Programming PLCs using Structured Text

Nieke Roos
Department of Computing Science
University of Nijmegen
Toernooiveld 1, NL-6525 ED, Nijmegen, The Netherlands
nieker@sci.kun.nl
Keywords: Programmable Logic Controller, PLC, Structured Text, ST,
Pascal
Abstract
An ever increasing part of the industrial and safety-critical
processes is presently controlled by Programmable Logic
Controllers (PLCs). As its name already suggests, a PLC can be
and has to be programmed to perform its task. Since its birth, a
wide range of different PLCs has been put on the market that
makes use of an even wider range of programming languages.
Recently, the IEC 1131-3 standard has been published in an effort
to reduce chaos to order in the universe of PLC programming
languages.
This paper takes a closer look at one of the programming
languages defined by the IEC 1131-3 standard, namely the
Structured Text programming language. A comparison is made
between Structured Text and the language upon which it was
based, the Pascal programming language. The primary differences
that result from this comparison are then discussed and some
conclusions are drawn about the IEC 1131-3 standard.
1 The Programmable Logic Controller
1.1 Introduction
As the Industrial Revolution stormed the civilized world late last century man started to create
machines that could alleviate the hard labour some existing tasks brought with them. This very first
large-scale reflection of mans desire to be able to automate was the incentive to an evolutionary
developement of industrially employed equipment. The first machines ran on steam, later on gasoline
was used as primary source of energy and at present the majority of the devices in the industry is
powered by electricity. At the same time the dimensions of these devices decreased from steam
engines the size of an average suburban dwelling to electrical appliances not bigger than a lunchbox
containing even smaller circuits and wiring.
The invention of the computer and its massive application following the Digital Revolution taking
place in the 1960s, gave birth to a new phenomenon: the art of programming. This revolutionary
ability to control the behaviour of an apparatus by precisely conveying the actions it has to perform
exerted enormous influences not only on the computer science domain but on other domains as well.
An important and interesting example of the influence programmability has had on other domains is
the Programmable Logic Controller (PLC), resulting from the synthesis of the electrical world and the
universe of computer science in an effort to combine this programmability with electrical circuitry.
1.2 The architecture
The primary purpose of a PLC is to control a particular process or collection of processes in industrial
or other safety-critical applications. This system of processes has certain properties, which are to be
controlled by the PLC. These properties are communicated to the PLC through digitized analogue
signals directly or indirectly emanating from the system. In response to these inputs the PLC produces
electrical control signals that are fed back to the system. The main task of a PLC, therefore, is to
convert the inputs received to particular related outputs.
1.2.1 The hardware
In essence, the PLC is a miniaturized version of a general-purpose computer. Like its big brother, its
structure can be divided into a hardware-part and a software-part. As depicted in Figure 1, the PLCs
hardware is composed of:
an input interface,
an output interface,
a program memory,
a data memory, and
a processor.
PLC
Program
memory
Processor
Output
interface
Data
memory
Input
interface
Instruction
s

m
i
i

1
n
o
o

Figure 1: PLC hardware architecture

The PLCs task commences as the input signals from the system of processes that has to be
controlled arrive at the input interface. These inputs are then stored in the PLCs data memory. Next,
the PLCs processor starts fetching and executing instructions from the PLCs software contained in
the PLCs program memory. As the processor is doing this, the data memory serves as storage of
intermediate and/or final results produced by the execution of a particular program instruction. After
the processor has executed all the instructions in the program memory, the output interface is
responsible for communicating the final result or results back to the system of processes that is under
the PLCs control. Finally, the PLCs task starts all over again as new input signals arrive at the input
interface. Hence, the PLC can be regarded as executing its task indefinitely.
Note that the before-mentioned task-loop usually is said to be executed not by the hardware but by
something that is a combination of hardware and software, called firmware. Firmware is software that
has been written onto read-only memory.
1.2.2 The software
When one takes a closer look at this task-loop, one can observe that it basically is the responsibility of
the PLCs software contained in the program memory to convert the inputs received at the input
interface to related outputs emanating from the output interface. Like every software program, the
program that resides in the PLCs program memory and that performs this conversion has to be written
in a particular programming language. However, which programming language should a PLC
manufacturer choose to program his/her products?
Until recently, this posed a huge problem. Because there was no such thing as a uniform language
for programming PLCs, every manufacturer developed and applied a programming language of
his/her own. This caused a virtually endless series of different programming dialects to be created. For
a company new to the field to pick one to use in its designs was the same as looking for a needle in a
haystack. In 1993, however, the International Electrotechnical Commission (IEC) decided to facilitate
the needle finding by publishing the IEC 1131 International Standard for Programmable Controllers.
The third part of this standard, entitled 1993 Programmable languages but commonly called IEC
1131-3, contains the definition of five languages for programming PLCs ([IEC]).
2 IEC 1131-3
2.1 The software model
Part three of the IEC 1131 standard defines an architecture for the PLC software called the software
model (see Figure 2). At the highest level, this software model describes the notion of a configuration.
In fact, this notion of configuration is just a fancy way to refer to the entire PLC software. On a lower
level, a configuration can contain one or more resources. A resource can be seen as providing a
processing environment for executing the program contained within the resource. A program is
composed of a collection of definitions of software elements called functions and function blocks and
their application within a sequence of statements. Functions and function blocks are the basic building
blocks of a program, able to exchange data. A function is a particular algorithm, i.e. sequence of
statements, with zero or more input values and one output value. A function always produces the same
output value for a given set of input values, as opposed to a function block that can produce differing
output values each time it is executed with the same set of input values. This is because, in addition to
the algorithm a function consists of, a function block also contains a data structure, in which values
can be stored, even after the function block has finished executing. These stored values can influence
the function blocks result, which accounts for the possibility of differing outputs when executed with
the same set of inputs.
2.2 The languages
The third part of the IEC 1131 standard also contains a definition of five different programming
languages, each of which can be used to create the program statements, both the statements that
constitute the algorithm of a function or function block and the sequence of statements that appear
stand-alone in the body of the program. These five languages are:
Instruction List, a textual programming language resembling assembler.
Structured Text, another textual programming language resembling Pascal.
Ladder Diagram, a graphical programming language based on the graphical presentation of Relay
Ladder Logic used by electrotechnical engineers.
Function Block Diagram, another graphical programming language resembling electronic circuit
diagrams.
Sequential Function Chart, a third graphical programming language derived from Petri Nets.
The next section will take a closer look at one of these programming languages, the Structured Text
programming language.
3 Structured Text
The Structured Text programming language is a high level programming language. Because the
Structured Text language is based upon another high level programming language, namely Pascal,
both languages bear strong resemblance syntactically. This resemblance is used to clarify the
Structured Text syntax by comparing Pascal code to equivalent program pieces written in Structured
Text. To be able to write pieces of Structured Text, the language constructs and the data types that
these pieces are composed of have to be introduced first. This is done in the next two paragraphs.
Configuration
Resource
Program
Resource
Program
Resource
Program
Figure 2: PLC software model
Function Block

Function Block
Function

Function
Statements
Function Block

Function Block
Function

Function
Statements Statements
Function

Function
Function Block

Function Block
3.1 The language constructs
One of the problems concerning Structured Text within the framework of the IEC 1131-3 standard is
determining exactly what is part of the language and what is not. The standard defines Structured Text
to be just the statements that can be used to construct function and function block algorithms and
program bodies. However, within the standard, elements common to all five programming languages
defined in the standard, such as elements that identify, for instance, a configuration, a resource or a
program and the declaration of variables, are written using a syntax that also resembles the Pascal
syntax in such a way that many of the people that have applied the standard have unjustly regarded
these common elements as being part of the Structured Text programming language as well, in spite of
the standard. Clearly, the standard isnt as clear on this as one would like a standard to be.
A comprehensive context-free description of the syntax of the language constructs that, according
to the IEC 1131-3 standard, do belong to the Structured Text programming language is included in
appendix A.
3.2 Data types
Like almost every other programming language, the IEC 1131-3 standard also allows the use of a
variety of different types of data in Structured Text programs, both elementary and derived.
3.2.1 Elementary data types
The elementary data types provided by the standard are the following:
Integers
Floating point numbers
Data types that represent time
Strings
Bit strings
3.2.1.1 Integers
IEC Data type Description Range
SINT Short integer -128 +127
INT Integer -32768 +32767
DINT Double integer -2
31
+2
31
-1
LINT Long integer -2
63
+2
63
-1
USINT Unsigned short integer 0 +255
UINT Unsigned integer 0 +2
16
-1
UDINT Unsigned double integer 0 +2
32
-1
ULINT Unsigned long integer 0 +2
64
-1
Table 1: Integer data types
Examples: -18, 0, 1977 (decimal), 2#00010010 (binary), 8#22 (octal), 16#12 (hexadecimal)
3.2.1.2 Floating point numbers
IEC Data type Description Range
REAL Real numbers 10
38
LREAL Long real numbers 10
308
Table 2: Floating point data types
Examples: 18.7, 18.7E-77, 0,19e+77
3.2.1.3 Data types that represent time
IEC Data type Description Examples
TIME The duration of time
after an event
T#18d7h19m7s7ms
TIME#18h7s
DATE Calendar date D#1977-07-18
DATE#1977-07-18
TIME_OF_DAY Time of day TOD#18:07:19
TIME_OF_DAY#23:59:59.99
DATE_AND_TIME Date and time of day DT#1977-07-18-18:07:19.77
DATE_AND_TIME#1977-07-18-12:00:00
Table 3: Time data types
3.2.1.4 Strings
IEC Data type Description Examples
STRING Character strings Hello world

Table 4: String data type

3.2.1.5 Bit strings
IEC Data type Description
BOOL Bit string of 1 bit
BYTE Bit string of 8 bits
WORD Bit string of 16 bits
DWORD Bit string of 32 bits
LWORD Bit string of 64 bits
Table 5: Bit string data types
Examples: 00010010, FALSE (equivalent to 0), TRUE (equivalent to 1)
3.2.2 Derived data types
As well as providing the elementary data types mentioned above the IEC 1131-3 standard also
presents the Structured Text programmer with the ability to build totally new data types. A new data
type is constructed in Structured Text by enclosing its definition in the keywords TYPE and
END_TYPE. The definition consists of the name of the data type under construction, followed by a
colon, followed by the kind of derived data type to be constructed. There are four kinds of derived data
types:
Structured data types
Enumerated data types
Sub-ranges data types
Array data types
3.2.2.1 Structured data types
A structured data type is a composite data type, constructed by enclosing the elements of the
composition in the keywords STRUCT and END_STRUCT. Each element consists of its name followed
by a colon followed by its data type. A structured data type itself can contain one or more structured
data types. For example:
TYPE Rectangle :
STRUCT
TopLeft : Point;
Height : INT;
Width : INT;
END_STRUCT;
END_TYPE
In the example above, the Point data type is a structured data type itself.
3.2.2.2 Enumerated data types
An enumerated data type is constructed by enclosing the elements of the enumeration in parentheses.
Each element consists of a different name. For example:
TYPE Color :
(Red, White, Blue);
END_TYPE
3.2.2.3 Sub-ranges data types
A sub-ranges data type is constructed by limiting the range of another data type, usually integers.
Limited versions of a data type consist of the name of the data type to be restricted, followed by a
lower and an upper bound separated by two dots and enclosed in parentheses. For example:
TYPE Angle :
INT(-180..+180);
END_TYPE
3.2.2.4 Array data types
An array data type is constructed by the keyword ARRAY, followed by a multi-dimensional vector of
ranges of indices, followed by the keyword OF, followed by the (elementary or derived) data type that
is going to be contained in the array. A multi-dimensional vector of ranges of inidices consists of a
comma-delimited list of ranges of indices, enclosed in brackets. A range of indices consists of a lower
and upper bound of the range, both integers, separated by two dots. For example:
TYPE Display :
ARRAY[1..768, 1..1024] OF Color;
END_TYPE
3.3 Usage
In order to clarify the Structured Text syntax, the next few paragraphs contain a comparison between
the following constructs that can be found in both the Structured Text programming language and
Pascal: assignments, function calls, conditional statement and iteration statements. Each time the
notable differences will be put forward.
3.3.1 Assignments
Structured Text Pascal
VarA := (VarB * 24 MOD 2 = 1) XOR
(VarC <= VarD + 34);
VarA := (VarB * 24 MOD 2 = 1) XOR
(VarC <= VarD + 34);
(* AnArray is an array containing
twenty integers *)
AnArray := 10(1), 5(2), 5(3);
(* AnArray is an array containing
twenty integers *)
FOR ElementNr := 1 TO 10 DO
BEGIN
AnArray[ElementNr] := 1;
END;
FOR ElementNr := 11 TO 15 DO
BEGIN
AnArray[ElementNr] := 2;
END;
FOR ElementNr := 16 TO 20 DO
BEGIN
AnArray[ElementNr] := 3;
END;
VarA := AND(Var1, Var2, ..., VarN); VarA := Var1 AND Var2 AND ... AND
VarN;
Table 6: Assignments in Structured Text versus assignments in Pascal
Assignment-wise one major dissimilarity can be observed. Notice the ease with which an array can be
initialized in the Structured Text language. Also notice the effort it requires to do the same in Pascal!
Operator-wise there is another difference between the two languages. In the Structured Text
language, some of the operators are extensible, i.e. some operators can have any number of arguments
(of the same type).
3.3.2 Function calls
Structured Text Pascal
FUNCTION Ave_REAL : REAL
VAR_INPUT
Input1, Input2 : REAL;
END_VAR
Ave_REAL := (Input1 + Input2) / 2;
END_FUNCTION
Average1 := Ave_REAL(5.0, 4.0);
Average2 := Ave_REAL(Input2 := 6.0);
(* Value 3.0 assigned to Average2 *)
FUNCTION Ave_REAL
(Input1, Input2 : Real): Real;
BEGIN
Ave_REAL := (Input1 + Input2) / 2;
END;
Average1 := Ave_REAL(5.0, 4.0);
Average2 := Ave_REAL(0.0, 6.0);
Table 7: Function calls in Structured Text versus function calls in Pascal
Note that, strictly speaking, the FUNCTION END_FUNCTION and the VAR END_VAR
constructs mentioned in this paragraph do not belong to the Structured text programming language, as
discussed earlier.
3.3.3 Conditional statements
Structured Text Pascal
IF VarB > 0 THEN
IF VarC = 3 THEN
VarA := TRUE;
ELSE
VarC := 3;
VarA := FALSE;
END_IF;
ELSIF VarB < 0 THEN
VarA := FALSE;
ELSE
VarC := 3;
VarA := TRUE;
END_IF;
IF VarB > 0
THEN
BEGIN
IF VarC = 3
THEN VarA := TRUE
ELSE
BEGIN
VarC := 3;
VarA := FALSE;
END
END
ELSE IF VarB < 0
THEN VarA := FALSE
ELSE
BEGIN
VarC := 3;
VarA := TRUE;
END;
CASE VarC OF
1, 2 : VarB := 1;
3 : VarB := 2; VarA := TRUE;
4..10 : VarB := 3;
ELSE
VarB := 0; VarA := FALSE;
END_CASE;
CASE VarC OF
0, 1 : VarB := 1;
3 : BEGIN
VarB := 2;
VarA := TRUE;
END;
4..10 : VarB := 3;
ELSE
VarB := 0; VarA := FALSE;
END;
Table 8: Conditional statements in Structured Text versus conditional statements in Pascal
The main difference between the Structured Text language and the Pascal language, when one looks at
the conditional statements in both languages, is the explicit use of the BEGIN and END keywords in
the Pascal code in cases that construct bodies contain two or more statements. Note the inconsistent
omission of these keywords in the ELSE part of the Pascal CASE construct.
Another, minor, difference is the necessary omission of the semicolon before the ELSE IF or
ELSE part of the IF construct in the Pascal programming language.
3.3.4 Iteration statements
Structured Text Pascal
FOR VarB := 1 TO VarC DO
VarD := VarD + 1;
VarE := 2 * VarE;
END_FOR;
FOR VarB := 1 TO VarC DO
BEGIN
VarD := VarD + 1;
VarE := 2 * VarE;
END;
FOR VarB := 100 TO 0 BY 2 DO
VarD := VarD + VarB;
END_FOR;
FOR VarB := 50 DOWNTO 0 DO
VarD := VarD + 2 * VarB;
END_FOR;
WHILE VarA AND (VarD <= VarC) DO
VarD := VarD + 1;
VarE := 2 * VarE;
VarA := VarE < 100;
END_WHILE;
WHILE VarA AND (VarD <= VarC) DO
BEGIN
VarD := VarD + 1;
VarE := 2 * VarE;
VarA := VarE < 100;
END;
REPEAT
VarD := VarD + 1;
VarE := 2 * VarE;
VarA := VarE < 100;
UNTIL NOT(VarA AND (VarD <= VarC));
REPEAT
VarD := VarD + 1;
VarE := 2 * VarE;
VarA := VarE < 100;
UNTIL NOT(VarA AND (VarD <= VarC));
Table 9: Iteration statements in Structured Text versus iteration statements in Pascal
As was the case with conditional statements, the primary difference between the Structured Text
language and the Pascal language, when one looks at the iteration statements in both languages, again,
is the explicit use of the BEGIN and END keywords in the Pascal code in cases that construct bodies
contain two or more statements. Note another inconsistent omission of these keywords in the Pascal
REPEAT construct.
Another difference is the possibility of including a BY part in the FOR construct of the Structured
Text language, indicating the magnitude of the step taken every time a FOR loop is executed. The need
of a similar construct in the Pascal language can, awkwardly, be circumvented by slightly adjusting
the statements within the FOR construct, possibly combining this with the use of the DOWNTO
construct available in the Pascal language, as illustrated by the example above. This DOWNTO
construct in Pascal is equivalent to using the TO construct in combination with a BY 1 in the
Structured Text programming language.
3.4 Fundamental differences between Structured Text and Pascal
In the last section a couple of minor, mostly syntactical, dissimilarities between the Structured Text
programming language and Pascal have been mentioned. There are, however, more fundamental
differences between the two. These are discussed in the following paragraphs.
3.4.1 Function blocks
The first fundamental difference is the fact that the concept of function block, as discussed earlier, has
no precise Pascal equivalent. This discrepancy can be accounted for, however, as everything that can
be done by function blocks in Structured Text can be achieved by the appropriate use of functions or
procedures and (global) variables in Pascal.
3.4.2 Directly represented variables
Another concept occurring in the Structured Text language but not found in Pascal, is the notion of
directly represented variables. Directly represented variables are variables that are not identified by a
name, but that are addressed by their position in the PLCs data memory. This concept has been
adopted from programming languages, such as the Ladder Diagram programming language defined by
the IEC 1131-3 standard, that make use of so-called ladder diagrams. Another reason why it has been
included in Structured Text is the fact that it is commonly used in many PLCs.
However, the kind of identification that directly represented variables advocate is extremely
implementation-dependent and therefore should have no place in a standard such as IEC 1131-3 that
addresses implementation-independent issues.
3.4.3 Default initial values
Another difference between the Structured Text language and the Pascal language is the fact that
Structured Text variables do not have to be initialized explicitly. When an explicit initialization of a
variable is omitted that variable will default to the value all variables of its type default to. Integers,
floating point numbers, times and bit strings default to zero, dates default to 0001-01-01 and strings
default to the empty string. Derived data types default to the default values of the elementary types
they are derived from, unless a default value is contained within the type definition, in which case the
default values of the elementary data types are overridden.
This concept of default initial values allows parameters to be omitted from function calls, in which
case the unused input parameter will take the default value as given in the function type definition. An
example of this can be seen in paragraph 3.3.2, where the value 3.0 is assigned to Average2.
3.4.4 Several output values
A fourth fundamental difference between the two languages is the fact that a function in the Structured
Text language can have more than one output value, without them having to be packed together in
some kind of structure. If a function has more than one output value, these extra output values are in
the output variable section (labelled VAR_OUTPUT) or in the section intended for variables that serve
as input as well as output (labelled VAR_IN_OUT).
In the IEC 1131-3 standard, one particular example of the construction mentioned above can be
found to play a part in another phenomenon common to the Structured Text language, but unknown to
the Pascal programming language: bonus parameters for free. With every function, two bonus boolean
parameters are supplied gratuitously: an EN input parameter indicating whether the function body is
executed or not and an ENO output parameter indicating whether the execution of the function has
resulted in an error of some kind. In case the latter parameter indicates the occurrence of an error, the
functions other result(s) is (are) undetermined.
3.4.5 Overloading
A fifth difference is the fact that the Pascal language does not support overloading, contrary to the
Structured Text programming language. Pascal does, however, support the concept of coercion. It is
because of this, and not because of the support of overloading, that it is possible in Pascal that the
square function, which, according to its definition, requires a real argument, can be supplied with an
integer argument.
Note that the Structured Text programming language does not allow one to create an overloaded
function oneself. The only overloaded functions existing within the Structured Text language are
functions offered by Structured Text itself.
3.4.6 Recursion
The sixth and final fundamental difference, and maybe the most important one, is the total absence of
the notion of recursion within the Structured Text programming language. The IEC 1131-3 standard
specifically prohibits functions, function blocks and programs to be recursive, that is, their invocation
should not cause the invocation of another function, function block or program of the same type. One
can only guess at the why and wherefore of this decision, because when it comes to explaining the
reasons of banning recursion from within the confines of the Structured Text language, the standard
remains silent. One possible reason could be that a program written in the language as defined by the
IEC 1131-3 standard can simply be viewed as one big collection of macro-expansions, thus rendering
the processing of a Structured Text program less complex. Introducing recursion into the game would
totally annul this view. The question is, however, whether the reduction in complexity is as significant
as the advocates of the omission of recursion claim it to be, compared to the increase entailed by the
introduction of other language elements such as the concept of overloading mentioned above. With all
the concepts that have been included in it, the Structured Text language has long outgrown the
characterising traits of a basic programming language and one could wonder why stop here and not
include recursion? Other possible reasons against allowing recursive Structured Text are the difficulty
of testing and the unpredictability of its real-time performance. These are both perfectly valid reasons,
but, strangely enough, none of them are used to give an explanation for the absence of recursion in
IEC 1131-3. In conclusion, one can say that the unaccounted omission of recursion definitely
constitutes one of the standards major flaws.
4 Conclusions
After a brief introduction to the universe of the Programmable Logic Controller, this paper discussed
one of five ways defined by the IEC 1131-3 standard to program this device: the Structured Text
programming language. A comparison has been made between Structured Text and the language upon
which it was based, the Pascal programming language. This comparison revealed a great similarity
existing between the two. Of the few distinctions that have been uncovered, the absence of the ability
to write recursive programs in the Structured Text language is the most remarkable one, especially
because of the undocumented nature of this omission, as a result of which the choice of what
programming constructs to include in the language and what not gives the impression of being a fairly
arbitrary one.
References
[HAB] Wolfgang A. Habing and Krzystof M. Sacha. Programmable Logic Controllers. Real
Time Systems: Implementation of Industrial Computerized Process Automation.
[IEC] International Electrotechnical Commission, Technical Committee No. 65: Industrial-
process Measurement and Control, Sub-committee 65B: Devices, Working Group 7:
Programmable Controllers. Working Draft = IEC 61131-3, 2
nd
Ed. Programmable
Controllers Programming Languages.
[LEW] R.W. Lewis. Programming Industrial Control Systems using IEC 1131-3.
[TOU] Konstantinos Tourlas. An assessment of the IEC 1131-3 Standard on Languages for
Programmable Controllers.
Appendix A:
The syntax of the Structured Text programming language
A.1 Expressions
Production rules:
Expression ::= XOR_Expression {'OR' XOR_Expression}
XOR_Expression ::= AND_Expression {'XOR' AND_Expression}
AND_Expression ::= Comparison {('&' | 'AND') Comparison}
Comparison ::= EquExpression { ('=' | '<>') EquExpression}
EquExpression ::= AddExpression
{ComparisonOperator AddExpression}
ComparisonOperator ::= '<' | '>' | '<=' | '>='
AddExpression ::= Term {AddOperator Term}
AddOperator ::= '+' | '-'
Term ::= PowerExpression
{MultiplyOperator PowerExpression}
MultiplyOperator ::= '*' | '/' | 'MOD'
PowerExpression ::= UnaryExpression {'**' UnaryExpression}
UnaryExpression ::= [UnaryOperator] PrimaryExpression
UnaryOperator ::= '-' | 'NOT'
PrimaryExpression ::= Constant | Variable | '(' Expression ')'
| FunctionName '(' [ST_FunctionInputs] ')'
ST_FunctionInputs ::= ST_FunctionInput { ',' ST_FunctionInput}
ST_FunctionInput ::= [VariableName ':='] Expression
Table 10: Production rules of expressions in Structured Text
A.2 Statements
Production rules:
StatementList ::= Statement ';' {Statement ';'}
Statement ::= NIL
| AssignmentStatement
| SubprogramControlStatement
| SelectionStatement
| IterationStatement
Table 11: Production rules of statements in Structured Text
A.3 Assignment statements
Production rule:
AssignmentStatement ::= Variable ':=' Expression
Table 12: Production rule of assignment statements in Structured Text
A.4 Subprogram control statements
Production rules:
SubprogramControlStatement ::= FB_Invocation | 'RETURN'
FB_Invocation ::= FBName
'(' [FB_InputAssignment
{',' FB_InputAssignment}] ')'
FB_InputAssignment ::= VariableName ':=' Expression
Table 13: Production rules of subprogram control statements in Structured Text
A.5 Selection statements
Production rules:
SelectionStatement ::= IfStatement | CaseStatement
IfStatement ::= 'IF' Expression 'THEN' StatementList
{'ELSIF' Expression 'THEN' StatementList}
['ELSE' StatementList]
'END_IF'
CaseStatement ::= 'CASE' Expression 'OF' CaseElement
{CaseElement}
['ELSE' StatementList]
'END_CASE'
CaseElement ::= CaseList ':' StatementList
CaseList ::= CaseListElement {',' CaseListElement}
CaseListElement ::= Subrange | SignedInteger
Table 14: Production rules of selection statements in Structured Text
A.6 Iteration statements
Production rules:
IterationStatement ::= ForStatement
| WhileStatement
| RepeatStatement
| ExitStatement
ForStatement ::= 'FOR' ControlVariable ':=' ForList 'DO'
StatementList
'END_FOR'
ControlVariable ::= Identifier
ForList ::= Expression 'TO' Expression
['BY' Expression]
WhileStatement ::= 'WHILE' Expression 'DO'
StatementList
'END_WHILE'
RepeatStatement ::= 'REPEAT'
StatementList
'UNTIL' Expression
'END_REPEAT'
ExitStatement ::= 'EXIT'
Table 15: Production rules of iteration statements in Structured Text