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

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

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

Technical

Manual

Norman

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

List Items 27 Special

Is

Organized

Compiler

Compilation

of

Quantities

Constants

Prec ision

Constants

Complex

Constants

Expressions

Operators

STATEMENTS

GO

IF

Statement

List

Constants

Hierarchy

STATEMENT

ents

GO

TO

Statement

GO

TO

IF

Statement

Lists

Considerations

- X

Cards

TO

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

Input

Field-Termination

and

Stored

Input/Output

Access

Input/Output

FILE

Control

icting

Storage

Statement

Using NAME

(Fixed

Decimal

(Normalized,

(General)

(Integer)

(Logical)

(Alphanumeric)

(Alphanumeric,

(Hexadecimal)

(Machine

(Hollerith)

(Skip;

(Tab) 49

(Scale

(Record

FORMAT

Numeric

Input

Statement

IN OUT

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

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

of

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

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/,

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

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

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

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

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

FACTOR

is is is

to

of

to

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

line

source

program.

four times,

on

- X

TO

an

assignment

statement,

on

unit

FORMAT

statement

informs

if

lines

Cards

statement

108,

corresponding

correspond

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

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

2,

are

type

of

or minus

Constants

constants

147,483,647.

data

precision

are

approximations

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.

that

or IIfalse

1 may

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

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

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

greater

Array

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

parentheses

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

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

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

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 Double Precision

Double Complex

Complex Complex

specifications

exists.

The

order

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

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

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)

arithmetic

operator

expressions

type.

comparisons

than

«)

than

or

to

(=)

equal

than than

is

true.

is

false.

is

always

is

always

2

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

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

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"

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

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

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

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

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

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

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

of

the

of

the

e is

the

value

variable

set

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

is

set

to

9238E-l)

or

decreased

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

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

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

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

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

statement,

statement

22

1234

99999

1),

v

JUMP

TO M

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

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

second.

IF

statements, which

of

IF

statements:

program

where

e is

The

arithmetic

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

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

statement is represented as

(e) s

where:

e is a

The

statement

lowing

after

the

Examples:

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

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

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

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

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

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

e3 is

not

(i.

2.

va

lue

of

not

be

any

range.

repetitively.

constitute

label

k.

present,

e.,

to

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

[

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

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

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

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,

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,

88

statement

example:

does

not

cause

the

of

compiler the

CONTINUE

to

generate

machine

statement

is

instruc-

to

allow

[

CONTINUE

PAUSE

where

Statement

statements

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

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

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

Sigma XDS 5, XDS 7 Reference Manual

Specifications and Main Features

Frequently Asked Questions

User Manual