
Return to return to test the strategySay for a software development item, the test set of the item atCarry out the process of the test in will keep with the example the test developIn" the test uses a database", and as to it's carry on the maintenance and managements.When get aThe 基 line edition of a software hour, the havings that useds for the 基 line edition test measureTry out the example becomes the 基 line test to use a database.At need to carry on back to returnTime of test, can according to choose of return to return to test the strategy, fromThe 基 line test uses to withdraw the fit test to constitute back to return to measure with the example in a databaseTry the pack, pass to circulate back to return the test to wrap to carry out back to return the test to protect the existenceThe 基 line test use the test within a database and may be automatic test feet with the exampleOriginally, also probably is a test to carry out the process with the handicraft of the exampleReturn to return the test to need time, budget and manpowers to plan, implement andManage.For the sake of under the budget and degree of progresses that give settle, possibly efficient andCarry on back to return the test effectively, need to carry on the 维 to the test with a databaseProtect also the certain strategy choice of basis correspond of return and return the test wrap. 2.1 test with the maintenance of a databaseFor satisfying customer's demand with maximum limit and wanting appliedlyBeg, the software will be modify multifariously and push continuously in its life cycleA new edition, after the modification of or new the software of the edition will increase aSome new function or produce some varieties on the software function.Along withThe change of the software, the function and applications of the software connect the and software solidTook place to turn into now, the test uses some tests within a databases to use example possibilityWill lose to aim at sex and usefulnesses, but another some tests use the example is possibleBecome dated, have some the test to use the example and will can't circulate completely forGuarantee the test use a database in test use the valid flight of steps leading to a palace hall of the example, must to testCarry on the maintenance with a database.At the same time, was modify of or the software 功 that add to addAbility, only depend re- circulate the test of the past use example combine shortage with announce to publicThe problem that is among them, have the necessity to annex the new test to test these with the exampleNew function or characteristic.Therefore, the test returns with the maintenance work of a database shouldInclude to develop the new test to use the example, these new tests use to test with the exampleThe new characteristic of the software perhaps overlay the existing test to use the example can't overlay softA function or characteristicThe test uses the maintenance of the example is the process that a dissimilarity break, usually canWith be the line of 基 that the software develop basis, the main contents of the maintenance wrap括 Descend say a few aspects. 2.1. the I deletes the dated test to use the exampleBecause demanding change etc. the reason may make a 基 line testUse the example to no longer suit to is test system, these tests would be dated with the example. For example, a certain changed the boundary of the quantity to take place the change, aim at the boundary value originallyOf test can't complete to test to the new boundary.So, in the software of eachA modification should delete the dated test behind use example2 1.2 improvements be free from the test of the control to use the exampleAlong with the progress of the software item, the test uses a database in of with regular meetingIncrease continuously, among them will appear some to the importations or the movement appearances veryThe test of the sensitive use example these test not easy repetition and result hardControl, will affect to return to efficiency of return the test, need to carry on the improvement. make itAttain re-usable and the request that can control. 2.1.3 delete the test of the redundancy to use the exampleIf exist two perhaps more tests to aim at with the example a mutuallyThe input and output that is together carries on the test, so these tests use the example is a 冗Remaining of.The redundancy test lowered to return to efficiency of return the test with the existence of the exampleSo need the periodical sorting test to use a database, and delete redundancy of with the exampleRemove. 2.14 increase the new test to use the exampleIf a certain procedure segment, piece of 构 or keys connect at existingTest medium did not is test, should develop the new test so use example afreshAs to it's carry on the test to merge lately- develop test to line of 基 with the example alsoThe test wrap medium. The maintenance that passes to use a database to the test not only improves the test to use the exampleOf can use sex, and also raised the test database and can believe sex, can be also at the same timeWith keep a 基 line test with the efficiency and effects of a database at aOn the higher Class
马上就要出征海外了,除了恶补一下Spoken English以外,还得把本职的一些工作补补了... ----- 软件测试术语 ----- 按照颗粒度划分 接口测试:Interface Testing 单元测试:Unit Testing 系统测试:System Testing 集成测试:Integration Testing 封闭测试(封测): closed Testing,一般游戏界比较常用,它类软件我喜欢称之为“封版测试” 内部测试(内测):Internal Testing 外部测试(公测): Public Testing 验收测试:Acceptance Testing, 一般由用户&客户进行确认是否可以接受一个产品的验证性测试按照侧重点划分 功能测试:Function Testing 性能测试:Performance Testing 稳定性测试:Stability Testing 压力测试:Stress Testing 安全测试:Security Testing按照方法手段分类 冒烟测试:Smoke Testing 回归测试:Regression Testing 健壮性测试:Robustness Testing 黑盒测试(alpha测试):Black-Box Testing 白盒测试(beta测试):White-Box Testing 灰盒测试(灰度测试):Gray-Box Testing按照特定专项分类 界面测试:UI Testing 容量测试:Volume Testing 保护测试:Protectiion Testing 验收测试:User Acceptance Testing按照效率分类 人工/手动测试:Manual Testing 自动化测试:Automation Testing其它: 测试对象:Test Objectives 测试范围:Test Scope 测试策略:Test Strategy 测试方法:Test Approach 测试过程:Test Procedures 测试环境:Test Environment 测试完成标准:Test Criteria 测试用例:Test Cases 测试进度表:Test Schedule 测试风险:Test Risk 测试覆盖:Test Coverage
Regression testing strategy For a software development project, the project in the test group In the process of the implementation of testing by the development of the test cases will be kept to "The test case", and its maintenance and management. When I get The baseline version of the software, a test version of the baseline for measuring all Trial on the case of a test case for the baseline. In the need for reunification When testing, we can according to the chosen regression testing strategy, The baseline test case to extract suitable test case of regression test Test kits, by running the regression testing to achieve regression testing kits stored in The baseline test case in the test case may be automatic test feet This may also be a test case of manual implementation process Regression testing needs time, funding and manpower to plan, implement and Management. In order to set the budget and to the progress, as efficiently as possible and Effective and regression testing, the need for the conduct of peacekeeping Test Case Guarding based on certain strategies and select the appropriate regression testing kits. 2.1 Maintenance of the test case In order to maximize meet customer needs and applications to Demand software in its life cycle will be changed frequently and continuously push A new version, the revised or new versions of the software will add a Some new features in the software function or have some changes. With Software changes, software functionality and the application interfaces, and software is Now had evolved, and some of the test case may test case Will lose relevance and effectiveness, while others may test case Become obsolete, there are some test cases will be completely unable to run Test Case guarantee the effective Majesty in the test case, the need for testing The use of maintenance. At the same time, be modified or new additions to the software work Can only rely on re-running the test case and not previously revealed The problem, it is necessary additional new test cases to test these New features or characteristics. Therefore, the test case for the maintenance work should be Including the development of new test cases, these new test cases to test The new software features or overwrite an existing test cases can not be covered by the soft Function or characteristics of Test Case maintenance is a different fault of the process, which can be Will be the development of software to baseline as a benchmark, and safeguarding the main contents of packages Includes the following aspects. 2.1. I delete outdated Test Case Because of changes in demand and other reasons might make a baseline test Use Case is no longer suitable for testing systems, and these test cases will be obsolete. For example, the boundaries of a given variable has changed, the original boundary value The test will not be able to complete testing of the new border. Therefore, in every software , Should be revised to delete outdated Test Case 2 1.2 improve uncontrolled Test Case With the progress of software projects, in the test case for the use of meeting Continues to increase and there will be some input or running very Sensitive test case of these tests is not easy and the results are difficult to duplicate Control, it will affect the efficiency of regression testing, and the need for improved. It Achieve repeatable and control requirements. 2.1.3 delete redundant Test Case If there are two or more test cases against a group of With the input and output to test, then the test case is redundant Balance. Redundancy Test Case reduce the existence of the efficiency of the regression testing Therefore, the need for regular finishing the test case, and delete redundant use case Get rid of. 2.1 4 add a new test case If a program, the key component of the interface or the existing Tests have not been tested, then they should re-development of new test cases To test and the development of a new test case to the baseline merger Testing package. Through the test case not only for the maintenance of improved test case Availability, but also enhanced the credibility of the test, but can also To be a baseline test case for maintaining the efficiency and effectiveness of a Higher levels
1、从是否关心内部结构来看
(1)白盒测试:又称为结构测试或逻辑驱动测试,是一种按照程序内部逻辑结构和编码结构,设计测试数据并完成测试的一种测试方法。
(2)黑盒测试:又称为数据驱动测试,把测试对象当做看不见的黑盒,在完全不考虑程序内部结构和处理过程的情况下,测试者仅依据程序功能的需求规范考虑,确定测试用例和推断测试结果的正确性,它是站在使用软件或程序的角度,从输入数据与输出数据的对应关系出发进行的测试。
(3)灰盒测试:是一种综合测试法,它将“黑盒”测试与“白盒”测试结合在一起,是基于程序运行时的外部表现又结合内部逻辑结构来设计用例,执行程序并采集路径执行信息和外部用户接口结果的测试技术。
2、从是否执行代码看
(1)静态测试:指不运行被测程序本身,仅通过分析或检查源程序的语法、结构、过程、接口等来检查程序的正确性。
(2)动态测试:是指通过运行被测程序,检查运行结果与预期结果的差异,并分析运行效率、正确性和健壮性等性能指标。
3、从开发过程级别看
(1)单元测试:又称模块测试,是针对软件设计的最小单位----程序模块或功能模块,进行正确性检验的测试工作。其目的在于检验程序各模块是否存在各种差错,是否能正确地实现了其功能,满足其性能和接口要求。
(2)集成测试:又叫组装测试或联合,是单元测试的多级扩展,是在单元测试的基础上进行的一种有序测试。旨在检验软件单元之间的接口关系,以期望通过测试发现各软件单元接口之间存在的问题,最终把经过测试的单元组成符合设计要求的软件。
(3)系统测试:是为判断系统是否符合要求而对集成的软、硬件系统进行的测试活动、它是将已经集成好的软件系统,作为基于整个计算机系统的一个元素,与计算机硬件、外设、某些支持软件、人员、数据等其他系统元素结合在一起,在实际运行环境下,对计算机系统进行一系列的组装测试和确认测试。
在系统测试中,对于具体的测试类型有:
(1)功能测试:对软件需求规格说明书中的功能需求逐项进行的测试,以验证功能是否满足要求。
(2)性能测试:对软件需求规格说明书的功能需求逐项进行的测试,以验证功能是否满足要求。
(3)接口测试:对软件需求规格说明中的接口需求逐项进行的测试。
(4)人机交互界面测试:对所有人机交互界面提供的操作和显示界面进行的测试,以检验是否满足用户的需求。
(5)强度测试:强制软件运行在异常乃至发生故障的情况下(设计的极限状态到超出极限),验证软件可以运行到何种程序的测试。
(6)余量测试:对软件是否达到规格说明中要求的余量的测试。
(7)安全性测试:检验软件中已存在的安全性、安全保密性措施是否有效的测试,
(8)可靠性测试:在真实的或仿真的环境中,为做出软件可靠性估计而对软件进行的功能(其输入覆盖和环境覆盖一般大于普通的功能测试)
(9)恢复性测试:对有恢复或重置功能的软件的每一类导致恢复或重置的情况,逐一进行的测试。
(10)边界测试:对软件处在边界或端点情况下运行状态的测试。
(11)数据处理测试:对完成专门数据处理功能所进行的测试。
(12)安装性测试:对安装过程是否符合安装规程的测试,以发现安装过程中的错误。
(13)容量测试:检验软件的能力最高能达到什么程度的测试。
(14)互操作性测试:为验证不同软件之间的互操作能力而进行的测试。
(15)敏感性测试:为发现在有效输入类中可能引起某种不稳定性或不正常处理的某些数据的组合而进行的测试。
(16)标准符合性测试:验证软件与相关国家标准或规范(如军用标准、国家标准、行业标准及国际标准)一致性的测试。
(17)兼容性测试:验证软件在规定条件下与若干个实体共同使用或实现数据格式转换时能满足有关要求能力的测试。
(18)中文本地化测试:验证软件在不降低原有能力的条件下,处理中文能力的测试。
4、从执行过程是否需要人工干预来看
(1)手工测试:就是测试人员按照事先为覆盖被测软件需求而编写的测试用例,根据测试大纲中所描述的测试步骤和方法,手工地一个一个地输 入执行,包括与被测软件进行交互(如输入测试数据、记录测试结果等),然后观察测试结果,看被测程序是否存在问题,或在执行过程中是否会有一场发生,属于比较原始但是必须执行的一个步骤。
(2)自动化测试:实际上是将大量的重复性的测试工作交给计算机去完成,通常是使用自动化测试工具来模拟手动测试步骤,执行用某种程序设计语言编写的过程(全自动测试就是指在自动测试过程中,不需要人工干预,由程序自动完成测试的全过程;半自动测试就是指在自动测试过程中,需要手动输入测试用例或选择测试路径,再由自动测试程序按照人工指定的要求完成自动测试)
5、从测试实施组织看
(1)开发测试:开发人员进行的测试
(2)用户测试:用户方进行的测试
(3)第三方测试:有别于开发人员或用户进行的测试,由专业的第三方承担的测试,目的是为了保证测试工作的客观性
6、从测试所处的环境看
(1)阿尔法测试:是由一个用户在开发环境下进行的测试,也可以是公司内部的用户在模拟实际操作环境下进行的测试
(2)贝塔测试:是用户公司组织各方面的典型终端用户在日常工作中实际使用贝塔版本,并要求用户报告
软件测试的内容:
1 得到需求、功能设计、内部设计说书和其他必要的文档
2 得到预算和进度要求
3 确定与项目有关的人员和他们的责任、对报告的要求、所需的标准和过程 ( 例如发行过程、变更过程、等等 )
4 确定应用软件的高风险范围,建立优先级、确定测试所涉及的范围和限制
5 确定测试的步骤和方法 ── 部件、集成、功能、系统、负载、可用性等各种测试
6 确定对测试环境的要求 ( 硬件、软件、通信等 )
7 确定所需的测试用具 (testware) ,包括记录 / 回放工具、覆盖分析、测试跟踪、问题 / 错误跟踪、等等
8 确定对测试的输入数据的要求
9 分配任务和任务负责人,以及所需的劳动力
10 设立大致的时间表、期限、和里程碑
11 确定输入环境的类别、边界值分析、错误类别
12 准备测试计划文件和对计划进行必要的回顾
13 准备白盒测试案例
14 对测试案例进行必要的回顾 / 调查 / 计划
15 准备测试环境和测试用具,得到必需的用户手册 / 参考文件 / 结构指南 / 安装指南,建立测试跟踪过程,建立日志和档案、建立或得到测试输入数据
16 得到并安装软件版本
17 进行测试
18 评估和报告结果
19 跟踪问题 / 错误,并解决它
20 如果有必要,重新进行测试
21 在整个生命周期里维护和修改测试计划、测试案例、测试环境、和测试用具
参考资料:百度百科-软件测试
软件测试的方法有以下几种。
1、按是否查看程序内部结构分为:
(1)黑盒测试(black-box testing):只关心输入和输出的结果。
(2)白盒测试(white-box testing):去研究里面的源代码和程序结构。
2、按是否运行程序分为:
(1)静态测试(static testing):是指不实际运行被测软件,而只是静态地检查程序代码、界面或文档可能存在的错误的过程。
(5)动态测试(dynamic testing),是指实际运行被测程序,输入相应的测试数据,检查输出结果和预期结果是否相符的过程。
3、按阶段划分:
(1)单元测试(unit testing),是指对软件中的最小可测试单元进行检查和验证。(2)集成测试(integration testing),是单元测试的下一阶段,是指将通过测试的单元模块组装成系统或子系统,再进行测试,重点测试不同模块的接口部门。
(3)系统测试(system testing),指的是将整个软件系统看做一个整体进行测试,包括对功能、性能,以及软件所运行的软硬件环境进行测试。(4)验收测试(acceptance testing),指的是在系统测试的后期,以用户测试为主,或有测试人员等质量保障人员共同参与的测试,它也是软件正式交给用户使用的最后一道工序。
4、黑盒测试分为功能测试和性能测试:
(1)功能测试(function testing),是黑盒测试的一方面,它检查实际软件的功能是否符合用户的需求。
(2)性能测试(performance testing),软件性能测试分为:
一般性能测试:指的是让被测系统在正常的软硬件环境下运行,不向其施加任何压力的性能测试。
稳定性测试也叫可靠性测试(reliability testing):是指连续运行被测系统检查系统运行时的稳定程度。
负载测试(load testing):是指让被测系统在其能忍受的压力的极限范围之内连续运行,来测试系统的稳定性。
压力测试(stress testing):是指持续不断的给被测系统增加压力,直到将被测系统压垮为止,用来测试系统所能承受的最大压力。
5、其他测试类型:
回归测试(regression testing)是指对软件的新的版本测试时,重复执行上一个版本测试时的用例。
冒烟测试(smoke testing),是指在对一个新版本进行大规模的测试之前,先验证一下软件的基本功能是否实现,是否具备可测性。
随机测试(random testing),是指测试中所有的输入数据都是随机生成的,其目的是模拟用户的真实操作,并发现一些边缘性的错误。
扩展资料:
软件测试的基本流程:
1、单元测试、集成测试、系统测试和验收测试(确认测试)。
2、测试需求分析,测试计划编写,测试用例编写,测试,缺陷记录,回归测试,判断测试结束,测试报告提交。
3、测试流程依次如下:
a.需求:阅读需求,理解需求,与客户、开发、架构多方交流,深入了解需求。
b.测试计划: 根据需求估算测试所需资源(人力、设备等)、所需时间、功能点划分、如何合理分配安排资源等。
c.用例设计:根据测试计划、任务分配、功能点划分,设计合理的测试用例。
d.执行测试:根据测试用例的详细步骤,执行测试用例。
e.执行结果记录和bug记录:对每个case记录测试的结果,有bug的在测试管理工具中编写bug记录。
f.defect tracking:追踪leader分配给你追踪的bug.直到 bug fixed。
g.测试报告:通过不断测试、追踪,直到被测软件达到测试需求要求,并没有重大bug。
h.用户体验、软件发布等……
参考资料来源:百度百科—软件测试方法