Digital PDFs

AA-N383B-TK

May 1985

546 pages

Original

18MB

Document:	FORTRAN Language Manual Feb87
Order Number:	AA-N383B-TK
Revision:	000
Pages:	546
Original Filename:	AA-N383B-TK_FORTRAN_Language_Manual_Ver_10_May85.pdf

OCR Text

TOPS-10/TOPS-20
FORTRAN Language Manual
AA-N383B-TK

May 1985
This document describes the language elements of
FORTRAN-10 and FORTRAN-20.
This manual supersedes the TOPS-10120 FORTRAN Language
Manual, order number AA-N383A-TK.
OPERATING 8YSTEIVI:
SOFTWARE:

TOPS-10 V7.02
TOPS-20 V4.1, V5.1

FORTHAN-10 V10
FORTF~AN-20 V10

Software and manuals should be orderecl by title and order numbEr. In the United States. send orders
to the nearest distribution center Outside the United States. order~; should be directed to the nearest
DIGITAL Field Sales Office or representative

NortheasttMid-Atiantic Region

Central Region

Digital Equipment Corporation
PO Box CS2008
Nashua. New Hampshire 03061
Telephone:(603)884-6660

Digital Equipment Corporation
DlgHal Equipment Corporation
Accessories and Supplies Center Accessories and Supplies Center
1050 East Remington Road
632 Caribbean Drive
Schaumburg. Illinois 60195
Sunnyvale. California 94086
Telephone (312)64Q--5612
Telephone.(408)734-4915

Western Region

digital equipment corporation. marlboro. massachusetts

First Printing, February 1983
Revised, May 1985
© Digital Equipment Corporation 1983, 1985. All Rights Reserved.

The information in this document is subject to change without notice and should
not be construed as a commitment by Digital Equipment Corporation. Digital
Equipment Corporation assumes no responsibility for any errors that may
appear in this document.
The software described in this document is furnished under a license and may
only be used or copied in accordance with the terms of such license.
No responsibility is assumed for the use or reliability of software on equipment
that is not supplied by DIGITAL or its affiliated companies.
The following are trademarks of Digital Equipment Corporation:

~DmDDmDTM
DEC
DECmate
DECsystem-10
DECSYSTEM-20
DECUS
DECwriter
DIBOL

MASSBUS
PDP
P/OS
Professional
Q-BUS
Rainbow
RSTS

RSX
RT
UNIBUS
VAX
VMS
VT
Work Processor

The postage-prepaid READER'S COMMENTS form on the last page of this
document requests the user's critical evaluation to assist us in preparing future
documentation.

CONTENTS

CHAPTER 1
1.1
1.2
1.2.1
1.2.2
1.2.3
1.2.4
CHAPTER 2
2.1
2.2
2.2.1
2.2.2
2.2.3
2.2.4
2.3
2.3.1
2.3.2
2.3.3
2.3.4
2.3.5
2.4
CHAPTER 3
3.1
3.2
3.3
3.4
3.4.1
3.5
3.6
3.7
3.8
3.9
3.10
CHAPTER 4
4.1
4.2
4.3
4.3.1
4.3.2
4.3.3
4.4
CHAPTER 5

INTRODUCTION
OVERVIEW • • • • • • • • • • • • • •
•••••
ORGANIZATION • • • • • • • • • • • • • • •
FORTRAN Language Elements (Part I)
••
FORTRAN Statements (Part II)
•••••••
FORTRAN Language Usage (Part III)
••••
APPENDIXES • • • • • • • • • • • • • • • • • • •

M~NUAL

1-1
1-1
1-2
1-2
1-3
1-4

CHARACTERS AND LINES
CHARACTER SET
• • • • • • • • • • • • • • •
STATEMENT DEFINITION AND FORMAT • • • • • • • • •
Statement Label Field and Statement Numbers
••
Line Continuation Field
••••
• •
Statement Field
• • • • •
Remark Field • • • • . •
••••••••••
LINE TYPES • • • • • • • •
• • • • • • • • •
Initial and Continuation Lines •
• ••••
Multi-Statement Lines
••
Comment Lines and Remarks
•••••
• •
Debug Lines
•••.••••
•
Blank Lines • • • • • • • • • • • •
••• •
LINE-SEQUENCED SOURCE FILES
• •

2-1
2-2
2-3
2-3
2-4
2-4
2-4
2-4
2-5
2-5
2-6
2-7
2-7

CONSTANTS
INTRODUCTION
• • • • •
•
INTEGER CONSTANTS • • • • • • • • •
• •
REAL CONSTANTS • • • • • • •
• • • • • •
DOUBLE-PRECISION CONSTANTS • • • • • • • • • • • •
Comparison of Real, D-floating, and G-floating •
COMPLEX CONSTANTS
• • • • • •
•
CHARACTER CONSTANTS • • • • • • • • • • • • • • •
OCTAL AND DOUBLE-OCTAL CONSTANTS
•••••••
LOGICAL CONSTANTS
• • •
HOLLERITH CONSTANTS • • • •
• • •
s'rATEMENT LABEL CONSTANTS
• • • • •
• •

3-1
3-1
3-2
3-2
3-3
3-4
3-4
3-5
3-6
3-6
3-7

SYMBOLIC NAMES, VARLr\BLES, AND ARRAYS
SYMBOLIC NAMES •
·
VARIABLES
ARRAYS • • • •
••••
·
Array Element Subscripts •
•
Dimensioning Arrays
Order of Stored Array Elements •
CHARACTER SUBSTRINGS • • • •
• •

4-1

• • • 4-2
. . . .
• • •

• •

. 4-3
• 4- 3

• 4-4
• 4-5
• • 4-6

EXPRESSIONS

ARITHMETIC EXPRESSIONS
5.1
Rules for Writing Arithmetic Expressions.
5.1.1
Arithmetic Constant Expressions
5.1.2
Integer Constant Expression
5.1.2.1
CHARACTER EXPRESSIONS
• • • • • .
5.2
iii

5-1
• 5-2
• 5-5
• • • 5-5
• • 5-6

5.2.1
5.3
5.3.1
5.4
5.5
5.5.1
5.5.2
5.5.3
5.5.4
5.6
CHAPTER 6
6.1
6.2
6.3
6.4
6.4.1
6.4.2
CHAPTER 7
7.1
7.1.1
7.1.2
7.2
7.2.1
7.2.2
7.3
7.4
7.4.1
7.4.2
7.5
7.5.1
7.6
7.7
7.8
7.9
7.10
CHAPTER 8
8.1
8.2
8.3
8.4
CHAPTER 9

Character Constant Expression
LOGICAL EXPRESSIONS . • . • •
Logical Constant Expression
RELATIONAL EXPRESSIONS . . • .
EVALUATION OF EXPRESSIONS
Parenthetical Subexpressions .
Hierarchy of Operators • • • . • • • •
Mixed-Mode Expressions • . •
••••
Use of Logical Operands in Mixed-Mode
Expressions
• • •.
CONSTANT EXPRESSIONS . • • . • . • . • •

• • 5-7
• 5-7
5-10
5-11
5-13
5-13
5-14
5-15
5-15
5-15

EXECUTABLE AND NONEXECUTABLE STATEMENTS
6-1
EXECUTABLE STATEMENTS
.•.••••••••.
NONEXECUTABLE STATEMENTS . • •
· 6-2
ORDERING OF FORTRAN STATEMENTS • • . • • •
· 6-2
COMPILATION CONTROL STATEMENTS
• . 6-4
PROGRAM Statement
. • • • • • • • .. 6-4
INCLUDE Statement
• • • • • • • • • • •
• 6- 4
SPECIFICATION AND DATA STATEMENTS
DIMENSION STATEMENT . • • .
Adjustable Dimensions
·
Assumed-size Arrays
•
TYPE SPECIFICATION STATEMENTS
Numeric Type Specification Statements
Character Type Specification Statements
IMPLICIT STATEMENTS
• • • • • • . • •
COMMON STATEMENT . • . • . . • . • • • • •
Dimensioning Arrays in COMMON Statements
Character Data in COMMON • • • . • • •
EQUIVALENCE STATEMENT
• . • • • • •
EQUIVALENCE and Extended Addressing
EXTERNAL STATEMENT •
INTRINSIC STATEMENT
PARAMETER STATEMENT
DATA STATEMENT •
•
SAVE STATEMENT • • •
•

· 7-1
• • • .. 7-3
•

• •

.' 7 - 4

· 7-4
. • .. 7-5
• .. 7-6
•

.. 7-7

• .

•

7-8
7-10
7-10
7-11
7-17
7-17
7-18
7-20
7-21
7 -24

• .
• •
••
• •

8-1
8-3
8-3
8-4

•

ASSIGNMENT STATEMENTS
ARITHMETIC ASSIGNMENT STATEMENT
• • • . • . •
LOGICAL ASSIGNMENT STATEMENTS . . • • • • • •
ASSIGN (STATEMENT LABEL) ASSIGNMENT STATEMENT
CHARACTER ASSIGNMENT STATEMENT • . • • • . • •
CONTROL STATEMENTS

9.1
GO TO STATEMENTS •
Unconditional GO TO Statements .
9.1.1
Computed GO TO Statements
9.1.2
Assigned GO TO Statements
9.1.3
9.2
IF STATEMENTS
. . • • • • • • • .
Arithmetic IF Statements • •
9.2.1
Logical IF Statements
9.2.2
Logical Two-Branch IF Statements • •
9.2.3
Block IF Statements
. . ..
9.2.4
Statement Blocks . . • . • . . • .
9.2.4.1

· 9-1
• . 9-2
9-2
· 9-3
• • 9-3
• 9-3
· 9-4
• 9-4
9-5
· 9-8

•
Block IF Examples
••••
9.2.4.2
Nested Block IF Constructs
• • • • •
9.2.4.3
DO STATEMENT • • • • • • • • • •
9.3
Indexed DO Statement •
9.3.1
Executing an Indexed DO Statement
9.3.1.1
• •
DO Iteration Control
9.3.1.2
• •
DO WHILE Statement • •
9.3.2
The Range of a DO Statement
••••
9.3.3
Nested DO Statements • • • • •
9.3.4
Extended Range • • . • • • •
9.3.5
Permitted Transfer Operations
9.3.6
END DO STATEMENT • •
••• • • • • • •
9.4
CONTINUE STATEMENT • • • • •
9.5
STOP STATEMENT • •
9.6
PAUSE STATEMENT
• • • •
9.7
END STATEMENT • • • •
9.8
CHAPTER 10

• • • 9-8
• • • 9-9
9-10
9-10
9-12
• • 9-12
••
9-13
9-14
9-14
9-15
9-16
9-17
9-17
9-18
9-19
9-20

DATA TRANSFER STATEMENTS

DATA TRANSFER OPERATIONS •
10-5
10.1
DATA ACCESS
•••••••••
10-5
10.2
Sequential Access
10-5
10.2.1
Direct Access
10-6
10.2.2
FORMATTED AND UNFORMATTED DATA TRANSFERS • •
10-6
10.3
Formatted Data Transfers • •
10-6
10.3.1
Internal Files • • • • . •
10-7
10.3.1.1
Unformatted Data Transfers •
10-8
10.3.2
Unformatted Data Transfer to ASCII Devices • •
10-8
10.3.3
DATA TRANSFER STATEMENT FORMS
10-8
10.4
10.4.1
Data Transfer Statement Names
• • • • ••
10-9
Data Transfer Control-Information List • •
10-9
10.4.2
Unit References in Data Transfer Statements
• 10-10
10.4.3
FORTRAN LO'Jical Unit Identifier
• • • • • • 10-11
10.4.3.1
Internal File Identifier • • • • • • • • • • 10-13
10.4.3.2
Record Number References In Data Transfer
10.4.4
Statements • • • • • • • • • • • • • • • • • • 10-13
Format References in Data Transfer Statements 10-13
10.4.5
FORMAT-Statement Formatting
• • 10-14
10.4.5.1
List-Directed Formatting • • • • • • •
10.4.5.2
• 10-16
NAME LIST-Statement Formatting
••••
10.4.5.3
• 10-16
Optional End-of-File Transfer of Control
10.4.6
(END=) • • • • • • • • • • • • • • • • •
• 10-17
10.4.7
Optional Dat3 Transfer Error Control (ERR=)
• 10-18
10.4.8
Optional Error Variable For Error Reporting
(IOSTAT=)
•••••••••••••
• • 10-19
10.4.9
Data Transfer Statement Input/Output Lists • • 10-20
10.4.9.1
Simple List Elements.
• • ••
• • 10-21
10.4.9.2
Implied DO Lists. • • • • • • • •
• • • 10-22
10.5
READ STATEMENT. . • • • • • • • • •
• • 10-24
10.5.1
Formatted READ Transfers. • • • •
10-26
10.5.1.1
Sequential FORMAT-Statement READ. •
• 10-26
10.5.1.2
Direct-Access FORMAT-Statement READ
• • 10-27
10.5.1.3
Sequential List-Directed READ
• • • • • 10-27
10.5.1.4
Sequential NAMELIST-Statement READ.
• 10-28
10.5.2
Unformatted READ Transfers. • • •
• 10-29
10.5.2.1
Sequential Unformatted READ
• 10-29
10.5.2.2
Direct-Access Unformatted READ.
• 10-30
10.6
WRITE STATEMENT • • • • • • • • • •
• 10-30
10.6.1
Formatted WRITE Transfers
10-31
10.6.1.1
Sequential FORMAT-Statement WRITE
• 10-32
10.6.1.2
Direct-Access FORMAT-Statement WRITE
10-33
10.6.1.3
Sequential List-Directed WRITE. • • •
10-34
10.6.1.4
Sequential NAMELIST-Statement WRITE
• • 10-35
v

•
• •
•
· •
•
• •
••
•
• •
• •
• •
• •

10-35
10-36
10-36
10-37
10-38
10-39
10-40
10-40
10-41
10-42
10-43
10-44
10-45
10-45
10-46
10-47
10-48
10-48
10-49
10-50
10-51

FILE-CONTROL STATEMENTS
11.1
OPEN STATEMENT • • • . . •
11.2
Implicit OPEN • • • • .
11.2.1
OPEN on a Connected Unit. • • • • .
11.2.2
OPEN STATEMENT SPECIFIERS
11.3
ACCESS Specifier • • • • • .
11.3.1
ASSOCIATEVARIABLE Specifier
11.3.2
BLANK Specifier
••.•••
.••••
11.3.3
BLOCKSIZE Specifier
••.•••
11.3.4
BUFFERCOUNT Specifier
••
11.3.5
CARRIAGECONTROL Specifier
.•••••
11.3.6
DENSITY Specifier
•
11.3.7
DEVICE Specifier . • • • •
• •
11.3.8
DIALOG Specifier • • . • •
• •
11.3.9
DIALOG= Specifier
•••••
• ••
11.3.10
DIRECTORY Specifier (TOPS-I0)
••
11.3.11
DIRECTORY Specifier (TOPS-20)
••
11.3.12
DISPOSE Specifier . • • • • • • . • • • • • •
11.3.13
ERR Specifier
•.••.•••.•••••••
11.3.14
FILE Specifier • • . • . • • • • • •
•
11.3.15
FILESIZE (INITIALIZE) Specifier (TOPS-10 Only)
11.3.16
FORM Specifier •
•
11.3.17
IOSTAT Specifier
. • • .
• •
11.3.18
LIMIT Specifier
. • • •
•
11.3.19
MODE Specifier •
• •
11.3.20
NAME Specifier • • .
• •
11.3.21
PADCHAR Specifier
••
11.3.22
PARITY Specifier • •
•
11.3.23
PROTECTION Specifier (TOPS-10)
•••
11.3.24
PROTECTION Specifier (TOPS-20)
••
11.3.25
READONLY Specifier • • • • . •
11.3.26
RECL (RECORDSIZE) Specifier • • • • •
• •
11.3.27
RECORDTYPE Specifier • • • .
• •
11.3.28
STATUS (TYPE) Specifier
•••••••
11.3.29
TAPE FORMAT SPECIFIER • • . .
.••• • • •
11.3.30
UNIT Specifier (Required)
.•••••••••
11.3.31
VERSION Specifier (TOPS-I0)
••
11.3.32
CLOSE STATEMENT
• • •
11.4
Implicit CLOSE. • • • • • •
• •••••
11.4.1

11-1
11-1
11-3
11-4
11-4
11-6
11-8
11-9
11-10
11-10
11-11
11-12
11-13
11-13
11-14
11-14
11-16
11-16
11-17
11-18
11-18
11-19
11-20
11-21
11-21
11-23
11-24
11-24
11-25
11-27
11-28
11-28
11-29
11-30
11-32
11-33
11-34
11-34
11-34

10.6.2
Unformatted WRITE Transfers
10.6.2.1
Sequential Unformatted WRITE • •
10.6.2.2
Direct-Access Unformatted WRITE
••••
10.7
REREAD STATEMENT . • • . • . • . . • •
10.7.1
Sequential FORMAT-Statement REREAD.
10.7.2
Sequential List-Directed REREAD • • • • •
10.8
ACCEPT STATEMENT • • • • • • • . • • • . • •
10.8.1
Sequential FORMAT-Statement ACCEPT.
10.8.2
Sequential List-Directed ACCEPT
10.9
TYPE STATEMENT . . • • . • • • • • •
10.9.1
Sequential FORMAT-Statement TYPE • •
10.9.2
Sequential List-Directed TYPE
10.10
PRINT STATEMENT • • • • . . • • . .
10.10.1
Sequential FORMAT-Statement PRINT
10.10.2
Sequential List-Directed PRINT • . .
•
10.11
PUNCH STATEMENT • . • • • • • • • • •
10.11.1
Sequential FORMAT-Statement PUNCH
10.11.2
Sequential List-Directed PUNCH • • • •
10.12
INTERNAL FILES AND ENCODE/DECODE STATEMENTS
10.12.1
Internal READ and WRITE Statements.
10.12.2
ENCODE and DECODE Statements • • • • • • •
CHAPTER 11

•
•
•
•
•
•

•
•
•
•
•
•
•

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

11.5
11.5.1

CLOSE STATEMENT SPECIFIERS
11-34
DEVICE, DIRECTORY, FILE, NAME, and PROTECTION
Specifiers • • • . .
• 11-36
11.5.2
DIALOG Specifier. . • • . .
• 11-36
11.5.3
DIALOG= Specifier
• • • • 11-37
11.5.4
DISPOSE Specifier
• 11-37
11.5.5
ERR Specifier
. • • • • • • 11-39
11.5.6
IOSTAT Specifier.
11-39
11.5.7
LIMIT Specifier
. • • • •
11-39
11.5.8
STATUS Specifier. .
• 11-40
11.5.9
UNIT Specifier (Required)
. . • • 11-41
11.6
OPEN AND CLOSE STATEMENT EXAMPLES • . • • • • • 11-41
11.7
INQUIRE STATEMENT
• 11-42
11.7.1
INQUIRE by File
• • •.
• • • 11-42
11.7.2
INQUIRE by Lnit
11-43
11.7.3
Inquiry Specifiers.
11-43
11.7.3.1
ACCESS Specifier.
11-43
11.7.3.2
BLANK Specifier • •
• 11-44
11.7.3.3
CARRIAGECONTROL Specifier
• 11-44
11.7.3.4
DIRECT Specifier.
11-44
11.7.3.5
ERR Specifier
•••••
• 11-45
11.7.3.6
EXIST Specifier
. • 11-45
11.7.3.7
FORM Specifier. • • .
••••
• 11-45
11.7.3.8
FORMATTED Specifier
. • • • 11-45
11.7.3.9
IOSTAT Specifier.
• 11-46
11.7.3.10
NAME Specifier. • • • . • .
• • • • • 11-46
11.7.3.11
NAMED Specifier
. • • • • • • 11-47
11.7.3.12
NEXTREC Specifier
• • • 11-47
11.7.3.13
NUMBER Specifier.
•• • •
• • 11-47
11.7.3.14
OPENED Specifier.
• • • •
• 11-48
11.7.3.15
RECL (RECORDSIZE) Specifier
• 11-48
11.7.3.16
RECORDTYPE Specifier. . • • • •
11-48
11.7.3.17
SEQUENTIAL Specifier. • •
. • • • • • • 11-49
11.7.3.18
UNFORMATTED Specifier
• • • •
• 11-49
11.8
DEVICE CONTROL STATEMENTS • • • • •
• • 11-52
11.8.1
FIND Statem~nt • • • • . • • • • •
11-53
11.8.2
REWIND Stat~ment • •
• 11-54
11.8.3
UNLOAD Stat~ment • • •
. • ••
• 11-54
11.8.4
BACKSPACE Statement
11-55
11.8.5
ENDFILE Stat.ement
• • • • • •
11-55
11.8.6
SKIPRECORD Statement.
• 11-56
11.8.7
SKIPFILE Statement.
11-57
11.8.8
BACKFILE Statement. .
• • 11-57
CHAPTER 12
12.1
12.1.1
12.1.2
12.1.3
12.1.4
12.1.5
12.2
12.2.1
12.2.2
12.2.3
12.3
12.3.1
12.3.2

FORMATTED DATA TRANSFERS
FORMAT-STATEMENT FORMATTING
• • • • • • . • •.
12-2
Specifying a Format List in a FORMAT Statement 12-2
Specifying a Format Specification as a
Character E)pression • . • • • • • • • • • • • 12-3
Specifying a Format Specification in a Numeric
Array
• . • • • • • • • • • • • • • • • 12-4
Specifying a FORMAT Statement Using an ASSIGNed
Variable • .
• • • • • • • •.
• • • • 12-4
The Ordering and Interpretation of Format List
Items
• . • • • • • •
12-5
EDIT DESCRIPTORS • . • • • • • • • •
12-6
Repeatable Edit Descriptors
• • • •
12-7
NonrepeatabJe Edit Descriptors.
12-8
Carriage-Control Specifiers
12-9
INTERACTION or INPUT/OUTPUT LIST AND FORMAT LIST 12-11
General Description
. • • • • • ••
• 12-11
Formatted Input
. • • • • • •
• • • • • • 12-14
vii.

Formatted Output • • • • • • • • • •
12.3.3
• •
Embedded Format Specifications •
12.3.4
FORMAT EDITING • • • • • • • • •
12.4
•
Apostrophe (') Editing.
12.4.1
•
H Editing
.••••.•
12.4.2
•
positional Editing • • •
12.4.3
• •••••
T, TL, and TR Editing
12.4.3.1
•
X Editing
.•••••.
12.4.3.2
$ (Dollar Sign) Editing
12.4.4
• • •
/ (Slash) Editing
12.4.5
(Colon) Editing
12.4.6
S, SP,and SS Editing •
12.4.7
P Editing
••••••.••.
12.4.8
•
BN and BZ Editing
12.4.9
• •
Q Ed it i ng
•.•••
12.4.10
• ••••.
Numeric Editing
12.4.11
I Editing
12.4.11.1
F Editing
12.4.11.2
•
E and D Editing
••••
12.4.11.3
• • •
G Editing
••••••••
12.4.11.4
Complex Editing
••••
12.4.11.5
o ( Oc tal) Ed i tin g
• • • •
·
12.4.11.6
• •••.•
Z Editing
••••••••
12.4.11.7
• . •
L Editing
..••••.
12.4.12
• •
A Editing
12.4.13
R Editing
.••.
12.4.14
•
LIST-DIRECTED FORMATTING • . • • • •
12.5
NAMELIST-STATEMENT FORMATTING
• • • •
• ••
12.6
•
NAMELIST STATEMENT • • . •
12.7
NAMELIST-Controlled Data Input Transfer
•
12.7.1
NAMELIST-Controlled Data Output Transfers
12.7.2
CHAPTER 13

12-15
12-15
12-16
12-16
12-17
12-18
12-19
12-21
12-22
12-23
12-24
12-24
12-25
12-28
12-29
12-29
12-31
12-32
12-32
12-33
12-36
12-36
12-37
12-38
12-39
12-40
12-41
12-44
12-44
12-45
12-46

FUNCTIONS AND SUBROUTINES

INTRINSIC FUNCTIONS
13.1
. • ••
Using an Intrinsic Function
13.1.1
Character Intrinsic Functions
13.1.2
Character Comparison Functions •
•
13.1.3
Bit Manipulation Functions • •
•
13.1.4
·
STATEMENT FUNCTIONS
• • • • • •
13.2
Defining a Statement Function
••••
· •
13.2.1
Using a Statement Function • •
• •
13.2.2
Statement Function Restrictions
13.2.3
EXTERNAL FUNCTIONS . • • . • • • • • •
·
13.3
FORTRAN-Supplied External Functions
13.3.1
User-Defined External Functions
· •
13.3.2
Function Subprogram Restrictions • •
• •
13.3.3
Using a Function Subprogram
· • •
13.3.4
• •
SUBROUTINES
• • . • • • • • • •
13.4
FORTRAN-Supplied Subroutines.
•
13.4.1
· •
ALCCHR Subroutine
• • • • . • ••
13.4.1.1
CDABS Function . • . • . •
13.4.1.2
CDCOS Subroutine
13.4.1.3
• •••..
CDEXP Subroutine •
13.4.1.4
CDLOG Subroutine
13.4.1.5
CDSIN Subroutine •
·
13.4.1.6
CDSQRT Subroutine
·
13.4.1.7
CHKDIV Subroutine
·
13.4.1.8
CLRFMT Subroutine
• •
13.4.1.9
DATE Subroutine
•
13.4.1.10
DIVERT Subroutine
•
13.4.1.11
DTOGA . . • • • •
•
13.4.1.12
DUMP Subroutine
••••
·
13.4.1.13
viii

13-1
13-2
13-12
13-14
13-14
13-15
13-16
13-16
13-17
13-18
13-18
13-20
13-21
13-21
13-22
13-24
13-24
13-25
13-25
13-26
13-26
13-27
13-28
13-28
13-29
13-29
13-30
13-30
13-30

13.4.1.14
ERRSET Subroutine
•. ••
· 13-31
13.4.1.1S
ERRSNS Subroutine
• • • . • . . • . . 13-32
13.4.1.16
E XITS u b r c· uti n e
• 13-33
13.4.1.17
FFUNIT Subroutine
. • • • • . . • 13-33
13.4.1.18
GTODA Subroutine • •
• 13-34
13.4.1.19
ILL Subroutine • •
13-34
13.4.1.20
LEGAL Subroutine • •
· • 13-34
13.4.1.21
MVBITS Subroutine
• 13-3S
13.4.1.22
OVERFL Subroutine
· 13-3S
13.4.1.23
PDUMP Subroutine • .
. • • • 13-36
13.4.1.24
QUIETX Subroutine
..•.
· . 13-36
13.4til.2S
SAVFMT Subroutine
. . . • . • 13-37
13.4.1.26
SAVRAN Subroutine
· 13-37
13.4.1.27
SETRAN Subroutine
• • • •
• 13-38
13.4.1.28
SORT Subrcutine
· . 13-38
13.4.1.29
SRTINI Subroutine
..••.
• 13-39
13.4.1.30
TIME Subrcutine
• • 13-39
13.4.1.31
TOPMEM Subroutine
• . • 13-40
13.4.1.32
TRACE Subroutine • •
• • • . • • • • . . 13- 41
13.4.2
. • • • 13-42
User-Defined Subroutines • •
13.4.2.1
SUBROUTINE Statement • • • • • • • • . . . • 13-43
13.4.2.2
CALL Statement • • • . .
. • • . . . • 13-44
13.4.2.3
Execution of a CALL Statement
• • 13-44
13.4.2.4
Actual Arguments for a Subroutine
• • 13-4S
13.4.3
ENTRY Statement
•••• •
. .•.
· 13-4S
13.4.4
. • • 13-47
RETURN Statement • . • • . • • • . .
13.4.S
Dummy and Actual Arguments • • • • • •
• • 13-49
13.4.S.1
Length of Character Dummy and Actual
Arguments
••••. • • • • • . • . .
· 13-51
13.4.S.2
Character and Hollerith Constants as Actual
13-51
Argumen ts
HAPTER 14

14.1
HAPTER IS

BLOCK DATA SUBPROGRAMS
BLOCK DATA STATEMENT •

14-1

WRITING USER PROGRAMS

IS.1
lS.l.l
lS.1.2

lS-l
GENERAL PROGRAMMING CONSIDERATIONS
Accuracy and Range of Double-Precision Numbers
lS-1
Writing FORTRAN Programs for Use on Other
lS-2
Computers
•.•••••.•••••••••.
IS-2
lS.1.3
Using Floating-Point DO Loops
•••••.
lS.1.4
Computation of DO Loop Iterations
lS-2
15-3
Subroutines - Programming Considerations
lS.l.S
15-4
lS.1.6
Reordering of Computations • • •
. ••.
15-5
lS.1.7
Dimensioning of Dummy Arrays • • • • . • •
lS.2
. • • . • • .
lS-6
FORTRAN GLOBAL OPTIMIZATION
15-6
IS.2.1
Optimization Techniques
•••••••••
lS.2.1.1
Elimination of Redundant Computations
15-6
lS-7
lS.2.1.2
Reduction of Operator Strength • • •
Removal of Constant Computation from Loops .
15-7
lS.2.1.3
15-8
Constant Folding and Propagation
lS.2.1.4
lS.2.1.S
Removal of Inaccessible Code •
IS-9
lS-9
lS.2.1.6
Global Register Allocation • • . •
lS.2.1.7
lS-9
I/O Optimization • . • • • • • . •
15-9
Uninitialized Variable Detection • . • • • .
lS.2.1.8
15-9
lS.2.1.9
Test Replacement • •
• • • • •
Programming Techniques for Effective
lS.2.2
Optimization. . • • • • • • • . . •
• lS-10
FUNCTION SIDE EFFECTS
• • • • • • . • • . lS-10
lS.3
INTERACTING WITH NON-FORTRAN PROGRAMS AND FILES
lS-10
lS.4
ix

15.4.1
15.4.2
15.4.3
15.4.4
15.4.5
15.4.6
15.4.7
15.4.7.1
15.4.7.2
15.4.8
15.4.8.1
15.4.8.2
15.4.9
15.4.10
15.4.11

CHAPTER 16

Using The Sharable High-Segment FOROTS • • • • 15-10
Calling Sequences
• • • • • 15-11
Accumulator Usage
• • 15-11
Ar g urn e n t Lis t s . •
15 -1 2
Argument Types • • •
. • • • • • • • • • 15-14
Description of Arguments
• • • • • . • • • 15-14
Interaction with COBOL. •
• • • • 15-16
Calling FORTRAN Subprograms from COBOL
Programs • • • • • • . • • • • • • • • • • • 15-16
Calling COBOL Subroutines from FORTRAN
15-18
Programs . • • • • • • • • • • •
• 15-18
Interaction with BLISS-36 • • • •
Calling FORTRAN Subprograms From BLISS-36
Programs • • • • • • • • • • • • • • •
• 15-18
Calling BLISS-36 Routines From FORTRAN • • • 15-19
LINK Overlay Facilities . • • • • • • • • • • 15-19
FOROTS and Overlay Memory Management • • • • • 15-20
Extended Addressing Memory Layout (TOPS-20
only)
• • • • • • • • • • • • • • • • • • • • 15-22
USING THE FORTRAN COMPILER

USING THE FORTRAN-I0 COMPILER
TOPS-I0 COMPILE-Class Commands • •
RUNNING THE FORTRAN-IO COMPILER
TOPS-I0 Compiler Command Switches
• • • •
USING THE FORTRAN-20 COMPILER
TOPS-20 COMPILE-Class Commands • •
RUNNING THE FORTRAN-20 COMPI LER • • • • • • •
TOPS-20 Compiler Commands Switches
•••••
THE /DEBUG SWITCH • • . • • • • • •
• • •
THE /NOWARN SWITCH • • • . • • • • •
• • •
THE /EXTEND SWITCH (TOPS-20 ONLY)
/EXTEND and Applications with Large Arrays • •
/EXTEND and Applications with Large Executable
Code • • • • • • • • • • • • • • • • •
•
/EXTEND PSECT Placement
•• • • •
• •
16.5.2.1
16.5.2.2
Building Large-Code Applications..
••
Arguments to /EXTEND • • • • • • • • • • • • •
16.5.3
Linking With TWOSEG REL Files
•
16.5.4
THE /FLAG (/FLAG-NON-STANDARD) SWITCH
•
16.6
16.7
READING A FORTRAN COMPILER LISTING • • • • • • •
16.7.1
Compiler-Generated Variables • • • •
•
16.8
ERROR REPORTING
. • • • • • • • • •
•
Fatal Errors and Warning Messages
••
16.8.1
Message Summary
• • • •
• •
16.8.2
CREATING A SHARABLE HIGH SEGMENT FOR A FORTRAN
16.9
PROGRAM • • • • • • • • • • • •
• • • •••
16.1
16.1.1
16.1.2
16.1.3
16.2
16.2.1
16.2.2
16.2.3
16.3
16.4
16.5
16.5.1
16.5.2

CHAPTER 17
17.1
17.1.1
17.1.2
17.1.3
17.2
17.3
17.4
17.5
17.6
17.7
17.8

16-1
16-1
16-2
16-2
16-5
16-5
16-6
16-8
16-11
16-14
16-15
16-16
16-17
16-17
16-18
16-19
16-21
16-21
16-22
16-24
16-33
16-34
16-34
16-35

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)

..····· ····
····
·

INPUT FORMAT
Variables and Arrays
Constant Conventions
Statement Labels and Source Line Numbers
FORDDT AND THE FORTRAN /DEBUG SWITCH
LOADING AND STARTING FORDDT
SCOPE OF NAME AND LABEL REFERENCES
FORDDT COMMANDS
ENVIRONMENT CONTROL
FORTRAN /OPTIMIZE SWITCH
CALLING FORDDT

····
·· . . .
·····
. . .·
· · · ·
. ·
. · ·· · · ·· · ·
·
x

17-2
17-2
17-3
17-4
17-4
17-5
17-6
17-7
17-16
17-17
17-17

17.9
CHAPTER 18
18.1
18.2
18.3
18.3.1

FORDDT AND FORTRAN-20 EXTENDED ADDRESSING

• 17-17

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
FEATURES OF FOROTS • • • • • • • • • •
ERROR PROCESSING • • • •
• • • • •
INPUT/OUTPUT FACILITIES
• • •
Input/Output Channels Used By FOROTS

• • • •.
• • • ••
• • • ••
(TOPS-10

18-1
18-2
18-2

18-2
On 1 y)
• • • • • '. • • • •
• • • • •
18.3.2
File Access • • • • • • •
• • • • • 18-3
18.3.3
Closing Files After Non-standard Termination. 18-3
18.3.3.1
Sequential Access
••••••
• • • • 18-3
18.3.3.2
Direct (Random) Access Mode
• • • • • • • • 18-4
18.4
ACCEPTABLE TYPES OF DATA FILES AND THEIR FORMATS 18-4
18-4
18.4.1
ASCII Data Files • • • • • •
18.4.2
FORTRAN Binary Data Files • • • •
18-4
• • • ••
18-4
18.4.2.1
Format of Binary Files.
18.5
USING FOROTS . • • • • •
• • 18-14
18.5.1
FOROTS Entry Points
• • • • • • • 18-15
• • • 18-16
18.5.2
Calling Sequences • • • • • • •
18.5.3
MACRO Calls for FOROTS Functions.
• 18-17
18-18
18.5.3.1
Sequential-Access Calling Sequences
18-19
18.5.3.2
Internal Pile Calling Sequences • • • •
18.5.3.3
NAMELIST I/O, Sequential-Access Calling
Sequences • • • • • • • • • • • • • •
• 18-20
18-20
Array Offsets and Factoring
••.••
18.5.3.4
I/O Statenents, Direct-Access Calling
18.5.3.5
Sequences • • • • • • • • • • • . •
• • 18-22
Default Devices Statements, Calling
18.5.3.6
Sequences • • • • • • • • • • • • • • • • • 18-23
18.5.3.7
Statements to position Files. • • •
• • 18-24
18.5.3.8
List-Directed Input/Output Statements . • • 18-24
18.5.3.9
Input/Output Data Lists
•••••••
18-25
18.5.3.10
OPEN and CLOSE Statements, Calling Sequences 18-28
18.5.3.11
Memory Al:.ocation Routines • • • • • • • • • 18-29
18.5.3.12
Channel A:1ocation and Deallocation Routines 18-30
18.6
FUNCTIONS TO FACILITATE OVERLAYS. •
• 18-31
18.7
LOGICAL/PHYSICAL DEVICE ASSIGNMENTS
• • • 18-36
18.8
F'OROTS AND INQUIRE BY FILE STATEMENT • • • • • • 18-36
CHAPTER 19

USING THE FORr;"'RAN RE:AL-TIME SOFTWARE (TOPS-I0 ONLY)

INTRODUCTION
19.1
19.2
UlSING FORRTF
19.2.1
Memory
19.2.2
Modes
Priority-Interrupt Levels
19.2.3
Masks
19.2.4
SUBROUTINES
19.3
19.3.1
LOCK
RTINIT
19.3.2
CONECT
19.3.3
RTSTRT
19.3.4
BLKRW
19.3.5
RTREAD
19.3.6
19.3.7
RTWRIT
19.3.8
STATO
STATI
19.3.9
RTSLP
19.3.10
RTWAKE
19.3.11
DISMIS
19.3.12
DISCON
19.3.13

·.

.·
·

·
··

····
·········
····
·····
····
·
·
· · · ·· ·· · ·
·
·····
·· ·· ·· ·
·····
····
·· ·· ·· ··
····
·
· · ·· ·
·· ·· ··· ··· ··
·
· ·
····
···· ·········
····
····
xi

19-1
19-2
19-2
19-2
19-2
19-2
19-3
19-3
19-3
19-4
19-4
19-4
19-5
19-5
19-5
19-6
19-6
19-6
19-6
19-7

19.3.14
19.3.15

19-7
19-7

UNLOCK • • • • •
Error Messages •

APPENDIX A

SUMMARY OF FORTRAN STATEMENTS

APPENDIX B

ASCII-196S CHARACTER CODE SET

APPENDIX C

COMPILER MESSAGES

APPENDIX D

FOROTS ERROR MESSAGES

D.1

ALPHABETICAL DESCRIPTION OF FOROTS MESSAGES

APPENDIX E

INTERACTIVE DEBUGGER (FORDDT) ERROR MESSAGES

APPENDIX F

FORTRAN-SUPPLIED PLOTTER SUBROUTINES

• • • D-9

PLOTS SUBROUTINE
• • F-2
F.1
AXIS SUBROUTINE
• • • •
• F-2
F.2
F.3
CAXIS SUBROUTINE •
• • • • • • •
• F-3
LINE SUBROUTINE
. • • • •
. F-4
F.4
MKTBL SUBROUTINE •
F-5
F.S
Character Tables • • • • •
• • • • . F-S
F.S.1
Creating a Character Table •
• • F-6
F.S.1.1
Creating a Stroke Table
• • . F-6
F.S.1.2
Sample Character Stroke Table
. F-7
F.S.1.3
FORTRAN- and User-Defined Character Sets • • • F-9
F.S.1.4
F.6
NUMBER SUBROUTINE
• • ••
• • • • • . F-9
PLOT SUBROUTINE
• • • • • • • •
• • • • F-10
F.7
F.S
SCALE SUBROUTINE • • • • • • • • • • •
F-10
SETABL SUBROUTINE
• • • • •
. • ••
F-11
F.9
SYMBOL SUBROUTINE
F-12
F.10
WHERE SUBROUTINE •
F-12
F.11
INDEX
FIGURES
2-1
4-1
6-1
7-1
7-2
7-3
7-4
7-5
9-1
10-1
11-1
11-2
15-1
F-1
F-2
F-3

Fields Within a FORTRAN Line • •
A 3 x 3 x 2 Array
• • . • • • •
• • • •
Ordering of FORTRAN Statements.
Shared Storage using EQUIVALENCE Statement •
Equivalence of Substrings
.••••••••••
Equivalence of Character Arrays
• • • • • •
Valid Equivalencing
••••••
Invalid Equivalencing
•••••
Examples of Block IF Constructs
Components of Data Transfer Statements •
.
TOPS-10 File Protection Number • •
TOPS-20 Protection Number
••••
• • •
• •
Run-time Memory Layout for Section Zero
Plotter Character Table Entry
••••
Character Stroke Table Entry • •
Sample Character Stroke Table
.•••

xii

.. 2-3
..

4- 4

. 6-3
7-12
7-13
7-14
7-16
7-16
.. 9-7
10-S
11-25
11-27
15-21
.. F-6
• F-6
.. F-S

TABLES
2-1
3-1
4-1
5-1
5-2
5-3
5-4
5-5
5-6
5-7
8-1
10-1
10-2
10-3
10-4
10-5
10-6
10-7
10-8
10-9
10-10
10-11
11-1
11-2
11-3
11-4
11-5
11-6
12-1
12-2
12-3
12-4
12-5
12-6
13-1
13-2
15-1
15-2
16-1
16-2
16-3
16-4
16-5
16-6
16-7
16-8
17-1
18-1
18-2
19-1
D-l

FORTRAN Character Set
•• • • • • • • • • • • .
Comparison of Real, D-floating, and G-floating

• 2-2

Numbe r 5

•

••

•

3-4

Use of Symbolic Names
• • • • • • • • • • • • . • 4-1
Arithmetic Operations and Operators
• • • • • • • 5-2
~rype of the Rt~sul t Obta ined from Mixed-Mode
Operations • • • • • • • • • • • • • • . •
. • 5-3
Permitted Base/Exponent Type Combinations
. • • • 5-5
Logical Operators
• • • • • •
• 5-8
Logical Opera:ions Truth Table • • • •
5-9
Relational Operators and Operations
5-11
Hierarchy of PORTRAN Operators • •
5-14
Rules for Conversion in Mixed-Mode Assignments • • 8-2
FORTRAN I/O S':atement Categories.
10-2
Summary of Da~a Transfer Statement Forms •
10-3
FORTRAN Logical Device Assignments •
10-12
Summary of READ Statement Forms
• • • • 10-25
Summary of WRITE Statement Forms • .
. • • 10-31
Summary of REHEAD Statement Forms
• • • • • • • 10-38
Summary of ACCEPT Statement Forms
• • 10-40
Summary of TYPE Statement Forms
• 10-43
Summary of PRINT Statement Forms .
• . 10-45
Summary of PUNCH Statement Forms • •
. 10-47
Summary of Internal READ/WRITE and ENCODE/DECODE
Statement Forns
• • • • • • • • . •
. 10-49
Summary of OPEN Statement Specifiers and
11-5
Arguments
.••••••••
. • 11-23
DEVICE and MODE Cross-Table
'I'OPS-I0 Protection Code Values
· 11-26
'I'OPS-20 Protection Code Val ues
••••
· 11-27
Summary of CLOSE Statement Specifiers and
Arguments
• • • • • • • .•
• • • 11-35
Summary of Device-Control Statements •
. . . 11-51
Repeatable FORTRAN Edit Descriptors
12-8
12-9
Nonrepeatable FORTRAN Edit Descriptors
12-11
Carriage-Control Specifiers
•••••
Eecord, Format List" and I/O List Interaction
· 12-13
Default Field Widths for Numeric Edit
Descriptors
• • . • • • . • • • • . • . . • • 12-30
Effect of Data Magnitude on G-Format Output
Conversion • , • • • • •
• • • • . • • • 12-35
FORTRAN Instrjnsic Functions .
13-4
FORTRAN-Suppl~,ed Subroutines.
• . • • •.
• 13-23
Argument Types and Types Codes
. . 15-14
~[emory Allocations for /EXTEND and /NOEXTEND • • 15-23
FORTRAN-I0 Compiler Switches. . • . .
16-3
FORTRAN-20 Compiler Switches.
••••.
16-9
JI,rg umen ts to /DEBUG Swi tch • •
•....
16 -12
Legal Dummy and Actual Argument Associations • . 16-14
Arguments to /NOWARN Switch
••••
16-15
/EXTEND DefauJt Memory Layout
. • • • • 16-16
Arguments to /EXTEND Switch
••••
. • • 16-20
Arguments to /FLAG Switch
• • • • ••
. 16-22
FORDDT Commands
• • • • • • • .
17-1
FOROTS Entry Points
••• . •
. • • 18-15
Function Numbers and Function Codes
. • . 18-32
Error Messages - Code Format and Full Message
Format
. • • • • • • • •
19-8
. • D-2
FOROTS Error Codes • • • • •

xiii

PREFACE

The FORTRAN-IO/20 Language Manual reflects the software as of Version
10 of the FORTRAN-IO/20 compiler, the FORTRAN-IO/20 Object Time System
(FOROTS), and the FORTRAN-IO/20 debugging program (FORDDT).
This manual describes the FORTRAN language as
implemented
for
the
TOPS-IO operating system (FORTRAN-IO) and the TOPS-20 operating system
(FORTRAN-20).
Any differences between FORTRAN-IO and
FORTRAN-20 are
noted in this manual.
Since this is a reference manual, we assume that you have used FORTRAN
before.
If you haven't, you should read one of the many introductory
FORTRAN texts.
CONVENTIONS
The following conventions are used throughout the manual:
Braces {

indicate that a choice must be made from
of the enclosed lines.

Brackets [ ]

indicate an optional feature.

one

or.

indicate the omission of information from a
programming
example or that items in a
command line can be optionally repeated.

Lowercase letters

indicate variable information you supply in a
command string.

UPPERCASE LETTERS

indicate fixed (or literal) information that
you must enter as shown in a command string.

Ellipsis

indicates a blank.
The standard for FORTRAN is the American National Standards Institute
(ANSI)
FORTRAN,
X3.9-1978 (also known as FORTRAN-77).
FORTRAN-IO/20
extensions and additions to ANSI FORTRAN are in blue print
in
this
manual.

MANUALS REFERENCED
The following manuals are referenced from TOPS-IO pUblications:

•
•
•

TOPS-IO Operating System Commands Man
uaI
-SOS Reference Manual
TOPS-IO Moni tor Calls
- - Manual

• TOPS-IO Hardware Reference Manual
• TOPS-IO LINK Reference Manual
•
•

TOPS-IO SORT/MERGE User's Guide
TOPS-IO FORTRAN Installation Guide

The following manuals are referenced from TOPS-20 pUblications:

•

TOPS-20 Commands Reference Manual

• TOPS-20 EDIT Reference Manua I
•
•
•
•
•

TOPS-20 User's Guide
TOPS-20 Monitor Calls Manual
TOPS-20 Link Reference Manual
TOPS-20 SORT/MERGE User's Guide
--TOPS-20 FORTRAN Installation Guide

---

The following TOPS-IO/TOPS-20 manual are referenced:
•

FORTRAN-IO/20 and VAX-II FORTRAN Compatibility Manual

•

TOPS-IO/TOPS-20 FORTRAN Pocket Guide

•

TOPS-IO/TOPS-20 COBOL-74 Language Manual

•

TOPS-IO/20 BLISS Language Guide

•

TOPS-IO/20 Common Math Library Manual

xvi

CHAPTgR 1
INTRODUCTION

1.1

OVERVIEW

The FORTRAN language, as
implemented on the TOPS-IO and TOPS-20
operating systems,
is compatible with and encompasses the standard
described in "AmE~rican National Standard FORTRAN, X3.9-1978" (referred
to as the FORTRAN-77 standard) at the full-language level.
FORTRAN-10/20 provides many extensions and additions to the FORTRAN-77
standard that 9reatly enhance the usefulness of FORTRAN and increase
its compatibility with FORTRAN languages implemented by other computer
manufacturers.
The
extensions
and additions to the standard
FORTRAN-77 are printed in this manual in blue print.
A FORTRAN source program cons~sts of a set of statements constructed
using the language elements and the syntax described in this manual.
A given FORTRAN statement performs anyone of the following functions:
1.

It causes operations such as
branching to be carr led out.

It specifies the

It specifies the characteristics of the source program.

tYPE~

multiplication,

division,

and

and format of the data being processed.

FORTRAN statements are composE~d of keywords (words that are recognized
by the compiler) used with elements of the language set:
constants,
variables, and expressions.
~here are
two basic types of FORTRAN
statements:
executable and nonexecutable.
Executable statements
spec~.fy
the
actions
of
the
program;
nonexecutable statements describe the characteristics and arrangement
of data, editing information, statement functions,
and
the kind of
subprograms that may be incl.uded in the program.
The compilation of
executable statements results in the creation of executable code in
the object program.
Nonexecutable statements provide information only
to the compiler; they do not create executable code.

1.2

MANUAL ORGANIZATION

This

manual

is divided
into three parts:
Language
Elements,
and Language Usage.
Sections 1.2.1 through 1.2.3 contain
general descriptions for each of these three parts.
Section 1.2.4
contains a general description of the appendixes.
Stater~nts,

1-1

INTRODUCTION
1.2.1

FORTRAN Language Elements (Part I)

Part I of this manual describes the fundamental
programs, including (by chapter):

elements

FOHTRAN

Chapter 2, CHARACTERS AND LINES, describes FORTRAN source program
characters and lines.
The FORTRAN compiler
interprets your
source program and translates it into machine code
(executable
code) •
Chapter 3, CONSTANTS, describes FORTRAN data types and constants.
FORTRAN enables you to manipulate information (data) in a variety
of ways. This chapter describes the techniques for defining
FORTRAN constants of various data types.
Chapter 4,
SYMBOLIC NAMES,
VARIABLES,
AND ARRAYS, describes
symbolic names, variables, and arrays in FORTRAN.
The symbolic
name is used in a variety of ways
in a source program;
this
chapter describes the conventions for using symbolic names to
define both variables and arrays.
introduces
and
describes
FORTRAN
Chapter 5,
EXPRESSIONS,
operators.
Expressions in FORTRAN are formed using variables,
constants, and operators.

1.2.2

FORTRAN Statements (Part II)

Part II of this manual describes all the statements in the FORTRAN
language.
The following list describes (by chapter) the information
presented in Part II:
Chapter 6, EXECUTABLE AND NONEXECUTABLE STATEMENTS, defines the
term "FORTRAN Statement", and describes the "Compilation Control
Statements".
Chapter 7, SPECIFICATION AND DATA STATEMENTS, describes all the
statements
within
the
category
"Specification
and Data
Statements".
The specification statements
enable
you
to
explicitly define the data types of variables used within your
program; the DATA statement enables you to create initially
defined constants within your program.
Chapter 8, ASSIGNMENT STATEMENTS, describes all the statements
within the category "Assignment Statements". FORTRAN assignment
statements enable you to assign values to variables, and to
assign statement labels to symbolic names.
Chapter 9, CONTROL STATEMENTS, describes all the statements
within the category "Control Statements".
The default execution
sequence of lines in a FORTRAN program is each line from
left-to-right, and all lines from top-to-bottom.
You use the
FORTRAN control statements to alter
the default
execution
sequence,
to stop or pause during program execution, or to mark
the end of an executable program.
Chapter 10, DATA TRANSFER STATEMENTS, describes the data transfer
category of "FORTRAN Input/Output (I/O) Statements".
As the term
implies, a data transfer statement moves data from one place to
another.

1-2

INTRODUCTION
Chapter 11, FILE-CONTROL AND DEVICE-CONTROL STATEMENTS, describes
file-control
and devi~e-control categories of "FORTRAN I/O
Statements".
The file-control statements enable you to associate
a unit number with a file.
Device-control statements enable you
to position a storage medium (for example, magnetic tape)
on a
connected unit.
Chapter 12, FORMATTED DATA TRANSFERS, describes three types of
data
formatting.
During
certain types of data transfer
operations, you must specify the format of the data being
transferred.
FORTRAN provides three techniques for specifying
the format of data:
FORMAT-Statement,
List-Directed,
and
NAMELIST-Statement formatting.
Chapter 13, FUNCTIONS AND
SUBROUTINES,
describes
FORTRAN
functions and subprograms.
Functions and subprograms provide a
technique for producing clear and concise FORTRAN programs.
FORTRAN-IO/20 provides both predefined functions and subprograms,
and the statements for defining your
own
functions
and
subprograms.
Chapter 14, BLOCK DATA SUBPROGRAMS, describes the block-data
subprogram.
This type of subprogram enables you to define
initial values for variables in COMMON.

1.2.3

FORTRAN Language Usage (Part III)

Parts I and II of the manual contain complete descriptions of FORTRAN
elements and statements.
Part III of the manual contains explanations
of how you use FORTRAN-IO/20.
The following usage topics are covered
in Part III:
Chapter 15 , WRITING USER PROGRAMS, presents
some
general
considerations that you should follow when you are creating
FORTRAN source programs.
In addition, this chapter contains a
description of the FORTRAN optimizer.
Chapter 16, USING THE FORTRAN COMPILER, describes how to use the
FORTRAN compiler and contains descriptions on how to compile,
load, and execute a FORTRAN program.
In addition,
this chapter
contains descriptions of how to read a compiler-generated program
listing, and how to create a reentrant FORTRAN program.
This
Chapter also describes how to use FORTRAN-20 extended addressing.
Chapter 17, USING THE FORTRAN INTERACTIVE DEBUGGER
(FORDDT),
describes how to use the FORTRAN interactive debugging program
(FORDDT) to test and debug a running program. This chapter also
contains a brief explanation of how to debug a running FORTRAN
program using DDT, the system debugger.
Chapter 18, USING THE FORTRAN OBJECT-TIME SYSTEM
(FOROTS),
describes the FORTRAN Object-Time System (FOROTS).
This chapter
also contains descriptions of how you can use the FOROTS
software.
Chapter 19, USING THE FORTRAN lREAL-TIME SOFTWARE (TOPS-IO ONLY),
describes how to use the FORTRAN real-time software.
This
chapter is for TOPS-IO installations only.

1-3

INTRODUCTION
1.2.4

APPENDIXES

The appendixes describe various useful
topics are covered in the appendixes:

The

information.

following

Appendix A, SUMMARY OF FORTRAN STATEMENTS, summarizes the forms
of all FORTRAN statements and provides a section reference where
each statement is described in detail.
Appendix B, ASCII-1968 CHARACTER CODE SET,
lists the character
code set defined
in the X3.4-1968 version of the American
National Standard Code for Information Interchange (ASCII).
Appendix C, COMPILER MESSAGES,
messages.

describes

the

FORTRAN

compiler

Appendix D, FOROTS ERROR MESSAGES,
messages.

describes

the

FOROTS

error

Appendix E, FORDDT ERROR MESSAGES,
messages.

describes

the

FORDDT

E!rror

describes

the

Appendix F, FORTRAN-SUPPLIED PLOTTER SUBROUTINES,
FORTRAN-supplied plotter subroutines.

1-4

CHAPTER 2
CHARACTERS AND LINES

The basic elements of the FORTRAN source program are its characters
and lines.
Characters are l]sed to form statements, expressions, and
comments in FORTRAN source programs.
Lines, and fields within lines,
are used to define the context in which characters are interpreted by
the FORTRAN compiler.
program

This chapter describes the relationships among
source
characters, lines, and fields within source program lines.

2.1

CHARACTER SET

Table 2-1 lists the digits,
letters, and symbols recognized by
FORTRAN.
The remainder of the ASCII-1968 character set is acceptable
within character or Hollerith constants or comment text,
but these
characters cause fatal errors in other contexts.
NOTE
The complete ASCII chdracter set is defined
in the
X3.4-1968 version of the "American National Standard
Code for Information Interchange".
A summary of the
standard ASCII set ia also contained in Appendix B of
this manual.
NOTE

Lowercase alphabetic
upper-case
outside
character constants.

characters
are
the context of

2-1

treated
Hollerith

as
or

CHARACTERS AND LINES
Table 2-1:

FORTRAN Character Set
Letters

Uppercase:

ABC D E F G H I J K L M N 0 P Q R STU V W X Y Z

Lowercase:

abc d e f 9 h i j

kIm n

p q r stu v w x y z

Digits

o 1 234 567 8 9
Symbols
!

#
$
&

(
)

Exclamation point
Quotation Mark
Number Sign
Dollar Sign
Ampersand
Apostrophe
Left Parenthesis
Right Parenthesis
Asterisk
Plus

:
;

=
>
,.

Comma
Hyphen (Minus)
Period (Decimal Point)
Slant (Slash)
Colon
Semicolon
Less Than
Equal To
Greater Than
Ci rcumflex

Line Termination Characters
Line Feed

(LF), Form Feed

(FF), Vertical Tab (VT)

Line Formatting Characters
Carriage Return (RET), Horizontal Tab (TAB), Blank
Note that horizontal tabs normally advance the character position
pointer to the next position that
is an even multiple of 8.
An
exception to this is the initial tab, which is defined as a
tab that
either
includes or starts in character position 6.
(Refer to Section
2.3.1 for a description of initial and continuation line types.)
Tabs
within character specifications count as one character, even though
they may advance the character position as many as eight places.

2.2

STATEMENT DEFINITION AND FORMAT

Source program statements are divided into physical lines.
A line is
defined as a string of adjacent character positions, terminated by the
first occurrence of a
line termination character,
regardless of
context.
As shown in Figure 2-1, each source program line is divided
into four fields.

2-2

CHARACTERS AND LINES

t-ol
. . . .- - - · - - - - - -

Figure 2-1:
2.2.1

Line Character Positions

---------~·~I

2 !-~---___v--7-0--7-1-7-2-'/~
Statement Continuation
Statement Field
Remarks
Label Field
Field
Mf1.S-176181

Fields Within a FORTRAN Line

Statement Label Field and Statement Numbers

You can place a number ranging from 1 to 99999 in the statement label
field of an
initial
line to
identify the statement.
Any source
program statement that is referenced by another statement must have a
statement number.
Leading zel~OS and all blanks in the label field are
ignored; for example, the numbers 00105 and 105 are both accepted as
statement number 105.
You can assign the statement numbers in a source program in any order;
however,
each statement number must be unique with respect to all
other statements in the program or subprogram.
A main program and a subprogram can contain
identical statement
numbers.
In this case, references to these numbers are understood to
mean the numbers in the same program unit in which the reference
is
made. An example follows:
Assume that main module MAINMD and subprogram SUBI both
contain statement numbE~r 10~).
A GO TO 105 statement, for
instance, in MAINMD will refer to statement 105 in MAINMD,
not to 105 in SUBI.
A
GO TO 105 in SUBI will transfer
control to 105 in SUBI.
An initial tab may be used to skip all or part of the label field.
If
an
initial tab is encountered during compilation,
FORTRAN-IO/20
examines the character immediately following the tab to determine the
type of line being entered.
If this character is one of the digits 1
through 9, FORTRAN-IO/20 cons:.ders the line as a continuation line and
uses the second character after the tab as the first character of the
statement field.
If the character following thE! tab is not one of the digits 1 through
9,
FORTRAN-10/20 considers the line to be an initial line and the
character following the tab i~; considered to be the first character of
the statement field.
The character
following
the initial tab is
considered to be in character position 6 for a continuation line, and
in character position 7 for arl initial line.

2.2.2

Line Continuation Field

Any character of the FORTRAN character set (except a blank, a zero, or
an exclamation point) placed in the line continuation field (position
6) identifies the line as a continuation line
(see Section 2.3.1).
Whenever you use an initial tab to skip all or part of the label field
of a continuation line, the next character you enter must be one of
the digits I through 9 to identify the line as a continuation line.

2-3

CHARACTERS AND LINES
2.2.3

Statement Field

Any FORTRAN statement can appear
in
the
statement
field.
Blanks
(spaces)
and tabs do not affect compilation of the statement.
Blanks
and tabs may be used freely in this
field
for
appearance purposes,
with
the exception of textual data given within either a character or
Hollerith specification,
where blanks and
tabs are
significant
characters.

2.2.4

Remark Field

In lines consisting of 73 or more character positions, only the
first
72 character
positions are
interpreted by FORTRAN.
Note that tabs
generally occupy more than one character position,
usually advancing
the cursor to the next character position that is an even multiple of
8.
The pxcept:.ion is the tab in a
label
field,
which advances
the
"llrSOr
either
to column 6 or 7, deoendinq on the character following
1-. he' tab (see Section 2.2.1).
All other characters after
character
remarks and do not affect compilation.

position

are

treated

t-t1at remarks may also be added to a line jn character positions 1
L;-:r'onoh
72, provided the text of the remark is preceded by the symbol
nTH
(see Section 2.3.3).

:JOt-p

2.3

LINE TYPES

A line in a FORTRAN source program may be:

An initial line

A continuation line
i\

multi-statement line

A comment line
i\

debuo 1 j ne

A blank line

These lines are described in Sections 2.3.1 through 2.3.5.

2.3.1

Initial and Continuation Lines

A FORTRAN statement may occupy the
statement
fields
of
up
to
20
consecutive lines.
The
first
line
in a multi-line statement is
referred to as the initial line; the succeeding lines are referred
to
as continuation
lines.
Initial
lines may be assigned a statement
number and must have either a blank or a zero in character position 6.
'nltl~l
, ; S (Co
i j · ii

; prni--'rd.

tab may be used to skip all or part of the label field.
If
i nIt i a 1 tab for t his 0 u r po s e , yo u mus t i mm e d i ate 1 y follow
:wnrJUmerjc character; that is, the fjrst
character of
the
i leid must be nonnumeric.

.H)

2-4

CHARAc'rERS AND LINES
You cannot assign a statement label to a continuation line.
Instead,
you
identify a continuation line by placing a character from the
FORTRAN character set (except blank, zero, or exclamation point)
in
character position 6 of that line.
This position
is
the line
continuation field.
The label field of a continuation line must be
blank.
Note that blank lines, comments, and debug lines that are treated like
comments
(that is, debug line; that are not compiled with the rest of
the program) are legal continuation lines and do not terminate a
continuation sequence (see Section 2.3.4).
The following is an example of a 3-line FORTRAN FORMAT statement
two continuation lines:
105

with

FORMAT (IX, 'This eX.3mple shows how continuation lines I,
2 'are used to accommodate FORTRAN statements that do not'
3 'entirely fit on a single line.')

In this example the characters 2 and 3 in position
lines as continuation lines.

identify

those

Mul t i-Sta tE~ment Li nes

2.3.2

You may write more than one FO~TRAN statement in the statement field
of one line.
The rules for structuring a multi-statement line are:
1.

Successive statements must be separated by a semicolon (;).

Only the first statement in the series can have
number.

The last statement in a line is continued to the next line if
that next line is mad,? a continuation line.

statement

An example of a multi-statement line is:
450

DIST=RATE

* TIME; TIME=TIME+O.05; CALL PRIME(TIME,DIST)
NOTE

If
a
statem?nt
sequence
in
a
multi-statement
line
consists of a
logical IF (see Section 9.2.2)
followed
by any other executable statement, then
the statement f~llowing the IF will be
executed
in all cases, even if the IF
expression evallates as false.

2.3.3

Comment Lines and Remarks

Lines that contain descriptive text only are called comment lines.
Comment lines commonly identify and
introduce a
source program,
describe the purpose of a particular set of statements, and
introduce
subprograms.

2-5

CHARACTERS AND LINES
To structure a comment line:
1.

You must place one of the characters C (or c), *, $, /, or
in character position 1 of the line to identify it as a
comment line.

You place the text of the comment in
line.

You may place comment lines anywhere in the
including preceding a continuation line.

You may write a large comment as a sequence of any number of
lines;
however,
each line must carry the
identifying
character (C (or c), *, $, /, or !) in
its first character
position.

the

remainder
source

The following is an example of a comment that occupies more
line:
C
C
c
c

the

program,

than

one

SUBROUTINE - A12
This subroutine formats
and stores the results of
the HEAT-TEST program

Comment lines are printed on all listings, but are
by the compiler.

otherwise

ignored

You may add a remark to any statement field, in character positions
7
through 72, provided the symbol!
precedes the text.
For example, in
the line
IF(N.EQ.O)STOP

! Stop if card is blank

the text "Stop if card is blank" is identified as a
remark by the
preceding
symbol.
The compiler ignores all characters from the
exclamation point to the end of the line.
The characters
following
the exclamation point, however, appear in the source program listing.
To be treated as a remark symbol,
the exclamation point must not
appear in a Hollerith or character constant.
Note that characters appearing in character positions
73 and beyond
are automatically treated as
remarks, so that you need not use the
symbol
(see Section 2.2.4).

2.3.4

Debug Lines

As an aid in program debugging, a D (or d) in character position 1 of
any line causes the line to be interpreted as a comment line; that is,
not compiled with the rest of the program unless the /INCLUDE switch
is present
in
the compiler command string.
(See Chapter 16 for a
description of the compiler switches.)
When the /INCLUDE switch is present in the compiler command string,
the D (or d) in character position 1 is treated as a blank so that the
remainder of the line is compiled as an ordinary (noncomment) line.
A
debug
line can have a label following the D (or d).
Note that if the
debug statement is an initial line, all of its continuation lines must
contain a D (or d) in character position 1.

2-6

CHARAC~lERS

2.3.5

AND LINES

Blank Lines

You may insert lines consisting of only blanks, tabs, or no characters
anywhere in a FORTRAN source program.
Blank lines that contain
remarks only, are considered as blank lines. Blank lines are used for
formatting purposes only;
they cause blank lines to appear in their
corresponding positions in source program listings; otherwise, they
are ignored by the compiler.

2.4

LINE-SEQUENCE:D SOURCE FILgS

FORTRAN-IO/20
accepts
I ine--sequenced
fi les
as
produced
by
line-oriented text editors
(for E~xample, SOS on TOPS-IO or EDIT on
TOPS-20).
These sequence numb(~rs are used in place of the listing
line numbers normally generated by FORTRAN.
The listing line numbers
are not the same as FORTRAN statement: numbers.

2-7

CHAPTEH 3
CONSTANTS

3.1

INTRODUCTION

Constants are quantities that do not change value during the execution
of the object program.
The data types you can use for constants in
FORTRAN-10/20 source programs are:
1.

Integer

Real

Double-precision

Complex

Character

Logical

Octal

Double-octal

Hollerith

10.

Statement: label

The use and format: of each
Sections 3.2 through 3.10.

3.2

these

data

types

are

discussed

INTEGER CONSTANTS

An integer constant is a string of one to eleven digits that
represents a whole decimal number
(a number without a fractional
part).
Integer constants must be within the range of -(2**35-1)
to
(+2**35)-1 (-34359738367 to +34359738367).
positive integer constants
may optionally be signed; negative integer constants must always be
signed.
You cannot use decjmal points, commas, or other symbols in
integer constants (except for a preceding sign, + or -).
Examples of valid integer constants are:
345
+345
-345

3-1

CONSTANTS
Examples of invalid integer constants are:
+34S.
3,4S0
34.S

3.3

(use of decimal point)
(use of comma)
(use of decimal point; not a whole number)

REAL CONSTANTS

A real constant can have any of the following forms:
1.

A basic real constant:
a string of decimal digits followed
by
a decimal point,
followed
optionally by a decimal
fraction, for e~ample, LSS7.42.

A basic real constant followed by a decimal integer exponent
written in E notation
(exponential notation)
form,
for
example, lSS9.E2 or lSS9.e2.
The number following the E (or
e)
specifies a power of ten by which the basic real constant
will be multiplied.

An integer constant (no decimal point) followed by a decimal
integer exponent written in E notation, for example, lSS9E2
or lSSge2.

Real constants may be of any size; however, each will
fit the precision of 27 bits (7 to 9 decimal digits).

rounded

Precision for real constants
is maintained to approximately eight
significant digits;
the absolute precision depends upon the numbers
involved.
The exponent field of a real constant written in E notation cannot be
empty
(blank);
it must be either a zero or an integer constant.
The
range of magnitude permitted a real constant is from approximately
1.47 * LO**(-39) to 1.70 * 10**(+38).
The following are examples of valid real constants:
-98.76S
7.0E+0
.7E-3
SE+S
SOlIS.
SO.El

(=
(=
(=

7.)
.0007)
SOOOOO.)

SOO.)

The following are examples of invalid real constants:
72.6ES12
.37SE
SOO

3.4

(exponent is too large)
(exponent incorrectly written)
(no decimal point given)

DOUBLE-PRECISION CONSTANTS

Double-precision constants are similar to real constants written in
notation form; the differences between these two constants are:
1.

Double-precision constants, depending on their magnitude,
have precision from 16 to 18 places, rather than the 8-digit
precision obtained for real constants.

3-2

CONSTAN~rs

constant

occupies

storage

two

Each
double-precision
locations.

The letter 0
(or (1),
instead
of
E,
is
used
in
double-precision constants to identify a decimal exponent.

On KL model B systems, there a~e two forms of double-precision number
formats.
If the /GFLOATING compiler switch is specified (see Chapter
16), the double-precision numbE?r form.at is called G-floating.
If the
/OFLOATING compiler switch
(the default) is specified (see Chapter
16), the double-precision number format
is called O-floating.
See
Section 3.4.1 for
a comparison of the different double-precision
number formats.
On KS systems, only the O-floating double-precision number
provided.

format

You must use both the letter 0 and an exponent
(including zero)
in
writing a double-precision constant.
The range of magnitude permitted
a double-precision constant is from approximately:
1.47 * 10**(-39)

to 1.70 * 10**(+38)

for O-floating

or
2.78 * 10**(--309)

to 8.99 * 10**(+307) for G-floating

The following are examples of valid double-precision constants:
7.9003
7.90+03
7.90-3
79003
7900

(=
(=
(=
(=
(=

7900.)
7900.)
.0079)
79000.)
79.)

The following are examples of invalid double-precision constants:
7.90999
7.9E5

(exponent is too large)
("E" denotes single precision; "0" denotes double
prE~cision)

3.4.1

Comparison of Real, O-floating, and G-floating

For KL model B systems, G-floal:ing double-precision is provided as an
alternative double-precision number format.
You must specify the
/GFLOATING compiler switch (see Chapter 16) to invoke the G-floating
double-precision format.
If you specify the /OFLOATING compiler
switch (the default),
the O-floating format
is used.
Table 3-1
summarizes the comparisons among real, O-floating, and G-floating.

3-3

CONSTANTS
Table 3-1:

Comparison of Real, D-floating, and G-floating Numbers
Bits of
Exponent

Bits of
Mantissa

Real

1.47 * 10** (-39)
to 1.70 * 10**(+38)

8.1

D-floating

1.47 * 10** (-39)
to 1.70 * 10**(+38)

18.7

G-floating

2.78 * 10** (-309)
to 8.99 * 10**(+307)

17.8

Range

Digits of
Precision

3.5

COMPLEX CONSTANTS

You can represent a complex constant by an ordered pair of
integer,
real, or octal constants written within parentheses and separated by a
comma.
For example,
(.70712,
-.70712)
and
(8.763E3,
2.297)
are
complex constants.
In a complex constant, the
first
(leftmost)
constant of the pair
represents the real part of the number; the second constant reprE~sents
the imaginary part of the number.
Both the real and
imaginary parts
of a complex constant can be signed.
The constants that represent the real and imaginary parts of a complex
constant occupy two consecutive storage locations
in the object
program.

3.6

CHARACTER CONSTANTS

A character
constant
is a
string of printable ASCII characters
enclosed by apostrophes.
Both delimiting apostrophes must be present,
and the string must be at least one character in length.
The compiler
accepts control characters in character constants with the following
exceptions:

Character

Octal Value

- NUL
- LF
- VT
FF

"0
"12
1113
"14

- CR

"IS

"'@
"'J
"'K

"'L -

"'M

NOTE
The CHAR function (see Chapte~ 13)
can be used
to
build variables that contain these control characters.
The value of a character constant is the string of characters b(=tween
the delimiting apostrophes.
The value does not include the delimiting
apostrophes,
but does
include all spaces or
tabs
within
the
apostrophes.
Within a character constant, the apostrophe character
is
represented
by two consecutive apostrophes
(with no space or other character
between them).
3-4

CONSTANTS
The length of the character constant
is the number of characters
between the apostrophes, except that two consecutive apostrophes count
as a single apostrophe.
Each character in the string has a character position that is numbered
consecutively starting at one.
The number indicates the sequential
position of a character in a string, from left to right.
There is one
character storage location for each character in the string.
If a character constant appears in a numeric context (for example,
as
the
expression on the
right side of an arithmetic assignment
statement), it is considered a Hollerith constant (see Section 3.9).
Examples of valid character constants and their lengths are:
Length

Value

'WHAT? '

WHAT?

'TODAYS"S DATE IS:

TODAY'S DATE IS:

'He said, "hello'"

He said, "hello"

,,,,

Examples of invalid character constants are:
'HEADINGS

,,
"Now or Never"

3.7

(no trailing apostrophe)
(a character constant must contain
one character)

least

(quotation marks cannot be used in
apostrophes)

place

OCTAL AND DOUBLE-OCTAL CONSTANTS

You may use octal
numbers
(radix 8)
as constants
in arithmetic
expressions,
logical expressi~ns, and data statements.
Octal numbers
up to 12 digits in length are considered standard octal constants;
they are stored
right-justified
in one storage location.
When
necessary, standard octal constants are padded with leading
zeros
to
fill their storage location.
If you specify more
than
12 digits
in an octal
number,
it
is
considered a double-octal constant.
Double-octal constants occupy two
storage locations and may contain up
to
24
right-iustified octal
digits; leading zeros are adde·j to fill any unused digits.
If you assign a single-octal c~nstant to a double-precision or complex
variable,
it
is stored right-justified in the high-order word of the
variable.
The low-order porti)n of the variable is set to
zero.
If
you assign a double-octal c)nstant to a double-precision or complex
variable, it is stored right-jJstified in the two words.
All octal constants must:
1.

Be preceded by a double quote
octal, for example, "777

Be signed if negative, but optionally signed if positive

Contain one or more of the digits 0 through 7, hut not 8 or 9

3-5

(")

to identify the

digits

CONSTANTS
The following are examples of valid octal constants:
"123456700007
+"12345 (optional sign)
-"7777
"-7777
The following are examples of invalid octal constants:
"12368
7777

(contains an 8)
(no identifying double quote)

When you use an octal constant as an operand
in an expression,
its
form
(bit pattern) is not converted to accommodate it to the type of
any other operand.
For example, the subexpression (A + "202400000000)
has as
its result the sum of A with the floating point number 2.0;
while the subexpression (I + "202400000000) has as its result the sum
of I with a large integer.
You cannot use octal constants as stand-alone arguments for
library
functions
that require non-octal arguments.
MINO,
for instance,
requires integer arguments and cannot accept octal arguments.
When you combine a double-octal constant in an expression with
(or
assign it to) either an integer or real variable, only the contents of
the high order location (leftmost) are used.

3.8

LOGICAL CONSTANTS

The Boolean values of truth and falsehood are represented
in FORTRAN
source programs as the logical constants .TRUE.
and .FALSE ••
Always
write logical constants enclosed by periods, as in the preceding
sentence.
You may use logical quantities in arithmetic and logical statements.
Only the sign of a numeric value used in a logical IF statement is
tested to determine if it is true (negative) or false (nonnegative).

3.9

HOLLERITH CONSTANTS

A Hollerith constant

is a
string of alphanumeric and/or special
characters preceded by nH (for example, nHstring).
In the prefix nH,
the letter n represents a number that specifies the exact number of
characters (including blanks) that follow the letter H.
\JOTE

If a character constant appears

considered

in a numeric context
Hollerith constant (see Section

3.6) •

The following are examples of Hollerith constants:
2HAB
14HLOAD TEST #124
6H#124-A

3-6

CONSTANTS
NO'I'E
A tab in
a
Holleritt
constant
.A.B.
character; for examplE, 3H

counted

one

You may enter Hollerith constents into DATA statements as a string of:
1.

ASCI I
Up
to
ten
7-bit
characters
double-precision tYPE variables

Up
to
five
variables

7 -·b it

ASCI I

characters

for
for

all

complex

other

type

The 7-bit ASCII characters that comprise a
Hollerith constant are
stored
left-justified
(st2rting
in
the
first
word
of
a
double-precision constant (thE high-order word) or the real part of a
complex constant)
with blarks
placed in empty character positions.
Holler i th constan ts tha t occur·y more than one var iable are
stored
as
successive variables
in the list.
The following example illustrates
how the string of characters is stored in a six-element
array calle~
A:
DIMENSION A(6)
DATA A/27HA string of mary characters/
A (1)
A(2)
A(3}

A(4)
A (5)

A(6)

3.10

is
is
is
is
is
is

set
set
set
set
set
set

to
to
to
to
to
to

'A str'
'ing 0'
'f man'
'y chat
'racte'
'rs

STATEMENT LABEL CONSTANTS

Statement
labels are
statement numbers.

numerjc

identifiers

that

represent

program

You write statement label constants as strings of one to five decimal
digits,
that are preceded by either an asterisk (*), a dollar sign
($), or an ampersand (&).
Foz' example, *11992, $11992, and &11992 are
all
valid statement
label constants.
You
use statement label
constants only in the argument. list of CALL statements to identify the
number
of
the executable statement to return to in a multiple RETURN
statement (see Chapter 13).

3-7

CHAPTF~R

SYMBOLIC NAMES,. VARIABLES, AND ARRAYS

4.1

SYMBOLIC NAMES

Symbolic names consist of any alphanumeric combination of one to six
characters, the first of which must be a letter.
1 i
you use more than
six characters in a symbolic name,
the compiler prints a warning
message and
ignores all bu: the first six characters.
The compiler
interprets lowercase letters in symbolic names as uppercase letters.
The following arE! examples of legal symbolic names:
A12345
IAMBIC
ABLE
C

The following arE! examples of illegal symbolic names:
.AMBIC
SAB

( fir s t c h a r act E~ r i s not ale t t e r )
(first charactE~r is not a letter)

You use symbolic names to identify specific items of a FORTRAN source
program;
Table 4-1 lists these items, together with an example of a
symbolic name and text reference for each.
Table 4-1:

Use Clf Symbolic Names

Symbolic Names
Can Identi fy
1.
2.
3.
4.
5.
6.
7.
S.
9.
10.

Variables
Arrays
Array elements
Substrings
Functions
Subroutines
Intrinsic functions
PROGRAM Statement
PARAMETER Statement
COMMON block names
11. NAMELIST list

For 8xample
Pl r CONS1~ , LIMIT
TAX
TAX(3,5)
FOO(l:N)
MYFUNC, VALFUN
CALCSB, SUB2, LOOKUP
SIN, ATAN, COSH
TEST
VI, C2,K
DA~r1AR,
COMDAT
DA':'A3

4-1.

For a Deta iled
Description
See Section
4.2
4.3
4.3.1
4.4
13.2
13.4
13.1
6.4.1
7.S
7.4
'
~

.".:

....

;: 1

SYMBOLIC NAMES, VARIABLES, AND ARRAYS
4.2

VARIABLES

A variable is a data storage location identified by a symbolic name; a
variable
is not a constant, an array, or an array element.
Variables
specify values that are assigned to them in such ways as assignment
statements
(Chapter 8),
DATA statements (Chapter 7), or at run time
through I/O data transfers (Chapter 10).
Before you assign a value to
a
variable,
its value is undefined; and you should not reference it
except to assign a value to it.
The value you assign to a variable can be either a
constant or the
result of a calculation that is performed during the execution of the
object program.
For example, the statement IAB=5 assigns the constant
5 to the variable lAB.
In the statement IAB=5+IB, however, the value
assigned lAB depends on the value of variable
IB at
the
time the
statement is executed.
The type of a variable determines the interpretation of its
Variables can be:
1.

Integer

Real

Logical

Double-precision

Complex

Character

contents.

The type of a variable is determined either implicitly, by the first
letter of the variable name
(described below), or explicitly, by
declaring the variable type
in a
type declaration statement
(see
Chapter 7).
FORTRAN uses the following default
types are not explicitly declared:

conventions

for

variables

whose

Variable names that begin with the letters I, J, K, L, M,
N are integer variables.

Variable names that begin with any letter other than I, J, K,
L, M, or N are real variables.

NOTE
These default conventions can be altered
use of the
IMPLICIT statement,
which
described in Section 7.3.
The following are examples of determining
according to the preceding conventions:

the

type

by
is
a

variable

Variable

Beginning Letter

Assumed Data Type

ITEMP
OTEMP
KAl23
AABLE

I
0

Integer
Real
Integer
Real

4-2

SYMBOLIC NAMES, VARIA.BLES, AND ARRAYS
4.3

ARRAYS

An array is an ordered set of data identified by an array name. Array
names are symbolic names and must conform to the rules for writing
symbolic names (see Section 4.1).
Arrays are made up of smaller units of data called array elements. As
with variables, you may assign a value to an array element. Before
you assign a value to an array element it has an undefined value. You
should not reference an alray element until you have assigned it a
value.
An array element is referenced by using the array name together with
some number of subscripts that describe the position of the element
within the array.

4.3.1

Array Element Subscripts

The general form of an array element name is AN (Sl, S2, ••• Sn), where
AN is the array name and Sl ttrough Sn represent 1 through n subscript
expressions.
You may use any number of
subscript quantities in an
element name; however, the number used must always equal the number of
dimensions (see Section 4.3.2) specified for the array.
A subscript can be any constant or expression
example:

(see

Chapter

S),

for

Subscript quantities may contain arithmetic expressions that
involve addition, sobtraction, multiplication, division, and
exponentiation.
For
example,
(A+B,C*S,D/2)
and
(A**3,(B/4+C)*E,3) ale valid subscripts.

l\rithmetic expressic.ns
(see Chapter 5)
used
in
array
subscripts may be of any type, but noninteger expressions
(including complex)
are converted to integer when
the
subscript is evaluatE!d.

A subscript may contain function references (see Chapter 13).
For example, TABLE
(SIN(A)*B,2,3) is a valid array element
identifier.

Subscripts may contain array element identifiers nested to
any level as subscripts.
For example,
in the subscript
(I(J(K(L»),A+B,C) the first subscript expression given is a
nested 3-level array reference.

Some examples of valid array elements are:
1.

IAB(1,S,3)

ABLE (A)

TABLEl(lO/C+K**2,A,B)

4•

~AT ( A , AH (2

* L) , • 3 * TAB A, M + 1,0) , 55 )
i

4-3

SYMBOLIC NAMES, VARIABLES, AND ARRAYS
4.3.2

Dimensioning Arrays

You must declare the size (number of elements) of an array to enable
FORTRAN to reserve the number of locations needed to store the array.
Arrays are stored as a
series of sequential storage locations.
Arrays,
however, are visualized and referenced as if they were single
or multi-dimensional,
rectilinear matrices dimensioned on a
row,
column,
and plane basis.
For example, Figure 4-1 represents a 3-row,
3-column, 2-plane array.

3 ROWS

.~
'l'-~

~'v

-------~-------/
3 COLUMNS
Figure 4-1:

MH S 11~~·81

A 3 x 3 x 2 Array

You specify the size of an array by an array declarator written as a
subscripted array name.
In an array declarator each subscript
quantity is a dimension of the array and must be either an
integer
expression, an integer variable, or an asterisk (*).
Only the upper bound in the last dimension declarator
in a
list of
dimension declarators can be an asterisk.
An asterisk marks the
declarators as an assumed-size array declarator (see Section 7.1.2).
NOTE
allowed
in
Variable array dimensions are
only
subprograms.
See adjustable dimension statements,
Section 7.1.1.
For example,
declarators.

TABLE(I,J,K)

and

MATRIX(lO,7,3,4)

are

valid

array

The total number of elements that comprise an array is the product of
the dimension quantities given in its array declarator.
For example,
the array lAB dimensioned as IAB(2,3,4) has 24 elements (2 * 3 * 4
24) •
You dimension arrays only in the specification statements DIMENSION,
COMMON, and type declaration (see Chapter 7).
Subscripted array names
appearing in any of the
these statements are array declarators;
subscripted array names appearing in any other statements are always
array element identifiers.

4-4

SYMBOLIC NAMES, VARIABLES, AND ARRAYS
In array declarators, the position of a given subscript quantity
determines the particular c:imension of the array (for example, row,
column, or plane) that it represents.
The first
three subscript
positions specify the number of rows,
columns, and planes that
comprise the named array;
each following subscript given
then
specifies a set comprised of n-number (value of the subscript) of the
previously defined sets.
For example:

SpE~cifies thE! Array(s)

The Dimension Declarator

TAB(2)

~~
~~

TAB(2,2)

TAB(2,2,2)

,.""".",*,'

--'

1,1,1

1,1,2

1,2,2

2,1,2

2,2,2

1,2,1

2,1,1
TAB(2,2,2,2 )

....... ........

2,2,1

-- -1,1,1,1

1,2,1,1

2,1,1,1

2,2,1,1

.... "''''

"'''' ....

",-

....- '"

1,1,2,1
I

.,.., '"

-_ --....

1,2,2,1

2,1,2,1

2,2,2,1

-'" -

.... "

-'-

1,1,2,2

1,2,2,2

2,1,2,2

2,2,2,2

_--

1,1,1,2

1,2,1,2

2,1,1,2

2,2,1,2

.,..,

---

MH·S 1762-81

NOTE
FORTRAN-IO/20 permits up to 127 dimensions in an array
declarator.
(The FORTRAN-77 Standard allows a maximum
of 7 dimensions.)

4.3.3

Order of Stored Array E:lements

The elements of an array are stored in ascending order.
The value of
the first (leftmost) subscript varies between its minimum and maximum
values most rapidly.
The value of the last
(rightmost)
subscript
increases to its maximum value least rapidly.
For example, the
elements of the array dimensioned as 1(2,3)
are stored
in
the
following order:
1(1,1)

1(2,1)

1(1,2)

] (2,2)

1(1,3)

1(2,3)

In the following list, the elE·ments of the three-dimensional array
(B(3,3,3»
are stored
row by row from left to right and from top to
bottom.

4-5

SYMBOLIC NAMES, VARIABLES, AND ARRAYS

8(1,1,1)

8(2,1,1)

8(3,1,1)--1

,------------------------ - 8(1,2,1)
8(2,2,1)
8(3,2,1)-,
,-------------------------8(1,3,1)
8(2,3,1)
8(3,3,1)-,
[:B(1~~2) - --B(2~1 ~2) - --a(3,1~2)==1

[------------ - ----------- 8(1,2,2)

8(2,2,2)

8(3,2,2)-,

~~~~,!~2I===~~~3~
===~~'~~}=J
8(2,1,3)
8(3,1,3)-,

-I - 8(1,1,3)

= ==

~ :~~,~~3I ~ = ~~~!~ =~(~,~~3~~;
L - 8(1 ,3,3)

8(2,3,3)

8(3,3,3)
MR·S·1756·81

Thus B(3,1,1)

is stored before B(1,2,1), and so forth.

Character array elements are stored in successive character positions,
and do not necessarily start on a word boundary.
Character array
elements are stored
five characters per word
(seven bits
per
character), and the low order bit is never used, for example:
CHARACTER*3 A(4)
The array A will be stored in the following way:

A(1)

A(2)

A(3)

A(4)

Unused
A.._

x Ixl

MR-S-2S28-83

where:
x

4.4

means bits are not used.

The value in bit 35 is zero.

CHARACTER SUBSTRINGS

A character substring is a contiguous segment of a character variable
or character array element.
A character substring is identified by a
substring name and can be assigned values and referenced.
A character substring reference has one of the following forms:
v([el]:[e2])
or
a(s[,s] •.. )

([el]:[e2])

4-6

SYl1BOLIC NAMES, VARIABLES, AND ARRAYS
where:
v

is a character variable name.

is a character array name.

is a subscript expression.

is an optional
nu~eric
expression that specifies
leftmost character position of the substring.

the

is an optional nuneric expression that specifies
rightmost character position of the substring.

the

Character positions within a character variable or array element are
numbered from left to right, beginning at 1.
For example, LABEL{2:7)
specifies the substring beginning with the second character position
and ending with the seventh character position of the character
variable LABEL.
If the value of the numeric
integer,
FORTRAN converts
fractional part before use.

Expression el or e2 is not of type
it to an integer value by truncating any

The values of the numeric expression el and e2 must meet the following
conditions:
1 .LE.

el .I.E.

e2 .LE.

len

where:
len is the length of the character variable or array element.
If el is omitted, FORTRAN assumes that el is 1.
FORTRAN assumes that e2 equals len.

omitted,

For Qxample, NAMES{1,3) (:7) specifies the substring starting with the
first character position and ending with the seventh character
position of the character array element NAMES{1,3).

4-7

CHAPTER 5
EXPRESSIONS

5.1

ARITHMETIC EXPRESSIONS

An arithmetic expression
is formed with arithmetic operands and
arithmetic operators.
The evaluation of such an expression produces a
numeric value.
Arithmetic expressions may be either simple or compound.
arithmetic expression consists of an operand that can be:
1.

A numeric constant

A numeric variable

A numeric array element

An arithmetic function reference (see Chapter 13)

An a r i t hme tic
parentheses

logical

expression

simple

written

within

Operands may be of integer, real, double-precision, complex,
octal, double-octal, or Hollerith type.

logical,

The following are examples of valid simple arithmetic expressions:
105
lAB
TABLE(3,4,5)
SIN (X)
(A+B)

(integer constant)
(integer variable)
( a r ray e 1 em E~ n t )
(function reference)
(a parenthetical expression)

A compound arithmetic expression consists of two or more operands
combined by arithmetic operators.
Table 5-1 lists the arithmetic
operations permitted in FORTRAN and the operator recognized for each
operation.

5-1

EXPRESSIONS
Table 5-1:

Arithmetic Operations and Operators

Operation

Binary Operator
+

A+B
A-B
A*B

** or "

A**B or A"'B

Unary Operator

Example

Addition
Subtraction
Multiplication
Division
Exponentiation
Operation

Identity
Negation

5.1.1

Example

AlB

+A
-B

Rules for Writing Arithmetic Expressions

Observe the following rules in structuring arithmetic expressions:
1.

The operands comprising an arithmetic expression can be of
different types.
Tables 5-2 and 5-3 illustrate all permitted
combinations of data types and
the type assigned
to the
result of each.
NOTE
All combinations of numeric data types except
double-precision with complex are allowed in
FORTRAN.

If you specify two adjacent operators, and the
second
is a
minus or a
plus, the second operator is considered a unary
operator and acts only on the term immediately following
it.
Thus,
in the example (A*X+B)*+C, the subexpression, *+C, is
interpreted as the binary operator * and the unary +.
You cannot, however, have two adjacent binary operators in an
expression.
For
example,
the expression A*/B is not
permitted.

All operators must be included; no operation is implied.
For
example, the expression A(B) does not specify multiplication,
although this is implied in standard algebraic notation.
The
expression A*(B)
is required to specify a mUltiplication of
the operands.

5-2

Table 5-2:

Type of the Result Obtained from Mixed-Mode Operations

Type of Argllment 2
For operators

+ ,-,',

Double
Precision

Real

Integer

LO~lical

Complex

Double
Octal

Octal

Literal

~------------------------~~--------------'-i----------------;--------------'----~---'------------~-------.--------_4----------------_+----- ----------+----------------~
1. Type of operation

Integer

used
2_ Type associated
with result
3_ Conversion on
Argument 1

1. Integer

Uouble preCIS!'")!'

2. Integer

(Jouble Preclslul'

J. From InleClN Ie
Doubie Pn~cISl(jp

J. from jmegc'
to ticaj

3. None

,
'L.

J. Holm InlNlc' 10
L(ln;:)I;:~

intr'q

InteGer

3. No",,'

3 Nor".

3. Non"

'.1,;'

us -~d as Heai

4. Conversion on
Argument 2

4. None

:l.

4. Non"

N( U'

4. Hig n order 1V0ro

is used directly'
low oraer worli

tilan ordPi \'\{(},

IS used dlleeti·,
tunner worGS aft"

IS Iqrlo.-eo

1. Type of operation

1. Real

1. Double PreCISion

2. Real

'-- Doubie PreCISI,)!,

3. None

3. Used direGtlv as
the high order

1. Heal

used

2_ Type associated

lie.,1

with result

3. Conversion on

Argument 1

Real

4. Conversion on
Argument 2

I-rom 'me,:"
to Rr-_!

4. None

Lsed dlrp.cll~ as

J. Non.-

None

tile Real n.ll'

wora: low orch!,
wora IS zen,

jf]aln~T 'j

D,~~7

4. Non,;

NOlii'

Norw

4. High orOel won;

IS used dw'ctly
lOW Ofoel wor,-j

IS Iqnoreo

4. HI9h order wor.i
is used dlrectlv
further word,
are Ignore,;

~--------------------------~----------------4_--------------~~-----------------------.-----------~----------------~----------------+_--------------_4----------------~

Double
Precision

1. Type of operation
used
2. Type associated
with result
3. Conversion on
Argument 1
4. Conversion on
Argument 2

1. Doubie Precis!. n

I. Double PreC"Wii'

2. Uoubl" t-'reclslClJ

:< None
4

f\h~:

Hom Imeoer tu

Used direcllv

1. Double Precision

1. lJout.le PrecIsion

1. Double I'r,,~ls'on

2. Douale t>recisloll

2. Double Precisiion

2. Double PrecIsIon

2. Double PrecIsion

2. Double PrecislO!'

3. None

J. No,,'

3. Non . . ·

NII'i!

4. None

4. Useo dllecTly a,;
thl' hiqh onh':
woro: lOW orde;

NOtl('

the nlun t.'rr:~"f

Double Preclsl' I

word: iew or!1:i"";

wc:c !~, lP.'"

Complex

1. Complex

4. I--rom Inteqer II.
(.omDI~)!

\I~IU'~

--------1---------.---1

!. L..omnlex

Compl~""

Lorfluh.:t,

Cornplf· ...

2. Complex

L. COlllllle!

N[;;

3. None

3. Nol'

Nnr,~'

3. No""

4. None

4. Used drrectlv .IS

nneC'tiv .::.
the HI"i";; p.u
InlaQm~t \f I I .... "

4. First two WC"'j,

1 Us("o d!rf>(:tl':
ttH!

USP.r1 as. Hpal p. 'r'

1. New

4. first two WOI(j,:
are used dir~~!!'
further \Nunj-:
are Ignorl'rl

woro IS zer':

~---------.---------_+--------__f----------~---------_+---------f------.----_+--------~-.-

1. Type of operation
used
2. Type associated
with result
3_ Conversion on
Argument 1
4. Conversion on
Argument 2

Douoj~ ~rcClsloP

1. Double PrecIsion

Hl:~ii

2. (~ompl""

me Heal par!;

..... 3. ':

Imaqmary p.;; 1

nnaqlnary pctn

2. Cornplex

USPfl

are used olrcr.tl'..:

Further WOf!;-..
are IQno,· ' ....

Logical

1_ Type of operation
used
2. Type associated
with result
3. Conversion on
Argument 1

4. Conversion on
Argument 2

~or1t;)

2. 1J0ubie Pree"'"''

2. Octal

Oc!~1

2. Octal

}

Octa;

f,.l . • "

J. Used directly ~'.

3. None

NOfl('

J. None

Non"

4. None

Hiah order wc"r.!

;-.tont:

Nnne

lJ sed alrrcll"

II e Heal

lhe high order
word; low order
wora IS zern
4

Integ'~r

1. lJouble tJreclslul

InICQ("

0,.,'

inleat~r

Ir laQlniU'" 0;1"

io,lon ..

-1

Hiqll order word

rurtner ;}IIII~;;
IS Iqnolt-'tl

I VDe OT ooeraV!n r .

"L.

I voe assucICll.-;(i

2. Real

1. Double .... reCISILJ'.

HC.JI

Inl"q~"

2. Oc.ai

Double t-'reC'IS!I;,

3. Used directly as
the hlqh ordet
word; low orcl."

Imeq'"

Uclal
Ncr-~

J. NOfn-

word IS .ler n .
4

l.·(l~;;;'-:r">itljl

,",;

Non-

j'urll'

Hiqh order ''II'ptd

4. Hioh order "'C>"

IS U:icd dlifL.i:b.

iow oloer 'J\'Cf{1

IS used drrcc;:'j
furtner woc'"

l!i lonOf~o:j

ar~

IqnOrf>rJ

~-------------------------+_--------------.~--------------~--------------~---------------~---------------4---------------4---------------~------------~
1. Uouole t-"reCISIUII

H,'.!'
L..
.1

"2.. LJouble t-"reCISlon

!rHeaE"

(;'U1V .. ' ...... , •••••

IS used di, l-cil
hJW OHler .fi{),,;

IS IJseri nl;f~1

H!Qtl urder ...-vor,j

IS used dlrectlv

low (Iral~r \Nil(fj
I~

IQllort.·U

4. Non,.

HIOh ararl"" wnni
IS used t1!rf~~.:tb.
lOW orner WO.I.

IS used ct, '

J. Hlqh orde, .f.- •. ,

·'-.!d; j

"'I;

IS used Qtrc-:';'"

lOW order W:.ii.
are Igoor""

lOW order y..i 1 1'.1

IS Iqnorc'r.

IS lunnrer1

4. NOlh:-

HIUh 'Jr':1e,

4. Hloh order WGf~l

IS used d""o.:tl •.

4. Hlch order ~",'(:f( I
IS used dli;.;(';;·~·
lOW orar.or ~·'..;f·!·.

low orner ~""_'f;1

are Iqnot(\(i

~---------------_4-----------

. r__---.-------+_-------.--_;_-.------------------. -------+----------11----------+---------1
;.

L iit .... :

t ~n:':.;::'!--;.ll;:r-::
.u'lHH~\,.nt

In~,_·('

...

IrT~'f"-

,r:'C;l'..

h.l.rtrit-"'f

.'d' :1~-"

IS IiSC'U (Hft.:' ,.
jupr~er

Hrst twu WO'lL
are USP.(~ Oirec:·~

arc l"Un'Jn-~~:
Nc·

i- !st ivvl) wurOb

a' e used dlr.~c\iv

J. Hlqn oroer word
IS u<;f,·d oirectlV'

further wm·,b

Wnf(1

are IOnore{l
-t

Nu,',,'

.I.

t"Han ~Jr..:1cr ":Vl"" i

Mlon Ordej .-vu'
IS

tunner WOIUS

rurther W(;I':';

dt~ 1l.!l1uretj

·t. None

usea OIP'C!:'}.

IS USeO tllret !.,

ctre IUfHHt'(1

4. Hlgtl order ""01·1

,; . •1100 Of(;€i

V.I."

!S usea ol:-e::::: .
furirler WOf{j

are IU"Cf'I~';

4. High oroer wOln

is used din-. ct~'!
low o·der word

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

-_..._----------\--------------"-----------------_.. _-----------_._-

--.-----------''-----------~

MR-SI751-Bl

,.................a. .I ........ I. . . . .IIII. . . . . . . . . . .IIIIIIII. . . . . . . . . . . . .~. . . . . . . . . . . . . . . . . . . . . . . .~. . . . . . . . . . . . . . . . . . . . . .~. . . .I . . . . . . . . . . . .I. . . . . .~. .B . .&

EXPRESSIONS
Table 5-3:

Permitted Base/Exponent Type Combinations
Exponent Operand

Base Operand

DoublePrecision

Integer

Real

Integer

Real

DoublePrecision

Complex

Real

DoublePrecision

Complex

DoublePrecision

DoublePrecisio n

DoublePrecision

(Illegal)

Complex

(I llega 1)

Complex

5.1.2

Complex

Arithmetic Constant Expressions

An arithmetic constant expression is an arithmetic expression in which
each operand is one of the following:
1.

A numeric constant

A symbolic name of a numeric constant

A call to the function
ICHAR
(see Chapter
argument is a character constant expression

arith~etic

constant expression enclosed in parentheses
13)

where

thp-

The exponentiation operator is not permitted unless the exponent is of
type
integer.
Note that variables,
array elements, and function
references are not allowed.
Example:
5+6*(ICHAR('Z')-ICHAR('A ' )+1)*4.1**3

5.1.2.1 Integer Constant Expression - An integer constant expression
is an arithmetic constant expression
in which each constant or
symbolic name of a constant is of type integer.
Example:
3+4**6+2

5-5

EXPRESSIONS
5.2

CHARACTER EXPRESSIONS

Character expressions consist of character operands and character
operators.
The evaluation of a character expression yields a single
value of character data type.
A character operand can be anyone of the following:
1.

A character constant

A symbolic name of a character constant

A character variable

A character array element

A character substring

A character expression, optionally enclosed in parentheses

A character function reference

The only character operator is the concatenation operator

(II).

A character expression has the form:
character operand

[Ilcharacter operand] •.•

The value of a character expression is a character string formed by
successive left-to-right concatenations of the value of the elements
of the character expression.
The length of a character expression
is
the sum of the lengths of the character elements.
For example, the
value of the character expression 'AB'II'CDE' is 'ABCDE', which has a
length of 5.
Note that the expression:

A=AIIB
has no effect on A, since the concatenation result is truncated to the
length of A.
Parentheses do not affect the value of a
character expression.
example, the following character expressions are equivalent:

For

('ABC'II'DE')II'F'
'ABC'II('DE'II'F')
'ABC'II'DE'II'F'
Each of these character expressions has the value 'ABCDEF'.
If a character element in a character expression contains spaces,
spaces are
included
in
the value of the character expression.
example, 'ABC 'II'D E'II'F' has a value of 'ABC D EF'.

5-6

the
For

E}(PRESSIONS
5.2.1

Character Constant Expression

A character constant expression is a character
each operand is one of the following:

expression

A character constant

The symbolic name of a character constant

A character constant expression enclosed in parentheses

A call to the function CHAR
(see Chapter
a r gum e n t i s ani n t e g E' r con s tan t ex pre s s ion

13)

which

where

the

Variables, array elements, substrings, and function references are not
allowed.
Example:
'HELLO' II:HAR (13)1 ICHAR (] 0)1I'GOODBYE'

5.3

LOGICAL EXPRESSIONS

Logical expressions can be either simple or compound. Simple logical
expressions consist of a logical operand, which can be one of the
following:
1.

A constant

A variable

An array element

A function reference (see Chapter 13)

An expression written within parentheses

Compound logical expressions consist of two or more logical or numeric
operands combined by logical operators.
The evaluation of a logical
expression produces a truth value (type logical, true or false)
as
determined by the resulting bit pattern.
Table 5-4 gives the logical operators permitted
description of the operation each provides.

5-7

FORTRAN

and

EXPRESSIONS
Table 5-4:
Operator

Logical Operators
Description

.AND.

AND operator.
Both of the logical operands combined
this operator must be true to produce a true result •

• OR.

Inclusive OR operator.
If either or both of the logical
operands combined by .OR.
are true, the result will be
true.

.NEQV.

Exclusive OR operator (also .XOR.).
If either but not
both of the logical operands combined by .NEQV.
is
true, the result will be true.

.EQV.

Equivalence operator.
If the logical operands being
combined by
.EQV.
are both the same (both are true or
both are false), the result will be true.

.NOT.

Complement
operator.
This
operator
specifies
complementation
(inversion)
of the item
(operand or
expression) that it modifies.
The original
item,
if
true by itself, becomes false, and vice versa.

Logical expressions are of the general form P .op. Q, where P and Q
are logical operands and .op.
is any logical operator except ".NOT.".
The .NOT.
operator complements the value of an operand;
it must
appear
immediately before the operand that it modifies, for example,
.NOT.P.
Table 5-5 is a
truth table illustrating all possible
logical
combinations of two logical operands (P and Q) and the result of each
combination.

5-8

EXPRESSIONS
Table 5-5:

Logical Operations Truth Table

When P is

And Q is:
t-----------. -r------

Then the Expression:

Is:

True

.NOT. P

False

.NOT. P

True

P .AND. Q

True

False

P .AND. Q

False

True

P .AND. Q

False

P .AND. Q

False

True

P • OR. Q

True

False

P .OR. Q

True

False

True

P .OR. Q

True

False

P .OR. Q

False

True

P .NEQV. Q

False

True

False

P • NEQV. Q

True

False

True

P .NEQV. Q

True

False

P • NEQV. Q

False

True

P .EQV. Q

True

False

P • EQV. Q

False

True

P. EQV. Q

False

P • EQV. Q

True

For example, consider the following variables:
Variables
PHETT, RUN
I,J,K
DP,D
L,A,B
CPX,C

Type
Real
Inteqer
Doubje-Precision
Logical
Complex

5-9

EXPRESSIONS
Examples of valid logical
variables are:

expressions

consisting

the

preceding

L.AND.B
(PHETT*I).NEQV. (DP+K)
L. AND. A. OR .• NOT. (I -K)
Logical
operations are performed
on
the
full
36-bit
binary
representation of the operands involved.
However, when an operand of
a logical expression is double-precision or complex,
only the
first
word
of a double-precision operand (the high-order word) or the real
part of
the complex
operand
is
used
in
the
specified
logical
operation.
The result of a logical operation is found by performing the specified
operation simultaneously
for
each of the corresponding bits in each
operand.
r'or example, consider the expression A=C.OR.D, where C="456
and
D="201.
The operation performed by the processor and the result
j s:
Word
Bits
0
Operand C 0
Operand D 0
Result A 0

.. 24

0
0
0

·0
.. 0

25 26
0 0
0 0
0
0

29
0
0

31
0
0

28
0

0
1

32 33
1

34 35
I
0
0 1
1 1

Table 5-5 also illustrates all possible logical
combinations of
two
(P
one-bit binary operands
and Q)
and gives
the result of each
Simply read 1 for true and 0 for false.
combination.
If a logical expression
is
used as an operand
in
an arithmetic
expression,
its value is not converted to accommodate it to the type
of any other operand.

5.3.1

Logical Constant Expression

A logical constant expression is a logical expression
operand is one of the following:

which

A logical constant

The symbolic name of a logical constant

A relational expression in which each operand is
expression

A logical constant expression enclosed in parentheses

each

constant

Variables, array elements, and function references are not allowed.
Example:
.NOT. (PARML1.NE.PARML2)
where PARMLI and PARML2 are specified in a
Sec t ion 7.8).

5-10

PARAMETER

statement

(see

EXPRESSIONS
5.4

RELATIONAL EXPRESSIONS

Relational expressions consist of two arithmetic expressions or
character
expressions
combined by a relational operator.
relational operator allows you to test the relationship between
arithmetic or two character expressions.

two
The
two

The result of a relational expression is always a
false value.

logically

true

You can write relational operators either as a
2-letter mnemonic
enclosed within periods
(for example,
.GT.) or use the symbolic
equivalent, for example, >, instead of .GT.
Table 5-6 lists the mnemonic and symbolic forms of the FORTRAN-10/20
relational operators and specifies the type of test performed by each.
Table 5-6:

Relational Operators and Operations
Operators

Relation Tested

Mnemonic

Symbolic

.GT.
.GE.
.LT.
.LE.
.EQ.
.NE.

Greater than
Greater than or equal to
Less than
Less than or equal to
Equal to
Not equal to

>:=

<
<~

- .-

Relational expressions are of the general form A .op. B, where A and B
represent arithml~tic or character operands, and .op.
is a relational
operator.
You can mix arithmetic operands of type
double-precision in relational expressions.

integer,

real,

and

A relational expression cannot be used to compare the value of an
arithmetic expression with the value of a character expression.
However, you can compare a num~ric expression to a character constant.
In this case, the character '~onstant is considered to be a Hollerith
(see Section 3.9).
You can compare complex operands using only the operators
.EQ.
(==)
and
.NE.
(#) •
Complex quantiti,es are equal if the corresponding
parts of both words are equal.
For example, assume the following variables:
Variables

Type

PHETT, RON
I,J,K
DP,D
L,A,B
CPX,C
CHR,RA

Real
Integer
Double-Precision
Logical
Complex
Character

5-11

EXPRESSIONS
Examples of valid
variables are:

relational

expressions

consisting

the

above

Examples of invalid relational expressions
variables are:

consisting

the

above

.EQ.

operands

and

(PHETT) .GT.lO
I

C.EQ.CPX
CHR.LT.RA

(PHETT) .GT 10

(closing period missing from operator)

C.GT.CPX

(complex operands can only
and .NE.
operators)

RA.EQ.RON

(you cannot compare
character operands)

compared

arithmetic

Examples of valid expressions that use both logical
operators to combine the preceding variables are:

and

relational

( I • GT. 10). AND. (J. L E. K)
( (I * RON) . EQ. (1/ J) ) • OR. L
(I. AND. K) # ( (PHETT) . OR. (RON) )
'-:#CPX. OR. RON
~

Logical expression
is used as an operand
in an arithmetic
its value is not converted to accommodate it to the type
i)f any other operand.

':~-"'DreSSlon,

In character relational expressions "less than" means "precedes in the
ASCII collating sequence," and "greater than" means "follows in the
ASCII collating sequence", for example:
, AB ' / / ' ZZZ'

. LT.

'CCCCC'

This expression tests whether 'ABZZZ' is less than
'CCCCC'.
Since
that relationship does exist, the value of the expression is true.
If
the relationship stated does not exist, the value of the expression is
false.
If the two character expressions in a relational expression are not
the same length, the comparison is performed as if the shorter one is
padded on the right with spaces until
the lengths are equal,
for
example:
, ABC'

• EQ.

'ABC

, AB'

• LT • ' C '

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

5-12

EXPRESSIONS
NOTE
The rule that character relationals extend the shorter
operand with spaces to match the length of the longer
operand has an interesting effect when the longer
string ends with characters in the range CHARCO) to
CHAR(31) (ASCII control characters such as 'bell'
and
line feed).
Since space is CHAR(32), the trailing spaces supplied
as filler by FORTRAN compare being greater than
trailing control characters. Thus, the string
'FOO'
is .GT. 'FOO"G' (FOO followed by a bell).

5.5

EVALUATION OF EXPRESSIONS

The following considerations determine the order of computation
FORTRAN expression:
1.

The use of parenthesEs

An established hierarchy for
the
relational, and logical operations

The location of operators within an expression

5.5.1

execution

arithmetic

Parenthetical Subexpressions

In an expression, all sUbexpressions enclosed within parentheses are
evaluated first.
When parenthetical subexpressions are nested (one
contained within another), the most deeply nested subexpression is
evaluated
first;
the next most deeply nested sUbexpression is
evaluated second;
and so on, until
the value
of
the
final
parenthetical expression is computed.
When more than one operator
is contained
in
a
parenthetical
subexpression,
the required computations are performed according to
the hierarchy assigned to operators by FORTRAN (see Section 5.5.2).
For example, the separate computations
expression A+B/«A/B)+C)-C are:
1.

Rl=A/B

R2=Rl+C

R3=B/R2

R4=A+R3

R5=R4-C

performed

evaluating

the

results

the

where:
Rl through R5

represent the interim and
computations performed.

5-13

final

EXPRESSIONS
5.5.2

Hierarchy of Operators

The following hierarchy
(order
classes of FORTRAN operators:
first,
second,
third,

execution)

assigned

the

arithmetic operators
relational operators
logical operators

Table 5-7 specifies the precedence
operators of the above classes.

assigned

the

individual

With the exception of exponentiation and
integer division, all
operations on expressions or subexpressions involving operators of
equal precedence are computed
in any order
that
is algebraically
correct.
A subexpression of a given expression may be computed
in any order.
For example,
in the expression (F(X) + A*B), the function reference
may be computed either before or after A*B.
Table 5-7:

Hierarchy of FORTRAN Operators

Class

Level

EXPONENTIAL

First

** or

ARITHMETIC

Second
Third
Fourth

-(negation) and + (identity)
* ,/
+,-

RELATIONAL

Fifth

.GT.,.GE.,.LT.,.LE.,.EQ.,.NE.
or

Symbol or Mnemonic

...

-'-

_.-

>,>=,<,<=,==,#

LOGICAL

Sixth
Seventh
Eighth
Ninth

.NOT.
.AND.
.OR.
.EQV.,.NEQV.

Operations specifying integer division are evaluated from left to
right.
For example,
the expression I/J*K is evaluated as if it had
been written as (I/J)*K), but this left-to-right evaluation process
can be overridden by parentheses.
I/J*K (evaluated as(I/J) *K) does
not equal I/(J*K).
When a series of exponentiation operations occurs in an expression, it
is evaluated in order from right to left.
For example, the expression
A**2**B is evaluated in the following order:
first Rl = 2**B (intermediate result)
second R2 = A**RI (final result).
As with other expressions, parentheses alter
the evaluation of the
above expression.
The expression (A**2)**B is evaluated in theSE! two
steps:
first RI = A**2 (intermediate result)
second R2 = RI**B (final result)

5-14

EXPRESSIONS
5.5.3

Mixed-Mode Expressions

Mixed-mode
exprE!ssions
arE~
evaluated
on
a
basis
of
subexpression-by-subexpression r
with the type of the results obtained
converted and combined with other results or terms according
to the
conversion procedures described in Table 5-2.
For example, assume the follow!ng variables and data types:
Variables

Type
Double-Precision
Real
Integer

I,J

The mixed-mode expression D+X*(I/J)
manner:

evaluated

the

following

I/J

X*Rl Rl is converted to type real and is mUltiplied by X
to produce H2

D+R2 R2 is convel:ted to
type
added to D to produce R3

Rl is

integE~r

double-precision

and

where:
Rl and R2,
and R3
repr(~sent
the
interim
respectively of the computations performed.

5.5.4

and

final

results

Use of Logical Operands in Mixed-Mode Expressions

When you use logical operands :_n mixed-mode expressions, the value
the log ical oper-and is not convertE~d in any way to accommoda te it
the type of the other operands in the expression.
For example,
L*R,
where L is type logical and R is type real, the expression
evaluated without convertinq L to type real.

5.6

of
to
in
is

CONSTANT EXPRESSIONS

A constant expression
is an arithmetic constant expression
(see
Section 5.1.2), a character constant expression (see Section 5.2.1),
or a logical constant expression (see Section 5.3.1).

5-15

CHAPTER 6
EXECUTABLE AND NONEXECUTABLE STATEMENTS

Each statement is classified as
executable
or
nonexecutable.
Executable statements specify actions and form an execution sequence
in a program.
Nonexecutable statements do the following:
1.

Specify characteristics, arrangement, and initial
data

Contain editing information

Specify statement functions

Classify program units

Specify entry points within subprograms

values

Nonexecutable statements are not part of the execution sequence.
Nonexecutable statements may be labeled, but such statement labels
must not be used to control the execution sequence.

6.1

EXECUTABLE STATEMENTS

The following statements are classified as executable:
1.

Arithmetic, logical, statement label (ASSIGN), and
assignment statements

Unconditional GO TO, assigned
statements

Arithmetic IF, logical IF statements, and two-branch logical
IF statements, IF THEN, ELSE, and ELSE IF THEN statements

CONTINUE statement

STOP and PAUSE statements

DO and DO WHILE statements

READ, REREAD, WRITE, and PRINT statements

OPEN and CLOSE statements

REWIND, BACKSPACE, ENDFILE, BACKFILE,
FIND and UNLOAD statements

10.

CALL and RETURN statements

6-1

TO,

and

character

computed

SKIPRECORD,

SKIPFILE,

EXECUTABLE AND NONEXECUTABLE STATEMENTS

6.2

11.

END, END IF, and END DO statements

12.

DECODE and ENCODE statements

13.

ACCEPT, PUNCH, and TYPE statements

14.

INQUIRE statement

NONEXECUTABLE STATEMENTS

The following statements are classified as nonexecutable:

6.3

PROGRAM,
FUNCTION,
statements

DIMENSION,
COMMON,
EQUIVALENCE,
IMPLICIT,
INTRINSIC, EXTERNAL, and SAVE statements

INTEGER,
REAL,
DOUBLE
PRECISION,
COMPLEX,
CHARACTER type-specification statements

DATA statement

FORMAT statement

Statement function statement

INCLUDE statement

NAMELIST statement

SUBROUTINE,

ENTRY,

and

BLOCK

DATA

PARAMETER,
LOGICAL,

and

ORDERING OF FORTRAN STATEMENTS

The order in which you place FORTRAN statements in a program unit
is
important.
Certain
types of statements must be processed before
others to guarantee that compilation takes place as you expect.
Figure 6-1 shows the required order of statements and comment
lines
within a program unit.
Horizontal lines indicate (from the top of the
diagram to the bottom) the order in which statements and comment lines
must appear in a program.
For example, a PROGRAM statement must occur
before FORMAT statements.
FORMAT statements,
in
turn,
must
occur
before an END statement.
Vertical
lines
in
the diagram
indicate how comment
lines
and
statements may be interspersed in the program.
For example, PARAMETER
statements must be placed after all PROGRAM, FUNCTION,
or SUBROUTINE
statements,
and
before
all
statement
function
and executable
statements.
PARAMETER statements can be placed before,
after,
or
between all
IMPLICIT and other
specification statements.
Comment
lines may be interspersed anywhere in a program.
Generally if FORTRAN encounters statements that are out of place,
it
prints
warning messages and
continues compilation.
In some cases,
however, out-of-place statements cause
the compiler
to
terminate
compilation or generate unexpected results.

6-2

EXECUTABLE AND NONEXECUTABLE STATEMENTS

PROGRAM, FUNCTION, SUBROUTINE, or
BLOCI( DATA 1 Statements

IMPLICIT
Statements
PARAMETER
Statements
Comment
Lines
and
INCLUDE:i
Statements

Other
Specification
Statements
FORMAT
and
Entry2
Statements

NAMELIST
and
DATA
Statements

Statement
Function
Definitions

Executable
Statements

E'~D Statement

BLOCK DATA subroutines cannot contain any executable statements, statement
functions, FORMAT statements, EXTERNAL statements, INTRINSIC statements,
or NAMELIST statements (See Section 14.1).

The ENTRY statement is allowed only in functions or subroutines. All executable
statements which reference an~1 dummy parameters must physically follow the
ENTRY statement unless the references appear in the FUNCTION statement, the
SUBROUTINE statement, or in a precedin9 ENTRY statement.

The placement of an INCLUDE :,tatement is dictated by the types of statements to
be included.
MR-S-3822-85

Figure 6-1:

Ordering of FORTRAN Statements

NOTE

In FORTRAN-IO/20,
a DATA statement can precede a
PARAMETER
statement
or
another
specification
statement.

6-3

EXECUTABLE AND NONEXECUTABLE STATEMENTS
6.4

COMPILATION CONTROL STATEMENTS

You use compilation control statements to
identify FORTRAN programs
and
to specify their
termination.
Statements of this type do not
affect either the operations performed by the object program,
or the
manner in which the object program is executed.
The three compilation
control statements are:
1.

PROGRAM statement

INCLUDE statement

END statement

The PROGRAM statement and the INCLUDE statement are described
in
the
following sections.
The END statement is described in Section 9.8.

6.4.1

PROGRAM Statement

This statement allows you to give the main program a name other
than
the compiler-assumed name "MAIN." The general
form of a PROGRAM
statement is:
PROGRAM name
where:
name

is a symbolic name
that begins with an alphabetic
character and contains a maximum of six characters.
(See Section 4.1 for a description of symbolic names.)

The PROGRAM statement must be the first statement in a program unit.
(see Section 6.3 for
a discussion of the ordering of FORTRAN
sta temen ts. )

6.4.2

INCLUDE Statement

This statement allows you to
include code segments or external
declarations in a program unit without having them in the same file as
the primary program unit. When an INCLUDE statement is encountered
during compilation,
the compiler replaces the INCLUDE statement with
the contents of the specified file.
The general form of the
INCLUDE
statement is:
INCLUDE 'filespec [/switch]'
where:
filespec

is a TOPS-IO or TOPS-20 file specification
(refer
to
the
TOPS-IO or TOPS-20 Operating System
Commands manliaT):-- The only restricti~is-- that
under---rfOps=lo you cannot speci fy subd i rectory
information.

switch

is one of the following optional switches:
/CREF

indicates the
included
statements
are
to be used
to augment the
cross-reference listing (default).

6-4

EXECUTABLE AND NONEXECUTABLE STATEMENTS
/LIST

indicates that the statement in the
specified
file
is to be listed in
t~e compilation source listing.
A
number
indicating
the depth of
rlesting of include
files
precedes
(>.3ch statement listed (default).

/NOLI3T

indicates
that
the
included
statements are not to be prInted in
tje compilation listing.

/NOCREF

indicates

,; ta temen t s
Llqmen t
listjng.

that
the
included
are
not
to be used to
the
cross-reference

The following rules govern the use of the INCLUDE statement:
1.

The INCLUDEd file can contain
any
legal
~ORTRAN
statement
except a statement that terminates the current program unit,
such as the END, PROGRAM, FUNCTION, SUBROUTINE, or BLOCK DATA
statements.

The INCL~DEd file can contain other INCLUDE statements.
This
is called
nestinq INCLUDE statements.
The number of nested
levels i'3 restricted to 12.
2.

The proper
placement
of
the
INCLUDE
statement within
a
program
unit depends
upon
the
types
of statements to be
INCLUDEd.
(See Section 6.3 for information on
the
orderinq
of FORTRAN statements.)

The file to be INCLUDEd must be on disk.

Note that an asterisk (*) is appended
to
the
line
numbers
of
the
INCLUDEd
statements on the compIlation listing.
The level ot nesting
is indicated following the asterisk.

6-5

CHAPTE1R 7
SPECIFICATION AND DATA STATEMENTS

Specification statements are used
to
specify
storage allocation,
and data arrangement.
specification statements:
1.

DIMENSION

Statements that explicitly specify type,
including
INTEGER,
REAL, DOUBLE PRECISION, COMPLEX, LOGICAL, or CHARACTER

IMPLICIT

COMMON

EQUIVALENCE

EXTERNAL

INTRINSIC

PARAMETER

DATA

10.

SAVE

Specification statements are nonexecutable and
ordering guidelines described in Section 6.3.

7.1

type characteristics,
There are ten types of

must

conform

the

DIMENSION STATEMENT

The DIMENSION statement provides FORTRAN with
information needed
to
identify and allocate the space required for arrays.
You may specify
any number of subscripted array names as array declarators
in a
DIMENSION statement.
The general form of a DIMENSION statement is:
DIMENSION a(d)

[,a(d) ••• ]

where:
each a(d)

is an array declarator.
An array declarator
provides
the name and dimension(s) of an array.
An array declarator is written
in
the
following
form:
a(d [,d •.• ])

7-1

SPECIFICATION AND DATA STATEMENTS
where:
a

is the symbolic name of the array.

is the array dimension declarator.
declarator is as follows:
dl:

The form of a

dimension

] d2

where:
dl

is an optional
integer expression or
integer
variable specifying
the lower dimension bound.
The lower dimension bound is the first element
in
that dimension of the array.
If dl is not
specified the default is 1.

is an integer expression or integer variable that
specifies the upper dimension bound.
The upper
dimension bound is the greatest element
in that
dimension of the array.
You must have at least
one d2 specification in each array declaration.
If both dl and d2 are specified, dl cannot have a
value greater than d2.
The values dl and d2 can,
however, be the same.
An asterisk (*) can also occur as an upper bound,
but only as the last dimension.
An asterisk marks
the declarator as an assumed-size array declarator
(see Section 7.1.2).
NOTE

A dimension bound expression must
function or array element reference.

not

contain

If the array is a dummy argument to a subprogram, then dl and d2 can
be
integer dummy arguments and d2 can be an asterisk; otherwise, they
must be constants.
If the symbolic name of a constant or variable that appears
in a
dimension bound expression
is not of implicit default integer type
(see Section 4.2),
it must be specified
integer by an
IMPLICIT
statement or a type-statement.
Examples:
DIMENSION EDGE (-1:1,4:8), NET (5,10,4), TABLE
DIMENSION TABLE (IAB:J,K,M,10:20)
where:
lAB, J, K, and M are of type integer.

7-2

(567)

SPECIFICATION AND DATA STATEMENTS
7.1.1

Adjustable Dimensions

When used within a subprogram, a declarator for an array that is a
dummy argument can use integer dummy arguments as dimension bounds.
The following rules govern the use of adjustable dimensions
in a
subprogram:
1.

The array name must be a dummy argument.
Each variable that
is used as a dimension bound must be either a dummy argument
or in COMMON (see Section 7.4).

For multiple entry suoprograms, if any variables that specify
dimension bounds are dummy arguments which do not occur in
the formal argument list of the entry point used,
the value
of the variables as passed
for a prevIous call are used.
However, this is only permitted if the subprogram has
not
changed those dummy arguments.
Futhermore, when overlays are
used, a SAVE statement that preserves the local variables of
the subprogram is neejed.

If the value of an array dimension variable is altered within
the program, the dimensionality of the array is not affected.

The size of an array within a subprogram cannot
size of the original array,
as defined
in
program.

exceed the
the calling

Example 1:

SUBROUTINE SBR (ARRAY,Ml,M2,M3,M4)
DIMENSION ARRAY(Ml:M2,M3:M4)
DO 27 L=M3,M4
DO 27 K=Ml,M2
ARRAY (K,L)=VALUE
CONTINUE
END

In the example above, the dimensions of ARRAY
each time the subroutine SBR is entered.

will

re-specified

Example 2:
SUBROUTINE SBI (ARR1,M,N)
DIMENSION ARRl(M,N)
ARRl(M,N)=VALUE
ENTRY SB2(ARRl,M)
ENTRY SB3(ARRl,N)
ENTRY SB4(ARRl)

END
In the example above, the first call made to the subroutine must be
made
to SBI so that all of the dimension bounds have defined values.
If a call is made to SBI with the values M=ll and N=13,
a
succeeding
call
to SB2 will use the value N=13, but will give M a new value.
If
a succeeding call is made to SB4,
the
last values passed
through
entries SBl, SB2, or SB3 are used for M and N.

7-3

SPECIFICATION AND DATA STATEMENTS
7.1.2

Assumed-size Arrays

An assumed-size array is a dummy array for which the upper bound
the last dimension is specified as an asterisk(*), for example:

SUBROUTINE SUB(A,N)
DIMENSION A(l:N,l:*)
Since storage for array A is allocated in the calling routine,
the
upper bound of the rightmost dimension of A does not affect the
subscript calculations or storage allocation for A.
Therefore, subroutine SUB can be written to handle arguments with any
rightmost dimension
(the last subscript is never range checked for
being too large,
even when the /DEBUG:BOUNDS compiler switch is
specified). Such a subroutine can declare assumed-size arrays.
The size of an assumed-size array, and the number of elements that can
be referenced, are determined as follows:
1.

If the actual argument corresponding to the dummy array is a
noncharacter array name, the size of the dummy array is the
size of the actual-argument array.

If the actual argument corresponding to the dummy argument is
a
noncharacter array element name, with a subscript value of
s in an array of size a, the size of the dummy array is:
a+l-s.

If the actual
argument
is
a
character-array
name,
character-array element name,
or character-array element
substring name, and begins at character storage unit b of an
array with n character storage units, the size of the dummy
array is INT«n+l-b)/y). Where y is the length of an element
of the dummy array.

Because the actual size of an assumed-size
assumed-size array name cannot be used as:

7.2

array

not

known,

An array name in the list of an I/O statement

A unit identifier for an internal file in an I/O statement

A format specifier in an I/O statement

A NAMELIST statement element

TYPE SPECIFICATION STATEMENTS

Type specification statements explicitly declare the data
type of
variables, arrays, or function names.
You can give an array name in a
type specification statement, either alone (unsubscripted) to declare
the type of all
its elements, or with dimension bounds, to specify
both its type and dimensions.
There are two forms of type specification statements:
numeric type
specification
(see Section 7.2.1)
and character type specification
(see Section 7.2.2).

7-4

SPECIFICATION AND DATA STATEMENTS
7.2.1

Numeric Type Specification Statements

The general form of numeric type specification statements is:
type v [ , v ••• ]
where:
type

can be anyone of the following declarators:
1.

INTEGE:R

REAL

DOUBLE: PRECISION

COMPLE:X

LOGICJ,L

is a variable, array, or function name to be declared
the specified type.
The names listed must be separated
by commas and can appear in only one type statement
within a program unit.

Examples:
INTEGER A,
REAL R, M,

a, TABLE, FUNC
ARRAY(5:l0,lO~20,5)

If a name that is the same as an intrinsic FORTRAN function name
appears in a conflicting type statement, it is assumed that the name
refers to a user-defined routine, variable, or array of the given
type.
If you place a generic FORTRAN function name in an explicit
type statement, it loses its qeneric properties.
NO'I'E
The data
type
size modifier,
*n,
is accepted
by
FORTRAN-IO/20
to
be
compatible
with
the
type
statements
used
by other manufacturers.
You may
append
this size mod-.fier to the declarators, causing
some to E~licit messagE~s warning users of the
form of
the variable specified by FORTRAN-IO/20:
Declarator
INTEGER*2
INTEGER*4
LOGICAL*1
LOGICAL*2
LOGICAL*4
REAL*4
REAL*8
COMPLEX*8
COMPLEX*16
REAL*16
COMPLEX*32

Forn of Variable Specified
Full word integE~r with warning message
Full word i n tegE~r
Full word logical with warning message
Full word logical with warning message
Full word logical
Full word real
Doubl e-pr~?c is ion real
Complex
Complex with warning message
Doubl e-prl?c is ion real with warning message
Complex with wal::'ning message

7-5

SPECIFICATION AND DATA STATEMENTS
addition, you can append
the data
size
modifier
to
individual
variables, arrays,
or
function names.
Its effect
is
to override,
for
the
particular element,
the size modifier
(explicit or
implicit)
of the primary
type.
For example,
fn

~ype

HEAL*4 A, 8*8, C*8(lO}, D
A ~nd D are single-precision (one word)
real,
and Band C are double-precision
(two words for each element) real.

7.2.2

Character Type Specification Statements

The form of the character type specification statement is:
CHARACTER [*len[,]]

v[*len]

[,v[*len]] •••

where:
v

len

is one of the following:
•

The name of a symbolic constant, variable,
function subprogram

•

An array declarator

array,

is the length of the character data item and is one of the
following:
•

An unsigned, nonzero integer constant

•

An integer constant expression enclosed in parentheses
and with a positive value

•

An asterisk enclosed in parentheses

If you specify CHARACTER*len, len is the default length specification
for
that list.
If an
item in that list does not have a length
specification, the item's length is len.
But if an item does have a
length specification,
it overrides the default length specified in
CHARACTER*len.
A length specification of asterisk
(for example, CHARACTER*(*»
specifies that a dummy argument or function name assumes the length
specification of the corresponding actual argument or
function
reference
(see Chapter 13).
A length specification of asterisk for
the symbolic name of a constant specifies that the symbolic constant
assumes the actual length of the constant that it represents.
If you do not specify a length, a length of one is assumed.
Note that
a
length specification of zero is invalid.
You can use a character
type declaration statement to define arrays by including array
declarators
(see Section 4.3.2) in the list.
If you specify both an
array declarator and a length, the array bounds precede the length,
the form is:
a [ (d) ] [*len]

7-6

SPECIFICATION AND DATA STATEMENTS
where:
a

is an array name, and a(d)

is an array declarator.

Examples of character type specification statements follow:
CHARACTER*32

NAMES(lOO)

SOCSEC(lOO)~9,

The above stateme'nt specifies an array SOCSEC comprising one hundred
9-character elements, and an array NAMES comprising one hundred
32-character elements.
PARAMETER (LENGTH=4)
CHARACTER* (4 +LENGTH)

LAS~I,

FIRST

The above statements specify two 8-character variables,
LAST
FIRST.
(The PARAMETER statemEmt is described in Section 7.8.)

and

SUBROUTINE SI(BUBBLE)
CHARACTER LETTER(26), BUBBLE*(·)
The above statements specify an array LETTER comprising twenty-six
I-character elements and a dummy argument, BUBBLE, which has a length
defined by the cCllling program.
CHARACTER*16 QUEST* (5*IN~1 (A»
The above statement is invalid.
an integer constant expression.

7.3

The length specifier for QUEST is not

IMPLICIT STATEMENTS

IMPLICIT statements declare the data type of variables and functions
according to the first letter of each symbolic name.
The IMPLICIT
statement has two forms:
IMPLICIT type (a[,a] ••• )[,type (a[,a] .•• )] •••
IMPLICIT NONE
As shown in the statement above, an IMPLICIT statement consists of one
or more type declarators sE~parated by commas.
Each type declarator
has the form:
type

(a[,a] .•• )

where:
type

can be anyone of the following declarators:
1.

INTEGER

2.,

REAL

3.,

DOUBLE PRECISION

4.,

COMPLEX

LOGICAL

CHARACTER [,'" len]

7-7

SPECIFICATION AND DATA STATEMENTS
a

is an alphabetic specification in either of the general
forms:
c or cl-c2, where c, cl, or c2 is an alphabetic
character.
The latter form specifies a range of
letters,
from cl through c2,
which must occur in
alphabetical order.

When you specify type as CHARACTER*len, len specifies the length for
character data type.
Len is an unsigned, nonzero integer constant or
an integer constant expression enclosed
in parentheses and with a
positive value.
If you do not specify a length, a length of one is
assumed.
Each letter in a type declarator list specifies that each symbolic
name
(not declared
in an explicit type specification statement)
starting with that letter is assigned
the data
type named
in the
declarator.
For example, the statement:
IMPLICIT REAL (R,M,N,O)
declares that all names that begin with the letters R, M, N, or 0 are
type REAL names,
unless declared otherwise in an explicit type
statement.
NOTE
Type declarations
given
in
an
explicit
type
specification override those also given in an IMPLICIT
statement.
IMPLICIT declaratinns do
not
affect
intrinsic functions.
The length is also overridden
when a particular name appears
in a CHARACTER or
CHARACTER FUNCTION statement (see Chapter 13).
You may specify a range of letters within the alphabet by writing the
first and
last letters of the desired range separated by a dash, for
example, A-E for A,B,C,D,E.
Thus, the statement:
IMPLICIT INTEGER (I,L-P)
declares that all symbolic names that
I,L,M,N,O, and P are of type INTEGER.

begin

with

the

letters

You may use more than one IMPLICIT statement, but they must appear
before any other declaration statement
in the program unit.
(See
Section 6.3 for a discussion on ordering FORTRAN statements.)
The same letter must not appear as a single letter, or be included
in
a
range of letters, more than once in all of the IMPLICIT statements
in a program unit.
You can use an IMPLICIT NONE statement to provide warning messages for
variables not explicitly declared,
including variables implicitly
declared by other IMPLICIT statements.
If you specify IMPLICIT NONE,
no other IMPLICIT statement should be included in the program unit.

7.4

COMMON STATEMENT

The COMMON statement enables you to establish storage that may be
shared by two or more programs and/or subprograms, and to name the
variables and arrays that are to occupy the common storage.
The use
of common storage conserves storage and provides a means to reference
the same data in different subprograms without passing arguments.
7-8

SPECIFICATIOlJ AND DATA STATEMENTS
COMMON statements have the
COMMON [/cb/]

fo~lowing

form:

nlist[[,]/[cb]/nlist] •..

where:
cb

is an optional common block name.
the rules for ~;ymbol ic names.)

nlist

names,
array names,
and array
that are
to occupy the common area.
The
items specified for a common area (block)
are ordered
within
storagc~
as
they are listed
in the COMMON
statement.

is a list of

(See Section 4.1 for

variablE~

dE~clarators

all
numeric data
type or all
The items in nlist must be (~ither
character data
type.
A common block cannot contain both numeric and
character data.
A symbolic name can be used to identify each block.
However, you can
This
omit
the symbolic name
fo~
one block
in a
program unit.
unlabeled block is called the blank common block.
The elements of a named commOll block can be assigned initial values by
DATA statements appearing ill the BLOCK DATA subprograms (see Chapter
14).
In standard-conforming programs,
the elements of the blank
common
block
may
not
be assigned
initial
values.
However,
FORTRAN-lO/20 allows any comml)n block elements to be defined in a DATA
statement in any program unit.
A given common block name may appear more than once in the same COMMON
statement,
and
in more
th.3n one COMMON statement within the same
program or subprogram.
When extended addressing is i~ effect, COMMON blocks
reside
in
the
large
data
area
by def3ult.
However,
the /EXTEND:COMMON or
/EXTEND:NOCOMMON switches can be used to explicitly allocate COMMON
blocks in the large data area or small data area (see Section 16.5).
During compilation of a source program, FORTRAN strings
together all
items
listed
for each commo~ block in the order in which they appear
in the source program.
For e~ample:
COMMON X,Y,Z/STl/A,B
COMMON/STl/TST(3,4)/ST2/rAB(2,2)
COMMON/ST2/C,D,E//P,Q
COMMON W
has the same effect as the si,gle statement:
COMMON X,Y,Z,P,Q,W/STl/A,B,TST(3,4)/ST2/TAB(2,2),C,D,E
All elements specified for a common block are placed
into one area.
Common block names must be unique with respect to all subroutine,
function, and entry point names.

7-9

SPECIFICATION AND DATA STATEMENTS
NOTE
If you use overlays, you can use the SAVE statement to
retain the value of variables in a named common across
overlays (see Section 7.10).
(Blank common is always
saved.)
For example:
Main Program

Subprogram

COMMON DELTA, LENGTH
COMMON /COMI/KILOS,PRICE

SUBROUTINE CALC
COMMON/COMl/N,A
COMMON Z,KOUNT

CALL CALC

The COMMON statements in the main program put DELTA and
LENGTH into
the blank common block, and put KILOS and PRICE into a common block
named COM1.
The COMMON statements in the subroutine make Z correspond to DELTA in
the main program, KOUNT correspond to LENGTH, N correspond to KILOS,
and A correspond to PRICE.
To prevent conflict with other COMMON blocks, LINK requires
largest definition for each common block be loaded first.

7.4.1

that

the

Dimensioning Arrays in COMMON Statements

Array names with dimension bounds can be given in COMMON statements.
However,
variables cannot be used as dimension bounds in a declarator
appearing in a COMMON statement;
adjustable dimensioning
is not
permitted in COMMON.
Each array name given in a COMMON statement must be dimensioned either
by the COMMON statement or by another dimensioning statement within
the program or subprogram that contains the COMMON statement, but not
both.
For examples,
COMMON /A/B(lOO), C(lO,lO)
COMMON X(5,15) ,Y(5)

7.4.2

Character Data in COMMON

Each character variable in a COMMON block is allocated to start at the
first available character position.
For example,
CHARACTER B*3,C*3,D(3)*2
COMMON B,C,D

7-10

SPECIFICATION AND DATA STATEMENTS
The COMMON block will be allocated in the following way:

0(1)

0(2)

0(3)

Unused

~ r,,------A----___,

rr----',A------..,,--~----...\f~~

Ix I
o

~!1

x I

MR-S-252783

where x means the bits are not used.

7.5

EQUIVALENCE STATEMENT

The EQUIVALENCE statement associates two or more
same storage location.
The format of the EQUIVALENCE
EQU I VALENCE (n lis t)

HtatemE~nt

variables

with

the

is:

[, (n Ii!; t ••• ) ]

where:
nlist

is a list of variablE~ names,
array elements,
array
names,
and chal~acter substring references separated by
commas and enclosed in parentheses.
You must specify
two or more of these items in each list.

In an EQUIVALENCE statement, each expression
in a subscript
substring reference must be an integer constant expression.
The EQUIVALENCE statement a:.locatE!s all of the
items
in
parenthesized list beginning at the same storage location.
example, the statements:

a
one
For

EQUIVALENCE (,A,B,C)
EQUIVALENCE (LOC,SHARE{3):
specify that the variables A, B, and C are to share the same storage
location, and that: the variablE! LaC and the array element SHARE(3) are
to share the same location.
The relationship of equivalence
is transitive.
following statements have the game effect:

For

example,

the

EQUIVALENCE ('A,B) , (B,C)
EQUIVALENCE (A,B,C)
The following EQUIVALENCE statE~ment makes the first character of the
character variables KEY and STAR share the same storage location.
The
character variable STAR is equjvalent: to the substring KEY (l:lO):
CHARACTER KEY*16, STAR*lO
EQUIVALENCE (KEY,STAR)
You can equivalence variables of different numeric data
types.
Character variables must not: be equivalenced to numeric variables.
For example, you can equivalence a
real
variable equivalent to a
complex variable.
In this case, since each complex variable occupies

7-11

SPECIFICATION AND DATA STATEMENTS
two words of storage, and each single-precision variable occupies one
word of storage, if you equivalence a real and a complex variable, the
real variable shares storage with
the real part of the complex
variable.
Figure 7-1 depicts
the shared storage when a complex
variable is equivalenced with a real variable.
Source Program Statements:
COMPLEX A
REAL B
EQUIVALENCE (A,B)

Memory Location
A or B

Stores: Real Part of Complex A
or Entire Real B

Second Part of
Memory Location A

Stores: Imaginary Part of
Complex A
~-------------------------~

1--36-8il word--!
MR-S-1764-81

Figure 7-1:

Shared Storage using EQUIVALENCE Statement

The EQUIVALENCE statement does not
imply
(or perform)
any type
conversions.
If you equivalence a
real
variable and an integer
variable, the data within the equivalenced location will be treated as
a real or integer number, depending on whether it is referenced by the
real or integer variable.
If you equivalence a real variable with a double-precision variable,
the data in the high-order word of the double-precision variable will
be
used
by
the
real
variable.
For
positive
D-floating
double-precision numbers (see Section 3.4), the high-order word is in
the same format as a single-precision number.
For G-floating double-precision numbers (KL model B only - see Section
3.4),
the high-order word is not in the format of a single-precision
number.
Thus,
equivalencing a
real variable and a
G-floating
double-precision variable will produce incorrect results.
Equivalencing a negative D-floating number and a real variable may not
produce correct results either, for example the number:

577000000000
000000000001
777777777777, almost 1.0)

(the

negative

200777777777

does not have a valid single-precision number in its high-order word.
If you equivalence an array and a variable, the array does not assume
any of the properties of the variable, and the variable does not
assume any of the properties of the array.
When you use an array element in EQUIVALENCE statements, it must have
either as many subscripts as dimensions of the array, or only one
subscript.
In either case, the subscripts must be integer constants.
Note
that
the
single subscript case
treats
the array as a
one-dimensional array of the given type.

7-12

SPECIFICATION AND DATA STATEMENTS
The following example shows
the
effect
I-dimensional and a 2-dimensional array:

equivalencing

DIMENSION A(3,2) ,B(6)
EQUIVALENCE (A(l,l) ,B(l»
or
EQUIVALENCE (A(l),B (1»
In this example, each array elE~ent of array A shares the same storage
area with an element of array B:
A(I,l)
A(2,1)
A(3,1)
A(1,2)
A(2,2)
A(3,2)

B(l)
B (2)

B(3)
B (4)

B(5)
B(6)

Specifying an array name wjthout subscripts in an
EQUIVALENCE
statement is the same as specifying the first element of the array.
When you make one character substring equivalent to another character
substring,
the EQUIVALENCE statement also sets equivalences between
the other corresponding characters in the character strings,
for
example,
CHARACTER NAME*16, 10*9
EQUIVALENCE (NAME (10: 13), ID (2: S) )
As a result of these statements, the character variables NAME
share space as illustrated in rigure 7-2.

NAME
Character
Position

1
2
3
4

5
6
7

10
Character
Position

8
9

1
2

10
11

3
4
5
6
7

1 '.I.1 ~I
1 t!~

H.
Hi

8
9
MR-S-2523-83

Figure 7-2:

Equivalence of Substrings

7-13

and

SPECIFICATION AND DATA STATEMENTS
The following statement also aligns the variables as shown
7-2:

Figure

EQUIVALENCE (NAME(9:9) ,ID(l:l))
the
If the character substring references are array elements,
between
the
other
EQUIVALENCE
statement
sets
equivalences
corresponding characters in the complete arrays.
Character elements of arrays can overlap at
for example:
CHARACTER FIELDS(6)*4, STAR(5)*5
EQUIVALENCE (FIELDS(l) (2:4), STAR(2)

any

character

posit.ion,

(3:5))

As a result of these statements, the character arrays FIELDS and
share storage space as shown in Figure 7-3.
STAR

Character
Position

Subscript

2
3
FIELDS

Subscript

Character
Position

2
3
2

2
3

4
5

4
6

2
3
4
MR-S-2S24-83

Figure 7-3:

2
3

Equivalence of Character Arrays

7-14

STAR

SPECIFICATION AND DATA STATEMENTS
General EQUIVALENCE Restrictions:
1.

You cannot cause two different elements of an array to become
equivalenced
to each other.
Thus, the following statement
sequence is prohibited because it specifies the same storage
location (B) for A(l) and A(2):
DIMENSION A(2)
EQUIVALENCE (A(l) ,B), (A(2) ,B)

An EQUIVALENCE statement must
not
specify
that
two
consecutive locations are nonconsecutive.
For example, the
following statement sequence is prohibited because B(l) takes
two storage locations,
the second of which would make A(2)
nonconsecutive to A(l):
INTEGER A(2)
DOUBLE PRECISION B(2)
EQUIVALENCE (A(l) ,B(l)),

(A(2) ,B(2))

An EQUIVALENCE statement in a SUBROUTINE
or
FUNCTION
subprogram must not refer to an argument of the subprogram.
For example, the following statement sequence is prohibited:
SUBROUTINE A(B,C)
EQUIVALENCE (B,X)

You cannot cause two different substrings of the same
character variables or arrays to become equivalenced to each
other.
For example,
the following
statement sequence
is
prohibited because it specifies the same substring B(1:3) for
A(1:3) and A(2:4):
CHARACTER A(3)*4,B*4
EQUIVALENCE (A(l) (1:3) ,B(1:3)),

(A(l) (2:4) ,B(1:3))

You also cannot use the EQUIVALENCE statement to assign
memory locations
in a way that
is inconsistent with the
normal linear storage of character variables and arrays.
For
example,
the following statement sequence
is prohibited
because it would require the character variable B(2:2) to be
equivalent to both A(l) (2:2) and A(l) (1:1):
CHARACTER A(2)*lO,B*lO
EQUIVALENCE (A(l) (1:3) ,B(1:3)),

(A(l) (4:6) ,B(5:7))

Restrictions on EQUIVALENCE and COMMON:
1.

You cannot use the EQUIVALENCE statement to equivalence two
elements
in different common blocks.
Thus, the following
statement sequence is prohibited:
COMMON /BLOCKl/A,B,F/BLOCK2/C,D,E
EQUIVALENCE (A,C)

You cannot set two quantities declared in a COMMON block to
be equivalent to one another.
Thus, the following statement
sequence is prohibited:
A,B,C
EQUIVALENCE (A,C)

COM~MON

7-15

SPECIFICATION AND DATA STATEMENTS
3.

Quantities placed in a common area by means of an EQUIVALENCE
statement are permitted to extend the end of the common area
forward.
For example, the statements:
COMMON/R/X,Y,Z
DIMENSION A(4)
EQUIVALENCE (A,Y)
cause the common block R to extend from Z to A(4) arranged as
shown in Figure 7-4.

Location X

Location Y and A(1)

}

~-----------t

Shared Locations

Location Z and A(2)

Location A(3)
Location A( 4)
MR-S-1746-81

Figure 7-4:

Valid Equivalencing

You cannot use EQUIVALENCE statements that cause the start of
a common block
to be extended backwards.
For example, the
invalid sequence:
COMMON/R/X,Y,Z
DIMENSION A(4)
EQUIVALENCE(X,A(3) )
would require A(l) and A(2) to extend the starting
location
of block R in a backwards direction as illustrated in Figure
7-5.

Location A(1)
Location A(2)

Location X and A(3)
t-----------I

}
Causes COMMON R to Extend Backward

Location Y and A(4}

Location Z
MR-S-1747-81

Figure 7-5:

Invalid Equivalencing

7-16

SPECIFICATION AND DATA STATEMENTS
7.5.1

EQUIVALENC:E: and Extended Addr1essing

When extended addressing is in effect, and an EQUIVALENCE statement
causes a
variable to be in COMMON, that variable resides in the same
psect as the rest of the COMMON block.
For variables not in COMMON,
if you equivalence a
large variable
(default 10,000 or more words)
with other variables
(including
scalars), all these variables will reside in
the
large data psect.
For example,
REAL A(20000),X
EQUIVALENCE (A(l),X)
causes variable X to be placed in the large data psect,
equivalenced with a large array.

since

Conversely, if each equivalence variable is small (default less
than
10,000 words),
they will reside in the small data psect, even if the
total size of the equivalence class is over the small variable limit.
For example,
REAL A(8000) ,B(8000)
EQUIVALENCE (A (8000), B (1»
The arrays will reside in the small data psect, because
smaller than 10,000 words.

each

one

See Sections
addressing.

extended

7.6

15.4.11

and

16.5

for

information

EXTERNAL STATEMENT

Any user subprogram name to be used as an argument to another
subprogram must appear
in an EXTERNAL statement
in the calling
subprogram.
The EXTERNAL statement declares names
to be subprogram
names to distinguish them from other variable or array names.
The subprograms mentioned in the EXT:E:RNAL statement cannot be FORTRAN
intrinsic
functions;
they can be only user-supplied
functions,
subroutines, or block data subprograms.
(The
INTRINSIC statement
discussed in Section 7.7 allows intrinsic function names to be used as
arguments.) The EXTERNAL state~ent has the following form:
EXTERNAL proc[,proc •.• ]
where:
proc

is the symbolic name of a user-supplied subprogram, the
name of a dummy argument associated with the name of a
subprogram, or a block data subprogram.

The EXTERNAL statement declares each symbolic name included in
it to
be the name of an external procedure, even if a name is the same as
that of an intrinsic function.
For example, if SIN is specified in an
EXTERNAL statement
(EXTERNAL SIN), all subsequent references to SIN
are to a
user-supplied
function name SIN,
not
to the
intrinsic
function of the same name.
The name specified in an EXTERNAL statement can be used as an actual
argument
to a subprogram, which can then use the corresponding dummy
argument in a function reference or a CALL statement.
7-17

SPECIFICATION AND DATA STATEMENTS
NOTE
Note that a complete function reference used as an
argument,
for
instance,
FUNC(B)
in CALL SUBR
(A(FUNC(B) ,C», represents a value, not a subprogram.
therefore,
A complete function reference is not,
defined in an EXTERNAL statement.
The interpretation of the EXTERNAL statement described above is
different from that of earlier versions of FORTRAN-lO/20.
:f the
INOF77 compiler switch is specified (see Sections 16.1.3 and 16.2.3),
lhe subprogram names can be intrinsic functions.
the names
Por compatibility with previous versions of FORTRAN-10/20,
of external subprograms can be preceded by an asterisk (* ) or an
ampersand (&) within an EXTERNAL statement.
For example,
EXTERNAL *SIN, &COS
declares SIN and COS to be user subprograms.
(If a prefixed
not an intrinsic function, then the prefix is ignored.)

name

Note that specifying an intrinsic function in an EXTERNAL statement
without a prefix (that is, EXTERNAL SIN) has no effect upon the usage
of the function name outside of actual argument lists.
If the name
has generic properties, they are retained outside the actual argument
1 ist.
(The name has no generic properties within an argument list.)
The names declared
in a program EXTERNAL statement are reserved
throughout the compilation of the program, and cannot be used in any
declarator statement other than a type statement.

7.7

INTRINSIC STATEMENT

The INTRINSIC statement allows you to use intrinsic function names as
arguments to subprograms.
See Section 13.1 for further information on
intrinsic functions.
The INTRINSIC statement has the form:
INTRINSIC fun[,fun] •••
where:
fun is the symbolic name of an intrinsic function.
The INTRINSIC statement declares each symbolic name included in it to
be the name of an intrinsic procedure.
This name can then be used as
an actual argument to a subprogram, which can use the corresponding
dummy argument in a function refe~ence or a CALL statement.
The appearance of a generic function name in an
INTRINSIC
does not cause that name to lose its generic property.

7-18

statement

SPECIFICATION AND DATA STATEMENTS
NOTE
You cannot use a generic-only name in an INTRINSIC
statement.
The generic name must be the same as an
instrinic function name.
For example,
INTRINSIC LOG
is illegal because there is no function named
'LOG'.
LOG is the generic name that selects between functions
such as ALOG, DLOG, or CLOG.
Only one appearance of a symbolic name is permitted in all of the
INTRINSIC statements of a program unit.
Also, a symbolic name must
not appear in both an EXTERNAL and an INTRINSIC statement in a program
unit.
An example of
follows:

the

use

the

EXTERNAL

and

INTRINSIC

statements

Main Program

EXTERNAL CTN
INTRINSIC SIN, COS

CALL TRIG(ANGLE,SIN,SINE)

CALL TRIG(ANGLE,COS,COSINE)

CALL TRIG(ANGLE,CTN,COTANT)

Subprograms
SUBROUTINE TRIG(X,F,Y)
Y = F(X)
RETURN
END
FUNCTION CTN(X)
CTN = COS(X)/SIN(X)
RETURN
END
In this example, when TRIG is called with a second argument of SIN or
COS, the function reference F(X) references the math library functions
SIN and COS; but when TRIG is called with a second argument of CTN,
F(X) references the user function CTN.

7-19

SPECIFICATION AND DATA STATEMENTS
7.8

PARAMETER STATEMENT

The PARAMETER statement allows you to
during compilation.

define

constants

symbolically

The general form of the PARAMETER Statement is:
PARAMETER

(p=c[,p=c) ••. )

where:
s~nbolic

is a

name.

is a constant expression (except for expressions involving
multiplication,
division,
or exponentiation of complex
numbers).
(See Chapter 3 for
a descr ipt ion of FORTRAN
constants.)

The constant acquires the same data type as the symbolic name.
If the
symbolic name is of type integer, real, double precision, or complex,
the corresponding expression
(c)
must be an arithmetic constant
expression
(see Section 5.1.2).
If the symbolic name is of type
character or logical, the corresponding expression
(c)
must be a
character constant expression
(see Section 5.2.1)
or a
logical
constant expression (see Section 5.3.1), respectively.
The data type of a symbolic name defined to be a constant is specified
by a
type-statement or
IMPLICIT statement preceding the defining
PARAMETER statement.
Also, if the length specified for
the symbolic
name of a character constant is not the default length of one, its
length must be specified by a
type-statement or
IMPLICIT statement
preceding the symbolic name of the constant.
NOTE
The form and
the
interpretation to the PARAMETER
statement described above are different from those of
the PARAMETER statement provided in earlier versions
of FORTRAN-IO/20.
The earlier version is described
below.
This form and
interpretation can still be
used,
however a warning message will be issued.
This
form of the PARAMETER statement is:
PARAMETER p=c[,p=c ••• ]
The symbolic name acquires the same data type
constant.

the

During compilation,
the symbolic names are replaced
by
their
associated constants, provided the following rules are observed:
1.

Symbolic names may appear only where
acceptable.

Symbolic name references
statement definition.

Symbolic names must be unique with respect to all other names
in the program unit.

must

7-20

appear

FORTRAN
after

constants
the

are

PARAMETER

SPECIFICATION AND DATA STATEMENTS
4.

Symbol ic names must r!ot be redef ined in subsequent
statements.

Symbolic names must not be used as part of another constant,
such as within a character constant or as the count for a
Hollerith constant.

Symbolic names
specification.

part

DATA statements are used to supply the initial values
arrays, array elements, substrings, and COMMON areas.

variables,

7.9

must

not

used

PARAMETER

format

DATA STATEMENT

The form of

th~

data statement. is:

DATA nlist/clist/ [[,]nlist/clist/] •..
where:
nlist

identifies a set of items to be initialized.

clist

contains the set of values to be assigned the items
nlist.

For example, the statement:
DATA IA/5/,IB/IO/,IC/15/
initializes variable IA to the value 5, variable IB to the value 10,
and variable IC to the value 15.
The number of storage locations you
specify in the list of variab]es must be equal
to the number of
storage locations you specjfy in its associated list of values.
If
not, a warning m€!ssage is output.
When the value list specifies more storage locations than the variable
list,
the excess values are ignored. When the value list specifies
fewer storage locations than the variable list, the excess variables
are not initialized.
The nlist portion of each nlist/clist/ set can contain
the names of
one
or more variables, array names,
array elements, character
substring names, or labeled COMMON variables.
You may specify an
entire array (unsubscripted array name) or a portion of an array in a
DATA statement nlist as an implied DO loop construct.
(See Section
10.4.9.2 for a description of implied DO loops.)
The form of an implied-DO list: in a DATA statement is:
(dlist,i=n1,n2[,n3])
where:
d1ist

is a list of array element names,
names, or impljed-DO lists.

is the name of an integer
index variable.

variable,

called

n1,n2,n3

are integer E~xpressions that can
constants and loop index variables.

contain

7-21

character

substring
the

loop

integer

SPECIFICATION AND DATA STATEMENTS
For example, the statement:
DATA (NARY(I) ,I=1,S)/1,2,3,4,S/
initializes the first
five elements of array
NARY(2)=2, NARY(3)=3, NARY(4)=4, and NARY(S)=S.

NARY

NARY(l)=l,

When you use an implied DO loop in a DATA statement,
the loop index
variable must be of type
INTEGER,
and the Initial, Terminal, and
Increment parameters of the loop must be of type INTEGER.
In a DATA statement, references to an array element or substring must
be integer expressions in which all terms are either integer constants
or indices of embracing implied DO loops.
These types of integer
expressions can include the exponentiation operator.
The clist portion of each nlist/clist/ set can contain one or more
numeric,
logical,
Hollerith,
octal,
hexadecimal,
or character
constants.
You may specify literal data as either
a
Hollerith
sgecific3tion,
for example, SHABCDE,
or a string enclosed
in single
quotes,
for example,
'Abcde'.
Each ASCII data
item is stored
left-justified and is padded with blanks if necessary.
When you assign the same value to more than one
item in nlist, a
repeat specification may be used.
The repeat specification has the
form:
n*d
where:
n

is an integer that specifies how many times the value d
is
to be used.
For example, a clist specification of /3*20/
specifies that the value 20 is to be assigned to
the first
three items named in the preceding list.
The statement:
DATA M,N,L/3*20/
assigns the value 20 to the variables M, N, and L.

When the specified data type is not the same as that of the variable
to which it is assigned, FORTRAN converts the data item to the type of
the variable.
The type conversion is performed using the rules 9iven
for
type conversion
in arithmetic assignments.
(See Table 8-1.)
Octal, logical, Hollerith, hexadecimal, and character constants are
not converted.
Sample Statement

Result

DATA PRINT,I,0/'TEST',30, 11 77/, (TAB(J),J=1,30)/30*S/

The first 30
elements of
array TAB are
initialized to
S.O.

DA TA ( (A ( I , J) , I = 1 , S) , J = I , 6 ) /30 * I . 0/

No conversion
required.

DATA( (A(I,J) ,I=S,10) ,J=6,15)/60*2.0/

No conversion
required.

7-22

SPECIFICATION AND DATA STATEMENTS
When character variables are initialized, length conversions are made.
The conversion is based on the following rules:
1.

If the constant contains fewer characters than the length of
the element in nlist, the rightmost character positions of
the element are initialized with spaces.

If the constant contains more characters than the length of
the element in nlist, the character constant is truncated to
the right.

Character constants and
HollErith constants can also be used
to
initialize numeric variableE.
The character string is stored left
justified
in
the numeric variable.
When
the character
string
specified
is longer than one numeric variable can hold, the string is
stored left justified across ES many variables as are needed
to hold
it.
If necessary, the last variable used is padded with blanks up to
its right boundary.
For character variables, each variable or array
exactly one character constant in the data list.
For example, assuming that X,
statement:

and

are

element

must

have

single-precision,

the

DATA X,Y,Z/'abcdefghijkl'/
causes:
X to be initialized to '2bcde'
Y to be initialized to 'fghij'
Z to be initialized to '~lbbb'

When a character string is to be stored
in double-precision and/or
complex variables, and the specified string is only one word long, the
second word of the variable iE padded with blanks.
For example, assuming that thE variable C is complex, the statement:
DATA C/'ABCDE','FGHIJ'/
causes the first word of C to be initialized to 'ABCDE' and its second
word
to be
initialized
to 'bbbbb'.
The string 'FGHIJ' is ignored.
The first word contains the rEal part of
the ~omplex
variable;
the
second word contains the imaginary part.
In addition, the following twc
in DATA statements:

forms of bit data constants are allowed

O'di ... dn'
Z'hi ... hn'
where di are octal digits and hi
are hexadecimal digits with A-F
representing
the decimal eGuivalent of 10-15.
These constants are
right-justified.
Note that you can also use the double quote (") form
of octal constants as describE'd in Section 3.7.

7-23

SPECIFICATION AND DATA STATEMENTS
7.10

SAVE STATEMENT

The SAVE statement retains the values stored in a variable, array,
or
common block after execution of a RETURN or END statement in a
subprogram.
The SAVE statement has the following form:
SAVE [a[,a] ••• ]
where:
a

is a named common block name (preceded and
slash), a variable name, or an array name.

followed

NOTE
Ordinarily, the values of all variables are retained
after
execution
of a
RETURN or END statement.
However, when overlays are used,
the SAVE statement
must be used to ensure retention of values.
An entity specified by a SAVE statement within a program unit does not
become undefined upon execution of a RETURN or END statement in that
unit.
If the entity is
in a common block, however,
it may be
redefined in another program unit that references that common.
Procedure names, the names of variables and arrays in a common
and dummy argument names cannot be used in a SAVE statement.

block,

A SAVE statement that does not explicitly contain a list is treated as
though it contained a list of all allowable items in the program unit
that contains the SAVE statement.
If a particular common block name is specified by a SAVE statement in
a subprogram of an executable program, it must be specified by a SAVE
statement in every subprogram in which that common block appears.
NOTE
It is not necessary to use the SAVE statement to
retain
the value of a blank common block;
the
definition status of blank common is automatically
retained after a RETURN or END statement.
Also, when the SAVE statement is used,
it is not
necessary to specify the LINK switch /OVERLAY:WRITABLE
when loading a program.

7-24

CHAPTER 8
ASSIGNMENT STATEMENTS

Assignment statements assign values to variables, array elements,
or
character substrings.
There are four kinds of assignment statements:

8.1

Arithmetic assignment statements (see Section 8.1)

Logical assignment statements (see Section 8.2)

Statement Label assignment (ASSIGN)
8.3)

Character assignment statements (see Section 8.4)

statements

(see

Section

ARITHMETIC ASSIGNMENT STATEMENT

You use statements of this type to assign numeric values
to numeric
variables or array elements. Arithmetic assignment statements have
the form:
v=e
where:
v

is the name of the numeric variable or array element that is
to receive the specjfied value.

is an arithmetic expression.

In assignment statements, the equal symbol (=) does not imply equality
as
it would
in algebraic E?xpressions; it implies replacement.
For
example, the expression v=e is interpreted as "the contents of the
location identified as v are t:o be replaced by the value of expression
e; the previous contents of v are lost."
When the type of the specified variable or array element name differs
from that of its assigned value, FORTRAN converts the value to the
type of its assigned variable or array element.
Table 8-1 describes
the type conversion operations performed by FORTRAN for each possible
combination of variable and value types.

8-1

ASSIGNMENT STATEMENTS
Table 8-1:

Rules for Conversion in Mixed-Mode Assignments

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

Expression
Type (e)

VariabLe Type (v)
~------------------------------------------.--------------------~

REAL

INTEGER

COMPLEX

DOUBLEPRECISION

LOGICAL

CHARACTER

~---------------+----------------------------------------------------------------~

Real

R,I

H,L

Integer

R,F,I

Complex

C,R

R,L

Doubleprecision

H,I

Logical

R,I

H,L

Octal

R,I

H,L

Hollerith

Character

DoubleOctal*

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

Legend
D = Direct replacement
C = Conversion from real to integer with truncation, overflow i ~
possible
F = Conversion from integer to real with rounding
R = Real part only
I = Set imaginary part to 0
H = High-order only
L = Set low-order part to 0
M = Convert with no truncation and no rounding
N = Convert with rounding; truncation can occur and overflow is
possible
0 = Round to one word of precision, overflow is possible
X = Not allowed
Notes
*

Octal numbers with 13 to 24 digits are
termed double-octal.
Double-octals require two storage locations.
They are stored
right-justified and are padded with
zeros to
fill
the
locations.

8-2

ASSIGNMENT STATEMENTS
&

Use the first two words of the Hollerith constant.
If the
Hollerith constant
is only one word long, the second word is
padded with blanks.

Use the first word of

To convert double-octal ~umbers to complex, the low-order octal
digits are assumed to be the imaginary part, and the high-order
digits are assumed to be the real part of the complex value.

8.2

thE~

HollE!rith constant.

LOGICAL ASSIGNMENT STATEMENTS

Statements of this type are used to assign values to variables and
array elements of type logical.
Logical assignment statements have
the following form:
v=e
where:
v

is the name of a variable or array element

is a logical expression

For example, assuming that the variables L, F, M, and G
logical, the following statements are valid:
Sample Statement

type

replaced

Results

L=.TRUE.

The contents of L are
logical truth.

F=.NOT.G

The contents of F are replaced by
the complement of the contents of
G.

M=A.GT.T
or
M=A)T

L=«I.GT.H) .AND. (J<=:K)

8.3

are

If A is greater
than
T,
the
contents
of M are replaced by
logical truth; if A is less than or
equal
to T, the contents of Mare
replaced by logical
false.
This
can also be read:
If A is greater
than T, then M is true,
otherwise,
M is false.
The contents of L are replaced by
either the true or false resultant
of the expression.

ASSIGN (STATEMENT LABEL) ASSIGNMENT STATEMENT

The ASSIGN statement is used to assign a statement label constant (a 1
to 5 digit statement number)
to a variable name.
The form of the
ASSIGN statement is:
ASSIGN s TO i

8-3

ASSIGNMENT STATEMENTS
where:
s

is a statement number in the current program unit.

is a variable name.

For example, the statement:
ASSIGN 2000 TO LABEL
specifies that the variable
2000.

LABEL

references

the

statement

number

with the exception of complex, double-precision, or character, you can
use any type of variable in an ASSIGN statement.
Use the ASSIGN statement in conjunction with assigned GO TO control
statements
(see Chapter 9), or as a
format identifier in an I/O
statement (See Chapter 10). The ASSIGN statement sets up statement
label variables that are then referenced in subsequent GO TO control
statements, or in format specifiers in I/O statements.
The following
sequence illustrates the use of the ASSIGN statement:
555 TAX=(A+B+C)*.05

ASSIGN 555 TO LABEL

GO TO LABEL

8.4

CHARACTER ASSIGNMENT STATEMENT

The character assignment statement assigns the value of the character
expression on the right of the equal sign to the character variable,
array element, or substring on the left of the equal sign.
The form of the character assignment statement is:
v=e
where:
v

is a character variable, array element, or substring.

is a character expression.

If the length of the expression on the right side of the assignment is
greater than the length of the variable on the left side, the
character expression is truncated on the right.
If the length of the expression on the right side of the assignment is
less than the length of the variable on the left side, the character
expression is filled on the right with blanks.

8-4

ASSIGNMENT STATEMENTS
FORTRAN-IO/20 allows overlap between the character expression and the
character variable, array element,
or substring.
(That is, the
character positions defined in the character variable, array element,
or substring can be referenced
in
the character expression.) For
example, the following assignments are allowed:
CHARACTER *4 A,B
DATA A/'abcd'/,B/'efgh'/
A(1:3)

A(2:4)

B(2:4)

B(1:3)

After the above assignment statements,

is 'bcdd', and B is 'eefg'.

The expression must be of character data type.
You cannot assign
numeric value to a character variable, array element, or substring.

Note that assigning a value to a character substring does not affect
character positions in the character variable or array element not
included in the substring.
If a character position outside of the
substring has a value previously assigned, it remains unchanged.
If
the character position is undefined, it remains undefined.
Examples of valid and invalid character assignment statements follow.
All variables and arrays
in the examples are assumed
to be of
character data type.
Valid
FILE = 'PROG.2'
REVOL(l) = 'MAR'//'CIA'
LOCA(3:8) = 'PLANTS'
TEXT ( I , J +1) (.2: N-1 )

NAME//X

Invalid
'ABC'

CHARS

(the left side must be a character variable,
array element, or substring reference)

CHARS

(expression on the right must be
data type)

8-5

character

CHAPTEn 9
CONTROL STATEMENTS

FORTRAN object programs normally execute statement by statement in the
order
in which they were p~esented to the compiler.
The following
control statements, however, enable you to alter the normal sequence
of statement execution:
1•

CAL L (Sec t ion 1 3 • 4 . 2 • :2)

CONTINUE

3•

DO WHILE

ELSE (Section 9.2.4)

ELSE IF

END (Section 9.8)

END DO

END IF (Section 9.2.4)

(Section 9.5)

( Sec t ion 9. 3 )
(Section 9.3.2)

~~HEN

(Section 9.2.4)

(Sec t ion 9.4)

(SE~ction

10.

GO TO

11.

IF (Section 9.2)

12.

IF THEN

13.

STOP (Section 9.6)

14.

PAUSE

1 5•

RET URN ( Sec t ion 13. 4 • '0

9.1)

(Section 9.2.-1)

(SE~ction

9.7)

The CALL and RETURN statements are described in Sections 13.4.2.2 and
13.4.4,
respectively.
The rernainin9 statements are described in this
chapter.

9.1

GO TO STATEMENTS

A GO TO statement causes the statement that
it identifies
executed next, regardless of i~s position within the program.
Unconditional
There are three kinds of GO TO statements:
9. 1. 1),
Compu ted
(see Sec t ion 9. 1.2) ,
and Assigned
9.1.3).
9-1

(see Section
(see Section

CONTROL STATEMENTS
9.1.1

Unconditional GO TO Statements

An unconditional GO TO statement
specified statement label.

transfers

program

control

the

You can position an unconditional GO TO statement anywhere in
source program, except as the terminating statement of a DO loop.

the

The form of the unconditional GO TO statement is:
GO TO s
where:
s

is a statement label of an executable statement.

For example:
GO TO 300

9.1.2

Computed GO TO Statements

The form of a computed GO TO statement is:
GO TO ( s

[, s] ••• ) [ ,] e

where:
(s[,s] ••• )

is a list of statement labels.

is an integer expression.

You may include any number of statement labels
in the list of a
computed GO TO statement.
However, each statement label must appear
within the program unit containing the GO TO statement.
The same
statement label can appear more than once in the list.
The value of the expression must be an
integer value
(it will be
truncated if necessary) that is greater than 0 and less than or equal
to the number of statement labels given in the list.
If the value of
the expression is not within this range, the next sequential statement
is executed.
When a computed GO TO statement is executed,
the value of the
expression is computed first.
The value of the expression specifies
the position of the label (within the given list of statement labels)
that
identifies the statement to be executed next.
For example, in
the statement sequence:
GO TO ( 2 0, 10 , 5) K
CALL XRANGE(K)
the variable K acts as a switch, causing a transfer to statement 20 if
K=l, to statement 10 if K=2, or to statement 5 if K=3.
The subprogram
XRANGE is called if K is less than 1 or greater than 3.

9-2

CONTROL STATEMENTS
9.1.3

Assigned GO TO Statements

The form of an assigned GO TO statement is:
GO TO i

[[,]

(s [, s] ••• ) ]

where:
i

is a variable name and the optional parenthesized list is a
list of statement labels.
The statement labels specified
must appear within the program unit containing the GO TO
statement.

Assigned GO TO statements must be logically preceded by an ASSIGN
statement
(see Section 8.3) that assigns a statement label value to
the variable i.
The assigned GO TO statement transfers program
control to the label that has been ASSIGNed.
The statement label value assigned must appear within the same program
unit as the GO TO statement that uses that value.
In statements with
a specified list, if i is not assigned one of the statement label
values given in the list, the next sequential statement is executed.
Examples:
ASSIGN 300 TO STAT1
GO TO STATI
GO TO STATl, (177,300,777)

9.2

IF STATEMENTS

There are four kinds of IF statements:
arithmetic
(see Section
9.2.1),
logical
(see Section 9.2.2), logical two-branch (see Section
9.2.3), and block IF (see Section 9.2.4).

9.2.1

Arithmetic IF Statements

The form of the arithmetic IF statement is:
IF

(e)

sl, s2, s3

where:
e

is an expression enclosed within parentheses and sl, s2, and
s3 are statement labels of three executable statements
appearing within the program unit containing
the
IF
statement.
The expressIon e must not be of type complex.
The same statement label can appear more than once in the IF
statement.

This type of IF statement transfers control of the program to one of
the given statements according to the computed value of the given
expression.
If the value of the expression is:
1.

Less than 0, control
identified by label sl.

transferred

the

sta tement

Equal to 0, control
identified by label s2.

transferred

the

statement

9-3

CONTROL STATEMENTS
3.

Grea ter than 0,
con trol
identified by label s3.

transferred

the

sta tE~ment

Examples:
IF(ETA)4,

Transfers control to statement 4 if
ETA is negative, to statement 7 if
ETA is 0, and to statement 12 if
ETA is greater than O.

IF ( KA P PA- L (1 0) ) 2 0 , 14 , 14

9.2.2

Transfers control to statement 20
if KAPPA is less than the 10th
element of array L and to statement
14 if KAPPA is greater than or
equal to the 10th element of array
L.

Logical IF Statements

The form of the logical IF statement is:
IF (e)

where:
e

is any expression.
complex.

The

expression

is an executable statement.

must

not

type

If the value of the expression is true
(negative),
control
is
transferred
to the executable statement within the IF statement.
If
the value of the expression is
false
(nonnegative), control
is
transferred
to
the next sequential executable statement.
The
statement you give in a logical IF statement may be any executable
statement except a DO statement, an END statement, or another logical
IF statement.
Examples:
IF(T.OR.S) X=Y+l

Performs
an
arithmetic
assignment operation if the
result of the IF is true.

IF(Z.GT.X(K»

Performs a subroutine call
if
the result of the IF is true.

CALL SWITCH(S,Y)

IF(K.EQ.INDEX) GO TO 15

Logical Two-Branch IF Statements

9.2.3
'[''lP

Performs
an
unconditional
transfer
if the result of the
IF is true.

format of a logical two-branch IF statement is:
[l-'

( e)

5,1,

9-4

CONT~OL

STATEMENTS

where:
c

is any expression,
and sl and
s2 are statement labels
appearing
within
the program unit containing the
IF
statement.
The expression must not be of type complex.

Logical two-branch IF statemer;ts transfer control to either statement
sl or s2, depending on the corrputed value of the given expression.
If
the value of the given logical .expression is true (negati.ve),
control
is transferred
to statement
51.
If the value of the expression is
false (nonnegative), control is transferred to statement s2.
Examples:
Transfers control to statement
LO
if
LOGl
is
true
(negative) ;
0therwise transfers
control
to
statement 20.

IF (LOG1) 10,20

IF (A.LT.B.AND.A.LT.C)

9.2.4

31,32

Transfers control to statement
if
A is
less
than both B and
transfers control to statement
i f
A is qreater than or equal
I~ i ther B or C.

C;
32
to

Block IF Statements

Block IF statements conditionally execute blocks
statements.
The four block IF statements are:
•

IF THEN

•

ELSE IF THEN

•

ELSE

•

END IF

These statements are used
in block
construct has
the followinQ
form,
statements are optional:
-

(or

groups)

IF constructs.
The block IF
where the ELSE IF THEN and ELSE

IF (e) THEN
block
ELSE IF (e) THEN
block

ELSE
block
END IF
where:
e

block

is a sequence of zero or more complete. FORTRAN statements.
This sequence is called a statement block.

logical expression.

9-5

CONTROL STATEMENTS
Each block
IF statement,
except
the END
IF statement,
has an
associated
statement block.
The statement block consists of all the
statements following the block IF statement up to (but not
including)
the next block IF statement in the block IF construct.
The statement
block is conditionally executed based on
the values of
logical
expressions
in
the preceding block IF statements.
A statement block
can be empty.
The IF THEN statement begins a
block
IF construct.
The block
following it is executed if the value of the logical expression in the
IF THEN statement is true.
The first statement of
the block cannot
directly follow the THEN on the same line.
For example, the following
is ilLegal:
IF (T.LT.X) THEN T

The correct form is:
IF (T.LT.X) THEN
T = X
The ELSE statement specifies a statement block to be executed
if no
preceding statement block in the block IF construct was executed.
The
ELSE statement is optional.
The ELSE IF THEN statement is similar to the ELSE statement, except it
requires an additional
condition
for
execution.
The ELSE IF THEN
statement specifies a statement block to be executed if both the value
of the logical expression in the ELSE IF THEN statement is true, and
no preceding statement block in the block IF construct was executed.
A block
IF construct can contain any number of
ELSE
IF THEN
statements.
The ELSE IF THEN statement is optional.
The END IF statement terminates the block IF construct.
Figure 9-1 describes the flow of control for four examples of block IF
constructs.

9-6

CONTROL STATEMENTS

CONTROL STATEMENTS
Flow of Contml

Construct

IF (e) THEN
block
END IF
Execute
block

IF (e) THEN
block1
ELSE
block2
END IF

IF (e 1) THEN
block 1
ELSE IF (e2) THEN
block2
END IF

IF (e 1) THEN
block1
ELSE IF (e2) THEN
block2
ELSE IF (e3 ) THEN
block3
ELSE
block4
END IF

False

Execute
block1

Execute
block2

Execute
blo,ck2

False

Execute
block1

Execute
block2

Execute
block3

Execute
block4

MR-S-2525-83

Figure 9-1:

Examples of Block IF Constructs

After the last statement in a statement block is executed, control
passes
to the next executable statement following
the END IF
statement. Consequently, at most one statement block in a block IF
construct is executed each time the IF THEN statement is executed.

9-7

CONTROL STATEMENTS
these
ELSE IF THEN and ELSE statements can have statement labels, but
labels cannot be referenced.
The END IF statement can have a
statement label to which control can be transferred,
but only from
within the block IF construct.
Section 9.2.4.1 describes restrictions on statements in a statement
block.
Section 9.2.4.2 describes examples of block IF constructs.
Section 9.2.4.3 describes nested block IF constructs.

can
contain
any
9.2.4.1 Statement Blocks - A statement block
executable FORTRAN statement except an END statement (see Section
statement block, but you
9.8).
You can transfer control out of a
Note that you cannot
cannot transfer control back
into the block.
transfer control from one statement block into another.
DO loops cannot ove~lap statement blocks.
When a statement block
contains a DO statement (see Section 9.3), it must also contain the DO
loop's terminal statement or END DO statement.
Conversely, if a block
IF construct appears within the range of a DO loop, the corresponding
END IF statement must also appear within the range of that DO loop.

9.2.4.2 Block IF Examples - The simplest block IF construct consists
of the
IF THEN and END IF statements; this construct conditionally
executes one statement block.
Form

Example

IF (e) THEN
block
END IF

IF (LOWER.LE.UPPER) THEN
MIDDLE=(LOWER+UPPER)/2
END IF

The statement block consists of all the statements between the IF THEN
and END IF statements.
The IF THEN statement first evaluates the logical expression
(e),
(LOWER.LE.UPPER).
If the value of e is true, the statement block is
executed.
If the value of e is false, control transfers to
the next
executable statement after the END IF statement; the block is not
executed.
The following example contains a block IF construct with
THEN statement:
Form

Example

IF (el) THEN
blockl
ELSE IF (e2) THEN
block2
END IF

IF (ITEM.LT.A(MIDDLE»
THEN
UPPER=MIDDLE-I
ELSE IF (ITEM.GT.A(MIDDLE»
THEN
LOWER=MIDDLE+l
END IF

ELSE

Blockl consists of all statements between the IF THEN and the ELSE IF
THEN statements;
block2 consists of all the statements between the
ELSE IF THEN and the END IF statements.
If ITEM is less than A(MIDDLE), blockl is executed.
If ITEM is not less than A(MIDDLE),
A(MIDDLE), block2 is executed.

9-8

but

ITEM

greater

than

CONTROL STATEMENTS
If ITEM is not less than A(MIDDLE)
and
ITEM is not greater than
A(MIDDLE),
neither blockl nor block2 is executed; control transfers
directly to the next executable statement after the END IF statement.
The following example contain5 a
statement:

block

construct

Form

Example

IF (e) THEN
block1
ELSE
block2

IF (ITEM.GT.A(MIDDLE»
LOWER=MIDDLE+1
ELSE
S8ARCH=MIDDLE
RETURN
END IF

END IF

with

ELSE

THEN

Block1 consists of all the statements between the IF THEN and the ELSE
statements; block2 consists of all the statements between the ELSE and
the END IF statements.
If ITEM is

greatE~r

than A(MIDJ)LE), block1 is executed.

If ITEM is not greater than A(MIDDLE), block2 is executed.

9.2.4.3 Nested Block IF Constructs - A block IF construct can be
included
in a statement bloc~ of another block IF construct.
But the
nested block IF construct m~st be completely contained within a
statement block; it must not ~verlap statement blocks.
The following example contains a nested block IF construct.
Form

Example
FUNCTION SEARCH(A,N,ITEM)
CHARACTER*(*) A(N),ITEM
INTEGER SEARCH,N,LOWER,MIDDLE,UPPER
LO'~ER=l

UPlI?ER=N
IF(LOWER.LE.UPPER) THEN
MIDDLE=(LOWER+UPPER)/2

IF (e) THEN

blockl

IF (e) THEN
10
blocka
ELSE IF (e) THEN
blockb
ELSE
blockc

IF (ITEM.LT.A(MIDDLE»
THEN
UPPER=MIDDLE-1
ELSE IF (ITEM.GT.A(MIDDLE»
THEN
LOWER=MIDDLE+1
ELSE
SEARCH=MIDDLE
RETURN
END IF

END IF

GOTOIO
END IF

END IF
20 SEARCH=O
RE'rURN
END

9-9

CONTROL STATEMENTS
If LOWER is less than or equal to UPPER, blockl is executed.
Block1
contains a nested block IF construct.
If ITEM is less than A(MIDDLE),
blocka is executed.
If ITEM is greater than A(MIDDLE)
blockb is
executed.
If ITEM is equal to A(MIDDLE), blockc is executed.
If LOWER is greater than UPPER, control is transferred
to the first
The nested IF
executable statement after the last END IF statement.
construct is not executed.

9.3

DO STATEMENT

The two types of DO statements are:
1.

Indexed DO

(DO statement)

Pretested indefinite DO (DO WHILE statement)

The indexed DO statement is described in Section
WHILE statement is described in Section 9.3.2.

9.3.1

9.3.1,

,=Fld

the

Indexed DO Statement

DO statements simplify the coding of iterative procedures;
that is,
the statements in the DO statement range are executed repeatedly a
specified number of times.
The form of an indexed DO statement is:
Indexing Parameters

---=--TERMINAL

STATEMENT
LABEL
INDEX
VARIABLE

i = el,

\~:-::-IN~C~R=E~M=E~NT=\
(OPTIONAL)
TERMINAL
PARAMETER
PARAMETER

INITIAL
PARAMETER

Mns1760·S1

where:
s

Terminal statement label s identifies the last statement of
the DO statement range.
The statement must follow the DO
statement in the same program unit.
s
is
omitted,
then
the loop must be terminated by an END DO statement (see
Se c t ion 9. 4) •
The terminal statement can be any executable statement
than one of the following:
•

Unconditional or assigned GO TO statement

•

Arithmetic IF :n logical two-branch IF statemE-?nt

•

Block IF, ELSE IF, ELSE, or END IF statement

9-10

other

CONTROL STATEMENTS
•

RETURN statement

•

STOP statement

•

END statement

•

DO statement

If the terminal statement is a logical IF, it can contain any
executable statement except one of the following:

•

DO statement

•

Block IF, ELSE IF, ELSE, or END IF statement

•

END statement

•

Another logical IF statement

Index variable i is an unsubscripted numeric variable whose
value is defined at the start of the DO statement operations.
The index variable must not be of type complex.
The index variable is available for
use
execution of the range of the DO statement,
value within the DO loop does not change the
the DO loop will execute.
The DO loop index
available for use in the program when:

throughout each
but altering its
number of times
variable is also

Control is transferred outside the range of the DO loop
by a GO TO, IF, cr RETURN statement located within the DO
range

Control is transferred outside the range of the DO loop
by an I/O statement with either or both the options END=
or ERR= (see Chapter 10)

A subprogram is executed from within the DO statement
range having the
index variable as an argument or in
COMMON

Initial parameter el assigns the index variable i its initial
value.
This parameter can be any expression, but cannot be
of type complex.

Terminal parameter e2 provides the value used
to determine
how many repetitions of the DO statement range are performed.
This parameter can be any expression, but cannot be of type
complex.

Increment parameter e3 specifies the value to be added to the
initial parameter (e!1) on completion of each cycle of the DO
loop. The increment parameter is optional.
If e3 and
its
preceding comma are omitted, e3 is assumed to be equal to 1.
This parameter can be any expression, but cannot be of type
complex.

9-11

CONTROL STATEMENTS
9.3.1.1 Executing an Indexed DO Statement - The
indexing parameters
el, e2 or e3 can be any expressions.
Their values are calculated only
once, at the start of each DO loop operation, to determine the values
for
the
initial,
terminal, and increment parameters.
If necessary,
the initial, terminal, and increment parameters are converted,
before
use, to the data type of the index variable.
The number of times that a DO loop will execute, called the
count, is specified by the formula:

iteration

MAX(INT«e2-el+e3)/e3) ,0)
If the iteration count is less than or equal to zero, the body of the
loop is not executed.
The index variable retains its assigned value
(el) •
NOTE
The interpretation of the
iteration count described
above
is different
from that of earlier versions of
FORTRAN-I0/20.
If
the /NOF77 compiler switch
is
specified
(see Sections 16.1.3 or 16.2.3), and the
iteration count is less than or equal
to
zero,
the
body of the loop is executed once.
Since the iteration count is computed at
the start of a
DO loop
operation, changing
the value of the loop index variable within the
loop cannot affect the number of times that the loop is executed.
At the start of a DO loop operation, the index value
is set to
the
value of the
initial parameter
(el);
and
the iteration count is
established.

9.3.1.2 DO Iteration Control - At the end of each DO loop cycle,
following steps are executed:
is

(e3)

added

the
the

The value of the increment parameter
index variable.

The iteration count is decremented.

If the
iteration count
is greater
than
zero,
control
transfers
to
the
first executable statement after the DO
statement for another iteration of the loop.

If the iteration count
is
less
than
execution of the DO loop terminates.

equal

zero,

Exit from a DO loop upon completion of the number of
iterations
specified by the loop count is referred to as a normal exit.
If no
other DO loop shares the
terminal
statement,
or
if this DO loop
statement
is outermost,
control passes
to the
first
executable
statement after the terminal statement of the DO loop.
The final value of the index variable is the value determined by
1.

9-12

step

CONTROL STATEMENTS
NOTE:
The interpretation of the
index variable described
above
is different
from that of earlier versions of
FORTRAN-10/20.
! f
tile
/NO~"77
compiler
switch
is
specified
(see Sections 16.1.3 or 16.2.3), the final
value of the index var:.able of the DO statement
is
undefined after a normal loop exit.
8xit from a DO loop may also bE! accomplished by a transfer of control
by a
statement within
the DO loop range to a statement outside the
range of the DO statement.
Th~.s is called an extended range
DO loop
(see Section 9.3.5).
When execution of a DO loop terminates, and other DO loops share its
terminal
statement,
control
transfers outward
to
the next most
enclosing DO loop in the DO nesting structure (see Section 9.3.4).
Examples of DO Iteration Control:
100

DO 100 I = 1,10
J=I

After execution of these statements, 1=11 and J=lO.
switch is specified, I is undefined and J=lO).

200

L=O
DO 200 K
L=K

the

/NOF77

(If

the

/NOF77

5,1

After execution of these statements, K=5 and
switch is specified, K is undefined and L=5).

9.3.2

(I f

L=O.

DO WHILE Statement

The DO WHILE statement is simiJ.ar to the DO statement described
in
Section 9.3.1.
Instead of executing a fixed number of iterations, the
DO WHILE statement. executes
for
as
long as a
logical expression
contained in the statement continues to be true.
The form of the DO WHILE statement is:
DO

[s[,]] WHILE

(e)

where:
is the label of an executable statement that must physically
follow in the same program unit.

is a logical express:on.
'rhe DO WHILE statement tests the logical expression at
the
beginning
of each execution of the loop, including the first.
If the value of
the expression is true, the statements in the body of the
loop are
executed;
if
the expression
is
false,
control
transfers
to the
statement followinq the loop.
If no Label appears in a DO WH:LE statement, the DO WHILE loop must be
terminated with an END DO statement (see Section 9.4).

9-13

CONTROL STATEMENTS
The following example demonstrates the use of the DO WHILE statement:
CHARACTER*132 LINE
1=1
LINE (132:) = IX I
DO WHILE (LINE(I:I)
I = I + 1
END DO

9.3.3

.EQ.

The Range of a DO Statement

The range of a DO statement is defined as the series of statements
that follows the DO statement,
up to and including the specified
terminal statement or END DO statement.
If another DO statement appears within the range of a DO statement,
the range of that statement must be entirely contained within the
range of the first DO statement. More than one DO statement may have
the
same labeled terminal statement but not unlabeled END DO
statement.
(See Section 9.3.4, Nested DO Statements.)
If a DO statement appears within an IF block, ELSE IF block, or ELSE
block
(see Section 9.2.4),
the range of the DO statement must be
contained entirely within that block.
If a block IF statement appears within the range of a DO statement,
the corresponding END IF statement must also appear within the range
of the DO statement.

9.3.4

Nested DO Statements

One or more DO statements can be contained within the range of another
DO statement.
This is called nesting.
The following rules govern the
nesting of DO statements:
1.

The number of nested levels
restricted to 79 DO loops.

(DO

loop

The range of each nested DO statement must be entirely within
the range of the containing DO statement (such as, they
cannot overlap).
For example:
Valid

Invalid

DO 1

DO 2

~
c=
C

DO 3

9-14

The ranges of
loop DO 2 and
DO 3 overlap
MR·S·1758·81

within

loop)

CONTROL STATEMENTS
3.

More than one DO loop within a nest of labeled DO loops can
end on the same statement. When this occurs, the terminal
statement is considered
to belong to the innermost DO
statement that ends on that statement. Only a statement that
occurs within the range of the innermost DO statement can use
the statement label of the shared terminal statement for
transfer of control.
For example:
DO 4

- - - ..- Outermost DO Loop

004 _ _

00_4_ _

[I~..- Innermost DO Loop
- - - . . - Terminal Statement
MR-S-1759-81

Although all four DO loops share the same terminal statement,
the terminal statement "belongs" to the innermost DO loop.
4.

Nested loops cannot share an unlabeled END DO statement.
Each unlabeled END DO terminates exactly one DO loop.
For example:
Correctly Nested
DO Loops

Incorrectly Nested
DO Loops

DO 10 1= 1,20

DO 10 1=1,5

DO J=1,!i

DO J= 1,10

DO K=1,10

CONTINUE

END DO

[

END DO
10

9.3.5

CONTINUE

Extended Range

By following certain rules, it is possible to transfer out of a DO
loop,
perform a series of statements elsewhere in the program, and
then transfer back into the DO loop.
The statements that are executed
after a transfer out of a DO loop and before a transfer back into the
same DO loop are collectively known as the "extended range." A DO loop
that permits transfer
in and out of its range is called an extended
range DO loop.

9-15

CONTROL STATEMENTS
NOTE
;'his feature makes the flow of a program difficult
to
follow,
does
not conform to the FORTRAN-77 standard,
;Ind is therefore discouraged.
"'he following rules govern the use of extended range DO loops:
The statement that causes the transfer out of
the DO
loop
'oust be contained within the most deeply nested DO (innermost
loop having the same terminal
statement).
This
loop must
~lso
contain
the
statement
to WhICh
the extended range
r-eturns.

A transfer into the ranqe of a DO statement is permitted only
if
the
transfer
is made from the extended range of that DO
~; ta temen t .
The extended range of a DO statement must not contain another
s ta temen t .

)")0

The extended range of a DO statement cannot change the
v~riable or indexing parameters of the DO statement.

You can call a subprogram within an extended range.

The following example illustrates the use
loop:

extended

index

range

DIMENSION TABLE(10,5), VALUE(10)
LOGICAL LOGARR(lO)
1)0 1000 I = 1, 10
An extended range DO loop
If (LOGARR(I»
GOTO 500
Test logical array item
I=K
CALL SUBROT(K)
Invoke subroutine using
current index value
DO 2 () 0 J = 1, 5
Nonextended range loop
TABLE(I,J)
0
CONTINUE
:.;00

GOTO 2000
VALUE(I)

)000

CONTINUE

:: 000

2100

Extended range invocation
Invoke function GETVAL with
current index
Terminal statement for outer
loop

GETVAL(K)

STOP
TYPE 2100, I
FORMAT (' I = " I 2)
LOGARR(I) = .TRUE.
GOTO 500

Extended range starts
Extended range ends and
r.::eturns

END

9.3.6

Permitted Transfer Operations

The following rules govern the transfer of program control from within
a DO statement range or the ranges of nested DO statements:
1.

A transfer out of the range of any DO loop
is permitted
at
any
time.
When
such a
transfer occurs, the value of the
controlling DO
loop's
index
variable
is defined as
the
current value.
9-16

CONTROL STATEMENTS
2.

A transfE~r into the r,3.nge of a DO statement is permitted
it is made from the ectended range of the DO statement.

You can call a subprogram from within the range of any:
a.
b.
c.

DO loop
nested DO loop
extended ranqe 101)p (in which you leave the loop through
a GO TO,
execu~e statements in the extended range, and
return to the oriqinal loop)

The following examples illustrate the transfer
from within the ranges of nested DO statements:

Valid
Transfers

Invalid
Transfers

operations

permitted

~"""I--_e_xt_ended~rangle
MR-S-17,7-81

9.4

END DO STATEMENT

The END DO statemE'nt terminatE!s the range of a
DO or DO WHILE
statement.
The END DO statemE~nt must be used to terminate a DO block
if the DO or DO WIULE statement defining the block does not contain a
terminal-statement
label~
The END DO statement may also be used as a
labeled terminal statement if the DO or DO WHILE statement does
contain a terminal-statement label.
The form of the END DO statemerlt is:
END DO

9.5

CONTINUE STATEMENT

The form of the CONTINUE statement is:
CONTINUE
Execution of the CONTINUE statE~ment has no effect.
the terminating statement of a DO loop.

9-17

It may be used

CONTROL STATEMENTS
In the following example, the labeled CONTINUE statement
legal termination for the range of the DO loop.

provides

DIMENSION STOCK(lOO)
DO 20 1=1, 100
STOCK(I)=O
CALL UPDATE (STOCK(I»
IF(STOCK (I) .EQ. 0) GO TO 30
20 CONTINUE
STOP
30 TYPE 35
35 FORMAT (' UPDATE ERROR')
END

9.6

STOP STATEMENT

Execution of the STOP statement causes program execution to be
terminated.
A descriptive message may optionally be included in the
STOP statement to be output to your
terminal
immediately b«~fore
program execution is terminated.
The form of the STOP statement is:
STOP [n]
where:
n

is an optional decimal integer constant of up to 6 digits,
or a character constant.
The constant is printed at the
terminal when the STOP statement is executed.
You can have any number of characters
in the character
constant.
You can use continuation lines to accommodate
large character strings.
The constant is printed without
leading zeroes, unless they are specified in the statement.
NOTE
The word STOP is not printed when the STOP statement
is executed unless the word STOP is included in the
statement as a character constant.

The following examples show the results of executing STOP statements
that contain a 6-digit decimal string and a character constant.
10

PROGRAM TEST
STOP 123456
END

EXECUTE STOP1.FOR
FORTRAN: STOPl
TEST
LINK:
Loading
[LNKXCT TEST execution]
123456
CPU time 0.1
Elapsed time 0.3

9-18

CONT~OL

STATEMENTS

PROGRAM TEST
STOP 'The prograrr has stopped'
END

EXECUTE STOP2.FOR
FORTRAN: STOP2
TEST
LINK:
Loading
[LNKXCT TEST execution]
The program has stopped
CPU time 0.1
Elapsed time 0.3

9.7

PAUSE STATEMENT

Execution of a PAUSE statement suspends the execution of the object
program and gives you the option of continuing execution of the
program, exiting from the program, or beginning a TRACE operation.
The form of the PAUSE statement is:
PAUSE [n]
where:
n

is an optional integer constant of up to 6 digits,
or a
character constant.
The constant is printed at the terminal
when the PAUSE statement is executed.
You can have any number of characters in the character
constant.
You can use continuation lines to accommodate
large character strings.
The constant is printed without
leading zeros, unless they are specified in the statement.

If execution of the program is resumed after a PAUSE, program control
continues as if a CONTINUE had been executed.
Execution of the PAUSE
statement causes the word PAUSE, the optionally specified constant,
and the following prompt to be printed at the terminal:
Type G to Continue, X to Exit, T to Trace
The responses to this prompt are:
G

continues program execution at
following the PAUSE statement.

causes program termination.

produces a trace back list at the terminal.
This list
consists of invoked routine names and locations, plus the
location and module names of the callers of those routines.
Using this
information you can track the active path of
execution from the main program to the PAUSE trace routine.
(See Section 13.4.1.32 for a detailed description of this
feature.)

9-19

the

statement

immediately

CONTROL STATEMENTS
PROGRAM PTEST
PAUSE
PAUSE 234
PAUSE 'Character String'
END
EXECUTE PTEST.FOR
FORTRAN: PTEST
PTEST
LINK: Loading
[LNKXCT PTEST execution]
PAUSE
Type G to Continue, X to Exit, T to Trace.
G

PAUSE
234
Type G to Continue, X to Exit, T to Trace.
G

PAUSE
Character String
Type G to Continue, X to Exit, T to Trace.
X

CPU time 0.3

9.8

Elapsed time 18.8

END STATEMENT

This statement signals FORTRAN that the physical end of a program unit
has been reached.
END is an executable statement.
The general form
of an END statement is:
END
The following rules govern the use of the END statement:
1.

This statement must be the last physical statement
source program unit (main program or subprogram).

When executed in a main program, the END statement has the
effect of a STOP statement;
in a subprogram, END has the
effect of a RETURN statement.

An END statement may be labeled, but it must not be continued
(that is, it must appear only on an initial line).

9-20

CHAPTER 10
DATA TRANSFER STATEMENTS

FORTRAN I/O statements are divided into three categories by
as follows:

function,

Data Transfer Statements
transfer data between memory and
files.
The "files" can be devices such as TTY: or MTA:.
Internal files and ENCODE/DECODE statements are used for
memory-to-memory data transfers.

File Control Statements
associate and disassociate files
and
FORTRAN
logical
unit
numbers,
and can specify
characteristics of such an association.

Device Control Statements
position files.
For example,
using the device control statements you can position magnetic
tape to a particular file or record.

This chapter describes data transfer statements.
file-control and device-control statements.

Chapter 11 describes

Table 10-1 lists the three categories of I/O statements,
the
statements within each category, and the sections in which each I/O
statement is further described.

10-1

DATA TRANSFER STATEMENTS

Table 10-1:

FORTRAN 110 Statement Categories

Categories

Statements

Sections

Data Transfer

READ
WRITE
REREAD
ACCEIYf
TYPE
PRINT
PUNCH
ENCODE
DECODE
Internal READ
Internal WRITE
OPEN
CLOSE
INQUIRE
FIND
REWIND
UNLOAD
BACKSPACE
ENDFILE
SKIPRECORD
SKIPFILE
BACKFILE

10.5
10.6
10.7

File Control

Device Control

Table 10-2, on the
statement forms.

tab-divider,

10.8
10.9

10.10
10.11
10.12
10.12
10.12
10.12
11.2
11.4
11.7
11.8.1
11.8.2
11.8.3
11.8.4
11.8.5
11.8.6
11.8.7
11.8.8

summarizes

10-2

all

the

data

transfer

DATA TRANSFEd STATEMENTS

Table 10-2:

Summary of Data Transfer Statement Forms

Data Access

Statement Construct

Section

Sequential Formatted
(FORMAT Statement)

READ(UNIT un,FMT fI ,END = sll ,ERR= sll ,IOSTAT- iosilliolistl
READ(
un,FMT -- fI ,END - s II ,ERR = s II ,IOSTAT = ios III iolist I
READ(
un.
fI ,END - sII,ERR= sll ,JOSTAT = ioslli iolistl
READ fI ,iolist I
READ(UNIT-*,FMTo /l,END-sll,ERR -sll,IOSTATc. iosilliolistl

10.5.1.1

WRITE(UNIT, un,FMT -fi.ERR-sll,IOSTAT=iosDliolistl
WRln.:(
un,FMT - flERR - sll ,lOSTAT = ioslHiolistl
WRITE(
un,
fI ERR-slI,lOSTATc iosilliolistl

10.6.1.1

W({ITI': f1.iolist I

WRITE (UNIT - *,FMT - fi ,FRR - sll ,IOSTAT ~ iosPliolistl
1<i<:I{I':'\I)IF1\l'J' t[.l-:NJ)
In':I{EAJ) {'1.lolist I

sl.EHf{

sll,lOSTAT

ioslJliolistl

;iJ.i.]

\(TI',I'TIFMT f'lYNJ)
\('( '1':\ ''I' f'I.iolisll

sl.EH,I{

sllJOSTAT

iosiJiiolistl

; iI.h.1

TYI'Eil·iV1T· n.Elm
TYPE f'I.iolis! I

"II.J(JSTA'!'

ioe,llliolistl

I'I.EIn{

,.:II.H)STAT

iosliJiolistl

sll.IUS,\,AT

io"liJiolis!1

II{IN'J'!VMT

III i

PRINT fI ,iolist I
l'l!:\'(,IIIFMT fl.J-:I{H
j'llNCII l'I.iollstl
1';~(:OJ)E(l'.f'.al.EI{f{

1)/':<'( )DEIl'.Lal.l~KJ-{

Sequential Formatted
(List Directed)

.-;II.J(S'I'AT
,.: II.JUSTAT

I'.) ] I. J

ioslJiiolistl
ios ill iolis! I

III

READ(UNIT - un,FMT ~- *I,END - sll,ERR = sll ,lOSTAT = iosilliolistl
READ(
un,FMT - *I.END - sll ,ERR = sll ,lOSTAT= iosDliolistl
un.
*1.END~sll,ERR=sll,lOSTAT=ioslliiolistl
READ(
READ *1 ,iolist I
READ(uNIT- *,FMT ~ *I,END sll,ERR-sll,IOSTAT-iosIHiolistl

10.5.1.:3

WRlTE(uNJT - un,FMT = *1 ,ERR - s II ,lOST AT = ios DI iolist I
un,FMT- *1,ERR~sll,IOSTAT=iosDliolistl
WRI'I'E(
WRITE(
un,
~I ,ERR = sll ,lOSTAT= ioslJl iolistl

10.6.1.3

\\'I{ITE 'I.ioli,,! I

WRITE(UNIT = ';',FMT - *1 ,ERR- sll ,lOSTAT = ioslHiolistl
i{I·:I<I-:ADIF1\1T'I.i<:l\'[)
I~ F H I':A D"I.IOII:"t I

~ II.Elm

,.:I.JUSTAT

iosliJlOiI,.:t I

\('('!':I'TI1'1\l'I' 'I.END
\(.(' 1·:I>'I"I.iolist I

o-II.EHI\

,,11.lOSTAT

io"liJio]i"tl

TYPI'>v:vn
I.EHH
TY 1'1':'I.iolist I

,,11.l1)STAT

Im;lllioli"tl

10.9.2

'UINTIFMT

sll.OS'I'AT

lOslljiolist I

10.10.2

s: .IUS'I'AT

io:"l!liolistl

'I.":HH

:(i 1-' ~

PRINT *1 ,iolist I
!'l'NCIIIFMT 'IYIW
I'll'~ (' III. i III i st I

10-3

1('

11.2

DATA TRANSFER STATEMENTS
Table 10--2:

Summary of Data Transfer Statement Forms (ConLl

Data Access

Statement Construct

~;('qu('nti<ll Formatted
,:\Al\H:J.lST Statement!

/{EA /)1 I J N IT
HEAD( l.TNIT

/IUi./.4

]{E.\/)(

un.FMT lI<lnwl.END sll.Elm· sll.IOSTAT iosll
\lfl.:\':\1 L namP! .END ;.;11.Elm· s IIJOSTAT - ins il
llllYi\1T nanH'I.E]\;J) sHElm sll.lOSTAT - ioslJ
IITl.NML nall1l'l./<:ND sll.ERR s II.JOSTAT io;.; II
1111.
nanwl.END sll.Elm ;.;II.IOSTAT iosll

WI{ITEJ 11NI'I'
WHITEIUNJT
WHITEI
WHITEr
\VHJTEI

Illl.F!\1T 1Ii1111('I.EHH sll.IOSTAT insll
I1I1.Ni\lL 1l<lIll<'l.EHH s 1I.Ji )STAT iosll
lIrlYMT nanwl.EJ{H ;.;II.l0STAT io;.;IJ
I1I1.NML naml'I.EJm ;.;II.H)STAT insll
nallll'l.EJU{ sll./OSTAT illsl J
lin.

/IUi /.1

I\E.\/)(

I{EADI

Sequential Unformatted

Direct Formatted

READ( UNIT - unl ,END ~ s II ,ERR ~ s II ,IOSTAT = ios III iolist I
READ(
unl,END = sll ,ERR - sll JOSTAT- ioslliiolistl

10.5.2.1

WRITE (UNIT = unl ,ERR = sll ,IOSTAT- ioslll iolistl
WHITE(
unl,ERR -sll,IOSTAT~iosllliolistl

10.6.2.1

HEAD(UNIT- un,FMT ~ f,REC - rnl,ERR sll,lOSTAT~ ioslliiolistl
READ(
un.FMT = f,REC - rnl ,ERR - sll ,IOSTAT~ ioslHiolistl
READ(
un,
f,HEC-rnl,ERR- sll,IOSTAT~iosllliolistl
IU:ADI
un'rn.F\rr I'
I.EJu{ ;.;II.IOS'I'AT iosillinlisli
HEADI
11111'11.
r
I,EHH sll.lOSTAT iosliliolisll

10.5.1.2

WRITE(UNIT~un,FMT

10,6.1.2

WRITE(
WHITE(
\VJ{IT/':I
\rJ{]TEI
Direct Unformatted

Section

f,REC rnl,EHR-sll,IOSTAT
un,FMT ~ f,HEC - rnl ,ERH - sll,IOSTAT
un,
f,REC=rnl,ERR sll,IOSTAT
un'rn.FMT f
I.EJm ,1i.IOSTAT
1I1l'1'll,
r
I,Elm 'ill.lOSTAT

iosilliolistl
ioslliiolistl
iosilliolistl
ins III inlist I
io;.;ll/iolist I

READ(UNIT - un,REC ~ rnl ,ERH ,'co sll ,lOSTAT - ioslliiolistl
READ(
un,REC - rnl ,ERR - s II ,IOSTAT = ios III iolist I
J{EADI
un'rn
I.Elm sll.IOSTAT io;.;ll/iolistl

10.5.2.2

un,REC - rnl ,ERR = sl/ ,IOSTAT= ioslliiolistl
un,REC ~ rnl ,ERR - sll ,lOSTAT~ ioslliiolistl
IIn'rn
1.f<:HH ;.;II.10STAT iosilliolistl

10.6.2.2

WRITE (UNIT
WRITE(

Key:
UNIT- un

is a FORTHAN logical unit number or internal file specifier (Section 10.4,3).

UNIT-*

is a default unit specification used with the READ Statement to read from CDR, and with
the WHITE Statement to write to LPT: (see Section 10.4.3).

REC~rn

is a direct-access record number (Section 10.4.4).

IIIl

an all(,1'nat(' W<l\' of specifying Logical Unit NumiJl'r and n'cord numhc'r of a riirectI (j ,-l.·1 I.

~1c('('"S transfer! SI'ct Ion

FMT~f

is FORMAT-statement formatting; iolist is optional (Section 1O.4.5.ll.

FMT=*

is list-directed formatting; iolist is optional (Section 10.4.5.2).

F!\IT

nan\('

I\' \1 I. .

is th(' ;1I1('rnal in' form of Ihl' NA:\1FLlST st;tll'nwnt f;1I'mal ;';J)('cifi!'1' ISection 1().·L'J,:~1

END"'-'s

is an optional end-of-file transfer specifier (Section 10.4.6),

ERR- s

is an optional error transfer specifier (Section 10.4.71.

IOSTAT= ios

is an optional 1/0 status specifier (Section lO.4.S).

iolist

is a data transfer 110 list ISection 10.4.9).

10-4

DATA TRANSFER STATEMENTS
10.1

DATA TRANSFER OPERATIONS

Data transfer statements are used to transfer data between memory and
files
or between memory and memory.
Data can be
transferred
sequentially (sequential access) or
randomly
(direct access).
The
areas
in memory from which data is to be taken during output (write)
operations,
and
into which data
is stored during
input
(read)
operations are specified by:
1.

A list in the data transfer statement

A list defined by a

FORMAT specifications
statement

~AMELIST

statement

referenced

the

The appearance and arrangement of transferred data
by:

data

transfer

specified

can

Format specifications located in either a FORMAT statement or
an array (FORMAT-statement I/O)

The contents of an I/O list

An I/O list defined

(list-directed I/O)

in a NAMELIST statement

(NAMELIST I/O)

These three methods are known collectively as formatted I/O.
In contrast to formatted I/O transfers, FORTRAN has
several methods
for
transferring data without regard for the type and arrangement of
the data being transferred.
These methods are known collectively as
unformatted
I/O.
Unformatted
I/O transfers are particularly useful
when you want the internal (memory) representation of the data being
transferred
to be
the same as the external (file) representation of
the data.
In addition, unformatted data
transfers are generally faster
than
formatted
transfers.
This
is because unformatted data transfers do
not convert the data to or from its ASCII
representation during
the
transfer.
The following sections describe the types of access available,
the
types of data
transfers available, and the statements used for I/O
transfer operations.

10.2

DATA ACCESS

There are two forms of access available
sequential
These forms are described in the following sections.

10.2.1

and

direct.

Sequential Access

If the data access is sequential, the data records are transferred
in
a
serial
fashion
to or
from
the external data
file.
Each
sequential-access input stateMent transfers the next record(s)
from
the accessed data file, such that data records are transferred in the
same order that they appear in the file.

10--5

DATA TRANSFER STATEMENTS
10.2.2

Direct Access

If the data access is direct, the data records are transferred
to or
from a
file in any desired order, as specified by a record number in
the data transfer statement.
(Section
10.4.4 describes specifying
records in data transfer statements.)
Direct-access transfers, however, can be made only to
files
residing
on disk that have been previously set up (using an OPEN statement) for
direct access.
Direct-access files must contain
identically sized
records that are accessed by a record number.
You must use the OPEN statement to establish direct access
(see
Section 11.2).
Execution of the OPEN statement must precede the first
data transfer statement for the specified logical unit.

10.3

FORMATTED AND UNFORMATTED DATA TRANSFERS

The term "formatted data transfer" describes an intermediate step that
occurs during a data transfer.
This intermediate step, which does not
occur in an unformatted data transfer,
converts the data
from
its
internal
(memory)
representation
to a different external
(file)
representation.
(Formatted data transfers are described
in Section
10.3.1.)
An unformatted data transfer refers to the transfer of data
with no
change to the data during
the
transfer.
In an unformatted data
transfer, the internal (memory) representation of the data and
the
external (file) representation of the data are the same.
(Unformatted
data transfers are described in Section 10.3.2.)

10.3.1

Formatted Data Transfers

In a formatted data transfer, the internal and external format of
data is controlled during the data transfer in one of three ways:
1.

the

FORMAT-Statement Formatting
The data
transfer
statement
contains
a
statement number,
a
numeric array name,
a
character expression,
or an
integer,
real,
or
logical
variable as a format identifier.
The statement number references a line that contains a FORMAT
statement.
The array name references an array that contains
a
format
specification.
The value
of
the
character
expression
is a format specification.
The integer, real, or
logical variable references a FORMAT statement number
that
was assigned with an ASSIGN statement.
In
the
following
example,
the data
transfer
statement
contains a
statement number of a FORMAT statement.
The
FORMAT statement, in turn,
contains edit descriptors that
control the formatting of the data during the transfer:
WR I T E

101

(2 2 , 1 01 ) X, J , Z

FORMA T ( 1 X, FlO. 5, I 5 , F 6 . 4 )

See Section 10.4.5.1 for more information on FORMAT-statement
formatting.

10-6

DATA TRANSFER STATEMENTS
2.

List-Directed Format~ing
The data
transfer
statement
contains an asterisk as thE~ format identifier.
The asterisk
signifies that the transfer is controlled by the data type of
the variables in the data transfer statement I/O list.
In the following example, the data transfer is controlled
the I/O list items X, J, and Z:
WR I ~~E

( 2 2, * ) X, J , Z

In this example, unless the data types of X, J,
and Z have
been set explicitly to a type other than the default data
type,
the transferred values of X and
Z
appear
in
floating-point
form,
and the transferred value of J appears
in integE~r form.
See Section 10.4.5.2 for more
formatting.
3.

information

list-directed

NAMELIST-Statement Formatting
The data transfer statement
contains a
NAMELIST name as the format identifier.
This
NAMELIST name associates the data transfer statement with a
NAMELIST I/O list defined in the NAMELIST statement elsewhere
in the same program unit.
Elements in the NAMELIST I/O list,
in turn. dictate th~ formatting of the data during the data
transfer.
In the following exam9le, the data transfer is controlled
the NAMELIST.

PROGRAM NAMLS'r
NAMELIST/VAR/X,Y,Z
READ (22, VAR)
WRI'[,E (5,VAR)
END
See Section 10.4.5.3 fornore information
formatting.

NAMELIST-statement

10.3.1.1 Internal Files - Internal files provide the capability to
perform formatted data transfers between character variables and the
elements of an I/O list.
T~eir use with formatted sequential READ and
WRITE statements reduces the need
to use the ENCODE and DECODE
statements for internal 1/0 (s~e Section 10.12).
An internal file consists of a character variable, a character array
element, a character array, or a character substring; a record In an
internal file consists of any of the above except a character array.
If an internal
file
is a character variable, array element,
or
substring,
that file comprises a single record whose length is the
same as the length of the variable, array element, or substring.
If an internal file is a character array,
that file comprises a
sequence of records, with each record consisting of a single array
element.
The sequence of records in an internal file is determined by
the order of subscript progression (see Section 4.3.2).
Every record
of the file has the same length, which
is the length of an array
element in the array.

10-7

DATA TRANSFER STATEMENTS
The character variable, array element, or substring that is the record
of the
internal
file becomes defined by writing the record.
If the
number of characters written in a record is less than the length of
the record, the remaining portion of the record is left-justified and
filled with blanks.
A record in an internal
file can be read only if the character
variable,
array element, or substring comprising the record has been
defined (that is, a value has been assigned to the record).
Prior to
data
transfer, an internal file is always positioned at the beginning
of the first record.

10.3.2

Unformatted Data Transfers

Unformatted data is transferred in two forms on TOPS-20
(BINARY
or
IMAGE), and
three
forms on TOPS-IO (BINARY, IMAGE, or DUMP).
In an
explicit OPEN statement (Section 11.2.1), you can specify one of these
forms as an argument
to
the MODE specifier.
(Section 11.3.20
describes the MODE specifier and its arguments.)
On disk devices and CORE-DUMP tapes,
numeric data
items
are
transferred directly as
36-bit words.
Character data
items are
t~ansferred as 7-bit
bytes.
Numeric and character
items can be
interpersed
in the same I/O list.
Nu~eric data items and, for BINARY
files, record markers (LSCWs) are always word-aligned
(see Section
18.4.2).
On
INDUSTRY tapes, numeric data items should not be used.
Character data items are transferred one character per
frame
(see
Section 11.3.30).

10.3.3

Unformatted Data Transfer to ASCII Devices

Unformatted data transfer can be done to and from ASCII devices
(such
as line printer, plotter, or terminal).
Character data is transferred
exactly as it appears in the input/output list, with no formatting
or
carriage control.
The method for transferring numeric data items depends on the device.
For non-terminal devices (such as, line printer or plotter), numeric
data
is treated
as
if
it
were
packed
(Hollerith)
data,
left-justifited, five characters per word.
For the terminal, the data
is treated as if it were right-justified, one character per word.

10.4

DATA TRANSFER STATEMENT FORMS

Table 10-2, on the tab divider,
summarizes the forms of all
the
FORTRAN data
transfer statements.
Figure 10-1 shows the three major
components of data transfer statements.
Statement
Name

(Control-Information List)

I/O List

'-----"

I'---..r---J

(See Section 10.4.1)

(See Section 10.4.9)

(See Section 10.4.2 - 10.4.8)
MF~·

Figure 10-1:

s· 1750-8i

Components of Data Transfer Statements

10-8

DATA TRANSFER STATEMENTS
10.4.1

Data Transfer Statement Names

In a data transfer statement, the statement name indicates whether the
operation is an input (read) or output (write) operation.
The FORTRAN data transfer statements described in this chapter are:
1.

READ (See Section 10.5)

WRITE (See Section 10.6)

REREAD (See Section 10.7)

ACCEPT ( Sc!e Section 10.8)

TYPE (See Section 10.9)

PRINT (See Section 10.10)

PUNCH (Se(;:! Section 10.ll)

ENCODE

DECODE (See Section 10.12)

10.

Internal READ (See Section 10.12)

11.

Internal 1(vRITE (See Section 10.12)

10.4.2

(S(?e Section 10.12)

Data Transfer Control-Information List

A control-information list is
included
in every data
transfer
statement.
Each control-information list (including those having an
implicit definition of device) can contain:
1.

One unit specifier (see Section 10.4.3)

One forma t specifier (see Section 10.4.5)

One record specifier (see Section 10.4.4)

One I/O status specifier (se1e Section 10.4.8)

One error specifier (see Section 10.4.7)

One end-of-file specifier (see Section 10.4.6)

The following rules govern the placement and inclusion of items
control-information list:
1.

If the keyword UNIT= is omitted from the unit specifier,
unit
specifier
must
be
the
first
item
in
control-information list.

10-9

a
the
the

DATA TRANSFER STATEMENTS
2.

If the control-information list contains a
format
specifier
(FMT= or
NML=),
the statement is a formatted data transfer
statement.
Otherwise, it is an
unformatted data
transfer
statement.
The NML= keyword is used for NAMELIST formatting
only, although you can also use the FMT= keyword for NAMELIST
formatting.
If the keywords FMT= or
NML=
is omitted
from
the
format
specifier,
the
format
specifier must be the second item in
the control-information list, and the first item must be the
unit specifier without the keyword UNIT=.

If the control-information list contains a
record
specifier
(REC=),
the statement
is a direct-access data
transfer
statement.
Otherwise,
it
is a
sequential-access
data
transfer statement.
If the keyword REC= is omitted from the record specifier, the
unit
specifier (without the keyword UNIT=) must appear first
in the control-information list, followed by a
single quote
( ' ) ! and then the record specifier.

A control-information list cannot contain
specifier and an end-of-file specifier.

If the format specifier is an asterisk or a NAMELIST name,
a
record
specifier
must
not
Je
included
in
the
~ontrol-information list.

A control-information list in an internal file or
ENCODE or
CECODE statement must contain a format specifier other than
dn ~sterisk or NAMELIST name, and must not contain
a
record
';0ecifier.

10.4.3

both

record

Unit References in Data Transfer Statements

The unit specifier is used to refer to a file or device.
a unit specifier is:
UNIT

The form

where:
un

is a logical unit identifier or an internal file identifier.

A logical unit identifier (see Section 10.4.3.1) is used to
refer
to
an external file.
An internal file identifier (see Section 10.4.3.2)
is used to refer to an internal file.
The keyword UNIT= is optional if the unit specifier is the first
in the control-information list.

10-10

item

DATA TRANSFER STATEMENTS
10.4.3.1 FORTRAN Logical Unit Identifier - The FORTRAN logical unit
identifier
is associated with the file to or from which data is being
transferred.
This identifier is an integer expression whose value
is
in the range of 0 to 99, or an asterisk.
For example, the following WRITE statement contains the
reference to
logical
unit number
22 as the first item in the control-information
list:
WRITE

(22,101)

Note
Table 10-3 lists the default logical unit number assignments.
that
22
logical
unit number
identifies the file as DSK:FOR22.DAT.
The unit
Thus,
the sample WRITE statement references a disk.
for the READ
identifier asterisk corresponds
the card
reader
to
statement, and to the line printer for the WRI'rE statement.
The compiler automatically assigns default logical
unit numbers
for
the REREAD,
READ,
ACCEPT, PRINT, PUNCH, TYPE, and WRITE statements.
Default unit numbers are negative integers that cannot be accessed.
For example:
where

OPEN(UNIT=n) or READ/WRITE (UNIT=n)
integer is illegal.

Assigning a negative decimal number to a
level is illegal.

device

negative

command

assign a
You can,
however,
from monitor command
level,
default device to another device.
For example, using the
TOPS-20 DEFINE command (or TOPS-IO ASSIGN command),
you can
assign LPT:
(line printer) to DSK:
(disk).
If you do this,
then any I/O statements
that reference the
line printer
actually reference the disk.
You can optionally make the logical device assignments at runtime,
or
you can use the default assignments contained by the FORTRAN Object
Time System (FOROTS).
Table 10-3 lists
the default logical device
assignments.
You should specify the device explicitly in an OPEN
statement (see Section 11.2) if you wish
to override the default
assignment.

10-11

DATA TRANSFER STATEMENTS
l allll' ill-:{:

FOHTHAN Logical Device Assignments

I)('fault Devices (inaccessible to til£' user)

--! I)~-ia~I~FiI;~~J~l~--------l- i~~);i~~ll;~it'-N-~~~I;~;- -- --.F()J{PI.T.DAT
Fi Ie last read

,'j H{

HH{( 'UR.nAT

F( )l{TTY.1 )AT
FOI{IJYr. [)AT

FOl{PTP.UAT
FORTTY ()AT

-7

-on

-A

._ 1__________ ~~__ ._______ _

-----------Usc
For use by FORPLT
REREAD statement
READ stat.emellt
ACCEPT statement
PRINT statement
PUNCH stat.ement
TYPE statement

Standard Devices"
i}pvicc

'I-D~t:~~lt ~~ii~n;~~-------TLOgiC-;ll~it N-~;b~;--------

FOJ{(Hl.DAT
FOIWl.DAT

00
01

02
03

i
i

· TY

f)4

I
I
I

"','
i
I

)'1'1'
"'.\ 1

0.')
f)(:;

fJl

{1.'I

-----------

Usc
Disk
Disk
Card Reader
Line Pnnter
Console Teletype
lJser's Teletype
Paper Tape Reader
Paper Tape Punch
Display'
DECtape

10
11

!",\-i

iT.\.-'
i';'.l,f;

l:{
14
1[,

\I! Au

.11'-'\1

! i

'-)TA~

11"1

, iHTl{

:;',1"

2fl

:1.--;'"

...-:.;';

· ,:-; f\

2:{
M

• )~!{

IW\, 1
! W\'2
· "F\':;
· ;;·:\'·t
,

FOH25.DAT

Magnetic Tape

Assignable DeVice
Disk

Assigna ble Devices

2fi

.. .

~
\ ' -:
: ! . \ oJ

I Di:-;k
I
I
.

FOH99./lAT

_. _____.__.=~ ___ .________J_[)i'k _____.__.__

":1i' device table ean be altered when l'OHOTS is installed or bv the ,,;v~,telIl administrator. The supplied
q!! Hili,., are PHhpr value;; in the default tabll' pictured above. or all positive logical unit numhers default to disk.

;H'ck to sm' vvhich device taole is oein!! uH:d at vour installation.

10-12

DATA TRANSFER STATEMENTS
10.4.3.2 Internal File Identifier - The
internal
file
identifier
specifies the
internal file to be used.
This identifier is the name
of a character variable, character array, character array element, or
character substring.
Example:
CHARACTER*132 LINE
WRITE (UNIT=LINE,FMT= I (F) ') 3.14159

10.4.4

Record Number References In Data Transfer Statements

All direct-access data
transfer
statements must contain a
record
specifier, which is used in the transfer to identify the number of the
record to be accessed.
The form of the record specifier in the control-information list is:
REC=rn
where:
is a positive integer
record number.

expression

that

indicates

the

When you use the REC=rn form to specify the record
number,
you can
place the record specifier anywhere in the control-information list.
An alternative way for including the record specifier is:
un'rn
where:
is a positive
integer constant,
variable,
or
array
el'2men t
tha t represents the logical unit number of the
device to or from which
the data
transfer
is beinq
made.
When
you
use
this
form
for
specifying the
loqical unit nu~ber, you cannot use the UNIT= keyword.

is an apostrophe delimiting
from the record number.
rn

the

unit

is a positive
integer constant,
variable,
element that represents the record number.

When you use the alternative form for specifying
you cannot use the keyword REC=.

10.4.5

logical

the

record

number
array
number,

Format References in Data Transfer Statements

All formatted data transfer statements must contain a format specifier
in the control-information list.
The general
form of the format
specifier is:
FMT=f
or
NML= a NAMELIST name
10-13

DATA TRANSFER STATEMENTS
where:
FMT=

format
is the keyword used in the keyword form of
the
specifier.
Using
the
keyword
form
of
the
format
specifier makes
it
positionally
independent
in
the
control-information list.

NML=

is the keyword that can be used
instead
of
FMT=
for
NAMELIST
formatting.
Either FMT= or NML= can be used
for NAMELIST formatting.

is a format
identifier.
Depending
on
the
type
formatting chosen, f can be one of the following:
1.

A statement number

A numeric array name

A character expression

An integer, real, or logical variable

An asterisk

A NAMELIST name

If you do not use the keyword form of the format specifier,
you
place
the
format
specifier
as
the
second
item
of
control-information
list
(immediately
following
the
logical
number specifier) (see Section 10.4.2).
Sections 10.4.5.1 through 10.4.5.3 describe all forms
specifier.

10.4.5.1 FORMAT-Statement
Formatting specifier has the following form:

The

must
the
unit

the

format

FORMAT-statement

format

FMT=f
where:
FMT=

is the optional keyword in the format specifier.

is one of the following:
1.

The
statement
number
FORMAT
statement
of
a
appearing
in
the
same program unit as the data
transfer statement

The name of a numeric array

A character expression

An integer, real, or logical variable that has been
assigned
a
FORMAT statement number with an ASSIGN
statement (see Section 8.3)

(See
Section
12.1
for
FORMAT-statement formatting.)
10-14

information

DATA TRANSFER STATEMENTS
The following examples show all forms of the FORMAT-statement
format
specifier.
In
the
first
e~ample,
the
format specifier (FMT=lOl)
references the FORMAT statement 101 in the same program unit.

101

PROGRAM TEST
1=67
P=90.8
WRITE (UNIT=22,FMT=lJl) I,P
FORMAT (IX, 'FIRST VA::'UE IS: ' , I , ' SECOND VALUE IS:
END

',F)

In the second example, the saml~ format list used in the first example
is stored in an la-element ar~ay.
Note that the word "FORMAT" is not
included in the array.
PROGRAM TESTB
DIMENSION MYARAY(lO)
MYARAY (1) =' (IX,' , ,
MYARAY(2)='FJRST'
MYARAY(3)=' "ALU'
MYARAY(4)='E IS:'
MYA RA Y (5 ) =' '1', I , ,
MYARAY(6)=' ,q SEC'
MYARAY(7)='OND V'
MYARAY(8)='ALUE '
MYARAY(9)='IS: '"
MYARAY (10) =' "F) ,
1=67
P=90.8
WRITE
END

(UNIT=22,FMT=MYARAY)I,P

In the third example, the same format
list used
examples is stored in a character expression.

the

first

", I ,"

' //

two

PROGRAM TESTe
INTEGER I
REAL P
CHARACTER WORDl*5,WORD2*6
1=67
P=90.8
WORDl='FIRST'
WORD2='SECOND'
WR I 'r E (U NI 'r = 2 2 , FM T =' (1 X, " , / /WO R01/ /' VA L UE IS:
1 WORD2//' V.I\LUE IS: ",F)') I,P
END

In the fourth example, the format specifier (FMT=IFORMT)
variable that has been assigned a statement number.

101

PROGRAM TESTD
ASSIGN 101 TO IFORMT
1=67
P=90.8
WRITE (UNIT=22,FMT=IFORMT) I,P
FORMAT (IX, 'FIRST VALUE IS: " I,
END

references

'SECOND VALUE IS:

',F)

For more information on FORMAT-statement formatting, see Section 12.1.
10-15

DATA TRANSFER STATEMENTS
10.4.5.2 List-Directed Formatting - In list-directed formatting,
the
variables
in
the I/O list of the data transfer statement dictate the
formatting of the data during the transfer.
The form of the list-directed format specifier is:
FMT=*
where:
FMT=

is the optional keyword part of the
format
specifier.
Including
this
keyword
in
the
format specification
makes the specification positionally independent in the
control-information
list.
If you omit
the FMT=
keyword,
the
format
specifier must be
the
second
specifier (the unit specifier must he first).

is an asterisk that indicates that
list-directed.

In the following example, the variables
list-directed formatting.

and

the

formatting

are

formatted

PROGRAM TESTLD

1=67
P=90.8
WRITE (UNIT=22,FMT=*)
END

I,P

List-directed formatting is further described in Section 12.5.

Formatting - If
the
formatting
is
10.4.5.3 NAMELIST-Statement
\lAMELIST,
the
format
specifier
in
the control-information
list
reference
to
a
NAMELIST name defined
in
a
NAMELIST
contains a
statement
in
the
same
program unit.
Since the
NAMELIST name
defInition contains an
I/O list,
a
data
transfer
statement
that
contains a
NAMELIST name in the format specifier cannot also contain
'n1 I/O 1 ist.
The form of the NAMELIST format specifier is:
FMT=name
or
NML=name

!"MT=

is the optional keyword part of the
format
specifier.
Including the keyword in the format specification makes
i t positionally independent in the
control-information
list.
If
you do not include the keyword part of the
Format specifier, you must place the
format
specifier
:;econd (after the logical unit :1umber specifier) in the
control-information list.

~~ML=

is an alternative keyword that can be used in place
::'MT.

IS the NAMELIST name.
The NAMELIST name is defined
NAMELIST statement in the same program unit.

,;ame

10-16

DATA

TR~NSFER

STATEMENTS

In the following example, the data transfer statement uses a
name in its format reference:

NAMELIST

PROGRAM TESTNL
NAMELIST/MYIOLT/I,P
READ (UNIT=5,NML=MYIOLT)
WRITE (UNIT=5,FMT=MYIOLT)
END
'rhe execution of this sample program is as follows:
EXECUTE TES'I.FOR
LINK:
Loading
[LNKXCT TESTNL execution]
$MYIOLT I=675,P=34.71$
$MYIOLT
675, P= 34.71000
$END

CPU time

O.L~

Elapsed t:me 32.0

For further information on thE! NAMELIST statement, see SectIon 12.7.

10.4.6

Optional End-of-File Transfer of Control (END=)

The optional end-of-file transfer specifier
(END=)
specifies
a
statement number to which control passes if this statement attempts to
read past the last data record of a file.
If you
include an ERR= specifier
(Section 10.4.7)
and no
specifier, control passes to the statement indicated in the
specifier whenever an end-of-file condition occurs.
Note that an
specifier on any output statement and on an input statement
direct-access file is ignored.

END=
ERR=
END=
of a

If no END= specifier, IOSTAT= specifier, or ERR= specifier is included
in the data
transfer stab~ment,
and an end-of-file condition is
encountered,
an error message is displayed on the
controlling
terminal, and program execution is terminated.
The form of the END specifier is:
END=s
where:
The

END=

statement

END=

is the keyword part of the END= specifier.
portion of the END= specifier is required.

is the statement number of an executable
the current program unit.

transfer
In the following example, the end-of-file specifier causes a
of
control
to statement 50 after
the data
transfer statement
encounters an end-of-file on unit 22.

10-17

DATA TRANSFER STATEMENTS
PROGRAM TESTEN
READ (UNIT=22,FMT=30,END=50) A,B,C
FORMAT (FjFjF)
GO TO 100
WRITE (UNIT=5,FMT=75)
FORMAT (lX,'END-OF-FILE HAS BEEN ENCOUNTERED')
WRITE (UNIT=5,FMT=105)
FORMAT (IX, 'EXECUTION HAS ENDED')
END

30
50
75
100
105

The following
shows the sample program being executed and
the
end-of-file branch being taken.
In this example, the READ statement
reads from
the default filename,
FOR22.DAT.
To demonstrate
the
end-of-file branch,
FOR22.DAT is an empty file.
Thus, when the READ
statement attempts to
read
records
from FOR22.DAT,
an
immediate
end-of-file condition is detected.
EXECUTE TEST. FOR
FORTRAN:TESTEN
TESTEN
LINK:
Loading
[LNKXCT TESTEN execution]
END-OF-FILE HAS BEEN ENCOUNTERED
EXECUTION HAS ENDED
CPU time 0.2
Elapsed time 0.5

10.4.7

Optional Data Transfer Error Control

(ERR=)

The optional error specifier (ERR=) enables you to specify a statement
to which control passes if an error occurs during the data transfer.
If an error occurs other than for end-of-file, the file is positioned
after the record containing the error.
NOTE
If the program attempts to read
from
the same unit
after an ERR= branch occurs, the record following the
record containing the error will be read.
To read a
record containing the error, the program must execute
either a
REREAD statement
(Section 10.7)
or
a
BACKSPACE
(Section
11.8.4)
followed by a
READ
statement.
If no ERR= specifier or IOSTAT= specifier
is present and
occurs during the data transfer, the program is aborted.

error

The form of the error specifier is:
ERR=s
where:
ERR=

is the keyword portion of the error specifier.

is the statement number of an executable
the same program unit.

10-18

statement

DATA TRANSFER STATEMENTS
The following example shows the error specifier being used
to pass
control to the statement at ljne 85 if an error occurs during the data
transfer.

30
50
75
85
86
100
105

PROGRAM TESTEN
READ (UNIT=22,FMT=30,END=50,ERR=85) A,B,C
FORMAT (F/F/F)
GO TO 100
WRITE (UNIT=5,FMT=7~)
FORMAT (lX,'END-OF-E'ILE HAS BEEN ENCOUNTERED')
GO TO 100
WRITE (UNIT=5,FMT=86)
FORMAT (lX,'THE TRANSFER ENCOUNTERED AN ERROR')
WRITE (5,105)
FORMAT (' EXECUTION HAS ENDED')
END

TYPE FOR22.DAT
100.
200.
AAAA BBBB CCCC DDDD
EXECUTE TESTEN.FOR
FORTRAN: TESTEN
TESTEN
LINK:
Loac\ ing
[LNKXCT TESTEN execution:
THE TRANSFER ENCOUNTERED AN ERROR
EXECUTION HAS ENDED
Elapsed t::.me 2. B
CPU time 0.2
In this example, the error branch is taken when the
input routine
detects
a
nonnumeric
da1:a
item while attempting
to read a
floating-point number into val~iable C.
If the file FOR22.DAT contains
more than three records, the next READ accesses record 4 in the file.

10.4.8

Optional Error Variable For Error Reporting (IOSTAT=)

The optional I/O status specifier enables you to designate an
integer
variable which receives a value indicating the success or failure of
the data transfer.
When the data transfer statement is successfully executed,
the
variable is assigned a valu«? of ZE?ro.
If an error occurs during the
data transfer, the variable is assigned a positive value
indicating
which error occured
(see Appendix D).
In this case, if there is no
ERR= specifier, the program proceeds to the statement after
the data
transfer statement.
If an end-of-filE:?, occurs during the data transfer, the variable is set
to -1.
In this case,
if there is no END= or ERR= specifier, the
program proceeds to the statement after the data transfer statement.
The form of the error

variabl,~

specifier is:

IOSTAT=ios
where:
ios

is an
integer
specifier.

variable

10-19

that

the

I/O

status

DATA TRANSFER STATEMENTS
The following example shows the I/O status specifier being used to
report the number of the error on default unit 5 if the error branch
is taken.
10
30
40
50
75
85
86
100
105

PROGRAM TESTEN
READ {UNIT=22,FMT=30,END=50,ERR=85,IOSTAT=J)A,B,C
FORMAT (F4.1/F4.1/F4.1)
WRITE {UNIT=5,FMT=40)A,B,C
FORMAT (lX,'THE VALUES ARE: ',3F6.1)
GO TO 100
WRITE (UNIT=5,FMT=75)
FORMAT (lX,'END-OF-FILE HAS BEEN ENCOUNTERED')
GO TO 100
WRITE {UNIT=5,FMT=86)J
FORMAT (lX,'THE TRANSFER ENCOUNTERED AN ERROR; STATUS:
IF(J.GT.O) GO TO 10
WRITE (5,105)
FORMAT (' EXECUTION HAS ENDED')
END

',IS)

TYPE FOR22.DAT
100.
200.
AAAA BBBB CCCC DODD
80.
90.
95.
EXECUTE TESTEN.FOR
FORTRAN: TESTEN
TESTEN
LINK:
Loading
[LNKXCT TESTEN execution]
THE TRANSFER ENCOUNTERED AN ERROR; STATUS:
THE VALUES ARE:
80.0
90.0 95.0
EXECUTION HAS ENDED
CPU time 0.2

307

Elapsed time 1.5

In this example, the IOSTAT variable J is set when
the
first
READ
detects a nonnumeric data
item while trying to input the data for
variable C.
In
this case,
the value of 10STAT represents the
processor specific error number (the second value listed in the FOROTS
error messages in Section 0.1), and
indicates that an
illegal
character has been detected in the data.
After the error status has
been printed, the second READ successfully executes using
records 4,
5, and 6 from the file.

10.4.9

Data Transfer Statement Input/Output Lists

The I/O list in an input or output statement contains the names of
variables,
arrays,
array elements, or character substrings.
The I/O
list in an output statement can also contain expressions,
function
references, or constants.
An I/O list has the following form:
e[,e] .•.

10-20

DATA TRANSFER STATEMENTS
The variable i and the parameters el, e2, and e3 have the same forms
and the same functions that they have in the DO statement (see Section
9.3).
The list immediately preceding the DO loop control variable is
the range of the implied DO loop.
Elements in that list can reference
the index, but they must not alter it.
Some examples are:
WRITE

(3,200)

(A,B,C, 1=1,3)

The statement in this example functions as though you had written:
WRITE

(3,200) A,B,C,A,B,C,A,B,C

The following two statements are
WRITE

(3,200)

thE~

same:

(X(I),I=l,])

WR I T E ( 3 , 200) X (l) ,X ( 2) ,;(( 3 )
Another example is:
WRITE

(6)

(1, (J,P(I) ,Q(I,.J) ,J=l,L) ,I=l,M)

The I/O list in this example consists of an implied DO list containing
another implied DO list nested with it.
The implied DO lists together
write a total of (l+3*L) *M fields, varying values of J for each value
of I.
In a series of nested implied DO lists, the parentheses
indicate the
nesting
(see Section 9.3.4).
Execution of the innermost list is
repeated most often.
For example:
150

WRITE (6,150) «FORl1(K,L) •. L=l,lO), K=l,10,2)
FORMAT (FIO.2)

Because the inner DO loop is E~xecutE~d 10 times for each iteration of
the outer loop,
the second subscript, L, advances from 1 through 10
for each increment of the first subscript.
This is the reverse of the
order of subscript progression.
In addition, K is incremented by 2,
so only the odd-numbered rows of the array are output.
The entire list of an
implied DO list
is transmitted
control variable is increment(~d, fOl= example:
READ

( 5, 999)

(P (I),

before

the

(Q ( I , J), J := 1 , 10), I = 1 , 5 )

In this example, pel), Q(l,l), Q(1,2), ••• ,Q(1,10) are read before I is
incremented to 2~
When processing multidimensional an=ays, you can use a combination of
fixed subscripts and subscripts that vary according to an implied DO
list, for example:
READ

(3,555~»

(BOX(l,J), J=l,lO)

This statement assigns input values to BOX(l,l) through BOX(l,lO)
then terminates without affecting any other element of the array.

and

The value of the control variable can also be
example:

For

WR I T E (6, 1111)

(I,

output

directly.

1=1, 21)

This statement simply outputs the integers 1 through 20.

10-23

DATA TRANSFER STATEMENTS

READ
Statement

10.5

READ STATEMENT

The READ statement transfers data from a file into memory.
There are
two categories of READ statements:
formatted (see Section 10.5.1) and
unformatted (see Section 10.5.2).
Table 10-4 summarizes the various forms of the READ statement.

10-24

DATA
Table 10-4:

TRA~SFER

STATEMENTS

Summary of READ Statement Forms

Data Access

Statement Construct

Sequential Formatted
(FORMAT Statement)

READ(UNI'l = un,FMT = fI ,END = s H,ERR = sJ[ ,lOSTAT = ios])! iolist I
READ(
un,FMT = fI ,END = s 11 ,ERR ~ s 11 ,lOSTAT = ios 1)[ iolist 1
REA D(
un,
fl ,END = s If,ERR = s H,10STAT = ios])1 iolist 1

Sequential Formatted
(List Directed)

READ(UNI'l = un,FMT = *1 ,END = s 11 ,ERR = s 11 ,10STAT = ios])l iolist I
READ(
un,FMT = *1 ,END = s]l ,ERR --,- s 11 ,lOSTAT = ios III iolist I
READ(
un,
*1 ,END = s 11 ,ERR = s 11 ,lOSTAT-=- ios I)] iolist 1

Sequential Formatted
I NAMELIST Statement!

HEAD(UNl'I-un.FMT--namel.END- sll.EHH ~IIJOSTAT -iosl)
REAJ)(UNlT un.NML - namel.END sll.ERR sliJOSTAT -iosl)
HEAD!
un.FMT namel.END-slI.EHR sll.lOSTAT iosl)
READ!
un.NMLc namel.END' sll.ERH sll.lOSTAT ios])
READ(
un.
namel.END-- !:i11.ERR sliJOSTAT iosl)

Sequential Formatted
(Default Unit)

READ fI ,ioli.,t 1
READ *I,iolistl
READ(UNI'l = *,FMT =fl,END=s]l,ERR= sll,lOSTAT= iosllliolistl
READ( UNI'l = *,FMT = *1 ,END = s 11 ,ERR = s 11 ,lOSTAT =- ios 1)] iolist 1

Sequential Unformatted

READ(UNI'I = unl ,END = s 11 ,ERR = s 11 ,10STAT = ios])l iolist 1
READ(
unl ,END = sll,ERR= sJ!,10STAT= iosl)]iolistl

Direct Formatted

READ(UNI1 = un,FMT= f,REC = rnl,ERR= sll ,10STAT= iosl)liolistl
READ(
un,FMT = f,REC = rnl ,ERR ~ s 11 ,lOSTA T = ios 1)1 iolist 1
un,
f,REC = rnl ,ERR = s 11 ,lOSTA T --,- ios III iolist I
READ(
I{EADI
un'rn.Fl\1T f
I.ERH sliJOSTAT iosi!liolistl
un'rn.
f
I.EHH-sIiJOSTAT iosilliolistl
READI

Direct Unformatted

READ(UNI'I = un,REC = rnl ,ERR = s 11 ,lOSTAT = ios])1 iolist 1
un,REC =rnl,ERR= sll ,10STAT= ios])]iolistl
READ(
un'rn
I.EHH sil.lOSTAT - iosllJiolis11

Key:
UNIT=un

is a FORTRAN logical unit number (Section 10.4.3),

UNIT=*

is a default unit specification mection 1004.3).

REC=rn

is a direct-access record number (Section 10.4.4).

un 1'n

is an alternate way of specifying Logical Unit Number and record number for a direct-access
1ransfer (Section 10.4.4).

FMT=f

is FORMAT-statement formatt'ng; iolist is optional (Section 10.4.5.l>.

FMT-*

is list-directed formatting; iolist is optional (Section 10.4.5.2).

FMT -- name

is NAMELIST-siatell1cnt form[ tting; iolist is prohiuited (Section 10.4.5.3).

Nl\lL- name

is the altl'rnativ(' form of the !'.AMELlST-statement format spl'cifier ISection 10.4.5.:3)

END=s

is an optional end-of-file transf.~r specifier (Section 10.4.6).

ERR=s

is an optional error transfer spl~cifier (Section 10.4.7!.

IOSTAT=ios

is an optional 110 status specifier (Section 10.4.8).

iolist

is a data transfer 110 list (Section 10.4.9),

10-2 !3

DATA TRANSFER STATEMENTS
10.5.1

Formatted READ Transfers

A formatted READ transfer uses a READ statement that specifies that
the
transferred data
is edited during the transfer, such that the
external and internal representation of the data are different.
The
three
types of formatted
READ statements are:
FORMAT-statement,
list-directed, and NAMELIST-statement.
There are two types of access
to the device from which the READ
statement transfers data.
They are sequential and direct.
If you
want to perform a direct-access formatted READ from a device, you must
use
FORMAT-statement
formatting.
List-directed
and
NAMELIST
formatting can only be used with sequential-access
formatted
READ
statements.

10.5.1.1 Sequential FORMAT-Statement READ - This section descl~ibes
the sequential-access (FORMAT-statement) formatted READ statement.
This statement has the following forms:
READ (UNIT=un,FMT=f[,END=s] [,ERR=s] [,IOSTAT=ios]) [iolist]
READ

(un,FMT=f[,END=s] [,ERR=s] [,IOSTAT=ios]) [iolist]

READ (un,f[,END=s] [,ERR=s] [,IOSTAT=ios]) [iolist]
If an I/O list is included in these forms, it specifies that data
is
transferred
from
logical unit un,
formatted according
to
the
specification given by f, and transferred into the elements of the
specified I/O list.
If an I/O list is not included, the input record is skipped.
(If
the
FORMAT statement specifies slash editing, more than one record can be
skipped.
H or apostrophe editing can cause data transfers to occur to
the FORMAT statement itself.
See Section 12.4.)
The
following
example contains two READ
contains an I/O list; the second does not:
5

statements:

the

first

READ (22, 5) A, Z, J
FORMA T (2 FlO. 2 , I 5 )
READ (22,5)
END

In this example, the
first
READ statement reads one record
from
logical unit
22, formats the data according to the FORMAT statement,
and assigns the values to the variables A, Z, and J.
The second READ
statement skips one input record on logical unit 22.
The default unit forms of this READ statement operates in the same way
as the first
forms,
except that data transfers reference the card
reader, which is the default logical unit for these forms.
The default unit forms of this statement are:
READ f[,iolist]
READ (UNIT=*,FMT=f[,END=s] [,ERR=s] [,IOSTAT=ios]) [iolist]

10-26

DATA TRANSFER STATEMENTS
10.5.1.2 Direct-Access FORMAT-Statement READ - This section describes
the direct-access (FORMAT-statement) formatted READ statement.
The forms of this statement are:
READ (UNIT=un,FMT=f,REC=rn[ ,ERR:=s] [,IOSTAT=ios]) [iolist]
READ (un,FMT::f,REC=rn[,ERR=s] [, IOSTAT=ios]) [iolist]
READ (un,f,REC=rn[,ERR=s] [,IOSTi\T=ios]) [iolist]
If an I/O list is included in these
forms,
the data specified by
record rn is transferred, according to the format specifications given
in f, into the elements of the I/O list.
These forms can be used only
with disk files
that have been opened by an OPEN statement that
specifies ACCESS='DIRECT', ACCESS='RANDOM',
or ACCESS='RANDIN'
(see
Sec t i on 11. 3. 1) •
If the record specified by rn has not been written, an
(except for IMAGE mode files).

error

results

The following example shows this form of the READ statement.

OPEN(22,RECORDSIZE=25,ACCESS='DIRECT')
READ (22,5,REC=10)A,Z,J
FORMAT (2F10.2,I5)
END

In this example, the READ stab?ment lreads record 10 from logical
unit
22,
formats the data according to the FORMAT statement, and assigns
the values to variables A, Z, dnd J.
The alternative forms of this ~EAD statement operate in the same way
as the first forms.
The only difference between the forms is the way
in which the unit and record s~ecifications are expressed.
The alternative forms for this statement are:
READ (un'rn,FMT=f[,ERR=s] [,IOSTAT=ios]) [iolist]
READ (un'rn,f[,ERR=s] [,IO;:;TAT=ios]) [iolist]
In the alternative forms,
the unit and
record references do not
contain the keywords UNIT= and REC=.
Instead the unit number is
specified first; a single quot'? (') is specified next; followed by a
record number, a comma, and finally the format reference.

10.5.1.3 Sequential List-Directed READ - This section describes
sequential-access (list-directed) formatted READ statement.

the

This statement has the following forms:
READ

(UNIT=un,FMT=*[,END=:3] [,ERR=s] [,IOSTAT=ios]) [iolist]

READ (un,FMT:=*[,END=s] [,EHR=s] [,IOSTAT=ios]) [iolist]
READ (un,*[,END=s] [,ERR=sl [,IOS~PAT=ios]) [iolist]
with these forms, the data is transferred from logical device un and
is
formatted according
to the data types of the elements of the I/O
list.
If the I/O list is not included, a record is skipped.

10-27

DATA TRANSFER STATEMENTS
The default unit forms of this statement are:
REA 0 * [ , i 0 1 is t ]
READ

(UNIT=* ,FMT=* [,END=s] [,ERR=s] [, IOSTAT=ios]) [iolist]

With these forms, the data is transferred from the card
reader
(the
default device),
and is formatted according to the data types of the
elements in the specified I/O list.
The following example shows this form of the READ statement:
CHARACTER*14 C
DOUBLE PRECISION T
COMPLEX D,E
LOGICAL L,M
READ (l,*) I,R,D,E,L,M,J,K,S,T,C,A,B
END
The external record to be read contains the following:
4 6.3

(3.4,4.2),

(3,2 )

, T,F,,3*14.6 ,'ABC,DEF/GHI' 'JK'/

Upon execution of the program unit, the following values are
to the I/O list elements:
I
R
D
E
L
M
K
S
T
C

assigned

4
6.3
(3.4,4.2)
(3.0,2.0)
.TRUE.
.FALSE.
14
14.6
14.6DO
ABC,DEF/GHI'JK

A, B, and J are unchanged.

:0.5.1.4 Sequential NAMELIST-Statement READ - This section describes
t:)C sequential-access
(NAMELIST-statement) formatted READ statement.
~~lS

statement has the following forms:
,'~EAD

(UNIT=un,FMT=name[,END=sJ [,ERR==s] [,IOSTAT=ios])

l~EAD

(UNI'r=un,NML=namet,END=s] [,ERR=s] [,IOSTAT=ios])

i;;EAD

(un,FMT=name/,END=sJ [,ERR==s] [,IOSTAT=ios])

i~EAD

(un,NML=name[ ,END=sJ [,ERR=s] [,IOSTAT=ios])

!.;;r:AD

(un,name( ,END=s] [,ERR=s] [, IOSTAT=ios])

WIth these forms,

the data is transferred from the specified unit into
locations specified by the
NAMELIST list.
The formatting is
·»ntrolled by the implicit data types of the NAMELIST list items.
We
>;~lagest
that
you
use
the
NAMELIST
form of the READ statement to
'~-,lnsfer data from files created by the NAMELIST
form
of
the WRITE
:-. t-,d ternent
(Sect ion 10.6.1.4).
~he

10-28

DATA TRANSFER STATEMENTS
where:
e

is a simple input or output item (see Section
an implied DO list :see Section 10.4.9.2).

10.4.9.1)

The I/O statement: assigns values to, or
transfers values from,
the
list elements in the order in which they appear (from left to right).

10.4.9.1 Simple List Elements - A simple input list item can
of the following:
1.

A variable name

An array element

A character substring name

An array name

one

namE~

For example:
READ (5, 10) J, K (3) , CH (1 : ])
A simple output list item can be one of the above, or it can be one of
the following:
1.

A consta.nt

A function reference

An expression

For example:
WRITE (5,10) J,K(3),(L+4)/2,CH(1:3)
An input list itE~m cannot be an expression.
However, it
expressions as subscr ipts or ~)ubstr ing bounds.
I/O list items can be of the following types:
1.

In teger

Real

Double-precision

Complex

Logical

CharactE~r

Octal

Double Octal

Hollerith

10-21

can

contain

DATA TRANSFER STATEMENTS
When you use an unsubscripted array name in an I/O list,
an
input
statement reads enough data
to fill every element of the array; an
output statement writes all the values in the array.
Data
transfer
begins with the initial element of the array and proceeds in the order
of subscript progression, with the leftmost subscript varying most
rapidly from lower to upper bound.
For example, the following defines
a two-dimensional array:
DIMENSION ARRAY(3,3)
If the name ARRAY with no subscripts appears in a READ statement, that
statement assigns values
from the
input record(s) to ARRAY(I,I) ,
ARRAY(2,1) , ARRAY(3,1) , ARRAY(I,2), and so on through ARRAY(3,3).
In an input statement, variables in the I/O list can be used in
subscripts later in the list, for example:

array

READ (1,1250) J,K,ARRAY(J,K)
FORMAT (Il,IX,Il,IX,F6.2)

1250

The input record contains the following values:
1,3,721.73
When the READ statement is executed, the first input value is assigned
to J
and
the second to K, thereby establishing the actual subscript
values
for ARRAY(J,K).
Then the value 721.73
is assigned
to
ARRAY(I,3).
Variables that are to be used as subscripts in this way
must appear before (to the left of) their use as the array subscripts
in the I/O list.

10.4.9.2 Implied DO Lists - An implied DO list is an I/O list element
that functions as though it were part of an I/O statement within a DO
loop.
Implied DO lists can be used to:
1.

Specify iteration of part of an I/O list

Transfer part of an array

Transfer array elements in a
sequence
order of subscript progression

different

from

the

As in explicit DO loops, zero-trip implied DO loops are possible
Section 9.3).

(see

An implied DO list has the form:
(d 1 i s t , i =e 1, e 2 [ ,e3] )
where:
dlist

is an I/O list.
Dlist
implied DO lists.

is the index control variable that can represent a
subscript appearing in a preceding subscript list.

el,e2,e3

are
the
indexing
parameters
that
specify,
respectively, the initial, terminal, and increment
values that control the range of
i.
If e3
is
omitted
(with its preceding comma), a value of 1
is assumed.
10-22

can

also

contain

other

DATA TRANSFER STATEMENTS
The following example shows this form of the READ statement:
NAMELIST /DATA/A,Z,J
( 2 2 , DA TA)
END

l~ EAD

In this example, the NAMELIST 3tatement associates the NAMELIST name
DATA with a
list of three items.
The corresponding READ statement
reads input data and assigns v::dues to the specified namelist items.

10.5.2

Unformatted READ Transfers

Unformatted READ transfers move data
from a specified file to
locations in memory.
Unlike formatted READ transfers, unformatted
transfers do not involve any editing of the data.
The two types of unformatted data transfers enable
specified file either sequentially or directly.

you

access

NOTE:

The OPEN statement MJDE specifier enables you to
specify in which form the unformatted data file exists
(see Section 11.3.20).
If you execute an unformatted
READ statement without having first specified the MODE
in an OPEN statement, the data file is assumed
to be
BINARY.
(For additi·)nal
information on unformatted
data file forms, see S'?ction 11.2.)

10.5.2.1 Sequential Unformatted READ - This
sequential-access unformatted READ statement.

section

describes

the

This statement has the following forms:
READ

(UNIT=un[,END=s] [,ERR=s] [,IOSTAT=ios]) [iolist]

READ (un[,END=s] [,ERR=s] [,IOSTAT=ios]) [iolist]
If the I/O list is present, the data is transferred as one logical
record from the specified logical unit.
This type of read should only
be used to read files that have been created by unformatted WRITE
statements.
If you omit the I/O list portion of the statement, the statement skips
one logical record on input from the specified unit.
The following example shows this type of READ statement used both with
and without the I/O list:
READ (22) A, Z, J
READ (22)
END
In this example, the first
READ statement reads one record from
logical unit 22 and assigns values to variables A, Z, and J.
The
second READ statement skip one record from logical unit 22.

10-29

DATA TRANSFER STATEMENTS
10.5.2.2 Direct-Access Unformatted READ - This section describes
direct-access unformatted READ statement.

the

This statement has the following forms:
READ

(UNIT=un,REC=rn[,ERR=s] [,IOSTAT=ios]) [iolist]

READ

(un,REC=rn[,ERR=s] [,IOSTAT=ios]) [iolist]

If an I/O list is included in these forms, the data, in
the
form of
one
logical
record,
is transferred from the specified unit into the
elements of the I/O list.
Only files that have been output by an
unformatted WRITE statement should be transferred by this form of the
READ statement.
In addition, for FORTRAN binary files, if the record
specified by rn has not been written, an error results.
These forms can be used only with disk files that have been opened by
an OPEN statement that specifies ACCESS='DIRECT', qCCESS='RANDOM ' , or
ACCESS='RANDIN ' (see Section 11.3.1).
Tne alternative form of this READ statement functions the same as
the
first forms.
The only difference between the forms is in the way that
the unit and record are specified.
The alternative form of this statement is:
READ

(un'rn[,ERR=s] [,IOSTAT=ios]) (iolist]

In this form, the unit and
record
references do
not contain
the
keywords
UNIT= and REC=.
Instead the unit number is specified first;
a single quote (I) is specified
next;
then
the
record
numbe:r
is
specified last.
The following example demonstrates the use
statement:

the

unformatted

READ

OPEN (22,ACCESS='DIRECT' ;.·:CCORDSIZE=3)
READ (22,REC=10)A,Z,J
READ (22Il2)B,X,K
END
In this example, the first READ statement reads record 10 from logical
unit
22 and assigns values to the variables A, Z, and J.
The second
READ statement reads record 12 from logical unit 22 and assigns values
to the variables B, X, and K.

WRITE
Statement

10.6

WRITE STATEMENT

WRITE statements transfer data from memory to a
file.
The various
forms
of
the WRITE statement enable
it to be used in sequential,
append, and direct-access transfer modes for
formatted,
unformatted,
list-directed, Jnd NAMELIST-controlled data transfers.

10-30

DATA TRANSFER STATEMENTS
Table 10-5 summarizes all forms of the WRITE statement.
Table 10--5:

Summary of WRITE Statement Forms

Data Access

Statement Construct

Sequential Formatted
(FORMAT Statement)

WRlTE(UNIT =, un,FMT =£1 ,ERR = sll ,10STAT= iosilliolistl
un,FMT = £I ,ERR = s 11 ,lOSTAT ~ ios])l iolistl
WRITE(
un,
fl,ERR=slI,IOSTAT=ios111iolist1
WRITE(

Sequential Formatted
(List Directed)

WHITE(UNIT =, un,FMT = *1 ,ERR = s 1[ ,10STAT = ios DI iolist I
WRITE(
un,FMT = *[,ERR = sll ,10STAT = ios 1lI iolist I
WRITE(
un,
*[,ERR=sJ[,10STAT=iosl)liolist1

Sequential Formatted
(j\,"AMELIST Statement)

WEITE(UI'IT - un.FMT - namel.ERH - sll.IOSTAT, iosll
WHITE(LWIT- un.NML- namel.EHR - sll.lOSTAT iusll
WHITEr
un.FMT - namel.ERH s II.IOSTAT - ios i I
WHITEr
un.NML namel.EHR - sllJOSTAT - lOS!1
WHITEI
un.
namel.ERH sll.lOSTAT iosll

Sequential Formatted
(Default Unit)

w nnE fI.i )]jst I
WHITE ';-I.iolist I
WRITE(UNIT ~ *,FMT = £I ,ERR = s 1[ ,lOST AT = ios 1lI iolist I
WRITE (UNIT =, *,FMT = *1 ,ERR = s I[ ,IOSTAT = ios 1)1 iolist I

Sequential Unformatted

WRITE(UNIT =- unl ,ERR = sJ[ ,10STAT = ioslll iolistl
unI,ERR= sJ[ ,10STAT= iosllliolistl
WRITE (

Direct Formatted

WRITE( UNIT =, un,FMT = f,REC = rnl ,ERR = s J[ ,lOSTAT - ios l)j iol ist I
WRITE(
un,FMT = f,REC = rnl ,ERR = s 11 ,lOSTAT ~. ios 1)1 iolist I
WRITE(
un,
f,REC = rnl ,ERR= sll ,IOSTAT 'co iosDliolistl
WHIT~;I
un'rn,FMT f
I.EHR- sll.lUSTAT -ioslJliolistl
\\'R.IH)
un'rn.
r
I.ERH --sIiJOSTAT iosDliolistl

Direct Unformatted

WRITE (UNIT =, un,REC = rnl,ERR == s11,10STAT ios III iolist 1
un,REC = rnl ,ERR = s 11 ,lOSTAT - ios 111 iolist 1
WRITE(
Wf{JTE(
un'rn
I.ERR-sII.lOSTAT iosliliolistl
co-

Key:
UNIT=un

is a FORTRAN logical unit number (Seetion 10.4.3),

UNIT=*

is a detmlt unit specification (Section 10.4.3).

REC=rn

is a direct-access record number (Section 10.4.4).

un'rn

is an alternate way of sppcifYcng Logical Unit Number and record number fill' a direct-access
transfer (Section 10.4.4 J.

FMT=f

is FORMAT-statement formatting; iolist is optional (Section 10.4.5.11.

FMT=*

is list-directed formatting; iolist is optional (Section 10.4.5.2).

FMT-'-name

is NAMELIST-statement fornatting; iolist is prohihited (Section 10.4.5.:3).

NML

is the alternative form of the NAME LIST-statement format specifier (Section 10.4_5.3)

name

ERR=s

is an optional error transfer Epecifier (Section 10.4.7).

10STAT=ios

is an optional 110 status spec: fier (Section 10.4.8).

iolist

is a data transfer 110 list (Se<tion 10.4.~1).

10.6.1

Formatted WRITE Transfers

A formatted WRITE transfer USE'S a WRITE statement that specifies that
the transferred data
is edited during the transfer, such that the
external and internal representations of the data are different.
The
three types of formatted ~lRITE statements are:
FORMAT-statement,
list-directed, and NAMELIST-st.atement.
10-31

DATA TRANSFER STATEMENTS
There are two types of access to the device to which the WRITE
statement transfers data.
They are sequential and direct.
If you
want to perform a direct-access formatted WRITE to a device, you must
use FORMAT-statement formattin~.
List-directed and NAMELIST-statement
formatting can only be used for
sequential-access formatted WRITE
statements.

10.6.1.1 Sequential FORMAT-Statement WRITE - This section describes
the sequential-access (FORMAT-statement) formatted WRITE statement.
This statement has the following forms:
WRITE (UNIT=un ,FMT=f [ , ERR=s] [, IOSTAT=ios] ) [iol ist]
WRITE (un,FMT=f[,ERR=s] [,IOSTAT=ios]) [iolist]
WRITE (un,f[,ERR=s] [,IOSTAT=ios]) [iolist]
If the I/O list is included in these forms, the data specified by the
elements of the I/O list are output to a file on logical unit un.
The
output data
is formatted
in this file according to the FORMAT
specifications given in f.
A blank record is written if the I/O list is not specified, and one of
the following is true:
1.

The FORMAT statement is empty.

No slash, H, or apostrophe editing descriptors occur alone.

No slash, H, or apostrophe editing
first repeatable edit descriptors.

descriptors

precede

the

WRITE

that

See Section 12.4.
The following example contains a sequential
contains an I/O list, and one that does not:

5
15

formatted

A=11.4
Z=13.9
J=5
WRITE (22,5)A,Z,J
FORMAT (IX,2FIO.2,I5)
WR I T E (22, 15 )
FORMAT (' PAGE NO.1')
END

The following is written to logical unit 22:
11.40
PAGE NO. 1

13.90

The default unit forms of this statement are:
vvRITE f[,iolist]
WRITE

(UNIT=*,FMT=f[,ERR=s] [,IOSTAT=ios]) [iolist]

10-32

DATA TRANSFER STATEMENTS
If an I/O list is included in these forms, the data, specified by the
elements within the
I/O list, are transferred to the default device
(line printer).
The transferIed data are formatted according to the
FORMAT specification given by f.
The following example shows both forms of this WRITE transfer:

5
15

A=11.4
Z=13.9
J=5
WRITE ~>,A,Z,J
FORMAT (lX,2FIO.2,I~.)
WRITE 15
FORMAT (I PAGE NO. J. I )
END

The following is written to the default device (line printer):
11.40
PAGE NO. 1

13.90

10.6.1.2 Direct-·Access FORM1~T-Statement WRITE - The
direct-access
(FORMAT-statement)
formatted WRITE statement is described in this
section.
This statement has the following forms:
WRITE(UNIT=un,FMT=f,REC=l~n[,ERH=s]

[,IOSTAT=ios]) [iolist]

WRITE (un ,FM'l'=f, REC=rn [ , EHR=s] [" IOSTAT=ios] ) [iol ist]
WRITE(un,f,HEC=rn[,ERR=s:i [,IOSTAT=ios]) [iolist]
If you include an I/O list in these forms, the data in the I/O list is
written starting at record rn to a file on logical unit un.
The
formatting is controlled by the FOR1'1AT specifications given at f.
Only disk files that have been opened by an OPEN statement that
specifies ACCESS='DIRECT ' or ACCESS='RANDOM ' (see Section 11.3.1) can
be accessed by a WRITE statement of this form.
If you omit the I/O list portion of this statement, at least one blank
record (specified by REC=rn) is written to logical unit un.
The following example shows a direct-access formatted WRITE
that contains an I/O list, and one that does not:

5
15

A=11.4
Z=13.9
J=5
OPEN(22, RECORDSIZE=25,ACCESS='RANDOM ' )
WRITE (22,5,REC=10)A,Z,J
FORMAT (2FlO.2,I5)
WRITE (22,15,REC=11)
FORMAT (I PAGE NO. II)
END

10-33

statement

DATA TRANSFER STATEMENTS
The following is written to logical unit 22:
11.40
PAGE NO. 1

13.90

The alternative forms of this WRITE statement operate the same way as
the
first forms.
The only difference between the forms is in the way
that the logical unit and the record number are expressed.
The alternative.forms of this statement are:
WRITE

(un'rn,FMT=f[,ERR=s] [,IOSTAT=ios]) [iolist]

WRITE

(un'rn,f[,ERR=s] [,IOSTAT=ios]) [iolist]

In these forms, the unit and record
references do not contain the
keywords
UNIT= and REC=.
Instead the unit number is specified first;
a single-quote (I) is specified next, followed by a record
number,
a
comma, and finally the format reference.

10.6.1.3 Sequential List-Directed WRITE - This section describes
sequential-access (list-directed) formatted WRITE statement.

the

This statement has the following forms:
WRITE

(UNIT=un,FMT=*[,ERR=s] [,IOSTAT=ios]) [iolist]

WRITE

(un,FMT=* [, ERR=s] [, IOSTAT=ios]) [iolist]

WRITE

(un,*[,ERR=s] [,IOSTAT=ios]) [iolist]

These forms of the WRITE statement specify that the data identified in
the
I/O list is written to logical unit un.
Because the transfer is
list-directed (FMT=*), the data is formatted according to the implicit
data
types of the variables in the I/O list.
If the I/O list is not
included, a blank record is written.
The default unit forms of this statement are:
WRITE *[,iolist]
WRITE

(UNIT=*,FMT=*[,ERR=s] [,IOSTAT=ios]) [iolist]

The default unit forms function in the same way as the
first
forms,
except that
the output
is written
to
the default device (line
printer) •
The following example shows the list-directed WRITE statement:
DIMENSION A(4)
DATA A/4*3,4/
WRITE (1, *) 'ARRAY VALUES FOLLOW'
WRITE (1,*) A,4
END
The following is written to logical unit 1:
ARRAY VALUES FOLLOW
4*3,400000, 4

10-34

DATA TRANSFER STATEMENTS

10.6.1.4
Sequential NAMELIST-Statement WRITE - This sectIon describes
the sequential-access (NAMELJET-statement) formatted WRI'l'E statement.
This

statement has

the

followjng

forms:

VvRITE

(UNIT=un,FMT=namel ,ERR=sJ [,IOSTAT=ios])

WRITE

(UNIT=un,NML=namel ,ERR=sJ [,IOSTA'I'=ios])

WRITE

(un, FMT=name I, EHR=~;J [, IOSTAT=ios] )

WRITE

(un,NML=name[ ,ERR=~iJ

\'.7RITE

(un,namef ,ERR=:sll,JOSTAT=:ios])

I ,IOSTAT=iosJ)

These forms of the
WRITE
sti-:tement
transfer
data
defined
referenced
NAMELIST
statE:'ment
(FtvIT=name or NML=name) to the
the logical unit specified by un.
The following
statement:

example demonstrates

the

NAMELIST

CHARACTER*19 NAME(2)
DA TAN AM E /2 '*' I /
I~EAL PITCH, ROLL,
YAW, POSIT(3)
iJOGICAL DIAC:NO
INTEGER ITEPA'l'
NAMELIST /PP,RAM/ NAME, P TCH, HOLL,
ACCEPT (FMT=PARAM)
WRITE (UNIT=l,FMT=PARAM)
':ND
The

input contaLns

the

YAW,

form

POSIT,

DIAGNO,

the

in
the
f i l e on

WRITE

ITERAT

fcllow_nq:

bSPARAM NAME(2) (lO:)='HE:SENBEHG',
hPITCH=5.0, YAW=O.O, ROLL=5.0,
',)LH AGNO =. TRUE.
hITERAT=lO$END
'I' heW R I T Est ate men t

w r I tEO'S

SPARAM
NAME= '
5.000000, ROLL= 5.000000,
DIAGNO= T, ITERAT= 10
$END

10.6.2

t h (~

YAW=

toll ow i n q :

HE ISENBERG', PI TCH=
O.OOOOOOOE+OO, POSIT= 3*0.0000000E+00,

Unformatted WRITE Transfers

Unformatted WRITE transfers move data from memory to a
file.
formatted
WRITE transfers, u1formatted WRITE transfers do not
any editing of the data.
The two types of unformatted data transfers enable you
f i l e either sequentially or directly.

10-35

to write

Unlike
involve

DATA TRANSFER STATEMENTS
NOTE

The MODE spRcifier of the OPEN statemRnt enables you
:0 sORcify thR type of unformatted data file you want
~o create
(see Section 11.3.20).
If
you execute an
I.~nformatted
WRITE statRment without having
first
specIfied the MODE in an OPEN statement, the data file
IS
by
default BINARY.
For additional information on
unformatted data file forms, see Section 11.2.

10.6.2.1 Sequential Unformatted WRITE - This section
sequential-access unformatted WRITE statement.

describes

the

This statement has the following forms:
WRITE

(UNIT=un[,ERR=s) [,IOSTAT=ios)) [iolist)

WRITE

(un[,ERR=s) [,IOSTAT=ios)) [iolist)

If the I/O list is present in these forms, the data is written as one
logical record to the file contained on the specified logical unit un.
If you omit the I/O list in this statement, the statement writes one
blank logical record
to the file contained on the specified loqical
unit un.
The following example shows this form of the WRITE statement with
I/O list and without the I/O list:
WRITE
WRITE
END

the

(22)A,Z,K
(22)

In this example, the first WRITE statement writes a record to the file
connected to logical unit 22 containing the values of the variables A,
Z, and K.
The second WRITE statement writes one blank record
to
the
file connected to logical unit 22.

10.6.2.2 Direct-Access Unformatted WRITE - This section describes the
direct-access unformatted WRITE statement.
This statement has the following forms:
WRITE

(UNIT=un,REC=rn[ ,ERR=s) [,IOSTAT=ios)) [iolist]

WRITE (un,REC=rn[,ERR=s] [,IOSTAT=ios)) [iolist]
These forms can be used only with disk files which have been opened by
an OPEN statement that specifies ACCESS='DIRECT' or ACCESS='RANDOM'
(see Section 11.3.1).
If an I/O list is included in these forms,
the
data,
in the form of one logical record, is transferred from the
memory to record rn of the file on the specified logical unit.
If the I/O list is not specified, the statement
blank record.

10-36

outputs

one

logical

DATA TRANSFER STATEMENTS
The following example shows this type of WRITE statement with
list and without an I/O list:

I/O

OPEN (22,ACCESS= 'DIRECT' ,id:.:CORDSIZE=3)
WRITE (22,REC=10)A,Z,K
WRITE (22,REC=12)
In this example, the first WRITE statement writes the values of the
variables A,
Z,
and
K to record 10 on logical unit 22.
The second
WRITE statement writes one logical blank record
to
record
12 on
logical unit 22.
'rhe alternative form of this ~ype of WRITE statement operates
in
the
same way as the first forms.
The difference between the forms is in
the way that the unit and the record are specified.

The alternative form of this ::;tatement is:
WRITE

(un'rnr,ERR=sl [,IO:,TAT=ios]) [iolist]

In this form, the unit and
~ecord
references do not contain
the
keywords UNIT= and REC=.
In,;tead the unit number is sp2cified first;
a single quote (') is
specified next;
then
the record
number
is
specified last.

10.7

REREAD STATEMENT

The REREAD statenent causes
the
last
record
read
from
the last
sequential formatted READ or ~CCEPT statement to again be accessed and
processed.
You cannot use th,= REREAD feature until
an
input
(READ)
transfer has been accomplishe,j.
You can use the REREAD statement only
for sequential-access formatt,?d data transfers.
The REREAD statement
can be used with both FORM,\T-statement formatting a.nd list-directE:'d
formatting.
Once a record has been access,?d by a
formatted
READ statement,
the
record
transferred can be reread as many times as desired.
You can
use the same or a new format 5pecification for each successive
REREAD
statement.
Table 10-,6

summal~izes

all the forms of the REREAD statement.

10-37

DATA TRANSFER STATEMENTS

Tahlt, lO-():

Summary of IU~I{I'~AD Stait'nwllt "onns

r-'----. ~ )~~~ Access

:)L:tH'm('ui (

\_.__.._----

. ----"'-' ._-_.

i Sequential Formatted
r

onstrucl

'i l.l\ 1. \I) . I- 1\1'1"

1\1<: N I)~ Ii Y IW

s 11.1 ()STAT

. ..

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

ios III IOlist I

FORMAT Statt'JIlenU

f-.-.---------.---.- . -. .

I S(·quential Fornwttt'd
i

LI,,{ Din'("\.I'(jl

I--K-t~;;
ic: iist-riin'cipd IClrtllal j II~!'. 1I,j,,;t ic: (llll iOl1:1i ,:--;.'ci IIlI1 I iJ 4.;).! I.

FMT

;():-iTAT-

li1.7.1

IOC:

Sequential

FORMAT-Statement REREAD

Tllis section describes
:::;::a tcmen t •
'l' '1 e

fir s t

for m

I~EI~F:AD

0 f

the

secuentlal-dccess

t h 1 s s t ,:, t em e n t

(FORMAT-statement)

REREAD

J S :

(FMT=tl.END=sl 1.ERR=sll,lOS'l'AT=ioSI) i i o l i s t j

It' the I/O l i s t is snecifled in ttlis
Lorm,
t'-le
preVIOUS
record
is
l?:"C1nsterred trom the loalcaL urllt (soecllleeJ In ttle preVIOUS tormatted
READ statement) to the memory locatIons specItled by the
elements
in
tt1e
I/O
list.
The transterred record IS tocrndtted according to the
:~':)RMAT soecifications oiv(~n
in f.

rt you omit the 110 l i s t from thIS
stdternent,
the
Input
record
is
skipged.
(If the FORMAT st~tement specIfies 31ash editInq, more than
::)ne record can be skipped.
H or apostrophe
editIng
can
cause
data
transfers to occur to the t'vkf'i!AT statement Its(~lt.
See SectIon 12.4.)

Ule

second
first
t'ORMAT

form at this R~R~AU statement operates In the
form.
The diLletence beLweel1 tile two torms
soecitiers are exrnf.l.sspd.

],'10

second

form of

rile

ln~l~EAD

this

SLdL(.l.lI1l-:'IIT_

same
way
as
is In the way

I.S:

t[,iolistJ

not
this forrn, the keyword (oun oj thp i'ORMAT sJecltier (t'M'I'=)
IS
in
the
t'OkMAT ref ,,::,rence.
Whenever YOU use tne Keyword torm of
t:1is specifier, you must. enclose the keywurd l l s t in parentheses.
IlJ

;.c;c·d

10-38

DATA TRANSFER STATEMENTS
The following example shows tte formatted REREAD being used:

r,
10
15

CHARACTER J*5
DIMENSION J(5)
READ (20,5)A,X,I
FORMAT (2F10.2,I5)
REREAD 15,J
FORMAT (5A5)
END

In the above sequence, statemEnt 1 reads the two real varIables A and
and the integer I.
StatemEnt 10 rereads the last record input from
unit 5 as a character string cf 25 characters, five per word, and puts
five characters per element irto the array J.

10.7.2

Sequential List-Directed REREAD

'rhis section describes the
statement.

sE·quential-access

(list-directed)

REREAD

The first form of this statemE'nt is:
F~EREAD

(FMT=*r,END=s] [,Ef'R=s] [,IOSTAT=jos)) [iolist)

In this form, the last record read by a formatted
READ statement
is
transferred
from
the
loqic"l
unit (specified in the formatted READ
statement) into the memory locations identified by the elements of the
I/O
list.
Since the formatting is list-directed, the format of the
data is controlled by the date types of the elements in the I/O list.
If no I/O list is included, no data is transferred.
The second form of this statement operates in
the
same way as
the
first
form.
The difference between the two forms is the way in which
the formatting is specified.
The second form of this statement is:
I~ ERE AD

* [,i 0 1i s t )

The following example shows
statement:
HEAD

(L:O,*)
*, B

1:he

list-directed

form

the

REREAD

I~EREAD

END
In this example, the READ stai:ement reads data from
logical
unit
20
into variable "A.
The REREAD statement rereads the data from logical
unit 20 into variable B.

10-39

DATA TRANSFER STATEMENTS

r------------.---J

l__~~:em_=:____ _
ACCEPT

10.8

ACCEPT STATEMENT

The ACCEPT statement enables you to input data from your terminal into
memory.
You can use the ACCEPT statement only for sequential-access
formatted data transfers.
This
statement can be
used
with both
FORMAT-statement and list-directed formattina.
Table 10-7 summarizes all forms of the ACCEPT statement.

Summal'~' of ACCEPT Statement Forms

Table ]()-7:

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

Data Access
- - - - - - ----- ----- ----- --S('4Uenlial Formatted
I FURl\lAT Statement)

Statement Consh'ud

~--

ACCEPT(FMT fl,ENJ)
ACCEPT fI,ioli~tl

,,11.Elm

c.II.1()STAT

io" ill iol i"j I

I (List DirC'ded'

.1-:1\'1)

"11.Elm

;-;II.lOSTAT

io" III iolist I

ACCEPTI.ioli~t I

~ Sequential Forlllatt~J--- --- ---- ACCEPT( F\JT

"-'---"--1

--"---"1

Key:
FMT--{,

is FOf{:\lAT-statenwnt forma11inl!: inlist is optional I.S('ction lO.4:)])

is list-directed formatting; iolist is optional ISt'ctiol1 104 ..'1.2l.

END

is an optional end-of-fil(' transfer "pl'cifipr IS,'ction lO.4.f)).

EHH ;-;

is an optiol1a I error transfl'r spccifi('l' IS('ction 10...1-.71

IOSTAT - ios

is an optional I,n slatus specifier 'Section lO.4.H).

101 ist

is a dala transfer 10 list (S('ction lO49!

l ___ ...._._..._.___ .. __

10.8.1

Sequential FORMAT-Statement ACCEPT

This section describes the sequential-access
~~ t a temen t .

(FORMAT-st~tement)

The first form of this statement is:
ACCEPT

(FMT=f[,END=s] [,ERR=s] [,IOSTAT=ios]) [iolist)

ff you include the I/O list in this form, the data is taken
from
the
terminal
and
stored
in
the memorv locations identified in the I/O
list.
The transferred data
is
formatted
accordina
to
the
FORMAT
specifications given in f.

10-40

DATA TRANSFER STATEMENTS
vou omit the 1/0 list from this form, the input record is
skipped.
thE~ FORMAT statement soecifies slash editing, more than one record
CLln bf~ skipped.
H or apostropne editing can cause data
transfers
to
occur to the FORMAT statement itself.
See Section 12.4.)
If

(If

'['h0 spconri form of this statem,~nt operates in
the
same
way
as
the
first form.
The rlifference between the two forms is in how the FORMAT
l"p"!"prence is expressed.
"'11 ('

S (> con d

for III

'\lCF:PT

t his s tat e m ~ n t i s :

f r • i 01 is t 1

in this form,
omItted.
':'11('

0 [

the keyword portion of the FORMAT

specifier

(FMT=)

followinq example sho'Ns both forms of the FORMAT-statement ACCEP'r.
(FMT=3S)A,Z,J
FORMAT (2FlO. 2, IS)
t'.·,::::CEPT J 5, B
FORMAT (FIO.2)
ACCEPT

lr,

"\J D

)n this example, i::he first ACCEPT statement accepts the values of
the
variables
A~
Z,
and
J
from
the
terminal
in
th~~ form of FORMAT
~;tatement 35.
Thf~
second
AC(~EPT
statement
accepts
the
value
of
variable P, from the termi~')al i1 the form of FORMAT statement 15.

10.8.2

SequentiaJ List-Direct·:?d ACCEPT

'1'hEO' list-directed ACCEPT statement transfers
data
entered
from
the
t.crminal
into variables specified in the I/O list.
The formattinq of
the transferred dilta is controlled by the data types of the
items
in
t he I /0 1 i st.
T rH~

fir s t

form

!\ ( :C E f-' T

0 f

t his

s tat em el t i s :

( F M T= :', r , END = s 1 [ , ERR

=s] [, lOS TAT = i 0 s] ) [i ali s t ]

In this form, the data is
tra1sferred
from
the
terminal
into
the
memory
iocations
identiried
in the I/O list.
Since the transfer is
"i1,st-clirected, thE:' data is formatted according to the
data
types
of
lhe
items
in
the
I/O
list.
If no I/O list is inc]uded, a line is
:: k i ppea •

The second form oJ:: this statpm~nt operates in
the
same
:l,st
fcrm.
'rhe
dlff('renr,::>
in
the
two
forms
is
! ist-directed iormattina refer~nre is specified.
Thf> spcond
V'C E P'P

form of this

starpm,~nt:

is:

* I • i o:t is t 1

10-41

way
as
in
how

the
the

DATA TRANSFER STATEMENTS
In the following example,
both
forms
of
the list-directed ACCEPT
statement are used to take information, character-by-character, from
the terminal.
This example additionally shows the list-directed TYPE
statement being used to print the ACCEPTed data at the terminal:
PROGRAM ACCTST
ACCEPT *,I,J,K
TYPE *,K,I,J
ACCEPT (FMT=*)G,H,F
'P Y P E

*, H , F , G

END
EXECUTE ACCEPT.FOR
FORTRAN: ACCEPT
ACCTST
LINK:
Loading
[LNKXCT ACCTST execution]
23456 9876 12
12, 23456, 9876
12.34 98.16 789.67
98.16000, 789.6700, 12.34000
CPU time 0.2
Elapsed time 40.4

_ 0 _____0_]
__"__"
TYPE
Statement

10.9

TYPE STATEMENT

Use
The TYPE statement enables you to output data to your
terminal.
the
TYPE
statement
only
for
sequential-access
formatted data
transfers.
This statement can be used with both FORMAT-statement and
list-directed formatting.
Table 10-8 summarizes all the forms of the TYPE statement.

10-42

DATA TRANSFER STATEMENTS
Summar" of TYPE Stalenwnt FI' rt,,~;

Tahle' 1(1-8:

Data Ac("pss
'~'i 'qllPnlla I Form;l1l<·d

I "IJf'~

I-'\j'i

1'1'\'1'1,: n "d, ... i

Form:lttpd
,I.i;;! f)i r('I'j ('0 I

i'l\'I'I-:

,";"IJll!'nl iai

,;;II()~Tj,T

ill~llii"!i,,tl

I I'I{I~ ,..;11 I()STAT

ill"lllillli;;tl

Ii 1<1(f{

i'(lHMXr ~tatt'l1l(>ntl

l'I'Yf'I'; 1-1\1'1
I I,.! ~T:

:\1T

i()STAT

Ill;;

",Ii,,)

iO.9.1

Sequential

FORMAT-Statement

';' his seC' t :t 0 n
'; td temen t.

d E~ S C rIb e s t h (~

fir s t

t his

':' h p

tor mot

~.; po 0 tJ, > n t 1 a ] -- dec f~ s S

TYPE

(FORMAT-statement)

~; t (~ t (> in t 1'1 [ - I S :

( PM'T' = t , ,. P. R R::: s!

i' ':' P I<~

TYPE

! , I OS T A '1':: lOS I ) I 1 01 i s t: I

I f
you inc 1 u d e t h ,"
1 / (j lIS t I n t: i11 S t () r m ~ t he cj ,;j t a l S t r (1 nsf err e d f r om
t-hr:
I/O
1 i st
tr)
your
h->rmi nal.
'rhe transfprred data is formatted
ICCOrolnq to the t'ORMAT SPPcltl,'atlons aiven in t.

it' the· 1/0 -I ist

hlank

record is written
foLlo'vJinq ,is true:

'he

FORMAT

<'tHe>

staternE'nt

not-

st)pc;ifipd~

and

one of

emntv.

1'::;

:Jo

sldsr1.

or <-1L'()strool1e eciltinq descriptors occur alone.

; J0

s 1 .') S h.

0 r

,()f)

<1 D () S t

Ll.'-+

ior

r n 0 rl 1'''

e cj 1 t- j 1'1 q
d p s (~ r 1 p t- 0 r s .

;lIor'p

riescrlntors

Jnrormatl.on

i'ormat:

precede

the

editinq)

spcond form ot thIS statement operates in
the
same
way
as
the
irs t form.
The d i [ f EO rel1('e be h<Jpen t he two forms is in how the FORMAT

Thp
i

sC"C'ond

Th(='

torm at

this

statement,

lS:

I,iolistl
fn

this

'.)n1 i t

form,

the

keyword

portiul1

tpo .

10-43

the

FORMAT

specifier

(FMT=)

" ,~
I

DATA TRANSFER STATEMENTS
The following example shows both forms of the FORMAT-statement TYPE:
A=11.4
2=13.9
.1=5

K=lO
T YP E ( FM T =5 ) A , Z , J
FORMAT (IX,2FIO.2,I5)
TYPE 15,K
FORMAT (lX.IIO)

~:ND

T~e

following is typed on your terminal upon execution:
LINK:
Loading
[LNKXCT TEST40 execution]
ll.40
13.90
5
10
CPU t I me O. 22
Elaosed time 2.00

10.9.2

Sequential List-Directed TYPE

The list-directed TYPE statement transfers dat~ from a program to
the
terminal.
The formatting of the transferred data is controlled by the
ridta types of the items in the I/O list.
TrJ.e fIrst form of this statement is:
TYP~~ (FMT=* [, ERR=s]

[, IOSTAT= ios]) [iolist]

In this form,
the data
is
transferred
from
the
program
to
the
terminal.
Since the transfer is list-directed, the data is formatted
according to the data types of the items in the I/O list.
If
no
I/O
list is included, a blank record is written.
T~e

second form of this statement operates in
the
same way as
the
fLrst
form.
The difference between
the
~wo
forms is in how the
I lst-directed formatting reference is specified.

The second form of the statement is:
TYPE *[,iolistJ
'fhe following example shows
statement:

both

forms

the

list-directed

1\=11.4

~: l3 • 9
,1=')
~' YP E
( FM T = *) , A , Z , J
TYPE *,K

r:ND

'rho

following

is typed on the terminal upon

Loading
rLNKXCT TEST41 executionl
11.40000,
13,90000, 5

:,INK:

;()

CPU tIme 0.20

Elapsed time 0.87
10-44

ex~cution:

TYPE

DATA TRANSFER STATEMENTS

PRINT
Statement

10.10

PRINT STATEMENT

The PRINT statement transfers data from memory to the line printer.
You can use the PRINT statement only for sequential-access formatted
data transfers.
'This statement can be used with both FORMAT-statement
formatting and list-directed formatting.
Table 10-9 summarizes all forms of the PRINT statement.
Table 10-9: Summary of PRINT Statement Forms
Data Access

Statement Construct

Sequential Formatted
(FORMAT Statement)

:'IW\'T(FMT-lI.ERH

Sequential Formatted
(List Directed)

i'!?lI\T(f<'MT
I.I';IW
PRINT *[ ,iol: st 1

sll.lOSTAT- iosiilioiistl

PRINT f[,iolistl
;.;II,IOSTAT

ios!ilinlisti

Key:
FMT=f

is FORMAT-statement formatting; iolist is optional (Section 10.4.5.1).

FMT = *

is list-dirl~cted formatting; iolis', is optional (Section 10.4.5.2).

ERR = s

is an optional error transfer sp(,cifier (Section 10.4.7).

IOSTAT ~ ios

is an optional 110 status specifi.~r (Section 10.4.8).

ioIist

is a data transfer 110 list (Section 10.4.9).

10.10.1

Sequential FORMAT-Statement PRINT

This section describes the sequential-access (FORMAT-statement)
statement.

The first form of this statement is:
PRINT

(FMT=fr,ERR=-s] r,IOSrAT=-ios]) [iolist]

If the I/O list is included in this form, the data identified by the
I/O list is transferred
from memory to the line printer.
The
formatting of the transferred data
is controlled by the FORMAT
specifications given in f.

10-45

DATA TRANSFER STATEMENTS
A blank record is written if the I/O list is not specified, and one of
the following is true:
1.

The FORMAT statement is empty.

No slash, H, or apostrophe editing descriptors occur alone.

No slash, H, or apostrophe editing
first repeatable edit descriptor.

descriptors

precede

the

See Section 12.4.
The second form of this statement operates in the same way as the
first form.
The difference between the two forms is in how the FORMAT
specifier is expressed.
The second form of this statement is:
PRINT f[,iolist]
The following example shows two PRINT statements; one with an I/O list
and one without:

10
20

A=7.6
B=12.5
C=20.9
PRINT 10
PRINT 20,A,B,C
FORMAT (' Beginning of test')
FORMAT (' Values are:',3F)
END

The following is printed to the line printer upon execution:
Beginning of test
Values are:
7.6000000

10.10.2

12.5000000

20.9000001

Sequential List-Directed PRINT

This section describes
statement.

the

sequential-access

(list-directed)

The first form of this statement is:
PRINT (FMT=*[,ERR=s) [,IOSTAT=ios]) [iolist]
This form of the PRINT statement specifies that the data identified by
the elements of the I/O list is output on the line printer.
The data
is formatted according to the data types of the elements
in the I/O
list.
If no I/O list is included, a blank record is written.
The second form of the list-directed PRINT statement operates in the
same way as the first form.
The difference between the two forms is
in the way that the formatting is expressed.
The second form of this statement is:
PRINT *[,iolist]

10-46

DATA TRAI-JSFER STATEMENTS
The following
statement:

example

shows

the

use

the

list-directed

0=1

E=40
F=23.3
PR I NT *,0, E, l:!'
END
The following is printed to the line printer upon execution:
1.000000, 40.00000, 23.30000

i-PUNCH

L____:~ta:ment ____
10.11

PUNCH STATEMENT

The PUNCH statement transfers data
from memory to
the paper
tape
punch.
You can
use the PUNCH statement only for sequential-access
formatted data transfers.
This statement can be used with both
FORMAT-statement formatting and :List-directed formatting.
Table 10-10 summarizes all forms of the PUNCH statement.
Table 10-10:

Summary of PUNCH Statemenl Forms

Data Access

Statement Construct

Sequential Formatted
(FORMAT Statement)

PUNC H(Fl\lT - fl,ERH - s IIJOSTAT = ios III iolist I
PUNCH fI.iolistl

Sequential Formatted
(List Directed)

PUNCHd,'l\lT ~~I ,ERH"" s II ,10STA'I' - ios Iii iolist I
PUNCH *1. olistl

1---------------

- - - - - - - ""------

Key:
FMT-f

is FOHMAT-statement formattll1g; iolist is optional (Section 10.4.5.1).

FMT ~ *

is list-directed formatting; iolist is optional (Section 10.4.5.2).
is an optional error tram;fer s )ecifier (Section ]0.4.71.

IOSTAT-~ ios

is an optional 1'0 status ::'peci'ier I Section 1O.4.8!.

iolist

is a data transfer 1'0 list (Sec~ion I0.4.9).

10-47

DATA TRANSFER STATEMENTS
10.11.1

Sequential FORMAT-Statement PUNCH

ThIS section describes the seauential-access
,; i . , temen t.
TrlP

first

(FORMAT-statement)

PUNCH

form of this statement is:

PUNCH

(FMT=f[,ERR=s] [,IOSTAT=ios]) [iolistl

the I/O list is specified in this form, the data identified by the
in
the
I/O list are transferred to the paper tape punch.
The
formattinq of the data is
controlled
by
the
FORMAT specifications
(J I ven in f.
i c'_'ms

\ blank record is written if the I/O list is not specified, and one of
the followinSl is true:
1.

The FORMAT statement is empty.

No slash, H, or apostrophe editing descriptors occur alone.

No slash, H, or apostrophe editinq
first repeatable edjt descriptor.
~ee

descriptors

precede

the

Section 12.4.

The second form of this statement operates in
the
same way as
the
First
form.
The difference between the two forms is in the way that
the format specification is referenced.
rhe second
PUNCH

form of this statement is:
f[,iolist]

The followinq example shows the formatted PUNCH statement:
i?UNCH lO,A,B,C
F'ORMAT (iF)

to.ll.2

Sequential List-Directed PUNCH

Thls section describes

the

sequential-access

(list-directed)

PUNCH

:; t 'i t em e n t .
T~e

first
~')UNCH

form of this statement is:
(FMT=*[ ,ERR=s] [,IOSTAT=ios]) [iolist]

Phis form of the PUNCH statement transfers the data identified by the
,"oments of the I/O list to the paper tape punch.
Since the transfer
i:i list-directed, the formatting of the data is controlled by the data
:-Yges of the items within the I/O list.
If no I/O list is includE~d, a
~]ank record
is wrItten.

10-48

DATA

TR~NSFER

STATEMENTS

The second form of this staterr,ent operates in the same way as the
first
form.
The difference between the two forms is in the way that
the list-directed format reference is written.
The second form of this statement is:
PUNCH *[,iolist]
The following example shows the list.-directed PUNCH statement:
PUNCH *,D,E,F

INTERNAL FILES AND
ENCODE/DECODE
Statements

10.12

INTERNAL FILES AND ENCODE/DECODE STATEMENTS

Internal READ/WRITE statements dnd ENCODE/DECODE statements
for internal I/O.
Table 10-11 summarizes all the forms of the
ENCODE/DECODE statements.
Table 10-11:

internal

are

used

READ/WRITE

and

Summary of Internal READ/WRITE and
ENCODE'DECODE StatemeD1 Forms

Data Access

Statement Construct

Sequential Formatted
(FORMA T Statement)

l:t,C()OBk.f.al.BW{o s II,lOSTAT ioslJI iolist I
DECODE(! .Lal.BHJ{ -- s :1.l0STAT - ios III iolist I

READ(UNIT = un,FMT = fl ,END = s][ ,ERR = s If ,lOSTA T = ios ])[iolist]
READ(un,FMT = fl,END = s][,ERR = slI,IOSTAT= ios])[iolist]
READ(un,fl ,END = s]1 ,ERR = s]! ,IOSTAT = ios])[iolistl
WRITE(UNIT = un,FMT = fl,ERR = s][ ,IOSTAT = ios])[iolist]
WRITE( un ,FMT = fl ,ERR = s][ ,lOSTA T = ios])[ iolist1
WRITE(un,fl,ERR = sH,IOSTAT = ios])[iolist]

Key:
UNIT=un

is an Internal File identifier Section 10.4.3.2).

is the total number of characters being transferred.

is a FORMAT-statement formatting reference.

is the name of the array from which or to which data is being transferred.

END=s

is an optional END-of-file specifier.
is an optional error transfer Hpecifier (Section 10.4.7).

JOSTAT- iOB

is an optional 110 status specifier (Section 10.4.8).

iolist

is a data transfer 110 list (Se:::tion 1O.4.B).

10-49

DATA TRANSFER STATEMENTS
10.12.1

Internal READ and WRITE Statements

The internal READ statement transfers data from an
internal
file
to
I/O list elements.
The internal WRITE statement transfers data from
I/O list elements to an
internal
file.
Internal READ and WRITE
statements are always formatted.
NOTE
The DECODE statement can he used as an alternative
to
the
internal READ statement, and the ENCODE statement
can be used as an alternative to
the
internal WRITE
statement.
(See Section 10.3.1.1 for more information
on internal files.)
The internal READ statement has the following forms:
READ

(UNIT=un,FMT=f[,END=s] [,ERR=s] [,IOSTAT=ios]) [iolist]

READ (un,FMT=f[,END=s] [,ERR=s] [,IOSTAT=ios]) [iolist]
READ (un,f[,END=s] [,ERR=s] [,IOSTAT=ios]) [iolist]
In the above forms, un is an internal
10.4.3.2) •

file

identifier

(see

Section

If an I/O list is included in these forms, it specifies that data
is
transferred
from internal file identifier, un, formatted according to
the specification given by f, and transferred into the elements of the
specified I/O list.
If an I/O list is not included, the input record is skipped.
(If the
FORMAT statement specifies slash editing, more than one record can be
skipped.
Apostrophe or H editing can cause data transfers to occur to
the FORMAT statement itself.
See Section 12.4.)
The following example
statement:

10
20

demonstrates

CHARACTER*9 STRING
STRING = '3.14 6.02'
READ(STRING,lO) PI, A
FORMAT(F4.2, IX, F4.2)
WRITE(5,20) PI, A, PI+A
FORMAT(' PI=', F6.3, 5X,
STOP
END

the

use

the

internal

'A=', F6.3,

5X,

'PI+A=', F6.3)

READ

The READ statement in this example
is an
internal
file
read.
It
extracts the two numbers that are encoded in the character variable
STRING, converts the numbers to floating point, and then stores them
into the two variables PI and A.
The following is printed at the
terminal when the above program is executed:
EXECUTE IR.FOR
LINK:
Load ing
[LNKXCT IR execution]
PI= 3.140
A= 6.020
PI+A= 9.160
CPU time 0.19
Elapsed time 0.40

10-50

DATA TRANSFER STATEMENTS
The internal WRITE statement has the following forms:
WRITE (UNIT=un ,FMT=f [, ERR=s] [, IOSTAT=ios] ) [iol ist]
WRITE (un,FM'r=f[,ERR=s] [,IOSTAT=ios]) [iolist]
WRITE (un,f[,ERR=s] [,IOSTAT=ios]) [iolist]
If the I/O list is included in these forms, the data specified by the
elements of the I/O list are output to a file on internal file
identifier un.
The output data is formatted in this file according to
the FORMAT specifications given in f.
A blank record is written if the I/O list is not specified, and one of
the following is true:
1.

The FORMAT statement is empty.

No slash, H, or apostrophe editing descriptors occur alone.

No slash, H, or apostrophe editing descriptors
first repeatable edit descriptors.

preceded

the

See Section 12.4.
The following example demonstrates
statement:

10
20

the

use

CHARACTER*20 CHARS(3)
INTEGER PHNE(3)
PHNE(l) = 617
PHNE(2) = 481
PHNE(3) = 4054
WRITE(CHARS,10) (I, PHNE(I), 1=1,3)
FORMAT ( 'PHNE(', 11, ')=',14)
WRITE(5,20) (I, CHARS(I), 1=1,3)
FORMAT (' Record " 11, I of CHARS is'"
STOP
END

the

internal

WRITE

A20, I " ' )

The first WRITE statement in the above program is an
internal file
write.
Since the characteI
variable being written to is a three
element array, the internal file is a file of three records.
When
this program is executed, the following is output to the terminal:
EXECUTE IW.F'OR
LINK:
Loading
[LNKXCT IW execution]
Record 1
Record 2
Record 3
CPU time

10.12.2

of CHARS is "PHNE(l)= 617
of CHARS is "PHNE(2)= 481
of CHARS is "PHNE(3)=4054
0.2,4
Elapsed time 0.82

"
"
"

ENCODE and DECODE Statements

The DECODE statement can be u~;ed as an alternative to an
internal
READ,
and the ENCODE statement can be used as an alternative to the
internal WRITE.

10-Sl

DATA TRANSFER STATEMENTS
'r'he ENCODE statement transfers data from the variables of a
specified
rio list
into a
specified array.
ENCODE operations are similar to
those performed by a WRITE statement.
The DECODE statement transfers data from a specified array
into
the
variables of an
1/0
list.
DECODE operations are similar to those
performed by a READ statement.
ENCODE and DECODE statements have the following forms:
~;NCODE

(c,f,a[,ERR=s] [,IOSTAT=iosl) [iolist]

Jt-':CODE

(c,f,a[,ERR=s] [,IOSTAT=ios]) [iolistl

where:
specifies the number of characters
in each
internal
record
of the array.
This argument can be any integer
expression, and must be the first specification in
the
stntement.

NOTE
~ivp
~rray

characters per word
are
stored
in
the
without regard to the type of the array.

~pecifies

either a FORMAT-statement or a numeric array
1:hat contains format specifications.
This must b,e the
second specification.

specifies
the array,
array element,
variable,
or
character substrinq reference that is to be used in the
transfer operations, and it must contain
at
least c
characters.
More than one element of the array can be
used by the ENCODE/DECODE.
iolist

specifies an 1/0 list of the standard form.

multiple records are stored by ENCODE or read by DECODE, each new
record
starts c
characters nfter
the
previous
record;
no CR/LF
(carriage return/line feed) is inserted between records.

~hen

NOTE
the array contains fewer characters
than
required
the
format
and
the
I/O
list,
the variables
followinq the
array
in memory are used.
If
the
processing of the I/O list requires more characters in
1 single record
than are specified
by the character
count c, blanks are used.
rf
by

l'he fo1lowinq example shows how the ENCODE and DECODE
Jsed:

DIMENSION B(4) ,A(2)
A(l)=300.45
A(2)=3.0
C='l2345'

no 2 ,J=1, 2
:0

I'; NC ODE (1 6 , I 0 , B) J , A ( J )
FORMAT (IX. 'A ( I , I l, I)

10-52

FB.2)

statements

are

DATA TRANSFER STATEMENTS

s
11

'rYPE 11,B
FORMAT(4A5)

CONTINUE

DECODE(5,12,C)B
FORMJI.T(3Fl.0,.lX,F: .• O)

TYPE 13,B
FORMAT(4F5.2)
END

During the first iteration of the DO loop, the
ENCODE
statement has
transferred
the contents
o~
variable J and array element A(l) into
array B.
rrhe formatting of the data being transferred is specified by
the FORMAT statement at lineLO.
After the first iteration of:he DO loop, the contents of array Bare:
A (1) I

13(1)
R (2)
B (3)

'300.4 '

13 (4 )

I
I

The TYPE statement at line 5 :ypes array B on the terminal during
first iteration of the DO Loop.

the

During the second iteration of the DO loop, the data
IS
transferred
from variable J and array element A(2) into array B.
Atter the second
iteration, the contents of array Bare:
B (1)

A (2) ,

13(2)

I3 (3 )
13(4)

3 .. 0'

The TYPE statement at line 5 types array B on the terminal during
second iteratjon of the DO 10Jp.
The DECODE statement:
1.

Extracts the digits 1,

Converts them to fl03ting-point values

Stores them in

Skips the next character

Extract.:;; the digit 5 from C

Converts it to a

Stores the value in 8(4)

B( L),

2, and 3 from C

B(2), and B(3)
(the digit 4)

floating-point value

10-53

the

DATA TRANSFER STATEMENTS
The following shows what is printed at the
program is executed:
EXECUTE T.FOR
LINK:
Loading
[LNKXCT T execution]
A(l)
300.45
A(2) =
3.00
1.00 2.00 3.00 5.00
CPU time 0.1
Elapsed time 0.8

10-54

terminal

when

the

above

CHAPTER 11
FIL8-CONTROL AND DEVICE-CONTROL STATEMENTS

11.1

FILE-CONTROL STATEMENTS

Prior to transferring any data using one of the forms of data transfer
statements,
you can establish a connection between a logical unit and
a file by using the OPEN statement.
After the completion of
a data
transfer,
you can
terminate the connection between the logical unit
and the file before ending the program by using the CLOSE statement.
The OPEN statement enables you to explicitly connect a logical unit to
a file prior to the first data transfer, and also to specify a variety
of characteristics about the file and the data transfers.
After the last data transfer is completed, the CLOSE statement enables
you to explicitly disconnect
the
logical
unit from the file and,
optionally, to specify a variety of characteristics about the CLOSE.
If you do not precede an I/O statement with an OPEN statement,
FOROTS
automatically performs an "implicit OPEN" (see Section 11.2.1).
Similarly, if you do not specify a
CLOSE statement
to
explicitly
disconnect a
file
from a logical unit, FOROTS performs an "implicit
CLOSE" (see Section 11.4.1) when your program terminates.
You need not specify the OPEN and CLOSE statements
if
performed by the implicit OPEN or CLOSE are satisfactory.

the

actions

OPEN
Statement

11.2

OPEN STATEMENT

The OPEN statement is used to specify characteristics of a
file
you wish to read or write.
An example of an OPEN statement is:
OPEN

that

(UNIT=20,FILE='MYDATA.DAT')

The specifiers inside the parentheses give information about the
and determine how the file is opened.

11-1

file

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
The UNIT specifier (in the example above, "UNIT=20") is required in an
OPEN statement.
All other specifiers are optional, including the FILE
specifier in the example shown
above.
You
can supply many other
optional
specifiers
(see Section 11.3
for
a
description of OPEN
statement specifiers).
The order in which the specifiers appear
does
not affect the execution of the OPEN statement.
By using
the
OPEN specifiers,
you are
able
characteristics of each data transfer, including:
1.

The name of the data file

The type of access required

The data format of the file

The disposition of the data file

The data file record and block sizes

define

certain

In addition, a DIALOG argument permits you to establish a dialog
mode
of operation when the OPEN statement containing it is executed.
In a
dialog
mode,
interactive
terminal/program
communication
is
established, enabling the user to define or redefine the values of the
OPEN statement specifiers.
When a file is open for output (STATUS='NEW'
or
ACCESS='SEQOUT'),
a
null
file
is created on the device specified by FILE=
OEVICE=, or
if none, the first structure in the job's search list.
An OPEN statement is referred to as a
both of the following are true:

"deferred"

OPEN

•

The OPEN statement specifies STATUS='UNKNOWN'
specify a STATUS value).

•

The
OPEN
'SEQUENTIAL'

statement
(or

does

statement
specifies
ACCESS='SEQINOUT'
(or does not specify an ACCESS value).

if
not
or

The actual opening of the
file
is deferred
until
the
first
data
transfer
statement
(READ,
WRITE, PRINT, PUNCH, or SKIPRECORD).
The
actual opening of the file means the determination of
the
physical
device,
and
for
TOPS-20,
the generation number (if not explicitly
specified) .
If the first data transfer statement is
a
READ or
SKIPRECORD
the
first
file
that matches
the
file
specification given in the OPEN
statement is opened.
If
no
file
exists
that matches
the
file
specification given, a null file is created on the device specified by
FILE= or DEVICE=, or if none, the first structure in the job's search
list.
The file is positioned as if a READ or SKIPRECORD statement had
been executed, and an end-of-file error will be generated
(see
END=,
Section 10.4.6).
If the first data transfer statement
is
a WRITE,
PRINT,
or
PUNCH
statement,
a
new file
(with
a
new generation on TOPS-20) will be
created on the device specified by FILE= or DEVICE=, or if none,
the
first structure in the job's search list.
If the file specified in the OPEN statement does not exist, and either
a
CLOSE statement
is executed oi the program runs to completion, a
null file is created on the device specified by FILE= or
DEVICE=,
or
if none, the first structure in the job's search list.

11-2

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
11.2.1

Implicit OPEN

When the OPEN statement has not been executed before a data
transfer
that references the unit number, an implicit OPEN is performed.
An implicit OPEN has almost exactly the same effect as if you had
put
an OPEN statement with the following format in the program just before
the data transfer statement:
OPEN

(UNIT=un,STATUS='UNKNOWN' ,FORM=fm)

where:
nu~ber

is the unit
statement.

specified

the

data

transfer

is 'UNFORMATTED' if the data transfer statement
is an
unformatted
RE~D
or WRITE statement; otherwise fm is
'FORMATTED' •

In addition, if the data transfer statement has an ERR specifier,
the
implicit OPEN has this same qualifier included.
This is also true of
the IOSTAT specifier.
NOTE
The default
for
the BLANK specifier
is different
depending on whether the OPEN is implicit or explicit
(see Section 11.3.3).

11-3

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
11.2.2

OPEN on a Connected Unit

If the OPEN statement contains a
STATUS=OLD specifier
(see Section
11.3.29), then its action depends on whether a file is already OPEN on
the unit, and whether the file specified by the OPEN is the same
file
that
is currently on the unit.
If the file specified by the OPEN is
different from the OPEN file, the connected file is closed and the new
file
is opened.
If the file specified by the OPEN is the same as the
connected file, the file is not closed, and the file
pointer
is not
moved.
This action
is not
affected
by the /F66 compiler switch
(described in Chapter 16).

11.3

OPEN STATEMENT SPECIFIERS

All of the OPEN statement specifiers are optional,
except
the
UNIT
specifier,
which
is
required.
Some specifiers have default values
that can depend on the unit number or the values of other specifiers.
Table 11-1 summarizes the
type of value required by
to detailed description"s
specifiers are summarized

specifiers in the OPEN statement and
the
each.
A section number is provided to refer
of each specifier.
The CLOSE statement
in Table 11-5.

11-4

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
Table 11-1:

Summary of OPEN Statement Specifiers and Arguments

Argument

Possible Value

Section

ACCESS=

Character expression \\ ith one of the following values:
~;EqlN',
'SEQOLJT'. 'SEQINOUT', 'SEQUENTIAL',
'DIRECT', ru\ NJ)()M' 'I{ANIHN', 'APPEND'

11.3.1

,\SSO(,IATEV ARIABLE

i:Jll'gl'], variuble or intlg('r array element

11:3_2

BLANK ~

Character expression \\ ith one of the following values:
'NULL', 'ZERO'

11.3.3

HLOCKSIZE

1'1lcgcr ('xpre:.;sioll

11,;31

('haracter exp]'(-'s:.;ion \\ ith ol1e of the following valups:
"( mTRAN' , 'LIST', 'IIEVICE'

11_:U;

'Iwractcr t'xp)'('s~ion \\ iih "me of lhe following valw's:
'GGG'. 'BOO'. 'l(iW)', '6:250', 'SYSTEM'

II ;{'i

DEVICE

Character exprp:.;sion

I\,:U-I

IJ\:\LOG

('haractt'r expl'es~ion

D] HECTOI{Y
,TO I'S 10 i

l-h;ll'<:ictc'r expreS:--lOn

I I_:\. 11

! >I1{E('TOHY
,T()I'S,-2()\

('h:}r<}clt'r pxprcs~\()n

1 LT IL'

r )!SI'()S"~

(h;II'Llcj('r expn'sslOll

IH 1FFEI{(,OUNT( 'AlmlA(~ECONTI{OL

I )J':NSITY

',~()()',

\\ ith one of the following values:
<.\ VE', '[)I':LE'n:', 'I' {lNT', 'KEEP', 'UST, '!'llNClI'.

: 1 , j 1;1

I':XI'UNCE

EHR=

Statement number

11.3.14

FILE c-

Character expression

11.3.15
11,:1 Hi

I'-IIYSIZI':
INITIALIZE

FOHM-

Character expression v:ith one of the following values:
'FORMATTED', 'UNFORMATTED'

11.3.17

IOSTAT

Integer variable or intt~ger array element

11.3.18

Lll\lIT
l\1() I )J.:

('lwracll'J' ('xprl's<-.iol1 v'it h one- of the following value:-:
\:-->('11', 'LINE)), 'BHlA[{Y'. 'IMAGE', 'I)lIMP'

'\ ,\1\11':

('JWl'HctC'l' l'xpr'cs~i()11

1':\ 1)( '1IAI{

,\ character e:,;pl'l':';SIU:l 111 which thl' fir::;t charilctl'l' is

• , ::. ~!

",.:t,d
I',\I{ITY

(-haractc<]' ('xpn'S,'lOll \'"lth one of' the' following v:t1u(':--:
" )I)J)', EVEN'

I'[WTI-:( 'TION
/'()J'S 10\

11-5

1 • .:~2:\

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
'fable 11-1:

Summary of OPEN Statement Specifiers and ArJ,{uments (Coni.)
Possible Value

Argument

Section

!'!{(),[,H'TI()N

:::2.>

Teil'S :ZO'
1{1':-\\)()NLY

HECL-

Integer expression

11.3.27

HH '()IWSIZE
i 1.:\.:.!1i

STATUS-

Character expression with one of the following values:

TYPF

'OLD',
'NEW'.
'SCHATCII'.
'UNKNOWN', 1\ FI':!'· .. ()[':LET!':

I' \PI':F< )I{MAT

('hdl':lCtel' ('xpn'ssi(lll

with OIl(> .)1' till' f'()!!()\\'ing \',d\ll'~:
(·()In.: 1)['I\11' or INI>L'STI{Y

11.:UO

UNIT-

Integer expression

11.:3.:11

11..3.29

:'~\I'[l:\(;E'

li.:U:2

NOTE
Par
compatibility
with
previous
versions
of
PORTRAN-10/20, you can specify a numeric array name as
the value of each of the followinq specifiers:
DIALOG=
DIRECTORY
NAME
When a numeric array name is used, FOROTS assumes that
it
contains
a
string
of characters terminated by a
null character.
In addition, you can specify a numeric variable as the
value
of
the
DEVICE
and
FILE
specifiers.
If the
variable is single precision, FOROTS assumes
that
it
contains
5 characters;
if
it
is double precision,
FOROTS assumes that is contains 10 characters.
The use of numeric array names and
numeric
variables
in
place
of
character
variables
is
a nonstandard
feature.

ACCESS
Specifier

11.3.1

ACCESS Specifier

The ACCESS specifier describes the type of
data
transfer
statements
allowed.
Records within files can be accessed directly (randomly) and
sequentially.
The form of the ACCESS specifier is:
ACCESS

acc
11-6

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
where:
acc

is a character expression having a value equal to
the following:

one

'SEQIN'
'SEOOUT'
'SEQINOUT'
'SEQUENTIAL'
'DIRECT'
'RANDOM'
'RANDIN '
'APPEND '
ACCESS has a number of arguments, each of which specifies a method
of
data
access.
SEQUENTIAL
is
the default
access unless the device
(UNIT) opened is a read-only device, in which case
the default
is
SEQIN.
If
the device
ope~ed
is
a write-only device, the default
access is SEQOUT.
The arguments to the ACCESS specifier are:
SEQIN

(Implies ST~TUS='OLD') The specified
opened
tor
read-only sequential
ACCESS='SEOIN' is specified,
it
is
specifying
ACCESS='SEQUENTIAL'
and
Section 11.3.26).

SEQOUT

and
The specifi~d data file is opened
for
output
sequential
If
access.
the specified file already
exists,
it
is superseded
(TOPS-IO) ,
a
new
or
generat:ion is created (TOPS-20) .

SEQUENTIAL

The specified data file
is opened
for
sequential
access.
Records can be read from or written to the
file in sequential order.
However, when a record is
written
to
the file, it becomes the last record of
the file.
Any data following
that
record
becomes
i naccess i bl

data
file
is
access.
When
equivalent
to
READONLY
(see

,=.

Records can also be written to
the
file
and
then
read,
as
long
as
a
device-positioning statement
(BACKSPACE or REWIND, Section 11.8) is
used
before
the READ statement.
SEQINOUT

Same as SEQJENTIAL

DIRECT

The specified data file
may be
read
from
and/or
written
to
in
units of fixed-length records.
The
record to be accessed next is specified in the data
transfer statement by a record number.
The relative position of each record is
independent
of
the
previous READ or WRITE statement.
The RECL
specifier (see Section 11.3.27)
is
required
for
random-access operations.
You must specify a disk
device when the DIRECT argument is used.

11--7

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
:~'ANDI

(Implies STATUS='OLD ' ) The specified
data
file
is
,:)t·'neo
torread-only direct access.
More than one
j'-;P[
can read the same tile at the
same
time
wi th
·Ti.:!':SS=· RANDIN·.
When ACCESS= I RAND IN , is spec it i ed,
;s l~quivalent to specifying
ACCESS='RANDOM'
and
·L\LJUNLY (see Section 11.3.26).

i'ANDOM

Same as DiRECT

'\.PPEND

The

specItIed
tile
IS
opened
for
sequential
access.
APPEND
is
the same as SEQOUT
.~~~p~ tnat the file is positioned at its end
after
h,
id">EN statement.
Headinq an APPEND mode file is
;~eqal.
R~WIND and BACKSPACE are illegal for files
"t)t-'neu Wltfl APPEND access.
r J tl-'-on1v

ASSOCIATEVARIABLE
Specifier

L._._ ...__._._P,._P ____ .. __ .. ~ .._u . . _N".. __ .. __ ... __,__ ,_··_~._n·.

; '.3.2

ASSOCIATEVARIABLE SpecItier

~his specifier enables you to declare a variabie whose
value
is
the
11umber
ot
the
next
record that will be read from or written to the
L i 1e •
~or example, atter the execution of an OPEN statement and prior to the
[irst data transfer, the associate varIable is set to 1.

a data transfer atter the first record is transferred, the value of
he associate variable is 2.

'f'he form of the ASSOCIATEVARIABLE specItier is:
/\SSOCIATEVARIABLE= Inteqer varIable or integer array element
r~ you are using the ASSOCIATEVARIABLE specitier
in
a
program
that
makes
use
of
the
LINK overlay facility, please read the paragraphs
~ ~l a t
toll ow .
;;~

the variable yOU specify as the ASSOCIATEVAHIABLE

is declared in a
subroutine.
then
that subroutine must be loaded in the root
i ink of the overlay structure.
It the subroutine cannot be loaded
in
~lle root link of the overlay structure, declare your ASSOCIATEVARIABLE
l n a (.;Or-U40N statement
so
that
the
ASSOCIATEVARIABLE will
operate
:):'"operly.
~ORTRAN

The reasons tor these steps are:
~hen

the overlay facility is used to
load
FORTRAN modules,
va ria b 1 e s i n
the
mod u 1 (~ s
are g r 0 up e d wit h the
.-:, 'Jr_ 1 n e
In wh i c n they are declared.
,1 e

10 c a .i

j~;len
.'iE:"/

FuRTRAN subroutInes are loaded bv the overlay
are uivided into sets called overlay links.
'..;!' t:-

1 ink, the 0 n e s p e c 1 LL edt 0 bet her 0 0 t 1 ink,
reSIdent
in memory.
The other overlay links are
ff; 2' m(j r 'l a s
rea u i r. e d •
u

Ve r 1 d Y

i:W~VS
;

facility,

11-8

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
Accessing a file opened with arl ASSOCIATEVARIABLE changes the value of
the specified variable.
If tilis variable is in a nonresident overlay
link when the access is made, J>roqram execution produces unpredictable
res u 1 t s .
M0 reo v E' r ,
t: his
val' i a b 1 e
is
res e t
to Z e roe a c h tim e its
8verlay link is removed froIT! memory.
Only variables declared in routines loaded
into the
root
link are
always
resident.
Variables declared in COMMON statements and those
declared in the main proqram al'e always resident and can always be
l)Sea as an assocIate variable.

For more information Ofl the LINK overlay facility. see
t.he LINK ProgrammE~r' s F~eterence Manual. and Chapter 15

BLANK
Specifier

11.3.3

BLANK Specifier

The
BLANK
specifier
applj,es
only
when
reading
formatted
(FORMAT-statement)
numeric fjelds that have a field width specified.
BLANK enables you to specify how blanks in formatted numeric fields
are treated in a read transfer (either as zero or ignored).
The form of the BLANK specifier is:
BLANK = blnk
where:
blnk

is a character expression having a value equal
'NULL' or 'ZERO'.

either

The arguments to the BLANK specifier are:
NULL

specifies that all blank characters within numeric
formatted
input fields are ignored. The exception is
that a field of all blanks has a value of zero.

ZERO

specifies that all blanks are treated as zeros.

If an OPEN statement is executed and the BLANK specifier is not given,
the default is BLANK='NULL'.
If no explicit OPEN statement js executed before a data transfer on a
unit,
the default is BLANK='ZERO' for all devices except terminals.
For terminals,
the default
js always BLANK='NULL'
regardless of
whether or not the OPEN statement is given.
The BLANK specifier is overridden if a corresponding data transfer
statement references a format, list that contains either the BN or BZ
descriptor.
In this case, the BN or BZ descriptor in the format list
overrides the setting in the OPEN statement until the end of the
format list, or until the settjng is changed within the format list.
(The BN or BZ descriptors are described in Section 12.4.9.)
11-9

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
Example:
OPEN(UNIT=l,~EVICE='DSK',FILE='FOO.DAT',BLANK='ZERO')

READ(I,10)K
10 FORMAT(I5)
CLOSE(UNIT=I)
OPEN(UNIT=I,DEVICE='DSK',FILE='FOO.DAT' ,BLANK='NULL')
READ(I,10)L
CLOSE (UNIT=I)
END
In the above example, if FOO.DAT contains 123bb, K has the value 12300
and L has the value 123.

f""--' . ---. . -. . . . . --

~---''''''''----''-------''--'-'-l

BLOCKSIZE

Specifier

L.__ .._. _ . . _ ......... __. . . . __. .__. . . . . . _ ..... _. __ ......____..-1I

il.3.4

BLOCKSIZE Specifier

The BLOCKSIZE specifier enables you
block size for magnetic tape files.

specify

physical

storage

NOTE
'\LOCKSIZE specltles the physical
record
length,
and
specifies the logical record length.

lU~CL(RECORDSIZE)

The argument is an integer expression, and for CORE-DUMP tape
format,
the value assigned
represents
the
number of words in the physical
~lock.
For INDUSTRY tape tormat, the value represents the number of
hytes
in
the physical block.
(See the TAPEFORMAT specifier, Section
11.3.30.)

':'he form of the BLOCKSIZE specifier is:
;),LOCKSIZE= Integer; expression

BUFFERCOUNT
;pecifier

11.1.5
hE:

BUFFERCOUNT Specifier

BUFFERCOUNT specifier enables you to
llsed in the data transfer.

. lHLers

11-10

define

the

number

I/O

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
The BUFFERCOUNT is the number of pages used in disk transfers, and
ignored for nondisk transfers.

" (~
I .:>

The form of the BUFFERCOUNT specifier is:
BUFFERCOUNT= Integer expression
If a BUFFERCOUNT is not specified, or is assigned a value at zero,
buffercount is four pages.

the

NOTE
If MODE='DUMP'

is specified, BUFFERCOUNT is ignored.

The BUFFERCOUNT specifier
does
not
affect
the
operation
at
program, but it can affect execution time and memory req1lirements.
the
buffercount
specifies
the
maximum
For random I/O,
are
in memory
(not
yet written to dISk)
buffers
which
operations.

the

numbe r o t
during I/O

NOTE
For TOPS-20 extended addressing, all
fit in FOROTS's secticn.

I/O buffers

must

r~-·"---"-11

11.3.6

CARRIAGECONTROL
Specifier

CARRIAGECONTROL Specifier

The CARRIAGECONTROL specifier enables you
to
decide
how
the
first
character
of
each
record er~countered during an output data transfer
operation is
treated.
(Section
12.2.3 describes
carriage-control
specifiers.)
The form of the CARRIAGECONTR(IL speCIfier is:
CARRIAGECONTROL

= cc

where:
cc

is a character expression having a value e<1Ud.L
the following:
'FORTRAN'
'LIST'
I TRANSU\TED'
'DEVICE'

11-11

one

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
_~\~

arquments to the CARRIAGECONTROL specifier- are:
speclties that the FORTRAN data
file
attribute
is
.. ,
In
the file's access information, so that when
'H~
tIle is printed, the
fj_rst
character
of
each
r~~ord
IS
replaced
with a carriage return and the
]rrC~pondlng
printer-control
vertical
motion
'Jaracter(s)
(see
Table
12-3).
The
record
i l~rmindtor
at
the
end
of
the
record
will
be

r.I:-''1'

specifies that the first character is output with no
;- " pia c e men t .

TRANSLATED

speclties that the first character of each record is
:~Dlacea
wIth
a
carrIage
return
and
the
, orresponalng
printer-control
vertical
motion
-,en deLer (s) (see Table 12-3).
No record terminator
, written at the end of the record.
c,- .-} t e t hat the 1 a s t d a t a r e cor din the
f i 1e
has
no
"=lcJinq
carriaqe
return/line feed sequence unless a
'~aflk record
is written.
O[

i lxed-ienuth

tiles,

TRANSLATED

treated

,. c· 'T"
i
•
~.)

specifies that the
first
character
will
only be
i~pldce8
for a carriage-control device (such as LPT
() [

:'he default value

l'TY).

is CARRIAGECONTROL='DEVICE'

1 inp
nrintp.r
software
aSS1]meS
that
the
first
(;f1-':1rdcter
at
all
data
files
is a carriage-control
,:ildracter It the fIle has the extension .DAT or if the
:!TLE:FORTRAN switch is specified.

'!hp

DENSITY
Specifier

,3.7

DENSITY Specifier

';'he DE:NSITY specifier is ionored except when used with magnetic
tape;
lr
Germlts
you to specity the tape density.
If you do not specify a
'_Jpe densIty, FORTRAN assumes that you have set the density at monitor

ievel

"hp

tor~

that
ot the

you

are

D~NSITY

satisfied

specifier is:

11-12

with

the system default for the

FILE-CONTROL ANI> DEVICE-CONTROL STATEMENTS
where:
dens

is a character expression having a value equal. to
the f.ollowinq:

one

for

the

'200'
'51)6 I

'800'
'1600'
'G250'
. :;YSTEM'
SYSTEM specifies that the den:)lty
magnetic tape device being us!~d.

JLS

the

detault

denslty

0r:VICE
: ;[}ecifier

11.3.8

DEVICE Specifier

The DEVICE speciJ~ler enablt~s IOU to specity tne
name
oJ.:
Lfle
deVIce
involved in the data transter.
A loqlcal name always taKes precedence
over a physical name.
The DE~ICE arguments can
specity
1/0
devlces
located at remote statlons, a3 weil as loglcal deVIces.
The form of the DEVICE specltier is:
DEVICE= Character expreS.310n
It you omit this option, tne logical name un (wnere un IS tne
decimal
unit
number)
is
tried.
IE this is not successtul, the standard
( d e fa u 1 t) de vic e ass 0 cia ted wit h the un i t I S use d ( see '1 a ole 1 U- 3) •

;J~!'\LOG

~p:edfier

11.3.9

DIALOG Specifier

The DIALOG specifier enables you to type In additIonal
when the OPEN statement is actually executed.

UP~N

speciiiers

If the DIALOG specifier is found In an OPEN llSt, tnen eacn
time
the
OPEN
statement
is executed FUHU'l'S suspenejs executIon dnd prompcs t:.he
terminal with an asterisk.
The torm of the DIALOG specltier is:
DIALOG

11-13

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

Yuu
can
respond
to
the
asterisk
prompt
by
entering
a
file
;',pecification,
DIALOG
switches
(see below), or a file specification
followed bv DIALOG switches.
The file specification may
be
a
full
file specification including the device, directory name, and so on.
i'JUTE
i) I A L OG s wit chi san v 0 PEN s p e c i fie r
( ex c e p t
D I AL OG ,
CIALOG=, UNIT, NAME, FILE, IOSTAT, or ERR) preceded by
d
::; 1 ash
(/).

;"-,r example, when FOROTS enters DIALOG mode, you
:3uch as:

can

type

string

*DSK:FOO.BAR/MODE:BINARY/ACCESS:DIRECT

DIALOG=
Specifier

J 1 .3.10

DIALOG= Specifier

The DIALOG= specifier enables you to include all or a portion
of
the
·,:?EN
speclfiers
in
a
character
expression.
The
contents of the
character expression are interpreted as if yOll had
given
the
DIALOG
:pecifier
(see
above);
and, when the asterisk prompt was given, you
liad typed in
the
same
strinq
as
is
contained
in
the
character
0xpression, followed by a carriaqe return.
The form of the DIALOG= specifier is:
~) tAL 0(; =

C h a r act ere x pre s s ion

L:xample:
OPEN(UNIT=l,DIALOG='DSK:FOO.BAR/ACCESS:SEQOUT/MODE:ASCII')
L c:22

!')O

l'JnITE(1,100)I
FORMAT(I5)
;~ND

DIRECTORY (TOP5-10)
Specifier

1 j . L 11

DIRECTORY Specifier

(TOPS-IO)

'in TOPS-lO, the DIRECTORY specltier is iqnored except for disk
files.
!t
Sp8cltles
the
location
of
the
User
File Directory (UFD) and,
.) EJL. ion a 11 y, the Sub F i leD ire c tor y ( S FD), e i t h er a f wh i c h can
can t a i n
, }!'~
fi le snecified in the OPEN statement.

11-14

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

c •
1· ""

'l'he form of the DIRECTORY specitier
U1RECTORY:=:

Character express10n

The UFD is the di~ectory in which a user's files are stored;
the
SFD
exists
within
the
UFD.
An
SFD
is often used to group files into
::cpdrate subdirectories.
The following

is a sample of the UFD and SFD specification:

lO,7,SFDA,SFDB
In the sample specification, 10,7 is
the
project-programmer
number.
'I'his
is an adequate directory specification if the file is in the UFD
~;peci[ied by 10,7.
The SFDA and SFDB specify two
levels
of
subfile
(lirectories.
The complete directory specification indicates that the
! lIe is located in subtile directory SFDB.
As indicated, the path
to
SFDB is through the UFD lU,l and through the SFD SFDA.
NOTE
Hefer to
the
TOPS-IO
Mon1tor
Calls
o f- d ire c t o--r i e s
complete
description
dIrectory structures.
The follo\ving

Manual
tor
a
and multilevel

is an example of a character expression specification:

1'[ HECTORY= t 10,7, SFDl, SFD2, SFD3·
- ....-. ~.~

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

---

~---.

Project
SubFile
programmer Directory
Number
Path
f 0 11 0 win 9 i 5 a n ex am p 1 e
individual elements:

'1' he

0 tho w

to ass em b l e a

s pe c 1 tic a t ion

f r om

(:HARACTER* 1 0 PROJ, PROG, PATH.l, PATH2
(:HARACTER* 1 COMM
P R O,J ==

i?HOG:=

I I 0 I
17 1

:'/:.,TH== I SFDA I
PATH= 'SFDB'
l'OMM:= I
llPEN(UNIT=l,DIRECTORY=PRtJ//COMM//PROG//COMM//PATHl//COMM//PATH2)
Th~

above specification

equivalent

,:~ x pre s s ion:

l 0, 1 , SFDA, SFDB I

11-15

the

following

character

--------]

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

DIRECTORY (TOPS-20)
Specifier

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

11.3.12

DIRECTORY Specifier

(TOPS-20)

On TOPS-20, the DIRECTORY specifier is ignored except for disk
files.
The DIRECTORY specifier enables you to define the path through the
directory structure to a file specified in the OPEN statement.
The form of the DIRECTORY specifier is:
DIRECTORY=

Character expression

The argument to the DIRECTORY specifier
is
3
character
expression
whose elements comprise the directory path specification, for example:
OPEN(UNIT=22,DIRECTORY= 'GUEST')
!Looks for DSK:(GUEST)FOR22.DAT
or
CHARACTER*12 ID
1D= 'GUEST.CLASS3'
OPEN(UNIT=22,DIRECTORY=ID)
!Looks for DSK:(GUEST.CLASS3)FOR22.DAT

---···__· __···-_.'" .-. . . . ._. . ._. _. __._. . ."._DISPOSE
r
I

""J'

Specifier

L ___.__.__._____
11.3.13

DISPOSE Specifier

The DISPOSE specifier enables you to specify an action to
the file is closed.

occur

when

The form of the DISPOSE specifier is:
DISPOSE = dis
where:
dis

is a character expression having a value equal to
the following:
'KEEP'
'SAVE'
'l)ELETE'
'2XPUNGE'
'PRINT'
I

[.1ST I

'PUNCH'

11-16

one

FILE-CONTROL ANI> DEVICE-CONTROL STATEMENTS
The DISPOSE

specj~fier

must ha'le one of the following values:

KEEP

Specifies that the file is to be left where
the OPEN
statement specifies.
DISPOSE='KEEP' is the default.

SAVE

Same as KE:EP.

DELETE

Specifies on TOPS-I0 that, if the device
is either
a
DECtape or disk,
delete the file; otherwise, take no
action.
On TOPS-20, if the device involved is
the file; otherwise, take no action.

disk,

delete

EXPUNGE

On TOPS-I0, saIne as DELETE.
On TOPS-20, if the device
involved
is .3 disk, expunge the file; otherwise, take
!10 action.

Specifies that the file will be printed and kept.
file must be o~ disk.

LIST

Specifies that the file will be
The file must ~e on disk.

PUNCH

Specifies that the file will be punched on
the
tape punch and kept.
The file must be on disk.

printed

and

The

deleted.
paper

ERR
Specifier

11.3.14

ERR Specifier

The ERR specifier enables you to designate a statement number of an
executable statement,
in the current program unit, to which control
passes if an error occurs during the execution of an I/O statement.
If an error occurs and no ERR specifier or
IOSTAT specifier
(see
Section 11.3.18) is supplied, the program types an error message.
If
the program is running under batch, it is aborted.
If the program is not running under batch, it enters DIALOG mode after
processing
all
of
the other specifiers, as if you had supplied the
DIALOG specifier (see Section 11.3.9).
This
is true
regardless of
whether
or not the OPEN statement was explicitly executed or implied
by the execution of the first data transfer statement for a unit.
The form of the ERR specifier is:
ERR= s
where:
s

is the number of an executable statement to which program
control passes
if an error occurs during the execution of
the statement that includes the ERR specifier.

The subroutine ERRSNS can be called
to pinpoint the
Appendix D for FOROTS error values returned by ERRSNS.
11-17

error.

See

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

FILE
Specifier

11.3.15

FILE Specifier

The FILE specifier enables you to name the file involved in the
transfer operation.
You can specify a full file specification.

data

The form of the FILE specifier is:
FILE= Character expression
The value of the character expression is any legal TOPS-IO or TOPS-20
file specification.
(See the TOPS-IO Operating System Commands Manual
or the TOPS-20 User's Guide.)
-----If you omit the period and extension, the extension .DAT
is assumed.
If
just the extension is omitted, a null extension is assumed.
Thus,
if you want a filename without an extension,
remember
to
use
the
period.
If a filename is not specified, a default name is generated
the form:

that

has

FORxx.DAT
where:
xx

is the FORTRAN logical unit number (decimal) or the
logical
unit name
for the default statements ACCEPT, PRINT, PUNCH,
READ, WRITE, or TYPE.

FILESIZE
(INITIALIZE)
Specifier
(TOPS-10 only)

11.3.16

FILESIZE (INITIALIZE) Specifier

(TOPS-IO Only)

The FILESIZE (or INITIALIZE) specifier is
used
for
disk operations
only.
It enables you to estimate the number of words that an output
file is going to contain.
The form of the FILESIZE specifier is:
FILESIZE= Integer expression
The value assigned as a FILESIZE argument can be a integer expression,
and is rounded up to the next higher block boundary (multiple of 128).

11-18

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
The value specified by FILESIZE= is used as an estimate only.
The
effect of FILE~3IZE=
is to help the monitor try to choose the best
place on the disk to put the file.

FORM
Specifier

11.3.17

FORM Specifier

The FORM specifier designates whether the records in a data transfer
operation are formatted or unformatted.
You should not mix formatted
(character) and unformatted (binary) records in the same file.
The form of the FORM specifier is:
FORM

where:
ft

is a character expression having
, FORMATTED' 0 r 'UtJFORMA'1'TED'.

value

equal

The arguments to the FORM specifier are:
FORMATTED

specifies that the records being
contain character (formatted) data.

transferred

UNFORMATTED

specifies that the records being
contain bj.nary (unformatted) data.

transferred

If FORM is not specified and HaDE is 'ASCII' or 'LINED',
the default
value for
FORM is
'FORMATr~~ED'.
Otherwise,
if MODE is 'BINARY',
'IMAGE', or 'DUMP' (TOPS-IO only),
the default value for FORM is
'UNFORMATTED' •
If both FORM and MODE are specified and they are incompatible,
then
DIALOG
mode
is
entered,
and you are asked to correct the
incompatibility.
In the
following
example,
MODE='BINARY'
and
FORM='FORMATTED'
are specified in the same OPEN statement.
As shown
below, when the program is executed,
interactive DIALOG mode is
automatically
entered
to
enable
the
user
to
correct the
incompatibility.
PROGRAI'-l TRIMP
,FORM='FORMATTED')
ltoJRITE (UNIT=l ,FMT=lO:.)
FORMATflX,'This is il test .. '}
OPEN(UNIT=l,MODE='B~NARY'

lOl

END

EXECUTE TRINP
LINK:
Loading
[LNKXCT TRIMP execution]
?OPEN unit 1 DSK:FOROl.DAT at MAIN.+4 in TRIMP (PC InS)
?Incompatible attributes /MODE:BINARY /FORM:FORMATTED

11-19

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
[Enter correct file specs]
*/MODE:ASCII
CPU time 0.3
Elapsed time 11.4
TYPE FOROl.DAT
This is a test.
If neither FORM nor MODE is specified the default value for FORM
depends on the access.
If the access is SEQUENTIAL (or is defaulted),
the default for FORM is FORMATTED.
If the access is DIRECT or RANDOM,
the default for FORM is UNFORMATTED.
NOTE
For ASCII devices (line printer,
plotter,
terminal,
industry magnetic tape),
the FORM= specifier has no
meaning and is ignored; both formatted and unformatted
data transfers are legal (see Section 10.3.3).

IOSTAT
Specifier

11.3.18

IOSTAT Specifier

The IOSTAT specifier identifies an integer variable that
is used
store the I/O status code during the execution of a statement.

The form of the IOSTAT specifier is:
IOSTAT= Integer variable or integer array element
If no error occurs during the execution of the statement, the
variable is set to zero.

defined

If an error does occur during the execution of the statement,
the
defined variable is assigned a positive integer value that corresponds
to the number of the FOROTS error that occurred (see Appendix D for
FOROTS error codes).
When an error occurs, no error message is typed; instead, the program
either continues at the ERR= statement number (if the ERR specifier is
included), or continues at the statement immediately following
the
OPEN statement (if no ERR specifier is included).

11-20

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

[
11.3.19

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

_ __-_ _ _.]
•...

... ...

LIMIT
Specifier
--.- - - ...

..- - - - - - - - -

LIMIT Specifier

The LIMIT specifier designates the number of output units
(such as
pages)
for
spooled print or punch requests, which result from using
DISPOSE='PRINT',
DISPOSE='PUNCH',
or DISPOSE='LIST'
(see
Section
11.3.13).
The form of the LIMIT specifier is:
LIMIT= Integer expression

11.3.20

MODE Specifier

The MODE specifier defines the
record.

data

mode

external

file

is a character expression having a value equal to
the following:

one

After a MODE has been assigned (either explicitly or by default),
cannot be changed until the file is closed and then reopened.

The form of the MODE specifier is:
MODE

mod

where:
mod

'ASCII'
'LINED'
'BINARY'
I IMAGE'
'DUMP'

The default value of MODE depends on the values of FORM and ACCESS.
If FORM
is FORMATTED,
then
the default MODE is ASCII.
If FORM is
UNFORMATTED, then the default MODE is BINARY.
If ACCESS is SEQUENTIAL
and no FORM= is specified, then the default MODE is ASCII; if ACCESS
is DIRECT or RANDOM, and no FORM= is specified, then the default
MODE
is BINARY.
(See Section 11.3.17
for details on FORM, and Section
11.3.1 for details on ACCESS.)

11-21

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
Character data is supported in formatted BINARY and IMAGE mode
it is not supported in DUMP mode files.

tiles;

The possible values ot MODE are:
l\SCI I

specities

the
data
to
be
)-bit ASCII
characters.
are terminated with a line feed, form teed, or
\/ertlcal tab.
l~ecor(js

Takes effect on Input only.
For
output,
this
mode
defaults
to
ASCII.
LINED specifies
the data to be
7-bit ASCII
characters with
optional
line
sequen~e
numbers.
FORTRAN removes the line sequence numbers, if
present, before supplying the data to the
user.
(The
1 ine
sequence
number
can
be
obtained
by using the
lunction LSNGET, see Section 13.3.1.)
Note that
a
page mark
in
a
file
containing
line
::equence numbers is a separate record.
FORTRAN removes
the blank sequence number, and the carriage return/form
tRed sequence is read as a blank record.
BINARY

Specifies that data is formatted as
a
FORTRAN binary
data
tile.
A BINARY file is composed of 36-bit words
~ [-; C e Sec t ion 1 0 • 3 . 2).
The fir s t wo r d 0 f e a c h r e cor d i s
written
by
FOROTS
and specifies the beginning of the
;iinary record; this 36-bit value is
called
a
type
I
:.()uical Segment Control Word (LSCW).
i<rlct! binary record
i ,:~;CWs .
The
type

contain one
or
more
type
2
LSCW,
vir it ten
by
FOROTS und e r
::(~r ta in condi tions. is used to mark a record that spans
lnternal buffer boundaries.
i\
'..I f

can
2

third FOROTS-written word, the type 3 LSCW, is always
Itt en as the I a s t 36 - bit va III e i n e a c h BIN ARY r E~ cor d •

All data in a "BINARY" transfer remains
equivalent
its external form with its internal representation.

IMAGE

Specifies an unformatted binary mode.
Like the
BINARY
form
of
unformatted
transfers,
IMAGE specifies that
dAta is transferred as 36-bit values, with the internal
~nd
external
representation of the data remaining the
'~a me.
Unlike BINARY tiles, IMAGE files do not contain
record
~nformation
(LSCWs);
they contain only the data
~r~nsterred.
IMAGE files can be backspaced if a record
~; i ze i s s pec if i ed •

DUMP
TOPS-IO

corresponds to TOPS-IO DUMP mode I/O.
(See the TOPS-lO
Mon i to r
Ca 11 s
Man ua I .)
Reco rd
size
i s
igrl-or-ecf~
Cha~acterdafa can~cif "be written into or read from DUMP
1l1Ode
files.
Note
that
there
is
little
or
no
pertormance
advantage
to
uSlng
DUMP, as FOROTS uses
DUMP mode internally for all disk files.
NOTE

For ASCII devices (line
printer,
plotter,
termInal,
Lndustry magnetic
tape),
the MODE= specifier has no
meaning and is ignored; both formatted and unformatted
clata transfers are leqal (see Section 10.3.3).
11-22

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
:'d:)le 11-2 summarizes the different MODE arquments that are
n a iff ere n t de vic e s •

supported

Tahle 11-2:

DEVICE and MODE Cross-Tahl,~

'ASCII'

I)pvi('p

J)lc;k (sequential)

X
X
X
X
X
X
X
X
X
X

Disk Idin~ct.)
11I'~Ctapc

'j'f'nninal
l'vhl~tap('

I ,j lit' Print!'f

H.t·ader
<. ';11'(1 Punch
I 'aper Tapt' Reader
!';iper Tape Punch
I ';lrd

I, 'BINARY'

'LlNI~))'

X
X

I
I
I

x
x
x
x
X

'IMA(~E'

X
X
X

I
I

X
A
X
X
X
A

P(:~~~~011
I
I

I
i

~.J

NL\:VlE
~. pi-"clfier

I i.3.21

NAME Specifier

The NAME specifier is used to specitv a full t i l e speclflcatlon.
You
:-:an
use
this specifier instEad of the DEVICE, FILE, and/or UllH..:CTORY
...:;pecifiers.
~'he

form of

the NAME specifier

~. L"',M E:=

C h a r act ere x pre s sic n

~xamples

the NAME specifier

IS:

are:

: TOPS-l 0):

NAME= I DSK : FOCi. BAR I 1 () ,34.1 '

'T()PS-20):

NAME=ipS:<SMITH)BILLING.DAT i

rhe NAME specifier can not be qiven in DIALOG mode.
Also,
the
OPEN
,Latement cannot have both a I.·lALOG= specitier and a NAMEspecltler.

11-23

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

PADCHAR
Specifier

il.3.22

PADCHAR Specifier

The PADCHAR specifier is used onlv
lperations.
PADCHAR enables vou
,]sed to pad fixed-lenqth formatted
~pecified
record
lenqth (see the

with formatted output data transfer
to speclfv a character that will be
records, on output only,
to
their
RECL(RECORDSIZE) specltier, Section

:l.3.27).

'rho

torm tor the PADCHAR specltier is:
lADCHAR= A character expreSSIon In WhICh the tirst

character

lsed

ignored
if
The default pad character is space.
The pad character IS
is,
if the
tixed-lenqth
records
are
not
beinq
used
(that
HECL(RECORDSIZE) specitler is absent). or
if
formatted
I/O is not
:)e i nq done.

~o specity a null character tor the pad character, you
inList use the tunction CHAR (see Chapter 13), since the
;'::Clrnpiler does not allow null character constants,
for
l~xamp..i.e:

\,Pf·;N

(UNI'),=l. PADCHAR=CHAR (0))

PARITY
Specifier

11.3.23

PARITY Specifier

The PARITY specitier

')ermics
~ransfer

is onlv Llsed for maqnetic
tape operations.
It
you to specitv paritv to be oi)served (odd or even) during the
at aata.

The form ot the PARITY
'ARITY =

sPE~citler

IS:

par

'",here:
D~r

cnaraccer exuression navlnq a value equal to 10DD' or

, ~:VEN; •

11-24

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

PROTECTION (TOP5-10)
c::,pecifier

11.3.24

PROTECTION Specifier

(TOPS-lO)

This option specifies a protection code to be
assigned
to
the
data
file
being transferred.
The protection code is a 3-dIglt octal value
indicating the level of acces~ to the tile.
Each of tne three numbers
in
the
protection
code
ha3
a
specItic
meaning.
The
number
in
tte
leftmost position desIgnates the tile
owner's protection; the middlE- numbe·r designates the pro]ect
member's
protection; and the rightmost number designates the protection tor all
others
users on
the
system.
The
system default
tor
t.he
file
protection is assigned if a protection is not specitied.
On TOPS-IO, the form of the PLOTECTION specitler is:
PROTECTION= Inteqer expression
Figure 11-1 illustrates the
code.

1'OPS-IO

3-digIt

octal

tile

protectIon

When sett.ing the protection code for a file
open
tor
output,
be
sure
not to set the protection SUCt1 that
the file is protected against writing by the
program;
if you should do this
the OPEN statement WIll fail.
To protect the fj_le aqainst writing by the owner,
you
should give
the
PROTECTION
specIfIer
in t.he CLOSf<~
~3tatement •

r----·--

File Owner

PROTECTION = nnn ~.- Other Users

t_.____. . _Prolect Member~

File owner

An octal dlQlt In the tirst r'OSitiOn speCifies the file access for tile 111e owner. I he file owner
IS tle user whose pro'lrarnmer number maiches the directory !n wllietl the file IS
con·:alned.

PrOject Members

An actal digit in the seC::>nd position specltles the tile access tor me project members.
Protect members are us(~rs whose prolect number matches the Ijlrectory In which the file
IS comalned.

Other Users

An octal value In the thlr,j Dosltlon speCifies the file access for all users other than the file
owner or a protect rnerntJer_
-"1 p '!'
•• "

Fig ure 11-1:

TOPS-IO F i 1 e Pr,)tect ion Nurnbe r

11-25

f"'

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
":1ble 11-3 lists all
the possible values for
each field of the
"-ocectlon COde.
keter to the TOPS-IO Monitor Calls Manual for more
information.
T'l()TE

referred
to
in Table
extended file protection .
i" i 1 P.
Daemon allows any user
to
v
who
can and who cannot access
,t:lr tiles
(if applicable).
(Refer to
l'uP~-lU
Operating §ystem _~o_l!I_m_~r~d_s
]:,

; I:"

Daemon

D~ovides

• r

"f"

;, r-:; l a _L • )

Tahl~' 11--:~:

TOPS-IO Protection Cod£' Valm's

'!klHI
.VI~>aning in Owner Field

\' dill{'

':'~.1\' [ill' "WIlt'r C<Ill ex('cule.
i fl,'

/'('ad. append to. update. wntc. re!wme. or change the protection of

j'dp Tlw I'ill' I>al'rnon 1;'; not called Oil a protection failure,

:'f 1,· ,1 I.. ,;\\,111'1' {',In !'Xl'CUlt~_ n'ad. apPl'nd to. update. write. or ~'pname the fill', Thl' Fill' Daemon
i'-

nol

1',lilpt! 011 il prott'ctioll j~lIlun',

':'IH' fi I,' {)Wllt'!' ('(in I'xpcute. read. dDpend to. UpJcltt'. Of' write the file, The Vile Daemon is not
,::1< ,j ::!1 ,I pr!ll.I'('t\OI1 lailun'.

:'Iw Ii 1(' OWIH'1' C<1l1llot acce;.;;.; the hie. HUwl'\'l'r. tile uwner can use the RENAME monitor call to
'h,111l:l' IIII' 1111' prolL'lllOll. The hie Dal'mon 1:- nut called on a protection failure,
Tit!' ii I" OWIl,'!' call eXI'cute. read. append to. update. writl', rename. or change till' protection of
:111' iill', The mOl11tor calb thl' fIle Daemon on a protection fadUl'e,
i fl!' iii" "Wlwr ('all Px('cute. n·ad. apPl'nd to. update. or write the fill', The monitor calls the File
: :;il'iiliiil .1I1 ;,

['IH'

prlJlvclllJlI {'ad un',

fil(· O\Vlwr can ('xt'cut(' or !'l'ad thl' file, The monitor call~ the File Daemon on a protection

f"; I i I t 1 !'( ~ .

I'hl'

lilt' Il\VTlI'r call1lot atT,'SS the file, Howe\'er. the owner can use the 1{r.;NAME monitor call to
LfH' Idp pnll.t'l'tlOll. The Il)onitor call,.; the Fill' Daemon on a protection failure,

:,;IOll'"

,:ktal

Meaning in Proj('('t Member and Other User Fields

Value

I1Il'mtH'!' IO/' Olilel" c<ln l'xecute. read. append to. update. write. rename, or change the

• !'i'p ,11'1111'('t

iii

'!!'.ilj-'l'!jlll!

,:q

;;;('i('('l

dl" iilt"

IIIL'llllwl illl' "lIwl' I I'<tll ,'xl'cuLl'. 1"',1(1. appt'nd to. update. write. or rename the file,

! :". [)f'('ll'('[ 1T1t'lflilt'l' ;(Ii' olifl'l j ,'all (·Xl'cut!'. /'('ad_ appt'nd to. update. or write the file.

i'h., nl'OIl'I"

IflI'llltH'1' lor ottH'1'1 call t'Xl'Cutl'. rpad. appl'nd to. O!' update the file.

!'IH' (l1'1J1l't'1

IIWIlllwl' lor ntiw!,1 call I'XI'l'ute. I't'ad. or append to the file,

II II,' prlllVl't 11I1~fl1tlel' (or Ultlt'I'1 can execule or reaL! the tile.

TIlt' nrOf('1'1 11lt'1Il1H'1' lor other I call onlv execute the fIle,
'j';ii'

i'liP;'-> L ii.

pnlll'l.'i 1I1l'lllhe)' Ill)' oliler! has Ill) at'Ll'S" to the file.

PH,()'l'U:T LON

can be an integer expression.

If the

argument

zero vdlue or lS not specified, the default protection
i '.J [ Lile TOPS-10 Installation is used.

11-26

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

PROTECTION (TOPS-20)
-:recifier

11.3.25

PROTECTION Specifier

(TOPS-20)

The PROTECTION specifier enab:.es you to designate an octal
protectIon
code
for the file.
The protE!ction code is a 6-digjt octal value that
designates the access to the 1:ile of the owner,
group members,
and
other system users.
On TOPS-20, the form of the PHOTECTION specltler IS:
PROTECTION= Integer
Figure 11-2 illustrates a

-r· - -

expr(~ssion

TOP~;-20

pI:otection cone.

File Owner
! : ' - - G r o up Memter
r--Other Users

PROTECTION = nnnnnn

File Owner

The leftmost two digits jesiqnate the file access tor the 111e owner
user who is connected 10 the directory in which the file resl(ies.

Group Member

The middle two digits d,;slgnatc the file access tor group memt)ers l~roup memDers ar(~
established by the syst;m administrator. A group member~;hip enatdes a user to srlarl";
files among other users In the same group.

Other Users

The rightmost two digit~: designate the file access fOI all us~~rs on lilC systern. oltwr Ulan
the file owner or a qrouo member.

Figure 11-2:

I he tile owner IS the

TOPS-20 Protect Lon Number

Table 11-4 lists the possible protection values and their meaninqs
the TOPS-20 file protection code.
Table 11-4:
Octal
Value

TOPS-20 Protection Code Values

Meaning in 0" ncr, Group Member. or Other Ust'!" Field"

Permits fu II access to th(.' fi Ie.

Permits rC'ld-only access io the j Ie.

Permits write and delete access: 0 the file.

Permits execute-only access to tll' file.

Permits append-only ac('cs~ to the fill>.

Permits li;;,ting of the fil!' specification using the DIRE('TOI{Y (,()lllnwlld.

Permits nco access to the' fil('.

11-:27

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
TOPS-20, PROTECTION specities a protection code to be
assigned
to
data
tile belna transterred.
The protection code determines the
i0vel of access cnat three classes ot users have to the file.

~ne

takes an lnteaer expreSSIon as an argument.
If
PROTECTION
asslqnea
a zero value or is not specified. the default protection
>;rle establishE~d for vour connected directorv 1S used.

')!~OTECTION

READONLY
Specifier

~1.3.26

READONLY Specitier

READONLY specltier is used to speclty that the program will
only
from
the tile.
Output to the tIle is L_leqal and will cause an
'rro[ ~t execution tIme.

':"lC

-,-','10

~': e

form

the

READ a NL Y s p e cit i e r is:

'::f\iJON L Y

REel
(RECORDSIZE)
Specifier

11.3.27

RECL

. :::'>:'RDSIZE) Specifier

The RECL (0r KECORDSIZEI
of characters or words
for all files opened for
';~ANDIN').
(See Section

specifier enables you to specify the number
in each record transferred.
RECL is required
direct access (ACCESS='DIRECT', ~RANDOM'.
or
11.3.1.)

The form of the REeL specifier is:
REeL= Integer expression
an ASCII transter (MODE=fASCII
or 'LINED'). the value assigned
the number 0 f cha r acte rs in each record.
I

;n<CL spec 1 ties

For output to disk or CORE-DUMP tape
files,
in addition
to
these
characters,
FOROTS adds a carriage return/line feed to each record,
followed by enough null characters to fill the current word,
so that
records are word-aligned.
REeL is enforced on output by padding short
records with
the padding character
for
formatted
records.
Long
records are truncated.
FORTRAN enforces a specified REeL for all input operations.
If the
REeL specified in an OPEN statement is different from the actual size
of the records, FOROTS reads the number of characters specified by
REeL.

11-28

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
For input to disk or CORE-DUMP tape files, specifying a record size
directs FOROTS to read records that are word-aligned. The calculation
of the actual recordsize is the size specified, plus two for the
carriage return/line feed, plus the number of nulls necessary to
word-align the record.
1"0 r
INDUSTRY tapes, wi th RJ~COf~DTYPE=:' FIXED', RECL spec 1 t 1 es the
exact
ilumber
of
characters
In
edch
record;
there are no termInators or
;),ldd ing character s.
.b'or lNDU~i·l·H.Y
t:apes
WI th
RECORDTYPE=' VARIABLE' ,
f~ECL :;pecifies the maximum record size in the f i l e , excludIng the RCW.

When the record is read, regardless of the contents of the record, it
is interpreted as specified by the rules above; there are RECL
characters of data, and the rest are ignored.
No interpretation is
done of the characters in the data part of the record. These
characters appear in the FOROTS line buffer exactly as they appear in
the file, including nulls and control characters.
T ~1 the cas e 0 t
,: fld r ac te r s and

M0 1) E = ' LIN E: D', 1: h e
va Iu e
0 t
R E C L e x c 1 u (1 e s t h e
tab in each .lltte-sequence n umbe r •

t i ve

iJilldrV tran!::;ter (MODE::::'B_:NARY'" or 'IMAGE'), the
value
assigned
f-n:CL
specitles
the
fluml>er
ot 36-bit words in eacl1 record.
.For
MODE='BINARY', ttle value In RJ;CL excludes the LSCWs wrItten by t'UROTS.
fn

1-0

MOD E '" ' 1) U M P'

1 ssp e c i fie d"

R EeL

1 g nor e d •

RECORDTYPE
SDecifier

11.3.28

RECORDTYPE SpecltIer

The Rr~CORDTYPE specitier detilles
magnetic tape fIle.
The

torm of

the HECORDTYPE

lRCORDTYPE

the

Spl~cltIer

format.

is:

= rtype

tJhere:
Lype

l ~j

den d r d C !_ e r

t-' X pre S :::0. 1 0 n

11-29

the

records

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

ncr:

(CORE-DUMP) .
For
the standard ANSI
Tnu3, the
record data
is
r~Ym;~A[orS
or
carriaae control
7"""'''y"'1
·'11 r~C';-1.y
arter
another.
to
: x;·rj-- i ."!ui fl
>iocks.
For this record
OPEN
tne

c;l?rlrU-'(]

format

jh-oir:

tor

t -, l s i s

:1:

f n

,- f~, ,-'i t-

I.,"

~ i

'~I

'-: ,1Iil P

'/ f-'

n h v:::; 1 r. A l b -1 0 c k s () n t h f>
X" An, f () r the last b Joe k .

thi~

!:!

format
sDecifies
standard stream record
pnd
of

r~~()rrl

lIes.
,. :

f,',:.;;r
•

~t?lnn'1l"'C1

_OJ

-;,

:::;rrprlm

rj,~ i 1mi t

t)'"
""-

,ii'[ldV

[()r

I;;

rerminAi.or

records.

For
irjpntical to that

the
_.,'. ::.

,'x '"",

~',(~

,; l

r or

the

i.ASr:.

nl.ocK.

: '; i (~. :-.... ;.:; ..

,-- ! ;--'

r r

- ,-:

-- r ) ! l M P i
1

-..-: 1

f!!, , i

;:nii:-;;'-;f':~

n('),

c1eflne(1

tor

r t1 p h v t- p s i n
r .~ c~ r:- p r ot dat:.a

_
,,_

; ,,, (i

,;,

,if">

1 t

i p

t n P A C t: II i'I i r e cor d

Rr~CORDS I Z E

CRLr'.

rlrit'i'I.

formAt

t-Anp

11-,-hlt

WIlt

1 ,--i

chis

the

And

(COR f<---:-D1JMP) •

is the standard

recon1s
are
(latA IS wr'lttpn

Tnus.
rp("r)rO

rn-"'lY',"lcrprs.
, .. / r

(.....l.

r ", ,...
,. , ~
---".

•.:..:- 1

() t

""7 £.,l.

exceed the

block
i t

tl"
" :ti

; -) : i. i'.

f:·
! -:

.~~ ;"'; ( .. C"

this

i i , f.."ofi
t .. ~

• ~rJ

rOijr

rpcord
tess

cnaracters

STATUS
;-; 'f:~E)

Specifier

11.3.29

STATUS

Specifier

The STATUS
specifier lets you specify whether
or
not
the
file
being opened must exist, or what to do with the opened file when
it is closed.

11-30

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
The form of the STATUS specifier is:
STATUS = sta
where:
sta

is a character expression whose value is equal to
the following:

one

'F:XPUNGE'
'OLD'
'NEW'
'SCRATCH'
'UNKNOWN'
'KEEP'

'DELETE'
The arguments to STATUS are:
I';XPUNGE

On TOPS-lO, thjs specifies that
the
fIle
IS
~eletej
\!lhen it is clo~,ed.
On TOPS-20, this spec fies that the
f iJp is deletec. and expunged when it is ctosed.
NOTE
On TOPS-IO
any delete
also
expunges
a
t J I.e
from
st.orage.
On TOPS-20, a DELETE operatIon
marks the 1ile as deleted; an EXPUNGE c)peratlon
immediat:.ely erases the file from storaqe.
I

OLD

Specifies that
the
file
being
opened must
already
exist.
If the file does not exist, an error results.

NEW

On TOPS-10, STATUS='NEW' specifies that the
file must
not exist.
]f the file does exist, an error results.
An error also occurs if you specify STATUS='NEW'
with
ACCESS=" ·::O:N', 'SEQUENTIAL',· ~'i:(;TN()UTi
(to a read-only
device), .Il~ '[U,NDIN' (see Section 11.3.1).
acts
On
TOPS-20,
the
STATUS='NEW'
specifier
in
the
differently,
depending
on what you have
directory befoI·e STATUS='NEW' is executed.
Also,
the way you specify the
file
in
the
OPEN
statement which contains
the STATUS='NEW' specifier
influences the way the STATUS='NEW' specifier operates.
The
following
list describes
the ways
that
this
specifier can operate when used on TOPS-20.
1.

If the file specified in the OPEN statement does
not currently exist
in
the directory,
and
no
generation
number
is
specified,
then
the
STATUS='NEW'
specifier creates the specified file
and gives jt a generation number of 1.

If
the
fj.le
specified
in the OPEN
statement
contains a
name, extension, and generation number
that does not exist, the specified file is used.

If the
flle
specified
in the OPEN or
CLOSE
statement
contains
a
name,
extension,
and
generation number that is exactly the
same
as an
existing
file in your directory, then STATUS='NEW'
causes an error, and no file is created.
11-31

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
4.

If you did not specify a generation number, but the
file
specified has the same name and extension as
an existing file in your directory, then
the
file
with
the
same name and
extension and the next
highest generation will be created.

Specifies that the file will be automatically deleted
when
the
file
is closed.
STATUS='SCRATCH' implies

SCRATCH

The
A SCRATCH file is always given a name by FOROTS.
the FOHTRAN
name of the
file
is
inaccessible
to
program.
If STATUS='SCRATCH'

is used, you must not specify FILE,
'l(
Vr:RSl()N.
If your program is
writing a file with STATUS='SCRATCH', and the
file
is
being written to disk, you can retain it by executing a
CLOSE statement that renames the file
to
a
specified
name.
!-\

i-i,i;',,,',i!()N,

If a file is opened with STATUS='SCRATCH',
the access
must
be
ACCESS=' SEQUENTIAL'
',.;JUT· 1
or
ACC ES S = ' DIRE CT '
, " , ',,, til . ) ( see Sec t ion 11. 3 • 1) •
,I

Specifies that a file opened
for
an
input operation
must exist.
When a
file
is opened for output with
STATUS='UNKNOWN', if the file exists, it is superseded;
if it does not exist, it is created.

UNKNOWN

UNKNOWN is the default for STATUS.
This value is
used
unless you specify STATUS or unless the value of STATUS
is otherwise determined by the ACCESS specifier.
, ; ;-:; c '-: ; i i e ,c:;

t, tl a t

;'~ h (;

:- i i e

i s

no',:

del e ted •
to

)i..Ja~ent

.1fi': :-,'l'/\'J'lJS=' UNKNOWN' •

",.j',

lj n TOP S - 1 O.

S U t' elL 1 e s t II a L the

UidL
I

! ..:::

tlle

,~ 1 () :, pd.

T' ~i,PEFORMAT
S!J(::ci iier

I l.3.JO

:nrrn

TAPE FORMAT SPECIFIER

1 ne

iAi"r;i'UHMAT

Soecifyinq
specifvinq

snecltler

11-32

IS:

elI e

w i 11 bee r a sed w hen

tile
will
be
deleted
t i l e IS erased when a
(]lVt~n.
'J'o
undelete
a

TIl e

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
where:
tmode

is a character expression
'CORE-DUMP' or 'INDUSTRY'.

having

value

equal

The values for the TAPEFORMAT specifier are:
CORE-DUMP specifies the "DIGITAL-compatible" tape
format,
which
is 36-bits stored
in five frames on a 9-track tape.
The SET TAPE RECORDSIZE
(TOPS-20)
or SET BLOCKSIZE
(TOPS-IO)
command
is
interpreted to be the number of
36-bit words in the magnetic tape blocks on the tape,
and
is used if no BLOCKSIZE specifier is given in the
OPEN statement.
If a BLOCKSIZE specifier is given
in
the OPEN statement, it is interpreted to be the number
of 36-bit words
for
both
formatted and
unformatted
files.
INDUSTRY

specifies characters are read or written
in standard
industry tape format, one character per tape frame.

UNIT
Specifier
(required)

11.3.31

UNIT Specifier (Required)

The UNIT specifier defines the FORTRAN logical unit number to be used.
FORTRAN devices are identified by assigned decimal numbers within the
range 0-99 (see Table 10-3).
UNIT must be an integer expression.
The form of the UNIT specifier is:
UNIT= Integer expression
If the unit specifier
is the
first
specifier given
in
the OPEN
statement,
the
keyword UNIT= is optional.
For example the following
statement opens a file on unit 20:
OPEN (20,FILE='MYFILE')
NOTE
The FORTRAN standard logical
unit assignments are
described
in Section 10.4.3.1. Although the range of
logical unit numbers shown in Table 10-3 is 0-99,
the
range of UNIT numbers
is an
installation-defined
parameter.

11-33

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

VERSION (TOP5-10)
Specifier

11.3.32

VERSION Specifier (TOPS-IO)

Use the VERSION specifier for disk operations only; it enables you
assign a 12-digit octal version number to an output file.

The form of the VERSION specifier is:
VERSION=Integer expression

11.4

CLOSE STATEMENT

The CLOSE statement disassociates an active file from a
logical
unit
and releases the memory occupied by I/O buffers and other unit-related
data.
The CLOSE statement can also change some of the
file
characteristics that were assigned during the OPEN, such as the name,
protection, directory, and disposition of the file.
Once a CLOSE statement has been executed, you must
use another OPEN
statement (or implicit OPEN) to regain access to the closed file.
The form of the CLOSE statement is:
CLOSE

(closelist)

where:
closelist

11.4.1

is a list of CLOSE statement specifiers.
This
list must contain the UNIT specifier (see Section
11.5.9)
and
can
optionally
contain
other
specifiers.

Implicit CLOSE

An implicit CLOSE occurs when FOROTS automatically closes a
logical
unit without execution of a CLOSE statement.
This can happen when a
program terminates, or when you execute an OPEN for
a
unit
that
is
already connected to another file.

11.5

CLOSE STATEMENT SPECIFIERS

All of the CLOSE statement specifiers are optional,
except the UNIT
specifier which is required.
Some CLOSE statement specifiers have the
same formats as the corresponding specifiers in the OPEN statement.
Table 11-5 summarizes the specifiers in the CLOSE statement,
and
the
typ~ of value required by each.
A section number is provide~ to refer
to detailed descriptions of each specifier.

11-34

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
Table 11-5:

Summary of CLOSE Statement SpecifiHs and Arguments

Argument

Pnssible Value

DLVICE

Section
~ : .S. I

DIALO(;

( 'haractcr cxpn'SSlOn

I 1IIn:( "I'< my

t 'harllctcl' exprcsslOn

J)1:--iI'OSE

lliaractl'l' l'XPI'l'''c-IOJ1 \nth ol1e of tlll' following values:
:-;1\ VE', 'IWU:Tl';', 'l'l{lNT', KEEP', 'LIST', Pl 'Nell',

i i.0 .. 1

EXPUNCE'

ERR-

Statement number

11.5.5

IOSTAT=

[nteger variable or integer array element

11.5.6

1.1\11'1'

I l.f) "

.\i,\.l\lE

II.j

I'IU),),ECTION

~ 1.'. 1

STATUS·

Character expression with one of the following values:
'KEEP', 'DELETE', :\!'UN<a.;'

11.5.8

TYI'E

UNIT-

Integer expression

11.5.9

QOTE
For
compat:.bility
with
previous
versions
of
FORTRAN-lOj20, you can sp(~cify a numeric array name as
the value of each of the :ollowing specifiers:
DIALOG=
DIRECTOHY
NAME
When a numeric array name is used, FOROTS assumes that
it
contains
a
strIng
~f characters terminated by a
null character.
In addition, you can specify a numeric variable as the
value
of
the
DEVICE
and
FILE
specifiers.
If the
variable is single precision, FOROTS assumes
that
it
contains
5
characters;
if
it
is double precision,
FOROTS assumes that is contains 10 characters.
The use of numeric array names and
numeric
variables
in
place
of
character
variables
is
a nonstandard
feature.

11-35

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

DEVICE, DIRECTORY]
FILE, NAME, and
PROTECTION
Specifiers

11.5.1

DEVICE, DIRECTORY, FILE, NAME, and PROTECTION Specifiers

The CLOSE statement file identification specifiers can be
used
when
you want
to rename the output file when it is closed.
Their formats
are the same as the corresponding specifiers in the OPEN statement.
If any of these specifiers are given in the CLOSE statement, the
file
is
renamed
when
it
is closed.
If some, but not all of the file
identification parameters are specified on a CLOSE statement, only the
specified parameters are changed when the file is renamed.
Example:
OPEN(20,ACCESS='SEQOUT' ,FILE='FOO.BAR')

CLOSE(20,FILE='NEWFIL')
The above sequence renames the output file to DKSB:NEWFIL.BAR.
Refer to the following sections under the OPEN statement:
FILE - see Section 11.3.15
NAME - see Section 11.3.21
DEVICE - see Section 11.3.8
DIRECTORY (TOPS-I0) - see Section 11.3.11
DIRECTORY (TOPS-20) - see Section 11.1.12
PROTECTION (TOPS-IO) - see Section 11.3.24
PROTECTION (TOPS-20) - see Section 11.3.25

DIALOG
Specifier

11.5.2

DIALOG Specifier

The DIALOG specifier
enables
you
to
type
in
additional
specifiers when the CLOSE statement is actually executed.

If the DIALOG specifier is found in the CLOSE list, then each time the
CLOSE statement is executed, FOROTS suspends execution and prompts the
terminal with an asterisk.
The form of the DIALOG specifier is:
DIALOG

11-36

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
respond
to
t.tle
a~terlSK
Drornot
by
enterLnq
a
tile
iJiALOG
switcres
(see below), or a t1le spec1fication
r,)11owed by DIALOG switches.
The file specification
cal.
be
a
full
" I p ~~nA"ltication 1ncludino the device.
directorv namE:. cmd so on.
~:an

',,11

")E:~cirication.

'/0 II

:,)eCli

': i E-' r

e nc. e r a i 1.1 e S Dec 1 t 1 cat 1 0 rl t h d t
IS
dlr:1:ererlt
lcacion
currencly a!:Osloned co trle [lIe. ruH\JiS
it- ! ~o. ci ()~~~~j to tile r1(~W name.

tlom
KE1"AMf~s

the
tile
the tile

:.,""',':
!'! 1\ LUG

. '.

'!/\.LOG=.
:_.1 ash

s WIt c 11

UNIT.

1 san \.

NAME:.

C L OS Esp e c 1 t 1 e r

foILf:o~.

[USTAT,

(e x c e p t J) J l\ LOG ,
preceded bv

t:RH)

(I).

;~:ie.~_::)G

-=-

;:"'--ifier

D[ALOG= Specitier
DTAL()(;= spec1tier enables vou to include all or a portion
ot
the
spec1lIers
1n
a
cl1aracter
expression.
l'ne C()[1tents at the
:j'~'ract:C'r expressIon are incerpreteri as if vou had
qiverl
the
DIALOG
';~"C'11:1er
(see
above);
and, when the asterIsk prompt was gIven, you
:jdd tVDed 1n
the
same
strirq
as
1S
contaIned
in
Lhe
character
'~Dression, followed by a carriaqe return.

"'lie

':_:JSE

:';1'2

the DIALOG= specif1er 1S:

form 01
! 1\ L OC;

= C h a rae t e r e x 0 res S Ion

'<xamOJe:
.()f;:~

(UNIT=20,ERR=lO,DIflLO(3=' /DISPOSE:DELETE')

.-Jhen DIALOG= 1 S a i ven in the l LOSE 1 i st,
-'Lher ::;wecitie[~. exceDt DIALOl.

:l t

processed

alter

all

'l~ J>OSE
-.,:",:,:ifier

DISPOSE Specifier
!' '--I

IH~:) PO:"; Esp e c 1 tIe r

en a 1) 1 e s

v 0 U to

;hp

tile

is closed.

!'IIt='

form ,)t the DISPOSE specliier
f'r~;p()sF.

is:

= dis

11-37

S Pe c 1 t Y

a n act 1 0 n t ()

occur

when

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
, 'er. e:
i
;

::'lc:l(=jcter

expression having a value equal to

on'=

! { ) i i OW 1 no:

flO"

X i-~U~·lG E '
f NT I

"j(

!,jJNCH I

'" ~1 C

iJ I S PO S E
':~p

spec 1 tie r

In u s t

have one

0 t

the to 11 0 win g va I u e s :

Epccitles that the file is to be left on the
connected
n: , 1, •
U I S PO S E = 'KE: E P ' i s
the
de fa u 1 t •
You can not
';:';~:clt\/ DISPOSE='KEEP'
if
in
the
corresponding
OPEN
~~;_lr.ement
yOU
specified STATUS='SCRATCH' (see Section
~ • ,:; ~ 1 •

~ame

as Kr:r:P.

:'~Dt::cifies

:; ''; (: tap e

on TOPS-IO that, if the device
is
either
a
dis k ,
del e t e the f i Ie; 0 the r w i s e, t a ken 0

0 r

·;n TUPS--20, if the device involved is
fk
tile: otherwise, take no action.

disk,

delete

On TOPS-IO, same as DELETE.
On TOPS-20, if the
device
;;"i'Jolved
is
a disk. expunge the file; otherwise, take
"j

c t 10 n •

iT NT

Specifies that the file will be printed and
kept
(the
tile
wlil
not
be
kept
if
you
specify
the
CLOSE
·;Latement STATUS='DELETE' or 'EXPUNGE').
The file must
H' on dlSK.

:~T

specifies that the file will be printed,
deleted,
and
'«;)~Inaed
(the
file will not be deleted if you specify
:-h,_, CL()Sr~ statement STATUS='KEEP').
The file
must
be
~
i 1s k •
. I

;;C;_~CltU?S
;, ,

tJ : H: c

that

the

,HI d

flie

k e pt.

will be punched
on
the
paper
The
f i 1 emu s t
be
0 n dis k •

value ot the
CLOSE
statement
DISPOSE
specifier
the
value
of
the OPEN statement DISPOSE
~ecifier and the OPEN statement STATUS
specifier
if
':'.'\TUS=· EXPUNGE', 'KEEP', and 'DELETE'.

[':-It::

'!)er :--;edes

11-38

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

ERR
Specifier

11.5.5

ERR Specifier

The ERR specifier enables you to designate a
statement label of an
executable statement,
in the current program unit, to which control
passes if an error occurs during the execution of an I/O statement.
The form of the ERR specifier is:
ERR= s
where:
s

is the number of an executable statement to which program
control passes
if an error occurs during the execution of
the statement in which the ERR specifier is included.

The ERR specifier works the same way when it is given in a CLOSE as it
does when given in an OPEN statement (see Section 11.3.14).

IOSTAT
Specifier

11.5.6

IOSTAT Specifier

The IOSTAT specifier identifies an integer variable that
is used
store the I/O status code during the execution of a statement.

The form of the IOSTAT specifier is:

- - - -.- .-._" - - -- - ]
1.

IOSTAT= Integer variable or integer array element

LIMIT
Specifier

11.5.7

___________________

LIMIT Specifier

The LIMIT specifier designates the number of output units
(such as
pages)
for
spooled
print or punch requests, which result from using
DISPOSE='PRINT',
DISPOSE='PUNCH',
or DISPOSE='LIST'
(see
Section
11.5.4) .

11-39

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
The form of the LIMIT specifier is:
LIMIT= Integer expression

STATUS
Specifier

11.5.8

STATUS Specifier

The STATUS specifier tells FOROTS what to do with a file when
it
is
closed.
In the CLOSE statement, STATUS values are a subset of the
DISPOSE specifier (see Section 11.5.4) values.
NOTE
The ANSI-77 standard does not have DISPOSE
allows STATUS='KEEP' or STATUS='DELETE'.

and

only

The form of the STATUS specifier is:
STATUS = sta
where:
sta

is a character expression having a value equal to
the following:

one

'KEEP'
'DELETE'
'EXPUNGE'
The arguments to STATUS are:
KEEP

Specifies that the file is not deleted.

DELETE

On TOPS-IO, specifies that the file is deleted.
On TOPS-20, specifies that the
file
is marked
for
deletion when the file is closed.
The file is erased
when a TOPS-20 EXPUNGE command is given.
To undelete a
deleted file, use the TOPS-20 UNDELETE command.

EXPUNGE

On TOPS-IO, the same as delete.
On TOPS-20,
specifies that the file is deleted and expunged.
NOTE

The value of the CLOSE statement STATUS specifier
supersedes the value of the OPEN statement DISPOSE
specifier and OPEN statement STATUS specifier
if
STATUS='EXPUNGE', 'KEEP', and 'DELETE'.

11-40

this

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

UNIT
Specifier
(Required)

11.5.9

UNIT Specifier (Required)

The UNIT specifier tells FOROTS which logical unit number
is to be
closed.
This specifier corresponds to the UNIT specifier in the OPEN
statement (see Section 11.3.31) and to the UNIT specifiers in the data
transfer statements (see Section 10.4.3).
The form of the UNIT specifier is:
UNIT= Integer expression
If the unit specifier is the
first
specifier given
in the CLOSE
statement,
the keyword UNIT=
is optional.
For example, to close a
file on unit 20, you can use the following command:
CLOSE (20)

11.6

OPEN AND CLOSE STATEMENT EXAMPLES

The following are examples of OPEN and CLOSE statements:
OPEN

(UNIT=l,DEVICE='DSK' ,ACCESS='SEQIN' ,MODE='BINARY')

causes a disk
file
named
FOROI.DAT
(since no
FILE= option was
specified) to be opened on unit 1 for sequential input in binary mode.
OPEN (UNIT=3,DEVICE='DSK' ,FILE='PAYROL.DAT',
1
ACCESS='RANDOM',MODE='ASCII',RECORDSIZE= 80,
2
ASSOCIATEVARIABLE=I,ERR=240)
Causes a disk file named PAYROL.DAT to be opened on unit 3 for
random
I/O operations
in ASCII mode.
The
records
in PAYROL.DAT are 80
characters long; the ASSOCIATE VARIABLE for this file
is
I.
If an
error occurs during
the execution of this OPEN statement, the OPEN
terminates, and control is transferred to the statement labeled 240.
CLOSE

(UNIT=3,DISPOSE='DELETE')

The above statement causes the file
closed and deleted.

11-41

associated

with

unit

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
11.7

INQUIRE STATEMENT

The INQUIRE statement inquires about specific properties of a file
name or of a logical unit number on which a file might be opened.
The
INQUIRE statement has two forms:
one inquires by file, and the other
inquires by unit.

11.7.1

INQUIRE by File

An INQUIRE by file is an INQUIRE statement containing the following:
•

A FILE= keyword

•

An associated file specification

•

No UNIT= keyword

It is used to obtain information about a file based on the file name.
INQUIRE by file can be used to get information on the following files:
•

Files that are "connected"; meaning files for which an OPEN
statement has been executed or for which a data transfer
statement has been executed.

•

Files that are not "connected."

INQUIRE by file has the following form:
INQUIRE (FILE=fi [, fl ist] )
where:
fi

is a character expression whose value specifies
of the file to be inquired about.

flist is a list that can contain at most one of
inquiry specifiers (see Section 11.7.3).

each

the

name

the

INQUIRE by file may be used any time during the execution of a
program.
It can be used before a file is opened to find out about the
existence of the file, or after the file is opened to find out other
attributes of the file.
It can also be used to find the unit number
on which the file is opened.
If the same file is opened on more than
one unit, the smallest number on which the file is opened is returned.
The determination of whether a file specified in an INQUIRE
is opened on a unit is the following:

statement

The file specification given in the INQUIRE statement is used
to lookup the file.

If the file exists, the file specification, expanded with the
physical device name and generation
(TOPS-20 only),
is
compared with the file specification for each open unit,
in
ascending order, until there is an exact string match.

11-42

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
3.

If the Eile does not exist, the specification given in the
INQUIRE statement (with a default of DSK: added if necessary
for the device name), is compared with the file specification
for
each open unit,
in ascending order, until there is an
exact string match.
Note that this match will only be
successful for 'deferred' OPEN files, since non-deferred OPEN
files are always established
in the specified directory
immediately.
Therefore, the file exists (see item 2 above).
NOTE
If a file exists, INQUIRE by file will
not generally
match the file with a unit for which a 'deferred' OPEN
has been done, since the file
specification for
the
unit has not been expanded.
For example, the file's
logical device name has not been replaced by a
physical device name.

(See Section 18.8 for information on FOROTS and INQUIRE by file.)

11.7.2

INQUIRE by Unit

INQUIRE by unit is an INQUIRE statement containing a UNIT= keyword and
no FILE= keyword.
It is used to find out information about the file
that may be "connected" to the specified unit.
INQUIRE by unit has the following form:
INQUIRE ([UNIT=]u,ulist)
where:
u

is the number of the logical unit to be
inquired about.
The unit need not exist, nor need it be connected to a
file.
If the unit is connected to a
file,
the
inquiry
encompasses both the connection and the file.

ulist is a list that can contain at most one of
inquiry specifiers (see Section 11.7.3).

each

the

If the optional UNIT= keyword if omitted, u must be the first item
the list.

INQUIRE by unit can be used at any time during
the execution of a
program.
It can be used before a file is opened to find out if there
is another file open on the unit, or after the file is opened to
find
out other attributes of the file.

11.7.3

Inquiry Specifiers

The specifiers described in the following
either form of the INQUIRE statement.

11.7.3.1
form:

sections

ACCESS Specifier - The ACCESS specifier

ACCESS

acc

11-43

may

has

the

used

following

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
where:
acc

11.7.3.2
form:

is a
character variable,
array element,
or
substring
reterence.
It
is assigned the value 'SEQUENTIAL' if the
file is connected for sequential access,
or
'DIRECT'
if
the
file
is connected for direct access.
If there is no
connection, acc is 'UNKNOWN'.

BLANK Specifier - The

BLANK

specifier

has

the

following

blk

where:
array element,
or substring
is a
character variable,
reference.
It
is assigned the value 'NULL' if the file
and
is assigned
the
was last opened with BLANK='NULL',
value 'ZERO' if the file was opened with BLANK='ZERO'.
If
the file is not open, blk is 'UNKNOWN'.

blk

11.7.3.3 CARRIAGECONTROL
has the following form:

Specifier - The

CARRIAGECONTROL

Specifier

CARRIAGECONTROL = cc
where:
~c

11.7.3.4
form:

is a
character
variable,
array element,
or
substring
reterence.
It is assigned the following values:
L

'FORTRAN' if the file has the FORTRAN carriage-control
attribute

'LIST' if the file has
attribute

'NONE'

'TRANSLATED' if the file has FORTRAN carriage-control
characters being
translated directly
into vertical
motion characters.

UNKNOWN
if
the CARRIAGECONTROL value
determined. or the file is not open.

the

implied

if the file has no carriage-control attribute

DIRECT Specifier - The DIRECT specifier

DIRECT

carriage-control

dir

11-44

has

cannot

the

following

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
where:
dir

is a character variable,
array element,
or
substring
reference.
It is assigned the following values:

11.7.3.5
ERR

'YES'
file

if DIREc'r is an allowed access

method

for

the

'NO' if DIRECT is not an allowed access method for the
file

'UNKNOWN' if tne processor
is unable to
whether DIRECT is an allowed access method

determine

ERR Specifier - The ERR specifier has the following form:

= s

where:
s

11.7.3.6
form:
EXIST

is the label of an executable statement.
ERR is a control
specifier;
if an error occurs during execution of the
INQUIRE statement, control is transferred to the statement
whose label is s.

EXIST Specifier - The

EXIST

specifier

has

the

following

where:
ex

11.7.3.7
FORM

is a logical variable or logical
array element.
It
is
assigned the value
.TRUE. if the specified file or unit
exists, and the value .FALSE.
if the specified
file or
unit does not exist.

FORM Specifier - The FORM specifier has the following form:

= fm

where:
fm

is a character variable,
array element,
or
substring
reference.
It
is assigned the value 'FORMATTED' if the
file is connected for formatted I/O, and 'UNFORMATTED'
if
the file is connected for unformatted I/O.
If there is no
connection, fm is 'UNKNOWN'.

11.7.3.8 FORMATTED
following form:
FORMATTED

SpecifieI' - The

= fmd

11-45

FORMATTED

specifier

has

the

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
where:
fmd

11.7.3.9
form:

is a
character variable,
array element,
or
substring
reference.
It is assigned the value 'YES' if formatted is
an allowed form for the file.
It is assigned
the value
'NO'
if formatted is not an allowed form of the file, and
the value 'UNKNOWN' if the form cannot be determined.

IOSTAT Specifier - The IOSTAT specifier

IOSTAT

has

the

following

ios

where:
ios

11.7.3.10

is an integer variable or integer array element.
It
is
assigned a
processor-dependent positive integer value if
an error occurs during execution of the INQUIRE statement,
or assigned the value zero if there is no error condition.

NAME Specifier - The NAME

specifi~r

has the following form:

NAME = nme
where:
nme

is a character
reference.
It
inquired about.

variable,
array element,
or
substring
is assigned
the name of the file being

The value assigned to nme is not necessarily identical
to
the value specified with FILE=.
For example, the value
that the processor returns may contain a directory name or
generation number (TOPS-20 only).
However, the value that
is assigned is always valid for use with FILE= in an OPEN
statement.
NOTE
FILE and NAME are synonyms when used with the OPEN
statement,
but not when
used with the INQUIRE
statement.
For INQUIRE by unit, FOROTS
returns the
full,
expanded
file specification if any of the following is true:
•

If there is a file open on the specified
STATUS other than UNKNOWN or SCRATCH.

unit

with

•

If there is a file open on the
ACCESS other than SEQUENTIAL.

unit

with

•

An I/O transfer statement has been executed using
specified unit and the unit has not been closed.

the

11-46

specified

FILE-CONTROL ANI> DEVICE-CONTROL STATEMENTS
FOROTS returns thE! string given in the OPEN for NAME= with
defaults applied j,f both of the following are true:
•

If the file is open on the specified unit as a
result
of an OPEN statement
in which STATUS='UNKNOWN' and
ACCESS='SEQUENTIAL' are specified or implied.

•

No I/O transfE!r statement has been executed us ing
specified unit.

the

If STATUS='SCRATCII', FOROTS returns blanks for NAME=.
If there has been no OPEN statement, and no
I/O transfer
statement has bE~en executed using
the specified unit,
FOROl'S returns for NAME:: the default file
specification
for that unit.
For INQUIRE by file, FOROTS returns the
full,
expanded
file
specification if the
file exists in the specified
directory.
If the
file does not yet exist on the
speci.fied directory,
but has been opened by a 'deferred'
OPEN, FOROTS returns the string given
in the
INQUIRE
statement,
with defaults applied
for
the device and
generation number (TOPS--20 only).
Otherwise,
blanks are
returned.

11.7.3.11
form:

NAMED Specifier - The NAMED

NAMED

specifier

has

the

following

nmd

where:
nmd

is a logical varidble or logical
array element.
It
is
assigned the value
.TRUE.
if the specified file has a
name, and the val~e .FALSE. if the file does not have a
name ..

11.7.3.12
form:

NEXTREC Specifier .- The NEXTREC specifier has the following

NEXTREC

where:
nr

11.7.3.13
form:
NUMBER

is an integer variable or integer array element.
It
is
assigned an
integer value that is one more than the last
record number read or written on the specified direct
access file.
If no records have been read or written, the
value of nr is one.
If the
file
is not connected
for
direct access, or if the position is indeterminate because
of an error condition, nr is zero.

NUMBER Specifier - The NUMBER specifier has
num

11-47

the

following

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
where:
num

is an integer variable or integer array element.
It
is
assigned
the number of a logical unit currently connected
to the specified
file.
If there
is no logical
unit
connected
to
the file, num is not defined.
If more than
one unit is connected to
the
file,
the smallest unit
number is returned.

11.7.3.14
form:

OPENED Specifier - The OPENED specifier has

OPENED

the

fo1101Ning

where:
od

is a logical variable or logical
array element.
assigned
the value .TRUE. if the specified file is
on a unit or if the specified
unit
is opened;
assigned
the value
.FALSE.
if the file or unit
open.

11.7.3.15 RECL
(RECORDSIZE)
Specifier - The
specifier has the following form:
RECL

RECL

It
is
opened
it
is
is not

(RECORDSIZE)

= rcl

where:
rcl

is an integer variable or integer array element.
If the
file
(or
unit)
is opened, rcl is the record length for
fixed-length record files.
In all
other cases,
rcl
is
zero.
If the
file
is opened for formatted I/O, rcI is
expressed in characters, and
in words
if the
file
is
unformatted.

11.7.3.16 RECORDTYPE Specifier - The
fallowing form:
RECORDTYPE

RECORDTYPE

specifier

has

the

rtype

where:
rtype is a
character variable,
array element,
or substring
reference.
It
is assigned the value 'FIXED' if the file
has fixed-length records, assigned the value 'VARIABLED if
the
file
has variable-length records, and assigned the
value 'STREAM' if the file is a stream file
(default
for
disk
and
maqnetic tape).
If the processor cannot
,Jetermine the record type, rtype
is assigned
the value
'UNKNOWN' .

11-48

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
11.7.3.17 SEQUENTIAL Specifier - The
following form:
SEQUENTIAL

SEQUENTIAL

specifier

has

the

seq

where:
seq

is a
character variable,
array element,
or
reference.
It is assigned the following values:
1.

'YES' if SEQUENTIAL is an allowed
the specified file

access

'NO'

'UNKNOWN' if the processor cannot determine
SEQUENTIAL is an allowed access method

method

for

if SEQUENTIAL is not an allowed access method

11.7.3.18 UNFORMATTED Specifier - The UNFORMATTED specifier
following form:
UNFORMATTED

substring

whether

has

the

unf

where:
unf

is a
character variable,
array element,
or
substring
reference.
It is assigned the value 'YES' if unformatted
is an allowed form for the file.
It is assigned the value
'NO'
if
unformatted
is not an allowed form of the file,
and the value 'UNKNOWN' if the form cannot be determined.

11-49

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
Table 11-6 summarizes the form and use of the FORTRAN
statements.
Table 11-6:

Summary of FORTRAN Device Control Statements
Statement Form

Section

FIND il~NIT 11l1.Ht:C rnl.EIW ~II.IOSTAT
FIND lun'rnl.Elm ~1I,lOSTAT j()~11

ioo;l!

IUU

HEWIND un
HEWIND (UNIT - unl ,ERR - s" ,IOSTAT = ios II
REWIND (unl,ERR ~ sll,IOSTAT'- iosl)

:I 1.8.2

l TNLOA D lIn
l'!'.'LOAD ((TNIT unl.EIW o;ll.IOSTAT
llNI.OAD (unl.EHH o;lI.IOSTAT io~11

ll.!-i.:l
io~1

BACKSPACE un
BACKSPACE (UNIT- unl,ERR sll,I08TAT-iosl)
BACKSPACE (unl,ERR=sll,IOSTAT-- ios))

I 1.8.4

ENDFILE un
ENDFILE (UNIT = unl ,ERR = s II ,lOST AT - ios))
ENDFILE (unl,ERR'=sll,IOSTAT~iosl)

11.8.5

SKII'J{JTOHf) un
11.S.G
SKIPRECOH]) dJNJT unl.END o;lIYlm 0;1I.108TA" iosll
SKII'HECOJ{I) iunl.r:ND o;lI.Elm 0;II.10STAT ioo;li
1 - - - - - - - - - - - - - - . - - - - - - - - - - - - - 1 - . - -. SKIPFILE un
llY'.I
SKIPFILE IT [NIT unl.EHH ~ II.IOSTAT ioo; II
8KII'FILE (unl.EHH.;II.IOSTA'1 iosll
I ;\('I(FI1.E lIn
IA(,KFII.E((TNIT unl.EHH 0;11.I08TAT
I ACKFII.E (lI11I.ERR 0;II.10STAT ius Ii

11.H.H

io',11

11-51

device

control

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
11.8

DEVICE CONTROL STATEMENTS

Device control statements enable you
to
position external
devices.
For
example,
when
performing data transfers with magnetic tape, you
use device control
statements to
position
the
tape.
The device
control
statements may be
used
for both formatted and unformatted
files.
The following list contains all of the device control statements,
the section in which each statement is described.

FIND

REWIND

(Section 11.8.2)

.J •

UNLOAD

(Section 11.8.3)

BACKSPACE

ENDFILE

SKIPRECORD

/'.

SKIPFILE

(Section 11.8.7)

BACKFILE

(Section 11.8.8)

]

and

(Section 11.8.1)

(Section 11.8.4)

(Section 11.8.5)
(Section 11.8.6)

NOTE
The results of the BACKSPACE and SKIPRECORD statements
are
unpredictable when
used
on
list-directed
and
NAMELIST formatted data.
The general form of every device control statement is:
keyword un
or
keyword

(UNIT=un [,END=s] [,ERR=s] r ,IOSTAT=ios])

or
keyword

(un[,END=s] [,ERR=s] [,IOSTAT=ios])

where:
keyword

is the statement name.

is the FORTRAN logical unit number.
Table
10-3
lists
the default
logical unit numbers.
If you
enclose
the
logical
unit
specification
In
parentheses,
you can include the keyword portion
(UNIT=) of the logical
unit
specification.
The
keyword
form
of
the
unit
specifier makes the
specification positionally
independent
in
the
parenthetical list.

END=s

is the optional end-of-file specifier
10.4.6) •

11-52

(see Section

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
ERR=s

is
the optional
10.4.7) •

IOSTAT=ios

is the optional I/O error
Sec t ion 1 'J • 4 • 8) "

error

specifier

(see

status

specifier

Section
(see

The operations performed by the device control statements are
usually
used
only for magnetic tape devices (MTA).
In FORTRAN, however, the
device control operations are simulated for disk devices.

[
11.8.1

- - - - --

--.------]

FIND
Statement
-----_.. -

FIND Statement

In earlier versions of FORTRA1-.J-IO/20, the FIND statement could be used
during
direct-access
READ/WRITE operations
to
reduce
the
time
necessary
to do data
tran:;fers
and
to
locate
records
in
a
direct-access
fi.le.
For FORTRAN-IO/20 Version 10, the FIND statement
has no effect, except for
se:ting
the
current
record
number
and,
therefore, the associate vari.:tble.
The first form of the FIND st.ltement is:
FIND

(UNIT=un,REC=rn [,ERH=s] [, IOSTAT=ios])

The second form of the FIND s~atement operates in the same way as
the
first
form.
The difference between the two forms is in the way that
the unit number and record number are specified.
The second form of this statement

c' •
1· ",

FIND(unl rn[ ,.ERR=s] [,IOSTAT=ios])
In this form, the unit number and record number do
not
contain
the
keywords UNIT= and REC=.
Ins·:ead, the unit number is specified first,
followed by a single quote ( I ' , and finally the record number.
The following example

demonst~ates

how the FIND statement is used:

PIND(UNIT=I,REC=lOO)
This statement positions the
rcads record 100.

~ile

that

the

READ

statement

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

REWIND
Statement

11.8.2

REWIND Statement

The REWIND statement, used for disk files, causes a specified file
to
be positioned at its initial point.
For magnetic tape, execution of
the REWIND statement positions the magnetic tape at its initial point.
The forms of the REWIND statement are:
REWIND un
or
REWIND (UNIT=un[ ,ERR=s] [,IOSTAT=ios])
or
REWIND (un[ ,ERR=s] [,IOSTAT=ios])
where:
un

I~EWIND

is the logical unit number of the device on which the REWIND
is being performed.
Table 10-3 lists the default logical
unit numbers.
is illegal for files opened with ACCESS='APPEND'.

r -- . _.-. -' ",-- -,-..--.._.......... -- --..._- -.-----,-.------- ..,-

UNLOAD
Statement

I
L,_______ .,. ,_.
I l.8.3

UNLOAD Statement

rne UNLOAD statement rewinds and unloads the magnetic tape
with the specified unit.
The forms of the UNLOAD statement are:
UNLOAD un
or
UNLOAD (UNIT=un[ ,ERR=s] [,IOSTAT=ios])
ur

LiNLOAD (un [ , ERR=s] [ , IOSTAT= ios] )

11-54

associated

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
where:
un

is the logical unit number of the device on whi.ch the UNLOAD
is being performe~.
Table 10-3 lists the default logical
unit numbers.

BACKSPACE
Statement

11.8.4

BACKSPACE Statement

Execution of a BACKSPACE statement causes the file
connected
to
the
specified unit to be positioned before the preceding record.
If there
is no preceding record, the position of the file is not changed.
If
the preceding
record
is an ENDFILE record (see Section 11.8.5), the
file is positioned before the ENDFILE record.
The BACKSPACE statement cannot be
used
for direct-access
files,
append-access files, or files that are formatted with list-directed or
NAMELIST-statement formatting.
The forms of the BACKSPACE statement are:
BACKSPACE un
or
BACKSPACE (UNIT=un [ ,ERR=~.] [ ,IOSTAT= ios] )
or
BACKSPACE (un [ ,ERR=s] [ ,IOSTAT= ios] )
where:
un

the
is the logical unit number of the device on which
BACKSPACE
is being performed.
Table 10-3 lists the default
logical unit numbers.

ENDFILE
Statement

11.8.5

ENDFILE Statement

The ENDFILE statement closes the
file on the specified
unit.
On
magnetic
tape,
an 'ENDFILE record' is written and is then positioned
after the end-of-file mark.
For disk, the file is closed and then positioned at the end of the
file, and an end-of-file status is set.
This status is the equivalent
of the file being positioned after an 'ENDFILE record'.

11-55

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
Thus, for both disk and magnetic tape,
a
BACKSPACE operation given
after an end file operation positions the file after the last data
record (that is, before the physical (or virtual) 'ENDFILE record').
The ENDFILE statement can be used only with sequential access files.
The forms of the ENDFILE statement are:
ENDFILE un
or
ENDFILE (UNIT=un [ ,ERR=s] [ ,IOSTAT= ios] )
or
ENDFILE (un[ ,ERR=s] [,IOSTAT=ios])
where:
un

is the logical unit number of the device on which the
Table 10-3 lists the default
ENDFILE is being performed.
logical unit numbers.

SKIPRECORD
Statement

11.8.6

SKIPRECORD Statement

The SKIPRECORD statement skips the record
immediately following
the
current
(last accessed)
record.
If the SKIPRECORD statement is
executed prior to accessing any records, then the first record in the
file is skipped.
You cannot use SKIPRECORD on direct-access files.
The forms of the SKIPRECORD statement are:
SKIPRECORD un
or
SKIPRECORD (UNIT=un [ , END=s] [ , ERR=s] [ , IOSTAT=ios] )
or
SKIPRECORD{un[ ,END=s] [,ERR=s] [,IOSTAT=iosj)
where:
un

is the logical unit number of the device on which the
SKIPRECORD is being performed.
Table 10-3 lists the default
logical unit numbers.

11-56

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS

-_.__.__ .=._-J-..
[

SKIPFILE
Statement
---. --------

11.8.7

SKIPFILE Statement

This statement is used only for magnetic tape operations.
Unless
an
end-of-file has been encountered, the SKIPFILE statement advances to
the
beginning
of
the next
file.
If
an
end-of-file has
been
encountered,
SKIPFILE skips the next file.
If the number of SKIPFILE
executions exceeds the number of files available
to be
skipped,
an
error occurs.
The forms of the SKIPFILE statement are:
SKIPFILE un
or
SKIPFILE

(UNIT=un [,ERR=s] [, IOSTAT=ios])

or
SKIPFILE

(un[ ,ERR=s] [,IOSTAT=ios])

where:
un

is the logical unit
number
of
the device on which
the
SKIPFILE
is
being performed.
Table 10-3 lists the default
logical unit numbers.

-_·_-_·_--J

BACKFILE
Statement

11.8.8

BACKFILE Statement

This statement is used only for
magnetic
tape operations.
If
an
end-of-file has been encountered, the BACKFILE statement positions to
the start of the file whose end-of-file was detected.
Otherwise,
the
BACKFILE
statement
positions
to the start of the file that precedes
the current (last accessed) file.

11-57

FILE-CONTROL AND DEVICE-CONTROL STATEMENTS
The forms of the BACKFILE statement are:
bACKFILE un
,~

F;ACKFILE

(UNIT=un r , ERR=s] [ , IOSTAT= ios] )

or
BACKFILE

(un [ , ERR=s] [, IOSTAT=s] )

\vhere:
is the logical unit
number
of
the
device
on which
the
BACKFILE
is
being performed.
Table 10-3 lists the default
logical unit numbers.

NOTE
On a magnetic tape with multiple files,
the
position
the
tape
after
an ENDFILE record of one file is
equivalent to the position at
the
beginning
of
the
next file.
of

11-58

CHAPTER 12
FORMATTED DATA TRANSFERS

Data transfers can be either formatted or
unformatted.
When the
internal
(memory)
representation of the data
is translated to a
different external (peripheral storage) representation during a data
transfer, that data transfer is considered formatted.
Conversely, when the internal and external representations of the data
are the same, that data transfer is considered unformatted.
A formatted data transfer
involves editing of data as
it
is
transferred
to and from memory.
FORTRAN provides you with three ways
for specifying how the data
is
formatted during a
formatted data
transfer.
These are:
1.

FORMAT-Statement Formatting

List-Directed Formatting

NAMELlST-Statement Formatting

Of the three types, FORMAT-statement formatting provides you with
the
most control over how the data is formatted.
Section 12.1 describes
FORMAT-statement formatting.
List-directed formatting means that the formatting
is controlled by
the data
types of the l/C
list elements.
Section 12.5 describes
list-directed formatting.
NAMELlST-statement formatting is the third method for
formatting
the
data;
the
formatting is controlled by the data types of the namelist
elements.
In this form,
the
I/O list
is defined
in a
NAMELIST
statement and referenced by the data transfer statement.
Section 12.6
describes NAMELlST-statement formatting.

12-1

FORMATTED DATA TRANSFERS

FORMAT-Statement
Formatting

12.1

FORMAT-STATEMENT FORMATTING

A FORMAT statement directs the editing of data during
its transfer
between internal and external storage.
Every formatted
(FORMAT
statement) data transfer statement contains a reference to one of the
following:
1.

A line containing a FORMAT statement with a format list

A numeric array containing a format list

A character expression containing a format list

An integer, real, or logical variable that has been
a FORMAT statement number with an ASSIGN statement

assigned

The format list is made up of format specifiers.
During execution of a formatted data transfer statement, items in the
I/O list are associated with specifiers in the referenced format list.
The specifiers dictate how the various data items are formatted.
Section 12.1.1 describes how to create a
format list in a
FORMAT
statement;
Section
12.1.2
describes
how to create a
format
specification as a character expression.
Section 12.1.3 describes how
to create a numeric array that contains a format list.
Section 12.1.4
describes how to specify a FORMAT statement using an ASSIGNed
variable.

12.1.1

Specifying a Format List in a FORMAT Statement

The general form of a FORMAT statement is:
n

FORMAT fs

where:
n

is the required statement number.
This number,
referenced
in the control-information list of an
I/O statement (see Section 10.4.2)
provides the
association between the data transfer statement
and the FORMAT statement.

is a format specification.
specification is:
([format list])

12-2

The form of

format

FORMATTI!D DATJll TRANSFERS
where:
format list

is a list of items
following forms:

which

may

take

any

the

[r] ed
or
ned
or
[r]

(fl)

where:
r

is a nonzel~O, unsigned, integer constant called
repeat specification.

is a
repeatable
12.2.1) •

edit

descriptor

(see

Section

ned

is a nonrepeatable edit
12.2.2) •

descriptor

(see

Section

is a nonempty format list.

The only placement restrictions for FORMAT statements are that they
follow PROGRAM,
FUNCTION,
SUBPROGRAM, or BLOCK DATA statements, and
that they precede the END stat.~ment.
The following example illustrates FORMAT-statement formatting.
The
FMT specifier
in the WRITE statement references the label of FORMAT
statement 101.
This FORMAT statement contains a
list of edit
descriptors
(X,
I, and F) that dictate the formatting of the data in
I/O 1 i s t ( va ria b 1 E~ S J, Y, and :~).

101

12.1.2

J=2
Y=3.0
Z=S.l
WRITE(UNIT=S,FMT=lOl)J,Y,Z
FORMAT(lX,I,F,F)

Specifying a Format Specification as a Character Expression

You can store format specifications in character variables, character
arrays,
character array elements, character substrings, or character
expressions.
A character format specification must be of the form described
in
Section 12.1.1. Note that the form begins with a left parenthesis and
ends with a right parenthesis. Charc~cter data may follow the right
parenthesis that ends the format sp.~cification, with no effect on the
format specification.
Blank characters may precede the
format
specification.
If the format
identifier
is a character array name,
the format
specification may be contained in more than the first element of the
array.
(A character array format specification is considered to be a
concatenation of all
the array .~lements of the array in the order
given by array element ordering (see Section 4.3.3).)
12-.3

FORMATTED DATA TRANSFERS
However, if a character array element name is specified as a format
identifier, the length of the format specification must not exceed the
length of the array element.
The following example shows the same format specification used in the
examples
in
Section
12.1.1.
This time, however,
instead of
referencing the format specification by
statement
number,
or
referencing the name of a numeric array, the data transfer statement
references the name of the character variable in which the format
specification is contained.
J=2
Y=3.0
Z=5.1
CHARACTER FORNAM*lO
FORNAM = '( 1 X, I , F, F) ,
WRITE(UNIT=5,FMT=FORNAM)J,Y,Z

12.1.3

Specifying a Format Specification in a Numeric Array

An alternative to using FORMAT statements
specification in a numeric array.

store

the

format

The format specifications are associated with a data
transfer
statement
by referencing the array name containing the format
specification, instead of a statement label of a FORMAT statement.
The following example shows the same format specification used in the
example in Section 12.1.1. This time, however, instead of referencing
the format specification by statement number,
the data transfer
statement references the name of the numeric array in which the format
specification is contained.
T~TEGER FORNAM(2)
FORNAM (I) = (IX, I
FORNAt-1 ( 2 ) = F, F)
I
I

J=2
y= 3.0
2::::5.1

WRITEfUNIT=5,FMT=FORNAM)J,Y,Z
END
In the above example the format specification is stored in both words
array FORNAM.
This is because the format contains ten characters:
the first five are in FORNAM(l); and the last five are in FORNAM(2).

NOTE
When storing a format specification in an array,
always include the outer most parentheses enclosing
the form~t specifiers.
Note that the word FORMAT
should not be included in the string.

12.1.4

Specifying a FORMAT Statement Using an ASSIGNed Variable

Integer, rcal, or logical variables that have been ASSIGNed FORMAT
statement numbers can be used as format specifiers.
(See Section 8.3
for information on the ASSIGN (statement label) assignment statement.)
12-4

FORMATTED DATA TRANSFERS
The variable is assigned a statement number by an ASSIGN statement.
that
refers to the
The
format
specifier
references
the variable
statement number it has been assigned.
The following example shows thE! same format specification used in
the
examples
in Sections 12.1.1, :2.1.2, and 12.1.3.
This time, however,
instead of referencing the forrlat specification by statement number,
the data
transfer
statement
references a variable
that has been
assigned a statement number by an ASSIGN statement.

101

12.1.5

ASSIGN 101 TO IFORMT
WRITE(UNIT=5,FMT=IFOHMT)J,Y,Z
FORMAT ( lX, I , F , F)

The Ordering and Interpretation of Format List Items

For standard conforming programs, all items within the
format
list
should
be separated by comMas, with the exception of the following
cases:
1.

Between a P edit descriptor and an immediately following
E, D, or G edit descriptor (See Section 12.4.11)

Before

after a sla:,h edit descriptor

(See Section 12.4.5)

Before or after a colon edit descriptor

(See Section 12.4.6)

Ol~

In FORTRAN-I0/20, the use of c.)mmas to delimit format edit descriptors
wit h i n a for mat 1 i s t i s 0 P t i 0:1 a 1 as 1 0 n gas no am big u i t Y ex i s t s .
Fo r
example,
FORMAT (3X,A2)
can be written as
FORMAT (3XA2)
Rut the specification
FORMAT (12215)
is ambiguous, since it can represent
FORMAT (122,15)

FORMi\ T (1,22 15)

and requires the comma to eliminate ambiguity.

12-5

FORMATTED DATA TRANSFERS

FORMAT-Statement
Edit Descriptors

12.2

EDIT DESCRIPTORS

Edit descriptors within the format list describe the manner of editing
performed on the data being transferred.
For example, when you transfer integers from a file to memory, you use
an
I
edit descriptor.
When the data transfer statement is executed,
an item in the I/O list is associated with the I
edit descriptor
in
the format list, and the following results:
1.

Before being stored in memory, the data is converted
to an
internal
integer
format
by the
I
edit descriptor in the
format list.

The memory location in which the data is stored is identified
by the I/O list element.

The following sample program demonstrates how an integer is read
from
the
terminal (external device) into the memory location identified in
the I/O list of the ACCEPT statement.
PROGRAM FORMAT

101
102

TYPE *,'Please enter a two digit number:'
ACCEPT 101,K
FORMAT(I2)
TYPE 102,K
FORMAT(lX,IS)
END

The sample output below shows what happens when the user executes the
above program.
The
user
enters 78
in
response
to
the ACCEPT
statement.
This causes the integer value to be stored in the variable
K according
to
the I edit descriptor in FORMAT statement 101.
Then
the type statement causes the value of variable K to be printed at the
terminal according to the I edit descriptor in FORMAT statement 102.
EXECUTE TEST
LINK: Loading
[LNKXCT FORMAT execution]
Please enter a two digit number:
78
78
Elapsed Time 8.7
CPU Time 0.1
The I edit descriptor is an example of a repeatable edit descriptor.
FORTRAN has
two
types of edit descriptors:
repeatable (Section
12.2.1) and nonrepeatable (Section 12.2.2).
The third
type of
item
that appears
in a
format
list
is
the carriage-control specifier
(Section 12.2.3).

12-6

FORMATTED DATA TRANSFERS

Repeatable Edit
Descriptors

12.2.1

Repeatable Edit Descriptors

A repeatable edit descriptor may be preceded by an optional, unsigned,
nonzero, integer constant that specifies a repeat count.
This integer
is called a repeat specification.
Using a repeat specification in an edit descriptor gives you a
shorthand way to specify multiple fields with a single specification.
For example, without using the repeat specification, if you wanted
to
specify four
fields,
each of which contain an integer value that is
six characters long,
you might construct the
following
FORMAT
statement:
101

FORMAT (lX,I6,I6,I6,I6)

If you use the repeat specification, however, you need only
the edit descriptor and field width a single time, as follows:
101

specify

FORMAT (lX,416)

These two FORMAT statements are equivalent.
Table 12-1 lists the repeatable edit descriptors.
Each descriptor
listed
in the table
is shown in its complete form.
The key at the
bottom of Table 12-1 describes all the optional elements in each edit
descriptor.
The
right-most column of Table 12-1 references the
section in which each edit descriptor is discussed.

12-7

FORMATTED DATA TRANSFERS
Table 12-1:

Repeatable FORTRAN Edit Descriptors

Edit Descriptor

Descriptor Type

Hefer to:

Irlllwl.mll

Integer

Section 12.4.11.1

IrJFI w.dl

Floating Point

Section 12.4.11.2

IrlEI w.dlEell

Scientific Notation

Section 12.4.11.3

IrlD! w.dl Eell

Scientific Notation

Section 12.4.11.3

IrlGlw.dlEell

General Conversion

Section 12.4.11.4

F,E,D,G
(Two successive)

Complex

Section 12.4.11.5

IrlOlwl.mll

Octal

Section 12.4.11.6

IrlZlwl.mll

Hexadecimal

Section 12.4.11.7

IrJLlwl

Logical

Section 12.4.12

IrJAlwJ

Character or Hollerith

Section 12.4.13

IrlRlwl

Hollerith

Section 12.4.14

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

_..-

_._.--

---

_.-

Key:
r

is a nonzero, unsigned, integer constant called a repeat specification.

is a nonzero, unsigned, integer constant which is equal to the total number of characters in the numeric field
being described. The numeric edit descriptors are described in Section 12.4.11.

.m is an unsigned, integer constant which specifies the minimum number of digits to be output to the field
being described. If necessary, leading zeros are output. The value of m must not exceed the value of w.
If m is zero and the value of the internal data item is zero, the output field consists of only blank characters,
regardless of the sign control in effect.
.d

is a nonzero, unsigned, integer constant which specifies the total number of digits to the right of the decimal
point in the numeric field being described. If .d is specified, w must also be specified. The maximum value is
63 digits.

is a nonzero, unsigned, integer constant which is equal to the total number of digits in the exponent field of
the numeric field being described. The maximum value is 15 digits.

Nonrepeatable Edit
Descriptors

12.2.2

Nonrepeatable Edit Descriptors

A nonrepeatable edit descriptor can not be preceded by a repeat
specification.
The nonrepeatable edit descriptors provide a vaI-iety
of editing possibilities, such as positioning within a record,
including character constants in a FORMAT statement, and delimiting
records within a single format descriptor.
Table 12-2 lists the nonrepeatable edit descriptors.
The format,
function,
and section number where each descriptor is discussed are
listed in the table.

12-8

FORMATT3D DATA TRANSFERS
Table 12-2:

Nonrepeatable FORTRAN Edit Dpscriptors
mction

Edit Descriptor
'h1. .. hn'

Character Data

Refer to:
Section 12.4.1

------

nHh

Hollerith Data

Section 12.4.2

Tc
TLc
TRc

In-Hecord Positio ning

Section 12.4.3.1

--------

InlX

In-Hecord Positio ning

Section 12.4.3.2

(Do dar sign) Pre" ents recl lrd from terminating with END
OF LINE

Section 12.4.4

------!

(Slash) Record De limiter

Section 12.4.5

(Colon) Format-C ontrol TI !rmination

Section 12.4.6

--------

S
SP
SS

PIll!; sign control for outpl It of positive numeric fields

Section 12.4_7

Scal ing Factor fo r Numer c Fields

Section 12.4.8

BN
BZ

Specifies the ha ndling 0 f blanks during the input of
Numeric Fields

Section 12.4.9

Input Only Descriptor - returns the number of characters left in the cu rrent re< ord.

Section 12.4.10

--------

Key:
n

is a nonzero, unsigned, integer consta nt whicr is equal to a number of spaces (X descriptor) or the total
number of characters (H descriptor).

is a character capable of representation by the p rocessor. This type of character is described in Appendix B.

is a nonzero, unsigned, in teger constan t which i s equal to a number of characters within a record relative to
the current position.

is an optionally signed integer constan t which ::!eclares the scaling factor for the field being described.
--------

Carriage-Control
Specifiers

12.2.3

Carriage-Control Specifiers

In a data output transfer, the first character of each record can be
used for carriage control. A carriage-control specifier dictates the
action of the printing mechanism on output devices.
For example,
carriage-control
specifiers
determine the vertical spacing for
line-printer output.

12-9

FORMATTED DATA TRANSFERS
NOTE
The CARRIAGECONTROL specifier of the OPEN statement
enables you to decide how the first character of each
record is treated. Depending on the value of the
CARRIAGECONTROL specifier, the first character can be:
the

printer-control

Replaced with
character (s) .

appropriate

Disregarded as a carriage-control character and,
instead, be transferred as part of the record.

For more information on the CARRIAGECONTROL specifier,
see Section 11.3.6
The carriage-control specifier may be written as a character constant.
The following example shows the blank carriage-control character in a
FORMAT statement:
WRITE(5,101)
101

FORMAT('

,'This is a string')

END
When this example is executed,
the string
printed on unit 5, the terminal, as follows:

the

format

list

This is a string
If you omit the carriage-control specifier from a data output transfer
format list, FOROTS interprets the first character to be output to the
record as the carriage-control character. using the example above, if
we omit the blank specifier, FOROTS assumes that the first character
encountered (in this case, the "T" in "This") is the carriage-control
character.
Executing
this
example,
after
removing
the
carriage-control specifier, causes the first character to be stripped
from the character constant. Thus, the output at the terminal is:
his is a string
The carriage-control characters are summ~rized in Table 12-3.
The $
(dollar sign)
output edit descriptor modifies the action of the
carriage-control specifier (see Section 12.4.4).

12-10

FORMATTED DATA TRANSFERS
Table 12-3:

Carriage-Control Specifiers
--------.~-------

Specifier

Format List Form

Pr inter Ct·aracter

-r--"---

.. - - ...

Octal Value

Effect on Carriage Control

012

Skip to next. line (form feed after
60 lines on printer).

- - - - - - . - - - - - - - - - .. - ... ·. r-·-·--. . _-·
blank

... _ - - - - I - - -- - - - ---'--"-'- ~--. - - - - - - - - + - - - . -

plus

'+'

zero

'0'

one*

'1'

-_._ -_._---

..-

.....

-----_._---

'2'

three'"

'3'

minus

asterisk'"

'*'

period*

' ,

comma'"

' ,

slash*

'/'

...

012,012

I-----------~.--

014

- - - - - - I---- - - - _ . _ - -

two*

~-------

LF,LF

Suppress line feed: overprint tl1<'
line .
............ _ _--Skip a line.

-- - - - - -_
...

DLE

020

013

... _ - - - - I - - -- - - - - _ . _ - - - - - - - _ ..

---_ ..

012,012,012

....

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

Skip two lines.

- - - - - - - - - - - - - - - - f - - - - - -. - ....-----

DC3

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

__-

Space to next half page .
_--_ .._ __ _ __
Space to next one-third of a page .

.._ - - - - r---- - - - - . - - - - - - - - - - - - _ .. -

LF,LF,LF

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

Form feed to top of next page.
..........

023

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

Skip to next line; suppress form
feed. (Continous print)

- - - - - . - - - - - - - - - - - - f - - - - -......- - -- .-- - -. . -.-----

DC2
.._---- -

022

Triple space. with a form feed
after every 20 lines printed.

- - - - - - . - - -.--------t--------.-.-- ---------

DC1

021

Double space, with a form feed
after every 30 lines printed.

- - - - - . - - - - - - - - - - - - - - 1 - . _ - _ . _ _ ._- - - - - - - -

DC4

024

_ _ _ _ _ •_ _ _•...L- _ _ _ _ _

Space to next one-sixth of a page.
--'~

_ _ _ _ _ _ _ . _ ... _ _ _ _ .. _ _ _ _

Indicates carriage-control specifiers for which the effect on carriage control is device dependent. The effpct
described is for a line printer with a standa! d form set.up.

Note - This table assumes a st.andard form setlip for your line printer (or other output dcyic('l.

1/0 List & FORMAT List
Interaction

INTERACTION OF

12.3

INPUT/OU~'PUT

This section describes how
during a data transfer.

12.3.1

thE~

LIST AND FORMAT LIST

I/O list and the format

list

interact

General Description

Format control is initiated by execution of a formatted data transfer
statement.
The actions performed by format control depend on the
interaction of the edit descrjptors in the format specification and
the I/O list elements in the data transfer statement.
The following example shows how the I/O list elements in a data
transfer statement interact with the edit descriptors in the format
list in a simple data transfer.
100

READ (5, 100) N, X, Y
FORMAT (I5,FI2.0,F'10.O)

12-11

FORMATTED DATA TRANSFERS
In this example, the I/O list is
N,X,Y
and the format specification is
(I5,F12.0,FI0.0)
The variables in the I/O list and the specifiers
up as follows:

the

format

are

mat~hed

I5
F12.0
FI0.0

A formatted data transfer statement matches elements of the
I/O list
and specifiers
in the
format specification.
The matching proceeds
from left to right, one
I/O list element to one
repeatable edit
descriptor.
In the above example, there are three elements in the
I/O list and
three
format specifiers.
However,
the
interactions can be more
complicated than those in the example.
A format specifier can be
preceded by a repeat count, in which case it corresponds to more than
one element in the I/O list.
Also, an element of the I/O list can be
an array name, in which case it can correspond to more than one format
specifier.
The number of elements in the I/O list and in the
format
specification do not have to be the same.
Table 12-4 details what happens in these more complex cases.

12-12

FORMATTED DATA TRANSFERS
Table 12-4:
1.

Record, Format List, and 110 List Interaction

Record ends
Format specification continues
110 list continues

Action:
The transfer continues as if the record was extended with blanks.
Example:
10

READ (5 dO) A ,8 ,C
FORMAT (3F10.0)

Record contains:
40

Resulting values:
A = 40.0
B
10.0
C '"" 0.0
Note that this situation :is not applicable to an output transfer.

2. 110 list ends
Format specification continues
Action:
The format scan continues until it encounters 1 repeatable edit descriptor, a colon, or until the rightmost right
parenthesis of the format is reached.
Nonrepeatable edit descriptors up to and including the first colon, if present, are processed if they are encountered during the scan.
Example:

12
123
C
1234
D = 12345
B

WRITE (5,20) A,8,C
FORMAT (' A'= I ,F3.0,
1
D= I ,F6.0)

8= I ,F4.0,

C= I ,F5.0,

Resulting output:

A= 12., B = 123., C = 1234., D =
Note that the', D =' descriptor was processed. The format scan does not terminate until it encounters a descriptor which requires an 1'0 list element, and 1here is no 110 list element to supply. However, the colon edit
descriptor will cause the format scan to stop if there is nothing left in the 110 list (see Section 12.4.6J.

3. Format specification ends
110 list continues
Action:
A new record is started. The format scan contiIlues, starting at the beginning of the last complete parenthesized
h1TOUP within the format specification. If there is no parenthesized group within the format, the format is
restarted from the beginning (see Section 12.34).
Example:
A=123.
8=7654.125
C=1.6125

WRITE (5 dO) A ,8 ,C
FORMAT (F15.6)

12-13

FORMATTED DATA TRANSFERS
Table 12-4: Record, Format List, and 1/0 List Interaction (Cont.)
Resulting output:
123.000000
7654.125000
1.612500
4.

1/0 list ends

Format specification ends
Action:
On input, if there are any characters remaining in the record, they are ignored. On output, the record is simply
terminated without any extra characters added.
Example:
10

READ (5 ,10) GAMMA
FORMAT (F5.2)

Record contains:
12.34 value of GAMMA
Resulting values:
GAMMA = 12.34
The extra data (the comment) in the input record is ignored.

The execution of a formatted I/O statement proceeds by matching I/O
list elements and FORMAT edit descriptors.
The edit descriptors I, 0,
Z, R, F, E, 0, G, L, A, and Q each correspond to one element of the
I/O list. No I/O list element corresponds to H, X, P, T, :, $, S, SP,
SS,
BN,
BZ, or apostrophe edit descriptors.
If one of these
descriptors
is encountered,
it is executed and the
format scan
continues.

12.3.2

Formatted Input

A formatted input statement begins by reading a record
from the
specified unit.
The format is scanned from left to right.
X, P, T,
BN, BZ, and / edit descriptors are executed as they are encountered.
If an
I,
0, Z, R, F, E, 0, G, L, A, or Q descriptor is encountered,
data is read into the corresponding I/O list element as specified by
the edit descriptor.
If the I/O list contains no more elements,
execution of the READ
statement ends.
Additional
records will be read from the specified
unit when a slash occurs
in the
format,
or when the last right
parenthesis of the format is reached, and I/O list elements remain to
be filled.
When an input record is terminated by a slash or by the end of the
format,
any data left in the input record is discarded.
If the input
record is exhausted before the data transfers are completed,
the
remainder of the transfer is completed as if the record were extended
with blanks.

12-14

FORMATTED DATA TRANSFERS
12.3.3

Formatted Output

A formatted output statement begins by scanning the format.
The H, X,
P,
T, BN, BZ, :, /, $, s, SS, SP, and apostrophe edit descriptors are
executed as they are encountered.
If an I, 0, Z, R, F, E, D, G, L, or
A descriptor is encountered, data is translated from the corresponding
I/O list element, as specified by the edit descriptor, and placed
in
the output record.
If the I/O list contains no more elements,
the output record
is
written to
the specified unit, and execution of the WRITE statement
ends.
Additional records will be written to the specified unit when a
slash occurs
in the format,
or when the last right parenthesis is
reached, and I/O list elements remain to be transferred.

12.3.4

Embedded Format Specifications

Format specifications may contain embedded format specifications with
optional
repeat specifications.
If a repeat specification is used,
the entire format specification that
it precedes
is scanned the
specified number of times during the I/O transfer.
In the example:
100

WRITE (1,100) A,B,C,D,E,F,G,H,I,J
FORMAT (FlO.2,4(I5,lX,I3) ,18)

the variable A is matched with the
format
item FlO.2.
Then,
the
variable B is matched with IS, variable C with 13, variable D with IS,
and so on for four iterations of the embedded
format specification.
Finally, the variable J is matched with the format item 18.
If no repeat specification is used preceding
specification, a repeat count of 1 is implied.

embedded

format

When the last right parenthesis of the format is reached, and more I/O
list elements remain to be transferred, a new record is started and
format scanning continues.
The scanning continues at the beginning of
the
format specification whose right parenthesis is the next to last
right parenthesis in the format.
If there are no embedded
format
specifications,
format scanning continues at the beginning of the
format.
For example:
DIMENSION A(lOO)
INTEGER CASE
CASE=33
100

WRITE (l,lOO)CASE, (A{I) ,1=1,100)
FORMAT ('THIS IS CASE' ,I6,//,4{lX,FI0.5»

After A{l) through A(4) are written, a new line is started and
format
scanning
continues
at
the
beginning of the embedded format
specification. Thus, A(5) through A(8) are written; a new record
is
started, and so forth.

12-15

FORMATTED DATA TRANSFERS
The output file would appear as follows:
THIS IS CASE 33
A (1)

A (2)

A(5)
A(9)

A(6)
A (10)

A(3)
A(7)
A (II)

A(4)
A(8)
A (12)

Example:
DIMENSION A(5) ,DATl(100) ,B(4,100) ,DAT2(100)
WRITE (1,100) CASE, (A(K) ,K=1,5), (DATI (J), (B{I,J) ,I=1,4),
1 DAT2 (J) ,J=l, 100)
100 FORMAT ('CASE' ,I5,11,5(lX,FI0.3) ,1,{FI0.3,4(3X,F15.5)
1 ,lX,FI0.3»
In this example, after A{l)
through A(5),
DAT1(1),
B(l,l)
through
B(4,1),
and DAT2(1) are written, a new record is started, and format
scanning begins. The format scanning begins at the embedded
format
specification
following
the 'I'
(the specification whose right
parenthesis is the next to last right parenthesis) •
The output file would appear as follows:
CASE
A{l)
DATl(l)
DATl(2)
DATI (3)

12.4

33
A (2)
B{l,l)
B(1,2)
B(1,3)

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

A(4)
B{3,1)
B(3,2)
B{3,3)

A(5)
B(4,1)
B(4,2)
B{4,3)

DAT2 (l)
DAT2(2)
DAT2(3)

FORMAT EDITING

Tables 12-1 and 12-2 describe forms of all
the
FORMAT
edit
descriptors.
The edit descriptors enable you to specify the form of a
record and to specify the editing of the data as it is transferred.
The edit descriptors are described according to the character used to
accomplish a particular modification to the data or record in ~lhich
the data are stored.

APOSTROPHE (')
Editing

12.4.1

Apostrophe (') Editing

The apostrophe (')
edit descriptor
(single-quote)
include a character constant in a format list.

12-16

enables

you

FORMATTED DATA TRANSFERS
The form of the apostrophe edit descriptor is:
'hl ••• hn'
where:
'hl ••• hn'

is a character constant.

To include an apostrophe as part of the character constant,
use two successive apostrophes.
This descriptor is only used for output; the
the apostrophes are written.

characters

you

must

enclosed

Example:
10

TYPE 10
FORMAT (' That' 's the way!')

will output
That's the

I~ay!

H
Editing

12.4.2

H Editing

The H edit descriptor (also called the Hollerith descriptor)
you to include character strings in a format list.

enables

The form of the H edit descriptor is:
nHhl ••• hn
where:
n

is a nonzero, unsigned, integer constant that indicates
the total
number of ASCII characters included in the
string.

hI ••• hn

is a string of ASCII characters.
set is described in Appendix B.)

(The ASCII

character

You may transmit alphanumeric data directly from the FORMAT
using either the H or apostrophe specifiers.

statement

This descriptor is only used for output; the n characters that follow
the H are written.
For example, you can use the following statement
sequence to print the words PROGRAM COMPLETE on the printer:
PRINT 101
I
101 FORMAT (17HlpROGRAM~COMPLETE)

12-17

FORMATTED DATA TRANSFERS
The result of apostrophe editing is the same
For example, you may use the descriptors:

Hollerith

editing.

101 FORMAT (17H~PROGRAM~COMPLETE)
and
101 FORMAT ('~PROGRAM~COMPLETE')
in the same manner.
Apostrophes can appear anywhere within a
Hollerith edit descriptor
without having to be represented by two apostrophes.
However, if the
H edit descriptor occurs within a character constant,
the apostrophe
is written as two apostrophes, which are counted as one character.

POSITIONAL
Editing

12.4.3

positional Editing

The positional edit descriptors specify the position at which the next
character will be transmitted to or from the record.
The positional
edit descriptors are:
T, TL, TR, and X.
The T edit descriptor specifies the character position within a record
where the next character will be transmitted (see Section l2.4.3.1).
The TL and TR descriptors specify the number of character positions to
the left or
right,
respectively,
of the current position for the
character position of the next character (see Section 12.4.3.1).
The X descriptor specifies the number of character positions to the
right of the current position for the character position of the next
character (see Section 12.4.3.2).
NOTE
On output, a record is initially filled with blanks.
Therefore,
fields skipped by the positional editing
descriptors will be blank-filled.
However, the output
record
length is determined by actual output. Merely
specifying a positional editing descriptor with no
output will not change the record size.
Thus, the
record written with:
FORMAT (I6,50X,T10,I3)
will have a record length of 13 characters,
output was done after the 50X.
Examples:
The statement sequence:
PRINT 2
2 FORMAT (T50,'BLACK' ,T30,'WHITE')
12-18

since

FORMAT~(,ED

DATA TRANSFERS

causes the following line to be printed:
BLACK

lHITE
(print position 30)

(print position 50)

The statement sequence:
1 FORMAT (T35,'MONTH')
READ (2,1)
causes the first 34 characters of the
input data associated with
logical unit 2 to be skipped, and the next five characters to replace
the characters M, 0, N, T, and H in storage.
If an input record containing:

ABCP{~J6XYZ
is read with the format specification:
10 FORMAT (T7,A3,Tl,A3)
then the characters XYZ and ABC are read in that order.
You can use the field descriptor nX to introduce blanks into output
records or to skip characters of input records.
The letter X
specifies the operation, and n is a positive integer that specifies
the number of character positions to be either made blanks (output) or
skipped (input).
The statement:
FORMAT (5H~3TEP,I5,10X,2Hy=,F7.3)
may be used to print the line:

T, TL, and TR
Editing

12.4.3.1 T, TL, and TR Editing - The T edit descriptor specifies that
the transmission of the next character to or from a record is to occur
at the specified character position.
The form of the T descriptor is:
Tc
where:
c

is a positive, unsigned, integer constant that indicates the
character position to or from which the next character will
be transferred.

For example:
FORMAT (T20,I5,T8,I2)
12-19

FORMATTED DATA TRANSFERS
specifies that the characters read or written for
the 15 descriptor
will
start
in character
position 20,
and
the characters read or
written for the 12 descriptor will start in character position 8.
For
out put
to
car ria g e - con t r o l d e vic e s
( 1 i n e p r i n t era n d t e r min a 1), Tn
specifies that n-l will be
the next character written,
since
the
character position 1 is the carriage-control character position.
The TL edit descriptor specifies that the
transmission of
the next
character
to or from the record is to occur at a position which is a
specified number of positions backward from the current position.
The form of the TL edit descriptor is:
TLc
where:
c

is a positive, unsigned, integer constant that indicates the
character
position,
c
positions backward from the current
position, to or
from which
the
next character will
be
transferred.
If c
would cause transmission to start at a
position before
the beginning of
the current
record,
transmission will start instead at position one.

For example:
FORMAT (I5,T13,A5,TLI0,I2)
specifies that the characters read or written for
the A5 descriptor
will
start at character position 13, and that the characters read or
written for the 12 descriptor at will start at character position 8.
The TR edit descriptor specifies that the
transmission of
the next
character
to or
from a
record
will occur at a position that is a
The
specified number of positions forward from the current position.
function of
this
form is identical to that of the X descriptor (see
Section 12.4.3.2).
The form of the TR edit descriptor is:
TRc
where:
c

is a positive, unsigned, integer constant that indicates the
character
position,
c
positions
forward from the current
position, from which the next character will be transferred.

Example using all three types of T descriptor:
10

TYPE 10
FORMAT('

TYPE 20
FORMAT(T29,'BLACK' ,TIO,'WHITE')

234567890123456789012345678901234567890')

TYPE 30
FORMAT(TI0,'9012' ,TR5,'8901')
TYPE 10
TYPE 40
FORMAT(T20,'<SECOND)' ,TLI0,')FIRST<')

END
12-20

FORMATT~D

DATA TRANSFERS

EXECUTE TEST2.FOR
FORTRAN: TES~~2
MAIN.
LINK:
Loading
[LNKXCT TEST2 execution]
2345678901234567890123456 '78 90 1234 56 78 90
WHITE
BLACK
9012
8901
2345678901234567890123456'78901234567890
)FIRST<ND>
CPU time 0.1

Elapsed

tim~

0.5

In FORMAT 20, 'BLACK' is writt~n, then 'WHITE' is written to the left
of it.
In FORMAT 30,
fiv~
positions are skipped between the two
character strings being printed.
In FORMAT 40, '<SECOND)' is written,
the format goes back ten positions and writes ')FIRST<' over the
previously written character string.

X
Editing

12.4.3.2 X Editing - The X edit descriptor specifies that
the
transmission of characters to or from a record will occur a specified
number of characters forward from the current position.
The form of the X edit descriptor is:
(n]X
where:
n

is an optional, unsigned, positive,
integer constant that
indicates the number of characters forward from the current
position, at which the next character will be transmitted.
The default value is 1.

Example:
10

TYPE 10
FORMAT(' 1234567890123456789012345678901234567890')

TYPE 20
FORMAT(' A WORD OR TWO' ,10X,'OR THREE')
END

EXECUTE TEST3.FOR
FORTRAN: TEST3
MAIN.
LINK: Loading
[LNKXCT TEST3 execution]
1234567890123456789012345678901234567890
A WORD OR TWO
OR THREE
CPU time 0.2

Elapsed time 2.1
12-21

FORMATTED DATA TRANSFERS
In this example, ten positions are skipped between the printing of
'A WORD OR TWO' and 'OR THREE'.

$
(DOLLAR SIGN)
Editing

12 • 4 .4

(Do 11 a r S i g n) Ed i t i ng

The $ (dollar sign) output edit descriptor
suppresses all carriage
control at the end of the current record (for CARRIAGECONTROL='LIST')
or at the beginning of the next record (for CARRIAGECONTROL='FORTRAN'
or 'TRANSLATED').
This descriptor is used for interactive I/O; it leaves the terminal
position at the end of the text so that a response will follow the
output on the same line.
Example:
WRITE (5,10)
READ (5, *) N
FORMAT (' Number of samples:'
WR I TE ( 5 ,20 )
READ (5,*) X
FORMAT (' Mean value:
END

10
20

$)
$)

If the user enters 100 for Nand 1.23 for
will produce the typescript:

executing

the

program

the

several

Number of samples: 100
Mean value:
1.23
The $ edit descriptor can be used to append
statements into a single line.
For example:
DO 10 I = 1,10
WR I TE (5, 20) I
FORMAT (lX,I3,$)
WRITE (5,20)
END

10
20

will produce one line of output:
1

The $ edit descriptor is ignored for input.

12-22

output

FORMATTED

DAT1~

TRANSFERS

/ (SLASH)
Editing

12.4.5

(Slash) Editing

The / (slash) edit descriptor indicates the end of data transfer for a
record.
Two consecutive slashl!s indicate the transmission of an empty
record.
On input to a file connected
Eor sequential access,
the
remaining
portion of the current record is skipped, and the file is positioned
at the beginning of the next record.
This new record becomes the
current record.
On output to a file connected for sequential access,
the current record is terminated, and a new record is created, which
becomes the current and last rl!cord of the file.
A record that contains no char~cters may be written.
record may be skipped on input.

Also, an

entire

If the file is connected for direct access,
the
record number
is
increased by one, and the file is positioned at the beginning of the
record that has that record number.
This record becomes the current
record.
The following statements will write a record with no characters:
100

WRITE (1,100)
FORMAT(/)

To handle a group of
I/O records where different
records have
different
field descriptors,
use a slash to indicate a new record.
For example, the statement
FORMAT (308/15,2F8.4)
is equivalent to
FORMAT (308)
for the first record, and
FORMAT (15,2F8.4)
for the second record.
You may omit separating commas when you use a slash. When n slashes
appear at the beginning or end of a format, n blank records will be
written on output or skipped on input. When n slashes appear
in the
middle of a
format,
n-l blank records are written on output or n-l
records are skipped on input.

12-23

FORMATTED DATA TRANSFERS

: (COLON)
Editing

12.4.6

(Colon)

Editing

The:
(colon) edit descriptor terminates format control if there are
no more
items
in
the
I/O list of the corresponding data transfer
statement.
The colon edit descriptor has no effect if there are any
items left
in
the
I/O list of
the
corresponding data transfer
statement.
Example:
Xl
X2
X3
10
20

100
200
300

TYPE 10, Xl,X2
FORMAT(F6.1,F6.1,:,'THIS SHOULD NOT PRINT' ,F6.1)
TYPE 20,Xl,X2,X3
FORMAT(F6.1,F6.1, :,' BUT THIS SHOULD' ,F6.1)
END

EXECUTE TEST4.FOR
FORTRAN: TEST4
MAIN.
LINK:
Load ing
[LNKXCT TEST4 execution]
100.0 200.0
100.0 200.0 BUT THIS SHOULD 300.0
CPU time 0.1
Elapsed time 1.5

S, SP, and SS
Editing

12.4.7

S, SP,and SS Editing

The S, SP, and SS edit descriptors control the output of the optional
plus sign
in numeric output fields.
These descriptors are used as
follows:
S

indicates that the
system-defined
action
is
taken.
system-defined specification for Digital FORTRAN is SSe

indicates that the plus sign
numer ic output fields.

indicates that the plus sign
is not
printed
in positive
numeric output fields.
This descriptor is the default.
12-24

printed

all

The

positive

FORMATTED

DAT}~

TRANSFERS

The S, SP, and SS edit descrip~ors affect only
I,
F,
E,
D,
and
G
editing
during
the
executiOJl of
a data transfer output statement;
these edit descriptors have no effect during input
transfers.
These
descriptors
stay in effect un":il the end of the I/O transfer or until
another S, SP, or SS is encoun':ered.
Example:
Xl
X2
X3
X4
10

100
200
300
400

TYPE 10, XI,X2,X3,X4
FORMAT("
,S,F7.1, SP,F7.1, SS,F7.1, SP,F7.1)
END

EXECUTE TESTS.FOR
FORTRAN: TESTS
MAIN.
LINK:
Loading
[LNKXCT TESTS execution]
100.0 +200.,0
300.0 +400.0
CPU time 0.1
Elapsed time 0.2
In this example,
X2
and
X4
have
plus
signs
because
of
the
descriptors in front of the F descriptors that correspond to them.

p
Editing

12.4.8

P Editing

The P edit descriptor is used
and G edit descriptors.

":0

specify a scale factor for F,

representing

the

The form of the P edit descrip':or is::
kP

where:
k

is an optionally signed integer
scale factor.

constant

If a scale factor is not specified in a format list, a scale factor of
zero
is
assumed.,
Once a scale factor is specified in a format list,
that scale factor remains in effect for
all
F,
E,
D,
and
G edit
descriptors
unti1
a
new scale factor is specified, or to the end of
the execution of the current I/O statement.
Scale
factors
have
no
effect on I, Z, and 0 edit descriptors.

12-25

FORMATTED DATA TRANSFERS
The scale factor affects the F, E, D, and G data transfers as follows:
On input:
If there is an exponent field, the scale factor k has no effect.
If
there
is no exponent
field
on the number read in, the number is
multiplied by 10**(-k) before being assigned to the input variable.
On output:
The basic real constant part of the quantity, in E and D editing only,
is multiplied by 10**k and
the exponent
is reduced by k.
For G
editing, the scale factor has no effect unless the magnitude of the
data item to be edited is outside of the range that permits the use of
F editing.
If G editing is specified, and the magnitude of
the data
item to be edited is such that E editing is required, the scale factor
has the same effect as with E output editing (see Section 12.4.11.3).
The comma is optional between a
P edit descriptor
following F, E, D, or G edit descriptors.

and

For example, assume the data involved is the real number
edit descriptor

immediately
26.451;

the

F8.3
produces the external field

~~26.45l
The addition of the scale factor of -lP, as in
FORMAT (-lP,F8.3)
produces the external field

When you add a scale factor to D, E,
and G (external
field
not a
decimal fixed-point) edit descriptors, the scale factor multiplies the
number by the specified power of ten,
and
the exponent
is changed
accordingly.
In input operations, type F (and type G,
if the external
field
is
decimal
fixed-point)
conversions are the only ones affected by scale
factors.
When you add a scale factor to a D or E edit descriptor, it specifies
a power of 10 so that the e~ternal form of the number has its mantissa
multiplied by the specified power of 10;
its exponent
is adjusted
accordingly.
For example, assume the data involved is the real
edit descriptor
Ell.3
produces the external field
¥~0.125E+02

The addition of the scale factor 2P, as in
FORMAT (2P,E11.3)
12-26

number

12.49;

the

FORMATTED DATA TRANSFERS
produces the external field
Y/jS12.49E+00
With a scale factor of zero, the number of significant digits
by a format of the form:

printed

Ew.d
or
Dw.d
is the number of digits to the right of the decimal point.
For a negative scale factor nP,
for -d<n<O, there will be ABS(n)
leading zeros and d-ABS(n) significant digits after the decimal point
(for a total of d digits after the d,ecimal point).
If ns...-d,
there
will be d insignificant digits (zeros) to the right of the decimal
point.
If the scale factor nP is positive, for 0<n<d+2 there will be n
significant digits to the left of the decimal point and d-n+l
significant digits to the right of the decimal point (for a total of
d+l significant digits).
If n>d+2,
there will be d+l significant
digits and n-d-l insignificant trailing zeros on the left of the
decimal point.
If the data to be printed is 12.493, these formats produce results
follows:
FORMAT

OUTPUT

SIGNIFICANT
DIGITS

REASON

E15.3
IPE15.3
-lPE15.3
2PE15.3
-3PE15.3
4PE15.3
6PE15.3

J6J6J6J6;))1) 0 • 125 E+ 0 2

3
4

n=O
n<d+2
-d<n
n<d+2

J6t6J6)$,K>)S1.249E+Ol
)5JzS}6}6b))0.012E+03

~JzSt6}6J6J612.49E+OO

16)S)Sk):>,kS0.000E+05
J6t6,kSjSJzSjS12 4 9. E- 0 2

J6JzS X> pn 2 4 9 0 0 • E - :) 4

n~-d

n<d+2
n>d+2

Example:
10
20
30

TYPE 10
FORMAT(' Type in a real number')
ACCEPT 20,Xl
FORMAT(2P,F)
TYPE 30"Xl
FORMAT(' Number read with P=2 =' ,F,
1 I,' (Number read)*10**(-2) ')
TYPE 40~X1
FORMAT(/,' The above number written with P=2' ,I,
1 ' is:::', 2 P , F ,I "
( N urn b era b a v e) * 1 0 * * (2) , )
END

12-27

FORMATTED DATA TRANSFERS
EXECUTE TESTP.FOR
FORTRAN: TESTP
MAIN.
LINK:
Loading
[LNKXCT TESTP execution]
Type in a real number
5.
Number read with P=2 =
(Number read)*10**(-2)

0.0500000

The above number written with P=2
is =
5.0000000
(Number above)*10**(2)
CPU time 0.2
Elapsed time 5.1
The number the program receives is (5.)*(10**(-2))
out is (.05) * (10** (2)) •

and the value typed

BN and BZ
Editing

12.4.9

BN and BZ Editing

The BN and BZ edit descriptors specify how blanks other
than leading
blanks are interpreted only for numeric input fields where a width has
been specified.
These edit descriptors have no effect on numeric
output fields.
The BZ descriptor specifies that blanks will be read as
zeroes.
The
BN descriptor specifies that blanks will not be read as zeroes.
The
use of the BN or BZ edit descriptors in a format overrides the BLANK=
specifier
in the OPEN statement for the duration of the use of that
format.
(The BLANK= specifier is described in Section 11.3.3.)
For example:
101

ACCEPT (FMT=101)A,B,C,D
FORMAT (BN,r5,FI0.2,BZ,FI0.2,F8.5)

reads the first two numbers of data, ignoring blanks embedded
in the
numbers.
Then the program reads the second two numbers, substituting
zeroes for blanks embedded in the numbers.

12-28

FORMATTED DATA TRANSFERS

------·-------r,
a
I
!
[,
Editing

_ _ _ ,, __ , __ , _ _ _ _ _- - - - JI

12.4.10

0 Editing

The 0 edit descriptor sets a cJrresponding integer variable in the I/O
list to the numbe~ of characters left in the record being transferred.
This descriptor is for use wit, input transfers only.
You can
use
multiple Q descriptors in the same format list.
The Q edit descriptor
is useful when you need to kno~ the number of characters remaining
in
a record.
For example:
TYPE *,'Enter text:'
ACCEPT 1.00,L,Jl
FORMAT (A5,Q)

100

when used to read the data
Enter text:
HELLO THIS IS A TEST
would yield
~haracters

the
value
15
for
variable Jl,
in the data, and A5 reads 5 of them.

since

there

are

Numeric
Editing

12.4.11

Numeric Editing

The I, F, E, D, G, Z, and 0 edit descriptors are used to
specify the
input
and
output of
integer,
real,
complex,
double-precision,
hexadecimal, and octal data.
The numeric edit descriptors are repeatable, and can be
used
without
specifying
size.
For
output,
if you use a numeric edit descriptor
without specifying a field width, the defaults shown in Table 12-5 are
used.
For input, the data is scanned until
a
blank,
comma,
or
character
illegal for the specified edit descriptor is encountered, except for A
format, which uses the defaults shown in Table 12-5.

12-29

FORMATTED DATA TRANSFERS
I'ahle 12-5:

))dault Field Widths for NUJlwrie Edit D('scriptors

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

~=t~~.~~.~:~or _~ _!" ~;ia~l:;,

....

. .. _-_.

-_. __ . _.

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

Field Width
..

F ':-'lnglp pr(>c )

r-;::-:lio~;bk,-~;~e~·~---

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

F15 -;

---

;·F25.1H

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

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

_._-_ ..

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

\.: isingle prec. 1
E (double prec.)

E2f>.lH

D isingle prec. 1

])1 f>. 7

[) Idouble prec.)

D2f>.IH
- -.-

-.- - ...- - -------.-- - - - - - l

(; (single prec. I
(; (double prec.1

o lsi ngle prec. I

OJ ;j
.... -- .. -··---···--······-------------------l

() (double prec. 1
LUi

Z I smgle prec.)
.. -... --.- ....--.-- - - - - - - - - - 1

Z Idouble prec. 1
. ···-·-·-·---·····-·-··-------·-----l

A:"i

.-\ Isingle prec. 1
A 'double prec.)
··-I-{(-si-ngT(-:-p~·~~-·- . ------

t----

All)
[{5

L..._R_(_d-_~_lb_l(> ~};;~--- - =~-=t~~-_: HI 0

_.... - .- ... _---_._- ._._----------------'
If t}ll' dd~lUlt Eeld width for F furmat is too small li)r the data, the field width expands to fit the data.

The following conventions apply to all I/O transfers using the numeric
edit descriptors:
1.

The interpretation of blanks is determined by a combination
of any BLANK= specifier in the corresponding OPEN statement
(see Section 11.3.3), and any BN or BZ edit descriptor
(see
Section 12.4.9)
that is currently in effect in the format
list. A field of all blanks is always equal to zero.

On input transfers, with F, E, D, and G editing, a decimal
point appearing in the input field overrides the portion of
the edit descriptor that specifies the location of the
decimal point.

On output transfers, the representation of a positive or zero
value in the field may be prefixed with a plus, as controlled
by the S, SP, and SS edit descriptors (see Section 12.4.7).
The representation of a negative value in the field is
prefixed with a minus sign.

On output transfers, the representation of the transferred
datum is right-justified in the specified field.
If the
number of characters produced by the editing is smaller than
the specified field width, leading blanks are inserted in the
field.

12-30

FORMATTED DATA TRANSFERS
5.

On output transfers, if the number of characters produced
exceeds the field width, or if an exponent exceeds its length
(as specified in the Ew.dEe or Gw.dEe
forms),
the entire
field width, represented by w, is filled with asterisks.

The next sections describe the individual numeric edit descriptors.

I
Editmg

12.4.11.1
editing.

Editing - The

edit

descriptor

specifies

integer

The form of the I edit descriptor is:
[r]I[w[.m]]
where:
r

is an optional,
nonzero,
unsigned,
integer
constant
indicat.ing how many fields of I are being specified.
The
default is one field.

is an cptional, nonzero, unsigned, integer constant that
is
equal
to
the
total
number of digits in the integer field
being described.
If w is not
specified,
for
output,
the
value
is
15
(the default); for input, the data is scanned
until a blank, commil, or character illegal for
the
I
edit
descriptor is encountered .

•m

is an optional, uns~gned, integer constant (separated from w
by a period) that indicates the minimum number of digits to
be output to the integer field being described.
The default
is one digit
(Il~j.l).
If necessary, leading zeroes are
output .'
The value of m must not exceed the value of w.
If m is zero
and
the value of the internal data item is zero, the output
field consists of only blank characters, regardless of the
sign control in effect.

On input, the Iw.m and the Iw forms
treated the same.

the

edit

descriptor

are

Example:
10

FORMAT(I,I8,2I9.5)

The first data item is output as a
1- to
IS-digit
right-justified
integer
in the first 15 columns.
The second item is a 1- to 8-digit
integer occuPyin9 the next 8 columns.
The third and fourth items are
5to 9-digit integers occupying 9 columns each, with leading zeroes
appended to the data to make ~hem 5 digits if necessary.

12-31

FORMATTED DATA TRANSFERS

F
Editing

12.4.11.2 F Editing - The
(floating-point) editing.

edit

descriptor

specifies

real

The form of the F edit descriptor is:
[r]F:w.dJ
where:
r

is an optional, nonzero,
unsigned,
integer
constant
indicating the number of fields of F being specified. The
default is one field.

is an optional, nonzero, unsigned, integer constant equal to
the total number of digits in the F field being described.
This total includes the digits to the right and left of the
decimal point, the decimal point itself, and (if included)
the sign. On input, if the decimal point is omitted, the
rightmost d digits of the string, with leading zeros assumed
if necessary, are interpreted as the fractional part of the
value represented.
If w is not specified,
for output,
the
value
for
single-precision is 15
(the default), and the value for
double-precision is 25 (the default).
For input, the data
is scanned until a blank, comma, or character illegal for
the F edit descriptor is encountered •

•d

is an optional, nonzero,
unsigned,
integer constant that
specifies the total number of fractional digits in the field
of width w. The default for single-precision is 7 digits;
the default for double-precision is 18. The maximum is 63
digits.
NOTE

If the default field width for F format (with no width
specified)
is too small for the data, the field width
expands to fit the data.

E and 0
Editing

12.4.11.3 E and D Editing - The E
editing of real data.

and

The form of the E edit descriptor is:
[r]E:w.d[Ee] :
12-32

edit

descriptors

specify

FORMAT~'ED

DATA TRANSFERS

The form of the D edit descriptor is:
[r]D[w.d[Ee]]
where:

is
is an optional, non~ero, unsigned, integer constant that
equal
to
the numbE~r of E: or D fields being described.
The
one
defaults are one single-precision
E
field
and
do ubI e-·prec i s ion D field.

is an optional, non~ero, unsigned, integer constant equal to
the
total
number of digits
in the
E or D field being
described.
The total for both types of fields is equal to:
1.

ThE! total number of digits on both sides of the
point

ThE! decimal point

ThE! sign for

ThE~

The digits in the exponent

thE~

decimal

itsE~lf
numbE~r

(if included)

exponent character (i f

included)

If w is not specifi«~d, for
output,
the value
for
single
precision
is 15
:the default),
and the value for double
precision is 2!:' (th(~ default).
For
input,
the data
is
scanned until a blank, comma, or character illegal for the E
or D edit descripto~ is encountered.
d

is an optional, non~ero, unsigned integer constant equal
to
the
total
number of fractional digits in the field being
described (unless a scale factor greater
than one
is
in
effect).
The defaults for both the E and D edit descriptors
are 7 digits
if single precision,
and
18
if
double
precision.
The maximum is 63 digits.

is an optional, non~ero, unsigned, integer constant equal to
the
total
number of digits
in
the
E or D field being
described.
The default for both types of edit descriptors
is two digits.
The maximum is 15 digits.
For KL model B systems, if the program is compiled with
the
/GFLOATING switch
(see Section 16.1.3 or 16.2.3), you may
want to specify three digits
to
accommodate
the exponent
field of double-precision numbers.

G
Editing

12.4.11.4 G Editing - The G edit descriptor allows editing
of
integer,
real, double-precision, logical, complex, or character data.
With the exception of real, double-precision, and complex data,
the
type of conversion performed by the G edit descriptor depends on the
type of the corresponding variable in the I/O list.

12-33

FORMATTED DATA TRANSFERS
The form of the G edit descriptor is:
[r] G [w.d [Ee] ]

where:

is an optional, nonzero, unsigned, integer constant that is
equal to the number of G fields being described. The
default is one.

is an optional, nonzero, unsigned, integer constant equal to
the total number of digits in the G field being described.
The total for both types of fields is equal to:
1.

The total number of digits on both sides of the
point

The decimal point itself

The sign for the number (if included)

The exponent character (if included)

The digits in the exponent

decimal

If w is not specified, for output, the value for single
precision is 15
(the default), and the value for double
precision is 25 (the default).
For input,
the data is
scanned until a blank, comma, or character illegal for the G
edit descriptor is encountered.
d

is an optional, nonzero, unsigned integer constant equal to
the total number of fractional digits in the field being
described (unless a scale factor greater than one is in
effect). The defaults are 7 digits if single precision, and
18 if double precision. The maximum is 63 digits.

is an optional, nonzero, unsigned, integer constant equal to
the total number of digits in the G field being described.
The default is two digits. The maximum is 15 digits.
For KL model B systems, if the program is compiled with the
/GFLOATING switch
(see Section 16.1.3 or 16.2.3), you may
want to specify three digits to accommodate the exponent
field of double-precision numbers.

For input, in the case of real, double-precision, and complex data,
the G-format conversion is the same as for E-format conversion. For
output, however, the type of conversion performed depends on the
magnitude of the data items. Table 12-6 illustrates the conversion
performed for various ranges of real, double-precision, and complex
data.

12-34

FORMATTED DATA TRANSFERS
Table 12-6:

Effect of Dalla Magnitude on G-Format Output
Conversions

Data Magnitude (m)

Effective Cor,version

m .LT. 0.1.
0.1 .LE. m .LT. 1.0
1.0 .LE. m .LT. 10.0

Ew.d
F(w - nl.d,n(x)
F(w - nUd-l ),n(x)

10**d-2 .LE. III .LT. 10**d-J
10**d-l .LE. III .LT. 10**d
m .GE. 10**d

F(w - nl.1,n(x)
F(w - nl.O,n(x)
Ew.d

where:
x

is a blank

is 4 for Gw.d and e + 2 for Gw.dEc

where:
x

is a blank.

is 4 for Gw.d and

e~2

for Gw.dEe
NOTE

I naIl n urn e ric fie 1 d co n v e r s ion s, the fie 1 d wid t h
( w)
you specify should be large enough to include the
decimal point, sign, and, where applicable,
the
exponent character
(E),
the exponent sign, plus the
exponent digits. This is in addition to the number of
digits in the number to be represented.
If the specified width is too small to accommodate the
converted number,
the field will be filled with
asterisks (*).
If the number converted occupies fewer
character positions than specified by w, it will be
rig h t - j us t i fie din t h I~ fie 1 d , and 1 e a din g b 1 an k s will
be used to fill the field.
If the numeric data u~presentation cannot fit into the
field width F(w-n),
the n spaces (n(x»
are removed
from the right, and the numeric data representation is
again processed into :he field width Fw.
Examples of G output conversions (where the
Format

Inter:1al Value

G13.6
G13.6
G13.6
G13.6
G13.6
G13.6
G13.6
G13.6
G13.6

-1.1234S678
1.23456789
12.34567890
123.45678901
-1234.56789012
12345.67890123
123455.78901234
-1234567.89012345

I) • 0 1 2 3 4 5 6 7

12-35

¥ signifies a blank) are:
External Representation
~0.123457E-Ol

-0.123457J6)6;)~

)6)51. 23457l6)'5~J6

;.sp512.3457}6J6~~

J6t6 12 3.457j25.kS.l6'»

t6-1234.57)5J6)S~
J6}612 3 4 5 • 7,t6kS~t6

PSj25123457.j6}25pJpJ
-0.123457E+07

FORMATTED DATA TRANSFERS
F~r comparison, consider the following
example of the
same
output under the control of an equivalent F field descriptor.

values

Format

Internal Value

External Representation

F13.6
F13.6
F13.6
F13.6
F13.6
F13.6
F13.6
F13.6
F13.6

0.01234567
-0.12345678
1.23456789
12.34567890
123.45678901
-1234.56789012
12345.67890123
123456.78901234
-1234567.89012345

~J6)6});S0 • 012346
,lzSJ6)S;S-0.123457
~J61t5JzS;S1.234568
jSJt5~~12.345679
~p5PS123.456789

J'.'-1234.567890
~12345.678901

123456.789012

*************

Complex
Editing

12.4.11.5 Complex Editing - A complex number consists of a
pair of
separate
real numbers.
The first number of the pair is the real part
of the complex number; the second number is the imaginary part.
The editing of a complex number involves specifying two successive
F,
E,
D,
or G edit descriptors.
The edit descriptors need not be the
same.
A sample format list description for a complex number is:
101

FORMAT (FIO.2,EIO.2)

In this sample, the FIO.2 edit descriptor
provides editing
for
the
real
part of the complex number; the ElO.2 edit descriptor provides
editing for the imaginary part.
You may include any nonrepeatable edit descriptors between
and imaginary edit descriptors for a complex number.

[

the

real

~----.----.J
o
(Octal)
Editing

________ ._ _, _ •. _., __._._ _ _ _ _ _ _ _ _ M._

12.4.11.6 0 (Octal)
octal editing.

Editing -

The 0

The form of the 0 edit descriptor is:
··rjO[w[.m]]

12-36

(octal)

edit descriptor specifies

FORMATTED DATA TRANSFERS
\.yflere:
i oS

()escribed.
'.,

non z e r 0 ,
un s i 9 ned ,
1 n t e \.1 t r
co fj oS tan L
the
number
ot
successive octal
tielas beinq
The detault IS one octal tleld.

d n a p t ion a 1 ,

~;pecifying

1 san
opt ion aI,
f1 0 n z e r 0 ,
un s 1 9 ned , I n t e q £:' r
con!::) t d II C
::-pecifying
the
total
number
ot dlglts In ttlt' octal tield
l)einq described.

It
w
is
not
specified,
for
output,
the
value
tor
d
c.ingie-precision
octal
field
is l~ (the detault), and the
vdlue
tor
a
double-precision
octal
field
IS
2S
(the
clefault).
For
inI=ut,
the
data is scanned untll a blank,
comma, or character illeqal for the
0
edIt
descrIptor
is
.:~ncountered •
'1.1

j s an optional, unsiqned, integer
constant
speci tyinq
the
min i murn
n urn b e r o t
dig its
to be 0 u t put t o t net] e 1 a •
The
cl e fa u] t s are 1 2. tor sin 9 1 e - pre cis ion 0 c tal v 0 .iu e san d / 4 tor
double-precision octal values.

1
i

(Hexadecimal)

::::~:d;inq

12.4.11.7 Z Editing - The
Z edit descriptor
speclties
lnput
and
out put
0 f hexadE~c ima 1 val ues.
Hexadec ima l i s u ba se J f) n umoe r system
where the characters 0-9 and A-F (or a-f) represent
the
numbers
0-9
and 10-15, respectively.
(On output, A-F only.)
The form of the Z edit descrIptor

is:

ir)Z[w[.m]]
where:
r-

Ls a n 0 p t ion aI,
il n s i 9 ned ,
non z e r 0 , I n L e q (~ r
c () 11 oS Ld n t
:; p e cit yin 9
the
nun b e r
0 f
con sec UtI ve
hex ad e c 1 mal tie 1 d s
beinq specified.
The default is one hexadeclmal tleJd.
i.s
an
optional,
unsigned,
nonzero,
inleger
constant
ot digItS in tile 11exaciec Imal
.:;pecltying
the
to::al
number
field being describE~d.
w
is
not:
spe(:lfied"
for
output,
tne
\1a.lue
Lor
a
:;ingle-precision
h(~xadec:lmal fie_Ld
is .is (tile detaul C), dnd
:~ h e val u e for at do u I) 1 e - pre cis ion
hex ad e c 1 mcLl
tIe J d
1S
L S
(the
default).
FIH
input,
the
data
i,"; scanned untll a
)lank, comma, or ch,Hacter illegal for the Z ecilt (jescrlptor
is encountered.
[f

is an optional, uns lqned, integer
constant
StJ-=Cl LYlnq
tne
.ninimum
number
at
digits
to be output In the tleld.
The
iefault for sinqle-~recision hexadecimal tlclds is 9 dlgits;
the detault for double-precIsion fields IS Hi dlqlCS.

12-37

FORMATTED DATA TRANSFERS

L
Editing

12.4.12

L Editing

The L edit descriptor provides editing of logical data.
The form of the L edit descriptor is:
[r]L[w]
where:
r

is an optional,
unsigned,
nonzero,
integer
constant
specifying
the
number of consecutive logical fields being
described.
The default is one logical field.

is an optional,
unsigned,
nonzero,
integer
constant
specifying
the
total
number of characters in the logical
field being described.
If w is not specified, for output,
the value
is 15
(the
default)
For
input,
the data
is scanned until a blank,
comma, or character illegal for the
L edit descriptor is
encountered.

You may transfer logical data under format control in a manner similar
to numeric data transfer by use of the field descriptor
Lw
where:
L

is the control character and w is an integer specifying
the
field
width.
The data
is
transmitted as the value of a
corresponding logical variable in the associated I/O list.

The input field consists of optional blanks, optionally followed by a
period,
followed
by a T for true or F for false, optionally followed
by any series of characters (such as, .TRUE.
for true or .FALSE.
for
false) •
If the entire input data field is blank or empty, a value of
false is stored.
On output, w minus 1 blanks followed by T or F will be output
value of the logical variable is true or false, respectively.

12-38

the

FORMATT3D DATA TRANSFERS

A
Editing

12.4.13

A Editing

The A edit descriptor specifies the editing of character or
Hollerith
data.
The data
are stored left-justified in a word and padded with
blanks to the right.
NOTE
The R edit descriptor performs the same
function
for
Hollerith
data,
only
it
stores
the
data
right-justified in a word with leading nulls.
The
R
edit descriptor
is not supported for character data.
For a description of the R edit descriptor,
see
Section 12.4.14.
The form of the A edit descriptor is:
[r]A[w]
where:
r

is an optional, unsigned, integer constant specifying
the
number of consecutive A fields being defined.
The default
is one A field.

the
is an optional, unsigned, integer constant specifying
total
number of characters in the field being defined.
The
default for single-precision values
is
5 characters;
the
default
for double-precision
and
complex values
is 10
characters.

Depending on the
I/O operation,
the A edit descriptor
transfers
character or Hollerith data into or from a variable in an I/O list.
A
list variable may be of any type.
For example,
READ (6,5) V
5 FORMAT (A4)
causes four character or Hollerith characters to be read from
and stored in the variable v.

unit

The A descriptor deals with
variables containing
left-justified,
blank-filled characters.
The following list summarizes the result of
character or Hollerith data
transfer
(both
internal
and
external
representations)
using
the A descriptor.
These explanations assume
that w represents the field width and m represents the total number of
characters
possible
in
the variable.
Double-precision and complex
variables contain 10 characters (m=lO)i
integer,
real,
and
logical
variables contain 5 (m=5).

12-39

FORMATTED DATA TRANSFERS
A Descriptor
1.

INPUT, where w > m -- The rightmost m characters of the field
are read in and stored in the corresponding variable.

INPUT, where w < m -- All w characters are read in and stored
left-justified
and
blank-filled
in
the
corresponding
variable.

OUTPUT,
where
w > m
m characters
are
output
and
right-justified
in the field.
The remainder of the field is
blank-filled.

OUTPUT, where w < m -The
leftmost
corresponding variable are output.

characters

the

._--_._---:-------------

[

Editing

12.4.14

R Editing

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

The R edit descriptor specifies the editing of
Hollerith
data.
The
Hollerith data
are stored right-justified with leading nulls.
The R
edit descriptor is not supported for character data.
>,JOTE
The A edit descriptor, described in
~ection
12.4.13,
performs
the same function as the R descriptor except
that
it
left-iustifies
the
data
in
storaqe
with
t:cailing spaces.
The form of the R edit descriptor

is:

[r] R [w]

where:
r

is
an
optional,
unsigned,
nonzero,
integer
constant
specifying the number of consecutive R fields being defined.
The default is one R field.

is an optional, unsigned,
nonzero,
integer
constant
that
specifies the total number of characters in the R field.
If
w is not specified,
for
output,
the
value
for
a
single-precision field is 5 (the default), and the value for
a double-precision or complex field is 10.
For
input,
the
data
is
scanned until a blank, comma, or character illegal
for the R edit descriptor is encountered.

12-40

FORMATT8D DATA TRANSFERS
The R descriptor deals with variables containing
right-justified,
zero-filled
characters.
The following list summarizes the result of
Hollerith data transfer (both internal and
external
representations)
using
the
R descriptor.
These explanations assume that w represents
the field width and m represents
the
total
number
of characters
possible
in
the variable.
Double-precision and complex variables
contain 10 characters (m=lO); integer,
real,
and
logical variables
contain 5 (m=5).
NOTE
When more than five characters are stored, bit zero of
the
low-order
word
is
skipped.
Thus,
a
double-precision or complex variable
filled
by an
R-format data transfer is of the form:

R Descriptor
1.

INPUT, where w > m -- The rightmost m characters of the field
are read in and storej in the corresponding variable.

INPUT, where w < m -- All w characters are read in and stored
right-justified
and
zero-filled
in
the
corresponding
variable.

OUTPUT,
where w > m
1m
characters
are
output
and
right-justified
in the field.
The remainder of the field is
zero-filled.

OUTPUT, where w < rn -- The
rightmost
corresponding variable are output.

characters

the

List-Directed
Formatting

12.5

LIST-DIRECTED FORMATTING

The use of an asterisk in a data transfer
statement
in place of a
FORMAT statement label specifies list-directed formatting.
For this
type of
formatting,
the
type of each transferred data
item
is
specified by the types of respective elements in the I/O list.
List-directed input data transfers are performed without
regard
for
column,
card,
or
line boundaries.
List-directed output transfers
produce records with a maximum length of 72 characters
(the default)
or the length specified by the RECL specifier (see Section 11.3.27 for
devices other than the terminal).
Otherwise, the maximum length of
the current terminal width is used.
The following is a sample list-directed data transfer statement:
READ (5,*)I,IAB,M,L

12-41

FORMATTED DATA TRANSFERS
You may use list-directed transfers to read data from any acceptable
input
device,
including
a
terminal.
However,
do
not
use
device-positioning commands in conjunction with list-directed data
transfers.
If you do, the results are unpredictable.
Data for list-directed transfers should consist of alternate constants
and
delimiters.
The
constants
used
should
have
the
following
characteristics:
1.

Input constants must be of a form acceptable to FORTRAN.

Character constants must be enclosed within single quotes,
for example, 'ABLE'.
Each apostrophe in a character constant
must be represented by two apostrophes.

The end of a record is equivalent to a blank except when
it
occurs in a character constant.
In this case, the end of the
record is ignored and the character constant
is continued
with
the next reco~d.
The first character of the continued
record must be blank, which is ignored.

If the string of a character constant exceeds the
length of
the data
item,
the
string is truncated.
If the string is
shorter than the data item, the string is left-justified
and
remaining character positions are blank filled.

Blanks are
used
as delimiters
in list-directed
input.
not permitted
in any
Embedded
blanks
are,
therefore,
list-directed data item,
with
the exception of character
constants.

Decimal points may be omitted from real constants that do not
have a fractional part.
In this case, it is assumed that the
decimal point follows the rightmost digit of a real constant.

Complex constants must be enclosed within parentheses.

Octal constants must be preceded with a double quote (II).

A numeric data
item can correspond only to a
numeric
constant,
and a character data item can correspond only to a
character constant.

A delimiter in a list-directed list of data items separates one data
item
from another.
Delimiters in data for list-directed input must
comply with the following:
1.

Delimiters may be commas, blanks, or slashes.

Delimiters may be either
preceded by or
followed
by any
number of blanks, carriage return/line feed characters, tabs,
or line terminators; any such combination
is
treated as a
single delimiter.

A null
item
(the complete absence of a data
item)
is
represented
by
two
consecutive
commas
that have no
intervening constant (s).
You may place any number of bl anks,
tabs,
or carriage
return/line
feed characters between the
commas of a null item.
Each time you specify a null item
in
the
input data,
its corresponding list element is skipped
(unchanged) •

12-42

FORMATTED DATA TRANSFERS
The following illustrates the effect of the input of
item:
I/O List

A,B,IAB,N

Data input

101,'A' ,,20

null

Resulting contents of I/O list items:

A
B
IAB

101.
'A'
unchanged
20

Slashes (/) cause the current input operation to
terminate
even
if all
the items of the I/O list are not filled.
The
contents of items of the I/O list that either are skipped (by
null
items)
or have not received an input data item before
the transfer is terminated remain unchanged.
Once the
I/O
list of the data transfer statement is satisfied, the use of
the / delimiter is optional.

Once the I/O list has been satisfied
(values have been
transferred to each item of the list), any items remaining in
the input record are skipped.

Constants or null
items
in data
for
list-directed
assigned a repeat count so that an item is repeated.

input

may

A constant with a repeat count is written as:
r*K
where:
is an integer constant that specifies the number of
times
the constant is repeated, the asterisk delimits
the repeat count from the constant,
and K represents
the constant.

A null item with a repeat count
is written as an
specifies the repeat count, followed by an asterisk.

integer,

The following are examples of constants and null items:

10*5
3*'ABLE'
3*

represents 5,5,5,5,5,5,5,5,5,5
represents 'ABLE' ,'ABLE' ,'ABLE'
represents null,null,null
NOTE

The asterisk form representing nulls must be delimited
by a
comma or slash; in this case spaces are ignored
and not treated as delimiters.

12-43

which

FORMATTED DATA TRANSFERS

~O!/\!\':'ELlST·Statement
- -,rn'li:Htmq

NAMELIST-STATEMENT FORMATTING
transter statements descrIbed in Chapter 10
usually
include
i~;
a list ot varIable, array, or array element
::-' ,-1 r
data beinq transferred.

'jata
" (-~
" ... ~

~ltrrn~tive

way ot creating
1/0
lists
is
to
use
the
NAMELIST
t his
In e tho d ,
v0 u
can
s p e c i f v the I/O 1 i s t i n a
·1\.~L'LIf~,],
:~t,v-Rment
and
then
reterence
the
list
by
name
in
the
"'!)[)fi,-iiP
elilla
r.ransrer statement.
i

U c; 1 n:.l

"'Ii (' r: r- •

u~e
'·1 ~
,do.

NAf'1ELlST-statement
tormatt inq,
as
opposed
to
! ~ ~ t -- rl ire c ted for mat tin q, V0 U nee rl 0 n 1 y ret ere nee
NAi"1t:LIST name In a data transter statement.

c: '"'1. :~ r t:i'/

G r-

N.~MELIST

Slatement

J /'-. ~.: F: r . T S r r

/ n a me / 1 i s t r .1 n am e I .1 i s t J • • •

L::~:

:ld!!l!~

'::~,.::!:eci
",'

!':'':;\

! ~ 1i

j t. e ITl S com p r i sin a the N AM ELI S T I /0
tIle
1 i s t may bE' va ria b 1 e n arne s a r
:jC'Jd[rlte fTlUitlPie list items with commas.
: :;

••J

!i .,>

_,::rFl
rn (' s

'l' !\l~ L t:

()!

L t: 1 1 n

i :"; t

'·.r'!l~~!rj

! ~1

t.lw NAMELIST I/O list.
This is the name
In ddta transter statements.
Each NAMELIST
')C
cnc.1used in slashes.
OL

1 i st.
a r ray

,) t a N 1\ MEL 1ST
s tat e Ilt e n t I S
ide n t i fie d a n d
by the name immediately precedinq the list.

an example ot creatinq two NP,MELIST I/O lists
and SUM S •

having

C(),4.l ,TOTAL(lU)

'~I~,'!I,''\!<-:T!)N

. '- :',":'/\:..\L~~~/.Il,,~!C/SUMS/TOTAL
,i ..::
j

d (1 (1

'-'l~:::'

TABLE l(jentlties thE list consistinq of
the
d r r a v C. and the n am e SUM Sid en t i fie s the 1 i s t

ndIlle:'
f _ r: ':'

dr(d'i

l"Yl'.A,L.

12-44

FORMATT~D

DATA TRANSFERS

Once a list has been defined ill a NAMELIST statement, one or more
statements may ref:erence its name.

I/O

The rules for structur ing a NAr1ELIST statement are:
1.

You may use a maximum of six characters for a NAMELIST name.

You must begin the li:,t name with an alphabetic character.

You must enclose the

You should use
program.

You may define a NAMEL.IST name only once, and you mllst detine
it
by
a
NAMELIST s~atement.
Once defined, you may use the
name only in I/O tran!,ter statements.

You must define the NAMELIST name before
"tatement.s in WhICh ii: is used.

You must define
any
dimensioned
variable
contained
in
a
NAMELIST
statement
In
an
array declaration
statement
precedinq the NAf'lELIST statement.

12.7.1

NAMELI~3T

NAMELI~)T

NAMELIST-Controlled

Da~a

names

name in slashes.
that

are

unique

the

wi thin

data

the

transfer

Input Transfer

During data input
transfers
~n
which
a
NAMELIST-detined
name
is
referenced,
records are read Imtil a record is found that begins with
a blank, then $ (dollar sign), and then
the
desired
NAr'1ELIST
name.
The
dollar sign must be the s,~cond character in the rE~cord; the first
character in the record must b,~ a blank.
NOTE
You may use "&"
input.

instea(l of

II C" II
~)

NAMELIST-controlled

Da t a i tern s 0 f r e cor d s to be i n put (r e ad) us i n g NAM ELI S T - d e fin e d
must be separated by commas and may be of the following form:

lists

V=Kl,K2, ••• ,f(n

where:

may be a variable, array, or array element name.

K 1 , ••• , Kn

are constants.
A series of
ident.ical
constants
may
be
rj~presented as a single constant preceded
by a repetLtion count (5*5 represents
5,5,5,5,5).
You
can
:;pecify more than one constant only if V
is an arra'/.
If V is a scalar, then you may
have
only Kl.

The input data is always conve~ted to the type of
the
list
variable
when
there is a conflict of t'lpes.
A character constant 1S truncated
from the right, or extended on the right with blanks, if necessary, to
yield
a
constant
of
the
same
lE~ngth
as
the variable, array, or
substring.

12-45

FORMATTED DATA TRANSFERS
~he input operation continues until another $ symbol is detected.
If
variables appear
in
the
NAMELIST record that do not appear in the
NAMELIST list, an error condition will occur.
i\
character constant must have del imi ting
apostrophes.
If
an
apostrophe
is part of a character constant, it must be represented by
two consecutive apostrophes, which must
be contained
in
the
same
record
(one
apostrophe cannot end a record, and the other apostrophe
:~tart a
record).

For example, assume:
1.

A is a 2-dlmensional real array

7..

B is a I-dimensional integer array

C is an inteqer variable

0 is a character variable of length 5.

The program contains the NAMELIST declaration:
NAMELIST /FRED/ A,B,C,D

The input data is as follows:
~$FRED

A(7,2)=4, B=3,6*2.8, C=3.32, D='RON'$

A READ statement referring

the

NAMELIST-defined

name

FRED

will

result in the following:
1.

The inteqer 4 will be converted to floating point and
in A(7,2).

The integer 3 will be placed in B(l).

The inteqer 2 (after
,~ \ 2)

being

truncated)

placed

, B ( 3) , • • • , B ( 7) •

The floating point number
3.32
integer 3 and placed in C.

The character strinq 'RON~~'

12.7.2

will

placed

will

converted

the

will be placed in D.

NAMELIST-Controlled Data Output Transfers

When a
WRITE statement
reters
to
a
NAMELIST-defined
name,
all
variables
and arrays and their values belonging to the named list are
written out, each according to
its
type.
Character constants are
written with delimiting
apostrophes.
Arrays are written out by
columns.
Output data is written so that:
1.

The fields for the data will be large enough to
the significant digits.

The output can be read by an input
~AMELIST-defined list.

12-46

statement

contain

all

referencing

FORMAT,]'ED DATA TRANSFERS

For example, if ARRAY is a 2 X 3 real array, Al is a real variable, Kl
is an
integer variable, and D is a character variable containing the
five characters }\B'CD, the statements:
REAL ARRAY (L~ , 3)
CHARACTER D"~ 5
DATA ARRAy,Al,KI,D/-6.75~

0.234E-04, 680.0, -17.8,0.0,00,
AB' 'CD' /
NAMELI ST/NAMI/ARRAY, AI, K"., D
WRITE (u, NAtvll)
1

7 3 • 1,

generate the following form

output:

io!umn 1
~$NAMl

ARRAY=
Al=
$END

-6.750000,

0.2340000E-04,

680.0000,

-17.80000,

2*0.0000000,

73.10000, Kl= 3, D='AB' 'CD'
NOTE
Do not
use device-positioning
commands
such
as
BACKSPACE or
SKIPRECORD with NAMELIST-contro.lled I/O
operations.
If you do, the results are unpredictable.

12-47

CHAPTER 13
FUNCTIONS AND SUBROUTINES

Procedures you use repeatedly in a program can be written once and
then referenced each time you need the procedure.
Procedures that may
be referenced are either contained within the program in which they
are
referenced,
or self-contained executable procedures that can be
compiled separately.
The kinds of procedures that can be
referenced
are:
1.

Intrinsic functions

Statement functions

External functions

Subroutines

(FORTRAN-defined functions)

The first three of these categories are referred to collectively as
functions or function procedures; procedures of the last category are
referred to as subroutines or subroutine procedures.
Intrinsic functions perform a predefined computation with a
specific
number and type of arguments.
These functions are provided by FORTRAN
(see Section 13.1).
Statement functions are user-defined, single statement procedures that
resemble
assignment statements.
The appearance of a
statement
function
reference
in an
expression
causes
the
user-defined
computation to be performed (see Section 13.2).
External functions are separate program units that generally compute a
single value using one or more parameters.
There are two types of
external functions available:
user-defined and
FORTRAN-supplied.
A
user-defined external
function is defined with a FUNCTION statement.
Both types of external functions are invoked by including a
function
reference in an expression (see Section 13.3).
Subroutines are external program units that are used
to perform
multiple computations or alter variables.
There are two types of
subroutines
available:
user-defined
and
FORTRAN-supplied.
User-defined subroutines are defined with a SUBROUTINE statement (see
Section 13.4.2.1). Both types of subroutines are invoked with a CALL
statement (see Section 13.4.2.2).

13.1

INTRINSIC FUNCTIONS

Intrinsic functions are supplied with the
FORTRAN software.
Each
intrinsic function performs a predefined computation.
There are two
types of intrinsic functions:
specific and generic.
13-1

FUNCTIONS AND SUBROUTINES
Specific functions have an implicitly defined data type.
Each
specific function requires arguments of a particular type and returns
results of a predefined type. The IMPLICIT statement cannot be used
to change the type of a specific intrinsic function.
The data type of the return value of a generic function is determined
by the data type of its arguments. The FORTRAN generic functions are:
ASS
ACOS
AINT
ALOG
ALOGIO
AMAXI
AMINI

ANINT
ASIN
ATAN
ATAN2
CMPLX
COS
COSH
DSLE

DIM
EXP
INT
LOG
LOGlO
MAX
MIN
MOD
NINT
REAL
SIGN
SIN
SINH
SQRT
TAN
TANH
NOTE

Table 13-1 lists all the specific
and
generic
intrinsic functions. For ease of identification, each
generic function name in the table is indicated by an
asterisk.

13.1.1

Using an Intrinsic Function

An intrinsic function is used in a FORTRAN expression by referencing
the name of the function in an expression. For example, the following
program contains two intrinsic functions: ASS (returns the absolute
value of the argument)
and SQRT (returns the square root of the
argument) •
PROGRAM TEST
Y = -64
A = ASS (Y)
TYPE ,A

S = SQRT (A)
TYPE

END
When the preceding program is executed, variable Y is assigned the
value -64.
The ABS function in the second expression calculates the
absolute value of -64.
Next, in the third expression, the SQRT
function calculates the square root of the absolute value of Y, which
is A. The square root of A is assigned to B in the third expression.
Executing the program yields the following resulta:

13-2

FUNCTIONS AND SUBROUTINES
EXECUTE TEST. FOR
FORTRAN: TEST
TEST
LINK:
Load i ng
[LNKXCT TEST execution]
64.00000
8.000000
CPU time 0.1 Elapsed time 4.0
The following example contains specific and generic functions.
In the
example,
the generic function SQRT is used to find the square root of
the double-precision value 64.0. Next, the specific function DSQRT is
used to find the square root of the double-precision value 64.0.
If
the argument supplied to DSQRT was not a double-precision number, a
fatal compilation error would I"esult.
PROGRAM TESFUN
DOUBLE PRECISION A,B,AR,BR
REAL C,eR
A
B
C
C
C

64.00DO
64.00DO
64.00

GENERIC SQRT RETURNS DP
SQRT BECAUSE ARG TYPE IS DP
AR = SQRT(A)

C
C
C

SPECIFIC DSQRT PERFORMS THE
SAME FUNCTION WHEN GIVEN A DP
ARGUMEN'1'
BR

C
C

= DSC!RT (A)

SPECIFIC SQRT RETURNS A REAL
VALUE RESULT
CR = SQRT(C)
TYPE, ]\.R,BR,CR
END

Executing the program above yields the following results:
EXE TESFUN
LINK:
Load ing
[LNKXCT TESFUN Execution]
8.0000000000000000, 8.0000000000000000, 8.000000
CPU time 0.1
Elapsed time 0.7
Table 13-1 lists the FORTRAN intrinsic functions.
This table gives
function definitions,
argument and function types, and ranges of
acceptable values.. Each function contains a description of the range
for valid arguments(s) and the range within which the function returns
valid results.
If function arguments do not fall within the specified
range, the result of the function is undefined.
For more information on the precision and accuracy of the FORTRAN
intrinsic functions, refer to the TOPS-IO/TOPS-20 Common Math Library
Manual.

13-:3

FUNCTIONS AND SUBROUTINES
Table 13-1:

FORTRAN Intrinsic Functions

Name

Argument
Type

Definition

Argument Restrictions

Function
Type

Result
Range

Exponential
EXP*

y = e**x

Real

·89.415 .LE. x
.LE.88.029

Real

y .GT. 0

DEXP**

y = e**x

Double

D-floating:
-89.415 .LE. x
.LE. 88.029

Double

D·-floating:
y .GT. 0

G·-floating:
710.475 .LE. x
.LE. 709.089
CEXP

w = e**z

Complex

-89.415 .LE. REAUzl
.LE. 88.029
IAIMAG(zll .LE. 36394.429

G-floating:
y .GT. 0

Complex

All COMPLEX Numbers

Logarithm (LOG, LOG 10 =Generic Functions)
ALOG*

y = log(x) [base eJ

Real

x .GT. 0

Real

-89.415 .LE. y .LE. 88.029

DLOG**

y = log(x) [base eJ

Double

D-floating:
x .GT. 0

Double

D-floating:
89.415 .LE. y .LE. 88.029

G- floating:
x .GT. 0
CLOG

w = log(z) [base eJ

Complex

z .NE. (o,m

G-floating:
-710.4 75 .LE. y .LE. 709.08B
Complex

-89.415 .LE. REAUw) .LE. 88.029
-PI .LT. AIMAG(w) .LE. PI

ALOGlO*

y = log(x) [base 1O[

DLOG10** y = log(x) [base 1O[

Real

x .GT. 0

Real

-38.832 .LE. y .LE. 38.230

Double

D-·floating:
x .GT. 0

Double

D-floating:
-38.832 .LE. y .LE. 38.320

G-floating:
x .GT. 0

G-floating:
-308.555 .LE. y .LE. 307.953

Square Root
SQRT*

y = SQRT(x) = x**12

DSQRT**

y = SQRT(x) = x**L'2

_.-

Real

x .GE. 0

Real

y .GE. ()

Double

D-floating:
x .GE. 0

Double

D-floating:
y .GE. 0

G-floating:
x .GE. 0
CSQRT

w = SQRT(z)

z**112

Complex

Any COMPLEX Number

13-4

G-floating:
y .GE. 0
Complex

All COMPLEX Numbers
REAL(w) .GE. 0

FUNCTIONS AND SUBROUTINES

Table 13-1:

FORTRAN Intrinsic Functions (Cont.)

Name

Definition

Argument
Type

A rgument Restrictions

Function
Type

Hesult
Hange

Trigonometric
SIN*

y ,= sin(x)

.;in(x) (degrees)

:-iJND

DSIN**

y ,= sin(x)

Real

1,,1 .LE. 210828714

Real

Heal

1:1 .LE. --l7IH!)919

Heal

Double

D-floating:
I> i .LE. 6746~j18852

Double

(-floating:
1>·1 .LE. 1686()29713
CSIN

w = sin(z)

COS*

(;()SD

:".

DCOS**

y = cos(x)

= cos(x)
cos(xl (degn'cs)

w = cos(z)

TAN*

DTAN**

[)COTAN'

LE. I

Complex

IHEAUzlj .LE. 210828714
11\IMAG(z)1 .LE. 88.895

Complex

All COMPLEX Numbers

Real

I>.j .LT. 210828714

Real

1 .LT. y .LE. 1

J{eal

;: .LT. 4711-\5919

Heal

1 .1.T. \ . LE. 1

Double

ii-floating:
1;:1 .LT. 6746518852

Double

D-floating:
1.LE.y.LE.l
G-floating:
1 .LE. y .LE. 1

Complex

IImAL(zll .LE. 210828714
IAIMAG(zlj .LE. 88.895

Complex

All COMPLEX Numbers

tan(x)

Real

1:..:1 .LE. 363913

Real

All REAL Numbers

y = tan(x)

Real

II-floating:
I:{i .LE. 3373259426

Double

D-floating:
All D-·FLOATING Numbers

(f-floating:
I{i .LE. 843314856
C<JTAN

I .LE ..\

D-floating:
-1 .LE. y .LE. 1
G-floating:
. 1 .LE. y .LE. 1

G-floating:
Ixl .LT. 1686629713
CCOS

-1 .LE. y .LE. 1

U:. :Hi:!96

eoUx l

Ilpal

cotlxl

(l()uble

tloating:
x .LK :1:373259426
)

; !1(1ating:
x .LE. 8433.148[,6

13-5

G-floating:
All G-FLOATING Numbers
Ileal

All HKA.L Numbers

Ilouhlc

Df10ating
All D FLOATING Numbers
G· floating:
All G FLOATINU Numbcrs

FUNCTIONS AND SUBROUTINES
Table 13-1:
Name

FORTRAN Intrinsic Functions (Cont.)
Definition

Argument
Type

Argument Restrictions

Function
Type

Result
Range

Inverse
Trigonometric

ASIN*

y = arcsin(x)

Real

- 1 .LE. x .LE. 1

Real

- PJ/2 .LE. y .LE. PII2

DASIN**

y = arcsin(x)

Double

D-floating:
-1 .LE. x .LE. 1

Double

D-floating:
-PJ/2 .LE. y .LE. PII2

G-floating:
-1 .LE. x .LE. 1

G-floating:
-PJ/2 .LE. y .LE. PII2

ACOS*

y = arccos( x)

Real

-1 .LE. x .LE. 1

Real

DACOS**

y = arccos( x)

Double

Dfloating:
-1 .LE. x .LE. 1

Double

o .LE. y .LE. PI
D-floating:

o .LE. y .LE. PI

G-floating:
I .LE. x .LE. I

G-floating:

o .LE. y .LE. PI

ATAN*

y = arctan(x)

Real

x ~ any
REAL Numbers

Real

-Pl/2 .LE. y .LE. PII2

DATAN**

y = arctan(x)

Double

D-floating:
x = any D-FLOATING
Numbers

Double

D-floating:
-Pl/2 .LE. y .LE. PI/2

G-floating:
x = any G-FLOATING
Numbers
ATAN2*

y = arctan(argliarg2)

DATAN2** y = arctan(argllarg2)

G-floating:
-Pl/2 .LE. y .LE. PII2

Real

argl,arg2 = any
REAL Numbers

Real

-PI .LE. y .LE. PI

Double

D--floating:
arg I ,arg2 = any
f)-·FLOATING Numbers

Double

D-floating:
-PI .LE. y .LE. PI

G-floating:
argl,arg2 = any
G--FLOATING Numbers

G-floating:
-PI .LE. y .LE. PI

Hyperbolic
SINH*

y = sinh(x)

Real

Ixl .LE. 88.722

Real

All REAL Numbers

DSINH**

y = sinh(x)

Double

D--floating:
Ixl .LE. 88.722

Double

D-floating:
All D-FLOATING
Numbers

G-floating:
Ixl .LE. 709.782

G-floating:
All G-FLOA TING
Numbers

COSH*

y = cosh(x)

Real

ixl .LE. 88.722

Double

y .GE. 1

DCOSH**

y = cosh(xl

Double

D-floating:
!xi .LE. 88.722

Double

D-floating:
y .GE. 1

G-floating:
ix! .LE. 709.782

G-floating:
y .GE. 1

TANH*

y = tanh(x)

Real

Any REAL Numbers

Real

-1 .LE. y .LE. 1

DTANH**

y = tanh(x)

Double

D-floating:
Any D-FLOATING
Numbers

Double

D--floating:
-1 .LE.y.LE. 1

G-floating:
Any G-·FLOATING
Numbers

13-6

G-floating:
-1 .LE. y .LE. 1

FUNCTIONB AND SUBROUTINES

Table 13-1: FORTRAN Intrinsic Functions (Cont.)
Name

Definition

Argument
Type

Argument RE!strictions

Function
Type

Result
Range

Absolute Value
ABS*

Y = Ixl

Real

ArlY REAL Numbers

Real

Y .GE. 0

lABS

Y = Iii

Integer

ArlY INTEGER Numbers

Integer

Y .GE. 0

Y = Ixl

Double

D-floating:
AllY D--FLOATING
Numbers

Double

D--floating:
Y .GE. 0

DABS**

G-floating:
AllY G--FLOATING
Numbers
CABS

Y = Izl

Complex

AllY COMPLEX Numbers

G--floating:
Y .GE. 0
Real

Y .GE. 0

Truncation

AINT*

Sign of arg *
largest integer
.LT.largl

Real

AllY REAL Numbers

Real

All REAL Numbers

INT*

Sign of arg *
largest integer
.LT·largl

Real

AllY REAL Numbers

Integer

All INTEGER Numbers

IDINT

Sign of arg *
largest integer
.LT·largl

Double

AllY DOUBLE
PRECISION Numbers

Integer

All INTEGER Numbers

DlNT**

Sign of arg*
largest integer
.LT·largl

Double

D-floating:
AllY D--FLOATING
Numbers

Double

D·-floating:
All DFLOATING
Numbers

G-floating:
AllY G--FLOATING
Numbers

G-floating:
All G--FLOATING
Numbers

---.
Nearest Whole Number
ANINT*

Y = int(x + .5) if
x .GE. 0 else
y=int<x-.5)

Real

AllY REAL Numbers

Real

All REAL Numbers

DNINT**

Y = int(x + .5) if
x .GE. 0 else
Y = int<x -.5)

Double

D- ·floating:
AllY D--FLOATING
Numbers

Double

Dfloating:
All D--FLOATING
Numbers

G--floating:
AllY G--FLOATING
N'lmbers

G-floating:
All G--FLOA TING
Numbers

Nearest Integer
NINT*

Y = int(x + .5) if
x .OE. 0 else
Y = int(x -.5)

Real

x LE. 12"*35)·-1
x GE. -(2"*35)

Integer

All INTEGER Numbers

IDNINT

Y = int(x + .5) if
x .GE. 0 else
Y = int(x -.5)

Double

x LE. 12**35) 1
x GK (2"*:~5)

Integer

All INTEGER Numbers

13-7

FUNCTIONS AND SUBROUTINES

Table 13-1:
Name

FORTRAN Intrinsic Functions (Cont.)
Definition

Argument
Type

Argument Restrictions

Function
Type

Result
Range

Remaindering
AMOD

Remainder when
argl is divided
byarg2

Real

arg2 .NE. 0

Real

o .LE. y .LT. arg2

MOD*

Remainder when
argi is divided
byarg2

Integer

arg2 .NE. 0

Integer

o .LE. y .LT. arg2

DMOD**

Hemainder when
argi is divided
byarg2

Double

D-floating:
arg2 .NE. 0

Double

0-· floa ting:

o .LE. y .1.1'. arg2

G- floating:
arg2 .NE. 0

G-floating

o .LE. y .LT. arg2

Maximum Value (MAX = Generic Function)

AMAXO

Argument with
greatest value

Integer

Any INTEGER Numbers

Real

All HEAL Numbers

AMAXl*

Argument with
greatest value

Real

Any REAL Numbers

Real

All REAL Numbers

MAXO

Argument with
greatest value

Integer

Any INTEGER Numbers

Integer

All INTEGER Numbers

MAXI

Argument with
greatest value

Real

Any REAL Numbers

Integer

All INTEGER Numbers

DMAXI*'"

Argument with
greatest value

Double

D floating:
Any D FLOATING
Numbers

Double

D-floating:
All D-FLOA TING:
Numbers

G- floating:
Any G-FLOATING
Numbers

G-floating:
All G-FLOATING
Numbers

Minimum Value (MIN = Generic Function)

_.-

AMINO

Argument with
least value

Integer

Any INTEGER Numbers

Real

All REAL Numbers

AMINI"

Argument with
least value

Real

Any REAL Numbers

Real

All REAL Numbers

MINO

Argument with
least value

Integer

Any INTEGER Numbers

Intl'ger

All INTEGER Numbers

MINI

Argument with
least value

Real

Any REAL Numbers

Integer

All INTEGER Numbers

DMINI**

Argument with
least value

Double

D floating:
Any D-·FLOATING
Numbers

Double

D-floating:
All D-FLOATING
Numbers

G-floating:
Any G-FLOATING
Numbers

13-8

G-floating:
All G-FLOATING
Numbers

FUNCTIONS AND SUBROUTINES
Table 13-1: FORTRAN Intrinsic Functions (Cont.)
Name

Definition

Argument
Type

Argument Restrictions

Function
Type

Result
Range

Transfer of Sign
SIGN*

If arg2 .GE. 0
then larg11
else -!arg11

Real

Any REAL Numbers

Real

All REAL Numbers

ISIGN

If arg2 .GE. 0
then larg11
else -Iargl!

Integer

Any INTEGER Numbers

Real

All INTEGER Numbers

DSIGN**

If arg2 .GE. 0
then iarg11
else -Iargll

Double

D-floating:
Any D-·FLOATING
Nlmbers

Double

D-·floating:
All D-FLOATlNG
Numbers

G-floating:
Any G-FLOATING
Nlmbers

G-floating:
All G-FLOATING
Numbers

Positive Difference
DIM'"

If argl .GT. arg2
then arg1 - arg2
else 0

Real

Any REAL Numbers

Real

y .GE. 0

IDIM

If arg1 .GT. arg2
then argl - arg2
else 0

Integer

Any INTEGER Numbers

Integer

y .GE. 0

DDlM**

If arg1 .GT. arg2
then argl - arg2
else 0

Double

D-floating:
Any D-FLOATING
Nlmbers

Double

D-floating:
y .GE. 0

G- floating:
Any G-FLOATING
Nlmbers

G-floating:
y .GE. 0

Double Preeision Product
DPROD

argl *arg2

~Y REAL Numb",

Double

ALL DOUBLE PRECISION
Numbers

Conversion Routines
CONJG

arg = x ·t i"'y,
CONJG = x·- i*y

Complex

Any COMPLEX Numbers

Complex

All COMPLEX Numbers

REAL*

arg = x + i*y
returns x

Complex

Any COMPLEX Numbers

Real

All REAL Numbers

AIMAG

arg = x + i*y
returns y

Complex

Any COMPLEX Numbers

Real

All REAL Numbers

CMI'LX"

Returns
argl I i"arg~

Real

Any HEAL Numhers

Complex

All COMPI.EX Numbers

DFLOAT

lnte~er to double
preCISIOn

Integer

An\' INTEGEIi~ Numbers

Double

iy: .LT. ~"" :~5

DBLE*

Real to double
precision

Real

Any REAL Numbers

Double

All DOUBLE PRECISION
Numbers

SNGL

Double precision
to real

Double

Any DOUBLE
PRECISION Numbers

Real

All REAL Numbers

FLOAT

Integer to real

Integer

Any INTEGEF: Numbers

Real

Iyl .LT. 2**35

IFIX

Real to integer

Real

Ix! ,LT, 2**35

Integer

All INTEGER Numbers

ICHAR

Character to
Integer

Character

FiJ st character of
character value

Integer

o ,LE. y .LE. 127

CHAR

Integer to
Character

Integer

o . '~E. y .LE, 127

Character

All Single Character

13-9

FUNCTIONS AND SUBROUTINES
Table 13-1:

Name

FORTRAN Intrinsic Functions (Cont.)

Definition

Argument
Type

Argument Restrictions

Function
Type

Result
Range

Length
LEN

Length of
character entity

Character

Any CHARACTER Value

Integer

y .GE. 1

Integer

y .GE. 0

Index of a Substring
INDEX

Return location of
arg2 within argl
if not found
return 0

Character

Size of character
string argl must
be larger than
or equal to arg2

LGE

argl .GE. arg2

Character

Any CHARACTER Value

Logical

.TRUE. or .FALSE.

LGT

argl .GT. arg2

Character

Any CHARACTER Value

Logical

.TRUE. or .FALSE.

LLE

argl .LE. arg2

Character

Any CHARACTER Value

Logical

.TRUE. or .FALSE.

LLT

argl .LT. arg2

Character

Any CHARACTER Value

Logical

.TRUE. or .FALSE .

Character Comparisons

Bit Manipulation

lAND

IOH

Pf'rforms a logical
AND on corresponding
hit,.. of argl and arg2

Integer

arg I .arg2 o. any
Inte~er Numbers

Integer

All Integer Numbers

Pl'rf()rm~ an inclu~ive

Integer

arg l.arg2 = any
Integer Numbers

Integer

All Integer Numbers

01{ on corresponding

hits of argl ;md arg2
IEOR

Pertl)rm;.; an exclusive
Oft on corresponding
bits of argl and arg2

Integer

arg l.arg2 = any
Integer Numbers

Integer

All Integer Numbers

NOT

Complements each bit
of argument

Integer

Anv Integer Number

Integer

All Integer Numbers

ISHFT

L()~icall:v shiftt< argl

Integer

argJ
any
Integer Number
:16 . I.E. arg2 .LE. 36

Integer

All Integer Numbers

argl

Ir;teger

All Integer Numbers

leit arg2 bits if arg2
is positive; argl is
logically shifted right
aq~L ~,its if arg2 is
nt~~atlve

ISHFTC

Circularlv shifts (rotates) Intpger
rightmost arg:3 bits of
argl by arg2 places.
If ilrg2 is positive.
the rutation is to the
left: :f arg2 is negative.
the mtation is to the
nght

anv

Inte~er Numbpr

:16 .LE. arg2 .U: 36
I .LE. arg3 .I.E. 36

13-10

FUNCTIONS AND SUBROUTINES
Table 13-1: FORTRAN Intrinsic Functions (Cont.)

Name

Definition

Argument
Type

Arl!Ument Restrictions

Function
Type

Result
Range

Bit Manipulation
IBITS

Extracts bits arg2
through arg2 + arg:l-]
from argl ***

Integer

IBSET

Returns the value of
argl with bit arg2 of
argl set to 1

integer

IBCLR

Returns the value of
argl with bit arg2 of
argJ set to 0

Integer

ar~ 1 ~

Returns .TRUE. if bit
arg2 of argl equals 1;
returns .FALSE. if
hit arg2 of argl is 0

Integer

BTEST

any

Integer

All Integer Numbers

arl'l = any
Integer Number
o .LE. arg2 .LE. 35

Integer

All Integer Numbers

any

Integer

All Integer Numbers

Logical

.TRUE. or .FALSE.

ar~ 1

Int'~ger Number

(] .LE. arg2 .LE. :{5
ar~ 2 -+ arg3 .LE. 36

Int~ger Number

o .LE. arg2 .LE. 35
ar!!l

any

Int'~ger NumbE-r

o .LE. arg2 .LE. 35

Notes:
* = Generic function

** =- G-floating double-precision functions (KL model B only) -- are used if IGFLOA TING compiler switch is
specified (see Section 16.1.3 or 16.2.3l.

*** ,= See also the MVBITS Subroutine. Section 13.4.Ul
(2**35)-1 = 34359738367
-(2**35) = -34359738368

13-11

FUNCTIONS AND SUBROUTINES
13.1.2

Character Intrinsic Functions

Character
intrinsic functions
are
functions
that
take character
arguments or return character values.
Character comparison intrinsic
functions are functions
that
take character arguments and
return
logical values.
FORTRAN provides four character intrinsic functions:
1.

LEN
The LEN function
returns
the
length
of
a
character
expression.
The LEN function has the following form:
LEN(arg)
where:
arg

is a character
indicates how
expression.

returned
expression.
The value
many characters there are in the

The following example illustrates the LEN function:
C
C

This subroutine reverses an entire character
string.
SUBROUTINE REVERS(S)
CHARACTER T, S*(*)
J = LEN(S)
DO 10 I=1,J/2
T = S(I:I)
S(I:I) = S(J:J)
S(J:J) = T
J

= J

CONTINUE
RETURN
END

INDEX
The INDEX function searches for
a
substring
(arg2)
in a
specified character string
(arg1),
and,
if it finds the
substring, returns the substring's starting position.
If
arg2 occurs more than once in argl, the starting position of
the first (leftmost) occurrence is returned.
If arg2 does
not occur
in arg1, the value zero is returned.
The INDEX
function has the following form:
INDEX(argl,arg2)
where:
arg1

is a character expression specifying the string
to be
searched
for the substring specified by
arg2.

arg2

is a
character
expression
substring that is searched for.

13-12

specifying

the

FUNCTIONS AND SUBROUTINES
The following example illustrates the INDEX function:
C
C
C
C

This subroutine places the symbol # into the
variable MARKS at places corresponding to the
beginning of all occurrences of the substring
SUB within the string S.
SUBROUTINE FINSTR(SUB,S)
CHARACTER*(*) SUB, S
CHARACTER*132 MARKS
INTEGER I,J
1=1
MARKS

= INDEX(S(I:), SUB)
IF (J .NE. 0) THEN

1 + J

MARKS(I-l:I-l) = 1#1
IF (I .LE. LEN(S»
GO TO 10
END IF
91
3.

WRITE (5,91) S, MARKS
FORMAT (2(/lX,A»
END

ICHAR
The ICHAR function converts a character expression to
equivalent ASCII
code and
returns
the ASCII value.
ICHAR function has the following form:

its
The

ICHAR(arg)
where:
arg

is the character to be converted
to an ASCII
code.
If arg is longer than one character, only
the value of the first
character
is
returned;
the remaining characters are ignored.

CHAR
The CHAR function returns the single character whose ASCII
code
is the
integer or octal argument.
The CHAR function
has the following form:
CHl\R(arg)
where:
arg

is an integer expression.

The following example illustrates the CHAR and ICHAR functions:
CHARACTER C*l
INTEGER I
C
C

Convert number between 0 and 9 in I to a character
digit
C = CHAR(I+ICHAR('O'»
END
13-13

FUNCTIONS AND SUBROUTINES
13.1.3

Character Comparison Functions

The four character comparison functions provided with FORTRAN are:
LLT, where LLT{argl,arg2) is equivalent to (argl.LT.arg2)
LLE, where LLE{argl,arg2)

is equivalent to (argl.LE.arg2)

LGT, where LGT{argl,arg2) is equivalent to (argl.GT.arg2)
LGE, where LGE{argl,arg2)

is equivalent to (argl.GE.arg2)

The comparison functions have the following form:
func{argl,arg2)
where:
arg is a character expression.
The character comparison functions defined by the FORTRAN-77 standard
are guaranteed to make comparisons according to the ASCII collating
sequence, even on non-ASCII processors. On TOPS-IO/20 systems, the
character comparison functions are identical to the corresponding
character relationals.
An example of the use of a character comparison function follows:
CHARACTER*lO CH2
IF (LGT{CH2,'SMITH'»

STOP

The IF statement in this example is equivalent to:
IF (CH2.GT.'SMITH') STOP

:3.1.4

Bit Manipulation Functions

intrinsic bit manipulation functions are used for manipulation of
the
in
the
binary patterns that represent integers.
Integer data
Lypes are represented internally in binary two's complement notation.
!3it
positions
in
the
binary representation are numbered from right
~least siqnificant bit)
to left (most significant bit); the
rightmost
0it
position is numbered 0, and the leftmost bit position is numbered
~5.
A bit in a binary pattern has a value of 0 or 1.
;;i ts

The intrinsic functiDns lAND,
lOR,
lEaR,
and
NOT perform bitwise
')perations on all the bits of their arguments.
Bit 0 of the result is
-~e result of applying the specified logical operation to bit 0 of the
.lrguments.
Bit
1 of
the
result
is
the
result of applying the
.~pecified logical operation to bit 1 of the arguments, and so
on
for
.j ~ l
the bi ts of the resul t.
shift
functions
lSHFT and
lSHFTC shift
binary patterns.
A
:)ositive shift
count
indicates a left shift, while a negative shift
<aunt indicates a right shift.
A shift count of zero means no
shift.
_SHFT
specifies a logical shift; bits shifted out of one end are lost
",nd zeros are shifted in at the other end.
ISHFTC performs a circular
.hift;
bits
shifted
out at one end are shifted back in at the other
'''hi':'

13-14

FUNCTIONS AND SUBROUTINES
The function IBITS and the subroutine MVBITS (see
Section
13.4.1.21)
operate on bit
fields.
A bit field is a contiguous group of bits
within a binary pattern.
Bit fields are specified by a
starting
bit
position
and a length.
A hi: field must be entirely contained in its
source operand.
For example, the integer 79 i:; represented
pattern:

the

following

binary

o ••• 0101111
n ••• 6543210

(bi t

posi t ion)

where:
n

is 35

,:the number oE bit positions in an integer).

You can refer to the bit field
contained
in
bits
3
through 6
by
specifying
a
starting position of 3 and a length of 4.
In the above
example, the selected bit pat~ern would be the following:

o ••• 000101
Negative integer~; are represented in two's complement notation.
integer -79 is represented by the following binary pattern:

The

1 .•• 1010001
n ••• I) 5 4 3 21 0

( bit po sit ion)

where:
n

is 35

': the number

0,: bi t

posi tions in an integer).

NOTE
is
negative
The value of bit pos i': i on
1
a
for
35
,~
number
for
non--negative.
Also,
all the
and
0
high-order bits of th,~ pa t tE~ rn
to
the
left of
the
value up to bit 35 ar·~ the same as bit 35.
IBITS and MVBITS operate on bLt fields.
Both the starting position of
a
bit field and its lenqth a~e arguments to these intrinsics.
IBSET,
IBCLR, and BTEST operate only on one bit.
Thus, they do not require a
length argument.

13.2

STATEMENT FUNCTIONS

A statement function is a procedure specified by a single statement
that
is similar
in form 1:0 an arithmetic, character, or logical
assignment statement. The statement function enables you to define a
single-line computation once in your program, give it a name, and have
that calculation performed each time you reference the statement
function
in the program.
A statement function is classified as a
nonexecutable statement.

13-15

FUNCTIONS AND SUBROUTINES
13.2.1

Defining a Statement Function

Statement functions have the following form:
fun

[d [, d] ••• ] )

= e

where:
fun

is the symbolic name
for
the statement function.
The
function
name
follows the rules for forming symbolic names
in FORTRAN (see Chapter 4).

is an optional dummy argument.
Separate multiple dummy
arguments with commas.
(Dummy arguments are described in
Section 13.4.5.) The parentheses are still
required
if no
dummy arguments are specified.

is any type of FORTRAN expression.
The expression part of
the statement function
(to
the
right of the equal sign)
defines the computation performed using the dummy arguments.

The relationship between fun and e must conform to
the assignment
rules
in Sections 8.1,
8.2,
and 8.4.
Note that the type of the
expression may be different from the type of the statement function
name.
The
following
functions:

rules

govern

the

formation

and

use

statement

The dummy argument list in the statement function serves only
to
indicate the order, number, and type of arguments for the
statement function.

The dummy arguments used in a statement function are local to
that statement function.
It is valid to use the same names
in multiple statement functions in the same program unit.
A
dummy argument name may also be used elsewhere in the program
unit to identify a variable of the same type,
including
its
appearance as a dummy argument in a FUNCTION, SUBROUTINE, or
ENTRY statement.

Each dummy argument in a statement
function dummy argument
list must be unique; the same dummy argument cannot appear
twice in a single list.

Each variable reference in the function can be either a reference to a
dummy argument of the statement function, or a reference to a variable
that appears within the same program unit as
the statement function
statement.
If a statement function dummy argument name is the same as the name of
another entity,
the appearance of that name in the expression of a
statement function statement is a reference to the statement function
dummy argument.

13.2.2

Using a Statement Function

Statement functions are used in FORTRAN programs by referencing
the
name of the statement function in an expression that is in the same
program unit as the statement function definition.
If a
character
function
is
referenced
in a
program unit,
the
function
length
specified in the program unit must be an integer constant expression.
13-16

FUNCTIONS AND SUBROUTINES
For example, the following program uses a statement function
(called
PROFIT)
to determine
the ?rofit for
a product.
In the statement
function definition, PROFIT
is defined as
the difference between
wholesale and retail prices minus .05 sales tax.
PROGR)\M STAFUN
PROFrl' (A,B) ==
100

((A - B) -- (A * .05»

WRITE (UNIT=5,FMT==lOO)
FORMAT(lX,'Enter Mlolesale Price and Retail Price')
ACCEPT *, WHOSAL, RET.\I L

150

C = PROFIT(RETAIL,WHOSAL)

101

WRITE(UNIT=5,FMT=lQl) C
FORMA'r(lX, 'The Profit (minus sales tax)

is:

',F8.2)

END
When the program is executed, the
retail
and wholesale values are
entered at the terminal.
Next, the expression at statement number 150
uses the values of RETAIL and WHO SAL to calculate
the
profit as
defined
in the PROFIT statement function.
A sample execution of this
program yields the following results:
EXECUTE STATE
LINK:
Load i ng
[LINKXCT STAFUN executio~]
Enter Wholesale Price and Retail Price
31.67
45.95
The profit (minus sales tax) is:
11.98
CPU time 0.2
Elapsed time 18.5
When a FORTRAN expression that contains a statement function reference
is executed, the following happens:
1.

The actual arguments contained
reference are evaluated.

The actual arguments in the statement function are associated
with
the
dummy
arguments
in the statement function
definition.

The expression portion of the statement function is evaluated
using the actual arguments.

If necessary, the value of the expression is converted to the
type of the statement function.
Finally, the value resulting
from
the expression evaluation
is substituted
in
the
expression containing the statement function reference.

13.2.3

statement

function

Statement Function Restrictions

The following rules and restrictions must be
statement functions:
1.

the

adhered

when

using

The actual arguments in a function reference must agree
in
type and number with the corresponding dummy arguments in the
statement function dummy argument list.

13-17

FUNCTIONS AND SUBROUTINES
2.

An actual argument in a statement function reference can
itself be an expression; all actual arguments must be defined
when a statement function reference is evaluated.

A statement function can only be referenced in the
unit that contains the statement function statement.

A statement function must not contain a reference to another
statement function that appears later in the program unit,
but can contain a reference to another statement function
that appears earlier in the program.

The symbolic name used to identify a statement function must
not appear as a symbolic name in any specification statement
except a type statement (to specify the type of the function)
or as the name of a common block in the same program unit.

An external function reference (see Section 13.3)
in the
expression part of a statement function statement must not
cause a dummy argument of the statement function to become
undefined or redefined.

The symbolic name of a statement function may not be an
actual argument. It must not appear in an EXTERNAL statement
( see Se c t ion 7. 6) •

A statement function statement in a function subprogram (see
Section 13.3.4.) must not contain a function reference to the
name of the function subprogram or an entry name in the
function subprogram.

An actual argument in a statement function reference can be
any expression, ~ncluding a character expression involving
concatenation of an operand whose length specification is an
asterisk in parentheses.

10.

The length specification of a character
must be an integer constant expression.

13.3

statement

program

function

EXTERNAL FUNCTIONS

An external function is a procedure that is defined externally to the
program unit that references it. FORTRAN offers two types of external
functions: FORTRAN-supplied and user-defined.
The FORTRAN-supplied
external functions are described in Section 13.3.1; the user-defined
functions are described in Sections 13.3.2 through 13.3.4.

lJ.3.1

FORTRAN-Supplied External Functions

ORTRAN-supplied external functions
are
similar
to
intrinsic
unctions. To use a FORTRAN-supplied function, you reference its name
n an expression.

13-18

FUNCTIONS AND SUBROUTINES
The following are the FORTRAN-supplied external functions:
x=DTOG (y)

returns a G-floating double-precision number
in
the
range
1.47 x 10**-39
to
1.70
x
10**+38.
The
argument y
is aD-floating
double-precision
number.
(Also,
see
the
0TOGA subroutine, Section 13.4.1.12.)

x=GTOD (y)

returns a D-floating double-precision number
in
the
ranqe
1.47 x 10**-39
to
1.70
x
10**+38.
The
argument
y
is a G-floating
double-precision
number.
(Also,
see
the
GTODA subroutine, Section 13.4.1.18.)

x=LSNGET(unit}

returns the last line number read in
a
line
sequenced
file.
LSNGET returns a positive
integer if the last line
has
a
valid
line
number;
returns
zero
if the last line is a
page mark; or returns -1
If
the
last
line
number
is invalid (such as, AAAA with bit 35
se t) •
It al so
returns -1
It
the
tile
contains no line number, or was opened with a
mode other than, LINED (see Section 11.3.20).

x=RAN(O}

returns a pseudo random floating-point number
in the range of D.LT.x.LT.1.
The argument is
a durrmy (not used) and may
be
any number.
Refer
to the related subroutines SETRAN (see
Section 13.4.1.27) and
SAVRAN
(see
Section
13.4.1.26) •

x=RANS(O)

returns a pseudo random floating-point number
in
the
range of
O.LT.x.LT.l.
RANS
is a
primE-modulus random number generator
with
shuffling
capability.
It calls RAN
to
generate
its
initial
table
of
random
devi2tes.
Refer to related subroutins SETRAN
(see
Section 12.4.1.27)
and
SAVRAN
(see
S~ction 12.4.1.26).

y=SECNDS(x}

returns the
system
tIme
in
seconds as
a
single-precision, floating-point value, minus
the
value
of
its
sinqle-precision,
floating-point
argument.
The argument y is
set equal
to
the
time
in
seconds since
midnjght, minus the user-supplied value of x.

x=TIM2GO (0)

returns the number of seconds
remaining
in
the
-job's run-time limit.
The time limit is
set by the /TIME switch when
submitting
the
hatch
iob.
The
argument
is
a dummy (not
used) and may be any number.
ntay also specify a
time
limit
for
an
interactive
job by using the SET TIME-LIMIT
COffiffictnd on TOPS-20, 0 r
the SET TIME command
on TOPS-IO.
You

FORTRAN-supplied external functions are treated in the same manner
as
user-defined functions.
Impljcit or explicit type declarations affect
these
funct ions,
and
no
a I"g umen t
check ing
(type or
numbe r)
is
performed at compile time.

13-19

FUNCTIONS AND SUBROUTINES
13.3.2

User-Defined External Functions

An external user-defined function is a procedure that is external
to
the program unit that references it.
The function subprogram enables
you to define a multiline function.
By referencing the name of that
function in an expression, the lines of the function are automatically
executed.
The FUNCTION statement is always the first
statement
subprogram.
The form of the FUNCTION statement is:
[type]

FUNCTION fun

([arg1

function

[,arg2] .•• ])

where:
type

is an optional
type
specification for
the external
function.
This may be INTEGER, REAL, DOUBLE PRECISION,
COMPLEX, LOGICAL, or CHARACTER (plus the optional
size
modifier *len).
For CHARACTER, len is the length specification of the
result of the character
function.
If you specify
CHARACTER*(*), the function assumes the length declared
for
it in the program unit that invokes it.
If len is
an integer constant, the value of len must agree with
the length of the
function specified in the program
unit that invokes the function.
If a
length
is not
specified
in a CHARACTER FUNCTION statement, a length
of one is assumed.
function
If you do not specify a type, the type of the
subprogram
name determines the data
type of the
external function.

fun

is the
symbolic name of
the
externa~
function
subprogram.
Unless the optional data type is specified
in the FUNCTION statement, the
type of the
function
name
determines
the
data
type of the
function
subprogram.

arg

is an optional dummy argument.
Arg may be a variable
name,
array name,
or dummy procedure name.
Separate
multiple dummy arguments with commas.
The parentheses
are optional if no dummy arguments are specified.

You must define the symbolic name assigned a function subprogram as a
variable name in the function.
During each execution of the function,
this variable can be redefined.
The value of the variable at the time
of execution of any RETURN statement is the value of the function.
NOTE
The RETURN statement returns control to the statement
that
referenced
the function subprogram (see Section
13.4.4).
Additionally,
you may desire
to have a
function start executing at a statement other than the
first executable statement in the function subprogram.
The
ENTRY statement (see Section 13.4.3) enables you
to define an alternate entry point
in the function
subprogram.

13-20

FUNCTIONS AND SUBROUTINES

13.3.3

Function Subprogram

R~strictions

The following rules govern the structuring of a function subprogram:
1.

You can not use the ~;ymbolic name of a function subprogram in
any nonexecutable
~;tatement in the subprogram except in the
initial FUNCTION statement or a type statement.

Dummy argument names cannot appear
in any EQUIVALENCE,
COMMON, or DATA statement used within the subprogram.

The function subprogram can define or redefine one or more of
its arguments so as
to
return results in addition to the
value of the function.

The function subprogram can contain any FORTRAN statement
except BLOCK DATA, SUBROUTINE, PROGRAM, or another FUNCTION
statement.

The function subprogram should contain at least one RETURN
statement and must be terminated by an END statement.
The
RETURN statement signifies a
logical
conclusion of the
computation made by the subprogram, and returns the computed
function value and control
to
the calling program.
A
subprogram can have more than one RETURN statement.
The END statement specifies the
subprogram and implies a return.

physical

end

the

If the
type of a
function
is specified
in a
FUNCTION
statement,
the
function name must not appear
in a type
statement.
Note
that a
name must not have
its
type
explicitly specified more than once in a program unit.

A function specified in a subprogram may be referenced within
any other procedurl~ subprogram or the main program of the
executable program.
A function subprogram must not reference
itself, either directly or indirectly.

If the name of a func:tion subprogram is of type character,
each entry name
il1 the function subprogram must be of type
character.
If the name of the
function subprogram or
any
entry
in the subprogram has a length of (*) declared, all
such items must have a length of (*) declared; otherwise, all
such
items must have a
length specification of the same
integer value.

13.3.4

Using a Function Subprogram

After defining a function subprogram, you use it by referencing
the
name
of
the
function
s~bprogram
in an expression.
Function
subprograms are referenced in expressions using the following form:
fun

([argl,a.rg2, ••• argn])

where:
fun

is the function subprogram name.
This is the same name
that appears
in the corresponding FUNCTION statement.
The length of the character
function
in a
character
function
reference must be the same as the length of
the character function in the referenced function.

13-21

FUNCTIONS AND SUBROUTINES
arg

is an optional list of actual arguments.
These
arguments must agree in type and number with the dummy
argument list of the corresponding FUNCTION statement.
If the actual and dummy arguments do not agree, no type
conversion is done; and the results are unpredictable.
The
parentheses
are required even if no actual
arguments are specified.

When an expression that contains a function
executed, the following happens:

subprogram

reference

Evaluation of actual arguments that are expressions

Association of actual arguments with the corresponding
arguments

Execution of the referenced function subprogram

dummy

Section 13.3.3 describes all the restrictions that must be adhered
when using function subprograms.

13.4

SUBROUTINES

A subroutine subprogram is a procedure that is external to the program
units that reference it.
FORTRAN offers two types of subroutines:
user-defined dlllj FORTRAN-supplied. FORTRAN-supplied subroutines are
described in Sections 13.4.1 - 13.4.1.32; user-defined subroutines are
described in Sections 13.4.2 through 13.4.2.4.
NOTE
The FORTRAN-supplied subroutines are described in this
manual in two sections. Section 13.4.1 describes the
~eneral ~ORTRAN subroutines; Appendix F describes
the
~;ubroutines that enable you to use a plotter.
Program units reference subroutines with the CALL statement (see
Section 13.4.2.2).
The subroutine reference in the CALL statement
contains the unique name of the subroutine, as defined in a SUBROUTINE
statement (see Section 13.4.2.1). The SUBROUTINE statement is always
the first statement in a subroutine.
When a CALL statement reference is made to a subroutine subprogram,
program execution transfers from that CALL statement to the referenced
subroutine subprogram.
By including the ENTRY statement (see Section 13.4.3) within the body
of a subroutine subprogram, you can enter the subroutine at a point
other than the first statement in the subroutine. In this case, the
CALL statement used to reference an entry point in a subroutine
contains a reference to an entry point name, as opposed to the
subroutine name.
Return of program control from the subroutine to the calling program
unit occurs when the RETURN statement is executed (see Section
13.4.4). The RETURN statement is always the last statement executed
in a subroutine subprogram.

13-22

FUNCTIONS AND SUBROUTINES
'fable 13-2:

FORTRAN-Supplied Subroutines

'---"--- --"------------_._-----.....----------.
Form

Se(~tion

Function

~----------------------------.--~-------------------.-----------------.------~~------------CALL ALCCllR(size)

Allocates span' for dynamic character concatenation operatiolls.

13.4.1.1

rlpres - CDABS(dparg)

Heturns tIll' dO'lble··precision absolute valUl' of
the spPcifipd do Ible-precu:iion complex number.

1 :3.4.1.2

('ALL CDCOSldparg.dpresl

Finds the complex co~ine of the
double-prccisioJI compll'x number.

~pecifil'd

1 :~.4.1 :\

('ALL CDEXPldparg.dpres)

Finds the comllex l'xponenlial of the sp('citied
doubk-prl'cisio 1 complt'x number.

1 :3.4.1.4

('ALl. CDLOG(dparg,rlpres)

R<.'turns lhl' C( mplex logarithm of a
duubll'-pn'cisiol ('ompll'x numher.

~pecified

('ALL CDSIN(dparg,dprps)

Returns thl' 20mplex Sll1e of tlw
precision comp ex number spt'cificci.

doubll'-

1:1.<1.16

('ALL CDSQRT(dparg.dpres)

i{eturns the complPx ~quan' root of the double
preciSIOn complex number sp<'cificd.

1:1.4.1. 7

CALL CHKDIV(unitvarl

I{eturns thl' numb!.'r of t1w unit to which error
messages an' t'l'ing written.

1:IA.l.H

('ALL CLRFMT(arraynanH')

Discards the H)I{M AT statement .-;<lvpd by the
('xpcutioT1 of' tll' SAVFMT ~;ubroutine.

1:1.4. l.!::l

('ALL DATEII1amc)

I'lacl's thp cU!'l'('nt data. left-justified. in a characll'r "'arlahle

1:1.4.1.10

CALL 1)\ VERT(un)

Enables you 0 n'dll'l'ct error messages from
tIll' current dtvin' to an open file on a sp<'cified
devictc'.

};3.4.1.11

('ALL IrrO(;A(sname.dname,n \

Convert" del wnts of doublp-precision arrays
fi'om D float II1g double-prl'cision format to
(; .floating do .lbl(L.pr<'cisioJ1 format.

\:U.1.12

CALL IJ LJ 1\1 p( LB I . U H 1.limned I
1 ...• LBn.UBn,fi)rmatn II

Causl'1' spl'ci ied portions of memory to be
dumped to th,' line prillter (LPT:I.

(:ALI, El{HSETln)
('ALL EHH~ET(n.ll
CALL EHHSET(n.l,subri

Controls the output or arithmetIC error m('~
sage~ dlll'ing program ('xl'l'ution.

1:1.4.1.14

('ALL EHH~NS( II
('ALL ERI{SNS(] .•J)
CALL EI{]{SNS!I.,J.MSG I

J)('t<.'I·l11lfH'S t 1l' f('as()n for <In error trapped by
ERR on an OPEN. CLOSE, or datil tran.",fl'r
oPP),;,tiOll.

1:1.4.1.15

('ALL 1'~XlT

'I't'rminat.l's t w program and l'l'turns control to
the monitor.

1:1.4.1.1 ()

Heturm the IHlrntwr or thp first
FOHTRAN l'lgical unit

availabll'

n.<1 1 17

COIl\'('rb pit-ments of doublL'-precislOn arra.vs
from (;t1o;lting double·precision l(lrJnat to
I) lloatll1g d lUbll'-pn'cision format.

U.4 1 I H

CALL ILL

Sl't~ th(' lLU:(; flag.

1:U 1.1!)

CALL LE( ;AL

('lear~ the lLU:(;

J:U.1.20

CALL MVBJTS(m.Uen.n.jl

Tral1sfi'rs a bit lipid from one storag(' location
to a sl'('ond ,torClgt' location.

CA LL ()" EJ{FLIIANS I

I{('turns iniJrll1atlOll about ov('rf1ow. unckrflow. and di'ide' che'ck.

CALL (;T(H)A(sllame.dname.ni

nag.

('ALL PIHTMP(LB1.UD1JorJlIatl

Function,; l'xactly Ilk(, DUMP suhroutllw ('X-

I ... LBn.U Bn.limnatn I)

l'{'pl that COllt rol returns to the call1l1g program

1:1.1.1.21

;II'Il'r tlw dump has bn'n t'xt'l'uted.

QlJIETX

Suppn'Si-'es .111 SUlllll1ar~' typ<' out wlH'1l tlw proI-!'rarn
tl'rmi
WIt's.
___
____
_ _ ._ _ _ _ ._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

13-23

I;U.l.~4
~

_ _ _ _ _ _ _ __ _ J

FUNCTIONS AND SUBROUTINES
Tah/£' 1:J-2:

FORTHAN--Supplipd Suhroutines !Cont.)

,----------Form
I

- - 1--- - - - - - - - -

Function

('AIJ. SAVFMT1n;lnH'i,arravsiZ<'!!

Section

Directs rOBOTS tl) ('ncodc format sp('('ificain the slwcifipd charart(']' vari-

(lOllS l'ontain('rl

dhl(' (/1' :lrrav,

<'ALI. SAVHAN,tl'

j (',\I,L SOHTI'sort string',

Sav('s the last internal integer spcd valup gen,·rat(·d hv the nAN function,

I:1A.L26

Specifi('s the intl'rnal ill(('gn :-;('f'd valup filr th('
I{AN function,

1:3A,1.27

Sorts onl' or mon' tih's
progT;lnl,

CALL SHTINllnl

using the

SOHT

Din'c(s FOf{OTS to "tart allocating memory

13.4, 1.29

from top downward to account itJr large overla,v
programs and pn'allocatc's pages (jOO:677 for
SOH'/',

CALL TIME, x I
CALL TIMI'~'x,yl

I{eturns the curn'llt ! ill1e of day in left-justifipd

CALL TOPMEMI n'

Direch FOROTS !o start allflenring rnpll10ry
from top rlowllward to account for largp overla:v

ASCII
13A.1.31

prllgr,lIl1S,

,'ALL THACE

,-------------13.4.1

(;pnpr;] tes a Ii sf. of act i ve subprograms Oil the
(('rrninaL

____,____ ,__ -----------.1----------

FORTRAN-Supplied Subroutines

The FORTRAN software includes a set of predefined
subroutines.
Th.
section describes the general FORTRAN subroutines (and a function thi
is
similar)
in
alphabetical
order.
(See
Appendix
F
for
t!
tORTRAN-supplied plotter subroutines.)
NOTE
Sections
13.4.1.2
through
13.4.1.7
describe
subroutines
(and
a
function)
that
are
used
for
calculations on double-precision complex numbers.
You
must
supply
your
own
subroutines
for
performing
addition, subtraction, multiplication, and division of
double-precision complex numbers.
In
addition,
FORTRAN
does
not
support
the
double-precision
complex
data
type
(called
COMPLEX*16).
These numbers are
kept
as
arrays
and
cannot
be
used in expressions or as the arguments of
generic routines.

ALCCHR
Subroutine

]

L_____________ _
13.4.1.1

ALCCHR Subroutine - The ALCCHH
subroutine
allocates
spaCE
for
dynamic
character concatenation operations.
You do not normall~
need to allocate space for this purpose
unless
you
are
doing
very
large character concatenation operations.

13-24

FUNCTIONS AND SUBROUTINES

The form of the ALCCHR subroutine is:
CALL ALCCHR(size)
where:
size

is the integer size in characters for
or expanding the character stack.

either

CDABS
Function

creating

]

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

CDABS
Function - The
function
returns
the
13.4.1.2 CDABS
double-precision
absolute value of the specified double-precision
complex
number.
(Although CDABS
is a
function
and
is not
a
subroutine,
it
is
included here because
it
is similar
to some
subroutines. )
The form of the CDABS function is:
dpres = CDABS(dparg)
where:
dparg

is a 2-element double-precision array containing
the
complex value whose absolute value you want calculated.
The first element of dparg contains the
real
part of
the double-precision complex number; the second element
contains the imaginary part.

dpres

is a double-precision variable
that
absolute value of the complex number.

set

the

DCOS

Subroutine

---------_.

13.4.1.3 CDCOS Subroutine - Tje CDCOS subroutine
finds
cosine of the specified double-precision complex number.

the

complex

The form of the CDCOS subrouti,e is:
CALL CDCOS(dparg,dpres)
where:
dparg

is a 2-element jouble-precision array containing
the
complex value whose cosine you want calculated.
The
first element of dparg contains the real
part of
the
double-precision complex
number;
the
second element
contains the im3ginary part.

13-25

FUNCTIONS AND SUBROUTINES
dpres

is a 2-element double-precision array
in which
the
subroutine
returns the result of the calculation.
The
first element of dpres contains the real
part of the
double-precision complex
number;
the
second element
contains the imaginary part.

Example:
DOUBLE PRECISION dparg(2) ,dpres(2)
dparg(l) = IDa
!arg is
dparg(2) = -IDa
CALL CDCOS(dparg,dpres)

(1,-1)

C~EXP

Subroutine
-------~

13.4.1.4 CDEXP Subroutine - The CDEXP subroutine
finds
the
exponential of the specified double-precision complex number.

complex

The form of the CDEXP subroutine is:
CALL CDEXP(dparg,dpres)
where:
dparq

is a 2-element double-precision array that contains the
complex
argument to the subroutine.
The first element
of dparg contains the real part of the double-precision
complex
number;
the
second element contains the
imaginary part.

dpres

is a 2-element double-precision array that stores the
result of the calculation.
The first element of dpres
stores the real part of the result; the second element
stores the imaginary part.

Example:
DOUBLE PRECISION dparg(2) ,dpres(2)
dparg(l) = aDO
dparg(2) = IDa
!arg is (0,1)
CALL CDEXP(dparg,dpres)

13.4.1.5 CDLOG Subroutine - The CDLOG subroutine returns the
logarithm of a specified double-precision complex number.

13-26

complex

FUNCTIONS AND SUBROUTINES
The form of the CDLOG

subrout~ne

is:

CALL CDLOG(dparg,dpres)
where:
dparg

is a 2-element double-precision array that contains the
double-precision complex
number
whose
logarithm you
want calculate(i.
The first element of dparg
contains
the real part of the complex number; the second element
contains the inaginary part.

dpres

is a 2-element double-precision array that
stores
the
result
returnt~rl
by CDLOG.
The first element of dpres
contains the rt~al part of the double-precision
complex
number; the second element contains the imaginary part.

Example:
DOUBLE PRECISION dparg(2: ,dpres(2)
dparg(l) = IDO
!arg is
dparg(2) = ODD
CALL CDLOG(dparg,dpres)

(1,0)

------l
[
eOSIN
Subroutine

_____________ -.J

13.4.1.6

CDSIN Subroutine - ~~he CDSIN subroutine returns the
sine of the double-precision complex number specified.

complex

The form of the CDSIN subrout:.ne is:
CALL CDSIN(dparg,dpres)
where:
dparg

is a 2-element double-precision array that contains the
number
whose
sine
you
want
calculated.
The first
element of dparg
contains
the
real
part of
the
double-precision
complex
number;
the
second element
contains the inaginary part.

c]pres

is a 2-element double-precision
array
in
which
the
result of
thf!
calculation
is
returned.
The first
element of dpr~!s contains the real part of the
result;
tr..e second ele!11ent contains the imaginary part.

Example:
DOUBLE PRECISION dparq(2; ,dpres(2)
dparg(l) = -lDO
!arg is
dpar9(2) = lDO
CALL CDSIN(dparg,dpres)

13-27

(-1,01)

FUNCTIONS AND SUBROUTINES

CDSQRT
Subroutine

13.4.1.7 CDSQRT Subroutine - The CDSQRT subroutine
returns
the
complex square root of the specified double-precision complex number.
The form of the CDSQRT subroutine is:
CALL CDSQRT(dparg,dpres)
where:
dparg

is a 2-element double-precision array that contains the
riouble-precision
complex
number whose square root you
want calculated.
The first element of dparg
contains
the
real
part of the double-precision complex number;
t::.he second element contains the imaginary part.

dpres

is a 2-element double-precision array that contains the
result of the calculation.
The first element of dpres
contains the real part of the complex square root;
the
second element contains the imaginary part.

SXClmple:
:JOUBLE PRECISION dparg(2) ,dpres(2)
ciparg(l) = lODO
dpres(2) = -lODO
!arg is
CALL CDSQRT(dparg,dpres)

(10,-10)

ICHKDIV

Subroutine

13.4.1.8

CHKDIV Subroutine - The CHKDIV subroutine returns the number
the
unit
to
which
error messages
are
being
written.
This
c;ubroutine returns the value -1 if the messages are being sent to
the
'__ erminal.

The form of the CHKDIV subroutine is:
~ALL

CHKDIV(unitvar)

where:
unitvar

is the variable in which the unit number

13-28

is stored.

FUNCTIONS AND SUBROUTINES

r·-----·
i
i

CLRFMT
Subroutine

L.
13.4.1.9 CLRFMT Subroutine - The
CLRFMT subroutine discards
the
encoded
form
of
the
FORMA']' statement saved by the execution of the
SAVFMT subroutine (see Sectior 13.4.1.25).
The form of the CLRFMT subroutine is:
CALL CLRFMT(arrayname)
where:
is the nan!e of the array that contains the encoded
form
of
the
FORMAT specifications saved by the
SAVFMT sut·routine.

arrayname

r-------"-- --- -- --------]
~

I
!

DATE
Subroutine

L_____

____ ._

13.4.1.10 DATE Subroutine - 'l'he DATE subroutine
stores
the
current
date
as
a
left-justified AE;CII string in a character variable.
The
date is in the form:
dd-mmm-yyb
where:
is a 2-digit day (if the first digit is 0, it
is
converted
to
a blank), mmm i~. a 3-1etter month abbreviation (such as,
Jan,Feb), yy is a 2-letter year, and b is a blank.
The data
is stored in ASCII, left-iustified, and blank filled.

rid

The form of the DATE subroutir.e is:
CALL DATE(name)
where:
name

is the name of a character variable.

The date returned

by the subroutjne is stored in this variable.
NOTE
For
compatibility
with
previous
versions
of
FORTRAN-IO/20, you car:. specify a numeric array name as
the argument for the
DATE subroutine.
The current
date
is
stored as a left-iustified ASCII strinq in a
2-word
array or data
item
(double-precision
or
complex).
The array ITlust have at least two elements.

13-29

FUNCTIONS AND SUBROUTINES

DIVERT
Subroutine

13.4.1.11 DIVERT Subroutine - The DIVERT subroutine enables you to
redirect error messages from the current device to an open file on a
specified device.
The form of the DIVERT subroutine is:
CALL DIVERT(un)
where:
un

is the logical device number of the file on which
file resides.

the

open

----]
[
DTOGA
~broutine

13.4.1.12 DTOGA - The
DTOGA
subroutine
converts
elements
of
double-precision arrays
from D-floating double-precision format to
G-floating double-precision format.
The form of the DTOGA subroutine is:
CALL DTOGA (sname,dname,n)
where:
sname

is the name of the source array.

dname

is the name of the destination array.

is the number of elements to convert.

(See Section 13.4.1.18 for the GTODA subroutine.)

DUMP
Subroutine

13.4.1.13 DUMP Subroutine - The DUMP subroutine
portions of memory to be dumped to the line printer

13-30

causes
(LPT:).

specified

FUNCTIO~IS

AND SUBROUTINES

The form of the DUMP subroutire is:
CALL DUMP(LBl,UBl,tormatJ [ .•• ,LBn,UBn,formatn])
where:
LBl,UBl

are the integer values of the upper
addresses to bE dumped.

formatl

is an integer that
indicates
possible specifications are:

the

and

lower

dump

format.

memory
The

octal
real
inteqEr
/l.SC I I

0
1
'")

-3

If no arguments are supplied, all of user memory is dumped
In
octal.
If
only the bounds arguments are specified, or if the format value is
out of range, the dump format is octal.
If
only the
first
bound
argument
i s ' specified, all locations from that address to the end of
memory are dumped.
The dump is terminated by a c2l1 to EXIT.

r'------I

I,
~

L ___ _

13.4.1.14 ERRSET' Subroutine - The ERRSET subroutine
output of warning messages during program execution.

controls

the

The ERRSET subroutine has thrEe forms:
1.

CALL ERRSET(n)

CALL ERRSET(n,i)

CALL ERRSET(n,i,subr)

where:
n

is the maximum number of error messaqes to type.
is the code to which error the call applies; one of the
followinq:
-1
r)

1
~)

3
4
~

G
"
R

C'1nv 01 the followinG
inLeq('r overflow
jnteQ('r divide
input integer overflow
input floating overflow
float ng overflow
fluatJnq divide check
floatinG underflow
in~ut floating underflow
1 ihrarv routine error

13-31

FUNCTIONS AND SUBROUTINES
3

output field width too small
FORLIB warnings
nonstandard usage warnings
Bounds check warnings

2
1

is not specified, -1 is assumed

:3ubr

is the name of the user-defined error handling
routine
be invoked each time any of the above errors occur.
The e~fect is as if
[0

CALL SUBR

(I,IPC,N2,ITYPE,UNFIXD,FIXED)

were placed in the program just after
causing the trap.
I

the

instruction

= error number of trap, same as above

IPC
PC of
trap
instruction
(if
trapped, IPC = PC of the IOLST. call
N2 = second error number

code

(reserved for Digital)

ITYPE = data type of value
UNFIXD

= value returned by the hardware

¥IXED = value after fixup
\/Alue)

(SUBR

can

change

If SUBR is not specified, no routine is called
the APR trap.

this
on

,--------------i

ERRSNS
Subroutine

t_. . _______. ._.___._______._ .__. . _ _.__._. _. ________.
l3.4.l.l5
ERRSNS Subroutine - The ERRSNS subroutine
returns
integer
values
that describe the status (success or failure) of the last I/O
nperation (see Appendix D).
This subroutine can be used to
determine
the
reason
for
an
error trapped by ERR= on an OPEN, CLOSE, or data
transfer operation.
Both return values are
always
cleared
after
a
successful data transfer operation.
The forll1s of the ERRSNS subroutine are:
CALL ERRSNS

(I)

::::ALL ERRSNS

(I, J)

:ALL ERRSNS

(I,J,MSG)

13-32

FUNCTIONS AND SUBROUTINES
where:
I

returns a FORTF:AN-supplied number
class of failure that occurred.

describes

the

a
processor-specific
number
that
or qualifies
the type of the last

optionally returns
further
descrjbes
error.
MSG

that

the
If present, is a character variable used to return
text of the last error message.
ASCII
It the variable
is
the
text
for MSG
is
less
than 80
characters,
if
than
truncated;
the
variable
is greater
80
with
right
is
padded
the
characters, th{~
text
to
blanks.
NOTE
For compat .. bility with
previous versions of
FORTRAN-lO;'20,
you can specify a numeric array
as the MSG argument.
The numeric array is used
as
a 16-word array to return the ASCII text ot
the last error message.

13.4.1.16 EXIT Subroutine - The
EXIT subroutine
terminates
the
program and returns control to the monitor.
The EXIT subroutine takes
no arguments.
The form of the EXIT subrolltine is:
CALL EXIT

--·-···_- --- .'- ---·------------1
r
FFUNIT
I
Subroutine
L__.__________ ....__._______ J
13.4.1.17
FFUNIT Subroutine - The
FFUNIT subroutine
returns
lowest available FORTRAN logic:al unit number (see Table 10-3).
The form of the FFUNIT

subrou~ine

is:

CALL FFUNIT (n)

13-33

the

FUNCTIONS AND SUBROUTINES

1··----·.--.----------------GTODA
Subroutine

_______. _p ______

-I
_...1

i3.4.1.18 GTODA Subroutine - The GTODA subroutine
converts elements
double-precision arrays from G-floating double-precision format to
D-floating double-precision format.

The form of the GTODA subroutine is:
CALL GTODA(sname,dname,n)
v..'here:

~See

sname

is the source array name.

dname

is the destination array name.

is the number of array elements to convert.

Section 13.4.1.12 for the DTOGA subroutine.)

13.4.1.19
ILL Subroutine - The ILL subroutine sets
the
ILLEG
flag.
[f
this
flag
is
set
and
an
illegal
character is encountered in
floating-point, double-precision input, the corresponding value is set
to
zero
and no error message is issued.
The ILL subroutine takes no
~rguments.
The ILLEG flag is not set initially.
rhe form of the ILL subroutine is:
CALL ILL
r-·····-·--~---·--·---·-·---··-

I
I

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

LEGAL
Subroutine

L. . . _. . ._. _. _._. ___ . . _.
11.4.1.20
LEGAL Subroutine - The LEGAL subroutine
clears
the
ILLEG
[lag
sec by
the
ILL subroutine.
The
LEGAL subroutine takes no
1'rq umen ts .
T'iL'

form of the LEGAL subroutine is:

13-34

FUNCTIONS AND SUBROUTINES

[

-----'-------1
MVBITS
Subroutine

._.__. .'.. _.. ___, ______ J

13.4.1.21 MVBITS Subroutine -. The MVBITS subroutine transfers a
bit
field
from
one
storage
location
(source)
to
a field in a second
storage location (destination).
The form of the r.'iVBITS subrout:ine is:
CALL MVBITS(m,i,len,n,j)
where:
m

is an integer expre~;sion that represents the source location
from which a bit fiE!ld is transferred.

is an integer
exprE!ssion
that
identifies the
position in the SOUI-ce field transferred from m.

first

len is an integer expre~;sion that identifies the length
field transferred fl-om m.

bit
the

is an
integer
val-iable or
array element
that
is
the
destination location to which a bit field is transferred.

is an integer
expression
that
identifies
the
first
position in the c1estination field transferred from m.

bit

The MVBITS subroutine transfel~s
len
bits
from
position
i
through
i+len-l
of
the source locat~on (m) to positions j through j+len-l of
the destination location (n).
Other bits of the destination
location
and
all
of
the bits of the source location remain unchanged.
The
values of i+len and j+len musi: not be greater than 36.
(See Section

l3.J~.4

for

infornation on bit manipulation functions.)

[

-----.----.-.--J.
OVERFL
Subroutine

---

'-.'~

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

13.4.1.22 OVERFL
Subroutin,~ _. The
OVERFL
subroutine
information about overflow, ulderflow, and divide check.
The form of the OVERFL subrou':ine is:
CALL OVERFL(IANS)

13-35

returns

FUNCTIONS AND SUBROUTINES
wnere:
is an integer variable whose value specifies whether an
overflow, underflow, or divide check has occurred since
the last call to OVERFL.
The values returned are:

~ANS

at least one overflow,
check occurred.

underflow,

divide

none occurred.

I-----s~~~~~ne

L_. ___________._._____________. . . ___ . . __ . . _
t

13.4.1.23 PDUMP Subroutine - The PDUMP subroutine
functions
exactly
like
the
DUMP subroutine (see Section 13.4.1.13) except that control
returns to the calling program after the dump has been executed.
The form of the PDUMP subroutine is:
CALL PDUMP(LBl,UBI,formatl[ .•. ,LBn,UBn,formatn])
where:
LBI,UBI

are the integer values of the upper
addresses to be dumped.

formatl

is an integer that
indicates
possible specifications are:

the

and

lower

dump

format.

memory
The

octal
real
integer
ASCII

0
_L

2
)

_.1

no arguments are supplied, all of user memory is dumped
in octal.
If only the bounds arguments are specified, or if the format value is
aut ot range, the dump format is octal.
If
only the
first
bound
3rgument
is specified, all locations from that address to the end of
memor yare dumped.
[f

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

'-1
I

a U I E T X II'
Subroutine

L._ . _________________. ____________.___ . _J
l~.4.1.24
QUIETX Subroutine - The QUIETX subroutine
suppresses all
:>llmmary typeout when the program terminates, including library error
:,ilmmaries and CPU times.
The OUIETX subroutine takes no arguments.

"'he form of the OUIETX subroutine is:
"ALL OUIETX

13-36

FUNCTIONS AND SUBROUTINES

-- -_._- _. -------.- --_._------j
[

SAVFMT
Subroutine
---_._-----------

13.4.1.25 SAVFMT Subroutine _. The SAVFMT subroutine directs FOROTS to
encode
the FORMAT specificat~ons contained in the specified character
variable or array, and to SaVE! the encoded form.
This action improves
the
performance of
any
fuLure
1/0 statements which that character
variable or array is the format identifier.
FOROTS saves the encoded form of the format until
a
call
to
CLRFMT
(see
Section 13.4.1.9)
is
executed
for that variable or array, or
until another call to SAVFMT -os executed for that variable or array.
NOTE
After a call to SAVFMr~, you must not change the value
of
the
variable
or array_
If the value is changed,
the new value may be _.gnorecl.
A call to SAVFMT with a
variable or
array whose address is identical to that
in a previous call, does an implied call to CLRFMT.
The form of the SAVFMT subroutine is:
CALL SAVFMT(name[ ,arrays:.ze])
where:
name

is the nane of the
character
variable or
array
that contains the FORMAT descriptors that you want
encoded.

urraysize

is the nunber of array elements
specified.

array

NOTE
For
compatibility
with
previous
versions
of
FORTRAN-10/20, you can specify a numeric array name as
the argument for the ::;AVFMT subroutine.

[
13.4.1.26
internal
function,
time
it
to SETRAN

-------

-----------]

SAVRAN
Subroutine
------.--------

SAVRAN Subroutine -- The SAVRAN subroutine
saves
the
last
integer
seed
valu'~ generated by the RAN function.
The RAN
described in Sectio'1 13.3.1, returns a
random
number
each
is called.
This value can be used at a later time in a call
to reestablish the .3ame random number sequence.

13-37

FUNCTIONS AND SUBROUTINES
The form of the SAVRAN subroutine is:
CALL SAVRAN(n)
where:
n

is an integer variable
into
which
the
SAVRAN subroutine
stores the last internal integer seed value generated.

--~

SETRAN

Subroutine

13.4.1.27 SETRAN Subroutine - The
SETRAN subroutine
specifies
the
internal
integer
seed value
for
the
RAN function.
If the SETRAN
argument is zero, RAN uses its own default starting value.
The form of the SETRAN subroutine is:
CALL SP,TRAN(n)
where:
n

is a nonnegative integer constant
2**31) .

variable

(less

than

SORT
Subroutine

13.4.1.28 SORT Subroutine - The SORT subroutine
sorts one or more
files
using
the
SORT program.
You can
successfully use
this
sub r 0 uti n eon 1 y i f the SORT so f twa r e
has been
ins tall e d
( see
the
FORTRAN Installation Guide).
The SORT software is sold as a separate
product and may -i1O-t-be-avaTlable at your installation.
The form of the SORT subroutine is:
CALL SORT('sort string')
where:
sort string

is a command line containing
file
specifications
and SORT switches.
For specific information about
the SORT command line, see the SORT/MERGE
User's
Guide.

13-38

FUNCTIONS AND SUBROUTINES
NOTE
The sert string must
be compatIble with
the
current version of
the
standalone
SORT.
Therefore, the string
is
not
the
same
for
TOPS-I0
and
TOPS-20
(see the
'§~Q?_T / r-' ~~g_~ ~~~ r 's

G~_~~_~) •

13.4.1.29 SRTINI Subroutine - The SRTINI subroutine directs FOROTS to
start allocating memory from t.op downward to account for large overlay
programs and preallocates pagE·s 600: 677 (octal) for SORT.
The form of the SRTINI routinE

is:

CALL SRTINI(n)
where:
n

is top page number t.o use when allocating memory.

Note that FOROTS will not alle,cate pages (including
DDT pages)
have been marked as unavailabJe at memory initialization.

13.4.1.30 TIME Subroutine - ~'he TIME subroutine returns
time of day in left-justified ASCII.

the

that

current

The TIME subroutine has two forms:
CALL TIME(x)
or
CALL TIME(x,y)
where:
x i s a c h a r act e r va r i (\ b 1 e •
In the
0 n e a r 9 urn e n t
tor m ,
TIM E
r~eturns
the
time :_n x in the form:
hh:mm, where hh is the
h 0 u r ( 2 4 - h 0 u r tim e) and mm i s the min ute s •

is a character variable.
When the two argument form of
the
TIME subroutine is used, the forms of the time returned in x
is the same as the one arqument form, and the value returned
in
y has
the
f 0 ]- m :
b s s • t, wher e b i s a b 1 an k, s s i s the
current seconds, and t is the current tenths 01' seconds.
13-39

FUNCTIONS AND SUBROUTINES
HOTE
For
compatibility
with
previous
versions
of
FORTRAN-IO/20,
you can specify a numeric variable or
array element as an argument of the TIME subroutine.
The following example demonstrates using the one and the two
forms of the TIME subroutine in a program.

argument

PROGRAM TIMTST
CHARACTER*lO X,Y
CALL TIME(X,Y)
WRITE(UNIT=5,FMT=101)X,Y
CALL TIME(X)
102

WRITE (UNIT=5, FMT=102)X
FORMAT(lX,'THE ONE ARGUMENT TIME RETURNS:

',A)

SND
EXECUTE TIMTST
LINK:
Loading
[LNKXCT TIMTST execution]
THE TWO ARGUMENT TIME RETURNS: 09:00 20.9
THE ONE ARGUMENT TIME RETURNS: 09:00
CPU time 0.1
Elapsed time 0.2

S:~~:~i~e ]

L______________________________
13.4.1.31 TOPMEM Subroutine - The TOPMEM subrJutine directs FOROTS to
start allocating memory
from a
specified
page number downward to
account for large overlay programs.
The form of the TOPMEM subroutine is:
CALL TOPMEM(n)
where:
n

is the top page number to use in allocating memory.

Note that FOROTS will not allocate paqes (including
DDT pages)
have been marked as unavailable at memory initialization.

13-40

that

FUNCTIONS AND SUBROUTINES

I---~
I

TRACE

I . SUb~Uline

_____

13.4.1.32 TRACE Subroutine - The TRACE subroutine generates a list of
active
subprograms on the terminal.
An active subprogram is one that
has been called but has not YEt returned.
The main program is always
active.
The trace listinq st2rts at the currently active routine (the
one containing the call to
TFACE)
and
proceeds
back
to
the main
program.
The form of the TRACE subroutine is:
CALL TRACE
The information produced by tte TRACE routine consists
subprogram:

of,

for

each

The name of the routjne
The address of the rcutine (in octal)
J.

The address of the slbprogram call
offset)

(expressed as routine-name

-I-

The address of the slbprogram call
fset with the calling routine)

(expressed

label

()f

The number of argumerts passed to the called routine

The types of the argLments passed including:

'i
l"

(; ~ : :.
i)

:{ -

Character string descriptor
D-tloating couble precision
Heal
G-floating couble precision
Integer or couble integer
Literal string
Logical
Octal or dOlble octal
Statement 12bel
Unknown arglment type
Complex

If there are too many arguments to display, the
contain '

'types'

column

will

If local
symbols are
loadec
with
the
program,
the
label+offset
information will be much more informative.
A label of the form 12345P
refers to FORTRAN statement nLmber 12345; a label of the
form
56789L
refers
to
line
number
56789
in the compiler listing.
Line number
labels only appear if the program was compiled with /DEBUG:LABELS (see
Chapter 16).
The traceback listing is sent to
the
error-message
unit,
which
is
normally the
terminal.
YOl
can use the DIVERT subroutine (Section
13.4.1.11) to change where thE listing is sent.

13-41

FUNCTIONS AND SUBROUTINES
Example:

10
20

PROGRAM MAIN
TYPE 10
FORMAT (I' Calling 8UBl:')
CALL 8UBl
TYPE 20
FORMAT (I' Calling 8UB2:')
CALL 8UB2 (A,B)
END
SUBROUTINE SUBl
Y = F(X)
END
SUBROUTINE SUB2 (G,H)
CALL SUB3 (G,H,I)
END
SUBROUTINE SUB3
CALL TRACE
END

(Al,A2,A3)

FUNCTION F(Y)
CALL SUB2 (Y,O)
1:"=2.

END
EXECUTE TRC.FOR
Load i ng
LINK:
[LNKXCT TRC execution]
Calling 8UBl:
Name
TRACE.
SUB3
SUB2

(Loc)
(426373)
(256 )
(232)

( 30"1)

SUBl

(214)

«--«--«--«--«--«---

Caller
8UB3+2
8UB2+6
F+20
8UB1+2
MAIN.+7

(Loc)
(SUB3+2)
(SUB2+6)
(F+20)
(SUB1+2)
(MAIN.+7)

Args
0
3
2
1
0

Types

«--«--«--«---

Caller
8UB3+2
8UB2+6
MAIN.+14

( Loc)
(SUB3+2)
(SUB2+6)
(MAIN.+14)

Args

Types

3
2

FFI
FF

FFI
FI
F

Calling SUB2:
Name
TRACE.
SUB3
SUB2

( Loc)
(426373)
(256 )
(232)

CPU time 0.4

13.4.2

Elapsed time 3.1

User-Defined Subroutines

A subroutine subprogram is a separate program unit.
The FORTRAN CALL
statement
is used in a program unit to call a subroutine subprogram.
The CALL statement contains the name of the subroutine to which
control
passes when the CALL statement
is executed.
The CALL
statement can also optionally contain actual arguments that are passed
to
the called subroutine.
The CALL statement is described in Section
13.4.?'.2.

13-42

FUNCTIONS AND SUBROUTINES
The SUBROUTINE statement is always the first statement in a subroutine
subprogram.
The
SUBROUTINE
statement
defines the name and,
optionally,
any dummy arguments used by the
subroutine.
The
SUBROUTINE statement is described in Section 13.4.2.1.
The ENTRY statement enables yeu to enter a subroutine subprogram at a
statement other than the first statement of the subroutine. The ENTRY
statement is described in Section 13.4.3.
The last logical statement of a subroutine subprogram is always the
RETURN statement.
By default,
the RETURN statement always returns
control to the first executable statement in the calling program that
immediately follows the CALL statement.
Optionally, you may use the
alternate return form of the RETURN statement to return control
to a
statement other than the default.
Both forms of the RETURN statement
are described in Section 13.4.4.

13.4.2.1 SUBROUTINE Statement - The SUBROUTINE statement is always
the
first statement in a subroutine subprogram.
It is used to define
the name of the subroutine and, optionally, to define dummy arguments
that are used by the subroutine.
The form of the SUBROUTINE statement is:
SUBROUTINE sub [([dl[,d2) ••• ])]
where:
sub

is the symbolic name of the subroutine or dummy procedure.

is an optional dummy argument for the subroutine subprogram.
This argument can be a variable name, an array name, a dummy
procedure name, or any combination of these separated by
commas.
The parentheses are optional if no dummy arguments
are specified.

The
following
subprogram:

rules

control

the

structuring

subroutine

You can not use the symbolic name of the subprogram in any
statement within the defined subprogram except the SUBROUTINE
statement itself.
The symbolic name of a subroutine
is a
global name and must not be the same as any other global name
or any local name in the program unit.

The symbolic name of a dummy argument is local to the program
unit and must not appear in an EQUIVALENCE, PARAMETER, SAVE,
INTRINSIC, COMMON, or DATA statement except as a common block
name.

The subroutine subprogram may define or redefine one or
of its dummy arguments so as to return results.

If the actual argument is a constant or expression,
subroutine must not change the value of that argument.

the

The subroutine subprogram may contain any FORTRAN statement
except BLOCK DATA,
FUNCTION,
PROGRAM,
another SUBROUTINE
statement, or any statement that
either
directly
or
indirectly references the subroutine being defined or any of
the subprograms in the chain of subprogram references leading
to this subroutine.
13-43

FUNCTIONS AND SUBROUTINES
6.

Dummy arguments that represent statement labels may be either
an *, $, 0 r &.

The subprogram should contain at least one RETURN statement
and must be
terminated by an END statement.
The RETURN
statement indicates a logical end of the
routine;
the
END
statement signifies the physical end of the subroutine.
If
no RETURN precedes the
END statement,
then
the RETURN
statement is implicit.

Subroutine subprograms can have as many entry points as
desired
(see description of ENTRY statement given in Section
13.4.3) •

A character dummy argument whose length specification
is an
asterisk
in parentheses may appear as an operand
for
concatenation.

13.4.2.2 CALL Statement - The CALL statement is used
in a
program
unit
to
reference a
subroutine.
Execution of the CALL statement
causes a transfer of program control to the subroutine
referenced
in
the CALL statement.
The CALL statement can also contain a list of arguments that
by the computation performed in the referenced subroutine.

used

The form of the CALL statement is:
CALL sub [([al[,a2] ••• ])]
where:
sub

is the symbolic name of a subroutine or dummy procedure.

is an optional
actual argument
that
is
used by
the
subroutine.
The actual arguments in the CALL statement must
agree in position and type with the dummy arguments
in the
referenced
SUBROUTINE
statement.
The parentheses are
optional if no actual arguments are specified.

13.4.2.3 Execution of a CALL Statement - When
executed, the following results occur:

CALL

statement

Any actual arguments in the CALL statement argument list that
are expressions are evaluated.

The actual arguments are
then associated with
arguments in the referenced SUBROUTINE statement.

Control passes to the subroutine and it is executed.

Return of control
calling
program
statement.

the

dummy

from the
referenced subroutine
to the
completes
the execution of the CALL

A subroutine can be referenced within any other procedure or the main
program of the executable program.
A subprogram cannot, however,
reference itself, either directly or indirectly.

13-44

FUNCTIONS AND SUBROUTINES
13.4.2.4 Actual Arguments for
a
Subroutine - Actual
arguments can
appear
in
the
CALL statement argument list.
Actual arguments must
agree
in order,
number,
and
type with
the corresponding dummy
arguments
in the dummy argument list of the referenced subroutine.
The exception to the agreement rules between actual and dummy argument
lists is the use of a subroutine name or an alternate return specifier
as an actual argument.
Actual arguments in CALL statements can be any of the following:
1.

Any expression, including a character expression whose length
specification is an asterisk in parentheses.

An array name

An intrinsic function name

4..

An external proced ure name

A dummy procedure name

An alternate return label

An actual argument in a subroutine reference may be a dummy argument
name
that appears
in a
dummy argument list within the subprogram
containing the reference.
An asterisk dummy argument must not be used
as an actual argument in a subprogram reference.

13.4.3

ENTRY Statement

The ENTRY statement provides you with a method for entering a function
or subroutine subprogram at any executable statement.
The ENTRY
statement can appear anywhere within a function subprogram after
the
FUNCTION statement,
or
within a
subroutine subprogram after the
SUBROUTINE statement.
An ENTRY statement cannot appear between a
DO
statement and
the
terminal
statement of
its DO-loop or inside a
block-IF statement.
A subprogram may optionally have one or more ENTRY statements.
ENTRY statement is classified as a nonexecutable statement.

The form of the ENTRY statement is:
EN TR Yen [( d 1 [, d 2 ••• ] ) ]
where:
en

is the symbolic name of an entry in a function or subroutine
subprogram.
This symbolic name is called an entry name.
If
the entry name appears in a
subroutine subprogram,
it
is
referred
to as a subroutine name; if the entry name appears
in a function subprogram, it is called an external
function
name.

is a variable name, array name, dummy procedure name, or the
symbols:
*, $, or & (these symbols can represent a dummy
argument which is an alternate return
label).
The
symbol
references
(asterisk,
ampersand,
and dollar
sign)
are
permitted only when
the
ENTRY statement appears
in a
subroutine subprogram.

13-45

FUNCTIONS AND SUBROUTINES
If you do not specify any parentheses after the entry name,
you need
not specify any dummy arguments.
If,
however,
you
include the
parentheses, you must specify at least one dummy argument.
The
rules
for the use of an ENTRY statement follow:
1.

The ENTRY statement allows entry into a subprogram at a place
other
than
that defined
by the
SUBROUTINE or
FUNCTION
statement.
You may include more than one ENTRY statement
in
an external subprogram.

Execution begins at the first executable statement
the ENTRY statement.

Appearance of an ENTRY statement in
a
subprogram does not
negate
the rule that statement functions in subprograms must
precede the first executable statement.

ENTRY statements are nonexecutable
execution flow of a subprogram.

You can not use an ENTRY statement in a main program or
a subprogram reference itself through its entry points.

You can not use an ENTRY statement
statement construction.

The dummy arguments in the ENTRY statement need not agree
in
order, number, or type with the dummy arguments in SUBROUTINE
or FUNCTION statements or any other ENTRY statement
in
the
subprogram.
However, the arguments for each call or function
reference must agree with
the dummy arguments
in
the
SUBROUTINE, FUNCTION, or ENTRY statement that is referenced.

Entry into a subprogram initializes only the dummy
of the referenced ENTRY statement.

You can not reference a dummy argument unless it appears
in
the
dummy list of the
ENTRY,
SUBROUTINE,
or
FUNCTION
statement by which the subprogram is entered.

10.

The source subprogram must be ordered such that references to
dummy
arguments
in
executable
statements
follow the
appearance of the dummy argument
in
the dummy list of a
SUBROUTINE, FUNCTION, or ENTRY statement.

11.

Dummy arguments that were defined for a
subprogram by some
previous
reference
to
the
subprogram are
undefined
for
subsequent entry into the subprogram.

12.

The value of a function
current entry name.

13.

If an entry name
in
a
function
subprogram
is of type
character,
each entry name
and
the
name of the function
subprogram must be of type character.
If
the name of
the
function
subprogram or
any entry
in the subprogram has a
length of (*) declared, all such items must have a
length
specification of the same integer value.

must

13-46

and

the

returned

not

following

affect

range

use

the
have
a

arguments

the

FUNCTIONS AND SUBROUTINES
13.4.4

RETURN Statement

The RETURN statement returns control to the referencing
and may appear only
in a
function
subprogram or
subprogram.

program
unit
a
subroutine

The RETURN statement has two possible forms, depending on whether
it
is
the
last statement
in
a
function
subprogram or a subroutine
subprogram.
The form of the HETURN stateml;nt in a function subprogram is:
RETURN
The form of the HETURN stateml;nt in a subroutine subprogram is:
RETURN

[e]

where:
e

is an integer constant, variable, or expression.
This
form
of
thE~ RETURN statl;ment is called an alternate return.
The
alternate return fonn enables
you
to
select any labeled
statement
in
the calling
program
unit as a return point
after I'~xecution of the program unit in which
the
alternate
RETURN statement appenrs.
The value of e shou1d be a positive integer that is equal to
or
less
than
the
number of statement labels given in the
argument list of th(; calling statement.
If e is less than 1
or
is larger than ":he number of available statement labels,
a
standard
return
is performed.
(A
standard
return
transfers control
back
to
the first executable statement
immediately following the calling statement in
the
calling
program unit).

A dummy argument for a statE:!ment label must be either
an
asterisk
(*),
a dollar sign ($), or an ampersand
( &) •

You may use more than one RETURN (standard return)
subprogram.
The
use of
the
alternate
return
statement is restricted to subroutine subprograms.

13-47

statement
in
any
form of the RETURN

FUNCTIONS AND SUBROUTINES
For example,
program:

assume

the

following

statement

sequence

main

Each occurrence of RETURN returns control to the statement GO
in the calling program.

101

CALL EXAMP(1,*10,K,*15,M,*20)
GO TO 101

alternate return #1

alternate return #2

alternate return #3

END
SUBROUTINE EXAMP (L, *,M, *,N,*)

RETURN

RETURN(I+J)

END

If, on the execution of the RETURN(I+J)
is:

statement, the value of

Value

The following is performed:

less than one

a standard
return
to
the GO TO
statement is made
the return is made to statement 10
the return is made to statement 15
the return is made to statement 20
the GO TO
a standard
return
to
statement is made.

1
2
3
Greater than 3

13-48

(I+J)

101

FUNCTIONS AND SUBROUTINES
13.4.5

Dummy and Actual Argwnents

Since you may reference subprograms at more than one point throughout
a
program, many of the val~es used by the subprogram may be changed
each time it is used.
Dummy ~rguments in subprograms represent
the
actual values
to be used, w~ich are passed to the subprogram when it
is called.
For example, shown below is a subroutine (TEST) being called from
the
main program by a CALL statement.
In this example, the variables in
the CALL statement, A, B, and C(2), represent actual values
in
the
main program.
They are therefore called actual arguments.
On the other hand, the variables in the SUBROUTINE statement,
R,
X,
and
Z,
do not represent any values until they have values passed to
them from
the CALL stateme~t.
They are
therefore called dummy
arguments.
(The CALL, SUBROUTINE, and RETURN statements are described in Sections
13.4.2.2, 13.4.2.1, and 13.4.4, respectively.)
CALL TEST (A,B,C(2»

END
SUBROUTINE TEST (R,X,Z)

RETURN
END
Functions and subroutines use dummy arguments to indicate the type of
the actual
arguments they represent and whether the actual arguments
are variables, arrays, subroutine names,
or
the names of external
functions.
Each dummy argument must be used within a function or
subroutine as if it were a variable, array,
subroutine,
or
external
function identifier.
Dummy arguments are given in an argument list associated
with
the
identifier assigned
to the subprogram; actual arguments are normally
given in an argument list associated with a call made to
the desired
subprogram.
The position, number, and type of each dummy argument in a
subprogram
list must agree with the position, number, and type of each argument
list of the subprogram reference.
NO'TE
If the /DEBUG:ARGUMENTS compiler switch
is specified
(see
Section
16.3),
optional
type checking
is
performed at load time on dummy and actual arguments.

13-49

FUNCTIONS AND SUBROUTINES
Dummy arguments may be:
1.

variables

array names

subroutine identifiers

function identifiers

the symbols *, $, or & that are used to denote
of alternate return labels

the

position

When you reference a subprogram, its dummy arguments are replaced by
the corresponding actual arguments supplied in the reference.
All
appearances of a dummy argument within a function or subroutine are
related
to the given actual arguments.
Except for subroutine
identifiers and character constants, a valid association between dummy
and actual arguments occurs only if both are of the same type;
otherwise,
the results of the subprogram computations will
be
unpredictable.
Argument association may be carried through more than one level of
subprogram reference if a valid association is maintained through each
level. The dummy/actual argument associations established when a
subprogram is referenced are terminated when the desired subprogram
operations are completed.
The following rules govern the use and form of dummy arguments:
1.

The number and type of the dummy arguments of a procedure
must be the same as the number and type of the actual
arguments given each time the procedure is referenced.

Dummy argument names may not appear in EQUIVALENCE, DATA,
COMMON statements.

A variable dummy argument should have a variable, an array
element identifier, an expression, or a constant as its
corresponding argument.

An array dummy argument should have either an array name or
an array element identifier as its corresponding actual
argument.
If the actual argument is an array, the length of
the dummy array should be less than or equal to that of the
actual array.
Each element of a dummy array is associated
directly with the corresponding elements of the actual array.

A dummy argument representing a subroutine identifier
have a subroutine name as its actual argument.

should

A dummy argument representing an external function must
an external function as its actual argument.

have

A dummy argument may be defined or redefined in a referenced
subprogram only if its corresponding actual argument is a
variable.
If dummy arguments are array names, then elements
of the array may be redefined.

Additional information regarding the use of dummy and actual arguments
is given in the description of how subprograms are defined and
referenced.

13-50

FUNCTIONS AND SUBROUTINES
13.4.5.1 Length of Character Dummy and Actual Arguments - The
length
of a dummy argument of type character must not be greater than the
length of its associated actual argument.
Note that if the
character
dummy argument's length
is specified
as *(*), the length used is
exactly the length of the asscciated actual argument.
This
is
known
as a passed length character 6rgument.
The length of the dummy argument
is determined
each time control
transfers
to
the function.
The length of the actual argument can be
the length of a
character v~riable,
array element,
substring,
or
expression.
A character array dummy argumE'nt can have passed length.
of each element in the dummy argument is the length of the
the actual argument.
The passed
length and
the array
together determine the size of the passed length character

The
length
elements in
declaractor
array.

The following example of a function subprogram uses a
passed
length
character argument.
The function finds the position of the character
with the highest ASCII code value; it uses the length of the
passed
length character
argument
t.o control the iteration.
(Note that the
intrinsic function
LEN
is
used
to determine
the
length of
the
argument.
See Table 13-1 for a description of the LEN function.)

INTEGER FUNCTION ICMAX(CVAR)
CHARACTER*(*) CVAR
ICMAX =: 1
DO 10 I = 2, LEN (CV]\R)
IF (CVAR(I:I) .GT. CVAR(ICMAX:ICMAX))
RETURN
END

ICMAX =1

Each of the following functiorl references specifies a different length
for the dummy argument:
CHARACTER VAR*lO, CARRAY:3,5)*20

11
12
13
I 4
IS

ICMAX (VAR)
ICMAX(CARRAY(2,2))
ICMAX(VAR(3:8))
I CMAX (CARRAY (1,3) (5 15))
ICMAX (VAR(3:4)//CAHRAY(3,.5))

13.4.5.2 Character and
Hollerith
Constants as Actual
ArgumentsActual arguments and their cOJ~responding dummy arguments must agree in
data type.
If the actual
al~gument
is a
Hollerith constant
(for
example, 4HABCD),. the dummy aJ~gument must be of numeric data type.
In FORTRAN-IO/20, if an actua: argument is a character
constant
(for
example
'ABCD') r
the corresponding dummy argument can have either
numeric or character data tYPE!.
If the dummy argument has a
numeric
data type, the character cons~ant 'ABCD' is, in effect, converted to a
Hollerith constant by the FOR'rRAN compiler and the loader.
An exception to this occurs when the function or subroutine name
is
itself a
dummy argument.
It is not possible to determine at compile
time or load timE~ whether a charactEH constant or
Hollerith constant
is
required.
In this case, a character constant actual argument can
correspond only to a characte~ dummy argument.
13-51

CHAPTEFt 14
BLOCK D}\TA SUBPROGRAMS

Block data subprograms provide initial values for variables and
elements in named common blocks.

array

A block data subprogram must st:art with the BLOCK DATA statement.
The
only
valid statements within a
block data subprogram are
the
specification and DATA statemE~nts
(COMMON,
DIMENSION,
EQUIVALENCE,
IMPLICIT, PARAMETER, SAVE, type statements, and DATA statements).
The
last statement of a block data subprogram must be an END statement.
You can enter initial values illto more than one labeled
in a single subprogram of this type.

common

block

An executable program can contain more than one block data subprogram.

14.1

BLOCK DATA STATEMENT

The form of the BLOCK DATA sta":ement is:
BLOCK DATA [sub]
where:
sub

is the optional symbolic name of a block data subprogram
which the BLOCK DATA statement appears.

The name sub is a global name and must therefore be a unique
symbolic name within the executable program.
The following is an example of a block data subprogram:
BLOCK DATA TEST
COMMON /SQUARE/ CIRCLE,RECTAN,PI
DATA CIRCLE,RECTAN,PI/1.,2.,3.14159/
END
This example initializes the COMMON variables CIRCLE, RECTAN,
to 1., 2., and 3.14159 respectively.

14-1

and

CHAPTER 15
WRITING USER PROGRAMS

This chapter is a guide for writing effective programs with FORTRAN.
It contains techniques for optimization, interaction with non-FORTRAN
programs, and other useful programming hints.

15.1

GENERAL PROGRAMMING CONSIDERATIONS

The
following
paragraphs describe programming considerations you
should observe when preparing
a
FORTRAN program to be compiled by
FORTRAN.

15.1.1

Accuracy and Range of Double-Precision Numbers

Floating-point and real numbers may consist of up to 16 digits
in a
double-precision mode.
Their range is specified in Chapter 3, Section
3.2.
You must be careful when testing the value of a
number within
the
specified
range,
since,
although numbers up to 10**38 may be
represented, FORTRAN can only test numbers of up to eight significant
digits (REAL precision) and 16 significant digits (DOUBLE precision).
NOTE
For KL model B systems,
if the
/GFLOATING compiler
switch
is specified
(see Section 16.1.3 or 16.2.3),
double-precision numbers
up to
10**307
can
be
represented.
You must also be careful when testing floating-point computations for
a
result of O.
In most cases the anticipated result, that is, 0 will
be obtained; however, in some cases the result may be a
very small
number
that approximates
O.
Such an approximation of 0 will cause
tests for equality to 0 to fai1.
To increase the accuracy of its compile-time arithmetic, the compiler
does
all
floating
point arithmetic in double precision (it converts
back to single precision when necessary).
For KL model B systems,
if
the
/GFLOATING compiler switch
is specified (see Section 16.1.3 or
16.2.3), the compiler does its compile-time arithmetic in G-floating
double-precision.
If the /DFLOATING compiler switch is specified (the
default), the compiler does its compile-time arithmetic in D-floating
double-precision.
Compile-time arithmetic done in G-floating double-precision may not
overflow or
underflow as
it. might with D-floating double-precision
arithmetic.
G-floating double-precision has
a greater
range
than
D-floating double-precision (see Section 3.4).

15-1

WRITING USER PROGRAMS
15.1.2

Writing FORTRAN Programs for Use on Other Computers

If you prepare a program to run on both TOPS-I0 or TOPS-20 and another
manufacturer's computer, you should:
1.

Avoid using
any blue print language extensions
in this
manual.
(All information in blue print represents aspects of
FORTRAN that are extensions to the FORTRAN-77 Standard.)

Consider the accuracy and size of the numbers that
manufacturer's computer is capable of handling.

another

You can use the /FLAG compiler switch to
invoke the compatibility
flagger.
This feature provides warning messages for language elements
that are extensions to the FORTRAN-77 standard
or
that
are
incompatible with VAX FORTRAN (see Section 16.6).

15.1.3

Using Floating-Point DO Loops

FORTRAN permits you to use noninteger,
single- or double- precision
numbers as the parameter variables in a DO statement.
This lets you
generate a wider range of values for
the DO loop index variables,
which may, in turn, be used inside the loop for computations.
WARNING
If a noninteger
index
is used,
accumulation
of
rounding errors may lead to unexpected values for the
loop variable.

15.1.4

Computation of DO Loop Iterations

The number of times through a DO loop is computed outside the loop and
is not affected by any changes to the DO index parameters within the
loop.
The farmula far the number of times a DO loop is executed is:
DO 10 I=Ml,M2,M3
Number of cycles=MAX (INT((M2-Ml+M3)/M3) ,0)
The values of the parameters Ml, M2, M3 can be of any type except
complex.
If the
iteration count is less than or equal to zero, the
body of the loop is not executed.
The
index variable
retains
its
assigned value (Ml).
(See Section 9.3.)
NOTE
The interpretation of the
iteration count and the
index variable described above is different from that
of earlier versions of FORTRAN-IO/20.
If the /NOF77
compiler switch
is specified (see Section 16.1.3 or
16.2.3), and the iteration count is less than or equal
to
zero,
the body of the loop is executed once.
In
addition, the final value of the index variable of the
DO statement is undefined after a normal exit.

15-2

WRITING USER PROGRAMS

15.1.5

Subroutines - Programming Considerations

Consider the following items when preparing and executing subroutines:

During execution, no check is made
to see
if the proper
number of arguments
is passed (unless the /DEBUG:ARGUMENTS
compiler switch is specified, see Section 16.3).

If the number of actual arguments passed to a
subroutine
is
less than the number of dummy arguments specified, the values
of the unspecified
3rguments are undefined
(unless the
/DEBUG:ARGUMENTS compiler switch
is specified, see Section
16.3) •

If the number of actual arguments passed to a
subroutine
is
greater than the number of dummy arguments given, the excess
arguments are ignored (unless the /DEBUG:ARGUMENTS compiler
switch is specified, see Section 16.3).

If an actual argument is a constant and
its corresponding
dummy argument
is set to another value (an illegal usage),
all references made to the constant in the calling program
may be changed to the new value of the dummy argument.

No check is made to see if the arguments passed are of the
same type as the dummy arguments (unless the /DEBUG:ARGUMENTS
compiler switch is specified, see Section 16.3).
NOTE
An exception is that a check is always made
for G-floating and D-floating type mismatches
regardless of the /DEBUG:ARGUMENTS switch
being specified.
In addition, w~en you pass character data
to
a
subroutine or
function
that is compiled
wi th the /EXTE.'1D switch, the calling program
must
also be? compiled with the /EXTEND
switch.
If an actual parameter is a constant and
the corresponding
dummy is of type real, be sure to include the decimal point
in the constant.
If t~e dummy is double-precision,
be sure
to specify the constant with a "D".
NOTE
You are given no warning
if any of the situations
described
in
items 1,2,3,4,
or 5 occur (unless the
/DEBUG:ARGUMENTS compiler switch
is specified,
see
Section 16,.3).

Examples:
If a function F has a single dummy arqument of
function is called with:

F(2)

15-3

type

real,

and

that

WRITING USER PROGRAMS
F interprets the integer 2 as an unnormalized
In this instance, F(A) should be called with:

floating-point

number.

F(2.0)
Similarly, if the function Fl(D)

is called with:

Fl (2.5)
and D is double-precision, Fl assumes that its parameters have
been
specified with
two words of precision and picks up whatever follows
the constant 2.5 in memory.
The proper method is to use:
Fl(2.5DO)

15.1.6

Reordering of Computations

computations that are not enclosed within parentheses may be reordered
by the
compiler.
Sometimes
it
is necessary to use parentheses to
ensure proper results from a specific computation.
For example, assuming that:
1.

RLI represents a large number, such that RLl*RL2
an overflow condition, and

will

cause

RSI is a very small number, that is, less than 1, the program
sequence:

A=RSl*RLl*RL2
B=RS2*RL2*RLI

will not produce an overflow when evaluated
left
to
right,
since
the
first
computation
in each expression (that is,
RSl*RLl and RS2*RL2) will produce an interim result
that
is
smaller than either large number (RLI or RL2).
However, the compiler may recognize RLl*RL2 as a common sUbexpression
(see Section 15.2.1.1) and generate the following sequence:
temp
A
B

RLl*RL2
RSl*temp
RS2*temp

The computation of temp will cause an overflow.
You should write the program as follows to
results are obtained:

ensure

that

the

desired

A=(RSl*RLl)*RL2
B=(RS2*RL2)*RLI
computations may be reordered even when
selected.
15-4

global

optimization

not

WRITING USER PROGRAMS
15.1.7

Dimensioning of Dummy Arrays

When you specify an array as a dummy argument to a
subprogram unit,
you must
indicate to the compiler that the parameter is an array by
dimensioning the array in a specification statement.
This is the only
way the compiler is able to distinguish a reference to such an array
from a function reference.
A dummy array can be dimensioned
the
following ways:
1.

Assumed size

Adjustable dimensioned

Fixed dimension bound

Dimensioning the array with
a size of 1
dangerous,
practice.
The alternative
to
assumed-size arrays (see Section 7.1.2).

is a
common,
although
this practice is to use

Example:
SUBROUTINE SUBl(A,B)
DIMENSION A (l)
There are disadvantages to using the above technique because
it may
prevent the compiler from diagnosing illegal programs, specifically:
1.

Reading or writing the array by name
DIMENSION ARRAY
READ (l) ARRAY

(10)

The above is a binary read that
ARRAY.

will

read

ten

words

into

SUBROUTINE SUBl(A)
DIMENSION A(l)
READ (l) A
This binary read will cause one word to be read into A.
2.

Using the array as a format
SUBROUTINE SUB2(FMT)
DIMENSION FMT(l)
READ (l, FMT)
Only the first word of the format specification contained
FMT is used.

Using
the /DEBUG:BOUNDS compilation switch
(see Section
16.3),
the dimension
information used
is that which is
specified in the array declaration
SUBROUTINE SUB3(A)
DIMENSION A(l)
A(2)=0
The reference to A(2) will cause
message to be generated.

15-5

the

out-of-bounds

warning

WRITING USER PROGRAMS
15.2

FORTRAN GLOBAL OPTIMIZATION

You have
the option of
invoking
the global
optimizer
during
compilation.
The optimizer treats groups of statements in the source
program as a single entity.
The purpose of the global optimizer is to
prepare a more efficient object program that produces the same results
as the original unoptimized program,
but takes significantly less
execution time.
The output of the lexical and syntactic analysis phase of the compiler
is developed into an optimized source program equivalent (in results)
to the original.
The optimized program
is then processed by the
standard compiler code generation phase.

15.2.1

Optimization Techniques

15.2.1.1 Elimination of Redundant Computations - Often the
same
sUbexpression will
appear
in more than one computation throughout a
program.
If the values of the operands of such a common expression
are not changed between computations, the sUbexpression may be written
as a separate arithmetic expression.
Also, the variable
representing
its resultant may then be substituted where the subexpression appears.
This eliminates unnecessary recomputation of the subexpression.
For
example, the instruction sequence:
A=B*C+E*F

H=A+G-B*C

IF((B*C)-H)

10,20,30

contains the sUbexpression B*C three times when it really needs to
computed only once.
Rewriting the preceding sequence as:

T=B*C
A=T+E*F
H=A+G-T
IF(T-H) 10,20,30
eliminates two computations of the sUbexpression B*C from the
sequence.

overall

Decreasing the number of arithmetic operations performed in a
source
program by the elimination of common sUbexpressions shortens the
execution time of the resulting object program.

15-6

WRITING USER PROGRAMS
15.2.1.2 Reduction of Operator Strength - The time
required
to
execute arithmetic operations varies according
to the operator(s)
involved.
The hierarchy of arithmetic operations according
to the
amount of execution time required is:
MOST TIME

OPERATOR
**

/
LEAST TIME

+,-

During program optimization,
the global optimizer
replaces,
where
possible
some arithmetic operations that require the most time
with operations that require less time.
For example,
consider the
following DO loop that is used to create a table for the conversion of
from 1 to 20 miles to their equivalents in feet:
10

DO 10 MILES=1,20
IFEET(MILES)=5280*MILES

The execution time of the loop would be shorter if the time-consuming
multiply operation, that is, 5280*MILES, could be replaced by a faster
operation.
Since you increment MILES on each pass,
you can replace
the multiply operation by an add and total operation.
In its optimized form, the
equivalent to:

loop

would

replaced

sequence

K=5280
DO 10 MILES=1,20
IFEET(MILES)=K
K=K+5280

In the optimized form of the loop, the value of K is set to 5280
for
the
first
iteration of the loop, and is increased by 5280 for each
succeeding iteration of the loop.
This situation occurs frequently
implicitly contain multiplications.

subscript

calculations

that

15.2.1.3 Removal of Constant Computation from Loops - The speed with
which
a given algorithm may be executed can be
increased
if
instructions and/or computations are moved out of frequently traversed
program sequences into less frequently traversed program sequences.
Movement of code is possible only if none of the arguments
in the
items to be moved are redefined within the code sequences from which
they are to be taken.
Computations within a
loop consisting of
variables or constants that are not changed in value within the loop
may be moved outside the loop.
Decreasing the number of computations
made within a
loop greatly decreases the execution time required by
the loop.
For example, in the sequence:
10

DO 10 1=1,100
F=2.0*Q*A(I)+F

INumerical analysis considerations severely limit the
cases where this is possible.
15-7

number

WRITING USER PROGRAMS
the value of the computation
2.0*Q,
once calculated on
the
first
iteration, will remain unchanged during the remaining 99 iterations of
the loop.
Reforming the preceding sequence to:

QQ=2.0*Q
DO 10 1=1,100
F=QQ*A(I)+F

moves the calculation 2.0*Q outside
the
scope of
movement of code eliminates 99 multiply operations.

the

loop.

This

In addition, it is possible to
remove entire assignment
statements
from
loops.
This action can be easily detected
from the macro
expanded listings.
The internal
sequence
number
remains with
the
statement and appears out of order in the leftmost column of the macro
expanded listing (LINE).

15.2.1.4 Constant Folding
and
Propagation - In
this
method
of
optimization,
expressions containing determinate constant values are
detected and the constants are replaced, at compile
time,
by their
defined or calculated value.
For example, assume that the constant PI
is defined and used in the following manner:

PI=3.14159

X=2*PI*Y

At compile time, the optimizer will have used the defined value of PI
to
calculate the value of the
sUbexpression 2*PI.
The optimized
sequence would then be:

PI=3.14159

X=6.28318*Y

thereby eliminating a multiply operation from the object code program.
The evaluation of constant expressions at compile
time
is called
"folding";
the replacement of variables with their constant values is
called "constant propagation".
NOTE
For KL model B systems, use of the /GFLOATING compiler
switch
(see Section 16.1.3 or
16.2.3)
may affect
compile-time arithmetic.

15-8

WRITING USER PROGRAMS

15.2.1.5 Removal of Inaccessible Code - The optimizer detects and
eliminates any code within the source program that cannot be accessed.
In general, this will not happ8n since programmers do not normally
include such code
in their ]Jrograms; however, inaccessible code may
appear in a program during the debugging process.
The
removal of
inaccessible code by the o]Jtimizer
reduces the size of the object
program.
A warning message is generated for each inaccessible line removed.

15.2.1.6 Global Register Allocation - During the compilation of a
source program, the optimizer controls the allocation of registers to
minimize computation time
in the optimized object program.
The
allocation process
is designed
to minimize the number of MOVE and
MOVEM machine instructions that will appear
in the most
frequently
executed portions of the code.

15.2.1.7 I/O Optimization - Every effort is made to minimize the
number of required calls to the FOROTS system.
This is done primarily
through extensive analysis of implied DO loop constructs on I/O data
transfer
statements.
The
formats of these special blocks are
described in Chapter 18.
These optimizations reduce the size of the
program
(argument code plus argument block size
is reduced) and
greatly improve the performanc8 of programs that use implied DO loop
I/O statements.

15.2.1.8 Uninitialized Variable Detection - A warning message may be
generated when a scalar variable is referenced before it has received
a value (only when optimizing).

15.2.1.9 Test Replacement - If the only use of a DO loop index is to
reduce operator strength (see Section 15.2.1.2) and the loop does not
contain exits (GO TOs out of the loop),
the DO loop index
is not
needed and can be replaced by the reduced variable.
For example:

DO 10 1=1,10
K=K+7*I
CONTINUE

Reduction of operator strength and test replacement together transform
this loop into:

DO 10 1=7,70,7
K=K+I
CONTINUE

This situation occurs frequently
in
execution of these statements, 1=11.

15-9

subscript

computation.

After

WRITING USER PROGRAMS
lS.2.2

Programming Techniques for Effective Optimization

Observe the following recommendations during the coding of a FORTRAN
source program.
They will improve the effectiveness of the optimizer:
1.

Do not use DO loops with an extended range.

Specify label lists when using assigned GO TOs.

Nest loops so that the innermost index is the
largest range of values.

Avoid the use of associated I/O variables.

Avoid unnecessary use of COMMON and EQUIVALENCE.

lS.3

one

with

the

calls

FUNCTION SIDE EFFECTS

Unpredictable results can occur
if a statement
includes
different functions that modify the same variables.
Consider the following example:
COMMON A
A=S.
P=F(1.)+Q(2.)
END
FUNCTION F(X)
COMMON A
A=O.
F=X+l
END
FUNCTION Q(Y)
COMMON A
Q=A
END

In the preceding sequence, if P is evaluated by calling F before Q,
the value of P will be 2.
If P is evaluated by calling Q before F,
the value of P will be 7.

lS.4

INTERACTING WITH NON-FORTRAN PROGRAMS AND FILES

1S.4.1

Using The Sharable High-Segment FOROTS

If your program does not contain a FORTRAN main program module,
and
you wish to have the sharable FOROTS GETSEGed at run time, you must do
the following:

Force the loading of the FOROTS initialization routine RESET.
by declaring it as an external.

Define the symbol FOROT% as a global with a positive, nonzero
value before FORLIB.REL is searched.

IS-l0

WRITI1~G

USER PROGRAMS

Initialize FOROTS by the appropriate initialization call:
JSP 16, RESET.
EXP 0

15.4.2

Calling Sequences

The following paragraphs describe the standard procedures for
subroutine calls.
1.

writing

Procedure
a.

The calling program must load
the accumulator
(AC)
16
with
the addre::;s of the first argument in the argument
list.

The subroutine i ::; then
using AC 17.

The return will be made to
the
after the PUSHJ 17 instruction.

The FOROTS trace facility requires the
to be:

called

PUSHJ

instruction

immediately

calling

sequence

XMOVEI 16,AP
PUSHJ 17,F
where AP is the ?ointer to the argument list and F is the
subprogram name.
The word preceding the first word of an
entry point should have its name in SIXBIT.
2.

15.4.3

Restrictions
a.

Skip returns are not permitted.

The contents of the pushdown stack located before the
address specified by AC 17 belong to the calling program;
they cannot be rl~ad by the called subprogram.

FOROTS assumes
that
it has control of the
stack;
therefore,
you must not create your own stack.
The
FOROTS stack is initialized by the call to RESET.
(See
Section 15.4.1).

Accumulator Usage

The specific functions performed by accumulators (AC) 17,16,0,
are:
1.

and

Pushdown Pointer - AC 17 is always maintained as a
pushdown
pointer.
In sectio~ zero, its right half points to the last
location in use on the stack, and its left half contains the
negative of the n~mber of words allocated to the unused
remaindE~r of the staGk.
In non-zero sections,
the pushdown pointer contains
global address of the last location in use on the stack.

15-11

the

WRITING USER PROGRAMS
2.

Argument List Pointer - AC 16
is used as the argument
pointer.
The called subprogram does not need to preserve its
contents.
The calling program cannot depend on getting back
the address of the argument list passed to the called
subprogram.
AC 16 cannot point to the ACs or to the stack.

Temporary and Value Return Registers - AC a and 1 are used as
temporary registers and
for
returning values.
The called
subprogram does not need to preserve the contents of AC 0 or
1
(even if not returning a value).
The calling program must
never depend on getting back the original contents of the
data passed to the called subprogram.

Returning Values - A subroutine
results by modifying arguments.

subprogram

may

pass

back

A numeric function subprogram always returns the value of the
function in AC a (or ACs 0-1 if the value is double precision
or complex).
A function subprogram may also pass back
additional
results by modifying the arguments.
(See Section
15.4.4 for a description of character functions.)
5.

Preserved ACs - FORTRAN function subprograms preserve
through 15; subroutine subprograms do not.

ACs

The design of the called subprogram cannot depend on the contents of
any of the ACs being set up by the calling subprogram, except for ACs
16 and 17.
Passing
information must be done explicitly by the
argument list mechanism.
Otherwise, the called subprograms cannot be
written in either FORTRAN or COBOL.

15.4.4

Argument Lists

Since the FORTRAN compiler uses the indirect bits on argument lists
(note that this permits shared,
pure code argument lists), it is
essential for code that accesses parameters to take this into account.
Specifically, sequences that obtained the values of parameters through
use of operations such as:
HRRZ R,1(16)
to pick up the address of the second argument should be changed to
XMOVEI R,@1(16)
The latter operation will work when interfacing with FORTRAN.
The format of the argument list is as follows:
Arg count word
Arg list addr.---First arg entry
Second arg entry

Last arg entry
The format of the arg count word is:
bits 0-17
bits 18-35

These contain -n, where n
is the number of arg
entries.
Reserved for future DIGITAL development, and must be

15-12

WRITING USER PROGRAMS
The format of an arg entry is as
wo rd) :
bit

bits
bits
bit
bits
bits

1-8
9-12
13
14-17
18-35

follows

(each

entry

single

IFIW (Instruction Format Indirect Word) flag, must be
1.
Reserved for future DIGITAL development, must be o.
Arg type code.
Indirect bit if desired.
Index field, must be O.
Address of the argument.

For character functions, the
first
argument points to
the
return
value,
which
is a character string descriptor (see Section 15.4.6).
The actual arguments to the function start
in the second argument
entry.
The following restrictions shculd be observed:
1.

Neither the argument list nor the arguments themselves can be
on the stack.
The same restriction applies to any indirect
argument pointers.

The called program may not modify the argument
list
itself.
The argument list may be in a write-protected segment.
Note that the arg count word is at position -1 with
respect
to
the contents of AC 16.
This word is always required even
if the subroutine does not handle a variable number of
arguments.
A subroutine that has no arguments must still
provide an argument list consisting of two words,
that
is,
the argument count word with a 0 in it and a zero argument
word.

Example:
XMOVEI 16,AP
PUSHJ 17,SUB

AP:

;SET UP ARG POINTER
;CALI. SUBROUTINE
;RETURN HEHE

;ARGUMENT LIST
-3, ,0
IFIW 4,A
IFIW 4,B
IFIW 4,C

;SUBROUTINE TO SET THIRD ARG TO SUM OF FIRST TWO ARGS
SUB:

MOVE
ADD
MOVEM
POPJ

T,@O 16)
T,@l 16)
T,@2 16)
17,

15-13

;GET FIRST ARG
;ADD SECOND ARG
;SET THIRD ARG
;RETURN TO CALLER

WRITING USER PROGRAMS
15.4.5

Argument Types

Table 15-1:

Argument Types and Types Codes

Type Code
(Octal)

Description

FORTRAN Use
COBOL Use
I--------------+-------------+---------,-----.--......j

o
1
2

3
4
5
6
7
10
11
12
13

14
15
16
17

Unspecified
FORTRAN Logical
Integer
Reserved
Real
Reserved
Octal
Label
Double real
(D-floating)
Not applicable
Double octal
Double real
(G-floating)
Complex
Character
Reserved
Hollerith

Unspecified
Not applicable
I-word COMP
Reserved
COMP-l
Reserved
Reserved
Procedure address
Not applicable
2-word COMP
Not applicable
Not applicable
Not applicable
Byte string descriptor
Reserved
Not applicable

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

Literal arguments are permitted, but they must reside
in a writable
segment.
This
is because the FORTRAN compiler makes a local copy of
all non array elements and may copy dummy arguments back to the actual
arguments.
All
unused
type codes are reserved for future DIGITAL
development.

15.4.6

Description of Arguments

The types of the arguments that may be passed are:
1.

Type 0 -

Unspecified

The calling program has not specified the type.
The called
subprograms should assume that the argument is of the correct
type if it is checking types.
If several types are possible,
the called subprogram should assume a default as part of its
specification.
If none of the above conditions is true,
the
called subprogram should handle the argument as an integer
(type 2).
2.

Type 1 -

FORTRAN logical

A 36-bit binary value containing 0 or positive
.FALSE.
and negative to specify .TRUE .•
3.

Type 2 -

Integer and l-word-COMP

A 36-bit 2's complement signed binary integer.

15-14

specify

WRITING USER PROGRAMS
4.

Type 4 - Real and COMP-l
A 36-bit floating-point number.
bit 0
bits 1-8
bits 9-35

sign
excess 128 exponent
mantissa

Type 6 - Octal
A 36-bit unsigned binary value.

Type 7 _. Label and pl-ocedure address
The address of the parameter is the address of
return label or a subprogram.

Type 10 - Double real

number

represented

Type 11 - 2-word COMP
A 2-word (72-bit)
word
word
word
word

alternate

(D-f1oating)

A double-precision floating-point
D-floating form.
(See Section 3.4.)
8.

1,
1,
2,
2,

2'~;

bit 0
bits 1-35
bit 0
bits 1-35

complement signed binary integer.
sign
high order
same as word 1, bit 0
low order

Type 12 - Double octal
A 72-bit unsigned binary value.

10.

Type 13 - Double real

(G-floating)

(KL model B only)

A double-precision floating-point
G-floating form.
(See Section 3.4.)
11.

number

represented

Type 14 - Complex
A complex number represented as an ordered pair of 36-bit
floatin9-point
numbers.
The first represents the real part,
and the second repre:,ents the imag inary part.

12.

Type 15 - Character string descriptor
The format of the character string descriptor is:
word 1:
word 2:

ILDB-type pointer,
that
is,
aimed at
the
character preceding the first character of the
string
EXP character count

The character descriptor may not be modified by the called
program.
The character string
itself must consist of a
string of contiguous 7-bit ASCII characters.

15-15

WRITING USER PROGRAMS
13.

Type 17 - Hollerith
A string of contiguous 7-bit ASCII characters left
justified
on the word
boundary of the first word and terminated by a
null character in the last word.
The FORTRAN compiler emits constants that are
padded with
spaces to a word boundary, followed by a full-word containing
zero.

15.4.7

Interaction with COBOL

FORTRAN programs
conversely,
the
subprograms.

can call COBOL programs as subprograms,
and,
COBOL
programs can call
FORTRAN programs as

Note that I/O operations can be performed only in subprograms that are
written
in the same language as the main program.
Also note that APR
trap handling will be done in a manner consistent with
the
language
used in the main program.

15.4.7.1 Calling FORTRAN Subprograms from COBOL Programs - COBOL
programmers may write subprograms in FORTRAN to use the conveniences
and facilities provided by this language.
The CbBOL verb ENTER
is
used to call FORTRAN subroutines.
The form of ENTER is as follows:

EWfER FORTRAN program name IuSING
\lprocedure-name-1\
~~~~~!i~~r-l I l~!{procedure-2
~~~~;!i~~~::!J,
[
The USING clause names the data within the COBOL program that is to be
passed to the called FORTRAN subprogram.
The passed data must be in a
form acceptable to FORTRAN (see Table 14-1).
NOTE
G-floating double-precision does not exist as
type in COBOL.

data

The calling sequence used by COBOL in calling a FORTRAN subprogram is:
MOVEI 16, address of first entry in argument list
PUSHJ 17, subprogram address
If the USING clause appears
in the
ENTER statement,
the compiler
creates an argument list that contains an entry for each identifier or
literal in the order of appearance
in the USING clause.
It
is
preceded by a word containing, in its left half, the negative number
of the number of entries in the list.
If no USING clause is present,
the argument list contains an empty word, and the preceding word is
set to O.
Each entry in the list is one 3n-bit word of the form:
0-8

9-12

13-35

type

address

Bits 0-8 are reserved for DIGITAL.

15-16

WRITING USER PROGRAMS
Bits 9-12 contain a type code that indicates
argument.

the

USAGE

the

Bits 13-35 contain the address of the argument of the first
of the argument; the address can be indexed or indirect.

word

Following is a description of the types generated by COBOL,
their
codes,
how the codes appear in the argument list, and the locations
specified by the addresses.
1.

For I-word COMPUTATIONAL items
CODE:
IN ARGUMENT LIST:
ADDRESS:
FORTRAN 'rYPE:

For 2-word COMPUTATIONAL items
CODE:
IN ARGUMENT LIST:
ADDRESS:
FORTRAN 'rYPE:

11
XWD 440, address
that of the high-order
argument
Not allowed

word

the

For COMPUTATIONAL-l items
CODE:
IN ARGUMENT LIST:
ADDRESS:
FORTRAN TYPE:

XWD 100, address
that of the argument itself
INTEGER

XWD 200, address
that of the argument itself
REAL

For DISPLAY-6 and DISPLAY-7 items
CODE:
IN ARGUMENT LIST:
ADDRESS:
WORDl:
WORD2 :

FORTRAN TYPE:

15
XWD 640, address
that of a
2-word descriptor
for
the
argument
a byte pointer to the
identifier or
literal
bit 0 is 1 if the item is numeric
bit 1 is 1 if the item is signed
bit 2 is 1 if the item is a
figurative
constant (including ALL)
bit 3 is 1 if the item is a literal
bits 4 through 11 are reserved
for
expansion
bit 12 is 1 if the
item has a PICTURE
with one or more Ps just before the
decimal point, that is, 99PPV.
bits 13 through 17 are the number of
decimal places.
If bit 12 is 1, this
is the number of Ps.
bits 18 through 35 contain the size of
the
item in bytes.
If FORTRAN is called, the string must be
DISPLAY-7,
nonnumeric,
and either a
figurative constant or literal. The bits
0-17 of words must be zero.
The FORTRAN
type is then character.

15-17

WRITING USER PROGRAMS
5.

For procedure names (which cannot be used for calls to
subprograms)
CODE:
IN ARGUMENT LIST:
ADDRESS:
FORTRAN TYPE:

COBOL

7
XWD 340, address
that of the procedure
External subprogram name

The return from a subprogram (through POPJ 17,)
after the call.

15.4.7.2 Calling COBOL Subroutines from FORTRAN
COBOL subprograms use the CALL statement:

the

statement

Programs - To

call

CALL sub (args •.• )
where sub is a COBOL subprogram.

15.4.8

Interaction with BLISS-36

FORTRAN programs can call
BLISS-36
routines as subprograms,
and,
conversely, BLISS-36 programs can call subprograms written in FORTRAN.
BLISS routines called by FORTRAN programs must be able
to coexist
compatibly with FOROTS.
For instance, they must use FUNCT.
functions
(see Section 18.6) for dynamic memory management within
the section
that
FOROTS
is
in, since FOROTS assumes that it has control of that
section.
One problem that the BLISS routines can encounter is stack
The size of the program stack as set up by FOROTS may be too
BLISS routines which have
several STACKLOCAL variables.
supported way to
increase the size of the stack is to use
FORLIB that has been assembled with a larger stack.

overflow.
small for
The only
a copy of

15.4.8.1 Calling FORTRAN Subprograms From BLISS-36 Programs - To call
a FORTRAN subprogram from a BLISS-36 program, the FORTRAN subprogram
must be declared an EXTERNAL ROUTINE in any module
that contains a
call
to
the subprogram.
In addition, if the FORTRAN subprogram is a
subroutine, it must be declared with a linkage
type of FORTRAN SUB.
If the FORTRAN subprogram is a function, then it must be declared-with
a linkage of FORTRAN_FUNC.
For example:
EXTERNAL ROUTINE FOO:
BAR:

FORTRAN SUB,
FORTRAN=FUNC;

declares FOO to be the name of a FORTRAN subroutine which will
be
called
in this module, and declares BAR to be the name of a FORTRAN
function.
After
the
FORTRAN
subprogram
has
been
declared
appropriately,
it can be called
just like any function written in
BLISS.

15-18

WRITING USER PROGRAMS
15.4.8.2 Calling BLISS-36 Routines From FORTRAN - A BLISS-36
routine
that
is to be called
from
a FORTRAN program must have either the
FORTRAN SUB linkage attribute (if the routine
is to be used as a
subroutTne)
or the FORTRAN FUNC linkage attribute (if the routine is
to be used as a funct ion) •
The programmer that wishes to write a BLISS-36 routine
to be called
from FORTRAN must be familiar with the calling sequence used by
FORTRAN to call subprograms (see Section 15.4.2),
FORTRAN argument
lists
(see Section 15.4.4),
and FORTRAN argument descriptors (see
Sections 15.4.5 and 15.4.6). This knowledge is necessary because the
values of the formal arguments of the BLISS-36 routine are the FORTRAN
argument list entries that correspond to actual arguments of the BLISS
routine.
compiled
with
In
general,
the
BLISS
routines
must
be
ADDRESSING MODE (INDIRECT) ,
or MACHOP calls must be used to generate
any instruction that references formals since all
FORTRAN arguments
must be accessed through
indirect addressing.
This must be done
indirect bit
in
because the FORTRAN compiler frequently sets the
argument lists (see Section 15.4.4).
See the BLISS-36

.~anguage

Guid~

for more information.
NOTE

ADDRESSING_MODE (INDIRECT)
can
have
far
reaching
effects on your program, which you should understand
fully before using.

15.4.9

LINK Overlay Facilities

LINK provides several routines that are accessible directly from a
FORTRAN program.
These routines are presented here briefly, together
with the FORTRAN specification of their parameters.
In general,
LINK
performs these functions automatically.
These routines are available
only for your convenience.
Full details of the use of the overlay
facilities can be found in the LINK Reference Manual.
NOTE
Overlays are not allowed
addressing is used.
The following terms are
overlay routines.

used

when

TOPS-20

describe

the

extended

LINK

File spec

A
character
expression
consisting
'dev:file.ext[directory] ,
(TOPS-IO) ,
'dev:(directory>file.typ.gen' (TOPS-20) •

of
or

Name

A quoted string giving a
link name,
or an
integer constant or variable giving a link
number.

List of link names

A sequence of name items separated by commas.

15-19

parameters

WRITING USER PROGRAMS
The routines available are:
CLROVL

Specifies a non-writable overlay.

GETOVL(List of link names)

Used
change
to
structure in core.

INIOVL(File spec)

Used to specify the overlay file to
be
found
if
the
load
time
specification is to be overridden.

LOGOVL(File spec)

Used to specify where the log
file
is to be written.
If no arguments
are given, the log file is closed.

REMOVL(List of link names)

Removes the
core.

specified

links

from

RUNOVL(Name)

Loads the
specified
transfers to that LINK.

link

and

SAOVL

Specifies a writable overlay.

the

For a full description of these routines, refer to the LINK
Manual.

overlay

Reference

NOTE
The SAVE statement retains
the values stored
in
a
variable,
array, or common block after execution of a
RETURN or END statement
in a
subprogram.
When
overlays are used, the SAVE statement must be used to
ensure retention of values.
When the SAVE statement
is used,
it
is not necessary to specify the LINK
switch /OVERLAY:WRITABLE when loading a
program
(see
Section 7.10).

15.4.10

FOROTS and Overlay Memory Management

For sharable FOROTS, the FOROTS static data area
is several
pages
located at
the top of FOROTS.
FOROTS dynamic memory is allocated at
runtime below FOROTS and in a downward direction
(toward the
user's
program) .
For nonsharable FOROTS (FOROTS loaded from FORLIB),
the
FOROTS data
area
is
located
in the low segment, so that it will be linked with
variables used by the
user's program.
FOROTS dynamic memory is
allocated at
runtime starting at the page designated by the symbol
STARTP
in FORPRM.MAC,
downward
toward
the
user
program.
The
distributed value for this page number is 577.
If the user's program
has two segments, FOROTS allocates memory down
to
the user's high
segment,
skips over
the high segment, and begins allocating memory
below the user's high segment toward the user's low segment.
For both sharable and nonsharable FOROTS, when FOROTS can no
longer
allocate memory downward toward the user's low segment, it allocates
memory starting at the top of memory downward.
When FOROTS can no
longer allocate any memory, it reports:
?Memory full
and returns to the monitor after attempting to close all files.
15-20

WRITING USER PROGRAMS
Figure 15-1 illustrates the run-time memory layout.

Page

777
Reservl:!d for SORT, DBMS, and DDT

600
FOROTS

FOROTS Static Data

500
FOROl S Dynamic Data

-1- - - -

-1- - - - - - - Used fl)r Dynamic
Character Concatenations

User Program

o
MR-S-3878-85

Figure 15-1:

Run-time Memory Layout for Section Zero

FOROTS has a separate memory manager specifically designed for use by
OVRLAY.
This memory managE'r allocates memory at the top of the
users's low segment.
In general,
user programs that
use overlays
should not use the FUNCT. calls GAD, COR, and RAD.
OVRLAY expects to
be able to use memory beginning at the top of the user's low segment,
allowing
for
a special FOROTS scratch space allocated by the /SPACE
switch.
Under certain circumstances,
notably when concatenating character
expressions
whose
length
is not known until
runtime,
FOROTS
dynamically allocates a special scratch area
using
the FUNCT.
COR
function
(see Section 18.6).
This area is allocated the first time a
dynamic concatenation is performed.
such
The /SPACE switch to LINK determines the maximum size for
The default value
"dynamic concatenations"
if overlays are used.
(See the LINK Reference
given by LINK for this value is 4000 (octal).
Manual.)

15-21

WRITING USER PROGRAMS
You can allocate space for dynamic character concatenation
ALCCHR subroutine (see Section 13.4.1.1).

15.4.11

using

the

Extended Addressing Memory Layout (TOPS-20 only)

The FORTRAN compiler must determine which psect every word of data or
code should reside in.
For non-extended addressing compilations, the
data and code reside in either the low segment or the high segment.
For
extended addressing compilations,
there are three segments
(psects) in which the data and code can be allocated:
1.

The code psect corresponds to the non-extended high
The default name is .CODE ••

segment.

The data psect corresponds to the non-extended
The default name is .DATA ••

low

segment.

The .LARG. psect is where large data resides.
name of this psect cannot be changed.)

(Note that the

A data item can be placed
following ways:

the

.LARG.

psect

either

the

The data item is an array or any character scalar whose size
is greater or equal to the value of the /EXTEND:DATA switch
(default 10,000 words).
(See Section 16.5).

The data item is placed in a COMMON
group that is in the .LARG. psect.

block

Table 15-2 describes the various memory allocations for
non-extended compilations.

15-22

EQUIVALENCE
extended

and

WRITING USER PROGRAMS
Table 15-2:

Memory Allocations for !EXTEND and /NOEXTEND

Item

/NOEXTEND

!EXTEND

User subprogram

Hiseg

Code

FORLIB

Hiseg

.CODE.

Argument blocks

Hiseg

Code

Compile-time constant
character descriptors

Hiseg

Code

Array dimension information

Hiseg

Code

EFIWS

N/A

Code

Symbol tables (from LINK)

Lowseg or Hiseg

.DATA.

FORMAT statements

Lowseg

Data

Constants

Lowseg

Data

Small arrays and scalars

Lowseg

Data

Large arrays

Lowseg

.LARG.

COMMON variables

Lowseg

.LARG.

Variables EQUIVALENCED to
large arrays

Lowseg

.LARG.

PDV

N/A

.DATA.

(by default)

NOTE
When the sharable FOROTS is used,
LINK loads the
/NOEXTEND high-segment
into the low-segment.
This
is done so that the sharable FOROTS can be used as
the high-segment (see Section 16.9)

15-23

CHAPTER 16
USING THE FORTRAN COMPILER

This chapter explains how to use the FORTRAN compiler.
Section 16.1
describes using
the FORTRAN-I0 compiler and Section 16.2 describes
using the FORTRAN-20 compiler.

16.1

USING THE FORTRAN-IO COMPILER

This section describes how use the FORTRAN-I0 compiler. You should be
familiar with the TOPS-IO operating system.
The TOPS-IO Operating
System provides commands that enable you to compile, execute, and
debug FORTRAN programs.
These commands are known as the COMPILE-Class
commands.

16.1.1

TOPS-IO COMPILE-Class Commands

You can invoke the FORTRAN-IO compiler by using TOPS-IO COMPILE-Class
commands.
These commands enable you to compile, execute, and debug a
program by giving the commands at TOPS-IO command level.
The COMPILE-Class commands are:
COMPILE
LOAD
EXECUTE
DEBUG
Example:
.EXECUTE ROTOR. FOR
The compiler switches OPTIMIZE,
CREF, and DEBUG may be specified
directly in COMPILE-Class commands and may be used globally or
locally.
(See Section 16.1. 3 for a description of
FORTRAN-IO
switches.)
Example:
.EXECUTE/CREF Pl.FOR,P2.FOR/DEBUG
The other compiler swi tches
specific source file.

mu~~t

be passed

16-1

parentheses

for

each

USING THE FORTRAN COMPILER
Example:
.EXECUTE Pl.FOR(INCLUDE)
Refer to the TOPS-IO Operating System Commands
information about the COMPILE-Class commands.

16.1.2

Manual

for

further

RUNNING THE FORTRAN-IO COMPILER

On TOPS-lO, the command to run the FORTRAN compiler directly is:
.R FORTRA
The compiler responds with an asterisk
(*), and is then ready to
accept a command string.
The form of the FORTRAN compiler command
string is:
object filespec, listing filespec=source filespec(s)
You are given the following options:
1.

File specifications consist of an optional device name, a one
to six character filename, an optional one to three character
file extension, and an optional directory path specification.
The path may include SFDs.

You may specify more than one source file in the compilation
command string.
These files will be logically concatenated
by the compiler and treated as one source file.

More than one program unit
source file.

A program unit may consist of more than one source file.

If no object file is specified, no relocatable binary file is
generated.

If no listing file is specified, no listing is generated.

If no extension is given, the defaults are the following
the respective files:

may

contained

single

for

.LST (listing)
.REL (relocatable binary)
.FOR (source)
.CRF (cross
reference)
if the
specified (see Section 16.1.3)

16.1.3

/CROSSREF

switch

TOPS-IO Compiler Command Switches

anywhere
in the
Switches to the FORTRAN-I0 compiler are accepted
command
string.
They are
totally position and file independent.
Table 16-1 lists the switches.

16-2

USING THE FORTRAN COMPILER
Table 16-1:

FORTRAN-IO Compiler Switches

Switch

Meaning

De fa ul ts

/CROSSREF

Generates a file with extension .CRF
that can be input to the CREF program.

OFF

/DEBUG

Includes debugging information in your
program (see Section 16.3).

NONE

/DFLOATING

Indicates that double-precision
numbers are stored in D-floating
format. (See Sectio:1 3.4.)

/EXPAND

Includes the octal-formatted version
of the object file in the listing.

OFF

/F66

The FOHTRAN-66 standard rules apply for
DO loops and EXTERNAL statements. (Same
function as the /NOF77 switch.)

OFF

/F77

The FOHTRAN-77 stan<lard rules apply for
DO loops and EXTERNAL statements.

/FLAG

Invokes the compatibility
Section 16.6).

OFF

/GFLOATING

Indicates that double-precision
numbers are stored in G-floating
format .. (See Section 3.4.)

OFF

CompilE~s

OFF

/INCLUDE

flagger

(see

a D in card column 1 as

space.
/LNMAP

Produces a line number/octal location
map in the listing only if /MACROCODE
was not specified.

OFF

/MACROCODE

Adds the mnemonic translation of the
object code to the :.isting file.

OFF

/NOFLAG

Indicates that no compatibility
flagging will be done (seE! Section
16.6) •

/NOF77

The FORTRAN-66 standard rules apply for
DO loops and EXTERNAL statements. (Same
function as the /F6(1 swi tch.)

OFF

/NOERRORS

Does not print error messages on the
terminal.

OFF

/NOWARN

Suppresses warning nlessages (see
Section 16.4).

NONE

/OPTIMIZE

Performs global optjmization.

OFF

/SYNTAX

Performs syntax check only.

OFF

16-3

USING THE FORTRAN COMPILER
Each switch must be preceded by a slash (/).
Switch names need only
contain those letters that are required to make the switch name
unique.
You are encouraged to use at least three letters to prevent
conflict with switches in future implementations.
Example:
.R FORTRA
*OFILE,LFILE=SFILE/MAC,S2FILE
The /MAC switch will cause the
S2FILE to appear in LFILE.LST.

~ACRO

code

generated

for

SFILE

All switches, used or implied, are printed at the top of each
page.

16-4

and

listing

USING THE FORTRAN COMPILER
16.2

USING THE FORTRAN-20 COMPILER

This section describes how to use the FORTRAN-20 compiler.
You should
be familiar with the TOPS-20 Operating System.
The TOPS-20 Operating
System provides commands that ~nable you to compile,
execute,
and
debug FORTRAN program.
These commands are known as the COMPILE-Class
commands.

16.2.1

TOPS-20 COMPILE-Class Commands

The TOPS-20 COMPILE-Class commands enable you to initiate compilation,
execution,
and debugging of FORTRAN programs from TOPS-20 command
level.
The TOPS-20 COMPILE-Class commands are:
COMPILE
LOAD
EXECUTE
DEBUG
Example:
@EXECUTE ROTOR. FOR
The following FORTRAN compiler switches (see Section
specified directly in a COMPILI:-Class command:

16.2.3)

can

/ABORT
/BINARY
/CROSS-REFERENCE
/DEBUG
/LIST
/MACHINE-CODE:
/NOWARNINGS
/OPTIMIZE
/WARNINGS
NOTE
When you specify the svllitches /BINARY,
/DEBUG,
/LIST,
and
!NOWARNINGS
directly
in
a
COMPI LE-Class comrland, the swi tches cannot have
arguments as they can when running the compiler.
All other switches must b(~ specified by using
as shown in the following example:

Language-switches,

@COMPILE TEST.FOR/OPT/LANG:"/NOWARNINGS"
Refer
to the TOPS-20 Commands Reference Manual
information about the COM1)ILE-Class commands.

for

NOTE
You cannot use long TOPS-20
COMPILE-Class commands.

16-5

filenames

with

the

USING THE FORTRAN COMPILER
16.2.2

RUNNING THE FORTRAN-20 COMPILER

On TOPS-20, the command to run the FORTRAN compiler directly is:
@FORTRA
The compiler responds with the following prompt:
FORTRAN>
and is then ready to accept a command string.
You can use the question mark to list the commands beginning with
a
specific letter or
letters.
Type
the letter or letters
followed by a
question mark.
(Refer
to
the TOPS-20
User's
Guide.)
You can type commands to the system by using either
full
input,
recognition
input,
abbreviated input, or a combination of these
methods.
To give a command using full input,
type
the complete command
name,
arguments, or switches (if any), using a space to separate
the fields.
To give a command using recognition input, type a portion of the
switch or
filename and press the ESC key.
You must type enough
of the switch or filename to make it unique.
Continue typing and
pressing
the
ESC key until the switch or filename is complete.
Recognition input requires less typing than full
input,
so you
are
less likely to make a mistake.
You can use recognition in
typing switches, switches arguments,
and
file
specifications.
When
typing
file
specifications,
you can also use CTRL/F to
complete the rest of a partial file specification.
To give a command using abbreviated input, type only enough of
the switch or filename to distinguish one switch or filename from
another.
(Refer to the TOPS-20 User's Guide.)
You should enter a command string in one of the following forms:
1.

[COMPILE]<source-file-spec>

[switches]

If no switches are specified, the compiler
produces a
.REL
file,
with
the same filename as the source file.
The user
must use a /LISTING switch to get a listing file (see Section
16.2.3 for
a description of FORTRAN-20 compiler switches).
COMPILE is optional if the command begins with a
switch or
begins with a file specification that cannot be confused with
the words EXIT, HELP, TAKE, or COMPILE.

16-6

USING THE FORTRAN COMPILER
2.

[COMPI LE] < so urce- f i 1 e--s pec>+<source- f i 1 e-s pec>+ •••
[swi tches]
The source files are ':reated as if they were
together prior to the beginning of compilation.

concatenated

If no switches are specified, the compiler produces a
.REL
file, with the same filename as the last source file in the
list. The user must Ilse a /LISTING switch to get a listing
(see Section 16.2.3 for a description of FORTRAN-20 compiler
switches). COMPILE is optional if the command begins with a
switch or begins with a file specification that cannot be
confused wi th the words EXI~{" HELP, TAKE, or COMPILE.
3.

TAKE <file-spec> [/ECIlO]
The compiler reads the file specified as the command
input
stream.
The TAKE command is legal within 'take' files.
The
ma x i mum n est i n g de p t his 1 0 't a k e ' f i 1 e s •
The
on
can
the

/ECHO switch optionally causes commands to be displayed
TTY:
as they ar~ executed. The optional /NOECHO switch
be used on a nestl~d takE~ command to cancel the affect of
/ECHO switch while processing that nested command file.

RUN <file-spec> [/OFF:3ET:<integer>]
This command runs another program (for example, LINK).
It
causes an exit from the FORTRAN compiler and the start of
execution of the prog~am indicated by the file specification,
with the additional option of starting that program at an
OFFSET relative to thl~ normal starting address.

HELP
This command prints il1formation on the user's terminal
how to use the FORTRAN compiler.

about

EXIT
This command exits from FORTRA.

You are given the following options:
1.

Filename specificatio:1s consist of the following:
An optional device name (the default device is DSK:)
An optional

dir~ctory

name

An up to 39 alphanumeric character filename
An optional up to 39 alphanumeric character file type
An optional g~neration
version of the Eile

number

that

identifies

An optional file attribute to specify
characteristics of a file specification
(Refer to the TOPS-20 User's Guide)

16-7

the

distinctive

USING THE FORTRAN COMPILER
NOTE
LINK is restricted
to
3-character extensions.

6-character

filenames

and

You may specify more than one source file in the compilation
command string.
These files will be concatenated by the
compiler and treated as one source file.
The name of the
last source
file is used as the default name of the object
and listing files.
If the last source file does not have a
name
(such as, TTY:), FORTRAN-OUTPUT is used as the default
filename.

More than one program unit may
source file.

A program unit may consist of more than one file.

If no /LISTING switch is specified (see Section 16.2.3),
listing is generated.

If no extension is given, the defaults are the following for
the respective files:
.LST (listing)

contained

single

if the /CROSSREF switch is not specified

.CRF (cross reference)
if the
specified (see Section 16.2.3)

/CROSSREF

switch

.REL (relocatable binary)
.FOR (source)

16.2.3

TOPS-20 Compiler Commands Switches

Switches to the FORTRAN-20 compiler are accepted anywhere
in the
command string.
They are totally position and file independent.
Table 16-2 lists the switches.

16-8

USING THE FORTRAN COMPILER
Table 16-2:

FORTRAN-20 Compiler Switches

Switch

Meaning

Defaul ts

/ABORT

Causes the compiler to exit at the
end of a compilation that contains
errors.

OFF

/BINARY[:relfile]

Indicates that a relocatable binary
file is qenerated. You can
optionally specify the file
specification.

/CROSS-REFERENCE

Generates a file with extension
.CRF that can be input to the
CREF proqram.

OFF

/DEBUG [keys: ]

Includes debugging information in
your proqram (see Section 16.3).

NONE

/DFLOATING

Indicates that double-precision
numbers are stored in D-floating
format. (See SE~ction 3.4.)

/ECHO-OPTION

Echo switches selected from the
SWITCH. Irn filE~.

OFF

/EXPAND

Includes the octal-formatted version
of the object file in the listing.

OFF

/EXTEND[keys:]

Indicates extended addressing.
Programs can have up to 30
sections of code and data (see
Sec t ion :~ 6 • 5) •

OFF

/F66

The FORTHAN-66 standard rules apply
for DO loops and EXTERNAL statements. (Same function as the /NOF77
switch.)

OFF

/F77

The FORTHAN-77 standard rules apply
for DO loops and EXTERNAL
statements.

/FLAG-NON-STANDARD

Invokes t.he compatibility flagger
(see Section 16.6).

OFF

/GFLOATING

IndicateE; that double-precision
numbers are stored in G-floating
format. fSee Section 3.4.)

OFF

/INCLUDE

Compiles a D in card column 1 as
space.

OFF

/LISTING[:listfile]

Indicates a list file will be
generated. You can optionally
specify the file specification.

OFF

16-9

USING THE FORTRAN COMPILER
Table 16-2:

FORTRAN-20 Compiler Switches (Cont.)

Switch

Meaning

Defaults

ILNMAP

Produces a line number/octal
location map in the listing only
if IMACHINE-CODE was not specified.

OFF

IMACHINE-CODE

Adds the mnemonic translation of
the object code to the listing
file. This command will cause
a default ILISTING.

OFF

INOBINARY

Indicates that no relocatable
binary file is generated.

OFF

INOF77

The FORTRAN-66 standard rules apply
for DO loops and EXTERNAL
statements. (Same function as
the IF66 switch.)

OFF

INOFLAG-NON-STANDARD

Indicates that no compatibility
flagging will be done (see Section
16.6) •

INOERRORS

Does not print error messages
on the terminal.

OFF

INOEXTEND

Indicates extended addressing
is not in effect (see Section 16.5).

INOWARN

Suppresses warning messages (see
Section 16.4).

NONE

IOPTIMIZE

Performs global optimization.

OFF

IOPTION [:option]

Only read lines from the SWITCH.INI
file that start with FORTRA:option.

OFF

ISYNTAX

Performs syntax check only.

OFF

~----

____________________L -________________________________________- L__._ _ _ _ _ _

Each switch must be preceded by a slash (I).
Switch names need only
contain
those letters
that are
required
to make the switch name
unique.
You are encouraged to use at least three letters
to prevent
conflict with switches in future implementations.
NOTE
When
using
switches
in control
files,
you
encouraged to type the full name of the switch.

16-10

are

USING TEE FORTRAN COMPILER
Example:
@FORTRA
FORTRAN)SFILE+S2FILE/MAC/LIST:LFILE
The /MAC switch will cause the MACRO code generated
for
SFILE and
S2FILE to appear
in LFILE.LST.
An relocatable binary file will be
created with the name S2FILE.REL.
All switches, used or implied, are printed at the top of each
page.

16.3

listing

THE /DEBUG SWITCH

The /DEBUG switch tells FORTRAN to compile a series of debugging
features
into
your
program.
Several of these
features are
specifically designed to be used with the
FORTRAN debugging
program
(FORDDT).
Refer
to Chapter
17 for more information.
By using the
DEBUG switch arguments listed in Table 16-3, you can include specific
debugging features.
The form of the /DEBUG switch is:
/DEBUG:arg
or
/DEBUG: (argl,arg2, ••• )

16-11

USING THE FORTRAN COMPILER
Table 16-3:

Arguments to /DEBUG Switch

Arguments

Meaning

~----------------+-----------------------------------------------------------~

DIMENSIONS

Includes dimension information in
FORDDT.

.REL

file

for

TRACE

Generates references to FOR DDT required for
its
trace features (automatically activates LABELS).

LABELS

Generates a label for each statement of the form
<line-number>L.
(This option can be used without
FOR DDT .)

INDEX

Forces DO LOOP indexes to be stored at the
beginning of each iteration rather than held in a
register for the duration of the loop.
In addition,
this switch forces all function
values to be stored in memory prior to return from
the function.
If this switch is specified, you
can set a FOR DDT pause on the RETURN statement
(see Section 13.4.4) and then examine the value to
be returned.

BOUNDS

Generates the bounds checking code for all array
references
and
substring references.
Bounds
violations will produce run-time error messages.
Note that the technique of specifying dimensions
of 1 for subroutine arrays will cause bounds check
errors.
(You may use this option without FORDDT.)

ARGUMENTS

Generates type checking information at load time
for actual argument types and associated dummy
argument types.
Type violations will produce
non-fatal load-time error messages.
This switch
also performs type checking at compile-time for
statement functions.

NONE

Do not include any debug features.

ALL

Enable all debugging aids.

' " -_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _--L-_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _----J

Options available with the /DEBUG arguments are:
1.

No debug features - Either do not specify the
or include /DEBUG:NONE.

All debug features - Either /DEBUG or /DEBUG:ALL.

Selected features - Either a
that is:
/DEBUG:BOU/DEBUG:LAB
or a list of modifiers
/DEBUG: (BOU, LAB, ..• )

16-12

series

/DEBUG

modified

switch

switches,

USING TilE FORTRAN COMPILER

Exclusion of features - If you wish all but one or
modifiers and do not wish to list them all, you can use
prefix "NO" before the swi.tch you wish to exclude.
exclusion of one or more features implicitly includes all
others,
that
is,
/DEBUG:NOBOU
is
the
same
/DEBUG:(DIM,TRA,LAB,:ND,ARG) •

If you include more than one statement on
first
statement will
receive a
label
reference (/DEBUG:TRACE).

two
the
The
the
as

a
single line,
only the
(/DEBUG:LABELS)
or
FORDDT

NO'I'E
If a source file contains line sequence numbers that
occur more than once
in the same subprogram, the
/DEBUG option cannot be used.
Also, the /DEBUG option
and
the /OPTIMIZE option cannot be used at the same
time.
The following formulas may be used
to determine the
increases
in
program size
that will occur as at result of the addition of various
/DEBUG options.
DIMENSIONS

For each array, 3+3*N words where N is the number
of dimensions, and up to three constants for each
dimension.

'I'RACE

One instruction per executable statement.

LABELS

No increase.

INDEX

One
instruction
per
inner
loop
plus
one
instruction
for
some of the references to the
index of the loop.
Also one
instruction per
subprogram.

BOUNDS

For each array,
DIMENSIONS.

the

formula

the

same

For each reference to an array element, 5+N words
additional
words are generated, where N is the
number of dimensions in the array.
If you do not
specify BOUNDS, approximately 1+3*(N-l) words will
be used.
For each reference to a substring, add 5
words.
ARGUMENTS

increa~;e.

If the /DEBUG:ARGUMENTS switch argument is specified, type checking is
performed at LINK time for calls to external programs and at compile
time for calls to statement functions.
Non-fatal error messages are
issued at LINK time for the following cases:
1.

If the number of arguments in the called subprogram
calling program unit are not equal.

If the length of an array or character scalar actual argument
is less than that of the corresponding dummy argument.
(This
is checked only if the length of the actual
is known at
compile time.)

16-13

and

the

USING THE FORTRAN COMPILER
3.

If the associations 'of actual argument data types with dummy
argument data
types are other than those indicated as legal
in Table 16-4.

routine name
is
If a non-routine name is passed where a
expected,
or
a
routine name is passed where a non-routine
name is expected.

Non-fatal error messages are issued at compile time for the
cases (only for statement functions):

following

routine name
is
If a non-routine name is passed where a
expected,
or
a
routine name is passed where a non-routine
name is expected.

If the length of the actual character expression being passed
to
a
statement
function
is less than that of the
corresponding character expression dummy argument.

Table 16-4:

Legal Dummy and Actual Argument Associations
Actual Argullwnt Type

Alternate
Heturn
Label

Iloublp
Logical Intcger Heal D-tloating (i-tloating Complex Character (1<'1'11 IlolI('rith Oetal

Alternate
Return
III

Label

;.,

Eo<

Logical

E
~

Integer

;.,

Heal

III

E
E
;:s
Q

D-tloating
G-floating

X
\"

Complex

:'O.~

Character

X mdicates legal associations. All others will cause a warning to be issued if DEBUG:ARGUMENTS is specified.

16.4

THE /NOWARN SWITCH

The /NOWARN switch is used to suppress compiler warning messages.
If
this switch
is
used with no arguments,
all warning messages are
suppressed.
The /NOWARN switch may also be
used
with arguments as
shown in Table 16-5.

16-14

US ING THE
Table 16-5:

FOR~rRAN

COMPI LER

Arguments to /No\'lARN Switch

Arguments

Meaning

ALL

Suppress all warning messages.

NONE

Do not suppress warning messages.

xxx

Where xxx i~; the three character error mnemonic
for
the el~ror mE~ssage to be suppressed.
This is
example
the three letters that follow %FTN,
for
%FTNABD.

For example,
/NOWARN:LID
will suppress all warnings of
characters.

identifier

A list of arguments is also allowed.

having

than

six

For example,

/NOWARN: (LI D, DIM)
would suppress both LID and DIM types of warning messages.
(See Appendix C for a list of all compiler error mnemonics.)

16.5

THE /EXTEND SWITCH (TOP!;-20 ONLY)

Support for extended addressing is almost completely transparent;
a
program compiled with the /EXTEND switch will use extended addressing
without requiring! changes to the FOHTRAN source program.
When /EXTEND is specified, an-ays and COMMON blocks can extend across
multiple sections.
Executable code can also
reside
in multiple
sections, with the restriction that a single subprogram must not cross
a section boundary.
The /EXTEND switch can be
spE~cified
without arguments
to
use
the
default extended address space layout.
This is suitable for most
applications in which
the executable code fits
within a
single
section, but which may employ data structures that require more than a
section of memory.
When such an application
is compiled /EXTEND
without
further
arguments,
a default memory layout
is used that
depends on the default size settings for arrays and strings:

16-15

USING THE FORTRAN COMPILER
Table 16-6:

/EXTEND Default Memory Layout

Section

Pages

Contents

0-477

Executable code
Argument blocks
Literals
Non-COMMON scalars
Non-COMMON arrays smaller than
10,000 words
Non-COMMON strings smaller than
10,000 words

SOO-S77

FOROTS

600-777

Reserved

_.
_.

2-31

0-777

COMMON blocks
Arrays larger than 10,000 words
Strings larger than 10,000 words
-.

Additional arguments to /EXTEND (COMMON:, DATA:, PSECT:, and CODE), in
conjunction with the /SET switch to LINK, can be supplied to override
these defaults and direct specific
placement of COMMON blocks,
non-COMMON arrays and scalars, and executable code.
For example, a
decimal argument can be specified
for
the
DATA:
keyword,
which
overrides the 10,000 word default minimum size for large arrays and
character scalars.
/EXTEND:COMMON can be
used
to direct specific
named COMMON blocks
to non-default sections of extended memory.
(Section 16.S.3 describes the /EXTEND arguments.)

16.S.1

/EXTEND and Applications with Large Arrays

The simplest usage of extended addressing is for applications in which
the executable code fits in one section.
These applications may have
very large arrays or common blocks.
In many cases, these applications
can be compiled with the /EXTEND switch with no arguments, using the
default memory layout in Table 16-6.
In some cases, you may need to use the DATA or COMMON arguments
to
/EXTEND to
redirect the placement of variables in areas when default
placements cannot be used.
Specifying a smaller /EXTEND:DATA size may
be necessary if the total size of non-COMMON scalars and arrays causes
them to overlap FOROTS pages.
If such overlap occurs,
program
execution will terminate with the error messages:
"?Can't get FOROIO.EXE"
and
"?Can't overlay existing pages"
A smaller DATA size will force these data structures into a non-FOROTS
section.

16-16

USING TEE FORTRAN COMPILER
16.5.2

/EXTEND and Applications with Large Executable Code

A more complex use of extended addressing
is one
where
the
application's executable code is larger than a section.
In order to
build an application that has more than one section of executable
code, you have to specify which program units will be linked together
in each section.
When /EXTEND is specified, the extended code will be in three PSECTS,
one analogous
to the
present hiseg,
one analogous to the present
lowseg, and one containing the large variables.
If the user does not
specify the extended addressing switch, FORTRAN will generate a TWOSEG
REL file as it always has.
The three PSECTS generated
under /EXTEND
are:
1.

The "large data area" psect (.LARG.) has a maximum size of 30
sections.
It will consist of user-specified COMMON blocks,
arrays and character scalars that are larger
than
a
user-specified size (or default).
There is no restriction on the size of an individual array or
COMMON block, beyond the restriction on the total size of the
large data area.

The "small data area"
psect
(default name
.DATA.)
of a
program consists of Dser-specified COMMON blocks, and scalars
and arrays that are smaller than a
user-specified size
(or
default).
If FORDDT is used, it will reside in the • DATA.
psec t.

The "executable code and sharable data area"
psect
(default
name
.CODE.)
of a program consists of all executable code,
argument blocks anc
literals.
Library
functions
and
subroutines
used by the program are placed in the .CODE.
psect.
This does not
include space used
by SORT, which
occupies its own section.

For each "executable code area"
(code psect)
there will
be a
corresponding "small data area" (data psect).
The combined size of
these two areas must not exceed 256K words.
The default psect names
.DATA.
and .CODE. can be changed at compile time by the /EXTEND:PSECT
command switch (see Section 16.5.3).

16.5.2.1 /EXTEND PSECT Placement - The three psects will
according to the table below:

set

Psect

De fa ul t
Origin

Attributes

data

1001000

code

1300000

• LARG.

2000000

Sing IE section, Non-zero section,
Concatenated, Writable
Single section, Non-zero section,
Concatenated, Read-only
Non-ZEro section, Concatenated, Writable

-------

16-17

USING THE FORTRAN COMPILER
You can alter the default
psect origins when loading
programs by
giving a /SET switch to LINK (see the LINK Reference Manual).
This
may be necessary if a program has' too much code or local data
to
fit
in the regions allocated for them by the default psect origins.
This
would cause the psects to overlap, and would be indicated
by a
LINK
warning message
(%LNKPOV).
Altering the defaults is also useful if
the user wants to reserve one or more sections for some purpose
(such
as telling the monitor where dynamic libraries should be loaded).
The first page (locations 000-777) of any section that contains code
is
reserved
for
use by FORDDT.
The
LINK /SET switch should not
specify a psect origin less than 1000 for any such section.

16.5.2.2 Building Large-Code Applications - When compiling large-code
applications, the following considerations apply:
1.

The code and small data psects for a single program unit must
always reside together in the same section, since the PC will
not advance across section boundaries,
and
the
small data
area is assumed to be in the same section.

keyword.
This
You will
need
to
use
the /EXTEND:CODE
specifies that
the object code emitted by the compiler
assumes that all subprograms that
it calls may be
in a
separate section.

You must also use the PSECT argument to /EXTEND to specify
the psect names for code and small data.
If you wish to link
the program units in several different compilations
(source
files)
togeth~r in the same section, you should use the same
psect names for those program units.

At link-time, you must specify the starting address
for
using the /SET switch, according to the following rules:
1.

The code and small data psects for a
program
linked in the same section with each other.

If a section contains any executable code,
section is reserved for FORDDT and FOROTS.

16-18

each

unit

page

psect

must

that

USING THE FORTRAN COMPILER
3.

You must always allo~ate space for
the
.CODE.
and
.DATA.
psects, since FORLIB routines will be linked in these psects.

Pages 500-577 of the section that contains .CODE. and
are reserved for FOROTS.

Pages 600-777 are reserved for FOROTS I/O buffers and DDT.

You must always allocate space for the .LARG. psect.

16.5.3

.DATA.

Arguments to /EXTEND

By using the /EXTEND switch al'guments listed in Table
include specific extended addlessing features.
The form of the /EXTEND switch is:
/EXTEND:arg
or
/EXTEND: (argl,arg2, ••• )

16-19

16-7,

you

can

USING THE FORTRAN COMPILER
Table 16-7:

Arguments to /EXTEND Switch

Arguments

Meaning

CODE

Specifies that the object code produced by the
compiler
has to assume that any subprogram that
it calls could be in a separate section.
NOCODE
is the default.

COMMON [ : name]
or
COMMON: (name, ••• )

Without a
common block name specified,
causes
all common blocks
that
have not already been
allocated by /EXTEND: [NO] COMMON to be allocated
in the
.LARG.
psect.
This
is the default.
Individual
common
blocks
can
be
placed
explicitly in .LARG. by putting their names in a
list after COMMON:.
When you explicitly place
an individual common block in .LARG., any common
blocks that have not already been allocated
by
/EXTEND: [NO] COMMON are placed in the small data
psect.

DATA[:decimal number]

Specifies a decimal argument that is the minimum
si ze
(in words)
for
non-common arrays and
character scalars, which will
be allocated
to
the .LARG. psect.
The default is 10,000.

NOCODE

Allows the compiler to assume that all
of the
code will
be
in
the same section.
A program
compiled with the NOCODE argument cannot call
any subprograms compiled with the CODE argument.
This is the default.

NOCOMMON [:name]
or
NOCOMMON:(name, •.• )

Without a common block name specified,
causes
all common blocks to be
allocated in
the data
psect.
Individual common blocks can be
placed
explicitly in
the data psect by putting their
names in a list after COMMON:.
COMMON is the
default.

NODATA

Specifies that all
non-common variables will
reside
in
.DATA.
This
is equivalent to
DATA:I073741823, which excludes all
variables
from .LARG.

PSECT[: [data psect]
Allows
users to
set the code and data
psect
[:[code psect]]] names explicitly (the large data psect is always
called
.LARG.)
If
PSECT is specified with one
argument, that argument becomes the name of the
small data area
psect.
Any second argument
becomes the name of the code psect.
This allows
separate
program
units to be put in separate
psects,
then the
psect can be
placed
in
different sections with the /SET switch at LINK
time (see the LINK Reference Manual).
NOTE
When using the PSECT argument, the
small data
psect
and code
psect
for
any given program unit must be
loaded into the same memory section.

16-20

USING THE FORTRAN COMPILER
16.5.4

Linking

~rith

TWOSEG

RE~L

Files

If a main program unit compiled with /EXTEND is linked with subprogram
units
(FORTRAN or not) that were compiled for non-extended use, then
LINK will automatically place the LOWSEG of non-extended units in the
.DATA. psect, and the HISEG of such units into the .CODE. psect.
A program compiled with /EXTEND can call a
subprogram that
is not
compiled with /EXTEND; however, it is illegal for a subprogram that is
not compiled with /EXTEND to call a subprogram that is.
Programs that were compiled by old versions of FORTRAN-IO/20 (prior to
Version 7) will not work if loaded in a non-zero section.
Most MACRO routines written
for
non-extended
use will
require
conversion to run in non-zero sections.
Data structures accessed with
IS-bit address fi.elds, indexed and indirect words, stack pointers and
some monitor calls may need modification
to perform correctly in
extended sections.
See the TOPS-20 Monitor Call User's Guide for more
information concerning extend(~d MACHO progra~

16.6

THE /FLAG (/FLAG-NON-STANDARD) SWITCH

The /Ii'LAG switch invokes the compatibility flagger.
NO"fE
For TOPS-20 systems,
the
full
switch names
are
/FLAG-NON-STANDARD and /NOFLAG-NON-STANDARD; however
/FLAG and /NOFLAG work.
This feature provides warning messages for language elements used that
are the following:
•

Extensions to the ANSI FORTRAN-77 standard

•

Features not found in VAX FORTRAN

•

Features that could ,::ause logically
used on the VAX FORTRAN system

different

results

when

No'rE

VAX FORTHAN is used

0:1

the VAX/VMS operating system.

This allows the flagging of a~y element that could cause conversion
problems
for
programs written on the TOPS-IO/20 system that might be
compiled and executed on a
VAX/VI~S
system or an ANSI-compatible
system.
This
includes
problems that could occur at object time, as
well as compilation incompatibilities.

16-21

USING THE FORTRAN COMPILER
By using the FLAG switch arguments
specify which features to flag.

listed

Table

16-8,

you

can

The form of the /FLAG switch is:
/ F LA G [ : a r g ]
or
/ FLAG [ : (a rg 1 ,a rg 2, ••• ) ]
Table 16-8:

Arguments to /FLAG Switch

Arguments

Meaning

ALL

Gives warning
messages
for
language
elements
incompatible with both FORTRAN-77 and VAX FORTRAN.

ANSI

Gives warning messages whenev~r a language element is
an extension to the FORTRAN-77 standard.

NOANSI

Does not flag FORTRAN-77 extensions.

NONE

Does not flag.

NOVMS

Does not flag VAX incompatibilities.

VMS

Gives warning messages whenever a language element is
incompatible with VAX FORTRAN.

If no /FLAG switch is specified, no flagging is done.
If no arguments
are given with the /FLAG switch, then flagging is done for both
FORTRAN-77 and VAX incompatibilities.
The /NOFLAG switch indicates that no flagging will be done.

16.7

READING A FORTRAN COMPILER LISTING

When you request a listing from the FORTRAN compiler, it may contain
the
following
information, depending on the switches used at
compilation time:
1.

A printout of the source program plus an internal sequence
number assigned to each line by the compiler.
This internal
sequence number
is referenced
in any error or war.ning
messages generated during the compilation.
If the input file
is line-sequenced, the number from the file is used.
If code
is added by means of the INCLUDE statement, all INCLUDEd
lines will have an asterisk
(*)
appended
to
their
line-sequence number.

A summary of the names and relative program locations
(in
octal)
of
scalars
and arrays
(including unreferenced
character scalars and arrays)
in the source program plus
compiler-generated variables.

16-22

USING THE FORTRAN COMPILER
3.

All COMMON blocks and the relative locations
the variables in each COMMON block.

(in

A listing of all equivalenced variables or arrays and
their
relative locations.
Note
that all equivalenced variables
that are also in COM~ON are listed only as being in COMMON.

A listing of the subprograms referenced
and FORTRAN-defined library functions).

A summary of temporary locations generated by the compiler.

A heading on each page of the listing containing the
program
unit name
(MAIN.,
.BLOCK, program, subroutine or function,
principal entry), the input filename, the list of compiler
switches, and the date and time of compilation.

If you used the /MACRO switch, a mnemonic
printout of the
generated code (in a format similar to MACRO) is appended to
the listing.
This section has four fields:

(both

octal)

user-defined

LINE:
This column contains the internal sequence number
of the
line corresponding
to
the mnemonic code.
It
appears on the
first
line of the
code
sequence
associated witt.
that
internal
sequence number.
An
asterisk indicates a compiler inserted line.
LOC:
The relative location in the object program of the
instruction.
LABEL:
Any program or compiler
generated
label.
Program labels have
the letter "pH appended.
Labels
generated by thE compiler are
followed
by the
letter
"M".
Labels generated by the compiler and associated
with the /DEBUG:LABELS switch consist of the
internal
sequence number followed by an "L".
GENERATED CODE:

The MACRO mnemonic code.

If you use the /LNMAP switch and do NOT use
the /MACRO
switch,
a
line number/octal location map is appended to the
listing.
This section lists the line numbers
In
increments
of 10 on subsequent lines and each number from 0 through 9
for each line in adjacent columns.
The numbers appearing
inside
the matrix are
the relative octal locations of the
statements in the FORTRAN program unit.
For example, to find the
relative octal
location of line
number 001043,
find the row marked 001040 and then column 3
on that line.
The number
in
that place
is
the desired
relative location.
This listing can be very large and sparse
for line-numbered files with large increments, such as those
produced by the edit.or SOS on TOPS-IO (or the editor EDIT on
TOPS-20) •
NOTE
A single FOR1'RAN line can produce mUltiple
machine
inst.ructions.
In this case the line
number map Ijsts only the first location.

16-23

USING THE FORTRAN COMPILER
9.

A list of all argument blocks generated by the compiler.
A
zero
argument appears first followed by argument blocks for
subroutine calls and function references (in order of
their
appearance
in
the
program).
Argument blocks for all I/O
operations follow this.

10.

FORMAT statement listings.

11.

A summary of errors
during compilations.

16.7.1

detected

warning

messages

issued

Compiler-Generated Variables

In certain situations
the
compiler generates
internal
variables.
Knowing
what
these variables
represent can help you read the macro
expansion.
The variables are of the form:
.letter digit digit digit digit
The function of these variables can be determined by the first
of the variable name as described below:
Letter

letter

Function of Variable

Compile-time constant character descriptor

Arithmetic statement function formal parameters.

Result of a
DO
LOOP
initial
value
expression
parameter of an adjustable dimensioned array.

Result of
a
computation.

Temporary storage for expression values.

Result of reduced operator strength expression.

Result of the DO LOOP step size expression of
iteration count for a loop.

common

sUbexpression

constant

computed

For example:

.SOOOI
You may find these variables on the listing under SCALARS and ARRAYS.
The following examples show listings where all of these
pointed out.

16-24

features

are

USING TEE FORTRAN COMPILER

Example 1:
Program
Name
MAiN.

Compiler Version

Source
Name

TI~l.FoR

FORTRAN V.l0(1601l)/F77/~

5-AUG-82

10:26

MACRO code listing included
00001
00002
00003
00004
00005
00006
00007
00008
00009
00010
00011
00012
00013
00014
00015
00016
00017
00018
00019
00020
00021

100
C
10

IMPLICIT INTEGER (A-Z)
DIMENSION A(100,200)
COMMON I~ ( 100 ,200)
oPEN(UNIT=22,FILE='TIM.oAT')
SUM1=0
SUM2=0
00 1 00 ~,= 1 ,200
00100]=11100
Kl=I*J
IF (K1 .Ll. 500 .0R. Kl ,GT. 1500) K1=0
A(I,J)=K1
K2=I+J
IF (K2 .EQ. 100 .0R. K2 .EQ. 200 .0R. K2 .EQ. 300) K2=K2+1
B(I,J)=K2
SUM1=SUM1+Kl
SUM2=SUM2+K2
CONTINUE
WRITE(22 110)SUM1 ,SUM2
FORMAT ( , SUM 1 = " 19, '
END

SUM2= ' 119)

COMMON BLOCKS
/. CoMM. / (+47040 ) ___ --Relative addres:;es of each. variable - - - - - - - - - - . . . . ,

*K 1
*SUM2

47045

~oIEXpLICIT

2
47046

"EFINITIONL "z"
"

47047

.SOOOl
*SUMl

47043
47050

.SOOOO 47044

Relative address of each character data descriptor
CHARACTER DATA
NAME
'TIM1.DAT'

U*U NO EXPLICIT DEFINITION]
OESeR I PToR ADDRESS
START OF DATA
ADDR ( POS I TI ON)
.HSCHD+O

4~054~

LENGTH
8 - # of characters

Relative address of first character of string

Character position of first character of string

16-25

USING THE FORTRAN COMPILER
Internal sequence number on first instruction
for this source Ii ne

LINE

Octal displacement of instruction

~OC

lABEL

GENERATED CODE

>(MOI.JE I
PUSHJ
SETZB
MOI.JEM
MOI.JE
HlREM

0,0
16,RESET.
0,0
16,2M
17,OPEN.
2,SUM1
2,SUM2
2,[777470000001]
2 , • SOOOO

HRRZM

2,J

MOI.JE

2.[777634000001]

MOI.JE
I MULI
MOI.JEM
CAll
CAllE
JRST
JRST

3,J
3,0 (2)
3 ,1\ 1
3,764
3,2734
0,7M
0,6M

SETZB

4 .1\1

MOI.JE I
IMUl
ADDI
MOI.JE
MOI.JEM
MOI.JE
ADDI
MOI.JEM
MOI.JE
CAIE
CAIN
JRST

Compiler generated label
3.144
3,J
3,0 (2)
4 .1\1
4.A-145(3)
3,J
3.0 (2)
3.1\2
5,1\2
5 .144
5.310
0.8M

CAIN

5.454

ADS

3.1\2

MOI.JE I
IMUL
ADDI
MOI.JE
MDI.JEM
ADDM
ADDM

3.144
3.J
3.0 (2)
5,1\2
5.B-145(3)
4,SUM1
5.SUM2
Program label
2,5M
2, I
2,J
o , . SOOOO
0,4M
16tl1M
17,DUT.
16 tl2M
17 , I DlST •
16 tl M
17,E>(IT.

JFCl
JSP

0
2

3
4
5
6
7
10
11

3M:

4M:

5M:

5
6
7

14
15
16
17
20
21
7M:

10
11

6M:

24
25
26
27
30
31
32
33
34
35
36
37

10M:

8M:

15
16
17

42
43
44
45
46
47
50

12
13

51
52
53
54
55
56
57
60
61
6'"'

lOOP: •
AOBJN
HRRZM
ADS
AOSGE
JRST
:-:MDI.JE I
PUSHJ
:-:MOI.JE I
PUSHJ
:-:MDI.JE I
PUSHJ

16-26

USING THE FORTRAN COMPILER
ARGUMnn BLOCKS : _ - - - Function, subnutine, and FOROTS argument blocks
63
64
65
66
67
70
71
72
73
74
75
76

0, ,0
0, ,0
777776, ,0
436000 , ,26

1 M:
2M:

4066tW, , • H~;CHD+O
777776, ,0
401000,,26
4023tW, dOll
777775 , ,0
401100, ,SUrll
401100, ,SUr12
4000, ,0

11 M :

12M:

FORMAT STATEMENTS (11\' lOW SEGMENT) :
20

47056
47057

(' SUM1= ',IB,'

lOP:

SUM2= 'dB)

No err 0 r s de t e c ted J .....o---------Summary of errors

MAIN.

Example 2:
MAIN.
00001
00002
00003
00004
00005
00006
00007
00008
OOOOB
00010
00011
00012
00013
00014
00015
00016
00017
00018
0001B
00020
00021

TIM1.FoR

100
C
10

FORTRAN I,J. 10 ( 16(4) IF77 I l

5-AUG-82

10:28

IMPLICIT INTEGER (A-Z)
DIMENSION A(100,200)
COMMON 5(100,200)
oPEN(UNIT=22 ,FIlE= 'TIM .DAT')
SUM1=0
SUM2=0
DO 100 _'=1,200
DO 100 1=1,100
Kl=I*J
IF (Kl .IT. 500 .0R. Kl
GT. 1500) Kl=O
A(I,J)=Kl
K2=I+J
IF (K2 .EO. 100 .0R. K2
EO. 200 .0R. K2 .EO. 3(0) K2=K2+1
B(I.J)=K2
SUM1=SUM1+Kl
SUM2=SUM2+K2
CONTINUE
WRITE(22 dO)SUMl ,SUM2
FoRMAT(' SUM1= 'dB,'
END

SUM2=' dB)

COMMON BLOCKS
I.CoMM./(+47040)
B
+0
SCALAR!:; AND ARRAYS

"*" NO

E>~PLICIT

CEFINITION - "'X," NOT REFERENCED

47043
47050

.SOOOO

47044

Requested with ILNMAP

LI NE NUMBER 10CTAl loCATI ON MAP_.---

: 0

.SOOOl
*SUMl

3
47047

47046

47045

------:---------------------------- .. -----------------------------------------00000
00010
00020
MAIN.

30
23
61
[ No errors detected
16

3
41

5
46

6
47

7
50

13
55

Line #11 starts at octal offset 23 (from the prevous example, note that line 11 uses locations 23 through 27, but
only the first location is shown here).

16-2:7

USING THE FORTRAN COMPILER
Example 3:
MAIN.
00001
00002
00003
00004
00005
00006
00007
00008
00008
00010
00011
00012
00013
00014
00015
00016
00017
00018
00018
00020
00021

FORTRAN t,! .10 (1604) IF77/0PT 1M

TIM1.FOR

5-AUG-82

10:30

IMPLICIT INTEGER (A-Z)
DIMENSION A(100,200)
COMMON B(100,200)
OPEN(UNIT=22,FILE='TIM1.DAT')
SUM1=0
SUM2=0
DO 100 J= 1 ,200
DO 100 1=1,100
K1=I*J
IF (K1 .LT. 500 .OR. K1 .GT. 15(0) K1=0

100

K2=I+J
IF (K2 .EO. 100 .OR. K2 .EO. 200 .OR. K2 .EO. 3(0) K2=K2+1
B(Id)=K2
SUM1=SUM1+K1
SUM2=SUM2+K2
CONTINUE

WRITE(22 dO)SUM1,SUM2
FORMAT(' SUM1= ' d8,'
END

SUM2= ' d8)

COMMON BLOCK S
I.COMM./(+47040)
B
+0
SCALARS AND ARRAYS

"*" NO E:<PLICIT DEFINITION - "'X," NOT REFERENCED]

,.------Optimizer created variables - - - - - - - - ,
*K1
• SOOO 1 470t!5
*K2
47052

.R0001
.SOOOl
*SUM1

2
47046
47053

.ROJoo 3
*SUM2
47047

CHARACTER DATA [ "*" NO D(PLICIT DEFINITION
NAME
DESCR I PTOR ADDRESS
START OF DATA
AD DR ( POS IT I ON)
'TIM1.DAT'

.HSCHD+O

47060(1)

16-28

*J
*1

5
4
A
47050 .00001
47051

LENGTH
8

USING THE FORTRAN COMPILER
LINE

LOC

LABEL

GENERATED (ODE
JFCL
JSP

2
LI
6

*
7

3
LI
5
6
7
10
11
12
13

ill

9
10

16
17
20
21

)<MOI.JE I
PUSH~J

8M:

7M:

26
27
30
31
32
33

11 M:

3L1

10M:

ill

9M:

15
16

36
37
L10
L11

*
*

*
*
*
*19
*
21

L12
L13
LIL1
L15
L16
L17
50
51
52
53
5L1
55
56
57
60
61
62

7,1
1 ~ t • SOOOO

MOI.JE
MOI.JE

2 t [77763L100000 1 ]

Asterisks indicate opt.imizer
generated statements

1~t310

6t7

6M:

*17

SETZB
MOI.JEI
MOI.JEM
MOI.JN I
MOI.JE I
MOI.JEM
5M:

23
2L1

OtO
1 '3 tRESET •
o to
li3tLlM
17tOPEN.
1 ,) t 11
1~ ,1 LIL1
1.~ t • ROOO 1

MOI.IE I
ADD
MOI.IE
CAlL
CAlLE
JRST
JRST

L1t(l(2)
L1t.ROOOl

MOI.JE I

5tO

MOI.JH1
MOI.JE
ADDI
CAIE
CAIN
JRST

5tA-1L15(LI)
3 t7
3 to (2)
3 t 1 LIL1
3 t31 0
o t 10M

CAIN

3tLl5L1

ADDI

3,1

MOI.JEM
ADD
ADD
ADD

3tB-1L15(4)
11 t 5
1) t 3
6 t7

AOBJN
HRRZM
MOI.JE I
ADDM

2t6M
1 ~ ,1 LIL1
1 ~ ,1L1L1
1 ~ t • ROOO 1

ADDI
AOSGE
JRST
MOI.IEM
MOI.IEM
MOI.IEM
MOI.JEM
MOI.IEM
)-<MOt.JE I
PUSHJ
)-<MOI.IE I
PUSHJ

7,1
Ot.SOOOO
Ot5M
7 tJ
I I tSUMl
1)tSUM2
5 tK1
3tK2
13 t 12M
17tOUT.
1,3,13M
17tIDLST.

)<MOI.JE I
PUSH,J

13t3M
17 t D<IT.

5t6
5t76L1
5t273L1
Ot8M
Ot7M

100 P :

1 M:

2M:

16-29

USING THE FORTRAN COMPILER

ARGUMENT BLOCKS:
63
64
65
66
67
70
71
72
73
74
75
76

0, ,0
0,,0

3M:
4M:

12M:

13M:

777776, ,0
436000, ,26
406640, ,.HSCHD+O
777776, ,0
401000 , ,26
402340,,10P
777775, ,0
401100"SUM1
401100"SUM2
4000, ,0

FORMAT STATEMENTS (I N LOW SEGMENT) :
20

MAIN.

47062
47063

lOP:

[ No err a r s

6
(' SUM1= ',18,'

SUM2= ',18)

oj e te c t e oj

16-30

USING THE FORTRAN COMPILER
Example 4:
MAIN.
00001
00002
00003
00004
00005
00008
00007
00008
00008
00010
00011
00012
00013
00014
00015
00018
00017
00018
00018
00020
00021

TIM1.FoR

100
C
10

COMMON BLOCKS

FORTRAN I,J.1Cl(253C )/F77/M/E>:T

24-APR-85

18:27

IMPLICIT INTEGER (A-Z)
DIMENSICN A( 100 ,2(0)
COMMON B( 100 ,200)
oPEN(UNIT=22,FILE='TIM1.DAT')
SUM1=0
SUM2=0
DO 100 J= 1 ,200
001001=1,100
K 1 =I*J
IF (Kl .LT. 500 .0R. Kl .GT. 15(0) Kl=O
A(I,J)=kl
K2=I+J
IF (K2 .EO. 100 .0R. K2 .E:). 200 .(JR. K2 .EO. 300) K2=K2+1
B<I,J)=k2
SUM1=SUl"ll+Kl
SUM2 = SUI"12+K 2
CONTINUE
WRITE (2:'::: dO)SUMl ,SUM2
FORMAT ( , SUM 1 = ' d ~~ , '
END

SUM2= 'd8)

[ " I " STORED IN • LARG.

I. CoMM. / (+47040 ! )
+0
tL._________ Large com mon block

SCALARS AND ARRAYS

"*" NO E>: PLI C IT [JEF I N IT I ON - "'X," NOT REFERENCED
II ! II
l,.l A R I A B L EST 0 r~ E 0 IN. LA R G t
]

•
*K
.SOOOl
*1

*J
2
.SOOOO 4
*K2
7

]

Large variable

!A
*SUM2
*SUMl

0
5
10

CHARACT!::R DATA [ "*" NO E><PLICIT DEFINITION - "!" I,JARIABLE STORED IN .LARG.
DESCR 1 PTOR ADDRESS
START OF DATA
LENGTH
NAME
ADDR ( PoS IT I ON)
'TIM1.DAT'

.HSCH[)+O

LINE

LABEL

LoC

5
8
7

><Mol,JE I

4M:

5M:

0,0
18,RESET.
0,0

i8,2M
PUSHJ
SETZB
Mol,JEM
Mol)E
HLREM

17,oPEN.
2,SUMl
2,SUM2
2,[777470000001]
2 , • SOOOO

HRRZM

Mol,JE

2,[777834000001]

Mol)E
I MULI
Mol,JEM
CAlL
CAlLE
JRST
JRST

3d
3,0 (2)
3 ,K 1
3,784
3,27:::4
0,7M
0,8M

SETZB

4 ,K 1

3M:

GENERATED COOL
JFCL
JSP

2
3
4
5
8
7
10
11

14 ( 1 )

14
15
18
17
20
21
7M:

16-31

]

USING THE FORTRAN COMPILER
11

6M:

24
25
26
27
30
31
32
33
34
35
36
37

10M:

8M:

15
16
17

42
43
44
45
46
47
50

MoVEI
IMUL
ADDI
MOVE
Mol,JEM
Mol,JE
AoDI
Mol.'EM
Mol.'E
CAIE
CAIN
JRST

3.t44
3d
3,0 (2)
4,Kl
4 ,@[ .EFIW A-145(3)]
3,J

CAIN

5,454

ADS

3,K2

Mol,JE I
IMUL
ADDI
Mol.'E
Mol.'EM
ADDM
AD OM

3,144
3,J
3,0 (2)
5,K2
5 ,@[ .EFIW B-145(3)]
4,SUMl
5,SUM2

3,(H2)

3,K2
5,K2
5,144
5,310
0,8M

lOOP:
AoBJN
2,5M
HRRZM
2,1
ADS
2,J
AoSGE
O,.SOOOO
JRST
(l,4M
){Mol,JEI16.tlM
PUSHJ
17,oUT.
XMol.'EI16,12M
PUSHJ
17,IOLST.
){MOI.'E I
16 t1 M
17,E){IT.
PUSHJ

51
52
53
54
55
56
57
60
61
62

ARGUMENT BLOCKS:
63
64
65
66
67
70
71
72
73
74
75
76

0, ,0
1 M:

2M:

11M:

12M:

0,,0
777776 , ,0
436100,,[000000000026]
1I06GlIO, , • HSCHD+O
777776, ,0
1I01000, ,26
1I02340, ,lOP
777775, ,0
401100, ,SUMI
401100, ,SUM2
1I000, ,0

FORMAT STATEMENTS (I N LOW SEGMENT) :
20
MAIN.

17
20

lOP:

(

SUM 1 =

tI 8, I

SUM2=

[ No errors detected]

00001
00002
00003

16-32

tI8)

USING THE FORTRAN COMPILER
NOTE
Note that in the scalars and arrays list, 'NO EXPLICIT
DEFINITION'
indicates that the variable was never
explicitly defined,
as
in a TYPE
or
DIMENSION
statement.
Also, 'NOT REFERENCED' indicates that the
variable was declared, but never used, and therefore
was never allocated any storage in the program.
Character variables that are declared,
but
not
referenced,
appear
under the scalars and arrays
section of the listing.
No storage is allocated
for
either the character descriptor or the character data.

16.8

ERROR REPORTING

If an error occurs during the initial pass of the compiler (while the
actual source code is being read and processed), an error message is
printed on the listing immediately following the line
in which the
error occurred. When pertinent and possible, the error references the
internal sequence number of the incorrect line.
The error messages
along with the statement in error are output to the user terminal.
Example:
TYPE DAY. FOR
01000
01100
01200
01300
01400
01500
100
01600
C
01700
200
01800
01900

1=10
IMPLICIT INTEGER (X)
J=I**4
Kl
X=I+J+Kl
CONTNUE
TYPE 200,X
FORMAT(lX,I8)
END

COMPILE DAY.FOR
FORTRAN: DAY
IMPLICI,]' INTEGER (X)
01100
IMPL]CIT statement out of order
%FTNSOD LINE:OI100
01300
Kl
?FTNNRC LINE:01300 Statement not recognized
100
CONTNUE
01500
?FTNMSP LINE:01500 Statement name misspelled
01600
?
?FTNICL LINE:01600 Illegal character C in label field
?FTNFTL
If errors
appear in
output to
reference

MAIN.

3 fatal errors and 1 warning

are detected after the initial pass of the compiler,
they
the list file after the end of the source listing.
They are
your terminal without the statement in error, but they may
its internal sequence number.

16-33

USING THE FORTRAN COMPILER
16.8.1

Fatal Errors and Warning Messages

There are two levels of messages, warning and
fatal
error.
Warning
messages are preceded
by "%" and indicate a possible problem.
The
compilation will continue, and the object
program may be correct.
Fatal
errors are preceded by a "?".
If a fatal error is encountered
in any pass of the compiler, the remaining passes will not be called,
and no relocatable binary file will be generated.
Additional errors that would be detected in later compiler passes may
not become apparent until the first errors are corrected.
It is not
possible to generate a correct object program
for
a
source
program
containing a fatal error.
The format of messages is:
?FTNxxx Line:n text
or
%FTNxxx Line:n text
where:

?
%
FTN
xxx
Line:n
text

indicates a fatal message
indicates a warning message
is the FORTRAN mnemonic
is the 3-letter mnemonic for the error message
is the optional line number where error occurred
is the explanation of error

The printing of fatal errors and warning messages on your terminal can
be suppressed by the use of the /NOERRORS switch; however, messages
will still appear on the listing.
The /NOWARN switch will
suppress
warning messages on both the
user
terminal
and
in the listing.
Specific warnings can be suppressed by using options
to
the /NOWARN
switch (see Section 16.4).

16.8.2

Message Summary

At the end of the listing file and on the terminal, a message
summary
is printed after each program unit is compiled.
This message has two
forms:

When one or more messages were issued
?FTNFTL
%FTNWRN

name
name
name
name

1 fatal error and no warnings
2 fatal errors and no warnings
no fatal errors and 1 warning
no fatal errors and 2 warnings

or
2.

When no messages were issued
name

[No errors detected]

where name is the program or subprogram name.
Appendix C
complete list of fatal errors and warning messages.

16-34

contains

USING THE FORTRAN COMPILER
16.9

CREATING A SHARABLE HIGH SEGMENT FOR A FORTRAN PROGRAM

For non-extended addressing
programs,
the FORTRAN compiler always
generates two segment code for a program unit.
However, by default,
the linking loader loads all FORTRAN code
into the low segment to
allow the sharable run-time system to be bound to the program at
run-time rather than at load-time.
This default action of the loader can be overridden by using the LINK
switch /OTS:NONSHAR when loading
the program.
If this switch is
given, the linking loader loads the impure code (the data areas)
in
the low segment, loads the pure code (the machine instructions) in the
high segment, and binds a private copy of the run-time system to
the
program at load-time.
A program loaded with the /OTS:NONSHAR switch can be saved in order to
produce an executable
file
with a sharable high segment using the
TOPS-IO SSAVE command or the TOPS-20 EXEC SAVE command.
(The
LINK
switches /SSAVE
(TOPS-lO)
or /SAVE
(TOPS-20)
can also be used to
produce the sharable executable file.) This is an advantage if a large
application program is to be run by several users simultaneously.
The
code unique to the program and the copy of the FORTRAN run-time system
that
is private to the program is shared between all the program's
users.
The possible benefits gained by the users of the application program
sharing
the high segment of their program containing both user and
FOROTS code must be weighed against the loss of not sharing the common
copy of FOROTS as all other users.
(See the LINK Reference Manual.)

16-35

CHAPTER 17
USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)

FORDDT is an interactive program used to debug FORTRAN programs and
control their execution.
By using the symbols created by the FORTRAN
compiler, FORDDT allows you to examine and modify the data and FORMAT
statements
in your program, set breakpoints at any executable
statement or routine, trace your program statement-by-statement, and
make use of many other debugging techniques described in this chapter.
Table 17-1 lists all the commands available to the user of FORDDT.
Table 17-1:

FORDDT Commands

Command

Purpose

Control Commands

CONTINUE

Continues
execution
breakFoint.

DDT

Enters DDT.

GOTO

Transfers control to some program statement
within the open program unit.

Traces execution of the program.

START

Begins execution of the FORTRAN program.

STOP

Terminates the
monitor mode.

after

program

and

FOR DDT

returns

Data Access Commands
ACCEPT

Modifies variables or FORMAT statements.

TYPE

Displays variables.

CHARACTER

Defines dimensions of character arrays for
FORDDT
references.
(This
command
is
unnecessary if /DEBUG is specified
at
compile time.
See Table 16-3.)

17-1

Declarative Commands

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
Table 17-1:

FORDDT Commands (Cont.,

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

Command

Purpose

~------------------------~-----------------------------------------------,------

DIMENSION

Defines dimensions of real
and
integer
arrays
for
FORDDT
references.
(This
command
is
unnecessary
if
/DEBUG
is
specified
at compile
time.
See Table
16-3.)

DOUBLE

Defines dimensions of double-precision and
complex
arrays
for
FORDDT references.
(This command is unnecessary if /DEBUG
is
specified
at compile
time.
See
Table
16-3.)

GROUP

Defines indirect lists for TYPE statements.

MODE

Specifies format of typeout.

OPEN

Accesses program unit symbol table.

PAUSE

Sets FORDDT breakpoints.

PAUSE ON ERROR

Sets FORDDT breakpoints (for errors such as
arithmetic overflows).

REMOVE

Clears FORDDT breakpoints.

REMOVE ON ERROR

Clears PAUSE ON ERROR breakpoints.

Other Commands
LOCATE

Lists program unit names in which
symbol is defined.

STRACE

Displays
routine
program status.

WHAT

Displays current DIMENSION,
GROUP,
FORDDT breakpoint information.

traceback

given
current
and

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

The FORDDT commands are described in detail in Section 17.5.

17.1

INPUT FORMAT

FORDDT commands consist of alphabetic
FORTRAN-like
identifiers and
need consist of only those characters required to make the command
unique.
If you wish
to
specify parameters,
a
space
is
required
following the command name.
Comments may be appended to command lines
by preceding the comment with an exclamation point (!).

17.1.1

Variables and Arrays

FORDDT allows you to access and modify the data in your program using
standard
FORTRAN symbolic names.
Variables are specified simply by
name.
For example:
name
17-2

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
where:
name is a variable name.
E~lernents

Array

are specified in the following formats:

name
n am e ( s 1 , ••• s n)
j'

where:
name

is the name of the array

(sl, ••. ,sn)

are the
sub:5cripts of a
particular
array.
The
subscripts mllst be integer constants or variables.

You may reference an entire ar~ay simply by typing
the
array name
(without subscripts).
You may specify a range of array elements by
typing the first and last element in the chosen range, separated by a
dash (-).
The following examples show
and arrays to FORDDT:

thl~

various ways of

specifying

variables

ALPHA
ALPHA(7)
ALPHA(PI)
ALPHA(2)-ALPHA(5)

17.1.2

Constant Conventions

FORDDT accepts optionally signed numeric data in the standard
input formats:

FORTRAN

INTEGER - A string of decimal digits.

REAL - A string of decimal digits optionally
including
a
decimal
point.
Standard engineering
and double-precision
exponent formats are ~lso accepted.

OCTAL - }I.,. string of octal d:igits
double quote (").

COMPLEX - An ordered
pair of
integer or
real
separated by a comma and enclosed in parentheses.

LOGICAL - A Boolean argument, either .TRUE.

CHARACTER - A string of printable ASCII
by apostrophes.

HOLLERITH
A stri~g
of
alphanumeric
and/or
special
characters
delimited
by
any
alphanumeric or special
character, excluding the
space character,
which does not
occur
with the strin'J itself.
Such as, # 12AB#, where # is
the delimiting character.

17-3

optionally

preceded

constants

or .FALSE.

characters

enclosed

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
17.1.3

Statement Labels and Source Line Numbers

FORTRAN statement labels are
input and output by straightforward
numeric reference, such as 1234.
However, source line numbers must be
input to FORDDT with a number sign (#) preceding them.
This mandatory
sign distinguishes statement labels from source line numbers.

17.2

PAUSE #3

!This causes a pause at source line number 3.

PAUSE 3

!This causes a pause at the statement labeled 3.

FORDDT AND THE FORTRAN /DEBUG SWITCH

Most facilities of FORDDT are
features.
However, if you do
a FORTRAN program, the trace
available, and several of the

available without
the
FORTRAN /DEBUG
not use the /DEBUG switch when compiling
features
(NEXT command)
will
not be
other commands will be restricted.

Using the /DEBUG switch tells FORTRAN to compile extra information for
debugging.
(See Chapter 16 for more information.) These features are:
1.

/DEBUG:DIMENSIONS, which generates dimension
information
in
the
.REL file for all arrays dimensioned in the subprogram.
The dimension
information
is automatically available
to
FORDDT if you wish to reference an array in a TYPE or ACCEPT
command.
This
feature
eliminates the need
to
specify
dimension
information for
FOR DDT by using
the DIMENSION
command.

/DEBUG:LABELS, which generates labels
for
every executable
source line in the form <line-number>L.
If these labels are
generated, they may be used as arguments with
the
FORDDT
commands PAUSE and GOTO.
This switch also generates labels at
the
last
location
allocated
for
a
FORMAT statement so that FORDDT can detect
the end of the
statement.
These
labels have
the
form
<format-label>F.
If they are generated, you can display and
modify FORMAT statements by means of the TYPE and
ACCEPT
commands.
Note that the :LABELS switch is automatically activated with
the
:TRACE switch, since labels are needed to accomplish the
trace features.

/DEBUG:TRACE, which generates a reference
to
FORDDT before
each executable statement.
This switch is required for the
trace command NEXT to function.
Note that if more than one FORTRAN statement is placed
on a
single
input
line,
only the
first statement has a FORDDT
reference and line-number label
associated with
it.
This
also applies to the :LABELS switch.

/DEBUG:INDEX, which forces
the compiler
to store
in
its
respective data
location,
as well as a register, the index
variable of all DO loops at
the beginning
of each loop
iteration.
You will
then be
able to examine DO loops by
using FORDDT.
If you modify a DO loop index using FORDDT, it
will
not affect
the number of loop iterations because a
separate loop count is used.

17-4

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
In addition, this switch forces all
function values
to be
stored
in memory prior to return from the function.
If this
switch is specified, you can set a FORDDT pause on the RETURN
statement
(see Section 13.4.4) and then examine the value to
be returned.

17.3

/DEBUG:BOUNDS, which generates the bounds checking
code
for
all
arralY references.
Bounds violations produce run-time
e r ro r
messag es.
No te
tha t
the
techn ique 0 f
spec i fyi ng
dimensions of
1
fOl"
subroutine arrays causes bounds check
errors.
(You can use this option without FORDDT.)

/DEBUG:ARGUMENTS, which performs type checking at load
time
for
actual
argument
types and
associated dummy argument
types.
Type violations produce non-fatal,
load-time error
messages.
This swj,tch also
performs type checking
at
compile-time for statement functions.

LOADING AND STARTING FORDDT
1.

On TOPS-10, the simplest method of debugging with FORDDT is:
.DEBUG filespec{DEBUG)
On TOPS-20, the

corre~;ponding

command is:

@DEBUG filespec /DEBUG
On both systems, FORDDT responds with:
STARTING FORTRAN DDT
Program name:
When FOR DDT prompts you for a program name,
type
the same
name specified in th~ PROGRAM statement of the program being
debugged.
If the
PROGRAM sta temen t i s not
used
in
the
program being debugged,
FORDDT uses MAIN.,
and will not
prompt for a program name.
FORDDT next prints its command prompt:

»
The angle brackets indicate that FORDDT is ready to receive a
command.
2.

If you are on TOPS-20, you can type a question mark
to
prompt to get a list of all FORDDT commands, as follows:
»? One
ACCEPT
GOTO
OPEN
TYPE

of the following:
CHARACTER
CONTINUE
GROUP
HELP
REMOVE
PAUSE
WHAT

DDT
LOCATE
START

Also on 'I'OPS-20, you can use· the ESCape key
of FORDD'I' commands.
For example:
»CON<ESC>TINUE

17-5

the

DIMENSION DOUBLE
MODE
NEXT
STOP
STRACE
for

recognition

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
On both systems, you need only type the
of a specific FOR DDT command.

unique

abbreviation

program and
FOR DDT
You may wish
to
load
your compiled
directly with
the linking
loader.
(Loading with LINK is
The
accomplished implicitly in the
DEBUG command string.)
command sequence is as follows:
On TOPS-la, to start LINK, type:
.R LINK
On TOPS-20, type:
@LINK
On both systems, when LINK prompts you with an asterisk,
can type a command string in any of the following forms:
*filespec /DEB/G

(loads DDT)

*filespec /DEB:{FORDDT}/G
{FORTRA}

(loads FORDDT)

*filespec /DEB: (DDT,{FORDDT})/G
{FORTRA}

(loads DDT
and FORDDT)

*filespec /DEB: ({FORDDT} ,DDT)/G
{FORTRA}

(loads FORDDT
and DDT)

you

In the last two command forms
shown,
the
first debugging
program specified
(FORDDT or DDT) in the command string is
the one you communicate with after the LINK command string
is executed.
See Section 17.9
for
addressing programs.

17.4

information

extended

SCOPE OF NAME AND LABEL REFERENCES

Each program unit has its own symbol table.
When you initially enter
FORDDT,
you automatically open the symbol table of the main program.
All references to names or labels through FOR DDT must be made with
respect to the currently open symbol table.
If you have given the main program a name other than MAIN.
by using
the PROGRAM statement (see Section 6.4.1), FORDDT asks for the defined
program name.
After you enter the
program name,
FORDDT opens the
appropriate symbol
table.
At this point, symbol tables in programs
other than the main program can be opened by using the OPEN command.
References to statement labels,
line numbers,
FORMAT statements,
variables,
and arrays must have labels that are defined in the
currently open symbol table.
However, FORDDT will accept variable and
array references outside
the currently open symbol table, providing
the name is unique with respect to all program units in the given load
module.

17-6

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
17.5

FORDDT COMMANDS

This section gives a detailed description of all commands
The commands are given in alphabetical order:
ACCEPT

FORDDT.

Allows
you to change
the contents of a
FORTRAN
variable, array, array element, array element range, or
FORMAT statement.
The command format is:
ACCEPT namE'[/mode] value
where:
name

is the variable, array, array element,
range,
or
FORMAT statement
modified.
eleme~t

array
to be

mode

is
the
format of
the data value
to
be
entered.
The mode keyword must be preceded
by a slash (/) and
immediately follows
the
name.
Intervening
blanks are not allowed.
(Note that /mode does not apply to
FORMAT
modification.)

value

is the new value to be assigned.
The
format
of the
input value must correspond to the
specified mode.
DATA LOCATION MODIFICATION
Data Modes

The following data modes are accepted:
Mode
A
C
D
F
I
0

L
R
S
X

Meaning

Example

ASCII (left-justified)
CHARACTEF(
DOUBLE-PF:EC IS ION
REAL
INTEGER
OCTAL
LOGICAL
RASCII (right-justified)
SYMBOLIC
COMPLEX

/FOO/
'ABC'
123.4567890
123.45678
1234567890
7654321
• TRUE. or .FALSE •
\BAR\
PS I ( 2,4)
(1.25,-78.E+9)

If not specifie6, the default mode is REAL (F).
1.

Two-Word Va]ues
For the data modes
(R),
and SYMBOLIC
modifier on the
indicates that the
interpreted as two

17-7

ASCII
(A),
OCTAL
(0),
RASCII
(S), FORDDT will accept a "/BIG"
mode switch.
This
modifier
variable and the value are to be
words long.

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
Example:
ACCEPT VAR/RASCII/BIG '1234567890'
assumes that VAR is two words long and
given 10-character literal into it.

stores

the

The /BIG modifier can also be used to display more
than
the
first
256
characters of long character
strings.
2.

Character Variables
A character variable can be initialized by using an
ACCEPT command of the following form:
ACCEPT VAR/C 'string'
Note that length fo the variable is that which
is
specified
in the source program.
If the string is
longer than the variable, the rightmost characters
are
truncated.
If the string is shorter than the
variable, it is stored left-justified and padded on
the right with blanks.

Initialization of Arrays
If
the
name
field
of an ACCEPT contains
an
unsubscripted
array name or a
range of array
elements,
all
elements of the array
or
the
specified range are set to the given value.
Example:
ACCEPT ARRAY/F 1.0
or
ACCEPT ARRAY(5)-ARRAY{10)/F 1.0
Note that this applies only
ASCII and RASCII.

modes

other

than

For
character
arrays,
if the value has
fewer
characters than
the
length of the array element,
the rightmost character positions of the element
are initialized with spaces.
If the value has more
characters than the length of the array element,
the value is truncated to the right.
4.

Long Literals
When the value
field
of an ACCEPT contains an
unsubscripted
array
name
or
range of array
elements, and the specified data mode is ASCII, the
value
field
is expected to contain a long literal
string.
ACCEPT stores the string linearly into the
array or array range.
If the array is not filled,
the remainder of the array or range is filled
with
zeroes.
If the literal is too long, the remaining
characters are ignored.

17-8

USING

~rHE

FORTRAN

INTERAC~rIVE

DEBUGGER (FORDDT)

Example:
ACCEPT ARRAY/ASCII
'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
5.

FORMAT

Stat(~ment

Modification

When the name field of an ACCEPT contains a
label,
FORDDT expt~cts this label to be a FORMAT statement
label and that
the value
field
contains a new
format specification.
Example:
ACCEPT 10 (lHO,FIO.2,3(I2»
The new specification cannot be longer than
the
space originally allocated
to
the FORMAT by the
compiler.
'rhe remainder of the area is cleared
if
the new specification is shorter.
Note that FOROTS performs some encoding of FORMAT
statements when
it processes them for the first
time.
If any I/O statement referencing
the given
FORMAT has been executed, the FORTRAN program has
to be restal~ted (re-initializing FOROTS).
CHARACTER

character array.
The
Defines the dimensions of a
result of this command is the same as for the DIMENSION
the
array so dimensioned
is
command
except that
understood
by FORDDT to be a character array.
The
command fo rma t .~ s :
CHARACTER arrayname ([Ll:]Ul[,[L2:]U2, ••. ])
NOTE
This declal-ator cannot be used
to
specify
element length.
The length specified in the
user progran will be used.

CONTINUE

Allows the progl-am to resume execution after a
FORDDT
pause.
After a
CONTINUE
is executed,
the program
either runs to completion or
until
another
pause
is
encountered.
The command format is:
CONTINUE [n]
where the n is optional and, if omitted, is assumed
to
be 0 n e •
If
a val u e i s pro v ide d, i t can be a n urn e ric
constant or proqram variable, but it is treated
as an
integer.
When
the value n is specified, the program
continues execution until the nth occurrence of this
pause.
For exarlple,
CONTINUE 2()
continues execution until the 20th occurrence of
pause, or until a different pause is encountered.

. DDT

the

Transfers control of the program to DDT,
the
standard
system debugging
program.
Any files currently opened
by FOROTS are unaffected, and a
return
to
FORDDT
is
possible so that program execution may be resumed.

17-9

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
%FDDT is the global symbol used to
FORDDT.
The command format is:

return

control

%FDDT<ESC)G
Your program will be in the same condition as before
unless you have modified your core image with DDT.
DIMENSION

Sets, displays, or removes the user-defined dimensions
of
an
array for
FORDDT access purposes.
These
dimensions need not agree with those declared
to
the
compiler
in the source code.
FORDDT allows you to
redimension an array to have a larger scope
than
that
of the source program.
If this is done, a warning is
given.
NOTE
The DIMENSION command cannot be used to declare
double-precision,
complex, or character arrays
(see the CHARACTER and DOUBLE commands) •
The command format is:
DIMENSION name ([Ll:]Ul[,[L2:]U2, .•• ])
where:
name

is the name of the array

([Ll:]Ul .•. )

specifies the bounds of the
array,
where
L is the lower
bound and U is the
upper
bound.
The default value of L
is
1.
The bounds must be
integer
constants
or
variables.

For example:
DIMENSION ALPHA(7,5:6,10)
FORDDT remembers the dimensions of the array until they
are redefined or removed.
The command:
DIMENSION
gives a full list of all
for all arrays.

the

user-defined

dimensions

DIMENSION ALPHA
displays the current information for
only.

the

array

ALPHA

DIMENSION ALPHA/REMOVE
removes any user-defined
array ALPHA.

17-10

array

information

for

the

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
DOUBLE

Defines the dimensions of a double-precision or complex
array.
The
result of this command is the same as for
the DIMENSION command
except that
the
array
so
dimensioned is understood by FORDDT to be an array with
two-word elements.
The command format

1" c· •
~)

DOUBLE arrayname ([Ll:]Ul[,[L2:]U2, ••• ])
GOTO

Allows you to continue your program from a point other
than
the one at which it last paused.
The GOTO allows
you to continue at a statement label or
source line
number provided that the IDEBUG:LABELS switch has been
used or the contents of a
symbol
previously ASSIGNed
during the program execution has been used.
Note that the program must be STARTed before this
command can be used, and also note that a GOTO is not
allowed after
the <CTRL/C>,
REENTER sequence
(see
Section 17.6).
The command format

•
1" C"
~

GOTO n
GROUP

SE'ts up a str ing of text for input to a
TYPE command.
You can store TYPE statements as a list of variables
identified by the numbers 1 through 8.
This feature
el iminates
the neE!d
to
retype
the same 1 ist of
variables each time you wish to examine the same group.
Refer
to the ~'YPE command for the proper format of the
list.
The command format is:
GROUP [n list]
where:
n

is the group number 1-8

list

is a list of group numbers preceded by a
slash
(I)
and/or variable names to be
typed when you give the command:
TYPE
In, where n is the group number. The
validity of the list is not checked.

GROUP
with no arguments causes FOR DDT to type out the current
contents of all the grroups.
GROUP n
types out
requested.

thE!

contents

the

particular

Note that one group may refer to another.
Fa r exampl e :
GROUP 2 VAR2,VAH3
GROUP 3 I ~'. ,/2
17-11

group

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
LOCATE

Lists the program unit names in which a given symbol is
defined.
This is useful when the variable you wish to
locate is not in the currently open program unit and is
defined
in more
than one program unit.
The command
format is:
LOCATE n
where n may be any FORTRAN variable, array, label, line
number, or FORMAT statement number.

MODE

Defines the display format for succeeding
FORDDT TYPE
commands.
You need
type only the first character of
the mode
to
identify it
to
FOR DDT
because
all
characters after the first are ignored.
The modes are:
Mode
C
F
D
X
I

o
A
R
L

Meaning
CHARACTER
REAL
DOUBLE-PRECISION
COMPLEX
INTEGER
OCTAL
ASCII (left-justified)
RASCII (right-justified)
LOGICAL

Unless the MODE command is given, the
mode is the REAL (F) format.

default

typeout

The command format is:
MODE list
where list
separated
changed by
one mode
order:
C,

contains one or more of the mode identifiers
by commas.
The current setting
can be
issuing another MODE command.
If more
than
is given,
the values are typed out in the
F,

0, A,

A typical command string might be:
MODE A,I,OCTAL
NEXT

Allows you to
cause
FOR DDT to
trace source lines,
statement
labels,
and
entry point names during
execution of your program.
This command only provides
trace
facilities
if the program is compiled with the
FORTRAN /DEBUG switch.
If this switch is not used, the
NEXT command
acts as a CONTINUE command.
The command
format is:
NEXT [n] [/sw]
where:
n

is a program variable or integer numeric
value

is one of the following switches
/S= statement label
/L= source line
/E= entry point

17-12

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
The default starting value of n
is
1,
statement trace.
The default switch is /L.

single

The command
NEXT 20/L
traces the execution of the next 20 source line numbers
or until another pause is encountered.
Note
that
if no argument
argument given is used.

specified,

the

last

For example:
NEXT /E
changes the
tracing mode
to
trace only subprogram
entries using the numeric argument previously supplied.
OPEN

Allows you to open a particular program unit of the
loaded
prograrr so that the variables are accessible to
FORDDT.
Any previously opened program unit
is closed
automatically when a new one is opened.
Only global
symbols, symbols in the currently open unit, and unique
locals are available at anyone time.
Note that
starting FORDDT automatically opens the main program.
The command format is:
OPEN name
where name
is the subprogram name.
arguments reopens the main program.

PAUSE

OPEN

with

Allows you to place a FORDDT breakpoint at a
statement
number,
source line number, or subroutine entry point.
Up to ten breakpoints may be set at anyone time. When
a
breakpoint is encountered, execution is suspended at
that point and control
is
returned
to
FORDDT.
The
symbol
table of that subprogram is also automatically
opened.
The command formats are:
PAUSE
PAUSE
PAUSE
PAUSE
PAUSE
PAUSE
PAUSE

P
P
p
p
P
P

AFTER n
IF condition
TYPING /g
AFTER n TYPING /g
IF condition TYPING /g

where:
P

is the point where
the breakpoint
inserted
is an
integer constant, variable,
array element
is a group number

PAUSE 100

17-13

is
or

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
sets a
breakpoint at statement label 100, causes
execution to be suspended,
and
causes FORDDT to be
entered on reaching 100 in the program.
PAUSE #245 AFTER MAX(5)
causes a break to occur at source line number 245 after
encountering
this point the number of times specified
by MAX(5).
Note that AFTER can not be abbreviated.
PAUSE DELTA IF LIMIT(3,1) .GT.2.5E-3
causes a
FORDDT break to occur
if the
variable
LIMIT(3,1)
is greater than the value 2.5E-3.
The IF
can not be abbreviated,
and
the
following
FORTRAN
logical connectives are allowed:
.GT.,

.GE.,

.LT.,

.LE.,

.EQ.,

.NE.

Double-precision comparisons and arithmetic operations
are not allowed.
However,
comparisons can be made
between variables,
constants,
and logical constants
(such as .TRUE.
and .FALSE.).
PAUSE 505 TYPING /5
sets a
FORDDT breakpoint at
label
505,
and
the
variables
in group 5 are displayed.
The TYPING
specification can not be abbreviated.
PAUSE #24 AFTER 16 TYPING /3
causes a break at source line number 24 after 16 times
through; however, the contents of group 3 are displayed
every time.
the TYPING option is used with the PAUSE command,
control
can be transferred
to FORDDT at the next
typeout by typing any character on the terminal.

Whe~

Note
that pause requests remain after
a
(CTRL/C)
REENTER sequence, a START command, or a (CTRL/C) START
sequence.
PAUSE ON ERROR Causes the program to enter FORDDT whenever
an error
occurs
(such as an arithmetic overflow).
It has the
same command format as the PAUSE command.
REMOVE

Removes the previously
command format is:

set

FORDDT

from

the

breakpoints.

The

REMOVE [p]
Fo r exampl e,
REMOVE L#123
removes a
breakpoint
number 123.
REMOVE ALPHA

17-14

program

source

line

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
removes a
ALPHA.

breakpoint

from

the

subroutine

entry

your
FOR DDT
REMOVE with no arguments removes all
breakpoints,
and,
in this case,
no abbreviation of
REMOVE is allowed.
REMOVE ON ERROR Removes a PAUSE ON ERROR breakpoint.
It has the same
command format as the REMOVE command.
START

Starts your program at the normal FORTRAN main
entry point.
lhe command format is:

program

START
STOP

Terminates the program, closes all
files opened by
FOROTS,
and causes an exit to the monitor.
The usual
command format is:
STOP
STOP IRETURN
allows a
return to
devices or closing
issued.

STRACE

monitor mode without releasing
files
so that a CONTINUE can be

Displays a subprogram level traceback of the
state of the program.
The command format is:

current

STRACE
TYPE

Displays FORTRAN defined variables, arrays,
or array
elements on YOlr terminal.
The command format is:
TYPE list
where list may be one or more variables or array
references anc/or group numbers.
These specifications
must be separated by commas, and group numbers must be
preceded by a slash (I).
The command with no arguments
uses the last argument list submitted to FORDDT.
An array elemert range
example:

can

also

specified.

For

TYPE PI(5)-PI(13)
displays the values from PI(5) to PI(13) inclusive.
If
an unsubscripted array name is specified, the entire
array is typed.
There are several methods of choosing
the
typeout in conjunction with the MODE command:
1.

If you do not specify
real.

You can specify a format through the
described jn this chapter.

17-15

format,

the

form

default

MODE

command

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
3.

You can change the format(s) previously designated
by the MODE command by including print modifiers in
the TYPE or GROUP string.
The print modifiers are:
A
B
C
D
F
I
L

R
X

ASCII(left-justified)
LONG
CHARACTER
DOUBLE-PRECISION
REAL
INTEGER
LOGICAL
OCTAL
RASCII(right-justified)
COMPLEX

(C) ,
the
in mode CHARACTER
If you type a variable
is equal to the length
number of characters printed
declared in the FORTRAN source prog ram, up to a max imum
The IB switch can be used to
of 256 characters.
override the 256 character max imum.
The B switch may be used in conjunction with the A,
0,
and R switches.
This modifier
indicates that the
variable is to be interpreted as two words long.
The B
switch can also be used with the C switch to display
more than the first 256 characters of long character
strings.
The B switch can not be used alone.
The first print modifier specified
in a
string of
variables determines the mode for the entire string
unless another mode is placed directly to the right of
a particular variable.
For example, in:
TYPE II K,L/O,M,N/A,/2
the typeout mode
is
integer
specified.
Therefore,
K, M are integer integer
L is OCTAL
N is ASCII

WHAT

until

another

mode

the default mode for group 2 is

Displays on your terminal the name of the currently
open program unit,
any currently active breakpoints,
any group specifications, and any user-set
array
dimensions.
The command format is:
WHAT

17.6

ENVIRONMENT CONTROL

If a program enters an infinite loop, you can recover by typing a
<CTRL/C> (twice)
REENTER sequence.
This action causes FORDDT to
simulate a pause at the point of reentry and allows you to control
your run-away program.

17-16

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
Most commands can be used once the program has been reentered;
howevE!r,
GOTO,
STRACE, TYPE, and ACCEPT cause transfer of control to
routines external to FORDDT.
No guarantee can be made to ensure that
any of these commands following a (CTRL/C) REENTER sequence will not
destroy the program integrity.
The program must be returned
to a
stable state bE'fore any of these four commands can be issued.
In
order to restore program integrity, you should set a pause at the next
label and then CONTINUE to it:.
If the /DEBUG:TRACE switch is used, a
NEXT 1. command can be issued to restore program integrity.

17.7

FORTRAN /OPTIMIZE SWITCH

You should never attempt to m;e FORDDT with a program that has been
compiled with the /OPTIMIZE switch.
The global optimizer causes
variables to be kept in ACs.
For this reason, attempts to examine or
modify variables in optimized programs will not work.

17.8

CALLING FORDDT

FORDDT can be called directly
appropriate statement is:

from

user

FORTRAN

program.

The

CALL FORDDT
where no argument is required.
FORDDT must be loaded and
initialized
before a CALL to FORDDT is made.
This is done by starting the program
in debug mode prior to the first call (see Section 17.3, item 1). All
FORDDT commands are allowed.
A CONTINUE will resume normal execution
of thE~ user program (similar ":0 a RETURN from a subroutine).

Since FORDDT is defined as a global symbol,
users
should be careful
if they decide to use FORDDT as a
program, subroutine, or function name.

17.9

FORDDT AND FORTRAN-2a EXTENDED ADDRESSING

FORDDT VIa has been modified to be able to run
in any section and
access data and code in all sl~ctions.
The user interface to FOR DDT is
the same regardless of whether or not a
program uses extended
addressing.
FORDDT VIa is section independent.
The same FORDDT.REL will
either section a or a non-zerl) section.

work

If a program is loaded with the /DEBUG:FORDDT option,
LINK loads
FORDDT.REL with the prograln.
FORDDT.REL is a single-segment module
(it has only low segment code); therefore, when loaded with a
FORTRAN
object program that was compiled with the /EXTEND switch (see Section
16.5), FORDDT, by default, is redirected to the .DATA. psect.

17-17

USING THE FORTRAN INTERACTIVE DEBUGGER (FORDDT)
FORDDT Version 10 will
not
versions of FORTRAN-IO/20.

guaranteed

work

with

FORDDT and FORLIB must be in the same section.
Since they would
by
default go into the .DATA./.CODE. section, the user normally would not
need to be concerned about this.
However, you should be cautious when
you use the LINK /REDIRECT switch.
NOTE
The first page of any section that
reserved for FOR DDT and FOROTS.

17-18

contains

code

CHAPTER 18
USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)

This chapter describes the facilities that
the
FORTRAN Object Time
System
(FOROTS) provides for the FORTRAN user.
FOROTS implements all
standard FORTRAN I/O operations as set
forth
in the
FORTRAN-77
standard
In addition
it provides the
user with capabilities and
programming features beyond those defined in the ANSI standard.
The primary function of FOROTS is to act as a direct interface between
user-object programs and the TOPS-IO or TOPS-20 monitor during input
and output operations.
Other capabilities include:

Job initialization

Channel and memory management

Error handling and reporting

File management

Formatting of data

Mathematical library

User library (nonmathematical)

Specialized applications packages

Overlay facilities

FOROTS runs on any TOPS-IO or 1'OPS-20 system.
all TOPS-IO or TOPS-20 peripheral devices.

18.1

FOROTS interfaces

with

FEATURES OF FOROTS

The following list briefly describes many specific features of FOROTS;
more detailed
information concerning
the
implementation of these
f eat u r E! s i s g i v e n 1 ate r i nth i~; c hap t e r •
1.

Your program can run :.n either batch or timesharing mode
without
requiring a program change.
All differences between
batch-mode and timesharing-mode operations are
resolved by
FOROTS.

2.,

Your proglrams can access both
devices in the same manner.

3..

FOROTS helps provide complete data file compatibility between
all system devices.

18-1

directory

and

nondirectory

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
4.

FOROTS treats devices located at remote stations in the
way it treats local devices.

Programs written
for magnetic
tape operations will
run
correctly on disk under FOROTS supervision.
FOROTS simulates
the commands needed for magnetic tape operations.

You may change or specify object program device
specifications with a FOROTS interactive dialogue.

and

Non-FORTRAN binary data files may be read in
FOROTS.

mode

FOROTS
provides
interactive
program/operating
system
error-processing
routines.
These
routines permit you to
route
the
execution
of
the
program
to
specific
error-processing routines whenever designated types of errors
are detected.

An error traceback facility for
fatal
errors provides the
active execution path (by subroutine calls) between the main
program and the subroutine where the fatal error occurred.

10.

FOROTS provides a
trap-handling system
for
arithmetic
functions, including default values and error reports.

11.

FOROTS permits your program to switch from READ to WRITE
the same I/O device without loss of data or buffering.

12.

Although primarily designed for use with
the
FORTRAN-I0/20
object programs,
you can also use FOROTS as an independent
I/O system, and as an I/O system for MACRO object programs.

18.2

IMAGE

same

file
by

ERROR PROCESSING

Whenever a run-time error is detected,
the
FOROTS error-processing
system takes control of program execution.
This system determines the
class of the error and either outputs an appropriate message at the
controlling
terminal or
branches the program to a predesignated
processing routine.

18.3

INPUT/OUTPUT FACILITIES

On TOPS-I0, FOROTS uses monitor-buffered I/O for SEQUENTIAL, SEQINOUT,
SEQIN,
and
SEQOUT files
access,
and
uses dump mode
I/O for
DIRECT(RANDOM), RANDIN, and dump mode files access.
On TOPS-20, FOROTS uses PMAP monitor calls for disk files access other
than APPEND,
and
uses monitor-buffered
I/O for
all other
file
accesses.
The following sections describe I/O data channel and access modes.

18.3.1

Input/Output Channels Used By FOROTS (TOPS-I0 Only)

FOROTS uses extended channels starting
operations.
User programs can
request
through the ALCHN. and FUNCT. routines.
18-2

at channel
20
for
I/O
I/O channels 0 through 17

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
When a request is made for an I/O channel, a table is scanned until a
free channel
is found.
ThE!
first free channel is assigned to the
requesting program.
On completion of the assigned transfer,
control
of the I/O channel is returned to FOROTS by using the DECHN. routine.

lS.3.2

File Access

Data can be transferred between processor storage and
devices using either sequential or direct (random) access.

lS.3.3

peripheral

Closing Files After NOll-standard Termination

When a FORTRAN program is aborted by <CTRL/C> or an error, open files
cannot be closed with the monitor command CLOSE.
The following
command should be used:
REENTER
FOROTS then asks if you want the files to be closed.
YES, then, the files are closed.

you

answer

lS.3.3.1 Sequential
Access -- In
a
sequential-access
transfer
operation,
the I"ecords involved are transferred in the same order as
they appear in the source file.
Each I/O statement executed transfers
the
record immediately following the last record transferred from the
accessed source file.
A type of the sequential access
is available for
output
(write)
operations.
This type of access is called APPEND and is specified by
the OPEN statement specifier ACCESS='APPEND'
(see Section 11.3.1).
APPEND lets you write a
rE!cord immediately after the last logical
record of the accessed file.
During APPEND transfer,
the
records
already in the accessed
file
remain unchanged; the only function
performed is the appending of the transferred records to
the end of
the file.
You must specify transfer type!; (other than SEQINOUT) by setting the
ACCESS option of a FORTRAN OPEN statement to one of several possible
arguments.
For the sequential access, the arguments are:
ACCESS='SEQIN'
ACCESS='SEQOUT'
ACCESS='SEQUENTIAL'
ACCESS='SEQINOUT'
ACCESS='APPEND'

(file is opened for read-only access)
(file is opened for output)
(file is opened for input or output)
(same as SEQUENTIAL)
(sequential append access)
NOTE

A common way to append data to a
file
opened with
SEQUENTIAL access is to read past the end of file, and
then begin writing.
The FORTRAN-77 standard
requires
that a BACKSPACE operation be done to back over the
'end file reco rd' precE!d i ng the WRITE.

lS-3

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
18.3.3.2 Direct (Random) Access Mode - Direct access permits records
to be accessed and transferred from a source file in any desired
order. Direct access can only be used with disk files that have been
set up for direct access.
Direct-access files must contain a
specified number of identically sized records that may be individually
accessed by a record number.
Direct-access transfers may be done in either a read/write direction
or a special read-only direction.
You must specify random transfer
direction by setting the ACCESS option of an OPEN statement
(see
Section 11.3.1) to one of several possible arguments.
ACCESS='DIRECT'
ACCESS='RANDOM'
ACCESS='RANDIN'

18.4

(direct read/write access)
(same as DIRECT)
(direct read-only access)

ACCEPTABLE TYPES OF DATA FILES AND THEIR FORMATS

The following sections describe the
acceptable to FOROTS.

18.4.1

types

data

files

that

are

ASCII Data Files

Each record within an ASCII data file consists of a set of contiguous
7-bit characters. A vertical paper-motion character (that is, a form
feed, a vertical tab, or a line feed) terminates each set.
Logical
records may be split across physi~al blocks. There is no maximum
length for logical records.

18.4.2

FORTRAN Binary Data Files

Each logical record in a FORTRAN binary data file contains data that
the executing program can reference with either a READ or a WRITE
statement. A logical record is preceded and ended by a control ~Nord
and may have one or more control words embedded within it.
In FORTRAN
binary data files, there is no relationship between logical records
and physical device block sizes. There is no implied maximum length
for logical records.

18.4.2.1 Format of Binary Files - A FOROTS binary file can contain
three f~rms of Logical Segment Control Words (LSCW). These LSCWs give
FOROTS the ability to distinguish ASCII files from binary files.
The
value in the high-order 9 bits of an LSCW tells what kind of LSCW it
is: START, CONTINUE, or END.
LSCW
START
CONTINUE
END

001+ the number of words in the segment

including
the START LSCW word
(exclusive of the END
LSCW)
002+ the number of words in the segment
including
the CONTINUE LSCW
in the whole record
003+ the number of words
including all LSCWs

18-4

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
If the access you specify for
a
file
through
the OPEN statement
ACCESS=:
argument
is
'SEQIN'
'SEQOUT', 'SEQUENTIAL', or 'SEQINOUT',
all three LSCWs can appear in a record.
If you specify a record size,
all records are of the same length, and there are no CONTINUE LSCWs.
j

The following examples illustrate the LSCW.
file contains only 001 and 003 LSCWs.
C
C

The direct-access

LOOK AT A BINARY FILl: AND SEE THE LOGICAL SEGMENT
CONTROL WORDS.
OPEN (UNI T= 1, ACCESS= 'I>I RECT' ,MODE=' BINARY' ,
RECORDSIZE=100)
1
1=5
WRITE(l'l)
J=7
WRITE(1'2)
END

(I, J=I,10)
(J,K=l,IO)

18-5

binary

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
0/
1/
2/
3/
4/
5/
6/
7/
10/
11/
12/
13/
14/
15/
16/
17/
20/
21/
22/
23/
24/
25/
26/
27/
30/
31/
32/
33/
34/
35/
36/
37/
40/
41/
42/
43/
44/
45/
46/
47/
50/
51/
52/
53/
54/
55/
56/
57/
60/
61/
62/
63/
64/
65/
66/
67/
70/
71/
72/
73/
74/
75/
76/
77/

001000,000145~Number

000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000

of
words in
record
counting

START LSCW

100/
101/
102/
103/
104/
105/
106/
107/
110/
111/
112/
113/
114/
115/
116/
117/
120/
121/
122/
123/
124/
125/
126/
127/
130/
131/
132/
133/
134/
135/
136/
137/
140/
141/
142/
143/
144/
145/
146/
147/
150/
151/
152/
153/
154/
155/
156/
157/
160/
161/
162/
163/
164/
165/
166/
167/
170/
171/
172/
173/
174/
175/
176/
177/

18-6

000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
003000,000146~END

001000,000145
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000
000000,000000

LSCl..v

containing
the number
of words in
the record
including
all LSCWs.

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)

200/
201/
202/
203/
204/
20:)/
206/
207/
210/
211/
212/
213/
214/
215/
216/
217/
220/
221/
222/
223/
224/
225/
226/
227/
230/
231/
232/
233/
234/
235/
236/
237/
240/
241/
242/
243/
244/
245/

246/
247/
250/
251/
252/
253/
254/
255/
256/
257/
260/
261/
262/
263/
264/
265/
266/
267/
270/
271/
272/
273/
274/
275/
276/
277/
300/
301/
302/
303/
304/
305/
306/
307/
310/
311/
312/
313/

18-7

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
On TOPS-I0, in the sequential-access binary file,
the second record
crosses the disk block boundary and contains an 002 (CONTINUE) LSCW.
On TOPS-20, the CONTINUE LSCW occurs on buffer boundaries, whose size
is determined by the BUFFERCOUNT keyword in the OPEN statement (see
Section 11.3.5) (default is four pages, 4000 octal words).
C
C

LOOK AT A BINARY FILE AND SEE THE LOGICAL SEGMENT
CONTROL WORDS.
OPEN(UNIT=l,MODE='BINARY')
1=5
WRITE(l)
J=7
WRITE(l)
END

(I, J=l,lOO)
(J,K=l,lOO)

18-8

001000,000145
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000n05

OOOOOO,OOOOOS
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005

100/
]01/
102/
103/
104/
105/
106/
107/
110/
111/
112/
113/
114/
115/
116/
117/
120/
121/
122/
123/
124/
125/
126/
127/
1 30/
131/
132/
133/
134/
135/
136/
137/
140/
141/
142/
143/
144/
145/
146/
147/
150/
151/
152/
153/
154/
155/
156/
157/
160/
161/
162/
163/
164/
165/
166/
167/
170/
171/
172/
173/
174/
175/
176/
177/

18-9

001000,000032~Number

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
200/
201/
202/
203/
204/
205/
206/
207/
210/
211/
212/
213/
214/
215/
216/
217/
220/
221/
222/
223/
224/
225/
226/
227/
230/
231/
232/
233/
234/
235/
236/
237/
240/
241/
242/
243/
244/
245/

002000,000114~Continuc

000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007

LSCW

246/
247/
250/
251/
252/
253/
254/
255/
256/
257/
260/
261/
262/
263/
264/
265/
266/
267/
270/
271/
272/
273/
274/
275/
276/
277/
300/
301/
302/
303/
304/
305/
306/
307/
310/
311/
312/
313/
313/

18-10

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
Image files contain no LSCWs. You can only backspace
that is created with a record size.
C
C

IMAGE

LOOK AT AN IMAGE MODE FILE AND SEE NO LOGICAL SEGMENT
CONTROL WORDS.
OPEN{UNIT=l,MODE='IMAGE')
I=5
WRITE(l)

(I, J=l,lOO)

J=7
WRITE(l)
END

(J,K=l,lOO)

18-11

file

000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005

000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,(100005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000005
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007
000000,000007

244/
245/
246/
247/
250/
251/
252/
253/
254/
255/
256/
257/
260/
261/
262/
263/
264/
265/
266/
267/
270/
271/
272/
273/
274/
275/
276/
277/
300/
301/
302/
303/
304/
305/
306/
307/

18-13

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
The following example illustrates the
binary files.
C
C

LSCWs

for

character

data

LOOK AT BINARY MODE FILE WITH CHARACTER DATA AND SEE THE
LOGICAL SEGMENT CONTROL WORDS
OPEN (UNIT=l,MODE='BINARY')

0/
1/
2/
3/
4/
5/
6/
7/
10/
11/
12/
13/
14/

WRITE (1)

3, 'ABCDEF' ,4,'GHIJKL'

WRITE (1)
END

'MNOPQR' ,'STUVWX'

001000,000007
000000,000003
406050,342212
430000,000000
000000,000004
436211,145226
460000,000000
003000,000010
001000,000004
466351,750242
512472,452654
536600,000000
003000,000005

The following example illustrates the format
image files.
Image files contain no LSCWs.
C
C

character

data

LOOK AT IMAGE MODE FILE WITH CHARACTER DATA AND SEE
NO LOGICAL SEGMENT CONTROL WORDS
OPEN (UNIT=I,MODE='IMAGE')
WRITE (1) 3, 'ABCDEF' ,4, 'GHIJKL'
WRITE (1)
END

0/
1/
2/
3/
4/
5/
6/
7/

18.5

'MNOPQR' ,'STUVWX'

000000,000003
406050,342212
430000,000000
000000,000004
436211,145226
462331,647640
506452,352252
532573,000000

USING FOROTS

FOROTS has been designed to lend itself for use as an I/O system for
programs written in languages other than FORTRAN. Currently, MACRO
programmers may employ FOROTS as a general I/O system by
writing
simple MACRO calls that simulate the calls made to FOROTS by a FORTRAN
compiler. The calls made to FOROTS are to routines that implement
FORTRAN I/O statements such as READ, WRITE, OPEN, or CLOSE.
FOROTS will provide automatic memory allocation, data conversion,
buffering, and device interface operations to the MACRO user.
18-14

I/O

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
18.5.1

FOROTS Entry Points

FOROTS provides the following entry points for
FORTRAN compiler or a non-FORTRAN program.
contained in FORLIB.REL.
Table 18-1:

calls from either a
These entry points are

FOROTS Entry Points

Entry Point

Function

ALCHN.

Allocates ar
I/O channel
for
use
subroutine (see Section 18.5.3.12)

ALCOR.

Allocates memory (see Section 18.5.3.11)

CLOSE.

Closes a fil~.
In a FORTRAN program, this call
is
made when the CLOSE statement is ex~cuted (see
Section 18.5.3.10)

DBMS.

DBMS interface.

DEC.

DECODE routine.
This call, coupled with an
call, handles decoding.

DECHN.

Deallocates an I/O channel that was
ALCHN (see SE'ction 18.5.3.12)

DECOR.

Deallocates memory that was allocated by
call (see Section 18.5.3.11)

ENC.

ENCODE rout i rle

EXIT.

Closes all
files,
clears
interrupt system,
and
terminates
program
execution.
In a
FORTRAN
program, this call is made when an END statement is
executed in t.he main program.

EXITI.

Writes out buffers, closes and unmaps all files

FIN.

I/O list termination routine (see Section 18.5.3.9)

FIND.

FIND statement

FORER.

Err 0 r pro c e s ~; 0 r

FOROP.

Miscellaneou~;

FUNCT.

OTS-independE~nt

interface
to
provide
common
functions
(J.ike memory and I/O channel management)
for programs such as overlay handler and DBMS.
FUNCT. is an interface that works the same way with
FORTRAN, COBOL, and ALGOL run-time systems
(see
Section 18.6).

IFI.

In t ern a 1 f i 1 E~ i n put ( see Sec t ion 18. 5 • 3 • 2 )

MACRO

IOLIST

obtained
an

from
ALCOR

FORO'I'S utilities

18-15

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
Table 18-1:

FOROTS Entry Points (Cont'd)

Entry Point

Function

IFO.

Internal file output (see Section 18.5.3.2)

IN.

Formatted input
routine
(see Sections
18.5.3.5, 18.5.3.6, and 18.5.3.8)

IOLST.

I/O list routine (see Section 18.5.3.9)

MTOP.

REWIND,
BACKSPACE,
Section 18.5.3.7).

NLI.

NAMELIST input routine (see Section 18.5.3.3)

NLO.

NAMELIST output routine (see Section 18.5.3.3)

OPEN.

Opens a file.
Connects FORTRAN Logical Unit Number
to a file for I/O (see Section 18.5.3.10)

OUT.

Formatted output routine
(see Sections
18.5.3.5, 18.5.3.6, and 18.5.3.8)

RESET.

Job initialization entry

TB.

Binary input routine
18.5.3.5)

TRACE.

Traces subroutine calls

WTB.

Binary output routine (see
18.5.3.5)

and

(see

ENDFILE

18.5.3.1,

statements

(see

18.5.3.1,

Sections

18.5.3.1

and

Sections

18.5.3.1

and

~_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _L -_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _~

18.5.2

Calling Sequences

You must use the following general form for all calls made to FOROTS:
XMOVEI
PUSHJ

16,ARGBLK
17,Entry Point
(control is returned here)

where:
ARGBLK

is the address of a
specifically
formatted
argument block that contains information needed by
FOROTS to accomplish the desired operation.

Entry Point

is an entry point identifier (see Table 18-1) that
specifies the entry point of the desired FOROTS
routine.

With three exceptions, all returns from FOROTS will
be made
to
the
program instruction immediately following the call (PUSHJ 17, entry
point instruction).
The exceptions are:
1.

An error return to a specified statement number,
that
READ or WRITE statement ERR=option (see Section 10.4.7),

18-16

is,

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
;~.

An end-of-file return to a statement number, that is, READ or
WRITE statement END=option (see Section 10.4.6),

A fatal error that returns to the monitor.

Sections 18.5.3.1 through 18.~i.3.12 give the MACRO calls and
required
argumE~nt
block
formats
needed
to
initialize FOROTS and FOROTS I/O
operations.

18.5.3

MACRO Calls for FOROTB Functions

The following sections describe the forms of the MACRO calls to FOROTS
that are made by the FORTRAN compiler.
The calls described are
identified by the
language statement that they implement.
The
following
terms and abbreviations are used in the description of the
argument block U\RGBLK) of each call:
- -....~ =

pointer to the second word in the argument block
(This
is the address pointed to by the argument ARGBLK in the
calling sequence.)

FORTRAN logical unit number

count of ASCII characters

FORMAT statement address

list

name

a NAMELIST naml;

a variable specifyinq the
random access mode

list-directed [/0 (the FORMAT statement is not used)

type

type specification of a variable or constant

Input/Outpu~

list

logical

record

number

for

The argument block for all I/O statements is a
sequence of keyword
specifiers.
Bits 2-8 of each argument specify which argument is being
supplied, as follows:
1
:2
3

4
5
6
7
10
11
12
13

UNIT
FMT address
FMT size (in words)
END= address
ERR= address
IOSTAT= address
REC=
NAMELIST table addre3s
File-positioning function code
ENCODE/DECODE array 3ddress
Internal record length

18-17

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
The format of ARGBLK is:
0-1

2-S

9-12

14-17

(-count - negative of number of words in
block not including this one)

---. ARGBLK:

lS-35

kwd

type

· ··
1· a

kwd

type

··

where:
I =
y =
kwd

indirection bit
IS-bit address or data
keyword number
NOTE

Future versions of FOROTS will
not support argument
blocks with
index
registers specified either in the
arguments or in memory locations referenced indirectly
by these arguments.
Arguments must not reside in the
ACs.
In addition,
so-called
'immediate'
arguments
(those
with
a
type code of zero)
will
not be
supported.

18.5.3.1 Sequential-Access Calling Sequences - The READ and WRITE
statements for formatted sequential data transfer operations and their
calling sequences are:
READ (u, f)

1 is t

XMOVEI 16, ARGBLK
PUSHJ 17, IN.
and
WR I T E ( u , f)

1i st

XMOVEI 16, ARGBLK
PUSHJ 17, OUT.
The following arguments must be specified in ARGBLK:
1

2
3

UNIT
FMT address
FMT size

The following may also appear:
4
5
6

ERR
END
IOSTAT

IS-IS

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
The READ and WRI'rE statements for unformatted sequential data transfer
operations and their calling sequences are:
READ(u) list
XMOVEI 16, ARGBLK
PUSHJ 17, R'rB.
and
WR I T E ( u) 1 i s t
XMOVEI 16, ARGBLK
PUSHJ 17, w'rB.
The following arguments must be specified in ARGBLK:
1

UNIT

The following may also appear:
4

5
6

END
ERR
IOSTAT

18.5.3.2 Internal File Calling
Sequences - The READ and
WRITE
statements for formatted sequential data transfer operations using
internal files and their calling sequences are:

READ ( u , f ) 1 i s t
XMOVEI 16, ARGBLK
PUSHJ 17, I1PI.
,and
WR I T E ( u , f ) lis t
XMOVEI 16, ARGBLK
PUSHJ 17, IlPO.
The following arguments must be specified in ARGBLK:
1

2
3

UNIT (must be a character scalar or array)
FMT address
FMT si:ze

The following may also appear:
4
5
6

ERR
END
IOSTAT

18-19

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
18.5.3.3 NAMELIST I/O, Sequential-Access Calling Sequences - The READ
and WRITE statements
for NAMELIST-directed sequential data transfer
operations and their calling sequences are:
READ (u, name)
XMOVEI 16, ARGBLK
PUSHJ 17, NLI.
and
WRITE (u, name)
XMOVEI 16, ARGBLK
PUSHJ 17, NLO.
The following arguments must be specified in ARGBLK:
1
10

UNIT
NAMELIST address

The following may also appear:
4
5
6

END
ERR
IOSTAT

The NAMELIST table is generated from the FORTRAN NAMELIST statement.
The first word of the table is the NAMELIST name; following that are a
number of 2 word entries for scalar variables, and a number of
(N+4)
word entries for array variables, where N is the dimensionality of the
array.
The names you specify in the NAMELIST statement are stored, in SIXBIT
format,
first
in the
table.
Each name
is followed by a list of
arguments associated with the name.
The NAMELIST table is terminated
by a zero entry.
The name argument list can be in either a scalar or
an array form.

used
18.5.3.4 Array Offsets and Factoring - Address calculations
to
reference a given array element
involve factors and offsets.
For
example:
Array A is dimensioned
DIMENSION A (Ll:Ul,L2:U2,L3:U3, •.• Ln:Un)
The size of each dimension is represented by:
Sl
S2

UI-Ll+1
U2-L2+1

In order to calculate the address of an element referenced by:
A (Il,I2,I3, ••• In)
the following formula is used:
A+(II-Ll)+(I2-L2)*Sl+(I3-L3)*S2*Sl+ .•• +(In-Ln)*S[n-l]* •.• *S2*Sl
18-20

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
The terms are factored out depending on the dimensions of
not on the element referenced, to arrive at the formula:

the

array,

A+(-LI-L2*Sl-L3*S2*Sl ••• )+Il+I2*Sl+I3*S2*Sl ..•
The
parenthesized part of this
formula
is the offset
for
single-precision array; it is referred to as the Array Offset.

For each dimension of a given array, there is a corresponding
factor
by which a
subscript in that position will be multiplied.
From the
last expression, one can determine the factor for dimension n to be:
5[n-l]*S[n-2]* ••• *S2*Sl
For double-precision and complex arrays, the expression becomes:
A+2*(II-Ll)+2*(I2-L2)*Sl+2*(I3-L3)*S2+S1+ •••
Therefore, the array offset for a double-precision array is:
2*(-LI-L2*Sl-L3*S2*Sl .•• )
and the factor for the nth dimension is:
2*S[n-l]*S[n-2]* ••• *S2*Sl
The factor for the first dimension of a double-precision array is
always
2.
The
factor for the first dimension of a single-precision
array is always 1.
For character arrays, the offset is calculated
in bytes
instead of
words.
The byte offset
from
the start of a character array whose
elements are of length X is:
X* ( (II-Ll) + (I2-L2) *S2+ ••• )
This offset is X times the offset of a single-precision numeric array.
NAMELIST Table
35

NAMELIST name in SIXBIT
NAIv1.ELIST entry 1
NAIv1.ELIST en try 2

NAJ.1ELIST entry n

4000,,0

(FOROTS FIN. word)

18-21

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
SCALAR ENTRY in a NAMELIST Table

o 1

9-12

18-35

SIXBIT/SCALAR NAME/

1 0

Scalar addr

ARRAY ENTRY in a NAMELIST Table
0-1

1 2-8

9- 12 1 13

1 14-17

18-35

SIXBIT/ARRAY NAME/
1 0

tDIMS

type

I I

ARRAY SIZE

BASE ADDR

OFFSET
Factor 1
Factor 2

·
·
· n
Factor
18.5.3.5 I/O Statements, Direct-Access Calling Sequences - The READ
and WRITE statements for formatted direct-access data transfers and
their calling sequences are:
READ (u'r,f)

list

XMOVEI 16, ARGBLK
PUSHJ 17, IN.
and
WR I T E ( u ' r , f )

1i st

XMOVEI 16, ARGBLK
PUSHJ 17, OUT.
The following arguments must be specified in ARGBLK:
1
2
3

UNIT
FMT address
FMT size
REC

18-22

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
The following may also appear:
4
5

END
ERR
IOSTAT

The READ and WRITE statements for unformatted direct-access
and their calling sequences a~e:

transfers

READ (u'r)
XMOVEI l6,ARGBLK
PUSHJ P,RTB.
and
WRITE (u'r)
XMOVEI l6,ARGBLK
PUSHJ P,WTB.
The following argument must be specified in ARGBLK:
1
7

UNIT
REC

The following may also appear:

4
5

END
ERR
IOSTAT

18.5.3.6 Default Devices Statements, Calling Sequences - The FORTRAN
statements that
require
thE! use of a reserved system default device
and their calling sequences are:
Default Device
ACCEPT f, list
READ f, list
REREAD f, list

UNIT=-4

(TTY)
(CDR)
(REREAD)

UNIT~-5
UNIT~-6

XMOVEI 16, ARGBLK
PUSHJ 17, IN.
and
PRINT f, list
PUNCH f, list
TYPE f, list

(LPT)
(PTP)
(TTY)

UNIT~-3
UNIT~-2

UNIT~-l

XMOVEI 16, ARGBLK
PUSHJ 17, OUT.
The arguments for these calls are
the same as
formatted sequential READ and WRITE statements.

18-23

for

the

standard

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
18.5.3.7 Statements to Position Files - The formatted and unformatted
FORTRAN statements that can be used to control the positioning of
files and their calling sequences are:
Function
(FORTRAN Statement)

FOROTS Code

SKIPFILE (u)
BACKFILE (u)
BACKSPACE (u)
ENDFILE (u)
REWIND (u)
SKIPRECORD (u)
UNLOAD (u)

7
3
2
4

o
5
1

CALL:
XMOVEI 16, ARGBLK
PUSHJ 17, MTOP.
The following arguments must be specified in ARGBLK:
1
11

UNIT
FOROTS code

The following may also appear:
4
5
6

END
ERR
IOSTAT
NOTE
For disk files, UNLOAD is the same as REWIND; BACKFILE
and SKIPFILE are ignored.

18.5.3.8 List-Directed Input/Output Statements - You may write any
form of a sequential I/O statement as a list-directed statement by
replacing the referenced FORMAT statement number with an asterisk (*).
The list-directed forms of the READ and
calling sequences are:

WRITE

statements

and

their

READ (u, *) list
XMOVEI 16, ARGBLK
PUSHJ 17, IN.
and
WR I T E ( u,

1i s t

XMOVEI 16, ARGBLK
PUSHJ 17, OUT.
The arguments for these calls are the same as for the standard
formatted sequential READ and WRITE statements, except that the FORMAT
statement address and FORMAT statement size must be specified as zero.

18-24

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
18.5.3.9 Input/Output Data Lists - The compiler generates a
calling
sequence
to
the
run-time system for the I/O list in a READ or WRITE
statement.
The argument block associated with
the calling
sequence
contains the addresses of the variables and arrays to be transferred
to or from an I/O buffer.
The general form of an I/O list calling sequence is:
XMOVEI 16, ARGBLK
PUSHJ 17, IOLST.
Any number of elements may be included in the ARGBLK.
The end of
argument block is specified by a zero entry or a FIN entry.
Mnemonic Name

the

FOROTS Value

DATA
SLIST
ELIST
FIN
F77 SLIST
F77 ELIST

1
2

3
4
5
6

The elements of an I/O list are:

DATA
The DATA element converts one sing1e- or double-precision or
complex
item from external
to
internal
form
for a READ
statement and from internal to external
form
for
a WRITE
statement.
Each DATA element has the following format:

~EE
~bb
2.

14-17

18-35

SCALAR

SLIST
The SLIST argument converts an entire array from internal
to
external
form or vice versa,
depending on the
type of
statement (that
is,
READ or WRITE)
involved.
An SLIST
consist of a table of arguments that has the following form:

0-1

9-12

2-8

14-17

18-35

,--I--

SLIST

tyt: e

#ELEMENTS

a
1 a

tyt: e

INCREMENT

type

BASE ADDRI

1
1

18-25

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
For example, the sequence:
DIMENSION A(100) ,B(100)
READ(-,-)A
or
READ(-,-) (A(I) ,1=1,100)

!only when the /OPT switch is used

develops an SLIST argument of the form:
0-1

2-8

9-12

14-17

18-35

1
1
1
0

2
0
0
4

2
2
2
0

0
0
0
0

[144]
[1 ]
A
0

0
0
0
0

More than one base address may appear in a SLIST as
the increment is the same. The sequence:
DIMENSION A(100), B(100)
WRITE (-,-) (A(I) ,B(I) ,1=100)

long

!only when the /OPT
switch is used

develops an SLIST argument of the form:

0-1

2-8

9-12

14-17

18-35

1
1
1
1
0

2
0
0
0
4

2
2
2
2
0

0
0
0
0
0

[144 ]
[1 ]
A
B
0

0
0
0
0
0

ELIST
The SLIST format permits only a single increment to be
specified for a number of arrays, while the ELIST permits
different increments to be specified for different arrays.
An ELIST consists of a table of arguments that has the
following form:
The format of the ELIST is:
0-1

2-8

9-12

14-17

1 0

ELIST

type

No. Elements to
transfer
increment 1

1 0

type

Base ADDR 1
increment 2

1 0

type

Base ADDR 2
increment N

1 0

type

Base AD DR N

18-26

18-35

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
For example, the FORTRAN sequence:
DIMENSION IC(6,100), IB(lOO)
WRITE(-,-) (IB(I) ,IC(l,I) ,1=1,100)
produces the ELIST:
0-1

2-8

9-12

14-17

18-35

1
1
1
1
1
0

3
0
0
0
0

2
2
2
2
2
0

0
0
0
0
0
0

[144]
[1 ]
IB
[6]
IC
0

This could

0
0
0
0
0
0

The increment may be zero.
sequence:
WRITE(-,-) (K,I=l,lOO)

produced

the

!cnly when the /OPT switch is used

Produces the ELIST:

0-1

2-8

9-12

14-17

18-35

1 0

[144 ]

1 0

[0 ]

1 0

0 0

FIN
The end of an I/O list is indicated by a FIN
the I/O processor interprets this element, it
to FIN to terminate the I/O.
This call must
each I/O initialization call, including calls
list.

element.
When
performs a call
be made after
with a null I/O

The FIN routine may be entered by an explicit call or by an
argument in this I/O list argument block.
Both calls can not
be used. The FIN element has the following format:

I :-:~,=t._3_--L.._l_4_:_1_7--,_1_8_:_3_5_j....l
EXPLICIT CALL:
PUSHJ 17, FIN.

18-2:7

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
5.

F77 SLIST
if the numb€!r
This is the same as SLIST except that
elements is less than or equal to zero, no I/O is done.

F77 ELIST
is the number
This is the same as ELIST except that
elements is less than or equal to zero, no I/O is done.

18.5.3.10 OPEN and CLOSE Statements, Calling Sequences - The form and
calling sequences for the OPEN and CLOSE FORTRAN statements are:
OPEN statement call:
XMOVEI 16, ARGBLK
PUSHJ 17, OPEN.
CLOSE statement call:
XMOVEI 16, ARGBLK
PUSHJ 17, CLOSE.
where ARGBLK is:
0-1

9-12

2-8

14-17

Negative of
the number
of words in
block not
including
this one.

18-35

1 0
1 0
1 0

G
G
G

type I
type I
type I

X
X
X

H
H
H

·· ·
·
··
1· 0·

·
·
·
·
G

·· ·
·
·· ·
type I·

·
·
··
X

··

The G field (bits 2 through 8)
contains a value that defines the
argument name;
the H field (bits 18 through 35) contains an address
that points to the value of the argument.
Note that the G field
values for
OPEN and CLOSE statements are not the same as those for
other I/O statements.

18-28

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
The numeric codes that may appear in the G field are:
G Field

Open Argument

01
02
03
04
05
06
07
10
11
12
13
14
15
16
17
20
21
22
23
24
25
26

G Field

DIALOG
ACCESS=
DEVICE=
BUFFERCOUNT=
BLOCKSIZE=
FILE=
PROTECTION=
DIRECTORY=
LIMIT=
MODE=
FILESIZE=
RECORDSIZE=
DISPOSE=
VERSION=
Reserved
Reserved
IOSTAT=
ASSOCIATEVARIABLE=
PARITY=
DENSITY=
BLANK=
CARRIAGECONTROL=

27
30
31
32
33
34
35
36
37
40
41
42
43
44
45
46
47
50
51
52
53

Open Argument
FORM=
BYTESIZE=
PADCHAR=
RECORDTYPE=
STATUS=
TAPEFORMAT=
READONLY=
UNIT=
ERR=
EXIST=
FORMATTED=
NAMED=
NEXTREC=
NUMBER=
OPENED=
SEQUENTIAL=
UNFORMATTED=
NAME=
Reserved
Reserved
DIALOG=

lS.5.3.11 Memory Allocation Routines - The memory management module
is called to allocate or deallocate memory blocks.
There are two
entry points (ALCOR. and DECOR.) that control memory allocation and
deallocation.
When TOPS-20 extended addressing is in effect, ALCOR. and DECOR.
can
be used; however, memory will be allocated in FOROT's section instead
of
in the user's section.
You
can
use
the
LINK
switch
/OTSEGMENT:NONSHARABLE to put FOROTS in the user's section.
Use the ALCOR.
entry to allocate the number of words specified in the
argument block variable.
Upon return, AC 0 will contain either the
address of the allocated memory block or -1, which
indicates that
memory is not available.
The calling sequence for an ALCOR. call is:
XMOVEI 16, ARGBLK
PUSHJ 17, ALCOR.
where ARGBLK is:

r-0__-_l____

~__2_-S______~1_9_-_1_2_~
-1

l~served type

E
lS-29

14-17

1S-35
0

Address of
number of words

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
Use the DECOR. entry to deallocate a
previously allocated block of
memory;
the argument variable must be loaded with the address of the
memory block to be returned.
If the number of desired words is N, ALCOR. actually removes N+2 words
from
free
storage.
The pointer
returned points to the third word
(word 2 as opposed to word 0) removed from free
storage.
The
first
two words are
used
by FOROTS to maintain linked lists of allocated
(using ALCOR.) and free storage, and must not be modified.
The calling sequence for a DECOR. call is:
XMOVEI 16, ARGBLK
PUSHJ 17, DECOR.
where ARGBLK is:
0-1

2-8

_r-

14-17
-

-I

1 0

18-35

--0

Reserved type

Pointer to word
containing
address of block
to be returned

.18.5.3.12 Channel Allocation and Deallocation Routines - You may
allocate software channels in MACRO programs by means of calls to the
ALCHN. routine and deallocate them by calls
to
the DECHN.
routine.
Values are returned in AC O.
Use the ALCHN. entry to allocate a
particular channel
or
the next
available channel.
The channel
to be
allocated
is passed as an
argument to ALCHN.
Zero is passed as an argument to allocate the next
available channel.
Allowed channels are 1 through 17 (octal).
If the
channel requested is not available, or all channels are in use, ALCHN.
returns with a -1
in AC
O. In normal returns, AC 0 contains the
assigned number.
The calling sequence of an ALCHN.

routine is:

XMOVEI 16, ARGBLK
PUSHJ 17, ALCHN.
where ARGBLK is:
0-1

2-8

9-12

14-17

18-35

---1

---... 1 0

Reserved type

18-30

Pointer to a word
containing
the channel #
or zero

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
Use the DECHN. entry to deallocate a previously assigned channel.
The
channel
to be
released
is passed as an argument to DECHN.
If the
channel to be deallocated was not assigned by ALCHN. and
thus cannot
be deassigned, AC 0 is set to -Ion return.
The calling sequence for a DECHN. routine is:
XMOVEI 16, ARGBLK
PUSHJ 17, DECHN.
where ARGBLK is:
I

0-1

9 12
1 -

2-8

14-17

-1

1 0

18.6

18-35
0

Reserved type

[

Pointer to a word
containing
the channel #
to be released

FUNCTIONS TO FACILITATE OVERLAYS

FOROTS provides a subroutine (FUNCT.) to serve as an
interface with
the LINK overlay handler.
This subroutine consists of a group of
functions that allow the overlay handler to perform I/O, memory
management, and error message handling.
These functions have only one
entry point, FUNCT.; and they are called by the sequence:
XMOVEI 16, ARGBLK
PUSHJ 17, FUNCT.
The format of the ARGBLK is:
ARGBLK:

-(n+3>"O
IFIW 2,address of integer function code
IFIW 17,address of 3-letter ASCII error prefix
IFIW 2,address of status code on return
IFIW type,address of first argument
IFIW type,address of nth argument

where:
type
function code
error prefix
status

is the FORTRAN argument type (see Chapter 14)
is the number of one of the required functions
is ignored by FOROTS
is undefined on the call and set on the return
with cne of the values below.
-1

l •••• n

Function not implemented
Successful return
Specific error message

When TOPS-20 extended addressing is in effect,
FUNCT.
can be used;
however,
memory will
be allocated in FOROTS's section instead of in
the
user's
section.
You
can
use
the
LINK
switch
/OTSEGMENT:NONSHARABLE to put FOROTS in the user's section.

18-31

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
Table 18-2:
Function
Number
(Octal)

Function Numbers and Function Codes

Function
Mnemonic

0
1
2

ILL
GAD
COR

3
4
6

RAD
GCH
RCH
GOT

7
10

ROT
RNT

IFS

CBC

13
14
15
16

RRS
WRS
GPG
RPG

17
20
21
22

GPSI
RPSI
MPG
UPG

L-____________- L_ _ _ _ _ _ _ _ _ _ _ _ _ _

Function Description
Illegal function; returns -1 status
Gets a specific segment of memory
Gets a given amount of memory from
anywhere
in
the space allocated to the
overlay handler
Returns a specific segment of memory
Gets an I/O channel
Returns an I/O channel
Gets memory from the space allocated to
the object-time system
Returns memory to the object-time system
Gets
the
initial
run time
from
the
object-time system
time
file
Gets
the
initial
run
specification
from
the
object-time
system
Cuts back memory (if possible) to reduce
job size
Reads RETAIN status (DBMS)
Writes RETAIN status (DBMS)
Allocates memory on a page boundary
Deallocates memory obtained by function
15
Gets TOPS-20 PSI channel
Returns TOPS-20 PSI channel
Gets a contiguous set of pages
Returns a contiguous set of pages
~

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _._________ ._ _ _ _

Each function of the
FUNCT.
subroutine
is described
below.
The
arguments described in each of the following functions are what should
be in the addresses pointed to by the argument block described above.
ILL (0) FUNCTION - This function is illegal.
ignored, and the status returned is -1.

The

argument

block

GAD (1) FUNCTION - This function gets memory from a
specific address
in the space allocated to the overlay handler.
The arguments are:
arg 1
arg 2

address of requested memory
address of number of words of memory to allocate

A call to GAD with arg 2 equal to -1 requests all available memory.
On return, the status is one of the following:

o
1
2
3

successful allocation
not enough memory available
memory not available at specified address
illegal arguments (such as, address + size
than 256K)

18-32

greater

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
COR (2) FUNCTION - This function gets memory from any available
allocated to the overlay handler.
The arguments are:

space

undefined (address of allocated memory on return)
address of siz~ of requested allocation
On return, the status is one of the following:

memory allocatE'd
not enough memory available
illegal argument (that is, size is greater than 256K)

RAD (3) function - This
funct.ion
returns
address within the space allocated
to
arguments are:
arg 1
arg 2

memory at
the specified
the overlay handler.
The

address of memory to be return
address of si ZE! of memory to be returned

(in words)

On return, the status is one of the following:

successful return of memory
me-mory cannot be returned
illegal argument (that is,the address or
g reate r than 2~i6K)

3
GCH (4)
are:

FUNCTION - This function gets an I/O channel.

arg 1
arg 2

undefined
return)
ignored

(address

channel

number

the

size

The

arguments

allocated

On return, the status is one of the following:
()

sllccessful channel allocation
no I/O channels available

RCH (5) FUNCTION arguments are:
arg 1
arg 2

This

function

returns

I/O

channel.

The

address of number of channel to be returned
isrnored

On return, the status is one of the following:

o
1

channel releasl~d
invalid channel number

GOT (6) FUNCTION - This function gets memory from the space
to the object-time system.
The arguments are:
arg 1
arg 2

allocated

undefined (address of allocated memory on return)
address of sizl~ of mE~mory requested

On return, the status is one of the following:

o
1
3

successful allocation
not enough memory available
illegal argument (such as, size is greater than 256K)

18-33

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
This function differs from function
1
in
that
if the object-time
system has two free memory lists, then function 1 is used to allocate
space for links, and this function is used to allocate space
for
I/O
buffers.
Function 1 uses the free memory list for LINK, and function
6 uses the list for the object-time system.
ROT (7) FUNCTION - This function returns
system.
The arguments are:
arg 1
arg 2

memory

the

address of memory to be returned
address of size of memory to be returned

object-time

(in words)

On return, the status is one of the following:

successful return of memory
memory cannot be returned
illegal argument (such as, the address or the
greater than 256K)

size

RNT (10) FUNCTION - This function returns the initial
run
time
from
the object-time system.
(At
the beginning of the program, the
object-time system will have executed a RUNTIM UUO; that result is the
time returned by RNT.) The arguments are:
arg 1
arg 2

undefined
return)
ignored

(contains address

initial

run

On return, the run time is in arg 1, and the status is O.
is

time

The

status

IFS (11) FUNCTION (TOPS-IO only) - This function returns
the
initial
run-time
file
specification
from
the object-time system.
(This
initial file specification is the one used to begin the program;
that
is, it was given with the system RUN command.) The arguments are:
arg I
arg 2
arg 3

undefined (address of SIXSIT device on return)
undefined (address of SIXSIT filename on return)
undefined
(project-programmer number on
return
address of path block if run from SFD

On return, the status is one of the following:

successful return
error

This function tells the overlay handler which file to read
initial RUN command.

after

CSC (12) FUNCTION - This function cuts back memory if possible,
reduces the size of the user job.
There are no arguments.

the
which

The returned status is:

a
RRS
WRS

(13)
(14)

always
FUNCTION
FUNCTION - These functions are reserved for use by DBMS.

GPG (15) FUNCTION - This function gets memory on a page boundary.
arguments are the same as for FUNCTION 6, GOT.
arg 1
arg 2

ignored
address of number of words of memory to allocate
18-34

The

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
On return, arg 1 has the address of the allocated memory.
It will
be
on a page boundary.
(That is, it will be a multiple of 1000 octal.)
On return, the status is one of the following:

o
1

RPG
15.

successful
not enough memory available
illegal argument

(16) FUNCTION - This function returns memory obtained by
The arguments are the sarre as for FUNCTION 7, ROT.
arg 1
arg 2

FUNCTION

address of memcry to be returned
address of SiZE of memory to be returned (in words)

On return, the status is one of the following:

o
1

successful return of memory
was not allocated
i 11 ega 1 a r g urn e r, t

GPSI (17) FUNCTION - This function gets the TOPS-20 PSI
channel.
It
assigns a
software
interrupt channel
number.
GPSI provides only
controlled access to the PSI tables.
It
arranges that
the
tables
exist and that SIR and EIR have been done, but does not do AIC or any
other JSYS necessary to
set up the channel
(ATI or MTOPR,
for
example) •
arg 1
arg 2
arg 3

address of channel
number
to allocate,
or -1
to
allocate any user-assignable channel
address of lev~l number
address of int~rrupt routine (if the FORTRAN program is
running
on a system that supports extended addressing,
the interrupt Ioutine address may be a
30-bit address
in any section,
including section o.
Otherwise, the
interrupt addr~ss must be an 18-bit address.)

On return, the status is one of the following:

o
1
2

allocated OK (arg 1 is the channel
number
sent)
requested channel was already assigned
no free channe]s
arg ument error

-1

was

RPSI (20) FUNCTION - This function returns the TOPS-20 PSI
channel.
It returns a PSI channel allocated by FUNCTION 17. RPSI provides only
controlled access to the PSI tables.
It
removes the given channel
from
the
tables.
This
function does not do DIC or any other JSYS
necessary to remove an interrupt condition from a channel.
arg 1

address of charlnel number to return

On return, the status is one () f the following:
0

01<

channel was not. allocated
argument error

MPG (21) FUNCTION - This function gets a contiguous set of pages.
The
pages
requested are always allocated from the section FOROTS is in.
The user cannot depend upon this call to either create or destroy the
pages.

18-35

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
arg 1

first page number to allocate.
in the range
to 777.

arg 2

number of pages to allocate

The page number must be

On return, the status is one of the following:
successful allocation of all given pages
one or more pages were already allocated
illegal argument (bad page number or count)

°3
1

UPG (22) FUNCTION- This function returns a contiguous set of pages.
The pages
returned are considered to be in the section FOROTS is in.
The user cannot depend upon this call to either create or destroy the
pages.
arg 1

first page number to deallocate.
be in the range 9 to 777

arg 2

number of pages to deallocate

The page number

must

On return, the status is one of the following:

°1

successful deallocation of all given pages
one or more pages was not allocated by MPG
illegal argument (bad page number or count)

18.7

LOGICAL/PHYSICAL DEVICE ASSIGNMENTS

You make FORTRAN logical and physical device assignments at run time,
or standard system assignments are made according to a FOROTS Device
Table, that is,
DEVTB.
Table 10-3
in Section 10.4.3 shows the
standard assignments contained by the Device Table.

18.8

FOROTS AND INQUIRE BY FILE STATEMENT

See Section 11.7 for a description of the INQUIRE statement.
If no device is given for the FILE= specifier, FOROTS uses DSK: as the
default.
If no extension is given, FOROTS uses .DAT.
For TOPS-20, if
no generation number is given, FOROTS uses a generation number of 0.
FOROTS determines if the device specified is a disk.
a disk, the following happens:
•

If the device is

FOROTS determines
if a
file
exists
with
the
file
specification given in the INQUIRE statement.
It returns the
answer (either .TRUE.
or .FALSE.) in the variable specified
by the EXIST= specifier,
if any.
If such a file exists,
FOROTS 'expands' the file specification as follows:
•

A logical name is translated into a physical device name.

•

For TOPS-20, the file specification, which is overlaid by
the
user-specified directory,
filename, extension, and
generation.

•

For TOPS-20,
an
actual
file generation
number
substituted for a generation number of 0, -1, or -2,
18-36

USING THE FORTRAN OBJECT TIME SYSTEM (FOROTS)
The resultant file specification, in string form,
the 'full (expanded) file string.'
•

called

FOROTS searches for a match between the
file
specification
given
in
the
INQUJ:RE statement and
a file specification
associated with
a
logical
unit
for
which
there
is
a
"connection."
This is to determine the values to be returned
for the INQUIRE spec~.fiers OPENED= and NUMBER=
(see Section
11.7.3). FOROTS looks at all FORTRAN logical units for which
there is a connection in ascending order, starting with zero.
FOROTS compares the j~ile specification given in
the
INQUIRE
statement
(with
FILE= defaults applied) with the exact file
specification given
in
the
OPEN statement
(with
FILE=
defaults applied) if the following is true:
•

The file does not exist on the directory.

•

An OPEN statement: has been executed and STATUS='UNKNOWN'
and ACCESS='SEQUI:NTIAL' (see Section 11.3.1).

•

No data transfer statements have been executed using
unit.

the

If the file exists, FOROTS compares the
full
file
string
associated with the unit wi.th the full (expanded) file string
given in the INQUIRE statement.
The
file
exists
if
the
following is true:
•

An OPEN has bE!en executed wi th STATUS other
'UNKNOWN' or ACCESS other than 'SEQUENTIAL'.

•

An I/O transfer

~;tatement

than

has been executed.

If neither of these two comparisions are
successful,
FOROTS
returns
the curren1:
unit
number in the variable specified
with the NUMBER= specifier,. and returns 'YES' in the variable
specified with
the
OPENED= specifier.
If the same file is
connected on several units, the matching technique described
will
rE!turn
the
snallest
unit number for which there is a
match.
For non-disk devices specifiecl in the INQUIRE statement
file
string
specification, FOROTS looks a1: all the FORTRAN logical units for which
there is a connection in ascending order, starting with zero.
If the device in the file str~ng specified in the INQUIRE statement is
not
the
user's controlling
terminal,
FOROTS expands
the
file
specification given in the IN(~UIRE statement by translating a
logical
name given as
the device :.nto its corresponding physical name.
It
then compares the device part of this expanded file specification with
the dE!vice part of the full (E!xpanded) file string associated with the
logical unit.
If the device is
the
user's controlling
terminal
(device
'TTY'),
FOROTS determines
if
the device associated with the logical unit is
also the user's controlling terminal.

18-37

CHAPTEIR 19
USING THg FORTRAN REAL-TIME SOFTWARE (TOPS-IO ONLY)

19.1

INTRODUCTION

The FORRTF library subroutines are designed to allow the
timesharing
FORTRAN user
to do real-time programming on TOPS-IO systems.
These
subroutines, described in Section 19.3, are listed below:
LOCK
R'rIN IT
CONECT
R'rSTRT
BLKRW
RTREAD
RTWRIT
STATO
STATI
RTSLP
RTWAKE
DISMIS
DISCON
UNLOCK
With these subroutines, the timesharing job can dynamically connect
real-time devices
to
the priority Interrupt (PI) system, respond to
these devices at interrupt level,
remove the devices
from
the
PI
system,
and change their
PI level.
Use of these routines requires
that you have real-time privileges and are able to lock your
job
in
core.
The privilege bits required are:
JP.RTT
JP.LCK

(bit 13) - real-time privileges
(bit 14) - locking privileges

The number of real-time devices that can be handled at one time is an
assembly-time
constant
(RTDEVN)
in
the
FORRTF source.
The
DIGITAI...-d i str ibuted so ftware h,3S RTDEVN equal to
2,
but
it can be
changed
(up to 6) by editing the statement IRTDEVN==2" in FORRTF.MAC
and reassembling.
The error messages output by F~RRTF can be
in either
full
message
format or coded
format
(refer to Table 19-1).
Use of the code and
format saves over 100 words of run-time core.
If core
is limited,
reassembly of FORRTF.MAC with the assembly-time constant SHORT changed
from the DIGITAL-distributed a (full
format)
to -1
(coded
format)
accomplishes the core saving.
On multiprocessor systems, the
real-time
traps apply only to
the
processor
specified
by
t~e
job's CPU specification.
If the
specification indicates more tnan one processor, the specification
is
changed to indicate CPUO.
Note that the priority interrupt channel is
only for the indicated CPU.

19-1

USING THE FORTRAN REAL-TIME SOFTWARE (TOPS-IO ONLY)
19.2

USING FORRTF

Users of FORTRAN-IO real-time software must consider the following:
1.

Use of memory

Device control in block or single mode

Priority-interrupt levels

Masks

19.2.1

Memory

The job being
executed must be
locked
in memory with
the
LOCK
subroutine
(see Section 19.3.1).
Any data being read into memory can
only be read into the low segment and above
the
protected
job data
area
(the
first
140 locations).
The BLKRW subroutine (see Section
19.3.5) tests the validity of the locations specified to receive data
in block reading to ensure that no overwritings occur.
However, when in block mode, the block pointer must be
reset before
dismissing
the end-of-block interrupt; otherwise, all memory could be
overwritten.

19.2.2

Modes

Real-time jobs can control their devices in one of two ways:
block
mode or
single mode.
In block mode, an entire block of data is read
or written before the
user-interrupt
routine
is
run;
whereas,
in
single mode,
the user-interrupt program is run every time the device
interrupts.
fast-block mode and
normal-block
There are two types of block mode:
mode.
A device
in
fast-block mode
requires that a PI channel be
dedicated entirely to itself.

19.2.3

Priority-Interrupt Levels

Priority-interrupt levels 1 through 6 are legal depending
on the
system configuration.
The lower the number of the level, the higher
the priority of that level.
Programs that execute
for
a
long
time
should
not be put on high-priority interrupt levels, since they could
cause other
real-time programs on lower
levels to
lose
data.
Specification of the PI level as zero for a particular device causes
the device to be removed from the PI system.

19.2.4

Masks

For a description of the bits included
in
the startmsk and
intmsk
parameters of RTSTRT and the status word in STATO and STATI, see the
DECsystem-10 Hardware Reference Manual.

19-2

USING THE FORTRAN REAL-TIME SOFTWARE (TOPS-I0 ONLY)
19.3

SUBROUTINES

Each of the 14 subroutines associated with FORTRAN real-time software
is described
briefly
in
Sections 19.3.1 through 19.3.14.
These
subroutines have been
programmed
to
be compatible with
programs
written according to the TOPS-10 Monitor Calls Manual.

19.3.1

LOCK

LOCK locks the
job
in memory and
allocates and
initializes the
internal
controlling
tables for all real-time devices.
LOCK must be
called before any of the other real-time routines, and must be called
exactly once.
The form of the LOCK subroutine is:
CALL LOCK

19.3.2

RTINIT

RTINIT initializes the internal tables controlling a real-time device.
RTINIT must be called for each individual device being used.
The form of the RTINIT subroutine is:
CALL RTINIT (unit, dev, pi, trpadr, intmsk)
where:
unit

is the real-timE device unit number (any number from
I
to
RTDEVN).
This number is not connected in any way
with the FORTRAN logical unit number.

dev

is the device code for the real-time device
DECsystem-l_Q Hardware Reference Manual) •

is the priority-interrupt level on which the
real-time
device
is
to be
run.
Each
individual device
in
fast-block mode must have a level dedicated to
itself.
If the level
is equal
to
zero, the device will be
removed from the priority-interrupt system altogether.

(see

the

If it is necessary to connect one device
to
several
levels simultaneously,
a
negative value for PI tells
the system not to remove any other occurrences of the
device
from
any other (or the same) PI level.
(Note
that this count~. as another real-time device.)
tr pad r

is the address of a FORTRAN entry to
which
real-time
interrupts are
to
trap.
This can be a function or
subroutine subprogram.
Any variables that must
be
shared
betwe~n
the
user-level
code
and
the
interrupt-level routine must
be passed
by means of
COMMON.
Passing
them as parameters causes disastrous
results.

19-3

USING THE FORTRAN REAL-TIME SOFTWARE (TOPS-IO ONLY)
intmsk

is the mask of all interrupting flags for the real-time
This
is actually set up by RTSTRT and should
be zero whenever the real-time device is inactive (that
is,
in
a call
to
RTINIT,
except
in
the case of
fast-block mode).
In fast-block mode, intmsk must be
set to -1.

device.

19.3.3

CONECT

CONECT tells the system to connect a real-time device to the proper PI
level and sets up several elements of the device-controlling tables.
Every device must be CONNECTED.
The form of the CONECT subroutine is:
CALL CONECT (unit, mode)
where:
unit

is the real-time device unit number

mode

is either:

(see RTINIT).

-2, write a block of data, fast mode; then interrupt.
-1, write a block of data, normal mode; then interrupt.
0, interrupt every word
+1, read a block of data, normal mode; then interrupt.
+2, read a block of data, fast mode; then interrupt.

19.3.4

RTSTRT

RTSTRT can be used to start a real-time device, as well as to stop it
and
zero its interrupt mask.
A device must be started to be used and
should be stopped before it is disconnected.
The form of
the
RTSTRT
subroutine is:
CALL RTSTRT (unit, startmsk, intmsk)
where:
unit

is the real-time device unit number

startmsk

is the flags necessary to start
the device
(see
the
DECsystem-lO
Hardware Reference Manual).
If
the device is being stopped, this parameter should
be zero.

intmsk

is the mask of all
interrupting
bits
for
the
particular device
(see the DECsystem-lO Hardware
Reference Manual).
If the device is in fast-block
mode
and
being
started, intmsk should equal -1;
if, however, the device
in
any mode
is being
stopped, the parameter must be o.

19.3.5

(see RTINIT).

BLKRW

BLKRW is used with either of the block modes.
It sets up the size and
starting
address of
the data block being handled.
A new count and
starting address must be set up each time the current one runs out.
19-4

USING THE: FORTRAN REAL-TIME SOFTWARE (TOPS-I0 ONLY)
The form of the BLKRW subroutine is:
CF.LL BLKRW (unit, count, blkadr)
where:
real-tim~

unit

is the

count

is the number of words to be read or written.

bl kad r

is the array into which the data is to
fro m wh i chi t i~; tab e' rea d •

19.3.6

device unit number

(see RTINIT).

written

RTREAD

RTREAD, used with a device in single mode, reads a single word of data
from the device.
The form of the RTREAD subroutine is:
CALL RTREAD (unit, datadr)
where:
unit

is the

datadr

is the address of the location in which
data read.

19.3.7

real-tim~

device unit number

(see RTINIT).
to

store

the

RTWRIT

RTWRIT sends a single word of data to a
mode.

real-time

device

single

The form of the RT'WRIT subrouti ne is:
CALL RTWRIT (unit, datadr)
where:
real-tim~

unit

is the

datadr

is the location of the data word
device.

19.3.8

device unit number

(see RTINIT).

sent

the

STATO

STATO sends the specified status word to
real-time device.
The form of the STATO subroutine is:
CALL STATO (unit, statadr)

19-5

the

status

USING THE FORTRAN REAL-TIME SOFTWARE (TOPS-IO ONLY)
where:
unit

is the real-time device unit number

statadr

is the location of the word of status bits to
to the real-time device.

19.3.9

(see RTINIT).
be

sent

STATI

STATI reads the current device status bits into the location specified
for inspection by the FORTRAN program.
The form of the STATI subroutine is:
CALL STATI

(unit, adr)

where:
unit

is the real-time device unit number (see RTINIT).

adr

is the location into which the device status
to be read.

19.3.10

bits

are

RTSLP

RTSLP is called from the timesharing level and causes the FORTRAN job
to
sleep until
RTWAKE is called from interrupt level.
The program
goes to sleep for the specified number of seconds (up to 60). When it
wakes
up,
it checks to see if RTWAKE has been called from interrupt
level.
If RTWAKE has been called,
RTSLP returns
to
the calling
program; otherwise the job goes back to sleep again.
The form of the RTSLP subroutine is:
CALL RTSLP (time)
where:
time

19.3.11

is the length of sleep time in seconds.

RTWAKE

RTWAKE is called at interrupt level to wake up the FORTRAN program.
The form of the RTWAKE subroutine is:
CALL RTWAKE

19.3.12

DISMIS

DISMIS dismisses the
interrupt
user-interrupt
routine must be
causes its execution to begin.

currently being
processed.
The
sure
to dismiss the interrupt that

The form of the DISMIS subroutine is:
CALL DISMIS
19-6

USING THE FORTRAN REAL-TIME SOFTWARE (TOPS-IO ONLY)
19.3.13

nISCON

nISCON disconnects a real-time device from its PI level.
All devices
should be disconnected through calls to nISCON before the job is
terminated.
The form of the nISCON

subrout~ne

is:

CALL nISCON (unit)
where:
unit

19.3.14

is the real-time device unit number (see RTINIT).

UNLOCK

UNLOCK unlocks the
job from core.
When execution of a
job is
complete,
the job is automatically unlocked before the return to the
monitor.
The UNLOCK subroutine provides a method
to
unlock a
job
before execution is complete.
Note that all real-time device handling
must be finished before the job is unlocked.
The form of the UNLOCK subroutine is:
CALL UNLOCK

19.3.15

Error Messages

Table 19-1 lists real-time software error messages, including the code
format,
the full message format,
and the subroutine in which the
messag(~
occurs.

19-7

USING THE FORTRAN REAL-TIME SOFTWARE (TOPS-IO ONLY)
Table 19-1:

Error Messages - Code Format and Full Message Format

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

Code Format

Full Message Format

Subroutine in
which message
occurs

?ILLEGAL UNIT NUMBER.
TO HANDLE MORE DEVICES,
REASSEMBLE FORRTF WITH A
LARGER
?ERROR COMES FROM THE
SUBROUTINE "subroutine name"

"RTDEVN"

?RTINIT MUST BE CALLED BEFORE
CONECT

CONECT

~------------~--------------------------'----------'-------r-----------------

?CONECT MUST BE CALLED BEFORE
RTSTRT OR BLKRW

RTSTRT,BLKRW

~------------~-------------------------------------------+-------------.----

?REAL TIME BLOCK OUT OF BOUNDS
BLKRW
?END OF BLOCK TOO HIGH
[such as, overwrites some program
or in high segment]
?END OF BLOCK TOO LOW,
such as, start address less
than 140

4
A

~------------~-----------------------------------------.--+--------------.----

?JOB CANNOT BE LOCKED IN
CORE
?JOB NOT PRIVILEGED
?NOT ENOUGH CORE AVAILABLE
FOR LOCKING

A
B

LOCK

?APR ERROR AT INTERRUPT
LEVEL
?PDL OVERFLOW
?ILLEGAL MEMORY REFERENCE

6
A
B

~------------~------------------------------------'-------r-------------'----

A
B
C
D
E

o
1
8

A
B

?RTTRP ERROR
realtime trap error of the
following sort
?ILLEGAL PI NUMBER
PI channel not available
?TRAP ADDRESS OUT OF BOUNDS
?SYSTEM LIMIT FOR REALTIME
DEVICES EXCEEDED
?JOB NOT LOCKED IN CORE OR NOT
PRIVILEGED
?DEVICE ALREADY IN USE BY
ANOTHER JOB
?OCCURRED IN THE DISCON
ROUTINE
?OCCURRED IN THE CONECT
ROUTINE
?NOT ENOUGH CORE AVAILABLE
FOR THE CONTROL BLOCKS
?NOT ENOUGH CORE AVAILABLE

DISCON
CONECT

LOCK

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

19-8

l\PPENDIX A

SUMMARY OF FORTRAN STATEMENTS

This appendix summarizes the
forms of all
FORTRAN statements and
provides a
section referenCE! where each statement is described in
detail.
Form

Section Reference

ACCEPTIFMT= n,END = sll,ERR= sl[,IOSTA':' = iosliliolistl
ACCEPTIFMT= *[,END = s][,ERR = sJ[,IOSTA'f = iosjHiolistj
ACCEPT f1,iolist]
ACCEPT *r,iolistl

Section 10.8

ASSIGN s to i

Section 8.3

BACKFILE un
BACKFILE (UNIT = unr ,ERR = ]r ,IOSTAT = io:; l)
BACKFILE (un[,ERR ,= s][JOSTAT = ios1)

Section 11.8.8

BACKSPACE un
BACKSPACE (UNIT = un[,ERR= s][,IOSTAT c: iosl)
BACKSPACE (unf,ERR=s][,IOSTAT=:iosj)

Section 11.8.4

BLOCK DATA [sub]

Section 13.1

CALL sub r( fa1 [,a2]. .. ])]

Section 13.4.2.2

CHARACTER[*LEN[,]]v[*LEN]r,vf*LENII

Section 7.2.2

CLOSE (doselist)

Section 10.16

COMMON [lrcb]/1nlistf[,]/[eb]/nlist1...

Section 7.4

COMPLEX v Lv ... ]

Section 7.2

CONTINUE

Section 9.4

DATA nhst/clistl [[,]nlist/clist/J...

Section 7.9

DECODE(c,f,a/,ERR= sll ,IOSTAT = ios])liolistl

Section 10.12

DIMENSION a(d) [,a(d) ... l

Section 7.1

s [,1

i = e1,e21,e31

Section 9.3.1

s r,1

WHILE (e)

Section 9.3.2

DOUBLE PRECISION v [, v ... ]

Section 7.2

ELSE

Section 9.2.4

ELSE IF (e) THEN

Section 9.2.4

ENCODE:(c,f,a[,ERR = s]1 ,IOSTAT = ioslHiolist

Section 10.12

END

Section 9.7

A-I

SUMMARY OF FORTRAN STATEMENTS
Section Reference

Form
~_:"-IJ)

Section 9.4

END IF

Section 9.2.4

ENDFILE un
ENDFILE (UNIT = un[,ERR = s][,IOSTAT = ios])
ENDFILE (un[,ERR = s][,IOSTAT = ios])

Section 11.8.5

ENTRY en [(d1 [,d2 ... ])]

Section 13.4.3

EQUIVALENCE (nlist) [,(nlist) ... ]

Section 7.5

EXTERN AL proc [,proc]. ...

Section 7.6

FIND (UNIT"'un.REC-"'rnl.ERR sll JOSTAT '" iosl)
FIND (un'rnl.l<:RR - sll.IOSTAT~ iosl)

Section 11.8.1

FORMAT (fs)

Section 12.1.1

fun ([arg1,arg2 ... argnJ)

Section 13.3.4

[typ 1 FUNCTION fun

aarg1 r,arg21. .. 1)

Section 13.3.2

GO TO i [[,](s [,s] ... )]

Section 9.1.3

GO TO s

Section 9.1.1

GO TO (s [,s]. ..)[,] e

Section 9.1.2

INCLUDE filespec. switch

Section 6.4.2

IF (e) st

Section 9.2.2

IF (e) sl, s2, s3

Section 9.2.1

IF (e) sl, s2

Section 9.2.3

IF (e) THEN

Section 9.2.4

IMPLICIT type (a [,a ... m,type (a[,a ... ]) ...
IMPLICIT NONE

Section 7.3

INQUIRE (FILE = fil,flist1)
INQUIRE ([UNIT =] u,ulistl

Section 11.7.1
Section 11.7.2

INTEGER v [,v ... ]

Section 7.2

INTRINSIC fun[ ,fun1

Section 7.7

LOGICAL v Lv ... ]

Section 7.2

;'\iAMELIST ifIameilistl!namc.lisil. ..

Section 12.7

OPEN (openlist)

Section 10.14

PARAMETER (p=c [,p=c ... ])
PARAMETER p=c[.p"--cl

Section 7.8

PAUSE [n]

Section 9.6

PHINT( FMT = fI.ERR - s II JOSTAT =, ins I Ii iolist I
I'RINT(FMT -;'I.ERR - sll.lOSTAT-c iosilliolistl

Section 10.10

PRINT n,iolist]
PRINT *f,iolist1
PROGRAM name

Section 6.4.1

F :NCH(FMT'- f1.ERR-slI.lOSTAT=iosilliolistl
P 'NCH(F!\lT",-,c"'I.ERR-slI.lOSTAT josilliolistl
" T!'JCH fI.iolistl
I' "NCII;'I.iohstl

Section 10.11

A-2

SUMMARY OF FORTRAN STATEMENTS
Form

Section Reference

READ(UNIT = un,FMT = fl ,END = s][,ERR = ~;][ ,IOSTA'T = ios l)[iolist]
READ(
un,FMT = fl ,END = s][ ,ERR = ~;]l,IOSTA 'T = ios])[iolist]
READ(
un,
fl ,END = s]l,ERR = s]l,IOSTA'T = ios])[iolist]
READ(UNIT = un,FMT = *f ,END = s]f ,ERR = .,][,IOSTAT = ios])[iolist]
un,FMT = *[ ,END = s][ ,ERR = ., H,IOSTAT = ios])[iolist]
READ(
READ(
un,
*LEND = s][,ERR = .,][,IOSTAT = ios])[iolist]
READ( UNIT = un,FMT = name I,END = slf ,ELR sH ,IOSTAT= ios 1)
READ(
un,FMT = namel ,END = s If ,ELR .=.c sll ,IOSTAT = ios])
un,
namel ,END = 8 ]f,Ef:,R -=c s 1f,IOSTAT -= ios]j
READ(
READ f[ ,iolist1
READ *[ ,iolist1
READ(UNIT = * ,FMT = fl ,END = s][ ,ERR = s]l ,IOSTAT = ios])[iolist]
READ(UNIT= *,FMT= *[,END = s][,ERR= s][,IOSTAT= ios])[iolist]
READ(UNIT = un[,END == sl[,ERR = s][,IOSTAT = ios])[iolist]
un[,END =:= s][,ERR = s][,IOSTAT = ios])[iolist]
READ(
READ(UNIT = un,FMT = f,REC = rn[,ERR = s ILIOSTAT = ios])[iolist]
READ(
un,FMT = f,REC = rnl,ERR = s II ,lOSTAT = ios])[iolist]
READ(
un,
f,REC = rn[,ERR = s][ ,IOSTAT = ios])[iolist]
un'rn.FMT = f
[,ERR = 811 ,10STAT =. iosllliolistl
READ(
READ(
un'rn,
f
[,ERR=sll,IOSTAT=:ios1)[iolistl
READ(UNIT = un,REC = rnl ,ERR = sJ[ ,IOSTAT = ios])[iolist]
READ(
un,REC = rnl,ERR = s][,IOSTAT = ios])[iolist]
READ(
un'rn
I,ERR=sl[,IOST.AT =ios])liolistl

Section 10.5

REAL v [,v ... 1

Section 7.2

REREAD(FMT = fl ,END" s][,ERR= sl[,IOST.\.T = ios1lliolist1
REREADWMT = *1 ,END·~ sll ,ERR -- f',11 ,IOST-\T ,= ios1j[iolist1
REREAD fl ,iolist1
REREAD *[ jolist1

Section 10.7

RETURN lel

Section 13.4.4

REWIND un
REWIND (UNIT = unl,ERR = s][ ,lOST AT = ios))
REWIND (unl,ERR = s]l,lIOSTAT = ios])

Section 11.8.2

SAVE la Lal ... l

Section 7.10

SKIPFILE un
SKIPFILE (UNIT = unl ,E RR = s ]I ,I OSTAT -'-' i )8 I)
SKIPFILE (unl.ERR - sll,IOSTAT - ios])

Spction 11.8.7

-0.

SKIPRECORD un
SKIPRECORD (UNIT =- lllli ,ERR -=- s 11 JOSTA r
SKIPRECORD (unl.ERR·'-. sll .lOSTAT= ios1)

Section 11.8.6
ios I)

STOP In]

Section 9.5

SUBROUTINE sub l( Id1 [,d21. .. l)]

Section 13.4.2.1

TYPEIF'MT= fl ,ERR - s]1 ,IOSTAT = i()sl)fioli~t1
TYPE(FMT = *[ ,ERR --' s II ,IOSTAT _oc i08 111 ioli:it I
TYPE f1 .iolist]
TYPE *I.iolist1

Section 10.9

v = e

Section 8.2

UNLOAD un
UNLOAD (UNIT = unl ,ERR = s 11 .lOSTA T ~ it·s 1)
UNLOAD (unl, ERR -- s II ,lOST A T = ios I)

Section 11.8.3

A-3

SUMMARY OF FORTRAN STATEMENTS
Form

Section Reference

WRITE(UNIT = un,FMT = f1 ,ERR = s]!,I OST A T = ios 1)1 iolist I
WRITE(
un,FMT = fl ,ERR-' sll ,10STAT = ioslHiolist]
WRITE(
un,
fl ,ERR - sJ[ ,10STAT = ios111iolist1
WRITE(UNIT= un,FMT= *I,ERR _c sH,IOSTAT-=-- ioslliiolistl
WRITE(
un,FMT = *1 ,ERR - sJ[ ,10STAT= ios])(iolistl
un,
*1 ,ERR = s]1 ,10STAT = ios])1 iolist 1
WRITE(
WRITE( UNIT =- un.FMT = namel ,ERR -=. s 1I ,10STAT = ios 1)
WRITE(
un,FMT = namel.ERR '-- sll ,10STAT= iosl)
WRITE(
un.
namel,ERR-sll.IOSTAT=-iosj)
WRITE fl ,iolist 1
WRITE*I.iolist1
WRITE(UNIT = *,FMT = fl ,ERR= sll ,IOSTAT = iosj)[iolistl
WRITE(UNIT = *,FMT = *1 ,ERR = sll ,10STA1' = iosj)[iolistl
WRITE(UNIT = unl ,ERR= s](,10STAT = ioslliiolistl
WRITE(
unl ,ERR = sll ,10STAT= iosl)liolist]
WRITE(UNIT = un,FMT = f,REC = rnl ,ERR = sJ[,10STAT = ioslHiolistl
WRITE(
un,FMT = f,REC = rn(,ERR·cc sl[ ,IOSTAT = ioslJ(iolistl
un,
f,REC - rnl ,ERR= sJ/ ,10STAT = ios111iolist1
WRITE(
WRITE(
un'rn,FMT'-'- f
I,ERR'"" sl/ ,10STAT:.;o ios]ll iolistl
WRITE(
un'rn,
f
I,ERR = 81/ ,10STAT = ioslliiolistl
WRITE(UNIT= un,REC -'- rnl,ERR=sll,lOSTAT= iosJ)liolistl
WRITE(
un,REC - rnl ,ERR = sll ,IOSTAT = iosJHiolist]
WRITE(
un'rn
[,ERR=slIJOSTAT""iosj)lioli::3tl

Section 10.6

A-4

APPENDIX B
ASCII-1968 CHARACTER CODE SET

The character code set defined
in
the
X3.4-l968 Version of the
Information
Interchange (ASCII) is
American
National
Standard
for
given in this appendix.

B-1

ASCII-1968

CHARACTER CODE SET

ASCII CODE
Control Characters
Even
Parity
Bit

7-Bit
Decimal

7-Bit
Octal

Character

000

NUL

001

SOli

Start of heading [SaM, start of message] . Control A.

002

STX

Start of text [EOA, end of address]. Control B.

003

ETX

End of text [EOM, end of message] . Control C.

004

EaT

End of transmission: shuts off TWX machines and disconnects some
data sets. Control D.

005

ENQ

Enquiry [WRU, "Who arc you?"]. Triggers identification ("lIere
is ... ") at remote station ifso equipped. Control E.

006

ACK

Acknowledge [RU, "Arc you ... ?"]. Control f.

007

BEL

008

010

Backspace. Control II.

009

011

liT

Horizontal tabulation. Control I.

010

012

Lf J

011

013

012

014

FE
FE
FE

Vertical tabulation. Control K.

VT 3
fF 3

013

015

Carriage return (to beginning of line). Control M.

014

016

015

017

016

020

OLE

017

021

DCl

Device control I, turns transmitter (reader) on. Control Q (X ON).

018

022

DC2

Device control 2, turns punch or auxiliary on. Control R (TAPE,
AUX ON).

Class l

Remarks
Null, tape feed. Control (Ii (control shift p2).

Bell (audible or attention signal). Control C.

Line feed. Control J.
Form feed (to top of next page). Control L.
Shi ft out; change character set or change ribbon color to red.
Control N.
Shift in; ret urn to standard character set or color. Control O.

Data link escape [DeO]. Control P.

019

023

DC3

Device control 3, turns transmitter (reader) off. Control S (X OFF).

020

024

DC4

Device control 4 (stop), turns punch or auxiliary off. Control T
(TAPE, AUX OFF).

021

025

NAK

Negative acknowledge [ERR, error]. Control U.

022

026

SYN

Synchronous idle [SYNC] . Control V.

023

027

ETB

024

030

CAN

025

031

End of medium [SI]' Control Y.

026

032

SUB

Substitute [S2]' Control Z.

End of transmission block [LEM, logical end of medium]. Control W.
Cancel [Sol. Control X.

027

033

ESC

028

034

File separator [S4]' Control \ (control shift L2).

029

035

030

036

Croup separator [Ss]' Control] (control shift M2).
Record separator [S 6]' Control ~ (control shift N2 ).

031

037

Unit separator [S?]. Control - (control shift 0 2).

Escape, prefix [S3]' Control [ (control shift K2).

1 CC communication control, FE format effector, IS information separator.
20n LT33, LT35 and similar units.
3 Includes a carriage return on some equipment, but not on standard DEC units.

8-2

ASCII-1968 CHARACTER CODE SET

Graphic Characters
Figures

Upper Case

Lower Case

Even
7-Bit 7-Bit
Parity
Bit Decimal Octal Character
032
040
SP
1
033
041
o
042
034
o
035
043
#
$
044
036
o
045
037
%
1
038
046
&
1
047
039
o
040
050
o
051
041
042
052
*
+
043
053
o
044
054
1
045
055
o
056
046
o
047
057
1
fIJI
048
060
o
049
061
1
2
1
050
062
3
063
o
051
4
052
064
1
5
053
065
o
6
054
066
o
7
067
055
070
8
1
056
057
071
9
o
058
072
o
073
059
060
074
o
<
075
061
062
076
>
077
063
o

Even
Parity
7-Bit 7-Bit
Bit Decimal Octal Character
064
100
(a
1
065
101
A
o
B
066
102
o
067
103
1
C
104
068
D
o
E
105
069
1
F
106
070
1
107
071
G
o
II
110
o
072
073
III
074
112
075
113
K
o
114
L
1
076
115
M
077
o
116
078
N
o
079
117
1
a
080
P
o
120
Q
081
121
082
122
R
123
o
083
S
124
1
084
T
125
o
085
U
V
o
086
126
W
127
1
087
1
088
130
x
y
089
131
o
o
090
132
133
1
091
[
\2
134
o
092
135
1
093
I
094
1
136
137
095
o

Even
Parity 7-Bit
7-Bit
Bit Decimal Octal Character 3
o
096
140
,2

o
I

o
o
o
o
I

o
I

o
1

o
o
I

o
I

o
o
1

o
1

o
o

097
098
099
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127

141
142
143
144
145
146
147
150
151
152
153
154
155
156
157
160
161
162
163
164
165
166
167
170
171
172
173
174
175
176
177

a
b
c
d

g
h
j
k

1
m
n

o
p
q

x
y

.........,2,5

I Zero-slash absent on many units.

2 Under study by

responsible American National Standards Committee for possible change at next revision of ASCII

(ca. 1982).
3 Codes 140-173 first defined in 1965. For a full ASCII character set the operating system accepts codes 140-176 as

lower case. For a program requiring a character ;;et that lacks lower case, the operating system translates input codes
140-174 into the corresponding upper case codes (100-134) and translates both 175 and 176 into 033, escape.
Early versions of the DECsystem-l0 Monitor l(sed 175 as the escape code and translated both 176 and 033 to it.
4 Unassigned

control character (usually ALT MODE) before 1965. Code generated by ALT MODE key on some DEC
units, especially earlier ones; on some more reccnt units, the ALT key generates the standard escape code, 033.

5 Control character ESC before 1965; code generated by ESC key on some DEC units designed at that time.
6 Delete, rub out (not part oflower case set).

8-3

ASCII-1968 CHARACTER CODE SET

Remarks on Special Graphic Characters

SP Space - normally nonprinting.

Question mark.

Exclamation point.

Quotation mark, diaeresis.

Number sign. £ on some (non-DEC) units.

Dollar sign.

Percent.

Ampersand.

Circumflex, upward arrow head. t before 1965,
but used until 1972 on DEC units.

Opening parenthesis.

Plus.

Underline, underscore. - before 1965, but used
until 1972 on DEC units.
Grave accent, opening single quotation mark.
(a 1965-67, but never on DEC units.

Comma, cedilla.

Opening brace.

Hyphen, minus.

Vertical line. Control character ACK before
1965;--'1965-67, but never on DEC units;
: in appearance 1968-19-77, but generally not
on DEC units.

Period, decimal point.
Slant, slash, solid us.

Closing brace. Unassigned control
(usually ALT MODE) before 1965.

Colon.

character

Overline, tilde, general accent. Control char·
after ESC before 1965; I 1965-67, but never
on DEC units.

Semicolon.

Reverse slant. ~ 1965-67, but never on DEC
units. Shift L on LT33, LT35 and similar units.
Closing bracket. Shift M on LT33, LT35 and
similar units.

Closing parenthesis.
Asterisk.

Commercial at. ' 1965-67, but never on DEC
units.
Opening bracket. Shift K on LT33, LT35 and
similar units.

Apostrophe, closing single quotation mark,
acute accent. " in appearance on some DEC
units.

Greater than.

Less than.
Equals.

8-4

APPENDIX C
COMPILER MESSAGES

The FORTRAN compiler issues tvlO tYPE~S of messages: warning and fatal
error.
While compiling thE! progrram,
if the compiler encounters a
situation that does not prevent it from completing the compilation,
but does warrant your knowing about, it prints a warning message and
continues compilation.
If, however ,. the problem in your program is
such
that
compilation cannot continue, the compiler prints a
fatal-error message and stops compilation of the program. Whenever a
fatal error is generated a relocatable object module will not be
prod uced •
Compiler messages are printed in the following form:
?FTNxxx LINE:n text
or
%FTNxxx LINE:n flag: text
where:
?

indicates a fatal message
indicates a warning message
FTN
is the FORTRAN mnemonic
xxx
is the 3-letter mnemonic for the error message
Line:n is the optional ljne number where the error occurred
text
is the explanation of error
flag:
is the prefix for warning messages generated when the
compatibility flagger is invoked.
This prefix describes
the type of incompatibility the message refers to, and can
be one of the fol]owing:
~;

•

Extension to Fortran-77:

•

VMS incompatibility:

•

F'ortran-77 and VMS:

See Section 16.6 for more information on the compatibility flagger.
Square brackets ([ ]) in this appendix signify variables and
output on the terminal.

C-l

are

not

COMPILER MESSAGES
Fatal Errors
Each fatal error in the following list is preceded by ?FTN on the user
terminal and on listings.
They are presented here in alphabetical
order.
ABD

[symbolname] has already been defined [definition]
The usage given conflicts with current information about the
symbol.
For example, a symbol defined in an EQUIVALENCE
statement cannot be referenced as a subprogram name.

AHE

IF' at line [number] already has ELSE

AOA

Assumed-size arrays only allowed in subprograms

ASA

Assumed-size arrays cannot be used

ATL

ARRAY [name]

too large

The total amount of memory necessary
array is greater than 5l2P.
AWN

accommodate

Array reference [name] has wrong number of subscripts
The array was defined to have more or fewer dimensions
the given reference.

BOV

this

than

Statement too large to classify
To determine statement type, some portion of the statement
must be examined by the compiler before actual semantic and
syntactic analysis begins. During this classification, the
entire portion of the required statement must fit into the
internal statement buffer (large enough for a normal 20-line
statement) •
This error message is issued when the portion of a given
statement required for classification is too large to fit in
the buffer. Once FORTRAN-IO/20 has classified a statement,
there is no explicit restriction on its length.

CER

Compiler error in routine [name]
Submit an SPR for any occurrence of this message.

CEL

Character entry points must have the same length

CEN

Character expression used where numeric expression required

CEX

Constant or constant expression required

CFF

Cannot find file
The file referenced in an INCLUDE statement was not found.

CFL

Reference to character function [name]
length

CNE

Character and numeric entry points cannot be mixed

CPE

Checksum or parity
[ name]

error

C-2

must

have

fixed

[source/listing/object]

fil e

COMPILER MESSAGES
~haracter

CQL

No closing quote in

CSA

Can't split string

CSF

Illegal statement function reference in CALL statement

DDA

[symbolname]

DDN

DO loops too deeply nested -

DFC

Variable dimension [name] must be scalar, defined as
or in COMMON

DFD

Double [type]

numeric and character variables

is duplicate dummy argument
reduce nesting
formal

name illegal

Duplicate
fields
specification.
DIA

a~ross

constant

were

encountered

DO index variable [name]

INCLUDE

file

is already active

In any nest of DO loops, a given index variable may
defined for more than one loop.
DID

Cannot initialize a dummy parameter in DATA

DLN

Optional data value list not supported

not

The extended FORTRAN statement form that allows data
to be defined
in
type
specification statements
supported by FORTRAN-IO/20.
specification

without

associated

values
is not

DNL

Implied
DO
variables

list

DPR

Dummy parameter [name]

DSF

Argument [name]

DTI

The dimensions of [arrayname]

DVE

Cannot use dummy variable in EQUIVALENCE

ECS

[variable]
statement

ECT

Attempt to enter [symbolname]

EDN

Expression too deeply nested to compile

EID

ENTRY statement illegal inside a block IF or DO loop

ElL

Expression illegal in an input list

ElM

ENTRY statement illegal in main program

ENF

Label [number] must refer to an executable statement, not
FORMAT

ETF

Enter failure [number]

EXB

EQUIVALENCE extends COMMON block [name]

FEE

Found [symbol]

referenced before definition

is same as FUNCTION name

EQUIVALENCE-d

must be of the type integer

COMMON

illegal

SAVE

into COMMON twice

[filespec]
backwards

when expecting either [symbol]

General syntax error message.

C-3

[symbol]

COMPILER MESSAGES
FER

[file error text]
An
error
has
specification.

occured

when

processing

command

FID

Can't initialize character function name

FNE

Label [number] must refer to a
statement

FWE

Found

IAC

Illegal ASCII character [character]

IAL

Incorrect argument type for library function

IAN

Illegal assignment between character and numeric data

IBD

Illegal substring bound in DATA statement

IBK

Illegal statement in BLOCK DATA subprogram

ICL

Illegal character

ICN

Illegal combination of character and numeric data

IDN

DO loop at line:

[symbol]

when expecting

[character]

[ n urn be r ]

FORMAT,

not

file

executable

[symbol]
in source
[name]

in label field

i s i l l eg all y n est e d

You are attempting to terminate a DO loop before terminating
one or more loops defined after the given one.
IDS

Implicit DO indices may not be subscripted

IDT

Lllegal or misspelled data type

IDV

Implied DO index is not a variable

IED

Inconsistent EQUIVALENCE declaration
The given EQUIVALENCE declaration would cause some
name to refer to more than one physical location.

IFD

INCLUDEd files must reside on disk

IFE

[INCLUDE file error]
This error occured
while
trying
INCLUDE file on the DECSYSTEM-20.

open

IFS

Illegal format specifier

I ID

Non-integer implied DO index

IIP

Illegal implicit specification parameter

IIS

Incorrect INCLUDE switch

ILF

Illegal statement after logical IF

the

symbolic

specified

Refer to Section 9.2.2 for restrictions on logical IF object
statements.
IND

Improper nesting:

DO at line [number]

C-4

has not terminated

COMPILER MESSAGES
INI

Improper nesting:

]"F at line [number]

has not terminated

INN

INCLUDE statements may not be nested

IOC

Illegal operator fOl character data

IOD

Illegal statement used as object of DO

ION

Numeric operand of concatenation operator

IOR

Substring bound out of range

IQB

INQUIRE -

IQN

INQUIRE - neither

ISD

Illegal subscript expression in DATA statement

both UNIT and FILE keywords were specified
U~IT

nor FILE keywords were specified

Subscript expressions may be formed only
indexes and constants combined with +, ISN

[symbolname]

with

implicit

*, or I.

is not [symbol type]

The symbol cannot be used in the attempted manner.
ISS

[variable]

ITL

Illegal transfer into loop to label

IUT

Program units may not be terminated within INCLUDEd files

IVC

Invalid character constant

IVH

Invalid hollerith constant

IVP

Invalid PPN

IXM

Illegal mixed mode arithmetic
Complex and
expression.

illegal in SAVE statement

double-precision

[OPEN specifier]

[number]

cannot

appear

the

IXS

Illegal

IZM

Illegal [datatype]

FORTRAN will not run on a KA

KAS

FORTRAN can not compile for a KA

LAD

Label

LED

Illegal list directed [statement type]

LFA

Label arguments illegal in FUNCTION or array reference

LGB

Lower bound greater than upper bound for array [name]

LLS

Label too large or too small

[number]

same

specifier

size modifier [number]

al ready defined at line:

[number]

Labels cannot be 0 or greater than five digits.
LND

Label
[number]
declaration

must

refer

C-5

[statement] ,

not

COMPILER MESSAGES
LNI

List directed I/O with no I/O list

LTL

Too many items in list -

reduce number of items

In rare instances, a combination of long lists in
statement can exhaust the syntax stack.

single

formal

label

MCE

More than 1 COMMON variable in EQUIVALENCE group

MSP

Statement name misspelled

MST

[OPEN specifier] must be [integer or array]

MWL

Attempt to
arguments

NCC

Can't store numeric constant in character variable

NCF

Not enough core for
[number]

NEX

No exponent after D or E in constant

NFS

No filename specified

define

multiple

the

file

RETURN

without

specs.

Total

needed

The INCLUDE statement requires a filename.
NGS

Cannot get segment [segment name] code]

error code

[GETSEG error

This message means the system is unable to GETSEG one of the
compiler segments on the DECsystem-lO.
NIF

No matching IF

NIO

NAMELIST directed I/O with I/O list

NIR

Repeat count must be an unsigned integer

NIU

Non-integer unit number in I/O statement

NLF

Wrong number of arguments for library function [name]

NLS

[variable] may not be declared length star

NMD

No matching DO

NNA

NAMELIST not allowed in ENCODE, DECODE, and REREAD

NNF

No statement number on FORMAT

NNN

NML= must specify a NAMELIST

NRC

Statement not recognized

NUO

.NOT.

NWB

Numeric variable must be aligned on word boundary

NWD

Incorrect use of

NYI

Not yet implemented

OAG

Octal or logical argument illegal to generic function

is a unary operator

* or?

in [filespec]

C-6

COMPI LER fJIESSAGES
OBO

[variable] may only be specified once

OPW

OPEN/CLOSE parameter" [name]

OUB

Only upper bound of last dimension
asterisk

PD6

FORTRAN will not rurl on PDP6

PIC

The DO parameters of" [index name] must be integer constants

PN4

Project number must be 4 in ppn

is of wrong type

This error is for DI:CsystE!m-IO file
DECSYS'I'EM-20.
PRF

Pro t e c t ion fa i I u rei n urn b e r ]

PTL

Program too large
The program unit

ta~es

[arrayname]

may

specifications

the

real

[ f i I e s pe c ]

up more than 5l2P.

QEF

Quota exceeded or disk full

RDE

Rib

RFC

[function name]

RIC

Complex constant cannot be used to represent
imaginary part of a complex constant

RUS

Relational expression illegal as UNIT specifier

SAD

Array [name] - signed dimensions may appear only as constant
range limits

SMC

Size modifier conflict for variable [name]

SNC

Substring of non-character variable

SNL

[statement name]

SOR

Subscript out of range

STD

Statement [number]

TDO

[symbol type]

TFL

Too many FORMAT labEls specified

UCE

User core exceeded at location [address]
while processing statement [number]

UEC

Label [number]

previously used in executable context

UFC

Label [number]

previously used in FORMAT context

UKW

Unrecognized keyword

UMP

Unmatched parentheses

UNS

UNIT may not be specified

0 r

[number]

d ire c tor y err 0 r [ n urn be r ]
is

[filespec)

[ f i I e s pe c ]

recursive function call
the

statements may not be labeled

is a declaration

type declaration out of order

C-7

in phase

[segment]

COMPILER MESSAGES
USI

[symbol type]

[symbol name]

used incorrectly

The given symbol cannot be used in this way.
VNA

Subscripted variable in EQUIVALENCE, but not an array

VSE

EQUIVALENCE subscripts must be integer constants

VSO

Variable dimension allowed in subprograms only

WIF

[I/O type]

ZLD

Zero-trip DO loop illegal in DATA statement

is illegal with internal files

Warning Messages
Each warning message in the following list is preceded by %FTN on the
user
terminal
and
on listings.
They are
presented
here
in
alphabetical order.
ACB

Argument out of range of CHAR, high order bits ignored

ADS

Variable [name]

AGA

Opt - object variable, of
list, was never assigned

AIL

Illegal length argument for ICHAR, first character used

AIS

Extension to Fortran-77:

ANS

VMS incompatibility: ASSOCIATEVARIABLE not
OPEN

CAl

COMPLEX expression used in arithmetic IF

CAO

Consecutive arithmetic operators illegal

CAP

Extension to Fortran-77: Consecutive arithmetic operators

CCC

Fortran-77 and VMS: Carriage control character

CCN

CHARACTER constant used where numeric expression required

CIS

Conflicting INCLUDE switches

CNM

Character and numeric variables mixed

CNS

Extension to
Fortran-77:
non-specific length

COS

Extension to Fortran-77: Comment on statement

COV

Extension
to
Fortran-77:
character expression

CSM

Extension to Fortran-77: Comma field separator is missing

already declared in SAVE statement

C-8

assigned

GOTO

without

optional

Apostrophe in I/O specifier

Concatenation

Assigned

set

with

variable

VMS

variable

appears

COMPILER MESSAGES
CTR

Complex terms used

in a relational other than EQ or NE

The result of the other relational
operands is undefined.
CUO

operators

with

complex

Constant underflow or overflow
This message
is
issued when overflow or
underflow is
detected
as the result of building constants or evaluating
constant expressions at compile time.

DEB

Extension to Fortran-77: DEBUG lines

DFN

VMS incompatibility: Default file name on VMS
Fortran-IO/20

011

Previous declaration of intrinsic function is ignored

DIM

possible DO index modification inside loop

differs

from

A program that does this may be incorrectly compiled by the
optimizer, since it assumes that indexes are never modified.
Note that the number of
iterations
is calculated at the
beginning of the loop and is never affected by modification
of the index within the loop.
DIS

Opt -

program is disconnected - optimization discontinued

Submit an SPR if this message occurs.
DOW

Extension to Fortran-77: DO WHILE statement

OPE

VMS incompatibil i ty: Di ffE~rent precedence in exponentiation

OWE

Fortran-77 and VMS:
[descriptor]

OWL

Extension to Fortran-77: DO without statement label

DXB

DATA st.atement

EDD

Extension to Fortran-77: END DO statement

EDS

Extension to Fortrall-77: DECODE statement

EDS

Extension to Fortran-77:

EDX

Fortran-77 and VMS: FORMAT edit descriptor [descriptor]

EOC

Fortran-77 and VMS: Exponential operator

EXD

Extension to Fortran-77: Transfer of control into DO loop at
label [label]

FAR

Extension to Fortran-77: F'ormat in numeric array

FIF

Extension to Fortran-77:
function in Fortran-77

FIN

Extension to Fortran-77: FIND statement

FMR

Multiple RETURNs defined in a FUNCTION

FMT

VMS incompatibility: Keyword FMT instead of NML

Default

exceE~ds

widths

with

edit

descriptor

bounds of array [name]

ENCODE statement

[function name]

C-9

is not an intrinsic

COMPILER MESSAGES
FNA

A function without an argument list

FNG

Extension to Fortran-77:
function in Fortran-77

FNS

Extension to
subroutine

FOO

Statement function
dimensioned

HCP

VMS incompatibility: Hollerith constant padded with spaces

HCN

Hollerith constant used where numeric expression required

HCU

Extension to Fortran-77: Hollerith constant

IAT

Illegal type for argument [number]

ICC

Illegal character, continuation field of initial line

[function name]

Fortran-77:
declared

not

generic

[name]

not

FORTRAN-77

out

order

array

not

for statement function

Continuation lines cannot follow comment lines.
ICD

Inaccessible code.

Statement deleted

The optimizer will delete statements that cannot be
during execution.
ICS

Illegal character in line sequence number

ION

Opt -

illegal DO nesting -

reached

optimization discontinued

A GO TO within a DO loop goes to the ending statement of an
inner, nested DO loop.
The line number printed out with the
warning message is that of the OUTER DO.
DO

GO TO

CONTINUE

IFL

Opt -

infinite loop.

Optimization discontinued

IMN

IMPLICIT NONE

INC

Extension to Fortran-77:

INS

VMS incompatibility: /NOCREF switch

INS

VMS incompatibility: /CREF switch

INS

VMS incompatibility: Default for VMS is /NOLIST

INCLUDE statement

C-IO

COMPILER MESSAGES
IUA

Illegal use of an array - use scalar variable instead

KIS

Obsolete switch /KI

KWU

Fortran·-77 and VMS: Keyword [keyword name]

KWV

Fortran-77 and VMS: Keyword value for

LID

Identifier [name] more than six characters

[keyword name]

The remaining characters are ignored.
LNC

Fortran·-77 and VMS: Non-numeric operand in numeric context

LNE

VMS
incompati bi 1 i ty:
EQUIVALENCE-d

LOL

VMS incompatibility: List .of labels

LSP

Extension to Fortran-77:

MBD

IMPLICIT NONE -

MLN

Fortran·--77 and VMS: Mixing logical and numeric

MSL

Fortran-77 and VMS: Multi-statement line

MVC

Number of variables
[is less than/is
numbers of constants in DATA statement

NAM

Extension to Fortran-77: NAMELIST statement

NDP

Fortran-77 and VMS: No decimal places with edit descriptor

NEC

Extension to Fortran-77:
context

NED

No END statement in program

NIB

Extension to Fortran-77: Non-integer substring bounds

NIG

Extension to Fortran-77: Non-integer as
GOTO

NIK

Extension to Fortran-77: Non-integer used with [keyword]

NIS

Extension to Fortran-77: Non-integer subscript

NIX

Extension to Fortran-77: Non-integer as index to RETURN

NLC

Fortran·-77 and VMS: Non-lo9ical operand in logical context

NLK

Extension to Fortran-77: Use of NAMELIST

NOD

Global optimization not supported with /DEBUG -

NOF

No output file given

NPC

VMS incompatibility: Null padding before [symbolic name]

NPP

Extension to Fortran-77:
list

NSC

Fortran-77 and VMS: Non-standard character in column 1

L.og ical

and

[data type]

numer ic

variables

length specifier

[variable] must be explicitly declared

Numeric

C-ll

greater

expression

parentheses

index

than]

the

character

computed

/OPT ignored

around

PARAMETER

COMPILER MESSAGES
OCU

Fortran-77 and VMS: Octal constant

OHC

Octal or hexadecimal constant

010

Extension to Fortran-77:

010

Fortran-77 and VMS:

PAV

PARAMETER used as associative variable

PLP

PARAMETER list must be enclosed in parentheses

PPS

PROGRAM statement parameters ignored

[statement name]

statement

Used for compatibility purposes.
PSR

Pound sign (#)
instead

PWS

Fortran-77 and VMS: PRINT (Specifiers)

RDI

Attempt to redeclare implicit type

RIM

RETURN statement illegal in main program

RLC

Extension to Fortran-77: & used with return label

RLC

Fortran-77 and VMS:

$ used with return label

RLX

Fortran-77 and VMS:

Return label [label]

SBC

Extension to Fortran-77: Substring bounds not constant

SEP

VMS incompatibility:
name or entry point

SID

Slash (I)

SMD

Extension
to
Fortran-77:
Single
multi-dimensioned array [array name]

SNN

VMS incompatibility:
name

SOD

[name]

SOR

Fortran-77 and VMS: Subscript out of range for array
name]

SPN

VMS incompatibility:
PARAMETER name

SRO

Fortran-77 and VMS: Symbolic relational operator [operator]

SVN

VMS incompatibility:
function name

TLF

Fortran-77 and VMS:

TSI

Type of symbolic constant ignored

VAl

[name]

in random access - use REC= or apostrophe (')

[symbolic name]

in dimension bound -

statement

is the same as

use colon (:)

[symbolic name]

program

instead
with

subscript

is the same as NAMELIST

statement out of order

[symbol ic

name]

[symbolic name]

C-12

same

is same as variable or

Two-branch logical IF

already initialized

the

[array

COMPILER MESSAGES
VFS

VMS incompatibility:
routine on VMS

[function name]

VGF

VMS incompatibility:
on VMS

[function name]

is a

VIF

VMS incompatibil i ty:
function on VMS

[function

VND

FUNCTION return value is never defined

VNF

VMS incompatibility:
function on VMS

[function name]

VNG

VMS incompatibility:
f unct ion on VMS

[func t i on

VNI

Opt -

variable [name]

name]

Fortran-supplied
generic

function

intrinsic

name]

not
is

not

generic

is not initialized

The optimizer analysis determined that
the given variable
was never initialized prior to its use in a calculation.
VNS

VMS incompatibility:
subroutine

VSD

VMS incompatibility: Subroutine [subroutine name]

WDU

Fortran·-77 and VMS: wRITE wi th defaul t

WNA

Wrong number of

WOP

Opt - warnings given in phase
incorrect

[subroutine name]

argu~ents

is not a VMS-supplied
may differ

uni t

for statement function
1.

Optimized

code

may

One or more of the messages issued
prior
to
this message
resulted
from
situations that violate assumptions made by
the optimizer, and thus may cause it to generate code
that
does not execute as 1esired.
XCR

Extraneous carriage return
Carriage return was not immediately preceded or followed
a line termination character.

XEN

Fortran--77 and VMS:

XOR

Extension to Fortran-77: Logical

ZMT

Size modifier

[* or ,&]

[number]

with external name
.XOR. operator

treated as [data type]

This message is issued
when one of the data
type
size
modifiers that is accepted only for compatibility is used.
Internal Compiler Errors
An internal compiler error is an attempt by either the compiler or the
monitor
to document an
err~r
inside
the
FORTRAN compiler.
An
occurrence of an internal compiler error signifies that
something
is
wrong with the FORTRAN compiler.

C-13

COMPILER MESSAGES
Monitor-detected internal errors are of the form:
?
?
?

Internal compiler error
[message] at location [address] in phase [segment]
While processing statement [line-number]

where [message]

can be one of the following for TOPS-IO:

Illegal memory reference
A read or write was attempted to a non-existent page
Stack exhausted
Monitor detected PDL overflow
Memory protection violation
Illegal reference to high segment
or where [message] can be one of the following for TOPS-20:
Illegal memory reference
A read was attempted to a non-existent page
Non-existent memory write
A write was attempted to a non-existent page
Illegal memory read
A memory read failed
Illegal memory write
A memory write failed
Stack exhausted
Monitor detected PDL overflow
Compiler-detected errors are of the form:
? Internal compiler error-processing statement [line-number]
? Call to [routine-name] from [address]
Submit an SPR if you receive an internal compiler error.
At the end of program unit compilation, the compiler prints
summary line, which is one of the following:
[ No error detected ]
%FTNWRN
no warning(s)
%FTNWRN
[warning count] warnings(s)
?FTNFTL
[fatal count] fatal error(s)
?FTNFTL
[fatal count] fatal error(s)
warning (s)

C-14

and no warning(s)
and [warning count]

error

APPENDIX D
FOROTS ERROR MESSAGES

Errors detected
categories:

run

time

FOROTS

fall

into

the

following

System errors (SYS) - errors internal to FOROTS

Open errors (OPN) and CLOSE

I/O errors that occur during a file

Arithmetic
fault
calculations

errors

Library errors (LIB) routines

Data errors (DAT) - errors in data conversion on I/O

Device

Compatibility errors
compatibility flagger

I~rrors

(DEV)

(APR) -

errors

generated

OPEN

numer ic

FORLIB

library

- I/O hardware errors
(COM) -

errors

generated

the

The messages generated by FOROTS contain the following elements:
1.

A 3-letter code that identifies the type of message
only)

The message
encountered

For I/O errors, two integer values which are retrieved by the
ERRSNS subroutine

For compatibility errors, a prefix precedes the message
that
describes the type of incompatibility the messages refers to;
one of the following:

itself,

which

•

Extenstion to FORTRAN-77:

•

VMS incompatibility:

•

FORTRAN-77 and VMS:

describes

See Section 16.6 for more information
flagger ..

D-1

what

the

(TOPS-IO

FOROTS

has

compatibility

FOROTS ERROR MESSAGES
The 3-letter code (TOPS-I0 only) and
the message are,
by default,
printed at your terminal when an error occurs; the two ERRSNS values
are stored within the arguments you have supplied
for
the
ERRSNS
subroutine.
If you do not include a call to the ERRSNS subroutine in
your program, your program cannot have access
to
the
two
ERRSNS
values.
(For
instructions on how to use the ERRSNS subroutine, see
Section 13.4.1.15.)
Table D-l contains a list of all the 3-letter message
ERRSNS values that are generated by FOROTS.
Table D-l:
1st
Value

codes

and

the

FOROTS Error Codes
2nd
Value(5)

Code(l)

Meaning
-

No error detected
No error detected

0
0
IDC ( 3 )

Arithmetic trap
Integer divide check

IOV (3 )

Input Conversion Error
Integer overflow

n (7)

FOV (3 )

Input Conversion Error
Floating overflow

n (7 )

FOV (3 )

Arithmetic trap
Floating overflow

n (7)

FDC ( 3 )

Arithmetic trap
Floating divide check

n (7 )

FUN

Arithmetic trap
Floating underflow

n (7 )

FUN (3)

1
n

(7 )

---

2
n (7)

---

3
4

5
6
(3)

Input Conversion Error
Floating underflow
---

Output Conversion Error
field
width
too
Output
small

FTS

(3 )

21
104

IDU

105
106

UNO
NOF

107

CWU

108

CLE

109

ICE

110

NCS

FORLIB errors and warnings
DIVERT:
illegal to divert
to unit
DIVERT:
unit not open
DIVERT:
unit not open
for
formatted I/O
DIVERT:
Can't write
to
unit
Concatenation result longer
than expected
Illegal
length
character
expression
stack
No
character
allocated

D-2

FOROTS
Table D-l:
1st
Value

ERROI~

MESSAGES

FOROTS Error Codes (Cont'd)
2nd
Va 1 ue (5)

Code (1)

Meaning

112

AQS

113
114
11 ~;

SSE (3)
SRE (3)
TMA

116

CGP

117

CRP

(1)

lIE:

NSS

(2)

119
120
121

CFS
CGS
CPP

(2)
(2)
(3)

122
123

IPN
CCS

124

ECS

First argument of SORT must
be a quoted string
Substring range error
Subscript range error
Too many arguments in call
to SORT
Can't get pages 600:677 for
SORT
Can't return pages 600:677
after call to SORT
No free
section available
for SORT
Can't find SYS:SORT.EXE
Can't get SYS:SORT.EXE
Can't
preallocate
pages
600:677 for SORT
Illegal page number
Not
enough
memory
for
creating character stack
Not
enough
memory
for
expanding character stack
Argument less than or equal
to zero
Must give lower and
upper
bounds
to dump in non-zero
sections

ALZ

126

DMA (2,3)

509

ETL ': 3)

532
583

ARC ': 3)
FVM (3)

584

RIF (3)

590

DQW (2,3)

I/O warnings
Attempt to WRITE
beyond
fixed-length record
Ambiguous repeat count
Format and variable type do
not match
Reading
into
FORMAT
non-standard
Disk full or quota exceeded
- Please EXPUNGE, then type
CONTINUE

113
114

SSE f 3)
SRE f 3 )

FORLIB
bounds
check
warnings
Substring range error
Subscript range error

-1

EOF

End of file
End of file

24
25

302
510

BBF
RNR

512
517

IRN
RTL

536

CBI

536

CSI

Record or record
number
error
Bad format binary file
Attempt to
read a
record
that has not been written
Illegal record number
Record too large memory
full
Cannot backspace image file
with no RECORDSIZE
Cannot
skiprecord
image
file with no RECORDSIZE
D-3

FOROTS ERROR MESSAGES
Table D-l:
1st
Value

FOROTS Error Codes (Cont.)
2nd
Val ue (5)

Code(l)

Meaning

570

ICD

572

RSM

573

FCL

576

WBA

577

SLN

Non-digit
in
record
delimiter
Record size different
from
that specified in OPEN
Unexpected
continuation
LSCW found
Attempt
to WRITE beyound
variable or array
Record length negative or
zero

~------+------~------~------------------.---I

26
502

CSF (2,3)

535

BS I

541

UOA (3)

542

NCK (3)

550

CQF (1,3)
OGX (1,3)

595

(3 )

OPEN/CLOSE warnings
Can't set FORTRAN carriage
control attribute
BLOCKSIZE ignored:
device
is not a magnetic tape
Unknown
OPEN
keyword,
ignored
OPEN-only keyword specified
in CLOSE, ignored
Cannot QUEUE file
Galaxy
version
2
not
supported

~-------+--------~r-'----------4-------------------------------~

28
J
J

250+n
250+n
250+n

C LF

(2)

RNM
CLS
DEL
RNM

(1)
(1)

CLOSE error
Cannot CLOSE file
Cannot RENAME file
"Close" FILOP. error n (4)
"Delete" FILOP. error n (4)
"Rename" FILOP. error n (4)

(2)

(1)

OPEN error
Cannot set up to append
to
magnetic tape file
240
FRR
/RECORDTYPE:FIXED requires
/RECORDSIZE
240
RRI
Random
I/O
requires
RECORDSIZE
specifier
in
OPEN statement
240
RRR
Random
I/O
requires
/RECORDSIZE
242
NFC (1)
Too many OPEN units
243
CIR
/CARRIAGECONTROL:TRANSLATED
illegal
with
this
/RECORDTYPE
244
RLB
/RECORDSIZE
larger
than
/BLOCKSIZE
245
NSD
No such device
248
lAC
Specified ACCESS
illegal
for this device
249
IDM
Specified MODE
is
illegal
for this device
250+n
OPN (1)
Cannot OPEN file
405
PPN (2)
JSYS error - PPN cannot be
translated
503
CEF (2)
End
of
command
file
_ _ _ _ _ _- i_ _ _ _ _ _L -_ _ _ _ _ _ _ _L -_
___________________________
encountered
J

APP (2)

D-4

FOROTS ERROR MESSAGES
Table 0-1:

FOROTS Error Codes (Cont.)
··----r-

1st
Value

2nd
Va 1 ue (5)

Code(l)

Meaning

506
540

ICA
SDO

569

TFM

585

IAV

Incompatible attributes
Same device open on another
unit
with
conflicting
spec i fiers
Tape format conflicts
with
OPEN statement or default
Illegal value
for
OPEN
spec i fier

31 ~)

CDR

31 ~)

CDS

593

POI

594

CDF

239

IUN

321

CFC (3)

322

CFF (3)

323

CFR

(2,3)

323

CFX

(2,3)

323

CFR

(1,3)

323

CFX

(1,3)

32~1

CFK

(3)

32~i

CFT

(3)

326

CFO (3)

327

CFG

310

RBR

Mixed ACCESS modes
Can't do direct
I/O
to
sequential file
Can't do sequential I/O to
direct file
Illegal for DIRECT (RANDOM)
files
Can't
determine
whether
formatted or unformatted

Illegal logical unit number
Illegal unit number

(3)

Compatibility error
FORTRAN-77
and
VMS:
Carriage control character
VMS
incompatibility:
Intrinsic
routine
invoked
incompatibly
FORTRAN-77
extension:
FORTRAN-20 supplied routine
invoked
FORTRAN-77
and
VMS:
FORTRAN-20 supplied routine
invoked
FORTRAN-77
extension:
FORTRAN-IO supplied routine
invoked
FORTRAN-77
and
VMS:
FORTRAN-IO supplied routine
invoked
FORTRAN-77 and VMS: Keyword
[keyword]
FORTRAN-77
and
VMS:
Trailing
spaces
in output
record
extension:
FORTRAN-77
Overlap
of
character
assignments
FORTRAN-77:
and
VMS:
G
format descriptor used with
character
REREAD error
REREAD not
proceeded
READ

FOROTS ERROR MESSAGES
Table D-l:
1st
Value

FOROTS Error Codes (Cont.)
2nd
Val ue (5)

Code(l)

Meaning

OPEN/CLOSE statement syntax
errors
Unknown
or
ambiguous
keyword
Unknown swi tch
Ambiguous switch
Dialog string too long
Error parsing DIALOG string
No device specified with

241

ESV

241
241
533
539
544

USW (1)
ASW (1)
DLT
EDS/EDA (2)
NDI (1)

545
546
547
548

IPP
TMF
NS I
IDD

(1)
(1)
(1)
(1)

Illegal PPN
Too many SFDs
Null SFD
Illegal character in DIALOG
str ing

551

NQS

(1)

PADCHAR must
be
single
character in double quotes

263

CDT

554

CWL

301
306

ILF
DLF

524

RIC

538
552
553
575

IRC
IHC
IFW
UDT

583

FVF

307

ILC

501
508
574
579
581
582

UNS
IOL
IMV
I DI (1 )
DLL (1)
IWI

599

ICE

.. : "

WRITE on READ-only file
Cannot WRITE to
READ-only
file
Cannot write a
file with
MODE=LINED

Syntax error in FORMAT
Illegal character in FORMAT
Data in I/O list but not in
FORMAT
Reading
into
character
format illegal
Illegal repeat count
Illegal Hollerith constant
Illegal field width
Undefined data
type
or
internal FOROTS error
Format and variable type do
not match

Input conversion error
Illegal character in data

FOROTS calling errors
Unit not specified
Bad I/O list
Illegal MTOP value
Illegal DUMP mod~ I/O list
Dump mode I/O list too long
Illegal to initiate another
I/O statement
Illegal
length
for
character expression

D-6

FOROTS ERROR MESSAGES
Table 0-1:
1st
Value

FOROTS Error Codes (Cont.)
2nd
Va 1 ue (5)

Code(l)

Meaning

530
537

I LM
UMO
UTE
UTO

537

UTO (1,3)

587

ITE (1)

309
513

VNN
NEQ

514
515

NRP
ILN

516
519

I LS
CCC

521
522

RPE
SNV

580
596

NLS
NEC

597

ISS

598

SNQ

J
J

(2)

(2,3)
(1)

(1)

NAMELIST data errors
Variable not in namelist
"=" not found in namelist
data
Missing right paren
Variable or namelist does
not start with letter
Illegal subscript
Cannot convert constant to
correct type
Illegal repeat count
Sign with null value
Null string illegal
Found
character
when
expecting ":"
Substring
descriptor
illegal
String not within single
quotes

98
J
J

250+n
250+n
400
586
590

ISW
IOE
OSW
ISW
OSW
IOE
MFU
DQE

Error
in
magnetic
tape
operations
Unexpected MTOPR% error (2)
Error trying to set tape
Unexpected TAPOP. error
Unexpected
TAPOP.
error
trying to set parameters
Unexpected
TAPOP.
error
trying to set parameters
Tape is not usable by your
job

Unclassifiable
device
errors
Cannot switch to input
General purpose I/O error
Cannot switch to output
Cannot switch to input
Cannot switch to output
General-purpose I/O error
Memory full
Disk full or quota exceeded

(2)
(2)

(2)
(1)
(1)
(1)

(2)

No ERRSNS values
B LK

(3,6)

CCP
CDP
CGD
DBM
DST

(1,6)
(1,6)
(6)
(6)
(1,6)

D-7

Blocks allocated
but
deallocated
Cannot create page
Cannot destroy page
Can't get DBMS
DBMS not loaded
Error in dialog string

not

FOROTS ERROR MESSAGES
Table D-1:
1st
Value

FOROTS Error Codes (Cont.)
2nd
Val ue (5)

Code(l)

Meaning

EFS (6 )
FFX (6 )

Enter correct file specs
FOROP function code exceeds
range
Error in memory management
"Impossible" JSYS error
Memory full
Error number out of range
Pages allocated
but
not
deallocated
Deallocating
more
pages
than allocated
PDL overflow
Internal FOROTS error
Trap occured during
trap
processing

IEM
IJE
MFU
NOR
PAG

(6 )
(2,6)
(6 )
( 3 ,6)
(3,6)

PGD (6 )
POV (6)
SNH (6)
TDT (6)

(1) TOPS-I0 only
(2) TOPS-20 only
(3) This is a warning, not an error.
The error cannot be
trapped with an ERR= branch, but IOSTAT and ERRSNS will be
set.
(4) See the TOPS-IO Monitor Calls Manual
FILOP. error codes and their meanings.
( 5 ) " J " mea n s the TOP S - 2 0 J S YS err 0 r cod e •
be between 600000 and 610000 (octal).

for

the

list

Th i s n urn b e r will

(6) No ERRSNS values
(7) Where n is the number of times the error occurs
APR and LIB errors are usually reported as warnings, and
the program
continues.
The number of APR and LIB errors listed on the user's
terminal can be changed by the FORTRAN Library Subroutine ERRSET
(see
Section 13.4.1.14).
The
I/O errors either cause messages to be
printed on the terminal or can be trapped by an error exit argument
(ERR=statement number) on OPEN, READ, WRITE, and CLOSE.
The FORTRAN Library Subroutine ERRSNS (see Section 13.4.1.15)
aLlows
you to find out which I/O error occurred. When called, ERRSNS rE!turns
one or two integer values that describe the status of the last I/O
operation
performed
by FOROTS.
(The second
integer value
is
optional.) For example,
CALL ERRSNS (I,J)
calls this subroutine.

J, the second integer value, is optional.

D-8

FOROTS ERROR MESSAGES
D.l

ALPHABETICAL DESCRIPTION OF FOROTS MESSAGES

This section contains alphabetical descriptions of each warning and
fatal
error message
that
is generated
by FOROTS during program
execution.
Each message is first listed; then is followed by a brief
description of how it
is generated; followed by, in some cases, a
recovery procedure; and finally, where applicable,
followed by the
ERRSNS values associated with the message.
ALZ?

Arg umen t less than or
Cause:

to zero

E~qual

An argument (sllch as a memory size) was specified
a value less than or equal to zero.

with

Recovery: Specify the correct value for the argument.

APP?
onl y)

= 21

Second Value

= 125

file

error)

(TOPS-20

The MTOPR% JSYS failed trying to position the
tape with the function .MOFWF or .MOBKR.

magnetic

ERRSNS val UE!S:

First Value

Can't setup to append to magtape
Ca use:

<JSYS

Recovery: Use the information provided to determine a recovery.
ERRSNS valUES:

First Value

= 30

Second Value

= JSYS error
number

AQS?

First argument to SORT must be a quoted string
Cause:

The SORT routine detected that the first argument was
of type CHARAC~'ER, but the string was not word aligned.

Recovery: If the first argument to the SORT program
is of type
CHARACTER, it must be word aligned and terminated by an
ASCII null.
The most. reliable way to generate such an
argument is to use a character constant.
ERRSNS values:

First Vc:lue

Second Value

= 112

ARC % Ambiguous repeat count
Cause:

In a FORMAT statement,
a number between two
format
specifiers can be considered belonging to either one.

Recovery: Insert a comma before or after the number, depending on
which specifier the number belongs with.
ERRSNS values:

Fir s t

Va 1 ue

D-9

Second Value = 532

FOROTS ERROR MESSAGES
ASW?

Ambiguous switch /<sw>
Cause:

(TOPS-I0 only)

In dialog mode, a switch was specified, but not enough
of
the
switch was given
to
uniquely specify which
switch was intended.

Recovery: Retype the line, completely specifying which switch you
mean.
ERRSNS values:
BBF?

First Value

Second Value

241

Bad format binary file
Cause:

The control information stored
in a
binary file
is
incorrect.
The
file
cannot
be
read
using
MODE='BINARY'.
This error can be caused when the
file
you are
reading
was not written by FORTRAN using
MODE='BINARY' •

Recovery: Make sure that you are using the correct file and
mode.
Any
file
can be read with MODE='IMAGE'.
Section 11.3.19)
ERRSNS values:

First Value

Second Value

data
(See

302

BLK % Pages allocated but not deallocated
Ca use:

Internal FOROTS error in memory management.

Recovery: Submit an SPR and include your program.
BSI % Blocksize ignored:
Cause:

A BLOCKSIZE specifier was given in an OPEN statement
(see
Section 11.3.4)
but was not
used because the
device being opened is not a magnetic tape.

ERRSNS values:
CBI?

device is not a magnetic tape

First Value = 26

Second Value = 535

Can't BACKSPACE IMAGE file with no RECORDSIZE
Cause:

An OPEN statement with the MODE='IMAGE' specifier
(but
not
the RECORDSIZE= specifier) was executed prior to a
BACKSPACE statement that referred to the open unit.

Recovery:

If you are
using
fixed-length
records,
specify
the
RECORDSIZE parameter in the OPEN statement (see Section
11.3.27).
Otherwise, the BACKSPACE cannot be done.

ERRSNS values:

First Value = 25

D-I0

Second Value = 536

FOROTS ERROR MESSAGES
CCC?

Can't convert constant to correct type
In NAMELIST input, a variable was assigned a value that
does not match.
For example,
if C is a complex
variable, the input:

Ca use:

C=.TRUE.
is in error,
number.

since

.TRUE.

not

legal

complex

Recovery: Correct the error in the source program.
ERRSNS values:
CCP?

Can't create page <n>

(PAGE.

error <n»

(TOPS-IO only)

FOHOTS attempted to use a page of memory for some task,
but was unabl~
to.
The monitor error code gives the
reason.
This can be caused by
erroneous
MACRO
subroutines.
If no such cause
is found, it is an
internal FOROTS error.

Cause:

CCS?

Second Value = 519

First Value = 97

Not enough memory for creating character stack
A character stack was requested that was larger than 36
sections
(larqer
that the maximum virtual memory
aVcdlable) •

Ca use:

Recovery: Specify correct call argument.
ERRSNS values:
CDF?

The specified
file has had
both
formatted
and
unformatted I/O operations (or OPENs) performed on it.

Recovery: Use I/O operations
specifier.

and

OPENs

ERRSNS values:

Second Value

Can't

destl~oy

Ca use:

CDR?

124

Second Value

Can't determine whether formatted or unformatted
Cause:

CDP?

First Value = :21

First Value
page <n>

(PAGE.

error <n»

with

the

same

FORM=

594

(TOPS-IO only)

FOHOTS attempted to use a page of memory for some task,
but was
unabll~
to.
The monitor error code gives the
reason.
This can be caused
by
erroneous
MACRO
subroutines.
If no such cause
is
found, it is an
internal FOROTS error.

Can't do direct I/O to ::iequential file
Cause:

An attempt was made to perform I/O to
already open in a conflicting mode.

Recovery:

file

Second Value

315

Open file in the appropriate mode.

ERRSNS values:

Fir s t Va:!. u e = 3 1
D-ll

that

FOROTS ERROR MESSAGES
CDS?

Can't do sequentail I/O to direct file
Cause:
Recovery:

An attempt was made to perform I/O to a
already open in a conflicting mode.

First Value = 31

Second Value

315

Can't <read/write> an <input/output>-only file
Cause:

An attempt was made to perform I/O to a file,
but the
file is not open for I/O in the appropriate direction.

Recovery:

Open the file with ACCESS='SEQINOUT'
or
'RANDOM',
as
appropriate.
It
is not
possible to open a file for
APPEND access and then
read
from
it.
(See
Section
11.3.1.)

ERRSNS values:
CEF?

that

Open file is the appropriate mode.

ERRSNS values:
CDT?

file

Fi rst Val ue

End of command file encountered

Second Value
(TOPS-20 only)

Ca use:

An indirect file was specified as
and
the end of the
file
was
terminator character (line-feed).

Recovery:

Edit the file and insert a line-feed.

ERRSNS values:

First Value = 30

CFC % Fortran-77 and VMS:

263

a
DIALOG argument,
encountered before a

Second Value

503

Carriage control character

Cause:

A carriage-control
character
was
used
that
incompatible with ANSI FORTRAN and VAX FORTRAN.

Recovery:

If you want the program
to
be compa ti bl e
FORTRAN
or
VAX
FORTRAN,
use
a
carriage-control character.

ERRSNS values:

First Value = 33

CFF % VMS incompatibility:

Second Value

wi th
ANSI
compatible

321

Intrinsic routine invoked incompatibly

Ca use:

An
intrinsic
routine
was
invoked
in
a
method
incompatible with VAX FORTRAN
(such as
use of an
EXTERNAL statement for an intrinsic function).

Recovery:

If you want the
program
to
be compatible with VAX
FORTRAN,
change
to a method of
invoking intrinsic
routines that is compatible with VAX.

ERRSNS values:

First Value = 33

D-12

Second Value

322

FOROTS ERROR MESSAGES
CFG % Fortran-77 and VMS: G format descriptor used with character
Ca use:

The G format descriptor was used with character data,
which is an extension to ANSI FORTRAN and VAX FORTRAN.

Recovery:

If you want the program compatible with ANSI FORTRAN or
VAX FORTRAN, do not use the G format descriptor to edit
character data.

ERRSNS values:

First Value

CFK % Fortran-77 and VMS:

Second Value

327

Keyword [keyword]

Cause:

An OPEN or CLOSE: keyword was used that is
wi th ANSI FORTRJ,N and VAX FORTRAN.

Recovery:

If you want the program compatible with ANSI FORTRAN or
VAX FORTRAN, USE a compatible OPEN or CLOSE keyword.

ERRSNS values:

First Value = 33

CFO % Fortran-77 extension:

incompatible

= 324

Second Value

Overlap of character assignments

Cause:

A character assignment statement was used in which
the
character
positions defined in the character variable,
array element or substring on the
left of the equal
sign are refer~nced in the character expression on the
right of the equal sign.
This
is
incompatible with
ANSI FORTRAN.

Recovery:

If you want the program to be compatible with ANSI
FORTRAN, use a character assignment statement that does
not overlap the character expression and the character
variable, array element, or substring.

ERRSNS values:
CFR % Fortran-77
(TOPS-20 only)
CFR % Fortran-77
(TOPS-IO only)

Fir s t Va 1 ue = 33

Second Value

326

extension:

FORTRAN-20

supplied

routine

invoked

extension:

FORTRAN-IO

supplied

routine

invoked

Ca use:

A FORTRAN-IO/20-supplied subroutine was invoked that is
not available with ANSI FORTRAN.

Recovery:

If you want your
program to be compatible
FORTRAN, use a compatible subroutine.

ERRSNS values:
CFS?

First Value

Can't find SYS:SORT.EXE Ca use:

= 33

Second Value

(TOPS-20 only)

Recovery: Use
the
information provided
by the
determine the proper course of action.
First Value

= 21

D-13

ANSI

323

The file SORT.EXE cannot be found on SYS:.
A
supplied error message will give more detail.

ERRSNS values:

Second Value

monitor

119

FOROTS ERROR MESSAGES
CFT % Fortran-77 and VMS:

Trailing spaces in output record

Cause:

Your program contains a FORMAT that specifies
trailing
blanks
(X format and $
format).
In this case,
FORTRAN-IO/20 preserves the trailing spaces.

Recovery:

If you want the program compatible with ANSI
FORTRAN
and VAX FORTRAN,
do not use this form of the FORMAT
statement.

ERRSNS values:

First Value

Second Value

325

FORTRAN-20 supplied
routine
invoked
CFX % Fortran-77 and VMS:
(TOPS-20 only)
FORTRAN-IO supplied routine invoked (TOPS-IO
CFX % Fortan-77 and VMS:
only)
Cause:

A FORTRAN-IO/20-supplied subroutine was invoked that is
not available with ANSI FORTRAN or VAX FORTRAN.

Recovery: If you want the program
to be compatible with ANSI
FORTRAN or VAX FORTRAN, use a compatible subroutine.
ERRSNS values:
CGP?

First Value = 33

Second Value = 323

Can't get pages 600:677 for SORT
Cause:

The SORT subroutine brings the SORT program
into core
in pages
600 through 677.
Some of these pages were
already occupied by programs or data at the
time
that
SORT was called.

Recovery: Decrease the size of your program.
Having fewer
files
open
or
using a
BUFFERCOUNT=l specifier
in OPEN
statement may help (see Section 11.3.5).
If this does
not help,
you can segment the program by using LINK's
overlay facility (see the LINK Reference Manual).
ERRSNS values:
CGS?

First Value = 21

Can't get SYS:SORT.EXE Ca use:

Second Value

(TOPS-20 only)

The file SORT.EXE was found on SYS:, however
for
some
reason
it could not be merged into your program in
order to sort files.
A monitor supplied error message
will give more detail.

Recovery: Use
the
information provided
by the
determine the proper course of action.
ERRSNS values:
CIR?

= 116

First Value = 21

Second Value

monitor
120

/CARRIAGECONTROL:TRANSLATED illegal with this /RECORDTYPE
Ca use:
Recovery:
ERRSNS values:

First Value

D-14

Second Value

243

FOROTS ERROR MESSAGES
CLE?

Concatenation result larger than expected
Cause:

The specified substring bounds are out of range.

Recovery: Specify legal substring bounds.
ERRSNS values:
CLS?

First Va]ue = 21

CLOSE failed, <I/O error message>
Cause:

Second Value

108

(TOPS-IO only)

A CLOSE UUO. or FILOP. CLOSE function failed.

Recovery: Use the informat.ion provided to determine a recovery.
ERRSNS values.:

First Value = 28

Second Value = 250+n

CPP % Can't preallocate pages fiOO:677 for SORT
Ca use:

A call
to
preallocate
allocated.

the
SRTINI
subroutine was
SOHT's pages because
they

unable
to
were already

Recovery: Decrease the si~e of your program.
Having fewer
files
open or
using
a
BUFFERCOUNT=l specifier in the OPEN
statement may help (sE!e Section 11.3.5).
ERRSNS values:
CQF % Can't queue file:
Cause:

First Value = 21

Second Value

QUEUE. VUO ERROR <N>

(TOPS-lO only)

This error may occur \lihen executing a
CLOSE statement
in which the DI~)POSE specifier is given with one of the
values:
'LIST',
'PRINT',
or
'PUNCH',
and
GALAXY
release 4 is running.
(See Section 11.5.4.)

Recovery: Refer to
the
TOPS-H) Monitor
Calls Manual
ex pIa nat ion 0 f ~: h e QUE UE. err 0 r n urn be r < n> •
ERRSNS values:
CRP?

121

Fir s t Va :~ ue = 26

for

Second Value = 550

Can't return pages 600:1)77 after call to SORT (TOPS-IO only)
Cause:

Before the SORT subroutine
returns
to
the
user,
it
tries
to deallocate
the
pages that the SORT program
resides in.
Th,= pages could not be deallocated.
This
is not expected to occur.

Recovery: Submit an SPR.
ERRSNS values:

First Value

D-15

Second Value

117

FOROTS ERROR MESSAGES
CSF % Can't set FORTRAN carriage control attribute
Cause:

(TOPS-20 only)

You are running a TOPS-20 system that does not have
Edit
2981,
which allows support
for
the FORTRAN
carriage-control attribute.

Recovery: Autopatch the monitor to include Edit 2981.
ERRSNS values:
CSI?

First Value = 26

Second Value = 502

Can't skiprecord image file with no RECORDSIZE
Cause:

A SKIPRECORD statement was executed for a
file opened
with
MODE=IMAGE,
and
the
file did not have a
RECORDSIZE(RECL) specified.
Since there are no
record
markers
in
IMAGE mode files, FOROTS cannot determine
how far to skip.

Recovery: Add a RECL= specifier (see Section 11.3.27) to the OPEN
statement.
ERRSNS values:
CWL?

First Value

536

Can't write a file with MODE='LINED'
Cause:

The program attempted to execute an output operation
such as a WRITE statement after an OPEN statement for
the same
unit.
The OPEN statement
contained
a
MODE='LINED' specifier.

Recovery: Change the OPEN statement
(See Section 11.3.20.)
ERRSNS values:
CWU?

Second Value

DIVERT:
Cause:

First Value = 47

specifier

Second Value

MODE='ASCII'.
554

Can't write to unit <n>
DIVERT file is not opened for output.

Recovery: Open DIVERT file for output.
ERRSNS values:
DBM?

First Value = 21

Second Value

107

DBMS not loaded
Cause:

A DBMS call to a sharable FOROTS was attempted
DBMS.

without

Recovery: The system manager must build DBMS into sharable FOROTS
or
remove
the
file from FORLIB that directs the DBMS
call to the sharable FOROTS.

D-16

FOROTS ERROR MESSAGES
DEL?

Can't delete file:(FILOP.
Ca use:

error message>

The "DELETE" FILOP.

failed.

(TOPS-IO only)

The file is not deleted.

Recovery: Usually you can correct the problem when the program is
finished
and delete
the file with the monitor DELETE
command.
ERRSNS values:

First Value

Second Value

250+n

DLF % Data in 10 list but not in format
Cause:

An I/O statement has requested data to be
but
the
FORMAT
statement does not
descriptor that would translate the data.

Recovery:

Fi~ the FORMAT statement.
following descriptors:

transferred,
specify any

It must contain one

the

A,B,E,F,G,I,L,O,Q,R,Z
ERRSNS values:
DLL?

= 62

First Value

DUMP mode I/O list too long
Ca use:

Second Value

(TOPS-IO only)

The I/O list sfecified for a DUMP mode READ or WRITE
statement is too long for the FOROTS internal DUMP mode
control list.

Recovery: Split the I/O list across two or
statements.
ERRSNS values:

First Value

READ

Second Value

581

DMA % Must give lower and upper bounds to dump
(TOPS-20 only)
Ca use:

306

non-zero

WRITE

sections

A call to PDUMP or DUMP was made without specifying
memory bounds.
In section zero, this is interpreted as
'all of memory'.
For extended addressing,
lower
and
upper bounds must be specified.

Recovery: Specify lower and upper bounds for memory_
ERRSNS values:
DQE?

Disk full or quota
Cause:

= 21

First Vclue

Second Value = 126

exc~eded

The disk quota for the disk on which a
file
is being
written
is e~hausted, or the entire disk structure is
full.
If this error is encountered while running under
batch, the program is aborted and an attempt is made to
close all files.
If this error
is encountered while
timesharing,
t.he
user is requested to type an EXPUNGE
command and then a CONTINUE command.

Recovery: EXPUNGE or create
structure.

First Value

ERRSNS values:

D-17

room

the

Second Value

specified
590

disk

FOROTS ERROR MESSAGES
DQW % Disk full or quota exceeded (TOPS- 20 onl y)
Cause:

Please EXPUNGE, then type CONTINUE

The file or files being written on the disk have either
exhausted your disk quota or filled the structure.

Recovery: FOROTS leaves the terminal at EXEC level
so
that you
have more options to use to provide space on the disk
structure.
You can type CONTINUE to resume
processing
if you don't reset the current fork.
ERRSNS values:

First Value = 22

DSS % DISPOSE='SAVE' assumed Cause:

device is not a disk

First Value

Error in dialog string
Cause:

590

A DISPOSE value other than 'SAVE' was specified
fa r
a
Section
file
on a device other
than disk.
(See
11.3.13.)

ERRSNS values:
DST?

Second Value

549

(TOPS-IO only)

A syntax error in the DIALOG= specifier.

Recovery: Correct the error in the program
(See Section 11.5.3.)

DIALOG

mode.

DTL % Dialog string too long
Cause:

The argument to DIALOG= cannot be parsed because it
too long.

Recovery: Use a shorter string.
ERRSNS values:
ECS?

First Value = 45

Second Value

533

Not enough memory for expanding character stack
Cause:

More memory than is available on a KLI0 was
requested
by either a
dynamic concatenation (concatenation of
character variables of length*) or by a
call
to
the
ALCCHR subroutine.

Recovery: Reduce the size of your concatenation or
argument
to
ALCCHR.
If
you are running extended addressing, you
can present parts of the character stack by invoking
SORT and dynamic
libraries at
the beginning of the
program.
ERRSNS values:
EDS/EDA?
Ca use:

First Value

Error in DIALOG string -

Second Value

124

(TOPS-20 only)

A syntax error in the DIALOG= specifier.
11.5.3.)

(See

Recovery: Retype the specifier correctly in DIALOG mode.
ERRSNS values:

First Value = 45

D-18

Second Value = 539

Section

FOROTS ERROR MESSAGES
EFS [ Enter correct file specs ]
Cause:

Dialog mode.
You should respond
to
this message by
entering
any
information you wish to change about the
indicated file.
This can include the device, filename,
directory,
or any OPEN parameter.
The
form of a
response is:

TOPS-IO
TOPS-20

DEV:FILE.EXT[DIRECTORY] /SWITCH:VAL /SWITCH:VAL
DEV:<DIRECTORY)FILE.EXT.GEN /SWITCH:VAL /SWITCH:VAL
All parts of this specification are optional.

EOF?

End of file
Cause:

ESV?

An input statement has attempted to read more data than
the file contains.

Recovery: Use an
END= specifier
lengthen the file.

ERRSNS values:

Second Value

Fi rst Va I ue

the

READ

statement,
-1

<unknown/ambiguous> keyword value /<switch>:<value>
only)
Ca use:

A switch entered in dialog mode was not
was not specified uniquely.

(TOPS-20

recognized

Recovery: Retype the line, specifying the correct switch.
ERRSNS values:

First Value = 45

Second Value = 241

ETL % Attempt to write beyond fixed-length record
Cause:

In an ENCODE statement,
the
format specified more
characters than the string will
hold.
The excess
characters are ignored.
(See Section 10.12.)

Recovery: Shorten the format or lengthen the string.
ERRSNS values:
FCL?

First Value = 22

Second Value = 509

Found unexpected continuation LSCW
Cause:

A RECL(RECORDSIZE)
has been specified
in an OPEN
statement,
and
FOROTS has encountered a continuation
(type 2) logical segment control word
(LSCW).
This
type
LSCW is never written
in
fixed-length binary
records.

Recovery: Remove the RECL(RECORDSIZE)
statement.

specification in

ERRSNS values:

Second Value

First Value

D-19

573

the

OPEN

FOROTS ERROR MESSAGES
FDC % Floating divide check
Ca use:

The program contains a floating-point division in which
the divisor
is
too small compared to the dividend to
yield a result that is in the floating-point range.

Recovery: Correct
program
so
floating-point range.
ERRSNS values:

FFX?

that

First Value = 5

division

Second Value

wi. thin

n
where n is
the number
of times the
error occurs

FOROP. function code exceeds range
Cause:

A library routine has called for an operation
that
is
not
available.
This can be caused by using
a
mismatched FOROTS and FORLIB, or by an erroneous MACRO
subroutine.

Recovery: Make sure the versions of
FOROTS.EXE and
FORLIB.REL
that
you are
loading from are matching versions.
If
this doesn't help,
find
the
subroutine causing
the
problem and alter or remove the erroneous call.
FOV % Floating overflow
Cause:

A REAL or DOUBLE PRECISION number was read that is too
large
in magnitude
(see
Chapter 3).
This is only a
warning and does not stop execution of the
program.
The
results of a calculation that overflows are set to
the largest representable number with the sign of the
correct result.

Recovery: Modify the data so that its values fall in the range of
values that can be represented for the data type.
ERRSNS values:

FRR?

First Value = 3 or 4 Second Value = n
where n is
the number
of times the
error occurs

/RECORDTYPE:FIXED requries /RECORDSIZE
Ca use:

A RECORDTYPE='FIXED' was specified in an OPEN statement
without a RECORDSIZE (RECL) specifier.

Recovery: Specify RECL
11.3.27) •
ERRSNS values:

the

First Value

D-20

OPEN

statement

(see

Second Value

240

Section

FOROTS ERROR MESSAGES
FTS % Output field width too small
Ca use:

The field width specified in a FORMAT statement was not
large enough
to allow the printing of the value being
output.
For example, this error would occur
if the
number 100 is output with the format specifier "12".

ERRSNS values:

First Value

= 9

Second Value

= a

FUN % Floating underflow
Cause:

A REAL or DOUBLE PRECISION number was read that is too
small
in magnitude
(see
Chapter 3).
This only a
warning and does not stop execution of the
program.
The
result of a calculation that underflows is set to
zero.

Recovery: Modify the data so that its values fall in the range of
values that can be represented for the data type.
ERRSNS values:

First Value

= 6 or 7 Second Value = n
where n is
the number
of times the
error occurs

FVF?

Format and variable type do not match
Ca use:

An attempt was made to READ or
with other than A or G format.

WRITE

character

Recovery: Specify A or G edit descriptors when reading
data.
ERRSNS values:

First Value

Second Value

data

character

583

FVM % Format and variable type do not match
Cause:

An I/O statement has been executed that uses a
format
edit descriptor with a type that does not match the I/O
list item being processed.

Recovery: Specify the appropriate format edit descriptor for
I/O list item.
ERRSNS values:
lAC?

First Value

Second Value

the

583

/ACCESS illegal for this device
Ca use:

An attempt was made to OPEN a device
for
access specified (or implied) is illegal.

which

Recovery: Change the ACCESS specifier in the
OPEN statement
data transfer statement.
(See Section 11.3.1.)
ERRSNS values:

First Value

= 30

D-21

Second Value

= 248

the
or

FOROTS ERROR MESSAGES
IAV?

Illegal value for OPEN specifier
Cause:

An OPEN statement specifier has
that specifier.

value

illegal

for

Recovery: Specify a legal value for that OPEN specifier.
ERRSNS values:
ICA?

First Value

= 30

Second Value

= 585

Incompatible attributes
Ca use:

An illegal combination
specified.

open

attributes

has

been

Recovery: Change one or more of the conflicting specifiers.
ERRSNS values:
ICD?

First Value

= 30

Second Value = 506

Non-digit in record delimiter
Cause:

The
format of the
tape being
read
is
not
'D'
(DELIMITED).
The Record Control Word (RCW) contained a
non-digit, or the data on the tape is incompatible.

Recovery: Specify the correct TAPEFORMAT in the OPEN statement.
ERRSNS values:
ICE?

First Value

= 25

Second Value

= 570

Illegal length for character expression
Ca use:

A program has specified a
zero
length or negative
length character substring as an I/O list element.

Recovery: Fix program to specify a positive length substring.
ERRSNS values:

First Value

= 81

Second Value

= 599

IDC % Integer divide check
Cause:

Program contains an integer division by O.

Recovery: Correct division in program.
ERRSNS values:

First Value

= 1

Second Value

where n is
the number
of times the
error occurs
IDD?

Illegal character <chr>
Cause:

(TOPS-IO only)

An illegal character was encountered in dialog mode.

Recovery: Retype the response without illegal characters.
ERRSNS values:

First Value = 45

D-22

Second Value

= 548

FOROTS ERROR MESSAGES
IDI?

Illegal DUMP mode I/O list
Cause:

(TOPS-IO only)

An I/O list entry has been specified whose entry size
(number of words)
is different
from its increment.
This can only happen if an implied DO loop is specified
for the I/O list, the index increment is set to a value
other
than
1,
and
the
program
is compiled
with
/OPTIMIZE.

Recovery: Use an index increment of 1,
program with /OPTIMIZE.
ERRSNS values:
10M?

First Value = 81

not

Second Value

the

579

/MODE:<mode> illegal for this device
Ca use:

Not all devices can do I/O in all modes.
terminals cannot do binary I/O.

For

Recovery: Change the MODE= specifier or the device.
11.3.19.)
ERRSNS values:
IOU?

compile

DIVERT:
Cause:

Fi rst Va 1 ue

Second Value

example,

(See Section
249

illegal to divert to unit <n>
Unit specified is an input-only device.

Recovery: Specify a unit for which output is legal.
ERRSNS values:
IEM?

First Value

= 21

Second Value = 104

FOROTS internal error in memory management
Cause:

This is an internal
error
that
is not expected
to
occur.
It means
that the memory management routines
have detected a problem with their control information.

Recovery: Submit an SPR.
IFW?

Illegal field width
Ca use:

An illegal (negative)
FORMAT statement.

field width was

specified

Recovery: Specify a legal field width in the FORMAT statement.
ERRSNS values:
IHC?

First Value

= 62

Second Value

= 553

Illegal Hollerith constant
Ca use:

A Eormat specification contains an H edit descriptor
that
is not
preceded by a length or does not contain
enough characters.

Recovery: Use the correct format for an H edit descriptor.
Section 12.4.2.)
ERRSNS values:

First Value

62
0-23

Second Value

552

(See

FOROTS ERROR MESSAGES
IJE?

"Impossible" JSYS error at <PC) Ca use:

This is an internal FOROTS
to occur.
A monitor call
to.
A monitor-supplied
assistance in avoiding the

<JSYS ERROR)

(TOPS-20 only)

error that is not expected
failed that was not expected
error message may be
of
problem.

Recovery: Submit an SPR.
ILC ?

Illegal character in data
Ca use:

A format descriptor that
nonnumeric character.

requires

number

found

with

rerun

the

operation

that

Recovery: Fix the input data or FORMAT statement.
ERRSNS values:
ILF ?

First Value

= 64

A format specification contains
defined meaning.

Recovery: Correct the error in the
program.
ERRSNS val ues:

First Value

fo rma t

character

list

Second Value

An error was encountered during a file
FOROTS did not expect.

301

Recovery: This type of error should not happen.
SPR.

Please submit an

First Value

JSYS error
number

ERRSNS values:

Second Value

Variable or namelist does not start with letter
Ca use:

NAMELIST input contains something other
than a
legal
variable or NAMELIST name in a context where a variable
or NAMELIST name is required.

Recovery: Correct the source program with a
(See Section 12.6.)
NAMELIST name.
ERRSNS values:
ILS ?

and

Unexpected MTOPR% ERROR (TOPS-20 only)
Cause:

ILN?

307

Illegal character in format
Ca use:

ILM ?

Second Value

First Value

= 97

legal

Second Value

variable

515

Illegal subscript
Ca use:

In NAMELIST I/O, an illegal subscript was given for
array.

ERRSNS values:

First Value

D-24

Second Value

516

FOROTS ERROR MESSAGES
IMV?

Illegal MTOP value
Ca use:

A MARCO program has issued an MTOP call to FOROTS
an illegal valuE for the function.

with

Recovery: Specify a legal function value in the call.
ERRSNS values:
IOE?

First Value

= 81

Second Value

= 574

<IO error message>
Cause:

An I/O error has occurred.
The monitor error code
is
given,
along
with an
interpretation of the probable
meaning of the error bits.
This message normally
indicates that the data recorded on an external device
has been damaged and cannot be read correctly.

ERRSNS values:

First Va]ue

= 98

Second Value = 400

(TOPS-IO)

JSYS error
number
(TOPS-20)
IOL?

Bad format IO list
Cause:

The code generated by the compiler for an I/O list
is
not.
understood
by this version of FOROTS.
The
erroneous entry in
the
I/O list
is
ignored.
This
probably indicates an internal error in the compiler or
in FOROTS.
prob~em

Recovery: Locate the
it.

area of the I/O

Fir s t Va ::. ue

ERRSNS values:

list

and

Second Value

508

simplify

IOV % Integer overflow
Cause:

IPP?

An attempt was made to read data that was out of
for an integer variable.

ERRSNS values:

First Value

Illegal PPN

(TOPS-IO only)

Ca use:

= 2

Second Value

range

A directory specification starts with something that is
not a legal PPN specification.
The forms of legal PPNs
a rE~ :

[n,n],

[n,l,

[,n], or [,]

where n represents a 1- to 6-digit octal number.
Recovery: Use a legal
ERRSNS values:

dir~ctory

First Value

D-25

specification.

= 45

Second Value

545

FOROTS ERROR MESSAGES
IPN?

Illegal page number <n>
Cause:

A call to TOPMEN or SRTINI has specified a page
outside the range 1:777.

number

Recovery: Specify a correct page number
ERRSNS values:
IRC?

First Value = 21

Second Value

122

Illegal repeat count
Ca use:

An
illegal
statement.

repeat

count

was

given

FORMAT

Recovery: Correct the FORMAT statement.
ERRSNS values:
IRN?

First Value

= 62

Second Value

538

Illegal record number <n>
Ca use:

A direct-access I/O statement has
number that is zero or negative.

specified

record

Recovery: Correct the invalid record number in the program.
ERRSNS values:
ISS?

First Value

= 25

Second Value = 512

Illegal substring descriptor
Cause:

An
I/O statement refers
to an
illegal
substring
delimiter (substring not within bounds of string).

Recovery: Correct the substring specifier.
ERRSNS values:
ISW?

First Value

= 97

Second Value

597

Can't switch to input
Cause:

A file that was being written cannot be open
for
output.
The file is either protected against reading,
or has been deleted
before
the
OPEN for
read
is
executed.

Recovery: Specify correct protection for OPEN write.
ERRSNS values:

First Value = 98

Second Value

= 250+n
(TOPS-IO)
JSYS error
number
(TOPS-20)

ITE?

Tape is not usable by your job (TOPS-IO only)
Cause:

A tape unit was specified or implied that is not owned
by your job, and is probably owned by another job.

Recovery: ASSIGN the drive or MOUNT the tape.
ERRSNS values:

First Value

= 96
D-26

Second Value

587

FOROTS ERROR MESSAGES
IUN?

Illegal unit number <n>
Cause:

An I/O statement has specified a unit
negative or too large.

number

that

Recovery: Change the UNIT specifier value and rerun the program.
ERRSNS values:
IWI?

First Value

Illegal to initiate
<I/O statement)
Cause:

= 32

another

I/O

Second Value
statement

= 239

while

processing

An I/O statement, STOP statement, or PAUSE statement
has
been
initiated while processing
another
I/O
statement (such as in a function reference used as an
I/O list element), or while within a subroutine called
as a result of an I/O error through ERRSET.

Recovery: Remove the offending I/O statement, STOP statement,
PAUSE statement.
ERRSNS values:
MFU?

First Value

Second Value

582

Memory full
Cause:

There is insufficient memory to complete
the program.

execution

Recovery: Some memory can be saved by opening fewer
files at a
time, by using BUFFERCOUNT=l in OPEN statements, and by
using minimal tape block sizes.
If these techniques do
not help,
you can segment the program using LINK's
overlay facility (see the LINK Reference Manual).
NCA?

No memory available for character stack
Cause:

For non-overlay programs, this message
indicates that
the- memory manager has allocated all available space
bet.ween the uselo's 10\<1 segment and FOROTS.
For overlay
programs,
this messages
indicates that /SPACE:O has
be€~n speci fied to LINK.

Recovery: For non-overlay programs, LINK with /OTS:NONSHARE.
For
overlay programs,
specify at least 1000 to the /SPACE
switch in LINK (see the LINK Reference Manual).
ERRSNS values:

First Value = 21

NCK % <keyword> in CLOSE is medningless Ca use:

Second Value

= III

ignored

Options have be~n included in the CLOSE statement
are meaningless for closing the file.

Recovery: Use valid CLOSE options.

(See Section 10.17.)

ERRSNS values:

Second Value = 542

First Value = 26

D-27

that

FOROTS ERROR MESSAGES
NCS?

No character stack allocated - compiler error
Cause:

An internal compiler error has occurred.

Recovery: Submit an SPR.
ERRSNS values:
NOI?

Second Value

110

(TOPS-I0 only)

An OPEN statement has specified a null device name.

ERRSNS values:

First Value

= 45

Second Value = 544

Found "<chr>" when expecting ":"
Cause:

Substring parameters not separated by":"

Recovery:

Insert a ":" between substring parameters.

ERRSNS values:
NEQ?

No device specified with ":"
Ca use:

NEC?

First Value

First Value = 97

Second Value

= 596

Found "<chr>" when expecting "="
Ca use:

NAMELIST input found an illegal character in a
that requires an equal sign.

context

Recovery: Replace illegal character with equal sign.
ERRSNS values:
NFC?

First Value

Too many open units
Cause:

= 97

Second Value = 513

(TOPS-IO only)

On TOPS-I0 monitors before version 7.00, at most,
16 units can be open at the same time.

only

Recovery: Arrange the program so that it never needs to have more
than 16 simultaneously open units.
ERRSNS values:
NLS?

First Value = 30

Second Value

242

Null string illegal
Cause:

An attempt was made to input to a
during list-directed input.

zero

length

string

Recovery: Insert characters into the string, or remove the string
delimiting quotes.
ERRSNS values:
NQS?

First Value

Second Value

PAOCHAR must be single char in double quotes
Cause:

580

(TOPS-IO only)

In dialog mode, the PAOCHAR specifier must be
by the pad character in double quotes.

ERRSNS values:

Fir s t Va 1 ue = 45

0-28

Second Value

551

followed

FOROTS ERROR MESSAGES
NRP?

Missing right paren
Cause:

In NAMELIST or list-directed complex input, the closing
right parenthesis that ends a complex number was not
found.

ERRSNS values:
NSD?

First Value

Second Value

514

No such device <dev>
Ca use:

The specified device does not exist.

Recovery: Change the device name to one that does exist.
ERRSNS values:
NSI?

= 30

Second Value

= 245

Null SFD (TOPS-IO only)
Cause:

A directory specification contains a null SFD.

ERRSNS values:
NSS?

First Value

= 45

Second Value

= 547

No free section available for SORT (TOPS-20 only)
Cause:

SORT runs it its own section on machines that support
free
sections
extended
addressing.
There are no
available.

Recovery: There are 31 sections normally available when a
FORTRAN program runs.
If your application
trying to use e~tended addressing,
this error
not occur, and you should submit an SPR.
ERRSNS values:

First Va]ue = 2:1

OGX % Galaxy version 2 not supported

= 118

(TOPS-IO only)

Cause:

Your system is using an unsupported version of GALAXY.

Recovery:

Inform the system administrator
supported version of GALAXY.

ERRSNS values:
OPN?

Second Value

simple
is not
should

First Value

= 26

Second Value

upgrade

the

595

Can't OPEN file
Ca use:

The specified fj.le could not be opened.
The
reason
given
is
taken
from
the monitor error code (see the
TOPS-IO Monitor Calls Manual) •

ERRSNS values:

First Va:.ue

= 30

D-29

Second Value

250+n

FOROTS ERROR MESSAGES
OSW?

FILOP.
Ca use:

error n - can't switch to output
An attempt to open a file for write
access which
previously been open for read-only access failed.

has

Recovery: Change protection code; remove other file access.
ERRSNS values:

First Value

= 98

Second Value = 250+n
(TOPS-IO)
JSYS error
number
(TOPS-20)

PAG % Pages allocated but not deallocated
Ca use:

Internal FOROTS error in memory management.

Recovery: Submit an SPR and include your program.
PGD?

Deallocating more pages than allocated
Cause:

Internal FOROTS error in memory management.

Recovery: Submit an SPR and include your program.
POI?

<file positioning operation) Illegal for DIRECT (RANDOM)
Cause:

A file
positioning operation
(such
BACKSPACE)
was
attempted
on
a
DIRECT(RANDOM) access.

file

as
REWIND
file open

or
for

Recovery: Remove the file positioning statement.

PPN?

= 31

ERRSNS values:

First Value

(TOPS-20 only)

Cause:

593

A TOPS-20 OPEN statement has specified a PPN instead of
a
directory
name,
but
the
PPN has no matching
directory.

Recovery: Specify the correct PPN, or
directory name instead.
ERRSNS values:
RBR?

Second Value

= 30

First Value

better

yet,

Second Value

specify

the

405

REREAD not preceded by READ
Cause:

A REREAD statement was encountered
before any READ
statement.
A READ must be executed first so there is
something to reread.

Recovery: Cause a READ statement to be executed first.
ERRSNS values:

First Value

= 39

D-30

Second Value

= 310

FOROTS ERROR MESSAGES
RLB?

/RECORDSIZE larger than /BLOCKSIZE
Cause:

A RECORDSIZE was specified in an OPEN statement that is
larger than the specified or implied BLOCKSIZE.

Recovery: Correct either RECORDSIZE or BLOCKSIZE,
BLOCKSIZE if it is not specified.
ERRSNS values:
RIC?

First Value

= 30

specify

244

Second Value

Reading into character format illegal
Ca use:

An attempt was made to READ into a character format.

Recovery: Correct program to avoid this construct.
READ into a
(perhaps
character variable and
use
this variable
concatenated with other character expressions)
for
a
modifiable format.
ERRSNS values:

Fi rst Va] ue

524

Second Value

RIF % Reading into FORMAT nonst.andard
Cause:

A READ statement. was executed that reads data
into a
Hollerith or quoted string in a FORMAT statement.
This
is a practice contrary to
the ANSI standard and
is
likely not to work in future releases of FORTRAN-IO/20.

Recovery: READ into character variables
and
use
(perhaps
concatenated wit.h character constants) as the FORMAT.
ERRSNS values:
RNM?

First Va]ue

= 22

Second Value

= 584

Can't rename file
Cause:

An attempt to rename the specified file failed.

Recovery: Change file protection, remove other file access.
ERRSNS values:

First Value

Second Value

250+n
(TOPS-IO)
JSYS error
number
(TOPS-20)

RNR?

Record <n> has not been written
Ca use:

In direct-acces~; input, an attempt was made to
read
a
record
that was never written.
This may indicate the
USE~ of an incorrect rE~cord number.

Recovery: Make sure you
ERRSNS values:

al~e

First Va:Lue

requesting the correct record.

= 25

D-31

Second Value

= 510

FOROTS ERROR MESSAGES
RPE?

Illegal repeat count
Cause:

In NAMELIST or list-directed input, a repeated constant
was
found,
but
the
repeat count
is not a positive
integer.

Recovery: Correct the input and try again.
ERRSNS values:
RR1?

First Value = 97

Second Value

521

Random I/O requires RECORDSIZE specifier in OPEN statement
Cause:

Direct-access I/O was attempted to a file that has not
been opened
with
the RECL or RECORDSIZE specifier to
give the size of the record(s).

Recovery: Specify a
record
size.
The
record
size
is
in
characters
for
formatted files, words for unformatted
files.
(See Section 11.3.27.)
ERRSNS values:
RRR?

First Value = 30

Second Value

240

Random IO requires /RECORDSIZE
Ca use:

An
OPEN statement was
attempted
that
specified
ACCESS=DIRECT(RANDOM) with no record size specified

Recovery: Specify a record size.
ERRSNS values:
RSM?

First Value = 30

Second Value

240

Record size different from that specified
Ca use:

A record size found in a
binary record
is
than that specified in the OPEN statement.

different

Recovery: Specify the correct record size in the OPEN statement.
ERRSNS values:
SDO?

First Value = 25

Second Value = 572

Same device open on unit with conflicting specifiers
Ca use:

An OPEN statement was attempted for a device for
which
another
OPEN or data
transfer
statement had
been
already executed, and the file specifications were
in
conflict.

Recovery: Change file specifications
ERRSNS values:
SLN?

First Value = 30

Second Value

540

Record length negative or zero
Ca use:

An ENCODE or DECODE statement was initiated that has a
negative or zero value for the record (string) length.

Recovery: Correct program to specify a legal record length.
ERRSNS values:

First Value = 25
D-32

Second Value = 577

FOROTS ERROR MESSAGES
SNH?

Internal FOROTS error at <PC>
Ca use:

This error is Dot expected
to occur.
consistency check has found a bug.

in te rnal

Recovery: Please submit an SPR if you get this message.
SNQ?

String not within
Cause:

quotes

A character date\ item read as list-directed or NAMELIST
input is not enclosed in single quotes.

Recovery: Enclose
ERRSNS values:
SNV?

singl~

charactE~r

data item in single quotes.

First Va:.ue

= 97

Second Value

= 598

Sign with null value
Cause:

List-directed Ol~ NAMELIST input contains a + or not followed by a value.

sign

Recovery: COI-rect the input and try again.
ERRSNS values:

Fir s t

Va :~ u e

SRE % Subscript range error <n>
Cause:

= 97

Second Value

subscript <n> of array

522
<name>

line

An illegal subs~ript or range has been specified for an
array reference.

Recovery: Specify a legal array reference.
ERRSNS values:

First Value

= 23

SSE % Substring range error <var(bound»
Cause:

Second Value

114

on line <n>

An illegal substring bound or range has been
in a character expression.

specified

Recovery: Specify a legal reference.
ERRSNS values:
TFM?

First Value

= 23

Second Value

113

Tape format conflicts with OPEN statement or default
Cause:

The actual format of the
tape
(either
CORE-DUMP or
INDUSTRY)
conflicts with
the format specified in the
OPEN statement or by the monitor.
At this
point,
the
file is already opened in the wrong format.

Recovery: Specify the correct TAPEFORMAT in the OPEN statement or
with the EXEC.
ERRSNS values:

First Value

0-33

Second Value

569

FOROTS ERROR MESSAGES
TMA?

Too many arguments in call to SORT
Cause:

When the
first argument
in a call
to
SORT is a
character constant, the argument list must be copied in
order to convert the argument to a Hollerith constant.
At most, 10 arguments can be copied.

Recovery: Change the first argument to a Hollerith
use less than 10 arguments.
ERRSNS values:
TMF?

Too many SFDs
Cause:

First Value

= 21

Second Value

constant,

115

(TOPS-IO only)

A directory specification contains more than five SFDs.

Recovery: Specify the correct directory.
ERRSNS values:
UDT?

First Value = 45

Second Value

546

Undefined data type or internal FOROTS error
Cause:

Internal FOROTS error.

Recovery: Submit an SPR.
ERRSNS values:
UME?

Fi rst Val ue

Unexpected MTCHR error <n>
Cause:

Second Value

575

(TOPS-I0 only)

This message is not expected to occur.
It
indicates
that a MTCHR UUO has
failed.
The monitor-supplied
error code may give some indication of the reason.

Recovery: Submit an SPR.
ERRSNS values:

First Value

Second Value

531

UMO % <JSYS error> trying to set tape <density/parity/data mode>
(TOPS-20 onl y)
Cause:

It was not possible to set the indicated
parameter of
the tape.
The monitor error message gives the reason.

Recovery: Make sure you are
using
requested operations.
ERRSNS values:

UNO?

DIVERT:
Ca use:

First Value

= 96

drive

that

Second Value

supports

the

JSYS error
number

unit <n> is not open
The file to which error messages are diverted
opened for output before DIVERT is called.

Recovery: Open the file for output before calling DIVERT.
ERRSNS values:

First Value

= 21

D-34

Second Value

= 105

must

FOROTS ERROR MESSAGES
UNS?

Unit not specified
Cause:

A call was made to FOROTS which did not contain a
number.

unit

Recovery: Correct calling code.
ERRSNS values:

First Value

= 81

Second Value

501

UOA % Unknown OPEN keyword <n>, ignored
Cause:

The compiler has generated an OPEN call
that
an unknown keyword.
'rhe keyword is ignored.

Recovery: Make sure you are using the correct
compiler, FORLIS, and FOROTS.
ERRSNS values:
USW?

First Value

Unknown switch /<sw>
Cause:

= 26

contains

versions

Second Value

the

541

(TOPS-IO only)

In dialog mode, an unknown switch was specified.

Recovery: Retype the line, specifying the correct switch.
ERRSNS values:
UTE?

First Value

Unexpected TAPOP error <n>
Ca use:

= 45

Second Value

= 241

(TOPS-IO only)

This message is not expected to occur.
It
indicates
that a
TAPOP UUO has failed.
The monitor-supplied
error code may give some indication of the reason.

Recovery: Submit an SPR.
ERRSNS values:

First Value

Second Value

UTO % UnexpectedTAPOP.
error <n> trying to set
mode/blocksize> (TOPS-IO only)
Ca use:

<density/parity/data

It is not possible to set the
indicated
parameter of
the tape.
The monitor error message gives the reason.

Recovery: Make sure you are using
requested operation.
ERRSNS values:
VNN?

530

First Value

= 96

drive

that

supports

Second Value

the

537

Variable <var> not in namelist
Ca use:

NAMELIST input contains an assignment
that is not in the namelist.

Recovery: Correct the input and try again.
ERRSNS values:

First Vc,lue = 97

D-35

Second Value

309

variable

FOROTS ERROR MESSAGES
WBA?

Attempt to WRITE beyond variable or array
Ca use:

An attempt was made
to
write
beyond
the end
of a
character variable or array with an internal file WRITE
statement.

Recovery: Correct program to
variable or array.
ERRSNS values:

First Value

stay
25

D-36

within

limits

Second Value

576

character

APPENDIX E
INTERACTIVE DEBUGGER (FORDDT)

ERROR MESSAGES

FORDDT responds with two levels of messages - fatal error and warning.
Fatal
error messages indicate that the processing of a given command
has been terminated.
Warning messages
provide helpful
information.
The format of these messages is:
?FDTxxx text
or
%FDTxxx tex t
where:
?

indicates a fatal message
indicates a warning message
FDT is the FOR DDT mnemonic
xxx
is the 3-letter mnemonic for error message
text is the explanation of error
%

Square brackets ([ ]) in this section signify variables
output on the terminal.

and

are

not

?FDT

Fatal Errors
The fatal errors in the following list are each preceded
by
the user terminal.
They are listed in alphabetical order.
BDF

[symbol]
[symbol]

BOI

Bad octal input

is undefined
is multiply defined

An illegal
CCN

characte~

was detected in an octal input value.

Cannot continue
A pause has been placed on some
form
of
causing
FORDDT to
loop;
should
never
FORTRAN-compiled programs ..

CFO

Core file overflow
The storage area

CNU

skip instruction
be encountered in

The command

fOl~

[name]

GROUP text has been exhausted.
not unique

More letters of the command are required to
from the other commands.

E-l

distinguish

INTERACTIVE DEBUGGER (FORDDT)
CSH

ERROR MESSAGES

Cannot START here
The specified entry point is not an acceptable FORTRAN
program entry point.

DNA

Double-precision comparisons not allowed

DTO

Dimension table overflow
FORDDT does not have the space
to
dimensions until some are removed.

FCX

record

any

main

array

Format capacity exceeded
An attempt was made to specify a FORMAT statement
requiring
more space
than was originally allocated by the FORTRAN
compiler.

FNI

Formal not initialized, please retype
There was a
reference
to a
formal
subprogram that was never executed.

FNR

[array name]

parameter

some

is a formal and may not be redefined

Formal parameters may not be DIMENSIONed.
IAF

Illegal argument format [rest of user line]
The parameters to
the given command
were not specified
properly.
Refer to the documentation for correct format.

IAT

Illegal argument type

= [number]

An
unrecognized
subprogram argument
Submit an SPR if this message occurs.
ICC

type

was

detected.

Comparison of two constants is not allowed
A conditional test involves two constants.

IER

Internal FORDDT error [number]
Internal FORDDT error -

IGN

please report through an SPR.

Invalid group number
Group numbers must be integers and in the range one
eight.

INV

through

Invalid value [rest of user line]
A syntax error was detected in the numeric parameter.

IPN

Illegal program name

IRS

Illegal range specification
The particular range specified for an array is not defined.

ITM

Illegal type modifier - S
The mode S is only valid for ACCEPT statements.

E-2

INTEHACTIVE DEBUGGER (FORDDT)
IWI

ERROR MESSAGES

I/O within I/O error
An attempted TYPE or ACCEPT command
cannot be executed
because a
fatal
"I/O within I/O" error from FOROTS would
result, since the user program is currently processing
an
IOLST call.

JSE

[JSYS error message]
Error reading program name
input) •

LGU

[array name]

(on TOPS-20 or some other

lower subscript .GE.

FORDDT

upper

The lower bound of any given dimension must be less than
equal to the upper bound.
LNF

[label]

MCD

Compile program with the
statement

MLD

[array name]

is not a FORMAT statement
DEBUG

switch

type

FORMAT

dimensioned more
than once
(by
construct) in a single command.

More subscripts needed
The array is defined
to
have more
specified in the given reference.

NAL

multi-level array definition not allowed

The same array cannot be
means of the [dimensions]
MSN

dimensions

than

were

Not allowed
An attempt has been made to modify something other than data
or a FORMAT.

NAR

Not after a reenter
The given command is not allowed until program integrity has
been restored by means of a CONTINUE or NEXT command.

NDT

DDT not loaded

NFS

Cannot find FORTRAN start address for

[program name]

Main program symbols are not loaded.
NFV

[symbol]

is not a FORTRAN variable

Names must be 6-character
with a letter.

alphanumeric

NGF

Cannot GOTO a FORMAT statement

NPH

Cannot insert a PAUSE here

strings

beginning

An attempt has been made to place a breakpoint at other than
an executable statement or subprogram entry point.
NSP

[symbol]

no such PAUSE

An attempt has been made to REMOVE
never set up.
E-3

breakpoint

that

was

INTERACTIVE DEBUGGER (FORDDT)
NUD

[symbol]

ERROR MESSAGES

not a user-defined array

An attempt has been made to remove dimension information for
an array that was never defined.
PAR

Parentheses required, please retype
Parentheses are required for
the
statements and complex constants.

PRO

specification

FORMAT

Too many PAUSE requests
The PAUSE table has been exhausted.

The

maximum

limit

10.
RGR

Recursive group reference
A group may not reference itself.

SER

Subscript error
The subscript specified is outside the range of its
dimensions.

STL

[array name]

defined

size too large

An attempt has been made to

define

array

larger

than

256K.
TMS

Too many subscripts [dimensions]
The array is defined
to
have
fewer dimensions
specified in the given element reference.

URC

than

are

Unrecognized command [command]

Warning Messages
Each warning message is
preceded
by %FDT on your
terminal.
warning messages are listed here in alphabetical order.
ABX

[array name]

The

compiled array bounds exceeded

FORDDT has detected another symbol defined in the
specified
range of
the
array.
Note that this will occur in certain
EQUIVALENCE cases and can be ignored at that time.
CAB

Cannot allocate buffer for help file

CHI

Characters ignored:

"[text]"

The portion of the command string
included
thought to be extraneous and was ignored.
ECI

Buffer full -

excess characters ignored

EOH

Error opening help file

IOE

I/O error reading help file

E-4

"text"

was

INTEHACTIVE DEBUGGER (FORDDT)
IWI

ERROR MESSAGES

I/O within I/O
FORDDT has PAUSEd at a breakpoint while the user program
is
currently
processing
an
IOLST call.
TYPE and ACCEPT
commands cannot be PJ~ocessE~d at this breakpoint.
aJ~ray

NAA

[symbol]

is not an

NHF

Cannot find help file;

NSL

No symbols loaded

I'm sorry I can't help you

FORDDT cannot find the symbol table.
NST

Not STAHTed
The specified command requ:lres that a
START be
previously
issued to ensure tha~ the program is properly initialized.

POV

Program overlayed by [program name]
Program overlayed by ***
The symbol table is different from the last time FOR DDT had
control.
The
program name
is printed only if
it has
changed" otherwise ,.***, is printed.

SCA

Supersedes compiled

~rray

dimension

The FowrRAN generated dimension is being superseded for
g iven al~ray.
SPO

Variable is single-precision only

WSP

Writing to shared page

XPA

Attempt to exceed

pr~gram

area with [symbol name]

An attempt has been made
to
access
currently defined pr~gram space.

E-5

the

memory

outside

the

APPENDIX F
FORTRAN-SUPPLIED PLOTTER SUBROUTINES

The FORTRAN subroutine library contains a set of subroutines that are
used with
plotting devices.
To successfully use these routines, a
plotter must be connected
to
your system.
The
FORTRAN software
contains the following plotter subroutines:
PLOTS

The PLOTS subroutine initializes the plotter or
reports if
the plotter
is not available.
This must be the first
plotter subroutine specified.
(See Section F.I.)

AXIS

The AXIS subroutine ciraws an axis with tic marks and
scale
values at I-inch increments.
An identifying label may also
be plotted along the axis.
(See Section F.2.)

CAXIS

The CAXIS subroutine performs the same functions as the AXIS
subroutine.
(See Section F.3.)

LINE

The
LINE subroutine draws a
specified.
(See Section F.4.)

line

MKTBL

The MKTBL subroutine
(See Section F.S.)

NUMBER

The NUMBER subroutine causes floating-point
plotted as text.
(See Section F.6.)

PLOT

The PLOT subroutine lnoves the plotter pen to a new position.
Raising
and
lowering the pen is also specified in the PLOT
subroutine.
(See Se=tion F.7.)

SCALE

The SCALE subroutine selects scale values for
CAXIS subroutine.
(See Section F.8.)

SETABL

The SETABL subroutine
Section F.9.)

SYMBOL

The SYMBOL subroutine raises the plotter pen, moves
it to
the position specified by x and y, and plots a string of
characters.
(See Section F.IO.)

WHERE

The WHERE subroutine reports on the current position of
plotter pen.
(See Section F.II.)

specifies

F-l

through

special

the

points

character

set.

numbers

character

the

AXIS

set.

or
(See

the

FORTRAN-SUPPLIED PLOTTER SUBROUTINES

PLOTS
Subroutine

F.l

PLOTS SUBROUTINE

The PLOTS subroutine initializes the plotter or reports if the plotter
is
not available.
This must be the
first
plotter subroutine
specified.
The form of the PLOTS subroutine is:
CALL PLOTS(i[,steps])
where:
is an integer variable, which
is set to -1
if the
plotter is not available, or set to a if the plotter is
available.

The plotter may not be available because:
the system
does not have a
plotter;
the plotter is in use by
another user; the plotter is turned off; or the plotter
is being spooled, but you are trying to write to it
directly.
steps

is an optional floating-point variable or constant that
specifies the number of steps (per inch) used by the
plotter.
The default is 100 steps per inch, and may be
changed by the installation.
Commonly, plotters plot
100 steps per inch, 200 steps per inch, or
100 steps
per centimeter (about 254 steps per inch).

AXIS
Subroutine

F.2

AXIS SUBROUTINE

The AXIS subroutine draws an axis with tic marks and scale values at
I-inch increments.
An identifying label may also be plotted along the
axis.
NOTE
The AXIS subroutine is provided for compatibility with
previous
versions
of
FORTRAN-IO/20.
The AXIS
subroutine uses a numeric array to contain the label
that
is plotted along the axis.
The CAXIS subroutine
(see Section F.3) allows a character expression to
contain the label.

F-2

FORTRAN-SUPPLIED PLOTTER SUBROUTINES
The form of the AXIS subroutine is:
CALL AXIS{x,y,asc,nasc,size,theta,xmin,dx)
where:
x,y

is a variable or constant
starting point of the axis.

pair

that

specifies

asc

is the name of a numeric array that
that is plotted along the axis.

nasc

is an integer constant or variable that specifies the
number of characters contained in array asc.
If nasc
is negative, the label in the array is placed on the
clockwise side of the axis.
If nasc is positive, the
tic marks, label, and scale values are placed on the
counterclockwise side of the axis.

size

is a constant or variable that specifies the length
the axis in inches.

theta

is a constant or variable that specifies the angle at
which the axis
is plotted.
The value of theta is
usually 0.0 or 90.0.

xmin

is a variable or constant that specifies the
the scale at the beginning of the axis.

value

is a variable or constant that specifies the change
scale for a I-inch increment.

contains

the
label

NOTE
The proper values
for
xmin and dx may be
determined by calling the SCALE subroutine (see
Section F.8).

CAXIS
Subroutine

F.3

CAXIS SUBROUTINE

The CAXIS subroutine draws an axis with tic marks and scale values at
I-inch increments.
An identifying label may also be plotted along the
axis.
The form of the CAXIS subroutine is:
CALL CAXIS{x,y,asc,sign,size,theta,xmin,dx)

F-3

FORTRAN-SUPPLIED PLOTTER SUBROUTINES
where:
x,y

is a variable or constant
starting point of the axis.

pair

that

specifies

the

asc

is the name of a character expression that
label that is plotted along the axis.

sign

If nasc is negative, the label in the array is placed
on the clockwise side of the axis.
If nasc
is
positive, the tic marks, label, and
scale values are
placed on the counterclockwise side of the axis.

size

is a constant or variable that specifies the length
the axis in inches.

theta

is a constant or variable that specifies the angle at
which
the axis is
plotted.
The value of theta is
usually 0.0 or 90.0.

xmin

is a variable or constant that specifies the
the scale at the beginning of the axis.

value

is a variable or constant that specifies the change
scale for a I-inch increment.

contains

NOTE
The proper values
for
xmin and dx may be
determined by calling the SCALE subroutine (see
Section F.8).

LINE
Subroutine

F.4

LINE SUBROUTINE

The LINE subroutine draws a continuous line through a set of points.
The form of the LINE subroutine is:
CALL LINE(x,y,n,k)
where:
x

is the name of an array that contains the
x-coordinates of the points to be plotted.

floating-point

is the name of an array that contains the
y-coordinates of the points to be plotted.

floating-point

F-4

FORTRAN-SUPPLIED PLOTTER SUBROUTINES
n

is an integer constant or variable that specifies the
number of points to be plotted.

total

is an integer constant or variable that equals the number of
elements of x and y.
Since single-precision one-dimensional
arrays are usually used, this value is usually 1.

MKTBL
Subroutine

F.S

MKTBL SUBROUTINE

The MKTBL subroutine defines a special character set to be
used when
plotting;
the SETABL subroutine (see Section F.9) enables you to use
the character set defined by the MKTBL subroutine.
The form of the MKTBL subroutine is:
CALL MKTBL(setnumber,tableaddress)
where:

F.S.l

setnumber

is an integer variable from 1 to 10 that specifies
the numeric
identifier of the character set, for
example, the number of the ASCII character set is
1.
If the character set cannot be defined by a
call to MKTBL, a value of zero is returned in this
variable.

tableaddress

is a constant or variable that specifies the
starting address of a character table that has 128
(200 octaJ)
consecutive words.
Each character
table word contains the number of strokes (line
segments) for the character in the left half,
and
the addr~ss of the table of strokes in the right
half.
Se~ Section
F.S.l
for
a description of
these tabJes.

Character Tables

The next sections describe how to define and organize a character
table.
The charc3lcter set, called by the MKTBL subroutine, enables
to create and use a character set other than the default character
that
is
used to plot characters.
(Usually the default character
is ASC I I.)

set
you
set
set

To create your own character set, you need to create a character table
and
a character stroke table.
These two tables are descr ibed in
Sections F.S.l.l and F.S.l.2, respectively.

F-5

FORTRAN-SUPPLIED PLOTTER SUBROUTINES
F.S.l.l Creating a Character Table - A character table contains
128
entries (200 octal).
Each entry in the character table, regardless of
whether the table is defined in the plotter subroutine library or by
you through a call to the MKTBL subroutine, indicates the character to
be plotted for the ASCII character that has that numerical value.
Figure F-l is a diagram of an entry word in a character table.

Strokes

Address

~~
18 bits
18 bits
MnS1752·81

Figure F-l:

Plotter Character Table Entry

As shown in Figure F-l, each entry in the character table contains the
number of
strokes
(line segments) required to plot the character in
its left half and an address reference in the right half of the
word.
The
address in the right half of the entry references an entry in the
character stroke table for the character set.
The character
stroke
table is described in the next section.

F.S.l.2 Creating a Stroke Table - Each character
in
the character
table
has a corresponding character stroke table.
The purpose of the
character stroke table is to define the number
and
type of strokes
(drawn line segments) it takes to produce a character in the character
table.
Figure F-2 is a diagram of an entry in a character stroke table.

I
\

pen

Three 5-bit bytes

Fig ure F-2:

1~'<~81

Character Stroke Table Entry

As shown in Figure F-2, each stroke
is
stroke
table by three
5-bit bytes.
three bytes are:

described
in
the character
The possible values for these

Value and Meaning

Byte
1

MH c;

o
1

pen is raised (off the paper)
pen is lowered (on the paper)

x-coordinate value at end of stroke

y-coordinate value at end of stroke

F-6

FORTRAN-SUPPLIED PLOTTER SUBROUTINES
When determining the height and width of each character to be plotted,
consider the following conventions:
1.

Characters are drawn within a grid that is IS units
eight units wide.

high

Characters are generally plotted six
units above the base
line.
Two
units are generally left blank to the right of
each character; one cnit is generally left blank on top of
each character.
The spacing above, below, and on either side
of each character
provides
adequate
spacing
between
characters and between lines.

The plotter starts drawing each character at the
lower left
corner of the character grid.
If the grid is set up in the
conventional manner, the lower left corner is grid
position
(0,6).
Normal width characters end at the lower right corner
of the character grid
after
allowing
spacing
between
characters.
The last coordinate in each character grid is usually
(8,6).
The character grid for the next character has as its origin
(0,0) at the ending coordinate in the previous grid (8,6).

Accents, circumflexes, underscores, and other characters that
are to be plotted
in the same character grid as another
character should end at the same grid position as they began.
By
doing
this,
for
example,
the ASCII character
that
represents the accent character
is plotted before
the
character that is thE! letter to be accented.

F.S.l.3 Sample Character Stroke Table - In this sample, you want to
plot the Greek letter beta as a normal sized character.
When plotted,
the letter is drawn on a IS x 8 unit grid
above
the base line,
allowing for spacing between c:haracters and between lines.
The strokes
(line segments)
for
beta,
including
the
invisible
segments,
are then determined.
The character will be plotted more
smoothly if as many of the liIle segments as possible are connected,
and
if doubled segments are avoided.
Note that stroke 9 (the return
stroke for the crossbar of the beta) and stroke IS (moving the pen to
ending point (8,6) are not visible; the pen is raised.

F-7

FORTRAN-SUPPLIED PLOTTER SUBROUTINES

15
4

Y
2

~
6

8,9

N°

~4
1/

1V 15
~

Base Line

°
Figure F-3:

rv'In S 1754·81

Sample Character Stroke Table

If the address at the beginning of the
table that
includes the
character
in Figure F-3 is called GREEK, then the FORTRAN call to set
up the table could be:
IGREEK=4
CALL MKTBL (IGREEK, GREEK)
The above call to the MKTBL subroutine defines the GREEK character
table
to be the fourth character table.
To use the GREEK table for
plotting, use the FORTRAN call:
CALL SETABL (IGREEK, IFLAG)
The entry in the character table for beta is:
RADIX
BETA:

DIO
iVALUES IN DECIMAL
IS"TBETA iCHARACTER TABLE ENTRY FOR BETA

Note that the character table entry for beta contains the starting
address of the character stroke table (TBETA) in its right half.
The character stroke table for beta is:

TBETA:

SEEN==1
iIF 'SEEN' THEN THE STROKE MARKS THE PAPER
UNSEEN==O
iIF 'UNSEEN' THEN THE STROKE IS INVISIBLE
RADIX ~DIO
iALL VALUES ARE IN DECIMAL
BYTE{S) SEEN,2,8,SEEN,2,13
BYTE{S) SEEN,3,14,SEEN,S,14
BYTE{S) SEEN,6,13,SEEN,6,12
BYTE{S) SEEN,S,II,SEEN,2,11
BYTE{S) UNSEEN,S,11,SEEN,6,IO
BYTE{S) SEEN,6,9,SEEN,S,8
BYTE(S) SEEN,3,8,SEEN,2,9
BYTE{S) UNSEEN,8,6
F-8

FORTRAN-SUPPLIED PLOTTER SUBROUTINES
F.S.I.4
FORTRAN- and User-Defined Character Sets - The standard ASCII
character set is always defined and is character set I for calls to
the SETABL subroutine, unless character set 1 is redefined by a
user
call
to
the MKTBL subroutine.
If SETABL is not called, the ASCII
character set is the default.
The Cyrillic (Russian) character set is available as character set
2,
and
the
Feanorian character
set
is available as character set 3
(unless character sets 2 or 3 has been redefined by a user call to the
MKTBL subroutine).
In order
to use these character sets, the user
program must contain an
EXTERNAL statement for
variable
PLOTF
(Feanorian) or PLOTC (Cyrillic).
User-defined character sets should use character sets 4 through 10
avoid conflicts with the standard character sets.

NUMBER
Subroutine

F.6

NUMBER SUBROUTINE

The NUMBER subroutine causes floating-point numbers to be
text.

plotted

The form of the NUMBER subroutine is:
CALL NUMBER(x,y,size,fnum,theta,ndigit)
where:
x,y

are variables or constants that specify the x and
y
coordinates of the point to be plotted.
The specified
point is the lower left corner of the
first
character
to be plotted.

size

is a variable or constant that specifies the
size
(in
inches)
of the digits
to be plotted.
The specified
value
should
be a multiple of
.08
inches
(or
centimeters if plotter is metric) if a small value is
used.

fnum

is a variable or constant that
number to be plotted.

theta

is a variable or constant that specifies the direction
(in degrees)
of the base line on which the characters
are plotted.

ndigit

is a variable or constant that specifies the number of
digits to be p~.otted to the right of the decimal point.
If ndigit is negative, only the
integer
part of the
n urn be r i s plot ted; the res u 1 t i ng in t eg e r i s r 0 u nd ed •

F-9

the

floating-point

FORTRAN-SUPPLIED PLOTTER SUBROUTINES

PLOT
Subroutine

F.7

PLOT SUBROUTINE

The PLOT subroutine moves the plotter pen to a new position.
Raising
and lowering the pen is also specified in the PLOT subroutine.
NOTE
The plotter is not released after
specified movement.

completion

the

The form of the PLOT subroutine is:
CALL PLOT(x,y,penup/down)
where:
x,y

are the variables or constants that specify the x
and
y coordinates for the point to which the pen
will be moved.

penup/down

is an integer constant or variable that
specifies
whether
the
pen
is on the
paper or above the
paper.
The possible values for this variable are:
3
2
1

raise pen before movement
lower pen before movement
leave pen in current state (raised or lowered)

-1, -2, or -3

same as corresponding
positive
values except that on completion
of the indicated motion,
the new
pen position
is taken as a new
origin and the output buffer
is
sent
to
the
plotter.
Using the
negative values is helpful if you
are
plotting
consecutive
characters in the same program.

SCALE
Subroutine

F.B

SCALE SUBROUTINE

The SCALE subroutine scales values for the AXIS subroutine.
The form of the SCALE subroutine is:
CALL SCALE(x,n,s,xmin,dx)
F-IO

FORTRAN-SUPPL1ED PLOTTER SUBROUTINES
where:

is an array nam~ of a
one-dimensional
floating-point
array to be scaled for the AXIS subroutine.

is an integer C(lnstant or variable that
length of the at"ray in words (36-bit).

is a constant or variable that specifies the length (in
inches) of the desired axis.

xmin

is a constant or variable that specifies the
smallest
element
in an"ay x."
The value of xmin will be the
value of the scale at the beginning of the axis.

i s a c 0 n s tan t 0 I" va ria b 1 e t hat equals the change
in
scale
for
a
I-inch
interval
so that array x can be
plotted in inches.

specifies

the

SETABL
Subroutine

F.9

SETABL SUBROUTINE

The SETABL subroutine enables
used to plot characters.

fOU

to select the character set that

The form of the SETABL subroutine is:
CALL SETABL (setnum,status)
where:
setnum

is an integer constant or variable that equals the
number of the character
set.
The standard ASCII
character set is defined
to be set 1 and
is the
default.
Character sets are defined by the MKTBL
subroutine (see Section F. 5) •

status

is an integer variable whose value after
the call
to
the SETABL subroutine is either 0, if the character set
specified in setnum is valid, or -1, if the character
set specified by setnum is invalid.
NOTE
If you use
a character set other
than
the
character
sets defined
by default
in the
plotter subroutine library, you must call
the
MKTBL subroutine before calling
the SETABL
subroutine.

F-II

FORTRAN-SUPPLIED PLOTTER SUBROUTINES

SYMBOL
Subroutine

F.IO

SYMBOL SUBROUTINE

The SYMBOL subroutine plots a specified string of characters
(from
either the default character set or the character set specified by the
last successful call to the SETABL subroutine) .
The form of the SYMBOL subroutine is:
CALL SYMBOL(x,y,size,asc,theta[,nasc])
where:
x

is a constant or variable that equals the x coordinate
of the lower left corner of the first character to be
plotted.

is a constant or variable that equals the y coordinate
of the lower left corner of the first character to be
plotted.
The plotter pen is raised and moved to
pos i t ion
x,y before the string of characters is
plotted.

size

is a constant or variable that specifies the height (in
inches)
of the character to be plotted.
The specified
value should be a multiple of
.08
inches
(or
c en tim e t e r s i f yo u h a v e a met ric plot t e r) •

asc

is the name of a character expression or numeric array
that contains the ASCII characters to be plotted.

theta

is a constant or variable that specifies the direction
(in degrees)
of the base line on which the characters
are to be plotted.

nasc

is an integer constant or variable that is equal to the
number of characters in numeric array asc that are to
be plotted. This is ignored if a character expression
is specified for asc.

WHERE
Subroutine

F.Il

WHERE SUBROUTINE

The WHERE subroutine reports on the current position
pen, in inches, relative to the origin.

F-12

the

plotter

FORTRAN-SUPPLIED PLOTTER SUBROUTINES
NOTE
The plotter origin is set by a
call
to
the
AXIS
subroutine or a call to the PLOT subroutine that has a
negative value for the penup/down variable.
Also, the
WHERE subroutine does
not allow you to determine
whether the plotter pen is raised or lowered.
The form of the WHERE subroutine is:
CALL WHERE(x,y)
where:
x

is a variable
in which the subroutine
coordinate of the current print position.

returns

the

is a variable
in which the
subroutine
coordinate of the current pen position.

returns

the

F-13

INDEX

A ed i t i ng, 12 - 3 9
ABORT switch, 16-9
ACCEPT
seq u e n t i alL i s t - d ire c ted, 1 (! - 4 1
ACCEPT statement, 10-40
sequential FORMAT-statement,
10-40
Access
direct, 10-6, 18-4
sequential, 10-5, 18-3
ACCESS specifier
in INQUIRE, 11-43
in OPEN, 11-6
Accumulator usage, 15-11
Actual arguments, 13-49
Length of character, 13-51
Adjustable dimensions, 7-3
ALCCHR subroutine, 13-24
ALL with DEBUG, 16-12
ALL with FLAG, 16-22
ALL with NOWARN, 16-15
.AND. logical operator, 5-8
ANSI standard, 1-1
ANSI with FLAG, 16-22
Apostrophe editing, 12-16
Argument lists, 15-12
Argument types, 15-14
Argumen ts
Actual, 13-49
Dummy, 13-49
Arguments to FLAG, 16-22
ARGUMENTS with DEBUG, 16-12
Arithmetic assignment statement,
8-1
Arithmetic constant expressions,
5-5
Arithmetic expressions, 5-1
writing, 5-2
Arithmetic IF statement, 9-3
Arithmetic operations, 5-2
Arithmetic operator3, 5-2
Array declarator, 4-4, 7-1
Array element subscript, 4-3
Array elements
order of stored, 4-5
Array names, 4-3
Arrays, 4-3
assumed-size, 7-4
dimensioning, 4-4
Dimensioning dummy, 15-5
ASCII character code, B-1
ASCII data files, 18-4
ASSIGN statement, 8-3
Assigned GO TO statement, 9-3
Assignment statement, 8-1
ar i thmetic, 8-1
character, 8-4
Assignment statements
logical, 8-3

Assignments
mixed~mode, 8-2
ASSOCIATEVARIABLE specifier
in OPEN, 11-8
Assumed-size arrays, 7-4
AXIS subroutine, F-2

BACKFILE statement, 11-57
BACKSPACE statement, 11-55
Binary data files, 18-4
BINARY switch, 16-9
Bit data constants, 7-23
Bit manipulation functions, 13-14
Blank lines, 2-7
BLANK specifier
in INQUIRE, 11-44
in OPEN, 11-9
BLISS-36
calling routines, 15-19
Interaction with, 15-18
BLKRW subroutine, 19-4
BLOCK DATA statement, 14-1
Block data subprograms, 14-1
Block IF constructs
indexed, 9-9
Block IF statements, 9-5
BLOCKSIZE specifier
in OPEN, 11-10
BN editing, 12-28
Boolean values, 3-6
BOUNDS with DEBUG, 16-12
BUFFERCOUNT specifier
in OPEN, 11-10
BZ editing, 12-28
CALL statement, 13-44
Calling BLISS-36 routines, 15-19
Calling COBOL subroutines, 15-18
Calling FORDDT, 17-17
Calling sequencEs
FOROTS, 18-16
Carriage-control specifiers, 12-9
CARRIAGECONTROL specifier
in INQUIRE, 11-44
in OPEN, 11-11
CAXIS subroutine, F-3
CBC function, 18-34
CDABS function, 13-25
CDCOS subroutine, 13-25
CDEXP subroutine, 13-26
CDLOG subroutine, 13-26
CDSIN subroutine, 13-27
CDSQRT subroutine, 13-28
Channel allocation routines,
18-30
Channel deallocation routines,
18-30
Character assignment statement,
8-4

Ind ex--l

Character code
ASCII, B-1
Character comparison functions,
13-14
Character constant expressions,
5-7
Character constants, 3-4
Character expressions, 5-6
Character intrinsic functions,
13-12
Character set, 2-1
FORTRAN-defined, F-9
User-defined, F-9
Character substrings, 4-6
Character tables, F-5
Character type specification
s ta temen ts, 7-6
CHARACTER*len, 7-6
CHKDIV subroutine, 13-28
CLOSE
DEVICE specifier in, 11-36
DIALOG specifier in, 11-36
DIALOG= specifier in, 11-37
DIRECTORY specifier in, 11-36
DISPOSE specifier in, 11-37
ERR specifier in, 11-39
FILE specifier in, 11-36
Implicit, 11-34
IOSTAT specifier in, 11-39
LIMIT specifier in, 11-39
NAME specifier in, 11-36
PROTECTION specifier in, 11-36
STATUS specifier in, 11-40
UNIT specifier in, 11-41
CLOSE specifiers
Summary of, 11-35
CLOSE statement, 11-34
CLOSE statement specifiers, 11-34
CLRFMT subroutine, 13-29
CLROVL in LINK, 15-20
COBOL
calling subroutines, 15-18
Interaction with, 15-16
CODE with EXTEND, 16-20
Colon editing, 12-24
Commands
COMPILE, 16-1, 16-5
DEBUG, 16-1, 16-5
EXECUTE, 16-1, 16-5
FORDDT, 17-1, 17-7
LOAD, 16-1, 16-5
Comment lines, 2-5
Common block, 7-9
Common block name, 7-9
COMMON statement, 7-8
Common storage, 7-9
COMMON with EXTEND, 16-20
Compilation control statements,
6-4
COMPILE command, 16-1, 16-5
Compiler
Running FORTRAN-I0, 16-2
Running FORTRAN-20, 16-6

Compiler (Cont.)
using, 16-1
Compiler errors
internal, C-13
Compiler generated label, 16-23
Compiler generated variables,
16-24
Compiler listing, 16-22
Compiler messages, C-l
Compiler switches
FORTRAN-I0, 16-2
FORTRAN-20, 16-9
Compiling programs, 16-1, 16-5
Complex constants, 3-4
Complex editing, 12-36
Computed GO TO statements, 9-2
Concatenation operator, 5-6
CONNECT subroutine, 19-4
Constant expressions, 5-15
arithmetic, 5-5
character, 5-7
integer, 5-5
logical, 5-10
Constant folding, 15-8
Constant propagation, 15-8
Constants, 3-1
bit data, 7-23
character, 3-4
complex, 3-4
double-octal, 3-5
double-precision, 3-2
Hollerith, 3-6
integer, 3-1
logical, 3-6
octal, 3-5
real, 3-2
statement label, 3-7
Continuation lines, 2-4
CONTINUE LSCW, 18-4
CONTINUE statement, 9-17
Control information list, 10-9
Control statements, 9-1
COR function, 18-33
CROSSREF switch, 16-3, 16-9
D editing, 12-32
D-floating double-precision
format, 3-3
Data access, 10-5
Data magnitude on G-format, 12-35
DATA statement, 7-21
Data transfer operations, 10-5
Data transfer statements, 10-1
Data transfers
Formatted, 12-1
formatted, 10-6
unformatted, 10-8
DATA with EXTEND, 16-20
DATE subroutine, 13-29
DEBUG
and FORDDT, 17-4
DEBUG arguments
ALL, 16-12

Index- 2

DEBUG arguments (Cont.)
ARGUMENTS, 16-12
BOUNDS, 16-12
DIMENSIONS, 16-12
INDEX, 16-12
LABELS, 16-12
NONE, 16-12
TRACE, 16-12
DEBUG command, 16-1, 16-5
Debug lines, 2-6
DEBUG switch, 16-3, 16-9, 16-11
Debugger
FORDDT, 17-1
Debugging programs, 16-1, 16-5
DECODE statement, 10-51
Default memory layout
/EXTEND, 16-15
Deferred OPEN statement, 11-2
DENSITY specifier
in OPEN, 11-12
Device assignments
logical, 10-12
Device control statements, 10-1,
11-52
DEVICE specifier
in CLOSE, 11-36
in OPEN, 11-13
Devices
real-time, 19-1
DFLOATING switch, 16-3, 16-9
DIALOG specifier
in CLOSE, 11-36
in OPEN, 11-13
DIALOG= specifier
in CLOSE, 11-37
in OPEN, 11-14
Dimension declarator, 7-2
DIMENSION statement, 7-1
Dimensioning arrays, 4-4
Dimensioning dummy arrays, 15-5
DIMENSIONS with DEBUG, 16-12
Direct access, 10-6
direct access, 18-4
Direct FORMAT-statement READ,
10-27
Direct FORMAT-statement WRITE,
10-33
DIRECT specifier
in INQUIRE, 11-44
Direct unformatted READ, 10-30
Direct unformatted WRITE, 10-36
DIRECTORY specifier
in CLOSE, 11-36
in OPEN, 11-14, 11-16
DISCON subroutine, 19-7
DISMIS subroutine, 19-6
DISPOSE specifier
in CLOSE, 11-37
in OPEN, 11-16
DIVERT subroutine, 13-30
DO iteration control, 9-12
DO lists
implied, 10-22

DO loop extended range, 9-15
DO loop iterations, 15-2
DO loops
Floating-point, 15-2
DO statements
nested, 9-14
DO WHILE statement, 9-13
Dollar sign editing, 12-22
Double-octal constants, 3-5
Double-precision constants, 3-2
Double-precision format
D-floating, 3-3
G-floating, 3-3
DTOG function, 13-19
DTOGA subroutine, 13-30
Dummy arguments, 13-49
Length of character, 13-51
DUMP subroutine, 13-30
E ( e x po n e n t i a 1 not a t ion), 3 - 2
E editing, 12-32
ECHO-OPTION switch, 16-9
Edit descriptors
FORMAT-statement, 12-6
Nonrepeatable, 12-8
numer ic, 12-29
Repeatable, 12-7
Ed i ting
A, 12-39
Apostrophe, 12-16
BN, 12-28
BZ, 12-28
Colon, 12-24
complex, 12-36
D, 12-32
Dollar sign, 12-22
E, 12-32
F, 12-32
G, 12-33
H, 12-17
I, 12-31
L, 12-38
Nurn e ric, 1 2 - 2 9
o (octal), 12-36
P, 12-25
Positional, 12-18
Q, 12-29
R, 12-40
S, 12-24
Slash, 12-23
SP, 12-24
SS, 12-24
T, 12-19
TL, 12-20
TR, 12-20
X, 12-21
Z (hexadecimal), 12-37
ELSE IF THEN statement, 9-5
ELSE statement, 9-5
Embedded format specifications,
12-15
ENCODE statement, 10-51
END DO statement, 9-17

Index-3

END IF statement, 9-5
END LSCW, 18-4
END statement, 9-20
END= specifier, 10-17
ENDFILE statement, 11-55
Entry points
FOROTS, 18-15
ENTRY statement, 13-45
EQUIVALENCE and extended
addressing, 7-17
EQUIVALENCE statement, 7-11
.EQV. logical operator, 5-8
ERR specifier
in CLOSE, 11-39
in INQUIRE, 11-45
in OPEN, 11-17
ERR= specifier, 10-18
Error codes
FOROTS, D-2
Error messages
FORDDT, E-l
FOROTS, D-l
real-time, 19-7
Error processing
FOROTS, 18-2
Error reporting, 16-33
Errors
fatal, C-2
Fatal compiler, 16-34
ERRSET subroutine, 13-31
ERRSNS subroutine, 13-32
ERRSNS values, D-2
Evaluation of expressions, 5-13
Executable statements, 1-1, 6-1
EXECUTE command, 16-1, 16-5
Executing programs, 16-1, 16-5
EXIST specifier
in INQUIRE, 11-45
EXIT subroutine, 13-33
EXPAND switch, 16-3, 16-9
Expressions
ar i thmetic, 5-1
arithmetic constant, 5-5
character, 5-6
character constant, 5-7
constant, 5-15
evaluation of, 5-13
integer constant, 5-5
log ical, 5-7
logical constant, 5-10
mixed-mode, 5-15
relational, 5-11
EXTEND
applications with large arrays,
16-16
applications with large code,
16-17
EXTEND arguments, 16-19
CODE, 16-20
COMMON, 16-20
DATA, 16-20
NOCOMMON, 16-20
NODATA, 16-20

EXTEND arguments (Cont.)
PSECT, 16-20
EXTEND psect placement, 16-17
EXTEND switch, 16-9, 16-15
Extended addressing, 16-15
and EQUIVALENCE, 7-17
and FORDDT, 17-17
default memory layout, 16-15
memory layout, 15-22
Extended range DO loop, 9-15
External functions, 13-18
FORTRAN-supplied, 13-18
user-defined, 13-20
EXTERNAL statement, 7-17
F editing, 12-32
F66 switch, 16-3, 16-9
F77 switch, 16-3, 16-9
Fatal compiler errors, 16-34
Fatal errors, C-2
FFUNIT subroutine, 13-33
Field
line continuation, 2-3
remark, 2-4
statement, 2-4
statement label, 2-3
File control statements, 10-1,
11-1
File identifier
internal, 10-13
FILE specifier
in CLOSE, 11-36
in OPEN, 11-18
Files
ASCII data, 18-4
binary data, 18-4
internal, 10-7, 10-49
linking TWOSEG REL, 16-21
non-FORTRAN, 15-10
FILESIZE specifier
in OPEN, 11-18
FIND statement, 11-53
FLAG a rg umen ts
ALL, 16-22
ANSI, 16-22
NOANSI, 16-22
NONE, 16-22
NOVMS, 16-22
VMS, 16-22
FLAG switch, 16-3, 16-21
FLAG switch arguments, 16-22
FLAG-NON-STANDARD switch, 16-9
Floating-point
DO loops, 15-2
FORDDT
and extended addressing, 17-17
calling, 17-17
load ing, 17-5
starting, 17-5
FORDDT and DEBUG, 17-4
FORDDT commands, 17-1, 17-7
FORDDT debugger, 17-1
FORDDT error messages, E-l

Index-4

FORM specifier
in INQUIRE, 11-45
in OPEN, 11-19
format control, 12-11
Format editing, 12-16
Format list
I/O list interaction, 12-11
format list, 12-3
Format specifications
embedded, 12-15
Format specifier, 10-13
FORMAT statement, 12-2
FORMAT-statement ACCEPT
sequential, 10-40
FORMAT-statement edit descriptors,
12-6
FORMAT-statement formatting,
10-14, 12-2
FORMAT-statement PRINT
sequential, 10-45
FORMAT-statement PUNCH
sequential, 10-48
FORMAT-statement READ
direct, 10-27
equential, 10-26
FORMAT-statement REREAD
sequential, 10-38
FORMAT-statement TYPE
sequential, 10-43
FORMAT-statement WRITE
direct, 10-33
sequential, 10-32
Formatted data transfers, 10-6,
12-1
Formatted input, 12-14
Formatted output, 12-15
Formatted READ transfers, 10-2~
FORMATTED specifier
in INQUIRE, 11-45
Formatted WRITE transfers, 10-31
Formatting
FORMAT-statement, 10-14, 12-2
list-directed, 10-16, 12-41
NAMELIST-statement, 10-16,
12-44
FOROTS, 18-1
and INQUIRE by file, 18-36
MACRO calls for, 18-17
RESET in, 15-11
sharable, 15-10
using, 18-14
FCROTS calling sequences, 18-16
FOROTS entry points, 18-15
FOROTS error codes, D-2
FOROTS error messages, D-l
FOROTS error processing, 18-2
FOROTS features, 18-1
FOROTS input/output facility,
18-2
FOROTS memory management, 15-20
FORRTF subroutine, 19-1
FORTRAN-I0
compiler switches, 16-2

FORTRAN-I0 (Cont.)
running the compiler, 16-2
FORTRAN-I0/20, 1-1
]~ORTRAN-20

Running the compiler, 16-6
FORTRAN-20 compiler
using, 16-5
FORTRAN-20 compiler switches,
16-9
FORTRAN-77 standard, 1-1
FORTRAN-defined character set,
F-9
FORTRAN-supplied external
functions, 13-18
FORTRAN-supplied plotter
subroutines, F-l
FORTRAN-supplied subroutines,
13-24
Function effects, 15-10
FUNCTION statement, 13-20
Function subprogram, 13-20
Function subprograms
using, 13-21
Functions, 13-1
bit manipulation, 13-14
CBC, 18-34
CDABS, 13-25
character comparison, 13-14
character intrinsic, 13-12
COR, 18-33
DTOG, 13-19
External, 13-18
FORTRAN-supplied external,
13-18
GAD, 18-32
GCH, 18-33
Generic, 13-2
GOT, 18-33
GPG, 18-34
GPSI, 18-35
GTOD, 13-19
IFS, 18-34
ILL, 18-32
Intrinsic, 13-1
LSNGET, 13-19
MPG, 18-35
RAD, 18-33
RAN, 13-19
RANS, 13-19
RCH, 18-33
RNT, 18-34
ROT, 18-34
RPG, 18-35
RPSI, 18-35
RRS, 18-34
SECNDS, 13-19
Spec i fic, 13-2
Statement, 13-15
TIM2GO, 13-19
UPG, 18-36
user-defined external, 13-20
Using intrinsic, 13-2
Using statement, 13-16

Index--5

Functions (Cont.)
WRS, 18-34
Functions for overlays, 18-31
G editing, 12-33
G-floating double-precision
format, 3-3
G-format
Data magnitude on, 12-35
GAD function, 18-32
GCH function, 18-33
Generic functions, 13-2
GETOVL in LINK, 15-20
GFLOATING switch, 16-3, 16-9
Global optimization, 15-6
Global register allocation, 15-9
GO TO statement
assigned, 9-3
GO TO statements, 9-1
computed, 9-2
unconditional, 9-2
GOT function, 18-33
GPG function, 18-34
GPSI function, 18-35
GTOD function, 13-19
GTODA subroutine, 13-34
H editing, 12-17
Hierarchy of operators, 5-14
High segment
sharable, 16-35
Hollerith constants, 3-6
I editing, 12-31
I/O
optimization, 15-9
I/O 1 ist, 10-20
I/O list and format list
interaction, 12-11
IF statement
arithmetic, 9-3
logical two-branch, 9-4
IF statements, 9-3
block, 9-5
logical, 9-4
IF THEN statement, 9-5
IFS function, 18-34
ILL function, 18-32
ILL subroutine, 13-34
Implicit CLOSE, 11-34
Implicit OPEN, 11-3
IMPLICIT statements, 7-7
Implied DO lists, 10-22
INCLUDE statement, 6-4
INCLUDE switch, 16-3, 16-9
Index variable, 9-11
INDEX with DEBUG, 16-12
INIOVL in LINK, 15-20
Initial lines, 2-4
Initial tab, 2-3
INITIALIZE specifier
in OPEN, 11-18

Input
formatted, 12-14
NAMELIST-controlled, 12-45
Input/output facility
FOROTS, 18-2
Input/output statements, 10-1
INQUIRE by file and FOROTS, 18-36
INQUIRE specifiers
ACCESS, 11-43
BLANK, 11-44
CARRIAGECONTROL, 11-44
DIRECT, 11-44
ERR, 11-45
EXIST, 11-45
FORM, 11-45
FORMATTED, 11-45
IOSTAT, 11-46
NAME, 11-46
NAMED, 11-47
NEXTREC, 11-47
NUMBER, 11-47
OPENED, 11-48
RECL, 11-48
RECORDSIZE, 11-48
RECORDTYPE, 11-48
SEQUENTIAL, 11-49
UNFORMATTED, 11-49
INQUIRE statement, 11-42
by file, 11-42
by unit, 11-43
INQUIRE statement specifiers,
11-43
Integer constant expressions, 5-5
Integer constants, 3-1
Interaction with BLISS-36, 15-18
Interaction with COBOL, 15-16
Internal compiler errors, C-13
Internal file identifier, 10-13
Internal files, 10-7,· 10-49
Internal READ
statement, 10-50
Internal WRITE statement, 10-50
Intrinsic functions, 13-1
character, 13-12
Using, 13-2
INTRINSIC statement, 7-18
IOSTAT specifier, 10-19
in CLOSE, 11-39
in INQUIRE, 11-46
in OPEN, 11-20
Iteration control
DO, 9-12
Keywords, 1-1
L editing, 12-38
Labels
compiler generated, 16-23
LABELS with DEBUG, 16-12
Large arrays
and extended addressing, 16-16
Large code applications, 16-18

Index-6

Large executable code
and extended addressing, 16-17
LEGAL subroutine, 13-34
Length of character actual
arguments, 13-51
Length of character dummy
a r g urn en t s, 1 3 -- 51
Length specification, 7-6
LIMIT specifier
in CLOSE, 11-39
in OPEN, 11-21
Line continuation field, 2-3
LINE subroutine, F-4
Line types, 2-4
Line-sequence number, 16-22
Line-sequenced source files, 2-7
Lines
blank, 2-7
comment, 2-5
continuation, 2-4
debug, 2-6
initial, 2-4
multi-statement, 2-5
LINK
CLROVL subroutine, 15-20
GETOVL subroutine, 15-20
INIOVL subroutine, 15-20
LOGOVL subroutine, 15-20
REMOVL subroutine, 15-20
RUNOVL subroutine, 15-20
SAOVL subroutine, 15-20
LINK overlay facilities, 15-19
LINK overlay handler, 18-31
Linking TWOSEG REI... files, 16-21
List-directed ACCEPT
sequential, 10-41
List-directed formatting, 10-16
list-directed formatting, 12-41
List-directed PRINT
sequential, 10-46
List-directed PUNCH
sequential, 10-48
List-directed READ
sequential, 10-27
List-directed REREAD
sequential, 10-39
List-directed TYPE
sequential, 10-44
List-directed WRITE
sequential, 10-34
LISTING switch, 16-9
LNMAP switch, 16-3, 16-10
LOAD command, 16-1, 16-5
Loading FORDDT, 17-5
Loading programs, 16-1, 16-5
LOCK subroutine, 19-3
Logical assignment statements,
8-3
Logical constant expressions,
5-10
Logical constants, 3-6
Logical device assignments, 10-12
Logical expressions, 5-7

Logical IF statements, 9-4
Logical operators, 5-8
Logical segment control words,
18-4
Logical two-branch IF statement,
9-4
Logical unit identifier, 10-11
LOGOVL in LINK, 15-20
LSCW, 18-4
CONTINUE, 18-4
END, 18-4
START, 18-4
LSNGET function, 13-19
MACHINE-CODE switch, 16-10
MACRO calls for FOROTS, 18-17
MACROCODE switch, 16-3
Memory allocation routines, 18-29
Memory layout
extended addressing, 15-22
Memory management
FOROTS, 15-20
Over 1 a y, 1 5 - 2 0
Message summary, 16-34
Messages
compiler, C-l
warning, 16-34, C-8
Mixed-mode assignments, 8-2
Mixed-mode expressions, 5-15
Mixed-mode operations, 5-3
MKTBL subroutine, F-5
MODE specifier
in OPEN, 11-21
MPG function, 18-35
Multi-statement lines, 2-5
MVBITS subroutine, 13-35
Name
Common block, 7-9
NAME specifier
in CLOSE, 11-36
in INQUIRE, 11-46
in OPEN, 11-23
NAMED specifier
in INQUIRE, 11-47
NAMELIST statement, 12-44
NAMELIST-controlled input, 12-45
NAMELIST-controlled output, 12-46
NAMELIST-statement formatting,
10-16, 12-44
NAMELIST-statement READ
sequential, 10-28
NAMELIST-statement WRITE
sequential, 10-35
Names
array, 4-3
symbolic, 4-1
variable, 4-2
.NEQV. logical operator, 5-8
Nested Block IF constructs, 9-9
Nested DO statements, 9-14
NEXTREC specifier
in INQUIRE, 11-47

Index·-7

NOANSI with FLAG, 16-22
NOBINARY switch, 16-10
NOCOMMON with EXTEND, 16-20
NODATA with EXTEND, 16-20
NOERRORS switch, 16-3, 16-10
NOEXTEND switch, 16-10
NOF77 switch, 16-3, 16-10
NOFLAG switch, 16-3
NOFLAG-NON-STANDARD switch, 16-10
non-FORTRAN files, 15-10
non-FORTRAN programs, 15-10
NONE with
DEBUG, 16-12
NONE with FLAG, 16-22
NONE with NOWARN, 16-15
Nonexecutable statements, 1-1,
6-2
Nonrepeatable edit descriptors,
12-S
.NOT. logical operator, 5-S
NOVMS with FLAG, 16-22
NOWARN arguments
ALL, 16-15
NONE, 16-15
NOWARN switch, 16-3, 16-10, 16-14
NUMBER specifier
in INQUIRE, 11-47
NUMBER subroutine, F-9
Numeric edit descriptors, 12-29
Numeric editing, 12-29
Numeric type specification
statements, 7-5

o (octal) editing, 12-36
Object time system, lS-l
Octal constants, 3-5
OPEN
ACCESS specifier in, 11-6
ASSOCIATEVARIABLE specifier in,
11-S
BLANK specifier in, 11-9
BLOCKSIZE specifier in, 11-10
BUFFERCOUNT specifier in, 11-10
CARRIAGECONTROL specifier in,
11-11
DENSITY specifier in, 11-12
DEVICE specifier in, 11-13
DIALOG specifier in, 11-13
DIALOG= specifier in, 11-14
DIRECTORY specifier in, 11-14,
11-16
DISPOSE specifier in, 11-16
ERR specifier in, 11-17
FILE specifier in, II-IS
FILESIZE specifier in, II-IS
FORM specifier in, 11-19
Implicit, 11-3
INITIALIZE specifier in, 11-18
IOSTAT specifier in, 11-20
LIMIT specifier in, 11-21
MODE specifier in, 11-21
NAME specifier in, 11-23
PADCHAR specifier in, 11-24

OPEN (Cont.)
PARITY specifier in, 11-24
PROTECTION specifier in, 11-25,
11-27
READONLY specifier in, 11-28
RECL specifier in, 11-2S
RECORDSIZE specifier in, 11-2S
RECORDTYPE specifier in, 11-29
STATUS specifier in, 11-30
TAPEFORMAT specifier in, 11-32
TYPE specifier in, 11-30
VERSION specifier in, 11-34
OPEN statement, 11-1
deferred, 11-2
on connected unit, 11-4
OPEN statement specifiers, 11-4
Summary of, 11-6
OPENED specifier
in INQUIRE, 11-48
Operations
arithmetic, 5-2
Operators
arithmetic, 5-2
concatenation, 5-6
hierarchy of, 5-14
log ical, 5-S
relational, 5-11
Optimization
Global, 15-6
I/O, 15-9
Optimization techniques, 15-6
OPTIMIZE switch, 16-3, 16-10,
17-17
OPTION switch, 16-10
.OR. logical operator, 5-8
Order of stored array elements,
4-5
Ordering of statements, 6-2
Output
formatted, 12-15
NAMELIST-controlled, 12-46
OVERFL subroutine, 13-35
Overlay facilities
LINK, 15-19
Overlay handler
LINK, 18-31
Overlay memory management, 15-20
Overlays
Functions for, 18-31
P editing, 12-25
PADCHAR specifier
in OPEN, 11-24
PARAMETER statement, 7-20
Parenthetical subexpressions,
5-13
PARITY specifier
in OPEN, 11-24
PAUSE statement, 9-19
PDUMP SUbroutine, 13-36
PLOT subroutine, F-I0
PLOTS subroutine, F-2
Positional editing, 12-18

Index-S

PRINT
sequential FORMAT-statement,
10-45
sequential List-'directed, 10--46
PRINT statement, 10-45
Priority interrupt levels, 19-2
Priority interrupt system, 19-~.
PROGRAM statement, 6-4
Prog rams
compiling, 16-1, 16-5
debugg ing, 16-5
executing, 16-5
load ing, 16-1, 16-5
non-FORTRAN, 15-10
source, 1-1
wr i t i ng, 1 5-1
Programs debugging, 16-1
Programs executing, 16-1
PROTECTION specifier
in CLOSE, 11-36
in OPEN, 11-25, 11-27
Psect placement
/EXTEND, 16-17
PSECT with EXTEND, 16-20
PUNCH
sequential FORMAT-statement,
10-48
sequential List--directed, 10--48
PUNCH statement, 10-47
Q editing, 12-29
QUIETX Subroutine r 13-36
R editing, 12-40
RAD function, 18-33
RAN function, 13-19
RANS function, 13-19
RCH function, 18-33
READ
direct FORMAT-statement, 10-17
direct unformatted, 10-30
sequential FORMAT-statement,
10-26
sequential List-directed, 10·-27
sequential NAMELIST-statement,
10-28
Sequential unformatted, 10-29
READ statement, 10-24
internal, 10-50
READ transfers
formatted, 10-26
unformatted, 10-29
READONLY specifier
in OPEN, 11-28
Real constants, 3-2
Real-time devices, 19-1
Real-time error messages, 19-7
Real-time software, 19-1
RECL spec i f ier
in INQUIRE, 11-48
in OPEN, 11-28
Record specifier, 10-13

RECORDSIZE specifier
in INQUIRE, 11-48
in OPEN, 11-28
RECORDTYPE specifier
in INQUIRE, 11-48
in OPEN, 11-29
Register allocation
Global, 15-9
Relational expressions, 5-11
Relational operators, 5-11
Remark field, 2-4
Remarks, 2-6
REMOVL in LINK, 15-20
Reordering computations, 15-4
Repeat specification, 7-22
Repeatable edit descriptors, 12-7
REREAD
sequential FORMAT-statement,
10-38
sequential List-directed, 10-39
REREAD statement, 10-37
RESET in FOROTS, 15-11
RETURN statement, 13-47
REWIND statement, 11-54
RNT function, 18-34
ROT function, 18-34
Routines
calling BLISS-36, 15-19
channel allocation, 18-30
channel deallocation, 18-30
memory allocation, 18-29
RPG function, 18-35
RPSI function, 18-35
RRS function, 18-34
RTINIT subroutine, 19-3
RTREAD subroutine, 19-5
RTSLP subroutine, 19-6
RTSTRT subroutine, 19-4
RTWAKE subroutine, 19-6
RTWRIT subroutine, 19-5
Running the FORTRAN-10 compiler,
16-2
Running the FORTRAN-20 compiler,
16-6
RUNOVL in LINK, 15-20
S ed i ting, 12-24
SAOVL in LINK, 15-20
SAVE statement, 7-24
SAVFMT subroutine, 13-37
SAVRAN subroutine, 13-37
SCALE subroutine, F-I0
SECNDS function, 13-19
Sequential access, 10-5, 18-3
Sequential FORMAT-statement
PUNCH, 10-48
Sequential FORMAT-statement
ACCEPT, 10-40
Sequential FORMAT-statement PRINT,
10-45
Sequential FORMAT-statement READ,
10-26

Index--9

Sequential FORMAT-statement
REREAD, 10-38
Sequential FORMAT-statement TYPE,
10-43
Sequential FORMAT-statement WRITE,
10-32
Sequential List-directed ACCEPT,
10-41
Sequential List-directed PRINT,
10-46
Sequential List-directed PUNCH,
10-48
Sequential List-directed READ,
10-27
Sequential List-directed REREAD,
10-39
Sequential List-directed TYPE,
10-44
Sequential List-directed WRITE,
10-34
Sequential NAMELIST-statement
READ, 10-28
Sequential NAMELIST-statement
WRITE, 10-35
SEQUENTIAL specifier
in INQUIRE, 11-49
Sequential unformatted READ,
10-29
Sequential unformatted WRITE,
10-36
SETABL subroutine, F-ll
SETRAN subroutine, 13-38
Sharable high segment, 16-35
SKIPFILE statement, 11-57
SKIPRECORD statement, 11-56
Slash editing, 12-23
Software
Re a 1- tim e, 1 9-1
SORT subroutine, 13-38
Source files
line-sequenced, 2-7
Source program, 1-1
SP editing, 12-24
Specific functions, 13-2
Specification statements, 7-1
Specifiers
Carriage-control, 12-9
CLOSE statement, 11-34
END=, 10-17
ERR=, 10-18
format, 10-13
INQUIRE statement, 11-43
IOSTAT, 10-19
OPEN statement, 11-4
record, 10-13
SRTINI subroutine, 13-39
SS editing, 12-24
START LSCW, 18-4
Starting FORDDT, 17-5
Statement block, 9-8
Statement definition, 2-2
Statement field, 2-4
Statement format, 2-2

Statement functions, 13-15
Using, 13-16
Statement label constants, 3-7
Statement label field, 2-3
Statement numbers, 2-3
Statements
ACCEPT, 10-40
arithmetic assignment, 8-1
arithmetic IF, 9-3
ASSIGN, 8-3
assigned GO TO, 9-3
ass i g nm e nt, 8 -1
BACKFILE, 11-57
BACKSPACE, 11-55
BLOCK DATA, 14-1
block IF, 9-5
CALL, 13-44
character assignment, 8-4
character type specification,
7-6
CLOSE, 11-34
COMMON, 7-8
compilation control, 6-4
computed GO TO, 9-2
CONTINUE, 9-17
control, 9-1
DATA, 7-21
data transfer, 10-1
DECODE, la-51
Device control, 10-1, 11-52
DIMENSION, 7-1
DO WHILE, 9-13
ELSE, 9-5
ELSE IF THEN, 9-5
ENCODE, la-51
END, 9-20
END DO, 9-17
END IF, 9-5
ENDFILE, 11-55
ENTRY, 13-45
EQUIVALENCE, 7-11
executable, 1-1, 6-1
EXTERNAL, 7-17
.
File control, 10-1, Ll-l
FIND, 11-53
FORMAT, 12-2
FUNCTION, 13-20
GO TO, 9-1
IF, 9-3
IF THEN, 9-5
IMPLICIT, 7-7
INCLUDE, 6-4
indexed DO, 9-10
Input/output, 10-1
INQUIRE, 11-42
internal READ, 10-50
internal WRITE, la-50
INTRINSIC, 7-18
logical assignment, 8-3
logical IF, 9-4
logical two-branch IF, 9-4
NAM ELI S T, 1 2 - 4 4
nested DO, 9-14

Index-l0

Statements (Cont.)
nonexecutable, 1-1, 6-2
numeric type specification, '7-5
OPEN, 11-1
order ing of, 6-2
PARAMETER, 7-20
PAUSE, 9-19
PRINT, 10-45
PROGRAM, 6-4
PUNCH, 10-47
READ, 10-24
REREAD, 10-37
RETURN, 13-47
REW I ND, 11- 54
SAVE, 7-24
SKIPFILE, 11-57
SKIPRECORD, 11-56
specification, 7-1
STOP, 9-18
SUBROUTINE, 13-43
Summary of, A-I
TYPE, 10-42
type specification, 7-4
unconditional GO TO, 9-2
UNLOAD, 11-54
WRITE, 10-30
STATI subroutine, 19-6
STATO subroutine, 19-5
STATUS specifier
in CLOSE, 11-40
in OPEN, 11-30
STOP statement, 9-18
Storage
common, 7-9
Subexpressions
parenthetical, 5-13
Subprogram
function, 13-20
subroutine, 13-22
Subprograms
Block data, 14-1
Subroutine calls
writing, 15-11
SUBROUTINE statement, 13-43
Subroutine subprogram, 13-22
Subroutines, 13-1, 13-22
ALCCHR, 13-24
AXIS, F-2
BLKRW, 19-4
calling COBOL, 15-18
CAXIS, F-3
CDC 0 S , 1 3 - 2 5
CDEXP, 13-26
CDLOG, 13-26
CDS IN, 13-27
CDSQRT, 13-28
CHKDIV, 13-28
CLRFMT, 13-29
CONNECT, 19-4
DATE, 13-29
DISCON, 19-7
DISMIS, 19-6
DIVERT, 13-30

Subroutines (Cont.)
DTOGA, 13-30
DUMP, 13-30
ERRSET, 13-31
ERRSNS, 13-32
EXIT, 13-33
FFUNIT, 13-33
FORRTF, 19-1
FORTRAN-supplied, 13-24
FORTRAN-supplied plotter, F-l
GTODA, 13-34
ILL, 13-34
LEGAL, 13-34
LINE, F-4
LOCK, 19-3
MKTBL, F-5
MVBITS, 13-35
NUMBER, F-9
OVERFL, 13-35
PDUMP, 13-36
PLOT, F-10
PLOTS, F-2
programming considerations,
15-3
QUITEX, 13-36
RTINIT, 19-3
RTREAD, 19-5
RTSLP, 19-6
RTSTRT, 19-4
RTWAKE, 19-6
RTWRIT, 19-5
SAVFMT, 13-37
SAVRAN, 13-37
SCALE, F-10
SETABL, F-l1
SETRAN, 13-38
SORT, 13-38
SRTINI, 13-39
STATI, 19-6
STATO, 19-5
SYMBOL, F-12
TIME, 13-39
TOPMEM, 13-40
TRACE, 13-41
UNLOCK, 19-7
user-defined, 13-42
WHERE, F-12
Subscript
array element, 4-3
Substrings
character, 4-6
Summary of CLOSE specifiers,
11-35
Summary of OPEN statement
specifiers, 11-6
Summary of statements, A-I
Switches
ABORT, 16-9
BINARY, 16-9
CROSSREF, 16-3, 16-9
DEBUG, 16-3, 16-9, 16-11
DFLOATING, 16-3, 16-9
ECHO-OPTION, 16-9

Index--ll

Switches (Cont.)
EXPAND, 16-3, 16-9
EXTEND, 16-9
F66, 16-3, 16-9
F77, 16-3, 16-9
FLAG, 16-3, 16-21
FLAG-NON-STANDARD, 16-9
FORTRAN-20 compiler, 16-9
GFLOATING, 16-3, 16-9
INCLUDE, 16-3, 16-9
LISTING, 16-9
LNMAP, 16-3, 16-10
MACHINE-CODE, 16-10
MACROCODE, 16-3
NOBINARY, 16-10
NOERRORS, 16-3, 16-10
NOEXTEND, 16-10
NOF77, 16-3, 16-10
NOFLAG, 16-3
NOFLAG-NON-STANDARD, 16-10
NOWARN, 16 - 3, 16 -1 0, 16 -1 4
OPTIMIZE, 16-3, 16-10, 17-17
OPTION, 16-10
SYNTAX, 16-3, 16-10
SYMBOL subroutine, F-12
Symbolic names, 4-1
SYNTAX switch, 16-3, 16-10

UNFORMATTED specifier
in INQUIRE, 11-49
Unformatted WRLTE
direct, 10-36
sequential, 10-36
Unformatted WRITE transfers,
10-35
Unit identifier
logical, 10-11
UNIT specifier
in CLOSE, 11-41
UNLOAD statement, 11-54
UNLOCK subroutine, 19-7
UPG function, 18-36
User-defined character set, F-9
User-defined external functions,
13-20
User-defined subroutines, 13-42
Using FOROTS, 18-14
Using FORRTF, 19-2
Using function subprograms, 13-21
Using intrinsic functions, 13-2
Using statement functions, 13-16
Using the compiler, 16-1
Using the FORTRAN-20 compiler,
16-5

T editing, 12-19
TAPE FORMAT specifier
in OPEN, 11-32
TIM2GO function, 13-19
TIME subroutine, 13-39
TL editing, 12-20
TOPMEM subroutine, 13-40
TR editing, 12-20
TRACE subroutine, 13-41
TRACE with DEBUG, 16-12
TYPE
sequential FORMAT-statement,
10-43
sequential List-directed, 10-44
Type declaratars, 7-5
Type specification statements,
7-4
character, 7-6
n urn e ric, 7 - 5
TYPE specifier
in OPEN, 11-30
TYPE statement, 10-42
unconditional GO TO statements,
9-2
Unformatted data transfers, 10-8
to ASCII devices, 10-8
Unformatted READ
direct, 10-30
sequential, 10-29
Unformatted READ transfers, 10-29

Variable names, 4-2
Variables, 4-2
compiler generated, 16-24
VERSION specifier
in OPEN, 11-34
VMS with FLAG, 16-22
Warning messages, 16-34, C-8
WHERE subroutine, F-12
WRITE
direct FORMAT-statement, 10-33
direct unformatted, 10-36
sequential FORMAT-statement,
10-32
sequential List-directed, 10-34
sequential NAMELIST-statement,
10-35
sequential unformatted, 10-36
WRITE statement, 10-30
internal, 10-50
WRITE transfers
formatted, 10-31
unformatted, 10-35
Writing arithmetic expressions,
5-2
Writing programs, 15-1
Writing subroutine calls, 15-11
WRS function, 18-34
X editing, 12-21
Z (hexadecimal)

Index-12

editing, 12-37

TOPS-10/TOPS--20
FORTRAN Language Manual
AA-N383B-TK

READER'S COMMENTS
NOTE:

This form is for documen1 comments only. DIGITAL will use comments submitted on
this form at the company's discretion. If you require a written reply and are eligible to
receive one under Software Performance Report (SPR) service, submit your comments on an SPR form.

Did you find this manual understandablE, usable, and well-organized? Please make suggestions for improvement.

Did you find errors in this manual? If so, specify the error and the page number.

Please indicate the type of reader that you most nearly represent.

[J Assembly language programmer
[J Higher-level lan~]uage programmer
[J Occasional programmer (experinnced)
[J User with little programming experience

o Student programmer

[J Other (please specify)~~~_~~~~~~~~~~~~.~~~~~~~
Name

Date ~____________

Organization ~_ _~ _ _ _ _ _~ _______ Telephone ___~_ _ _ _ __
Street ___________
City _ _ _ _ _ _ _ ,~_ __

_ ____~ State ____ Zip Code _ _ __
or Country

--~.--0 ~.ot -F~HDere
00

oTear

and Tape - - - - - - - - - - - - - - - - - - - - - -

rfl-ll1--------~~~~;:;~~-

if Mailed in the
United States

BUSINESS REPLY MAIL

••

FIRST CLASS PERMIT NO. 33 MAYNARD MASS.
POSTAGE WILL BE PAID BY ADDRESSEE

SOFTWARE PUBLICATIONS

- - - -

200 FOREST STREET

MR01-2/L 12

MARLBOROUGH, MA

01752

•

Do Not Tear - Fold Here and Tape - - - - - - - - - - - - - - - - - - - - - - - - - - - -- - - - - - - - - - - - - -- -