Sigma XDS 5, XDS 7 Reference Manual

X1ClS

SIGMA

5/7

FLAG

Scientific

A XEROX

Data

COMPANY

Systems

Reference Manual

FLAG

Pri

ce:

$4.

00

REFERENCE

MANUAL

for

XDS

SIGMA

PRELIMINARY

September 1 969

5/7

90

COMPUTERS

EDITION

16

54A

Xerox

© 1969, Xerox Data Systems, Inc.

Data

Systems/701 South

Aviation

Boulevard/EI Segundo,

California

Printed

90245

in U.S.A.

ACKNOWLEDGMENT

The FLAG (Fortran Load And

Phillip

Horton,
of

the

Vanderbi

In

cooperation

piler

available

Sherrod,

It

University

with Vanderbi It

to

those customers who

Go)

and

Douglass Henry

Computer

University,

FORTRAN.

reference

The

Reference

IV-H
cerpted
all

the
extended
programs
Intrinsic
Reference

material

from

the

material

input/output

that

are

and

Basic External

Manua

in this manual was

Manual (XDS

XDS Sigma

in

Chapter 6 dealing

5/7

capabilities.

available

to FORTRAN
Functions",

I.

90

RELATED

Title

XDS Sigma 5
XDS Sigma 7 Computer Reference
XDS Sigma
XDS Sigma
XDS Sigma
XDS Sigma
XDS Sigma
XDS Sigma

Computer

5/7

Batch Processing

5/7

Batch Processing Monitor

5/7

FORTRAN IV-H

5/7

FORTRAN IV Reference

5/7

FORTRAN IV

5/7

FORTRAN IV Library

Reference

Operations

Compiler

0966)

FORTRAN IV

was

developed

under

Center,

Nashvi

Xerox

require

the

largely

though several

Reference

with FORMAT

by

the

direction

lie,

Tennessee.

Data

Systems is making

characteristics

taken

from

important

Manual (XDS

statements

Furthermore, FLAG provides most

IV

users. These subprograms

taken

from

the

previously

PUBLICATIONS

Manual
Manual

Monitor

Reference

Reference
Operations

Manual

Manual

Manual

Technical

Manual

Manual

Manual

Norman

of

of

the

XDS Sigma

sections

and

mentioned

Wheeler,

Dr.

George

the

FLAG

a fast

load-and-go

5/7

have

900956),

specifications

of

the

function

are

listed in

FORTRAN IV

Publication

900959
90
900954
901198
900966
900956
90
90

Robert

Haynam

Com-

FORTRAN

been

ex-

notably

and

with

"Table

No.

0950

11

43

15 24

sub-

9

The

specifications

depend

on a

specific

of

the

software system

configuration

of

equipment

described

such as

in this

ii

publication

additional

tape

are

units or

NOTICE

subject

to

change

larger

without

memory. Customers should

notice.

The

availability

consult

or performance

their

XDS sales

of

some

representative

features

for

details.

may

CONTENTS

INTRODUCTION

l.

How This

The FLAG
FLAG Programs 1
Conditional

DATA

2.

Limits

Constants

Identifiers
Variables

Functions

EXPRESSIONS

3.

Arithmetic

Mixed
Relational
Logical

ASSIGNMENT

4.

CONTROL

5.

Labels

GOT 0 Statem

ASSIGN
IF

CALL
RETURN

DO

CONTINUE
PAUSE
STOP
END

INPUT/OUTPUT

6.

Input/Output

Manual

on

Values

Integer
Real

Constants
Double
Complex
Double
Logical
Literal

Constants

Sca

lars 8

Arrays

Evaluation

Expressions

Expressions
Logical

Evaluation

Unconditional

Assigned

Computed

Statement

Statements

Arithmetic
Logical

Statement

Statement

Statement

Statement

Statement

Statement

List Items 27
Special

Is

Organized

Compiler

Compilation

of

Quantities

Constants

Prec ision

Constants

Complex

Constants

Expressions

Expressions

Operators

STATEMENTS

GO

IF

Statement

List

Constants

Constants

Hierarchy

Hierarchy

STATEMENT

ents

GO

TO

Statement

GO

TO

IF

Statement

Statement

Lists

Considerations

- X

Cards

TO

Statement

Statement

10
10

10
12
13
14
14
15

16

18
18

18
18
18
19

19
20
20
20
21
22
22
25
25
25
26

27
27

28

Input/Output

Formatted

1

3

5
5

5
5
6
6
6
7
7
7
8
8

8
9

7.

Acceptable
Intermediate
END

NAME

Simplified

OUTPUT

INPUT

FORMA T

F

Format

- E Format
D

Format
G Format
I

Format

L Format

A

Format

R Format
Z Format
M Format

H

Format
'Format

X

Specification

T

Specification

P

Specification

/

Specification
Parenthesized
Adjustable

Numeric

Widthless

Numeric
Comma
FORMA T
FORMATs
Extended
Memory-to-Memory

ENCODE

DECODE
Direct

BUFFER

BUFFER
Random
Auxi I iary

REWIND

BACKSPACE

END

Carriage

DECLARATION STATEMENTS
Classification

Implicit
Explicit
Confl

Array

Declarations
Array
References

DIMENSION

IMPLICIT

Statements

Input/Output

FORTRAN

Input/Output

and

ERR

Forms

LIS T Statement

Input/Output

Statement

Statement

Statement

Input

Field-Termination

and

Stored

Input/Output

Input/Output

Access

Input/Output

FILE

Control

icting

Storage

Statement

Using NAME

(Fixed

Decimal

(Normalized,

(Normalized,

(General)

(Integer)

(Logical)

(Alphanumeric)

(Alphanumeric,

(Hexadecimal)

(Machine

(Hollerith)

(Hollerith)

(Skip;

(Tab) 49

(Scale

(Record

FORMAT

FORMAT

Numeric

Input

Statement

Statement

IN
OUT

Statement

Statement

Declarations
Declarations

Specifications

Strings

with

List

Interfac

in

Arrays

Data

Input/Output

Statement

for

Printed

of

Identifiers

and

Redundant

to

Array

Statement

Statements

II

Statements

Statements

of

the

READ

Statements

LIST

Point) 38

with

E Exponent) 39

with D Exponent)---

--

Right-Justified)

Dependent)

Space

or

Backspace)

Factoror

Separator)

Specifications

Input

Width

ing 57

Capabilities

Conversion

Statements

Elements

Power

Specified

Statements

Output

Declarations

___

of

10) 49

__

28
29
30
31
32
32
32
33
34
37

40
40
42
42
43
44
45
46
47
47
48

51
52
53
54
55
56
56

58
59
59
60
60
61
62
62
63
63

64

64
64
64

65
65

65
65
65
66
66
66
67
67

iii

Explicit
Optional
Storage
COMMON

EQUIVALENCE
Interactions
EXTERNAL
BLOCK DATA Subprograms
DA

Placement

PROGRAMS

8.

Main
Subprograms

Arguments

Library Subprograms

OPERATIONS

9.

Running a FLAG

Job
F

INDEX

Type

Statements

Size

Specifications

Allocation

Statement
Labeled
Blank

Arrangement

Referencing

T A

DATA
DA

Statement
FUNCTION
SUBROUTINE Subprograms

Dummy
Dummy

Adjustable

Dummy Subprograms

Basic
Additional

The FLAG

Setup

LA G Debug

COMMON

COMMON

of

Statement

Statement

Variable

T A

Constant

and

AND

Programs

and
Scalars
Arrays

External

Examples

of

Storage

Order

Functi ons

Dummies

Dimensions

Library Subprograms

Control

Mode

Statements

of

COMMON

Data

in

COMMON

Statement

Allocation

List

Li

st

of

Declaration

SUBPROGRAMS

Subprograms

Functions

Job

Card

Statements

Statements

68

APPENDIXES

69
70
70
71

A.

B.

STATEMENTS

FLAG
DIAGNOSTIC

MESSAGES

98

100

71
72
73
73

_

75
76
76
77
77
78
79

80

1.

2.

3.

4.

Sample
Array

FLAG
FLAG

Processing

XDS FLAG Program 2

Storage

Job
Job

ILLUSTRATIONS

Setup -Single
Setup -Multiple

Mode

Program

66
96

Programs in Batch

97

80
80
80
81
82
83

84
84

85
86
86
86
86

93
93

93
95

96

104

Sample

l.

Mode

2.

Valid

3.

4.

Evaluation

r:::.

f\,~ixed

-'.

Standard

6.

FORTRAN II/FORTRAN IV

7.

Permissible

8.

9.

Intrinsic

10.

FLAG

Program

of

Mixed

Operators

Type

Combinations

of

VOiiob!e

Unit

Correspondences

and

Dummies

and

Option

Basic External

TABLES

Expressions Using

+ - * /

for

Logical

Assignments

Codes

Expressions

Types

and Expression

Equivalent

Exponentiations

"-Aodes

Statements_

Between

Functions

Arguments

__

3

12
12
14

16

28
30

83
87
94

iv

How

This

Manual

Most

of

the

material

several

though
convenience
the

FORTRAN IV-H Manual

The

remainder

of

general

contains

interest

the

important

of

those

of

this

essential

is

Organized

in this manual was

sections

who may

Chapter

to

the

operations

already

is

indicated

summarizes

new user.

information for

are

Chapters

taken

from

be

1.

the

familiar

by a

the

most

INTRODUCTION

from

the

Sigma

5/7

FORTRAN IV-H

Sigma

5/7

FORTRAN IV

with XDS FORTRAN

bracket

2 through 8

in

important

compiling

the

left

hand margin

features

are a detailed

and

running FLAG programs.

Reference

IV-H,

of

FLAG

description

Reference

Manual (XDS 90

material

of

and

the

page.

then

of

in this manual

briefly

the

Manual

presents

FLAG

(XDS 90

09

56C).

that

information

language.

09

For

differs

Chapter

66B),

the
from

9

l

already

Users
areas

marked by

able

to

The

FLAG

The FLAG (FORTRAN Load And Go) system for XDS Sigma
designed
operating

•

Comprehensive

• Fast compi

•

Significant

•

Special

FLAG may
oping his program. Further,
ment

where

FLAG

Programs

FLAG programs
cution
nal

to

program

familiar

brackets,

start

running FLAG programs

Compiler

to

be

compatible

characteristics:

diagnostic

lation.

reduction

accounting

be

used in

the

job stream

are

comprised

of

the

program and

the

program and

can

be

written

with

XDS

and

then

with

messages

in

total

and

processing

preference

it

contains

of

the

data

read

into

out

while

FORTRAN

other

processing

to

should

an

to

the

processing

IV-H

will

probably

read

Chapter

with

compilers

at

features

standard

be

numerous small programs, many

ordered

be

processed by

computer

9 for

minimum

compile

the

delay.

of

this

and

time

for

small-to-medium

to

minimize

FORTRAN

primary FORTRAN

set

of

statements

during

program

continues.

an

class.

execute

compilers

the

perfer

to

explanation

5/7

computers is

However, FLAG

time.

Monitor

program. Some

Statements

system

when

compiler

of

that

describe

execution.

belong

scan

Chapters 1 through 8 noting

of

FLAG

operations.

sized

the

system in

which

data

essentially

provides

programs.

overhead.

user

are

the

procedure

values

Similarly,

to

one

a FORTRAN IV-H

to

is

in

the

debugging

the

typical

written

of

to

data

two

by

to
be

general

Such users wi

the

user a

phase

university

novice

values

programmers.

be

followed during

processed may

generated

classes:

the

II

compiler

unique

of
environ-

changed

thus

set

devel-

be

exter-

by

be

of

exe-

the

1.

Executable

2.

Nonexecutable
program form, as well as
data

Statements

I

ine

on

the

sequence

The first
by

other

trailing)

t See

Appendix

statement/,

statement/,

initial

ization.

defining

form consists

numbers

field,
portions

ignored.

a FLAG program follow a

and

columns 1 through 5,

of

the

Chapter

A.

of

have

program. Labels

that

perform

that

provide

providing

80

spaces

no

effect

is

5, IIControl

computation,

information

information

prescribed

or

columns;

on

the

program. Col umns 1 through

used for

Statements

are

statement

written

input/output

to

the

format.

however,

as

decimal

II,

contains

to

the

compi

program

the

last

labels.

a more

operations,

ler

about

during

Figure 1 is a sample

eight

Statement

integers,

extensive

and program flow

storage

execution

col umns

72

are

labels

with

all

discussion

assignments,

about

FORTRAN

are

used

used for

allow

blanks

input/output

Coding

only

for

the

statements.

statements

(leading,

of

statement

control.

data

types

formats and

Form. Each

identification

to

be

referenced

embedded,

labels.

Introduction

and

or

or

The

body

of

each

statement

be

continued.

acter

other

zero

in column 6. If a

ing

on

continuation

Column
upon
spersed

1 may

the

with

Statements
constants

The

set

of

FLAG

than

blank

contain

program.

continuation

may

have

and

in FORMAT

characters

accepts

or

zero

statement

I ines

are

the

Comment

lines).

blanks

statements).

acceptable

Letters t :

Digits:
Speci

al

characters:

(useful)
Special

t

characters:

(other)

This

character

set

conforms

Figure 1 illustrates a sample

is

written
an

in columns 7

unl

imited

in column 6. The

is

labeled,

ignored.

character C to

lines may

inserted

to

ABC

indicate

appear

as

desired

FLAG

is

D E F G H I J K L M N 0 P Q R

0123456789

+ - * / =

to

the

Extended

FLAG program. An

through

number

of

continuation

initial

the

label

that

anywhere

to improve

()

• , $ I &

Binary-Coded

explanation

line

must

the

in

blank

Decimal

72,

but

of

each

appear

line

is

the

program,

readabil

II

is

given

if

additional

I ines. Each

statement

on

the

initial I ine

to

be

treated

except

ity,

except

STU

Interchange

in Table

space

is

required, a statement

continuation I ine

contains

as a

only

of

the

comment

the

statement;

only,

within a statement

within

literal

fields

V W X Y Z

Code

(EBCDIC)

1.

must

characters

with

(i.e.,

(e.

g.,

standard.

may

contain a char-

blank

or

labels

appear-

no

effect

inter-

in

Hollerith

PROBLEM

PROGRAMMER

I-r--.-.,--r---t"""""+-"',

~-.--,.--r-+-!

hI.O~J,F,

1 1 I

1 3

Sa

m pi e

____________________

II

NT EGE R

T I • , •

iK

=

iii

, • • • • !

.....,lR.;;E=;A,D,

,(

,F,A,C,T,~R,)

lK

= K • F AC T

lJFACT6R

1 2

WRI T E

Iii

,(

'

ii'

,1,

(1

iii

FACTOR,

Ii'

~.

•

,5,

),

,

= '

O'S , 6)

i 1

iii

_

• • , i

.,.

'f~-,.~TR-,-T-·,--y---r--,.__,__.-,---r---.--r-......,........-r--T"'"

,l-1r,J-r3.,J

(j

R

'K~'-'

I i * ---,---"'"

FORTRAN

K

.·-,.--r··--~"T"-.,._r_r_l--y--r--rj-,--,--.,..--,r--r--.--.--...--.---.--,--

,'--'

,---r--r--r---r-...---r--.---r--r

.1,

lr-~T-r--T-r-.--,---rl---r-.,...--,.

.... , ...

.,--

.

-'~·TT-'~-,.-

·--TM-"...-----r--

S!:I:»ls

~

SCIENT:FIC

CODING

Identification

FORTRAN STATEMENT

..-, -,..---r-.

......

,

'·~·-T--~--

,---1'

..

-T----r-r---.--'

"'l"'--,.

, i , i

T~.--.--r-r-r--T

DATA

'---

SYSTEMS

FORM

......

·

-,-,........,-,

-r,---.----.--r--.---.--..

--r""'-'--T-"-r-~

",-,---'---r--~..---r-,--r-r-I

y-1-.----.-.-

-r·-r-r-.-·

T-,"'·"-

...

I-..--

•

Iii

~-;-'-""---'

T

....

'-'-'---'-"--,---"

¥

,.-.~

STC:SP

5 F

eRMA T (t'6)

t--.-·-,-.--r-=6~f-'-+1~_.~R,M:

t--r-T-,-----r-+-+I,!=-r--"'-

'~';jEND

~

-.=....-,-,

--.--.----r-.......-r,--r-,

T .

(::~

:",~-,_T"~

.--.--.----.,

-.,

:-1""':

~-r_~-Q:

f:=:-~=~·~..,...-y-~,---,--.--r.

ir---r-",

~"-r""""-'"

"r

··Y·

....

,...........,.--....,.._·......--.-,

~,

.

..,--.--......--,-._

- ..

------'

-r........-.,.--.,...-r.

.

....,.......,.,

--.--.-.-

--_

~-r:---,---r-...,.....,r-T--r--,--,

•

..,.....-rj

-.-...,..---,.-~

....

_-----------

dollar

sign

($)

tThe
be

used in FORTRAN

Chapter

7), $ follows Z in

character

identifiers,

the

is

accepted,

such as

set

of

2 FLAG Programs

Figure

$,

letters.

1.

though

FIVES,

Sample

not

recommended,

or

$300. For

XDS FLAG Program

as a

letter

the

purposes

of

of

the

the

alphabet.

IMPLICIT

It

may

statement

therefore

(see

Table

1.

Sample

Program

Line

1,2

3

4

5

6

7

8

9

Meaning

The

character

nonexecutabl e statement

A

An

assignment

An

input

from

unit

fied by FORMAT 5

Statement

If FACTOR <

If
If FACTOR >

Statement
factor.
by

the

statement

The
and

one

The C in
the

value

C in

column 1 defines

statement

command

1.

10 tests

FACTOR =

11

In

other

value

continuation

column6causesline9

of

that

The form in

(line

the

0,
0,
0,

is

another

words,

of

FACTOR.

appearing

the

current

that

that

causes

which

14).

value

of
control
control
control

assignment

the

current

on

lines 8 and

line.

value

these I ines

defines

sets K equal

the

the

FACTOR

is
is
is

to

of

to

value

value

and
transferred
transferred
transferred

statement

valueofK

9 is

bea

continuation

FACTOR minus

the

to

of

FACTOR

of

FACTOR

transfers
to
to
to

that

an

as

comments.

compiler

1.

control
statement
statement
statement

assigns

is

replaced

assignment

of

1.

the

to

be

appears

12.

13.

11.

to K the

by

line

variabl

es FACTOR

read

into

external

to

statement

value

the

currentvalue

statement,

8.

This

storage.

to

the

11, 12,

of

the

written

statement

and

K as

integers.

The

value

is

computer

expression

of K multiplied

as

assigns

is

or

13 as follows:

an

initial

to

K times

FACTOR

read

speci-

line

10

11

12

13

14

15

16

Conditional

When

the

GO

(I

ine

6)

occurs;

Statement

The

written

for

The

FORMAT

FORMA T

The END
the

In

this program,
executed
statements

Compilation

12,

WRITE

output

out

designated

control

statement

statement

line

source

program.

four times,

on

- X

TO

an

assignment

statement,

on

unit

FORMAT

statement

informs

if

lines

Cards

statement

108,

corresponding

correspond

the

the

4,

which

statement).

STOP

causes

the

processor

value

statements

5, 12,

is

statement,

executed,

13,

causes

is

normally

execution

to

i ng to

during

of

FACTOR is

and

13 will

an

assigns

the

name

assigned

READ

statement

WRITE

statement

compilation

initially

on

line 7 through

be

unconditional

the

value

of

the

to

the

of

the

program

on

line

on

that

3 as

10 wi

executed

once

transfer

zero

variable K and

Printer

line

it

read

II

to

to

be

5.

12.

has

.by
be

each.

of

the

variable

(see

terminated.

reached

line

5,

executed

control

its

value

statement

the

statement

three

to

statement

K.

to

6,

line

physical

10 will

times,

be

15

end

and

10

of

be

the

FLAG

provides

is

processed as a

Otherwise,

a means for

statement

the

card

is

treated

only

conditional

when

as a

comment.

compilation

the

CX

option

of

statements.

is

specified

on

Any

the

line

FLAG

that

contains

control

Conditional

an X in

card

(see

Compilation

column

Chapter

- X

9).

Cards

1

3

This

feature
intermediate
to

be

removed from
treated
checkout

as comments

procedures.

enables

output

the

programmer to

and

special

the

program. Instead,

only.

Also,

error

These

they

include

checking.

statements

may

easily

in his program

When

the

program is

remain in

be

reinstated

checkout

compiled

the

listing,

at

additional

has

been

without

however,

any

time.

statements

completed,

the

CX

and

may serve as

for

checkout

these

option

purposes, such as

statements

and

the

statements

documentation

do

not

have
are
or

Continuation
is

also

an X card.

lines

for X cards should

For

C

FOR

~

-lsTATEMENT
NUMBER

1 5

X·

X

3 FeRMAT (

x

example:

COMMENT

J

6 7

'l!r

:c~PRINT

10 15

••

_,

PR I NT

also

be X cards;

A.

3 •

X.

3 • U .

6F

furthermore, a normal

E •

Y I

30

F

~

20

B

.J

...•

I I

C.

25

. .

D.

.

7')

. I

V.

W

.

, X,

I

line

may

have a continuation

FORTRAN STATEMENT

35

40

.

.

I

-.~

....

~--

..

--

line

that

45

I

4

Conditional

Compi

lotion

- X Cards

2.

DATA

Numerical

the

numerical
ample,
numerical
the
is

All

lim

Integer
that

Real
of
positive
nificant
rate,

the

execution

identified

data

Integer
Reai
Double

Complex

its

data

is,

-2

data

which

depending

quantities -constants

values

encountered

integer

quantity

processed by a FLAG program

on

31

(sometimes known as

are

or

negative

digits.

5 is

that

of

the

program, a

and

referenced

precision

Values

are

precise

31

to

+2

within

the

values
That is,
on

the

represented

is

of

Quantities

representations

- 1.

range

the

value

and

in a program. A

as

referenced

variable

by

an

identifier.

Doubl e Compl
Logical
Literal

Integer

data

floating-point

5.398

within

this

sixth

most

assigned to the

variables

11511;

by

name

may

can

of

may

79

x

10-

range

significant

the

rather

take

be

classed

ex

the
only

data)
to

or

datum.

- as

distinguished

constant

number

on many

range

be

assigned

can

7.237

the

value

digit

IT,

than

as

of

be

x 10

will

is a

to

three

by its

values

one

of

integers

integral

assigned

75

(i.e.,

zero.

be

in FLAG

quantity

decimal

explicit

rather

seven

types:

from

-2,

values

approximations

65

16-

Real

data

accurate,

are

a means

whose

value

places,

appearance

than

being

147,483,648

within

63

to

16

have

an

while

the

of

identifying

is expl

as

113.

14211.

in a program

restricted

to +2,

this

range.

of

real

numbers,

).

A real

associated

seventh

the

icitly

stated.

A

variable

statement.

to

one. A variable

147,483,647;

the

magnitudes

datum

may

acquire

precision

will sometimes

of

nature

For

is

6+

sig-

beaccu-

of

ex-

a

During

Double
have

Complex
real
imaginary
val ues for real

Double
precision

Logical

Literal

ues. Any

precision

an

associated

data.

complex

data

Data

data

The first

part

values.

are

of

Constants

Constants
each
pi

us

Integer

Integer

2,

are

type

of

or minus

Constants

constants

147,483,647.

data

precision

are

approximations

of

of

the

data.

data

can

acquire

character

the

characters

data

that

data.

AI

operators.

are

may

approximate

of

15+

the

two real

complex

have

though

represented

number. The

the

same form

only

the

strings of up to 255

discussed in

do

not

vary

numeric

The

operator

the

significant

of

complex

data

approximates

as

values

IItrue

Section

in

value

constants

is

not

by strings of

identical

numbers. These

values

complex

'l

characters.

and

considered

digits.

digits.

that

or IIfalse

1 may

are

are

considered

set

the

may

data

ll

appear

referenced

part

The

of

values

approximations

real

part,

be

assigned

except

that

•

Like

logical

in

literal

by naming

as

being

of

the

magnitude

as real

and

the

to

both

data,

data.

unsigned,

constant,

of

an

data.

take

second

each

the

real

literal

their

however.

integer

However;

the

form

real

part

are

and

data

values.

they

constant

double

of

an

datum

identical

imaginary

do

not

There

are

may

be

preceded

(See

Chapter

must not

precision

ordered

approximates

have

to

the

parts

are

numeric

constants

3.)

exceed

pair

set

by

data

of
the
of

double

val-

for

the

Examples:

382

13

997263

1961

1000000000

323344224

000546
382437

8

o

Data

5

Rea I Constants

Real

constants
the
the

numeric
power

value

of

are

ten

by

represented
and

consists of

which

the

by strings

the

letter

numeric

of

value

digits

with a

decimal

E,

followed by a signed or unsigned 1-or

is

to

be

multiplied.

point

Thus,

and/or

the

following forms

an

exponent.

2-digit

The

exponent

integer

are

permissible:

that

follows

represents

n.m

n. mE±e

where

n,

m,

and e are

The plus sign

For

example,

forms:

0.567E+05

567000.

The val

ue

of

constant,

Since

any

but

real

Examples:

5.0

0.01

n.

E±e

n.

strings

preceding e is

. 567E5 has

5. 67E4

E-1

a real

onl y

constant

567E02

constant

the

7 most

may

7.6E+5

6.62E-37

.m

nE±e

of

digits.

optional.

the

meaning.

567

x 105 and

56700.

56700.000E-00

may not

significant

be

written

exceed

digits

in a

the

are

variety

3.141592265358979323846
.58785504

can

also

I imits for real

retained.

of

ways,

be

represented

data.

Any number of

the

user has freedom

by

any

of

of

the

digits

choice

following

may

be

regarding

equivalent

written

form.

in a real

Double Precision Constants

Double
the
double

precision

exponent.

precision

n. mD±e

constants

To

denote a constant

constant

n. D±e nD±e

where

n,

m,

and e are

D

signifies a double

The plus sign

The

value

of a double

may

be

written

preceding

in a

strings

e is

precision

double

Examples:

1. 2345678765432D1
.9963D+3

Complex Constants

Compl

ex

constants

are

expressed as

are

formed

may

be

of

digits

precision

optional.

constant

precision

576.3D+01
.1254D-02

exactly

specifically

written

constant

constant,

an

ordered

in

may

like

any

not

312.

885.

pair

as

of

exceed

but

D-4
D+3

real

constants,

double

the

precision,

following forms:

the

only

the

15 most

of

constants

except

limits for

significant

in

the

that

the

exponent

double

form

the

letter

must be

precision

digits

are

D is used

present.

data.

retained.

instead

Any

of

Thus, a

number of

E in

digits

6

Constants

where

c 1

and

c2 are

signed

or

unsigned,

real

constants.

The

complex

constants

(1.

(98. 344E
( - 1 . , - 1

Neither

Double Complex

Double
part

is a

Examples:

(.757D6,3D-4)
(7. ,ODO)
(-4.

Neither

Logical

Logical

constant

have

34,52.01)

000.

part

of a complex

complex

double

286DO,

part

of a double

Constants

constants

(c1,c

values

as

indicated:

11,34452E-3)

)

Constants

constants
precision

757000.0DO

1.

3)

complex

may assume

) is

interpreted

2

983.

constant

are

formed

constant,

7.0DO

-4.286DO

either

1.

34
44

-

1.

0 1

may

exceed

in

exactly

the

complex

+

+
+ 1.

constant

of

the

as

+

52.0li

+

34.

000.

.0003DOi

O.ODOi

3DOi

may

two forms

meaning

452i

Oi

the

value

the

same

constant

exceed

c

+ c

i,

where

2

1

limits

established

way

as

complex

becomes a double

the

value

limits

i =

/=1.

for

real

constants.

complex

established

Thus,

data.

If

either

constant.

for

double

the

following

the

real

precision

complex

or

imaginary

data.

.

TRUE.

where

these

forms

Literal

Constants

A

literal

constant

where

is a string of up to
character

Within a literal
apostrophe).

by

blanks

are

Examples:

'ALPHANUMERIC
'IIDON"T!'"

Literal

constants

1. An

2. A constant

3.

argument

A PAUSE

.

FALSE

have

has

strings.

constant,

For

example,

not

considered

can

to a

item in a

statement

.

the

logical

the

form

255

two

consecutive

'AB"CD'

to

be

INFORMATION

appear

in

three

function

DATA

('Sl

form only)

or

statement

values

"true"

alphanumeric

quotation

represents

consecutive.

'

contexts:

subroutine

and

"false",

and/or

special

marks may be used

the

five

characters

respectively.

characters.

to

AB'CD.

Note

that

blanks

represent a single

However,

quotation

are

significant

quotation

(') marks

in such

mark (or

separated

A I

iteral

constant

cannot

appear

as

an

element

of

an

expression.

Constants

7

[

Identifiers

Identifiers
as subprograms

are

strings

and

of

letters

COMMON

and

decimal

blocks.

digits,

(See

Chapters 7 and

the

first

of

which

8 for discussions

must be a

letter,

of

COMMON

used to name

and

subprograms.)

variables

as well

Identifiers
ignored;
for

Examples:

in FLAG may consist

therefore,

clarity,

A345Q

X

ELEVAT

Variables

Variables
ables
a I

If

a

following

1.

2.

These
Consequently,

such (see
established

Scalars

are

may

be

iteral

string.

variable

implicit
Variables
Variables

classifications

"Explicit

for

it

data

any

has

is

whose
whose

double

the

ON

advisable

STRESS

L9876

whose

of

the

data

Normally,

not

been

typing

identifiers
identifiers

are

precision,

Type

applicable

of

TIME

and

not

to

use

J3

DIFFER

values

may

types.

integer

assigned

conventions

begin
begin

referred

Statements"

to as

complex,

data

up

to

six

alphanumeric

ONTIME

variables

to a particular

types.

are

identifiers

MELVIN

SETUP

vary

during

(There

is

are

assumed:

with

the

with

any

the

IIUKLMN

double

in

Chapter

are

identical.
that

program

no such

used.)

data

letters

other

complex,

7). The

characters.

There

correspond

QUANTY

execution

entity

as a

type

(see "C

I,

J,

K,

letter

are

rule".

and

values

are

no

to

FLAG

and

literal

lassification

L,

M, or N

classified

logical

assigned

Blank

restricted

statement

are

referenced

variable;

are

as real

variables

to

variables

characters

identifiers

types.

any

of

Identifiers",

classified

variables.

must be

may

embedded

with

an

type

of

as

integer

explicitly

not

in

in FLAG;

identifier.

variable

Chapter

variables.

exceed

identifiers

however,

Vari-

may

contain

7), the

declared

the

limits

are

as

A

scalar

Examples:

Jl

Arrays

An

array

arrays

FLAG

arrays

on

Array

EI

ements

array

An
identifier,

where:

v is
si
n is

Subscripts

variable

is

see

el

is a subscript (see below)

is a

single

datum

NAME SCALAR EQUATE

data

in

which

the

data

may

have

up

to

seven

"Array

Declarations"

ement

is a member

followed by a

the

array

name

the

number

list

of

subscripts,

of

the

of

subscripts

entity

accessed

form

an

ordered

dimensions

in

Chapter

set

of

data

enclosed

which

E

and

7.

comprising

must

be

via

an

NEW

set.

Associated

are

referenced

in

parentheses

equal

identifier

DHO

an

array.

to

the

number

with

by

of

an

Array

the

appropriate

XXX8

an

array

identifier.

el ements

of

dimensions

type.

is

the

property of dimension.

For a

complete

are

referenced

of

the

array

discussion

by

the

array

(0 < n ~ 7)

be

any

[

A subscript may
not

integer

Identifiers/Variables

8

it

is

converted

expression

to

that

integer

mode (by

has a

resultant

truncation).

mode

of

integer,

real,

or

double

precision;

if

the

result

is

The

evaluated

nated

as ALPHA(K-4),

Examples:

result

Subscripts

for a

the

subscript

value

of

must

K must

always

be

be

greater

greater

Array

than

than

4.

Elements

zero.

For

example,

if

an

array

element

is

desig-

[

MATRIX

CUBE

DATA
J35Z
BOB

Functions

Functions
quantities,
tion
A

function

in

where

Functions

IJKLMN

are

called

reference

reference

parentheses

is

is

a.

I

IIArguments

are

rule

(3,9,5,7,6,1,2)
(5*

J,

P,

3)

(I,

J, K,

L,

M,

N)

(1+4,

6*KRAN-2,

(3,IDINT(DSQRT(D)))

subprograms

arguments,

constitutes a reference

the

name

an

argument.

classified
applies.

is

denoted

of

and

in
The

that

the

Dummies

the

are

and

function

Arguments may

same

type

ITEMP)

referenced

produces a single

to

the

value

by

the

identifier

,

Chapter

way

be

as

variables;

ll

of a function

MATRIX(3,
CUBE(5*

DATA(I,
J35Z(I+4,
BOB(3,IDINT(DSQRT(D)))

as

basic

that

constants,

8).

is

not

9,5,7,6,

J,

P, 3)

J,

K, L,

6*KRAN-2,

elements

quantity,

produced

names

variables,

that

affected

1,2)

M,

N)

ITEMP)

in expressions. A

called

by

the

function,

the

function,

expressions,

is, unless

by

the

the

the

type

function

when

followed

type

is

of

its

function

value.

operating

by a list

or

array

specifically
arguments.

acts

upon

one

or

more

The

appearance

on

the

of

arguments

or subprogram names (see

declared

of a func-

given

argument.

enclosed

otherwise,

the

[

Examples

SIN(A+B)

Many

library

of

function

CH ECK(7.

functions

references

3,

are

provided

are:

J,

ABS(Y)) KOST(ITEM)

in FLAG. In

addition,

the

user may

define

his own

functions

(see

Chapter

8).

Functions

9

3.

EXPRESSIONS

Expressions

ences.

may
operands.
and

An

be

unary -that

the

evaluation

are
expression

Expressions may

Arithmetic

An

arithmetic

stant,

variabl

The

arithmetic

Operator

+

*

/

**

Arithmetic

expressions

A

-TERM

1.

2607

ACE -

DEUCE

W90ML * DE

F(5.8E2) = A /

strings

of

operands

may

contain

is,

they

may

be

of

any

expression

Expressions

expression

e,

or

operators

is a

function

Operation

Addition
Subtraction
Multiplication

Division
Exponentiation

are:

may

references

{binary}

be

+ W9CMI / XKA9RU

B9J

(L)

separated

subexpressions; subexpressions

operate

classed

sequence

or

(binary)

of a

relatively

on a

as

arithmetic,

has a

unique

of

integer,

connected

Positive

or

Negative

by

operators.

single

(unary)

simple

result.

real,

by

operand.

logical,

double

arithmetic

(unary)

form

Operands

are

They may

or

relational.

preCISion,

operators.

may

be

constants,

expressions

also

be

All expressions

complex,

enclosed

binary,

and/or

variables,

in

parentheses.

operating

are

double

or

on

single

complex

function

Operators
pairs
valued,

refer-

of

con-

or

the

more compl

x + (112 *

-B

+ SQRT(B ** 2 - 4 * A * C) + T *

(X

+

Y)

-((M + N) * (Z -

Evaluation Hierarchy

The expressi on

A +

B/C

might

be

eval

(A + B)

or

as

A +

(B / C)

10 Expressions

(G)

** 3 +

uated

/ C

icated

**

L(3)

O.

7352986E-7

Q(J»)

as

form

+ N / SDS) -

(H)

(S

+ B / I * (K(J) /

(V1

-

VO) + (Z

1 -

ZO»)

Actually,

that

it

is

the

latter

necessary

form is

to

formulate

the

way

rules

the

expression

for

expression

is

interpreted

evaluation

without

so

that

explicit

such

grouping.

ambiguities

do

This

exampleillustrates

not

occur.

Subexpressions
sions

as

x *

where

(I +

sion. The

1.

The innermost subexpressi
eval

2.

The

Operation

Exponentiation

Multiplication
and

Additi
Subtracti

Some

additional

1.

At

anyone

evaluated

2.

Consecutive

have

in

(Z

+ Y *

L)

may

evaluation

uated.

arithmetic

Division

on

and

on

from

(H

be

conventions

level

exponentiations

been

defined

- G / (I +

called

the

hierachy

operations,

Operator

**
*

/

+

of

evaluation,

left

to

right.

as

expressions

L) -W) + M(8))

innermost

is,

therefore,

on,

followed

in

the

following

Order

1 (highest)

2

3

are

necessary.

operations

Consequently,

are

performed

enclosed

subexpression,

as

follows:

by

the

next

order

of

of

the

1/

J / K / L is

left

to

in

parentheses.

and

innermost

precedence:

same

order

right.

(H

- G /

subexpressi

of

equivalent

Thus

It is

(I + L) -W)

on,

precedence

to

also

unti I

(except

((I /

possible

is

the

next

all

subexpressi ons

for

J)

/

K)

to

have

nested

innermost

exponentiation)

/

L.

subexpres-

have

subexpres-

been

are

I

L

3.

4.

A ** B ** C

is

interpreted

The use
differentl

The

sequence

As

an

algebraic

written

as

Example:

The expressi on

is

evaluated

as (A **

of

parentheses

y.

"operator

(A + B)

A *

(B

F + G

r =

1

D - r - H

r = C * r

4 3

is

notation,

/

c.

+ C *

(D

in

the

2

B)

** C

recommended,

operator"

parentheses

- E /

sequence

(F

is

+ G) -

as

many FORTRAN systems

not

permissible.

are

used

to

define

H) + P(3))

Therefore,

evaluation

interpret

A * -B must

sequences

consecutive

be

expressed

explicitly.

exponentiation

as

A * (-B).

A+B

Thus,

-C-

is

r 5 = B + r 4 +

r = A * r

6 5

where

the r are

the

i

P(3)

vari ous

level s of

eval

uati

on.

Arithmetic

Expressions

11

Mixed

V/hen

elements

expression

specifications

Expressions

an

arithmetic
may

contains

not

of

expression

appear

more

its

elements.

contains

in

an

arithmetic

than

one

type

Table 2 illustrates

Table

2.

elements

of

Mode

expression

element,

of

Mixed

of

more

than

except

the

mode

how

the

mode for

Expressions Using

one
as

of

type,

function

the

expression

mixed

IT

IS Known

arguments

expressions

Operators

as a mixed

(see

is

determined

+ - * /

rule

by

is

determined.

expression.

2,

,below).

the

type

When

and

Logicai

an

length

+-*/

INTEGER

REAL

DOUBLE
PRECISION

COMPLEX

DOUBLE
COMPLEX

It

can

Type
Doubie
Complex
Real
Integer

be

seen

that a hierarchy

Compiex

or

Double

INTEGER

Integer

Real Real

Double
Precision Precision

Complex Complex

Double

Complex

of

type

Precision

REAL

Real

Double

Double Double
Complex

and

length

Precedence

1

(highest)

2

3

4

DOUBLE
PRECISION COMPLEX

Double
Precision

Double
Precision

Double Double
Precision

Double
Complex

Complex Complex

specifications

exists.

The

order

Complex

Complex

Complex

Complex

Double

of

precedence

DOUBLE

COMPLEX

Double

Complex

Double

Complex

Double

Complex

Double

Complex

Double

Complex

is:

I

L

12

During

evaluation

resultant

The

1.

2.

Mixed

mode

foil owi ng rul es al so

Subscripts
ated

in

their

Only

expression

Expressions

of

mixed

of

each

operation

appl y to

and

arguments

own mode (i.

elements

Base

T

...

+o,..,or

£

...

~~~.

Real
Double

Complex

Doubl e Compl

Precision

expressions

are

e.,

of

Table

the

will

be

mi

xed

expressi

independent

integer)

the

types

3.

Valid

mode

as

shown in

of

and

shown

Type

of

an

ons:

the

expression

neither

in

Table 3 may

Combinations

operand

Table

affect

2.

nor

in

be

will

which

are

for

combined

"

} {

}

ex

**

**

be

converted,

they

appear.

affected

Exponentiations

by

with

,

the

Exponent

T_,,~~~~

1I11"'~"'"

Real
Doubl e

Integer

if

necessary,

These

mode

of

an

exponentiation

Preci

so

expressions

the

outer

operator.

si

on

that

the

are

evalu-

expression.

4.

5.

6.

The mode
arithmetic

Complex
these

types,

that

is

higher

Integer,
to

have

Values

of

of

the

expression.

of

the

results

of

an

exponentiation

operations

and

double

it

than

real,

imaginary

expressions, subexpressions,

acquires

(or

and

double

parts

(see

Table

precision

double

different

precision

of

zero.

elements

complex
from)

1).

have

type.

all

values

and

operation

the

This is

its

constituents.

that

appear

elements

same

may

can

level

the

in

not

be

determined

of

precedence.

only

case

complex

exceed

in which

or

double

the

value

in

the

same manner as

If

an

an

complex

limits

expression

expression may

expressions

associated

contains

that

for

both

have a type

are

assumed

with

the

other

of

mode

Relational

The form

where

Evaluations

pressions
Relational

Examples:

of a relational

e1 and

e

is a

of

are

operators

Operator

.IT.

.LE.

.EQ.

.NE.
.GE.
.GT.

l.LT.6

O.GT.8

O.

LT.

(2.

O.

LT.

- (2. **

Expressions

expression is

are

logical

cause

Meaning

less

Less

Equal
Not
Greater
Greater

N)

N)

arithmetic

operator

expressions

type.

comparisons

than

«)

than

or

to

(=)

equal

than
than

is

true.

is

false.

is

always

is

always

2

relational

relational

of

**

expressions whose mode is

(see below)

result

in

either

equal

to

(I)

or

to

equal

between

(~)

to

arithmetic

(:;:::)

(»

true,

while

false.

of

the

integer,

two

values

expressions.

real,

IItrue

or

double-precision

"false

ll

ll

or

,

i.e.,

relational

ex-

When two

each
mode expression is

A

test

these
amount in
binary

It

is

where

relational

arithmetic

in its own mode. The comparison is then made in

for

equality

quantities

their

representations

not permissible

(X

(L.LT.

(X

.GT.

expressions requires.

expressions

converted

between

are

only

binary

to

.GT.

O.2345E6))

O.

2345E6) is a

approximations

representations.

will

nest

are

to

the

real or

produce

relational

relational

compared,

mode of

double

It

these results.

expressions such as

using a

higher

precision

to

most

values,

can

only

subexpression,

relational

the

mode of

precedence.

quantities

numbers

be

said

that

rather

operator,

higher

may

not

be

that

are

computations

than

an

arithmetic

the

two expressions

precedence;

meaningful on a

"essentially"

whose

operands

expression,

i.

e.,

binary

equal

Relational

the

may

and

as

are

first

evaluated,

value

of

machine.

differ

by a small

results

have

the

definition

Expressions 13

the

lower

Since

exact

of

logical

Expressions

Logica! expressions are expressions of the

where

e.

are

I

c.

I

Evaluations of logical expressions result in

Logical elements

1. a ·Iogical

2. a logical

3. a relational

4.

any

5. a logical expression

6.

any

Logical Operators

There

are

Operator

.NOT.

.AND.
.OR.

Table 4

1. .

NOT. e is

2.

e1 .AND.

3. e1 .OR. e2 is

logical elements.

are

the

variable
constant

of

the

above

of

the

above,

three

logical operators:

illustrates

e2 is

binary

are

defined

or

expression

encl osed in

enclosed

preceded

Type
unary
binary
binary
the

meanings

"true"

only when e is

"true"

"true"

logical

when

operators

as

one

of

the following

logical

function

parentheses

in

parentheses

by

the

unary

of

the

logical operators.
"false".

only when both e1 and

either

or both e1 and

form

(see below).

either

of

reference

logical

the

entities:

operator

e2 are

e2 are

two

"true".

values,

.NOT.

"true".

"true"

or

"false".

Expression

e True

e False

e1 False e2 False

e1 True e

e 1

False

e1 True e2 True

14 Logical Expressions

Values

---

---

2

E2

False

True

Table

4.

Evaluation

.NOT.

False

True

---

---

---

---

of

Logical Expressions

e

---

Evaluation

e

1

.AND.

e

2

e

1

---

---

False False

False

False

True True

True

True

.OR. e

2

Evaluation Hierarchy

Parentheses
expressi ons.

A

.AND. B .OR.

does

not

have

A

.AND.

where

(B

.OR.

The

evaluation

1.

arithmetic

2.

relational

3.

the

innermost

4.

the

logical

Operator

.NOT.

.AND.

are

used to

Consequentl

the

same

(B

.OR.

0(3)

hierarchy

expressions

expressions

logical

operations,

Order

1

(highest)

2

define

y,

0(3)

meaning

0(3)

.NE.

X)

for

(the

subexpression,

evaluation

.NE.

as

.NE.

may

logical

relational

in

the

sequences

X

X)

be

called a logical

expressions

operators

followed

following

order

explicitly,

subexpression.

is

are

all

by

the

of

precedence:

of

next

ina

manner

equal

precedence).

innermost

similar

logical

to

that

discussed

subexpression,

etc.

for

arithmetic

.OR.

example,

For

L

is

interpreted

L

Note:

3

the

expression

.OR

•.

NOT. M .AND. X .GE.

as

.OR.

«.NOT.

It

is

permissible

other

e

not

is

e1 .AND

is

legal.

ator

.NOT.

if X and Z are

are

not

.NOT. X .AND.

.NOT.

words

.AND

1

valid,

equival

M)

Two

applies

(X

.AND.

to

have

..

OR. e

while

..

NOT.

consecutive

should
logical
ent:

.AND.

(X

.GE.

two

contiguous

2

e

2

.NOT.
be

variables

Z

Z)

Y

Y))

operators

enclosed

having

logical

are

in

parentheses

the

values

operators

not

permissible.

if

TRUE

only

it

contains

and

when

The

FALSE,

the

second

logical

two

or

more

respectively,

operator

expression

quantities.

the

is

.NOT.

to

which

For

following

i in

the

oper-

example,

expressions

the

first

In
Z

(also,

In

the

is

NOTed,

expression

FALSE), results in

second

expression X .AND.

resulting

.NOT.

in

the

X is

the

value

evaluated

value

Z is

TRUE

first

FALSE for

evaluated

for

the

and

produces

the

expression.

first

and

expression.

the

produces

value

the

value

FALSE. This,

FALSE. Then

Log i ca

when

ANDed

the

with

value

I Express ions 15

FALSE

4.

ASSIGNMENT

STATEMENT

Many kinds
ment,
matical

A

si

mpl e assignment

where

This

not

declare

of

which

notation.

v = e

v is a
e is

statement

that

statements

defines a computation

variable

an

means, lIassign

N = N + 1

is

not a contradiction:

The expression
expression is
verted

to

evaluated

the

type

statement

arithmetic

v

.!.:.

equal

need

not

of

the

are

recognized

(a

scalar

or

logical

to

it

increments

be

the

in

its

own mode,

variable

has

the

or

to v the

e,

but

same

according

Table

by

the

FLAG

to

be

performed

form

an

array

element

expression. (v must

value

rather

it

the

current

type

as

the

independent

to

5.

Mixed

and

of

the

expression

sets v

equal

value

variable,

Table 5 and

Variable

compiler.

is used in a

of

any

be a logical

to

of

N by

although

of

the

Types

The most

type)

e.

II

It is not

e.

Thus,

1.

type

of

assigned

and

Expression Modes

manner

variable

the

in

practice

the

variable.

to

the

basic

of

similar

only

an

equation

statement

it

usually

variable.

these
to

if

e is a

Then,

is

the

equations

logical

in

the

is. When

if

permissible,

assignment

in normal

expression)

true

sensei

it

state-

mathe-

it

does

is

not,

the

it

is

con-

Variable

Type

integer

real

double
precision

complex

double
complex

logical

The symbols used in

Symbol

X

F

Meaning

Direct
The

the
truncated

will

The

cision

Expression

integer

X

F

real

I

1

X P

double

precision

I I

F P X

R

D D D P

R

R X

N N N

Table 5 have

assignment

value

greatest

be

truncated

variable

is less

is

truncated

integer

to

4274,

is assigned
than

the

following meanings:

of

the

exact

to

integer.

less

than

or

and

-0.6

to 0).

at

the

high-order
the

real

that

of

integers,

value.

The

truncated

equal

to

Values

end

as

or

double-precision

conversion

the

that

well.

Mode

complex

value

absolute

are

greater

Results in this

to

real

complex

I

I

R

D D

i

I

N N

is

equal

to

the

val

ue

of

the

expression

than

the

maximum

case

generally

approximation

precision

may

of

cause

double

I

R

P

X

sign of

the

value.
a loss of

logical

the

expression times

(e.

g.,

4274.9983

size

of

an

are

not meaningful.

Since

significant

N
N

N

N

N

X

integer

real

is

pre-

digits.

16 Assignment

Statement

Meaning

The

P

R

D

N

Examples:

precision

The

real

part

of

The

real

imaginary

Not

allowed.

part

the

variabl

part

part

of

of

of

the

the

of

the

e is

the

value

variable

set

variable

variable

A = B

Q(I) = Z

L = F

CRE(8,

PI

**

2 + N *

.OR

..

NOT. C .AND.

ED)

= R (ALL, MEN)

= 4 * (ATAN(O. 5) + ATAN(O. 2) + ATAN(O. 125))

(L

- J)

(R.

GE.

is

increased
is

to

zero.

is

23.

assigned

assigned

is

set

to

9238E-l)

or

decreased

the

the

zero.

real

approximation

double

accordingly.

precision

approximation

of

the

expression.

of

The

the

expression.

imaginary

The

Assignment

Statement

17

Each

executable

less

this

Labels

If

program

identified

be

referenced.

sequence

control

by

labels.

statement

is

interrupted

is to

be

Nonexecutable

5.

CONTROL

in a FLAG program is

or

modified

transferred

by

to a particular
statements

executed

a control

statement,

may

have

STATEMENTS

in

the

order

of

its

appearance

statement.

that

labels,

statement

but,

except

must
for FORMATs,

be

identified.

in

the

the

source

Statements

labels

program,

should

un-

are

not

Statement
zeros

Statement
compilation

GO

GO
statements:

Unconditional

This

where
one

labels

are

not

857

labels

TO

Statements

TO

statements

statement

GOTO

k is a

whose

98

consist

significant.

00857

may

be

or

execution.

transfer

unconditional,

GO

TO

Statement

has

the

k

statement

label

is k. For

GO

TO 502

X = Y

of

up

to

The following

8 5 7

assigned in

control

assigned,

form

label.

example,

The

five

085

result

decimal

any

from

and

in

digits

labels

7

order;

one

point

computed.

of

the

and

are

equivalent.

their

numerical

in a program

execution

must

of this

be

greater

values

to

statement

than

have

another.

zero.

no

FLAG

is

that

effect

the

Embedded

on the

includes

next

blanks

sequence

three

forms

statement

and

leading

of

statement

of

GO

executed

is

TO

the

502

statement

Assigned

The format of

r..n

....,...., •

where

v is a nonsubscripted

kj is a statement

18

Control

A = B

502

will

GO

TO

Statement

the

assigned

Tn

"rtk

...., • 'L\

l'

the

statement

Statements

be

executed

k

2'

GO

I.,

"3'

...

labels

label

immediately

TO

is

I.,

\1

,

"n'J

integer

k 1 - kn·

(the

list

variable

after

enclosed

that

in

the

GO

has

been

brackets

TO

statement.

assigned

is

optional).

{via

an

ASSIGN

statement,

see

below} one of

Each

label

be

the

label

corresponds

Examples:

appearing

of

a program

to

the

current

in

the

statement).

assignment

I ist must

be

This

Of

defined

statement

the

variabl

in

the

program in

causes

e (v). I

control

which

to

the

be

transferred

GO

TO

statement
to

the

statement

appears

label (k.)

(i.

e.,

must

that

ASSIGN

GO

The

GO

ample) must

Computed

The

computed

where

k. is a

I

v is a nonsubscripted

This

statement
variable
ing

the

but

is no

Examples:

5371

TO

LOC, (117,56, 101,5371)

TO

statement

have

GO

TO

GO

statement I abel

causes

v,

for 1

computed

longer

so

TO

LOC

transfers

been

set

Statement

TO

statement

control

~ j ~

n.

GO

TO

considered.

control

bya

previously

integer

to

If

j is

statement.

to

the

executed

is expressed

variable

be

transferred

not

between 1 and

In most

statement

ASSIGN

as

whose

to

the

previous

labeled

value

statement

n,

5371.

statement

determines

whose

no

transfer

FORTRAN systems,

Note

prior

to

label

occurs,

that v (the

to

its

which

of

is kj,

and

control

this

variable

execution

the

k.

where

situation

in

control

I

j is

passes

has

11

LOCI

the

the

to

been

in

the

GO

TO

will be

integer

the

statement

considered

above

ex-

statement.

transferred.

val

ue

of

the

follow-

an

error,

Statement

GO

TO

(98, 12,405(3), N

GO

TO

(1,8,7,562),

GO

TO

(4,88,

GO

TO

(63,9,3,2),

ASSIGN

The

where

Examples:

Statement

ASSIGN

ASSIGN k TO

k is a
v is a nonsubscripted

ASSIGN 5 TO

ASSIGN

ASSIGN

ASSIGN

statement,

statement

22

1234

99999

1),

v

JUMP

TO M

TO

TO

I

N

J

used

to

label

integer

IRETURN

IERROR

Expression

3

2

o

8

assign a

variable

label

Value

to a variable,

Transfer

405

8

next

statement

next

statement

has

the

to

form

A

variable

that

has had a

label

assigned

to

it

may

be

used

only

in

an

assigned

GO

TO

statement.

ASSIGN

Statement

19

A

variable

a

variable

ill ustrates improper usage:

that

has most

that

has not been assigned a label may not

recently

had a label assigned

to

it

should not be used as a numeric

appear

in

any

context

requiring a

quantity.

label.

Conversely,

The following

case

ASSIGN
A =

This usage is not permissible
is

loaded.

703 TO

HI/LOW

Furthermore,

HI

M = 5

cannot

be

substituted for

ASSIGN 5

or

vice

versa,

IF

Statements

Very often

may be

arithmetic

Arithmetic IF Statement

The format for

called

to

M

because

it

is

desirable

conditional transfer statements,

and

logical.

arithmetic

the

to

IF

because

integer

constant

change

statements is

the logical flow

the

value

of

HI

is

"5" is implied in the first

of

are

used for this purpose. There

indeterminate,

a program on

since its

case,

the

basis

and

of

value

depends on where

the

label "5" in

some test.

are

two forms

the

the

second.

IF

statements, which

of

IF

statements:

program

where

e is

The

arithmetic

IF

IF

IF

If

e is a real

the

expression involves

For this reason, floating

for

zero.

Examples:

Statement
IF
IF
IF
IF

an

e < 0,

GO

e =

0,

GO

e>

0,

GO

or

(K)

1,2,3

(3

* M(J)

(C(J,

10)

(NEXT +

expression of

IF

statement is

TO

TO

TO

double precision expression, a

-7)

+ A / 4)

LAST)

integer,

are

statement

interpreted

k1

k2

k3

any

amount of computation, a very small number is more

point

arithmetic

76,4,3

23,

12,

3,156,3

12

real,

labels.

to mean

IF

Expression
47802

-6

0.0002

o

or

double

test

for

statements

value

precision modes.

exact

zero

may not be meaningful on a binary machine.

generall

y should not be programmed

Transfer

3

76
12

156

to

likely

to result than

to

have a unique branch

an

exact

If

zero.

Logical IF Statement

The logical

20

IF

Statements

IF

IF

statement is represented as

(e) s

where:

e is a

The

statement

lowing

after

the

Examples:

IF

IF

IF

CAll

This

statement,

foil owi ng forms:

CALL
CALL p (a1, a2, a3,···

logical

is

any

executable

s is

the

logical

statement

(FLAG

.OR.

(OCTT *

(.NOT.

Statement

P

mode

expression

statement

executed

IF

s,

TRR

SWITCH2) REWIND 3

used

statement

unless

L)

GO

.L T.

5.334E4)

to

call

if

the

expression e has

is

executed.

the

statement s causes a transfer.

TO

3135

CALL THERMAL

or

transfer

,an)

except a DO

The

control

to a subroutine

statement

the

value

statement

or

another

IItrue";

following

subprogram

logical

otherwise,

the

logical

(see

IF

statement

the

next

IF

will

Chapter

executable

be

executed

8),

may

take

statement

in

any

either

of

fol-

case

the

[

where

P is

is

a.

i

arithmetic
subprogram

A

subroutine
used in
For

Arguments
many results
store

A

A

to
not

Other

an

example,

CALL CHECK

the

resulting

CALL INVERT(A,

complete

subroutine

be

executed

cause

it

examples

the

identifier

an

argument,

is

similar

expression.

that

are

scalars,

as

desired.

matrix

discussion

name

has

as

a resul t

to

take

of

CALL

of
which

or

logical

names.

to a function

Furthermore,

array

The

in

K,

B,

D(J))

of

the

no

type

of

on

any

implicit

statements

the

subroutine
may

expressions,

(See

"Arguments

elements,

following

the

array

usage

and

(e.g.,

the

CALL.

subprogram.

be

any

of

the

statement

and

except

type.

are

that

it

while a function

or

arrays

call

might

B,

and

set

D(J)

forms

of

arguments

real,

integer)

Therefore,

given

below.

following:

label

Dummies",

does

not

must

may

be

be

used

to

equal

to

associated

the

appearance

Statement

Constants;

arguments

Chapter

necessarily

have

at

least

modified
invert

to

the

determinant

supprograms

with

labels

subscripted

(&a.,

where

8.i

return a value,

one

argument, a subroutine

by a

subroutine,

the

matrix

A,

of B ..

is

contained

it;

it

merely

of a subprogram

are

identifies

identified

or

nonsubscripted

a.

is

the

I

and

effectively

consisting

in

Chapter

name

by a

statement

must

not,

of K rows

the

block

in

a CALL

preceding

label),

therefore,

may

returning

and

8.

of

statement

ampersand.

variables,

or

have

none.

as

columns,

instructions

does

array

be

or

CALL ENTER(&44, N)

CALL RX23A(X

CALL EVALUE

** Y -

7,0,

SQRT(A * A + B *

B)

/ DIV,

TEST)

CALL

Statements

21

RETURN

The

Statement

RETURN

RETURN
RETURN v

statement

causes

an

exit

from a

subprogram.

It

takes

one

of

the

forms

[

[

where v is
number
and

A RETURN
physically.
appear

The
subroutine

it

which

The

is
(counting
the

compile-time

Examples:

of

Dummies" in

in a

first form,

causes

the

the

second

used

to

determine

from

entry

to

Calling
33

CALL IT

66 X(8) =

an

integer

asterisks

statement

There

main

to

the

latest

function

form, RETURN v, is used

the

diagnostic

Program

constant

that

Chapter

must

may

program.

RETURN

first

executable
value
reference

which

left

to

right

subprogram

(LOCK,

Y(C,

K)

or

appear

be

8 for a

be,

chronologicall

any

(without

assigned

appeared.

statement

in

the

did

to

be

produced.

RET,

QR, & 11, &883)

in

number

+ CHEBY(Z, Y)

INTEGER

the

SU

BROUTINE

discussion

of

RETURN

the

v) is

the

statement

to

the

function
(See

to

provide

label

SUBROUTINE

not

contain

variable

on

the

y,

the

statement

following

also,

an

in

the

calling

statement)

any

asterisks

whose

value

statement

use

of

asterisks

last

statement

statements

usually

the

CALL

name

to

be

IIFUNCTION

alternate

argument

in

must

be

(see

11

SUBROUTINE Subprograms"

in SUBROUTINE

executed

in a

subprogram.

used.

In a

statement

returned,

exit

corresponds

the

as

Subprogramsll,

from a SUBROUTINE

list

will

to

dummy

list,

Subprograms
SUBROUTINE IT (i,

RETURN 1

greater

in

any

A RETURN

subroutine,

that

call

the

function

be

used

the

statement

the

than

zero,

statements).

subprogram,

it

ed

the

value,

Chapter

subprogram.

as

the

RETURN

but

statement

returns
subroutine.

to

8.)

return.

label

that

statement

X,

P,

no

greater

and

but

it

control

the

The

The

will

will

*,

*)

than

11

Arguments

need

not

should

not

from

In a

functi

expression

value

vth

asterisk

be

used.

cause

the

of

a

the

be

in

last

on,

v

If

When

subroutine
statement
the

return

DO

Statement

These

statements

depends

DO
DOkv=e1,e

where

k
v

el'

e2'

In

the

second

IT

is

called

883

if

the

RETURN 2 exit

from

the

function

are

used

on

the

value

of a variable.

k v = e

value

,e2,e

1

2

is a statement
is a nonsubscripted

and

e3

is

positive.

form,

e3

,

3

are

and

to

or

label

integer

the

by

statement

is

to

the

control

not

integer

preceding

is

executed.

point

the

repetitive

The

defined

variable.

constants

33,

of

ca

DO

statement

before

greater

comma

return

When

II

in 66.

the

are

is

to

statement

the

execution

may

DO

statement.

than

zero

omitted;

function

of a group

be

written

or

unsigned

in

this

RETURN 2
END
FUNCTION

RETURN
END

11

if

the

RETURN 1

subprogram

of

statements.

nonsubscripted

case

the

CHEBY (ARG, EXP)

exit

CHEBY

value

is

The

1 is assumed

is

called

number

integer

executed,

by

statement

of

repetitions

variables

for

e3.

or

to

whose

66,

22

RETURN/DO

Statements

A

DO

statement

called

a DO

DO

statement.

The

execution

1. The

2.

3.

4.

5.

6. The

The

ly. For this

FORTRAN systems may
The

than

Logical

variable v is

The

range
After
The

value

If

v

is

ing

the

actua

where

range

terminal

one

DO

statement

GO
Arithmetic

IF

loop,

of a DO

of

each

iteration,

ofv

greater

one

I number

max

([

the

brackets

of a DO

reason,

statement

of

the

following:

TO

statement

IF

statements

indicates

and

The

statements

is

than

whose

2

e

-

~

loop will

statement

that

all

last

statement

loop

assigned

the

then

e2'

label
of

iterations

1

e

] )

+ 1, 1 for e

represent

it

is

recommended

interpret
of a DO

are

specifically

the

block

statements

in a

proceeds

the

value

is

executed

value

compared

control

is

k).

the

always

this

range

of

statements

within

DO

loop

in

the

following

of

e1.

for

one

of v is

incremented

with

the

is

passed

Otherwise,

defined

be

situation

-I

3

largest

executed
that

(i.

e.,

allowed

by

0

initially

it,

except

is

the

iteration.

terminal

to

the

statement

the

process

the

DO

integra I va

at

least

satisfied

differently.

the

statement

as

terminal

following

for

the

terminus

manner:

by

the

value

(e

following

is

repeated

statement

lue

not

once,

whose

RETURN
STOP
PAUSE

statements

it

are

opening

and

bears

value

).

2

is

given

exceeding

even

DO

loops

label

statement

of a DO

to

be

executed

DO

statement,

the

of

e3.

If

the

terminus

from

step

by

the

if

the

conditions

should

is

k)

may

statement

statement

statement

e3 is

not

(i.

2.

va

lue

of

not

be

be

any

range.

repetitively.

constitute

label

k.

present,

e.,

to

the

the

expression.

for

termination

used,

especially

executable

Such a block

the

range

the

value 1 is

statement

are

since

statement

of

met

is

the

used.

follow-

initial-

other

other

[

[

[

[

Example:

22

25 SUM = SUM +

54

12 L = Y(I)

the

In
unless
passed

The

value

of v during

e 1 + (i -

where i is

is

made

tion

during

Caution:

DO

IF
SIGMA
IF

CONTINUE

example

the

arithmetic

to

statement

of

the

anyone

the

out

of

which

1)

number

the

If

undefined

54

I = 1,

15

O(I)

(SUM.

LT.

0.0) SUM = 0.0

= SUM + R(I)

(SIGMA

that

* e

- H **

3/

begins

with

statement

IF

statement

12

at

the

end

variable v appearing

iteration

is

3

of

the

current

range

of a DO

the

the

transfer

entire

number

when

occurred.

program

T)

54,54,

12

22,

the

causes a transfer

of

the

fifteenth

in a

DO

statement

iteration,

before

of

iterations

all

iterations

control

and

passes

range

to

statement

iteration.

e1

and

have

specified

out

of

statements

depends

e3

have

been

for a
of

the

25

through

12.

If

a

II

on

the

number

the

meanings

completed,

DO

loop ~ executed,

DO

loop.

15

iterations

the

54

wi

II

are

of

iterations

discussed

value

of v will

the

be

executed

completed,

completed.

previously.

value

of v becomes

be

15

that

times,

control

The

If a

transfer

of

the

is

value

itera-

The

value

of

the

indexing

modified

by a

subprogram

parameters

called

within

(v,e1,e2'e

the

range

)

3

cannot

of

the

be

DO.

modified

within

the

range

of

the

DO,

nor

DO

can

they

Statement

be

23

A transfer out
only

occur

if

)

1,e2

,e

3

(v,e

of

the

range

of a DO

there

has been a prior transfer

are

changed

DO

GO

25 H =

TO 8605

outside

K,

Y,

loop

is

the

range

permissible

out

of

the

at

any

time; however, a transfer into

of

the

DO

range (assuming

DO). For example:

that

none of

the

range of a

the

indexing parameters

DO

may

1

[

24 A = H
25

8605 R = SIN(G(H» + JSU

8606

is

permissible; in fact,

11

32 T(J)

is

not val

id

because

A

DO

loop may

of

DO

loops,
fers to this
transfer
made.

Examples:

statement

is

made out

In

this

/8

JGU

= Y(H) ** 3

GO

TO

24

the

GO

TO

11

DO

32 J =

R(J) =

47.

= Q

no transfer could possibly

include

the

same

statement

can

of

the

case,

the

statements 8605 through 8606

2,36,2

E-7

* T(J)

occur

another

return transfer must be to

DO

loop.

may be used as the terminal

be made only from

range

of

the

innermost

Do

loops may be nested; however,

the

innermost

out

DO

the

are

considered

of

the

DO

loop

can

innermost

part

of

DO

range.

they

statement

loop.

a return transfer into

DO

for

Up

loop.

any

to 25

III egal

the

DO

range.

cannot

be

overlapped.

number of

DO

DO

ranges; however,

ranges may be nested.

the

range of nested

The

sequence

Only

DO

In

a nest

trans-

if

loops be

a

DO 1000
DO 100 J = 1,

~O

10

10

CONTINUE----.J

DO 100 L = 1,

~OlM=l'MMJ

A=B

100

1000

The terminal
ments 200 and 201 in

~ONTINUE

THIS

= DO

statement

I =

1,

II

------.

JJ-----.

K = 1, K K l

LL

----LJ

END~

of a range may not

the

illegal

example

above.

I

physically

precede

200
201

2000
20

2

the

D0200W=

D0200X=

DO 20 Y =

CONTINUE

DO 200 Z = 1,

D02U=l,

Q =

R-------+----

1,

1,

1,

UU-

WW---

XX-

yy--+

ZZ.-J

__

........

~ONTINUE-T

IT = WRONG-.J

DO

statement, as

is

shown

in

the

case

of

state-

24 DO Statement

CONTINUE

This

statement
CONTINUE

and

must

tion

and,

the

insertion

72
88

Statement

appear

consequently,

of a label

DO

IF
CONTINUE

H(33) = T(3,

is

72,

(X

written

in

that

1=

** I +

as

form. The

has no

at

any

point

1,20

O.

9999E-5)

R,

L,

E)

/22.5

CONTINUE

effect

on a running program. The purpose

in a program. For

72,

72,

88

statement

example:

does

not

cause

the

of

compiler
the

CONTINUE

to

generate

machine

statement

is

instruc-

to

allow

[

CONTINUE

PAUSE

PAUSE

where

Statement

statements

PAUSE

PAUSE

PAUSE

n is

's'

This

statement

puter

operator

beginning

If

an

integer

played

STOP

to

Statem

statements

n

's'

an

unsigned

is a

literal

causes
to

with

the

or a literal

the

computer

ent

are

most

are

written

integer

constant.

the

program

perform some

statement

constant

operator

often

used

as

constant

to

cease

specified

immediately

is

appended

when

the

as

the

terminal

of

up to

five

execution

action.

The

after

the

PAUSE.

to

the

program pauses;

statement

digits

(1 ~ n ~ 99999).

temporarily,

operator

PAUSE

statement,

otherwise,

presumably

can

then

of a DO

signal

the

word

the

word

range,

for

the

the

PAUSE

PAUSE

as in

purpose

program

and

is

displayed.

the

of

to

continue

this

example

allowing

value

above.

the

com-

execution,

will

be

dis-

STOP

statements

STOP

STOP

where

n is

an

This

statement
returned
before

to

termination.

n

unsigned

terminates

the

calling

are

written

integer

the

program.

in

the

form

constant

execution

If

an

of

a running program.

integer

is

appended

to

If

the

it

appears

STOP

within

a subprogram,

statement,

CONTINUE/PAUSE/STOP

it

will

be

output

control

immediately

Statements

is not

25

[

END

Statement

,A,n

END

ment

must

END

If program

of

a STOP

The

following

If

the

an

END

This

limitation
dicated
as

by

though

statement

appear

control

statement.

restriction

compiler

statement

is

the

following

they

were

is used

in

the

reaches

has

due

END

to

form

an

applies

encountered

and

will

to

an

historic

examples,

statements.

inform

END

to

act

the

FLAG

statement

any

statement

only

the

accordingly.

FORTRAN

certain

statements,

compiler

during

the

that

characters

An

feature;

that

execution

begins

END

END

namely,

although

it

has

reoched

of

with

the

at

the

statement

the

legitimate

the

physical

any

program (or subprogram)

character

end

may not

way

in

of a line,

string

appear

which

continuation

FORTRAN

it

end

of

END:

assumes

on a

continuation

statements,

a program. The

the

effect

that

the

statement

line.

is

specified.

will

be

processed

is

As

state-

that

is

in-

Processed

column:

Similarly,

6

x

x

illegal

as

EN D Statements

7

•••

END

FILE

2

END

END

statements

RATE

(I,

J,

of

= A * B

K)

=

.NOT.

the

same

Q

nature

as

those

in

Not

6

x

x

1 N

2 D

the

first

Processed

7

.....

END

FILE

2

END

RA

END (I, J
,

K) = .NOT.

E

column

as

will

END

Q

be

Statements

TE

treated

= A * B

as

END

statements.

26

END

Statement

6.

INPUT

jOUTPUT

The FORTRAN
sion

between

typewriters,

1.

READ
one

2.

FORMAT
editing

3.

Auxiliary

The

data
binary-coded
of

transmission

specificati

computer

card

and

WRITE

of

the

statements

information

I/O

transmitted

strings

the

on

statements.

Input/Output

An

input/output

of

their

transmission. These lists

where

m.

are I ist

I

language

units,

statements

group

of

statements

by

of

I/O

lists

list

represents

items

provides a series

storage

characters
statements

and

and

line

that

external

used in

that

conjunction

specifies

for

input/output

an

separated

external

printers.

devices

positioning

or
refer

ordered

have

by

of

statements

data

These

cause

specified

with

their

internal

and

statements

unformatted

to

external

group

the

form

commas,

that

handling

statements

lists

the

input/output

and

demarcation

are

transmitted

binary

values

devices,

of

data

names

as

shown.

determine

devices,

are

of

data

external

of

lists

that

the

such

as

of

three

types:

to

be

transmitted

of

formatted

representation

external

as

records

in

a form

similar

of

data

names,

identify

control

magnetic

files

(as on

of

the

of

and

tape

between

records

magnetic

sequential

to

internal

and

- for

data

to

condition

and

computer

to

provide

information

storage.

formatted

be

transmitted

for

paper

conversion

tapes)

data

transmis-

tape

handlers,

storage

consisting

For

either

data -to

and

the

and

and

of

type

format

order

List

Items

A list item may

A

single

datum

Examples:

A

MAT

RIX(25,

Multiple

1.

2.

data

An

array

the

array.

Example:

If B is a

where

j

DO-implied

and

and

B(l,

k

an

2-dimensional

expression

be

either a single

identifier

L)

identifiers

name

1), B(2, 1), B(3, 1),

are

items

is

are

appearing

the

dimension

are

lists

of

one

the

ALPHA(J,N)

in

array,

or

multiple

name

B

one

of

in a

list

the

•.. , B(l,2),

I imits of B

of

one

of

the

forms

datum

of a scalar

two forms:

without

---

list

item

B(2,2),

or

more

identifier.

variable

subscripts

B is

equivalent

.••

, B(j,

identifers

or

an

is

considered

to

k)

or

other

array

element.

equivalent

DO-implied

to

items

the

listing

followed

of

by

a comma

each

element

character

in

v

= e

1,e2

enclosed

in

parentheses.

Input/Output

27

The

elements

parentheses

: ...

,..J",,,:

... ,,, ...

'II,","""I"~

t'''--I'-oIII

Examples:

DO-impl

(X(I),
(A(I), I = 1,
«C(I,J),D(I,J),

with a

"'

..

'·

..

1=

v,

e1'

DO

·n",+", ..

""

.......

_.oJ

ied List

1,4)

10,2)

e2'

and e3

implication

c c

TI

-1'

J =

1,3),

have

are

nnrl c mn"

-2' _ .. --3

1=1,4)

the

same meaning as

considered to be in

nnn",nr

-1""'.--*

in

....... -

...

_,

Equival

X(l),

X(2), X(3), X

A(l),

A(3),

thic::

ent

Lists

A(5),A(7),

defined
the

rnnn'"

--"\::)-_ ... / --

for

range of

nnlv

(4)

A(9)
C(l,l),D(1,1),C(1,2),D(1,2),C(1,3),D(l,3)
C(2,

1), D(2, 1), C(2, 2),

C(3,

1), D(3, 1),

C(4,

1), D(4, 1),

Since

J is

the

C(3,2),
C(4,

innermost

D(2,2),

D(3,2),

2), D(4,
index,

the

DO

the

nc::

C::llhc::rrintc::_

------.r·-

2),

statement.

DO

implication.

...

C(2,

3), D(2, 3)

C(3,

3), D(3, 3)

C(4,

3),

D{4,

3)

it

varies more rapidl y

The items

For

input

than

enclosed

lists

I.

in

the

Special

2.

3.

4.

5.

Input/Output

All
integer

device

List

Considerations

1.

The ordering of a list is from

accompanied

when
An

unsubscripted

Constants may

For input lists
For

example,

(I,

J,

A (I), I = 1, J,

I,

J,

(A(I), I = 1, J,

As

an

output

(I,J,A(I),

The number

array

appear

the

as

an

list

1=

of

items in a single list is limited

Statements

input/output

variable

in

1.

The number may be assigned

statements specify a

reference

one

of two ways:

by

controlling

name in a list implies

in

input/output

DO-implying

input list

2)

2)

l,J,2)

whose

2. The number may be a standard
These standard assignments may

left

to

right with

DO-implied

repetition

indexing parameters.

the

lists

only

as subscripts

index

parameters (v, e

is not

allowed

is

allowed

is

allowed

only

device

value

to a device

unit

be

unit

number, u. This number may be

then

identifies

at

program run time.

number assignment, which is

overridden

by

of items enclosed in

entire

array.

, e2, e3) may not

1

by

the

statement

or

as indexing parameters.

length

parentheses

appear

within

specifications.

either

the

unit. This unit number corresponds to

run-time

recognized

assignments,

as referring to a

if

necessary.

(other than subscripts)

the

parentheses

an

integer

constant

an

actual

particular

as list items

or

an

physical

device.

Table 6 shows standard

access

random

devices.

device

assignments for FLAG. There

Unit Number

5,

i05

6,108

7,106

If

nonstandard
control
standard

28

Input/Output

cards

unit

unit

numbers
placed
number

Statements

are

used in a program

in front of

is

OUTIN (scratch mode).

the

FLAG control

Table

6.

they

card

If

Standard

must
(see

a unit is

are

no standard

Unit

Assignments

Standard Assignments

Card

reader

Line

printer

Card punch

be

assigned

Chapter

to

be

to

9). The

assigned

unit

assignments for

the

desired

default

to

an

device
function
input

device

by use

of

an

magnetic

of

assigned,

(e.

g.,

tapes

ASSIGN

non-

card

or

reader
ments

follow.

!ASSIGN

To

assign

!ASSIGN F :2,

To

assign

!ASSIGN

To

assign

!ASSIGN

or

magnetic

To

assign

F:l,

unit

number 2 to

unit

number 3 to a labeled

F:3,

unit

number 4 to a scratch

F:4,

tape}

the

IN

option

unit

number 1 to

(DEVICE, 51),

(DEVICE,

the I ine

LO,

(IN)

printer:

L)

magnetic

(LABEL, DATAFILE),

disc

(FILE, TEMP4)

should

the

(INSN,

file:

be

card

tape

PHS),

specified
reader:

with

serial

(IN)

on

number

the

ASSIGN

'PHS'

control

from

card.

which

data

Some

will

sample

be

read:

assign-

Formatted

Formatted

Input/Output

I/o

READ{u,

WRITE(u,

statements

f)k

f)k

Statements

are

used

to

process

binary-coded

(BCD)

records.

These

statements

have

the

forms

where

u is a

[

A

formatted

k

FORMAT
equivalent
values

into

Each

formatted

record

using

if

specifically

record

than

acters

are

A BCD

record
records.
which

case

been

specifically

as

magnetic

Sigma

Monitor

device

is

a FORMAT

is

an

READ

specified,

simple

character

unit

input/output

statement

into

list,

if k contains a DO-implication.

strings

input/output

more

than

one

requested

are

(or

can

be)

lost,

on

input

has a maximum

For

example, a punched

it

is

considered

written

or

paper

tape,

reference

number

statement

list,

{unsigned
label

which

causes

binary

values.

and

outputs

statement

READ

or

by

the

FORMAT

physically

they

are

treated

size

of
card

to

be

blank

is

indistinguishable

the

FORMAT

and

operations

the

character

begins

WRITE

present

132

contains
or

integer

or

an

unsubscripted

may

be

These

them.

processing

statement.

statement.

does

as

blanks.

characters.

empty.

(within

statement

manuals

or

integer

variable}.

array

omitted. A comma

string

in

the

external

are

then

assigned

Conversely, a formatted

with a new

More

than

However,

not

cause

Certain

80

characters.

In

other

the

may

determine

for a complete

attempting

processing

devices

A

words, a record

program)

description

name.

may

to

the

record.

one

record

from
the

optionally

record

variables

record

of a new

may

may

an

empty

actual

to

It

is

may

to

read

impose

contain

into

which

size

of

BCD

precede

be

converted,

appearing

WRITE

not

possible

be

processed

(or

write)

record;

other

as

any

record.

of

record

records).

the

according

in

the

statement

to

by

more

on

output

restrictions

few

as

zero

number

However,

written

list

k.

to

the

list

k,

or

the

converts

internal

process a particular

these

statements

characters

the

on

characters,

of

blanks

on

(see

extra

the

size

devices

the

on

char-

have

XDS

a

of

in

such

The

list k may

record

(on

more

than

information

IIH

Format

input)

one

is

Codes

be

or

record

either

ll

omitted

writing

used,

read

under

from a

into

IIFORMAT

Examples:

READ(105,

READ(5,

6)X,

FORM)

Y,

(A{I),

T(3, 5)

WRITE(N,FMT)(MASS(J,3),J=l,
WRITE(102,93)

MESAGE,

one

if

1=

formatted

blank

the

FORMAT

or

written

1,40),

ERR

record

from

Statements

H,Q
100,

NO

input/output

(on

output).

statement

the

locations

ll

}.

1)

statement.

However,

begins

with

in

Hollerith

storage

Normally,

information

or

occupied

this

has

may

slash

specifications,

by

the

the

effect

actually

be

FORMAT

Input/Output

of

skipping

processed,

in

which

statement

Statements

one

and/or

case

{see

29

Acceptable

The

following

sical

device,

FORTRAN

FORTRAN

as shown in

II

Statements

II

statements

Table

Table

are

accepted

7.

7.

FORTRAN II/FORTRAN IV

by FLAG. Each

of

these

Equivalent

statements

Statements

designates a specific

phy-

READ

Statement

This FORTRAN

READ

f, k

where

is a

k

is

The

READ

statement

according

list k,

PUNCH

This FORTRAN

to

or

the

Statement

FORTRAN
Statement

READ

f, k

PUNCH f, k

PRINT f, k

II

input

statement

an

input

list

causes

the

FORMAT

equivalent

II

output

II

statement

label

or

as

described

the

specified,

simple list

statement

FLAG
Equivalent

READ

(105, f)k

WRITE

(106, f)k

WRITE

(108,

has

the

form

an

array

name

of

the

FORMAT

earlier

character

into

if k contains a DO-implication.

has

stri ng in

binary

the

form

in

this

the

values

chapter

external

which

f)

k

statement

record

are

then

describing

to

be

assigned

Standard

Assignment
Card
Card
Line

read

from

to

reader
punch

Printer

the

devi

the

data

ce

105 and

variables

converted,

appearing

in

the

PUNCH

where

f is a
k

This

statement

statement,

Statement

PRINT

The form

PRINT f, k

where

f is a
k

The PRINT

FORMAT

Chapter).

f,

k

statement

is

an

output

causes

and

to

of

the

FORTRAN

statement

is

an

output

statement

statement,

be

label

list

described

internal

output

label

list as

causes

and

to

or

data

on

devi

II

PRINT

or

described

internal
be

an

array

earlier

to

ce

statement

an

array

output

name

be

converted

106.

name

earlier

data

on

of

in this

is

of

in this

to

be

device

the

FORMAT

chapter

into

character

the

FORMAT

chapter

converted

108 (see

statement

statement

into

also

strings, as

character

"Carriage

describing

specified

describing

strings, as

Control for

the

data

by

the

data

specified

Printer

the

applicable

Output"

by

FORMAT

the

in this

applicable

30

Input/Output

Statements

II

FORMA T -Free"

READ

and

PRINT

Statements

"FORMAT-free"

READ, k

PRINT, k

where

k is

an

for

"FORMAT-free"

to

be

read

by

actual

FORMAT

Intermediate

These
netic

Input/Output

statements

tapes,

discs,

READ(u) k

and

WRITE(u) k

where

forms

of

input/output

READ

one

READ

specifications

process

and

the

READ

variable

should

statement

used

Statements

information

drums.

and

list

be

the

by

They

PRINT

of

the

separated

values

should

"FORMAT-free

in

internal

have

the

statements

usual

form.

by

either a comma

be

punched 8 per

ll

(binary)

form

are

Output

READ

form

also

and

and

provided.

values

or

one

card

PRINT

are

designed

or

will

are

The

be

more

using

(8G)

to

general

forms

printed 8 per
blanks;
as

many

and

provide

if

more

cards

(8*

(2XG.6))

temporary

are

as

follows:

line;

input

than 8 values

as

necessary.

respectively).

storage

values

are

(The

on

mag-

u is a

k is

The

binary

size

of

the

of a binary

A

logical
programmer
format

of

ment

must

the

record.

no

limit

to

record

will

The

records
included
mation

produced

ment.

Other

produced

If

the

list

Unlike

formatted

written,

it

Examples:

device

an

input/output I ist,

READ/WRITE
items

in

READ/WRITE

record

is

concerned.

intermediate

be

input

If

the

the

number

be

read

produced

in

the

records

by

FORTRAN

on

other

k is

omitted

input/output

can

only

unit

statements

the

list k (see

may

consist

binary

by

one

input

list

of

or

written,

by

binary

to

an

intermediate

systems

machines

from a

be

processed

number

statement

of

(See

The

information.)

and

only

requests

items

that

regardless

WRITE

facilitate

or

by

binary

statements,

which

process

"Allocation

are

contained

several

SDS

one

more

can

statements

reading

binary

will

not

other

programs

READ/WRITE

by a READ

may

be

data

of

physical

Sigma

Monitor

This

READ

statement.

data

from a

be

processed

of

the

or

backspacing

WRITE

necessarily

no

data

statement

omitted
as a string

Variable

in

one

records;

(see

below)

of

Types",

logical

however,

binary

reference

means

that

the

It

is

permissible

binary

record

by a single

amount

do

not

statement

interpret

cannot,

of

data

consist

the

is

in

general,

to

of
proper

meant

the

statement, a record

transfer

with

can

no

occur
list

digits,

Chapter

record.

it

is

and

operations

information

than

is

READ/WRITE

be

transferred.

just

the
number

to

be

records

and,

be

input

is

skipped,

in

such

therefore,

in

arranged

7).

All

treated

manuals

output

to

read

less

present,

statement,

data

to

be

of

physical

read

subsequently

the

same

using a binary

or

an

operation.

into

words,

the

items

as a single

for a description

by a single

information

an

error

since

generated.

records.

way.

an

empty

has

little

depending

appearing

record,

binary

than

will

occur.

only

Control

Thus"..

by a binary

Similarly,

READ

statement.

record

If

an

empty

purpose.

as

WRITE

is

one

binary

is

written.

in

the

far

as

of

present
There

logical

words
the

READ

record

on

list

the

the

state-

infor-

state-

tapes

is

the

in

is

are

READ(3) E 1 (K),

READ

(N)

WRITE(MIN)R(J),

(M(K,

ARRAY

L), L =

G(J)

WRITE(3)VALUE

1,22)

Input/Output

Statements

31

END

and

Both

the

performed

ERR

Forms

formatted

if

an

error

of

the

and

intermediate

occurs

READ

or

an

Statements

binary

end-of-fi!e

READ

mark is

statements

read.

may

The

optionally
statements

include a specification
are

written

of

action

to

be

READ(u, f, END=s

READ(u,

where

sl

and

either

may

If

an

end-of-file

diately

NAME

The NAME
form

where

When
statement.

A NAME
to
since
to

NAME

an

be

placed

it

input.

to

statement

LIST

the

v.

array

LIST

enables

l'

END=Sl'

s2

appear

ERR=s2)k

are

each a statement

first.

mark is

sr

Statement

LIST

statement

LIST v l' v 2' v 3'

are

scalar

the

or

appears

name

user

I

name

statement

in

the

ERR=S2)k

encountered

If

an

error

is used

to

...

, v

array

identifiers.

in a NAME

with

no

identifiers

list;

i.

e.,

to

input

any

label.

during

occurs,

define

n

all

variable

Both

the

control

the

variables

Dummy

LIST

declaration,

following

nondummy

without

the

END=sl

processing

will

be

variables

it

causes

scalars

knowing

and

of

the

transferred

that

may

may

all

elements

all

and

arrays.

at

the

ERR=s2

READ

statement,

to

statement

be

processed

not

appear.

of

the

appropriate

This

compile

time

are

array

variables

can

optional;

control

s2'

by

INPUT

be

helpful

which

may

be

that

variables

if

both

will

be

transferred

statements.

processed

appear

during

program

it

will

are

in

present,

It has

by

an

the

checkout,

be

desirable

imme-

the

INPUT

program

The NAME
Each program has its own NAME LIST. This means

1.

2. If two

Examples

Simplified

This
FORMAT
can

process
editing
input
trolling

A

variable

for

example,

It

simply

LISTs.

NAME

NAME

NAME

is

the

or

in a

the

LIST

variables

that

appears

the

or

more programs
depends

of

NAME

LIST

LIST

T,

G,

LIST

NORM,

LIST

Input/Output

most

straightforward

statements

every

formatting.

very

input

or

type

free

form

operation.

defined

variable

on

which

statements:

I,

F,

RATE,

BOB, PHIL,

other
of

variable

Values

without

in

one

is in

have

form

kinds

may

program

COUNT,

the

in a program

program

COMMON

separate

DOUG

of
of
and

be

may

variables

is

doing

ITEM

input/output.

input/output

is

suitable

written

usual FORTRAN

unit

and

QUANTITY

out

the

not

the

which,

for

in a

are

be

aiso

with

inputting.

independent

following:

processed

appears

the

same

It

does

not

although

almost

any

natural

requirement

form

of

those

by

an

INPUT

in

the

other

name,

There is

require

FORTRAN

more

chosen

that

the
versatile,

the

defined

statement

program.

it

is

possible

no

;onflict

programmer

application

by

the

user

know

in

any

to

between

to

are

also

that

compiler.

exactly

other

in

another

input

the

learn

more

does

Similarly,

what

program

program unless,

into

either

separate

anything

complicated.

not

require

data

FORMAT

units.

of

them.

NAME

about

special
can

is

con-

It

be

32 NAME

LIST

Statement/Simplified

Input/Output

OUTPUT

Statement

The OUTPUT

OUTPUT k
OUTPUT, k
OUTPUT(u) k
OUTPUT(u), k

where

k is
u

The

name

pri

ate

For exampl

OUTPUT

might

X
Y =
X + Y =
SQRT(X**2 + Y**2) =

an

defines
unsigned
unit

of

each

format.

e,

produce

=.500000

1.

20000

statement

108,

X,

1.

may

have

any

of

the

output

list,

consisting

the

logical

integer

which

list

item

Y, X + Y, SQRT(X**2 + Y**2)

the

following

70000

unit

or

integer

is

the

will

lines

1.30000

number

variable.

standard

be

output,

of

output:

of

following

variables,

of

the

If

print

unit.

followed

forms:

expressions,

device

no

unit

by

an

on

which

number is

equal

and/or

the

specified,

sign

and

character

output

then

data

the

the

strings.

is

to

be

output

value

written.

will

automatically

of

the

It

may

item, in

an

be

any

be

appro-

on

Complex
the

which might

values

are

following

DOUBLE PRECISION D
COMPLEX C

LOGICAL

OUTPUT(6), I,

I =

R =
D =
C =

example:

L

result

79

4370.72

99301.

3922310385

(-56.2234,334.882)

L = T

Each

value

is

written

maximum

If

the

comma~er

When

Storage

width

of

first

item

in

the

word OUTPUT.

OUTPUT, (A +

an

unsubscripted

",

Chapter

DIMENSION
OUTPUT M, A

output

as

R,

D,

C,

in

these

lines:

on a separate

any

line

is 132

the

output

list

B),

X(I)

array

name

7). For

M(2,2), A (2)

example,

complex

L

line,

characters;

begins

Otherwise,

appears

constants;

beginning

excess

with a left

the

in

the

the

other

in

column

characters

parenthesis,

list

item

output

will

list,

data

types

2 (so

will

and

appear

the

that

be

no

entire

are

also

no

lost.

unit

to

be a unit

array

output

carriage

number

is

output

in

natural

control

is

number. For

will

specified,

in

storage

forms,

as

shown in

take

place).

there

must

be

exampl~

order

(see

"Array

The

a

OUTPUT

Statement

33

r

which

To

output

a

literal

as

in

could

M =

A

print

553

-4

o

11245

.472962

-33.0000

headings

constant).

the

following:

or

other

In

alphanumeric

this

case,

no

information,

equal

sign

or

value

a list item may

will be

~eneratedi

be a character

only

the

string

enclosed

character

string

in

quotes

itself

is

output,

(i.

e.,

OUTPUT

which

generates

FINAL COORDINATES

Implied
therefore

OUTPUT, (K, A(K)/B(K), K = N 1, N2)

which

could

K=3

A(K)/B(K) =

K

A(K)/B(K) =
etc.

Another
ment.
put on a

OUTPUT

The

actual

need

not

'FINAL

DO

loops may

preceded

produce

=4

feature

An

OUTPUT list item may consist

record.

(4)

format

be

concerned

COORDINATES'

the

line

be

used, as in

by

a comma.

this

output:

14.6135

15.0873

is

provided which

This wiii

X,

specifications

I,

cause

J,

AA, *

about

them,

the

enables

an

iNPUT

used to

since

following

OUTPUT records

of a single

statement

output

they

example.

asterisk

the

various types

are

provided

to

terminate

to

Note

that

be read

(*),

automatically.

which

reading.

of

data

subsequently

the

list begins with a

will

cause

For

exampie,

are

shown
Note

that

parenthesis

by a "NAMELIST" INPUT

the

characters

below,

all

although

the

*END* to

formats

the

programmer

are

widthless.

and

state-

be

is

out-

Data

Type
integer
real
double

precision
complex
double

complex

logical

INPUT

The INPUT
of

where

34

variabl

INPUT

INPUT(u)

u is

INPUT

Statement

statement

es. The forms

the

unit

number, as

Statement

Using NAME

Using

is

the

of

Format

lPG.6
lPG.15
lP,

lP,

L

NAME

LIST

counterpart

the

statement

described

LIST

lH(G.6,
lH(G.

Specifications

-X,

15,

-X,

of

the

OUTPUT

are

shown

under

the

lH,G.6,-X,

lH,

G.

statement,

below.

OUTPUT

lH)

15,

-X,

lH)

except

statement.

that

When

it

is

not

specified,

written

without a specific

it

is

assumed to be 105.

list

This form
identified
the

user
to

run.

run

of

to

the

by

the

decide

INPUT

input

at

run

statement

itself,

rather

time

which

is

designed

than
variables

to do

being

self-identified

named

(if any)

explicitly

are

to

be

input.

in

an

input

input, and

That is,

list

to

select

the

within

different

variables

the

program. This

being

input

input

enables

variables

are

from

statement

This

where

the

r.

are

or commas thay

are

of

the

v = c

first form,

v

ignored.

replacements

is a scalar

blanks

Each

In

the

c is a

The

second

a is
s.

I

c

form

is

is

the

an

processes records

each

a form

be

used to

separate

r. may

I

variable

constant

a subscript,

of

is

used to

identifier

an

input

of

which

appropriate

of

the

of

replacement

the

take

one

and

appropriate

into

an

an

array,

may be a signed

constant.

form

(similar to

r.. There may

I I

of

the

an

assignment

be

any

number

following forms:

type.

array

element.

or

Here,

unsigned

integer,

statement).

of

r. on a

and

Either

record.

semicolons

Except

within

(;),

as

constants,

indicated,

The third form

IIInput/Output

(see

a
c.

J

m

When

an
in which
of

constants as

the

separating

specifies

is

an

is a constant

is

equal

entire
the

array

there

comma

Example:

=

ALPHA

1.
4E-5
.87,

100000000000. ; X =

This

is

the

only

Variables
that
no

that

their

names

identifier

are

list

input

ll

Lists

).

In this

array

name,

of

an

to

the

number

array

is

specified

is stored (i.

are

elements

is

optional.

7302,

-67,

24.0983281640957
0;

case

in which a

to

be

processed by

can

be

recognized

after

it, which causes

into

an

entire

case,

appropriate

of

elements

as

above,

e.,

columnwise;

in

the

etc.

single

an

at

run

array,

in

the

same manner as

type,

and

in

the

array.

the

constants

see

array.

The

INPUT item (an r.) may

INPUT

statement

time.

Note

a"

permissible

c.

Figure J 2,

constants

I

must

that

this

scalars

are

assigned to

Chapter

may

overlap

have

can

and

described

7).

appear

from

been

be

done

arrays to

for

explicit

successive

Note

on

separate

one

referenced

that

record

array
there

input

by a NAME

by using a NAME

be

placed

in

input/output

elements
must

be

records, in which

to

another.

LIST

LIST

the

name list.

in

the

the

same number

statement

statement

lists

order

case

so

with

The permissible forms
constant

is to

be

of a constant

assigned.

that

appears

on

an

input

record

depend

on

the

INPUT

type

of

Statement

the

variable

Using NAME

to which

LIST

the

35

r

I

1.

For

integer,
specification.
widthless formats. This includes

constants
stants
within

2.

For

complex

where

complex

real,

(except

are

terminated

constants.

variables,

c 1

and

constant

and

The

constant

Hollerith).

at

If

this rule is followed, INPUT

c2 may

as

described

double

the

the

each

precision

may

therefore

all

Note,

first

blank

constant

be

any

in

Chapter

variables,

take

the

forms discussed in

howevei,

that

follows a

must

be

of

the

forms discussed

2.

any

that

replacements

of

the

the

of

since

digit

form

constant

the

forms

Chapter

they

or

decimal

above

is

scanned

described

2 for

are

scanned

may

be

for real

integer,

point.

written

data.

according

under

with

A good rule

to a G.O

"Numeric

real,

a widthless format,

exactly

The meaning

{width

less}

field

Input Strings" for

and

double

is

as assignment statements.

not

is

to

the

precision

these

con-

embed blanks

same as for a

3. For

Note

and arrays,

INPUT processing terminates when an asterisk
example,

or

I ogi cal

TRUE.

.

or

the

the

letter

Thus,

that

T =

55E-2;

J

= -3746E 02 i LGL(12,

CPX = (7. 32D-2,

0, 14,

R(55,-2) = 55.349384531062851907

J=9

SOOLE = T i FLAG = F ORTRAN4

variabl

or

constant

Tor F appears.

T,

F,

TRUE,

if

the

items in

the

resultant

A(l,

3.71*

es,

.

FALSE

may be

and

an
output

1,3) = 4,

I

3)

the

constant

•

any

character

Such a field is

FALSE

are

OUTPUT

records

DSL

-2)

= .

k'=()

,"

.....

, ,

statement

= 2

FALSE.

ARRAV=()

.........

may

string

all

permissible

can

,...

take

either

that

terminated

are

restricted

be

processed by

character

-,

of

the

forms

can

be processed by a widthless format;

by

the

input

strings for

an

is found in

first comma

to

scalars,

INPUT

the

or

nonleading

logical

statement.

variables.

array

elements

input string,

blank.

(with

wherever

that

is,

constant

that

may be. For

one

in which

subscripts),

C

= (2,

7.08364724286E-03)

ARRAY

*END*

The *END*
enables
ments. For exampl e

OUTPUT

generates

36 INPUT

=

1.

0

2.718281828459046

3.

141592653589793

-3944483

at

the

end

the

user

to

(4)

A,

records

Statement

that

Using NAME

of

the

separate

S,

C,

*,

can

second

OUTPUT records

ARRAY,

be

*, CPX, J(3), *

processed by

LIST

example

three

is

into

the

form

"files"

separate

of

end record

that

can

INPUT

generated

later

be processed

statements.

by

the

by

OUTPUT

one

or more INPUT

statement,

and

state-

FORMAT

Statement

The FORMAT

input/output

(BCD)

gram.

where

and

In

general,

ments

are

expressed as

FORMAT

n

~

0

is

S.

I

tions in

where

The commas
fications.
For

example,

123F27.13X

wi

II

be

interpreted

123

, F27.13 , X

not as

12

, 3F27.1 , 3X

To

obtain

the

statement

or

conversion

(Sl,S2,S3"",Sn)

either

a format

the

form

r(S

l'

S2'

m

~

o.

is a

S. is as

J

between

In

the

the

absence

the

specifications

as

latter

interpretation,

is

used

to

specify

DECODE/ENCODE

performed during

specification

S3'

•..

,S

m)

repeat

count

as

described

described

Si

(and Sj)

of

a comma,

above;

are

the

the

commas

the

conversion

operations.

of

one

in

other

optional

compiler

output

of

the

below.

words,

except

are

required.

to

It is

is

the

forms

repetitions

where

attaches

be

performed on

nonexecutable

reverse

described

of

may

ambiguities

as much

below

as

data

being

and

may

that

performed during

or a

repeated

be

nested (to

would

arise

possible

to

transmitted

be

placed

ten

from not

the

left-hand

anywhere

input.

group

levels).

during formatted

in

the

FORMAT

of

such

specifica-

separating

specification.

speci-

pro-

state-

Every FORMAT

ments. An
the

use

the

array

Format

of

data

an

input/output

secutive

where

entire

of

assignment

itself

specifications

values,

format

rFw.d rlw

rEw.d
rDw.d
rGw.d

the

characters

of

conversion,

is

an

is

omitted,

is

equivalent

statement

FORMAT (the

statements

is

referenced

describe

and

editing

list is processed

specifications.

rLw
rAw
rRw

F,

E,

data

optiona

316

16,16,16

its

I,

value

to

unsigned

should be

parentheses
or

by

the

the

to

be

by

Format

rZw

rMw

r1s

I

nHs

D,

G,

I,

generation,

is

assumed

labeled

input

input/output

kind

executed.

a sing

L,

integert that

so

statements.

or

type

Ie

format

specifications

iX
Tw
iP
r/

A,

R,

scal ing,

to

be

that

and

of

Each

Z, M,

indicates

1.

the

references

items

In this

statements.

conversion

integer,

specifi

cation,

may

H,

quotation

editing,

that

For

example,

may be made

they

enc

lose) may

case,

as

to

be

performed,

real,

double

whi Ie

be

any

of

the

mark

and FORMAT

the

specification

to

it

by formatted

be

stored in

described

complex

under

specific

precision,

data

following forms:

(I),

X,

control.

is

input/output

an

array

"FORMATs Stored in

data

to

be

or

T,

to

logical

are

P,

be

datum

operated

and slash

repeated

state-

variable

generated,

on by two

(/)

r times. When r

through
Arrays",

appearing

define

the

scaling

con-

in

type

tSee

also

"Adjustable

Format

Specifications".

FORMA T

Statement

37

r

I

w is

an

optional

braic

signs,

and

a

specification,

cOiiveis;on pSifoimed. This is

d for F,

n is

F

Format

Form:

rFw.d

Integer,

Double precision values

a

lIows for the

E,

D, and input G

tional

digits

appearing

sumed

to

be

zero,

are

equivalent.

For

output G specifications, d is

number

is

is a signed

(Fixed

real,

of

significant

an

unsigned, decimal

a string

of

integer

Decimal

double

appropriate

unsigned

blanks}
the

and

the

characters

t

{plus signs

Point)

preCISion,

are

converted

number

intege/

of

the

size

of

specifications,

in

the

the

decimal

digits

integer

or

of

digits

that

defines

external

the

external

discussed

magnitude portion

that

that

acceptable

are

either

with full precision

representation

field depends on

individually

is

an

point

character

also

an

unsigned

appear

in

the

defines

to

the

optional}. The function

part

of

complex

in

the

fractiona

width in

the

characters

of

the

data

the

characteristics

under

eCich

optional,

of

external

number

FLAG processor (see

if

I portion

unsigned

the

external

preceding

integer/

field;

of

characters

of

data

may

sufficient

of

it should not

therefore,

width

the

{including

being processed.

specification.

intege/

field.

but

in this

i is

described

be

processed by this form

is

field.

digits,

of

that

If

d is

not

appear.

context

its

value

being processed.

Chapter

1).

under X

specified

decimal

the

data

specifies

present, its

it

should not be

by w, and

points,

If

w is not present in

and

the

the

number

value

That is,

is

and P specifications.

used

of

the

Ew.O

to

conversion.

type of

and

define

zero.

value

alge-

of

is

of

frac-

as-

Ew

the

d

Output.
The field is right justified with as many leading blanks as necessary.
sign.

When no width

to express the

and

where 1) represents

If

a
lost wi
for

In order

where n

Internal values

Consequently,

273.4

-.003

-442.30416

349.5203
70000

-22

for

the

specification

.03359,

value

the

-1.22315

w

-67

requires more positions than

II

be from

specification

432034.
to

insure

~

d+2+n

is

the

7

is

speGified

value,

the

the

that

number

for
is
is
is
is

plus
is
is
is

is

left

F4.4,

is

is

are

converted

the

specification
converted
converted
converted
converted

converted
converted
converted

2F.4,

converted

character

or

converted
converted
such a loss of

of

digits

to

to
to
to

(i.e.,

w is not present),

one

blank

to

to
to

the

to .0336'O-67.000Ob
blank.

most

significant

to
to

to

the

to real constants, rounded

F11.

4,

273.4000

-,

r\nr\1"\

I.UVVV

-.0030

-442.3042

the

to

the

right of the

349

•

.%

70000.00

-22.Ob

output

list

are

2232
0000

digits

left of

allowed

does not

by

portion

of

occur,

the

decimal point.

the

the

converted

field.

Therefore, for

magnitude

field.

the

following

This

to d decimal

Negative

field

contains

of

w,

only w digits

is

not

treated

relation

places

with

values

are

only

the number of

the

specification

as

must hold true:

an

wi

error

an

overall

preceded

F.

i,

II

appear,

condition.

length

of

with a minus

digits

necessary

and

the

Thus,

w.

digits

also

II

t See

38 FORMAT

Adjustable Format

Statement

Specifications

n.

Input.
Input
point
the

If

the
of

is

Input
Strings".
is

present

value

will

33

802142

.34562

-7.001

the

width

first comma

the

next

333,

.003

converted

3.33

strings may

Each string

in

be

is
is
is
is

w is not

or

field

begins.

to

the

.003

take

the

input

explicitly

converted
converted
converted
converted

specified,

blank

that

va

lues

will

string,

defined

to

follows a

For

any

of

be

of

the

to
to

to

conversion

the

specification

E Format (Normalized, with E Exponent)

Form:

rEw.d

Integer,
precision

the

real,

va

lues

appropriate

double

are

number

preCISion,

converted

of

digits

or

with

in

the

the

integer,

length w with d characters

va lue

by

that

.033

802.142
.34562

-7.001

starts

digit

either
full

precision

fractiona I portion

real,

of

d is

decimal

with

or a decimal

2F.2,

part

of

complex
if

or

double

ignored,
point.

the

first
point.

the

string

sufficient

of

and

For

non-blank

data

width

the

precision

in

the

fractional

the

number

the

specification

The comma

may

be

processed

is

specified

fie

Id.

constant

of

character

or

blank

portion

digits

FlO.3,

in

by

by

wand

forms

discussed

of

the

in

the

the

input

is bypassed

this form

the

under

"Numeric

value.

fractiona I portion

string

of
value

If

and

ends

before

conversion

conversion.

of d allows

a

decimal

Double

of

with

for

Output.

where

Internal
right
E15.8:

When
ters
used,

is

where

Internal

.ddd

•.•

.ddd

•••

ddd

... d represents d digits,

lO±ee

values

justified

90.4450

-435739015.
.000375

the

width

necessary

the

output

-774.119,1.00001977

converted

-.77412En03n.l0000EhOlu

1)

represents

dE

ee

dE-ee

are

and

-1
.2

0.0

to

to

values

w is not

express

list

are

rounded

preceded

is

co

is

converted

is

converted

is

converted

is

converted

is

converted

present

the

the

character

converted

to d digits,

by

the

nverted

in

value

blank.

to

real

constants

while

ee

or

-ee

is

and

negative

appropriate

to
to -.43573902E
to
to
to
to

the

format

of

the

number

.90445000E

.37500000E-03

-. 1 OOOOOOOE
.20000000E

.OOOOOOOOE

specification,

data,

plus

of

the

interpreted

values

are

of

blanks.

02
09

01
00
00

the

one

blank

forms

as a

multiplier

preceded

The

converted

to

the

right

by a minus

following

field

of

the

of

the

sign.

are

examples

contains

field.

forms

only

If

The

external

for

the

number

the

specification

the

specification

of

field

charac-

is

2E.5

is

The

field,

counted

tude

digits,

the

from

decimal

the

point,

right,

includes

and

the

sign

the

of

exponent

the

value

digits,

(minus

the

sign

or

(minus

space).

or
If a

space),

width

the

letter

specification

FORMA T

E,

the

is

Statement

magni-

of

39

insufficient
the

left

or most

i=V',......,·u

..

'I.lor.

_"

.....

IIIt-'I

.....

"".

magnitude

significant

to

allow

portion

expression of

of

the

field.

an

entire

This

value,

is

not

only w digits

treated

as

an

will

error

appear.

condition.

The

digits

lost

are

from

EllA

-.2014E
.3619E 00
.1340E-03

of

To

Value

-2013.55
.361887
.000134

prevent

a loss

w ~ d+6

satisfied

is
which is

by

an

indicationof

60255.034

0.0000072

the

specification.

is
is

Input. The discussion

characters.

need

not

Conversion

have

exponents

Examples:

Input

Value

-113409E2

-409385E-03
849935E-02
6851

04

this kind,

magnitude
converted
converted

"Numeric

is

identical

specified.

Specification

E9.6
E.2
E10.5

E.O

E8A

2014E 04
3619E 00

~

1340E-03

it

is

necessary

Note

range for
to
to

-05

Input Strings

to F format

to ensure

that

the

05

ll

Converted

-11.340900

-4.09385
.0849935

6851.0

E6A

14E
19E

40E-03

above

any

datum. For

containsa

conversion.

to

04
00

that

the

feature

description

In

relation

can

be

example,

of

particular,

used

for

the

intentiona

the

forms
input

Ily to

specification

permi~sible

fields

for conversion in E format

obtain

the

E3.0,

for strings of

exponent

input

field,

First,

the
determine
is

interpreted

decimal

the

point

is

positioned

actual

position

as -.113409E02;

of

the decimal

which,

o Format (Normalized, with 0 Exponent)

Form:

rDw.d

similar

This format is
character

E.

for E12.6,

and

for D

12.6,

Input under D format

G Format (General)

Form:

rGw.d

G format is

the
performs depends on the
format

that

is

to E format, with

For

example,

-667.334

-667.334

only

is

is

is

the

format

type

converted

converted

same as for E

that

may

of

the

used when processing list items

according

to

the

point.

when

evaluated

the

exception

to

-.667334E

to

-.667334D

and

F formats.

be

used

with

any

list items. For a

of

the

specification;

In

the

first

(i.e.,

-.113409 x 102),

that

for

output,

03

03

type

of

data,

Gw.d

specification,

various types.

then,

example,

the

including

the

value

of

the

-113409E2 with a
becomes

character

logical.

the

following

-11.340900.

D will

The form of conversion

exponent

specification

be

present

table

shows

is

applied

instead

the

equivalent

of

to

E11.6

of

the

it

40

FORMAT

Statement

List Item Type

Input

Output

integer

real

double

precision

logical

Note

that,

as

parts

require

For

integer

and

the

only

is

case

When so used,

Output

of

Real
precision
way;
small.

number
ma I point

appended.

not
To

its).

values

that

is,

Specifica

of

digits

anywhere

If,

be

used;

express this

Select

i-1

lO

:s

with

all

individual

logical

in

which

the

equivalent

and

Double

depends

they

are

Ily, d is

that

wi

within

however,

instead

the

algebraically,

an

integer

M < 10i

Iw Iw
Fw.d
Fw.d
Lw

other

formats,

specifications,

list items,

d is not assumed

formats shown

Precision

on

the

magnitude

expressed

in

interpreted

II

be

output.

or

at

either

preceding

number will

let M represent

i such

that

(if

M =

0.0,

(see

below)
below)

(see
Lw

complex

the

d (in

Data

F format

as

If

or

trailing

be

then

values

are

and

conversion

Gw.d)

need

to

be

zero

if

above

become

Under G Format. The form

of

the

values.

whenever
indicating
the

va

lue

end

of

those d

normalized

the

i =

O)

the

of

zeros

and

magnitude

possible,

the

digits,

would

processed

occurs

not

be

not

specified.
width

G format

number

number

be

output

of

as

as

shown

specified;

less

but

in E format for

of

significant

is

such

that

is

required

with

a fol.lowing

the

value

two

separate

above

if

G format is

also

(see

of

output

attempts

that

what

wi

to

express

to

for

real

it

is

present,

"Numeric

conversion

to

express

values

digits

desired,

it

can

II

be

done,

the

exponent.

be

output

items;

the

or

double

it

very

useful in a

Input Strings ").

used

numbers in

that

be

expressed

and

va lue

(rounded to d

real

and

precision

wi

II

be

widthless

with

real

the

are

too

and

this is

by

no

exponent

correctly,

imaginary

data.

ignored.

and

double

most

large

exactly

placing

wi

F format

significant

This

form.

natural

or

too
the

the

deci-

II

be

will

dig-

Assuming a
ing

to

specification

the

specification

Fn.m,4X

If

i is less

than

Ew.d

Note

that

when

would

otherwise

a more

readable

Value

.02639
.2639

2.639

26.39

263.9

2639.

Note

that

the

are

lost

at

the

should hold true:

w

~

d+6

of

Gw.d,

0 or

greater

than

F format is used, four

be.

In this

way,

way.

The following

G

10.

3

.264E-Ol
.264

2.64

26.4

264.
04

.264E

choice

of

F or E format is

left

as in

other

numeric

let

d,

the

numbers

G

.3E-Ol
.3

3.
.3E 02
.3E

.3E

n =

w-4

specification

blanks

are

that

examples

10.1

03

04

independent

conversions.

and

m =

used is

output

are

output

illustrate

of

To

d-i.

Then,

following

in columns

the

effect

the

value

ensure

if

O:s i

the

number, in

will

of G format

of

the

that

this will

:s

tend

width

d,

conversion

the

to

line

output

w.

not

positions

up

on

If

w is not

happen,

takes

place

where

underneath

values

of

large

the

following

accord-

an

exponent

each

various

enough,

other

sizes:

digits

relation

in

When no

followed

Scale
This has
that

factors

the

values

width
by

four

effect

output

is

blanks.

(see

specified,

lip

Specifications

that

all

in F form

values

the

with

number

")

apply

output

a P

scale

will

be

followed

to

G format

in G format

factor

are

cannot

by a

single

only

when

unchanged

subsequently

blank;

the

E form is

(except

be

values

for

input

output

used,

rounding).

using

the

in F form will

not

w}1en

the

It

also

same FORMAT;

FORMA T

not

be

F form is used.

has

the

effect

the

scale

Statement

41

r

factor

will

take

a power

Note

rounding

In
but
value

giving

have

I Format {Integer}

Form:

of

10.

that

the

rounding

that

is

PRINT

~

FORMAT(lP,

principle,

when rounded

been

F form

of

M is

normalized

9.98E 01, which is

1.00E 02, which is less

rIw

Integer,
width
In
truncation

real,

double

specification w is

other

words,

of

the

effect

appl

ied when

5,99.76

G.2)

is

to

two

precision,

values

most signifi

during

applied

to

be

used

digits

(.9976E 02),

the

of

sufficient

greater

cant

input

but

to M {above}

the

number

if

the

it

becomes

way

it

is

accurate.

or

either

magnitude,

than

the

digits

not

during

output.

to

determine

is

actually

value

100.,

then

the P scale

printed.

part

of

maximum permissible

that

is

lies in

which

If

the

complex

real

normally

output.

the

range

is

factor

first rounding had

and

associated

Thus,

whether

Consider

outside

is

data

double

size

the

new

to

use E or F format

the

.1

:5

M < 100. The

the

range;

applied

may

be

processed by this form of conversion.

precision

of

integer

with

integer

value

will

be

following

so E form

(9.976E 01), and

been

values

data

case:

value

is

used throughout,

are

converted

may

be

operations.

different

is

not

necessarily

99.76

does

used. First

finally

the

processed,

from

lie

the

this

final

in full

without

the

old by

the

in

this

unrounded

value

value

precision.

same

range,

is

rounded

would

If

the

the

Output.
values;
a minus sign,
tion

The
specified.

is

converted

where

If

the
in
for

Input. External
be
As

processed by

Internal

however,

16

impl ies

273.4

-.003

-44204.965

converted

345.9,

345b700000-2b~30300

-IS

represents

magnitude

the

external

the

specification

-778801

identical

noted

above,

4570000000000000000000000.942

4.57

x 102

values

are

the

integers

and

the

field

wi

that

is

converted

7 is

converted

is

converted

is

converted

field

occupies

This minimum

70000,

to

of

string, and

input

to F format processing, with

however,

an

I (widthless) format,

numberofdigits

-2,

-.999,

the

character

data

requires more positions

the

12,

is

converted

strings may

the

4

converted

may

II

to
to
to 0
to

the

3030.3030

digits

to

most

to

contain

be

right-justified

273

-44204

minimum number

blank.

lost

take

any

significant

into

integer

as many

7

is

followed by

are

the

01

of

the

the

exception

digits

a real

constants.

digits

and

of

positions required to express

than

is

most

forms discussed under

will not

or

double

Real and

as

are

specified

preceded

one

permitted

significant.

thatfractional

by

blank.

by

be

truncated.

precision

the

Forexample,

the

This is not

variable,

double

by w.

appropriate

value

"Numeric

portions

precision

Negative

number

the

for

the

of

the

treated

Input Strings"

of

a val

For

example,

wou

Id

data

are

values

of

blanks.

data

value

specification

width

w,

as

an

error

ue

are

lost through

the

produce

the

truncated

are

preceded

The

whenever w is

51

the

only w

digits

condition.

and

conversion will

truncation.

input

field

internal

to

integer

by

specifica-

un-

output

list

appear

Thus,

value

L Format

Form:

Only

42

FORMA T

rLw

logical

(Logical)

data

Statement

may

be

processed with this form

of

conversion; any

other

data

type

causes an

error

to

occur.

Output.

The T

and F characters

.TRUE.

.FALSE.

where

15

Logical

represents

values

is

converted

is

converted

the

are

converted

are

preceded

character

to
to

bl

by

ank.

to

w-l

either

blanks.

a T

or

For

an F character

examples,

for

using

the

values

the

specification

II

true

II

and

L4,

"false",

respectively.

Specifications

.TRUE

• FALSE. is

Input.
"true"
field

If

a

or

"falsel~

has

the
for commas,
fields,

processed

True

T

TRUE

.

TRUE.
RIGHT
STAFF
24T+T42

For

widthless

first

non-blank

terminate
appears,

comma

the

are

the

ignored.

in

•

width

value

which

logical

character

field.

field

which w is

is

converted
converted

is

specified,

respectively.

"false".

terminate

by

an

L7

format,

False

F

FALSE

•

FALSE

•

READ

LEFT

(blank)

input,

is

The first T

has

the

value

undefined

to

to

the

first T

If

no

Tor

Characters

the

input

have

the

field

terminates

a comma,

or F encountered

IIfalse".

will

cause

or F encountered

F is found

appearing

string

(see

the

indicated

at

it

terminates

As

above,

the

following

before

between

IIComma

values:

the

first comma

the

field;

within

characters

conversions:

in

the

next w characters

the

end

of

the

the

Tor F and

Field-Termination

or

non-leading

if

it

is

not

the

field

determines

appearing

field,

the

the

end

II).

a comma,

the

between

determines

value

is

"false".

of

the

field

For

example,

blank.

the

~alue.

the

next

If

first

In

blank

neither

Tor F and

whether

are
the

other

the

Thus a

ignored,

following

words,

or comma

a T

nor

the

value

blank

except

input

if

the

wi

an

F

blank

is

II

or

A Format (Alphanumeric)

Form:

rAw

Output.
mal
The number
item

ples

Internal

digits)

per

of

(see IIStorage

illustrate

Data

Type

integer,

or

logical

double

precision

binary

character.

characters

Allocation

the

form

real,

values

are

converted

The most

produced

significant

by

an

Statements

of

A format

conversion:

Interna I Binary/Hexadecima

1100

1001 1101 0101

item

ll

,

Chapter

to

character

digits

depends

7).

1110

are

on

Assuming

I

0011 0101

C 9 D 5 E 3 5 C

1100

0100 1101 0110

1110

0100

C 4 D 6 E 4 C 2

1101 0011

1100

0101 0111 1011 1111 0010

D 3 C 5 7 B F 2

strings

converted

the

number

1100

at

the

first.

of

words

standard

1100

0010

rate

That

size

of

eight

binary

is,

conversion

of

storage

allocated

specifications,

Aw

A4
A2
A6
A

A8
A6
All
A

Externa

INTi,

IN

1mINT~',

INT*

DOUBLE=2
DOUBLE

1m1'5DOUBLE=2

DOUBLE=2

digits

is from

the

I S tri

(two

hexadeci-

left

to

for

that

following

ng

right.

type

exam-

of

where

1)

represents

the

character

blank.

FORMA T

Statement

43

As

with

all

In

each

of

the

data

fully,
integer

with
to

express

is

not

preceded

When

to

fully express

AI

phanumeric

lowing ways:

1.

the

treated

by

the

field

Previously

other

format conversions,

examples

and

therefore

variabies.

data

value

as

an

error

the

appropriate

width

is

the

character

conversions

input

using

above,

In

not

the

has

the

the

exampies,

completely,

condition.

number

specified,

value

are

normally

an

alphanumeric

complex

first A format

same

effect

note

the

When w has a

of

bl

ank

the

of

the

used

data

as

the

that

external

characters.

external
data.

to

output

format

are

treated

specifies

widthless

when

field

value

character

The

external

Hollerith

(A

or

either

exactly

form.

the

magnitude

is

shortened

greater

string consists

character

information

R)

as two

the

Normally,

from

than

real

number

necessary,

of

w does

the

string

right

of

that

of

only

or

as two

characters

alphanumeric

not

provide

(least

significant)

the

external

the

number

is

not

followed by a

has

been

double

created

precision

required

information is used

for

to express

enough

portion.

character

of

positions

blank.

in

one

values.

positions

string

necessary

of

the

the

This
is

fol-

2. Using a I

It

is

not

The

reason

non-printable

Input. When
ters

associated

For

example,

ABCDEFGHIJ is

However, when

the

most

filled

with

iteral

recommended

for this

characters

the

with

if

the

the

significant

blanks.

UVWXYZ

where

-6

represents

Naturally,

the

Widthless
number
an

integer

cessing would

is

converted

if

the

list item

ABCDEFGHIJKLMNOP

ABCD EFGH

is

completely

specifications

of

characters

variable,

be

to

constant

that

this form

is

that

not

may,

width w is

the

data

list item

converted

value

of w is
positions
Consequently,

is

converted

the

character

width

has a

filled

cause

associated

and a double

in

the

following manner:

IJKLMNOP

(i. e., in a DATA

of

conversion

all

the

256 possible

however,

larger

type
is

integer,

of

the

bl

value

with

the!

with

than

of

the

to

less

than

list item

when

to

ank.

equal

the

1st

item

the

precision

be

useful in

necessary

corresponding

and

GHIJ

the

the

UVWXYZb'1:>

to
external

to

data

statement,

be

characters

(that

the

specification

number

are

filled

list

item

the

number

field.

be

fi

II

type

of

variable,

or

passed as

used with random

that

can

other

contexts

is, when its

list

item),

the

A 10

is

of

characters

with w characters,

is

double

of

ed by

the

in

precision

characters

the

next n characters

list item.

that

order,

an

argument)

numeric

be

produced

(e.

g.,

on cards, or in ENCODE

magnitude

list item is

used,

associated

and

and

associated

If

a list

and

a field

values

can

is

greater

filled

with

the

the

remainder

the

field

with

from

contained

specification

created

actually

than

the

with

the

data

type
of

specification

the

data

the

input

references

of

other

than

be

printed.

operations).

number

rightmost

of

the

the

positions

is

type

of

string,

to

a real

3A

were

as

above.
The

of

charac-

characters.

list item,

are

A6,

the

list item,

where n is

variable,

used,

the

pro-

A

general

fields

rule

are

considered

for this

type

of

to

be

right-justified.

conversion

R Format (Alphanumeric, Right-Justified)

Form:

rRw

This form
internal

justified

Output.

acters

format

44

of

values

with

When

from

output.

FORMA T

conversion

are

considered

trailing

the

the

rightmost

This

Statement

is

similar

Hollerith

size

of w is

(least

difference

to A conversion,

to

be

right-justified

blanks.

insufficient

significant)

is

illustrated

is

that

to

portion

in

internal

allow

the

values

but

the

with

leading

expression

of

the

examples

rule

of

of

internal

at

the

are

considered

internal

binary

the

value.

top

justification

zeros,

complete

In

all

of

the

to

be

left-justified,

whereas

internal

other

following

is

reversed.

with A format

value,

R format

respects

page.

it

is

while

external

In

other

they

are

takes

identical

words,

left-

char-

to

A

Data Type

Internal
Character

Va lue

w

A Format R Format

integer,

or

logical

double

where

1) represents

Input.

As
characters
most) portion

remaining
A format

list
Data

left

Item

Type

integer,

or

logical

real,

precision

on

output,

associated

of

the

portion.

justifies

real,

DOUBLE=2

the

bl

ank

character.

R format

with

the

differs

type

I ist item with w

In

other

words, R format

the

characters

Externa I String

XYINT*

from A format

of

the

input

characters

right

and

inserts

4

2

6

none

8

6

10

none

I ist

from

justifies

trailing

w

4

6

2

none

INT*
IN

f>1>INT*

INTi~

DOUBLE=2
DOUBLE
1>f:>DOUBLE=2

DOUBLE=2

only

when

the

specified

item.

In this

case,

R format

the

input

string,

preceded

the

characters

Hollerith

blanks.

Interna I after
A

Conversion

XYIN

INT"'~

XYf1I)
XYIN XYIN

INTi~

T*

f1DINT~~

INTi~

DOUBLE=2
UBLE=2

1')"oDOUBLE=2

DOUBLE=2

width

(w)

fi

lis

the

by

enough

and

inserts leading

For

example,

Interna I after

R

Conversion

XYIN

INT~'~

zzXY

is less than

least

significant

binary

zeros

binary

the

number of

to

zeros,

(right-

fi

II

whi

the

Ie

double

precision

where

1>

z represents

Note

that

external

represents

the

field

the
eight

Hollerith

OOABAB

were

processed

Hollerith

by

constants

the

4HOOAB

mat should be used.

Z Format (Hexadecimal)

Form:

rZw

Z

conversion

the

external

is

similar

field

consists

85DOUBLE=2

Hollerith

binary

character

format

character

zeros.

zero

specifications

and

2RAB,

to R conversion,

of

hexadecimal

is not

which

except

digits,

8

6

10

none

blank

and

represented

A4,R2

into

are

not

that

which

85DOUBLE
85DOUBfrl)
DOUBLE=2

85DOUBLE

interna

two

integer

equivalent.

the

internal
are:

lIy as

variables,

For

data

85DOUBLE
zz85DOUB

DOUBLE=2

85DOUBLE

eight

binary

the

resulting

input

as

true

is

processed 4 bits

zeros.

Consequently,

values

right-justified

at a time

if

the

would

be

the

integers, R for-

instead

of

8,

and

o

123456789ABCDEF

FORMA T

Statement

45

Output.

ber

'of

"Storage

If

w

is

number

Wher:l

field

Internal

characters

Allocation

not

specified

of

positions

width

binary

values

produced

Statements",

large

necessary

permits,

are

by

an

enough,

to

all

of

converted

item

depends

Chapter

the

express

the

digits

to

hexadecimal

on

the

7). For exampl

leftmost

the

data,

in

digits

an

the

item

number

e,

an

are

lost, as in

digits

are

output,

digit

strings

of

words

integer

are

right-justified

including

at

the

rate

of

storage

allocated

produces 8 di gits, a

other

numeric

in

leading

formats.

the

zeros.

of

4 bits

double

field,

for

If

with

per

digit.

that

type

precision

w

is

larger

preceding

The num-

of

item (see

number, 16.

than

the

blanks.

When w is

blank

is

Examples:

Data

integer,
real,

logical

double

precision

wheref>

Input.

When

ciated

with

not

specified,

to

facilitate

Type

or

subsequent

Internal

0000 0000

a a a 8 E 3 5 C

0100

4 1 3 2 4 3 F 6

1010

A 8 8 8 5 A 3 a

represents a blank

the

width w

the

data

type

of

the

full number

of

rereading

Binary/Hexadecimal

0000

1000

0001

0011

0010

1000

1000 1000

character.

is

larger

than

necessary

the

corresponding I ist item),

digits

of

the

necessary

value

1110

0100

0101

(i.

0011

0011

1010

e.,

to express

(see

below).

0101

1111

0011

when its
the

I ist item

the

1100

0110

0000

magnitude

is

fi

II

ed

value

is

greater

with

is

Zw

z8

Z6

Z10

Z

Z16

Zl1

z18

Z

the

output,

followed

Externa I Stri ng

0008E35C

08E35C

f>f>0008E35c

0008E35C

413243F6A8885A30

3F6A8885A30

f>n413243F6A8885A30

413243F6A8885A30

than

the

number

rightmost

characters

by a blank.

of

digits

in

the

The

asso-

fi

eld.

When w
corresponds

There is,
al
a widthless Z
string
at

The following

is

too

however, a significant

phanumeri

terminates

the

left.

External Input

small,

the

digits

to

the

number

c.

Therefore, commas may

input

at

(Note

are

examples

string
the

that,

of

is

first

when w

Field

3A70049B
f:>f:>b3A7iJ'b
1:>D68,470 19

DCBA987654321

52CA91,

123456789 ABC,

1:>1)4%1>3

where

15

represents a blank.

M Format (Machine Dependent)

M format
format most

is

intended

appropriate

primarily

to

are

right-justified

digits

associated

difference

not

dependent

~mma

of

or

is

specified,

Z format

Format

Z5

Z8
Z8
Z12

Z

Z

Z

for

output.

the

machine

with

between Z and

be

used

to

on

the

non-leading

blanks

input

(assuming

Internal

0003A700
0003A700
00000D68
98765432
0052CA91
56789ABC
00000049

It

provides a machine-independent

on

which

in

the

I ist item, as

the

list

item,

wi

th R format.

the

item

R format on
terminate a hexadecimal
size

associated

blank,

are

treated

an

Hexadecimal

the

program

like

all

integer

with

other

as

zeros.)

list

Value

is

running. Thus, on

is

input.

the

list

numeric

item):

As

completely

Z format

input

string.

item;

formats. Excess

method

usual, when

filled

is

the

with

the

a numeric format,

Furthermore,

a widthless

hexadecimal

digits

of

dumping information in

an

octal

machine

width

external

the

will

it

exactly

not

length

input

be

would

field.

of

lost

the

46

FORMAT

Statement

be

interpreted

the

same as Z

as 0 format, and on a

(hexadecimal)

H Format (Hollerith)

Form:

nHs

where

n

:s

255

character

format. Thus,

machine,

it

could

also

as A

be

format.

used for

On

input,

the

Sigma

though this is

5/7,

it

is

interpreted

not

recommended.

exactly

Output.

The n

Specification

IRE
8R'b1>VALUE:
5R$3.95

9HX(2,5)'b=1)

where1S
Care

be

Input. The n

~s

represents

should

created,

as shown in

Specification

3H123
1

OHNOWf>IS'bTHE
5HTRUEf>
6H'b1>f>f>f>f>

where1S

that

If

n is

are

not

represents
to

present,

characters

be

taken

and

so

characters

appear

the

that

the

the

on

its

that

an

in

the

External

E

'b1WALUE:

$3.95

X(2,5)1>='b

character

the

characters

in

the

following

Input String

ABC
1:>T
FALSE

RANDOM

character

output

value

is assumed

string s

String

blank.

character

string s

string s contains

from

other

are

examples:

IMEf>FOR1>

blank.

record

This

generated

to

are

transmitted

format

replaced

feature

be

1.

to

the

external

exactly n characters,

specifi

cations

by

the

next n characters

Resultant

Specification

3HABC
lOHf>TlMEf>FORf>
5HFALSE
6HRANDOM

can

be

used

to

by

the H specification.

record.

are

not used as

change

from

the

For

so

that

the

titles,

instance,

the

part

of

input

dates,

desired

the

string.

record.

column

external

This

replacement

headings,

field

will

etc.,

, Format (Hollerith)

This is

an

alternate

requiring

Form:

The
ever,
specifi

Output.

is

Within

string

note

cation.

output

ABLE

the

characters

s may

the

The string s

as

the

BODIED

a I

string

restrictions

format for

in

contain

stri ng

not more

below

is

transmitted

the I character

Hollerith

the

string

transmission

to

be

than

255

concerning

to

the

external

is

represented

similar

counted.

characters.

the I character. A repeat

device

by

two

to

that

Any

Hollerith

in a

manner

adjacent I characters;

done

by H format. This

characters

count,

similar

may

r, may

to

that

thus,

has

the

advantage

appear

for H format. Thus,

Ip

(see

optionally

ILL

TAKE

FORMAT

Chapter
precede

FIVE

Statement

is

output

1);
the

of

how-

not

as

47

r

Inp·ut. The

quentiall

is

the

y from

used

to

MATRIX

specification

'MATRIX'

characters

process

the

the

itself

input

input

is

appearing

string.

field

changed

between

Therefore,

to

the

quotes

if

the

are

specifi

replaced
cation

by

the

same

number

of

characters

taken

se-

Blanks in FORMAT

X Specification (Skip; Space or

The form

This

If i is

lar

of a record

Output.

below}.

of
iX
specification

positive,

to

that

of

is

For

In

'WXYZ' ,

generate

where b represents

_A~

to

are

the

WXYZfr&Of>IJKL

negative

the

equivalent

value

left,

assuming

'FORTRAN' , -3X,

'FORTKNOX'

statements

the X specification

causes

no

it

has

an

effect

the

backspace

equivalent

positive

other

words, a

4X,

following

of i causes

to

the

to

values

'IJKL'

external

the

character

that

this

specification

are

significant

Backspace)

is

conversions

similar

control

backspacing ~ the

field

'KNOX'

on a typewriter.

of

i,

the

of i blanks

string:

blank.

processing to "beck up" in the record. The next

is

not

beyond

only

in

to

occur.

to

that

next i positions

will

of a space

beginning

be

the

beginning

Instead,

created.

Hand'

bar

In

particular,

of

in

the

specifications.

it

causes i positions

on a typewriter;

an

the

record.

output

record

For

example,

of

the

record.

attempt

if

to

will

the

specifications

field

For

example,

of

the

it

is

negative,

backspace

be

blanks

vIi!!

external

then begin I i I

the

field

it

has

beyond

---

(normally,

specifications

to

be

"skipped

an

effect

the

beginning

however,

ChOicctCiS

".

simi-

see

48

Note

that
previously
output,
characters

The
produced

K

generate

As

equivalent:

it

ability

= 13

and

PRINT 5, K ,

5 FORMAT(

Q (13) =

illustrated

XXXX

4X

FORMA T

when

either

produced

is

not

necessarily

have

been

to

specify a negative

at

the

end

Q(13) + 350.8,

'Q('

the

string

350.

80

in

the

Statement

backing

in

the

generated

of

external

Q(K)

above

up

positions

true

that

in

count

fields

the

statements

example,

or

moving forward by

being

skipped

the

specification

those

positions,

in

an X specification

by

widthless

-X

if

i is

') F.2)

not

specified

means

are

not

4X

since

numeric

of

destroyed.

will

produce

all

output

makes

formats (i.

it

is assumed

an X specification,

Thus, in

four

records

it

possible
e.,

to

be

blanks.

are

D,

1.

the

initially

to

backspace

E,

F, G,

Thus,

characters

exampl e given

It

will,

however,

set

to

blanks.

over

and

I). For

the

following

that

may

above

under

if

the

blank

example,

specifications

have

no

that

been
X

other

is

for

are

Input.

example,

and

the

will

The

F5.3,

the

input

76.4lIGNORE697

characters

IGNORE

not

be

next i characters

with

the

specifi

6X,

13

stri ng

processed.

from

cations

the

input

string

are

ignored

whenever

i is

positive

(that

is,

they

are

skipped).

For

Negative
tions

and

are

and

T Specification (Tab)

The form

This

less

X

It
specification

when

three

val

ues

13,

-lX,

the

string

123456

equiva

lent

13,

E4.1

the

stri ng

1233456

of

the T specification
Tw
specification

of

the

position

specification;

1 FORMAT( 5X

FORMAT(

2

can

be

seen

width
columns

3 FORMA

provides a capability

less formats

of i cause

E4.l

to

no

from

of

numbers

T(G.

causes

in

transfer

T6

the

7,

T21,

iii

processing

the

record

AS

AS

above

are

being

beginning

G.

characters

is

that

of

data

-2X
Tl2 , 17

example

7, T 41,

from

(either

was

being

occurs.

17

that

that

an X specification

used

and

in

positions

G.

7)

the

input

For

it

the

input

or

output)

processed

example,

is

permissible

character

1,

21,

string

before

position

and

to

be

processed

to

begin

at

the T specification.

the

following

to

tab

either

does

not,

is

thus

41,

the

following

character

FORMATs

forward

namely

unknown.

again.

position

or

that

of

FORMAT

Consequentl

w in

It

functions

are

equivalent:

backward.
tabbing

For

example,

statement

y,

the

the

record,

exactly

Furthermore,

to a given

to

could

specifica-

print

print

be

regard-

like

position

(or

read)

used:

an

a T

Note

that

backward

be

processed

As

with X specifications,

If

no

w is

P Specification

The form

again.

given,

of

the P specification

it

(Scale

tabbing

it

is assumed

Factor

iP

A P

specification

of

the

forms

i

10

for

and

-i

10 for

output

input

causes

the

can

is

not
to

or

Power

value

cause

possible

be

is

of

1.

of 10)

the

previously

to

tab

That

is, T is

scale

factor

output

information

to a position

the

same

to

be

set

previous

as

T1.

to

i,

to

be

to

where

overwritten,

the

beginning

the

scale

or

of

factor

previously

the

record.

is

treated

FORMA T

read

input

to

as a multiplier

Statement

49

r

At

the

beginning of

zero.

Any number of P

scale

factor

to

be chanQed several times during a formatted

within a single

the

value

Scale

factors

put G conversions.

input/o~tput

of

the

scale

are

each

specifications

factor

effective

formatted

operation

is not reset to

only

with F,

input/output

may be present in a FORMAT

due to

E,

operation,

the

number

zero.

and

D conversions,

before

input/output

of

items in a list (see "FORMAT

any

processing occurs,

statement,

operation.

floating-point

the

scale

factor

thereby

input G conversions, and

causing

If

a

the

FORMAT

and

List

value

is

re-scanned

Interfacing"),

is

set

of

the

E-type

to

out-

Output.
shifted right i
Thus, for D-and E-type
for F format
examples

The examples for E conversion

sion. When G conversion uses the F form, however, sca
always represents the internal

The

value

of

the

list item is

places.

output

illustrate

Format External field when internal

2PFI0.3

1 PFI0.3

OPFI0.3

-1

PFlO.3

-2PFI0.3

-3PFI0.3

-4PFI0.3

2PE14.3 27.183E-Ol

1 PE14.3

OPE14.3 .272E

-1

PE

14.3

-2PE14.3

-3PE14=3

-4PE14.3

On D-and E-type conversions, the

output,

it

is

not (unless i

output

scaling:

2.71828

271.828

27.183

2.718
.272
.027
.003
.000

2.718E

.027E
.003E
.OOOE
.OOOE

value.

the

above

scaled

external

is

0).

Scale

00
01
02
03
04
05

are

simi lar to those

by

the

multiplier

exponent

number is equal to

factors

do

not

-2.71828

-271.828

-27.183

-2.718

-.272

-.027

-.003

-.000

-27.183E-Ol

-2.718E

-.272E

-.027E

-.003E 03

-.OOOE

-.OOOE

that

Ie

factors do not

lOi. This scaling

the

affect

00
01
02

04
05

would result from D conversion and E-type G

field (±ee)

internal

value

numbers whose

apply.

causes

the decimal point to be

is

correspondingly reduced by

(except

value

0.00000 0.09999
.000

.000
.000
.000
.000
.000
.000

.OOOE
.OOOE
.OOOE
.OOOE
.OOOE
.OOOE
.OOOE

Thus, a number output in G format

for rounding),

value

is

zero.

is:

00 99.990E-03
00 9.999E-02
00 .100E 00
00
00
00
00

while

The following

9.999

1.000
.100
.0lO
.001
.000
.000

.010E
.00lE
.OOOE
.OOOE

conver-

1.

01
02
03
03

Note

that

when a

case

of

E format, this may cause

the

discontinuity

Input. During
has no

effect.

is

scaled

by 10-

scaling

It
internal

50 FORMA T Statement

during

External Field

-71.

436

-71.

436E

can

be

seen

value;

external

scale

in

F,

E,

':'owever, when

1

;

that

an

input

00

that,

if

it

value

= internal

factor

is

in

effect,

additional

the

way

the

exponent

D,

and G input conversions, if

the

input string does not

is,

the

decimal point

operation:

Scale

Factor

OP

3P

-lP

3P

-lP

on both input and output, if

does not,

then

value x Wi

output

scaling

changes.

is

Effective

-71.

436

-.071436

-714.36

-71.

436

-71.

436

rounding takes

to be required, as shown

the

input string

moved

Value

the

contain

left i places.

external

an

number has

place

after

the

scaling--has been performed.

above

in

contains

exponent

The following examples

field,

an

exponent

an

exponent

the

the

output

field,

value

indicate

specified,

of

of

0.09999.

the

scale

the

external

the

it

is

equal to the

factor

field

effect

In

the

Note

of

Once

a sca

ation,

unless redefined

a

OP

specification.

A,

and

the

FORMAT(2(F

A,

K,

and B

as

illustrated

List

'Item

A F.3

K 2PF.3
X

B -2PF.3

Ie

factor

K,

X,

B

FORMAT

are

below:

Effective
Format

Speci

has

For

.3,

2P), E

all

converted

fi

cat

2PE12.4

been

established during

by

an

additional P specification.

the

list

12.4,

-2P)

using

the

F.3 format

ion

an

input/output

specification,

To

reset the

operation,

scale

but

all

it

remains in

factor

three

to

have

effect

zero,

it

different

throughout the

is

necessary

scale

factors in

to

oper-

write

effect,

When i is not

P is

/

Specification

The

form

r/

Each

slash(/)

/ /

//

••. / or
input,
fication

the

final right parenthesis in a FORMAT

Output.
record is
statements

10

specified,

equivalent

of

the / specification

or

r/),

and blank records

and

either

Whenever a slash

begun.

WRITE

FORMAT (F5.

to

OP

(Record

Separator)

/

specified

since

no conversion occurs

of

the

If

no

(5, 10) X, Y

3//113)

its

value

causes

are

generated

parenthesis

specification

conversion has

is assumed

is

another

to

be

zero.

Therefore,

record to be processed.

between

during

characters

statement

is

encountered,

been

performed when

each

output

operations. The same

surrounding

is

not ignored.

the

of

the slash

the

current

the

In

the

case

of

contiguous

specifications,

condition

field

specifications

record being processed is

slash is

encountered, a blank

slash

specifications

records

are

are

,ignored during

can

occur

when a slash

contiguous; a slash

output,

record is

(i .e.,

preceding

and

another

created.

speci-

The

are

processed in

1.

A record

2. The first slash
record is begun.

3. The second slash is
no conversion

4.

The

value

encountered,

the

following manner:

is

begun, and X

is

encountered,

occurred

of

the

variable Y is

and

the

encountered,

third record is

is

converted

between

with

the

record

the

second record is

the

terminations

converted

terminated.

the

specification

containing

of

with

the

F5.3.

the

external

terminated,

the

first and second records,

113

specification,

representation

and a third record is

the

closing

of X is

the

second

right

terminated,

started.

record was

parenthesis

FORMAT

and

another

Note

that

blank.

character

Statement

since

is

51

r

I

If

a th i

rd

i tern Z were

WRITE

(5,

the

following

5.

A fourth record

additional

6. The first slash is

7.

Again,
started.

8.

Since

or

Note

since
The

original

or

both of which would cause

The two statements

4

cause
specification

the

there

sixth record, which

that

the processing

10

FORMAT

the

specification

FORMAT

FORMAT

10

FORMAT

10

WRITE

(M,4)

FORMAT

the

generation

EA), followed by

added

10)

X, Y, Z

is

begun, and Z

re-encountered,

second slash is processed; the fifth record, which

are

no

more list items, the

(F5.3,//)

113

statement

(F5.3, 2/113)

(F5.3,2/,113)

X

(3/EA/)

of three blank records, followed by a record

to

the

output Ii

steps would occur:

is

also

blank,

of

Z in steps 5 through 8 is

was not

identical

utilized.

could

another

is

converted

the

fourth record

is

also

have

effects.

blank

st,

specification

output.

as in

using the

been

record.

specification

is

terminated,

113

is not processed, a termination occurs, and

equivalent

written as

F5.3.

and

a fifth record is begun.

is

blank,

to processing with

containing

is

terminated;

the

and

the

statement

value

of X (converted

the sixth record

the

by

the.

is

final

effect

Input. The
input, records

READ

4

FORMAT

cause

three
assigned to X,
ated

with a

ment, which

Parenthesized

Within a FORMAT statement
preceded

where r

is

equivalent

by

and

3 FORMAT (3(A4,

FORMAT (A4,

3

of slash

ary

ignored in the cases where

(M,4)

X

(3/EA/)

records to be bypassed, a va lue from

and

a fifth record to be bypassed. This means

FORMAT

is

statement

not true

in

FORMAT

an

the

to

optional

S.

I

are

defined

F.

F.

2, 3X, A4,

repeat

2, 3X),

specifications

containing

FORTRAN systems

during input operations is similar to

slash

Specifications

any

number of

count,

previously, and m

31)

F.

in the form shown on

2, 3X, A4,

blank

records

the

fourth record to be

specifications

that

ignore a final slash.

specifications

~

O.

For example,

F.

2, 3X,

31)

are

may

created

that,

can

during output. For example,

converted

as with

be input by use of

be

repeated

the

following

the

the

effect

(with the

the

last example for

by

enc

page.

statement

for output,

the

identical

losing them in parentheses,

except

the

specification

output,

FORMAT

that

statements

EA) and

recordscre-

state-

for

There

is

no limit to

During

52 Parenthesized FORMAT

input/output

the

number

processing

Specifications

of

repetitions

each

repetitive

of

this form

specification

that

can

be present in a

is

exhausted in turn, as

FORMAT

is

each

statement.

singular

specification.

The following

34

FORMAT (4X, 2(A8,

1125 FORMAT (I,

8 FORMAT (2(18, 2(3X, F 12.9), F12.

In

the

last

example

The

presence
scanned
cular,
the
of

if

more list items

when

last

one

this process is

are

additiona I examples

R4,

above,

of

parenthesized

one

or

more such groups

encountered

contained

X,7G.3),

F.7,

5(E

14.8,

repetitions

groups

are

specified

prior

to

the

in "FORMAT

of

repetitive

I4,3(D,

2/),

have

within

than
have
final

and

specifications:

L5))

E 14.8)

9), A 16)

been

nested.

a FORMAT
are

processed the first time through

appeared,

right

parenthesis

List

Interfacing".

statement

the

rescan

Nesting

begins

of

the

of

this

affects

with

FORMAT

type

is permissible

the

manner

the

the

statement.

FORMAT

group

in

which

whose

A more

to a depth

the

FORMAT

statement.

right

parenthesis

complete

of

ten

levels.

is

re-

In

parti-

was

discussion

Adjustable

The

adjustable

statements

form

Any
tion.
merely

N.

of

quantities

is a

The following
the

1.

2.

3.

4.

in

is

highly

of

the

quantities

When

an

a form of

Also,

there

32

FORMAT

valid

statement,

values

are

Integer,

data

wi

II

wand

d (width

or

(skip

The

resultant

When N
prior

to

3,4,

and

the

3 FORMAT

which

are

A,B,C,D

and

the

3 FORMAT (3E4.1, 12X,

FORMAT

FORMAT

order

to

handle

variable,

r,

N is

encountered,

specification

is no limit to

replaced

set

utilized:

real,

be

scale

appears

th e items

1,A,B,C,

statement

statement

by N in a format

(NX,

FNA,

and

of

rules

double

truncated

and

factor

value

one

(if

(NEN.N,

equivalent

Specifications

specifications

slightly

and

which

W,

d,

or i (see "FORMAT

and

the

N(3X,

seven

values

defines

precision,

to

integer

decimal

count)

(negated

12,-2,D

point)

specifications

or

more times in a

any)

that

NX,

to

the

list

different

could

its

value

does not

number

E.5),

will

the

manner in

or

value.

if

preceded

are

NP,

-2P,

G 14.8)

feature

not

of N characters

specification.

NP,

either

specifications

to

G 14.8)

often

el

situations.

be

processed

Statement")

is

obtained

conflict

NGN.N)

be

taken

which

part

may

by a minus sign) may

single

be processed by

from

with

For

form

the

of

complex

may

be

replaced

specification,

iminates

Furthermore,

without

may be

the

next

any

variable,

that

may

example,

the

list.

value

of

data

be

replaced

by N

the

specification.

the

need

to

it

facilitates

this

feature.

replaced

input

or

output

subprogram,

be

used in a FORMAT

N must

be

its

may

be

only

or

-N.

negative

values

be

specified

supplied

by

must

An

write a great

the

by

the

letter

list item. The

etc.,

whose

statement

in a list

as

values

N,

whereas r (repeat

only

when N is used

appear

sequentially

example

is

number

input

of

records whose

N in a format

identifier

and

for

N.

the

list

of

FORMAT

letter

or

to

the

the

way

Non-integer

count)

to

replace

in

the

specifica­N is
may

be

number

in

which

and

i

i.

list

and

5.

Whenever N is

Format

Specifications"),

sequently,

7, A,

are

equivalent

A,

B, C and

used

the

difference

B, C and

to

with a specification

3FN.2

3F7.2

one

value

between

must

the

that

is

be

supplied

following

enclosed

for

two

examples

in

repetition-type

each

repetition

should

7, A,

are

equivalent

A,

B,

7,

C

B,

be

7, C

to

parentheses

of

the

specifications

noted:

and

3(FN.2)

and

3F7.2

Adjustable

(see

FORMAT

"Parenthezised

enclosed.

Specifications

Con-

53

r

6.

In

the

above

the

value

count.
a.

b. However, when

The same rule

When

ing in

are

equivalent

specifications

are

equivalent

of

the

that

0,4,

Y

0, Y

y

the

example,

Y

it

repeat

count;

can

repeat

count

specification

and

to

and

the

appearing

and

to

and

was noted

whereas,

be

extended

(r)

of a single

must

NG20.N,

F8A

repeat

count

within

N(G20.N},

F8.4

that

in

the

specification

in 3(FN.2},

to

include
specification

be

supplied. For

F8.4

of a parenthesized

the

parentheses

F8.4

3FN.2,

the

number

repeat

counts whose

is

replaced

example,

are

the

group is

bypassed, inc luding

one

of

values

by N

following

replaced

value

of

N is

required for N is

values

are

zero:

and

its

value

combination

by N and its

any

Ns

required,

equal

is

zero,
of

list and FORMAT,

value

that

may

regardless

to

any

is

zero,

appear.

the

Ns

repeat

appear-

all

the

Thus,

of

In both of
cation

The
certain

the

within

abi

lity

to

specifications

above

parentheses

specify

T=O

IF

(BOOlE)

F = 1 - T

PRINT 17,

17

FORMA

outputs

the

7. The
this purpose in

As

an

example

205

The

value
versions to be performed by
uous to
be

correctly

the

n in

an H specification,

value

of
READ(101,205)
FORMAT(

input

the

last

processed by

T = 1

BOOlE,

T(Ll,

N(3HRUE),

strings

TRUE

of N may be supplied by

an

input list is not

the

flexibility

I,

NE , A4}

for K

defines

field

input

examples, no

can

zero

repeat

within a FORMAT

T,

or

FALSE

provided by

K, K,

(A(J),

not only

the E specification.

into

A, regardless of

the

above

value

be

used

counts in this way

F

N'

AlSE'}

depending

the form shown in

considered

J=l,K},

the

statements:

was supplied for

to

determine

statement.

an

expression in

adjustable

CODE

number

the

the G specification;

whether

gives

on

the

value

statement

either

to be a true

format

of

va lues

At

the

same time,

number of such fields. Thus,

or not

the

For

example,

of

BOOlE.

17

above

an

input

specifications,

to

be

input

the

programmer

can

input

or

list item.

into

the

alphanumeric

value

the

Note
be

used.

an

output

consider

the

however, enclosing the

of

N will

be

supplied.

facility

that

array

all

of

selecting

although

list, but

the

A,

value

the following input records

an N cannot

an

statements

but

also

of CODE

expression used for

the

or

skipping

number
can

be

specifi-

replace

of

con-

contig-

can

1,

67.49,

HOPA

5

-14.3

37

.09711623 0 3E12 JASU

,NONE

This

example

nation

Numeric

The permissible kinds
D,

G,

In

other

be

integers,

54

Numeric

illustrates not

(see below).

Input

Strings

and I conversion. Any fie

words, numbers for input with E format need not have

etc.

Input Strings

only

of

input strings

adjustable

that

Id

that

format

specifications,

may be processed by numeric conversions

can

be

read using

one

but

also

of these formats

exponents,

numbers for input with I format need not

widthless formats

are

can

be read using

exactly

and

comma field

the

same for

any

of

the

termi-

F,

E,

others.

A

numeric

exponent.

E±e

where

the

in

which

input

An

plus

case

string

exponent

sign is

the

plus sign is

consists

is

normally

optional

of a string

specified

and

e is a

not

optional.

of

one-

digits

as

or

two-digit

Thus, a

with

variety

or

without a leading

number.

of

forms may

The form

be

sign, a decimal

±e

is

also

accepted

used

to

express

point,

data

and/or a trailing

(without

for

numeric

the

input:

E),

plus signs

fields

the

format

of

the

be

substituted

that

data

is

to

be

type

value

is used (with

be

of

specified).

±n.m

±n.m±e

are

contain

specification

to

field.

is

(if

any), a numeric

assigned

of

wi

II

be

used to

embedded

±n
±nE±e ±n.mE±e
±n±e

where

the

When

input

the

d in

placed d positions
the

end

A D may

indicate
form

of

exponent

which

it

Any

numeric

the

input

specification

A comma may
interpretation
(no

width

Width less Numeric Input

optional

the

left

Note

for

double

list

item

processed

terminate

±n.
±n.E±e
±n.±e

no

decimal

of

that

the

E in

precision,

is

double

may

any

type

and

±.m
±.mE±e
±.m±e

except

the

the

trailing

in

point

(as in Ew.d).

beginning

exponent

an

exponent

nor

string

precision.

be

used

in

floating-point,

of

list

any

numeric

blanks

an

(as in

may

is

it

will

with

item),

exponent

the

first

If

none

of

the

exponent,

begin

field,

necessary

be

converted

any

numeric

if

the

fractional

field,

depends

field

without

column

is

specified

with

either

with

no

to

use a D format. Regardless

with

type

necessary,

as

described

on

whether

above)

or

if

change

full

of

and

portion

below.

or

an

it

is assumed

no

exponent

a D,

in

double

format

then

of

not

E (as

described

the

decimal

to

E,

+,

or-.

meaning

precision

specification.

converted

the

value

Leading

the

format

above).

point

is

positioned

be

zero.

is

present, d positions

or

value.

to

integer.

is lost.

blanks

specification

of

if

the

the

If

are

It

the

The

is not
format used or

input

When

always

decimal

list

list

item

used

according

point

to

the

necessary

item

is

integer,

the

I format

ignored.

is

widthless

to

left

to

the

to

is

of

The

The

principle

that

the

number

ber

at

that

1. Leading

2.

Any

number

a.

Between

b.

Between

c.

Between

3. A

blank

4.

When a widthless
as

blanks.
found.
number

For

clarity,

the

field.

of

the

field

The

following

behind

widthless

is

finished. A blank

point.

Although

Thus:

blanks

do

not

of

blanks

the

leading
the E and
the

plus

that

follows a

(or

any

Normally, a widthless

Special

of

numbers should

it

numeri c

has

terminated.

is a

typical

provision is

values

the

terminating

input

cause

termination;

may

appear

plus

or

the

pi

us

or

minus sign in

digit

or

other)

made,

may

generally

Therefore,

widthless

is

that

also

in

the

minus sign

or

minus sign

decimal

field

format

however,

be

read

be

blank

numeric

the

field

indicates

they

following

the

exponent

point

runs

off

does

from a

written

or

comma

the

next

input

ends

that

are

and

the

or

first

terminates

the

end

not

terminate

to

terminate

blank

without

does

field

line

when

the

number

ignored.

places:

first

digit.

digit

and

the

the

of

the

record,

any

embedded

not

affect

begins

consisting

the

number

is

of

the

exponent.

first

digit

field.

input
until

at

widthless

and

they

the

with

the

of

eight

is

finished,

of

record,

least

fields

wi

II

blanks.

value

character

values:

finished.

if

it

the

exponent.

the

extra

one

non-blank

at

the

all

be

The first

of

the

following

A comma

is

meaningful

characters

end

of

zero.

blank

number,

character

the

record.

will

it

is

the

blank

always

to

finish

will

be

has

Thus,

then

considered

or

indicates

the

num-

interpreted

been

any

terminate

part

comma.

73

2E-4

.0007

-35.4

0 0 -16 27.08614E

12

Numeric

Input

Strings

55

r

I

The

but

following

3E

~ou

Id

be

-300.

is

not a

typical

2

+

interpreted

+

3.7

widthless numeric

3.7

as five

values,

-4.

4

namely,

17E

1700.

input

2

line:

5-

.005

03

Numeric Input with

When a width

apply

rules

1. Leading blanks

2.

Once

3. Any string of

For a format

produce

the

others

Note,

in

cause

the

justified

and/or

is

specified,

to

blanks in numeric

any

nonblank

specification

the

value

are

permissible,

the

fourth

trailing

in

their

fields.

widthless formats.

Width

Specified

are

ignored,

digits

shown in

example

blank

the

field

terminates

fields

with a width

except

character

that

is

omitted

such as F10.0, with no P

the

top

but

less

that

has

been

has

line

of

readable.

-

.00.4

they

found,

an

the

- 4 E - 3

-

.004

4

-

40

- 8

middle column

to a zero.

do

this

is

a common

- .

-4

of

is

equivalent

Failure

-

the

to

only

when

the

specified:

are

counted

all

blanks

assumed

column. The first

scale

value

factor,

of

7'.5
7'50+13

.

750E10

4

Care

75

.

above,

should

75

+

75

that

pitfall

width

is

as

part

of

beyond

zero.

all
three

that

the

E 12

E'l 0

o 1

E 16

the

exponent

always

be

that

can

exhausted

the

field

point

input

strings in

I ines in

OE

+

is

interpreted

taken

also

be

or

a comma

width.

are

treated

each

column

0.0

+

-0

-

to

assure

avoided

is

found. The following

as zeros.

each

of

the

columns

are

typical

0

as 10

rather

that

exponents

by using comma

below

numeric fields;

than

as 1,

be-

are

right-

termination

Input strings
the

presence

currently
regardless
the

are

The string

Two
tion

is

converted

t

For

thesis

being

character

3450,88412,

equivalent

345088412

contiguous
516,

the

303,

, - 1, ,

303

consistency

are

being

of

a comma in
processed is

of

the

value

following

to

F4.3,

containing

comma

string

000450

to

the

0

-1

with symbolic

also

accepted

processed

of

the

F5. 3

the

commas would

characters

values

0

450

as

under

the

string.t In

considered

w in

the

comma. For

and

indicate

input

field

terminators.

control

ended,

field

specification.

also

(via

of

F,

other

and no

example,

be

correctly

an

empty

the

INPUT

Use

E,

D,

G,

I,

or L specifi

words,

whenever

additional

The comma is not processed,

th-e

specification

processed by

field,

statement),

of

the

comma is recommended,

a comma

characters

which has

the

cations

appears

are

2F13.3

and

the

specifications

the

value

characters

may

be

in such

converted.

and

the

string

zero.

Therefore, for

semicolon,

however.

terminated

an

input

This

termination

the

next

2F. 3

or

2F8. 3.

asterisk,

at

any

point by

string,

the

field

occurs

field begins with

the

specifica-

and

right

paren-

56

Comma

Field-Termination

and

the

stri ng

0, ,

'"

is

converted

o 0 0 0 0

to

the

values

The comma must,
F4.0

were

used

1234,

the

value

would

giving

it a value

FORMAT

Formatted
time a formatted

processor

1. When

2.

3. During

and

input/output

operates

control

Subsequent
been

terminated)

more

characters

on

output

an

final

right

and

no I ist items

on

occurrence

of

course,

to

process

already

of

zero.

List

Interfacing

READ

in

the

is

initially

records

or

on a

the

extra

input

operation,

parenthesis

remain

of

the

fall

the

input

be

terminated

operations

or

WRITE

following

received, a new

are

started

the

final

record

characters

of

the

to

same

conditions.

within

string

because

are

controlled

statement

manner:

only

right

parenthesis

than

are

are

lost, on

processing

FORMAT

be

processed.

the

is

after

(or

of

is

field

it

is

of

field

by

executed,

input

a slash

of

can

be)

input

an

input

sensed,

Construction

meant

to

width,

the

FORMAT
control

record

is

read,

specification

the

FORMAT
physically
they

are

record

is

or

when

of

terminate.

and

the

comma would

requested

is passed

present

treated

terminated
the
an

output

to

or

construction

has

been

has

been

does

as

blanks.

whenever

FORMAT processor

record

For

example,

by

each

the

FORMAT processor. The FORMAT

processed

sensed.

not

cause a new

terminates,

if

terminate

READ

of a new

(and
Attempting

a slash

requests

the

format

the

or

WRITE

output

the

preceding

to

record

specification

an

item

and

the

specification

following

statement.

record

read

to

be

from

record

field,

is

begun.

record
(or

write)

begun;

or

the

the

is

written

Each

has

list

4.

5.

Every

time a conversion
cessed,
forms
because
that
exists.

would

When

version
following

When

made
been

statement
another
scan

a.

the

the

of a conflict

does

Thus, for

WRITE(6, 12)

12

FORMAT(///4HABCD)

produce

there

specification

the

to

determine

processed.

list

takes

If

there

is

rescanned.

FORMAT

appropriate

not

require

example,

three

are

no more items

the

READ

final

right

If

following

item

is

place

as follows:

are

no

specification
processor
conversion
between a specification

a I ist

blank

has

or

WRITE

parenthesis

if

all

the

list

the

present,

parenthesized

requests

item

(i.

the

statements

records

remaining

been

found,

statement

of

list items

has

been

RE~r

an

additional

and

proceeds

e.,

and

a FORMAT

(including

exhausted,

WRITE

groups

(i.

e.,

a I

isf

H, " X,

one

record

in

the

the

current

that

statement

record

of

specifications

F,

E,

D,

item.

with

and a data

T,

containing

list

and

record

initiated

statement

those

to

the

that

is

begun,

G,

If

one

the

next

P,

or

the

the

be

current

initiated

I,

L,

A,

or

more items

field

type,

an

/),

it

is

ABCD

final

right

is

terminated,

input/output

is

encountered

used

as

values

record

the

and

the

within

the

R,

Z,

M, or N

remain

specification.

error

occurs.)

processed

before

reaching

parenthesis

and

control

operation.

by

the

of

N in

is

terminated,

il1put/output

FORMAT

FORMAT

specifi

in

the

(If

conversion

If

the

whether

the

has

is passed

FORMAT

adjustable

and

operation.

statement

statement,

cation)

I ist,

the

processor

is

next

specification

or

not

another

final

right

been

reached

to

the

processor, a test

specifications)

control

However,

is

rescanned.

the

is passed

entire

is

to

be

not

possible

I ist

parenthesis.

or a con-

statement

to

if

The

FORMAT

pro­per-

is

item

is
have
the

re-

one

b.

If

one

or

more

parenthesized
parenthesis
the

following

was

the

example,

last

groups

one

encountered

the

rescan

do

appear,

begins

prior

at

however,

to

the

point

the

the

final

right

indicated.

rescan

parenthesis

is

started

with

the

of

the

FORMAT

FORMA T

group

and

whose

statement.

list

Interfacing

right

In

57

FORMAT{3X,{F7.2,A5),{X

I I I I I i

'ABC'3(3I4,2(G

15.7

II

//),R3)),E20.12,3HXYZ)

c.

6.

Each list
with

7. Each
least

statement

8.

The same rules
pie

If

the

group

peat

count

value

new
statements

5 FORMAT( A4 /

are

equivalent

5 FORMAT(

item

the

exception

READ

or

one

adjustable

wi

records in

is used

of

PRINT

PRINT

to

WRITE

II

be

apply

these

at

which

again

N is

not

5,

CODE,

to

the

5,

CODE,

A4/

be

converted

of

complex

statement

or

conversion

processed,

to

DECODE

cases

1 1 1

rescan

begins

here.

the

rescan

begins has a

for

each

rescan. In

supplied when

5,

(A(J),

N(G20.8)

statements

(A(J),

5(G20.8)

is processed by
data,

containing a non-empty

but

an

and

is

described

J=

)

J=l,

)

whichare

(see

step 4 above)

error

wi

ENCODE

under

the

50)

last closing
parenthesis

of

internal

group.

repeat

count

particular,

rescan

takes

place;

1,50)

one

specification

processed

1\

occur.

operations

"Memory-to-Memory

by

two

list must

specification.

as

final
parenthesis

of

FORMAT.

(r) in front

if

the

the

or

one

such-specifications.

refer

to

READ

Data

of

repeat

count

old va lue is used. Thus for

iteration

to a FORMAT

If

this

condition

and

WRITE.

Conversion".

right

it,

the

previous

was

specified

of a repeated

statement

is

The

value

of

with

example,

specification,

that

contains

not

met,

the

interpretations

the

re-

an

N,

the

FORMAT

of

multi-

a

at

FORMATs

As

mentioned
specifications
a string of

READ
with

WRITE

while

WRITE

If

the

the

externa

Stored

characters),

or

WRITE

no subscripts. For

(4,

(4,

variable M is

in

previously,

enclosed

statements

R)

E,

F, G

R(l»

E,

an

I stri ng

(F8.5,4HNAIL,I3)

and

the

statements

READ

(N,90)(M(I),I = 1,4)

90

FORMAT

Alternatively, M could
Care

must

In

these

cases,

be

(4A4)

taken

all

when storing

characters

Arrays

a FORMAT,
therein,
usually

that

example,

F,

G

integer

be

a dummy

including

may

be

stored in

by

use

of

an

refer

to a FORMAT stored in

refers

to

a FORMAT stored in

the

following

FORMAT

refers to a

array,

1

J

array

corresponding to a

into

an

array

in

the

string s,

the

an

input

statement

are

a FORMAT

including

beginning

array.

statement

are

two methods

equiva

lent

containing

blank

left

parenthesis,

The FORMAT must

or

an

assignment

an

array

must

reference

an

array

R,

whose

location

that

may

the

TO

f

1

literal

constant

characters,

argument (see IIArguments and Dummies

specifications

are

the

final

right

parenthesis,

be

stored as a

statement.

only

has

been

be

used to store a FORMAT in M:

statements

M(l)

'1'

/'l'

1"11.L.)

M(3)
M(4)

of

significant,

Hollerith

the

identifier

ASSIGNed

4H(F8.

I.TTe

'.'n

"'tn.J,"'tn

4HNAIL
4H,I3)

the

nHs

while

to R(l).

and's'
blank

and

string

(i.e.,

of

the

array,

forms.

characters

the

ll

).

58 FORMATs Stored in Arrays

are

insignificant

instead

Element

M(l)

M(2)

14(3)

M(4)

which

(F8.5,

where'G

of

integer,

is not

the

4Ho'G'O'O,

represents

in

all

the

Storage

(F 8 •

5,4Hf>f)"of>

NAIL

,I3)'M>M>

desired

NA,

the

other

specifications.

following

after

'Df>'Df>

1>f>f:l'f)

result,

I, L,I3)

character

results

READ

since

blank.

it

would

is

For

example,

occur:

equivalent

to

if

the

M in

the

FORMAT

above

READ

statement

were

double

precision

Even though a FORMAT may

(I8)

it

must

be

stored in

able

(or

an

array

rather

than

as

Extended

The

statements

"Memory-to-Memory
"Direct
"Random

are

optional

that

the

Input/Output

Input/Output"

Access

features

FLAG system

Memory-to-Memory

The

statements

an

ENCODE/DECODE

an

input/output
integer
in
FORMAT

These

array.

an

ENCODE/DECODE

being

statements

ENCODE(c,
or
ENCODE(c,

an

element)

if

the

scalar

described

Input/Output
within

avai

ENCODE

operation,

list

and

an

Whereas

used, records

have

the

f, s,

f,

an

s}

array.

below

Data

and

operati

n}

be

quite

short,

It

may

will

be

treated

contained

the

Capabilities

under

the

Data

Conversion"

Statements"

the

FLAG

lable

for

their

Conversion

DECODE

however,
internal
external

form

k

k

on may

after

buffer

record

the

such

as

not

be

stored

as

though

FORMAT.

headings

language.

use

includes

are

similar

no

actual

area.

This

has a certain

be

specified

first

record

DECODE(c, f, s,

or

DECODE(c, f,s) k

in a

the

scalar

Programmers

these

optional

to

formatted

input/output

buffer

area

physical

by

the

programmer. When

follow

each

n}

scalar

were

wishing

(BCD)

other

k

variable,
assigned

to

statements.

WRITE

takes

place;

is

specified

length,

in memory in

since a reference

the

location

use

any

of

these

and

READ

data

by

the

programmer

the

length

of

mu

Itiple

order

of

a FORMAT

statements

statements,

conversion

the

simulated

records

of

increasing

to a

scalar

statement,

should

respectively.

takes

place

and

is

usually

internal

are

specifi

storage

vari-

ensure

between

an

record

ed

by

address.

In

the

where

c

defines

subscripted

specifies

array

indicates

element,

n is

of

k is
a comma may

the

a FORMAT

in

which

the

or a scalar

an

optional

characters

an

input/output

optionally

number

variable.

a FORMAT has

first

integer

actually

of

characters

statement.

element

variable.

variable

processed

I ist

of

precede

been

or

starting

the

usual form,

the

Extended

per

internal

It may

stored.

location

into

which

(generated

list k.

Input/Output

be

and

record.

the

statement

of

will

or

scanned).

the

be

It

internal

stored,

Capabi

may

be

an

integer

label

of

a FORMAT

buffer.

upon

completion

lities/Memory-to-Memory

It

constant

may

of

or

statement

be

an

array

the

operation,

an

or

name,

Data

integer

the

non-

name

of

an

array

the

number

Conversion

an

59

Thus,

the

ENCODE/DECODE

statements

can

be

illustrated

as

r

r ENCODEl

lOECODE

Characters

fied as
record;

ENCODE

The ENCODE
specified

If

the
extra
necessary
operations,
{generated}

For

example,

3 FORMAT(2H(F

in

the

starting

in

other

Statement

by

number of

characters

to

K

=

12

L=5

ENCODE(l2,

I

the

words,

statement

f,

and

characters
are

fill

the

the

first

into

it.

the

(characters,

buffer

are

location.

at

causes

to

be

placed

lost;

they
record,
thing

following

3,

M)

K, L

,

I,

processed

When a new

character

the

in

generated

are

it

is

filled

done

with

statements

1H. ,

format,

not

start,

at

the

c +

1.

list items to

storage

by

the

filled

out

each

record

might

I,

1H)

count) jist

rate

of

record

is

It

is

recommended

be

beginning

FORMAT

into

the

with

trailing

is

be

used

)

four

per

begun,

converted

at

location

statement

following

blanks.

to fill

it

to

create,

word

without

it

starts

at

that c be

to

BCD

character

s.

is

greater

record.

In

fact,

with blanks;

for

later

regard

to

the

type

of

the

first

character

an

integral

than

If

fewer

on

this

use, a FORMAT stored in

multiple

strings,

the

specified

characters

ENCODE

is

done

following

according

are

operations,

before

any

of

size

the

variable

the

four

characters.

to

the

of

the

generated

as on

characters

the

array

speci-

previous

FORMAT

record,

than

are

WRITE

are

stored

M:

the

BCD

FORMAT so

The

(F

12n.

where b represents

DECODE

The DECODE
specified

As

with formatted

ified

by

the

new record

If

n

is

specified,

can

be

very
INTEGER KARD(80), DAVE/DAVE
READ

4 FORMAT(20A4)

DECODE(80,5,

5 FORMAT(A4,

IF

6 FORMAT(NX, NF)

created

would

56)niSiS

the

character

Statement

statement

by f,

and

count

is

begun

useful. The following

4,

KARD

(KODE

stoied

READ

(c),

only

it

will be

I)

.EQ.

causes

operations,

the

when

KARD,

DAVE)

into

occupy

the
the

extra

specifically

set

to

NC)

bl

characters

the

first

ank.

character
items in list k.

if

the

example

string

the

FORMAT

are

requested

number

of

makes use of this

1/

KODE, J

DECODE(80,6,

three

elements

beginning

statement

assumed to

by

characters

KARD)

of M and

at

location

requires more

be

blanks;

the

FORMAT (see "FORMAT and

scanned.

feature:

NC,

J, (A(I), 1=1, J)

would

s to

they

When

appear

be

characters

are

not

scanning

as

decoded,

obtained

according

from a record

from

list

with widthless formats, this

to

than

the

next

Interfacing").

the

FORMAT

are

record. A

spec-

The

above

statements

DAVE

2,

Note

that,

in

DECODE

way

60

Memory-to-Memory

essentially

because

it

1.

75,

the

above

initially

could

be

used to

80.91
example,

provides

fills its buffer with blanks.

Data

Conversion

the

capability

the

read

records

first DECODE

of

"reading

of

the

statement

the

form

card

is

used
twice".

to

decide

how to

ENCODE

interpret

cannot

be

the

used in

rest

quite

of

the

the

card.

same

Direct

FLAG

provides

BUFFER

which

have

1.

They
matted
binary,
environment.
extension
into
tions
giving

Input/Output

the

facility

IN

and

BUFFER

two

major

capabilities.

permit

the

processing,

READ

and

WRITE

READ

and

WRITE

That

of

computer

fixed-length

manual).

the

physical

The

programmer

to

perform

statements

statements

is, in FORTRAN,

memory,

BUFFER

complete

OUT

both

and

records

IN

and

asynchronous

on

that

control

input/output

input

and

output,

handle

records

process

as

BUFFER

information

binary

information

such,

binary

comprise a logical

OUT

over

the

of

only

records

routines,

data

and

operations

records

up

to

that

is

considered

have

record

however,

enabling

of

132

is

intended

special

(see

through

arbitrary

characters

for

to

be

ized

the

XDS

process

him

to

the

two

length

and

in

length,

communication

a form

of

control

information

do

Sigma

such

information

5/7

things

library

routines

format.

and

unformatted,

only

intermediate

FORTRAN IV

exactly

as:

Ordinary

with

a FORTRAN

storage,

and

are

as

specified,

for-

or

an

broken

Opera-

2.

The

and

where

interpret
read
in

conjunction

BUFFER
ables
output
the

BUFFER

CALL

CALL

u is

m

w

n is

IN

the

is

program.

IN

BUFFER

BUFFER

an

device

is

if

m=O,

indicates

also

specifies

of

positive

is

an

is

indicated

optional,
the
in progress. In
n is

binary

and

write

and

BUFFER

programmer

taking

place,

and

BUFFER

IN(u,

OUT{u,

integer

on

which

an

integer

the

the

be a scalar.

the

val

integer

as

1 =

operation

2 =

successful

3 =

end-of-fi

4 =

operation

number

of

normally

tapes

produced

binary

cards,

with

the

ENCODE

OUT

to

initiate

and

to

OUT

m,

s, w, i,

m,

s,

w,

constant

constant

mode is BCD;

starting

number

ue.

scalar

follows:

but

words

general

equal

or

the

operation

or

location

It may

of

into

incompl

compl

I e

encountered

complete

when

specified

actually

to

w (see below)

this

on

and

proceed

subroutines

n)

i,

an

be
words to

which

etion;

in

an

operation,

test

the

n)

integer

is

to

an

integer

otherwise,

of

of

any

is

ete

no

but

error

is

input

count

other

machines

and

DECODE

parallel

continue

status

of

the

are

called

nonsubscripted

be

performed.

nonsubscripted

the

mode is

the

internal

type,

although

be

input

or

dynamically

errors

has

occurred

an

integer

or

output.

is

significant

statements,

with

the

BUFFER

in

the

buffer.

output,

stored

scalar

It

is

only

or

by

program

with

IN

following

variable

variable

binary.

integer

starting

an

into

not

for

other

programs,

process

and

computation

or

BUFFER

fashion:

that

that

(An

integer 0 or

Normally,

is

recommended

at

s,

indication

which

is

continuously

BUFFER

IN

long

formatted

other

input/output

and

other

OUT

specifies

determines

s is

the

and

must

of

the

status

stored,

upon

set

up

operations;

processing

operation

the

logical

the

1 is

customary.)

identifier

for

ease

be

an

of

completion

lion

the

in

records.

operations.

at

some

unit

mode

of

of

an

of

manipulation.

integer

the

operation.

flyll as

BUFFER

while

later

number

the

operation;

array,

constant

of

the

the

operation

OUT

This

en-

the

input/

point

in

of

the

but

it

may

or

scalar

The

status

operation,

operations,

is

Thus,

the

BUFFER

CALL

BUFFER

{

BUFFER

IN/BUFFER

IN

}

OUT

OUT

(unit,

calls

mode,

can

start,

be

illustrated

words,

as

indicator,

count)

Direct

Input/Output

61

I

I

A

BUFFER
into,
ing

location

It

is

permissible

device,

or

written

including

IN

of

or

from,

the

to

BUFFER

consecutive

buffer.

intermix

intermixing

OUT

operation

That

asynchronous

on

always

words

in

iSI the

variable s merely

the

same

results

memory

operations

unit~

with

and

in

the

no

specifies

standard

processing

regard

to

the

READ/WRITE

of

the

type

starting

only

of

one

physical

the

variable

location.

operations

in

any

record.

specified

order

Data

as

and

is

the

on

read

start-

any

BUFFER

A

in

words
pearing
words

When

of-file.

The
despite

Example:

The
record

IN

call

on

BUFFER

the

specified

are

specified

in

the

are

present

an

end-of-file

No

error

status

the

following

length:

INTEGER

1 CALL

2

GO

3 M = MIN

PRINT

GOTO

4

STOP

5 FORMAT (/X, ILENGTH = I

IN

causes

mode.

by w than

record

are

on

the

is

data

will

will

be

error,

and

can

statements

BUFFER

BUFFER

TO

IN

(2,3,4,3,),

(N,5000)

5,

M, (BUFFER(K), K =

1

data

The

actual

changed

record

read,

be

read

indicated

be

could

(5000)

(5,

1,

J

to

number

are

actually

in memory,

than

i is

set

into

memory

when

used

if

be

used

BUFFER,

1,

1/

be

read

of

present

are

speci

to

3;

magnetic

upon

an

irrecoverable

the

programmer

to

list

5000,

J,

M)

(X, lOZlO))

into

memory from

words

in

and

that

fi

ed

by

encountering

binary

N)

entered

the

fact

w,

the

tape

chooses

tapes,

the

into

physical

will

extra

units

will
an

error

occurs.

to

in

hexadecimal,

specified

memory is

record

be

reflected

words wi II

remain

end-of-file.

ignore

the

unit,

the

minimum

that

is
by
be

positioned

However,

error.

ten

beginning

read,

only

the

value

lost,

and

immediately

the

data

words

of

per

at

wand

the

of

n wi

will

line,

the

number

n.

II

be

be

preceded

location

n.

That

of

However,

greater

following

read

into

specified,

is,

if

words

if

more

than

the

memory

by

the

more

ap-

w.

end-

62

BUFFER

A

calion

fied

mode.

record

on

attempt

If

the

indicator

has,

nonetheless,

As

mentioned

statement.

records

it

expects

logical

Note

that

ment

using

size

of

magnetic

Direct

OUT

BUFFER

The

number

the

device

to

write

30

variable

above,

This

has

created

record

by

to

find

{which may

the

output

IAI

format.

tape

Input/Output

OUT

being

words

been

records

one

BUFFER

the

control

produced

records.

causes

of

words

used,

on a card,

{i}

has

.written

written

important

OUT.

words

consist

However,

data

to

requested

in

which

been

on

the

ramification:

It

can be used

generated

of

several

in

BCD

this

be

written

in BCD,

set

to

specified

in

binary

by

BUFFER

does

from memory,

is

always

case n will

would

indicate

device.

by

BUFFER

the

to

backspace

by a binary

physical

OUT

not

include

written,

be

less

result

in n

an

error,

OUT

BACKSPACE

WRITE

records}.

is

virtually

carriage

beginning

unless

it

than w (it

being

set

an

irrecoverable

are

not

statement

over

any

statement

identical

control

is

can

the

kind

on

with

larger

never

to

20.

same

may

of

so

to

printed

the

specified

than

the

be

greater).

write

error

as

those

not

be

BCD record Sf

that

it

may

that

produced

output

location,

maximum

has

produced

used

to

but

backspace

by a

or

any

size

For

example,

occurred.

by a binary

backspace

in

binary

over

BCD

restrictions

in

the

speci-

of a physical

an

The

data

WRITE

over

binary

records

the

entire

WRITE

state-

on

the

Example:

The

statements

INTEGER M(40)

1

READ

2

FORMAT (20A4)

CALL

3

IF

below

(105,2)

BUFFER

(INDIC < 2)

could

M

OUT

GO

be

(106,

TO

used

1,

3

to

M,40,

read

INDIC)

BCD

cards

and

pack

them

onto

binary

cards.

GOTO

Note

that

FORMA T

Random

FLAG a Ilows full use

READ

where

u is

k is

Random
They

differ

1. They

and

the

reference

to

be

Access

DISC

an

the

disc.

is

also

(see

below).

an

access

from

refer

ENDFILE

1

scanned

u,

s, k

integer

an

input/output

input/output

the

to

random

operations

to

the

array

twi

ce.

Input/Output

of

random

constant

integer

standard

access

WRITE

constant

I ist,

operations

binary

access

are

or

nonsubscripted

as

files

not

M in

the

READ

Statements

devices

DISC

or

nonsubscripted

described

are

READ/WRITE

rather

applicable

through

u,

s, k

previously.

performed

statements,

than

to

to

statement

util

integer

integer

in

binary,

sequential

them.

causes

ization

variable

variable

however,

files.

of

whose

and

two

the

therefore

cards

(40 words)

following

value

specifies

whose

value

do

not

in

two

ways:

Consequently,

to

be

read

two

statements.

the

logical

defines

the

starting

reference

the

a FORMAT

REWIND, BACKSPACE,

and

unit

causes

number

disc

address

statement.

the

of

2. Information is
control
specified
bound

by

one

and

As

an

analogy,
element
to-Memory

In

READ

DISC

by

s. Reading is from

With

WRITE

at

the

location

The

value

of

Auxiliary

The

following

not

thought

words or

only

or

DISC

s may

record

in

the

input/output

the

binary
one

READ

the

group

of

Data

Conversi

statements,

the

statements,

defined

be

considered

READ/WRITE

disc

elements

words

appropriate

by

s.

Input/Output

set

of

statements

of

as

being

boundaries.

statement.

may

be

on

II)

are

the

to

As

list.

With a knowledge

restriction

thought

in

any

random

.

read

into

device.

binary

word

be

an

Statements

enables

the

broken

into

unit

many

locations

that

the

of

as a one-dimensional

order,

much

the

items

defined

values

of

the

address

relative

programmer

to

records.

of

the

disc

of

the

required

data

written

as

in

by

list

items

to

the

start

manipulate

Random

Data

by

array,

an

ENCODE

the

are

of

magnetic

Access/

is

processed

or drum

sizes
one

from

list k,

written

the

are
of

WRITE

which

or

starting

on

user1s

tapes

Auxi I

exactly

used

various

statement

DECODE

from

the

file.

and

iary

as

specified,

as

are

required

items

the

must

it

is

possible
statement

the

disc

appropriate

sequential

Input/Output

with

for

the

items

programmer is

be

input

with

to

select

an

(see

II

Memory-

location

device,

disc

defined

starting

files.

Statements

no

not

63

REWIND

This

statement

REWIND i

where

i is

Statement

is

an

unsigned

expressed

integer

as

constant

or

integer

variable.

Execution

BACKSP

The

where i is

When a BACKSPACE
record.
record

REWIND

END

This

where
mark

Sometimes,
cal

wind

the
such

of

a REWIND

ACE

Statement

BACK

SPACE

BACKSPACE i

an

unsigned

For

binary

is

interpreted

and

BACKSPACE

FILE

Statement

statement
END

i is

is

unit

their

monitor will

devices.

FILE

an

to

be

it

number

tapes

causes

i

unsigned
written.

is

statement

desirable

to

at
recognize

statement

has

the

integer

statement

tapes, a logical

as

end-of-file

integer

some

the

end

all

the

statements

to

take

other

of

this

constant

is

information

constant

device,

the

anomaly

causes

the

unit

whose

form

or

integer

executed,

record

that

are

marks

to

be

or

integer

a program

such as a I

job, it is permissible to

and

variable.

the

unit

may consist

output

by

executed

written

variable

that

has
ine

handle

the

referenced

one

on

been
printer.

logical

of

more

binary

for

tapes

the

specified

whose

written

specify

situation

unit

number is i

by

the

integer

than

one

WRITE

statement.

already

unit,

value

determines

for

output

Since

such programs

an

ENDFILE

appropriately.

to

value i is

physical

positioned

and

has

on

magnetic

or

It

be

rewound.

backspaced

record.

the

often

REWIND

is

In this

at

IIload pointll

the

form

unit

on which

tape

write

operation

not permissible to BACKSPACE

one

case a logical

have

an

and

assign

end-of-fi

on

any

logical

no

end-of-file

that

Ie and

effect.

logi-

re-

device;

64

Carriage

The first
certain

character

characters:

Character

o

If

one

of

these

left

one

position.

Any

other

record

is

printed;

printing)

Carriage

cannot

Control

in

characters

For

character

the

be

Control for

for

Printed

an

output

Effect

to

Skip
Space

two

is

example,

appearing

record

performed

Printed

Output

record

first I

lines

present,

the

as

the

remains

without

Output

that

ine

of

page

before

it

is

replaced

second

first

unchanged.

hampering

is

intended

before

printing

character

character

the

for

printing

printing

by a

blank

before

is

printed

in a record causes

This

includes

printing

the

speed on

may control

the

record

in column 2.

the

carriage

11+11

character,

all

lines.

the

is

printed.

printer

to

be

whose

carriage

The record

single

spaced

traditional

by

containing

is

before

function

not

shifted

the

(over-

7.

DECLARATION

STATEMENTS

Declaration

allocation,

Note:

[

Classification

An

The
which
way

Implicit Declarations

Unless

statements

1. When appl

2.

3.

4.

5.

6.

7.

statements

initial

All

declaration

order

given

identifier

scalar
array
subprogram
COMMON

category

the

other
An
An

parentheses.

An

list

otherwise,
An

statement

An

does

Library functions
type

may

into

the

identifier

identifier

specifically

are

implicitly

icable,

letter
identifier
identifier

identifier

enclosed

identifier

(i.e.,

identifier

appear

is not

block

(implicit

it

equivalent

are

used

values

of

statements

at

the

end

of

Identifiers

be

classified

which

an

appears

is to

be

classified.

declared

classified

an

identifier

type

that

is

called

is a function

This does

is a

statement

in

parentheses.

is

an

error.

is

classified

other

is

implicitly

in a

COMMON,

have

to

to

variables,

discussed in

of

the

as

referring

identifier

in

the

to

be

classification

with

subprogiam

not

apply

function

Again,

as a scalar

than

followed

classified

an

inherent

implicit

define

the

data

and

to

provide

this

chapter.

to

any

is

placed

and

program. These

in a

particular

according

is

a CALL

to

It must

EQUIVALENCE,

type.

to

the

integer

if

it

may

statement

if

it

appeOis in

declared

definition

also

comply

this

does

not

variable

by a

left

as a scalar

type

associated

Chapter 8 contains a complete

type

of

similar

chapter

of

the

following:

the

type

appearances

category

following

begins

be

altered

is a subprogram.

arrays.

if

it

appears

with

apply

if

it

makes

parentheses

if

it

does

or

NAMELIST

with

variables

information.

are

subject

(if

any)

constitute

or

type,

set

with

I,

by use

an

expression,

to

the

rules

to

declared

any

or in a CALL

not

them,

and

to

associated

identifiers

of

rules.

J, K,

L,

of

the

the

left

given

arrays.

other

appear

statement.
as

shown

functions,

the

rules

with

explicit

that

M,

or

N.

IMPLICIT

followed

of

an

equal

in

Chapter 8 under

appearances

statement).

in

an

executable

in

Table

description

the

dimensions

for

statement

it

depend

or

implicit

appear

by

in

It

is

real

statement).

an

argument

sign,

within

6 (see

of

these

placement

on
declarations

executable

if

it

followed

11

Statement

an

executable

or

DATA

Chapter

functions.

of

arrays,

the

contexts

begins

jist

enciosed

by a dummy

statement,

8).

storage

and

in

of

the

or

DATA

with

any

in

FunctionslJi

or

DATA

but

Inherent

Explicit Declarations

All

other

declarations

in

any

way

other

array

declarations

type

dec

larations
storage a !location
subprogram
subprogram

Conflicting and Redundant Declarations

Except

and

may not

real

where

conflict.

type,

specifically

or

are

than

those

declarati

dec

larations

definitions

For

example,

defined

explicit

noted

as

declarations.

described

a subprogram in more

ons

to

an

the

above.

contrary,

identifier

Explicit

may

Explicit

declarations

definitions

not

be

than

one

declarations

and

declarations

both

a subprogram

place,

etc.

are

include

required

of

the

name

in

order

classification

and

an

array

to

classify

of

an

name,

Declaration

an

identifier

identifier

both

integer

Statements

65

Array

Declarations

Array

dec

larations

explicitly

define

an

identifier

as

the

name

of

an

array

variable

and

have

the

form

v(dl, d2,

where

v is

the

n

is

the

d.

is

an

I

to

seven

may

Array

dec

larations

DIME

NSION
Explicit
C

aMMON

Examples:

x (10)

ARRAY (5, 15, 10)
CUBE

(4,7)

DATA

(4,3,6,12)

Array

Storage

Although

array
varies

(i.

an

elements

with

the

e.:

2-dimensiona I arrays

d3,···,

identifier
number

unsigned

be

type

array

dn)

of

dimensions (see

scalar

may

statements

statements

statements

may

in

sequence

highest

of

the

array

dimensions

integer

variables

that

instead

appear

have

frequency,

in

several

(from low address

are

stored

associated

defines

"Arrays"

of

dimensions,

the

next

"column-wise").

with

the

array

the

maximum

in

Chapter

integers

storage

leftmost dimension

3).

(see

"Adjustable

it

is

placed

toward high address

value

of

When v is

in

varies

Figure 2

the

corresponding

a dummy

Dimensions" in

storage

as a linear

with

illustrates

array

storage),

the

next

array

dimension. Arrays may

in a subprogram, d1 through d
Chapter

string.

such

highest

storage.

8). n

This string

that

the leftmost dimension

frequency,

have

contains

and

so forth

up

the

References

References
clared

for the

ber

of subscripts

to

to

Array

array

array

are

Elements

elements

(except

treated

array

A(3,

2.

Element

A(l,l,1)
A(2,l,l)
A(3,
A(1,2,l}
A(2,
A(3,
A(l,3,l)
A (2,
A(3,
A(1,1,2)
A(2,l,2)

A(3,l,2)

A(1,2,2)

A(2,2,2)
A(3,
A(l,3,2)

A(2,3,2)

A(3,3,2)

Array

of

subscripts

Item

1

2

3

I

4

5
6
7

8

9

10
11
12
13
14

15
16
17
18

Figure

must

contain
as discussed for EQUIVALENCE
as errors.

the

number

3,2)

1,

1)

2,1)

2,

1)

3,

1)

3,

1)

2,2)

Storage

corresponding

statements).

to

References

the

that

number

contain

of

dimensions
an

incorrect

de-

num-

66

Array

Declarations

Furthermore,
the

array

declaration.

the

array.

the

value

of

each

subscript should

Otherwise,

the

references

be

within

may

not

the

be

range
to

data

of

the

belonging

corresponding

to

the

dimension,

set

of

elements

as

specified

that

in

comprise

[

[

[

DIMENSION

This

statement

DIMENSION

where
a

lIocation

DIME

DIMENSION

IMPLICIT

This

statement

Dec

larati

IMPLICIT

where

each

and

c.

J

An IMPLICIT

be

overridden

Statement

the

Vi

are

of

the

NSION

Statement

l1.

ons

C.

is a type

I

type(c

type

is

INTEGER
REAL
COMPLEX

LOGICAL
DOUBLE PRECISION
DOUBLE COMPLEX

is a

character

Z,

declaration

is used

only

to

v1' v2' v3'···,

array

declarations

arrays

dec

lared.

MGO(l7), L TO(15),
AD(184),

is used

It has

l'

c2, c3,

one

single

A-G,

by

an

must

X(2,3,

to

alter

the

form

convention

...

of

the

six

alphabetic

follow

M-N,

may

explicit

,c

define

v n

the

conventions

of

m)

type

character

the

S

override

type

declaration

the

dimensions

as

described

For

example:

BB(36,22, 34)

4,5,

10), PETROL(5,6)

the form

declarations/

or

first

in

alphabetic

the

normal

of

arrays,

previously. A DIME

for

implicit

two

such

sequence.

(IJKLMN)

(see

below).

and

typing

characters

rule

As

has

from

separated

For

of

an

example,

the

form

NSION

the

IJKLMN

example,

implicit

statement

rule

by

a dash (minus

type

classification.

the

statement

does

not

discussed

affect

under

sign);

the

It, in

the

type

I1Implicit

second

turn,

or

may

IMPLICIT

would

cause

1.

Identifiers

2.

Identifiers

3.

Identifiers

integer.

4.

Identifiers

5.

All

The

statement

IMPLICIT REAL(A-Z)

would

cause

t

..

Optiona I Size

complex

other

types.

the

identifiers

all

COMPLEX(C),

following
beginning
beginning
beginning

beginning

identifiers

Specifications"

implicit
with C are
with
with

with K are

are

LOGICAL(T,

type

T, F,
H,I,J,

rea! (normal

to

be

real

later

conventions

complex.

L,

M,

or N are

or

Ware

integer

convention).

unless

in this

F,

L-N),

logical.

integer.

(normal

explicitly

chapter

INTEGER(H-J,

to

be

in

force:

The I and J are

convention).

declared

describes

otherwise.

the

W)

redundant

declaration

here,

because

of

double

DIMENSION/IMPLICIT

precision

these

and

are

normally

double

Statements

67

Whilean
For

exampl

implicit

e,

the

type

declatation

statement

may

be

redundant,

it

must not

conflict

with

any

other

implicit

type

declaration.

iMPliCIT

is ill

egal

An IMPLICIT

Explicit

These

statements

where

type

[

S.

I

REAL(A-Z)

because

Type

REAL

enclosed

N is

declared

statement

does

Statements

are

used

to

is

one

of

the

declaration/

INTEGER

DOUBLE PRECISION

is a type

ration.

type

specification

Optionally, a scalar,

in

slashes,

specification

I iNTEGER(N)

to

be

both real

not

affect

the

define,

for

may

explicitly,

that

the

purpose

take

any

is

array,

and

types

of

the

COMPLEX

LOGICAL

DOUBLE

either

the

or

of

defining

of

the

following

integer.

basic

external

type

of

COMPLEX

identifier

array

declaration

initial

an

identifier.

of a scalar,

values

forms:

library

may

functions.

They

array,

be

followed

for

the

have

the

function,

by a

variables.

form

DATA
In

or

other

is

an

constant

array

words,

decla-

list

each

For a

descdption

Note

that

REAL

initializes

DA

initializes

Exampl es

COMPLEX

LOGICAL

INTEGER

INTEGER

identifi
array
identifier/DATA
array

X,

Y,

only

Z,

TAX,

Y,

X,

Y,

of

expl i ci t type

GEORGE,

ROOT,

er

declaration

constant I ist/

declaration/DATA

of

DATA

constant

Z/3.7

/

while

Z/3.7,

3.7,

and

Z.

statements:

C3,ALPHA,CARRY(5,5),

BINARY, BOOLE(4,

NETRTE(9)/O,l,l,2,3,5,8,

PP

lists,

3.7/

4,

constant

and

their

XYZ

4,

4),

list/

function,

TRUTHF

see

13,

21/,MASS/O/

"DATA

Statement"

later

in this

chapter.

tSee

also

"Optional

68

Explicit

Type

Size

Statements

Specification

II

in

this

chapter.

An

explicit

in

combination

indicated:

type

declaration

IMPLICIT LOGICAL{L-P)
REAL

LEVEL,

with

PERCNT

the

standard

overrides

implicit

any

implicit

typing

declaration.

rule,

would

cause

Thus,

the

the

statements

following

identifiers

to

have

the

types

[

LEVEL3

LEVEL

KAPPA

POROUS

PERCNT -

X

Optional

In

addition

data.

This

*n

where

n is

the

case

of

option

is

used

Type

logical

-

-

real

-

integer

logical

-

real

real

-

Size

Specifications

to

the

standard

option

takes

the

number

integer

to change

of

and

type

the

form

bytes

logical,

ieci

to

Standard
Size

(bytes)

declarations,

occupied

only

the

double

by

the

standard

piecision

Optional
Size

an

data

size

and

(bytes)

optional

(there

is

permitted,

complex

are

form

four

to

is

provided

bytes

and

double

that

in a

word,

the

option

complex,

specifies

and

has

as

shown

eight

no

the

bits

effect.

below.

exact

in a

byte).

However,

size

of

In

this

the

Integer
Real
Complex

Logical 4

Double

cal

The
in
fies,

1.

precision

to

complex

INTEGER*4
REAL*4

REAL*8
COMPLEX*8
COMPLEX*16

LOGICAL*4

*n

modifier

Chapter

depends

In

the

data

may

8),

and

on

the

IMPLICIT

IMPLICIT REAL*8{I-K),

data

with

appear

explicit

statement,

statement,

4
4 8
8

are

identical

size

INTEGER
REAL
double
COMPLEX
doubl

LOGICAL

in

type

to

specification

precision

e compl

ex

three

kinds

statements.

as

follows:

the

*n is

INTEGER*4(A-H),

real

of

appended

16

data

with

of

16

bytes.

statements:

This

size

podtion

to

the

specification

Thus,

IMPLICIT

of

type

LOGICAL(L,

statements,

the

*n

declaration

of 8 bytes;

relative

N)

FUNCTION

to

the

word,

double

type

as

in

complex

statements

declaration

data

that

are

identi-

(discussed

it

modi-

Optional

Size

Specifications

69

2.

3.

In

the

In

explicit
both.
vidual
over

FUNCTION

REAL

FUNCTION

COMPLEX

type

When

appended

size

specification

the

* n

appl

statement,

MULT*8(X, Y,

FUNCTION

statements,

to

of

yi

ng to

the

the

*n

CNVERT* 16(C)

the

*n

can

the

type

word,

their

own.

whol e

statement.

is

appended

Z)

be

appended

the

In

other

*n

For

to

the

to

holds

words,

exampl

the

for

name

all

the

e:

of

the

type

word,

identifiers

*n

appended

function,

or

the

listed,

to

an

rather

than

identifiers

excepting

identifier

to

being

those

takes

the

type

declared,

with

an

precedence

word.

or

indi-

In

the
the

second

element

Storage

These

statements
allocation
The

storage

COMMON

EQUIVALENCE

To

make

by

each

type

Tvne

"/1-

-

integer
real
double
complex
doubl e compl
logical

COMPLEX*8

LOGICAL

first

example
example

of

array

Allocation

are

information

allocation

statement

proper

use

of

of

variable.

precision

ex

CUM,

LAUDE*

FLAG(lO), TRUTH*4(1O)

CUM

and

FLAG

FLAG,

and 4 bytes

and

16

LAUDE

TRUTH

are

per

Statements

used

to

is

provided,

statements

statement

the

storage

The

arrange

variable
the

are

allocation

following

Words

2
2

4

compiler

table

both

are

arrays,

element

storage

statements,

indicates

of

type

each

are

in

special

allocates

it
the

complex;

having

required

ways,

all

is

often

standard

CUM

10

elements.

for

array

as

variables

necessary

size

has 8 bytes,

TRUTH.

required

within

to know

associated

Four

by
the

while

bytes

the

programmer.

program

the

amount

with

each

LAUDE

are

in

type.

has

required

an

arbitrary

of

storage

16. In

for

If

no

each

storage

order.

required

COMMON

The
storage

The

where

70

Storage

Statement

COMMON

provides

COMMON

COMMON

the

w.

have

I

/c/v

where

c is
Vi

Allocation

statement

a means

statement

w1 w2 w3

the

, v2'

1

either

is a scalar,

is used to

by

which

has

...

form

v3,···,

Statements/COMMON

v m

the

array

identifier

the

wn

assign

more

form

name,

variables

than

one

of a labeled

or

array

Statement

to a region

program or subprogram may

COMMON

declaration

of

storage

block

or

is

called

reference

absent,

COMMON

the

indicating

storage.

same

blank

COMMON

data.

COMMON

When
all

other

W1

(the

first

places,

specification

blank

COMMON

in

the

statement)

is

indicated

is to

by two

specify

blank

consecutive

COMMON,

slashes. For exampl

the

slashes may

e:

be

omitted.

In

COMMON

For

each

specification
MON.
GHIA,
block

Label

Labeled

other

Any
cutable
All
needed

The
type
of a labeled

The

and

GROUP3.

ed

COMMON

COMMON

and

blank

labeled

program (see

labeled

by

variables

or

number

SUBROUTINE A
REAL
COMMON

MARKET, SENSE

(wi),

variables

COLD

COMMON

COMMON

the

program

defined

COMMON

T, V, W, X(21)

are

are

assigned

blocks

COMMON.

Chapter

blocks

require

as

between

/SET1/T,

the

/GROUP3/X,

the

variables

assigned in

to

blank

are

discrete

block

may

be

8). References

need

not

definition.

being

in a

particular

program in which

block

must

be

V,

W, X

listed

are

the

order

they

COMMON,

sections

referenced

be

identical

of

are

defined

labeled
the

SUBROUTINE B
COMPLEX

COMMON/SET

Y, JUMP

assigned

appear.

while

the

COMMON

by

any

made by

in

anyone

COMMON

block

in

all

the

G,

//

to

Thus, in

X,

number

block

program; in

is

referenced.

programs in

F(ll)

l/G,

GHIA,

the

Y,

F

COLD

indicated

and

region

of

programs

name, which must

block

COMMON

the

above

JUMP

are

and,

or

fact,

do

not

However,

which

it

example,

placed

as

subprograms

only

necessarily

is

such,

are

be

identical

those

the

definition

defined.

in

block

or

to

blank

MARKET, SENSE,

labeled

independent

that

blocks

have

For

COMMON

comprise

in

all

references.

containing

to

correspond in

of

the

overall

example:

of

an

COM-

each

exe-

data

size

references

Both
SET1

as

containing

routine B declares

Reference
ences

may
COMMON
a

single

Block names must

1.

Subprogram names

2.

Other

labeled

A
however,

Blank

COMMON

The

section

there

is

only
labeled
executable
programs
program.

SUBROUTINE GAMMA

COMMON

to

may

be

occur

statements.

labeled

COMMON,

COMMON

block

COMMON

it

is

usually

of

the

one

program (see

define

the

COMMON

24

words;

12

items

made

to

in a singl e

In both

be

unique

defined,

names

block
preferable

COMMON

such

area,

blank

Chapter

blank

COMMON

E,

D(20, 10), S

block;

the

definition

of

complex

the

name

COMMON

cases

block

of

with

respect

explicitly

may

have

to

choose

region assigned

and

empty

COMMON

8), do
areas

SET

1:

in

type.

of a labeled

statement,

the

processor links

the

appropriate

to:

or

implicitly,

the

same name

block

block

areas,

not

have

of

different

SUBROUTINE
COMMON

correspond

subroutine A specifies

COMMON

or

names

to

blank

name

specifications

defined

to

correspond in

in

block

the

block

together

name.

to

be

as

an

identifier

that

are

(or

unlabeled)

in

the

sizes,

and

ETA

R(10), N(5)

size.

name may

all

external

totally

various programs

yet

That

24

items

more

than

variables,

references

in

any

unique.

COMMON

always

size.

both may

is, both subprograms

of

real

type,

and

once

in

any

program.

be

specified

classification

refer

to

and

For

instance,

be

defined

is

not

it.

Furthermore,

subprograms

portions

in

any

as

other

discrete;

the

of

define

the

definition

Multiple

number

of

being

in

the

than

those

in

other
as

that

comprise

following two

the

same

the

block

in

refer-

individual

block,

into

above;

words,

opposed to

an

sub-

executable

sub-

COMMON

Statement

7]

Subroutine GAMMA
COMMON

Any
a

single

blank

that

number

of

COMMON

COMMON

contains

references

will

defines

a maximum

may

statement

be

placed

a minimum

of

be

made

to

or in several

together

of

202 words in

60

words,

blank

COMMON

in

the

depending

COMMON

statements. In

blank

COMMON

blank

COMMON;

on

the

types

with a program. The multiple

either

area.

of

the

case,

subroutine

variables

all

ETA
E,

D, S,

references

variables

declares a blank

R,

and

N.

may

occur

defined

as being in

in

Variables

may (see "DATA

Arrangement

Each
abies
region
in a
their

cause

in

blank

of

labeled

declared

particular

normal

COMMON

COMMON

Item

are

the

following

COMMON

arranged

sequence

1

2

3

4

5
6

7

8

9

10

11

12
13

Block E

W

X(l,l)
X(2,
X(3,1)
X(1,2)
X(2,2)
X(3,2)

X(

X(2,3)
X(3, 3)
J

COMMON

Statement"

COMMON

to

be in the

from

section

/E/W,

K,M/E/Y

arrangement

1)

1,3)

Y

may

later

block

and the

labeled

low address

is

contained

(see "Array

X(3, 3)

//T,

//C(4),

Blank

not

be

in

this

chapter).

blank

block

storage

in

the

Storage")

B,

Q/E/J

H,

N(2),

of

COMMON:

COMMON

T

B

Q

K

M

C(l)
C(2)
C(3)
C(4)

H

N(l)
N(2)

Z

initialized

COMMON

or the

toward high address

low address word or words

within

(using a

unlabeled

the

COv'\MON

area

area.

DATA

contain,

storage.

Z

statement)

The

variables

of

block

that

while

in the order of

in

each

The first

section.

or

area.

variable

For example the statements

those in

their

section

Array

labeled

appearance,

of

the

to be

declared

variables

COMMON

the

COMMON

as being

are

stored in

vari-

72

Since

a segment

be

aware

of which items in a segment

SUBROUTINE TOM
COMMON

will

define

the

Item

1 A
2
3

52
53
54

102

COMMON

Statement

of

the

COMMON

IS/A,

B(101)

block

S as follows:

TOM

B(l)
B(2)

B(51) X(51)
B(52)
B(53) Y(2)

B(101) Y(50) Y(50)

DICK

A
X(l)

X(2) ALPHA(3)

Y(l)

region

contain

SUBROUTINE DICK

COMMON

COMMON

HARRY

ALPHA(l)
ALPHA(2)

ALPHA(52)

Y(l)
Y(2)

may

be

defined

certain

IS/A,

X(51)

/S/y(50)

differentl

variables.

y in

each

For

example,

SUBROUTINE
COMMON

COMMON

/S/ALPHA(52)

program,

HARRY

/S/y(50)

it

may be

quite

important to

which allows
to

access

the

assume

A,

B,

the
array

X, Y,

routine

variable

and

TOM

Y by

ALPHA

and

are

DICK

that

of

to

access

identifier,

the

same

the

and

type).

variable

yet

the

A by

integrity

that

of

identifier,

the

block S is

the

routines DICK

maintained

and

(these

HARRY

examples

Referencing

Incorrect
MON

the

reference
block
following

The FLAG system
ing a
variables
variables

the

of

Data

in

COMMON

referencing

blocks

must

number

to

the

first

chart:

Type

of

Variable

Integer
Real 4
Double Precision
Complex

Double Complex
Logical

variable

in

the

R,

I, and CPX

COMMON

displacement

be

constructed

associated

storage

automatically

of

any

length

same

block,

R,

I,

CPX

of

these

of

COMMON

with

location

Reference

to

be

it

may

are

REAL,

variables

data

so

that

the

of

the

Number

4

8
8
8

4

begins

the

first

be

necessary

INTEGER,

within

will

terminate

the

displacement

variable

variable).

every

COMMON

assigned

and

the

block

execution.

of

(displacement

The

reference

within a block.

to

insert

an

COMPLEX,

is

as

shown

each

variable

is

the

number for

block

as

unused

respectively,

below:

To

ensure

number

if

its

To

obtain

variable

correct

in

the

of

type

specification

the

in

and a COMMON

block

bytes

of

correct

the

referencing

is

from

variable

block.

of

an

integral

the

beginning

is

shown in

were

8,

thus

displacement

For

example,

block

data,

COM-

multiple

of

the

allow-

for

other

if

is

defined

of

the

the

as

-t

R

+

-!

displacement

4

bytes

displacement

4 bytes

displacement

1

I

8

displacement

of

length

of 4 bytes

bytes

for I

4),

Statement

CPX

_I

The

displacements

(instead

with a length

proper

displacement

EQUIVALENCE

and

the

displacement

would

for

CPX.

= 0 bytes

= 4

bytes

= 8

bytes

= 16

bytes

CPX

are

have

evenly

of

to

be

divisible

CPX

would

inserted

by

their

reference

be

12,

which

between R and

numbers.

is

incorrect.

I or

between I and

However,

In

that

CPX to

if R were

case,

provide

an

extra

the

REAL

word

*8

The EQUIVALENCE

assign more

than

EQUIVALENCE

statement

one

variable

sl'

s2'

to

s3'···'

controls

the

same

sn

the

allocation

storage

of

location

variables

or

locations.

relative

to

one

It

is expressed

another.

as

EQUIVALENCE

Generally,

it

is used

Statement

to

73

where

each

of

(v1' v2' v3'···' v m)

Each

equival

the

foil owi ng

1.

A

scalar

2.

An

array

DIMENSION

REAL

EQUIVALENCE

would

make

When

multiple

DIMENSION

3.

An

array

count

is

ately

following

two

elements

of

X (see IIAllocation

REAL

EQUIVALENCE (A(7),

would

make

the

s.

is

an

equivalence

I

eliCe

set

sPedfi

eS

three

forms: I

or

array

name.

element,

B,

C,

where

A(3,

A,

3)

X(ll)

(A(l,3),

Band

A(l,3)

subscripts

name

as

follows:

statement

followed

in

the

is position

beyond X and

of

B,

C,

A(3,3)

A(l,3)

and B equivalent.

that

all

For

arrays,

the

subscripts

B),

equivalent,

are

to

be

which

by

an

unsigned

location

2;

and

refers

Variable

B)

set

of

the

v.

the

(C, X(l»

and,

used

the

array

of

the

so

on.

to

X(3, 1),

Types

the

form

aie

to be assigned the same

location

are

referenced

unsigned

integers.

similarly, C and

in

an

EQUIVALENCE

is

declared.

integer
first

ll

Thus,

).

element

element

if

X is a 3 x 3

where

of

the

size

count

the

is

that
For

X(l)

equivalent.

statement,

enclosed

array

is

array,

(in words)

storage

of

the

example,

that

in

denoted

X(l)

of

an

location.

first

element.

the

statements

statement

parentheses.

as

position

means

the

element

The

must

The

1;

the

same

as

is

dependent

v.

may be one

I

be

preceded

meaning

element

X(l,

1); X(3) is
on

by a

of

this

immedi-

the

type

of

See

also

that

cannot

Example:

The

is to

Word

2

3

4

5

6

7

8

9

where

IIInteractions

be

implemented.

effect

of

the

statements

DIMENSION
REAL

W, X
INTEGER
REAL

* 8

LC,

ELSIE

COMPLEX C

EQUIVALENCE

cause

the

indicated

Variables -Set

LC(l)
LC(2)

=

W(l) = ELSI

LC(3) = W(2) =
LC(4) = W(3)

I

rlc\

L\..\J}

LC(6)
LC(7)

the

arrangement

of

Storage

W(3),

Allocation

X(3,3),

J

(W,

LC(2), ELSIE), (X(6),

equivalences:

1

E1

ELSI

E2

of

set

LC(7)

Variabl

X(l,
X(2,
X(3,
X(l,2)

X(3,2)
X(l,3)
X(2, 3)
X(3,3)

1 has no

Statements

es -

1)

1)

= C 1

1)

=

C2

= J

bearing

J,

Set

on

ll

,

below,

C(3»

2

the

arrangement

for

further

of

set

rules

2.

concerning

equivalences

74

EQUIVALENCE

Statement

The

statement

EQUIVALENCE (LC(2),

has

the

same

same

as

the

Interactions

No

storage

and

EQUIVALENCE

thi

s is

the

contradictory

In

all

EQUIVALENCE
allocations
blocks

It

is

upper

EQUIVALENCE

item

allocation

EQUIVALENCE

not

followed,

cases,

the

of

COMMON

can

not

permissible

bound

established

declared

COMMON

DIMENSION

to

results

as

set

(J, X(6))

of

Storage

declaration

statement

set

should

such

reference

storage

specifications.

be

for

declaration

be

/BLK1/A(5),

allocation

made

equivalent.

an

EQUIVALENCE

by

in

that

Z(lO),

W),

the

EQUIVALENCE

in

this

Allocation

is

determines

contain

references

storage;

V(5)

references

is

not

allowed.

Consequently,

that

the

last

to

cause a segment

segment.

B/BLK2/E(4),

(W(l),

case.

Statements

permitted

the

are

either

sequence

is,

two

to

cause a segment

item

defined

Both

conditions

ELSIE),

H,

(C(3),

statement

to

cause

allocation

to

more

than

redundant

specified

EQUIVALENCE

variables

to

be

to

be

lengthened

are

Y(2,2)

J), (J, X(6))

in

the

previous

conflicts

or

in a COMMON

in

in

demonstrated

in

of

the

variables

one

variable

contradictory.

statements

the

same

of

the

COMMON

that

segment.

beneath

the

example,

arrangement

referenced

than

has

The

statement

are

COMMON

However,

the

in

the

and

previously
redundancy

takes

not

allowed

block

region

lower

bound

exampl

the

of

storage.
in

or

to

it

is

es

below.

set

(J,

Each

them

Therefore,

been

allocated.

is normal I y

precedence

to

define

in

different

be

lengthened

not

permissible

established

X(3,

2)) is

COMMON

ignored;

over

conflicting

COMMON

beyond

by

the

any

for

the

no
When

the

an

first

EQUIVALENCE (A,

The

first EQUIVALENCE

an

extension

Item

of

BLK1

A(l)

2

3

4

5

6

7

8

A(2)

A(3)

A(4)

A(5)

B

the

block

==

Z(l)

==

Z(2)

==

Z(3)

==

Z(4)

==

Z(5)

==

Z(6)

Z(7)

Z(8)

Z),

(V(4), E(2))

set

is a permissible

BLK2.

The

B

LK2

E(l) + V(3)

E(2)

E(3)

E(4)

H

Y(l,

Y(2,

Y(l,2)

declared

(i

II

ega I extension)

V(l)

V(2)

==

V(4)

==

V(5)

1)

1)

extension

storage

of

the

block

allocation

BLK1,

would

whereas

appear

as

the

second

shown

below.

set

illegally

defines

9

10

Note:

Assume

all

Z(9)

Z(lO)

items

are

of

Y(2,

the

2)

same

data

type.

Interactions

of

Storage

Allocation

Statements

75

The

fact

that
requirement
programs

comprising

COMMON

that

labeled

segments

COMMON

portions

of

an

may

be

lengthened

blocks

executable

of

the

program,

by

EQUIVALENCE

same

name,
contain

which

the

dedarations

are

defined

identical

in

number

in

no

separate

of

words.

way

nullifies

programs or

the

sub-

EXTERNAL

The EXTERNAL

where

The EXTERNAL
so
example,

but

The EXTERNAL

Example:

Library
(e.
subprogram.

L

the

p.

that

they

if

CALL

appears

EXTERNAL F

functions

g.,

SIN)

Statement

statement

are

subprogram

I

statement

may

be

passed

the

subprogram

ALPHA(F)

in

no

other

statement

(see

Table

appears

in

has

the

identifiers.

declares,

as

arguments

name F appears

context

to

can

be

8) may

an

EXTERNAL or

form

as

a subprogram, names

to

other

in

the

indicate

used

not

to

avoid

appear

that

as

explicit

subprograms

statement

it

is a

subprogram,

this.

arguments

type

statement,

that

might

otherwise

(see

II

Arguments

it

would

to a subprogram.

it

must

refer

be

and

Dummies

be

implicitly

If

the

to a variable

classified

ll

classified

name

of a library

implicitly

in

Chapter

or a user-supplied

as

scalars,

8).

as a scalar.

function

For

[

BLOCK

FLAG
program,

and
subprogram must
programs,

BLOCK DATA subprograms must

Within
more DATA

When initiaiizing variabies

in

1. The

2.

Data

DATA

Subprograms

permits

BLOCK DATA

may

each

The

may

variables

which

begins

contain

be

they

have

a BLOCK DATA subprogram,

statements;

COMMON

position

size

of

each

be

entered

only

declaration

terminated

no

block,

within

COMMON

into

in

with a

names

type

the

labeled

statements

in

so

block

more

COMMON

statement

statements

with

an

nor

are

appear

initialization

END

they

of

the

executed

before

may

to

be

form

(described

statement.

the

main

of

not

be

used for

initialized

in

this

Since

in

the

usual

program

labeled

initialization.

in a

special

chapter)
BLOCK DATA subprograms may

sense.

and

all

COMMON

iabeied COMMON, compiete deciarations

that

of

those

block

than

will

one

variables

correspond

COMMON

that

to

block

are

the

in a

being

size

single

initialized

declared

BLOCK DATA subprogram.

program

and

DATA

other

variables

snouid

in

called

statements

subprograms.

must

be

be included

will

be

correctly

all

other

programs

a BLOCK DATA

described

not

accomplished

for

established.

below.

be

called

oJ;

the variables

that

use

by

it.

sub-

by

one

The

other

or

76 EXTERNAL

Statement/BLOCK

DATA Subprograms

DATA

The

DATA

where

Statement

statement

has

the

form

S. is a

I

The primary purpose

3.

141592653589793
instead
required.

Giving
With
ceives

Consider
written
through
array

Here,
to
to
practice

of

the

its

using a

must

C(O)
C(l) . 1243549945
C(2)

etc.

the
give
depend

data

variabl

the

longer

PI a value

DATA

statement

value

another

power

the

constants.

be

initial

0

.2447786631

DATA

"initial"

on

loading

dictates

at

exampl e

statement

values

that

set

specification

e-I i st/constant-I i st/

of

the

DATA

at

every

appearance,

form

of

the

with a

DATA

statement

the

value

execution

series

ized

and

such

time.

that

profits

expansion.

But

constants

each

time

can

to

variables

disallows

initialization

be

of

statement

constant.

is

assigned

even

cannot

into

used

to

that

restarting

the

form

is to

the

variable

This

also

is somewhat

when

more from

The

efficient

be

subscripted,

the

routine

great

advantage.

are

going

the

be

done

give

names

to

PI

can

be

simplifies modifying

different

the

program is

the

use

of

way

to

program this in FORTRAN is with a

and

using

assignment

It is

to

be

changed.
program
with

once

executable

constants:

given

from

giving

loaded;

the

DATA

the

timing

statements,

not

recommended

This

it

has

changed

statements,

for

that

value

the

it a value
with

statement:

of

causes

example,

with a

program,

the

the

routine

such

that

proper

these

e.g.,

instead

if

with

assignment

An ARCTAN

as:

initialization

values.

with assignment

DATA

a more

an

is

the

of

statement

accurate

assignment

statement,

DO

adversely

DATA

Good

referring

function

statement

of

programming

to
and used
value

statement.

PI

loop,

stepping

affected

the

program

statements.

re-

can

be

rr as

is

be

if

an

used

The

effect

set,

in

DA

is

equival

X =

A = 7

L = .

ALPHA = 0

except

Variable

DATA

Variable

A

DATA

scripted

of

the

order

TAX,

ent

3.5

TRUE.

that

the

and

variable

arrays.

the

DATA

listed.

A,

L/3.

to

the

DATA

constant

List

list is

It may

statement

For

example,

5,

7, •

TRU

assignment

statement

lists in

similar

not

contain

is

to

initialize
the

statement

E./ , ALPHA/O/

statements

is

not

executable;

DATA

statement

to

an

input

list

implied

the

variables

its assignments

may

be

constructed

(see

Chapter

DO

loops. Subscripts must

6),

in

in

each

take

as

described

that

it

data

place

may

be

set

to

upon

in

the

contain

integers.

the

values

loading.

following two

scalars

of

the

sections.

or

subscripted

constants

or

in

the

unsub-

DATA

Statement

77

OAT A Constant

List

[

A DATA

where

the

constant

C.

are

either

J

c

r*c

where

c

is a signed

is

is

The

constant

may

also

following

1.

Integer,

2.

Logical

3.

literal

is
variable
characters, a complex

be

rules

broken

may

used.

apply

real,

constants

constants

up on a

(see

list

is

of

the

constants

or

unsigned

an

unsigned

to

be

be

The

double

"Storage

any

in DATA

may

integer

repeated.

of

the

type

of

precision

may

be

be

used

character-by-character

Allocation
variable, 8 characters,

form

or

repeated

constant

repeat

forms

described

the

constant

statements:

and

complex

expressed as .

with

any

Statements"

groups

of

count,

in

must

TRUE.

type

basis

of

constants

an

appropriate

whose

Chapter

be

the

same

variables

and.

FALSE.

of

variable,

and

depends

earlier

and a double

value

2,

as

may

although

in this

in

the

following

type

(nonzero)

including

the

type

be

initialized

or

abbreviated

on

the

chapter).

complex

forms:

(see

below).

indicates

literal

constants.

of

the

variable

with

as T and

integer

is

recommended. A literal

nu~ber

of

That

variable,

the

constants

words

is,

an

16

characters.

number

integer

of

Hexadecimal

that

it

of

those

F.

of

storage

variable

times

is

initializing.

types.

occupied

the

group

constants

constant

by

the

requires

The

4

I

I

!

Variable
acters

4.

Numeric

initialize

5.

Hexadecimal

the

As

The maximum

ized.

If

the

number

digits

is

items

will

be

initialized

in

any

literal

constant

and

logical

successive

constants

character Z followed

0123456789ABCDEF

an

example,

The following list shows the maximum number

Type

of

Variable

LOGICAL
INTEGER
REAL
REAL*8
COMPLEX 16
COMPLEX*16

of

digits

less

than

the

the

number

is

maximum,

constants
list items

may

by from 1

hexadecimal

of

digits

greater

as

required

to

fill

the

last

may

not

be

used for more

or

successive

be

used

to

initialize

to

32

hexadecimal

constant

allowed

Maximum number
Hexadecimal

than

the

hexadecimal

ZBOD

in a

Digits

8
8
8

16

32

maximum,

zeros

to

use up

variable

elements

hexadecimal

any

represents

of

the

are

supplied

used,

of

type

of

leftmost

the

than

an

of

digits.

the

constant

digits

characters
it

will

be

one

variable

array

appearing

variable.

These

bit

string

depends

for

each

hexadecimal

to

the

left.

specified.

filled

out

list

as

a list

The form

digits

are

101100001101.

on

the

variable

digits

If

there

are

with

trailing

item;

one

literal

item.

of a hexadecimal

type

of

variable

type:

are

truncated;

insufficient

blanks.

constant

const.ant is

being

if

the

char-

may

initial-

number

of

78

DA

T A

Statement

The

following

INTEGER MM(3)

COMPLEX

DATA

The

above

examples

C1,

C2

MM/ABCDEF

DATA

statement

illustrate

1

,

IGH

causes

some

of

1

/,

C1,C2/(17.8,

the

following

the

features

-4.0),

assignments

described

(17.8,

-4.0)

to

above:

be

made:

MM(l)

MM(2)
MM(3)
C1
C2

The

constant

Dummy
labeled

Placement

The

following

or

subprogram when using

1.

All
program.

2.

Declaration

4HABCD
2HEF
2HGH
(17.8,
(17.8,

list must

variables

COMMON

and

and

rules

declaration

statements

subprogram
IMPLICIT
type

statements
DIMENSION
COMMON
EQUIVALENCE
EXTERNAL

DA

T A

statements

NAME

LIST

-4.0)

-4.0)

completely

variables

may

be

Order

govern

FLAG.

statements

declaration

statement

statements

statements

statements

statements

statements

satisfy

in

blank

initialized,

of

Declaration

the

placement

must

appear

(if

present) should

statement

the

variable

COMMON
but

only

in a BLOCK DATA subprogram.

Statements

and

order

prior

to

appear

list

and

cannot

of

appearance

the

appearance

in

the

there

be

initialized

following

may

not

of

declaration

of

the

order

be

with

first

within

any

remaining

the

DATA

statements

executable

a program:

unused

statement.

within

statement

constants.

Variables

a main program

within

in

a

Failure

3.

Identifiers
values
statements.

4.

Identifiers
initialization
initialized

to

follow

that

specified

appearing

specified

by

this

appear

within

means

order

both

the

in

within

of

one

may

type

type

result

in

type

statementi

statements

the

type

or

more

in

one

statements

they

within

statementi

DATA

statements

or more

compiler

and

in

EQUIVALENCE

may

be

a BLOCK DATA subprogram may

all

identifiers

(or

diagnostic

initialized

left

uninitialized).

Placement

messages.

statements

by

one

or

within

a BLOCK DATA subprogram must

and

may not

more

not

Order

have

subsequent

have

values

of

Declaration

initial

DATA

for

data

data-

be

Statements

79

A

complete

program

as

described

Main

set

of

program

consists

in

of

this

Programs

units

one

main program

section,

or

may

8.

PROGRAMS

executed

together

and

all

required

be

preprogrammed

AND

SUBPROGRAMS

as a single

subprograms. Subprograms may

and

job

is

contained

called

in

the

an

run-time

executable

be

or

program. An

defined

system

by

libraries.

executable

the

programmer,

A main
one

Main

Once
in

the

Subprograms

Subprograms

and

[

program

of

the

a

FUNCTION
a SUBROUTINE
a BLOCK DATA

programs may

an

main

subroutines.t

Functions

Statement

FUNCTION

Basic

Subrouti nes

is

following

executable

program.

are

programs

functions

external

comprised

statements,

statement

statement

statement

contain

These may

subprograms

functions

any

program has

which

of a set

and

statement

been

may

be

further

of

FLAG

statements,

the

last

of

which

except a FUNCTION,

loaded,

be

classified

called

execution

by

other

as

is

an

of

programs;

follows:

the

first

of

which

END

statement.

SUBROUTINE, ENTRY,

the

program

they

begins

fall

(other

into

with

the

than

the

two

comment

or

BLOCK DATA

first

executable

broad

classes

lines)

cannot

statement

of

functions

be

statement.

SUBROUTINE

[

A

function
Subroutines
library

Statement

Statement
the

form

where

tThe BLOCK DATA subprogram,

80

Program

functions

l'

f(d

f is
d.

is

I

e is

is

referenced

are

Functions

functions

d2, d3,

the
the
an

arithmetic

and

subroutines

by

referenced

and

subroutines

are

functions

•••

, d

n)

= e

name

of

the

identifier

Subprograms

of
or

the

appearance

with CALL

are

that

function

a dummy

logical

which

of

its

statements

included

can

scalar

expression

in FLAG. These

be

defined

variable

is

neither a function

identifier

and

do

in

(see

within

not

necessarily

a singl e

below)

nor a

an

expression

return a value

are

described

expression. A statement

subroutine,

at

is

and

the

also

returns a

end

provided

(see

of

value

Chapter

this

chapter.

function

(see

(see

Chapter

5).

A number

definition

Chapter

2).
of

has

7).

A

statement
scalars;
e should
this

would

fined

Examples:

F(X)

EI(THETA) = CMPLX(COS(THETA), SIN(THETA))

AVG(PT,

Since

same

as

X,

and
expression
the

same

ment

function

The

statement

does

not,

A

statement

definitions

function

they

cannotbedummy

contain

be a recursive

functions.

at

must

least

one
definition.

= A * X ** 2 + B * X + C

NUM,

TOT) = 3 *(PT +

each

di is

merely

a dummy

the

names

of

other

there

is

another

refers

name

is

dummies using

function

it

will

function

must

precede

variable

to

the

dummy. The onl y

that

they

itself

acquire

may

all

have

at I east

arrays

or

subprograms (see

reference

variables

will

both

explicit

is

typed

implicit

be

referenced

executabl e statements

NUM)/TOT

and

in

in

the

have

(or

like

type

to

References

does

the
program

implicit)

(see

only

one

dummy

each

not

actually

program.

relation
the

same

any

other

"Implicit

within

argument.

"Arguments

dummy. The

to

other

identifier

statement

+ 1

exist,

the

Note,

however,

called

X,

then

between a statement

type.

This

enables

type

statements.

identifier:

Declarations"

the

program

in

the

program

Statement

and

functions

names

the

appearance

it

may

in

unit

in

which

Dummies" in

f may

of

statement

that

if a statement

function

the

programmer

appear

Chapter

in

which

they

function

not

appear

may

be

of X within

dummy

in

an

explicit

7).

it

is

defined.

appear.

dummi es

this

chapter).

in

made

onl y

function

function

and

to

declare

are

the

expression,

to

previousl y de-

dummies may

dummy is

the

statement

any

other

the

type

statement;

Statement

treated

The

quantity

types

function

onl y

as

expression

since

be

the

named

function

with

of

state-

if

it

[

FUNCTION

Functions

grams

are

or

where

f

d.

I

type

Subprograms

that

cannot

be

defined

introduced

is

the

is a dummy

in

this

INTEGER

REAL

DOUBLE PRECISION

identifier

chapter.

is

an

by a

argument

optional

FUNCTION

of

type

in

a singl e

the

function.

of

any

specification,

statement

statement,

of

the

of

forms

which

COMPLEX

LOGICAL

DOUBLE COMPLEX

may

the

form

(except

may

be

defined

asterisk),

be

any

of

the

as

FUNCTION

described

following/

subprograms.

in

"Arguments

and

These

Dummies

subpro-

ll

later

Every

FUNCTION

FUNCTION

A

FUNCTION

statement

FUNCTION.

tSee

in a FUNCTION

also

"Optional

subprogram must

subprogram,

subprogram must

Size

with

Specifications"

have

certain

contain

subprogram;

at

least

restrictions

at

least

i.

e.,

it

in

Chapter

one

(see

one

RETURN

should

dummy.

"Arguments

be

the

7.

Values

statement. A RETURN

last

may

and

Dummies").

statement

be

assigned

executed

to

dummies

statement

for

each

within

should

execution

Subprograms

be

the

the

of

last

the

81

The

identifier

statement, a DO control

of the function must be assigned a

variablef the

statement) within the subprogram.

variable

value

on the

at

left

least

side of

once

in the subprogram as the argument of a

an

arithmetic

statement, or

in

an

input list

CALL

(READ

Within the function the
should be assigned a
one

assigned to its

A FUNCTION subprogram may

TION

statement, or a

identifier

value

identifier

BLOCK

during

prior to the

contain

DATA

FUNCTION statement examples:

INTEGER FUNCTION DIFFEQ
REAL

FUNCTION IOU

FUNCTION

EXTRCT

LOGICAL FUNCTION

(W,

(N,A,

VERDAD(E,

FUNCTION subprogram examples:

COMPLEX FUNCTION GAMMA
COMPLEX

Z

M = 1

GAMMA = Z

DO 5 J =

N,

10

M = M * J

of a FUNCTION subprogram
each

execution

execution

any

of

of a

FORTRAN

the function.

RETURN

statement

statement.

(R,

S,

N)

X,

Y, Z

1,

Z2)

B,

C,

V)

F,

G,

H,

P)

(Z, N)

is

treated

The

statement.

except

as though it were a

value

returned for a FUNCTION

scalar

a SUBROUTINE statement,

variable

is

another

and

the last

FUNC-

5 GAMMA = GAMMA * (Z + J)

GA,V"VA

...:.

M * N + Z / GAMtv'A

RETURN

END

SUBROUTINE

SUBROUTINE subprogramsr like

Subprograms

FUNCTION subprogramsr are

self-contained

programmed procedures. Unlike

FUNCTIONS, however SUBROUTINE subprograms do not have values associated with them and

enced

in

an

expression. Instead,

SUBROUTINE subprograms begin with a SUBROUTINE statement

,

SUBROUTINE p(d

,

d

1

2

d

,.·.,

3

SUBROUTINE

dn}

subprograms

are

accessed

of

the

form

by

CALL

statements (see

or

SUBROUTINE p

where

p

is

the

identifier

is a dummy

d.

I

Note

that

while a FUNCTION must have

A SUBROUTINE subprogram must
the

last statement in a SUBROUTINE subprogram (that is, it should be the last statement

of

the SUBROUTINE).

of

the subroutine

argument of

contain

Elny

of the

forms

described in IIArguments

at

least one dummy, a

at

least one

RETURN

SUBROUTINE

statement; a

and

Dummies

need have none.

RETURN

statement should be logically

ll

later

executed

may

Chapter

in this

for

not be

chapter.

each

refer-

5).

execution

A SUBROUTINE subprogram may return values to the

in

COMMON

storage. I

82 Subprograms

calling

program

by

assigning values to the d. or to variables

A SUBROUTINE

ROUTINE

arguments

in

[

Arguments

statement,
to

the

SUBROUTINE

subprogram

return

va

and

Dummies

and/or

lues

to

program.

may

contain

a BLOCK DATA

the

ca

Iling

any

program.

FORTRAN

statement.

The SUBROUTINE

statements

The SUBROUTINE

except a FUNCTION

name

subprogram

must

not

appear

statement,

may use

in

one

any

another
or

other

SUB-

more

of

statement

its

Dummy
Both

FUNCTION

however,

indicate

age

is

reserved

any

of

the

a

scalar

an

an

an

a

statement

a

constant

a

subprogram

A dummy

a

scalar

arguments

while a FUNCTION

the

following:

array

array

expression

itself

provide

and

type,

for

variable

element

name

label

of

any

name

may

variable

SUBROUTINE subprograms

number,
it;

it

type

be

a means

and

is

only a name

(including

classified

of

must

have

sequence

within

passing

at

of

used

literal)

the

information

may

least

one.

subprogram

to

identify

subprogram

between a subprogram

have

dummy

arguments.

Dummies

arguments.

an

as

argument

one

of

are

merely

A dummy

in

the

IIformal

the

calling

following:

and

the

program

A SUBROUTINE

ll

parameters

does

not

actually

program.

need

and

exist,

An

that
not

are

and

argument

ca

have

used

no
may

lied

any,

it.

to

stor-

be

[

Table

an

array

a

subprogram

an

asterisk

8,

below,

Argument

scalar

expression
statement
array

I

iteral

subprogram

correspondence

tA

denoting a statement

indicates

or

array

label

name

constant

name

Table

element

of

the

8.

this

labe

permissible

Permissible

kind

may

I

kinds

of

correspondence

Correspondences

scalar

yes yes
yes
no

t

yes

t

yes
no no

not

be

entirely

meaningful

between

Between

array

t

no
no no
yes
yes

an

Arguments

Dummy

subprogram

(see

"Dummy Arraysll).

argument

and

Dummies

no
no

no
no
yes

and

a dummy.

asterisk

no
no
yes
no
no
no

Arguments

and

Dummies

83

A

statement

where
(as

opposed

&k

k is

label

the
to

an

argument

actual

integer

is

written

statement

constant).

label

as

and

the

ampersand

distinguishes

the

construct

as a statement

label

argument

Within a subprogram,
fier

with

certain

restrictions;

COMMON

EQUIVALENCE

DATA

NAME

LIST

The

reason

for

the

as a scalar,

implicit

In

general,

situation

an

array,

and

explicit

dummies must

is in

error

COMPLEX C

LOGICA.L L

X = F (C,

Reversing
would

eliminate

the

L)

order

the

a dummy may

namely,

above

restriction

or

a subprogram

classifications

agree

because

the

FUNCTION

LOGICAL

COMPLEX

of

either

the

error

in

this

be

dummies may

is

identifier

(see

in

type

types

of

LL

CC

arguments

example.

used

in

much

that

dummies

occurs

"Classification

with

the

the

arguments

F (LL,

CC)

in

the

the

same

not

appear

do

not

in

the

of

arguments

and

calling

reference

way

as

in

the

actually

same

manner

Identifiers"

to

which

the

dummies

any

other

following

exist.

in

Chapter

they

do

or

the

scalar,

types

Furthermore,

as

with

other

7).

correspond.

not

agree.

dummies

array,

of

statements:

classification

(actual)

For

in

the

or

subprogram

identifiers,

example,

FUNCTION

of

a dummy

the

following

statement

identi-

in both

There

are

two

exceptions

i.

A

statement

2.

A SUBROUTINE

All

arithmetic

then

placed

argument

(this

is passed (this is

NOTE:

All

number passed

name

or

logical

in a temporary

action

is

referred

constants

to a constant

Dummy Scalars

Dummy

are

Dummy

A dummy
subprogram
array).

not

Arrays

scalars

declared

argument

(the

A dummy

are

single

(implicitly

may

fact

array

The subprogram assumes

actual

array

and

calculates

to

the

rul e

as

(as

opposed

expressions

storage

referred

are

in

to

as

passed by

the

calling

to

IIcail

valued

or

explicitly)

be

defined

that a calling

does

not

that

the

argument

subscripts

of

an

argument

to a

appearing

location.

as

"call

by

name;

sequence,

entities

as

an

argument

actually

from

type

correspondence:

has no

FUNCTION

by

value").

name

II).

therefore,

that

that

correspond

to

be

arrays

array,

by

is

an

occupy

supplied

that

location.

as

The

the
array

any

in

type.

name) has no

actual

address

For al!

if

the

constant

to a

or

subprograms

presence

does

storage,

the

calling

arguments

of

that

other

called

will

be

single

of

not

in

it

statement

type.

in

the

temporary

arguments

subprogram
changed.

element

are

treated

its

identifier

itself

merely

identifies

calling

storage

in

define

defines

program

location

the

actual

stores

into

Obviously,

the

calling

as

scalars.

in

any

array

the

corresponding

an

area

the

first (or base)

are

first

evaluated

is

then

passed

address

a dummy
this is not

of

the

argument

corresponding

recommended.

program. Dummies

declaration

within

dummy to

in

the

calling

program.

element

as

be

of

and

the

that

the

an

an

84

Arguments

and

Dummies

Normal!
necessary,

y, a dummy

however,

array

and

is

given

useful

the

same

operations

dimensions

can

often

as

the

be

performed

argument

by

array

making

to

which

them

it

corresponds.

different.

For

This is

example,

not

DIMENSION

CALL OUT

In

this

case,
elements
to

the

Arguments

secutive
in

the

A(l,6)

the

first

CALL OUT(A)

dummy

that

string

the

example

CALL FOR('PHILIP MORRIS')

following

M(l)

= 4HPHIL

M(2) = 4HIP-bM
M(3) = 4H ORRI
M(4) = 4HSb'h6
M(5) is

A(lO,

(A(l,6))

the

1-dimensional

through

element

array B would

of

the

are

literal

of

characters

correspondences

undefined

10)

dummy

A(lO,

10)).

array,

correspond

constants

in memory,

hold:

and

should

SUBROUTINE OUT

DIMENSION

array B corresponds

if

the

to

are

not

However,

calling

the

normally

be

since

statement

first

half

and

its

starting

SUBROUTINE FOR(M)
DIMENSION

referenced

(B)

B(50)

an

of

the

received

to

array

were

array

by

dummy

location

M (5)

the

last

name

used

A.

arrays. A literal

is passed

half

of

without

as

the

the

2-dimensional

subscripts

constant

argument

as

an

is

address.

array

argument

stored

For

A (i. e.,

refers

as a con­instance,

where"b
boundary
than
For this

If

element
For
of
be

Adjustable Dimensions

Since
to
program in
scalar
array

represents

(multiple

one

word

reason,

an

array

of

example,

the

array

taken

in

a dummy

allocate

variables

each

DIMENSION

= SUM(P,

X
Y =

SUM(Q,

the

of

per

elementt (e.g.,

integer

corresponds
the

array.

if

the

calling

other

than

creating

array

storage

the

time

for

normal

rather
the

subprogram is

P(lO,

10,5)

9,3)

character

four

arrays

to

This is

the

correspondences

does

them.

manner.

than

5),

blank.

characters).

are

recommended.

something

true

whether

argument

first el

ement

not

actually
Therefore,

Instead,

by

constants.

called.

Q(9,

3)

double

that

is a

of

occupy

Literal

Therefore,

precision)

is

not

an

the

calling

scalar

and

wi

II

correspond

this

nature

any

the

dimensions

any

or

all

This

permits

The

following

FUNCTION
DIMENSION

constants

passing such a

may

result

array

or a literal

argument

the

dummy is

to

whatever

since

they

storage,

of

a dummy

the

dimensions

the

calling

statements

SUM

R(N,

are

fill

in

is

may

its

dimensions

(R,

M)

ed

out

constant

dummy

constant,

an

array

an

array,

happens

depend

array

of

a dummy

program

demonstrate

N,

M)

with

trail

to

a dummy

elements

the

and

the

references

to

be

upon a

are

used

do

not

array

to

supply

adjustable

ing

of a type

that

are

latter

will

dummy is

in

stored

particular

only

have

to

may

the

blanks

to

that

only

partially

correspond

not,

the

subprogram

near

the

implementation.

to

locate

be

defined

be

specified

dimensions

dimensions:

the

nearest

occupies

or

vice

scalar.

its

elements,

within

by dummy

of

to

to

the

word

more

defined.

the

first

versa.

elements

Care

not

the

sub-

dummy

must

tSee

"Allocation

of

Variable

Types"

in

Chapter

7.

Arguments

and

Dummies

85

Only
scalars.
be

When
the

[

the
is

Dummy Subprograms

a dummy

The

changed.

running
subprogram
range

of

done.

array

variables

in

the

is

entered,

the

array.

can

be

used

"debug"

If

given

adjustable

as

adjustable

mode (see

and

subscripted

a

one-dimensional

dimensions,

dimensions

Chapter

elements

dummy

9)

may

the

of

array

and
be

size

of a variably

the

array
is

the

dimensions

referenced

are

checked

dimensioned

must

elsewhere

dimensioned

to

make

with

size

be

in

1 no

specified

the

subprogram

array

sure

that

subscript

by

is

calculated

the

dummy

but

subscript

range

integer

should

each

is in

checking

not

time

[

A dummy

that
the
specified
"Classification

Example:

A

it

Library

FLAG
special
are

subprogram

can

do

so.

The dummy

calling

subprogram
may

program.

as

the

argument.

of

Identifiers"

EXTERNAL

A

=

DIFF(SIN,

B = DIFF(SQRT, DSQRT, Y)

(The

be

SIN,

programmer

identifier,

classified

Subprograms

includes a number

also

machine

provided.

codes

must

correspond

name

merely

Therefore, a calion

A dummy

in

Chapter

DSIN,

SQRT, DSQRT

DSIN,

X)

must

provide

to

be

passed

as a scalar

for

of

them.

variable}.

library

Most

to

an

argument

serves

to

identify a closed

a dummy

subprogram

subprogram

is

classified

7).

the

functions

as

an

subprograms. These

of

argument,

the

library

SIN,

that

is a subprogram

subprogram

is

actually a calion

in

the

same

FUNCTION

DOUBLE

DIFF

= DABS(F(Z) - DF(DBLE(Z)))

RETURN

END

DSIN,

must

previously

are

specially

subprograms

DIFF(F, DF, Z)

PRECISION

SQRT,

and

recognized

are

functions,

DSQRT.)

appear

name,

manner

DF

in

and

it

whose

the

as

any

an

EXTERNAL

by

the

although

is

the

only

actual

subprogram

other

compiler;

several

kind

location

subprogram

is

whose

statement

which

uti

lity

of

dummy

defined

name

(see

{otherwise,

generates

subroutines

by
is

Basic

External

The

basic

a

special

implicit

Table

9 lists

Function",
to

clarify

Additional

In

addition

EXIT

Form:

CALL EXIT

[ The

86

effect

Library

Functions

external

type

typing

the

library

is

Subprograms

that

rules.

the

function

it

is

not

definition

to

the

identical

function

Subprograms

functions

is

known

The

necessari

of

to

subprograms

to

the

compiler.

arguments

subprograms

Iy

the

formula

function.

listed

in

that

of

the

evaluate

to

these

provided

Table

STOP

commonly

This

type

functions

by

FLAG.

that

is

actually

9,

the

following

statement.

used

mathematical

is

not

necessarily

must

have

the

When a formula

used in

subprograms

the

same

proper

type,

is shown in

implementing

are

supplied

functions.

as

the

as

shown in

the

function;

in

These subprograms

type

it

would

Table

the

column

the

it

FLAG

"Definition

is

library:

acquire

9.

intended

have

by

of

only

Intrinsic

Names

Number

Arguments

of

Table

Type

of

Argument

9.

Intrinsic

Type
Result

and

of

Basic

External

Definition

Functions

of

Function

ABS

ACOS

AIMAG

AINT

ALOG
ALOG10
AMAX1

AMAXO

AMIN1

AMINO

AMOD

ASIN

N ~ 2

N

~

N~2

N ~ 2

2

Real
Real

Complex

Real

Real
Real
Real

2

Integer

Real

Integer

Real

Real

Real

Real

Real

Real

Real
Real
Real

Real

Real

Real

Real

Real

Absolute
Arc

CACOS.
Imaginary

plex)
Integer

truncated)
Natural

Common

Maximum

verted

Maximum

verted

Minimum

verted

Minimum

verted
Arg1 (mod

i.
Function

Arc

CASIN.

cosine

expressed

arg

e.,

sine

value.

part

logarithm

logarithm

value.

to

and

value.

to

and

value.

to

and
value.

to

and

arg2 *AINT(arg1/arg

-

1
the

undefined

in

in

radians.

part

of

.

compared

compared

compared

compared

arg2)'

sign

radians.

For

of

argument

as a real

argument

(base

(base

All

All

All

All

Evaluated

is

the

if

complex,

For

complex,

(zero

value.

(fractional

e).

10).

arguments

as

real

arguments

as

integer

arguments

as

real

arguments

as

integer

same

as

arg2 = O.

For

complex,

see

values.

values.

as

)

2

arg1 .

if

are

are

values.

are

are

values.

CABS.

see

not

com-

part

con-

con-

con-

con-

see

ATAN }

{

ATAN2

CABS

CACOS

CASIN

CATAN

CCOS

1,2

Real

Complex

Complex

Complex

Complex

Complex

Real

Real

Complex

Complex

Complex

Complex

Arctangent
arg2 = abscissa
assumed
arg1/arg2

-

1T<

see

CATAN.

Complex

CABS(x

Complex

=

Complex

=

Complex

= u +

-

-1T<US1T.

Complex

= (e

in

1.

Resul t

quadrant

R S

1T.

absolute

+ iy)

arc

-i . CLOG(Z
arc

-i . CLOG(i
arctangent.

iv

= -

CLOG(l -iZ)),

cosine.

iZ

+ e

radians.

(x).

ATAN(O,O) =

cosine.

If

(R)

allocated

value

=.y

x

Arg1 =

arg2

is

arctangent

(i.

e., modulus).

2

+

y2

CACOS(Z)

+ CSQRT(Z2 - 1))

sine.

CASIN(Z)

. Z + CSQRT(l - Z2))

CATAN(Z)

~

(CLOG(l + iZ)

allocated

CCOS(Z)

-iZ)/2.

ordinate

is

not

in

O.

present,

the

For

such

(y),

of

range

complex,

that

Library

Subprograms

87

Table 9.

Intrinsic

and

Basic

External

Functions

(cont.)

r

Intrinsi c
Names

CCOSH

CDABS

CDACOS

CDASIN

CDATAN

CDBLE

CDC as

CDCOSH

CDEXP

CDLOG

Number
Arguments I Argument

1

1

1

1

1

1

1

1

1

1

of

Type

of

Complex

Complex*16

Complex*16

Complex*16

Complex*16

Complex

Complex*16 Complex*16

Complex*16

Complex*16

Complex*16

Type

I

of

Result

Complex Complex

Real*8

Complex*16

Complex*16

Complex*16

Complex*16

Complex*16

C-omplex* 16

Complex*16

I

Definition

Double
See

Double

Double

Double

Converts

Double

Double
CCOSH.

Double

Double

See

hyperbolic

Z

= (e + e

complex

CABS.

complex

complex

complex

complex

complex

complex

complex

complex

CLOG.

of

Function

-Z

cosine.

)/2.

absolute

arc

cosine.

arc

sine.

arc

tangent.

to

double

cosine.

hyperbolic

exponential.

natural

logarithm

CCOSH(Z)

value

See

See

See

complex.

See

CCOS.

cosine.

See

(modulus).

CACOS.

CASIN.

CATAN.

See

CEXP.

(base

e).

CDSIN

CDSINH

CDSQRT

CDTAN

CDTANH

CEXP

CLOG

CMPLX

CONJG

COS

1

1

1

1

1

I

1

1

2 Real

1

1 Real

Complex*16

Complex*16

Complex*16

Complex*16

Complex*16

I

Complex

Complex

Complex

Complex*16

Complex*16

Complex*16

Complex*16

Complex*16

Complex

Complex

Complex

Complex

Real

Double

Double

rC'

\.....)lI

Double

Double

Double

CTANH.

Complex

Complex

a

Converts
plex

Complex

Cosine

see

complex

complex

rio.

II

I

....

n.

complex

complex

complex

exponential

CEXP(x + iy)

EXP(x) . (COS(y) + i .

=

natural

CLOG(Z) = CLOG(x

+ iv =

such

two

noncomplex

conjugate.

of

angle

In

that

CMPLX(x,y) = x +

= u

/located

number.

CCOS.

sine.

See

CSIN.

hyperbolic

square

root.

tangent.

hyperbolic

logarithm

(e **

See

arg).

(base e)

+ iy)

Izi

+ i . ATAN (y,x)

-IT

<

v:s

numbers

CONJG(x

in

radians.

sine.

See

CTAN.

tangent.

SIN(y)).

IT

•

iy.

+ iy)

For

complex,

See

CSQRT.

See

I

to a com-

=x-iy.

COSH

CSIN

88

Library Subprograms

1

1

Real Real

Complex

Complex

Hyperbolic

Complex

= (e i Z _ e

cosine.

sine.

CSIN(Z)

-i

Z) /

For

(2i).

complex,

see

CCOSH.

Table

9.

Intrinsic

and

Basic External

Functions

(cont.)

Intrinsic
Names

CSINH

CSNGL

CSQRT

CTAN

CTANH

DABS

DACOS

DASIN

Number
Arguments

1

1

1

of

1

1

1

1

1

Type

Type

of

Argument

Complex Complex

Complex*16

Complex

Complex

Complex

Real*8

Real*8

Real*8

of

Result

Complex

Complex

Complex Complex

Complex

Real*8

Real*8

Real*8

Definition

Complex

=

Converts

Complex

= u +

that

=

_ ( i Z

- e

Complex
CTANH(Z) = CSINH(Z)/CCOSH(Z)

= (e - e ) / e + e

Double

Double

Double

of

Function

hyperbolic

Z

(e

u

CSIN(Z) / CCOS(Z)

-Z

- e )
double

square

.

IV

= e ,

~

O.

tangent.

-e

hyperbolic

Z

-Z

precision

precision

precision

sine.

/2.

complex

root.

(In Z)

/2

CTAN(Z)

-i

Z) / . ( i Z + - i Z)

lee.

tangent.

Z

absolute

arc

cosine

arc

sine

CSINH(Z)

to

complex.

CSQRT(z)

allocated

-Z
).

value.

in

in

radians.

radians.

such

{DATAN }

DATAN2

DBLE

DCMPLX

DCONJG

DCOS
DCOSH
DDIM

DEXP
DFLOAT

DIM

DIMAG

DINT

DLOG

1,2

1

2 Real*8

1

1
1
2

1 Real*8
1

2 Real Real

1

1

1

Real*8

Real

Complex*16

Real*8
Real*8
Real*8

Integer

Complex*16

Real*8 Real*8

Real*8

Real*8

Real*8

Complex*16

Complex*16

Real*8
Real*8
Real*8

Real*8
Real*8

Real*8

Real*8

Double
See

Argument
precision.

Converts
double

Double

Double
Double
Double

See

Double

Argument
Same
integer

Positive

precision

ATAN.

two

complex

complex

precision
precision

precision

DIM.

precision

as

DBLE,

arguments.

difference.

converted

converted

= x - min

Imaginary
ment,

Integer
a

double

Double
(base

part

expressedasadouble

part

precision

precision

e).

arctangent

to a val

noncomplex

number.

conjugate.

consine
hyperbolic
positive

exponential

to

but

generally

(x,y).

of a double

of

the

argument
value.

natural

in

ue

numbers
See

See

of

angle

cosine.

difference.

double

DIM (x,y)

complex
precisionvalue.

expressed

logarithm

radians.

with

to

CMPLX.

CONJG.

in

(e **

arg).

precision.

used

with

double

a

radians.

argu-

as

Library Subprograms

89

Table

9.

Intrinsic

and

Basic E;I(ternal

Functions

(cont.)

r

Intrinsic

Names

DLOG10

DMAXl

DMINl

DMOD

DREAL

DSIGN

DSIN 1

DSINH

I

I

Number

Arguments

1

N:::2

N:::2

2

1

2

1

of

Type
Argument

I

1*8

Rea

Real*8

Real*8

Real*8

Complex*16

Real*8

Real*8

Real*8

of

I

I

Type
Resuit

Real*8

Real*8

Real*8

Real*8

Real*8

Real*8

Real*8

Real*8

of

Definition

I

Double
(base

Double
ments
double

Double
ments
double

Double
See

Real
expressed

Double
sign
positive.

Double

Double

precision

10).

precision

are

precision

precision

are

precision

precision

AMOD

part

precision

of

arg

precision

precision

of

Function

common

maximum

converted

values.

minimum

converted

values.

arg

•.

of a double

as a double

magnitude

If

arg

.

2

sine

hyperbolic

to

and

to

and

(mod

1

complex

precision

is

zero,

2

of

logarithm

value.

compared

value.

compared

arg2).

argument,

of

the

angle

sine.

value.

argl

sign is

in

radians.

All

All

with

argu-

as

argu-

as

DSQRT

DTAN

DTANH

EXP

FLOAT 1

I

lABS 1

lAND

ICOMPL

IDIM 2

IDINT

IEOR

IEXCLR

1

1

1

1 Real

2

1

1

N

:::

2

:::

2

N

Real*8

Rea!*8 Rea!*8

Real*8

Integer

Integer

Integer

Integer

Integer Integer

Real*8

Integer

Integer

I

I

,

Real*8

Real*8

Real

Real

Integer

Integer

Integer

Integer

Integer

Integer

Double

[)nllhlp

I

-

---.-

I

Double

Exponential
see

CEXP.

Argument

Integer

Logical

Logical
INOT.

Integer

= j -

Argument

Logical

Logical

IEOR.

precision

nrpric;inn

1-·--'-'-"

precision

(e **

converted

absolute

AND

(extract).

NOT

(lis

positive

MIN(j,k).

converted

EOR

(exclusive

EOR

(exclusive

square

root

(positive

tnnnpnt

·_··v-···

hyperboli c tangent.

value.

difference.

..

arg).

For

complex,

to a real

complement).

to

an

integer

OR).

OR).

value.

IDIM(j,k)

Same

Same

value.

as

value).

as

INOT

{INT}

IFIX

90

library

Subprograms

1

1 Real

Integer

Integer

Integer

Logical

Argument

NOT

(lis

converted

complement).

to

an

integer

va

lue.

Intri nsi c
Names

Number

Arguments

of

Table

Type

Argument

9.

of

Intrinsic

and

Type
Result

Basic

of

External

Functions

Definition

(cont.)

of

Function

lOR

ISA

ISC

ISIGN

ISL

MAX

MAX

MINO

2

2

2

2

Integer

Integer

Integer

Integer

Integer

Integer

Integer

Integer

Logical

Integer
arithmetically
in
the

Integer
circularly
arg2'
right.

Integer

If

is

2

0

1

N ~ 2

Integer

Integer

Real

Integer

Integer

Integer

Integer

Integer

Integer

logically
arg2'
right.

Integer

Integer

Integer
converted

arg2'

right.

arg2

not

OR

(merge).

shift

arithmetic.

If

arg

shift

circular.

the

If

arg2

magnitude

is

zero,

converted

shift

logical.

the

number

If

arg2 is-

maxi

mum

maximum

minimum

to

and

the

number

negative,

is

2

number
is

negative,

of

the

sign IS

to

integer

negative,

val

value.

value.

compared

Arg1 is

Arg1 is

of

bits

g

) with sign

ar

Arg1 is

of

bits

ue.

All

of

bits

the

shifted

specified

the

shift is

positive.

.

shifted

specified

the

shift is

arguments

in

integer.

shifted
specified

shift is

left

in

to

of

arg2'

Arg2

left

in

to

are

left

to

the

the

MINl

MOD

REAL

SIGN

SIN

SINH

SNGL

SQRT

TAN

TANH

N

2

1

2

~

2

Real

Integer

Complex

Real

Real

Real

Real*8

Real

Real

Real

Integer

Integer

Real
Real

Real

Real

Real

Real

Real

Real

Integer
converted
precision.

Arg1 (mod

where

i.

Function
Real
Magnitude

arg2 is

Sine
see

Hyperbolic
CSINH.

Argument
(single)

Square
see

Tangent
see

Hyperbolic
CTANH.

minimum

argl -arg2 * [arg1/arg2J

the

e.,

the

part

zero,

of

angle

CSIN.

precision.

root

CSQRT.

of

CTAN

value.

to

and

compared

arg2)'

brackets

sign is

is
of a complex

converted

the

undefined

of

arg

1

the

sign is

in

radians.

cosine.

(positive

angle

in

tangent.

Evaluated

with sign

All

indi

cate

same

as

if

arg2 = O.

number.

positive.

For

complex,

to a value

va

lue).

radians.

For

complex,

arguments

in

double

as

integer

arg1'

of

arg2'

For

complex,

with

For

For

are

part;

If

see

real

complex,

complex,

see

li

brary

Subprograms

91

SLITET -Sense

Form:

where

[

CALL

n

SLITET

is

an

Light Test

(n,v)

integer

constant

or

scalar

variable

specifying

whi

ch

sense

light

is

to

be

tested

(1

:s n :s

4).

[

v is

Sense

light

stored.

SLITE -Set

Form:

where

If

OVERFL -

Form:

where

If

flow
Overflow

Following

CALL

n is

n is 0,

a

all

CALL OVERFL

floating

condition

an

n is

Sense

SLITE

an

sense

Floating

is

an

overflow

exists,

and

underflow

integer

tested.

the

Light

(n)

integer

lights

Overflow

(s)

integer

variable

If

the

test,

constant

will

variable

has

occurred,

s is set

are

in

sense

the

sense

or

be

turned

Test

into

to

3.

defined

which

light

light

scalar

which

s is
The

in

the

the

is

on,

will

variable

Offi

if

will

set

to

machine

Sigma

result

the

be

n is

be

1 i

if

of

the

value

turned

(O:s n :s

1,2,3,

stored

no

overflow

is

left

in a

computer

test

wi

1 will

off.

4).

or 4,

the

result

no

reference

II

be

stored.

be

stored

the

corresponding

of

the

condition
overflow

manual.

in

Vi

if

it

test.

exists,

s is set

(underflow)

is

off,

sense

light

to

condition

the

2.

value

2 will

will

be

turned

If

a

floating

following

the

be

on.

under-

test.

[

92

DVCHK -

Form:

CALL DVCH K

where

This is

another

library

Divide

Check

(s)

is

an

integer

entry

Subprograms

variable

to

the

into

OVERFL subprogram

which

will

be

stored

described

the

result

above.

of

the

test.

FLAG

operates

tion

or

combined

This is

one

programs.

under

compilation

of

FLAG's most

control

of

the

and

execution

attractive

Sigma

features,

9.

OPERATIONS

5/7

Batch Processing

is a simple

together

procedure

with

its foci

Monitor

requiring

lity

(BPM).

the

for

Preparing

preparation

rapid

compi

a FLAG job for

of

a few

control

lation

and

execution

compila-

cards.

of

The user has a number

on

the

FLAG

control

Running a FLAG

Figures 3
ASSIGN,

BPM

the

The

different

ted

data

FLAG

The

Every FLAG job must

and 4 later

EOD,

arrange

FLAG

control

ASSIGN

from

if

M:SI

were

indications

card,

FLAG

Control

and

for

the

card

cards

the

standard

not assigned

could

is

available

Card

Job

shown in

of

convenient

card.

in this

FIN

cards

computer

required

ones

be

read

in

be

preceded

Option

chapter

shown in

operator

for his job.

the

examples

provided

to

the

from

the

Sigma

processing

codes

are

show

sample

the

to

insert

are

at

card

reader.

magnetic

5/7

Batch Processing

by a FLAG

options

explained

deck

examples

these

necessary

his

installation.

tape

control

at

setups for

control

That is,

or

disc.

card.

his-

later

are

standard

only

if

the

Detailed

Monitor

Its format is

disposal,

in

this

compiling

cards in

the

programmer

The

source

program

Reference

all

of

which

chapter

BPM

information

the

program

decks

in

and

control

FLAG

together

can

Table

executing

cards.

deck,

requires

decks

on

all

Manual

be

controlled

10.

FLAG jobs. The

Many

installations

the

programmer

file

or

I/o

and

EOD

with

their

control

(SDS 90

cards,

09

by

option

supplying

device

cards

would

appropriate

except

54).

codes

JOB,

using

only

assignments

be

omit-

end-of-

the

The

exclamation

it

may

begin

of

the

program.

fault.

one

blank

only

that

DEFAULT

When

the

successively

will
stantially

Once

FLAG has

each

subjob.

:FLAG

On

this

on

every
standard!
account

not

actually

If

option

column.

the

indicates

BJ

reduces

[(account

card

:FLAG

FLAG

number,

mark must

in

any

Option

codes

list

of

option

compile

the

read

Figure 5

the

characters

card

option
name,

entered

be

placed

column

(The
options

which

code

after

codes

are

are

given,

brackets

is

not

options

is

specified

and

execute

processing

the

BJ

option

illustrates

number, name)]

:FLAG must

if

the

AC

option

codes

are

and

the

on

the

card.)

in column

the!

character.

not

requiredi

they

must

be

around

the

option
required.)
are

in

on

time

required

code

how

the

[,option1,option2,.·.

valid

list

of

The

effect

the

! FLAG

any

number

on

the!

:FLAG

appear

has

been

on a :FLAG

option

codes

1.

The FLAG

The

if

none

separated

list shown

option

unless

their

card,

of

separate

for

each

FLAG

card

is

in columns

specified

card

indicate

control

optioni

are

are

specified,

by commas,

above

codes

are

complementary

FLAG

FLAG programs,

program in

card

it

expects

used in

,optionnJ

1-5.

on

the

except

only

batch

option

and

must

given

enters

the

processing.

The

user's
! FLAG
for

NOBJ,

that

these

command

codes

FLAG

wi

the

first

not

be

in

Table

options

the

batch

job

stream.

to

find a :FLAG

account

card;

AC,

items

is

usually

that

control

II perform

code

must

entered

10. In
have

been

job mode.

or

IIsubjobs". Use

card

Format

number

otherwise

and

are

they

NOAC.

not

begun

in column 2, though

processing

certain

be

preceded

on

the

card:

Table

10,

selected.

In

this mode, FLAG

immediately

of

the

and

name

are

(The

required;

and

operations

by

they

the

notation

of

this

option

:FLAG

card

are

required

optional.

brackets

the

brackets

execution

by

de-

at

least

indicate

sub-

preceding

is

All

around

are

Operations

93

r

Any

The
work
deck.

The

I/O

source

but

batch

unit

number assignments in

program

:EOD

job

deck

is

preferred.)

stream is

terminated

for

each

subjob must be

If

no

data

by

effect

the

when

is

present,

first!

the

batch

separated

the

control

job

from its

:EOD

card

encountered

run

card

is

started

data

is

by

still

remain in

an

:EOD

required

(other

than

effect

card.

to

terminate

:EOD).

for

all

subjobs.

(An IEOD

the

card

source

will

also

program

Option

DB
NODB

GO

NOGO

LS
NOLS
BC

NOBC

LO

NOLO

AD

NOAD

CX

NOCX

M,!l

Code

Meaning
The program
The program is
Note

generated

too

large

The program is

detected
When

checking
Note

executed
A

printed

No

I isting

FLAG will
(EOD)

the

series will

This

option
compiled
followed by
vents

FLAG will

end-of-data
A

machine-language

mat similar

No

machine-language

All

REAL
ables,
analyzing

No

"automatic

Source

and

will

Source
FLAG.

(Where!!..
One

segment consists
other
variables

tenths
ample,
2/10

is

is

compiled
not

that

use

of

for

the

to

run

executed

during

compilation

of

the

that

if

neither

unless

listing

of

the

compile

indication,

be

is

mainly

as a

unit

an

it

from assuming

terminate

indicator.

to a Meta-Symbol

variables,

constants, and functions

the

improvement in

card-images

be

compiled

card-images

(DEFAULT)

is a digit

segment

of

that

the

available

M8

specifies

to

be used for

contains

were

Table 10. FLAG

and

compi led in

the

DB

option

program

with

compilation.

is

source

one

of

source

until two

executed,

but

EOD

(typically

"debug"

when

finished,

program

GO

nor

or

more serious errors

the

FORTRAN

statements

a series

of

successive
if

intended

have

been

record.

"end-of-program"

compilation

(DEFAULT)

I isting

of

I isting

constants,

double

precision"

containing

by FLAG.

containing

1 through

of

the

noninitialized

the

machine

initialized

memory

that

8/10

the

program

Option

executed

"debug"
will

may

compilation

the

NOGO

source

appropriate.

for use

The

the

listing.

is

and

are

accuracy

an

an

9)

in a

are

of

in

mode.

cause

30

to

be

able

program is

without

produced.

execution.

are

statements

is

produced.
programs,

end-of-data

where

stored on

BC

option

upon

encountering

instructions

functions
implicitly

that

is

invoked.

X in column

X in column 1 will

FLAG

divides

instructions

DATA

to

be

memory

code

and

Codes

"debug"

substantially

40%

specified,

have

when a

variables

statement.

used for

is

mode.

(DEFAULT)

more),

to

run

without

is finished,

not

executed.

the

been

detected

in

the

source

each

followed by a

indications

a main program

magnetic

code

(DEFAUL n

are

results from

tape,

alerts

single

an

generated

implicitly

COMPLEX*16. This

(DEFAUL n

1,

wi

II

have

be

total

available

that

generated

The!!..

the

noninitial

to

be

used for non

initialized

more

machine

and

that

it.

whether

This

compiled

during

program

and

and

FLAG

EOD

is

IEOD

card

by FLAG

REAL

double

the X replaced

treated

aie

used iii

for

the

value

initial

variables.

instructions

some programs

or not errors

option

program will still

compilation.

is

produced.

single

are

encountered.

subprograms

therefore

to

this

condition

encountered.

or

some

is

produced

*8;

all

option

precision

as comment cards by

memory

ized

into

the

program and also

specifies

data

ized

This

that

have

allows

syntax

(DEFAUL

end-of-data

are

each

program

other

in a

COMPLEX

is

useful for

calculations.

by a

two segments.

piogiam. The

how many
area.
variables

option

need not

to

been

Then

to

and

single

space,

any

For

and

be

are

be

n

be

is

pre-

for-

vari-

ex-

be

94 Running a FLAG

Job

Table

10. FLAG

Option

Codes

(cont.)

Option

M.!J.

(cont

BJ

NOBJ

AC

NOAC
TL = sss

PL=ppp

PS

Code

.)

=nnn

Meaning
specified

size
Error
ARRAYS

PROGRAM

DICTIONARY OVERFLOW

The
Enter

the!
Run

Punch

used,

No

Set
may run. This

Set
will

Note

BJ

low.
limit

Set
that

statement. The remainder

vqriables.
tion
variables

When

compilation
structions
gram
of

NOGO
is possible
the

unless FLAG

allocation

Message

TOO

default

"batch

FLAG

in

standard

separate

should
accounting
time

limit for

page

limit for

be

allowed.

that

option,

Furthermore,

that

may

program

may

be

of

memory

is

attempting

are

and

initialized

the

NOGO

is

PS

option

indicates

problem

LARGE

TOO

option

job" mode. A sub job

control

(nonbatch)

accounting

appear

records

time

if

either

the

value

be

size
used

The

PS

size.

given

using

stored

compilation,

not

specified.

to

run

set

LARGE

is

M7.

card.

on

wi

current

includes

current

the

TL

specified

the

specified

for

current

to

hold

option

The

at

the

to

run

the

NOGO
into
variable

the

program,

equal

has

the!

limit

end

very

to

occurred:

All

subsequent

mode.

records for

FLAG

II

be

punched

job or

compilation

job

or

or

PL

is used

specified

on a

job

the

program

of

avai!able

is

similar

actual

of

the

large

and

memory,

area.

is

correct

If

the

and

the

indicated

by

one

Action

increase

decrease

increase

beginning

(DEFAULT)

control

subjob,

subjob,

option

subjob's

or

sub job,

and any

to

amount

source

programs,
M9

hence

The

actual

the

of

the

following

to

Take

n

error

-n

n

with

a :FLAG

subjobs must

each

subjob in

card

for sub jobs. (DEFAULT)

where

and

execution.

where

appears

as

the

default

on a ! FLAG

:FLAG

where

variables

memory

the

Mn

option

of

memory used by

program

it

options.

substantially

actual

and

is

the

total

size

program should

size

of

the

that

initiates

sss

is

the

ppp is

on a

!FLAG

limit for

control

control

nnnn is

vIi!!

be

but

listing.

is

sometimes a

When

program

same as

is less

the

program

also

begin

batch.

number

the

number

card.

the

that

used

allows

NOGO
less memory

size,

the

than

be

submitted

messages

card

the

control

each

card

number

were

to

hold the

the

good

is

which

size

the

and

must

immediately

with

:FLAG

The AC
batch

of

seconds

of

output

card

of

the

becomes

of

initialized

for more
program

idea

specified

is

needed
is

of

the

available

again,

initialized

that

a memory

follow

cards.

option,

run.

that

subjobs

the

words

noninitia!ized

accurate

and

to

listed

program when

that

the

pages

also

maximum

of

memory

in a

initialized

make

no

machine

for

the

at

the

total

size,

this

time

variable

when

that

has

that

DATA

alloca-

the

job

the
fol-

first

in­pro­end

it
with
area.

Job

Setup

Figure 3 shows
beginning
SHERROD,

As

previously

quired
The!

lowing
gram
regardless
machine-language

The!
of

the

and

end-of-program

source

Examples

of

a new job

and

nonstandard

FLAG

card

the

FLAG command

to

be

compi led

of

errors

FLAG

card

deck

to

deck

and

the

deck

the

job has

explained,

assignments.

summons

(GO),

statements

is

followed

the

compiler.

could

its

terminating!

setup

to

and

BPM,

priority

the

the

are
executed
produce

by

be

required

and

specifies

1.

!ASSIGN

FLAG

compiler

explained

in

a listing

generated

the

source

If

the

source

indicated

EOD

for

compiling

card

(with assignment codes) would

in

Table

"debug

of

by

the

program

program

by some

card

would

that

the

to

begin

10. Briefly,

ll

mode (DB),

the

source

compi

deck,

did

other

be

and

executing a single

job

is

compilation

execute

program

ler

(LO).

in turn

not

require a data

terminator

omitted

to

be

run

of

the

the

codes

the

statements

followed

such as a ! FIN

if

M:SI

were

program. The

under

account

be

source

shown

program when compi

(LS),

by

an

deck,

not

number 1234,

present

assigned

only

program. The

in

the

example

and

produce

! EOD

card

the!

EOD

card

or a new

to

!JOB

if

card

the

the

option

lation

a listing

which

card

Job

card

signals

the

user

programmer

codes

request

is

finished

indicates

could

!JOB

reader.

Setup

the

of

be

card.

Examples 95

the

is

re-

fol-

pro-

the

the

omitted

end

The