Description
At the beginning of your lab period, the lab instructor will give an example of calculating function, branch,
and condition coverage for a program. Recall that branch coverage is also called path coverage.
Answer each of the questions below. For questions 3 and 4, you will need to download the auxiliary file
L6Q4.txt; look for the link on the moodle pages for the course. For all questions involving the use of
LINUX/UNIX commands, place the command or pipeline you used along with the resulting output (i.e.
copy-and-paste from your terminal window) in a file called lab6.txt. However, do not include extraneous or superfluous commands or output; only include content relevant and essential to the specified task.
Then, with a text editor, add to lab6.txt your solutions to question 1(c) and all the parts of questions
2 and 4. Also with the text editor add text and identifying information to clearly distinguish which commands/output/code/description correspond to each task/question. This lab is out of a total of 14 marks,
with each question or question major part (1a, 1b, 3, etc.) being worth one mark except for 4(b) and 4(c),
which are each worth 2 marks. Marks may be docked for extraneous, irrelevant, or superfluous content or
for not following directions. Your submission is due at 11:55 p.m. on Thursday, October 20. Note that this
lab specification is 3 pages in length.
This lab can be completed on either tuxworld or one of the Macintosh computers in S311 and S315.
However, clearly identify in your lab submission which (type of) machine you used. The answers to some of
the lab questions vary slightly depending on the (type of) machine used, and therefore the marker needs to
know which it was.
1. (a) Create a procedural C++ program called datatype_sizes.cc that will output the sizes in bytes
of the datatypes long int, float, double and float *. Make your program simple and and
its logic straightforward. For example, the program logic can consist solely of four printf()
statements. Upload your program to moodle as part of your Lab 06 submission when complete.
With the exception of warnings about parameters argc and argv of function main() not being
used, your program must compile without errors or warnings when the -Wall and -Wextra options
are given to the C++ compiler.
Hint: use the C/C++ function sizeof(). The return type of sizeof() is size_t. To have
printf(3) output a value of type size_t, use the “%zu” or “%zd” format specification. The
program example from class on October 13 would be a very good starting point for your program.
Note: overly and unnecessarily complex programs may be docked marks.
(b) Compile your C++ program two ways: first with the option “-m32” and then with the option
“-m64”. These options specify 32- and 64-bit architecture, respectively. Name the resultant
executables 32bit_sizes and 64bit_sizes, respectively. Show a log of your compilations in
lab6.txt. Have the C++ compiler perform extra checks for potentially problematic situations
by using the -Wall and -Wextra options. You can also use -Wpedantic if you wish.
1
(c) Run each of your executables in part (b) (include a log of this in lab6.txt), and then write a
sentence or two comparing their outputs; that is, comparing the size of datatypes used by the
C/C++ compiler when compiling for the two different architectures (32-bit or 64-bit).
(d) Use a UNIX command to determine the file types of 32bit_sizes and 64bit_sizes; i.e. use a
UNIX command to determine the kind of information within the two files.
(e) Use uname(1) with an appropriate option to produce, on the standard output, the “machine
hardware name” of the machine on which you completed the previous parts to this question.
Include in your lab submission a log of the command you used, and its output.
2. (a) For the following pair of data definitions, state which one conforms to the programming guidelines
given in lecture. Place your answer (either “i” or “ii”) in lab6.txt.
i.
// A food is one of the four food groups.
typedef enum { MEAT, DAIRY, VEGETABLE, GRAIN } food_group;
// … …(food_group food, …) {
// …
// switch (food) {
// case MEAT: …; break;
// case DAIRY: …; break;
// …
// default: assert(false); break;
// }
ii.
// A food is one of the four food groups.
typedef enum { MEAT, DAIRY, VEGETABLE, GRAIN } food_group;
(b) For the following pair of module templates, state which one conforms to the programming guidelines given in lecture. Place your answer (either “i” or “ii”) in lab6.txt.
i. // For a given person, determine which food group they eat the least.
food_group least_eaten(person *p) {
return MEAT;
}
ii. // least_eaten : For a given person pointed to by p,
// determine which food group they eat the least.
// In:
// p != NULL
// Out:
// a value from enumerated type food_group
food_group least_eaten(person *p) {
return MEAT;
}
3. Use the pr command to produce, on the standard output, a listing of L6Q4.txt that has the lines
numbered consecutively.
Hint: you will need two options to pr to get the proper output, one of which is -t. You will need to
determine the other option.
4. Consider the fragment of code found in L6Q4.txt. Suppose that this code was tested using the test
cases (i.e. values of myint) 0, 5, -8, and -100. Answer the following questions in lab6.txt.
2
(a) Evaluate the function coverage of these test cases. In other words, what proportion of the functions
are actually called in at least one of the test cases? Only include in your analysis add1(),
substract1(), double_it(), and halve_it()—do not consider scanf() and main().
(b) Evaluate the path coverage of these test cases. In other words, to what extent has every possible
path (of execution) through the program been executed? Expression your answer as a proportion.
(c) Evaluate the condition coverage of these test cases. In other words, what proportion of the
conditions in if statements evaluate to true in at least one of the test cases and to false in at
least one of the test cases? That is, to be counted as being “covered”, a condition must evaluate
to both true and to false as part of the testing.
(d) Given the results of parts (a), (b), and (c), comment on the adequacy of the four test cases (0, 5,
-8, and -100) in terms of providing a thorough test of L6Q4.txt.