Digital PDFs

AA-L315A-TM

May 1983

256 pages

Original

8.5MB

Document:	PASCAL Language Manual Ver 1 Sep83
Order Number:	AA-L315A-TM
Revision:	0
Pages:	256
Original Filename:	AA-L315A-TM_PASCAL_Language_Manual_Ver_1_Sep83.pdf

OCR Text

TOPS-20
PASCAL Language Manual
AA-L~31 5A-TM

September 1983
This document descnibes the elements of the PASCAL
language supported by TOPS-20 PASCAL.

OPERATING SYSTEM:
SOFTWARE:

TOPS-20 VS.1 (2040,2060)
TOPS-20 V4.1 (2020)

PASCAL V1.0
LINK VS.1
RMS V1.2

Software and manuals should be ordered by title and order number. In the United States. send orders
to the nearest distribution center. Outside the United States, orders should be directed to the nearest
DIGITAL Field Sales Office or representative.

Northeast/MId-Atlantic Region

Central Region

Western Region

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

Digital Equipment Corporation
Digital Equipment Corporation
Acc€tssories and Supplies Center Accessories and Supplies Center
1050 East Remington Road
632 Caribbean Drive
Schaumburg, Illinois 60195
Sunnyvale, California 94086
Telephone:(312)640--5612
Telephone:(408)734-4915

digital equipment corporation. marlboro massachusetts

First Printing, September 1983
~: Digital Equipment Corporation 1983. 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 accorcfance 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:

~D~DD~D'M
DEC
DECmate
DECsystem-10
DECSYSTEM-20
DECUS
DECwriter
DIBOL

MASSBUS
PDP
piOS
Professional
Rainbow
RSTS
RSX

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

PREFACE
CHAPTER 1

INTRODUCTION

OVERVIEW OF PASCAL
·
Data Types . . • . • • • • . •
• • • • • .
Structure of a PASCAL Program
. • . .
1•1•3
Definitions and Declarations . . . • . · . . .
1 • 1 .4
Executable Statements
1.1.5
Subprograms
•...••••
•
1.1.6
Compilation Units
....
·
1.2
LEXICAL ELEMENTS • . • . • • .
·
1.2.1
Character Set • . . . . • . . • . • • . • • •
1.2.2
Reserved Words •
• ••.
1.2.3
Identifiers
.•••.•••
1.2.3.1
Predeclared Identifiers
·
1.2.3.2
User Identifiers • • • • • •
• •
1.2.4
Special Symbols
• •
1 .3
DELIMITERS . • . • • • • • • • • • .
1 .4
DOCUMENTING YOUR PROGRAM •
·
1.5
THE %INCLUDE DIRECTIVE • . • . . • • • . · • . .
1•1

. 1-2

1.1.1
1 • 1 .2

• 1- 2

CHAPTER 2
2.1
2.2
2.2.1
2.2.1.1
2.2.1.2
2.2.1.3
2.2.1.4
2.2.2
2.2.2.1
2.2.2.2
2.2.2.3
2.3
2.3.1
2.3,.1.1
2.3.1.2
2.3.1.3
2.3.1.4
2.3.1.5
2.3.2
2.3.2.1
2.3.2.2
2.3.2.3
2.3.2.4
2.3.3
2.3.4
2.3.4.1
2.3.4.2
2.4
2.5
2.6

• 1-2
. 1-4
• 1-4
• 1-4

. 1-5
. 1-5
• 1- 5

· 1-0
· 1-7
. 1-7
• 1-8
• l-8

• 1-8
• 1-9
1-10

PASCAL DATA TYPES
DECLARING DATA TYPES
· • • • • •
SCALAR TYPES • • • •
Predefined Data Types
. • • • .
INTEGER Data Type
• • • • • •
REAL Data Type • •
• • • • • •
BOOLEAN Data Type
• • ••
CHAR Data Type • • •
• • • •
User-Defined Scalar Data Types •
Enumerated Data ~rypes
• •
Subrange Data Types
• • •
The ORD Function •
·
STRUCTURED DATA TYPES
Array Types
• • • •
Multidimensional Arrays
String Variables • • • •
• • • •
Initializing and Assigning Values to
Array Type Compatibility.
•
Array Examples • • • • • •
·
Record Types . • • • • • • •
Records with Variants
Assigning Values to Records
• • • .
Record Type Compatibil~ty
Record Examples
Set Types
• • • • • • • •
File Types • • • • • • • • •
Internal and External Files
Text Files • • • • • • • • • •
POINTER TYPES
. • • • • • • • • •
PACKED STRUCTURED TYPES
• • • • •
TYPE COMPATIBILITY • • • • • • • • • • •

iii

• • • .
·
. • • .

• 2-1
. 2-2
. 2- 2

•

• 2-2
•

•

2- 3

• • • • • 2- 4
• • • • • 2-4
• • 2-5
•

•

• 2- 5

• • • • • 2-6
• • • • • 2-7
· • 2-8
•

•

• 2- 8

• • • • • 2-9
• • ••
2-11
Arrays
2-12
•

2-14

• • ••
· • ••

•

2-15
2-15
2-17
2-19
2-19
2-21
2-22
2-24
2-25
2-25
2-26
2-28
2-29

• • ••
• • ••
• • • •

CHAPTER 3
3. 1
1. 1 . 1
3. 1 .2
1. 1 . 3

3.1.4
3 . 1 .5
3.2
CHAPTER 4
4•1
4 .2
4•3
4 .4
4.5
4.6

CHAPTER 5
5.1
5.2
5.3
5.3. 1
5.3.2
5.3. 3
5.4
5 .4. 1
5.4.2
5.4 . 3
5.5
5.6
5.7
CHAPTER 6
11.1
6. 1. 1
6.1.1.1
6.1.1.2
6.1.1.3
11.1.1.4
6.1.2
6.2
6.3
6.3.1
6.3.1.1
6.3.1.2
6.3.1.3
6.3.2
6.4
6.5
6.6
11.7
6.S
CHAPTER 7
7.1

EXPRESSIONS
OPERATORS
Arithmetic Expressions.
Relational Expressions .
Logical Expressions
... .
Set Expressions
... .
Precedence of Operators
SCOPE OF IDENTIFIERS . • . .

·
.

.
.
·

.
.
·
.

1-1
3-1
3-4
3-5
3-5
. 1-6
· 3-7

•

• 4- 3

.
·
.

.
.
.
·

.
.
.
.

4-4
4-5
4-6
4-7

.
·
· .
·
• • • •
. . .
· . . .

.
.
.
.
•
.
.
·

5-2
5-2
5-4
5-4
5- 5
5-6
5-8
5-S
5-10
5-11
5 -12
5-13
5-14

6-1
6-'-1
11- 5
6-6

PROGRAM HEADING AND DECLARATION SECTION
THE PROGRAM HEADING
LABEL DECLARATIONS . .
CONSTANT DEFINITIONS .
TYPE DEFINITIONS . . .
VARIABLE DECLARATIONS
VALUE DECLARATIONS . .

· 4-2

PASCAL STATEMENTS
THE COMPOUND STATEMENT .
THE ASSIGNMENT STATE~ENT •
CONDITIONAL STATEMENTS . . .
The IF-THEN Statement
The IF-THEN-ELSE Statement
The CASE Statement .
.
REPETITIVE STATEMENTS
. . .
The FOR Statement
. . ..
The REPEAT Statement .
The WHILE Statement
THE WITH STATEMENT .
THE GOTO STATEMENT .
THE PROCEDURE CALL .

·
.

.
·
.....
. . . . . .
. ...
. . . . .

PROCEDURES AND FUNCTIONS
PREDECLARED SUBPROGRAMS
. • . . .. . •
Predeclared Procedures
· •
.. • •
Input/Output Procedures
Dynamic Allocation Procedures
The MARK and RELEASE Procedures
•
Miscellaneous Predeclared Procedures.
Predeclared Functions
FORMAT OF A SUBPROGRAM
PARAMETERS . . ' . . .
Formal Parameters
Value Parameters.
.. . •
Variable Parameters
• •
Formal Procedure and Function Parameters
Conformant Arrays
. . . . . .. . •
DECLARING A PROCEDURE
· •
DECLARING A FUNCTION .
FORWARD DECLARATIONS . • . . • .
EXTERNAL SUBPROGRAMS .
MODULES FOR SEPARATE COMPILATION •

•
•
•
·

.
.
•
•

•
.
•
•

•

• 6-9

.
.

. .
• .

6-10
6-12
6-20
6-21
11-21
6-22
6-23
6-24
6 - 25
6-28
6-32
6-35
6-36
6-37

INPUT AND OUTPUT
FILE CHARACTERISTICS • . . . . . . . . . . • . • . 7-2
iv

7•1. 1
7. 1. 2
7•1. 3
7.1.4
7.2
7.2.1
7.2.2
7.3
7.4
7.5
7.6
7.7
7 .7.1
7.7.2
7.7.3
7 • 7 .4

7.7.5
7 •7 •6

7.8
7.9
7.10
7.11
7. 12
7.13
7.14
7.15
7.10
CHAPTER 8
8.1
8.2
8.3
8.4
8.4.1
8.4.2
8.4.3
8.4.4
8.4.5
8.4.6
8.4.6.1
8.4./s.2
8.4.6.3
8.5
8.5.1
8.ry

8.7
CHAPTER 9
9.1
9.2
9.3
9.4
9.4.1
9.4.2
9.4.3
9.4.4
9.4.5
9.4.6
9.4.7

Fi Ie Names . . . . . . . . • • . . . .'
·
Logical Names
File Organizntion
· .
Record Access
·
RECORD FORMATS . . .
·
Fixed-Length Records .
• . • .
Variable-Length Records . . . . .
· •
THE CLOSE PROCEDURE
·
THE FIND PROCEDURE . . •
. . • .
THE GET PROCEDURE • . . •
THE LINE LIMIT PROCEDURE
THE OPEN PROC EDURE • • .
. . . . .
File History -- NEW, OLD, RE~DONLY, or UNKNOWN
Record Length
. . . • • • • . .
Record Access Mode -- SEQUENTIAL or DIRECT
Record Type -- FIXED or VARIABLE . . . .
Carriage Control -- LIST, CARRIAGE, or
NOCARRIAGE . . . . • • • . • • • .
Examples . . . • .
THE PAGE PROCEDURE •
THE PUT PROCEDURE
THE READ PROCEDURE
THE READLN PROCEDURE .
THE RESET PROCEDURE
THE REWRITE PROCEDURE • • . • .
THE WRITE PROCEDURE
THE WRITELN PROCEDURE • • . .
TERMINAL I/O • • . • • •

.
·
.
.
.
.
.
.
.
·
·
·

7-2
7-2
7-3
7-3
7-3
7-4
7-4
7-4
7-4
7-0
7-7
7-8
7-10
7-10
7-10
7-11
7-11
7-12
7-14
7-15
7-10
7-18
7-19
7-20
7-21
7-25
7-28

USING PASCAL ON TOPS-20
PROGRAM DEVELOPMENT PROCESS
. 8-1
FILE SPECIFICATIONS AND DEFAULTS
• 8-2
CREATING A PROGRAM • .
. • . 8-3
COMPILING A PROGRAM
• • • .
• . . . . 8-4
The PASCAL Command • • • • .
• • • • • • . • 8-4
PASCAL Compiler Commands • • • • • . . • • • . . 8-5
PASCAL Compiler Switches. • .
. • • • . . • 8-6
Specifying Switches in the Source Code.
• 8-9
Specifying Output Files • • • .
8-11
Compiler Listing Format • • • • .
8-11
Source-Code Listing
8-14
Machine-Code Listing •
•••.
8-15
Cross-Reference Listing
8-10
LOADING A PROGRAM
8-16
The LOAD Command • •
• • • • . • • •
8-16
EXECUTING A PROGRAM
. • • •
8-17
EXAMPLES. • • • • •
8-17
PASDDT: THE PASCAL-20 DEBUGGER
RUNNING PASDDT
USING SYMBOLIC VALUES
SCOPE
• • • •
PASDDT COMMANDS
ASSIGN •
BREAK • • • •
CLEAR • • • •
DISPLAY • • • • • • • •
EXIT • • • • • • • • • • • •
HELP • • •
••••• •
PROCEED
• • • • • • • •
v

• • • 9-1
9-1
• 9-2
9-3
• • • 9-3
• • • 9-4
• 9-6
• • 9-7
• • 9-9
9-10
9-10

9.4.8
9.4.9
9.4.10
9.4.11

9-11
9-11
9-14
9-15

REMOVE •
SET
SHOW • •
TRACE

APPENDIX A

PASCAL .MESSAGES

APPENDIX B

ASCII CHARACTER SET

APPENDIX C

SYNTAX SUMMARY

C.l

BACKUS-NAUR FORM .

• .

• • • .

• • • • • • • • • C-l

APPENDIX D

SUMMARY OF PASCAL-20 EXTENSIONS TO PROPOSED ISO
STANDARD

APPENDIX E

ISO COMPLIANCE

E.l
E.2
E.3
E.4

IMPLEMENTATION-DEFINED FEATURES
IMPLEMENTATION-DEPENDENT FEATURES
ERROR HANDLING . . • . • . •
EXCEPTIONS AND RESTRICTIONS

• • • • E-l
• • E-3
• • • •
• E- 3
•

•

APPENDIX F

DIFFERENCES BETWEEN PASCAL-20 AND VAX-II PASCAL

APPENDIX G

PROCEDURE AND FUNCTION CALLING SEQUENCES

G.l
G.2
G.3
G.l.1
G.3.2
G.3.3
G.3.4
G.l.5
G.4
G.t)

RUN-TIME STACK . . • • • . • . .
• •
MECHANICS OF A PROCEDURE CALL
PARAMETER PASSING
Value Parameter Passed By Value
••..
Value Parameter Passed By Address
Reference (VAR) Parameter
.•.•
•
Procedure Or Function Parameter
•
Conformant Array Parameter.
• •
PARAMETER ACCESSING EXAMPLE
CONFORMANT ARRAY EXAMPLE • .
• • • • •

•

• E-8

• • . • G-l
• • G-2
•

•

• G- 3

G-3
· G-3
•
•
•

•
•
•

•

• G-4
• G-4
• G-4

• G-5

INDEX

• G- 5

Index-l

FIGURES
FIGURE

1-1
1-2
2-1
2-2
2-3
3-1
4-1
7-1
7-2
8-1
9-1

Structure of a PASCAL Program
%TNCLUDE File Levels • . . . . .
Two-DimensionaJ Array Two D
Three-Dimensional Array ChesslD
Storing Components in an Array
Scope of Identifiers • . . . . • .
PASCAL Data Types
..••
File Position After GET
...•
File position after RESET
Compiler Listing Format
..•.•.
Scope
. . . . . • . . . .
. .
vi

•

• 1- 3

1-12
. 2-9
2-10
2-11
• 1- 8

· 4-5
· 7-6
7-19
8-12
.

•

9-2

G-l
G-2

G-3
G-4

Status of Stack After PUSHJ
Stack Frame
•.•.....
External Proce~ure Declaration .
Conformant Array Parameter . . .

· G-1
•

• G-2

• G-5
· G-6

•

TABLES
TABLE

1-1
1-2
1-3
3-1
3-2
3-3
3-4
3-5
3-6
6-1
6-2
f)-3

7-1
7-2
7-3
7-4
8-1
8-2
8-3
8-4
A-I
A-2
B-1

C-l
F-I
F-2

Feserved Words • . . . . •
Predeclared Identifiers
Special Symbols
•. . .
• •
Arithmetic Operators . . .
•
Result Types for Arithmetic Expressions
Relational Operators .
Logical Operators
t:e tOpe ra to rs
Precedence of Operators
Predeclared Procedures.
Predeclared Functions
Library Routines •
Default Values for TOPS-20 External File
Specifications • • • • . •
Summary of File Attributes
Default Values for Field Width .
Carriage-Control Characters
·
File Specification Defaults
PASCAL Compiler Commands • . . • . •
PASCAL Compiler Switches.
·
Source Switches
.••.• • • • •
Run-time Errors
.•..••••
I/O Errors • . • . • • .
The ASCII Character Set
• • • • •
BNF MetaSymbols
••• .
Additional Language Elements •
•
Additional Predeclared Functions •
•

vi i

I-f)

• • 1-7
•

•

1- 8

• • • .

• 1-2
· 1-3
· 1-4

•
•

• 3-5
• 3- 5
· 3-6

•

• fi-2
f)-13

fi-17
• 7-9
·
·
•

•
.
•

•
•
•

•
.
•

7 -12
7-22
7 -26
· 8-3

•
•

·
•
•

• 8-5
• 8-7
8 -10

· A-I
A-2
•

•

• .

• B-1

. C-l

•

• F- 4

F- 2

PREFACE

MANUAL OBJECTIVES
This manual describes the PASCAL language and
the PASCAL ctebugger,
PASDDT, as they are implemented on the TOPS-20 operating system.
This
document is designed primarily for reference;
it is not
a
tutorial
document.
INTENDED AUDIENCE
This manual is intended for readers who know the PASCAL language.
The
reader need not have a detailed understanding of the TOPS-20 operating
system, but some familiarity with either is helpful.
For
information
about
the
TOPS-20
operating
system,
refer to the documents listed
below under "Associated Documents." For introductory information about
the
use of
the PASCAL language, refer to the TOPS-20 PASCAL Primer,
Order Number AA-L3l4A-TM.
STRUCTURE OF THIS DOCUMENT
This manual contains the following chapters and appendixes:
•

Chapter 1 provides an introduction to the use of
describes the format of a PASCAL program.

PASCAL

•

Chapter 2
introduces basic
concepts
including
variables, data types, expressions, and scope.

constants,

•

Chapter 3 describes the components of an expression.

•

Chapter 4
section.

•

Chapter 5 describes the statements that perform
of the program.

•

Chapter ~ explains the use of functions and
procedures,
and
summarizes
the predeclared functions and procedures supplied
with the PASCAL-20 language.

•

Chapter 7 provides detailed information on input
procedures.

•

Chapter 8 describes the compiling, loading, and executing
PASCAL programs on the TOPS-20 operating system.

•

Chapter 9 describes the PASCAL-20 debugger, PASDDT.

describes

the

program

heading

an~

and

declaration
the

and

actions

output
of

•

Appendix A lists the various messages you can receive.

•

Appendix B lists the ASCII character set.

•

Appendix C presents the PASCAL-20 language in the Backus-Naur
form and includes syntax diagrams.

•

Appendix D summarizes
PASCAL-20 language.

•

Appendix E describes how PASCAL-20 complies with the standard
proposed
by the
International Standardization Organization
(ISO) .

•

Appendix F summarizes the differences between TOPS-20
and VAX-li PASCAL.

•

Appendix G describes the calJing
sequences and
conventions
used by PASCAL for user-defined procedures and functions.

the

extensions

ASSOCIATED DOCUMENTS
Associated manuals are:

•
•
•

TOPS-20 PASCAL Primer
--TOPS-20 User's
Guide
---- --TOPS-20 Commands Reference Manual

• LINK Reference -Manual
--•
•

EDIT Reference Manual
--TV Editor Manual

incorporated

the

PASCAL

CONVENTIONS USED IN THIS DOCUMENT
This document uses the foJlowing conventions:
Convention

Meaning
A
horizontal
ellipsis
means
thnt
the
preceding
item can be repeated one or more
times.

A vertical ellipsis means that not all of the
statements in a figure or example are shown.

[ ]

Double brackets in statement and
declaration
format
descriptions enclose optional items,
for examp] e:
WRITE

r ]

([OUTPUT,]

print-list)

Square brackets show
where
the
syntax
requires
square brackets.
This notation is
used with arrays, for example:
ARRAyrl •• 5] OF INTEGER

{ }

Braces enclose
lists
from which
choose one item, for example:

you

must

This symbol indicates
RETURN key.

press

the

where

you

The notation <CTRL/x> indicates that you must
press
the
key
labeled
CTRL
while
simultaneously pressing another key (x),
for
example, <CTRL/Z>.

UPPERCASE LETTERS

Uppercase letters in a command Jine
indicate
information that you must enter as shown.

lowercase letters

Lowercase letters in a command line
variable information you supply.

Simple_Procedure

In programming
examples,
all
identifiers,
(names
created
by the
programmer),
are
printed in
lowercase
letters with
initial
capitals.

Contrasting Colors

Orange - where examples contain both
user
input and computer output, the characters you
type are in orange;
the characters printed
on the terminal are in black.

indicate

CHAPTER I
INTRODUCTION

PASCAL-20 is an extended implementation of level 1 of the standard
proposed
for the PASCAL language by the International Standardization
Organization (ISO).
This manual describes the use of PASCAL under the
TOPS-20
operating system.
PASCAL-20
includes all
the language
elements as defined in the PASCAL User Manual and Report by Jensen and
Wirth, as well as the following extensions:
•

Exponentiation operator

•

Hexadecimal, octal, and binary integers

•

Double-precision real data type

•

Dollar sign ($)

•

External procedure and function declarations

•

CARD, CLOCK, EXPO, SNGL, and UNDEFINED functions

•

REM operator

•

OTHERWISE clause in the CASE statement

•

OPEN and CLOSE procedures for file access

•

FIND procedure for direct access to sequential files

•

Optional carriage control for output files

•

DATE, TIME, HALT, and LINELIMIT procedures

•

Variable initialization

•

Separate compilation

•

%INCLUDE directive
compilation

•

Support Eor calling externally declared FORTRAN subroutines,
and for declaration of PASCAL subroutines that can be called
by FORTRAN

and underline ( ) characters in identifiers

alternate

for

input

files

during

Refer to Appendix E for more information on ISO compliance.
This chapter presents an overview of the important concepts in PASCAL
and
illustrates the structure of a PASCAL program.
It also describes
PASCAL's lexical elements
the character set,
reserved words,
identifiers, and special symbols. The final sections explain how to
document your program and how to include existing files.
1-1

INTRODUCTION
1.1

OVERVIEW OF PASCAL

A PASCAL program performs operations on data items known as constants,
variables, and function designators.
A constant is a quantity with an
unchanging value.
A constant to which you give a
name
is called
a
symbolic
constant.
A variable is a quantity whose value can change
while
the
program executes.
A function designator
causes
the
execution of a
group of statements
that
is associated with an
identifier and returns a value.
The function type
is determined
by
the type of the value it returns.

1.1.1

Data Types

Every PASCAL data item is associated with a data type.
A data
type,
which
is
usually indicated by a type identifier, determines both the
range of values a data item can assume and the operations that can be
performed
upon
it.
In
addition, the type implicitly indicates how
much storage space is required for all possible values of
the data
item.
PASCAL provides
identifiers
for many predefined
types.
Thus,
a
program's operations can involve integers, real numbers, Boolean and
character data,
arrays,
records,
sets,
and
pointers
to dynamic
variables.
PASCAL also
allows
you
to
create your
own types by
defining an identifier of your choice to represent a range of values.
The type of a constant is the type of its corresponding
value.
You
establish variable and
function
types when
you declare them.
In
general, they cannot change.
Although variables and
functions
can
change
in value any number of times, all the values they assume must
be within the range established by their type.
A variable does not
assume a
value
until
the
program assigns
it one.
A function is
assigned a value during its execution.
PASCAL associates types not only with data
items,
but also with
expressions.
An
expression
is the computation of a value resulting
from a combination of variables, constants, function designators,
and
operators.
In PASCAL,
you can
form expressions using arithmetic,
relational,
logical,
string,
and
set
operators.
Arithmetic
expressions produce
integer or
real
number values.
Relational,
logical, string, and most set expressions yield
Boolean
results.
Other set expressions form the union, intersection, and differences of
two sets.

1.1.2

Structure of a PASCAL Program

A PASCAL program consists of a
heading
and
a
block.
The heading
specifies the name of the program and the names of any external files
the program may use.
The block
is divided
into
two
parts:
the
declaration section,
which contains data declarations;
and
the
executable section, which contains executable statements.
Figure 1-1
points out each part of a sample PASCAL program.

1-2

INTRODUCTION
PROGRAM Calculator (INPUT, OUTPUT);
TYPE Yes_No = (Yes, No);
VAR Subtota 1, OpeT'arld ! REAL.;
EGuatior. ! BOOL.EAN;
Ope T'ato T' ! CHAR f
AnsweT' ! Ye!:; .... No ji

Declaration
Section

PROCEDURE InstT'ucticms;
BEe, IN
WFn TEL N (' This pro g r a ITI add s, sub t T' act s, lTIul tip 1 i e s, and');
WRITEL.N ('divides real nUlTlbers. Enter a nUlTlber in response');
WRITELN ('to the Operand! prolTlPt and enter an operator -- ');
Procedure WRITELN ( ' t f -, *, /, or = -- in response to the Operator!');
Block
WRITELN ('proITlPt. The progralTl keeps a running subtotal');
WRITELN ('until ~ou enter an e~ual sign (=) in response to');
WRITELN ('the Operator! prolTlPt.
You can then exit frolTl');
WF~ I TEl... N (' t h €-! F' T' 0 ~.;! T' a ITI 0 I' b f::' !:.I ina new s f~ t 0 f cal c 1•.11 at i C) n!:; • ' ) ;
ENI; (*end of procedure Instructions*)
BEGIN
WRITE ('Do YOU need instructions? r~pe yes or no.
RFADl..N (An~; we Y') ;
If Answer = Yes THEN Instructions;

');

F~FF'EAT
EC~l..Iat:i.on !"" FAL~:;F;
Subtota 1 : :: .. 0;
WR I TE (' ClF'e rand! ' ) ;
READLN (Subtotal);

WHILE (NOT [~uation) DO
BEGIN
WRITE ('Operator!');
READLN (Operator);
IF (Operator = '=') THEN
BEGIN
E~uation != TRUE;
WRITELN ('The answer is ',Subtotal!5!2)
END
ELSE
BEGIN
WRITE ('Operand!');
READLN (Operand);
CASE Operator OF
Subtotal _. Subtotal + Operand;
'+
'
'- ,
Subtotal ' .- Subtotal - Ope T'and;
,- Subtotal * Operand;
Subtotal ,'*'
IIi
Subtotal ,-- Subtotal / Operand
END;
WRITELN ('The subtotal is ',Subtotal!5!2)
END
END;

Executable
Section

.....

WRITE ('An~ more calculations?
READLN (Answer);
UNTIL Answer = No;

END.

Figure 1-1:

T~pe

~es

or no.');
MR·S-3150-83

Structure of a PASCAL Program

1-3

INTRODUCTION
Procedure and
function declarations have
the
same
structure as
programs.
Note, in Figure 1-1, the heading and block of the procedure
instructions.
This manual
uses
the
term
subprogram
to denote a
procedure or function.

1.1.3

Definitions and Declarations

PASCAL requires you to define every constant and user-created type and
to declare every label, variable, procedure, and function used in your
program.
The declaration section of
the
program contains
LABEL,
CaNST,
TYPE,
VAR,
VALUE, PROCEDURE, and FUNCTION sections, in which
you define and declare all the data your program uses.
All of
these
except
LABEL
introduce identifiers and indicate what they represent.
LABEL declares numeric labels that correspond to executable statements
accessed
by the GOTO statement.
PASCAL allows you to assign initial
values to variables you declare in a
VAR section.
An
initialized
variable assumes the given value when program execution begins.

1.1.4

Executable Statements

The executable section of a PASCAL program contains
the
statements
that
specify the
program's actions.
The executable section
is
delimited by the reserved words BEGIN and END.
Between BEGIN and
END
are conditional
and
repetitive statements,
statements that assign
values to variables and functions, and statements that control program
execution.

1.1.5

Subprograms

PASCAL provides several ways for you to group together definitions,
declarations,
and
executable statements.
One way is to group them
into procedures and functions, generically called
subprograms.
Both
kinds of subprograms are groups of statements associated with an
identifier.
Procedures are usually written to
perform a
series of
actions, while functions are written to compute a value.
Subprograms constitute a convenient way to
isolate
the
individual
tasks
that
the main
program
is to accomplish.
Subprograms do not
exist independently of the program;
they are called
either by an
executable statement known as a
procedure call
or by a function
designator appearing
within
an
expression.
PASCAL supplies many
predeclared
subprograms
that perform commonly used
operations,
including input and output.
A subprogram consists of a heading and a block.
The heading
provides
the
name of the subprogram, usually a list of formal parameters that
declare the input data for the program, and, in the case of functions,
the
type of the result.
The subprogram block consists of an optional
declaration section and an executable section.
When
the declaration
section
is present,
it declares data that is local to the routine
(that is, data that is unavailable outside the subprogram).
PASCAL is a block-structured language in that it allows
you
to
nest
subprogram blocks not only within the main program, but also within
other subprograms.
Each subprogram can make its own local definitions
and
declarations and
can even redeclare an identifier. that has been
declared in an outer block.
A subprogram declared at an
inner
level
has access to the declarations and definitions made in all blocks that
enclose it.

1-4

INTRODUCTION
1.1.6

Compilation Units

A program is sometimes calJed a
compilation unit
in
this manual
because
it can be compiled as a single unit (unlike a subprogram,
which cannot be compiled without the context of a program).
A program
consists of a
heading and a block, just as a subprogram does.
The
heading consists of the name of the program and
possibly a
list of
identifiers to indicate any external files that the program uses.
The
declaration section of the program block declares data
that
is
available at all program levels, including all nested subprograms.

1.2

LEXICAL ELEMENTS

A PASCAL program is composed
entirely of lexical
elements.
These
elements can be individual symbols, such as arithmetic operators;
or
they can be words that have special meaning to PASCAL.
The basic unit
of any lexical element is a character, which must be a member of the
ASCII character set, as described in section 1.2.1.
Some characters
are
special
symbols that PASCAL uses as statement delimiters,
operators, and elements of the language syntax.
Special
symbols are
listed in Section 1.2.4.
The words that PASCAL uses are combinations of alphabetic characters
and
occasionally a dollar sign (~;), an underscore ( ), or a percent
sign (%).
PASCAL reserves some words
for
the names of executable
statements,
operations,
and
some of the predefined data
types.
Reserved words are listed in Section 1.2.2.
Other words in a PASCAL
program are called
identifiers.
Predefined
identifiers represent
routines and data types provided by PASCAL.
Other identifiers can be
created by you to name programs, constants, variables, and any other
necessary program segment that is not already named.
Section 1.2.3
explains the use of both kinds o~ identifiers.

1.2.1

Character Set

PASCAL uses the
full American Standard
Code
for
Information
Interchange
(ASCII)
character set
(see Appendix B).
The ASCII
character set contains 128 characters in the following categories:
•

The uppercase and lowercase letters A through Z and a through
z

•

The numbers 0 through 9

•

Special characters, such as ampersand (&), question mark (?),
and equal sign (=)

•

Nonprinting characters,
such
carriage return, and bell

1-5

space,

tab,

line

feed,

INTRODUCTION
The PASCAL compjler does not distinguish
between
uppercase
and
lowercase characters, except in character and string const~nts and the
values of character and string variables.
For example,
the
reserved
word
PROGRAM has
the
same meaning when written
as any of the
following:

PROGRAM
PRogrAm
pro ~.=.1 r a III

The constants below, however,

represent different characters:

,b '
,BI

The following two constants represent different strings:

'BREAD AND ROSES'
'Bread and Roses'

1.2.2

Reserved Words

PASCAL reserves the words in Table 1-1 as names for
statements,
data
types,
and
operators.
This manual
shows these words in uppercase
characters.
Table 1-1:

Reserved Words

AND
ARRAY
BEGIN
CASE
CONST
DIV
DO
DOWNTO
ELSE

NIL
NOT
OF
OR
PACKED
PROCEDURE
PROGRAM
RECORD
REPEAT

END
FILE
FOR
FUNCTION
GOTO
IF
IN
LABEL
MOD

SET
THEN
TO
TYPE
UNTIL
VAR
WHILE
WITH

You can use reserved words in your program only in
the
contexts
in
which PASCAL defines
them.
You cannot redefine a reserved word for
use as an identifier.
In PASCAL, the following words are considered semi reserved words:
MODULE
OTHERWISE
REM
VALUE
Like the
words.
for your
used for
they are

reserved words, PASCAL also
predefines
these
semi reserved
However,
unlike reserved words, you can redefine these words
own purposes.
If you redefine them, they can no
longer be
their original purpose within the scope of the block in which
redefined.

I-f)

INTRODUCTION

1.2.3

Identifiers

PASCAL uses the term identifier to mean the name of a program, module,
constant,
type, variable, procedure, or function.
An identifier is a
sequence of characters that CRn include letters, digits, dollar signs
( $ ), and un de r 1 i n e s ym b 0 1 s ( ), wit h the foIl ow i ng res t ric t ion s :
•

An identifier can begin
dig it.

with

any

character

•

An identifier must be unique
in
its first
within the block in which it is declared.

•

An identifier must not contain any blanks.

other

than

characters

PASCAL places no restrictions on the length of identifiers, but scans
only the
first
31 characters for uniqueness;
the rest are ignored.
~he following are examples of valid and invalid identifiers:
Va 1 id

Invalid

Fo r2n8
Max Words
Upto
$CREMBX

4awhile (starts with a digit)
Up&to (contains the ampersand)

1.2.3.1 Predeclared Identifiers - PASCAL predeclares some identifiers
as names of
functions, procedures, types, values, and files.
These
predeclared
identifiers are listed
in Table 1-2 and appear
in
uppercase characters throughout this manual.
Table 1-2:
ABS
ARCTAN
BOOLEAN
CARD
CHAR
CHR
CLOCK
CLOSE
COS
DATE
DISPOSE
DOUBLE
EOF
EOLN

Predeclared Identifiers
-----EXP
EXPO
FALSE
FIND
GET
HALT
INPU1r
INTEGER
LINELIMIT
LN
MAXINT
NEW
ODD

OPEN
ORD
OUTPUT
PACK
PAGE
PRED
PUT
READ
READLN
REAL
RESET
REWRITE
ROUND
SIN

SINGLE
SNGL
SQR
SQRT
SUCC
TEXT
TIME
TRUE
TRUNC
UNDEFINED
UNPACK
WRITE
WRITELN

------

You can redefine a predeclared identifier to denote some other
item.
Doing so, however, means that you can no longer use the identifier for
its usual purpose within the scope of the block in which it is
redefined.
READ
For example,
the predeclared
identifier READ denotes the
procedure, which performs input operations.
If you use the word READ
to denote something else, such as a variable, you cannot use the READ
procedure.
Because you could lose access to useful language features,
you should avoid redefining predeclared identifiers.

1-7

IN'rRODUCTION
The directives FORTRAN,
FORWARD,
EXTERN,
and
EXTERNAL are also
predeclared
by the
PASCAL compiler.
However,
they
retain their
meanings as directives even if you redefine them as identifiers.

1.2.3.2 User
Identifiers - User
identifiers denote
the
names of
programs,
modules,
constants,
variables, procedures, functions, and
user-defined
types.
User
identifiers name all
significant data
structures, values, and actions that are not represented by a reserved
word, predeclared identifier, or special symbol.

1.2.4

Special Symbols

Special symbols represent arithmetic, relational, and
set operators,
delimiters,
and
other
syntax eJements.
PASCAL includes the special
symboJs listed in Table 1-3.
Table 1-3:

Special Symbols

Name

Symbol

Name

Plus sign

Period

Equal

Multiplication

*
<

Not equal

Less than

Exponentiation

Colon

Subrange
ope ra to r

Comma

Parentneses

(

Comment

( * *)

)

Minus sign

Square brackets

Division

Greater than

Semicolon

Assignment
operator

Pointer

Greater than or
equal

Less
equal

1.3

than

Symbol

(

. .)

DELIMITERS

PASCAL uses two special symbols as delimiters:
the semicolon (;)
and
the period (.).
The semicolon separates one PASCAL statement from the
next.
One line of your program can contain one or many statements,
Th.e period marks
but
the statements must be separated by semicolons.
the end of the PASCAL program.

1-8

INTRODUCTION
The semicolon and the period are
the only characters that PASCAL
recognizes as delimiters.
Spaces, tabs, and carriage-return/line-feed
combinations are separators and cannot appear within an identifier,
a
number,
or a
special
symbol.
You must use at least one separator
between consecutive identifiers, reserved words, and numbers;
but you
can use more than one if you want.
You could, for instance, put each
element of a PASCAL program on a separate line:

r-f<OGf<AM
~:; i IYt

(

OUTPUT)

BEGIN
Wr:~ITEL..N

'This is a slmple

pro~ralYt.'

END.
You could also put the entire program on one line:

PROGRAM Sim(OUTPUT);BEGIN WRITELN('This is a simple

pro~ram.')END.

As long as each complete statement is separated from
the next by a
semicolon, PASCAL interprets your input correctly.
However, including
spaces, tabs, and carriage-return/line-feed combinations make your
program easier
to
read
and understand.
For readability, you could
write it as follows:

PROGRAM Sim (OUTPUT);
BEGIN
WRITELN(JThis is a simple
END.

pro~ram.')

The reserved words BEGIN and END are also used as delimiters.
BEGIN
indicates the start of the executable section or a compound statement,
and need not be followed by a semicolon.
END indicates the end of one of the following:
•

A record definition

•

An executable section

•

A compound statement

•

A CASE statement (see Section 5.3.3)

Although PASCAL does not require one, you can use a
semicolon
immediately before END.
A semicolon in this position results in an
empty statement between the semicolon and the reserved word END.
The
empty statement implies no action.

1.4

DOCUMENTING YOUR PROGRAM

In addition to statements and delimiters, you can put comments in your
PASCAL program.
Comments are simply words or phrases that describe
what happens in the program.
You can enclose comments in braces { }, as follows:

< This is a comment.

}
1-9

INTRODUCTION
Also, you can start a comment with the left-parenthesis/asterisk
character
pair, ano end
it wi th the asterisk/right-parenthesis
character pair, as follows:

(* This is also a comment *)
You can also mix the type of comment characters you use.
For example,
you can use a left brace with an asterisk/right-parenthesis character
pair:
{ This is another comment

You can place a comment anywhere a space is legal.
comments are not delimiteo by semicolons.

Unlike statements,

A comment can contain any ASCII character because PASCAL
text of the comment.

ignores

the

NOTE
To turn off braces
as recognized
comment characters, use the /NATIONAL
switch.
See Section 8.4.3 for more
information on thjs switch.

1.5

THE %INCLUDE DIRECTIVE

The %INCLUDE directive allows you to ~ccess statements from a PASCAL
file,
calJed
the
included file,
during compilation of the current
file.
The contents of the included file are inserted
in the place
where the PASCAL compiler finds the directive. This directive can
appear anywhere in the PASCAL program.
Format
/L IST
%INCLUDE 'file specification [{ /NOLIST

}ll'
n

where:
'file specification'

is the name of the file to be
The apostrophes are required.

/LIST

indicates that the included file
is to
be printed in the listing. This is the
default.

/NOLIST

indicates the included file is not to be
printed in the listing.

included.

When the compiler finds the %INCLUDE directive, it stops reading from
the current file and begins reading from the included file. When the
compiler reaches the end of the included file, it resumes compilation
immediately following the %INCLUDE directive.
This directive can appear wherever a comment can appear. An included
file can contain any PASCAL declarations or statements. However, the
declarations in an included file, when combined with the other
declarations in the compilation, must follow the required order for
declarations.

1-10

INTRODUCTION
In the following example, the %INCLUDE directive specifies
CONDEF.PAS, which contains constant declarations.
Main

P~SCAL

file

Program

PROGRAM Student_Courses (INPUT, OUTPUT,
CON!:>T %INCLUDE

TYPE

the

SCHED)~

'CONDEF. PAb'

Schf.~du 1. (.:."~;)

r.:ECOF~D

Yeal"'

( F r', S D, ..J r 1I S r' ) ,

PACKED ARRAY [1 •• 30] OF CHAR,
F'aY'(~nts
PACKED ARRAY [1 •• 40] OF CHAR;
Collc~~ge :
(Eng i nC·:."f;' r i n!:j" Arch i tectu r'(-:~, A~~ r i cuI tl.H'f:~)
E:ND;

NalTlE.\

External File
CONDEF.P~S

MAX .... CLASS ::::

300;

N._PROFS :~ 140;
FROS~'I :::: 3000;
The %INCLUDE directive instructs the compiler to insert the contents
of
the
file CONDEF.PAS after
the
reserved word CONST in the main
program.
The main program Student Courses
is compiled
as
if
it
contained the following:

PROGRAM Student_Courses (INPUT, OUTPUT, SCHED);
CaNST Max_Class = 300,
N_.P T'ofs = 140;
Frosh ::.: 3000;
TYPE Schedules

= RECORD
Year: (Fr, SQ, Jr, Sr);
Name: PACKED ARRAY [1 •• 30) OF CHAR;
Parents
PACKED ARRAY [1 •• 40) OF CHAR,
College : (Engineering, Architecture, Agriculture)
END;

You can use the %INCLUDE directive in another included file;
however,
recursive %INCLUDE directives are not allowed.
If, for example, the
file OUT.PAS contains an %INCLUDE directive for the file IN.PAS,
then
IN.PAS must not contain an %INCLUDE directive for the file OUT.PAS.

l-Ll

INTRODUCTION
An included file at the outermost 1evel of a program is sa id t.o be
included at the
first
leve1.
A file
included by a first-level
included file is at the second level, and so on. There is no limit to
Figure 1-2
the number of included files you can nest in a program.
illustrates some levels of included files.
Main Program

A.PAS

F' T'()~1 rail! p,

CON~:)T

VAR %INCLUDE 'C.PAS'
(* Both Level 2 *)

TYF'E %INCL..UDE 'A.PAS'
(*

LE:.\V0~1

ZINCL..UDE 'B.F'AS'

C.Pl\S

D.PAS

VAR ZINCLUDE 'D.PAS'
(* L..evfd 3*)

FUNCTION ZINCL..UDE

~E.PAS'

PROCEDURE ZINCL..UDE 'F.PAS'
<* Both Level 4 *)
F.PAS

G.PAS

FUNCTION ZINCL..UDE 'G.PAS'
Figure 1-2:

%INCLUDE File Levels

1-12

CHAPTER 2
PASCAL DATA TYPES

This chapter describes PASCAL data types and how to define and declare
them
in
the TYPE and VAR sections of a PASCAL program.
This chapter
also provides general information about using each data type.
PASCAL uses three categories of data types:
1.

Scalar

Structured

Pointer

Scalar data types represent ordered groups of values.
The scalar data
types,
which are described in Section 2.2, consist of predefined and
user-defined data types.
Predefined data types include integers, real
numbers,
and
characters.
User-defined datn types include a range of
explicitly defined values and a subrange of another data type.
Scalar
data types are building blocks for the structured data types.
Structured data types are collections of data
types organized
in
specific ways.
Structured data types include arrays, record files,
and sets.
These are described in Section 2.3.
Pointer data types provide access to dynamic
are described in Section 2.4.

2.1

data

structures.

They

DECLARING DATA TYPES

PASCAL provides two methods of declaring variables of a
particular
type.
You can define
the type in the TYPE section, and then use a
declaration in the VAR section to declare one or more variables of the
newly defined type.
The general format is:
TYPE type identifier = type definition;
VAR variable name: type identifier;
Alternatively, you can declare a
variable by specifying
the
type
definition
in the VAR section and omitting the type identifier and
type definition from the TYPE section.
The general
format
for
this
method is the following:
VAR variable name

type definition;

If a data ~ype is used only once within the program, it is simpler
define it In the VAR section.

2-1

PASCAL

D~TA

TYPES

If a data type is used more than once
in
the
program,
it is more
efficient
to define
the data
type within the TYPE section.
This
creates a structure that can be accessed by more than one
identifier.
For example,
if a program uses an array three times, you can define
the array type in the TYPE section, and assign
three
identifiers
to
that array type in the VAR section.

2.2

SCALAR TYPES

Scalar data types consist of ordered sets of values with
the concept
of predecessor and
successor.
For example,
the scalar data type
INTEGER represents whole numbers that follow in a predefined sequence:
5
is less
than 300.
Scalar data types encompass two subclasses:
predefined and user-defined.
These are described
in the
following
sections.

2.2.1

Predefined Data Types

PASCAL provides the following predefined scalar data types:
1.

INTEGER

REAL

SINGLE

DOUBLE

BOOLEAN

CHAR

The
predefined
types
SINGLE
and
DOUBLE
provide
explicit
single-precision and double-precision real numbers.
Throughout this
manual, the term REAL refers to REAL, SINGLE, and DOUBLE types.
The following sections describe each predefined data type.

2.2.1.1
INTEGER Data Type - The type
INTEGER denotes positive and
negative whole number values ranging from (-2**35) to (+2**35)-1, or
-34359738368 to +34359738367.
The largest possible value of
the
INTEGER data
type
is
known by the predefined constant identifier
MAXINT.
You can indicate a decimal
integer
combined with plus and minus signs.
constants in PASCAL:

constant with decimal digits
The following are valid decimal

-333

+1
89324
A minus sign (-) must precede a negative integer value.
A plus sign
(+)
may precede a positive integer, but the sign is not required.
No
commas or decimal points are allowed.

2-2

PASCAL DATA TYPES
In addition to decimal notation, PASCAL allows you to specify integer
constants
in binary, octal,
and hexadecimal notation.
You can use
constants written in these notations anywhere that decimal
integer
constants are permitted.
To specify an
integer constant
in binary, octal,
or hexadecimal
notation,
place a percent sign (%) and a letter in front of a number
enclosed in apostrophes.
The appropriate letters, which can be either
uppercase or lowercase,
are B for binary notation,
0 for octal
notation, and X for hexadecimal notation.
An optional plus or minus
sign can precede the percent sign to indicate a positive or negative
value.
Note that regardless of which notation you use, the value can
not exceed MAXINT, for example:

-%B'111001'
%b'10000011'
%0'7712'
-%0'473'
+%X'53Al'

-%x'DEC'

2.2.1.2 REAL Data Type - The reserved words REAL, SINGLE, and DOUBLE
denote the real number types.
In PASCAL, a real number can range from
+-0.14*10**-38 through +-3.4*10**38, with a typical precision of eight
decimal
digits.
REAL
and SINGLE are synonymous;
both have
single-precision real number va]ues.
The type DOUBLE allows you to
declare double-precision
real
variables.
You can assign real and
integer values to a variable of type REAL, SINGLE, or DOUBLE.
If you
assign an
integer value to a variable of type REAL, PASCAL converts
the integer to a real number.
In a PASCAL program, you can write real numbers in two ways;
fixed on
floating
point. With fixed point notation, you write the number with
the decimal point exactly where it appears in the value.
The first
way is the following form:

2.4
893.2497
-0.01
8.0
-23.18
0.0
Note that, in this form, at least one digit must appear on each side
of the decimal point.
That is, a zero must always precede the decimal
point of a number between 1 and -1, and a zero must follow the decimal
point of a whole number.
Some numbers,
however,
are too large or too small
to
write
conveniently in the above format.
PASCAL provides scientific (also
known as exponential)
notation as a
second way of writing
real
numbers.
In scientific notation, you write the number as a positive
or negative value followed by an exponent, for example:

2.3E2
-0.07E4
10.0E-1
-201E+3
-2.14159£0

2-3

PASCAL DATA TYPES
The letter E after the value means that the value is to be multiplied
by a
power of 10.
Note that you can use an uppercase or lowercase
the
power of
10;
the
letter.
The integer following the E gives
integer can be positive or negative.
Using scientific notation, you
can write the real number 217.0 in any of the following ways:
237E'O
::.~. ::~7E2

O.000237E+6
2370E····1
O.()()00000237E10

This format is often called floating-point format because the
implied
position of
the decimal
point "floats"
depending on the exponent
following the E.
At least one digit must appear on each side of
the
decimal point, if the decimal point is present.
PASCAL provides single and double-precision
representation
for
real
numbers.
Single precision
typically provides eight significant
digits, depending on the magnitude of the
number.
Double precision
extends the number of significant digits to 18.
To indicate a double-precision value,
you must
use
floating-point
notation, replacing the letter E with an uppercase or lowercase D, for
example:

ono
4. 371 ~7j2Bf.)65D··"3
--812d2
4[1··-3

The integer following the D is an
exponent,
as
in
single-precision
floating-point
numbers.
All
the above values have approximately 18
significant digits.

2.2.1.3 BOOLEAN Data Type - BOOLEAN data types can have
the value
TRUE
or FALSE.
Boolean values are the result of testing expressions
for truth or validity.
The result of a
relational
expression
(for
example, A < B) is a Boolean value.
PASCAL defines Boolean data types as predefined identifiers and orders
them so
that
FALSE is less than TRUE.
For assignment purposes, the
type BOOLEAN is compatibJe with those variables and
expressions
that
yieln a BOOLEAN result.

2.2.1.4 CHAR Data Type - The value of data
type CHAR
is a
single
value
from
the ASCII
character
set,
as listed in Appendix B.
To
specify a character value, enclose an ASCII character in apostrophes.
The
apostrophe
character
itseJf
must be
typed
twice within
apostrophes.
Each of the following is a valid character value:

'A'
, z"

'0'

, ,

,,,,
'? '

2-4

PASCAL DATA TYPES
You can use strings such as 'HELLO' and '****', but you must represent
them
as packed arrays of characters (see Section 2.3.1.2).
When you
use the ORD function in an expression of type CHAR, the result is
the
ordinal
value in the ASCII character set of the character value.
See
Section 2.2.2.3 for an explanation of the ORD function.

2.2.2

User-Defined Scalar Data Types

User-defined scalar data types are those that you define,
as opposed
to those data types that PASCAL predefines for you.
PASCAL allows you
to define two kinds of scalar data types:
enumerated
and
subrange.
An
enumerated
type consists of an ordered list of identifiers.
The
subrange type is a continuous range of values of a
defined
scalar
type,
called a base type.
The following sections describe these two
user-defined types.

2.2.2.1
Enumerated Data Types - An enumerated data type is an ordered
list of identifiers.
To define an enumerated type, list in some order
all the identifiers denoting its values. With PASCAL, you can define
an enumerated data type in two ways:
Fo rma t I
TYPE identifier

(identifier

II,identifier,

•.• ] )

(identifier

I[ ,identifier,

•.• ] )

Fo rma t 2
VAR identifier
where:
identifier

is the name of the enumerated type.

For example:

TYPE Beverage = (Milk, Water, Cola, Beer);
This TYPE section defines the type Beverage and lists all
that Beverage can assume within a program.

the

values

PASCAL assigns an order to the items in your list from left to
right.
Thus,
the values of an enumerated type follow a left-to-right order,
so that the last value in the list is greater
than
the
first,
for
example:

TYPE Seasons = (Spring, Summer, Fall, Winter) ,
The relational expression (Spring <
precedes Fall in the list of values.

Fall)

TRUE

because

Spring

The only restriction on the values of an enumerated type is that you
cannot define
the
same value in more than one list in the same TYPE
section.
For example, the following is illegal:

TYPE Seasons = (Spring, Summer, Fall, Winter)
School~ear = (Fall, Winter, Spring);

2-5

PASCAL DATA TYPES
To initialize a variable of an enumerated type, specify a constant
value.
For example, you can assign the variable Quarter of type
Seasons as follows:
VAR Quarter:

Seasons:= Fall;

The variable Quarter takes on the initial va]ue Fall.
Examples
TYPE ColoY's "~.(Rf:~(h Ypllow, Gret~r" Purple, Blue);
Sport = (Swim, Run, Ski);
Beverage = (Milk, Water, Colay Beer);
VAR Cookie:
(Oatmeal, Choc-Chipy Peanut-Butter,
Exercise, Fun: Sport:= Ski;
Drink:
Bevera~e;

Su~ar)

The TYPE section defines the types Colors ann Sport, listing
values that variables of each type can assume.

Su~ar;

all

the

The VAR section declares the variable Cookie, which can have the
values Oatmeal, Choc-Chip, Peanut-Butter, and Sugar. The variables
Exercise and Fun are declared as type Sport, and Drink is declared as
type Beverage.
Initial values are established for the identifiers
and Fun in the VAR section.

Cookie,

Exercise,

2.2.2.2 Subrange Data Types - A subrange specifies a limited
portion
of another scalar type (called the base type) for use as a type.

Fo rma t 1
TYPE identifier

lower limit •• upper limit

Format 2
VAR identifier

lower limit •• upper limit

where:
identifier

is the name of the subrange.

lower limit

is the constant
subrange.

value

the

low

end

the

end

the

separates the limits of the subrange.
upper limit

is the "constant value
subrange.

the

high

The subrange type is defined only for the values between and including
the lower and upper limits. The limits you specify must be constants;
they cannot be expressions.
(See Chapter 3 for information on
expressions.)
The values in the subrange are in the same order as in
the base type.

2-6

PASCAL DATA TYPES
The base type can be any enumerated or predefined scalar
type
except
REAL.
You can use a subrange type anywhere in the program where its
base type is legal.
The rules for operations on a
subrange are
the
same as the rules for operations on its base type.
A subrange and its
base type are compatible.
The use of subrange types can make a program clearer.
For example,
integer values
for
the
days
of the year range from 1 to 365.
Any
value outside this range is obviously incorrect.
You could specify an
integer subrange for the days of the year as fo]lows:

VAF:

Da~,~····Of···· y €·~a r

J •• 366

By specifying a subrange, you indicate that the values of the variabJe
Day-Of-Year are restricted to the integers from 1 to 366.
Example

TYPE Months = CJan, Feb, Mar, Apr, Mas~ Jun,
Jul, Au~, Sep, Oct, Nov, Dec),
VAR Camp_Mos: Mas •• Oct;
Leaf_Mos: Sept.Nov;
First_Half:
'A' •• 'M';
We> l"'cf:

() •• t.l::.'.i5::~~;'

This example defines the variables Camp_Mos and Leaf_Mos as
subranges
of the enumerated type Months.
A Camp Mos value can be only May, Jun,
Jul, Aug, Sep, or Oct.
A Leaf Mos val~e can be only Sep, Oct, or Nov.
The variable
First Half
is
subrange of the ASCII characters, with
possible values uppercase A through uppercase M.
The variable Word is
a subrange of the integers from a to 05535.

2.2.2.3 The ORD Function - Each element of a scalar type (except
the
REAL
type) has a unique ordinal value, which indicates its order in a
list of elements of its type.
The ORD function
returns
the
ordinal
value as an integer, for example:

OR[lC '(~')
This expression returns 81, which is the ordinal value of uppercase
Q
in
the
ASCII character set (see Appendix B).
Note that the order of
the ASCII character set may not be what you expect.
The numeric
characters are in numeric order, and the alphabetic characters are in
alphabetic order.
All uppercase characters have lower ordina]
values
than all lowercase characters, for example:
ORDC'Q')
OF~D( 'A')
ORDC'Z')

is less than ORDC'a') and
is less than ORDC '7') but
is less than ORD('a')

You can use ORD on a value of an enumerated
are ordered starting at zero, for example:

type.

Enumerated

types

Of,[I CTuef.)da~)

Assuming that Tuesday is a value of type Weekdays
values
Monday,
Tuesday,
Wednesday,
Thursday,
expression returns the integer 1.

the
(which includes
this
and
Friday),

The ordinal value of an integer is the integer itself.
For examp]e,
ORD(O) equals 0, ORD(23) equals 23, and ORD(-1984) equals -1984.

2-7

PASCAL DATA TYPES
2.3

STRUCTURED DATA TYPES

A structured data type consists of a
collection of related data
components;
it is characterized by its method of structuring and its
components.
All structured data types consist of a
collection of
elements or components that are grouped together in a structure in
which they can collectively be manipulated.
PASCAL provides four structured data types:
•

ARRAY

•

RECORD

•

SET

•

FILE

An array is a group of components of a predefined size and of the same
type.
A record
consists of one or more named fields, each of which
contains one or more data
items.
Records can
include fields of
different data types.
A set is a collection of data items of the same
scalar type, the base type.
You can access a set as an entity,
but
you cannot access the set components as individual components or
variables.
A file is a sequence of data components that are of the
same type;
each component can be individually accessed.
A file can
be of variable length.
Section 2.3.1 describes arrays;
Section 2.3.2 describes records;
Section 2.3.3 describes sets;
and Section 2.1.4 describes files.

2.3.1

Array Types

An array is a group of components of the same type that share a common
name.
You refer to each component of the array by the array name and
an index (or subscript).
An array type definition specifies the
type
of the indices and the type of the components.
Format
ARRAY [index type

[,index type ••• ]

] OF component type

where:
index type

specifies the type of the index.
The
(ndex
type can be a subrange, CHAR, BOOLEAN, or
enumerated type;
but it cannot be a REAL
type.

component type

specifies the type of the components
array.

the

The components of an array can be of any type.
For example, you can
define an array of integers, an array of records, or an array of real
numbers. You can also define an array of arrays, which is known as a
multidimensional array.
The indices of an array must be of a scalar type, but cannot be real
numbers.
Note that you cannot specify the type INTEGER as the index
type.
To use integer values as indices, you must specify an integer
subrange,
unless you are using
conformant-array parameters.
If
necessary, PASCAL determines the subrange.
For example, if the
index

2-8

PASCAL DATA TYPES
is BOOLEAN, then PASCAL converts the subrange to
FALSE •. TRUE
because
the
type
BOOLEAN has only two legal values.
For more information
about conformant-array parameters, refer to Section G.3.2.
The range of the index type establishes the size of the array and
way it is indexed, for example:

TYPE Letters
VAR Letl:

ARRAY

[1~.10]

the

CHAR~

Letters~

The array variable LETl has 10 components,
referred
LETl r2], LET 1 [31, and so on, through LETI no].

LETlrl],

You can use array components
in expressions anywhere
you can use
variables of
the component type.
For the array as a whole, however,
you can use only the assignment statement (:=).
An exception to
this
rule
is made
for
character strings, which PASCAL defines as packed
arrays of type CHAR.
See Section 2.3.1.2.

2.3.1.1

Multidimensional Arrays - An array with components of an
array type is a multidimensional array.
An array can have any number
of dimensions, and each dimension can have a
different
index
type.
For example, the following declares a two-dimensional array variable:

VAR Two_D:

ARRAY [0 •• 4J OF ARRAY ['A' •• 'D'] OF INTEGER;

PASCAL allows you to abbreviate the definition by specifying
index types in one pair of brackets, for example:

all

the

ARRAY [O •• 4,'A' •• 'D'] OF INTEGER;
To refer to a component of this array, you specify two
indices,
one
integer
and
one character,
in the order
they were declared:
Two_D[O,'A'l,
Two DrO,'B'],
and
so on.
You can
a]so
specify
Two D[O]r'A'].
The
first index indicates the rows of the array, and
the-second index indicates the columns.
Hence, you can picture
the
array Two D as in Figure 2-1.

'A'

'8'

'C'

'0'

4
MR-S-3113-83

Figure 2-1:

Two-Dimensional Array Two D

2-9

P~SCAL

DATA TYPES

When referring to the components of Two D, the first component in
the
first
row
is Two Dro,'.a.'l.
The
second
component
in this row is
Two D ro,' B' 1.
The Ii rst component in the second row is Two D rl,' A'] •
The- last component
in
the
last
row is Two Dr4,'D'].
general,
element j of row i is Two_Dri,jl.
-

You can define arrays
fashion, for example:

VAF< Ch (.;.~ s s :3 II

three

: . f.l f< F< AY [1..;3 ~

:I... n,

dimensions

similar

nr::. •1\ F< ::I 0 F C h (.:.~ ~:; ~:; 1'1'1 e 1"1 ;

This declaration
specifies a
three-dimensional
chess game.
The
indices of
the array are the levels, the ranks, and the files of the
chessboard.
For example, the reference Chess3D fI, 1,
QR]
specifies
the
first level, first square in the upper left corner (bottom lev~l,
first rank, Queen's Rook file).
Figure 2-2
illustrates
the
three
levels of this array.
Chess3D(1,n,
Chessmen)
(bottom)

Chess3D(2,n,
Chessmen)
(middle)
1

OR ON OS

KS KN KR

Chess3D(3,n,
Chessmen)
( top)

Cf-IESS3D{3.n. CHESSMEN/

CHESS3D{2.n.CHESSMEN/

CHESS3D{I.n.CHESSMENj

lK 098 81

Figure 2-2:

Three-Dimensional Array Chess3D

When storing values in an array, PASCAL increments
the
indices
from
right
to left.
Thus, PASCAL increments the rightmost index until the
maximum value is reached, then moves to the left to
the
next
index,
and
so
on,
until all indices have been incremented to the specified
amoun t.

2-10

PASCAL DATA TYPES
In the three-dimensional array Chess3D, PASCAL starts by holding
the
first
two
indices constant while stepping
through
the values of
Chessmen.
Thus, the first values are assigned to components Chess3D
fl,I,QR]
through
Chess3D
fl, 1 ,KR] •
Next,
the second
index
is
incremented and values are assigned to the components Chess3D [1,2,QRl
through Chess3D[1,2,KR1.
After these eight elements are assigned, the
second index is again incremented, and values are assigned to Chess3D
rl,3,QR]
through Chess3D rl,3,KR].
The assignment process continues
with the first index held constant until the
second
index has been
incremented
from 1
to 8.
Then, the first index is incremented, and
the process is repeated.
Hence,
all
values for
the bottom
level
(denoted
by Chess3D[1,n,Chessmenl)
are stored before any values for
the middle level (denoted by Chess3D [2,n,Chessmen]).
The
top level
(denoted
by Chess3Dr3,n,Chessmen]) receives its values last.
Figure
2-3 illustrates this order.

OR ON as
-

~:-:~;:: ::::=::=-+f----+----+--..--1--_--1

~ . __ j 4 .....-,1--t--+-+-+--+-.-+---1
: - - --".5

-- -or'

""'-'--+-+--f---t-- -

[:: -.> 6 ~I--t-+-+--+--+--+----l

(: ~)7 I--+--+--+--f---t--- _ .

:_-. ~ -~> 8

OR ON os

KS KN KR

1 .....-,I--t--+- I---+--+--+---i

'---'---'---L-

OR ON

KB KN KR

os a

KS KN KR

~12 f------1--t--+-+--+-+--+--1
• - - . A.

~. -.-J

_:.'3 -f--

- ----. f---

.'.

f--f----jl--t--+-+--+-+-t

~ ~ ~ :.) 5 f----t-+-+-t---+-+--+-t
~ :-.:. ' 6 .....-,I--t-+-+--+--- f----

:->
~ -;"

7 f------I--t--+-+-+- +-_ +---u.

'-----J'----'-----L---'_--'---'---'--~II

CHESS3D (I ,n,CHESSM[N (
lbo)ttom)

CHESS3D (2.n,CHESSMEN (

CHESS3D (3,n,CHESSMEN]

(middle)

Itop)

MR-S-3115-83

Figure 2-3:

Storing Components in an Array

2.3.1.2 String Variables - A character string variable in PASCAL is
defined
as a packed array of characters with a lower bound of 1.
To
declare a string variable,
specify a
packed
array of the proper
length, for example:
VAR NAME

PACKED Af;:RAY [1 + + 20] OF CHAR;

This declaration allows you to store a string of 20 characters in the
array variable NAME.
The length of this string must be exactly 20
characters.
PASCAL neither adds blanks to extend a shorter string nor
truncates a longer string.
You can assign to a string variable the value of any string constant
or variable of the defined length.
You can also compare strings of
the same length with the relational operators <, <=, >, >=, =, and <>.
The result of a string comparison depends on the ordinal value (in the
ASCII character set) of the corresponding characters in
the strings,
for example:

'motherhood'

> 'apple pies'

This relational expression is TRUE because lowercase 'm'
comes after
lowercase
'a' in the ASCII character set.
If the first characters in
the strings are the same, PASCAL looks for differing characters, as in
the following:
Istrin~l'

< 'strin~2'

This expression is also TRUE because the digit I
in the ASCII character set.

2-11

precedes the digit

PASCAL DATA TYPES
To assign a string constant to an array in the executable section, use
an assignment statement.
The string variable must be of the same size
as the array;
otherwise, an error occurs.
The
following
example
shows an assignment statement in the executable section:
TYPE

Strin~

VAR

Word

Strin~;

PACKED ARRAY[1 •• 10J OF CHAR;

BEGIN
W0 l' (.j

: ~~

.' () T' a n ~.:.~ e

END;
The string constant 'orange'
of the array.

is padded with spaces to match the

size

The READ and READLN statements automatically pad the
string
constant
if necessary.
Thus,
it is not necessary to pad the string constant
with spaces to match the varjable size, when using the predefined file
INPUT or reading from a file defined as TEXT.

2.3.1.3
Initializing and Assigning Values
to Arrays - You should
assign values to an array either in the declarations or the executable
section before using the
array within
the
program.
As
with all
variables, the value for each component is undetermined, until a value
is specifically assigned.
To assign values to an array in the executable section, a
value must
be assigned to each component in the array.
One method of doing this
is to use a
FOR statement.
By using
the
FOR statement control
variable as
the
array
index,
it
is possible to step through all
components of the array, setting them to the same initial value.
Each
index for the array is incremented at the same time the counter in the
FOR statement is incremented.
An example of this is:

VAR Arrau_ExaffiPle : ARRAY [1 •• 10J OF INTEGER;
I ncip>: : INTEGER;
BEGIN

FOR Index := 1 TO 10 DO
A T' r a ~~ .... E ~.: a IT! P 1 €.~ [I n (j (., ~-: :1 : :::: 0;

END,
In this example, the array has been defined as being an array of
type
INTEGER.
As
the
FOR
loop executes each
time,
the counter
is
incremented by one.
Likewise, the index is incremented
by one.
On
each execution of the loop, the current component is assigned a value
of o.
An array can be initialized in the VAR section.
The following example
shows each component of the array Array_Example being initialized with
the value 0:
ARRAY [1 •• 10] OF INTEGER

2-12

:= (10 of 0);

PASCAL DATA TYPES
To assign values to a two-dimensional array in the executable section,
you can use two nested FOR statements to increment the two indices, as
shown in the following example:
Zf.~ ro :::: <>;
VAR Table: ARRAY [1 •• 10,1 •• 5J OF INTEGER;
i n (.l f:·~ >~ .... 1 , i :1".' (i (.:~ >( .... 2 : I NT E GE R ;

CON~:>T

BEGIN
F (] F~ i n d f:~ ~.( .... 1

: ::::

FOR index_2

:I. T ()

° DO

:= :l TO 5 DO

Table[index_:l,index_2]

:= Zero;

END~

Array_Example is
INTEGER.

:lefined

two-dimensional

array

type

The nested FOR statements assign the value of
Zero
(which has been
assigned
the value of "0" in the CaNST section) to each component in
the array.
To initialize a two-dimensional array in the VAR section,
specify a
constructor for each row, in parentheses.
The following example shows
a two-dimensional array that is initialized in the VAR section:

VAR Table

ARRAY [1 •• 10,1 •• 5] OF INTEGER

:= (10 OF (5 OF 0»;

To assign values
to arrays of
three or more dimensions
executable section, use three or more nested FOR statements:

the

CONST Zero::::
VAR Table: ARRAY [1 •• 5,1 •• 3,1 •• 2J OF INTEGER,
index_l"index_2,index_3 : INTEGER;
BEGIN

FOR index_1 := 1 TO 5 DO
FOR index_2 := 1 TO 3 DO
FOR index_3 := 1 TO 2 DO
TableCindex_l,index_2]

:= Zero;

END;
This example shows the initialization of a
three-dimensional
array.
The value of· Zero
is assigned
to each element in Table, from
Table[l,l,l] to Table(5,3,2].
To initialize an array of three or more dimensions in the VAR section,
specify a
constructor
for
each row.
The following example shows a
three-dimensional array being initialized in the VAR section~

VAR Table: ARRAY [1 •• 5,1 •• 3,1 •• 2] OF INTEGER :=
(5 OF (3 OF (2 OF '

2-13

'»);

PASCAL DATA TYPES
2.3.1~4
Array Type Compatibility - You
can
assign one
array
to
another only if the arrays are either identical or compatible.
Arrays
of the same type or equivalent types are
identical.
The
following
example demonstrates identical arrays:

TYPE
VAR

Salar~

ARRAY [1 •• 50J

Pay

= Salary;

WaSe~

Income

or REAL;

Money : Pay;
The arrays Wage and Income are identical
because
both are
of
type
SALARY.
The
array ~oney of type PAY is identical to Wage and Income
because the type PAY
is declared
equivalent
to
the
type
SALARY.
Identical arrays are always compatible.
Arrays that Are not identical
following criteria:

are compatible if they meet all

•

They have the same number of components.

•

Their elements are of compAtible types.

•

Their indices are of compatible types.

•

The upper bounds of their indices are equal.

•

The lower bounds of their indices are equal.

•

Both are packed or neither is packed.

•

For packed arrays
of
subrange
types,
the
subranges must be the same for both types.

The following

two array types,

bounds

the

though not identical, are compatible:

TYPE Grades = ARRAY [1 •• 28] OF 0 •• 4;
Feb_Temps = ARRAY [1 •• 28J OF INTEGER;
Both types define arrays with 28 components, indexed
from
1
to
28.
The integer subrange components of type GRADES are compatible with the
integer
elements of
type
Feb Temps.
Therefore,
you
can assign
variables of
type GRADES
to -variables of type Feb Temps, and vice
versa.
Note that, if the TYPE definition specified packed arrays, the
types GRADES and Feb_Temps would not be compatible.
PASCAL does not check for valid assignments to subranges that are part
of a structured type.
If you assign an array of type Feb Temps to one
of type GRADES, you must ensure that the values are
in -the
correct
range.
An
out-of-range assignment does
not
result
in
an error
message, even if the CHECK option is enabled at compile time.

2-14

PASCAL DATA TYPES
2.3.1.5

Array Examples -

Example 1

TYPE Times = 1 •• 10;
VAR Raceresults : ARRAY[1 •• 50J OF Times;
I : I NTEGEfn

BEGIN

:= 1 TO 50 DO
Raceresults[IJ != 0

FOR I

END;
This example decl~res the variable Raceresults as a 50-component array
of Times.
The FOR statement assigns zero to each component in the
array.
Example 2
T YP E f:) t, r i n 9 ,,: PAC KED AF~ RAY [1.. 1 () ] 0 F CH AFU
VAR Composer, Word, Empt~ : String;

BEGIN
Word := 'engrossing';
Composer := 'C.P.E.Bach';
EITJPt~:J

:::::

END;
This example declares
three string variables.
It assigns string
constants
to the variables Word and Composer, and assigns a string of
10 spaces to the variable Empty.
Example 3

CONST Da\.~~:; :::: 31;
TYPE Weather:::: (Rain, Snow, Sunn~, Cloud~,
Month = ARRAY C1 •• DAYS] OF Weather;

Fogg~);

This example shows how you can use a constant identifier in the
index
type.
The
indices of arrays of type Month range from 1 to the value
of the constant Days.

2.3.2

Record Types

The record is a convenient way to organize several related data
items
of different types.
A record consists of one or more fields, each of
which contains one or more data items.
Unlike the components of an
array,
the
fields of a record can be of different types.
The record
type definition specifies the name and type of each field.

2-15

PASCAL DATA TYPES
Fo rma t
RECORD
field id : type
[;field id
variant clause }

type ... ]

[variant clause]

END;
where:
field

specifies
the
names of one or more
fields.
The names must be identibfiers
and must be separated by commas.

type

correspond ing
specifies the type of the
field(s).
A field can be :=Iny type.

variant clause

variant
of
the
specifies
the
part
fo r the
record.
See Section 2.3.7..1
fo rma t of a variant c]ause.

The names of the fields must be unique within the record, but can
be
repeated
in different
record types.
For instance, you could define
the field NAME only once within
a
particular
record
type.
Other
record types, however, could also have fields called NAME.
The values for the fields are stored in the order in which the
are defined, for example:
VAf~

fields

TealYl ... HE'C ! F,[COHD
I,J:i.n'5: INTEGEH;
Losses : INTEGER;
Pe T'cent : F<EAL
END;

The values for these fields are stored
Percent.

the

order

Wins,

Losses,

To refer to a field within a record, specify the name
of
the
rec~rd
variable and
the
name
of
the
field,
separated by a period.
For
example, Team Rec.Wins, Team Rec.Losses, and Team Rec.Percent refer to
the
three
fTelds
of
the -record
Team Rec decTared above.
You can
specify a field anywhere in the program that a variable of
the
field
type is allowed.
Thus, you could write:

Records can include fields that are themselves records,

VAR Order : RECORD
Part : INTEGER;
Received : RECORD
Month

for example:

(Jan, Fe~, Mar, Apr, Ma~, Jun,
Jul, Au~, Sep, Oct, Nov, Dec);

: 1 •• 31;
Year : INTEGER
END;
Inventory : INTEGER
END;
Da~

2-16

PASCAL DATA TYPES
The
fields
in
this
record
are
referred
to
as
Order.Part,
Order.Received.Month,
Order.Received.Day,
Order.Received.Year,
and
Order. Inventory.
The WITH statement provides an abbreviated
notation
for specifying the fields of a record (see Section 5.5).

2.3.2.1 Records with Variants - To allow a
record
to
contain
different data
types at different
times,
you can define a record
variant.
To do this,
specify one or more variants
in
the TYPE
definition.
A variant is a field or group of fields that can contain
a
different
type or amount of data
at different
times during
execution.
Thus,
two variables of the same record type can contain
different types of data.
To specify a variant, include a variant clause
in
the
record
type
definition.
The variant clause must be the last field in the record.
Fo rma t
CASE tag fieJd OF
case-label list

( [field id

type]

type ..• ] );

[;field id

where:
tag field

indicates the current variant of the
record.
You can specify the tag field in two ways:
I .

tag name :

tag type

If you use this form,
the
tag
field
is a
field
in the
record
that is common to all
variants.
Tag name and tag type define
the
name and
type of this field.
The tag type
can be any scalar type except a
REAL
type.
You can use
the
tag field in the same way
that you use any other field in
the
record;
that
is,
you can use the record.fieldname
fo rma t.
2.

tag

tYPE~

If you use this form, you must keep track of
the currently valid variant.
The tag type
can be any scalar type except a REAL type.
case-label list

specifies one or more constants
field type.

field id

specifies the names of one or more
fields.
The
field names must be identifiers and must
be separated by commas. Note that,
instead
of
the
field
identifiers,
you can specify
another variant clause,
as in
the
last
example in this section.

type

~pecifies

the

the type of the variant field.
type cannot be a FILE type.

2-17

tag

The

PASCAL DATA TYPES
When you specify the tag field in the first
form
(tag name
tag
type), you should reference on] y the fields in the currently val i(]
variant. The following example shows the use of this form:

TYPE

Name ~ PACKED ARRAY [1 •• 20J OF CHAR,
D a '.:~ ;;;: (M 0 n ~ T 1..1 (~~ y Wp (j ~ T h IJ, F l' i ) ;
~~ t () C k ::" F< E C Cl F< 1:1
F'd r t : 1.. (;9 (J 9 ,
Stoc~_Quantit~

: INTEGER,

f:>upp J if:" r : Name;
CASE Onorde1' : BOOLEAN OF
T"HUE :(P1'omi~:;pd: Da~:I;
I NTEGEF<;
o T' d f.~ T' _. (J u ant i t ~~
Price: REAL);
Da~:~ ;
FALf:;C : (l..ast .... :~hipmf.·~nt
INTEGER,
F<pc .. _Ouant:i. t~~
Co~:;t
: F<EAI. . )

END,
Tn this example, the last three fields in the record
type vary
depending on whether the part is on order. The tag name Onorder is
defined in the variant clause. Records for which the value of Onorder
is TRUE will contain information about the current order. Records for
whi~h this variable
is FALSE will contain information about the
previous shipment.
In the second way of specifying the tag field,
type, as in this example:

TYPE

=
=

Name
PACKED ARRAY [1 •• 20J
Date ::" I NTEGEf<;
Sex
(Female, Male),
HOSF'

;:~

you

use

only

tag

OF CHAR;

RECOF<D
Pat i f:~nt

Nam~l;

Birthdatp : Date;
A~H?

CASE

I NTEGEI=i:;
Se~<

OF
1 •• 30) ;

Fe ITI ale : ( B i r t h s
Male: ()

END,
Tn this example, you must keep track of the currently valid variant.
You can define a variant onJy for the Jast field
in the record.
Variant fields can, however, be nested, as in the following example:

TYPE

Name

= PACKED ARRAY [1 •• 20] OF CHAR,

Date =" INTEGER;
Sex = (Female, Male);
HosF'

= RECORD

Patient : Name;
Bi1'thdate

Age :

Date;

INTEGER;

CASE Pa1'sex : Sex OF
Male :
Female

( )

,(CASE Births : BOOLEAN OF
FALSE: ();
TRUE :

(Nok i ds :

INTEGER»

END;

This record type contains the name, birthdate, age, and sex of all
patients.
In addition,
it includes a variant field for each woman
based on whether she has had any children. A second variant, which
contains the number of children, is defined for women who have given
birth.
2-18

PASCAL DATA TYPES
2.3.2.2 Assigning Values to Records - To assign values
to a
record
variable,
use an assignment statement to specify a value for each
field of the record.
The following example shows the declaration of a
record and the assignment of values to two of the fields.

TYPE ROSTER

- RECORD
Name: PACKED ARRAY[1 •• 30J OF CHAR;
Number: INTEGER;
Grade:: I:::EAL..;
END;

VAR Student
ROSTER;
Testl,Test2,Test3

INTEGER;

BEGIN
WRITEL..N ('Enter your name.');
READLN (Student.Name);
Stu(·.lf~'nt • GJ'i:H:.h:~ : =:: (T es t 1 + TE~S t2

+ T€.'S t3) /

:3;

If you are initializing a record with a
variant clause,
you must
always specify a value for the tag field, even if it has no tag name.
Specifying a value for the tag field, ensures that PASCAL
initializes
the correct variant, for example:

TYPE Sex = (Male, Female);
Pe rson ::;; RECORD
Birthdate: RECORD
Month: 1 •• 12;
Day:
1 •• 31 ;
Year: INTEGER;
END;
Ase: INTEGER;
CASE s~~~< of
(Bearded:BOOLEAN);
Male:
Female: (N_Children:INTEGER);
END;
VAR Dad: Person := «5, 15, 1921), 62, Male, TRUE);

2.3.2.3 Record Type Compatibility - Two
records are
their types are identical or equivalent, for example:

TYPE Life

compatible

= RECORD

INTEGER;
:Born
INTEGER
:Died
lEND;
Plantlife = Life;
VAR

Mom, Dad : Life;
Coleus : Plantlife;

The record variables Mom, Dad, and Coleus are all compatible.
Mom and
Dad are both of type Life, which is equivalent to type Plantlife.

2-19

PASCAL DATA TYPES
Records of differing types are compatible if they meet
criteria:
•

They have the same number of fields.

•

Corresponrling field types are compatible.

•

Both are packed, or neither is
packed, corresponding fields
equal bounds.

packed.
If
of subrange

the

following

the types are
types must have

The following type is also compatible with Life and Plantlife:

TYPE Coords

= RECORD
X

Y :

INTEGER,

O •• lOO

END;
The integer subrange O.. 100 is compatible with the
type INTEGER.
However, PASCAL does not check for valid assignments to fields of
subrange types.
If you assign a record of type Life to a record of
type Coord, you must ensure that the value of the field Died is within
the subrange O .• 100. An out-of-range assignment does not result in an
error message.
If the records have variants, these criteria also apply:
•

The records must have the same number of variants.

•

Corresponding variants must have the same number of fields.

•

Corresponding field types within corresponding variants
be compa t i bl e.

•

The case labels associated with the variants
number, but need not agree in value.

•

Corresponding variants in structurally
must have identical tag constant values.

•

The tag constant lists in each record must be identical.

2-20

must

compatible

agree

must
in

records

PASCAL DATA TYPES
For example,
definition:

assume

the

program

includes

'A/ •• /D/;:
I<[CORD
Siz(? : INTEGER;
Calories : INTEGER;
ProtE~in :: 0 •• 40;
ea rb : I NTEGEFU
CASE Vi ts : Lf:~ts OF
'A'!"C','D': ( ) ;
'B' : (Niacin, Thiamine
END,
'A' •• 'F',
GT'adps
F,[CORD
School
Studentno : INTEGER,
Clas~:;:
1 •• 5;

the

following

TYPE

TYPE I... (::~t~:;

Info

HCH.IT'S

BOOLEAN)

1 •• ~~O;

Incompletes : 1 •• 6;
CASE Average : Grades OF
'B','C','D' : ( );
'A' : (Sendlet, Firstsem
END;

BOOL.EAN)

The types Info and School are compatible.
If you assign a variable of
one type to the other, however, you must be sure that both contain the
same variant.

2.3.2.4

Record Examples -

Example 1
TYPE Taxes - RECORD
Year : INTEGER;
Gross : REAL;
Net : REAL;
Deductions : INTEGER;
Itemized
BOOLEAN;
Interest
ARRAY [1 •• 5] OF REAL
END;
VAR Fed

Taxes

:= (1981, 25234.12, 18789.00, 4, TRUE,
(5 of 0.05»;

This example declares and initializes the record Fed of type Taxes.

2-21

PASCAL DATA TYPES
Example 2
T Y PES t l' i n ~.l :::; F' (.1 C1\ E It A F~ F< A Y [: 1. • • 2 () ::I

Personal

CH A F~ ;

F~[COF<D

N a ITt e

~:; t T' i n ~.:.l ;

Add T'e~;~:;

~-::ECnF<D

Numbe1' : INTEGER,
Town :
Zip: 0 •• <;<;<;99

~:)t1'€~pt,

!:;trin~:j;

END,
A~lP

() •• :t '.:.:jO

END;
VAH Facl..ll t~~,

F'P r~:;ona 1 ;

Mascot, Stuclent

BEGIN

,
:= 'Niklaus Wirth
Facult~.Address.NulTtber := 5;
Facl..llty.Address.Strpot := 'Clausiusstrdsse
,

Facl..llt~tNalTte

,
A

END,
The type Personal contains the field Address, which
is of a
record
type.
To
assign values to each of the fields of the record Address,
you must also specify the record Faculty in the assignment statement.

2.3.3

Set Types

A set is a collection of data items of the same scalar type, which
is
known
as
the
base
type
of
the
set.
Unlike arrays and records,
elements in the set cannot be accessed individually.
In
PASCAL,
you
use
a
set as an individual unit.
The type definition specifies the
values for each element in the set:
Fo rma t
TYPE identifier

SET OF base type

where:
identifier

specifies the type identifier for the set.

base type

specifies the data type.
Each element in the
set
must
be of
this data
type.
You can use the
identifier or definition of any scalar type except
a real TYPE.

A set can have up to 256 members, and the value of each member must be
between 0
and
255.
Therefore, real numbers or integers outside the
range 0 to 255 cannot be set elements.
After defining a base type, you can
type, for example:

declare

set

variables

SET OF (Racauet, Shoes, Balls, Boots,
Skis, F'oles, Goggles, Swimsuit);

2-22

that

PASCAL DATA TYPES
Sets are compatible if their base types are identical
for example:

equivalent,

TYPE Vitamins = SET OF (A, B1, B2, B6, B12, C, D, E, K),
Nutrients = Vitamins;
VAR

Watersoluble, Fatsoluble
Def:i. c i (o:o)folt : Nut J' i f.o)nts;

Vitamins,

The VAR section specifies three mutually compatible sets.
compatible base types are also compatible, for example:
VAR

Sets

wi th

Specials

ASCII : SET OF CHAR;
S pee i a 1. ~:; : SET (] F I! 1 • + 1 / I;

These two sets ~re compatible because the base
compatible with the ASCII character set.

type

Packing has no effect on set compatibility except when passing sets as
VAR parameters.
An
unpacked set is compatible with a packed set if
both sets meet the criteria above.
You can build set expressions by using the set operators described
in
Chapter 3.
Set operators allow you to specify set intersection,
difference, union, inclusion, and containment.
In addition,
you can
assign a
set expression
to a
set variable.
The base type of the
variable must include all members of the set to which
the expression
eval ua tes.
Example 1

TYPE Caps = SET OF CHAR,
VAR Vowel: Caps:=
ConsDnant : Caps

[/A/,/E/,/I/,/O/,/U'];

:= £: 'B' + + '[I', 'F' + + 'H',' J'. + 'N',
'P' •• 'T','V' •• 'Z'];

These declarations specify the set type Caps and
two
set variables,
Vowel
and Consonant.
The set Vowel is initialized with the set of
vowel characters as initial values.
The set Consonant is
initialized
with the set of consonants.
Example 2
VAR Ages

This example declares and initializes a set with an integer base type.

2-23

PASCAL
2.3.4

D~TA

TYPES

File Types

A file is a sequence of data components of the same type. The number
of components in a file is not fixed;
a file can be of any length.
The file type definition specifies the type of the file components.
Fo rma t
TYPE identifier

FILE OF component type

where:
component type

specifies the type of the components of the
file.
The component type can be any scalar
or structured type except a file type or an
array
or record
type containing a file
element or field.

The arithmetic, relational, Boolean, and assignment operators do not
work on file variables or structures containing file components.
For
example, you cannot assign one file variable to another file variable,
nor can you initialize a file variable.
Type compatibility for files applies only to file parameters.
Two
file parameters are compatible if their components are compatible and
if both are packed or neither is packed. You can pass a file only as
a VAR parameter.
PASCAL automatically creates a buffer variable for each file variable
you declare. The type of the buffer variable is the same as the type
of the file components. To denote the buffer variable, specify the
name of the associated file variable fo] lowed by a circumflex (~), for
example:
TYPE Scores = FILE OF INTEGER;
Scores;

VAR Math_Scores:

PASCAL creates ~ath Scores~ as an integer buffer variable associated
with the file Math Scores. The buffer variable takes on the value of
the file at the cur~ent file position.
The predeclared input and
output procedures move the file position, thus changing the value of
the buffer variable.
Example I
VAF, T ruthva 1. s

FIL.E OF' BOOLEAN;

This declaration specifies a file of Boolean values.
variable for this file is denoted by Truthva]s~.

The

buffer

Example 2
TYPE Names = PACKED ARRAY [1 •• 20] OF CHAR;

Data .... Fi Ie :::: FILE
Names;
RejecL.List, Wait_List:

VAH Accept .... L..ist"

Data_File;

This example defines the array type Names and the file type DataFile,
which contains a list of names. The VAR section specifies three file
variables of type DataFile, with
associated
buffer
variables
Accept_List~, Reject_List~, and Wait List~.

2-24

PASCAL DATA TYPES
Example 3
F I 1...[ OF

RECOf~D

TY'ial

INTEGER;

Datf:'~

RECOF~D

Month: (Jan, Feb, Mar, Apr, Mas, Jun,
Jul, Aug, Sep, Oct, Nov, Dec);
Da ~:1
Y€·~ar

1.. 3 l ~

: INTEGER

END;

Temp, Pressure : INTEGER;
Yield, Purits : REAl...
END;
The VAR Declaration specifies a file of records.
To access the fields
of
the
record
components,
you
specify
ResultsA.Trial,
ResultsA.Date.Month, and so on.

2.3.4.1
Internal and External Files - A file
that
is local
to a
program or subprogram
is called an internal file.
You can use an
internal file only within the scope of the program or subprogram
in
which it is declared.
The system retains an internal file only during
execution of the declaring program or subprogram.
After execution the
file
is no longer accessible.
The system creates a new file variable
with the same name the next time it executes the declaring unit.
The
contents of the old file are not available.
Internal files are not
specified
in the program heading.
Only internal
files can be
components of structured types.
An external file exists outside the scope of the program in which it
is declared.
An external file can be created by the current PASCAL
program, another PASCAL program, or a
program written
in another
language.
The system retains the contents of external file variables
after the execution of the program. You must specify the names of
external file variables in the program heading.
External files cannot
be part of a structured type.

2.3.4.2 Text Files - A text file is a file with components of type
CHAR.
PASCAL defines a
file
type called TEXT.
To declare a text
file, specify a variable of type TEXT, for example:

VAR Poem

TEXT;

The text file variable POEM is a file of characters.
Text files are
divided
into lines.
Each line ends with a line-separator character.
You cannot use this character directly,
but you can refer
to
it
indirectly through the predeclared procedures READLN and WRITELN and
the predeclared function EOLN.
The pred~clared file variables INPUT and OUTPUT are files of type
TEXT.
These files are the defaults for all the predeclared text file
procedures described in Chapter 7.
Note that TEXT is not equivalent
to FILE OF CHAR.
Example

VAR Guide, Manual:

TEXT;

This example declares the Variables Guide and Manual as text files.

2-25

PASCAL DATA TYPES
2.4

POINTER TYPES

Normally, variables have the same lifetime as the program
or
subprogram in which they are declared.
Program-level vari~bles are
allocated
in static storage,
and
subprogram-level variables are
allocated on the stack.
Some applications, however, require different
lifetimes or an unknown number of variables of a certain type.
PASCAL
allows you to use dynamic variables to fill these requirements.
Dynamic variables are dynamically allocated as needed during
program
execution.
Unlike other variables, dynamic variables are not named by
identifiers.
Insie~d,
you must refer
to
them
indirectJy
with
pointers.
A pointer type thus allows you to declare any number of pointer
variables to
refer
to dynamic variables of a specified type.
Each
pointer variable assumes as
its value
the ~ddress of a dynamic
variable.
The pointer type definition specifies the type of the dynamic variable
to which pointers of the pointer type refer.
Fo rma t
TYPE identifier

Abase-type identifier

where:
base-type identifier

indicates the type of
the
dynamic
variable
to which the pointer type
refers.
The base type can be any type.

Note the following example:
TYPE

MYrec

= RECORD
A ~ B d::
ENIH

Ptr_To_Myrec
VAR M :

INTEGER

= ~MYrec~

Pt T' _.. T D .... M!:j Y'PC Y

Variables of a pointer type point to variables of the base
type,
and
are said
to be bound to that type.
To indicate a pointer variable,
specify its name.
To indicate the dynamic variable to which a pointer
is bound,
specify the pointer name followed by a circumflex(A).
For
example, M is a pointer variable bound
to
records of type Myrec.
Specify MA to denote the record variable to which M points.

2-20

PASCAL DATA TYPES
Pointer type definitions are the only place in a PASCAL program where
you can use an identifier before you define it.
PASCAL allows you to
use the base type identifier in a pointer type definition
before
you
define the base type, for example:

TYPE Ptr_To_Movie ~ ~Movie;
Nam(,~ :::: F',:':JCKED ARri:AY [:t •• 20:3 OF CHAFi:;
Mov :i. (.:~ :::: Fi:FCOI:;:D
Title, Director : Name;
Yf:'~ a Y'

I NT E C! E I~: ,

Stars : FILE OF Name;
Next : ptr_To_Movie

END;
The TYPE section specifies the type identifier Movie
the type Movie.

before

defining

The value of a pointer is the storage address of the variable to which
it points.
Thus, in the example above, the value of the field Next is
a pointer to (or address of) a dynamic record variable of type Movie.
Pointers assume values at initialization, by assignment,
and
through
the
NEW procedure.
The value of a pointer can be any legal storage
address.
The value NIL indicates that the pointer does not currently
specify an address.
Thus, a NIL pointer does not point to a variable.
PASCAL allows you to define pointers to types
example:
TYPE

containing

files,

for

x:::: "'Y;
Y::~

RECORD
P : I NTEGEf;:;
Q : ARRAY[:ttt3] OF TEXT

END;
VAR M:X;

The pointer type X points to record type Y,
which contains a
file
component
in
field Q.
The files denoted by Q are never closed until
execution of
the
program
terminates,
unless you
use
the
CLOSE
procedure.
For example,
to
close
the
files
defined in the TYPE
section above, you must call
CLOSE with
the
parameters MA.Q[l],
A

M .Q(2], M".Q[31.

You can assign the constant NIL to a pointer as follows:

As a result of the assignment, the pointer variable M does
not
point
to
a variable.
NIL is the only value you can specify to initialize a
pointer.

2-27

PASCAL DATA TYPES
Example
:;;: {~ F< ~-;; A Y [1.. 30 ::I 0 F C H A F< ~
f' t r' ... T (J .... H :i. t ~:; :::: ,', Hit ~; ,
H i. t <:; ::: F~ [ C n P I:I

T Y PEN d IT! f.~

Title, Artistv Composer: Name;
I NT [ GER ~

We p k. ~:; .... n n .... C h art y N.... ~:) 0 :I. (J :
F :i. T' ~:; t .... V p r ~:; :i. 0 n : B 0 [) I... E (I N

END,
VA f< To F·' t f..1 n

A R F.: A Y [1.. 1 () J

0 F F' t J" .... To .... Hit ':; ,

INTEGEF<;

BEGIN
F (] I:~

:r : ::: 1. TO :I. 0 II 0
ToptenlIJ := NII...~

END,

This example defines the record type Hits to which pointers of type
Ptr To Hits refer.
The array variable Topten has elements of the
pointer type Ptr To Hits.
E~ch element of the array is
assigned the
constant value NIL.
The array Topten could be used in creating a Jist
of ten records of type Hits.

2.5

PACKED STRUCTURED TYPES

You can pack any of the structured types by specifying PACKED in the
type or variable declaration.
Packed data items are stored as densely
as possible.
Fo rma t
PACKED type definition
where:
type definition

defines an array, record, set, or file type.

You can initialize all packed structures in the VALUE or VAR section
in the same way that you initialize an unpacked structure of the same
type.
In general, packed data items require less storage space than
unpacked data
items of the same type. However, execution is usually
slower with packed data items.
In PASCAL, a packed array of characters specifies a string variable.
Example 1
TYPE

I:<an~:jes

.::: PACI\ED RECORD
Wo rei : 0 •• 6:::j~)3:::j;
B~t(~
: O •• 32/(,)/;
Bit : BOOLEAN
END;

This example defines a record type with three fields, each of which is
packed as densely as possible.

2-28

PASCAL DATA TYPES
Example 2
VAR

: PACKED ARRAY [1 •• 25J OF 2500 •• 50000;
I NTEGEF: i;

Cit~_Census

BEGIN
FOI:<

I: : :~: :L TO 2~:.:j DO
C 1. t ~~ .... C(.:~ n sus I:: I] ! :=:

();

END;
This example declares the variable City Census as a
25-element array
of integer values in the subrange from ~500 through 50000.
A value of
a is assigned to each element of the array.

2.6

TYPE COMPATIBILITY

Type compatibility rules determine the operations and assignments that
you can perform with data items of different types.
Two scalar types
are compatible if their type identifiers are declared
equivalent in
the TYPE section.
In addition, a subrange type is compatible with its
base type, and two subranges of the same base type (or equivalent base
types) are compatible.
For structured and
pointer
types,
PASCAL
enforces
structural
compatibility.
Two
structured (that is, arrays, records, files, and
sets)
or pointer
types are compatible
if their
structures are
identical.
The way PASCAL determines structural
compatibility depends on the
record
types
involved.
For
instance,
the
requirements
for
compatibility differ from those for array compatibility.
PASCAL uses compatibility rules in the fo]lowing three contexts:
1.

Expression compatibility

Assignment compatibility

Formal and actual parameter compatibility

Expression compatibility determines the types of operands you can
in an expression.
See Chapter 3 for information on expressions.

use

Assignment compatibility determines the types of values you can assign
to variables of each type.
Assignment compatibility rules apply to
value initializations, assignment statements, and value parameters.
Assignment compatibility is described with the assignment statement in
Section 5.2.
Formal and actual parameter compatibility determines the types of data
you can pass in a parameter list. Value parameters follow the rules
for assignment compatibility.
Variable parameters follow somewhat
different rules.
Value and variable parameters are described in
Chapter 6.

2-29

CHAPTER 3
EXPRESSIONS

An expression is a
symbol
or group of symbols
that PASCAL can
evaluate.
These symbols can be constants, variables, or functions, or
any combination of constants, variables, and functions, combined with
operators.
The simplest expression is a single variable or constant.
This chapter lists the various operators that PASCAL provides along
with
the
rules
for forming arithmetic, relational, logical, and set
expressions.

3.1

OPERATORS

PASCAL provides the following types of operators:

3.1.1

•

Arithmetic operators (such as +, -

•

Relational operators (such as <, >, =)

•

Logical operators (such as AND, OR, NOT)

•

Set operators (such as IN)

ArithmetIc Expressions

An arithmetic expression usually provides a formula for calculating
a
value.
To
construct an arithmetic expression, you combine numeric
constants, variables, and function identifiers with one or more of the
operators from Table 3-1.

3-1

EXPRESSIONS
Table 3-1:
Operator

Arithmetic Operators
Example

Meaning
--------------------------

A+B

Add l\ and B

A-B

Subtract B from A

A*B

Multiply A by B

A**B

Ra ise A to the power of B

AlB

Divide A by B

DIV

A DIV B

Divide

MOD

A MOD B

Produce
the remainder after dividing A by
B; B must be greater than 0

by B and truncate the result

The addition, subtraction, muJtiplication, and exponentiation
(+,
,
*,
and **)
operators work on both integer and real values.
They
produce real results when applied to real values and
integer
results
when applied to integer values.
If the expression contains values of
both types, the result is a real number.
The only exception to
these
rules concerns exponentiation.
PASCAL defines
the
results of an
integer raised to the power of a negative integer as follows:
Base

Result

Exponen t

Negative or 0

Negative

Error
1

-1

Negative and odd

-1

Negative and even

Negative

Any o the r
integer

For example, the expression 1**(-3) equals 1;
(-1)**(-4) equals 1;
and 3**(-3) equals o.

(-1)**{-3)

equals

-1;

The division (I) operator can be used on both real and integer values,
but always produces a real result.
Use of the division (I) operator
can therefore cause errors
in precision
in expressions
involving
integers.
The DIV, MOD, and REM operators apply to integer values only.
DIV divides one integer by another, producing an integer result.
DIV
truncates the
result;
that is, it drops any fraction.
It does not
round the result.
For example, the expression 23 DIV 12 equals 1, and
(-5) DIV 3 equals -1.
The MOD and REM operators return
the
remainder after dividing
one
operand
by another.
Both operators can be used only with integer
values and always produce integer results.
The MOD operator can be used only when the divisor is greater than 0;
it always
returns a positive result.
For example, the expression 5
MOD 3 (5 modulo 3) returns a value of 2, and
(-5)
MOD 3
returns a
value of 1.
3-2

EXPRESSIONS
The REM operator can be used on integers of all sizes and retains
the
sign of
the dividend.
For example, the expression 5 REM 3 returns a
value of 2, the expression (-5) REM 3 returns a value of -2;
and
the
ex pre s s ion 5 REM ( - 3) ret urn s a val u E~ 0 f 2.
In arithmetic expressions, PASCAL allows you
to mix
integers,
real
numbers
(single and
double precision), and integer subranges.
When
you assign the value of an expression to a variable, you must
ensure
that
the types of the variable and the expression are compatible.
In
general, you can assign an integer
expression
to
a
real
variable.
However, you cannot assign a real expression to an integer variable.
Table 3-2 Jists the type of the result for all
of arithmetic operators and operands.
Table 3-2:

possible

combinations

Result Types for Arithmetic Expressions

Operator
(Opera t ion)

Type of First
Operand

Type of Second
Operand

Type of
Result

INTEGER

( e x po n e n t i a t ion)

INTEGER, REAL

REAL, DOUBLE

REAL

DOUBLE

INTEGER, REAL,
DOUBLE

DOUBLE

INTEGER

(mul tipl ication}

INTEGER

REAL

REJ~L

INTEGER, REAL

REAL

DOUBLE

INTEGER, REAL,
DOUBLE

DOUBLE
DOUBLE

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

RE)\L,

INTEGER

DOUBLE

RE1\L,

INTEGER

REAL,

(division)

DOUBLE
RE1\'L,

DIV, MOD, REM

INTEGER

3-3

INTEGER

REAL

INTEGER, REAL,
DOUBLE

DOUBLE

DOUBLe

INTEGER

EXPRESSIONS

TabJe 1-2: Result Types for Arithmetic Expressions (Cont.)
Type of F'i rst
Operand

Type of Second
Operand

Type of
Result

+,-

INTEGER

(addition,
subt.raction)

INTEGER

REAL

REAL, INTEGER

REAL

DOUBLE

REAL, DOUBLE,
INTEGER

DOUBLE

Operator
(Opera t ion)
(division with
truncation,
modulus,
and remainder)

REAL,

3.1.2

INTEGER

Relational Expressions

A relational expression or condition tests the
relationship between
two expressions.
A relational expression consists of two scalar or
string variables or arithmetic expressions, separated by one of the
relational operators listed in Table 3-3.
Table 3-3:
Operator

Relational Operators
Example

Meaning

A = B

A is equal to B

A <> B

A is not equal to B

A > B

A is greater than B

A >= B

A is greater than or equal to B

A < B

A is less than B

A <= B

A is less than or equal to B

Note that the two characters in each of the <>, >=, and <= operators
must appear in the specified order and cannot be separated by a space.
PASCAL produces a
Boolean
result when
it evaluates a
relational
expression.
Every relational expression therefore evaluates to TRUE
or FALSE.
For example, the condition 2 < 3
is always TRUE;
the
condition 2 > 3 is always FALSE.

3-4

EXPRES~: IONS

3.1.3

Logical Expressions

Logical
expressions
test
the
truth value
of
combinations
of
conditions.
A logical expression consists of two or more expressions
that have Boolean results, separated by one of the
logicnl
operators
in Table 3-4.
Table 3-4:

Logical Operators

---------------------------------------Operator

Example

Result

AND

A AND B

TRUE if both A and B are TRUE

A OR B

TRUE
if either A or B is TRUE, or if both
are TRUE

----------,------------

NOT
NOT A
TRUE if A is FALSE, and FALSE if A is TRUE
--------------------------,----The AND and OR operators combine two conditions
to
form
a
compound
condition.
The NOT operator reverses the truth value of a condition,
so that if A is TRUE, then NOT A is FALSE.
As with relational expressions, the result of a logical expression
is
a
Boolean value.
Note that the entire logical expression is always
evaluated, even if the expression value could be
uniquely determined
from only a part of the expression.

3.1.4

Set Expressions

You can use
constants.
Table 3-5:

the

operators

Table

3-5

with

set

variables

and

Set Operators

Operator

Example

Meaning

A+B

Union of sets A and B

A*B

Intersection of sets A and B

A-B

Set of those elements of A that are not
also in B

A=B

Set A is equal to set B

A<>B

Set A is not equal to set B

A<=B

Set A is a subset of set B

A>=B

Set B is a subset of set A

IN B

A is an element of set B

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

3-5>

EXPRESSIONS
The set operators (+, *, -, =, <>, <=, and >=) require both operands
to be set values.
The TN operator, however, requires a set expression
as its second operand and a scalar expression of the
associated
base
type as its first operand, for example:

The value of this expression is TRUE,
which is a member of the set rl •. 10].

3.1.5

because

2*3

evaluates

Precedence of Operators

The operators in an expression establish the order
in which PASCAL
evaluates
the expression.
Table 1-6 lists the order of precedence of
the operators, from highest to lowest.
Table 3-h:

Precedence of Operators

Operators

Precedence

NOT

Highest

I, DIV, MOD, REM, AND

OR
<>, <, <=, >, >=,

Lowest

PASCAL evaluates operators of equal precedence (such as + and -)
from
left
to
right.
You must use parentheses for correct evaluation when
you combine relational operators, for example:

A (= X AND B (= Y
Without parentheses, PASCAL attempts to evaluate
this expression as
A<=(X AND B)<=Y and generates an
error.
The expression needs
parentheses, as follows:

X) AND (8

To evaluate the rewritten expression, PASCAL compares the truth values
of the two relational expressions.

3-n

EXPRESSIONS
You can use parentheses in any expression to force a particular
of evaluation, for example:
Expression:

* 5 DIV 2 ... 4

* 5 DIV (2 - 4)

order

Evaluates to:
16
····20

PASCAL evaluates the first expression according to
the normal
rules
for
precedence.
First,
it multiplies 8 by 5 and divides the result
(40) by 2.
Then, it subtracts 4 to get 16.
The parentheses in the
second
expression,
however,
force
PASCAL to subtract before
multiplying or dividing.
Hence, it subtracts 4 from 2,
getting -2.
Then, it divides -2 into 40, with -20 as the result.
Parentheses can also help to clarify an expression.
could write the first example as follows:
«8

For instance, you

* 5) DIV 2) - 4

The parentheses eliminate any confusion about how the expression is to
be evaluated.

3.2

SCOPE OF IDENTIFIERS

The scope of an identifier is the part of the program
in which you
have access
to
the identifier.
In a PASCAL program, the scope of a
constant, type, variable, or subprogram identifier
is the block
in
which the identifier is declared.
Figure 3-1 illustrates the scope of
identifiers declared at various levels.
Declarations in the main program block specify global
identifiers,
which can be accessed
in the main program and
in all
nested
subprograms.
For example,
A and B in Figure 3-1
are
global
identifiers.
They can be accessed from any level in the program.
Declarations in subprogram blocks specify local identifiers.
You can
use a local identifier in the subprogram that contains its declaration
and in all its nested subprograms.
For example, the identifiers C and
D are local to procedure Le~el lA and its nested subprograms Level 2A
and Level 3A.
You can USE~ C and D i 11 any 0 f
these
subprog rams, but
not in the main program or in the subprograms Level_lB, Level_2B, and
Level 2C.

3-7

EXPRESSIONS

PROGRAM Level_.O (IN PUT, OUTPUT);
VAR A,B : ] NTEGER;

···
PROCEDURE Level_1A (Z,

y);

VAR C,D : INTEGER;

··
· N L-e-v-e-=I-_-=2:--A:--7'(X:-7"':"')---=---=I:-7N":'::l:::-:'[:::-.:C:::-'E=::=R-:";- . FUNCTIO
VAR E : REAL;

~F~'R~'O:-C~~'E~-I~IU~R~E--L~e~v-e-l-_-"3~A--<-W-)-;---------

VAR F : REAL;

END;

·
·
· <
END;

<*end procedureLevel_3A*)

------------------------------,----

·•
· <*end procedure Level_1A*)
END;
.
···

PROCEDURE Level_1B <V, U, T);
INTEGER;
VAR G

F'R OCEDURE Level_2B (S, R, Q);
VAR H : REAL;

··
·
ENI

F'R OCEDURE Level_2C (P, 0);
VAR B : BOOLEAN;
j
:
CHAR;

·
·
· I; (*end rocedu re Leve 1._2C* >
ENI
P

·
·
· <*en d procedure Level_1B*>
END;

··
· <*end prograrr
END.
Figure 3-1:

Scope of Identifiers

3-8

EXPRESSIONS
Similarly, local identifiers declared in Level 1B are accessible to
LevellB,
Level ~~B, and Level2C, but not to Level lA, Level ?A,
Level-3A, or the nain program.
In general, once you define an identifier,
it retRins its meaning
within the block containing
its declaration.
You can, however,
redefine an identifier in a subprogram at a lower level.
If you do
so,
the identifier assumes its new meaning only within the scope of
the redefining block. Outside this block, the identifier keeps its
original
meRning.
For example, B is declared at program level and
redefined in Level 2C. Within the scope of Level 2C, B denotes a
Boolean variable. ·Everywhere else in the program,-however, B denotes
Rn integer.
The identifiers accessible to each routine in Figure
below.

3-1

Routine

Variables

Main p:rog ram

A, B ( :i. n t f::' ~.:.! f~ T' )
A, B ( i ntt~~tp T')
r_. , 1:1
Dy E
A, B ( i nte5.!e T') y C,
Ay B ( i ntf.~~J(·? T' ) C, [I, [y
A, B ( :i. ntf.~~~e T' ) , G

----l..evf~ll .... :LA

L..f?vf?l .... :?A
Level.. .. 3A
LeVE' 1 __ :L B
L f? V t~ 1 .. _2 B
Level .... 2C

,
,
,
A, B ( integer) , (3y H
A, B (Boolean) , ,J

3-9

are

r..

listed

CHAPTEH 4
PROGRAM HEADING AND DECLARATION SECTION

The first two parts of a PASCAL program are the program heading and
the declaration section.
The program heading specifies the program
name and the input and output files.
The declaration section can contain the following sections:
•

LABEL

declares labels for use by the GOTO
statement

•

CONST

defines
identifiers
constant values

•

TYPE

defines user-defined,
and pointer types

•

VAR

declares and optionally initializes
variabJes of all types

•

VALUE

initializes variables

•

PROCEDURE and FUNCTION

declare subprograms

represent
structured,

Your program need not include all these
sections,
but the sections
that are present must appear in the order listed above.
Although you
can specify many labels, constants, types, variables, and subprograms,
each section can appear only once in each declaration section.
This chapter describes the program heading
(Section 4.1),
label
declarations
(Section 4.2),
and constant defini tions (Section 4.3) •
It also outlines type definitions (SE!ction 4.4), variable declarations
( Sec t ion 4 • 5), and val u e dec I a rat ion s
( Sec t ion 4 • 6) •
Ch apt e r 6
describes the use of procedures and functions in detail.

4-1

PROGRAM HEADING AND
4.1

DECL~RATION

SECTION

THE PROGRAM HEADING

The program heading begins the PASCAL program.
It gives the program a
name and lists the external file variables the program uses.
Format
[fOVERLAID1]

PROGRAM program name

[(filename [,filename ••• ] )]

[OVERLAIDl

specifies that
variables with
Section G.8.

program name

specifies the name of the program;
six characters are significant.

filename

specifies the
identifier associated
external fiJe that the program uses.

where:
the program can share
global
separately compiled modules. See
only the first
with

The program name appears only in the heading and has no other purpose
within the program. The program name cannot be redefined at program
level.
The file names listed in the program heading correspond to the
external
files that the program uses. The heading must include the
names of all the external file variables. The predeclared text file
variables INPUT and OUTPUT, by default, refer to your terminal (in
interactive mode) or the batch input and log files
(in batch mode).
You must declare file variables for all other external files in the
main program declaration section, and specify those variables in the
program heading. Refer to Chapter 7 for more information on files.
Example I
PROGRAM Test1;
The program heading names the program Testl, but omits the file
variable list. This program does not use the terminal or any external
file.
Example 2
PROGRAM Sauares (INPUT, OUTPUT);
The program heading names the program Squares
predeclared file variables INPUT and OUTPUT.

and

specifies

the

The program heading names the program Payroll and
external file variables Employee, Salary, and Output.

specifies

the

Example 3
PROGRAM

Pa~roll

(EmploYee,

Salar~,

4-2

Output);

PROGRAM HEADING AND DECLARATION SECTION
4.2

LABEL DECLARATIONS

A label makes a statement accessible from a GOTO statement.
The label
section must list all
the
labels
in the corresponding executable
section.
Format
LABEL label

[,label ... ]

where:
label

specifies an unsigned integer.
When you declare more
than one label,
you can specify the labels in any
order.

The label is an unsigned integer.
Labels can be listed in any order
if more than one label is defined.
A label can precede any statement
in the program, but can be accessed only by a GOTO statement.
Within
the
program,
a
colon
(:)
must be placed between the label and the
statement.
The scope of a label is the block in which it is declared.
Therefore,
you can transfer control from one program unit to another program unit
in which the former is nested, for example:

PROGRAM Trial (INPUT,OUTPUT);
LABEL

7~;

PROCEDURE Max;
LABEL 50;

BEGIN

WRITELN ('Testins fairness of tosses');

GO TO 75;
END;

(*end of procedure Max*)

WRITELN ('Not fair!

BEGIN

A weighted coin!');

END.
The GOTO statement in the procedure Max transfers control to the main
program statement that has the label 75.
However, you cannot use a
GOTO statement in the main program to
transfer control
into
the
procedure at label 50.
Example
LABEL 0, 6656, 778, 4352;
The label section specifies four labels:
0,
6656,
778,
and
Note that the labels need not be specified in numeric order.
4-3

4352.

PROGRAM HEADING AND DECLARATION SECTION
4.3

CONSTANT DEFINITIONS

The constant section defines identifiers to represent constant values.
Fo rma t
CONST constant name

[constant name

value;

where:
constant name

specifies the identifier to be
name of the constant.

used

the

value

specifies an
jnteger, a
real number,
a
string, a Boo]ean va]ue, or the name of
another constant that is already defined.
Note that the value assigned to a constant
identifier cannot be an expression.
String
values must be enclosed in apostrophes.

The use of constant identifiers generally makes a program easier to
read, understand, and modify.
If you need to change the value of a
constant, you can simply modify the CONST declaration instead of
changing each occurrence of the value in the program. This capability
makes programs simpler to maintain and easier to transport.
Example
CON~)T

Ra i n
Y to" a T'

~::
~~

TRUE,
20 O:L ,

Pi ::: 3 .l4:1.!:j92?'
C () ITlIT! a:;::

','

= 'United States';
Citizenship = Country;

Countr~

This CONST section defines six constants.
The identifier Rain is
equal
to the Boolean value TRUE.
The identifier Year represents an
integer, and Pi represents the real number 3.1415927. The identifier
Comma represents a character, and the identifier Country represents a
string. Characters and strings must be enclosed in apostrophes in the
CONST section.
The identifier Citizenship represents the symbolic
constant Country and thus repr~sents a character string.
Note that,
since Citizenship represents a symbolic value and not a string,
apostrophes are not used.

4-4

PROGRAM HEADING AND DECLARATION SECTION
4.4

TYPE DEFINITIONS

The type definition introduces the name and the set of values
for
a
type.
Chapter 2
describes data types and includes examples of type
definitions.
Fo rma t
TYPE type identifier
[ type identifier

type defi.nition;
type defi.nition; ••.

]

where:
type identifier

specifies the identifier to be
name of the type.

type definition

defines a
type.
Figure 4-1.

The

types

used
are

the

shown

Note that you can use the identifier for a previously defined type
in
place of
the
type definition for a new type.
In addition, you can
define packed types for arrays, records, sets, and files, as described
in Section 2.5.

scalar

predefine
scalar

[

stmctu ••d

user-defined
scalar

J
INTEGER

enumerated

ARRAY

REAL

sub range

RECORD

BOOLEAt

SET

CHAR

FILE

DOUBLE
MRS-3147-83

Figure 4-1:

PASCAL Data Types

4-5

PROGRAM HEADING AND DECLARATION SECTION
4.5

VARIABLE DECLARATIONS

The variable declaration creates a variable and associates
an
identifier and a type with the variable. Optionally, the variable
declaration can be used to assign an initial value to a variable.
Chapter 2 describes data types and shows how to declare and initialize
variables of each type.
Fo rma t
VAR variable name [,variable name .•. ]

[variable name [,variable name ... ]

: type [ : = va 1 ue n ;
: type [ .:= value] ; ... ]

where:
variable name

specifies the identifier to be used as the name of
the variable.

type

names or defines a type. The type can be
the types shown in Figure 4-1.

value

specifies the initial value associated with the
identifier.
The value must be of the same data
type as the identifier.

one

You can also declare packed array, record, set, and file variables, as
described
in Section 2.5. Note, however, that you cannot initialize
file variables.

4-0

PROGRAM HEADING AND DECLARATION SECTION
4.6

VALUE DECLARATIONS

The value section initializes variables that are declared in the main
program declaration section.
You can initialize scalar, array,
record, and set variables with constants of the same type as the
variable's type.
The description below presents
general
information
on
value
initializations. The exact format of the value initialization depen~s
on the type of variable being initialized.
For detailed
formats and
examples, refer to the section in Chapter 2 that describes the type of
variable you need to initialize.
Fo rma t
VALUE variable name "= value;
[variable name ,= value;

. '.. ]

where:
variable name

names the variable to be initialized.
specify a list of variable names.

You

cannot

value

specifies a constant of the same type as the
variable, or specifies a constructor for an array
or record variable.

You must specify a value of the correct type for each variable being
initialized.
You must not specify an expression.
Scalar variables
require scalar constants, and set variables require set constants.
For arrays and record variables, you specify the value to be assigned
to each element or field in a parenthesized list called a constructor.
An array or record constructor must contain one constant value of the
appropriate type for each component of the structure.
program
The value
initialization can appear only in the main
declaration section.
You cannot use a value section in procedures,
functions, or modules.

4-7

CHAPTER 5
P~SCAL

STATEMENTS

PASCAL provides several statements to perform the actions within
the
program.
Any of
these
statements can appear
anywhere
in
the
executable part of a PASCAL program, procedure, or function.
PASCAL
also
includes
the
compound
statement,
which allows
you to group
statements.
This chapter presents reference information on each of the statements,
organized as foJlows:
comp~und

•

The

statement

•

The assignment statement

•

Conditional statements:
IF-THEN
IF-THEN-ELSE
CASE

•

Repetitive statements:
FOR
REP~AT

WHILE

• The WITH statement
• The GOTO statement
• The procedure call
PASCAL includes simple and compound statements.
A simple statement
can be executed and is complete in itself;
that is, it is not made up
of other
statements.
The simple statements are
the assignment
statement, the GOTO statement, and the procedure call.
A compound statement is an arrangement of simple statements that
executes sequentially.
You can use a compound statement anywhere in
the program that a simple statement is allowed.
This manual uses
the
term statement to mean either a simple or a compound statement.

5-1

PASCAL STATEMENTS
Simple and compound statements can also be used
in structured
statements.
A structured stAtement is a group of statements that can
be executed either in sequence,
conditionally, or repeatedly.
The
structured
statements are
the conditional,
repetitive,
and WITH
statements.
A compound stAtement can also be considered a
type of
structured statement.

5.1

THE COMPOUND STATEMENT

The compound statement allows you to group
sequential execution as n singJe statement.

PASCAL

statements

for

Format
BEGIN
statementl

;statement?, ...

statement n;

END;
where:
statement

denotes a simple or compound statement.

You can create a compound statement using any combination of PASCAL
statements,
including other compound
statements.
You must use
semicolons to separate
the
statements that make
up the compound
statement;
however,
no semicolon
is
required
between the last
statement and the END delimiter.
PASCAL treats
the entire compound
statement as a simple statement everywhere in the program.
Examples
of compound statements appear throughout this chapter.

5.2

THE ASSIGNMENT STATEMENT

The assignment statement assigns a value to a variable.
Format
variable name

0= expression;

where:
variable name

specifies the name of a
variable of any type
(except a file).
It could be an array element, a
file buffer variable, a function, or a field of a
record.

expression

specifies a value,
variable
name,
function
reference,
Boolean expression, set expression, or
arithmetic expression.

Note that the assignment operator is := in PASCAL.
this operator with the equal sign (=) operator.

not

The expression on the right of the operator establishes the
be assigned to the variable on the left of the operator.

5-2

confuse
value

PASCAL STATEMENTS
You can use the assignment statement to assign a value to
a
function
identifier or
to a variable of any type except a file.
The variable
and the expression must be of compatibJe types, with the following two
exceptions:
•

You can assign an integer expression to a real variable.

•

You can assign an integer or single-precision real expression
to a double-precision variable.

For structured types,
assignments.

PASCAL

enforces

structural

compatibility

Example 1

x : :: 1

The variable X is assigned the value 1.
Example 2

Temp

:= Celsius(Fahrenheit);

The value returned by the function Celsius is assigned to Temp.
Exampl e 3
T t::::A<B,
The value of the Boolean expression A < B is assigned to T.
Exampl e 4
VoweL.. Set !:::: [JA J ,

JE/,

JJ',

JO·'"

JU J ];

The base
The set Vowel Set is assigned the string constants shown.
I I "
'0', and
type of VoweT Set must include the characters 'A', , E "

'U' .
Example 5
M~~_ .. AT'T'a!:~[lJ

::::::

M~L ... AT'r·a!:J["7]

+ yf)l..lT' .... AT'ra!.~[14];

The first element of My Array is assigned
the
sum of
the
element of My_Array and-the fourteenth element of Your_Array.

seventh

Exampl e '5

The value of each element of the array Your Array is assigned
corresponding element of the array My_Array-:-

the

Exampl e 7

Assume
that Awardrec
and
New Winner are
record
variables
of
assignment-compatible
types.
~his example assigns the value of each
field of New Winner to the corresponding field of Awardrec.

5-3

PASCAL STATEMENTS
Example q
Ase~5

::::: Asps····[ 10+".7J'

Assume that the base type of the set variable Ages
is the
integer
subrange 0 .. 255.
This exampJe assigns the value of the set expression
Ages- rlO+7] to the variable Ages.

5.3

CONDITIONAL STATEMENTS

A conditional statement selects a statement for execution depending on
the value of an expression.
PASCAL provides three conditional
statements:
•

IF-THEN statement

•

IF-THEN-ELSE statement

•

CASE statement

These are described in the following sections.

5.3.1

The IF-THEN Statement

The IF-THEN statement causes the conditional execution of a statement.
Fo rma t
IF expression THEN statement;
where:
expression

specifies a Boolean expression.

statement

indicates a simple or compound statement.

The statement is executed only if the value of the expression is TRUE.
If the value of the expression is FALSE, program control passes to the
statement following the IF-THEN statement.
The THEN clause can specify a compound statement.
However, note that,
if you use
the compound statement, you must not place a semicolon
between the words THEN and BEGIN.
The example below shows a semicolon
immediately following the word THEN:

IF Da~ = Thurs THEN,
BEGIN

(* misPlaced semicolon *)

statement

END;
As a result of the misplaced semicolon, the empty statement becomes
the object of the THEN clause.
In
this example,
the compound
statement following the IF-THEN statement is executed
regardless of
the value of Day.

5-4

PASCAL STATEMENTS
Example 1
IF

«X*37/Con~:;tant)

Answer

+ Factor) :> lO()O.O THEN

:= Answer -

Factor;

If the value of the arithmetic expression is greater
new value is assigned to the variable Answer.

than

1000.0,

the value of

Example 2

IF (A :> B) AND ( B ~ C) THEN
D ::::: A···· C;

If both relational

- c.

expressions are true, n is

assigne~

Exampl e 3

IF (Name
'SMITH') AND (INITIAL
BEGIN
Count

I ,.J I )

THEN

:= Count + 1;

SlTIi thadd[ Count]l

! :::: Add re~:;~:;;

WFnTELN ('..} SMITH NO.

I,Count,

L.IVE~)

This example counts the number of J SMITHs, prints each
address, and stores it in an array.

AddT'f.-~ss)

one's

street

Exampl e 4

IF Day = Thurs THEN
FOR I := 1 TO MaxEffip DO
Pay[IJ

:= Salary[IJ * (l-Tax_Rate-FICA);

If the current value of the variable Day is Thurs,
the
FOR loop
is
executed,
and values for Pay(I] are computed.
If the value of Day is
not Thurs, the FOR loop is not execute~;
and program control
passes
to the statement following the end of the loop.

5.3.2

The IF-THEN-ELSE Statement

The IF-THEN-ELSE statement is an extension of the
IF-THEN statement
that
includes an alternative statement, the ELSE clause.
The ELSE
clause is executed if the test condition is false.
Fo rma t
IF expression THEN statementl ELSE statement2
where:
expression

specifies a Boolean expression.

sta temen tl

sta tement
denotes
the
exprt=ssion is true.

executed

the

statement2

denotes
the
statement
expression is false.

executed

the

5-5

PASCAL STATEMENTS
The objects of the THEN and ELSE clauses can be any simple or compound
statement,
including another IF-THEN or IF-THEN-ELSE statement.
The
ELSE clause always modjfies the closest IF-THEN statement.
IF

A~~i

THEN

IF B<>l THEN

C:~l

E L ~; ED: c= 1 ;

By definition, PASCAL interprets this
BEGIN and END delimiters, as folJows:
IF A:::J. THEN
BEGIN
IF B<>l THEN
EI... SF It: :::: 1
END;

statement

included

C::"~i

The variable D is assigned the value 1 if both A and B are equal to 1.
Example 1
IF Ih Sf.~aSf:~ THEN
WRITELN ('This person is sick.')
ELSE WRITELN ('This person is health~.');

This example prints a different line of text depending on the value of
the
Boolean variable Disease.
Note
that Disease
is a
Boolean
expression, so you need not specify Disease = TRUE.
Exampl e 2
IF Balance < 0.0 THEN
BEGIN
WRITELN ('Overdrawn b~ ',ABS(Balance»;
WRITELN ('Loan of ',Loan,' at ',Rate,
, % automaticall~ deposited');
Balance := Balance t Loan;
BILL_AMT := Loan
(ltRate)
END
ELSE WRITELN ('No Loan issued this month ');
WRITELN ('Balance is ',Balance);

If the value of Balance
is negative,
the compound
statement is
executed.
The compound
statement prints two lines of notification,
adds Loan to Balance, and computes the amount of
the bill
for
the
loan.
A zero or positive Balance results in a message stating that no
loan was issued.
The WRITELN procedure that prints the final
Balance
is independent of the conditional statement and is always executed.

5.3.3

The CASE Statement

The CASE statement causes one of several statements
depending on the value of a scalar expression.
Format
CASE case selector OF
case-label l i s t :
statement
[jcase-label list
statement ••• D
[OTHERWISE statementi ••• D
END;

5-6

executed,

PASCAL STATEMENTS
where:
case selector

specifies an expression that evaluates to any
scal~r type except a real type.

case-label list

specifies one or more constants of
the
same
type as
the
case
selector,
separated
by
commas.

Each case-label list is associated with a
statement that may be
executed.
The list contains the value of the case-selector expression
for which the system executes the associated
statement.
You can
specify the case labels in any order.
However, the difference between
the largest and smallest labels must not exceed 1000.
Each case label
can appear only once within a CASE statement, but can appear in other
CASE statements.
At run time, the system evaluates
the case-selector expression and
chooses which statement to execute.
If the value of the case-selector
expression does not appear in any case-label list, the system executes
the statement in the OTHERWISE clause.
If the value of the case-selector expression does not match one of the
case labels and you omit the OTHERWISE clause, the status of the CHECK
run-time option determines the action that the system takes.
Refer to
Section 8.4.3
for run-time options.
If CHECK is enabled, the system
prints an error message and terminates execution.
If
CHECK
is not
enabled,
execution continues with
the statement following the CASE
statement.
Example 1
CASE Ase OF
5,6
IF Birth_Month> Se? THEN Grade l= 1 ELSE Grade != 0;
7
BEGIN
Grade := 2;
ReadinS_Skill := TRUE
END;
8
Grade:= 3
END~

At run time,
the
system evaluates Age and
executes one of
the
statements.
If Age
is not equal
to 5, 6, 7, or 8, and the CHECK
option is enabled, an error occurs and execution is terminated.
Example 2
CASE Ase OF
5,6
IF Birth_Month> Be? THEN Grade != 1 ELSE Grade
7
BEGIN
Grade != 2;
ReadinS_Skill := TRUE
END;
8 ! Grade := 3
OTHERWISE Grade := 0

:= 0;

An OTHERWISE clause is added in this example.
If the value of Age
not 5, 6, 7, or 8, the value a is assigned to the variable Grade.

5-7

PASCAL STATEMENTS
Example 1

CASE

Alphabetic OF
'A','E','I'p'O','U' :

END;

Alpha_Fla~

:= Vowel~

Alpha .... Fla~:,~ != SometimE-'s
OTHERWISE Alpha_Flag := Consonant
"~(I

This example assigns a value to Alpha Flag depending on the
the character variable Alphabetic.

5.4

value

REPETITIVE STATEMENTS

Repetitive statements specify loops, that is, the repetitive execution
of
one
or more
statements.
PASCAL provides three
repetitive
sta tements:
•

FOR statement

•

REPEAT statement

•

WHILE statement

These are described in the following sections.

5.4.1

The FOR Statement

The FOR statement specifies the repetitive execution of a
statement
based
on
the value of an automatically incremented or decremented
control variable.
Forma t
FOR control variable

-= initial value {TO

} final value DO statement;

DOWNTO
where:
control variable

specifies the name of a
simple variable
any scalar type except a real type.

initial value

specifies an expression of the same
the control variable.

type

final value

specifies an expression of the same
the control variable.

type

The control variable, the initial value, and the final value must all
be of the same scalar
type,
but cannot belong to one of the real
types.
The repeated statement cannot change the value of the control
variable.
The control variable must be a simple variable;
that is,
it cannot be an element of an array, a field of a record,
the object
of a pointer reference, or a file buffer variable.
At run time, completion tests are performed before
the
statement is
executed.
In
the TO form, if the value of the control variable is
less than or equal to the final value, the loop is executed,
and
the
control variable
is
incremented.
When
the value of the control
variable is greater than the final value, execution of the
loop is
complete.

5-8

PASCAL STATEMENTS
In the DOWNTO form, if the value of the control variable is
than or equal
to the final
value,
the loop is executed,
control variable is decremented.
When the value of the
variable is less than the final value,
execution of the
complete.
Because completion tests are performed before the
executed, some loops are never executed, for example:

greater
and the
control
loop is

statement

FOR Control := N TO N + Q DO
Wpek r: N] :::~ We (::)k [N] + Newpa~:I'
If the value of N + Q is less than the value of N (that is,
negative), the loop is never execut,ed.

When incrementing and decrementing the control variable, PASCAL uses
units of the appropriate type.
For numeric values,
it adds or
subtracts one upon each iteration.
For values of other
types,
the
control variable takes on each successive value of the type.
For
example, a control variable of type
'A' •. '2'
is incremented
(or
decremented) by one character each time the loop is executed.
If the FOR loop terminates normally, that is, if the loop exits upon
completion and not because of a GOTO statement, the value of the
control variable is left undefined. You cannot assume that it retains
a value.
Therefore, you must assign a new value to the control
variable if you use it elsewhere in the program. However, if the FOR
loop terminates because of a GOTO statement, it retains the value it
had at termination;
and you can use that value elsewhere.
Example I
FOR N := Lowbound TO HiShbound DO
Sum

:= Sum + IntArra~[NJ;

This FOR loop computes the sum of the elements of Int_luray with index
values from Lowbound through Highbound.
Example 2
FOR Year := 1899 DOWN TO 1801 DO
IF (Year MOD 4) = 0 THEN
WRITELN(Year:4,' IS A LEAP YEAR');
The DOWNTO form is used here to print a list of all the leap years
the nineteenth century.

Example 3
FOR I

:= 1 TO 10 DO
FOR J := 1 TO 10 DO
A[I,JJ := 0;

This example shows how you can directly nest FOR loops.
For each
value of I,
the system steps through all 10 values of J and assigns
the value 0 to the appropriate array element.

5-9

PASCAL

STATE~ENTS

Example 4

FOR

Effiplo~ee

BEGIN

:= 1 TO N DO

HRS

: :::: 40;

FOR

Da~

:= Mon TO Fri DO

Sick(Employee,Da~J

THEN

HR!:l : :::: HR!3····B,
F' a y [: [ IT! P 1 () y f.;> f:~ J : :::: Wa ~J f:'~ [ E III P 1 D ~J e f:.~ ]

END;

* HI:;: S

You can combine compound statements as in this example.
The inner FOR
statement computes
the number of hours each employee worked from
Monday through Friday.
The outer FOR statement resets hours to 40 for
each employee and
computes each person's pay as the product of w~ge
and hours worked.

5.4.2

The REPEAT Statement

The REPEAT statement groups one or more statements for execution until
a specified condition is true.
Fo rma t
REPEAT statement

[;statement ... D UNTIL expression;

where:
expression

specifies a Boolean expression.

Note that the format of the REPEAT statement eliminates the need for a
compound statement.
The expression
is evaluated
after
the
statements are executed.
Therefore, the REPEAT group is always executed at least once.
Example

REPEAT
READ(X);

IF (X IN (10 1 •• 1 9 1 ] ) THEN
BEGIN
Di~it_Count
Di~it_Sum

END
ELSE Char_Count

UNTIL EOLN(INPUT);

:= Di~it_Count t

:= Di~it_Sum t ORD(X) - ORD(IOI)
:= Char_Countt1

Assume that the variable X is of type Char,
and
the variables
Digit_Count,
Digit_Sum,
and Char Count are
integers.
The example
reads a character (X) from the terminal.
If X is a digit,
the count
of digits
is incremented by one;
and the sum of digits is increased
by the value of X.
The ORD function is used to compute the value of
X.
If X is not a digit, the variable Char Count is incremented by
one.
The example continues processing characters from
the
terminal
until it reaches an end-of-line (EOLN) condition.

5-10

PASCAL
5.4.3

STATE~ENTS

The WHILE Statement

The WHILE statement causes one or more statements to be executed while
a specified condition is true.
Fo rrna t
WHILE expression DO statement;
where:
expression

specifies a Boolean expression.

The WHILE statement causes the statement following the word DO
to be
executed while
the expression is true.
Unlike the REPEAT statement,
the WHILE statement controls the
execution of only one statement.
Hence,
to execute a group of statements repetitively, you must use a
compound
statement.
Otherwise,
PASCAL
repeats only the
single
statement immediately following the word DO.
The expression is evaluated before the statement is executed.
If
the
expression
is
initially false, the statement is never executed.
The
repeated statement must change the value of the
expression.
If
the
value of the expression never changes, the result is an infinite loop.
Example 1
l"'HILE NOT EOF(Filel) DO

READL.N(Fil(·?l) y
This statement reads through to the end of the text file Filel.
Example 2
WHILE NOT EOLN(INPUT) DO

BEGIN
/:;:EAD(X) ;
IF NOT (X IN ['A' •• 'Z','a' •• 'z','O' •• '9'J) THEN
IE r T'

END;

:::::

E In' +1

This example reads an input character from the current line on
the
terminal.
If the character is not a digit or letter, the error count
(Err) is incremented by one.
Example 1
SUITt != 0;
Ntests ::::: 1;
AV9 := 100;

WHILE

(AV9 >= 90) AND (Ntests <= Maxtests) DO
BEGIN
Sum := Sum + Test [Ntests];
AV9 := Sum DIU Ntests;
Ntests := NTests +1

END;
IF AVG <: 90 THEN

WRITELN ('Your ave raSe dropped below 90 as of Test "

Ntests:5);

After initializing Sum
to
zero,
this program fragment
repeatedly
calculates a
student's average
test score.
When the average score
falls below 90 (or NTESTS > MAXTESTS), the calculations cease.
If the
average score
is less
than 90,
the system prints an informative
message.

5-11

PASC~L

5.5

STATEMENTS

THE WITH STATEMENT

The WITH statement provides abbreviated
fieJds of a record.

notation

for

references

Fo rma t
WITH record variabJe

[,record variable .•. ]

DO statement;

where:
record variable

specifies the name of the record variable
which the statement refers.

The WITH statement allows you to refer
to
the
fields of a
record
directly instead of using the record.fieldname format.
In effect, the
WITH statement opens the scope of the field identifiers so
tha t
you
can use them as you would use variable identifiers.
Specifying more than one
record
variable has
the
same effect as
nesting
WITH statements.
The names must appear in the order of their
declaration.
Thus, the following two statements are equivalent:

WITH Cat, Dos DO
Bills := Bills + Catvet

+ Do~vet;

and

WITH Cat DO
WITH Dos DO
Bills := Bills + Catvet

+ Dosvet;

Note that if the record Dog is nested within the record Cnt, you
specify Cat before Dog.

must

Example 1

RECORD
Gross : REAL;
Net : REAL;
Bracket : REAL.;
Itemized : BOOLEAN;
Paid : REAL
END;

WITH TaNes DO
IF Net < 10000.0 THEN Itemized

:= TRUE;

This statement tests the value of the
field
Taxes.Net,
Taxes.Itemized to TRUE if Taxes.Net is less than 10000.0.

5-12

and

sets

PASCAL STATEMENTS
Example 2

TYPE

PACKED ARRAY [1 •• 20] OF CHAR;

NalTlf:~

r-~ECORD

Datf.~

(Janv Feb, Mar, Apr, Ma~y Jun,
Jul, Au~, Sep, Oct, Navy Dec)~

Month
[I a ~:~

Yf.,~ar'

1+ •31;
:

INTEGEf<

[NIH

VAR

RECORD

Hosp

F' a't i ent

: NaITIP;
Birthdate : Datp;
A~e

J NTEGEf<

[ND;

WITH HasP, Birthdate DO
BEGIN
Patient := 'THOMAS JEFFERSON

, 'I.

Month ::::: Ap T';
Ase : ":: 2;'56
ENIH

The program segment in this example shows how you can
use
the WITH
statement
to assign values
to
the
fields of a record.
The WITH
statement specifies
the
names of
the
record
variables Hosp and
Birthdate.
The
record
names must be in order;
that is, Hosp must
precede Birthoate.
The assignment statements need
only specify the
field
names,
for
example:
Patient
instead of Hosp.Patient, Month
instead of Hosp.Birthdate.Month, and so forth.

5.6

THE GOTO STATEMENT

The GOTO statement directs the program to exit from a
program segment before its normal termination point.

loop

other

Fo rma t
GOTO label;
where:
label

specifies a statement label.

Upon execution of the GOTO statement, program control
shifts
to
the
statement with
the specified label.
The statement can be any PASCAL
statement or an empty statement.
The label must be defined
in
the
declaration section.
The GOTO statement must be within the scope of the label
declaration.
In addition,
you cannot use a GOTO statement that is outside a
compound statement to jump to a label that
is within
that compound
statement.

5-13

PASCAL STATEMENTS
Example
FOR I

:= 1 TO 10 DO

BEGIN

IF Real_Array[IJ

0.0 THEN

BEGIN

Result := 0.0;
GOTO 10

END;
Result

END;
10:

:= Result + 1.0/Real_Array[IJ

Invertsum:= Result;

This example shows how you can use the GOTO statement to exit from a
loop.
The loop computes the sum (Invertsum) of the inverses of the
elements of Rea] Array.
If one of the elements is zero however,
the
sum
is set to-zero;
and the GOTO statement forces an exit from the
loop.

5.7

THE PROCEDURE CALL

procedure and
A procedure call specifies the actual parameters to a
executes
the procedure.
(See Chapter fi for a complete description of
procedures.)
Format
procedure name

[(actual parameter

[,actual parameter ..• ] )]

where:
procedure name

specifies the name of a procedure.

actual parameter

specifies a constant, an expression, the name
of a procedure or function, or a variable of
any type.

The procedure call associates the actual parameters in the list with
the formal parameters in the procedure declaration.
It then transfers
control to the procedure.
The formal parameter list in the procedure declaration determines the
possible contents of the actual parameter list.
The actual parameters
must be compatible with the formal parameters.
Depending on the types
of the
formal
parameters,
the actual parameters can be constants,
variables,
expressions,
procedure names,
or function names.
An
unindexed array name in a parameter list refers to the entire array.
PASCAL passes actual parameters by the mechanism specified
in the
procedure declaration.

5-14

PASCAL STATEMENTS
Example 1

Tollbooth (ChanSey 0.25, Lane[lJ);
This statement calls the procedure Tollbooth, specifying the variable
Change,
the
real
constant 0.25, and the first element of the array
Lane.
ExampJe 2

This statement calls
the
procedure Taxes,
with
the
expression
Rate*Income and the string constant 'Pay' as actual parameters.
Example 3

Halt;
This statement calls the predeclared
parameters.

5-15

procedure

Halt,

which

has

CHAPTER

PROCEDURES AND FUNCTIONS

Procedures and functions are program units
that perform
tasks
for
other
program units.
A procedure associates a set of statements with
an identifier;
the statements are executed as a
group.
A function
returns a
value.
Each
function
5s associated with a type and an
identifier.
Procedures and functions have
similar
structures and
restrictions.
This chapter
uses
the term subprogram in descriptions that apply to
both procedures and functions.
PASCAL allows you to use three kinds of subprograms:
•

Predeclared subprograms, described in Section 6.1.

•

User-declared subprograms written in PASCAL.
Sections 6.2
and ~.3 present
the general
format of subprograms and
describe how parameters are passed to subprograms.
Sections
6.4 through 6.6 describe how to declare PASCAL procedures and
functions.

•

External subprograms.
This category includes subprograms
written
in other
languages.
Sections ~.7 and 6.8 describe
external subprograms.

You can include subprograms in the main program compilation unit,
or
you can compile
them separately from the main program in modules.
Separately compiled subprograms are considered external
to
the main
PASCAL program, and special usage rules apply (see Section 6.8).

6.1

PREDECLARED SUBPROGRAMS

PASCAL provides predeclared
procedures and
functions
that perform
various commonly used tasks, such as input and output operations and
mathematical functions.
These predeclared subprograms are described
in the following sections.

6.1.1

Predeclared Procedures

PASCAL provides procedures that perform input and output, allocate and
destroy dynamic variables, supply the system date and time, pack and
unpack array variables,
and
halt program execution.
Table 6-1
summarizes these procedures.

6-1

PROCEDURES AND FUNCTIONS
Table 6-1:

Predeclared Procedures

Procedure

Parameter Type

CLOSE (f)

DATE(string)

string = variable of
type
PACKED ARRAY
r 1 •• 1 1 1 0 F C HA R 1

DISPOSE (p)

DISPOSE(p,
tl, ... ,tn)

p = pointer variable
tl, ... ,tn
tag
field constants

Releases
storage
occupied by p . . ; used
when
p . . is a record
with variants.
Tag
field
values
are
optional;
if
specified, they must
be
identical
to
those specified when
storage
was
allocated by NEW.

FIND (f ,n)

= file variable n

Moves
the current
file
position
to
component n of file

= file variabJe

= pointer variable

positive
expression

integer

Action
Closes file f.
Assigns current date
to string.
Dea.lJocates
for p ..... The
variable
p
undefined.

storage
po in ter
becomes

f •

GET (f)

HALT

None

Terminates execution
of the program.

LINELIMIT(f,n)

f
text
file
variable n
integer
expression

Terminates execution
of the program when
output
to
file
f
exceeds n lines. The
value for n is reset
to its default after
each call to REWRITE
for file f.

MARK (a)

a
a variable of
type . . INTEGER

Places a marker
for
use when allocating
memory for dynamic
variables.

NEW (p)

Allocates
storage
for p....
and assigns
its address to p.

file variable

pointer variable

~-2

Moves
the current
file position to the
next component of f.
Then GET (f) assigns
the value of
that
component to f . . , the
file
buffer
variable.

PROCEDURES AND FUNCTIONS

Table

~-l:

Predeclared Procedures (Cont.)

------------,--------------------------------------------------Procedure

Parameter Type

Action

-----------------------------,-------NEW ( p, t I , . . . , t n )

p
pointer variable
tl, ... ,tn
tag
field constants

Allocates
storage
for p~; used when p~
is a
record
with
variants.
The
optional
parameters
tl
through
tn
specify the values
for the
tag
fields
of
the
current
variant.
All
tag
field values must be
listed in the order
in which
they were
declared.
They
cannot
be changed
during
execution.
NEW
does
not
initialize
the
tag
fields.

OPEN(f, attributes)

f
file
attributes
'rable 7-2)

variable
(see

Opens the
file
f
with
the
specified
attributes.

PACK ( a , i , z)

a = variable of type
ARRAY fm •. n] OF T
i = starting
index
of array a
z = variable of type
PACKED ARRAY [u •• v]
OF T

Moves
(v-u+l)
elements
from array
a
to array z by
assigning
elements
afl]
through
a(l+v-u]
to
zful
through
z r vl.
The
upper bound of
a
must be greater than
or equal to (l+v-u).

PAGE (f)

text
f
variable

Skips to
the next
page of file f. The
next line written to
f
begins
on
the
second line of a new
page.
The default
for f is OUTPUT.

PUT (f)

file

file variable

6-3

Writes the value of
fA, the file buffer
variable,
into
the
file f and moves the
current
file
position to the next
component of f.

PROCEDURES AND FUNCTIONS

Table

~-l:

Predeclared Procedures (Cont.)

Procedure

Parameter Type

Action

READ(f, vl, ... ,vn)

f
file
vl, .•. ,vn
variables

RE~D

variable

For vI

through vn,
assigns
the
next value
in the
input file f to the
variable. You must
specify at least one
variable
(vI). The
default for f
is
INPUT.

READLN(f, vl, ... ,vn)

f
text
file
variable vl, ... ,vn
variables

Performs
the READ
procedure
for
vI
through
vn,
then
sets
the
current
file position to the
beginning
of
the
next
line.
The
variable
list
is
optional.
The
default or
f
is
INPUT.

RELEASE(a)

a
a variable of
type "INTEGER

Deallocates
memory
allocated by the NEW
procedure up to the
point set by
the
MARK procedure.

RESET(f)

f i 1 e variable

Enables read i ng from
file
f•
RESET (f)
moves
the current
file position to the
beginning of file f
and
assigns
the
first component of f
to the file buffer
variable,
f ".
EOF(f)
is set
to
FALSE
unless
the
file is empty.

REWRITE ( f)

file variable

Enables writing ~o
file f.
REWRITE(f)
sets the file f to
zero length and sets
EOF (f) to TRUE.

UNPACK(z,a,i)

z = variable of type
PACKED
ARRAyru •• vl
OF t
a = variable of type
ARRAY fm .. nl OF T
i = starting index
in array a

Moves
(v-u+l)
elements from array
z
to array a by
assigning
elements
z (u] through z (vl to
a[i]
through
ari+v-u]. The upper
bound of a must be
greater
than
or
equal to (i+v-u).

6-4

PROCEDURES AND FUNCTIONS

Table

~-1:

Predeclared Procedures (Cont.)

Procedure

Parameter Type

Action

----------,

TIME(string)

string = variable of
type
PACKED ARRAY

Assigns
the current
time to string.

WRITE (f ,pI, •.. ,pn)

f
fUe
pl, ... ,pn
parameters

variable
write

Writes the values of
pI through pn
into
the file f. At least
one parameter
(pI)
must be specified.
The default for f is
OUTPUT.

WRITELN(f,pl, ... ,pn)

f
tex t:
f i 1e
variable pl, ... ,pn
w'rite paramE!ters

Performs
the WRITE
procedure,
then
skips
to
the
beginning
of
the
next line. The write
parameters
are
optional.
The
default
for
f
is
OUTPUT.

r 1 •• 1 1 1 0 F C HA R

6.1.1.1 Input/Output
procedures are:

•
•
•
•
•
•
•
•
•
•
•
•

CLOSE -

Procedures - The

Section 7.4

FIND

- SE'ction 7.5

GET

- Section 7.0

LINELIMI'I' -

Section 7.7

OPEN -

Section 7.8

PAGE -

Section 7.9

PUT

- Section 7.10

READ

- Section 7.11

READLN

- Section 7.12

RESET - Section 7.13
REWRITE

- Section 7.14

WRITE -

Section 7.15

• WRITELN -, Section 7.16

6-5

PASCAL

input

and

output

PROCEDURES AND FUNCT10NS
6.1.1.2 Dynamic
Allocation
Procedures - PASCAL
provides
the
procedures NEW and DISPOSE for use with variables that are dynamically
allocated.
NEW
The predeclared procedure NEW allocates memory for a dynamic variable.
To refer to the dynamic variable, you must use a pointer variable.
Format
~EW(p)

where:
p

specifies n pointer varinble.

The NEW procedure sets aside memory for pA, that is, the variable that
p points
to.
The value of pA is undefined.
You cannot assume that
the allocated space is initialized.
For example, you declare a pointer variable as follows:
VAR PTR

~INTEGER:

This declares PTR as a pointer to an integer variable.
The
integer
variable and
its address,
however,
do not yet exist.
You use the
following procedure call to allocate memory for the dynamic variable:

NEW(PTR);
This cal] allocates a variable of
type
integer.
The variable
is
A
denoted
by PTR ,
that is, the pointer variable's name followed by a
circumflex(A).
This call also assigns the address of
the allocated
integer to PTR.
DISPOSE
The predeclared procedure DISPOSE deallocates memory for
a
dynamic
variable.
As for NEW, you must use a pointer variable to refer to the
dynamic variable.
Format
DISPOSE(p) i
where:
p

specifies a pointer variable.

For example, to deallocate memory for
the dynamic variable
above example, you can issue the following procedure call:

the

DISPOSE(PTR)~

A
As a result of this procedure call, the memory allocated for PTR
is
deallocated;
and the variable is destroyed.
The value of PTR is now
undefined.
Pointer types and dynamic allocation allow you to create linked data
structures.
An example of the use of pointer types and the NEW and
DISPOSE procedures follows.

6-6

PROCEDURES AND FUNCTIONS
This program constructs a Jinked list of records.
Each student record
contains data on one student, that is, a name and a student ID number.
Each record also contains a field
th~t
is a
pointer
to
the
next
record.
The
program
reads a number and a name and assigns each of
them to a field of the student
record.
Then,
it
inserts
the
new
record
on
the
beginning of the linked list by assigning the "Start"
pointer to that new record.

PROGRAM LinkedList (INPUT, OUTPUT);
TYPE f:>TUDENT ... PTF~ ;::: "'~3TUDENT .... DATA;
STRING
PACKED ARRAY[1 •• 20J OF CHAR;
NUMBER = 1 •• 99999;
~3TUDENT .,.DATI~

::::

F~E:COF~D

: SlRINLH
Stud_ID : NUMBER~
Next : STUDENT_PTR
END;
Nam(,~

VAR Startv Student : STUDENT_PTR;
New .... I D : NUMBER;
New_Name : STRING;
Count : INTEGER,
PF~OCEDURE

WF~ I

TE._.DATA (Studt~nt : STUDEN1._.PTH);

(*This procedure prints the list of students. Because the
starts at the beginnin~ of the linked list, the student
names and ID numbers are printed in reverse of the order in
which the~ were entered.*)
printin~

VAR I,J : INTEGER;
Next_Student : STUDENT_PTR;
BEGIN
WHITELN ('NAME:', 'STUDENT IDt:':29);
REPEAT
WRITELN(Student~.Name : 20, Student~.Stud_ID
Next_Student := Student~.Next;
DISPOSE (Student);
Student := Next_Student
UNTIL Student = NIL
END;
(* MAIN PROGRAM *)
BEGIN
Count ::= 0;
WRITELN ('T~pe a 5-di~it ID number,'
'and a name for each student.');
WRITELN('Press CTRL/Z when finished.');
Start := NIL;
WHILE NOT EOF DO
BEGIN
READLN <New_ID, New_Name);
NEW <Student);
Student-.Next := Start;
Student-.Narne := New_Name;
Student-.Stud_ID := NEW_ID;
Start := Student;
Count := Count + 1
END;
IF Count > 0 THEN
WRITE_DATA(START)
END.
6-7

7);

PROCEDURES AND FUNCTIONS
In the main program, the WHILE loop reads a number and a name for one
student.
The following procedure call allocates memory for a new
student record:
NEW (!3tudent) ;
The new record is inserted at the beginning of the list,
that is,
StudentA.Next points to the previous head of the list.
The value of
the new student record is assigned to the Start pointer.
The WRITE DATA procedure writes the name and student ID number for
each stu~ent in the linked list. After writing data for one student,
the procedure assigns the address of the next record in the list to
Next Student.
The following call deallocates memory for one student
record:
11 I SPOS[ (Sturjf.~nt) ;
After deallocating memory,
the procedure assigns the value
of
Next Student to Student.
When the current student record points to
NIL, the loop stops executing.
NEW AND DISPOSE -- RECORD-WITH-VARIANTS FORM
You can use the following forms of NEW and DISPOSE
dynamic variables of a record type with variants:

when

manipulating

NEW ( p , t 1 , . • • , t n )
DISPOSE(p,tl, ••. ,tn)
The parameter p must be a pointer variable pointing to a record with
variants.
The optional parameters tl
through tn must be scalar
constants. They represent nested tag field values where tl
is the
outermost variant.
If you create p without specifying the tag field values,
allocates enough memory to hold any of the variants in
Sometimes, however, a dynamic variable takes values
particular variant.
If that variant requires less memory
allocates, you should use the NEW(p,tl, ••. ,tn) form.

the system
the recard.
of only a
than NEW(p)

For example, the following record represents a menu selection:
TYPE Menu_Ptr =
Meat_T~pe

~Menu_Order;

= (Fish, Fowl, Beef);

Beef _Po rt i on ;:;; «()z _. 1 0, Oz_ .. 16, Oz_32);
Menu_Order = RECORD
CASE Entree : Meat_T~pe OF
( F i S tL T ~ P f.~ : ( S a 1m 0 n, Cod, P (-? r c h, T T' out ) ;
Fish
Lemon: BOOLEAN);
(Fowl_T~pe :
(Chicken, Duck, Goose);
Fowl
Sauce: (Orange, Cherr~, Raisin»;
( Bee f._ Ty p e : ( S tea k, R0 a~; t, P rim e _.. rib) ;
Beef
CASE Size : Beef_Portion OF
Oz_.10, Oz_:L6 : (BeeL.. vf?~.:j : (AT'tichoke, ML·~ed»;
Oz_32 : (Stomach_Cure : (Bicarbonate,
Antacid,None_Needed»;
END;
VAR Menu_Selection

6-8

PROCEDURES AND FUNCTIONS
You can allocate memory for only the Fish variant as follows:

The example below shows how to call NEW and specify tag
for nested variants:

The tag field values must be listed in the order in
declared.

field

which

they

values

were

The DISPOSE(p,tl, ... ,tn) procedure call releases memory occupie~ by p.
The
tag
field
values
tl
through
tn must be
identical to those
specified when memory was allocated with NEW, for example:
D I SPOHE (Menu .... nf:~ 1 f:~ct i on,

Beef,

(}::~ .... 32);

This call deallocates the memory allocated by the last
call shown above.

NEW

proce~ure

6.1.1.3 The MARK and RELEASE Procedures - The MARK and
RELEASE
procedures give you the opportunity to deallocate memory set aside by
several NEW procedures without using several DISPOSE procedures.
If you plan to allocate memory for use by several
dynamic variables,
you
should
consider
placing
a
marker at some
point.
The MARK
procedure places an index so
that
when
you use
the
corresponding
RELEASE
procedure you deallocate
the memory allocated after that
marker.
This saves you from having to DISPOSE several times, one
for
each NEW.
The format for MARK is:
Fo rma t
MARK(a)
where:
a

is a variable of type INTEGER representjng a marker.

The format for RELEASE is:
Format
RELEASE(a)
where:
a

is the variable specified with MARK.

f)-9

PROCEDURES AND FUNCTIONS
Note the following example:
MAf~l\

( A)

NE l,oJ ( Mf:~ n IJ .... ~:) f? 1 (~ c t ion , F i ~;; h ) ;
NEW ( Mf.~nu .... ::>E':I. f~C t i on, BE'ef , O~-::. .... :32 ) ,

MARK<B)
NEW(Menu .... Sf~'lE'ction,Fowl) f

RELEASE(A)
In
this example,
all
three NEW alJocations are deallocated
by
RELEASE(A).
If
RELEASE(B)
had
been specified, only the third NEW
allocation, NEW(Menu_Selection,Fowl), would have been deallocated.

6.1.1.4 Miscellaneous Predeclared Procedures - PASCAL provides
four
miscellaneous predeclared procedures:
PACK, UNPACK, DATE, and TIME.
The predeclared procedures PACK and UNPACK pack and
unpack arrays.
Packing means that the data items are stored as densely as possible.
PACK
You can declare arrays to be packed by specifying PACKED in the TYPE
or VAR declaration.
Sometimes, however, you might want to convert an
array to a packed array within the executable section of the
program.
The predeclared procedure PACK copies elements of an unpacked array to
a packed array.
Fo rma t
PACK(a,i,z)
where:

is a variable of type ARRAY [m .. nl

is the starting index of a.

is a variable of type PACKED ARRAY[u .• vl

OF T.

The number of elements in a must be greater
than or equal
to
the
number of elements
in
z.
PACK(a,i,z)
assigns the elements of a,
starting with a(i], to the array z, until all of the elements in z are
filled.
When specifying
i,
keep in mind that the upper bound of a
(that is, n) must be greater than or equal to i+v-u.
For example, you can read integers from a file into an unpacked array,
element by element, then pack the whole structure.

VAR A
P

ARRAY[1 •• 20] OF 0 •• 15;
PACKED ARRAY[1 •• 20J OF 0 •• 15;

:= 1 TO 20 DO
READ (A[IJ);
PACK (A,1,P);
FOR I

6-10

PROCEDURES AND FUNCTIONS
This program fragment assigns the elements AD] through Ar20] to pfll
through P[201;
that is, 211] the elements in A are packed into P.
You can pack part of the array A as in the following example.

VAR A
p

ARRAY[1 •• 25J OF 1 •• 1S;
PACKED ARRAY[1 •• 20] OF 1 •• 15;

( *P Y'()cedu T'e c a 11

PACI\(A,:L ,P) v
This procedure call moves elements of the
array A into
the
packed
array P.
The parameter 1 specifies that the packing starts with array
element An].
Thus, the elements AD1 through Ar20] are Clssigned
to
prll
through
pr20].
Packing need not start with the first element;
for example, PACK (A,S,P)
packs elements Arsl
through Ar241
into
elements P[l] through pr201.
UNPACK
You can convert a
packed array
predeclared procedure UNPACK.

unpacked

array

with

the

Format
UNPACK(z,Cl,i) ;
where:

is a variable of type PACKED ARRAY[u •• v]

is a variable of type ARRAY[m .. n1

is the starting index of a.

OF T.

For UNPACK, the restrictions on the array indices and the value
are the same as for PACK.

You cannot pass individual elements of a packed array to a
subprogram
as a
VAR parameter.
Therefore, you must unpack the array before you
can pass its elements as VAR parameters.
(*

Declarations *)

VAR P
A

PACKED ARRAY[1 •• 10] OF CHAR;
ARRAY[1 •• 10] OF CHAR;

PROCEDURE PROCESS_ELEMENTS (VAR CH

END;

(* Part of main program *)
READ (P);
UNPACK(P,A,1);
FOR I := 1 TO 10 DO
PROCESS_ELEMENTS (A[l]);

6-11

CHAR) ;

PROCEDURES AND FUNCTIONS
This program fragment reads characters into the packed array P.
The
procedure call
to
UNPACK assigns Prl] through PrIOl to the unpacked
array elements AD]
through ArlOl.
Then,
for
each
ca11
to
PROCESS_ELEMENTS, one element of A is passea to the procedure.
DATE AND TIME
The predeclared procedures DATE and TIME assign the current
time to a string variable.

date

and

ARRAyrl .. 11]

Fo rma t
DATE(string);
TIME(string) ;
where:
string

specifies a variable of
CHAR.

type

PACKED

The folJowing example demonstrates the use of DATE and TIME:
(*

Declarations *)

T()da~3S __ Da te, CI..I l' rent . _Time

PACKED ARRAY[1.+11J OF CHAR;

(* Procedure calls *)
DATE(Toda~5_Date)~

TIME(Current_Time);
These two calls return results in the following

format:

19-Jan--1980

14:20:25
The time is returned in 24-hour format.
Thus, the time shown above is
14 hours,
20 minutes, 25 seconds.
In the DATE procedure, if the day
of the month is a I-digit number, the leading zero does not appear
in
the result;
that is, a space appears before the date string.

6.1.2

Predeclared Functions

PASCAL provides functions that compute arithmetic values, test certain
Boolean conditions,
transfer data
from
one
type
to another, and
perform other miscellaneous calculations.
Predeclared
functions
return a
value
as a
resuJt.
Table 6-2 summarizes the predeclared
functions.
Arithmetic functions perform mathematical computations.
Parameters to
these
functions
can
be
integer,
real,
single-precision,
01
double-precision expressions.
The arithmetic
functions,
except
for
the absolute
and square root functions, return a real value when the
parameter is an integer, single-precision, or real
value.
When
the
parameter
is
a
double-precision
expression,
the
result
is a
double-precision value.
The absolute and square root functions return
a value of the same type as the parameter.

6-12

PROCEDURES AND FUNCTIONS
Boolean functions return Boolean values after testing a condition.
The EOF and EOLN functions operate on files:
EOF tests for the
end-of-file condition on a variable of any file type, and EOLN tests
for the end-of-line condition on a text file variable. The ODD
function tests whether an integer parameter is odd or even.
The
UNDEFINED function,
for which you must supply a real value, returns
the value TRUE when the argument is not in normalized floating-point
format.
Variables containing a value that is not in a normalized
floating-point format cause a reserved operand fault when used
in
arithmetic computations.
Transfer functions take a parameter of one type and
return the
representation of that parameter in ~nother type.
For example, the
ROUND function rounds a real number to an integer, and TRUNC truncates
a real number to an integer.
The miscellaneous functions include PRED and SUCC. The PRED and SUCC
functions operate on parameters of any ordered scalar type (that is,
all scalar types except one of the renl
types).
SUCC returns the
successor value in the type;
PRED returns the predecessor value.
Table 11-2:

Predeclared Functions

Category

Function

Parameter Result
Type
Type

Purpose

Arithmetic

ABS (n)

integer,
rea I,
double

same
as n

Computes
the
absolute value of
n. The type of n
must
be
either
integer,
real, or
double;
nnd the
type of the result
is the type of n.

ARCTAN (n)

integer,
real,
double

real
double

Computes
the
arctangent of n.

COS (n)

integer,
real,
double

real
double

Computes
cosine of n.

EXP (n)

integer,
real,
double

real
double

Computes e**n, the
exponential
function.

LN (n)

integer,
real,
double

real
double

Computes
the
natural
logarithm
of n. The value of
n must be greater
than o.

SIN (n)

integer,
real,
double

real
double

Computes
of n.

6-13

the

the sine

PROCEDURES AND FUNCTIONS

Table 6-2:
Category

Boolean

Predeclared Functions (Cont.)
Function

Parameter Result
Type
Type

Purpose

SQR (n)

in teg er ,
rea] ,
doub1e

same
as n

Computes n**2, the
square of n.
The
type of n must be
either
integer,
real,
or double;
and
the
type of
the result is
the
type of n.

SQRT (n)

in teg e r ,
rea 1,
double

real
double

Computes
the
square root of n.
If n is less than
zero,
PASCAL
generates
an
e r ro r .

EOF (f)

file
variable

Boolean

Indicates whether
the
file position
is at
the end of
the file f. EOF(f}
becomes TRUE only
when
the
file
position
is after
the last element
in the
file.
The
default for
f
is
INPUT.

EOLN (f)

text file Boolean
variable

Indicates whether
the
position of
file
f
is at the
end of a
line.
EOLN (f)
is TRUE
on]y when the file
position is after
the last character
in a
line,
in
which
case
the
valur
of
fA is a
space. The default
for f is INPUT.

ODD {n}

integer

Boolean

Returns
TRUE
if
the integer n
is
odd; returns FALSE
if n is even.

UNDEFINED {r}

real,
double

Boolean

if
Returns
TRUE
the value of r is
a
not
in
normalized
floating-point
fo rma t.

6-14

PROCEDURES AND FUNCTIONS

Table

~-2:

Predeclared Functions (Cont.)

Category

Function

Parameter Result
Type
Type

Transfer

CARD (s)

set

integer

Returns the number
of
elements
currently
belonging
to
the
set s.

CHR (n)

integer

char

Returns
the
character
whose
o r din a ]
n urn be r i s
n (if it exists).

ORD (n)

any
scalclr
type
except
a
real
type

integer

the
Ret.urns
( integer)
ordinal
value
corresponding
to
the value of n.

HOUND(n)

real,
double

integer

Rounds the real or
double-precision
value n
to
the
nearest integer.

SNGL(d)

double

rea]

Rounds
the
double-precision
real number d to a
single-precision
rea 1 n urn b e r •

1:'RUNC (n)

rea I ,
double

in teg e r

Truncates the real
or
double-precision
value n
to
an
integer.

CLOCK

none

integer

Returns an integer
value equal to the
central
processor
time used
by the
current
process.
The
time
is not
expressed
in
milliseconds.

EXPO ( r)

real"
double

integer

Returns
the
integer-valued
exponent
of
the
binary
floating-point
representation of
ri
for
example,
EXPO{8.0) is 4.

Miscellaneous

~-15

Purpose

PROCEDURES AND FUNCTIONS

~-2:

Table

Predeclared Functions (Cont.)

Category

Function

Parameter Result
Type
Type

PRED(a)

any
scalar
type
except
a
r eaJ
type

Purpose

same
Returns
the
type as
predecessor value
parameter in
the
type of a
(if a predecessor
exists). But it is
up to you to make
sure
there
is
pred/succ.
The
compiJer
will
always return with
the
next ordinal
value higher/lower
than a.
There
is
no
bounds
checking. Checking
0ccurs only if the
PRED(a) or SUCC(a)
is
in
an
ass ignmen t
statement,
for
example:
(*$CHE~K+*)

x: =PRED (a)
SUCC(a)

any
same
scaJar
type as
type
parameter
except
a
real
type

Returns
the
successor value in
the
type of a (if
a
successor
exists) .

arithmetic
routines available with
PASCAL also provides additional
PASLIB,
the PASCAL
library.
Use the following format to specify a
routine from the Common Math Library:
FUNCTION routine name (VAR parameter list):

reSUl t type;}
{ FORTRAN

where:
routine name

is one of the names listed in Table

~-3.

parameter list are acceptable arguments for
this
routine.
See
the TOPS-IO/TOPS-20 Common Math Library Reference
Manual.
result type

is the type of the function result.

Table ~-3
lists
these
routines and
their
purpose.
For
more
information concerning
these
routines, refer to the TOPS-IO/TOPS-20
Common Math Library Reference Manual.

6-16

PROCEDURES AND FUNCTIONS

Table h-3:

Library Routines

Routine
Name

Purpose

ARCOS

arc cosine

AINT

truncation to integer

ALOG

natural logarithm

ALOGlO

base-l0 logarithm

AMAXO

largest of a series

AMINO

smalJest of a series

AMINI

smallest of a series

AMOD

rE!mainder

ANINT

nearest whole number

ASIN

a roc sine

ATAN2

arc tangent of two angles

COSD

cosine (angle in degrees)

COSH

hyperbolic cosine

COTAN

contangent

DABS

double-precision D-floating-point absolute value

DACOS

double-precision D-floating-point arc cosine

DASIN

doubJe-precision D-floating-point arc sine

DATAN

doubJe-precision D-floating-point arc tangent

DATAN2

double-precision
t\ol70 angles

DBLE

conversion
from
single-precision
doubJe-precision D-floating-point format

DCOS

double precision D-floating-point cosine

DCOSH

double-precision D-floating-point hyperbolic cosine

DCOTAN

double-precision D-floating-point cotangent

DDIM

d 0 ubI e- prE! cis ion
differenCE!

DEXP

double-precision D-floating-point exponential

DEAXP2.

exponentiation
of
a
double-precision
D-floating-point number to the power of an integer

D-floating-point

arc

D-floating-point

6-17

tangent of
to

positive

PROCEDURES AND FUNCTIONS

Table G-3: Library Routines (Cont.)
Routine
Name

Purpose

DEAXP 1.

exponentiation
of
a
double-precision
D-floating-point number to the power or another
double-precision D-floating-point number.

DFLO,l\T

conversion of an integer
D-floating-point format

DIM

positive difference

)INT

double-precision D-floating-point truncation

double-precision

,OG

double-precision D-floating-point natural logarithm

DLOGIO

double-precision D-floating-point base-IO logarithm

DMAXl

double-precision
series

D-floating-point

largest

DMINl

double-precision
series

D-floating-point

smallest

DMOD

double-precision D-floating-point remainder

DNIN'T

double-precision
number

DPROD

double-precision D-floating-point prod uc t

DSIGN

d 0 ubI e- pre cis ion D-floating-point transfer of sign

DSIN

double-precision D-floating-point sine

DSINH

double-precision D- fl oa t i ng- po in t hyperbolic sine

DSQRT

double-precision D-floating-point square root

DTAN

d 0 u b 1 e- pre cis ion D- fl oa ti ng-po in t tangent

DTANH

double-precision
tangent

D-floating-point

EXPI.

exponentiation of
another integer

EXP2.

exponentiation of a single-precision number to the
power of an integer

EXP3.

exponentiation of a single-precision number to the
power of another single-precision number

FLOAT

conversion of an integer to single-precision format

lABS

integer absolute value

lDIM

integer positive difference

D-floating-point

6-18

integer

nearest

whole

hyperbolic
the

power

PROCEDURES AND FUNCTIONS

Table

~-3:

Library Routines (Cont.)

Routine
Name

Purpose

IDINT

conversion of a dobule-precision D-floating-point
number to integer format

IDNINT

integer nearest whole number for a double-precision
D-floating-point number

IFIX

conversion
fo rma t

of a single-precision number to integer

INT

conversion
format

of a single-precision number to integer

ISIGN

integer transfer of sign

MAXO

largest of a series

MAXI

largest of a series

MINO

smallest of a series

MINI

smallest of a series

MOD

integer remainder

MINT

integer nearest whole number for a single-precision
number

RAN

random number generator

RANS

random number generator with shuffling

REAL

conversion of an integer to single-precision format

SAVRAN

save teh last random number generated

SETRAN

set the seed value for the random number generator

SIGN

transfer of sign

SIND

sine (angle in degrees)

SINH

hyperbolic singe

SNGL

conversion of a
double-precision D-floting-point
number to single-precision format

TAN

tangent

TANH

hyperbolic tangent

0-19

PROCEDURES AND FUNCTIONS
6.2

FORMAT OF A SUBPROGRAM

Subprograms are similar in format to programs. A subprogram consists
of a heading and a block;
the block contains a declaration section
and an executable section.
The heading specifies the name of the subpYogram and lists its formal
parameters.
For a function, the heading also indicates the type of
the value returned.
The declaration section defines labels and
identifiers for constants, types, variables, procedures, and functions
that are used in the subprogram. The executable section contains the
statements that perform the actions of the subprogram.
The labels and identifiers declared in the subprogram are local data
and are unknown outside the scope of the subprogram. The system does
not retain the values of local variables after exiting from the
subprogram. The following is a sample subprogram:

PROCEDURE PRINT_SYM_ARRAY (VAR A: ARR; Side: INTEGER);
(*

This procedure prints arra~ A if it is !=j \;1 mIII e t T' i c (which :i.~:;
determined b~ the function SYMMETRY). The a T' r'a~:1 i ~:; F' T' i nted
in row order, one row per line. *)

VAR I, J, K, L : INTEGER;
FUNCTION SYMMETRY : BOOLEAN;
(* This function returns true if the
other'wisf~'.

arra~

s~mffietric;

false

BEGIN
SYMMETRY := TRUE;
FOR K := 1 TO Side DO
FOR L := K TO Side DO
IF A[K,LJ <> A[L,KJ THEN SYMMETRY

:= FALSE;

ENIH

BEGIN
IF SYMMETRY THEN
BEGIN
WRITELN (IArT'a~ entered: I );
FOR I := 1 TO Side DO
BEGIN
FOR J := 1 TO Side DO
WRITE (A[I,JJ : 4 );
WRITELN
ENIH
END

ELSE WRITELN (IThe
END;

aT'T'a~

is not

s~ffimetT'ic+I);

Subprograms can be nested within other subprograms.
In the previous
example,
the
function
SYMMETRY
is
nested in the procedure
PRINT SYM ARRAY. The order of nesting determines the scope of an
identTfier.

6-20

PROCEDURES AND FUNCTIONS
Data items declared in any particular block of a
PASCAL program are
considered global
to all
its nested subprograms.
Thus, d~ta items
declared in the main program block are global to alJ
subprograms.
A
subprogram can access
its global
identifiers.
For example, the
function SYMMETRY above has no local variables.
Tt
uses
the global
identifiers K and L, and the parameters A and Side, which are declared
in PRINT SYM ARRAY.

6.3

PARAMETERS

Subprograms communicate data with the main program and with each other
by means of parameters.
A subprogram can have any number
of
parameters, but need not have any at all.
The subprogram heading lists the forma] parameters, which specify the
type
of data that will be passed to the subprogram.
For example, the
formal
parameter
list
for
the
procedure PRINT SYM ARRAY
is
the
following:

(VAR A

ARR;

Side:

INTEGER)

This Jist specifies two parameters to be passed
to
PRINT SYM ARRAY:
the variable A of the previously defined type ARR, and the-~al~e Side
of type INTEGER.
Each formal parameter corresponds to an actual parameter, specified in
the
subprogram call.
For example,
a
valid
procedure call
to
PRINT_SYM_ARRAY is the following:

This procedure call passes the variable Current Arr and the
Current Side to PRINT SYM ARRAY.

value

The formal parameters are identifiers used in
the
subprogram;
they
represent the actual parameters in each subprogram call.
You can call
a subprogram several
times with different actual
parameters.
At
execution,
each
formal parameter represents the variable or value of
the corresponding actual parameter.
The formal-value parameters,
and
the
actual
parameters to which they correspond, must be of identical
types.

6.3.1

Formal Parameters

The formal parameter list specifies the identifier for each parameter
and the type of each parameter to be used within the subprogram.
Format
(
[

[mechanism specifier ]
[ mechanism specifier]

identifier list
identifier list

n-21

type;
type .•. ]

);

PROCEDURES AND FUNCTIONS
where:
mechanism specifier

indicates how PASCAL passes data to this
parameter.
The mechanism specifiers for
PASCAL subprograms are VAR,
PROCEDURE,
and FUNCTION.

identifier list

specifies one or
more
separated by commas.

type

specifies the type of the parameters
in
the
list.
You
can
pass values,
variables, procedures, and functions
to
a
subprogram written
in PASCAL,
as
described below.

PASCAL provides
subprograms.

two

methods

for

passing

identifiers,

parameters

PASCAL

Value -- the value of the actual
parameter expression
is
passed
to
the subprogram.
The subprogram cannot change the
actual parameters's value during execution.
(Value
is the
default method.)

Variable -- the address of the parameter variabJe
is passed
to the subprogram.
The subprogram can change the parameter's
value.
The VAR mechanism specifier
indicates
that
a
parameter is to be passed as a variable parameter.

Value parameters pass the value of the actual parameter expression to
the subprogram.
The subprogram does not change the actual parameter's
value during execution.
Therefore, after the subprogram executes, the
value of the actual parameter is the same as before the execution of
the subprogram.
The following example shows a formal
parameter list
that includes two value parameters:

INTEGER;

PROCEDURE EXAMPLE (Counter

NAME

Variable parameters pass the address of the
actual
parameter
expression to
the subprogram.
The subprogram can change the actual
parameter's value.
Therefore,
after
the
subprogram executes,
the
value
of
the actual
parameter
is the value assigned
to
the
corresponding formal parameter within the subprogram.
To specify a
variable parameter, use the reserved word VAR before the parameter in
the formal parameter list.
The
following
example shows the
same
formal
parameter list as in the previous example.
In this example,
the parameters have been defined to be variable parameters.

PROCEDURE EXAMPLE (VAR Counter

INTEGER;

VAR Name:

CHAR);

6.3.1.1

Value Parameters - By default, PASCAL passes value parameters
to PASCAL subprograms. When you specify a value parameter, the formal
parameter list does not include the reserved word VAR.

The actual parameter corresponding to a formal value parameter must be
a
compatible expression.
Value parameters follow the
rules for
assignment compatibility.
For example, the following list passes all
parameters by value:

INTEGER;

CHAR)

6-22

PROCEDURES AND FUNCTIONS
integer
The actual parameters corresponding
to A and B must be
expressions.
The actual
parameter corresponding
to C must be a
character expression.
If the subprogram changes the value of a value parameter,
the change
is not
reflected
in the calling program unit.
Thus, if you do not
want the value of an actuaJ parameter to change as a
result of the
execution of a subprogram, you pass it as a value parameter.

6.3.1.2 Variable Parameters - To pass a variabJe parameter,
use
the
reserved word V~R.
You must
use
the VAR specifier to pass file
parameters and to pass actual parameter variables with values that
change during
execution of the subprogram.
The corresponding actual
parameter must be a
variable;
it cannot be a
constant or an
expression.
When you pass a variable as a variable parameter, the subprogram has
direct access
to
the corresponding actual parameter.
Thus, if the
subprogram changes the value of the formal parameter, this change
is
reflected in the actual parameter in the calling program unit.
In
the example procedure PRINT_SYM_ARRAY,
the actual
parameter
corresponding
to A is passed using variable semantics.
It must be a
variable of the previously defined type, ARR.
The actual
parameter
corresponding
to Side
is passed by value and must be an integer
expression.
The VAR specifier must precede each identifier list that
is to be
passed
using variable semantics.
Thus, VAR can appear more than once
in the formal parameter list, for example:

(VAR SEA, BREEZE:

REAL;

WIND:

INTEGER;

VAR SICK

As a result of this
formal
parameter list,
the actual
parameters
corresponding
to SEA,
BREEZE,
and SICK are passed
as variable
parameters;
and the actual parameter corresponding to WIND is passed
as a value parameter (the default) •
Compatibility
Variables passed to a subprogram as actual variable parameters must be
of the same type as the corresponding formal parameters.
The following restrictions also apply to variable parameters:
•

You cannot pass an element of a packed structure with the VAR
specifier,
although you can pass the entire structure.
You
must unpack the structure or assign its elements to simple
variables before you can pass individual elements.

•

You cannot pass a variable of a packed set type to a
formal
parameter that is an unpacked set type, and vice versa.

•

You cannot pass a
tag
field
of a
record with the VAR
specifier
(see Section 2.3.2.1).
You can pass the entire
record or assign the tag field to another variable
in order
to pass it.

~-23

PROCEDURES AND FUNCTIONS
6.3.1.3 Formal Procedure and
Function Parameters - PASCAL allows
procedures
and
functions
to
be passed
as parameters
to
other
procedures or functions.
To do this, a
full
procedure or
function
heading
is given as one parameter of the procedure being declared.
For example, the following procedure declaration specifies one value
parameter and one function parameter:
F' f~ 0 CE DU I:~ E ACT U A L.

VAL.

(V Al..! J NT EGE r~ ,
FUNCTION FORMAL. (Fl:INTEGER):INTEGER);

:= FORMAL (VAL + 1),

The formal function parameter takes one value integer
parameter,
and
returns an
integer value.
When procedure ACTUAL is called, you need
to supply it with two parameters:
an integer value;
and, the name of
a
function which
takes one integer argument and returns an integer
value.
Note that ONLY the function name is passed to
the
procedure;
do NOT supply the function's parameters in the procedure call.
They
are supplied when the procedure calls the function.
The parameter list of the formal procedure or function may consist of
anything
that can be defined
in a
normal
procedure or function
declaration, including value parameters,
VAR parameters,
conformant
arrays, and even other procedures and functions.
These procedures and
functions obey the same rules as the formal procedure or
function of
which they are a parameter.
The following is an
example of
procedures nest~d to two levels:

procedure

heading

with

formal

PROCEDURE OUTER(PROCEDURE FORMAL1(FUNCTION FORMAL2:REAL, B:REAL»,
Procedure OUTER has one parameter, a procedure.
This procedure has
two
parameters, a real-typed function (with no parameter.s) and a real
value parameter.
When a subprogram is called with a procedure or function
parameter,
the parameter
lists of
the
formal
and
actual parameters must be
congruous.
This means that the parameter lists must have
the
same
number of parameters;
and each corresponding parameter must be of the
same kind (value, VAR, etc.) and of the same type.
In
the
following
example, procedure YOU BET could be passed as a parameter to procedure
OUTER above, because Y5U BET and
FORMALl
have congruous parameter
lists.

PROCEDURE YOU_BET(FUNCTION YOU: REAL, BET: REAL),
In the following example,
procedures PRINTHEX and
PRINTOCTAL have
congruous parameter lists.
Procedure PRINTBINARY's parameter list is
not congruous to either of the others because both of
its parameters
are value parameters.

PROCEDURE PRINTHEX (VAL: INTEGER, VAR SIZE: INTEGER),
PROCEDURE PRINTOCTAL. (I: INTEGER, VAR F: INTEGER),
PROCEDURE PRINTBINARY (NUM: INTEGER, WIDTH: INTEGER),

f)-24

PROCEDURES AND FUNCTIONS
An optional syntax for formal procedure and function parameters is
to
omit
the
parameter
list
in
the declaration.
If this is done, no
checking on the number or type of parameters
is possible,
and
only
value
parameters are
allowed
when calling the formal procedure or
function.
No conformant array, procedure, or function parameters can
be passed to a formal procedure or function unless they are explicitly
declared.
For information on calJing sequences for
procedures, refer to Appendix G.

6.3.2

user-defined

functions

and

Conformant Arrays

Some programming applications require general
subprograms
that can
process arrays
of varying
size.
PASCAL allows you to declare such
subprograms using conformant arrays.
You can call the subprogram with
arrays of different sizes, as long as their bounds are within those
specified by the formal parameter.
For example, you could write a procedure that sums the components of a
one-dimensional
array.
Each time you use the procedure, however, you
might want to pass arrays with different bounds.
Instead of declaring
multiple procedures using arrays of each possibJe size, you can use a
conformant-array parameter.
The procedure treats the formal parameter
as if its bounds were those of the actual parameter.
The format of a conformant-array parameter is:
array id

ARRAyridlow •. idup : scalar-type id]
OF conformant-array type;

where:
idlow

is the lower bound constant identifier

idup

is the upper bound constant identifier

array id

is one or more
array

identifiers

associated

with

the

Idlow and idup are bound identifiers.
They define the lower and
the
upper
limits,
respectively,
of
the array.
Each id is treated as a
constant value in the subprogram;
therefore, you cannot assign values
to
the id other than in the definition.
You cannot use a subrange to
define the limits.
Scalar type specifies the data type of the index.
Note that you must
use
a
type
identifier
to
specify the range of indices.
The type
identifier can be one of the predefined scalar types INTEGER or CHAR.
Conformant-array type can be either
conformant-array specification.

6-25

type

identifier

another

PROCEDURES AND FUNCTIONS
Multidimensional arrays can also use conformant array parameters.
fo rma tis:

The

array-id : ARRAY ridlow •• idup : scalar-type idl
OF ARRAY ridlow •. idup : scalar-type idl
OF conformant-array type;
An abbreviated form can also be
conformant arrays. The format is:

used

define

multidimensional

array-id : ARRAY [idlow •• idup : scalar-type id;
i dn •. idn : scalar- type idn; .•• ] OF
conformant-array type;
When a subprogram with conformant-array parameters is called, the IDs
(the bound identifiers) assume the values of the lower and upper bound
values of the actual parameters. These values are those specified
in
the definition section of the actual-array parameter.
A conformant-array parameter can have up to one conformant dimension
packed.
This means that a one-dimensional conformant-array parameter
can be packed or unpacked.
A multidimensional conformant-array
parameter
can be unpacked, or its last
(rightmost)
conformant
dimension can be packed. Note that this restriction applies only to
the conformant part of the parameter;
the conformant array type can
be of any type, packed or unpacked.
The components and indices of the actual and formal conformant-array
parameters
must
be
of
compatible
types.
The
rules
for
conformant-array compatibility are the same as those for other arrays
with one exception.
That is, the range of the index types of the
actual-array parameter must be within the range specified for
the
formal parameter.

6-26

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

PROCEDURES AND FUNCTIONS
Example 1
The following program shows how to declare
parameters.

and

use

conformant-array

PROGRAM D~narr(INPUT~ OUTPUT);
(* Th i~:; p Y' 0 !,:,! T' a IT! ill u E t rat f:'~ ~:; t h f.~ U S (~ 0 f e D n f 0 T' IT! ant .... a T' T' a ~:~
paraITlf.,-ters.
The PY'ocedure SUIT! is called from the main program with two
d i ffp r.:~nt actua I pa T'amet(,:,~ r~:;: AT' l':L and AT' T'2. *)
1 ~:-;O ,
TYPE Rn~,~
ARRAY[1 •• 5J OF INTEGER;
VAR AT'r1
ARRAY[7 •• 20J OF INTEGER,
AT'r2
K,J ! INTEGER,
t

•

PROCEDURE Sum (VAR InarT' : ARRAY [Low •• High : INTEGER] OF
INTEGER),
(* This proceduY'e accepts actual-arT'a~ parameters with intpgpr
components whose indices are within the range specified b~
Rng. *)
VAF, I, Ans

INTEGER,
BEGIN
Ans :== 0;
FOR 1:= Low TO High DO
An ~j

~~ I"'I~)

+ I naT' T' [ 1 :] ;

WRITELN('The sum of the components is: ',Ans)
END,
(*end Sum*)
(* MAIN PROGF~AM *)
BEGIN
WRITELN ('TYPE 5 INTEGERS'),
FOR K:= 1 TO 5 DO
READ (AT'r10\]);
Sum (A r 1':L ) ;
WRITELN('TYPE 14 INTEGERS');
FOR J:= 7 TO 20 DO
READU~rr2L.J]) ;.
Sum (Arl~2)
END.

This program sums the components of the arrays Arrl
and Arr2.
The
procedure Sum
includes a
I-dimensional conformant-array parameter,
Inarr, whose indices are of type Rng. Within the main program, Sum is
called with two different arrays:
Arrl with index type (1 •• 51, and
Arr2 with index type [7 •• 20].
The first procedure call, Sum(Arrl), passes Arrl to Sum.
Low assumes
the value of 1, and High assumes the value of 5.
The FOR loop
processes array components Inarr[ll to Inarrr5].
When Sum
is called
with Arr2,
Low assumes the value of 7, and High assumes the value of
20. When Sum
is called with Arr2,
the FOR loop processes the
components Inarr[7] to Inarr[20].

6-27

PROCEDURES AND FUNCTIONS
Example 2
A conformant-array parameter can have more than one ~imension,
this example:
TYPE Level_Ran~e

= 1 •• 6;

Ne 1 aSSE-:·~~:; ::" 1 •• 8,
Nstudents = 1 •• 40~
Names = PACKED ARRAY [1 •• 35J OF CHAR;
VAR ~)tudents

AF\f~AY

F' r~ 0 CEIl U R E 1\ i d .... C () U n t

[1 •• 6,1 •• 0,1 •• 40] OF NalllE'S'

(!; c h (.1 () 1 :

A F~ I:~ A Y [: L (-::' V f? 1 .... L.. 0 w. • L.. (.:~ vel .... H :i. ~.:.~ h

L..evel_Ran~e;

Nclasses_L()w •• Nclasses_Hi~h

N 5 t IJ d €.~ n t tIl... 0 w. • N ~:) t u d f:~ n t .... H i ~.~ h :
: NstudentJ OF Names);

N cIa s 5 f? S '

This example defines School as
a
three-dimensional
conformant-array
parameter.
Each array passed to School mjght contain the names of all
the students in a particular elementary school.
The
indices of
the
array denote the number of grades in the school, the number of classes
at each grade level, and the number of students in each class.
The actual-array parameters can have from one to six grades,
one
to
eight classes at
any grade level, and one to forty students in any
particular
class.
Furthermore,
the
indices of
the
actual-array
parameters must be within the ranges shown in the TYPE section.
For
example, a school with six grades must use integer indices from one to
six.
Indices of zero to five, for instance, cannot be used.

6.4

DECLARING A PROCEDURE

A procedure is a group of statements that perform a
set of actions.
The
use
of
procedures allows
you
to break a complex program into
several units, each of which
performs one
task.
For
example,
a
program that computes social statistics from survey data might contain
procedures to input and validate the data, select a random sample, and
print results.
To declare a procedure, specify its header and block in the
procedure
and
function
section.
The header consists of the word PROCEDURE and
the procedure name along with any parameters you want to include.
The
block
consists
of
its own declaration
section and the executable
section.
You can declare a
procedure
in
the main
program,
in a
module, or in another subprogram.

/5-28

PROCEDURES AND FUNCTIONS
Fo rma t 1
PROCEDURE procedure-identifier

[(formal parameters)]

label section;
constant section;
type sec t ion;;
variable section;
procedure-and-function section;
BEGIN
statement [;
END;

statement .... ]

Format 2

f[ r[.GLOBAL]
TI} PROCEDURE procedure id [(formal parameters)]
FORTRAN 1 lJ

[FORWARD; ]

Format 3
PROCEDURE procedure id

[(formal parameters)] ;{[EXTERN [AL] ; TIl.
FORTRAN;
lJ (

Format 1 shows the format for a procedure that is included within
program that calls it.

the

Format 2 shows
the
format
for
a
procedure
that can
be called
externally,
that
is,
it can be called from another program.
The
FORWARD declaration permits the
use
of
forward
references
in
the
declarations section.
Forward declarations are described in Section
6.6.
The FORTRAN declaration at the beginning .of the
1ine
indicates
that this procedure can be called externally by a FORTRAN program.
Format 3 shows the
format
for
a
procedure
that
is being
called
externally.
The procedure must be compiled separately from the source
program that calls it.
Placing the FORTRAN declaration at the end
of
the
line
indicates
that
the
procedure being called by PASCAL is a
procedure written in the FORTRAN language.
Refer to Section ~.7.
procedure id
specifies the identifier to be used as the name of the procedure.
formal parameters
contains the names and
types of
the
formal
parameters.
It
optjonally can
include
reserved words VAR, PROCEDURE, and
the
FUNCTION.
label section
declares local labels.
constant section
defines local constant identifiers.
type section
defines local types.

6-29

PROCEDURES AND FUNCTIONS
variable section
declares local

v~riables.

procedure-and-function section
declares local procedures and functions.
s ta temen t
specifies
procedure

an action to be performed by the procedure.
A
can contain any of the statements described in Chapter

A procedure consists of a heading and a block. The procedure block is
similar in structure and contents to the main program block, with the
following exceptions:
•

The declaration section cannot contain VALUE initializations.

•

The procedure block ends with END followed by a semicolon
(i),
rather than a period (.) as in the main program. The
procedure does not have a block if it is a
forward
declaration or is defined externally (EXTERNAL or FORTRAN).

to the procedure,
You must declare all the variables that are local
but you should not redeclare the formal parameters or the procedure
identifier as variable, type, or constant identifiers.
For the two examples that follow, assume that these declarations
been made:

CONST NUMBER = 20;
TYPE Range = 0 •• 100;
List = ARRAY[l •• NUMBERJ OF
VAR

have

Ran~e;

ARF~ : List;
Min i IlIU Ill' Ma ~·d III u tTl
Ave T'age : r~EAI...;

Example 1

PROCEDURE READ_WRITE (VAR A
List);
VAR I : INTEGER;
BEGIN
WRITELN ('Type a list of 20 integers',
'in the range of 0 to 100.');
FOR I := 1 TO Number DO
BEGIN
READCA[IJ);
WRITE(A(IJ!7);
WRITELN
END
END;
The procedure READ WRITE reads a list of
into the array A,-and writes the array.
parameter, the array A.

20 integers,
inserts them
READ WRITE uses one variable

Given the declaration of ARR, the following is a valid procedure call:

READ_WRITECARR);

6-30

PROCEDURES AND FUNCTIONS
As a result of this call, the list of integers is written in the array
ARR.
The value of
this array is then returned to the program unit
that called the procedure READ WRITE.
Example 2

PROCEDURE MIN_MAX_AVG (VAR Miny Max : Ran~e;
VAR Avg : REAL; A : List);
VAR Sum, NMax~ NMin~ J : INTEGER;
BEGIN
M3X := Ael); Min:= Max; Sum:= Max;
NMax := 1; NMin:= 1;
FOR J := 2 TO NUMBER DO
BEGIN
Sum := Sum + A[JJ~
IF A[J) > Max THEN
BEGIN Max := AeJJ;
NMa>: : =:: 1
END
ELSE IF A[JJ = Max THEN
NMax := NMax + 1;
IF AEJJ < Min THEN
BEGIN Min := A[J);
NMin !:=: 1
END
Min THEN
ELSE IF A[JJ
NMin : :~: NMin + 1
END;
AVG := Sum/NUMBER;
WF~ I TELN;
WRITELN('Maximum =',Max:4,', occurring',NMax:4, , times');
WFnTELN;
WRITELN('Minimuffi =', Min:4,', occurring', NMin:4,' times');
WFnTELN;
WRITELN ('Average value (truncated) =', TRUNC(Avg):10);
WRITELN ('Average value ='r Avg : 20)
END;
This procedure computes the minimum, maximum, and average values
in
array A.
It also counts the number of times the minimum and maximum
values occurred, and stores those numbers
in NMin and NMax.
The
WRITELN statements print the results of each of those computations.
Min, Max, and Avg are formal variable parameters.
Their values,
as
assigned
in
the
procedure MIN_MA.X_AVG, are returned to the calling
program unit and can be used for further computations in the
program.
A is specified as a value parameter because its value does not change
in the procedure.
The following is a valid procedure call to the procedure:

The values of the formal parameters Min, Max, and Avg are returned
to
the actual
parameters Minimum,
Maximum,
and Average,
which were
defined in the main block of the program.

6-31

PROCEDURES AND FUNCTIONS
6.5

DECLARING A FUNCTION

A function is a group of statements that compute a scalar or
pointer
value.
To declare a functjon, specify its heading and block in the
procedure and function section.
Format 1
FUNCTION function id

[(formal parameters)] :result type;

labe] sect jon;
constant section;
type section;
variable section;
procedure-and-function section;
BEGIN
statement
END;

[;statement ... ]

Format 2
[GLOBAL] 11 I( FUNCTION function id
{IT [FORTRANl
lJ
Fo rma t

[( forma1 parameters)]

[FORWARD;D

FUNCTION function id

[(formal

parameters)] ;{[EXTERN [AL]
FORTRAN;

Format 1 shows the format for a function that is included
program that calls it.

;ll}
lJ
within

the

Format 2 shows
the
format
for
a
function
that can be called
externally,
that
js, it can be called from another program.
Placing
the FORTRAN declaration at the beginning of the line indicates that a
FORTRAN program can call
this
procedure.
The FORWARD declaration
permits the use of forward references
in
the declarations section.
Forward declarations are described in Section h.n.
Format 3 shows the format for a function that is defined
externally;
the procedure
is compiled
separately from the source program that
calls it.
Placing the FORTRAN declaration at
the
end
of
the
line
indicates
that the procedure being called by PASCAL is written in the
FORTRAN language.
Refer to Section h.7.
function id
specifies the

identifie~

to be used as the name of the function.

formal parameters
contain the names and
types of
the
formal
parameters.
They
optionally can
include
the
reserved words VAR, PROCEDURE, and
FUNCTION.
result type
specifies the type of the function's result.
a scalar or pointer value.

f5-32

The result must

PROCEDURES AND FUNCTIONS
label section
declares local labels.
constant section
defines local constant identifiers.
type section
defines local types.
variable section
declares local variables.
procedure-and-function section
declares local procedures and functions.
statement
specifies an action to be performed by the function.
A function
can contain any of the statements oescribed in Chapter 5. A
function must contain a statement that assigns a value to the
function
ioentifier (for every potential path through the code).
If it does not, the value of the function could be undefined.
A function consists of a heading and a block.
The formal
parameter
list in the function heading is identical in format to the list in the
procedure heading. The function block is similar in structure and
contents to the main program, with the following exceptions:
•

The function cannot contain a value initialization section .

•

The function block ends with END followed by a semicolon (;),
rather than a period
(.)
as in the main program. The
function does not have a block if it is
a
forward
declaration, or if it is defined externally (EXTERN(AL) or
FORTRAN) •

You must declare all variables that are local to the function, but you
should not redeclare a variable, type, or constant with the same name
as a formal parameter or the function identifier.
Each function must include a statement or statements that assigns a
value of the result type to the function name. The last value that is
assigned to the function name is returned to the calling program unit.
To use the value, include a function call in the calling unit.
Unlike
a procedure call, a function call
is not a statement.
It simply
represents a value of the function's resuJt type.

6-33

PROCEDURES AND

FUNC~IONS

Side Effects
A side effect is an assignment to a non]ocal variable,
or
to a
VAR
parameter,
within
a
function block.
Side effects can change the
intended action of a program and therefore, should be avoided.
The
following program illustrates an example of a side effect.

PROGRAM Example (OUTPUT)~
VAR X,Y : INTEGER,
ANSI, ANS2 : BOOLEAN;
FUNCTION Positive (ThisVar
BEGIN
Positive := FALSE;
IF ThisVaT' :::- 0 THEN
BEGIN
X := ThisVar - 10;
Positive := TRUE
END
END;
(* 0~nd F'o~; i t i v(,.~*)

INTEGEI:;: )

(* Side effect on X *)

(* MAIN PROGRAM *)

BEGIN
Y ::::: ?;

BOOLEAN;

X: == 1 ~5 Y

ANS1 != Positive(X) AND Positive(Y),
WRITEI... N ('ANSI e(~uals',ANS1);
Y != '7; X !:::: 1:5,

ANS2 != Positive(Y) AND Positive(X);
WRITELN ('ANS2 eouals',ANS2)
END.
This example generates the following output:

ANS 1 e(~ua 1 5
ANS2 e~uals

TRUE
FALSE

Thus, the output depends on which function call
is evaluated
first:
PositivelY)
or Positive(X).
PASCAL does not guarantee which part of
an expression is evaluated first.
The resulting value of a
function
should
not depend
on when the function is called, as it does in the
example above.
Therefore, you shou]d avoid
side effects on global
variables.
Example I

FUNCTION COUPONS: REAL;
VAR ANS : (YES, NO);
AMOUNT, SUBT : REAL;
BEGIN
SUBT := 0;
WRITELN ('AnY coupons? Type ~es or no.');
READLN (ANS);
IF ANS==YES THEN
BEGIN
WRITELN ('Type value of each coupon, one per line,
CTRL/Z when finished?');
REPEAT
READLN (AMOUNT);
SUBT := SUBT + AMOUNT
UNTIL EOF
END;
COUPONS := SUBT
END;
(* End COUPONS* )

6-34

PROCEDURES AND FUNCTIONS
The function COUPONS computes the total value of a group of coupons.
It uses only the
three local variables, ANS, AMOUNT, and SUBT, and
requires no parameters.
The result of this function is the real totaJ
of the coupon values.
The assignment statement, COUPONS := SUBT,
assigns the result to the function identifier.
To use the function COUPONS, specify its name, as follows:

TOTAL

:= SUBTOTAL - COUPONS;

The function call is treated
as a
real-vaJued expression
in
this
statement.
Note
that you can use the function call in the same way
that you can use a value of its result type.
Example 2

FUNCTION SYMMETRY <VAR A : ARR) : BOOLEAN;
<*This function returns true if the
returns false otherwise.*)

VAR I, J : INTEGER;
BEGIN
SYMMETRY := TRUE;
FOR I := 1 TO SIZE DO
FOR J := I TO SIZE DO
IF A[I,JJ <> A[J,IJ THEN
END;
<* SYMMETRY *)

arra~

A is

s~mmetric;

SYMMETRY:= FALSE

the array A.
The function SYMMETRY uses one variable parameter,
if A is
result is TRUE
SYMMETRY returns a
Boolean value;
the
symmetric, and FALSE if A is not symmetric.

6.6

FORWARD DECLARATIONS

Normally, you must declare subprograms before you refer
to
them.
However,
a
subprogram can reference another subprogram that has not
yet been declared if you use a
FORWARD declaration.
The forward
declaration
provides
the
compiler with information about the
forward-declared subprogram's formal parameters, and
indicates that
the block of the subprogram follows later in the source file.
For example, a complete declaration is impossible if two
subprograms
call each other
recursively.
Omitting
the declaration
is also
impossible because PASCAL needs information about a
subprogram's
formal parameters before it can compile a reference to the subprogram.
Therefore, you must forward-declare one of the recursive subprograms.
A forward declaration consists of the subprogram heading
(including
the formal
parameter list,
if any, and the result type, if it is a
function) and the FORWARD directive without a
subprogram block,
for
example:

PROCEDURE CHESTNUT (BLD :REAL; DOC: CHAR; VAR ARC: REC);FORWARD;
This example declares the procedure CHESTNUT in a FORWARD declaration.
The FORWARD declaration
includes only the information shown in the
example.
It could also include FORTRAN or GLOBAL,
as in Format 2
above.

6-35

PROCEDURES AND FUNCTIONS
When you specify the block of a forward-declared subprogram, you must
not repeat the formal parameter list or the result type of a function.
Except for these omissions, declare
the
heading
and
block
in
the
normal way.
Example

FUNCTiON ADDER (OPi, OP2, OP3 : REAL) : REAL; FORWARD;
PROCEDURE PRINfER (STUDENT: NAME_ARRAY) ;

BEGIN

Z := ADDER CA,B,C)
END;
FUNe'fION ADDER

OPi, OP2, OP3

REAL

BEGIN

PRINTER ('Leonardo da Vinci');

This example forward-declares the function ADDER.
The block of
the
function appears after the declaration of the procedure PRINTER.
Note
that the heading of the ADDER block specifies
its
formal
parameters
and result type within comment delimiters.
Although you must omit the
parameter list and result type when you define
the
function block,
inserting them as a comment is a good documentation practice.

6.7

EXTERNAL SUBPROGRAMS

The FORTRAN and EXTERNAL directives indicate procedures and
functions
that are defined external to a PASCAL program.
With these directives,
you can declare subprograms written
in another
language
(such as
FORTRAN or MACRO) and PASCAL subprograms that are compiled separately.
In PASCAL, the FORTRAN directive should be used
only for
separately
compiled
routines written in FORTRAN or a language using the FORTRAN
subprogram calling conventions.
The EXTERNAL directive must be used
only for separately compiled external routines written in PASCAL.
If you declare separately compiled
PASCAL subprograms as EXTERNAL,
their
names must be unique
within
the
first six characters.
In
addition, an external subprogram cannot have the same name as the main
program.

6-36

PROCEDURES AND FUNrTIONS
Example 1

FUNCTION SCORE (RESULT:

REAL)

REAL;

EXTERNALf

The function SCORE is a procedure in a library that exists
This declaration declares SCORE as an external subprogram.

disk.

Example 2

PROCEDURE FORSTRCS : PACKED ARRAY[L •• U:INTEGERJ OF CHAR) FORTRAN;
This statement declares the
FORTRAN procedure FORSTR.
The
parameter list specifies S as a conformant-array parameter.

6.8

formal

MODULES FOR SEPARATE COMPILATION

By placing PASCAL procedures and
functions
in a
MODULE,
you can
compile
them separately from
the main program.
At load time, you
specify the compiled files containing the main program and the modules
to be loaded together in the executable image.
The executable image
can include any number of modules, and each module can contain any
number of subprograms.
Format
[ [OVERLAID1]
MODULE module name (program parameters);
label section;
constant section;
type section;
variable section;
procedure-ana-function section;
END.
[OVERLAID]
Specifies that the module shares global
values with
the main
program
that calls it.
If
the module is OVERLAID;
then the
constant, type, and variable sections must be identic~l to
those
in the main program.
module name
specifies the identifier to be used as the name of the module.
program parameters
lists the external files.
This list must be identical
and in content to the list in the main program heading.

order

label section
declares global labels.
PASCAL issues a warning-level message if
a module contains a label section, but ignores the labels.
constant section
declares global constant identifiers as in a main program.
type section
defines global types as in a main program.

6-37

PROCEDURES AND FUNCTIONS
variable section
declares global variables as in a main program.
procedure-and-function section
declares the procedures and functions contained in the module.
A module is similar to a main program, except that
it has
no value
initialization section and no executable section.
Modules can contain
the
label,
constant,
type,
variable,
and
procedure-and-function
sections.
(PASCAL issues a warning-level message if a nodule contains
label
declarations,
but
ignores
the
labels.)
If
the module
is
OVERLAID,
then
the
constant,
type,
and
variable sections and the
program parameters must be identical to those
in
the main
program.
The
procedure-and-function section defines the subprograms contained
in the module.
To ensure that the program parameters and
the
constant,
type,
and
variable sections are
identical
in
the main program and in all
modules, you can place them in a separate file.
Then, you can use the
%INCLUDE directive
to
insert the contents of the file into the main
program and
into
all modules,
instead
of
repeating
all
the
declarations and definitions.
If a module shares global variables with a
main
program,
both the
module and the program headings must include the attribute [OVERLAID].
If the module heading does not contain [OVERLAIDl, then all its global
variables are private to the module and cannot be accessed by the main
program or other modules.
Subprograms declared at the outermost level
of a module can be declared and called from the main program (or from
another module).
You must declare the subprogram with the
EXTERNAL
modifier
in
the calling program unit and with the [GLOBALl attribute
in the module.
Similarly, subprograms declared at the outermost level
of
the main program with the [GLOBAL] attribute can be declared as
EXTERNAL in a module.
Each subprogram in the module can access data declared either
to itself or by the main program.

locally

Examples

[OVERLAID] MODULE SEP (INPUT, OUTPUT);
VAR I : INTEGER;
PROCEDURE READER (N : INTEGER);
VAR K,P : INTEGER;
BEGIN
I

:= 0;

FOR K := 1 TO N DO
BEGIN
READ (P);
IF P=O THEN I
END
END; <* READER *)
END.
<*MODULE SEP *)

:= I + 1;

The MODULE SEP contains one procedure, READER.
You can declare READER
as an external
subprogram in another module or in the main program.
Because SEP contains the definition of a global data item,
I,
it is
declared
as an
[OVERLAID]
module.
If
you declare READER as an
external subprogram, you must declare READER as [GLOBAL] in the module
so that the main program can call it.

6-38

CHAPTER 7
INPUT AND OUTPUT

This chapter describes input and output (I/O) for PASCAL on TOPS-20.
PASCAL provides predefined procedures to perform input and output to
file variables.
These procedures are divided
into
the
following
categories:
General Procedures
•

OPEN -- associates

•

CLOSE -- closes a file

•

FIND -- performs direct access to file components

file with specified characteristics

Input Procedures
•

RESET -- opens a fiJe and prepares it for input

•

GET -- reads a file component into the file buffer variable

•

READ -- reads a file component into

•

READLN -- reads a line from a text file

specified variable

Output Procedures
•

REWRITE -- truncates a file to length zero
for output

•

PUT -- writes the contents of the file buffer
the specified file

•

WRITE -- writes specified values into a file

•

WRITELN -- writes a line into a text file

•

LINELIMIT -- terminates program execution after a
number of lines have been written into a text file

•

PAGE -- skips to the next page of a text file

In addition, you can use the predefined functions EOF
text files.
The following sections describe:
•

PASCAL file characteristics

•

PASCAL record formats

7-1

and

prepares
variable

and

it
into

specified

EOLN

with

INPUT AND OUTPUT
•

PASCAL input and output procedures

•

Terminal 1/0

The input and output procedures are

7.1

presente~

in alphabetical order.

FILE CHARACTERISTICS

This section describes the organization
accessing records.

records

and

methods

The term file organization app] ies to the way records are
physically
arranged
on a
storage device.
The term record cccess refers to the
method used
to
read
records
from
or write
records
to a
file,
regardless of
the
file's
organization.
A file's organization is
specified when the file is created
and
cannot be changed.
Record
access is specified each time the file is opened and can vary.

7.1.1

File Names

The file name indicates the system name of a file that is
represented
by a PASCAL file variable in an OPEN procedure (Section 7.7).
For the
file name, you can specify a character-string expression that contains
a
TOPS-20
file
specification or
a
logical name.
Apostrophes are
required to delimit a character-string constant or a logical name used
as a file name.

7.1.2

Logical Names

The TOPS-20 operating system provides the logical name mechanism as a
way of making
programs device and
file
independent.
If you use
logical names, your PASCAL program need
not
specify the
particular
device on which a file resides or the particular file that contains
data.
Specific devices and files can be defined at run time.
A logical name is an alphanumeric string that you specify in place of
a file specification.
Logical names provide great flexibility because
they can be associated not only with. a
complete
file
specification,
but also with a
device,
a device and a directory, or even another
logical name.
On TOPS-20 you can create a logical name
specification by means of the TOPS-20
program execution, you can associate the
with the file specifications appropriate

@DEFINE DATA:

and associate it with a
file
DEFINE command.
Thus, before
logical names in your program
to your needs, for example:

PS:<BENJAMIN)TEST.DAT.2

This command creates the logical names DATA:
and associates
it with
the
file specification PS:<BENJAMIN)TEST.DAT.2.
The system uses this
file specification when it encounters the logical name DATA:
during
program execution, for example:

OPEN (INDATA,

'DATA:', OLD);

7-2

INPUT AND OUTPUT
In executing
this PASCAL statement,
the system uses
the
file
specification PS:<BENJAMIN>TEST.DAT.2 for the logical name DATA:.
To
specify a different file when you execute
the
program again,
issue
another DEFINE command, for example:

@DEFINE DATA:

PS:<JENNIFER)REAl.DAT+7

This command associates the logical name DATA:
with a different
file
specification and replaces the previous logical name assignment.
The
OPEN statement above now refers to the
file PS:<JENNIFER>REAL.DAT.7.
For more
inform3tion about
the
use of logical names, refer to the
TOPS-20 User's Guide.

7.1.3

File Organization

PASCAL supports sequential
file organization.
Sequential
files
consist of records arranged in the order in which they are written to
the file.
The first record written is the first record in the
file;
the second record written is the second record in the file, and so on.
As a result, records can be added only at the end of the file.

7.1.4

Record Access

You specify record access mode as a parameter to the
PASCAL provides two ways of accessing records:
•

Sequential

•

Direct

OPEN

procedure.

If you select sequential access mode, records are written to or
read
from
the
file,
starting at the beginning and continuing through the
file one record after another.
Having sequential access
to a
file means
that you can read
a
particular
record
only after
reading all the records preceding it.
New records can be written only at the end of a file that is open
for
sequential access.
If you select direct access mode, you can specify the order
in which
records are accessed.
Each FIND procedure call must include the
relative record number indicating the record
to be read.
You can
directly access a
file only if it contains fixed-length records,
resides on disk, and is open for input (reading).

7.2

RECORD FORMATS

Records are stored in one of two formats:
•

Fixed length

•

Variable length

You can access fixed-length records in either sequential or direct
mode.
Variable-length records can be accessed only in sequential
mode.

7-3

I~PUT

7.2.1

AND OUTPUT

Fixed-Length Records

When you specify fixed-length records, you ~re
specifying
that all
records
in
the file contain the same number of bytes.
A file opened
for direct access must cont3in
fixed-length
records
to
allow the
record
location to be computed correctly.
All binary files (that is,
all files except TEXT files) must have fixed-length
records.
P~SCAL
does not support binary files with variable-length records.

7.2.2

Variable-Length Records

Variable-length records can contain any nUffiber of byt~s
up to
the
buffer
size specified when the file was opened.
TEXT files must have
variable-length records.
PASCAL does not
support TEXT
files
with
fixed-length rec~rds.

7.3

THE CLOSE PROCEDURE

The CLOSE procedure closes an open file.
Fo rma t.
CLOSE

(file variable);

where:
file variable

specifies the file to be closed.

Execut.ion of this procedure causes the system to close t.he file.
Each
file
is automatically closed upon exit from the procedure in which it
is declared, except those which have been dynamically allocated with
the
procedure NEW.
These files should be explicitly closed;
if not,
they are automatically closed when the program ends, or when they are
DISPOSED.
You can close only files that have been opened
explicitly with
the
OPEN procedure or
implicitly by the RESET or REWRITE procedure.
Therefore, you cannot close the predeclared file variables
INPUT and
OUTPUT.
Example

CLOSE (AlbulTls);
This procedure closes the file Albums to further access,
the file if it is internal to the current program.

7.4

and

deletes

THE FIND PROCEDURE

The FIND procedure positions a file pointer at a
in the file.
Fo rma t
FIND (file variable, integer expression);

7-4

specified

component

INPUT AND OUTPUT
where:
file variable

specifies a file that is open for direct
access.
The file must have fixed-length
records.

integer expression

specifies
the
positive
integer
expression
indicating
the component at
which
to
position
the
file.
The
component number must not be less than
or equal to zero.

The FIND procedure allows you to directly access the components of a
file.
You can use the FIND procedure to move forward or backward in a
file.
The file must be open for direct access.
That
is,
you must
have
specified
DIRECT
in
the
OPEN statement
for
that file.
In
addition, the file must have fixed-length records.
After execution of the FIND procedure, the file is positioned
at the
specified
component.
The
file buffer variable assumes the value of
the component, for example:
FIN D (A 1 b 1.1 III S, 4) ~

As the result of this statement, the file position moves to the fourth
component
in
the
file
Albums.
The
file
buffer variable Albumsassumes the value of the fourth component.
If you specify a component beyond
occurs.

the

end

the

file,

error

You can use the FIND procedure only when reading a file.
If the
file
is open for output (that is, with REWRITE), a call to FIND results in
a run-time error.
Example I
BEGIN
FIND (Albums, Current

+ 2);

END;

If the value of Current is 6, this statement causes the file
position
to move
to
the
eighth component.
The file buffer variable Albumsassumes the value of the component.
Example 2
BEGIN
FIN D (A 1 b U IT! S, C U l'r e n t . - 1 ) ;

END~

If the value of Current is 6, this statement causes the file
to move backward one component to the fifth component.

7-5

position

INPUT AND OUTPUT
7.5

THE GET PROCEDURE

The GET procedure reads the next component of a
buffer variable.

file

into

the

file

Fo rma t
GET (file variab1e);
where:
file variable

specifies the file to be read.

Before you use the GET procedure to read one or more file components,
you must have called
the RESET procedure to prepare the file for
reading (input).
RESET moves the file position to the first component
and assigns its value to the file buffer variable.
As a result of the GET procedure, the file position moves to the next
component of the file.
The file buffer variable takes on the value of
that component, for example:
r;: [ ~:) E T (B () 0 k ~:; ) Y
N P ItJ Y' pc: ::" li n 0 k ~:; ,., Y

GET (Buoks);

After execution of the RESET procedure,
the file buffer variable
Books
is set to
the first component of the file.
The assignment
statement assigns this value to the variable Newrec.
The GET
proce~ure
then assigns the value of the second component to Books
moving the file position to
the second component.
The next GET
procedure moves the file position to the third component, as shown in
Figure 7-1.
A

RESET
GET
(BOOKS) (BOOKS)

B.,'""'",

of File

. .I

EOF

• • t
I

RESET
GET
GET
(BOOKS) (BOOKS) (BOOKS)

FigurE~

7-1:

MR-S-3117-83

File position After GET

By repeatedly using GET statements, you can read sequentially
the file.

7-f)

through

INPUT AND OUTPUT
When you reach the end of the file and you request a
GET operation,
EOF automatically becomes TRUE, and the file buffer variable becomes
undefined.
When EOF is TRUE,
you cannot
use
the GET procedure.
P~SCAL signals a run-time error, and program execution is aborted.
Example

BEGIN
GET (Phones);

END;
This example reads the next component of the file Phones into the file
buffer variable Phones
EOF(Phones) must be FALSE;
if it is TRUE,
an error occurs.
A

•

7.6

THE LINELIMIT PROCEDURE

The LINELIMIT procedure terminates execution of the
program after
specified number of lines have been written into a text file.

Format
LINELIMIT

(file variable, n);

where:
file variable

specifies the text file to which
applies.

represents a
positive
integer
expression
indicating
the number of lines that can be
before
execution
the
file
written
to
terminates.

this

limit

When PASCAL initializes a text file, it specifies a large default line
limit.
You can override
this limit by calling LINELIMIT with your
desired value.
After the number of lines output to the
limit, program execution terminates.

file

has

reached

the

line

Example

BEGIN
LINELIMIT (Debts,100);

Execution of the program terminates after 100 lines have been
into the text file Debts.

7-7

written

INPUT AND OUTPUT
7.7

THE OPEN PROCEDURE

The OPEN procedure does not actually open a file but rather allows you
to
specify file attributes.
You cannot use OPEN on a file th~t has
had a RESET or REWRITE done, or on the predeclared file INPUT.
Fo rma t 1
OPEN (file variable [ ,file name E
[,history E
IT ,record length E
[,record-access-method E
[ ,record type E
[,carriage control E );
Format 2
OPEN

(FILE VARIABLE := file variable
[ ,FILE NAME := file name E
[ ,HISTORY := history status E
[ ,RECORD LENGTH := positive integer E
[ ,RECORD-ACCESS METHOD := record-access-mode
[ ,RECORD-TYPE :~ record type E
[ CARRIAGE CONTROL .= carriage control D
);

where:
file variabJe

specifies the PASCAL file variable associated
with
the
file.
You
cannot open
the
predeclared file variable INPUT.
(internal files)
This parameter is ignored for internal files.
The system creates a
unique name for each
internal file.
(external files)

file name

provides
information about
the
file
to
TOPS-20.
The file name can be a variable or
constant identifier defined
as type PACKED
ARRAY (l .• n] OF CHAR, or a file specification
enclosed
in
apostrophes
(for
example,
'PS:(MASELLA)BOOKS.DAT') .
If you omit the file name, PASCAL will
first
attempt
to
use the file variable identifier
as a logical
name.
If
that name
is not
defined,
PASCAL will use the defaults shown
in Table 7-1.

The file variable and the file name parameters designate the
file
to
be opened.
The remaining parameters specify attributes for the file
and are summarized in Table 7-2, in Section 7.7.5.
Except for
the
file variable name, all parameters are optional.
Any parameters you
specify, however, must be in the order shown above
unless you use
keywo rd syn tax.
You can specify either the value of the parameter or the
keyword
the value of the parameter.
You can also use a combination.

7-8

and

INPUT AND OUTPUT
To specify only the value without the keyword, place each parameter in
the
same order shown in the format.
If a particular parameter is not
used, then a comma may be
inserted.
PASCAL generates a
warning
message
if
the position of an unused parameter is not jndicated by a
comma.
However, the correct default for the missing parameter is used
in either case.
To use a keyword, specify the keyword and its associated value.
When
you
use
a
keyword, you have to specify only the parameters that are
used;
it is not necessary
to
insert ~
comma
to
indicate
unused
parameters.
Keyword
parameters can be placed in any order;
they do
not have to be in the same order as shown in the format.
You can also use a combination of values and
keywords with values.
However,
once
you
use
a
keyword
within the statement, subsequent
values must be associated with a keyword.
Table 7-1:

Default Values for TOPS-20 External File Specifications

-----------------------------------------------------------------------------------Element

Default

Device

Current user device

directory

User's current default directory

file type

Depends on usage:
Input to PASCAL compiler - PAS
Output from PASCAL compiler - REL
Input to LINK - REL
Output from SAVE command - EXE
Input to RUN command - EXE
Compiler source listing - LST
LINK map listing - MAP
Input to executing program - DAT
Output from executing program - DAT

gen

Input: highest existing generation
Output: highest existing generation plus 1

When compiling a PASCAL program, you need specify only the
if the file is:
•

Stored on the default device

•

Catalogued under the default directory name

•

A file type of PAS

file

name

If more than one file meets these conditions, the compiler chooses the
one with the highest generation number.
For example, assume that your default device
is PS:;
your default
directory is (CHEN>;
and you supply the following file specification
to the compiler:

@F'ASCAL
F'ASCAL>CIRCl.E
The compiler searches device PS:
in directory (CHEN>,
seeking
the
highest generation of CIRCLE.PAS.
The compiler then generates the
file CIRCLE.REL, stores it on device PS:
in directory (CHEN>,
and
assigns it a generation number
that is one higher than any other
generation of CIRCLE.REL currently in PS:(CHEN>.

8.3

CREATING A F'ROGRAM

The first step in creating a program is to design and plan it.
The
TOPS-20 PASCAL Primer describes the use of PASCAL for the novice
PASCAL programmer who
is
already
familiar
with
higher-level
programming
language concepts.
Many books exist that describe
programming techniques, methods, and algorithms.

8-3

USING PASCAL ON TOPS-20
After planning the progr~m, you use an editor to create a
file
that
contains
the
source
statements.
You use a text editor to create a
source file.
You can use EDIT, TV, or any other text editor to create
the source file.
For example, to create a PASCAL program that has the file name
EXAMPL
and a file type of PAS, you can issue the EDIT command as follows:
(~EDIT EXAMF'L... PAS ~~
Input: EXAMPL.PAS
O():lOO

If this is a new file, an additional message is displayed
indicating
that a
new file is being created.
Because the EDIT command does not
assume a file type, you must include the file type as part of the file
name.
The EDIT command
runs the TOPS-20 default editor EDIT.
The
line number (00100) prompt indicates that EDIT
is
ready to accept
input.
For
information on how to
use EDIT, see the TOPS-20 EDIT
User's Guide.
You can also use any other
editor
to
which you have access,
for
exampJe, the TV editor.
To use TV, you can either type TV to TOPS-20,
or you can define the logical name EDITOR:
to
be SYS:TV.EXE.
For
more
information about the use of TV, refer to the TOP§-20 TV Edit.or
Manual.
After the program is created and edited,

8.4

it is ready to be compiled.

COMPILING A PROGRAM

After creating a PASCAL source program, you compile
it.
At
compile
time,
you specify the source filers) and indicate any qualifiers you
wish to use.
Optionally, the compiler produces one or more object files, which are
input
to
LINK,
and
one or more listing files.
The listing files
contain source-code listings, information about compilation errors,
and optional items such as cross-reference listings.

8.4.1

The PASCAL Command

To compile a source program, specify the PASCAL command and press
the
RETURN key.
TOPS-20
then returns the PASCAL prompt, at which point
you specify the file name and any switches.
@PASCALG0
PASCAL)source-filename U/switch(es)TI

8-4

USING PASCAL ON TOPS-20
where:
source-filename

specifies the source file(s)
containing
the
program or module
to
be compiled.
If you
have one program split
into
several
source
files,
you can specify these source files at
the same time by separating
the
file
names
with a
plus sign (+).
If you specify more
than one source
file,
the
files
nre
concatenated and compiled as one program.

/swi tch (es)

indicates special processing to be
by the compiler.

performed

In many cases, the simpJest form of the PASCAL command
is sufficient
for
compilation.
For other
situations,
however,
PASCAL provides
compiler commands and switches to specify special processing.
PASCAL
compiler commands give special instructions to the compiler.
PASCAL
compiler switches modify the compilation of the program.
Section 8.4.2 describes the PASCAL compiler commands,
8.4.3 and 8.4.4 describe the PASCAL compiler switches.

8.4.2

and

Sections

PASCAL Compiler Commands

Table 8-2 lists the commands to the compiler.
Table 8-2:

PASCAL Compiler Commands

--------------------Command

Purpose

------------------------------/EXIT

Exits from the PASCAL compiler

/HELP

Displays a help messnge

/RUN:

Begins execution of the specified program

/TAKE:

Takes commands from the specified command file

-------------------------------/EXIT
The /EXIT command exits you from the compiler
that were opened by the compiler.

and

closes

all

files

/HELP
The /HELP command displays a help message.
/RUN:filespec
The /RUN:
command exits you from the compiler and begins execution of
the
specified
program.
Using
the /RUN:
command
is the same as
specifying the /EXIT command
to
the
compiler and
then
using
the
operating system command RUN to execute the specified program.

USING

P~SCAL

ON TOPS-20

/TAKE
The /TAKE command takes commands from the specified command file.
/TAKE command recognizes the default file type CMD.

The

Example 1
(~PAf:)CAL..

PASCAL)/TAKE:

PLAN

P{~SCAL)

Assume that the command file PLAN.CMD contains the following:
PLAN. PAS IN()FI...AG-··NON··-~JTANDAr~[I IL I ~JT I NG ~

The /TAKE:
command causes the contents of PLAN.CMD to be executed.
In
this example,
the
source fiJe PLAN.PAS is compiled;
display of
warning messages
for
nonstandard
features
is suppressed;
and a
listing
file
is generated.
Make sure the command file ends with a
carriage-return/line-feed.
After execution of the command
file,
you
can give another command to the compiler.
Example 2
PASCAL)SORTER.PAS
PASCAL)/RUN: LINK

*
The compiler compiles the
source
file SORTER. PAS,
and
the /RUN:
command
is then
used
to
run the LINK program.
The /RUN:
command
exits you from the compiler and
causes
the
LINK program
to begin
executing.
The asterisk (*) is the prompt displayed by LINK.
Example 3

@PASCAL
PASCAL:>AVER.F'AS
F'ASCAL:>/EX I T
@

The EXIT command exits you from the compiler and puts you
command level.

8.4.3

TOPS-20

PASCAL Compiler Switches

Table 8-3 lists the switches you can use with
the
PASCAL compiler.
You can specify the switches following the file name or in source code
comments.
This section describes the
effect of each switch on a
PASCAL program.

8-6

USING PASCAL ON TOPS-20
Table 8-3:

PASCAL Compiler Switches

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

Switch

Purpose

In
Source

Default

/ABORT

Causes the compiler
to
exist at the end of a
compilation that contains
errors

Off

/BINARyr[:filespecll

Produces a
file.

binary

object

/CHECK

Generates code to
for
various
conditions

check
error

Yes

/CREF or
/CROSS-REFERENCE

Produces a cross-reference
listing of identifiers

Yes

Off

/DEBUG

Produces
information in
the object file to be used
wi th PASDD'r

Off

/ERROR-LIMIT:n

Stops
compilation after
the specified number of
errors

/FLAG-NON-STANDARD

Issues
warning messages
for nonstandard features

Yes

/LISTING[[:filespecl]

Produces a source listing
during compilation

Yes

Off

/MACHINE-CODE

Lists generated assembly
language in source listing

Yes

Off

/NATIONAL

braces
off
Turns
comment characters

Yes

Off

/WARNINGS

Prints
diagnostics
warning-level errors

fo rYes

------------------------------------------------------/ABORT
The /ABORT switch causes the compiler to exit at the end of a
compilation that contains errors.
This is useful when used with the
/TAKE:
command.
The default is /NOABORT.
/BINARY [:filespecD
The /BINARY switch can be used when you want to specify
the object file.
The /BINARY switch has the form:
/BINARY [:filespecD

8-7

the

name

USING PASCAL ON TOPS-20
If you omit the file specification, the object file defaults
to
the
name of
the last source file, the default directory, and a file type
of REL.
You cannot specify this switch in the source code.
You can disable this switch to suppress object code, for example, when
you want to test only the source program for compilation errors.
The default is /BINARY.
/CHECK
The /CHECK switch directs the compiler to generate
co~e
to
perform
run-time checks.
This code checks for ilJegal assignments to sets and
subranges, out-of-range array indices and case labels, and
references
to NIL pointers.
The system issues an error message and terminates
execution if any of these conditions occur.
When this switch is disabled, the compiler does not generate code
run-time checks.
The default is /CHECK.

for

/CROSS-REFERENCE or /CREF
The /CROSS-REFERENCE switch produces a cross-reference listing of all
identifiers.
The compiler generates separate cross-references for
each procedure and function.
To get complete cross-reference listings
for
a
program,
the
switch must be in effect for all modules of the
program.
This switch
is
ignored
if no
listing
file
is being
generated.
The default is /NOCROSS-REFERENCE.
You can specify this switch in the source code.
Note,
however,
that
the
cross-reference
listing for a portion of a procedure or function
may be incomplete.
/DEBUG
The /DEBUG
switch
specifies
that
the
compiler
is
information that can be used with run-time debugging.

generate

The default is /NODEBUG.
/ERROR-LIMIT:n
The /ERROR-LIMIT switch terminates compilation after
the
specified
number of errors, excluding warning-level errors, have been detected.
The default
limit
is
30
errors.
If
this
switch
is disabled,
compilation continues
through
the
entire unit.
You cannot specify
this switch in the source code.
The default is /ERROR-LIMIT:30.
Note that, after finding 20 errors (including warning messages) on any
one source line, the compiler generates error 255, Too Many Errors On
This Source Line.
Compilation of the line continues, but no
further
error messages are printed for that line.
/FLAG-NON-STANDARD
print
The /FLAG-NON-STANDARD
switch
tells
the
compiler
to
program uses
warning-level
messages
at
each place where
the
nonstandard PASCAL features.

8-8

USING PASCAL ON TOPS-20
Nonstandard PASCAL features are the extensions to
the
proposed
ISO
standard
for
the PASCAL language that are incorporated in PASCAL-20.
Nonstandard features include VALUE declarations and the exponentiation
operator.
Appendix D lists all the extensions.
By default, /FLAG-NON-STANDARD is enabled.
/LISTING
The /LISTING switch produces a source listing file.

It has the form:

/LISTING [:fiJespecE
You can include a
file
specification
for
the
listing
file.
The
default
file
specification designates
the name of the first source
file, your default directory, and a file type of LST.
The compiler does not produce a
Jisting
file
in
interactive mode
unless you specify the /LISTING switch.
In batch mode, the compiler
produces a listing file by default.
In either case, the listing
file
is not automatically printed.
/MACHINE-CODE
The /MACHINE-CODE switch places in the listing file
of the object code generated by the compiler.
The compiler ignores
enabled.

this

switch

the

/LISTING

representation
switch

not

The default is /NOMACHINE-CODE.
/NATIONAL
The /NATIONAL switch causes the braces to have no speciaJ
meaning.
Therefore,
if you specify the /NATIONAL switch, you cannot use braces
as comment characters.
Instead, you must use (* *).
The default is /NONATIONAL.
jWARNINGS
The /WARNINGS switch directs
the compiler
to
messages in response to warning-level errors.

generate

diagnostic

By default,
/WARNINGS
is enabled.
A warning diagnostic message
indicates that the compiler has detected acceptable but unorthodox
syntax, or has performed some corrective action.
In either case,
unexpected
results
may occur.
To
suppress warning diagnostic
messages, disable this switch.
Note that messages generated when
the
/STANDARD switch is enabled appear even if /WARNINGS is disabled.

8.4.4

Specifying Switches in the Source Code

You can use switches in the source code to enable and disable special
processing during
compilation.
When specified in the source code,
switches have the form:
(*$switch + ,switch + , •••

;comment *)

8-9

USING PASCAL ON TOPS-20
The first character after the comment del imiter must be a doll~r
sign
(S);
the dollar sign cannot be preceded by a space.
Table 8-4 lists
the switches you can specify in your source program.
Note
that you
can optionally use
a I-character abbreviation for each switch.
The
abbreviation is simply the first character of the switch name,
except
for CROSS-REFERENCE, which has X for an abbreviation.
Table 8-4:

Source Switches

Abbreviation

Full

Command-Line Switch

CHECK

LTST

LIST

MACHINE-CODE

NATIONAL

STANDARD

FLAG-NaN-STANDARD

WARNINGS

CROSS-REFERENCE

CREF

To enable a switch,
specify a
plus sign
(+)
after
its name or
abbreviation.
To disable a switch, specify a minus sign (-) after its
name or abbreviation.
You can specify any number of switches.
You
can also include a text comment after the switches, separated from the
list of switches by a semicolon.
When specified in the source code, the LIST switch cannot contain a
file
specification.
The listing file has the default specification
described above.
For example, to generate check code
for only one procedure
in a
program, enable the CHECK switch before the procedure declaration, and
disable it at the end of the procedure, as follows:

<*$Ct ; enable CHECK for TEST1
PROCEDURE TEST1;

onl~

END
<*$C-;disable CHECK *)
Command line switches override source-code switches.
If, for example,
the source code specifies NOWARNINGS,
but you type /WARN on the
command line, warning messages will be generated.
NOTE
When specifying the NATIONAL switch in
the
source
code,
always
use
the
parentheses/asterisks combination
(* *)
and not braces { }.

8-10

USING PASCAL ON TOPS-20
8.4.5

Specifying Output Files

The PASCAL compiler can produce object files and
listing
files,
as
well
as compile
the source code.
You can control the production of
these files with the addition of various file names and
switches on
the PASCAL command line.
PASCAL produces an object file automatically, taking the name from the
source file and assigning it the file type REL.
To change the name of
the object file, specify the /BINARY switch with a file name.
To produce a listing file, you must specify the /LISTING switch on the
PASCAL command
line.
You have the option of giving a file name with
the /LISTING switch or taking the default, which is the name of the
source file
and the file type LST.
Note, however, if you run PASCAL
from a batch control file, you automatically receive a
listing
file.
In
this case, to suppress the creation of a listing file, specify the
/NOLISTING switch in the batch control file.
During the early stages of program development, it is often useful
to
suppress the
production of object files until your source program
compiles without error.
To suppress the production of an object file,
specify the /NOBINARY switch along with the source file.
You can specify more
than one source
file at a
time,
to be
concatenated and
compiled.
When specifying multiple source files,
separate each one with a plus sign (+).
Although you may specify more
than one source file, you still receive one object file to load for
execution.
By default, the object file
produced
from concatenated
source files has the name of the last source file on the command line.
All other file specification attributes
(device,
directory,
and
so
forth) assume the default attributes.
Example 1

@PASCAL
PASCAL)XXX+YYY+ZZZ
Source files XXX.PAS,
YYY.PAS,
and
ZZZ.PAS are concatenated and
compiled as one
file,
producing an object file named ZZZ.REL.
In
batch mode, this command also produces the listing file ZZZ.LST.
Example 2

@PASCAL
PASCAL><S.GRAVES)MNP/LISTING
The source file MNP.PAS in directory <S.GRAVES> is compiled, producing
an object file named MNP.REL and a listing file name~ MNP.LST.
The
compiler places the object and listing files in the default directory.

8.4.6

Compiler Listing Format

When you request a listing
file
(by specifying
the /LIST switch,
PASCAL produces a compiler listing.
This section explains the format
of the compiler listing illustrated in Figure 8-1.

8-11

•

AVERAGE SCORE
SOURCE LISTING
LINE
NUMBERS

ADDRESS
PROC
DATA INST NO

100

000041

% PAS456

ex>
I
~

17-Aug-1983
CBL20:<MASELLA)AVER.PAS

200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1800

LEVEL
PROC STMT

Nonstandard Pascal: n$n OR
2 000215
0
1
0
3 000215
0
1
4 000220
1
0
5 000221
0
0
6
0
0
0400000
7
400012
0
1
8
400033
0
1
9
400034
0
1
10
400052
0
1
11
400061
0
1
12
400061
0
2
13
400076
0
14
400101
0
15
400104
0
16
400105
0
17
400113
0
18
400142
0
19
400172
0

11:23:32

PASCAL-20 1(611)

Page

o
(1 )

STATEMENT.
PROGRAM Averaqe Score (INPUT,OUTPUT);

CEr'456---.

***

BEGIN
To ta 1 : = 0;
Count := 0;
WRITF,LN ('F,nter your scores. When done, type CTRL/Z.');
WHILE NOT EOF no
BEGIN
READLN (SCORE);
Total ;= Scarp + To tell ;
Count := Count + 1;
End;
AverageScore ::. Tolal / Count; (*to produce real results"")
WRITELN ('The average score is: " AverageScore:4:1);
END.

1.0

Z
GJ
'tl

Cf}

'tl

1 • ') 1 sec 0 nd s

( 7 5 5 1 in e s pe r min ute) .

17-Aug-1983
CBL20:<MASELLA)AVER.PAS

Cf)

I
10

11:23:32

PASCAL-20 1(1111)
(] )

INSTRUCTION

ADDRESS

OPCODE

00 a 00
16 0 00
17 0 00
16 0 00
05 0 00
17 0 00
17 0 00
17 0 00
00 0 00
05 0 00

000000
400012
000000*
400013
000002'
400014
000000 0400015
000041'
400016
000005
400017
000000*
400020
777777
400021
000017
400022
000041'
400023

JFCL
JSP
G) MOVEM
SETZ
MOVEI
PUSH
PUSHJ
ADJSP
MOVE
MOVEI

255
265
202
(t400
201
261
260
105
200
201

t-3

Active options at end of compilation:
NODEBUG,STANDARD,LIST,CHECK,WARNINGS,CROSS REFERENCF"G)
MACHINE_CODE, OBJECT, ERROR_LP'lIT
30

LINE

1 Nonstandard feature
Last error (warning) on 1 ine

AVERAGE SCORF.
GENERATED CODE

o
C)

t'V

Compilation time:

1 ===)

in identifier in AVERAGE SCOREG)
VAR
Score, Total, \.ount : INTEGER;
AverageScore : REAL;

OPERAND{S)
00
PASLD%
ACH,OO
AC17,02
C)AC16 ,
ACOS,000041
AC]7,05
AC17,OO
INP%IN
AC17,777777
ACOO,l?
AC05,000041

AVERAGE SCORE
CROSS REFERENCE

17-Aug-1983
CBL20:<MASELLA)AVER.PAS
4

AVERAGESCORE
AVERAGE SCORE
COUNT
INPUT •
OUTPUT
SCORE
TOTAL

11:23:32

PASCAL-20

1(~11)

Page

(1)

12
7

:3
3

GLOBALLY DEFINED IDENTIFIERS~e

AVERAGE SCORE
GENERATED CODE

17-Aug-1983

] 1:23:32

PASCAL-20

CBL20:<~ASELLA)AVER.PAS

1(~11)

(1)

c:::

LINE

(X)

I
I-'

INSTRUCTION

ADDRESS

OPCODE

523214520302
7331362607 Hi
625016361736
713124064746
351000000000
.427356462744
203635772744
203474367744
627465620256
643135620310
677354526100
723636062500
416512241)136
551340000000
000000000000

400153
400154
400155
400151)
400157.
40011)0
400161
400162
40011'13
400164
400165
40016')
400167
400170
400171

CONS'T'ANT
CONSTANT
CONSTANT
CONSTANT
CONSTANT
CONSTANT
CONSTANT
CONSTANT
CONSTANT
CONSTANT
CONSTANT
CONSTANT
CONSTANT
CONSTANT
CONSTANT

Figure 8-1:

Compiler Listing Format

OPERAND(S)

"0
;t>o
U1

o
Z

t-3

'"'0
U1
I

N
MR-S-3122-83

USING PASCAL ON TOPS-20
The compiler listing
sections:
•

Figure

8-1

cont~jns

Source-code listing - When you request a
source code is listed by default.

the

following

listing

three

file,

the

1 isting,
Machine-code listing - To generate the machine-code
or
the
you
must
specify
the
/MACHINE-CODE
switch,
MACHINE-CODE + source switch.

•

Cross-reference listing - To gener~te cross-references
for
all
identifiers
used
in
the program, you must specify the
/CREF switch.

The numbers throughout this section are keyed to the numbers in Figure
8 -1.
TITLE LINE - Each page of the listing
cont~ins
a
title
line.
The
title
line
lists
the module n~me
the nate and time of the
compilation
the PASCAL compiler name ~nd version number
the
listing page number
and the file specification of the source f il e

o·

8.4.6.1

Source-Code Listing - The
lines of
the
source code are
printed in the source-code listing.
In addition, the listing contains
the following information pertaining to the source code:

0 -

•

Editor line numbers
If you created or edited the source
file
with an editor that automatically inserts line numbers,
these line numbers appear
in
the
leftmost column of
the
source-code
listing.
Editor line numbers are irrelevant to
the PASCAL compiler.
Note these are not the line numbers you
specify to PASDDT for debugging purposes.

•

Line numbers
The compiler assigns unique
1 ine numbers
to
the
source lines
in a
PASCAL module.
The symbolic
traceback that is printed
if your
program encounters an
exception at
run
time
refers to these line numbers.
Note
these are the line numbers you use when working with PASDDT.
Refer to Chapter 9 for information concerning the debugger.

•

Address
The listing includes the octal address
instruction on that line of source code.

•

Procedure number C) - PASCAL numbers each procedure
in
the
listing
file as it progresses down the source code, starting
with 0 for the main program.

•

Procedure level
Each line that contains a
declaration
lists
th~
procedure level
of that declaration.
Procedure
level 1 indicates declarations in the outermost block.
The
procedure-level
number
increases by one for each nesting
level of functions or procedures.

0 -

8-14

the

USING PASCAL ON TOPS-20
•

Statement level
The] isting
specifies a
statement
level
for
each
line of source code after the first BEGIN
delimiter.
The statement level starts at a and increases by
1 for
each nesting
level of PASCAL structured statements.
Specifically, the CASE statement, the REPEAT statement,
Rnd
the
BEGIN/END block
increase
the
statement level.
The
statement level of a comment is
the
same as
that of
the
statement that follows it.

ERRORS and WARNINGS -- The source-code listing includes information on
any errors or warnings detected by the compiler.
The actual message
is printed beneath the error in the source code.
In addition,
the
following items appear in the listing:
•

A circumflex (A) that points to the character position in the
line where the error was detected t)

•

A numeric code, following the circumflex, that specifies
the
particular error
On the following lines of the source
listing, the compiler prints the
text
that corresponds
to
each numeric code

•

the
The line number 'where the error was detected
and
line number of
the previous line containing an error
You can use these error
line numbers
to
trace
the error
diagnostics back through the source listing.

G).
G)

SUMMARY -- At the end of the source listing, the compiler
tells you
how many errors or warnings were generated (if any), with the source
line number where the last one occurred
The compiler prints the
status of all
the compilation optjons
and how much time was
required for the compilation G)

CD.

8.4.6.2 Machine-Code Listing - The machine-code listing (if requested
with
the /MACHINE-CODE switch) follows the source-code listing.
The
machine-code listing contains:

CD -

•

Source line number
A source line number marks the
first
object instruction that the compiler generated for the
first PASCAL statement on that source line.

•

Octal representation of instruction _
representation of the object instruction.

•

Address
instruction.

•

Opcode
instruction.

•

Operand
This contains the mnemonic
and address field for the operation.

This

the

relocatable

This is the mnemonic

G -

This is the

operation

address

the

code

for

the

accumulator

For more
information on machine-code
instructions,
DECsystem-lO /DECS,xSTEM-20. Processor Reference Manua 1.

8-15

octal

refer

field
to

the

USING PASCAL ON TOPS-20
8.4.6.3
Cross-Reference Listing - The cross-reference
listing
(if
requested
with
the /CREF
switch)
appears after
the machine-code
listing.
It contains two s~ctions:

fa - This section lists all

•

User-specified identifiers
identifiers you declared.

•

Globally-defined identifiers
This
section
lists
PASCAL predefined identifiers that the program uses.

CD -

the
the

Each line of the cross-reference listing contains an identifier and
a
list of the source lines where the identifier is used
The first
line number indicates where the identifier
is declared.
Predefined
identifiers are
listed as if they were declared on ] in(~ 0
The
cross-reference listing does not specify pointer type identifiers that
are used before they are declared.

fa.

8.5

LOADING A PROGRAM

After compiling your PASCAL program, you
load
the
object module(s)
with
the
LINK program to produce an executable image file.
Loading
resolves all references in the object code and
establishes absolute
addresses
for
symbolic locations.
This section describes the use of
the LOAD command.

8.5.1

The LOAD Command

To load an object module, specify the LOAD command
general form:
LOAD filename

the

following

[/switch(es) .•• ]

where:
filename

specifies the input object file to be loaded.
The
file
must be a REL file, created with the PASCAL
compiler (or some other translator).

/swi tch (es)

specify input file options.

The file specification must be typed on the
same
line as
the
LOAD
command.
If the file specification does not fit on one line, you can
continue typing without pressing the RETURN key.
The LOAD command runs the LINK program,
which
reads
the
REL
file
specified
on
the
command
line and loads it into memory.
When LINK
exits, it leaves your program in memory, ready to be SAVEd as an
EXE
file or to be STARTed.
Th~

LOAD command can do a COMPILE command before running LINK.
If you
specify a
file
name to the LOAD command for which there is a source
file but no REL
file,
the
LOAD command
automatically performs a
COMPILE command first.

See the TOPS-20 -------Commands Reference
Manual
-------LOAD command.

8-16

for more information on

the

USING PASCAL ON TOPS-20
8.6

EXECUTING A PROGRAM

After you have compiled and linked your program, you can
execute
the
object file with the START command;
or you can save the file with the
SAVE command, and then execute it with the RUN command.
You ~an save time by using the EXECUTE command.
The
EXECUTE
command
acts by performing
a
LOAD command (which may start with a COMPILE
command) followed immediately by a
START command.
So,
instead
of
performing
severa]
separate steps of running the compiler, compiling
your program, loading your program, and starting your program, you can
compile, load, and start your program with a singJe command:
(~EXECUTE

f i 1 ename

Using the EXECUTE command does not save the executable
image of
the
object
file.
The SAVE command stores a copy of the executabJe image
in an executable file.
The default file type that is created is
EXE.
After
an
executable
image is saved, you can execute it with the RUN
command.
The SAVE and RUN commands have the form:
(~~~AVE

@f~UN

f i 1 ename ~
f i 1 ename ~

You must specify the file name;
default values are
omit optional
elements of the file specification.
type is EXE.

8.7

appl ied
if
you
Th e d e fa u] t f i 1 e

EXAMPLES

Example 1
(~F'ASCAL..

PASCAL> ~:;()HTEf~ /NOFLAG/L.. I ST I NG
PASCAI...:>/EXIT
@EXE SnRTEf~
f i] e
This example uses the
PASCAL
command
to
compile
the
source
SORTER.
The /NOFLAG
switch
prevents display of messages about
switch
nonstandard PASCAL statements in the
program;
the
/LISTING
generates a .LST file of the compilation in your disk area.

Example 2

@.PASCAL.
PASCAL:>MERGER/DEBUG
F'ASCAI, . >/EX I T
@.LOAD MERGER/DEBUG
@DEBUG MERGER
The PASCAL command is used to compile the program with
the debugger,
PASDDT.
The DEBUG
command is used to load the compiled program and
the PASCAL debugger automatically.
The DEBUG
command
executes
the
program with PASDDT.

8-17

USTNG PASCAL ON TOPS-20
Example 3
frl LOAD PLAN ICOMF'I LE / Cf~LF
@PFUNT PLAN. L.S"J
(~ !:)TAJ:~T

The LOAD command first compiles the program because of the /COMPILE
The
switch,
then,
if compilation js successful, loads the program.
c ross- refe renee
/CREF switch generates
a
listing
file
with
information.
The PRINT command prints the listing. The program is
then executed with the START command.

8-18

CHAPTER 9
PASDDT: THE PASCAL-20 DEBUGGER

The PASCAL-20 debugger, P~SDDT, provides
the means
to monitor
and
modify the
execution of a PASCAL program.
PASDDT provides symbolic
debugging capabilities that allow you to read and
modify the
values
associated with variables by referring
to
the PASCAL identifiers
within your program.

9.1

RUNNING PASDDT

To use the debugger,
you must
program.
First,
compile
the
switch in the command string:

compile and
load
PASDDT with
the
source program and include the /DEBUG

@PASCAL
PASCAL>filename/DEBUG
PASCAL>/EXIT
@

Then, load the program along
command:

with

PASDDT

specifying

the

The DEBUG command loads and starts the program
with the debugger, PASDDT.

currently

DEBUG

@DEBUG filename
memory,

If you want more control over where the debugger
is
placed,
or
you
have
the
need for more options, you can use the TOPS-20 LINK program
directly.
Refer to the LINK Reference Manual for TOPS-20 information.
When running PASDDT, the source file with the extension PAS should
be
located
in
the
same directory as the EXE version of the file.
This
should also be the directory to which you are connected
when
running
PASDDT.
If
you have
specified a file with the %INCLUDE directive,
PASDDT looks for a file with the same name.

9.2

USING SYMBOLIC VALUES

Symbolic values are the identifiers defined within the source program.
PASDDT allows access
to
identifiers available only in the current
scope, and performs recognition on these identifiers.
You can specify a location in the source code by using the line number
shown
in
the listing file created when the program is compiled.
See
Section 8.4.7.

9-1

PASDDT: THE PASCAL-20 DEBUGGER
9.3

SCOPE

Scope is the range within the program in which a specific definition
of an identifier exists.
The scope corresponds to the part of the
program in which the identifier can be used.
Figure 9-1 shows the
scope of three variables in program Modules.
The scope of the
identifier A (a global variable) is the entire program. The scope of
variable B, declared
in procedure Outer, is the entire procedure,
including procedure Inner. The scope of variable C is limited to
procedure Inner.
SCOPE
A

PROGRAM Modu 1 (;~S (I NF'UT, OUTPUT);
VAR A : INTEGER := 0;
B

F' t, 0 CE D LJ F~ E
VAF~

B :

(J 1..1 t f!' T' ;

I NTEGEF~;

PROCEDUHE Inner,
VAH C
INTEGEf<;
BEGIN (*beSin Inner*)

END; (*end I nne T'* )
BEGIN
(*besin Outf.·~r*)

Figure 9-1:

Scope

PASDDT uses the concept of dynamic scnpe
and
their values. Dynamic scope is the
of the program;
the dynamic scope refers
particular identifier is being used.
procedure (a procedure that calls itself)
same identifier.

when accessing
identifiers
scope in relation to the use
to the level at which a
For example, a
recursive
has multiple scopes for the

PASDDT represents the scope of an identifier with a positive integer.
Global data has a scope of 1.
Each nested level from that has a scope
of one greater than that of the level from which it is called.
The
following example is a recursive function that calculates factorials.

9-2

PASDDT: THE

PASCAL-~O

DEBUGGER

PF<OGI:<AM Ca 1 cu 1 ate (I NPUT, OUTPUT) ,
VAR

Answer~

Num : INTEGER;

FUNCTION Factorial (Number
BEGIN
WHILE Number > 0 DO
F i3 C tor i a 1. : ::" N1,,1 IT! b E\ T'

BEGIN
Answer

INTEGER)

INTEGEF< ,

* F act 0 r :i. a 1 ( NU IT! tl f.~ Y' .... 1),

:= Factorial(Num);

ENDt

When Factorial is called from the main program, the value of Num
is
passed
to Number.
Until Number is equal to 0, the function Factorial
continues to call itself;
the value of Number
is decremented
by 1
each
time
the
function calJs itself.
Assume that the value 3 is
passed to Number.
Number then has the value of 3, then of 2, then of
1, then of O.
Each call to itself represents another level of dynamic
scope.
Number has a scope of 1 the first time it is called;
a
scope
of 2 the second time it is called, and so on.

9.4

PASDDT COMMANDS

The following sections describe each of the
default radix for all purposes is decimal.

PASDDT

commands.

The

PASDDT uses the facility of recognition and guidewords.
Recognition
permits you to
type
enough of the word or identifier to be unique.
For example, you can type CL to specify the CLEAR command.
You can
also
use
recognition with identifiers you have used in the program.
For example, if you have defined the identifiers NewList and NewTABLE,
you
could
use NewL and NewT,
respectively,
to specify these
identifiers.
PASDDT also provides guidewords when you press the ESC key.
The guide
words indicate what you should enter next.
In the following sections,
guidewords are displayed in parentheses.
You do not need to type
the
guidewords or the surrounding parentheses in the syntax.
In addition,
PASDDT displays the options that you can enter when you type a
question mark
(:').
For more
information about
recognition and
guidewords, refer to the TOPS-20 User's Guide.

9.4.1

ASSIGN

ASSIGN assigns a value or virtual address to the specified identifier.
The format is:
ASSIGN (VARIABLE or ADDRESS) {user identifier}
octal address

(:=)

{constant
}
octal value

where:
user identifier

is a variable name in the active scope.

octal address

is a virtual address in the user program.

9-3

PASDDT: THE PASCAL-?O DEBUGGER
constant

is a value with a
simple data
type
INTEGER,
REAL,
CHAR,
BOOLEAN,
DOUBLE,
user-defined enumerated type.

octal value

is the value
address.

placed

the

of
or

octal

Example

PASDDT)ASSIGN New_Int

(:=)

This example assigns the value of 20 to user identifier New Int.

9.4.2

BREAK

BREAK sets a new breakpoint at a
specified
location or resets an
existing
breakpoint.
During a debugging session, when the program
reaches the location specified, it stops execution, thus allowing
you
to perform debugging
operations.
You can set a maximum of 20
breakpoints.
To set a new
command:
BREAK

breakpoint,

(AT)

use

line number

the

[(NAME)

following

format

the

BREAK

break identifierD

where:
line number

is a line number corresponding to a
line
in
the source code.
These
line numbers are
found in the listing file generated when you
compile the program.

break identifier

is the
name
you
associate
with
the
breakpoint.
Each breakpoint identifier can
contain any number of characters,
but the
first
nine characters must be unique.
The
break identifier can contain all
characters
except an underscore ( ) or dollar sign ($).

Note that you cannot set a breakpoint in the declaration section of a
program.
If you do, PASDDT sends you an error message and does not
set the breakpoint.
To change the status of a breakpoint from CLEAR (Section
the following format of the BREAK command:
BREAK

9.4.3),

use

(AT) {line number
}
break identifier

where:
line number

is the line number of the breakpoint that you
want to reactivate.

break identifier

is the name of the
reactivate.

breakpoint

you

want

Note that you can find out the status of breakpoints by displaying
them with
the DISPLAY command.
See Section 9.4.4 for information on
the DISPLAY command.

9-4

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

PASDDT: THE PASCAL-20 DEBUGGER
The following

rules apply to the use of BREAK:

If you specify a
following happens:
•

break

identifier,

but

not

line

If there is already a breakpoint with this
is used.

number,

the

identifier,

this

breakpoin~

•

If there is no breakpoint
occurs.

If you specify a
following happens:

line

number,

with
but

not

assigned

this

identifier,

error

break

identifier,

the

for

this

location,

that

•

If a
breakpoint
is
breakpoint is used.

•

If a breakpoint has not been assigned, a
established with the identifier NO NAME.

If the command specifies both a line number and
the following happens:

new
a

breakpoint

break

identifier,

•

If a breakpoint has already been assigned for
this
location
with
this
identifier,
the previously defined breakpoint is
used.

•

If a breakpoint has already been assigned for
with a different identifier, an error occurs.

•

If a breakpoint has already been assigned at another location
with
this identifier, and no breakpoint has been set at this
location, PASDDT prints a warning message.
PASDDT then asks
if you want
to override
the address associated with that
breakpoint.
If so, the previous breakpoint is discarded, and
a
new breakpoint
is declared with the given identifier and
location.

this

Note that you must use the PROCEED command after each break
program to continue.
PROCEED is described in Section 9.4.7.

location

for

Note the following example:

PASDDT)break 10 first
PASDDT)displa~ break
BREAK POINT(S):
Name
= FIRST
Break Address
000000400044
Line Number
10
Status
BREAK
PASDDT)proceed
»BREAK:FIRST
PC AT VIRTUAL ADDRESS
00000040044
PASCAL LINE NUMBER
10
LINE 9:
Count:= O~
LINE 10:
WRITELN ('Enter ~our scores. When done,
LINE 11:
WHILE NOT EOF DO
PASDDT)

9-5

t~pe

CTRL/Z.');

the

PASDDT: THE PASCAL-20 DEBUGGER
As shown in the previous example, PASDDT automatic~lly displays three
lines of source code at a breakpoint.
See the SET command, Section
9.4.9, for details of controlling the display of source code when a
breakpoint is reached.

9.4.3

CLEAR

CLEAR turns off breakpoints and tracepoints set
TRACE commands.
The format is:
CLEAR (AT)

jline number
}
\break identifier

with

the

BREAK

and

[(NAME) bre~k identifier]

whe re::
line number

is the line number in the source code
the breakpoint or tracepoint is set.

break identifier

is the name of the breakpoint or
associated with the line number.

where

tracepoint

You can specify either both the line number and associated
identifier, just the line number, or just the bre~k identifier.

break

The CLEAR command does not delete the breakpoint or tracepoint;
the
CLEAR command merely turns it off.
Also, although you have cleared a
breakpoint or tracepoint, it is still counted in the total 20 allowed.
To delete a breakpoint or tracepoint, use the REMOVE command, Section
9.4.8.
If you want to check the status of
breakpoints
and
tracepoints, use the DISPLAY command, Section 9.4.4.
Note the following example:
PASDDT:::bT'€:.'ak 11 onp
PASDDT)break 12 two
F' A S [t D T :::.0.1 :i. S p 1. a'.:l b rea k
BREAK POINT (~:» :
Nam'f.?
:= ONE
Break Add T'(-::'ss
Lin,:? Numb<::.' Y'

Status
Nc31T'E'

000000400062
:l1
I-:lF<LAt(

::.:: TWO

Break Acid T'€~SS
Line NlJlTlbe r
Statu~:;

000000400071
:t2
BRE(.lt(

PASDDT:::clear 11
PASDDT:::di SF' la~:I break
BREAI\ PO! NT ( S) :
:::: ONE
Name
B real>:. Add T'€:'S~:;
Line Number
Status

000000400062
11

Name
:::: TWO
Break Addrpss
Line Number
Status

000000400071
12

OFF

BRE/~K

9-tS

PASDDT: THE PASCAL-20 DEBUGGER
PASDDT)clear two
PASDDT)displaY break

PO I NT ( [» !

m~EttK

NalYt(·:·~
:::: ONE
B r f? a k A (! d T' f:~ ~; !:;
Line Nl..IlTtb€·~ l'

OO()()0040()062
1:1.

OFr

Status
~::
TWO
NalTtf.~
B T'E~ak Add rE.'S!:;
L.. i ne NUlYlt:.\f.-~ T'

000000400()71
1 ':)

OrF

L,t.atu5
PASDDT> t,l T'f·~ak 12
PASDDT>disp}ay break

BREAK F'DINT(S):
Na"ff.·~
:::: ONE
Break Add T'E.'!;;!:;
Line NUITtbe T'

000000400062
11

OFF

Status
Name
:::: TWO
BT'eak Addre!:;s
I... i ne Numbe T'

000000400071
12

Status
PASDDT) e;.~

9.4.4

BREAK

DISPLAY

The DISPLAY command has three functions:
1.

To show the status of breakpoints and tracepoints

To show the calling sequence of procedures and functions

To show a line (or range of lines)

To DISPLAY information about
following format:

breakpoints

DISPLAY B [REAK-IDENTIFIERD

of the source code
and

tracepoints,

use

the

[break identifierD

where:
B [REAK-IDENTIFIERD

indicates that you
want
breakpoints and tracepoints.

break identifier

is the name of a particular breakpoint or
tracepoint.
If you do not specify a break
identifier,
PASDDT displays
information
about all of them.

display

The information you receive includes the name of the breakpoint,
the
line number of the breakpoint, and the status of the breakpoint (ON or
OFF) •

9-7

PASDDT: THE PASCAL-20 DEBUGGER
Note the following example:

F't-lf; VI:! T>f."j:i ~:;F' 1. d ':1

r: F~ E f~ 1< POI NT ( ~:;) :
Name

LI r E' aV

FIRS1'

I: Y' f~ d k Ad (I l' f~ ~:; ':;
I... :i, f"tf~ Nl..llTil.'.l(':'~ Y'
~:; t, d t I..J ~:;

Name

o() 0 () 0 0 0 () () :5 ') ::,:j
.1.9
Bf~EAK

SECOND

:0 r E' a k f1 cli.".i Y' f~ ''', ':;
I... i. n e N UITI t . . C~ r

O()()O O()OOO 4 ~:i 2
23

~:;t,atl,I<:;

THAC[

= THIRD

N~m0

n r'eak
L i f"t(,~,

Ad(1 T'f?~:;S

O()()O()()0006:1.?
37

NUITI~:.le T'

PASDDT>displa~

break second

POINT (~;) :

BF,EAK

Name
= SECOND
Break Address
000000000452
I... i n f:~ N u mb c,~ T'
23
Status
TRACE

To DISPLAY the calling sequence of procedures and functions,
following format:
DISPLAY

(OPTION)

1 [NVOCA1'ION-STACK]

use

{integer} [{/STATIC
*
/DYNAMIC

the

where:
I [NVOCATION-STACK]

indicates that you want to display
calling sequence of your routines.

the

integer

is a
decimal
number
representing
many routine calls to display.

how

routine
indicates
that you want all
to
the main
calls
displayed
back
program.

/STATIC

causes the static level of the
routines
to be displayed as they were defined in
the user program.

/DYNAMIC

causes every invocation of a routine
be displayed.
This is the default.

The information you receive includes the calling sequence, the static
level of the procedure or function, and the address at the time of the
display.

9-8

PASDDT: THE PASCAL-20 DEBUGGER
Note the following eXAmple:

PASDDT)dispIBY invocation-stack */static
~:;

PI?OCEDUI:<E NAME

Gf< 0 CE f< Y .... D 11... 1...

PAf:;DDT>d i !:;F']. a\~ i nvoc

TAT J C I... E V[ I...

ADDr<Ff:;~:;

OOO:L40

:I.

*1 (l':~nalTl i c

~:;TATIC

r-f<OCE DUf<[ NAt'1E
C; F< [) C [ F< Y.... n J 1...1...

I... EVEL..

ADDf~'ES~:;

O()O:l.40

PA~:;DDT)

To DISPLAY lines of
following format:
DISPLAY

source

code

(OPTTON) S [OURCE-LINEn

the

current

line number

scope,

use

the

[/RANGE:integern

where:
S [OURCE-LINE n

indicates that you want to display a ] ine
the source code.

line number

is an integer
representing
a
line
in
the
source code.
You will find these numbers in
the listing file.

/RANGE:

is an optional value indicating that you want
to display more than one line of source code.

integer

is a decimal number indicating how many lines
of source code to display.

DISPLAY prints the source code of the line number you specify and
other lines included in the range.

any

Note the following example:
PrY:lDDT >d i Sp 1 a~:j ~)C)l.1 T'Cf? ~)I T'an9f?: 10
LINE ::.=;:
YP~;;, ... ND ::::
(YP~:;
, No),

LINE 6:
LINE 7:
I... INE B!

(* Defines data type Yes_No
with valups Yes and No *)

VAR

LINE 9:
LINE :LO:
LINE :L1:
LINE :1.2:

LINE 13:

ItplTI ...YT'ic(", , Total,
C () 1..1 F' [) n .. _A III C) U n t
An!:> : Yes .... Nc);

<*Declares three real variab:l.ps*)
<* Declares a variable, An~; ¥ of t~IPP
Yes .. _No

~:;ubt Dta:l.

CouF'ons: REAL.. : :=: O. ();

LINE :J.4:
PAf~DDT>

9.4.5

EXIT

EXIT halts execution of the program.
EXIT

9-9

The format is:

PASDDT: THE PASCAL-20 DEBUGGER
Note the following

ex~mp]e:

F' It.) S nn T:> F X I T
C F:' U t :i. IYI C-' :
U!

() () • :I. <:.)

J. d F' ~:; (.:.) (1 t i ITI P :

[

() 3 • :3"7

This exits you from PASDDT an~ puts you at TOPS-20 command level.
CPU time and elapsed time are ~lso displ~yed.

9.4.6

The

HELP

HELP displays a help message.
HELP

(with PASDDT)

The format is:

[command name]

Note the following example:

[HELP::!

H [ I... F'

[ C (] HlITI a n ,j n a ITI (.:.~

T h i. ~:;

C () ITIIT!

l ~~ F' f:~

wit h •

0 T'

T' P t u T' n :I

and a:l. low ~:; t h f~
foIl D I-J [~d

HE l.. F' /
T \:~ P e

i:l

1..1 ~:;

e l' t n

b ~~

the

G I.J est ion IT! a T' k

~:1 f.~ t

(/? /)

h (~) 1. F' wit h

C () !TIm

a n (.1

t (] Sf:) e

~~ C)

PAn 1:1 11 T •

u wan t h f;;> 1. p

t h (-?

co !TIm and s

that are available.
PA~)DDT> P;':

This displays help messages about all of the PASDDT commands.

9.4.7

PROCEED

PROCEED initiates or continues execution of
the
program until
a
breakpoint is reached or the user program terminates.
The format is:
PROCEED (with USER Program)
Example 1
i~ PASCAL.
PASCAL>TEST.PAS/DEBUG

FASCAL:>/EXIT
TEST

(~, DEBUG

FAt;DDT> F'f<OCEED

This example shows source program TEST.PAS being compiled, loaded, and
started.
PROCEED initiates the execution of the program.
Example 2

(;1 PASCAL
@PASCAl..>EXAMPL IDEBUG
f'ASCAL> lEX I T
@LOAD EXAMPL.REL,SYS:PASDDT.REL
(;l SAVE EXAMPL
(!) F<UN EXAMPL
PASDDT>PROCEED

9-10

PASDDT: THE PASCAL-20 DEBUGGER
This
example compiles and
loads
the
program
EXAMPL.
Loading
PASDDT.REL with the program EXAMPL causes a copy of PASDDT to become a
part of
the
executable
file
EXAMPL.EXE.
PROCEED
initiates
the
execution of the program.
Example 3

p f:':j ~:; II II T >P H() C E E D

This example causes the program to resume execution after stopping
a breakpoint.

9.4.8

REMOVE

REMOVE deletes breakpoints and tracepoints completely.
REMOVE

(AT) {Jine number
}
break identifier

[(NAME)

The format is:

break identifier]

where:
1 ine number

is the line number in the source code
you set the breakpoint or tracepoint.

where

break identifier

is the name of the breakpoint or tracepoint.

You can specify either a line number followed by the break
identifier
associated with
that
line number, just the line number, or just the
break identifier.
Note the following example:

PASDDT>b reak 10
PASDDT>displa'3 break
rmFAK POINT (S) :
Name
: <NONAME>
Break Address
000000400044
Line Number
10
Status
BREAK
PASDDT>remove 10
F'ASDDT>d i SF' 1 <3'3 b r'eak
NO BREAKPOINTS SET
F'ASDDT>E'~'~

9.4.9

SET

The SET command has four functions:

Turning
on automatic displaying
of source
breakpoint is executed.
This is the default.

Telling PASDDT how many lines to display when a breakpoint is
executed.
The default is three.

9-11

code

when

PASDDT: THE

PAS~AL-20

DEBUGGER

Telling PASDDT which separately compiled module
to
use
in
terms of setting
breakpoints and gaining information about
the user program. When PASDD~ starts up, the default is
the
main program.

Setting the level of jnformation displayed
default is VERBOSE.

To SET the automatic displaying of source code
following format:

PASDDT.

The

off,

the

use

SET (OPTION) {A [UTO-DISPLAY D
l
NO-A [UTO-DISPLAyD (
where:
A.UTO-DISPLAY

turns on the automatic displaying
of source
code at a breakpoint.
This is the default.

NO-AUTO-DISPLAY

turns off the automatic displaying of
code at a breakpoint.

source

When AUTO-DISPLAY is set, PASDDT displays the line of source code at
the breakpoint.
If your terminal is slow or you are using a dial-up
line, you may want
to
turn off displaying
the
source code with
NO-AUTO-DISPLAY.
Note the following example:

.10
set no-ooauto

PASD[lT>bT'f.~ak
PA~)DDT>

PAf:;DDT> p T'oc(~ed
»BF,EAK:

PC AT VIRTUAL ADDRESS
000000000416
PASCAL LINE NUMBER
10
PASDDT>break 12
PASDDT>set auto
PASDDT>PI'oC
Enter YOUI' SCOI'es. When done, t~pe CTRL/Z.
23
»BREAK:

PC AT VIRTUAL ADDRESS
000000000443
PASCAL LINE NUMBER
=
12
LINE 11:
WHILE NOT EOF DO
LINE 12:
BEGIN
LINE 13:
READLN (Score);
To SET the number of lines to automatically display, use the following
format of the SET command:
SET (OPTION) W [INDOwD

where:
lrlINDOW

indicates that you want to set the number of
lines
of
source code to display at a
breakpoint.

is a
decimal
integer specifying how many
lines of source code to display.
The default
is three.

The window is centered around the line on which the breakpoint is set.

9-12

PASDDT: THE PASCAL-20 DEBUGGER
Note the following

ex~mple:

PASDD1)set window 10
Pt,!:;ODT>p I'OC
4~:.:j

»Bf<EAI< ~

PC AT

VIRTUAl... ADDRESS

000000000443

I... J NF NUMBEI<
I... INE "?! BEGIN
I... 1 N[ n:
Tot a :I. !:::: () y
I... J N E (1':
C 0 u n t. ~:::: () y
1..1F~ I TEL. N (I E n tt-:~ l' ~~ 0 U J'
I...JNE :1.0:
LINE :I. 1
WHII... E NOT FOF DO

P(.)!:;C(~L

·
LINE
LINE 13:
LINE
·
LINE
:1.2:

F~ [ A DI... N

Count

:1. !:.=.; :

Wh €.~ n don e,

t ~:t F' peT RI... / 1 • I ) ;

BEGIN

:1.4

LINE 1 f.)

~:; C 0 J' e !,;.

(!:) COl' e ) Y

Total := Scorp + Total;
:= Count + 1;

Endy

To SET the name of the module where PASDDT can
find
information
relevent for debugging purposes, use the following format of the SET
command:
SET (OPTION)

M [ODULE]

module name

where:
MODULE

indicates that you want to specify
of a separately compiled module.

the

name

module name

is the identifier after the word PROGRAM or
MODULE
in the source file.
At start-up, the
default is the main program.

Note the following example:

PASDDT>spt module ? MODULE NAME AVERAGE_SCORE
PASDDT>set module average_score
F'ASDDT>
To SET the level of information displayed by PASDDT, use the following
format of the SET command:
SET (OPTION) V [ERBOSITY]

(OF TYPEOUT) {V [ERBOSEll }
BRIEF lJ

where:
VERBOSITY

indicates that you want to change
of information displayed.

VERBOSE

displays the most information.
default.

BRIEF

dispJays

the

This

level
is

minimal amount of information.

9-13

the

PASDDT: THE PASCAL-20 DEBUGGER
Note the following example:
PA~;DD T>~:;.,·:·.>t

VP r;·:.lo';; I TY

(OF TYPEOUT) b T' i. e'f

PA~:;DnT::F' rucp

::.:it:·,

:> >Br~FA~<:
I... I NE NUMBEI:~
:1.2
BEGIN
8:
Total:= 0;
Co u n t ; ~o () y
9:
:LO:
("II:~ I TEl... N (/ En t P l' ~:J (.") 1..1 r .; C () T' f' s.
It!
("IH TLE NOT FOF no

F'(~~:;C()/...

LINE "7!

LINE
I... INE

LINE
LINE
LINE :L?:
I... INE 13:
LINE 14:

LINE

Wh p n (Ion E',

t ~:~ p (.:.) C T f< l.. / Z • ' ) ,

BEGIN
f~ E A 1:1 L.. N

:l~::j:

Count

(S COl"' e ) ;

Total := Score t
:= Count t 1;

Total;

Enrj y
LINE 1,'.):
PAf;nItT>Sf::,t w:i nl.".luw ~.:J
F'(':)~:;DDT)1-"' roc

/U
»BF~EAK :

PAbCAL.. I... I NE NUMBER

LINE :LO!
LINE :1.1;

WRITELN ('Enter ~our scores.
~JH:r LE NOT [OF DO

When done,

t~pe

CTf<L/Z.');

BEGIN

LINE 12:

LINE 13:

F~E(.:IDLN

LINE 14:

(SCOT'f.~);

Tutal

:= Score t

Total;

PASDDT>t');.: i t

9.4.10

SHOW

The SHOW command prints the current
format of the SHOW command is:

value

identifier.

The

SHOW (VARIABLE OR ADDRESS) {user identifier}
octal 0r'ldress
where:
user identifier

is the name of
scope.

octal address

is A virtual address in the user program.
(This is not the address shown in the listing

variable

the

current

f i 1 e.)

Note that, when you set a breakpoint, PASDDT automatically displays
three lines of source code, unless you specified SET NO-AUTO-DISPLAY.
You can then see any identifiers that are declared for the current
scope of the user program with the SHOW command. You can also type a
question mark (?) to the SHOW command to see what identifiers are
available.

9-14

PASDDT: THE PASCAL-?O DEBUGGER
Note the following example:
P (.~ ~:; 11 D T ::- /..1 r (7.' a k

:I. ()

r'A~:;DDT:> F' rocc:'f.·~(i

:> :> nF;: F (~,.( :
PC AT

VIRfUAL ADDRESS
000000400044
PASCAl... I... I NE NUMBE'~
1 ()
LINE 9:
Count:= ()~
LINE 10:
WRITELN ('Enter ~our scores. When done, t~pe CTRL/Z.');
LINE 11:
WHILE NOT EOF DO
PASDDT:>show ? VIR1UAL ADDRESS
o J' F' A~; CAL. VA f~ J f~ [: l.. Eon e 0 f the f 0 1. low i n ~~ :
AVFF~(~GESCDf~E
AVERAGE .... SCURE
COUNT
INPUT
OUTPUT
SCORE
TOTAL
F'f.1SDDT> (~,how seo T'C';'>
()
VALUE
TYPE
INTEGER
ADDRESS
037060
PA~:;DDT> show count
ljALUE
o
TYPE
INTEGER
OT?062

PASDDT>show input
F :i. I (-? is OPEN fo l' T'E~ad i ng
File is a TEXT file
F :i. :I. p

i ~:; TTY:

File is the

standard

Fi If? name
F'ASDDT> e;<

TTY:

9.4.11

INPUT file

TRACE

TRACE sets a new tracepoint at a
specified
location or resets an
existing tracepoint.
When the program reaches the location specified,
PASDDT prints a message indicating that a trace was placed on this
line.
Unlike the BREAK command, TRACE does not halt program execution
or print source code.
You can set a
total
of 20
tracepoints and
breakpoints.
Clearing them does not delete them from the count.
To set a new
command:

tracepoint,

TRACE (AT)

use

line number

the

[(NAME)

following

format

the

TRACE

trace identifierD

where:
line number

is the line number associated with the
line
of source code that you want to trace.
This
line number can be found in the listing file.

trace identifier

is the name you want to associate with the
tracepoint.
Each identifier can contain any
number of characters, but the
first nIne
characters must be unique.
The identifier
can contain all characters except underscore
( ) and doLlar sign ($).

9-lS

P~SDDT:

THE PASCAL-?O DEBUGGER

To reset a tracepoint that you have turned
off with CLEAR
9.4.3), use the following format of the TRACE command:
TRACE

(AT)

J line number

(Section

}

tbreak identifier
where:
line number

is the line number of an existing
that you want to reset.

break identifier

is the name of an existing
you want to reset.

The following

tracepoint

that

rules apply to the use of TRACE:

If you specify a
following happens:

trace

identifier,

but

not

•

If there is already a tracepoint with this
tracepoint is used.

•

If there is no tracepoint
occurs.

If you specify a
following happens:

line

number,

with
but

not

number,

the

identifier,

this

identifier,

error

trace

identifier,

the

for

this

location,

that

•

If
a
tracepoint
is
tracepoint is used.

•

If a tracepoint has not been assigned, a
new tracepoint
is
established with
the identifier NO NAME.
You can have more
than one tracepoint with the name NO NAME.

If you specify both
following happens:

line

assigned

this

line

number

and

trace

identifier,

the

•

If a tracepoint has already been assigned for
this
location
with
this
identifier,
the previously defined tracepoint is
used.

•

If a tracepoint has already been assigned for
with a different identifier, an error occurs.

•

If a tracepoint has already been assigned at another location
with
this identifier, and no tracepoint has been set at this
location,
PASDDT prints a warning message.
PASDDT then
asks
if you want to override the address associated with the
previous tracepoint.
If so,
the
previous
tracepoint
is
discarded,
and
a
new tracepoint is declared with the given
identifier and location.

9-10

this

location

PASDDT: THE PASCAL-20 DEBUGGER
Note the following example:

PASDDT>trace 10
F'r:":)f;DDT> p 1'ocf.~pd

one

»Tr~ACF: ONE
PC AT VIRTUAL ADDRESS
PASCAL LINE NUMBER
Ent~:~ l'

~:IOU r

~:;co T'P~;;.

000000400044
:1.0

Whf.:'n (lOne,

t\:lf"e

CTF~I... /Z.

7:':;

44
Dr;
'·'7Thf~

ave ra!.:.!p

~:;co r'(-;-:o

:i~;:

69.3

9-17

APPENDIX A
PASCAL MESSAGES

._------------Table A-I:

Run-time Errors

? PRTOOI

Value not within subrange of variable in assignment.

? PRTOO2

Case selector out of range.

? PRTOO3

Array index out of bounds.

? PRTOO4

Conformant array index out of bounds.

? PRTOO5

Size of conformant arrays incompatible.

? PRToor;

NIL pointer value at runtime.

? PRTOO7

Attempt to divide by O.

? PRTOO8

Mod with 0 or negative value.

? PRTOO9

Set value out of range in assignment.

? PRTOIO

Set element out of range in assignment.

? PHTOll

Stack expansion faiJed - No
for stack space.

? PRT012

Memory expansion failed - No more memory available.

? PRT013

DISPOSE called with NIL pointer.

? PRT014

Page creation failed, error code

? PRT015

Page destroy

? PRTOII1

Insufficient initial memory -

? PRT017

Fatal error PC number.

? PRT018

Fatal error - illegal memory reference
user PC number.

? PRT019

Fatal
error
PC number.

? PRT020

Fatal
error PC number.

illegal memory write detected at user

? PRT021

Fatal
error numbe r.

non-existent page detected at user PC

f~iled,

ilJegal

memory

available

= number.

error code = number.
cannot start program.

instruction detected

detected

user
at

illegal memory read detected at user

A-I

PASCAL MESSAGES
Note that errors PRTOl7 through PRT021 are probably caused by improper
use of an array or pointer variable.
If you receive one of these
errors, try compiling the program with the /CHECK switch.
Table A-2:

I/O Errors

? PIOOOI

User buffer overflow. File in error: filespec

? PIOOO2

Line limit exceeded. File in error: filespec

? PIOOO3

File not open for reading. File in error: filespec

? PTOOO4

File not open for writing. File in error: filespec

? PIOOO5

Field width <= zero. File in error: filespec

? PIOOO'S

String write error. File in error: filespec

? PIOOO7

Integer write error. File in error: filespec

? PIOOO8

Field width < zero. File in error: filespec

? PIOOO9

Attempt to read past EOF. File in error: filespec

? PIOOlO

Integer read error. File in error: filespec

? PIOOlI

String read error. File in error:

? PIOOl2

Illegal character in
filespec

? PIOOI]

Attempt to RESET(output). File in error: filespec

? PIOO14

Attempt to RESET(input). File in error: filespec

PIOO15

? PIOO U:;

number

error:

to reset/rewrite uninitialized file. File in
error: filespec

~ttempt

Error
in
filespec

opening

binary

? PIOOl8

Error
in
fiJespec

writing

PIOO19

Error
in
filespec

closing

binary

File

Integer overflow. File in error: filespec

PIOOl7

I/O.

filespf~c

file.

object

File

error:

file. File in error:

filc.

File

error:

? PIOO20

Delete file error. File in error: filespec

? PIOO21

Include/Exclude file error. File in error: filespec

PIOO22

Attempt to use FIND on text
filespec

PIOO21

Attempt to
filespec

PIOO24

Field width too
error: filespec

FIND

file.

File

error:

a negative record. File in error:
small

A-2

for number output. File in

PASCAL MESSAGES

---------------------------------------------_._--------Table A-2:

I/O Errors (Cont.)

-------------? PIOO25

Scalar out of range. File in error:

filespec

? PIOO2f)

Attempt to open/close INPUT
filespec

file.

File

error:

? PIOO27

Attem~t

file.

File

error:

OUTPUT

? PIOO2R

Attempt
to
filespec

write

to READONLY file.

File in error:

? PIOO29

Attempt
to open already opened file.
filespec

File in error:

? PIOO30

At tern p t

toR EWR I T E ( 0 u t put). F i 1 e in err 0 r:

f i ] e s pe c

? PIOO31

At~::empt

to REWRITE (input) • File in error:

filespec

? PIOO32

Exponent too large. File in error:

filespec

? PIOO33

Exponent too small. File in error:

filespec

? PIOO34

Reset
error
filespec

? PIOO35

Scalar is all blanks. File in error:

filespec

? PIOO3(j

Attempt to read past enct of
filespec

File

error:

? PIOO37

Attempt
to readln/writeln to a binary file.
err-or: filespec

File in

? PIOO40

PAGE called with fjle
err-or: filE~spec

File

? PIOO42

EOLN called when EOF true.

? PIOO43

Testing
EOF
filespec

? PIOO44

Too many open files.

? PIOO45

Attempt
to
filespec

filespec

File

not

found.

line.

opened

unopened

File

NOCARRIAGE.

File in error:
file.

RESET new internal

in error:

File

file.

filespec
in

e r ro r:

File in error:

Compile-Time Errors

? PASOOI

Error in simple type

The declaration for a base type of a set or the index type of
array contains a syntax error.

PASOO~

Identifier expected

The statement syntax requires an
found.

A-3

identifier,

but

none

can

PASCAL MESSAGES
? PAS003

tlpROGRAM tI or tlMODULE tI expected

The statement
MODULE.
? PAS004

tI)

syntax

requires

the

reserved

word

PROGRAM

expected

The statement syntax requires the right-parenthesis character.
? PAS005

11:11

expected

The statement syntax requires a colon character.
? PASOOf)

I I I eg a]

symbol
as

misspelled

such

The statement contains an illegal symbol, such
reserved word or ilJegal character.
? PASOO7

Error in parameter list

The parameter list contains a syntax error,
comma, colon, or semicolon character.
? PASOO8

1I0F II

missing

expected

1'he statement syn ta x requires the reserved word OF •.
? PASClO9

II (II expected

The statement syn tax requires the left-parenthesis character.
? PASOIO

Error in type

The statement syn tax requires a data type, but no type identifier
is present.
? PASOll

II r II expected

The statement syntax requires the left square bracket character.
? PAS012

II 1 II expected

The statement syn ta x requires the right square bracket character.
? PAS013

IIENDII expected

The compiler cannot find the delimiter END, which marks
of a compound statement, subprogram, or program.
? PAS014

the

end

"., " expected

The statement syn tax requires the semicoJon character.
? PAS015

Integer expected

'rhe statement syn tax requires an integer.
? PAS016

11=11

expected

The statement syntax
requires the equal sign
to separate a
constant identifier from a constant value or to separate a type
identifier from a type definition.

A-4

PASCAL MESSAGES

? PASOl7

"BEGIN" expected

The compiler cannot find the delimiter
beginning of an executable section.

? PAS018

BEGIN,

marks

the

" expected

The compiler cannot find
the
symbol,
between the endpoints of a subrange.
? PAS019

which

required

Error in field list

The field list in a record declaration contains a syn tax error.
? PAS020

" , " expected

The statement syntax requires a comma.
? PAS021

Empty parameter (successive " , " )

not allowed

The parameter list attempts
to
specify
parameter, or contains an extra comma.

? PAS022

IJlegal

null

missing

(nonprintable) ASCII character

The program contains an illegal character that is not a printable
ASCII character.

? PAS023

"," or ")" expected

The
statement
syntax
requires
right-parenthesis character.
? PAS024

either

comma

'" expected

The statement syntax requires two quotation marks.

? PAS050

Error in constant

A constant contains an illegal character or is improperly formed.

? PAS051

":=" expected

The statement syntax requires the assignment operator.

? PAS052

"THEN" expected

The compiler cannot find the reserved word THEN to
IF-THEN statement.

complete

the

The compiler cannot find the reserved word UNTIL to complete
REPEAT statement.

the

? PAS053

? PAS054

"UNTIL" expected

"DO" expected

The compiler cannot find the reserved word DO to complete the FOR
statement or the WHILE statement.
? PAS055

"TO" or "DOWNTO" expected

The compiler cannot find the
FOR statement.

rE~served

A-5

word TO or

DOWNTO

the

PASCAL

? PAS05A

MESS~GES

Invalid expression

The statement syntax requires an expression, but the first symbol
the compiler finds is not Jegal in the expression.

? PAS059

Error in variable

A reference to an array element or record field contains a syntax
error.

? PAS060

"ARRAY" expected

The compiler cannot find the reserved
definition.

PASO~l

PASO~2

ARRAY

the

type

reserved

word

PROCEDURE

"PROCEDURE" or "FUNCTION" expected

The statement syntax requires
FUNCTION.

word

the

Internal compiler error

An internal error has been detected.

? PAS095

Functions "BIN", "OCT", or "HEX" not allowed in this context

The context does not allow the functions BIN, OCT, or HEX.
? PAS096

File component may not exceed

~5515

words

The file component is larger than 65535 words.

? PAS097

Error count exceeds error limit.

Compilation aborted.

The number of errors exceeded the limit you specified.

? PAS099 End of input encountered before end of program.
aborted.
? PAS100

Array size too large

A declared
array is larger
than
2,147,483,647
2,147,483,647 bits for a packed array.

? PAS101

bytes

Identifier declared twice

An identifier is declared twice within a
You can
redeclare
identifiers only in
sections.

? PASl02

Compilation

declaration section.
different declaration

Lowbound exceeds highbound

The lower limit of a subrange is greater than the upper limit of
the subrange, based on their ordinal values in their base type.

? PASI03

Identifier is not of appropriate class

The identifier names the wrong class of data.
For example,
it
names a
constant where
the syntax of the statement requires a
procedure.

? PAS104

Identifier not declared

The program uses an identifier that has not been declared.
A-6

PASCAL MESSAGES

? PASIOS

Sign not allowed

A plus or minus sign was found
nonnumeric type.

? PASI06

front

expression

Identifier previously used in this block.

You gave two items the same identifier within the same block.

? PASI07

Incompatible sub range types

The subrange types are not compatible according to the
type compatibility.

? PASIOS

File not allowed in variant part

A file type cannot appear in the variant part of a

? PASI09

rules

record.

Type must not be real or double

You cannot specify a rea] value here.
Real values cannot be used
as array subscripts, control values for FOR loops, tag fields of
variant records, elements of set expressions,
or houndaries of
subrange types.

? PASIIO

Tagfield type must be scalar or subrange

The tag field for a variant record must be a scalar
type.

? PASlll

field
are
compatible

of
incompatible
types.
according to the general

Index type must not be real or double

Array subscripts cannot be real values;
integer or integer subrange values.

? PASll3

subrange

Incompatible with tagfield type

The case label and the
tag
These
two
items must be
compatibility rules.

? PASll2

if numeric, they must be

Index type must be scalar or sub range

Array subscripts must be scalar or subrange values, and cannot be
of a structured type.

? PASl14

Base type must not be real or double

The base type of this set or subrange cannot be one of
types.

? PASllS

real

Base type must be scalar or subrange

The base type of this set or subrange must be scalar or
values, and cannot be of structured type.

? PASll6

the

subrange

Actual parameter must be a set of correct size

The actual parameter must be of correct size when passed as a VAR
parameter.

? PASl17

Undefined forward reference in type declaration:

Compilation aborted.
A-7

name

PASCAL MESSAGES
? PASll8

Value injtializatjon must be in main program

A VALUE initialization must be in the main program.
? PASl19

Forward declared:

repetition of parameter list not allowed

You cannot repeat the parameter
declaration of a subprogram.
? PAS120

list

after

forward

the

Function result type must be scalar, subrange, or pointer

The function specifies a result that is not a scalar, subrange,
or pointer type.
Function results cannot be structured types.
?

P~S121

File value parameter not allowed

A file cannot be passed as a value parameter.
? PAS122 Forward declared function:
allowed

repetition of

result

type

not

The result of the function appears in both
the
forward
declaration and
in the later complete declaration. Th€ result
can appear only in the forward declaration.
? PAS123

Missing result type in function declaration

The functjon heading does not declare the type of the
the function.
? PASl24

Fraction format for real and double only

You can specify two integers in the field width (such
for real, single, and double values only.
? PAS125

result

R:3:2)

Error in type of predefined function parameter

A parameter passed to a predefinen function is not of the correct
type.
? PAS126

Number of parameters does not agree with declaration

The number of actual parameters passed to the subprogram is
different from the number of formal parameters declared for that
subprogram. You cannot add or omit parameters.
? PAS127

Parameter cannot be element of a packed structure

You cannot
subprogram;
use it.

pass
you

one element of a packed structure to
a
must pass the entire structure if you want to

? PAS128 Result type of actual function parameter does not agree with
declaration

The result of an actual function parameter is
specified in the formal parameter list.
? PAS129

not

the

type

Operands are of incompatible types

Two or more of the operands in an expression are of incompatible
types.
For example, the program attempted to compare a numeric
and a character variable.

A-8

PASCAL MESSAGES

? PAS130

Expression is not of set type

The operators you specified are valid only for set expressions.

? PAS131

Type of variable is not set

The statement syntax requires a set variable.

? PAS132

Strict inclusion not allowed

You must use the <= and >= operators
to
test
set
inclusion.
PASCAL does not allow you to use the less than «) and greater
than (»
signs for this purpose.

? PASl33

File comparison not allowed

Relational operators cannot be applied to file variables.

? PASl34

Illegal type of operand(s)

You cannot perform the specified operation on data items
specified types.

? PAS135

the

Type of operand must be Boolean

This operation requires a Boolean operand.

? PAS136

Set element must be scalar or subrange

The elements of a set must be scalar or subrange
cannot have elements of structured types.

? PASl37

types.

Sets

Set element types not compatible

The elements of this set are not all of the same type.

? PASl38

Type of variable is not an array

A variable that is not an array type is followed by a left square
bracket or a comma inside square brackets.

? PASl39

Index type is not compatible with declaration

The specified array subscript is not
specified in the array definition.

? PAS140

compatible

with

the

type

Type of variable is not record

A period appears following a variable that is not a record type.

? PASl41

Type of variable must be file or pointer

A circumflex character appears after the name of a variable
is not a file pointer.

? PASl42

that

Illegal parameter substitution

The type of an actual parameter is not compatible with
of the corresponding formal parameter.

A-9

the

type

P~SCAL

MESS~GES

? PAS143

Loop control variable must be an
point scalar

unstructured

an
integer,
integer
it cannot be a real

The control variable in a FOR loop must be
subrange,
or user-defined
scalar
type;
variable.

? PAS145

Type conflict between control variable and loop bounds

The type of the control variable in a FOR loop
with the type of the bounds you specified.

? PAS146

incompatible

Assignment of files not allowea

You cannot assign one file to another.
used to give values to files.

? PAS147

non-floating

Output procedures must be

Label type incompatible with selecting expression

The type of case label Is incompatible with the type to which the
selecting
expression evaluates.
Case labels and
selecting
expressions must be of compatible types.

? PAS148

Subrange bounds must be sca]ar

You can specify subranges of scalar
specify a real or string subrange.

types

only.

You

The index type of a
nonconformant array
although it can be an integer subrange.

cannot

integer,

You cannot assign a value to an external or predeclared
identifier.

function

? PAS149

? PAS150

? PAS151

Index type must not be integer

Assignment to this function is not allowed

Assignment to formal function parameter not allowed

You cannot assign a value
parameter.

? PAS152

the

name

formal

function

No such field in this record

You attempted to access a record by an incorrect
field name.

? PAS153

cannot

nonexistent

Type of parameter must be character string (array of char)

The actual parameter passed to this function or procedure must be
a character string.

? PAS154

Type of parameter must be integer

The actual parameter passed to this function or procedure must be
an integer.

? PAS155

Recursive %INCLUDE not allowed.

? PAS156

Multidefined case label

Compilation aborted.

The same case label refers to more than one statement.
label can be used only once within the CASE statement.
A-IO

Each case

PASCAL MESSAGES

? PAS157

Case label

range exceeds 1000

The range of ordinal values between the largest and smallest case
labels must not exceed 1000.

? PASl58

Missing corresponding variant declaration

In a call
to NEW or DISPOSE,
more
tagfield
constants were
specified
than
the number of nested variants in the record type
to which the pointer refers.

? PAS159

Double, real or string tagfields not allowed

Tag fields cannot be
scalar.

? PAS160

real

string

variables,

but

Previous declaration was not forward

The reiteration of a procedure or function that was
declared is illegal.

? PASJGl

must

not

forward

Procedure/function has already been forward declared

The subprogram has already been forward declared.

? PAS162

Undeclared procedure or function:

name Compilation aborted.

You specified a procedure or function without declaring it in the
declaration section.

? PAS163

Type of parameter must be real or integer

The subprogram
parameter.

? PAS164

requires

real

integer

expression

This procedure/function cannot be actual parameter

The specified predeclared procedure or
function cannot be an
actual
parameter.
If you must use it in the subprogram, cal] it
directly.

? PAS165

Multiply defined label

A label appears in front of more than one statement in
executable section.

? PASl66

single

Multiply declared label

The program declares the same label more than once.

? PASl67

Undeclared label

The program contains a label that has not been declared.

? PAS168

Undefined label:

label number Compilation aborted.

You specified a label as an argument to the GOTO
the label is not defined.
? PASl69

statement,

but

and

Set element value must not exceed 255

The ordinal value of an element of a set must be
255.

A-II

between

PASCAL MESSAGES
? PAS170

Value parameter expected

A subprogram that is passeo as nn actual parameter can have
value parameters.
? PAS17l

Type of variable must be textfiJe (file of char)

The specified operation or subprogrnm
variable as an operand or parameter.
? PASl72

Undeclared external file:

You specified nn external file
declaration section.
? PASl73

only

requires

text

file

name Compilation aborted.
that

was

not

declared

the

Negative set elements not allowed

The valu€ of an integer set element must be between 0 and 255.
? PASl74

Parameter must be a file type

The specified subprogram requires a file as a parameter.
? PAS175

"INPUT" not declared as nn external file

The program makes an implicit reference to the file variable
INPUT, but INPUT is either not declared or has been redeclared at
an inner level.
? PAS176

"OUTPUT" not declared as an external file

The program makes an impJicit reference to the file variable
OUTPUT,
but OUTPUT is either not declared or has been redeclared
at an inner level.
? PAS177

Assignment to function identifier not allowed here

Assignment to a function jdentifier is allowed
function block.
? PAS178

within

the

Multidefined record variant

A constant tag
fieJd value appears
definition of a record variant.
? PAS179

only

than

once

the

File of file type not allowed

You cannot declare a file that has components of a file type.
? PAS181

Array bounds too large

The bounds of an array are too large to allow the elements of the
array to be accessed correctly.
? PASl82

Expression must be scalar

The expression must specify a scalar value;
are not legal.

structured variables

? PAS183
"[GLOBAL]" or "[FORTRAN]" may only precede
level 1

declarations

The words (GLOBALl or [FORTRAN] can be placed only in a
or procedure heading.
A-12

function

PASCAL MESSAGES

? PAS184

External procedure has same name as main program

Program and procedure names must be unique.

? PAS186

Formal procedures may not have conformant array parameters

You cannot pass a conformant array as a parameter to a
that is itself passed as a parameter.

? PASl87

IJlegal

conformant array assignment

The program attempts to perform an illegal
conformant arrays.

? PAS188

procedure

assignment

involving

Parameter must be scalar and not real or double

The parameters to the predeclared functions sucr and PRED must be
scalar types, and cannot be one of the real types.

? PAS189

Actual parameter must be a variable

When you use VAR with a
formal
parameter,
the
corresponding
actual parameter must be a variable and not a general expression.

? PAS190
"READLN" ,
textfiles

"WRITELN"

and

"PAGE"

are

defined

only

fo r

The predeclared procedures READLN, WRITELN, and PAGE operate only
on text files.

? PAS191

"READ" and "WRITE" require input/output parameter list

The READ and WRITE procedures require
you cannot omit the parameter list.

? PASl92

least

one

parameter;

Illegal type of input/output parameter

Arrays, sets, records, and pointers cannot be parameters
READ and WRITE procedures.

? PAS193

the

Field width parameter must be of type integer

The field width you specify must be an integer.

? PAS194

Variable must be of type PACKED ARRAYfl .• ll]

OF CHAR

The DATE and TIME procedures require a parameter of PACKED
[ 1 .. 11] OF CHAR.

? PAS195

ARRAY

Type of variable must be pointer

The statement syntax requires a variable of pointer type.

? PAS196

Type of constant does not agree with tagfield type

The type of a constant in a tag value list is
the tag field type.

? PASI97

incompatible

with

Type of parameter must be real or double

The statement syntax
precision) value.

requires

A-I]

real

(singIe-

double-

PASCAL MESSAGES

? PAS198

Type of parameter must be double

The statement syntax requires

? PAS199

double-precision value.

Parameter must be of numeric type

The procedure or function requires
value.

? PAS200

integer

real

number

Parameter must be scalar or pointer and not real

The procedure or
function
requires an
integer,
user-defined
scalar,
Boolean, integer subrange, user-defined scalar subrange,
or pointer parameter.

? PAS201

Error in real constant:

digit expected

A real constant contains a nonnumeric character where
is required.

? PAS202

numeral

String constant must not exceed source 1ine

The end of the Jine occurs before the apostrophe
that closes a
string.
Make
sure that the second apostrophe has not been left
out.

? PAS201

Integer constant exceeds range

An integer constant is outside the permitted
(that is, 2**31 to 2**31).

? PAS204

integers

Actual parameter is not correct type

The actual
parameter
is not
corresponding formal parameter.

? PAS205

range

compatible

type

with

the

Zero length string not allowed

You cannot specify a string that has no characters.

? PAS206

Illegal digit in binary, octal or hexdecimal constant

A binary, octal, or
digit.

? PAS207

hexadecimal

constant

contains

illegal

Real or double constant out of range

singleor double-precision
real
number
is outside
the
permitted
range
0.29*10**(-38) to 1.7*(10**38) for positive
numbers and -0.29*10**(-38)
to -1.7*(10**38)
for
negative
numbers.

? PAS208

Data type cannot be initialized

This variable contains a type, such as a
initialized.

? PAS209

file,

that

cannot

Variable has been previously initialized

You can specify only one VALUE declaration for a variable.

? PAS2l0

Variable is not array or record type

The VALUE initialization for a variable that is not a
an array contains a constructor.
A-14

record

PASCAL MESSAGES
? PAS?II

Incorrect number of values for this variable

The VALUE declaration contains too many or too few values for the
variable being initialized.

? PAS21?

Repetition factor must be positive integer constant

The repetition factor
in an
positive integer constant.

? PAS2I1

array

initialization

must

Type identifier does not match type of variable

The optional type identifier must be compatible with the t:"?e
variable to be injtialized.
? PAS2I4

Incorrect type of value element

A constant appearing in a VALUE initialization has a
type other
than
that of the variable, record field, or array element to be
initialized.
? PAS2l5

RMS record size must be a positive integer constant

The record size specified in the OPEN procedure call
positive integer constant.

? PAS216

"OLD"

was

not

is not allowed for this file

You cannot specify OLD for an internal file.

? PAS217

Assignment to Conformant Array Index is not allowed

You cannot make this assignment to a conformant array index.

? PAS2I8

Array must be unpacked

An array parameter to PACK or UNPACK is not unpacked correctly.

? PAS2I9

Array must be packed

An array parameter to PACK or UNPACK is not packed correctly.

? PAS220

Packed bounds must not exceed unpacked bounds

The bounds of the packed array exceed the unpacked bounds.

? PAS224

"rOVERLAID1" expected

The statement syntax requires the keyword
? PAS225

rOVERLAID].

Illegal file attribute specification

You specified an attribute in the
recognized by the compiler.

OPEN

statement

? PAS226 Positional parameter not allowed after first
parameter

? PAS227

that

non-posjtional

"OLD n , "NEW", nREADONLY", or "UNKNOWN" expected

The statement syntax requires either the keyword OLD or NEW.

A-I5

not

PASCAL MESSAGES
? PAS228

"SEQUENTIAL" or "DIRECT" expected

The statement syntax requires either the
DIRECT.

? PAS229

keyword

SEQUENTIAL

"FIXED or "VARIABLE" expected

The
statement
VARIABLE.

syntax

? PAS230

"NOCARRIAGE",
expected

requires

"NONE",

either

"CARRIAGE",

the

keyword

"FORTRAN",

The statement syntax requires one of
the
following
NOCARRIAGE, NONE, CARRIAGE, FORTRAN, or LIST.
? PAS231

FIXED

"LIST"

keywords:

II legal keyword

This keyword cannot be specified in this context.

? PAS232

Parameter has already been specified

You have specified the same parameter twice.

? PAS233

File variable must be specified

You forgot to specify the file variable.

? PAS234

Identifier or character string literal expected

You need to specify a

? PAS235

identifier or character string literal.

Parameter cannot be specified in this posjtion

You specified a parameter that did not belong in this place.

? PAS237

A "NEW" and "DIRECT"

file must have fixed-length records

You specified variable-length records for a file that
fixed-length records.
? PAS238

must

have

Record type may not be "VARIABLE" for "DIRECT" files

You cannot have variable length records for a DIRECT file.
? PAS239

%INCLUDE file not found.

Compilation aborted.

? PAS240

Include/Exclude file error.

? PAS250

Too many nested scopes of identifiers

CompiJation aborted.

You can have only 20 levels of nesting.
occurs with each block or WITH statement.

? PAS251

new

nesting

level

Too many nested procedures and/or functions

Subprograms can be nested no more that 20 levels deep.

? PAS252

Assignment to function
scope conflict

not

allowed

here.

Probable

A function is nested within a function with the same name.

A-l{)

name

PASCAL MESSAGES
? PAS253

Too many arguments in an "OPEN" statement

You specified too many arguments.

? PAS255

Too many errors on this source line

The PASCAL compiler diagnoses only the first 20
source line.

each

The expression is too deeply nested.
To correct this error,
should separately evaluate some parts of the expression.

you

? PAS259

? PAS261

errors

Expression too complicated

Declarations out of order or repeated declaration sections

The declarations must be
in
the
following
order:
constants,
types,
variables, values, and subprograms.
main program can contain value declarations.

PAS2~4

Only
routines

You cannot
routines.

? PAS265

"VAR"

parameters

specify

value

allowed

for

parameters

for

labels,
Only the

"rFORTRAN1"

declared

rFORTRANl

declared

Parameter not allowed for " [FORTRAN1" declared routines

This type of parameter is
routines.

? PAS266 Conformant part of
dimension packed

not

allowed

Conformant

for

arrays

rFORTRAN]
can

have

declared
on]y

one

You packed more than one dimension in a conformant array part.

? PAS267

Conformant arrays must be of the same type

When using relational operators with conformant arrays, the array
types must be equivalent.

? PAS300

Division by zero

The program attempts to divide by zero.

? PAS302

Index expression out of bounds

The value of the expression is out of range for the array element
to which you are assisning it.
? PAS303

Value to be assigned is out of bounds

The value to the right of the assignment operator is out of range
for the variable to which it is being assigned.

? PAS304

Set element expression out of range

The value of the expression is out of range for the
to which you are assigning it.

? PAS30S

Field width must be greater than zero.

A-17

set

element

PASCAL MESSAGES
? PAS306

Index type of conformant array parameter
declaration

exceeds

range

The index type of the actual conformant array parameter
beyond the range declared in the formal parameter list.
? PAS307

extends

Modulus with zero or negative value

The program tried to take the mod of zero.

? PAS309

Variable space
(octal)

? PAS310

exceeds

low

segment

memory

size,

Code space exceeds available address space, 777777

377777

(octal)

Compile-Time Warnings

% PAS401 Identifier exceeds, 31, characters
Identifiers can be any length, but PASCAL scans only the first 31
characters for uniqueness.

% PAS402 Error in option specification
A compiler option is incorrectly specified in the source code.

% PAS403 Source input after "END." ignored
The compiler ignores any characters after the END that terminates
the program.
% PAS404 Duplicate external procedure name

Two external procedures or functions have been declared with
the
same
name.
They
refer
to
the
same externally compiled
subprogram.
% PAS405 LABEL declaration in MODULE ignored

The compiler ignores label declarations at the outermost level in
a module.

% PAS407 Illegal option on include file specification;

LIST assumed

You specified an option that is not available.
% PAS408 One or more parameter values assumed before "param"

% PAS409 Alternative ordering of HISTORY and RECORDLENGTH parameters
% PAS4I0 Parameter type is not known by FORTRAN.
% PAS413 Case label out of range
% PAS450
% PAS451
% PAS452
% PAS453
% PAS454
% PAS455
% PAS456

Nonstandard
Nonstandard
Nonstandard
Nonstandard
Nonstandard
Nonstandard
Nonstandard

PASCAL:
PASCAL:
PASCAL:
PASCAL:
PASCAL:
PASCAL:
PASCAL~

Exponentiation
VALUE declaration
OTHERWISE clause
%INCLUDE directive
MODULE declaration
Label exceeds 9999
"$" or "" in identifier
A-I8

PASCAL MESSAGES
% PAS457

Nonstandard PASCAL:

% PAS458
% PAS459
% PAS460
% PAS461
% PAS4fi2

Nonstandard
Nonstandard
Nonstandard
Nonstandard
Nonstandard

PASCAL:
PASCAL:
PASCAL:
PASCAL:
PASCAL:

% PAS463
% PAS4fi4
% PAS465
% PAS466

Nonstandard
Nonstandard
Nonstandard
Nonstandard

PASCAL:
PASCAL:
PASCAL:
PASCAL:

% PAS467
% PAS468

NonstAndard PASCAL:
Nonstandard PASCAL:

% PAS469
% PAS470

Nonstandard PASCAL:
Nonstandard PASCAL:

% PAS4 7 1

Nonstandard PASCAL:

% PAS472

Nonstandard PASCAL:

% PAS473
% PAS474

Nonstandard PASCAL:
Nonstandard PASCAL:

% PAS475
% PAS476

Nonstandard PASCAL:
Nonstandard PASCAL:

% PAS477

Nonstandard PASCAL:

% PAS478
% PAS479

Nonstandard PASCAL:
Nonstandard PASCAL:

% PAS480
% PAS481

Nonstandard PASCAL:
Nonstandard PASCAL:

Conformant passed to value conformant
array
Directive "rGLOBAL)" or "rFORTRAN) "
Binary, octal or hexadecimal constant
Double precision constant
External procedure declaration
Binary, octal
or hexadecima]
data
output
Output of user-defined scalar
Input of string or user-defined scalar
Input/output of double precision data
Implementation-defined type, function,
or procedure
Directive " [OVERLAIDl"
Formal and actual
parameters not of
identical type
Control variable is not local
Formal and actual parameters not both
packed or unpacked
No parameter list declared
for
this
call
No parameter list declared
for
this
fo rma t
Nonstandard parameter declaration
Array or record types not identical in
assignment
Types not identical in comparison
VAR parameter is selector of variant
record
Parameter is pre-defined procedure or
function
NIL used as constant identifier
Case constants do not cover
range of
tag- type
Input/output of conformant string
Comparison of conformant strings

A-19

APPENDIX B
ASCII CHARACTER SET

Table B-1:

The ASCIl Character Set

-----------------------------------------ASCII

Character

Meaning

NUL
SOH
STX
ETX
EOT
ENQ
ACK
BEL
BS
HT
LF

Null
Start of heading
End of text
End of text
End of transmission
Enquiry
Acknowledgement
Bell
Backspace
Horizontal tab
Line feed
vertical tab
Form feed
Carriage return
Shift out
Shift in
Data link escape
Device control 1
Device control 2
Device control 3
Device control 4
Negative acknowledgement
Synchronous idle
End of transmission block
Cance]
End of med i urn
Substitute
Escape
File separntor
Group separator
Record separator
Unit separator
Spac '9 or blank
Exc]amation mark
Quotation mark
Number sign
Dollar sign
Percent sign
Ampersand
Apostrophe
Left parenthesis

Decimal
Number

o
1

2
3
4
5
6
7
R

10
11

V1:

FF'

CH
SO
S1
DLE
DC1
DC2
DC3

1<1

15
16

17
18
19
20

DC4

21
22
23

Nl\K

Cl~N

25
26
27
28
29

E~1

31
3~

SYN
ETB
SUB
ESC
FS
GS
RS
US
SP

33
34

36
37
38
39
40

%
%

B-1

ASCII CHARACTER SET
Table 8-1: The
ASCII
Decimal
Number

~SCII

Character

41
42

43
44
45

4fi

47
48
49
50
51
52
53
54
55
56
57
58
59
60
61

Character Set (Cont.)

/
0
1
2

3
4

5
f)

8
9

63
64
65
66
67
fi8
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95

?
@

8
C
D
E

F
G
H
I

J
K

L
M
N

P
Q

R
S
T

U
V

W
X
y

Z
[

or
or

Meaning

Right parenthesis
Asterisk
Plus sign
Comma
Minus sign or hyphen
Period or decimal point
Slash
Zero
One
Two
Three
Four
Five
Six
Seven
Eight
Nine
Colon
Semicolon
Left angle bracket
Equal sign
Right angle bracket
Question mark
At sign
Upper case A
Upper case 8
Upper case C
Upper case D
Upper case E
Upper case F
upper case G
upper case H
Upper case I
Upper case J
Upper case K
Upper case L
Upper case M
Upper case N
Upper case 0
Upper case P
Upper case Q
Upper case R
Upper case S
upper case T
Upper case U
Upper case V
Upper case W
Upper case X
Upper case Y
Upper case Z
Left squa re b racke t
Back slash
Right square bracket
Circumflex or up arrow
Back arrow 0 r underscore

B-2

ASCII CHARACTER SET
Table B-1: The ASCII Character Set (Cont.)
ASCII
Decimal
Number
96
97
98
99
100
101
102
103
104

Character

Meanjtng

accent
Lowe l~ case a
Lowe l~ case b
Lower case c
Lower case d
Lower case e
Lower case f
Lower case 9
Lower case h
Lower case i
Lower case j
Lower case k
Lower case 1.
Lowe l~ case m
Lower case n
Lower case 0
Lower case p
Lower case q
Lower case r
Lower case s
Lower,case t
Lower case u
Lower case v
Lower case w
Lower case x
Lower case y
Lower case z
Left brace
Vertilcal line
Right brace
GraVE~

a
b
c
d

e
f

9
h

lOS
106
107
108
109
110

k
1
m

III

112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127

P
q
r
s
t
u
v
w

x
y

z
{

I
}

Ti1dE~

DEL

Delete

B-3

APPENDIX C
SYNTAX SUMMARY

This appendix summarizes the syntAx of the PASCAL-20 language
Backus-Naur Form (BNF):

C.l

the

BACKUS-NAUR FORM

In the BNF, each element of the language
is defined
recursively
in
terms of simpler
elements.
The element being defined is written to
the left of the symbol: ::= and its definition is writt.en to the
right
of tha t symbol.
The BNF uses a group of metasymbols that differ from
the conventions
used
in
the
rest of
this manual
and
are not part of the PASCAL
language.
Table C-l ] ists the meta-symbols used in the BNF.
Table C-l:

BNF MetaSymbols

Symbol

Meaning

------------------------------------------------------------------separates
the
definition.

< >

element

being

defined

from

its

Encloses a definable language element.
Encloses an optional element.
Means "or"; separates possible elements.
Encloses elements
that may be repeated one or more
times, but need not be present

The remainder of this section lists PASCAL in BNF.
<compilation unit> ::= <program>
<module>

I <module>

<module heading> <global declaration part>
<procedure and function declaration part> END .

""=

•

C-l

SYNTAX SUMMARY

<global declaration part> ::= <label decJaration part>
<constant definition part>
<type definition part>
<variable declaration part>
<module word> ::= MODULE I r rOVERLAJDll MODULE
<program heading> ::= <program word> <identifier>
<program word> <identifier> ( <program param'2ters>
<program parameters> ::= <external file identifier>
{ , <external file identifier> }
<external file identifier> ::= <identifier>
<program word> ::= PROGRAM

r rOVERLAIDll

PROGRA~

I S}

<identifier> ::= <identifier head> {<Jetter or digit>
<identifier head> . -= <Jetter>
<letter> . -=

I C
I P
b I c
0
I p

a
n
<digit> : :=,0

I I

D
Q

E
R

e
r

I
I
f I
s I

I 2 I 3 I 4 I 5 I fi

I I
I V
h I
u I v

G
T

H
U

g
t
I

I ,} I K
I W I X
I j I k
I w I x

I
I
I
I

I M

Y I z
I I m

y I z

I 8 I 9

<special symbol> ::= <operator symbol> I <reserved word>
<semireserved word> I <incJude symbol>
<operator symbol> - -= + I - I * I /
I ) I r I 1 I

" I ** I

= I <> I < I > I <= I >=
I } I := I _ I , I
I: I
(* I * ) , (. , • )

<reserved word> - -= DIV , MOD I NIL
IN
OR I AND I NOT I IF
THEN I ELSE I CASE I OF I REPEAT I UNTIL I WHILE
DO I FOR , TO I DOWNTO I BEGIN I END I WITH , GOTO
CONST I VAR , TYPE I ARRAY I RECORD I SET , FILE
FUNCTION I PROCEDURE I LABEL I PACKED I PROGRAM
<semireserved word> ::= REM I OTHERWISE I MODULE

VALUE

<include symbol> ::= %INCLUDE
<blank> ::= <single blank> { <single blank>
<single blank> ::= <space character> I <tab character>
<block> ::= <declaration part> <value initialization part>
<procedure and function declaration part> <statement part>
<declaration part> ::= <label declaration part> <constant declaration
part> <type declaration part> <variable declaration part>
<label declaration part>

::=

<empty>

I LABEL <Jabel> { , <label> } ;

<label> ::= <unsigned integer>
<constant definition part> ::= <empty> I
CONST <constant definition> { ; <constant definition> }
<constant definition> ::= <identifier> = <constant>

C-2

SYNTAX Sm1MARY

<unsigned number> I <sign> <unsigneo number> I
<constant identifier> I <sign> <constant identifier>
<string> I NIL

::=

::=

<JettE!r 0>

:.=

' <octa] digit sequence> '

<letter 0> ::= 0 1 0
<octal digit sequence> ::= <octal digit> { <octal digit> }
<octal digit> - -= 0

I 1

I 2 I 3

::= X

I 4

I 5

I h

I 7

' <hex ("iigit sequence> '

I x

I A I B I
I elf

I DIE I F I a I b I c I d

<binary integer> "-= <letter b> ' <binary digit sequence> '
<letter b> ::= B

::=

sequence> • <digit sequence> D <scale
sequenc.e> • <digit sequence> d <scale
sequence> D <scale factor> I
d <scale factor>
<digit> }

<s ig n> :: = + I <constant identifier> ::= <identifier>
<string> ::= , <character> { <character> } ,
<character> ::= <any ASCII character except '> I
<type definition part>
<empty>
TYPE (type definition>

00:

C-3

<type definition> }

SYNTAX SUMMARY

::=

--=

<type>

I <structured type>

I <pointer type>

I <subrange type>

<type identifier> - -= <identifier>
<structured type> - -= <unpacked structured type>
PACKED <unpackeri structured type>
<unpacked structured type> ::= <array type>
< set type> I < f i 1 e type>

I <record type>

<array type> ::= ARRAY
<index type> { , <index type> } 1 OF
<component type>
<index type>

::=

<component type> ::= <type>
<record type> ::= RECORD <field list> END
<field list>

::= <fixed part>
< va ria n t pa r t>

I <fixed part>

<fixed part> ::= <record section>{

<variant part> I

<record section> }

<record section> ::= <empty> I
<field identifier> { , <field identifier> } : <type>
<variant part> ::= CASE <tag field> <type identifier> OF
<variant> { ; <variant> }
<tag field> ::= <field identifier> :
<variant> ::= <case label list> :

< empty>

( <field list>

I <empty>

00=

<pointer type> ::=

<variable declaration part> ::= <empty> I
VAR <variable declaration> { ; <variable declaration>
<variable declaration> ::= <id~ntifier list> : <type>
I <identifier list> : <type> := <value>
<identifier list> ::= <identifier> { , <identifier> }
<value initialization part> ::= VALUE <value initialization>
{ ; <value initialization> } ; I <empty>
C-4

SYNTAX

Sur~MARY

I <constructor>

r <constant element list> 1

<set constant> ::=

::=

<opt.ional type>

<constant> •• <constant>
( <value element>
{ i' <value element> } )

""=

<empty>

""=

<value>

."=

<procedure declaration> ::= <internal procedure declaration>
<external procedure declaration>
<forward procedure declaration>
<internaJ procedure declaration> ::= <procedure heading> <block>
rrGLOBAL]] <procedure heading> <block>
[rFORTRAN]] <procedure heading> <block>
<external procedure declaration> ::= <procedure heading> EXTERN I
<procedure heading> EXTERNAL I <procedure heading> FORTRAN
<forward procedure declaration> ::= <procedure heading> FORWARD
[rGLOBAL]] <procedure heading> FORWARD I
rrFORTRAN1] <procedure heading> FORWARD
<procedure heading> ::= PROCEDURE <identifier> ;
PROCEDURE <identifier> ( <formal pnrameter section>
{ ; <formal par"ameter section> } )
<formal parameter sect.ion> ::= <extended parameter group>
VAR <extended parameter group> I
FUNCTION <parameter group> I
PROCEDURE <identifier> { , <identifier> }
<procedure heading> I <function heading>
<extended parameter group> ::= <parameter group> I
<identifier> { , <identifier> } : <conformant array schema>
<conformant array schema> ::= ARRAY r <index type specification>
{ ; <index tYPE! specification> } ] OF <type identifier>
ARRAY ( <index type specification>
{ , <index type specification> } , OF <conformant array
schema> PACKED ARRAY r <index type specification> 1
OF <type identifier>
<index type specification> ::= <identifier> •• <identifier>
<scalar type identifier>
C-5

SYNT.AX SUMMARY

<i~entifier>

::=

, <identifi.er>
<type identifier>

<function declaration> - -= <internal function declaration>
<external function declaration>
<forward function declaration>
<internal function declaration> ::= <function heading> <block>
rGLOBAL) <function heading> <block>
rFORTRANl <function heading> <block>
<external function declaration> ::= <function heading> EXTERN I
<function heading> EXTERNAL I <function heading> FORTRAN
<forward function declaration> ::= <function heading> FORWARD I
rGLOBALl <function heading> FORWARD I
rFORTRANl <function heading> FORWARD
<function heading> - -= FUNCTION <identifier> : <result type> ;
FUNCTION <identifier> ( formal parameter section>
{ ; <forma] parameter section> } ) : <result type>
<result type> ::= <type identifier>
<statement part>
<statement>

::=

::=

::=

::=

::=

C-6

SU~1MARY

SYNTAX

<pointer variable> ::= <variable>
<expression> ::= (simple expression> I <simple expression>
<relational operator> <simple expression>
<relational operator> ::= = I <> I < I <= I >=

> I lN

<simple expression> ::= <term>
<sign> <term>
<simple expression> <adding operator> <term>
<adding operator> ::= + I -

I OR

* I /

I DIV I MOD I AND I

REM

** <factor>

<factor> ::= <vari.able> I <unsigned constant> I ( <expression> )
<function designator> I <set>
NOT <factor>
<unsigned constant> - -= <unsigned number>
<string>
<constant identifier> I NIL
<function designator> - -= <function identifier>
<function identifier> ( <actual parameter>
{ , <actual parameter> } )
< fun c t ion ide n t i fie r> - - = < ide n t i fie r>
<set> ::= [

<elemE~nt

list> ]

<element list> : :== <element> { ,

<el.~ment>

} I <empty>

<element> ::= <expression> I <expression> ._ <expression>
<procedure statement> ::= <procedure identifier>
<procedure identifier> ( <actual parameter>
{ , <actual parameter> } )
<procedure identifier> ::= <identifier>
<actual parameter) - -= <expression> I
<procedure identifier> I <function identifier>
<go to statement> ::= GOTO <label>
<empty statement> ::= <empty>
<empty> :: =
<structured statement> ::= <compound statement> I
<conditional statement> I <repetitive statement> I
<with statement)
<compound statement> ::= BEGIN <statement>

; <statement> } END

IF <expression> THEN <statement> I
IF <expression> THEN <statement> ELSE <statement>

::~

C-7

~YNTA.X

SUMMARY

::=

<otherwise part> ::= <empty>
OTHERWISE <statement,> {
; OTHERWISE <statement>
< end case> :: = END I

< empty>

END

<repetitive statement> - -= <while statement> I <repeat statement>
<for statement>
<while statement> ::= WHILE <expression> DO <statement>
<repeat statement> ::= REPEAT <statement,> { ; <statement>
UNTIL <expression>
<for statement> ::= FOR <control variable> := <for list> DO <statement>
<for list> ::= <initial value> TO <final value>
<initial value> DOWNTO <final value>
<control variable> ::= <identifier>
<initial value> ::= <expression>
<final value> ::= <expression>
<with statement> ::= WITH <record variable list> DO <statement>
<record variable list>

::=

<open statement> ::= OPEN ( <open parameters>
<open parameters> ::= <file parameters>
I <keyword parameters> { , <keyword parameters>
<keyword parameters> - -= FILE VARIABLE := <file variable>
T FILE NAME := <file name>
I HISTORY := <file status>
I RECORD LENGTH := <record length>
I ACCESS-METHOD := <record access mode>
I RECORD-TYPE := <record type>
I CARRIAGE CONTROL := <carriage control>
<file name> ::= <variable>

<constant identifier>
I I <file specification> '

<file status> ::= OLD I NEW I UNKNOWN I READONLY
<record length> ::= <unsigned integer>
<record access mode> ::= SEQUENTIAL I DIRECT
<record type> ::= VARIABLE I FIXED
<carriage control> ::= CARRIAGE I NOCARRIAGE I LIST

C-8

SYNTAX SUr-1MARY

""=

<RMS file specification> ::= ,

<file specification> '

< file status>

""=

::=

."=

I <empty>

C-9

I <empty>

APPENDIX D
SUMMARY OF PASCAL-20 EXTENSIONS TO PROPOSED ISO STANDARD

Category

Extension

Lexical and
syntactical

Semi reserved words:
REM

MODULE, OTHERWISE,

Exponentiation operator (**)
REM operator
Binary, hexadecimal, and octal
for integers

notation

Double-precision real data type
Dollar sign
($)
and
underscore
characters in identifiers

()

Identifiers of up to 31 characters
Predefined types

SINGLE, DOUBLE

Predeclared Procedures

CLOSE,
DATE,
FIND,
HALT,
OPEN, TIME, MARK, RELEASE

READ, READLN, WRITE,
and WRITELN extensions

Parameters of character string
and
enumerated types for READ and READLN

LINELIMIT,

Parameters of enumerated types for WRITE
and WRITELN
READ, READLN, WRITE,
and WRITELN extensions

Optional carriage-control specification
for text files with WRITE and WRITELN

Declarations

variable initialization

Statements

OTHERWISE clause in CASE statement

Procedures and Functions

External
procedure
declar,ation

and

function

Support for calling externally declared
FORTRAN subroutines and for declaration
of PASCAL subroutines that can be called
by FOR'TRAN
Compilation

MODULE
capability
for
combining
declarations
and
definitions to be
compiled
independently from
the main
progralltl
D-I

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

----

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

APPENDIX E
ISO COMPLIANCE

This appendix is a statement of the compliance of PASCAL-20 with
PASCAL standard ISO 7185.
It is divided into four -sections:

the

Implementation-defined features.
This is a
list of the
features
of
PASCAL
that
must
be
defined by each
implementation, but which may vary between implementations.

Implementation-dependent features.
This is a
list of the
features of PASCAL which mayor may not be defined by an
impl emen ta tion.

Errors. This is an explanation of how PASCAL-20 handles each
of the errors defined by ISO 7185.

Exceptions and restrictions. This is a list of violations of
the
standard
which
PA$CAL-20
does
not detect,
and
restrictions imposed by PASCAL-20.

This appendix does not list extensions to
the ISO 7185 standard
implemented by PASCAL-20.
These extensions are listed in Appendix D.

E.1

IMPLEMENTATION-DEFINED FEATURES

The ISO 7185 standard leaves the exact definition of a number of
language elements up to the implementation.
Following is a list of
those features, and the definition given them by PASCAL-20.
1.

The predefined type
charact€!r set.

CHAR

The identifier MAXINT denotes a value of 2**35-1,
the maximum integer value of a PDP-lO word.

In a WRITE or WRITELN statement, the default field-width
for
integers is 10, for real numbers is 16, and for Booleans is
16.
The number of digits written in the exponent of a
floating-point format real is 2.
On output to a text file,
the Boolean values TRUE and FALSE are translated
to the
uppercase character strings "TRUE" and "FALSE" respectively.

E-l

corresponds

the

7-bit

ASCII

which

ISO COMPLIANCE
4.

A program parameter that is a
file variable
is bound at
runtime
to an external (operating system) file.
An external
file may exist prior to execution of the program and
is not
deleted when
the program exits.
PASCAL-20 determines which
file specification to
use when opening
the
file
in
the
following manner:
•

If an OPEN statement is executed
prior
to a
RESET or
REWRITE of the file, the file specification given in the
OPEN statement is used.

•

If no OPEN statement is executed, the RESET or REWRITE
looks
for
a logical name with the same name as the file
va ria b 1 e .
I f i t fin d s the ]. og i cal
n arne,
t ha t
n arne
is
used as the file specification.

•

Otherwise, the variable name is used as
and an extension of .DAT is assumed.

the

file

name,

The ISO 7185 standard also states that the precise actions of the
I/O
procedures REWRITE,
PUT,
RESET,
GET,
and PAGE are implementation
defined.
In the following items, F is assumed to be a file variable,
and
the various
preand
post assertions
for each operation are
assumed to be true.
1.

REWRITE(F) creates or supersedes an
and opens the file for writing.

PUT(F), where F is a text file, writes the contents of the
file
buffer variable to the output buffer, and advances the
buffer pointer;
no file operation is performed.

PUT (F) , where F is not a text file, writes
the file buffer variable to the file.

RESET(F) opens an existing operating system file for reading,
reads
the
first
record,
and assigns a value to the file
buffer variable.

GET (F) , where F is a
text file,
advances the
file
input
buffer pointer.
If EOLN (F) is true, the next record is read
from the file,
and
the buffer pointer
is reset
to
the
beginning of the buffer.

GET (F) , where F is not a text file,
from the file.

PAGE(F) first flushes the output buffer to the
file.
Then,
it causes the file to skip to the top of the next page;
the
mechanics of this depend
on
the
setting of the
file's
carriage-control
attribute.
If this attribute is LIST (the
default), a form feed character (control-L) is written to the
file.
If
the attribute
is CARRIAGE or FORTRAN, a '1' is
written to the file.
If the attribute is NOCARRIAGE or NONE,
an error is generated.

E-2

operating

reads

the

system

file,

contents

record

ISO COMPLIANCE
E.2

IMPLEMENTATION-DEPENDENT FEATURES

These language elements are defined
by the
ISO 7185 standard as
"possibly differing between processors and not necessarily defined for
any particular processor." Following is a description of how PASCAL-20
treats each of these elements.

E.3

The order of evaluation of array
order (t.hat is, left to right).

indices

their

lexical

The order of evaluation of members of a
set constructor
their lexical order (that is, left to right).

The order of evaluation of the operands of a dyadic operator
(for example, '+') depends on the complexity of each operand.
Both operands are always evaluated
(even
in
Boolean
expressions which could be "short-circuited").

The order of evaluation, accessing,
and binding of actual
parameters to
formal
parameters in a procedure or function
call is their lexical order (that is, left to right).

The varjable in an assignment statement is accessed after the
evaluation of the expression.

The effect of reading a file at the point where a
PAGE was
executed when writing
the
file
is to
return either a
form-feed character (control-L) or the digit
'1', depending
on the status of the file when the PAGE was executed.
(See
above for a description of PAGE's actions.)

Program parameters which are not file variables
defined meaning in PASCAL-20, and cause an error.

have

ERROR HANDLING

This section describes how the PASCAL-20 compiler and run-time system
detect violations of level 1 of the standard proposed by the ISO for
the PASCAL language.
Errors detected at run time cause a
program to
terminate and
return appropriate error messages.
Errors described
here as "not detected" cause a program to produce unexpected results.
The type of an index value is not assignment compatible with the index
type of an array.
Explanation:

Detected at run time if checking was enabled during
compilation.

The current variant changes while a reference to it exists.
Explanation:

Not detected.
An example of a
reference
to a
variant is the passing of the variant to a formal
VAR parameter.

IlIJ'he value of a variable to which a pointer refers (p)
Explanation:

Detected if checking was enabled at run time.

The val ue of a va r iable to wh ich a po inter refers (p)
Explanation:

is NIL.

Not detected.

E-3

is undefined.

ISO

CO~PLIANCE

The DISPOSE procedure
is calJed
to dispose
of
variable while a reference to the variable exists.
Explanation:

heap-allocated

Not detected.
Examples of
such
references are:
passing
the variable or
a
component of it to a
formal VAR parameter, or using the
variable
in
a
WJTH statement (if the variable is a record).

The value of file f changes while a reference to
Explanation:

exists.

Not detected.
An example of a reference to
f~
is
the passing of f- by reference to a routine;
until
the routine has
ceased
execution,
you may not
perform any operation on file f.

The ordinal type of an actual parameter is not
assignment
with the type of the corresponding formal parameter.
Explanation:

compatible

Detected at run time if checking was enabled during
compilation of the called routine.

The set type of an actual parameter is not assignment compatible
the type of the corresponding formal parameter.
Explanation:

Detected at run time if checking was enabled during
compilation of the called routine.

A file is not in Generation mode when a PUT, WRITE, WRITELN,
procedure is attempted.
Explanation:

when

PUT,

WRITE,

WRITELN,

PAGE

WRITELN,

Detected at run time.

The value of
the
file
procedure is attempted.
Explanation:

PAGE

Detected at run time.

The result of an EOF function is not TRUE when a PUT, WRITE,
or PAGE procedure is attempted.
Explanation:

Detected at run time.

A file is in Undefined mode
procedure is attempted.
Explanation:

with

buffer

variable

undefined

when

PUT

Not detected.

A file is in Undefined mode when a RESET procedure is attempted.
Explanation:

Not detected.

A file is not in Inspection mode when a GET, READ, or READLN procedure
is attempted.
Explanation:

Detected at run time.

A file is in Undefined mode when a GET, READ, or READLN
attempted.
Explanation:

Detected at run time.

The result of an EOF function is TRUE when
procedure is attempted.
Explanation:

procedure

Detected at run time.
E-4

GET,

READ,

READLN

ISO COMPLIANCE
The type of the file buffer is not assignment compatible with the type
of the variable that is a parameter to a READ or READLN procedure.
Explanation:

Detected at run time if checking is enabled
compilation at the READ or READLN.

during

WRITE
or WRITELN
The type of the
expression being written by a
the type of the file
procedure
is not assignment compatible with
buffer variable.
Explanation:

Detected at run time if checking is enabled
compilation of the WRITE or WRITELN.

The current variant does not exist in the list of
with the NEW procedure.
Explanation:

variants

during

specified

Not detected.

The DISPOSE(p) procedure is called to deallocate a
pointer variable
that was created using the variant form of the NEW procedure.
Explanation:

Not detected.

specify the
The variant form of
the
DISPOSE procedure does not
disposal
of the
same number of variants that were created by the
variant form of the NEW procedure.
Explanation:

Not detected.

The variant form
of
the
DISPOSE procedure does not
specify the
disposal of the same variants that were created by the variant form of
the NEW procedure.
Explanation:

Not detected.

The value of the parameter to the DISPOSE procedure is NIL.
Explanation:

Detected at run time.

The value of the parameter to the DISPOSE procedure is undefined.
Explanation:

Usually detected at run time.

A variant record created by the NEW procedure is accessed as a
rather than one component at a time.
Explanation:

Not detected.

In the PACK(a,i,z} procedure, the type of the index
assignment compatible with the index of a.
Explanation:

when

not

the

value

least

one

greater

than

the

upper

Not detected.

The index value i in the PACK procedure
bound of the index of a.
Explanation:

value

Not detected.

The PACK procedure is attempted
component of a is undefined.
Explanation:

whole,

Not detected.

E-5

ISO COMPLIANCE
In the UNPACK(z,i. ,a) procedure, the type of the index value i
assignment compatible with the index type of a.
Explanation:

the

upper

the

value

least

one

Not detected.

The resulting value of SQR(X)
Explanation:

than

Not detected.

The UNPACK procedure is attempted when
component of z is undefined.
Explanation:

not

Not detected.

The inctex value i in the UNPACK procedure is greater
bound of the index type of a.
Explanation:

does not exist.

Detected at run time.

In the expression LN(X), the value of X is negative.
Explanation:

Detected at run time.

In the expression, SQRT(X), the value of X is negative.
Explanation:

Detected at run time.

The resulting va]ue of TRUNC(X) does not exist after
the
following
calculations have been done:
if the value of X is positive or zero,
then 0 <= X-TRUNC (X) <1;
otherwise, -1 < X-TRUNC (X) <= o.
Explanation:
The resulting
calculations
ROUND(X)
is
equivalent to

Detected at run time.

value of ROUND(X) does not exist after
the
following
have been done:
if the value of X is positive or zero,
equivalent
to TRUNC(X+O.5);
otherwise,
ROUND(X)
is
TRUNC(X-O.5).

Explanation:

Detected at run time.

The resulting value of CHR(X)
Explanation:

Not detected.

The resulting value of SUCC(X)
Explanation:

is called when the file f is undefined.
Detected at run time.

The function EOLN(f)
Explanation:

does not exist.

Detected at run time if checking was enabled during
compilation.

The function EOF(f)
Explanation:

does not exist.

Detected at run time if checking was enabled during
compilation.

The resulting value of PRED(X)
Explanation:

does not exist.

is called when the file f is undefined.

Detected at run time.

E-6

ISO COMPLIANCE
The function EOLN(f)
Explanation:

is called when the result of EOF(f)

is TRUE.

Detected at run time.

A variable is not initialized before it is first used.
Explanation:

Not detected.

In the expression X/Y,
ExplAnation:

the value of Y is zero.

Detected at run time.

In the expression I DIV J,
Explanation:

is zero.

Detected at run time.

In the expression T MOD J,
Explanation:

the value of J

is zero or negative.

Detected at run time if J
J is negative.

is zero;

An operation or function involving integers does not
mathematical rules for integer arithmetic.
Explanation:

conform

the

control

Detected at run time.

A function result is undefined when the function
the calling block.
Explanation:

not detected if

returns

Not detected.

The ordinal type of an expression is not assignment compatible with
the
type
of
the
variable or
function
identifier
to which it is
assigned.
Explanation:

Detected at run time if checking was enabled during
compilation.

The set type of an expression is not assignment compatible with
the
type of the variable or functjon identifier to which it is assigned.
Explanation:

Detected at run time if checking was enabled during
compilation.

None of the case labels is equal in value to the case
CASE statement.
Explanation:

selector

Not detected.

In a FOR statement, the type of the initial value
is not assignment
compatible with the type of the control variable, and the statement in
the loop body is executed.
Explanation:

Detected at run time if checking was enabled during
compilation.
Assignment
compatibility
is not
enforced if the statement
in
the
loop body can
never be executed.

E-7

ISO COMPLIANCE
In a FOR statement, thp type of the
final
value
is not assignment
compatible with the type of the control variable and the statement in
the loop body is executed.
Explanation:

Detected at run time if checking was enabled during
compilation.
Assignment
compatibility is not
enforced if the statement in the loop body can
never be executed.

When an integer is being read from a text file, the ~igits read do not
constitute a valid
integer value.
(Initial spaces and end-of-line
markers are skipped.)
Explanation:

Detected at run time.

When an integer is being read from a text file, the type of the value
read is not assignment compatible with the type of the variable.
Explanation:

Detected at run time.

When a real number is read from a text file, the digits read do not
constitute a valid
real
number.
(Initial
spaces and end-of-line
markers are skipped.)
Explanation:

Detected at run time.

The value of the file buffer variable is
READLN procedure is performed.
Explanation:

undefined

when

READ

Not detected.

A WRITE or WRITELN procedure specifies a
field width in which the
integers representing
the total
width and the number of fractional
digits are less than 1.
Explanation:

Detected at run time.

The bounds of an array passed to a conformant array parameter
outside the range specified by the conformant array's index type.
Explanation:

E.4

are

Not detected.

EXCEPTIONS AND RESTRICTIONS

PASCAL-20 is not able to detect a number of violations of the ISO 7185
standard. Violations that are not defined as errors are listed below.
1.

Statements threatening a FOR-loop control variable are not
detected.
A statement "threatens" a variable if it occurs
within the FOR-loop or within a prior procedure, and if it is
one of the following kinds of statements:
•

A READ or READLN statement containing the variable

•

An assignment statement with the variable to the left

the' :='

•

A procedure or function
formal VAR parameter

E-8

call

with

the

variable

ISO COMPLIANCE
2.

The concept of a "totally-undefined" variable has no meaning
in PASCAL-20.
Any violation of the standard which requires
"totally-undefined" to be meaningful is not detected.

Variant records in PASCAL-20 are
implemented
by overlaying
the
fields of the variants in the same memory.
Changing the
active variant does not alter the contents of
that memory,
and the fields of the newly-active variant are not undefined.
Also, note that variant parts of records cannot be of
the
type FILE.

The DISPOSE procedure does not cause its argument
undefined, but rather the argument is set to NIL.

No restrictions on the relative placement of labels and
statements are enforced.

Packed and unpacked sets are implemented in similar ways
in
PASCAL-20.
Therefore,
the compiler
is very "loose" about
interchanging packed and unpacked sets;
the only requirement
imposed
on compatibility of set
types is that their base
types must be compatible.

Under certain conditions, a procedure call may be bound
to
the wrong
procedure definition.
This can only happen when
two procedures are declared with the same name at different
levels,
and
the
inner procedure
is called
before it is
declared.
The following example illustrates this situation.

become
GOTO

PROCEDURE B;
BEGIN (* level 1 *)
END;
PROCEDURE C;
PROCEDURE D;
BEGIN
B

(* call bound to level 1 procedure *)

END;
PROCEDURE B;
BEGIN (* level 2
END;
BEGIN
END;

PASCAL-20 also imposes a number of restrictions
are:

programs.

These

Identifiers may be of any length,
but only the
first
31
characters are
stored.
Therefore identifiers in PASCAL-20
must be unique in the first 31 characters.

The ordinal values of the base type of a set type are limited
to
the
range 0 •• 255.
This means that no set may have more
than 256 elements.

File variables are not permitted as
part of a record.

The range of ordinal values of case
statement must be less than 1000.

E-9

fields

constants

the

variant

CASE

APPENDIX F
DIFFERENCES BETWEEN PASCAL-20 AND VAX-II PASCAL

PASCAL-20 VI and VAX-II PASCAL V2 are compatible languages;
that is,
you can compile and
run a PASCAL-20 program on a VAX/VMS operating
system, and you can compile and run a VAX-II PASCAL program on a
TOPS-20
operating system.
However, there are differences between the
languages.
The VAX-I] PASCAL language (V2) contains features that
the PASCAL-20
language does not.
Therefore,
if you plan
to
transport PASCAL
programs from the VAX/VMS operating system to
the TOPS-20
operating
system,
you should avoid
using
these additional
features.
This
appendix lists the additional features that VAX-II
PASCAL
(V2)
has
over PASCAL-20 (VI).

F-l

DIFFERENCES BETWEEN PASCAL-20

~ND

VAX-Il PASCAL

Table F-I Jists the language eJements that VAX-II PASCAL V2
PASCAL-20 does not:
Table F-I:

has

that

Additional Language Elements

Category

Language Element

Special Symbol

Type cast operator

Nonstandard
Reserved words

%DESCR
%IMMED
%STDESCR
%REF
VARYING

Predeclared Identifiers

~DD INTERLOCKED
ADDRESS
BITNEXT
BITSIZE
CLE~R INTERLOCKED
DBLE
DELETE
EST.~BL ISH
FINDK
INDEX
INT
LENGTH
LOCATE
LOWER
NEXT
PAD
QUAD
QUADTRUPLE
READY
RESETK

Data Types

UNSIGNED: a through 2**~2 - 1
o through 42949n7295

Formal Parameter List

foreign section

Structured Type

VARYING OF CHAR

Real Types

G_floating
QUADRUPLE

String Operators

Plus sign (+)

Predeclared Procedures

DELETE(f,e)
ESTABLISH(id)
FINDK(f,kn,kv,m,e)
LOCATE(f,n,e)
READV (s ,vI, .•• ,vn)
RESTK(f,kn,e)
REVERT
TRUNCATE(f,e)
UNLOCK ( f , e)
UPDATE(f,e)
WR I T EV ( s , pI, • • • , p n )
F-2

REVERT
SET INTERLOCKED
SIZE
STATUS
SUBSTR
TRUNCATE
UAND
UFB
UINT
UNLOCK
UNOT
UNSIGNED
UOR
UPDATE
UPPER
GROUND
UNTRUNC
UXOR
WRITEV

DIFFERENCES BETWEEN PASCAL-20 AND VAX-II PASCAL

Table F-I: Additional Language Elements (Cont.)
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - , , - - - -'.-- ---<"-,-------

Category
Language Element
- - - - - - - - - - - - - - - - - - , , - - , - - - - - - - - - - - - - - - - - - - - , - - , - -,
Predeclared Functions

See Table F-2

OPEN
Parameters

Procedure

DISPOSITION
SHARING
USER ACTION
ERROR
KEYED

CLOSE
Parameters

Procedure

DISPOSITION
USER ACTION
ERROR

-',"

The following predeclared procedures are contained
in both PASCAL
languages, however, VAX-II PASCAL offers an additional argument:
e

= error parameter

These predeclared procedures are:
FIND(f,n,e)
OPEN(f,parameters,e)
PAGE(f,e)
PUT(f,e)
READ(f,vl, ... ,vn,e)
READLN(f,vl, ... ,vn,e)
RESET(f,e)
REWRITE(f,e)
WRITE(f,pl, •.. ,pn,e)
Whenever you see rfattribute-listl] in a
syntax description
in
the
VAX-Il
PASCAL language,
it signifies
that you can add
certain
arguments to the syntax depending on the type of circumstances.
Note
the following syntax and attribute descriptions:
Attribute-list
({identifier

(

{

constant-expression }, ... ) ]

},.•.. ]

In the PASCAL-20 language,
only a
PROGRAM,
MODULE,
PROCEDURE,
or
The only attribute for a PROGRAM
FUNCTION can have an attribute-list.
or MODULE is OVERLAID.
The only attributes
for
a
PROCEDURE or
FUNCTION are GLOBAL and FORTRAN.
Note that in PASCAL-20 the GLOBAL
attribute can not be followed by an identifier.

F-3

DIFFERENCES BETWEEN PASCAL-20 AND VAX-II
Table F-2 lists additional predeclared functions
provides but P~SCAL-~O does not:
Table F-2:

P~SCAL

that

PASCAL

Additional Predeclared Functions

Category

Function

Boolean

UFB(f)

Transfer

DBLE(x)
QUAD (x)
UINT(x)
UROUND (r)
UTRUNC (r)

Dynamic Allocation

ADDRESS(x)

Character String

INDEX(sl,s2)
LENGTH(s)
PAD(s,fill,l)
SUBSTR(s,b,l)

Un signed

UNAD(ul,u2)
UNOT (u)
UOR(ul,u2)
UXOR(ul,u2)

Allocation Size

SIZE(x,cl, ... ,cn)
NEXT (x)
BITSIZE(x)
BITl\JEXT (x)

Low Level
Interlocked

ADD INTERLOCKED(e,v)
SET-INTERLOCKED(b)
CLEAR INTERLOCKED(b)

Miscellaneous

STATUS (f)

If you plan
to
transport VAX-II PASCAL programs
PASCAL-20 programs to VAX/V~1S, note the following:
•

VAX-II

TOPS-20,

The precision for INTEGER data type:
TOPS-20

( .- 2 * * ~ 5 ) t h r 0 ug h ( + 2 * * 3 5 ) -1
through +34359738367

-34~59738368

VAX/VMS
•

•

(-2**31)+1 through (+2**31)-1
-2147483647 through +2147483647

The maximum number of items in an enumerated type:
TOPS-20

Depends on amount of memory available

VAX/VMS

65,535

PASCAL-20 has two procedures MARK and RELEASE that are used
in conjunction with NEW and DISPOSE.
VAX-II PASCAL does not
have MARK and RELEASE.

If you have any further questions concerning VAX-II PASCAL,
the VAX-II PASCAL documentation set.
F-4

refer

APPENDIX G
PROCEDURE AND FUNCTION CALLING SEQUENCES

This appendix]
describes the calling sequences and
conventions used
by PASCAL for user-defined procedures and functions.
This information
is particularly useful in writing MACRO routines that
interface with
PASCAL programs.
Note that, in this appendix,
procedures and
functions.
explicitJy noted.

G.I

"procedure"
refers to both
the word
in the case of functions are
Exceptions

RUN-TIME STACK

The majority of
~un-time
information
is kept on the
stack;
in
particular, the stack contains all local variables, parameters, static
and dynamic links, and return addresses for
procedure calls.
The
stack pointer
is kept
in AC 17.
A frame pointer, used for most
variable accessing, is kept in AC 16.
The frame pointer is fixed over
the course of a procedure, while the stack pointer can change.
Figure
G-l shows the status of the stack just after a PUSHJ to a procedure.
=======~==================

AC 0 -)

1111111I111111111111111111
I previous stack frame

1111111I111111111111111111

(--

=========~====================

I parameter values,
I addresses, descriptors I

I
I

=======~==================

I saved stack pointer

)--

=======~==================

ACl7 -)

I caller's return addressl
=======~==================

Figure G-l:

Status of Stack After PUSHJ

The information in this appendix is
Moody @ 1982.
G-I

copyrighted

Scott

Arthur

PROCEDURE AND FUNCTION CALLING SEQUENCES
Figure G-2 shows the stack frame in its entirety,
first executable statement in a procedure.

1111111111111111111111111I

I previous stack frame

11111111111111111111111111
parameter values,
addresses, descriptors
saved stack pointer

I
I
I

the

procedure,

the

I
I
I

)--

I
I
I
I

caller's return addressl
dynamic link

prior

<----<-- I

==========================

AC16 -)

just

)-----

static link
function value
(if a
function)
2 words

I local variables

11111111111111111111111111
oversized value
parameters
saved ACs 2 -

size of conformant
value parameter space
conformant array value
parameters

AC17 -)

next available stack
space
Figure G-2:

G.2

Stack Frame

MECHANICS OF A PROCEDURE CALL

When PASCAL generates a
call
to
a
following general sequence is followed:

Save the current stack

Push the parameters onto the
Passing") •

Push the

s~ved

point~r

user-defined
in AC
stack

o.
(see

below,

stack pointer onto the stack.

G-2

"Parameter

PROCEDURE AND FUNCTION CALLING SEQUENCES
4.

Copy the current frame
dynamic link.

Determine the frame pointer of the block lexically
the called procedure, and copy it into AC
o

pointer

into

Th i s i s

enclosing

This is the static link.

Call the procedure by means of a PUSHJ.

After the called procedure returns,
the
stack
is
restored
to
previous state.
PASCAL expects the called procedure to preserve
accumulators except ACs a and~.
(Note particularly that AC
16,
f ram e
po i n t e r ,
mu s t
be
res tor e d . )
I n t h e cas e 0 f a f un c t ion,
function value is returned in AC a (P,Cs a and 1 if the function is
type DOUBLE).

G.3

the

its
all
the
the
of

PARAMETER PASSING

How a parameter is passed depends on what kind of parameter it is and,
for
a
value
parameter, how large it is.
The five different methods
used to pass parameters are:
1.

Value parameter by value

Value parameter by address

3•

Re fe renCE! (VAR)

4•

Procedure/function parameter

Conformant array parameter

parameter

Each of these methods is described be]ow.

G.3.1

Value Parameter Passed By Value

A value parameter passed by value is simply pushed
onto
the
stack.
This method
is
used for sets, DOUBLE, and any type that fits into a
single word, including integer, real,
char,
pointers,
scalars,
and
small packed arrays and records.

G.3.2

Value Parameter Passed By Address

Any value parameter larger than one word, other than a DOUBLE or
set
parameter,
is passed
by address.
The address of the parameter,
without indirection or indexing, is pushed onto the stack.
It is
the
responsibility of the called procedure to copy the parameter into its
own local storage.

G-3

PROCEDURE AND FUNCTION CALLING SEQUENCES

G.3.3

Reference (VAR) Parameter

All VAR parameters are
passed
by address.
The address of
the
parameter,
without indirection or indexing, is pushed onto the stack.
Thus, it is perfectJy safe (and recommended) to access the contents of
a VAR parameter using indirection.

G.3.4

Procedure Or Function Parameter

Procedures and
functions,
as
parameters,
are
passed
as
two-word
descriptors.
The
first word of the descriptor is the address of the
procedure's entry point;
the second is its static link.
The static
link
is needed when the procedure is actually called (see above under
"Mechanics of a Procedure Call").

G.3.5

Conformant Array Parameter

A conformant array can be of any size;
however,
in order
for
parameters
to have fixed offsets from the frame pointer, a fixed-size
descriptor must be passed.
This oescriptor contains a fixed
part of
two words,
and
a
variabJe part of three words for each conformant
dimension.
The fixed part contains the base address of the
array
in
the
first
word.
The second word contains the overall size of the passed array
in words.
Each three words of the variable part contain
information on one
conformant dimension.
The second and third words contain the lower
and upper bounds, respectively, on the index of
the dimension.
The
first
word
contains
the
size of the dimension, which is the upper
bound minus the lower bound plus one.
The conformant array descriptor is the same for
both
value and VAR
parameters.
It is the responsibility of the called procedure to copy
value conformant parameters into its own local storage.

G-4

PROCEDURE
G.4

~ND

FUNCTION CALLING SEQUENCES

PARAMETER ACCESSING EXAMPLE

Figure G-3
gives an example of a
typical
external
procedure
declaration, followed by MACRO code showing how the various parameters
can be accessed.
PROCEDURE Test(first: DOUBLE; VAR second! INTEGER; third:
ENTRY

TE~3T

TEST:

MOVE

1,,"':?(1./)

MOVEI

2,(~··~:·3(17)

MOVE
MOVE

29 .... :3(1.7)

DMOVE

4, .... !}(:L7)

MOVEM

:'3,t~""3(17)

PDP,j

1?~

Figure G-3:

:3¥(~"'<3(17)

INTEGER); EXTERNAL;

;Define as ~lobal symbol
;Save ACs in local stora~e
;Get value of THIRD
;Get address of SECOND
;Get address of SECOND (alternate method)
;Get value of SECOND
;Get (2-word) value of FIRST
;Perform computations
iReturn new value of SECOND
; F~f?S to re ACs
;Return to caller

External Procedure Declaration

Saving the ACs is not shown;
they are assumed to be preserved in some
static locations,
not on the stack.
The stack pointer is used for
accessing the parameters.
(If the ACs are saved on the stack, or
if
the stack is used for other purposes, then another AC should be set up
for accessing parameters.)
Offset a (from AC 17) is t.he return address;
offset -1 is the
saved
stack pointer.
Therefore, offset -2 is the last parameter location.
Since THIRD is an integer, it takes up only one word, and is accessed
with an offset of -2.
The next previous parameter, SECOND, is a VAR
parameter, passed by address.
The address is
in the next previous
word,
at offset -3.
FIRST is a double-precision real, taking up two
words.
Therefore, offset -4 is the second word of the parameter,
and
offset -5 is the first word.

G .5

CONFORMANT AR.RAY EXAMPLE

Figure G-4 shows an example of an external procedure decl~ration with
a
conformant array parameter, and a picture of the stack after the
call.
The base address of the array is at offset -6;
the length
in
words is at offset -5.
The lower and upper bounds on the index are at
offsets -3 and -2, respectively.
The number of elements is at offset

-4.

G-5

PROCEDURE AND FUNCTION CALLING SEQUENCES
PROCEDURE Confar(VAR a: ARRAyrJa .• ha:INTEGER] OF INTEGER); EXTERNAL;

-6 (l 7)

array base address

fixed

-5 (l 7)

size

part

-4 (17)

# elements

-3(17)

lower bound

( ] a)

(per dimension)

-2 (17)

upper bound

(ha)

pa rt

-1 (l 7)

saved stack pointer

0 far ray

(wo rd s)

(ha-la+l)

o (l7)

return C'tddress

Figure G-4:

Conformant ArrAY Parameter

G-6

variable

INDEX

-A/ABORT switch, 8-7
ABS function, 1-7, 6-13
Actual parameter, 5-14
AND operator, 1-6 3-5
ARCTAN function, ~-7, 6-13
Arithmetic expression, 3-1
Arithmetic functions, 6-13
Arithmetic operator, 1-8, 1-1
ARRAY, 1-0
ARRAY ~ata type, 2-8
Array examples, 2-15
Array type compatibility, 2-14
ASCII character set, 2-7, B-3
ASSIGN command, 9-3
Assigning
a string constant, 2-12
a string variable, 2-11
values to an array, 2-12
values to records, 2-19
Assignment operator, 1-8
Assignment statement, 2-12, 5-2
Asterisk, 1-8
j'

-BBACKUS-NAUR form, C-l
Base type, 2-6, 2-22
BEGIN, 1-0, 1-9
Binary notation, 2-2
/BINARY switch, 8-7
Block, 1-2, 1-4
BNF syn tax, C-l
BOOLEAN data type, 1-7, 2-2, 2-4
Boolean functions, 6-14
BOOLEAN results, 3-5
BREAK command, 9-4

-cCalling sequences
function, G-l
procedure, G-l
CARD function, 1-7, 6-15
Caret, 1-8
CARRIAGE, 7-11

Carriage control, 7-11
Carriage control characters, 7-20
Case selector, 5-0
CASE statement, ]-0,2-17,5-4,
5-h
Case-label list, 2-17, 5-0
CHAR data type, 1-7, 2-2, 2-4
Character set, 1-5
/CHECK switch, A-7, A-8
CHR function, 1-7, h-15
CLEAR command, 9-6
CLOCK function, 1-7, 6-15
CLOSE procedure, 1-7, G-5, 7-4
Collection of elements, ?-8
Colon, 1-8
Comma, 1-8
Command
ASSIGN, 9-3
BREAK, 9-4
CLEAR, 9-6
DISPLAY, 9-7
EXECUTE, 8-17, 9-1
/EXIT, 8-5
EXIT, 9-9
/HELP, 8-5
HELP, 9-10
LOAD, 8-ln, 9-1
PROCEED, 9-10
RE~OVE, 9-11
/RUN, 8-5
RUN, 8-1 7
SAVE, 8-1 7
SET, 1-6, 9-11, 9-12, 9-13
SHOW, 9-14
START, 8-17, 9-1
/TAKE, A-5, 8-6
TRACE, 9-15
Comment, 1-8, 1-9
Compilation unit, 1-5
Compile-time errors, A-3
Compile-time warnings, A-18
Compiler listing format, 8-16
Compiling a program, 8-4
Component type, 2-24
Components, 2-8
Compound statement, 5-1, 5-2
Conditional statement, 5-4

Index·-l

Conformant array, h-25
CONST, 1-6, 4-1, 4-4
Constant definition, 4-4
Constant identifier, 4-4
Constant name, 4-4
Control variable, 5-8
COS function, 1-7, 6-13
Creating a program, 8-3
/CREF switch, 8-7, 8-8
/CROSS-REFERENCE switch, 8-7, 8-8

-DData type, 2-1, 4-5, 4-f)
ARRAY, 2-8
BOOLEAN, 1-7, 7.-2, 2-4
C HA R ,

1 - 7,

2 - 2,

2- 4

DOUBLE, 1-7, 2-2
enume ra ted, 2-4
FILE, 2-8, 2-24
INTEGER, 1-7, 2-2
pointer, 2-1, 2-2n
predefined, 2-2
REAL, 1-7, 2-2, 2-3
RECORD, 1-6, 2-8, 2-15
scalar, 2-1, 2-2, 2-4
SET, 2-8, 2-22
SINGLE, 1-7, 2-2
structured, 2-1, 2-8
subrange, 2-6
Data types, 1-2
DATE procedure, 1-7, 6-5, 6-12
Deallocating memory, 6-6
/DEBUG switch, 8-7, 8-8, 9-1
Debugger, 9-1
Debugging a program, 9-1
Decimal notation, 2-2
Declaration, 1-4
Declaration section, 4-1
Declaring a data type, 2-1
Definition, 1-4
Delimiter, 1-8
Direct access, 7-3, 7-10
DISPLAY command, 9-7
DISPOSE procedure, 1-7, f)-5, 6-6
DIV operator, 1-6
Division, 1-8
DO, 1-6, 5-8, 5-11, 5-12
Documenting your program, 1-9
Dollar sign, 1-7
DOUBLE data type, 1-7, 2-2
Double precision, 2-4
DOWNTO, 1-6, 5-8
Dynamic allocation procedure, 6-6
Dynamic variable, 2-26, 6-6

-EELSE clause, 1-6, 5-5
END, 1-6, 1-9
Enumerated data type, 2-5
EOF function, 1-7, 6-14
EOLN function, 1-7, 6-14

Equal, 1-8
/ERROR-LIMIT switch, 8-7, 8-8
Examples, 8-17
EXE file, 8-17
Executable statement, 1-4
EXECUTE command, 8-17, 9-1
Executing a program, 8-17
/EXIT command, 8-5
EXIT command, 9-9
EXP function, 1-7, h-13
EXPO function, 1-7, 6-16
Exponentintion, 1-8, 3-2
Expression, 3-1
a r i t hm e tic, 3-1
logical, 3-5
relational, 3-4
set, 3-5
Expression compatibility, 2-29
Extensions, D-1
EXTERNAL, f)-36
External file, 2-25, 4-2
External subprogram, f)-I, n-3h
-F-

FALSE, 1-7
Field identifier, 2-15
FILE, I-f)
File
TEXT, l-7
File characteristics, 7-2
FILE data type, 2-8, 2-24
File examples, 7-12
File history, 7-10
File organization, 7-3
File specification defaults, 8-3
File specifications, 8-2
File status, 7-10
File variable, 7-16
Final value, 5-8
FIND procedure, 1-7, 6-5, 7-4
FIXED, 7-11
Fixed length, 7-3
Fixed-length record, 7-3
/FLAG-NON-STANDARD switch, 8-7,
8-8
Floating-point format, 2-4
FOR statement, 1-6, 5-8
Formal and actual parameter
compatibility, 2-29
Formal parameter, 6-20, 6-29,
6-32
FORTRAN, 6-29, 6-32, 6-36
FORWARD, 6-29, 6-32
FORWARD declaration, 6-35
FUNCTION, 1-6, 4-1
Function, 6-1
ABS, 1-7, 6-13
ARCTAN, 1-7, 6-13
CARD, 1-7, 6 -1 5
C HR, 1-7, 6 -15
CLOCK, 1-7, 6-15
COS, 1-7, 6 -13

Index-2

-L-

Function (Cont.)
EOF, 1-7, 6-14
EOLN, 1-7, 6-14
EXP, 1-7, 6 -13
EXPO, 1.-7, 6-16
LN, 1-7, 6 -13
ODD, 1-7, 6 -14
ORD, 1-7, 2-7, G-15
PRED, 1-7, 6-1 ()
ROUND, 1-7, 6-l~)
SIN, 1-7, f)-14
SNGL, 1-7, f)-IS
SQR, 1-7, 6-14
SQRT, 1-7, 6-14
SUCC, 1-7, 6-16
TRUNC, 1-7, 6-1~)
UNDEFINED, 1-7, f)-IS
Function calling sequences, G-l
Function declaration, 6-32
Functions
a r i t hm e tic, 6 -1 3
Boolean, 6-14
transfer, 6-15

LABEL, 1-6, 4-1, 4-1
Label, 5-13
Label declaration, 4-3
Less than, 1-8
Less than or equal, 1-8
Lexical elements, 1-5
LINELIMIT procedure, 1-7, 6-5,
7-7
LIST, 7-11
/LISTING switch, 8- 7 , 8-9, 8-11
LN function, 1-7, 6-13
LOAD command, 8-16, 9-1
Loading a program, 8-16
Local file, 2-25
Logical expression, 1-5
Logical operator, 3-1, 3-5
Loops, 5-8
Lower limit, 2-6

-M/MACHINE-CODE switch, 8-7, 8-9
Mark procedure, 6-5
MAXINT, 1-7, 2-3
Mechanism specifier, 6-21
'~inus sign, 1-8
MOD operator, 1-6
'~ODULE, 1-6, 6-37
I~odul es, t) -3 7
Multidimensional array, 2-9
Multiplication, 1-8

-GGET procedure, 1-7, 6-5, 7-6
GLOBAL, 6-29, 6-32
Global file, 2-2S
GOTO statement, 1-6, 4-3, 5-13
Greater than, 1-8
Greater than or equal, 1-8
-8-

-NHALT procedure, 1-7, 6-5
Heading, 1-2, 1-4
/HELP command, 8-5
HELP command, 9-10
Hexadecimal notation, 2-2

/NATIONAL switch, 8-7, 8--9
NEW procedure, 1-7, f)-5, 6-6
NIL, 1-6
NOCARRIAGE, 7-11
Not equal, 1-8
NOT operator, 1-6, 3-5

-I-

-0-

I/O errors, A-2
Identifier, 1-2, 1-7
IF statement, 1-6
IF-THEN statement, 5-4
IF-THEN-ELSE statement, 5-4, 5-5
IN, 1-6
%INCLUDE directive, 1-10
Included file, 1-10
Index, 2-8
Initial value, 5-8
Initializing
a variable, 4-6
an array, 2-12
Initializing a variable, 4-7
INPUT, 1-7, 4-2
Input, 7-1
Input procedure, 6-5
INTEGER data type, 1-7, 2-2
Internal file, 2-25

Octal notation, 2-2
ODD function, 1-7, 6-14
OF, 1-6
OPEN procedure, 1-7, 6-5, 7-8
Operator, 3-1
AND, 1-6, 3-5
arithmetic, 1-8, 3-1
ass ignmen t, 1-8
DIV, 1-6
logical, 3-1
MOD, 1-6
NOT, 1-6, 3-5
OR, 1-6, 3-5
relational, 1-8, 2-11, 3-1, 3-4
REM, 1-6
set, 1-8, 3-1, 3-5
sub range , 1-8
OR operator. 1-6, 3-5

Index--3

INDEX
ORD function, 1-7, 2-7, 6-15

o r d e red set, 2 - 5
Ordinal value, 2-7, 2-11
OTHERWISE clause, 1-6, 5-6
OUTPUT, 1-7, 4-2
Output, 7-1
Output procedure, 6-5
OVERLAID, 6-37
-p-

PACK procedure, 1-7, ~-5, ~-10
PACKED, 1-6, 2-28
Packed array, 2-11,2-28
Packed file, 2-28
Packed record, 2-28
Packed set, 2-28
PAGE procedure, 1-7, 6-5, 7-14
Parameter, 6-21
act ua 1, 5 -1 4
formal, 6-20, 6-29, 6-32
Parentheses, 1-8
PASDDT, 9-1
PASDDT command, 9-3
Passing parameters, 6-22
Period, 1-8
Plus sign, 1-8
Pointer, 1-8
Pointer data type, 2-1, 2-26
Pointer variable, 6-6
Precedence of operators, 3-6
PRED function, 1-7, 6-16
Predecessor, 2-2
Predeclared
fun c t ion, 6 -1 2
identifier, 1-7
procedure, 6-1
subprog ram, 6-1
Predefined data type, 2-2
PROCEDURE, 1-6, 4-1
Procedure, 6-1
CLOSE, 1-7, 6-5, 7-4
DATE, 1-7, 6-5, ~-12
DISPOSE, 1-7, 5-5, 6-6
FIND, 1-7, 6-5, 7-4
GET, 1-7, 6-5, 7-6
HALT, 1-7, 6-5
LINELIMIT, 1-7, ~-5, 7-7
MAHK, 6-5
NEW, 1-7, 6-5, 6-6
OPEN, 6-5, 7-8
PACK, 1-7, 6-5, 6-10
PAGE, 1-7, 6-5, 7-14
PUT, 1-7, 6-5, 7-15
READ, 1-7, 6-5, 7-16
READLN, 1-7, 6-5, 7-18
RELEASE, 6-5
RESET, 1-7, 6-5, 7-19
REWRITE, 1-7, 6-5, 7-20
TIME, 1-7, 6-5, 6-12
UNPACK, 1-7, ~-5, 6-11
WRITE, 1-7, 6-5, 7-21
WRITELN, 1-7, 6-5, 7-25

(Cont.)
Procedure call, 5-14
Procedure calling sequences, G-l
Procedure declaration, 6-28
Procedure identifier, 6-29
Procedure name, 5-14
PROCEED command, 9-10
PROGRAM, 1-5, 4-2
Program development, 8-1
Program heading, 4-1, 4-2
Program name, 4-2
Program structure, 1-2
PUT procedure, 1-7, 6-5, 7-15

-RRandom access, 7-3
READ procedure, 1-7, 6-5, 7-16
READ statement, 2-12
READLN procedure, 1-7, ~-5, 7-18
READLN statement, 2-12
REAL data type, 1-7, 2-2, 2-3
Record access, 7-3
Record access mode, 7-10
RECORD data type, 1-6, 2-8, 2-15
Record examples, 2-21
Record format, 7-3
Record length, 7-10
Recrrrd type, 7-11
Record type compatibility, 2-19
Record variable, 5-12
Records with variants, 2-17
Relational expression, 3-4
Relational operator, 1-8, 2-11,
3-1, 3-4
RELEASE procedure, 6-5
REM operator, 1-6
REMOVE command, 9-11
REPEAT statement, 1-6, 5-8, 5-10
Repetitive statement, 5-8
Reserved words, 1-~
RESET procedure, 1-7, 6-5, 7-19
Result type, 6-32
REWRITE procedure, 1-7, 6-5, 7-20
ROUND function, 1-7, 6-15
jRUN command, 8-5
RUN command, 8-17
Run-time errors, A-I

-sSAVE command, 8-17
Scalar data type, 2-1, 2-2
Scientific notation, 2-4
Scope of identifiers, 3-7
Semicolon, 1-8
Semi reserved words, 1-6
Separate compilation, 6-37
Sepa ra to r, 1-8
Sequential access, 7-3, 7-10
SET command, 1-6, 9-11, 9-12,
9-13
SET data type, 2-8, 2-22
Set expression, 3-5

Index-4

INDEX
Set operator, 1-8, 3-1, 3--5
Setting a breakpoint, 9-4
Shifting program control, 5-4
SHOW command, 9-14
Simple statement, 5-1
SIN function, 1-7, 6-14
SINGLE data type, 1-7, 2-2
Slash, 1-8
SNGL function, 1-7, 6-15
S P e cia 1 s ym bo 1 s, 1- 8
Specifying output files, 8-11
Specifying switches in source
code, 8-9
SQR function, 1-7, 6-14
SQRT function, 1-7, 6-14
Square brackets, 1-8
START command, 8-17, 9-1
Statement
assignment, 2-12, 5-2
CASE, 1-6, 2-17, 5-4, 5--6
compound, 5-1, 5-2
conditional, 5-4
FOR, 1-'5, 5-8
GOTO, l-6, 4-3, 5-13
IF, 1-6
IF-THEN, 5-4
IF-THEN-ELSE, 5-4, 5-5
READ, 2-12
READLN, 2-12
REPEAT, 1-6, 5-8, 5-10
repetitive, 5-8
simple, 5-1
WHILE, 1-6, 5-8, 5-11
WITH, 1-6, 5-12
String comparison, 2-1l
String variable, 7-11
Structured data type, 2-1, 2-8
Subprogram, 1-4
Subprogram format, 6-20
Subprogram heading, 6-20
Subrange data type, 2-6
Subrange operator, 1-8
SUCC function, 1-7, 6-16
Successor, 2-2
Switch
/ABORT, 8-7
/BINARY, 8-7
/CHECK, 8-7, 8-8
/CREF, 8-7, 8-8
/CROSS-REFERENCE, 8-7, 8-8
/DEBUG, 8-7, 8-8, 9-1
/ERROR-LIMIT, 8-7, 8-8
/FLAG-NON-STANDARD, 8-7, 8-8
/LISTING, 8-7, 8-9, 8-11
/MACHINE-CODE, 8-7, 8-9
/NATIONAL, 8-9
NATIONAL, 8-7
/WARNINGS, 8-9
/WARNINGs, 8-7
Symbolic values, 9-1

(Co n t. )
Syntax summary, C-l

-T'rag field, 2-17
'rag name, /,-1 7
'rag type, 2-1 7
/TAKE command, 8-5, 8-6
Test condition, 5-5
TEXT file, 1-7, 2-12, 2-25
Text file, 2-25
'rHEN, 1-t;
TIME procedure, 1-7, 6-5, 6-12
TO, 1-6, 5-8
TRACE command, 9-15
Transfer functions, ~-15
'rRUE, 1-7
TRUNC function, 1-7, 6-15
Truth value, 3-5
TYPE, 1-6, 4-1, 4-5
Type compatibility, 2-29
Type definition, 4-5
Type identifier, 4-5
TYPE section, 2-1

-uUNDEFINED function, 1-7, 6-15
Underscore, 1-7
UNPACK procedure, 1-7, 6-5, 6-11
UNTIL, 1-6, 5-10
Upper limit, 2-6
User identifier, 1-8
User-defined scalar data type,
2-5

-vVALUE, 1-6, 4-1, 4-7
Value declaration, 4-7
Value initialization, 4-7
Value parameter, 6-22
VAR, 1-6, 4-1, 4-6, 6-23
VAR section, 2-1
VARIABLE, 7-11
Variable declaration, 4-6
Variable length, 7-3
Variable name, 4-6
Variable parameter, 6-23
Variant clause, 2-15

-w/WARNINGS switch, 8-9
WARNINGS switch, 8-7
WHILE statement, 1-6, 5-8, 5-11
WITH statement, 1-6, 5-12
WRITE procedure, 1-7, 6-5, 7-21
WRITELN procedure, 1-7, 6-5, 7-25

Index-5

TOPS-20
PASCAL Language Manual
AA-L31 5A-TM

READEIR'S COMMENTS
NOTE:

This form is for document 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.

D Assembly language programmer

o Higher-level language programmer
D Occasional programmer (experienced)
D User with little programming experience
D Student programmer
D Other (please specify)~~~~_~~~~~~~~~~~~~~~~~~

Name

~~_~,

________________ Date _ _ _ _ _ _ _ _ _ _ __

Organization ___,________________ Telephone .,_ _ _ _ _ _ _ __
Street _____

______________________

City _ _ _ _ _ ,__________._ _ _ _

__________

State ____ Zip Code ______
or Country

- Do Not Tear - Fold Here and Tape

III II I

DmDoma

No Postage
Necessary
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/L12

MARLBOROUGH, MA

01752

Do Not Tear - Fold Here and Tape

I
I

.=
(1)

!
I....J

I~
It::
I 0

1°
I~
1.£
I~

I:;
IW
!
I