Academia.eduAcademia.edu

Figure 5 – uploaded by Yixiang Chen

See full PDFdownloadDownload figure
Figure5 shows how to compress an unstructured activity diagram and transform the unstructured module into a black box node. Eventually the unstructured activity diagram converts into an intermediate representation of IBM. The compressed black box node is shown in the following Figure6(b).

Figure5 5 shows how to compress an unstructured activity diagram and transform the unstructured module into a black box node. Eventually the unstructured activity diagram converts into an intermediate representation of IBM. The compressed black box node is shown in the following Figure6(b).

Related Figures (15)

The correctness of the path with respect to the pre-condition and the functional scenario is unlikely to be established by means of testing. This deficiency can be eliminated by repeatedly applying the axiom for assignment in Hoare logic. So, we need to form path triple and the definition are below: International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018 The specification states that the input mark must be in the range from 100 to 0 and the output pas: is a bool type. According to the definition proposed before, three functional scenarios can be derived from this formal specification, as shown below.
The correctness of the path with respect to the pre-condition and the functional scenario is unlikely to be established by means of testing. This deficiency can be eliminated by repeatedly applying the axiom for assignment in Hoare logic. So, we need to form path triple and the definition are below: International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018 The specification states that the input mark must be in the range from 100 to 0 and the output pas: is a bool type. According to the definition proposed before, three functional scenarios can be derived from this formal specification, as shown below.
Figure 1. TBFV-M usage scenario Model-Based Systems Engineering (MBSE) combines process and analysis with architecture. In the last decade, the model-driven approach for software development has gained a growing interest of both industry and research communities as it promises easy automation and reduced time to market [23]. Because of the graphical notation for defining system design as nodes and edge diagrams, SysML model addresses the ease of adoption amongst engineers [24].
Figure 1. TBFV-M usage scenario Model-Based Systems Engineering (MBSE) combines process and analysis with architecture. In the last decade, the model-driven approach for software development has gained a growing interest of both industry and research communities as it promises easy automation and reduced time to market [23]. Because of the graphical notation for defining system design as nodes and edge diagrams, SysML model addresses the ease of adoption amongst engineers [24].
The TBFV-M method takes the specification describing the users’ requirements and the SysML Activity Diagram model as input and verifies the correctness of the SysML model according to the specification. The procedure of TBFV-M is illustrated in Figure2
The TBFV-M method takes the specification describing the users’ requirements and the SysML Activity Diagram model as input and verifies the correctness of the SysML model according to the specification. The procedure of TBFV-M is illustrated in Figure2
International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018 First, we use transformation algorithm to compress the input Activity Diagram, which may contain unstructured module. The transformation is a cyclic process, dealing with loop module, concurrent module and the problem of multiple starting nodes separately. After compressing, we transform this unstructured activity diagram into an intermediate representation form Intermediate Black box Model (IBM). IBM consists of one basic module and a map from black box to the corresponding original actions. The third phase is the test path generation based on IBM. In this phase, two problems should be solved, which are basic module test path generation and black box test path generation. Details of automated test paths generation algorithm and implementation of unstructured SysML Activity Diagram has been developed in our previous work [30].
International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018 First, we use transformation algorithm to compress the input Activity Diagram, which may contain unstructured module. The transformation is a cyclic process, dealing with loop module, concurrent module and the problem of multiple starting nodes separately. After compressing, we transform this unstructured activity diagram into an intermediate representation form Intermediate Black box Model (IBM). IBM consists of one basic module and a map from black box to the corresponding original actions. The third phase is the test path generation based on IBM. In this phase, two problems should be solved, which are basic module test path generation and black box test path generation. Details of automated test paths generation algorithm and implementation of unstructured SysML Activity Diagram has been developed in our previous work [30].
International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018 For loop module, like the Figure6(a), since the infinite traversal loop is not possible, it is possible to propose a different expansion algorithm for different types of loops when processing the loop module. For simple loop, you can take the following test case sets (where n is the maximum number of passes allowed):
International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018 For loop module, like the Figure6(a), since the infinite traversal loop is not possible, it is possible to propose a different expansion algorithm for different types of loops when processing the loop module. For simple loop, you can take the following test case sets (where n is the maximum number of passes allowed):
International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018 For concurrent module, if it is a simple one, this is the classic concurrency module, which can be used fully arranged algorithm or other optimization algorithm for automatic test case generation. The concurrent module in figure6(b) is a No_FJ, which represents that there is no join node at the end of concurrent module. The following proposed an algorithm for generating the noJoin concurrent module test case:
International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018 For concurrent module, if it is a simple one, this is the classic concurrency module, which can be used fully arranged algorithm or other optimization algorithm for automatic test case generation. The concurrent module in figure6(b) is a No_FJ, which represents that there is no join node at the end of concurrent module. The following proposed an algorithm for generating the noJoin concurrent module test case:
International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018 According to the specification, we can construct a set of SysML model and the Activity Diagram is shown below. To show the ability of detecting errors, we deliberately make a mistake in this diagram. According to the specification, when the input price is over 100, we will give 20%-discount. However, in the below figure, we miswrite the final price with subtracting $10. And as the principle of TBFV-M, the proof based on Hoare Logic will help to discover the mistake, which will be presented later.
International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018 According to the specification, we can construct a set of SysML model and the Activity Diagram is shown below. To show the ability of detecting errors, we deliberately make a mistake in this diagram. According to the specification, when the input price is over 100, we will give 20%-discount. However, in the below figure, we miswrite the final price with subtracting $10. And as the principle of TBFV-M, the proof based on Hoare Logic will help to discover the mistake, which will be presented later.
First, we derive Functional Scenarios from specification and generate test paths from Activity Diagram. The result is shown as below.
First, we derive Functional Scenarios from specification and generate test paths from Activity Diagram. The result is shown as below.
International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018 At the same time, we can extract data constraints from each test scenario, which is used fo matching with functional scenario. Then, the matching process is shown below. The blow figure chose the second path and matched the first functional scenario as an example and shows the substitution process, from bottom to up. The second path is also the wrong path, which exists inaccurate with requirement.
International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018 At the same time, we can extract data constraints from each test scenario, which is used fo matching with functional scenario. Then, the matching process is shown below. The blow figure chose the second path and matched the first functional scenario as an example and shows the substitution process, from bottom to up. The second path is also the wrong path, which exists inaccurate with requirement.
Finally, we turn this verification problem into proving whether the pre-condition of specification can imply Ppre. If it can be proved, means that the path satisfies the requirement. From the above segment, we can see the implication is false. As the strategy of implication mentioned before, this implication uses simple data structure, so that we use testing to access the procedure of verification. We can simply generate a test data (price,150), and because of existing the counterexample, the implication can not be always true. And we can also be aware that this path exists errors and need to be modified.
Finally, we turn this verification problem into proving whether the pre-condition of specification can imply Ppre. If it can be proved, means that the path satisfies the requirement. From the above segment, we can see the implication is false. As the strategy of implication mentioned before, this implication uses simple data structure, so that we use testing to access the procedure of verification. We can simply generate a test data (price,150), and because of existing the counterexample, the implication can not be always true. And we can also be aware that this path exists errors and need to be modified.
And the result is that the supporting tool has the ability to figure out these mistakes, as out expectation. Table 2. Evaluation Result
And the result is that the supporting tool has the ability to figure out these mistakes, as out expectation. Table 2. Evaluation Result
International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018
International Journal of Software Engineering & Applications (IJSEA), Vol.9, No.4, July 2018
Related topics:
EngineeringComputer ScienceSoftware EngineeringSoftware Engineering educationActivity TheoryEmpirical Software EngineeringSoftware Quality EngineeringSoftware Project ManagementSoftwareSoftware Requirement Engineering
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved