Software Testing and Quality Control

Class testing (unit testing for classes) requires a driver - test code that instantiates the class, calls its methods, and verifies results. Classes typically DO need instances of other classes (dependencies) to test interactions, and transitions between states should be tested. Option A correctly identifies that drivers are essential; B and C are incorrect.

Build Verification Test (BVT) and smoke test are often used interchangeably - both are quick tests run after each build to verify basic functionality and catch major errors that would make further testing pointless. BVT ensures the build is stable enough for detailed testing. Both A and B correctly describe aspects of BVT.

Glass box testing (white box testing) provides three key benefits: focused testing on specific code paths or modules, measurable test coverage through code analysis, and visibility into control flow structures including decision points and branches. By seeing inside the code, testers can target areas that black box testing might miss, ensure all logic paths are exercised, and verify internal correctness rather than just input-output behavior.

Internal failure in testing context refers to addressing defects at their origin within the development process itself. Training programmers to write better code and introduce fewer bugs represents a preventive approach that targets the root cause of defects - the human element and development practices. This is fundamentally different from external failures that manifest during testing or in production, as it focuses on prevention at the source rather than detection downstream.

Structural testing examines the internal implementation details of software, including code structure, logic flow, algorithms, and data handling. It requires knowledge of and access to the internal code architecture, making it synonymous with white box testing. This contrasts with black box testing (also called functional testing) which only considers inputs and outputs without knowledge of internal workings.

A test driver is a control program that invokes and coordinates the execution of the software component being tested with various test data sets. It serves as the test harness that drives the testing process, captures output from each test case, and organizes results for evaluation. Test drivers are essential for automated testing and enable systematic execution of multiple test scenarios without manual intervention.

Exhaustive testing means testing all possible combinations of inputs and preconditions. While theoretically possible, it becomes impractical for real-world systems due to the combinatorial explosion of test cases. The answer correctly identifies this dichotomy - it's possible in principle but not feasible in practice.

White-box testing techniques require knowledge of internal code structure. Statement testing, decision testing, and condition coverage all examine internal code paths. Boundary value analysis is a black-box technique that only considers input/output boundaries without internal code knowledge.

Test closure activities occur after testing is complete and include handing over testware, archiving testware, and evaluating the process. The hand-over of test-ware (test scripts, data, logs) to maintenance or other teams is explicitly part of test closure.

Smoke testing (also called confidence testing) is the initial testing done to verify that the most crucial functions work. If the smoke test passes, you have confidence the build is stable enough for further testing. Sanity testing focuses on specific defect fixes, not initial confidence.