1. Using Use Case Scenarios and Operational Variables for Generating Test Objectives Javier J. Gutiérrez María José Escalona Manuel Mejías Arturo H. Torres Department of Computer Languages and Systems. University of Seville, Spain. escalona@lsi.us.es STV 2007. Paris, France

2. Introduction Context of the work Introduction Use cases & test cases generation. The approach Conclusions. Use cases are a testing artefact Use cases is a wide used technique to define functional requirements. The test phase must assure that the final software system covers these functional requirements. There are two very used techniques for extracting test cases from use cases: Scenario analysis, based on scenario identification. Test data analysis, based on category partition method

3. Introduction Objectives of the paper Introduction Use cases & test cases generation. The approach Conclusions. Objectives: Applying Cathegry Partition Method (CPM) and Extended Use Case Test Patterns (both very similar) to obtain test objectives. Test objectives are defined as combinations of the partitions of the operational variables (categories) identified.

4. Introduction Definitions Introduction Use cases & test cases generation. The approach Conclusions. Operational variable: any element that may vary between two use case scenarios ( categories of an use case ). Domain: all possible values that a variable operational can get. Partition: set of domain values which exposes the same behaviour in the use case.

5. Use cases is a technique to express requirements graphically. Graphics are not enough detailed to express all the funcional requirements Some alternatives or representation are defined. For instance, patterns. Use cases & test cases generation Definition of use cases (I) Introduction Use cases & test cases generation. The approach Conclusions.

6. Use cases & test cases generation Definition of use cases (II) NDT + language patterns. (The) [ACTOR] [ACTION] Introduction Use cases & test cases generation. The approach Conclusions.

7. Use cases & test cases generation Test objectives Test objective : Something to test ( with a test case ) A test objective is a test case without a result. Test objectives may be defined in serveral ways. A use case may be a test objective. Use case scenarions. Test value combination. Both 2 and 3. Introduction Use cases & test cases generation. The approach Conclusions.

8. The approach Generating use cases scenarios. STEP 1 STEP 1: From use cases definition to XML definition Use cases templates are extended with language patterns and they are translate into a XML document using the same structure than the template Introduction Use cases & test cases generation. The approach Conclusions. A use case searches for a set of links by their description and shows the results. The visitor asks the system searching for links by description. The system asks for the description. The visitor introduces de description. The system searches for the links which match up with the description introduced by the visitor. The system shows the found results. At any time, the visitor may cancel the search, then the use case ends. If the visitor introduces an empty description, then the system searches for all the stored links and step 5 is repeated. If the system finds any error performing the search, then an error message is shown and this use case ends. If the result is empty, then the system shows a message and this use case ends. DTD available

9. The approach Generating use cases scenarios. STEP 2 STEP 2: From formal pattern in XML to activity diagrams From formal description of pattern an activity diagram can be generated easily. (tool available: ObjectGen) Introduction Use cases & test cases generation. The approach Conclusions. If [CONDITION] then [ACTION] and this use case ends.

10. The approach Generating use cases scenarios. STEP 3 STEP 3: From activity diagram to use case scenarios Each path of the activity diagram is a possible use case scenario. Classical graph techniques may be used to generate scenarios. (tool available ObjectGen) Introduction Use cases & test cases generation. The approach Conclusions. Use case: Search link by description The Round-Trip Criterion Scenarios: 5 S 1: 1, 2, D01, End. S 2: 1, 2, D01, 3, D02, 4.1, 5, End. S 3: 1, 2, D01, 3, D02, 4, D03, 4.2, End. S 4: 1, 2, D01, 3, D02, 4, D03, D04, 4.3, End. S 5: 1, 2, D01, 3, D02, 4, D03, D04, 5, End. Use case scenario 1: 1: The visitor asks the system searching for links by description. 2: The system asks for the description. 3: The visitor cancels the search then the use case ends.

11. The approach Generating combinations of partitions. STEP 1 (I) STEP 1: Definition of operational variables and partitions Every decision node, in the activity diagram, evaluates a logical predicate to choose an output transition. Therefore, an operational variable is assigned to each decision node and logical predicate. The domain of the operational variable is divided into as many partitions as output transitions Introduction Use cases & test cases generation. The approach Conclusions.

12. The approach Generating combinations of partitions. STEP 1 (II) Introduction Use cases & test cases generation. The approach Conclusions. Variables: activity and decision nodes. Particiones: Decisions output flows.

13. The approach Generating combinations of partitions. STEP 2 (I) Introduction Use cases & test cases generation. The approach Conclusions. STEP 2: Construction of the dependency matrix and generation constraint 1 -> 2 -> 3 -> V_D01(Sí) One path with variables and the partitions in which variables takes values. We see, in dependency map, that from V_D01, variables V_D02, V_D03 and V_D04 may be evaluated. But those variables do not appear in the path, so, we have identifiyed a restriction. If V_D01 == P_1 ( cancels ) then V_D02 = V_D03 = V_D04 = no value V_D01: User cancels peration. Boolean V_D01V_D02V_D03V_D04 V_D010 111 V_D0200 11

14. The approach Generating combinations of partitions. STEP 2 (II) STEP 2: Construction of the dependency matrix and generation constraint These steps can be automatically generate with an available tool: ValueGen Introduction Use cases & test cases generation. The approach Conclusions.

15. The approach Results with the constraint construction Introduction Use cases & test cases generation. The approach Conclusions. Without constraints: 4 variables, 2 partitions for each variable: 16 combinations. With constraints: 4 variables, 2 partitions, 2 constraints: 7 combinations.

16. Conclusions General conclusions The approach introduced may be applied in the early stages of the development process, when use cases are being defined, because it does not need a set of complete use cases The full automatism of the process and the availability to support tools allow the easy re-generation of test objectives when use cases change The approach only uses main, alternative and erroneous steps which are present in all styles, so, it may be easily used without having to change the use case style and format of an organisation It allows the managing of the traceability of the test cases from a use case through their test objectives. (To know how many use cases were covered by tests) Introduction Use cases & test cases generation. The approach Conclusions.

17. Conclusions Future works To improve the tool support: mainly the interface. To integrate test tools with the use case supporting tool for NDT. To apply in practise with NDT. Introduction Use cases & test cases generation. The approach Conclusions.

18. Address Introduction Use cases & test cases generation. The approach Conclusions. Testing & Tools Javier Gutiérrez (javierj@us.es) www.lsi.us.es/~javierj/ Requirements María José Escalona (mjescalona@us.es) www.lsi.us.es/~escalona/

19. Using Use Case Scenarios and Operational Variables for Generating Test Objectives Javier J. Gutiérrez María José Escalona Manuel Mejías Arturo H. Torres Department of Computer Languages and Systems. University of Seville, Spain. escalona@lsi.us.es STV 2007. Paris, France