Software Testing · End Term · Jan 2026 · AN

**1. Identify the risk:** Checkout and payment previously worked, but refactoring and integration changes can accidentally break their existing behavior. **2. Recheck established behavior:** Regression testing reruns suitable tests to detect such unintended breakage. These may be functional tests, but regression describes why they are being rerun after a change. Performance and usability have different specific goals. **Answer: B.** A: **Incorrect:** Performance testing measures responsiveness or capacity rather than specifically detecting regressions. B: **Correct:** Regression testing checks that changes have not broken previously working behavior. C: **Incorrect:** Functional tests can be part of the regression suite, but regression is the more precise purpose here. D: **Incorrect:** Usability testing concerns users' ease of interaction.

**1. Define the target:** Usability concerns whether representative users can achieve their goals effectively, efficiently and with an understandable interaction. **2. Test with users and tasks:** Observing users attempt realistic tasks can reveal confusing labels, navigation difficulties or error-prone controls. Syntax checks, installation compatibility and database capacity do not directly answer that usability question. **Answer: C.** A: **Incorrect:** Code correctness and syntax are not the principal usability objective. B: **Incorrect:** Installation and compatibility testing concern deployment across environments. C: **Correct:** This directly measures ease and intuitiveness of user interaction toward goals. D: **Incorrect:** High transaction volume concerns performance and scalability.

**1. Start with S:** The rule S→aXa places one a at each end. Those outer letters are fixed and are not repeated by any rule. **2. Expand X:** Each use of X→bX adds one b. Eventually X→b terminates the derivation, so at least one b is required. This gives X⇒b to the power n for every integer n≥1. **3. Combine:** The language is \(\{ab^na:n\geq1\}\). For example, S⇒aXa⇒aba, or S⇒aXa⇒abXa⇒abba. Empty interiors and repeated aba blocks cannot be generated. **Answer: D.** A: **Incorrect:** The grammar does not recursively add matching a blocks at the two ends. B: **Incorrect:** Repeated aba blocks are not the derivation pattern. C: **Incorrect:** Allowing n=0 would include aa, but X has no empty production. D: **Correct:** There is exactly one a at each end and one or more b letters between them.

**1. Understand backlog:** A maintenance backlog contains unresolved reported problems or requested fixes. An index of incoming versus completed maintenance work indicates whether that backlog is accumulating or being reduced. **2. Match the metric category:** BMI therefore measures maintenance handling and quality, rather than properties of object-oriented class structure or only the initial product's static characteristics. Interpret a reported BMI using its specified period and numerator/denominator convention. **Answer: C.** A: **Incorrect:** BMI tracks handling of maintenance work rather than directly measuring an inherent product property. B: **Incorrect:** It is not a class/inheritance metric. C: **Correct:** Backlog handling is a maintenance-quality concern. D: **Incorrect:** This is not primarily a metric of activities during the original development process.

**1. Identify what is bypassed:** Direct URL access skips expected pages or workflow stages. For example, a user may attempt to reach confirmation without completing earlier checks. **2. Classify the bypass:** This attacks the application's control flow or permitted transition sequence. Parameter or value testing instead alters submitted inputs while session-based tests focus on session-related data. **Answer: B.** A: **Incorrect:** Parameter-level tests modify or omit parameters rather than primarily skipping workflow steps. B: **Correct:** The test deliberately bypasses the required sequence of application states. C: **Incorrect:** Value-level tests vary input values rather than navigation order. D: **Incorrect:** No particular session-data manipulation is described.

**1. Match access changes:** Access Modifier Change changes whether a field or method is public, protected, private or package-visible. This matches description B. **2. Match reference changes:** Reference Type Change changes compatible object references on the right-hand side of assignments, matching A in the supplied table. **3. Match construction changes:** Actual Type Change changes the concrete object type used in a new expression, matching C. Therefore the mapping is 1-B,2-A,3-C. **Answer: D.** A: **Incorrect:** This swaps access/reference/actual-type descriptions incorrectly. B: **Incorrect:** Access modifiers do not describe assignment-object changes in A. C: **Incorrect:** Actual type change should match new expressions in C, not assignment references in A. D: **Correct:** Each operator is paired with the corresponding mechanism in the table.

**1. Distinguish the user's submission:** Initially the user submits a form, but the question asks about the subsequent move after successful server processing. **2. Identify server-directed navigation:** The server instructs the browser to load the confirmation page automatically. In the application transition graph this is a redirect transition, rather than a user clicking an ordinary link or submitting another form. **Answer: C.** A: **Incorrect:** The described transition specifically involves automatic redirection. B: **Incorrect:** The original submission is a form transition, but the subsequent server-directed page load is different. C: **Correct:** A server-instructed automatic move is a redirect transition. D: **Incorrect:** A simple link transition would involve following a hyperlink.

**1. Identify the statement:** super(brand) explicitly invokes Vehicle's constructor from Car's constructor. **2. Match the targeted mutation:** Deleting this parent-constructor invocation is Parent Constructor Deletion. It is not deletion of an overriding method or a hiding field. **Compilation note:** In this exact example, deleting the call would introduce an implicit super(), but Vehicle has no no-argument constructor. Such a mutant can therefore fail compilation and must be classified as an invalid/stillborn mutant rather than executed as an ordinary valid program. **Answer: C.** A: **Incorrect:** General super keyword deletion is not the named operator specifically targeting this constructor call. B: **Incorrect:** No overriding method body is being deleted. C: **Correct:** The removed construct is the explicit parent-constructor invocation. D: **Incorrect:** No hidden instance variable is being removed.

**1. Count choices per characteristic:** The partitions have 2 A choices, 3 B choices, 2 P choices and 2 Q choices. One choice of each is already represented in the base test (A2,B3,P1,Q1). **2. Apply Base Choice Coverage:** Keep every other characteristic at its base value while varying one characteristic to each nonbase block. The additional-test count is \[(2-1)+(3-1)+(2-1)+(2-1)=1+2+1+1=5.\] **3. Distinguish total and additional:** Including the original base test, there are six tests; the question asks for the five additional tests. **Answer: C.** A: **Incorrect:** This misses two required nonbase variations. B: **Incorrect:** This misses one nonbase variation. C: **Correct:** The sum of nonbase blocks is exactly five. D: **Incorrect:** Six includes the already supplied base test.

**1. Notice two fields with the same name:** SmartBulb and DimmerBulb each declare their own brightness field. Fields are hidden, not dynamically overridden like methods. **2. Trace the state update:** DimmerBulb.powerOn sets the inherited isOn field true, but assigns 100 to DimmerBulb.brightness. **3. Trace the state observation:** The inherited SmartBulb.showStatus method reads SmartBulb.brightness, still 0. Thus it prints ON with brightness 0 instead of the intended 100. The update and observation refer to inconsistent state definitions, giving the named State Definition Inconsistency Anomaly. **Answer: B.** A: **Incorrect:** The more specific inconsistency arises from the two distinct hidden brightness definitions. B: **Correct:** The overridden update and inherited observation use different brightness fields. C: **Incorrect:** This is not a mismatch in the runtime type used at the call site. D: **Incorrect:** The principal defect is state-definition inconsistency, not changing a field's visibility.

**1. Use the actual object class:** obj is a C object. Each instance-method call uses the most specific inherited implementation for that signature. **2. Resolve the four methods:** C does not override u, so B.u runs. C overrides v, so C.v runs. w is inherited from B. C overrides x, so C.x runs. **3. List in call order:** B.u, C.v, B.w, C.x. Inherited methods are not automatically replaced simply because an object belongs to a deeper subclass. **Answer: D.** A: **Incorrect:** Both u and x are resolved to more specific implementations than A in this program. B: **Incorrect:** This ignores the B override of u and the C overrides. C: **Incorrect:** C provides v and x, so their inherited older implementations are not used. D: **Correct:** This is the correct most-specific implementation for each signature.

**1. Substitute symbolic inputs:** mode=m0, ambientTemp=a0 and targetTemp=t0. The local gap becomes t0-a0. **2. Negate the helper result:** isEcoMode(mode) returns mode==0. Therefore !isEcoMode(mode) requires m0≠0. **3. Follow the nested branch:** abort is inside if(gap>=3), adding t0-a0≥3. Both conditions must hold on the same execution, so the path constraint is \[(m_0\ne0)\land(t_0-a_0\geq3).\] For example, mode=1, ambient=20, target=23 reaches abort. **Answer: C.** A: **Incorrect:** This makes Eco mode true, so its negation prevents entry. B: **Incorrect:** This uses the wrong inequality for the inner branch. C: **Correct:** Both required branch conditions are correctly conjoined. D: **Incorrect:** A disjunction permits paths where one required condition is false.

**1. Compare only the changed token:** currentBalance in the condition becomes accountLimit. The comparison operator >=, the withdrawalAmount operand and the called method stay unchanged. **2. Classify the mutation:** One scalar variable reference is replaced by another compatible scalar variable reference. This is scalar variable replacement, not changing arithmetic, relational or assignment operators. **Answer: A.** A: **Correct:** The mutation replaces one variable operand with another. B: **Incorrect:** No arithmetic operator is changed. C: **Incorrect:** The relational operator >= is unchanged. D: **Incorrect:** No assignment operator is changed.

**1. Identify unconditional sections:** P1 initializes and obtains the data. P7 prints the closing output. Both surround the conditional body. **2. Follow the outer choice:** Invalid city data executes P2 only. Otherwise the code processes the available list: P3 repeats for its items, then P4 computes the average. **3. Follow the inner choice:** After P4, exactly one of P5 or P6 prints the high/normal result. With a dot for sequence, a bar for choice and a star for repetition, the expression is \[P_1\cdot\big(P_2\mid(P_3^*\cdot P_4\cdot(P_5\mid P_6))\big)\cdot P_7.\] Only P3 is repeated; the average calculation P4 happens after the loop. **Answer: A.** A: **Correct:** This preserves the outer choice, repetition of P3 alone, inner choice and final P7. B: **Incorrect:** This incorrectly repeats the average calculation P4. C: **Incorrect:** The nesting and alternatives do not match the code's invalid-data versus valid-data paths. D: **Incorrect:** This incorrectly sequences both mutually exclusive printing branches instead of choosing one.

**1. Identify definitions of x:** Node 1 defines x=a. Node 3 redefines x=b. Other relevant nodes either use x in predicates or print x at node 5. **2. Trace definition-clear routes:** The supplied tests cover routes 1-2-5, 1-4-5, 1-2-4-5 and 1-3-4-5. The first three carry the node-1 definition to its downstream predicate/output uses. The fourth visits the node-3 definition and carries that new value to the print. **3. Compare coverage levels:** These paths cover every definition-clear def-use route for x in this acyclic graph. Thus All-DU-Paths holds, which also implies All-Uses and All-Defs; the 'only' alternatives are too weak. **Answer: C.** A: **Incorrect:** All-Defs is achieved, but the given tests satisfy a stronger criterion. B: **Incorrect:** All-Uses is achieved, but not merely that weaker level. C: **Correct:** All definition-clear paths connecting x definitions to uses are covered. D: **Incorrect:** The graph does satisfy a listed coverage criterion.

**1. Count states, not edges:** The FSM has states 1 through 6, so state coverage has six state requirements. The number of labelled transitions is a separate measure. **2. Trace ABDWQ:** Start at 1. A goes to 2, B to 3, D back to 2, W to 5 and Q to 6. State 6 is accepting, so the sequence is accepted. **3. Trace AWVXY:** A goes 1→2 and W goes 2→5, but V is defined from state 1 to 6, not from 5. The sequence cannot continue as written. **Answer: B, C.** A: **Incorrect:** Eight confuses some other graph count with the six state requirements. B: **Correct:** The complete labelled path ends in accepting state 6. C: **Correct:** There is one requirement for each of the six states. D: **Incorrect:** After AW the machine is in state 5, where V is unavailable.

**1. Translate the policy:** Free shipping is allowed when total>=500.00, including the exact boundary 500.00. **2. Compare implementations:** A directly returns that predicate. D returns false below 500 and true otherwise, which is equivalent for ordinary finite order totals. C uses >500 and fails the provided exact-boundary test. **3. Inspect B beyond the tests:** total>499.99 passes all five listed tests, but is not equivalent over arbitrary double values; for total=499.995 it grants free shipping below 500. If valid currency inputs are strictly restricted to exact two-decimal amounts, this distinction disappears. The source intends the complete threshold policy over the method's numeric input domain, not merely passing the five examples. **Answer: A, D.** A: **Correct:** The inclusive comparison exactly encodes the free-shipping threshold. B: **Incorrect:** It incorrectly accepts values between 499.99 and 500 over unrestricted finite doubles, despite passing the shown tests. C: **Incorrect:** The strict comparison rejects 500.00 and fails a provided test. D: **Correct:** The below-threshold rejection and otherwise-true result match the policy for ordinary finite totals.

**1. Write the predicate:** \(p=(c\lor\neg b)\land\neg a\). Clause a determines p only when the other factor c or not b is true. **2. Apply restricted active coverage:** In each pair, change a while holding b and c fixed, and require the predicate to change. The fixed pairs (b,c)=(T,T),(F,T),(F,F) satisfy the active condition. **3. Map to row numbers:** With the table's row ordering, these yield (1,5), (3,7) and (4,8). For (b,c)=(T,F), p is false regardless of a, so (2,6) is inactive rather than active. **Answer: C.** A: **Incorrect:** This wrongly includes inactive pair (2,6) and omits active pair (4,8). B: **Incorrect:** Pair (2,6) keeps p false and therefore does not demonstrate active determination. C: **Correct:** All three pairs hold minor clauses fixed and flip p when a changes. D: **Incorrect:** Rows 1 and 6 also change minor-clause values and do not form a restricted active pair.

**1. Make a inactive:** For \(p=(c\lor\neg b)\land\neg a\), changing a cannot matter when c or not b is false. That requires c=F and b=T. **2. Hold the minor clauses fixed:** Rows 2 and 6 have those same b,c values and opposite a values. Both give p=F. Thus (2,6) supplies restricted inactive coverage with the predicate false. **3. Check predicate true:** If p=T, both factors must be true, so a=F and c or not b=T. Changing a to T would make p false. Hence a cannot be inactive while p remains true, and no corresponding true-output pair exists. **Answer: B.** A: **Incorrect:** These are not the inactive false-output pair, and p cannot remain true when a changes. B: **Correct:** Only (2,6) works for p false; no restricted inactive pair exists for p true. C: **Incorrect:** The listed pairs are active pairs, not the required inactive pair. D: **Incorrect:** This reverses the active/inactive interpretation and output conditions.

**1. Count graph components:** The CFG has N=6 vertices and E=8 directed edges. It is one connected control-flow component. **2. Apply the formula:** Cyclomatic complexity is \(V(G)=E-N+2=8-6+2=4\). **3. Interpret:** Four is the size of a basis of independent paths, not the total number of all possible start-to-finish paths. Counting branch alternatives also gives the same independent-path complexity. **Answer: C.** A: **Incorrect:** Six is the number of vertices, not the complexity. B: **Incorrect:** The required formula gives four, not three. C: **Correct:** E-N+2 gives 4 for this connected CFG. D: **Incorrect:** Five does not match the graph counts.

**1. Define an edge pair:** An edge-pair requirement is a reachable path of two adjacent edges, represented by three vertices. Use the given directed edges rather than arbitrary triples. **2. Enumerate by middle vertex:** Through vertex 2: (1,2,3),(1,2,5). Through 3: (1,3,4),(1,3,6),(2,3,4),(2,3,6). Through 4: (3,4,5). Through 5: (2,5,6),(4,5,6). **3. Count:** There are 2+4+1+2=9 distinct two-edge requirements. Every listed pair lies on a start-to-finish route in this graph. **Answer: B.** A: **Incorrect:** This misses three valid two-edge paths. B: **Correct:** Enumerating the compatible incoming/outgoing edges gives nine pairs. C: **Incorrect:** Eight is the number of edges, not the number of consecutive edge pairs. D: **Incorrect:** Seven omits two valid pairs.