JPH06149561A - Software developing device - Google Patents

Abstract

PURPOSE:To automate test data preparation in a test process and to improve productivity in the case of software development. CONSTITUTION:This device is provided with a specification description part 1 for applying a software specification described in a state transition model containing the concept of time as well, test case description part 2 for generating a test case describing the sequence of transitions in this specification, time sequential data generation part 3 equipped with an extracting function for extracting applied from the outside and time limit existent between these events concerning the software of a test object by using the specification of a software depending on the state transitions provided from the specification description part 1 for this provided test case and a generating function for generating time sequential data satisfying the time limit corresponding to the events extracted by this extracting function, and test execution part 4 for executing the software of the test object by using these generated time sequential data.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a software development device using a computer, and more particularly to a software development device capable of efficiently performing a system test process of control software.

[0002]

2. Description of the Related Art Software under development is tested in order to confirm whether the programming content operates according to specifications. This process is the test process, but in the system test process in software development using a computer, the test manager has conventionally created test cases (various states and conditions assumed for testing) based on the software specifications. Then, the test data (simulated information for testing) corresponding to the test case is created and given to the software under test to be executed, thereby performing the operation test.

At this time, in the case of control software, the test data actually given to the test target software is
It is time-series signal data, and in this case, it was a difficult task to create the signal data corresponding to the target test case.

Further, in the conventional system test data generation method, it is possible to generate input data in a form that covers the operation of the system from the system specifications and the control structure of the program, but the target is controlled. In the case of software for software, time-series data must be generated taking into consideration not only simple combinations of inputs but also how those inputs change over time. ,
You can't expect more than a unit test.

[0005]

In the system testing process, which is performed to confirm whether the software under development operates according to the specifications, conventionally, the person in charge of testing has a test case based on the specifications of the software. I created the test data corresponding to the test case, gave it to the software to be tested, and executed it.

However, if the object is control software, the test data is time-series signal data. Therefore, creating actual test data from test cases in the test data creation stage is not only labor-intensive, but also considers how the input changes over time and time series. , But it is extremely difficult to create it manually.

In order to ensure the reliability of software, it is necessary to confirm what behavior behaves in all actual situations and whether it conforms to the specifications. However, in the case where the change over time has to be taken into consideration, it is not easy to create such test data, and it is difficult to obtain sufficient reliability.

At the present stage, it is believed that the expected test data corresponding to the test case is created, the test is repeated, the bug is found from the test result, the bug is corrected, and the bug is resolved by testing again. I was going to ship it.

Therefore, the object of the present invention is to
It is possible to automate the test data creation work in the test process, and it is also possible to automatically create the test data of the time-series signal corresponding to the test case at a timing that matches the actual state of control. It is an object of the present invention to provide a software development device capable of improving the reliability and the reliability of a software product.

[0010]

In order to achieve the above object, the present invention is configured as follows. That is, there is a specification description means for giving a software specification described by a state transition model that also includes the concept of time, and based on the software specification given by this specification description means, for a certain event in a certain state in this specification, For the test case description means that executes the action and generates the test case that describes the sequence of transitions to a certain state, and the test case obtained from this test case description means,
An extraction function that extracts externally applied events and time constraints existing between the events from the software to be tested using the software specifications based on the state transitions obtained from the specification description means. Time series data generation means having a generation function for generating time series data satisfying the time constraint corresponding to the event, and a test for executing the software to be tested using the time series data from the time series data generation means And an execution unit.

[0011]

In the present apparatus having such a configuration, in the test process in software development, the specification description means gives the description of the software specification by the state transition model including the concept of time, and the test case description means From, a test case described in a sequence of transitions in this described specification (in a certain state, a certain event is executed for a certain event to reach a certain state) is given. Then, the time-series data generation means generates time-series data that is actually used for testing the system from this test case, and the test execution means executes the system test using the time-series data. The time-series data generation processing by the time-series data generation means uses the software specifications based on the state transitions obtained from the specification description means for the test case obtained from the test case description means to externally test the system under test. Extract the events you give and the time constraints that exist between them,
This is a process of generating time-series data that satisfies the time constraint and corresponds to the extracted event. By performing a test using the generated time-series data, it is possible to perform a system test in the actual operating condition of the software product.

Once the time of occurrence of the transition has been determined, the corresponding event causing the transition can be associated with that time of occurrence from the software specifications. Then, when the event is decided, what kind of test data is needed is decided, so if the test data decided by the corresponding event that causes the transition is generated at the occurrence time that causes the transition, it is necessary at the necessary timing. It becomes possible to generate various test data, and it becomes possible to automatically generate time-series data corresponding to the actual condition of control.

According to the present invention, it is possible to automatically create test data of a time series signal corresponding to a test case in a form that matches the actual condition of control, and it is possible to automate the optimum test data creation work in the test process. It not only saves a great deal of effort for testing and improves productivity in software development, but also makes it possible to perform sufficient testing due to the ease of testing work, thereby improving the reliability of software products. realizable.

[0014]

Embodiments of the present invention will now be described with reference to the drawings. FIG. 1 is a block diagram showing the system configuration of the present invention. Reference numeral 1 is a specification description unit, 2 is a test case description unit, 3 is a time series data generation unit, and 4 is a test execution unit.

Of these, the specification description unit 1 supplies the software specifications described by the state transition model including the concept of time, and the test case description unit 2 describes the specification description unit 1. A test case is described by a sequence of transitions in a specified specification (in a certain state, a certain event is executed for a certain event to reach a certain state). The time-series data generation unit 3 is for generating time-series data that is actually used for testing the system from the test case described by the test case description unit 2, and the test execution unit 4 is for the time-series data generation unit 3. The system test is automatically executed using the generated time series data.

In such a configuration, the software specification described based on the state transition model is input from the specification description unit 1. The specification description unit 1 may be a file storage device such as a disk device, or a console which is a man-machine interface equipped with a keyboard. In the case of a console, the operator inputs the software specification description by key input to give it to the time-series data generation unit 3. In the case of a file storage device, the software specification description stored in advance as a file is read out and read out. It is given to the series data generation unit 3.

A test case is given from the test case description unit 2 to the time series data generation unit 3. Test cases are given as a sequence of transitions on the specification due to state transitions. When these are given, the time-series data generation unit 3 extracts the time constraint regarding the occurrence time of those transitions from the sequence of transitions that is the given test case, and transitions so as to satisfy this time constraint. After determining the occurrence time, the data corresponding to each event that triggers each transition is generated as time-series test data (see FIG. 2). Then, the time-series test data generated by the time-series data generation unit 3 is given to the test execution unit 4, and the test execution unit 4 executes software using these time-series test data to perform a test.

Next, the test data generation process, which is an important part of the present invention, will be described. FIG. 2 shows a test case description unit 2 and a time-series data generation unit 3 that are the parts that implement this.
2 is a system configuration diagram of FIG. In the figure, reference numeral 21 is a test case creation unit that constitutes the test case description unit 2, and specifies the operation corresponding to the test case on the software specification based on the state transition model. Reference numeral 22 is a test case created by the test case creation unit 21. The software specification is obtained from the state transition software specification 24 given by the specification description unit 1. The time series data generation unit 3 includes a time constraint extraction unit 23, a transition time determination unit 25, and a time series signal generation unit 26.

The time constraint extraction unit 23 uses the test case 22.
And the state transition software specification 24, the time constraint regarding the time when the transition occurs is extracted. The transition time determination unit 25 determines the transition time so as to satisfy the constraint determined by the transition time determination unit 25. Also,
The time-series signal generation unit 26 generates an event associated with the transition as a time-series signal from the outside, based on the transition time determined by the transition time determination unit 25.

In such a configuration, first, the test case creation unit 21 specifies the operation corresponding to the test case on the software specification based on the state transition model. The designation of the operation will be expressed as a sequence of transitions on the state transition software specification 24.

As a method of designating the test case 22, a person (operator) may designate a transition sequence on the transition diagram, but it is also possible to automatically generate it according to a certain algorithm. .

The test cases, that is, the transition sequences generated in this way are 22. The time constraint extraction unit 23
From the test case 22 and the state transition software specification 24, the time constraint regarding the time when the transition occurs is extracted. Furthermore, the transition time determination unit 25 determines the transition time so as to satisfy the constraint. Finally, the time series signal generation unit 26 generates an event associated with the transition as a time series signal from the outside based on the determined transition time. Next, the time constraint extraction processing by the time constraint extraction unit 23 and the transition time determination unit 2
The determination processing of the transition time by 5 will be described. First,
The processing flow is shown in FIG. The time constraint extraction processing by the time constraint extraction unit 23 is performed by the time axis transition placement processing 310 and the time constraint extraction processing 311.

In the time axis transition arrangement processing 310, the transition sequence given by the test case is arranged on the time axis. The arrangement of transitions on the time axis means that the transition sequence given by the test case is arranged as a point sequence on the time axis. At this time, the time of transition on the time axis is undecided. The arrangement of the transitions on the time axis is as shown in FIG. 4, and this is an example in which each transition is arranged on the axis.

Next, there is a time constraint extraction process 311.
Specifically, the section of the transition in which the timer is operating on the time axis (called the time section) is extracted from the combination of events and actions related to the timer associated with the transition, and further, the section of the section from the set time of the timer is extracted. Find time constraints. This state is shown in FIG. As shown in Figure 5, "start, wait", "continue, timeout" of the timer
By combining the above, two time intervals are extracted, and the time constraint is obtained from the set time of the timer.

The transition time determination processing by the transition time determination unit 25 is realized by the time interval normalization processing 312 and the transition time determination processing 313. In the time interval normalization process 312, the redundant part in the time constraint obtained by the time constraint extraction process 311 is eliminated so that no contradiction occurs in the transition time determination process 313. Specifically, it corresponds to canceling the inclusion relation of time intervals as shown in FIG.

Since the transition time is determined in the transition time determination process 313, the occurrence time of each transition is actually set so as to satisfy the time constraint given by the time constraint extraction process 311 and the time interval normalization process 312. decide.

In the present embodiment, one of the following two methods is used as the method of determining the occurrence time. (1) ・ Method based on knowledge about products. (2) ・ A method to allocate time constraints.

Among these, in (1) “method based on knowledge about product”, when no time constraint is given when determining the time of a certain transition, from the knowledge about the product, the transition from a certain state Get how long the event to trigger occurs and determine the transition time from that time.

(2) In the "method of allocating time constraints", until an event occurs, which is obtained from the number of transitions existing in the time interval, the constraint time given to the time interval, and the knowledge about the product. The time until the transition occurs is distributed and the transition time is determined based on the three elements of the time normally required for the transition.

FIG. 7 shows how to determine the transition time. In this example, the transition time is determined by dividing the constraint time T given to the time interval by the ratio of the time (state retention time) until the event that causes each transition normally occurs. Further, globally, the time determination as a whole is advanced by propagating the time determination based on the constraint of each timer. FIG. 8 shows a state of global time determination by propagation of time determination.

By determining the transition time as described above, the generation time of the event that causes the transition is determined, and a series of operations for giving an input of the system corresponding to the event to the time is a time series. Signal test data can be generated.

As described above, in the present system, the software specification described based on the state transition model,
Using a test case given as a sequence of transitions on the specification due to state transitions, the temporal constraint on the occurrence time of those transitions is extracted for the sequence of transitions that is the given test case, and this temporal After the occurrence time of the transition is determined so as to satisfy the constraint, the data corresponding to each event that triggers each transition is generated as time-series test data (see FIG. 2).

In determining the occurrence time, first, referring to FIG.
, The transition Tr1, corresponding to the test case,
Arrange the columns of Tr2, ~ Trn, and then, as shown in FIG.
By focusing on the operation of the timer, we extract the time constraint between transitions from the software specifications. Specifically, during the transition from operating the timer to the end of the timer, paying attention to the part where the timer is operating, the elapsed time (sum of) is equal to the set time of the timer. It uses the relationship that there is. For example, in the example shown in FIG. 5, when the timer setting value is T, the time constraint T = T1 is obtained by finding and adding the time period in which the timer is operating.
+ T2 can be extracted.

In this way, after the constraint regarding the transition occurrence time is extracted, the time constraint is normalized as shown in FIG. 6 in order to eliminate the redundancy of the time constraint. Specifically, it is to eliminate duplication of the interval in which the timer is operating, which constitutes the time constraint.

Finally, the occurrence time of each transition is determined so as to satisfy the time constraint extracted as described above. As a determination method, as shown in FIG. 7, the average residence time to the state, which is one of the knowledge about the product, is used, and the time constraint given as the time width is distributed.

The determination of these transition times proceeds in a manner of propagating the relation of the time constraint for each timer as shown in FIG. In FIG. 8, the time determination is propagated in the direction of the arrow. If the transition occurrence time is determined in this way,
An automatic test is enabled by correlating the event that causes the transition at each occurrence time from the software specifications, and generating the data corresponding to the event at the determined occurrence time and giving it to the test target system.

When the event is determined, what kind of test data is necessary is determined. Therefore, if the test data determined by the corresponding event that causes the transition is generated at the occurrence time that causes the transition, at the necessary timing, Necessary test data can be generated, and time-series data corresponding to the actual condition of control can be automatically generated. The present invention is as described above,
Moreover, the present invention is not limited to the embodiments shown in the drawings, and can be appropriately modified and carried out within the scope of the invention.

As described above, the present invention uses the software specification described based on the state transition model and the test case given as the sequence of transitions on the specification due to the state transition, and the transition which is the given test case. For the columns of, extract the temporal constraints on the occurrence times of those transitions, determine the occurrence times of the transitions so as to satisfy this time constraint, and then data corresponding to various events that trigger each transition. Is generated as time-series test data, and in determining the occurrence time, first,
By arranging the sequence of transitions corresponding to the test case and then focusing on the operation of the timer from the software specifications, the time constraint between transitions is extracted. In this way, after extracting the constraint on the occurrence time of transition, in order to eliminate the redundancy of the time constraint, the constraint is normalized, and finally, the form that satisfies the time constraint extracted as above is satisfied. Then, the occurrence time of each transition is determined. As a decision method,
The average residence time to the state, which is one of the knowledge about the product, is used to determine the distribution of the time constraint given as the width of time. The determination of these transition times proceeds in the form of propagating the constraint relationship for each timer.

When the occurrence time of the transition is determined in this way, the corresponding event that causes the transition is made to correspond to the occurrence time from the software specifications, and the data corresponding to the event is generated at the determined occurrence time. It is possible to perform an automatic test by applying it to the system under test.

According to the present invention, when a test case is described, it becomes possible to automatically generate test data of a time series signal corresponding to the test case at a timing matched with the actual condition of the control. It saves a great deal of effort for the test, and allows the test to be performed in a form that matches the actual control situation in terms of timing.
The reliability of software can be improved.

[0041]

As described above in detail, according to the present invention,
In the test process in software development, it is possible to easily describe the test case and automatically generate the test data of the time-series signal that matches the actual control condition from the described test case. Therefore, the work involved in the generation of test data is significantly reduced, the productivity of software is improved, and sufficient tests can be performed from the quantitative margin of work, and the reliability of software is improved. A software development device can be provided.

[Brief description of drawings]

FIG. 1 is a system block diagram showing an embodiment of the present invention.

FIG. 2 is a system block diagram showing details of main parts in FIG.

FIG. 3 is a system block diagram showing details of main parts in FIG.

FIG. 4 is a diagram for explaining an embodiment of the present invention,
The figure for demonstrating the determination process of the occurrence time.

FIG. 5 is a diagram for explaining an embodiment of the present invention,
The figure for demonstrating the process which extracts the time restrictions between transitions.

FIG. 6 is a diagram for explaining an embodiment of the present invention,
The figure for demonstrating the process which normalizes a constraint.

FIG. 7 is a diagram for explaining an embodiment of the present invention,
The figure for demonstrating the process which determines the occurrence time of each transition.

FIG. 8 is a diagram for explaining an embodiment of the present invention,
The figure for demonstrating the process which determines the occurrence time of each transition.

[Explanation of symbols]

1 ... Specification description section 2 ... Test case description section 3 ... Time series data generation section 4 ... Test execution section 21 ... Test case creation section 22 ... Test case 23 ... Time constraint extraction section 24 ... State transition software 25 ... Transition time determination section 26 ... Time-series signal generation unit.

Claims (1)

[Claims] 1. A specification description means for giving a software specification described by a state transition model including the concept of time, and based on the software specification given by the specification description means, a certain event in a certain state in this specification On the other hand, a test case description means for executing a certain action to generate a test case describing a sequence of transitions to a certain state, and a test case obtained from this test case description means are obtained from the specification description means. An extraction function that extracts externally applied events to the software under test and the time constraints existing between those events using the software specifications based on state transitions, and time constraints corresponding to the events extracted by this extraction function. Time series data having a generation function for generating time series data satisfying A data generation unit, the software development apparatus characterized by comprising a test executing unit for executing the time-series data test software using the time-series data from the generating means.