VAR, ARRAY and STR <initialization> Parameter

C Test Script Language


In this documentation, the Component Testing <initialization> parameters for C specify the initial value of the variable.


INIT = <exp>

INIT IN { <exp>, <exp>, ... }

INIT ( <variable> ) WITH { <exp>, <exp>, ... }

INIT FROM <exp> TO <exp> [STEP <exp> | NB_TIMES <nb> | NB_RANDOM <nb>[+ BOUNDS]]

INIT FROM <exp> TO <exp> [STEP <exp> | NB_VALUE <nb> | NB_RANDOM <nb>[+ BOUNDS]]




The <initialization> expressions are used to assign an initial value to a variable. The initial value is displayed in the Component Testing report for C.

The INIT value is calculated during the preprocessing phase, not dynamically during test execution.

Initializations can be expressed in the following ways:

The INIT IN and INIT (<variable>) WITH expressions cannot be used with for arrays that were initialized in extended mode or for structures.

The INIT FROM expression can only be used for numeric variables.

The STEP syntax cannot be used when the same variable is tested by another VAR, ARRAY or STR statement.

The NB_TIMES, NB_VALUE, and NB_RANDOM keywords require that the target platform supports floating point numbers.

INIT == allows the variable to be left un-initialized. You can thus control the values of variables that are dynamically created by the service under test. The initial value is displayed in the test report as a question mark (?).

An initialization expression can still be used (INIT == <expression>) to include of expected value expression when using the INIT pseudo-variable is used. See Expected_Value Expressions.


The initialization expressions <exp> can be among any of the following values:

  1. Numeric (integer or floating-point), character, or character string literal values. Strings can be delimited by single or double quotes

  2. Native constants, which can be numeric, characters, or character strings

  3. Variables belonging to the test program or the module to be tested

  4. C or Ada functions

  5. The keyword NIL to designate a null pointer

  6. Pseudo-variables I, I1, I2 ..., J, J1, J2 ..., where In is the current index of the nth dimension of the parameter and Jm the current number of the subtest generated by the test scenario's mth INIT IN, INIT FROM or LOOP; the I and I1 variables are therefore equivalent as are J and J1; the subtest numbers begin at 1 and are incremented by 1 at each iteration. These pseudo-variables I, In, J, and Jn must not be declared as typedefs or variables in the source code.

  7. A C expression with one or more of the above elements combined using any operators and casting, with all required levels of parentheses, the + operator being allowed to concatenate character string variables

  8. For arrays and structures, any of the above-mentioned expressions between braces ('{}') for C or brackets ('[]') for Ada, including when appropriate:

    1. For an array element, part of an array or a structure field, its index, interval or name followed by '=>' and by the value of the array element, common to all elements of the array portion or structure field

    2. For structures you can test some fields only, by using the following syntax:

      1. For C: {<value>,,<value>}

      2. For all languages: [<fieldname> => <value>, <fieldname> => <value>]

  9. The keyword others (written in lower case) followed by '=>' and the default value of any array elements or structure fields not yet mentioned

  10. For INIT IN and INIT WITH only, a list of values delimited by braces ('{}') for C composed of any of the previously defined expressions.

Additional Rules

Any integers contained in an expression must be written either in accordance with native lexical rules, or under the form:

Note   Because of the way hexadecimal values are handled, the value range should not exceed half of the maximum range when the initialization is expressed in hexadecimal.

The number of values inside an INIT IN parameter is limited to 100 elements in a single VAR statement.

The number of INIT IN parameters per TEST LOOP block is limited to 7.

The number of INIT IN parameters per TEST block is limited to 8.

In Component Testing for C, if variables are used in the expression, then the test evaluates the the INIT value with variable values from after the execution.

All Euclidian divisions performed by the Test Script Compiler round to the inferior integer. Therefore, writing -a/b returns a different result than -(a/b), as in the following examples:



VAR x, INIT = pi/4-1, ...

VAR y[4], INIT IN { 0, 1, 2, 3 }, ...

VAR y[5], INIT(y[4]) WITH { 10, 11, 12, 13 }, ...

VAR z.field, INIT FROM 0 TO 100 NB_RANDOM 3, ...

VAR p->value, INIT ==, ...

ARRAY y[0..100], INIT = sin(I), ...

ARRAY y, INIT = {50=>10,others=>0}, ...

STR z, INIT = {0, "", NIL}, ...

STR *p, INIT = {value=>4.9, valid=>1}, ...


In the following example, the C test Script Compiler generates code that tests x against a then b after the execution of the code under test:

VAR y,        init in (1,2),            ev = init
VAR a,        init(y) with ( 10, 20 ),  ev = 50
VAR b,        init(y) with ( 30, 40 ),  ev = 70
VAR x,        init(y) with (a, b),      ev = init
#a := 50;
#b := 70;


Related Topics

<expression> parameterExpected_Value Expressions<variable> parameter (C)VAR, ARRAY and STR