 2 ,S

xl!

WRITE (u' r[ ,f» [k)

xl!

WRITE (u'r[,f) [,END=SI)
[ ,ERR=S2» [k)

x 2 ,S

xl!

DELETE (u' r[ ,ERR=s2)

xlO

[k)

x2,3
x

FORMAT statement:
Statement-number FORMAT
( format-spec)

1 Call DEFINE FILE with different format of arguments.
2 Unformatted only.
3 Record size measured in 16-bit words (= 1/2 storage unit) •
4 Record size measured in storage units.
S END= option not allowed.
6 Warning for i.
8 END=SI syntax used if record number is outside of file.
9 REC=r is syntax for direct access.
10 Also UNIT = u, REC = r.
l! Also UNIT = u, REC = r, FMT = f.

5-15

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-14
<-AMERleAN->
NATIONAL

Indexed I/O Statements
<--PDP-8-->

05/8 05i78
FORT FORT
IV
IV

<-PDP-11-> VAX!
TOPS-10
RSX
RSX
VMS
TOPS-20
RSTS/E RSTS/E
lAS
RT-ll
!AS
VAX/ll
(F77)
(FOR)
FORTRAN F-10

x2
x2

x3
x3

STANDARDS

FORTRAN
X3.9 -77 (Full
1966 Language)

05/8
FORT

READ (u[,f] ,KEY=k1[,KEYID=i])
[k]

READ (u[,f],KEY=k2[,KEYID=i]
[ ,END=Sl] [,ERR=S2] ) [k]
WRITE (u[ ,f]) 1 [k]

u = unit nunber
f = format specifier
kl = key specifier
i = key identifier
sl = statement label
s2 = statement label
k = an I/O list

WRITE (u[,f] [,END=sl]
[,ERR=s2] ) [k]

REWRITE (u[ ,f]) [k]

REWRITE (u[ ,f] [,END=sl]
[ ,ERR=2]) [k]

DELETE (u[ ,ERR=s2])

UNLOCK (u[ ,ERR=s2] )

1 Syntax identical to sequential write.
2 Also UNIT = u, FMT = f.
3 Also UNIT = u.

5-16

3
3

A COMPARISON OF FORTRAN L~NGUAGE FEATURES

Table 5-15

Format and Types of Conversion

<--PDP-8-->

<-PDP-ll->

VAX/

NATIONAL
STANDARDS
FORTRAN
X3.9 -77 (Full
1966 Language)

RSX
RSTS/E

VMS

as/8

as/78

FORT
II

FORT
IV

RSX
RSTS/E
RT-11
lAS
(FOR)

(F77)

FORTRAN

F-10

A (alFhammeric)

D (real D decimal exponent)

E (real E decimal exponent)

F (real, no exponent)

X·

G ( real)

H (literal)

, ... , (literal)

( integer)

L (logical)

0 (octal)

P (scale factor)

<-AMERleAN->

Q (record length)

TOPS-IO
TOPS-2O

lAS
VAX/ll

T (position indicator in
record)

X (skipped data or blank)

Z (hexadecimal data)

Format specification in
arrays

5-1 7

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-15

Format and Types of COnversion (COnt)

<-AMERlCAN-)
NATIONAL
STANDARDS
FORTRAN

<--PDP-8--)

<-PDP-ll-) vAX!
RSX
RSX
VMS

TOPS-I 0

RSTS/E

TOP5-2 0

X3.9 -77 (Full
1966 Language)

05/8
FORT
II

OS/8 OS(78 RT-ll
FORT FORT lAS
IV
IV
. (FOR)

Carriage Control

R (right-justified A)

S (control of optional + sign)

/ (record separator)

RSTS/E

lAS
(F77)

VAX/II
FORTRAN

F-IO

(format scan terminator)

(format separator)

BN (blank = null)

BZ (blank = zero)

SS (suppress optional + sign)

SP (print optional + sign)

TL (tab left)

TR (tab right)

3 Used as carriage control character also.

5-18

A COMPl\RISON OF FORTRAN LANGUAGE FEATURES

Table 5-16
<-AMERICAN->
NATIONAL
STANDARDS
FORTRAN
X3.9 -77 (Full
1966 Language)

Real

Integer

Library Functions
<--PDP-8-->

05/8 OS/78
FORT FORT
IV
IV

<-PDP-ll-> VNI./
TOPS-10
RSX
RSX
VMS
RSTS/E RSTS/E
TOPS-20
RT-ll
lAS
!AS
VNI./ll
(FOR)
(F77)
FORTRAN F-10

Double-precision

Quad precision

Complex to real

Double complex to
double-precision

Integer to real

Integer to double

-1

Integer to quad

Real to integer

Double to real (obtain
most significant part)

05/8
FORT

Absolute value:

Conversion:

Double to integer

Quad to real

Real to double

Real to quad

Byte or integer*2 to
integer*4 (zero-extend)

G-floating to D-floating

D-floating to G-floating

1 Available as implied conversion only.
2 Argument is a two-element double-precision array.

5-19

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-16
<-AMERlCAN->
NATIONAL
STANDARDS
FORTRAN
X3.9 -77 (Full
1966 Language)

Library Functions (Cont)
<--PDP-8-->

as/8 OS/78
FORT FORT
IV
IV

as/8
FORT

<-PDP-ll-> VAX/
TOPS-l 0
RSX
RSX
VMS
TOPS-20
RSTS/E RSTS/E
RT-ll
lAS
lAS
VAX/ll
(F77)
(FOR)
FORTRAN F-IO

Complex to real
(obtain real part)

Double complex to real
(real part)

Complex to real
(obtain imaginary part)

Double complex to double
real (imaginary part)

Real to complex

Double real to
double complex

Real (radians)

Real (degrees)

Double (radians)

Quad

Complex

Double complex

.x 4

Sine

Cosine

Tangent

Cosine:

Hyperbolic (real and double) :

1
2
3
4

Real only.
Also quad.
Radian and degree versions.
Arguments are two-element double-precision arrays.

5-20
- _ ..

-.--~---

.-------~~~~-

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-16

Library Functions (Cont)

<-AMERICAN->
NATIONAL
STANDARDS
FORTRAN
X3.9 -77 (Full
1966 Language)

Arc-sine (real and double)

Arc-cosine (real and double)

x2 ,3

Real

. x3

Ibuble

Quad

Quotient of two arguments

Real (radians)

Real (degrees)

Double (radians)

Quad

Complex

Ibuble complex

<--PDP-8-->
05/8
FORT

TOP5-l0
<-PDP-ll-> VAX!
RSX
VMS
RSX
"RSTS/E RSTS/E
TOP5-20
05/8 OS/78 RT-ll
lAS
FORT FORT
IAS
VAX/ll
IV
(FOR)
FORTRAN F-1O
IV
(F77)
x 2 ,3
X

Arc-tangent:

Sine:

1
2
3
4

Real only.
Also quad.
Radian and degree versions.
Arguments are two-element double-precision array.

5-21

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-16
<-AMERlCAN->
NATIONAL
STANDARDS
FORTRAN
X3.9 -77 (Full
1966 Language)

Library Functions (Cont)
<--PDP-8-->
05/8
FORT
II

05/8 OS/78
FORT FORT
IV
IV

<-PDP-ll-> Vl\I../
TOPS-10
RSX
RSX
VMS
RSTS/E RSTS/E
TOPS-20
RT-II
lAS
lAS
Vl\I../11
(FOR)
(F77)
FORTRAN F-IO

Tangent

Cotangent

lAND

IOR

lEOR

2
X

Nar

Logical functions:

Error function
Garrrna function

Switch register

Complex conjugate

positive difference
(ul-Min(u1 ,u 2»

Log garrana

1 Also double complex.
2 Available in all expressions as .AND., .OR., .XOR., and .NOT. operators for integer values.

5-22

A COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-16
<-AMERICAN->
NATIONAL
STANDARDS
FORTRAN
X3.9 -77 (Full
1966 Language)

Library Functions (Cont)
<--PDP-8-->
OS/8
FORT
II

<-PDP-ll-> VAX/
TOPS-l 0
RSX
RSX
VMS
RSTS/E RSTS/E
TOPS-20
OS/8 OS/78 RT-ll
lAS
FORT FORT lAS
VAX/ll
IV
IV
(FOR)
(F77)
FORTRAN F-IO

Bit set

Bit test

Bit clear

Shift

Circular shift

Insert bits

Move bits

Real

Double

Quad

Complex

Double complex

Real

Double

Quad

Complex

Double complex

Exponential:

Logarithm:

1 Arguments are two-element double-precision arrays.

5-23

A COMPARISON OF FOR'fRAN LANGUAGE FEATURES

Table 5-16
<-AMERlCAN->
NATIONAL
STANDARDS
FORTRAN
X3.9 -77 (Full
1966 Language)

Library Functions (Cont)
<--PDP-8-->

05/8 OS/78
FORT FORT
IV
IV

<-PDP-ll-> VAX/
TOPS-1 0
RSX
RSX
VMS
RSTS/E RSTS/E
TOPS-20
RT-ll
!AS
lAS
VAX/ll
(FOR)
(F77)
FORTRAN F-IO

05/8
FORT

Square root:
Real

Double

Quad

Complex

Double complex

Real to real

Real to integer

Double to integer

Quad to integer

Real

Integer

Double-precision

Truncation:

1 Arguments are two-element double-precision arrays.

5-24

COMPARISON OF FORTRAN LANGUAGE FEATURES

Table 5-16

Library Functions (Cont)

<-AMERICAN->
NATIONAL
STANDARDS
FORTRAN
X3.9 -77 (Full
1966 Language)

TOPS-l 0
<-PDP-ll-> VAX/
RSX
VMS
RSX
RSTS/E RSTS/E
TOPS-20
lAS
as/8 OS/78 RT-ll
FORT FORT lAS
VAX/II
FORTRAN F-I0
(FOR)
(F77)
IV
IV

Maximum value
(Number of arguments ~2
for all functions)

Minimum value
(Number of arguments ~2
for all functions)

Real

Integer

Double-precision

Quad precision

Test sense switch

Random number

Convert sign magnitude to
2s complement and vice versa

<--PDP-8-->
as/8
FORT

Transfer of sign:

Remainder of time limit
DIVERT run time error messages
to a file

ERRSET controls handling of
error conditions

DATE, TIME

TRACE

5-25

Reader's Comments

ANSI-77 FORTRAN Information Document
EY-1233E-ID-OOOl

Your comments and suggestions will help us in our continuous effort to improve the quality and usefulness of our
pUblications.

What is your general reaction to this manual? In your jUdgment is it complete, accurate, well organized, well
written, ctcJIs it easy to use? ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

What faults or errors have you found in the manual? _ _~~_~~~_ _ _~~_~_~_~~_

Does this manual satisfy the need you think it was intended to satisfy? _~~_~~_~~_~~~_
Does it satisfy your needs? _~~~~~~~~~_~~_

Why? ___________________________

Please send me the current copy of the Technical Documentation Catalog, which contains information on
the remainder of DIGITAL's technical documentation.

Name

______

_ _ _ _ _ _ __

Street _~~~~~~~~~~~~~~~~_
_ _ _ _ _ _ _ _ _ _ _ _ _ _ __
City
_~

Title
Company _________________ State/Country __________________
Department _ _ _ _ _ _ _ _ _~~~___ Zip

Additional copies of this document are available from:
Digital Equipment Corporation
ESD&P Order Processing
12A Esquire Road
North Billerica, MA 01862

Order No. EY-1233E-ID

MRO

Do Not Tear - Fold Here and Staple

III II I

No Postage
Necessary
if Mailed in the
United States

BUSINESS REPLY MAIL
FIRST CLASS

PERMIT NO.33

MAYNARD, MA.

POSTAGE WILL BE PAID BY ADDRESSEE
I

Digital Equipment Corporation
Educational Services/Quality Assurance
12 Crosby Drive (BUO/E08)
Bedford, MA 01730

Digital Equipment Corporation • Bedford, Ma. 01730