Software Testing · Quiz 1 · Jan 2025

**1. Identify the concept:** Every interior vertex of a chain has in-degree and out-degree one. A maximal chain cannot be extended into a larger chain. Thus i is false and ii is true. **Answer: B — Statement (i) is false but statement (ii) is true.** A: **Incorrect:** This reverses the two truth values. B: **Correct:** Interior degrees are one and maximality prevents extension. C: **Incorrect:** The first statement is false. D: **Incorrect:** The second statement is true.

**1. Identify the concept:** A fault is a static defect; an error is the resulting incorrect internal state; a failure is externally visible incorrect behaviour. The mapping is fault→2, error→3, failure→1. **Answer: B — Fault→2, Error→3, Failure→1.** A: **Incorrect:** This swaps fault and failure meanings. B: **Correct:** All three definitions are correctly matched. C: **Incorrect:** A fault is a static defect, not an external failure. D: **Incorrect:** This swaps the static defect and internal state.

**1. Identify the concept:** An FSM transition is an edge between states. Covering every graph edge therefore means covering every transition. **Answer: B — Transition coverage.** A: **Incorrect:** State coverage corresponds to node coverage. B: **Correct:** Transitions are graph edges. C: **Incorrect:** Transition-pair coverage corresponds to consecutive pairs of edges. D: **Incorrect:** Transition coverage is the direct equivalent.

**1. Count decisions:** The matrix-dimension if and the three nested for-loop conditions give four decisions. **2. Add the base path:** \(V=4+1=5\). **Answer: C — 5.** A: **Incorrect:** The derived count is 5, not 3. B: **Incorrect:** The derived count is 5, not 4. C: **Correct:** The derived count is 5, which matches this option. D: **Incorrect:** The derived count is 5, not 6.

**1. Identify the concept:** Detecting failures alone corresponds to Level 2, but cooperative reduction of software-usage risk raises the minimum to Level 3 in the testing-maturity model. **Answer: C — Level 3.** A: **Incorrect:** Demonstrating correct operation alone is a lower goal. B: **Incorrect:** The cooperative risk-reduction objective exceeds Level 2. C: **Correct:** Risk reduction with tester–developer cooperation defines Level 3. D: **Incorrect:** The scenario does not require Level 4’s broader quality discipline.

**1. Identify the concept:** Bottom-up integration tests lower-level clusters before their real callers exist, requiring drivers. Sandwich integration includes a bottom-up portion, so it also uses drivers. **Answer: C, D — Sandwich; Bottom-up.** A: **Incorrect:** The actual components are combined together rather than replaced by missing callers. B: **Incorrect:** Top-down integration generally uses stubs for missing callees. C: **Correct:** Its bottom-up portion needs caller replacements. D: **Correct:** Drivers invoke the lower-level components under test.

**1. Check test-path endpoints:** A complete test path must start at initial node 1, end at final node 6, and follow every listed edge. **2. Check each candidate:** [1,2,5,6] and [1,4,5,3,6] follow valid edges. The first candidate stops prematurely at 3, and the last requires nonexistent edge 2→4. **Answer: B, C — [1,2,5,6] and [1,4,5,3,6].** A: **Incorrect:** It ends at nonfinal node 3. B: **Correct:** It follows 1→2→5→6 and reaches the final node. C: **Correct:** Every edge exists and the path reaches node 6. D: **Incorrect:** There is no edge 2→4.

**1. Match the testing basis:** Unit tests and integration/system tests can derive cases from internal code, branches and interactions. The course classifies these as testing scopes that may use structural knowledge. **Answer: A, D — Unit testing; integration and system testing.** **Convention note:** Testing scope and white-box technique are independent. Profiling-based performance tests can also inspect internals in practice; A and D are the course’s intended classification. A: **Correct:** Unit tests can inspect internal logic. B: **Incorrect:** Performance tests primarily measure timing, throughput and resource use. C: **Incorrect:** Acceptance tests primarily check external user requirements. D: **Correct:** These scopes can use structural and source-code knowledge.

**1. Identify the faults:** allPrimes checks indices length−1 down to 1, skipping index 0. It also accepts numbers less than 2 because their divisor loop is empty. **2. Trace the tests:** Test 3 skips the initial composite 6 and accepts the checked −4, returning true. Test 4 skips its only composite 6 and returns true. Both assertFalse calls fail; tests 1 and 2 return the expected results. **Answer: C, D — testCase3() and testCase4().** A: **Incorrect:** All checked entries are prime, so assertTrue passes. B: **Incorrect:** The last element 12 is composite and returns false as expected. C: **Correct:** The skipped 6 and unchecked negative-number case cause an incorrect true return. D: **Correct:** Its only composite is skipped at index 0.

**1. Identify the concept:** The queue state is its occupancy. A capacity-three FIFO queue has occupancies 0,1,2,3, so there are four abstract states. **Answer: A — 4.** A: **Correct:** The four occupancies define the states. B: **Incorrect:** There is no separate fifth occupancy state. C: **Incorrect:** Errors do not add two stored queue occupancies. D: **Incorrect:** FIFO packing does not permit eight independent occupied/empty slot patterns.

**1. Distinguish queries and transitions:** isEmpty only reads occupancy. enqueue adds an element when !isFull, while dequeue removes an element when !isEmpty. **2. Apply the occupancy model:** Adding to an empty queue changes zero occupied positions to one; adding to a two-element queue fills it. **Answer: A, D — isEmpty leaves state unchanged; enqueue takes empty to one element.** A: **Correct:** isEmpty is a query and does not mutate the queue. B: **Incorrect:** The enqueue guard is !isFull, not !isEmpty. C: **Incorrect:** dequeue reduces occupancy rather than changing two elements to three. D: **Correct:** enqueue adds the first element to an empty queue.

**1. Locate x’s definition:** x=a at node 1 is the only definition of x. **2. Count reachable uses:** x is read on predicate edges (1,2), (1,3), (1,4), in z=x at node 3, and in print(x) at node 4. Each is reached without redefining x, giving five DU-pairs. **Answer: 5.**

**1. Track y’s definitions:** y=b at node 1, while y=z at node 2 overwrites it. **2. Inspect the candidate:** [1,2,3,4] has an interior redefinition at node 2, so it is not a definition-clear DU-path carrying node 1’s y to print(y). **Answer: C — [1,2,3,4].** A: **Incorrect:** Node 1’s y reaches its predicate and print uses without redefinition. B: **Incorrect:** It starts at the new definition at node 2 and reaches print(y). C: **Correct:** Its interior node 2 kills the starting definition of y. D: **Incorrect:** Node 3 redefines z, not y.

**1. Locate reads of z:** Node 2 reads z in y=z. Node 4 reads z in print(z). **2. Exclude definitions:** z=c at node 1 and z=x at node 3 write z rather than read it. Hence z has two uses. **Answer: 2.**

**1. Find DU-path obligations for z:** Definition 1 reaches print(z) directly on [1,4], and reaches the read at node 2 on [1,2]. Definition 3 reaches print(z) on [3,4]. **2. Embed them in complete tests:** The offered set {[1,4],[1,3,4],[1,2,3,4]} covers all three obligations; none of the smaller offered sets includes the direct [1,4] obligation. **Answer: D — {[1,4],[1,3,4],[1,2,3,4]}.** **Convention note:** Two tests, [1,4] and [1,2,3,4], would already suffice; D is the only sufficient set among the offered options. A: **Incorrect:** It omits definition 1’s direct print use and its use at node 2. B: **Incorrect:** It does not start at initial node 1 and requires a nonexistent self-loop. C: **Incorrect:** It omits the DU-path [1,4]. D: **Correct:** It contains complete tests covering every z DU-path.

**1. Check the cycle:** [3,5,3] is a simple cycle and a distinct rotation from the already listed [5,3,5]. **2. Check the exit path:** [5,3,4,6] is a maximal simple path. The other candidates repeat node 3 internally and cannot be prime paths. **Answer: A, C — [3,5,3] and [5,3,4,6].** A: **Correct:** This is the missing cycle rotation. B: **Incorrect:** Node 3 repeats internally, so the path is not simple. C: **Correct:** It is simple and cannot be extended while staying simple. D: **Incorrect:** Node 3 repeats internally, so the path is not simple.

**1. Establish the lower bound:** A test choosing 1→2 reaches 4→6 and cannot also visit node 3 or 5. At least two tests are needed even for node coverage. **2. Construct two sufficient tests:** [1,2,4,6] and [1,3,5,3,4,6] together cover all six nodes and all seven edges. Both minima are two. **Answer: B — 2 for both node and edge coverage.** A: **Incorrect:** One test cannot visit both mutually exclusive initial branches. B: **Correct:** Two tests are necessary and sufficient for both criteria. C: **Incorrect:** The displayed two-path construction already covers all edges. D: **Incorrect:** Three are unnecessary.

**1. Check nodes and edges:** The two tests together visit every node and traverse all seven edges. **2. Find missing edge pairs:** Neither test contains [1,3,4] or [3,5,3]. Thus edge-pair coverage fails, and direct prime-path coverage also fails. **Answer: B — Node and edge coverage, without edge-pair coverage.** A: **Incorrect:** All edges are covered, not just nodes. B: **Correct:** All nodes and edges are covered, but required consecutive edge pairs are missing. C: **Incorrect:** The pairs [1,3,4] and [3,5,3] are absent. D: **Incorrect:** The prime paths [1,3,4,6] and [3,5,3] are not directly toured.