1 / 28

IPC144 Session 5 Problem Analysis and Program Design Continued

IPC144 Session 5 Problem Analysis and Program Design Continued. 1. Objectives Design good test data Synthesize a Test Plan from test data Apply a rigorous approach to testing flowcharts using Walkthroughs. 2. Test Data. Test Data From the PDC: PDC: Test the Algorithm for Correctness

ailani
Télécharger la présentation

IPC144 Session 5 Problem Analysis and Program Design Continued

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. IPC144 Session 5 Problem Analysis and Program Design Continued 1

  2. Objectives • Design good test data • Synthesize a Test Plan from test data • Apply a rigorous approach to testing flowcharts using Walkthroughs 2

  3. Test Data Test Data From the PDC: PDC: Test the Algorithm for Correctness One of the most important steps Identifies major logic errors Walkthroughs are a technique for testing Test plans are used to document and plan the tests necessary to prove your algorithm performs as required 3

  4. Test Data, continued Test Data, continued There could exist an infinite number of possible values that a program can accept as input. For example, write a program that accepts three numbers, adds them together and displays their total and average. There are an infinite number of possible input values- does this mean that we must try all possible values to ensure that the program works correctly? There must be a reasonable set of test data we can use to prove the algorithm works empirically. 4

  5. Test Data, continued • Test Data, continued • There are two general types of testing: • Black Box Testing • White Box testing 5

  6. Test Data, continued Black box testing The system, or program, is tested without knowing what is inside it Tests are developed based solely on the requirements statement - based on the inputs, what are the expected outputs Used to test the system or a multi-module program as a whole Can be used to help clarify the requirements statement as you find test data that does not appear to be accounted for in the requirements 6

  7. Test Data, continued White box testing The algorithm is tested knowing what is inside it It can only be developed by a programmer (typically the programmer that created the software)‏ Everything about the program is known- therefore everything is tested for You must test every possible program instruction and every possible situation. 7

  8. Test Data, continued White box testing, continued Start identifying test data during the Design Phase As you build your flowchart & pseudocode look for test cases, document them as you find them Look for unusual conditions / situations to be added to your testing As you find conditions that are not handled by your algorithm- enhance your algorithm and add the test to your Test Plan. What is good test data for both Black Box Testing and White Box Testing? What do we look for? 8

  9. Test Data, continued • Test Data • Black Box • Look at requirements • Get test data and expected results from the user. • What if user ends program as their first action? • Unexpected values (user types words instead of a number)‏ • White Box • Look at the Program and how it was written • Define data that will result in all branches of all decisions being tested • Ensure that ranges are tested for their minimum and maximum values • Look for ways that a divide by zero may happen • Simulate all possible error conditions (even if you have to alter your program to ensure they occur)‏ 9

  10. Test Plans • Test Plans • Why all the trouble? • User groups and Audit departments look for them • Provides a concise description of WHAT was tested • Provides a concise description of HOW it was tested • Provides a concise description of the ENVIRONMENT it was tested in • A well design and documented test plan can break the 'finger pointing' 10

  11. Test Plans, continued • Test Plans, continued • Provides a concise description of WHAT was tested • At some time software comes under scrutiny: • The system has failed - was this scenario tested? • Somebody (maybe you) months later asks the question was ______ tested? 11

  12. Test Plans, continued Test Plans, continued Provides a concise description of HOW it was tested What were the test cases? What data was used? Was certain sequence of events tested for? Again, a system failure may precipitate the question: Was this sequence of events tested for? 12

  13. Test Plans, continued Test Plans, continued Provides a concise description of the ENVIRONMENT is was tested in What other programs where part of the package? What was the version of these other programs? What hardware was it tested on? What was the version of the operating system? It "NEVER" happens that after you have certified a system somebody changes a component and doesn't tell you! 13

  14. Test Plans, continued • Test Plans, continued • An example of the parts of a Test Plan: • Cover Page • Environment setup • Black Box Testing • White Box Testing (can include “negative tests” to ensure the program does not crash or have undesired side-effects) 14

  15. Test Plans, continued Test Plans - Cover Page A brief description of what the program or system does. Not too technical - just enough that somebody non-technical can understand what the program does An extract of the requirements statement may form part of the cover page 15

  16. Test Plans, continued Test Plans - Environmental Setup Optional – depending on nature of your work environment More technical description What hardware is being used (as specific as necessary)‏ What version of the operating system is being used What is the version of YOUR program that is being tested The files that are used to makeup your program What are the versions of the other programs that interface with yours Are there versions of test data files? 16

  17. Test Plans, continued Test Plans - Black Box Testing These tests are to ensure that the program meets the requirements statement Predominately positive testing is performed Hopefully the Business Analyst has provided test cases that can be incorporated into the Test Plan Concentrate on obtaining the correct outputs for the test inputs Each test has two parts: the test, the expected results 17

  18. Test Plans, continued Test Plans - White Box Testing These tests are to ensure that the program handles all expected and unexpected data Every path within the program must (at least in this course) be tested. Can become very complex with the inter-relationships between variables One philosophy behind White Box Testing is that "it would be nice that whenever a program prompts a user for a number, the users always enter a legitimate value that the program can handle" May require modifications to the code to simulate error conditions (be sure to perform these modifications and tests on a COPY of the program)‏ Each test has two parts: the test, the expected results 18

  19. Test Plans, continued Test Plans - Negative Testing The intent is to maliciously attack your code - to see where (if) is breaks Can be an easy task or frustrating task depending on how well constructed the code is You will normally have built into your code a means to detect error conditions The goal of your program is to shut down gracefully (return an error code or error message) as opposed to crashing or affecting other programs or systems 19

  20. Test Plans, continued Example Write a program to help a local Elementary School's library with the overdue fines on books. The librarians don't wish to always charge students for overdue books, however sometimes the number of books that are overdue is too much. When this happens the librarians enter into the program the student's name, the name of the overdue book they borrowed, and the date it was signed out (the librarian reads these fields from the sign-out card). The colour of the card is also entered, as this indicates what the normal sign-out period is for that book. White cards are due back in 14 days, Blue cards in 7 days, Green cards the next day, and Orange cards in 28 days. The colour code of the card is entered as a single character into the program (W, B, G and O). When this information has been entered, the program determines by how much the book is overdue (based on the date the program is run) and the overdue charge, ($0.05 per day). A list is produced that will show the student's name, the name of the book, the number of days the book is overdue and the fine that is payable. 20

  21. Test Plans, continued Library Fines Test Plan The program is a tool that will allow librarians to assess fines on overdue library books. The fine is calculated as $0.05 / day. The fines are based on the colour code of the sign-out cards, where: Orange Cards - 28 day sign-out period White Cards - 14 day sign-out period Blue Cards - 7 day sign-out period Green Cards - 1 day sign-out period The librarian is responsible for entering the student name, book name and the sign-out date. 21

  22. Test Plans, continued Environmental Setup Program and Version: library.c Version 1.0 Hardware: IBM RS/6000 O/S: AIX 4.3 Compiler: cc Other files: no other files are required 22

  23. Test Plans, continued Black Box Testing 1) Enter 'Stu1', 'Book1', 'O', today's date - 30 Expected result: overdue = 2, fine = $0.10 2) Enter 'Stu2', 'Book2', 'W', today's date - 30 Expected result: overdue = 16, fine = $0.80 3) Enter 'Stu3', 'Book3', 'B', today's date - 30 Expected result: overdue = 23, fine = $1.15 4) Enter 'Stu4', 'Book4', 'G', today's date - 30 Expected result: overdue = 29, fine = $1.45 5) Enter 'Stu5', 'Book5', 'G', today's date Expected result: overdue = 0, fine = $0.00 23

  24. Test Plans, continued White Box Testing 1) Enter a colour code other than 'O', 'W', 'B', 'G'. Expected Result: error message “Invalid card colour code entered”. Program to prompt for the colour code again. 2) Enter a future date (i.e. the book is not yet overdue) Expected Result: error message “Invalid Future date entered”. Program to prompt for the date again. 24

  25. Walkthroughs Walkthroughs Walkthroughs are a means to test the algorithm BEFORE it has been coded Walkthroughs are a tool to test that your algorithm behaves correctly Because the computer is a mindless automaton - in order to check the program YOU must become the mindless automaton! You 'play computer' - turn OFF your mind and follow the instructions EXACTLY as written Remember to follow the instructions and perform what is asked for, not what you want to happen 25

  26. Walkthroughs, continued Walkthroughs, continued It is hard to keep track of the many variables in a program, so take a sheet a paper: List the variables, and their values. Keep track of the new values of a program as you execute the instructions For output, reserve a space on your paper to track any information that is displayed to the user. This can be very tough, as it is very easy at this stage to execute the instructions as you WANT them to be executed. When we talk about pseudo code, walkthroughs can be used to test the INTEGRITY of the code after it has been translated from flowcharts to pseudo code. 26

  27. Walkthroughs, continued start num = 1 var = 0 var = num+ var DISPLAY var num = num + 2 num <= 7 stop Walkthroughs, continued TRUE FALSE 27

  28. Walkthroughs, continued i > 0 DISPLAY f DISPLAY “Done” j > 1 Start Walkthroughs, continued i = 5 FALSE TRUE j = i f = 1 FALSE TRUE f = f * j j = j - 1 i = i - 1 Stop 28

More Related