Unit Testing

SOFTWARE ENGINEERING NOTES PART 3 UNIT TESTING Unit testing is a method of software testing in which the smallest parts of the software application (which we call units) are tested. Simply put, "unit testing is such testing in which the program is broken into pieces, and each piece is tested closely." The main purpose of this testing is to ensure that the source code of each unit of the software is correct and can be used. Earlier adhoc tools were used to test units, but nowadays frameworks (Java Framework, .NET Framework and PHP Framework etc.) are used to unit test. Unit tests are written and performed by the developers and the WHITE BOX TESTING method is used to do this. The thing to remember here is that “unit testing is very time consuming and requires a lot of patience.” This testing is very effective because most of the defects are identified by its use. Generally unit testing is done after integration testing. Advantages of unit testing Unit testing has the following advantages:1: - Through this testing, we find defects and bugs in the software in the early stages, later it becomes very difficult to find defects and bugs. 2: - This testing makes the process of coding more effective and agile, so that we can add more and more features to the software. 3: - When we do unit testing, then the need for manual testing is very less. Anyway manual testing is very boring and expensive. 4: - Through this testing, we can improve the design of the software without breaking its design. 5: - When we detect the bugs in advance, it saves our time and cost. 6: - It increases the efficiency of the testing code and it becomes easy to maintain. 7: - One advantage of this testing is that it provides us with the updated documentation of the software. INTEGRATION TESTING Integration testing is a method of software testing in which two or more software units or modules are combined together and tested in a group. The main purpose of this testing is to find bugs and faults between integrated units and to identify work, performance and reliability between units. Normally intergation testing is performed by an integration tester. This testing is done after unit testing and before validation testing. When unit testing is done, each unit is integrated one by one and this process continues until all the units are integrated. INTEGRATION TESTING APPROACH There are many approaches to integration testing, some of them are as follows: 1:-Top-down approach 2:-Bottom-up approach 3:-Big bang approach 1: - Top-down: - In this integration testing, the integrated units of the top level are tested first and then the sub-units below it are tested. 2:-Bottom-up:-In this integration testing, first the sub-units of the bottom level are tested and then the main units above are tested. 3: - Big Bang: - In this type of testing, all the units are integrated together and after that all are tested in a group. In this integration testing, no single unit can be integrated until all the units are ready. This testing is very risky because if we do not do the documentation properly then the risk of failure increases. The advantage of this testing is that it is convenient for small systems. Its disadvantage is that it is very difficult to find faults and bugs in it, even if a fault is found, it becomes very difficult to find out the main reason for the fault and this testing takes a lot of time. .IMPORTANCE OF SOFTWARE TESTING While making the software, some mistake is made by us because we are humans and humans make mistakes continuously. Some mistakes are not harmful, while some mistakes can be very dangerous. To rectify all these mistakes, we have to test the software. So let's know why software testing is needed? 1: - Software testing is done to find all the possible defects and errors in the software. 2:- Software testing is also done to ensure the quality of the software product. If the quality is better then more number of customers will use the software product. 3: - To increase the performance of the software. 4: - To prove that there is no fault in the software. 5: - To increase the reliability of the software. 6:- To ensure that the software is made according to the customer's requirement or not? 7:- Software testing is also done to stay in business. ALPHA AND BETA TESTING Alpha Testing It can be easily understood on the basis of the following points:This testing is a type of acceptance testing. The main purpose of this testing is to detect possible defects and bugs in the software product before releasing it to the customers. This testing is always performed by the developers at the developer site (where the engineers work). This testing is done when the software is ready, if some changes have to be made due to defects after alpha testing, they are done. This testing is completed in two steps, in the first step this testing is performed by the developers while the second step is performed by the software quality assurance (SQA) team. This testing is closed for public means there is no involvement of public in this testing. This testing is always done in virtual environment only. Beta Testing Beta testing can be easily understood on the basis of the following points. This is also a type of acceptance testing and it is also called field testing. This testing is performed by end users in real environment. It is not performed by the testing team. The purpose of beta testing is to detect the defects and problems present in the software based on the perspective of the users and to get the feedback available from the users, so that these problems can be removed in the next updated version of the software. This testing is done in real time environment. This testing is done by the public. Beta testing is done after alpha testing and no other testing is done after beta testing. RECOVERY TESTING Recovery Testing can be easily understood on the basis of following points:1:-Recovery testing is a type of non-functional testing. 2: - This testing is done to ensure that when an application crashes or fails, how quickly and how well it recovers. 3:- To do this testing, the application itself is damaged or crashed and then recovery testing is done in it in every way. Due to which the tester gets to know whether the recovery is being done properly or not. SECUTITY TESTING The functionality of the data and software is secured through security testing. This is a non-functional testing i.e. functional requirements are not required in it. Security testing is a type of software testing in which it is determined whether the system will be able to protect data and functionality from potential intruders or not. The goal of security testing is to identify the threats present in the system and also to detect (mark) all possible risks. Following are the basic security concepts:1:-confidentially 2: integrity 3:-Authentication 4: availability 5:-authorization 6:-non-repudiation Confidentially means that the data or information that is there will not be disclosed to another person or party, that is, only the authorized user will be able to read the information. Integrity means that the information or data that is there can be modified, inserted and deleted only by the authorized user, that is, unauthorized user cannot do any modification. In authentication, the genuineness of the user and the product is confirmed, and it is ensured that the computer program is trustworthy. Availability ensures that the authorized user can access the information and resources at any time. Authorization means that the user will be allowed for any valid and requested operation. Non-repudiation message transimission is guaranteed, that is, when one person sends a message to another and the other person receives the message, later both the persons cannot deny that the message was sent and received. That means later the sender cannot say that I did not send the message and the receiver cannot say that I did not receive the message. WHAT ARE SOFTWARE RISKS AND WHAT ARE THEIR TYPES First of all we will understand what is risk? “RISK is a possible (uncertain) future event in which we may get negative and undesirable results. The probability of which can be from 0 to 100% but it is probability not truth. Software risk are of the following types: - 1:-Management risks 2:-technical risks 3:-financial or budget risks 4:- schedule risks 5:-contractual & legal risks 6:-operational risks 1: - Management risks: - The main reasons for this risk are the following: Failure to plan properly. Low level of management experience and lack of training. Lack of proper communication. Failure to find solution to organization issues. L ack of authority. 2: - Technical risks: - The main reasons for this risk are the following: Continuous changes in needs. Non-availability of new technology. Coding not being in high level language. Software not working properly. 3: - Budget risk: - Money is a very important factor in any project, so there are many risks in it. The main reasons for this risk are as follows:Software not completed within estimated budget. Continuous increase in the cost of resources. Growth of software projects. Delay in software project. 4: - schedule risks: - The main reasons for this risk are the following: Non-completion of the project within the estimated time. Allocation of resources going wrong. Sudden expansion of the project. Not being able to track the resources properly. 5: - Contractual and legal risks: - There are also legal and contractual risks in any project which are as follows: Software not being built within the government rules and regulations. A ny defect in the software within the warranty of the software. Change in software requirements. Safety and health issues etc. 6:-operational risks:- The main reasons for this risk are the following:Process not being implemented properly. Lack of team spirit. Lack of clarity in one's responsibilities means not performing one's responsibilities properly. Lack of practice. Lack of proper communication means lack of communication in the team. 7: - Other risks: - Apart from the above mentioned risks, there are other risks which are as follows: Change in the policy of the government. Inadequate tools. Lack of facilities. Non-availability of computer resources. Non-receipt of contracts or losing the contracts received. WHAT IS EQUIVALANCE PARTITIONING AND BOUNDRY VALUE ANALYSIS EQUIVALANCE PARTITIONING Equivalence partitioning is a software testing technique. In this technique, the input data of the software unit is divided into equivalence classes (portions). Equivalence classes means that similar type of data will be divided into one class and thus other similar data will be divided into different classes on the basis of similarity. In this technique we test only one condition from each class because we know that all the conditions of a class are same. If one condition of the class works, then all the conditions of that class will work. If a test condition of the class does not work then no other condition of the class will work. The advantage of equivalence partitioning technique is that it takes less time to test the software. It can be used on all levels of testing like:- Unit testing, security testing, alpha beta testing etc. For example: - Suppose a customer is given a discount in a shop according to his purchase. No discount is given if the purchase is between ?50 to ?100, 3% discount is given if the purchase is between ?100 to ?200, If the purchase is between ?200 to ?300 8% discount is given, if the purchase is above 300 then 12% discount is given. It is shown below that we have four valid classes and one invalid class. Equivalence partitioning is also called equivalence class partitioning. BOUNDRY VALUE ANALYSIS Boundary value analysis (BVA) is a black box testing technique based on testing the boundaries between classes (partitions). Can be used to find errors in the boundaries of the input data. There are two types of boundary, one lower boundary (from where the range starts) and the other (where the range ends). And each boundary has a valid value and an invalid value. Test cases are designed on the basis of both these valid and invalid values and we usually choose one test case from each boundary. For example:- Suppose there is a software in the college which works only on valid username and password and it accepts only minimum 6 and maximum 10 characters. valid range 5-10, invalid range 5 or less, and invalid range 11 or more than 11. BVA is also called range checking. Equivalence partitioning and boundary value analysis are related to each other and they are used together at all levels of testing. GREY BOX TESTING Gray box testing is a software testing technique which is a combination of white box and black box testing. In black box testing, the tester does not have knowledge of the internal structure of the software, and in white box testing, the tester has knowledge of the internal structure of the software, while in gray box testing, the tester has knowledge of both internal and external functions. The name of this testing is gray because in this the software for the tester is like a semi-transparent box in which he can see partially inside it. The advantage of gray box testing is that we can test the external part and internal part (coding) of the software through it. It is used in integration testing. It is mostly used to test software applications based on the Internet. This testing is also called translucent testing. For example:- Suppose the tester tests a website, if he finds broken links and other inappropriate links in it, then he can fix them by going to the HTML part of the website and check them at the same time.