The software test method is not fully standardized and unified, mainly because software industry products have a wide range of software. But there are still many commonly available concepts and methods that each software test method is basically available. The ideas and methods described here are available for most applications. Six basic concepts of software test methods are important: white box testing, black box testing, gray tape test, effective use case and invalid use case, boundary conditions, and equivalents test. White-Box Testing or Glass-Box Testing is tested through the source code of the program without using the user interface. This type of test requires the disadvantage or error of internal code from the code syntax to algorithm, overflow, path, condition, and the like, which is further corrected. Black-Box Testing is strictly tested by using the entire software or some software function, and does not understand the source code of the program or a source of software or some software function. How is the code program is designed. The tester knows how software works by entering their data and look at the result of the output. Usually testers are not only entered input data that is definitely out of the test, but also use challenging input data and input data that may be wrong to understand how software processes various types of data. Gray-Box Testing: Gray-Box Testing: It is like a black box test, but the tester has understood the software or some software function. of. Even some source code is also read. Therefore, testers can be targeted to perform a certain identified condition / function test. The meaning is that if you know the design of the product and the in-depth understanding of the product through the user interface, you can test its performance more effectively and in-depth from the user interface. Valid case (Valid case) or a legal input case: is a test case that known software programs can handle correctly. Generally mean the test cases for software input. For example, in Microsoft Excel, enter "= 1 1" with the keyboard, see the result "2". The effective use case input here is "= 1 1". Invalid use case (Invalid case is not legal input case) or error case: Yes, those test cases that do not support processing in advance. For example, in Microsoft Excel, enter "= a 1" with the keyboard, see the result "#NAME?". The "= A 1" input here is both an invalid use case and an error use case. Boundary Cases: Tests around the boundary value. It is usually meant to maximize, minimum or the longest string, or the like can be processed. For example, the font size of a software font is: from 8 to 72. Then the boundary test case should include: less than 8, equal to 8, equal to 72 and greater than 72. Equivalent Classes: Equivalent type test case refers to if there are many test cases to do more, no new defects are found. Because although the inputs and output results are different, they all passed the source code paths of the same software. Usually, as long as one source code program is used to handle all values within a certain numerical range, except for boundary values, all values within the boundary value range are generally equivalent. Because if the software program can handle a value correctly, the program can correctly handle any other valid input values other than boundary values within this range. Let's use the font size of the above software font to explain. The font size of the software support is: from 8 to 72. So all supported font numbers between 8 and 72 can be considered an equivalent class test case.
The basic method of software testing software test methods may have different categories in different books, different schemas and different interpretations. For example, from the perspective of testers, 馐 远 4 4 创 创 创 创 瓤 瓤 瓤 瓤 瓤 瓤 创 瓤 瓤 瓤 瓤 瓤 瓤 瓤 馐 治 洳 洳 洳馐 洳馐 洳馐 突 洳馐 洳馐 饫镆 饫镆 鄣 鄣 ㄖ饕 懿 懿 懿 馐 馐 椒 ㄖ饕 ㄖ饕 懿 懿 懿 懿 懿 懿 懿 懿 懿 懿 懿 懿 懿 懿 懿 懿馐 馐 `` u ion ion, user interface (User Interface or UI) test, Ad hoc generally refers to 'discussion or open' test, boundary condition, performance test (Performance TEST, regression test, strong test, configuration and installation test, compatibility test, international support test (INTERNATIONAL SUFFICIENCY), and localized language test ( Localization. Function Test: Verify that the test software function can work normally according to its design. Whether the expectation behavior when running software is in line with the original design. For example, the feature that tests Microsoft Excel insert-> symbols includes testing that can be correctly inserted into the symbol in the Microsoft Excel's selected unit and display the correct symbol? Can you correctly display symbols of different fonts? // Note The following test should be a Smoke Test Acceptability Test: It is a simple test of the most basic function before the test version of the version of the test department. Because the version of the test version of the test department should first verify that the version is basically more stable for the feature tested. Some minimum requirements must be met. For example, it will hang or crash if it is easy. If a new version does not pass tuned verification, it should block the test sector to test the test version. At the same time, it is also necessary to find the main defect that causes the instability of this version and urges to correct as soon as possible. User Interface Test: Analyze whether the design of the software user interface is based on the user's expectations or requirements. It often includes tests on all buttons, dialogs, and dialogs on all buttons, text, error prompts, menu and help contents. For example, test the size of the dialog box used in the Microsoft Excel, all buttons are aligned, string font size, error information content, and font size, toolbar position / icon, etc. 'Explore or open the' type test: Not in accordance with one of the formal test cases according to the class, nor is it limited to test case specific steps. This test is a variety of functions of using the software on the basis of understanding the software functionality, using flexible imagination and creativity to simulate users. This software is usually involved in many test cases or by more complex steps. Boundary condition test: It is a test of the boundary value. It is usually meant to test whether the functions of the software can correctly process the maximum, minimum, or the longest string that can be processed by the software. Performance test: Usually verify that the performance of the software is repeatedly used under normal environment and system conditions to meet performance indicators. Or when you perform the same task, the new version is not slower than the old version. Generally, the system memory capacity will not be lost when the program is running. For example, the verification program saves a huge file new version is not slower than the old version. Return test: Testing according to the repaired defects.
The purpose is to verify that the defects that have previously been repaired but have been repaired no longer reappear. Generally refers to the step of retesting the steps that a known correction is again surrounded by it. It is often difficult to determine the range of re-test required, especially when approaching the product release date. Because the source code must be changed when a defect is to be corrected, it is possible to affect the functions controlled by this partial source code. So when verifying the deficiencies, it is necessary to obey the steps of defects, but also all functions that may be affected. It should therefore be encouraged to automate all regression test cases. Powerful test: It usually verifies whether the performance of the software can work properly under various extreme environments and system conditions. Or is the ability to verify the performance of the software under various extreme environments and system conditions. For example, under the lowest hard drive space or system memory capacity conditions, the verification program repeatedly performs a huge file 1000 times will not crash or crash after 1000 times. Integration and Compatibility Test: Verify that this feature can be coordinated with other programs or components as expected. Compatibility often means coordination between old and old versions, including the use of products that are tested and compatible with other products. For example, when using the same product, it does not affect the operation of saving files, formats, and other data between the old versions of users. Assembly / Installation / Configuration Test: Verify that software programs are on hardware of different manufacturers, and software on different languages supporting, and software installed in different ways can operate as expected. For example, install the English version of Microsoft Office 2003 on the Korean version of Windows me, and then verify that all functions are running normally. International Support Test: Verify that the software program can also operate as expected on different countries or regions, and can also be respected and supported in accordance with the original design, font, text, text representation, special format, etc. For example, can the Arab string with the English version of Windows XP and Microsoft Word? Can Microsoft Word with arabic version of Windows XP and Arabic version? For example, whether the Japanese version of the Microsoft Excel dialog box displays Japanese correctly translated? Once the test personnel to implement international support tests often need to basically understand the language requirements and expectation behavior of these countries or regions. Localization Language Test: To verify that all the different language version softwares that have been planned to be published are properly translated into local languages as expected. Such tests typically include verification menus, dialogs, error information, all user interfaces, including all user interfaces, can display properly translated local text. The principle of software testing should put "software testing as early as soon as possible" as software developers, the test case should be used by test input data and corresponding expected output results. These two components should avoid checking their program in design testing. When using an example, it should include reasonable input conditions and unreasonable input conditions.
Software test object software test is not equal to program testing. Software Tests should run through software-defined documents obtained from the entire period of software definition, summary design, detailed design, and program coding, including requirements specifications, profile design specifications, detailed design specifications, and source programs, Should become the object of software testing
Software Test Classification White Box Test: How to test the software (how) to complete the function and do what kind of function (what), white box test is generally made of black box test by developers: I only know what WHAT is tested, black box test The unit test is generally made by the tester is a white box test, the purpose is to verify whether the designed class meets the needs JUnit is a unit test tool.
Test Framework Definition Framework defines some of the box boxes. The test code written must follow these bar frames: inherit a class, realize a certain interface Framework is what we are saying
