Ideia principal: A análise de valor de limite (BVA) complementa o teste de classe de equivalência (ECT), concentrando-se em testar as bordas das classes de equivalência, onde é mais provável que ocorram falhas, aumentando assim a precisão e a eficácia dos testes.

**Conteúdo

  • Integração do BVA com o ECT

    • Finalidade: O BVA é usado para direcionar os limites das classes de equivalência identificadas no ECT. Como os erros são frequentemente encontrados nas bordas dos intervalos de entrada, o BVA testa sistematicamente esses pontos de limite juntamente com os valores dentro e fora dos limites.
    • Metodologia: Após definir as classes de equivalência por meio do ECT, o BVA é aplicado a essas classes selecionando os valores extremos (mínimo e máximo) e os valores imediatamente adjacentes. Essa abordagem foi projetada para descobrir falhas relacionadas às condições de limite que podem não ser detectadas testando apenas valores representativos dentro das classes.
  • Abordagens de 2 e 3 valores:

    • Abordagem de 2 valores: Envolve o teste de dois valores de limite para cada classe de equivalência - normalmente os limites superior e inferior. Esse método garante que os casos extremos de cada classe sejam examinados, mas pode deixar passar problemas que ocorrem ligeiramente dentro dos limites.
    • Abordagem de 3 valores: Estende a abordagem de 2 valores, incluindo um valor adicional dentro de cada limite. Essa abordagem testa os valores de limite, além de um valor dentro do limite e outro fora dele, proporcionando um exame mais completo dos possíveis problemas de casos extremos.

Atomic Note: Integrating Boundary Value Analysis with Equivalence Class Testing

Title: Application of Boundary Value Analysis in Equivalence Class Testing

Main Idea: Boundary Value Analysis (BVA) complements Equivalence Class Testing (ECT) by focusing on testing the edges of equivalence classes, where faults are more likely to occur, thereby enhancing the precision and effectiveness of tests.

Content:

  • Integration of BVA with ECT:

    • Purpose: BVA is used to target the boundaries of equivalence classes identified in ECT. Since errors are often found at the edges of input ranges, BVA systematically tests these boundary points along with values just inside and just outside of the boundaries.
    • Methodology: After defining equivalence classes through ECT, BVA is applied to these classes by selecting the extreme values (minimum and maximum) and immediate adjacent values. This approach is designed to uncover faults related to boundary conditions that might not be detected by testing only representative values within the classes.
  • The 2-Value and 3-Value Approaches:

    • 2-Value Approach: Involves testing two boundary values for each equivalence class – typically the upper and lower limits. This method ensures that the edge cases of each class are examined, but it might miss issues that occur slightly inside the boundary limits.
    • 3-Value Approach: Extends the 2-value approach by including an additional value just inside each boundary limit. This approach tests the boundary values, plus one value just inside the boundary and one just outside, providing a more thorough examination of potential edge case issues.

Context & Linkage:

  • This note elaborates on the integration techniques of BVA within ECT as outlined in the “Boundary Value Analysis Applied to Equivalence Classes” section of the provided slides, with further reference to “Introduction to Software Testing” by Ammann and Offutt.
  • Related Notes: Can be linked to practical examples of BVA in various testing scenarios, discussions on the efficacy of the 2-value versus 3-value approaches, and additional boundary testing techniques that might be used in conjunction with ECT and BVA.

This atomic note explains how Boundary Value Analysis can be effectively integrated with Equivalence Class Testing, highlighting two common approaches for selecting boundary values. It emphasizes the importance of boundary testing in discovering potential defects at the most vulnerable points of the input domain.