Testing arrays with pseudo-variables

Component Testing for C

Another form of initialization consists of using one or more pseudo-variables, as the following example illustrates:

TEST 3

FAMILY nominal

ELEMENT

VAR x1, init = 0, ev = init

VAR x2, init = SIZE_IMAGE-1, ev = init

VAR y1, init = 0, ev = init

VAR y2, init = SIZE_IMAGE-1, ev = init

ARRAY image, init=(int)(100*(1+sin((float)(I1+I2)))), ev = init

ARRAY histo[0..200], init = 0, ev ==

ARRAY histo[201..SIZE_HISTO-1], init = 0, ev = 0

VAR status, init ==, ev = 0

#status = compute_histo(x1, y1, x2, y2, histo);

END ELEMENT

END TEST

I1 and I2 are two pseudo-variables which take as their value the current values of the array indices (for image, from 0 to 199 for I1 and I2). You can use these pseudo-variables like a standard variable in any C expression.

This lets you create more complicated test scripts in the case of very large arrays, where the use of enumerated expressions is limited.

For multidimensional arrays, you can combine these different types of initialization and test expressions, as the following example shows:

ARRAY image, init = {0 => I2, 1 => { 0 => 100, others => 0 },

& others => (I1 + I2) % 255 }