Software Engineering · End Term · Jan 2026 · AN

**1. Identify what is missing:** The question is not simply which code changed, but why a design decision was made. **2. Preserve the rationale:** Knowledge documentation records design choices, assumptions and reasons so future maintainers can understand them when the original author is unavailable. Commit messages can help, but merely having version control does not guarantee that rationale was recorded. **Answer: B.** A: **Incorrect:** Version control preserves revisions, but does not by itself document design reasoning. B: **Correct:** Documentation transfers the rationale beyond the original developer. C: **Incorrect:** Refactoring changes code structure; it does not recover undocumented reasons. D: **Incorrect:** Tools do not substitute for recording the relevant knowledge.

**1. Classify the activity:** Competitor study, customer demand and market needs concern product-market research before implementation. **2. Match the listed role:** Among the options, the Marketing Team is the primary fit. Product managers and analysts may also participate in real organizations, but they are not offered here. The engineers mainly turn the selected needs into a technical design and implementation. **Answer: B.** A: **Incorrect:** Engineering primarily designs and builds the technical solution. B: **Correct:** Market and competitor research match the activity described. C: **Incorrect:** Support primarily helps existing customers and handles issues. D: **Incorrect:** An engineering manager coordinates engineering delivery rather than primarily owning market research.

**1. Identify the immediate purpose:** The team wants to execute integration and system tests before accepting faulty code into a release. **2. Use the testing environment:** A controlled testing environment is used to run these automated checks. Pipeline automation can reject or revert a failed build. Staging often supplies a production-like final rehearsal, but the described primary activity is testing, not serving live customers. **Answer: B.** A: **Incorrect:** Development is primarily for implementing and locally exercising changes. B: **Correct:** This environment is dedicated to the integration/system validation described. C: **Incorrect:** Staging emphasizes production-like pre-release validation; the general dedicated test environment is the intended answer. D: **Incorrect:** Running this process on the live customer environment is not the primary testing setup described.

**1. Notice the exposure percentage:** Only 5% of users initially receive the new release. The team observes real behavior before expanding access. **2. Identify the rollout:** This is a canary deployment. A small initial audience limits the impact of defects and provides evidence before a wider rollout. Blue-green instead maintains two complete deployment environments and typically switches traffic between them. **Answer: B.** A: **Incorrect:** Blue-green is defined by parallel environments, not this gradual small-user rollout. B: **Correct:** The small monitored initial rollout is the defining canary feature. C: **Incorrect:** Version deployment is a generic phrase, not the specific strategy. D: **Incorrect:** Canary is explicitly available and matches the scenario.

**1. Identify the rented resource:** The startup obtains a virtual machine with specified CPU and RAM. **2. Identify management responsibility:** It installs and manages its own web server and database on that VM. The provider supplies infrastructure rather than a fully managed application platform or finished software. This is Infrastructure as a Service. **Answer: C.** A: **Incorrect:** A virtual machine is not a dedicated bare-metal machine. B: **Incorrect:** PaaS generally abstracts more of the server/runtime management. C: **Correct:** The provider rents computing infrastructure while the customer manages installed software. D: **Incorrect:** SaaS supplies an application to use rather than a VM to administer.

**1. Find the repeated work:** Many requests need the same product-page query result. Recomputing or refetching it every time wastes resources. **2. Reuse a saved result:** Keeping the result in fast memory lets later requests use a cached copy. This is caching. A practical cache also needs expiry or invalidation so product changes do not leave stale results indefinitely. **Answer: B.** A: **Incorrect:** Compression reduces representation size rather than reusing computed query results. B: **Correct:** Caching stores a reusable result for faster future reads. C: **Incorrect:** A task queue schedules work; it does not by itself reuse its result. D: **Incorrect:** Load balancing distributes requests across servers rather than storing query results.

**1. Identify the testing level:** Unit testing checks individual units, and integration testing checks interactions between combined components. **2. Move to the complete system:** System testing evaluates the fully integrated software against system requirements, such as complete user workflows and expected overall behavior. Its scope is not determined by which developer wrote the code. **Answer: D.** A: **Incorrect:** Individual-module testing is unit testing. B: **Incorrect:** Interface/intermodule behavior is the focus of integration testing. C: **Incorrect:** Developer identity does not define a testing level. D: **Correct:** System testing exercises the complete integrated system.

**1. Observe integration timing:** The modules are all combined at once after separate unit tests. No incremental subsystem integration is described. **2. Name the approach:** Big-bang integration combines everything together before testing the interactions. Because many interfaces become active simultaneously, a failure can be difficult to isolate to a particular module or interface. **Answer: C.** A: **Incorrect:** Top-down integrates progressively from higher-level modules. B: **Incorrect:** Bottom-up starts with lower-level modules and combines them progressively. C: **Correct:** Combining all modules in one step is big-bang integration. D: **Incorrect:** A mixed approach would combine deliberate top-down and bottom-up integration.

**1. Count decisions:** The function tests x>0 and, inside that branch, x modulo 2==0. These are two binary decisions. **2. Use the decision formula:** For this single connected routine, cyclomatic complexity equals number of decision points+1. Therefore V=2+1=3. Else clauses are alternative branches of existing decisions, not new decisions themselves. **3. Identify independent paths:** The paths produce A for positive even x, B for positive odd x, and C for nonpositive x. Inputs 2,1,0 illustrate the three paths. **Answer: B.** A: **Incorrect:** This counts only the two decision nodes and misses the baseline path. B: **Correct:** Two decisions plus one gives complexity 3. C: **Incorrect:** Else does not introduce an additional independent condition. D: **Incorrect:** There are not four decision points in the shown function.

**1. Identify behavior classes:** check(n) returns 'Even' when n modulo 2 is zero and 'Odd' otherwise. The relevant equivalence classes are even integers and odd integers. **2. Test each group:** 2,4,6 all take the Even branch. Every other proposed set mixes at least one even with one odd input, so its members do not all have the same branch behavior. **Answer: A.** A: **Correct:** All three inputs are even and follow the same behavior. B: **Incorrect:** 3 is odd while 2 and 4 are even. C: **Incorrect:** 1 and 3 are odd while 2 is even. D: **Incorrect:** 0 is even and 1 is odd.

**1. Identify the purpose:** The simplified working version helps users discover and clarify their needs before the final system is built. **2. Use the disposal detail:** Because this preliminary version is discarded, it is a throwaway prototype. Its main value is requirement discovery, not becoming the finished production implementation. **Answer: B.** A: **Incorrect:** Waterfall is a sequential lifecycle, not specifically a disposable requirement-discovery version. B: **Correct:** A throwaway prototype is built to learn and then discarded. C: **Incorrect:** Spiral is a risk-driven iterative lifecycle rather than the specific prototype described. D: **Incorrect:** Build-and-fix lacks the deliberate requirement-learning purpose stated here.

**1. Identify the missing scope:** 'Track orders' does not say which order states are included, such as shipped, returned or in-transit orders. **2. Distinguish defects:** The problem specifically highlights omitted coverage details, making the requirement incomplete. Ambiguity means multiple interpretations; inconsistency means contradictory requirements. In real reviews incompleteness can also create ambiguity, but incomplete is the best match to the omission emphasized here. **Answer: C.** A: **Incorrect:** Ambiguity may result, but the stated defect is missing coverage information. B: **Incorrect:** No two mutually conflicting requirements are supplied. C: **Correct:** The list of covered order states is missing. D: **Incorrect:** Nothing establishes that order tracking itself is invalid.

**1. Identify the need:** Stakeholders have differing priorities, so the team needs a shared discussion that reveals both agreement and conflict. **2. Use a facilitated group:** Focus groups allow participants to respond to one another, compare needs and expose differences such as daily versus monthly summaries. Individual interviews and questionnaires gather views but do not themselves provide that direct group interaction. **Answer: C.** A: **Incorrect:** Interviews primarily collect individual views rather than this shared discussion. B: **Incorrect:** Questionnaires collect structured responses without live negotiation. C: **Correct:** A facilitated discussion among stakeholders exposes consensus and conflicting preferences. D: **Incorrect:** Documents show existing recorded information rather than facilitating the described conversation.

**1. Identify the obstacle:** The story is too large to finish within the available sprint capacity. **2. Make the work feasible:** Splitting it into smaller deliverable stories addresses Achievable: each piece should be realistic within the team's time and resources. Sprint length also supplies a time boundary, but among the listed SMART terms the feasibility problem is Achievable. **Answer: D.** A: **Incorrect:** Specific concerns clarity of the objective, not primarily its feasibility. B: **Incorrect:** Measurable concerns how success is assessed. C: **Incorrect:** Relevant concerns alignment with broader goals. D: **Correct:** The split makes the work realistically completable in the sprint.

**1. Identify the risk:** Close can end the editing session and may risk loss of work. Placing it next to Save makes an accidental destructive click more likely. **2. Reduce user error:** Separating the controls is a safety measure. It prevents or reduces harmful mistakes rather than mainly teaching the interface, helping memory or reducing the number of actions. **Answer: C.** A: **Incorrect:** Learnability concerns how easily a new user learns the system. B: **Incorrect:** Memorability concerns returning users remembering it. C: **Correct:** The change reduces the chance of an accidental harmful action. D: **Incorrect:** The primary aim is error prevention, even if the mouse movement becomes longer.

**1. Compare the interaction demands:** A command interface requires remembering a command's name and syntax. Dragging visible file icons lets users identify objects already on screen. **2. Name the heuristic:** Recognition over recall minimizes memory demands by making choices and actions perceptible. The interface relies on recognizing the file and destination rather than recalling a typed command. **Answer: C.** A: **Incorrect:** Exploration is not the main distinction between these two interaction styles. B: **Incorrect:** Consistency is about uniform patterns rather than the specific memory demand. C: **Correct:** Visible objects support recognition instead of recalling command syntax. D: **Incorrect:** The scenario does not primarily concern feedback about system activity.

**1. Identify the evidence source:** The estimates come from senior developers' experience with similar systems. **2. Distinguish formal methods:** Expert judgment uses experienced practitioners' informed assessments. COCOMO uses a specified parametric model. Delphi is a structured multi-round process, typically involving independent or anonymous estimates and convergence; those additional steps are not stated here. **Answer: C.** A: **Incorrect:** Heuristic is broad; the explicit use of experienced people matches expert judgment more specifically. B: **Incorrect:** No COCOMO size-based formula or model is described. C: **Correct:** Experienced developers are asked to provide their informed assessments. D: **Incorrect:** No structured anonymous rounds or Delphi convergence process is mentioned.

**1. Identify the audience and activity:** The team shows completed product features to stakeholders at the end of a sprint. **2. Name the event:** A Sprint Review inspects the completed increment and gathers feedback about product progress and next steps. A retrospective instead examines how the team worked and how to improve its process. **Answer: B.** A: **Incorrect:** Sprint Planning prepares the upcoming sprint rather than demonstrating its result. B: **Correct:** The Review is the stakeholder inspection and feedback event. C: **Incorrect:** The Retrospective focuses on the team's process and improvement. D: **Incorrect:** The Daily Scrum is the short daily coordination event.

**1. Identify decomposition:** The team starts with major project activities and divides them into smaller assignable pieces. **2. Name the planning artifact:** A work breakdown structure organizes project scope hierarchically into manageable work packages. Dependency sequencing, resource allocation and critical-path calculation can then use the defined work; they are different planning steps. **Answer: D.** A: **Incorrect:** Critical-path analysis needs activity durations and dependencies, not just decomposition. B: **Incorrect:** An activity network expresses dependencies among defined activities. C: **Incorrect:** Resource allocation assigns people or equipment to tasks. D: **Correct:** Breaking project scope into smaller work packages creates the WBS.

**1. Identify the intended interaction:** Seller uses an Inventory object by calling viewInventory(). In the course-level object-model description, that link between interacting objects is treated as an association. It is not inheritance or a whole-part ownership relation. **2. Keep the UML distinction clear:** A method call alone strictly establishes usage/dependency, not necessarily a persistent structural association. Association is the source's intended answer if Seller is linked to Inventory as assumed; the stem omits whether such a stored structural link exists. **Answer: C.** A: **Incorrect:** Calling another object's method does not make Seller a subclass of Inventory. B: **Incorrect:** A call does imply usage dependency in precise UML; the source instead intends the broader interacting-object association interpretation noted above. C: **Correct:** This is the source's intended linked-object relationship, under its association assumption. D: **Incorrect:** There is no whole-part collection relationship described.

**1. Identify constructor overload growth:** Many optional settings create too many combinations of constructors. **2. Build step by step:** Builder lets the caller supply mandatory choices and then configure optional features through separate steps before constructing the final object. This avoids one constructor overload for every combination and can validate required properties at the end. **3. Match the scenario:** Processor and RAM are required; graphics, storage, OS, warranty and software are independently chosen options. This is a natural builder use case. **Answer: B.** A: **Incorrect:** A factory selects/creates objects, but does not primarily address this stepwise configuration problem. B: **Correct:** Builder supports controlled incremental construction with many optional features. C: **Incorrect:** Adapter reconciles incompatible interfaces. D: **Incorrect:** Facade presents a simpler interface to an existing subsystem.

**1. Find the concrete dependency:** Seller directly constructs BankAccount and calls its pay method. This ties a high-level selling workflow to a particular payment implementation. **2. Introduce an abstraction:** Seller should depend on a Payment interface, with BankAccount, UPI and other methods implementing it. An implementation can then be supplied without hard-coding it into Seller. **3. Name the principle:** Dependency Inversion says high- and low-level modules should depend on abstractions. The need to modify Seller for each provider is a symptom of that missing abstraction; it also motivates open/closed design. **Answer: B.** A: **Incorrect:** The issue is not clients being forced to implement an excessively broad interface. B: **Correct:** Seller depends directly on a concrete payment class instead of a payment abstraction. C: **Incorrect:** No invalid subtype substitution behavior is described. D: **Incorrect:** The specific highlighted violation is the direction and concreteness of dependencies.

**1. Select the destination branch:** To include feature1 in main, first check out main, for example with git switch main. **2. Merge the feature branch:** Running git merge feature1 while on main integrates that branch's changes into the current main branch, subject to resolving any conflicts. It may fast-forward or create a merge commit depending on history. **3. Distinguish commands:** Checking out feature1 only changes the working branch; it does not integrate it into main. The listed merge command is the intended integration operation. **Answer: C.** A: **Incorrect:** Pull combines remote fetching and integration and does not use this local branch as the intended normal argument. B: **Incorrect:** Checkout changes branches rather than merging changes into main. C: **Correct:** Merge feature1 integrates the feature branch into the currently checked-out destination. D: **Incorrect:** Rebase rewrites a line of commits onto another base and is not the stated straightforward merge operation.

**1. Inspect the parameter's meaning:** flag=0 tells Account to display information, while flag=1 tells it to update information. **2. Classify the dependency:** User is not merely passing business data such as an account number; it sends a control value deciding which internal action Account performs. That is control coupling. Separate display and update methods can make the interface clearer by removing the mode switch. **Answer: B.** A: **Incorrect:** The parameter specifically directs behavior, which is more than ordinary data coupling. B: **Correct:** The flag controls which operation the called module executes. C: **Incorrect:** Common coupling would involve shared global data. D: **Incorrect:** Content coupling would require accessing another module's internals directly.

**1. Define a system client:** A client is an external person or component that uses the system's services across the boundary chosen for the analysis. It need not be a human. **2. Apply the definition:** End users initiate wallet transactions. An inventory service or payment gateway can also be a client when it invokes or exchanges services with the wallet through an interface. **3. Separate development from use:** Writing the implementation makes someone a developer, not automatically a runtime client. A developer can separately be a client when actually using the service, but that is not established merely by writing code. **Answer: A, B, C.** A: **Correct:** People performing transactions directly use the wallet services. B: **Correct:** A separate service can consume the wallet interface and be a client relative to that boundary. C: **Correct:** API interaction can make a payment gateway a system client in the scenario. D: **Incorrect:** Writing code alone is not the client role being analyzed.