Black Box Testing Methods Revealed

1. The important role of black box testing in the rapid application development (rad) environment

Software testing methods are generally divided into two types: white box testing and black box testing. White box testing, also known as structural testing, logic-driven testing or testing based on the program itself, focuses on the internal structure and algorithm of the program and usually does not care about functions and performance indicators. Black box testing, also known as functional testing, data-driven testing or specification-based testing, actually stands on the standpoint of the end user to check whether the input and output information and system performance indicators meet the functional requirements and performance requirements in the specification.

As the RAD environment evolves, software engineering faces new challenges, including:

The scale of application systems is becoming larger and larger, and the structure is becoming more and more complex;

The development team has more and more people, and the division of labor is becoming more and more detailed;

The increasing project investment leads to greater investment risks.

In such a context, software quality faces a greater crisis, and the key to solving the problem is black box testing. However, since traditional black box testing is often limited to manual testing, which is carried out spontaneously based on the experience of engineers and lacks a strict testing management mechanism, the effect is not obvious.

Before distributing an application system, if it is not subjected to scientific and thorough black box testing, it is equivalent to delivering a large number of hidden defects to the end user. This is irresponsible for the development team itself, the project investors and the end users, and will seriously damage the interests of all three parties.

Today, software quality requirements are receiving more and more attention. Black box testing plays an important and irreplaceable role in software quality supervision. With the progress and development of software development platforms and software design ideas, especially the development of RAD technology, more specific requirements are put forward for black box testing. People have found that it is necessary to follow certain testing theories and rely on excellent testing tools to conduct scientific and complete testing.

2. Black Box Testing Operation Steps

In the traditional software development life cycle, testing is often postponed to the later stage of the entire development process. That is, testing is started only when the coding of the application is basically completed. The disadvantages of this are:

a) As the application is large and complex, the testing work is complicated and it is difficult for testers to organize scientific and comprehensive test cases, which greatly increases the testing cost and seriously affects the comprehensiveness and effectiveness of the test;

b) Since the time interval between the development and testing of the module involved in the defect is long, the programmer has to pay a higher cost for modification and maintenance;

c) Due to the constraints of the distribution date, testing work is often completed in a hurry, leaving a large number of defects to the end users. In other words, the real testing work is actually done by the end users.

Therefore, in order to ensure that the testing work is carried out scientifically, accurately, comprehensively and orderly, a strategy of developing and testing at the same time should be adopted, so that the development work and testing work can be carried out in parallel. This is the concept of "the earlier the test, the better" as the saying goes. [page]

A complete set of tests should consist of five phases:

1. Test plan

First, based on the specifications of functional requirements and performance indicators in the user demand report, define the corresponding test requirement report, that is, formulate the highest standard for black box testing. All future testing work will be carried out around the test requirements. Applications that meet the test requirements are qualified, otherwise they are unqualified. At the same time, it is necessary to appropriately select the test content, and reasonably arrange the test personnel, test time and test resources.

2. Test Design

Decompose and refine the test requirements formulated in the test planning phase into several executable test processes, and select appropriate test cases for each test process (the quality of test case selection will directly affect the effectiveness of the test results).

3. Test development

Establish repeatable automated testing processes.

4. Test Execution

Execute the automatic testing process established in the test development phase and track and manage the defects found. Test execution generally consists of unit testing, combination testing, integration testing, system joint debugging and regression testing. Testers should conduct tests step by step in a scientific and responsible manner.

5. Test Evaluation

Combined with quantified test coverage and defect tracking reports, a comprehensive evaluation is conducted on the quality of the application software and the work progress and efficiency of the development team.

Obviously, black box testing can only ensure the quality of the application if it is carried out strictly according to the steps. However, if there is no excellent testing tool to help, relying solely on manual testing will not only consume a lot of manpower, material and financial resources, but also make many testing tasks difficult or even impossible to achieve.

3. Comparison between manual testing and automatic testing

Manual testing cannot guarantee the scientificity and rigor of black box testing because:

Testers are responsible for the preparation and compilation of a large number of documents and reports, which may be overwhelming;

Due to the limitations of software distribution date, development cost, personnel, resources and many other factors, it is difficult to conduct comprehensive testing;

If the time it takes to fix a defect is considerable, regression testing becomes extremely difficult;

There is a lack of scientific and effective management methods for the large number of defects found during the testing process. Responsibilities become unclear and no one can provide accurate data to the decision-makers to measure the current work progress and efficiency.

Fatigue and other human factors caused by repeated testing make the testing standards inconsistent. The longer the test takes, the lower the rigor of the test.

It is difficult to test invisible objects or invisible properties of objects.

Therefore, automatic testing becomes the best solution. The so-called automatic testing is actually to hand over a large amount of repetitive work to the computer to complete. An excellent automatic testing tool can not only meet the basic requirements of scientific testing, but also save a lot of time, cost, personnel and resources, and the test scripts can be reused (including being used by different projects).