Method and apparatus for judging necessity of performing integration test

The present invention relates to a necessity determination method and an apparatus for determining the necessity of connection inspection of software parts.

In recent years, in embedded systems, which are computer systems embedded in machines and devices to realize specific functions, many variations are prepared in one product series in order to meet the diverse demands of users.

In software development of such embedded systems, the software product line (SPL), which is one of the development methodologies that systematize technical elements for various types of products, is attracting attention.

In SPL, the information obtained by product analysis specifies the required specifications according to the feature model that defines the relationships between the individual specifications (also called features) that are subdivided, and is prepared in advance according to the specified required specifications. Individual products (application software) are produced by appropriately combining the core assets (reusable software components) (see, for example, Patent Document 1 and Non-Patent Document 1).

<p><patcit num="1"><text>Japanese Unexamined Patent Publication No. 2012-8660</text></patcit></p>

<p><nplcit num="1"><text>Yoshimura, Kikuno, "Introduction of Software Product Lines in Embedded Systems" Information Processing Vol.50 No.4 Apr.2009 P295-P302</text></nplcit></p>

<p num="0007"> By the way, in SPL, it is necessary to guarantee that software components provided as core assets operate according to specifications even when used alone or in combination with other software components. For that purpose, it is necessary to perform a joint inspection between software components in advance.</p><p num="0008"> Here, considering that all of any two combinations of n kinds of software parts are subjected to the connection inspection, it is considered._nC₂There are street combinations. In practice, it is necessary to inspect even when three or more software parts are combined.</p><p num="0009"> That is, there is a problem that as the number of software parts prepared as core assets increases, the labor required for the integration inspection increases remarkably. Therefore, in reality, it is necessary to take measures such as omitting the connection inspection for the combination of software parts that cannot be selected at the same time, but for each combination of a huge number of software parts, each time. However, if the worker was checking the necessity of the joint inspection one by one, there was a problem that the labor was increased rather than reduced.</p><p num="0010"> An object of the present invention is to reduce the time and effort required for integration inspection of software components provided as core assets in a software product line in order to solve the above problems.</p>

<p num="0011"> In the method for determining the necessity of coupling inspection of the present invention, in the first step, the correspondence between each feature constituting the feature model expressing the characteristics of the product group manufactured by the software product line and the type and variation of the software component is determined. Acquire the corresponding information to be represented.</p><p num="0012"> In the second step, the type of software component is i (i = 1, 2, ... M: M is the total number of types of software component), and the variation is j (j = 1, 2, ... N)._i: N_iIs the total number of variations of software parts specified by type i), and one software part specified by i, j is S_ijAs represented by, a state variable prepared for each feature, which is true when the feature is selected and false when the feature is not selected, is used, and the software component S is used according to the correspondence information acquired in the first step._ijFor each software component S_ijSet of state variables of all features associated with C_ijTo generate.</p><p num="0013"> In the third step, the software component group consisting of Ni variations of the software component specified by i is S._ix, S_ixN other than the variation specified by j_i-1 variation is represented by t (t j), C_ijThe logical expression obtained by combining all the elements of H (C)_ij), H (C_ij) Is logically denied, and the logical expression obtained is NH (C)._ij), The software component S_ijEvery time, H (C_ij) And N_i-1 NH (C_it) Is combined by a logical product, and the individual discriminant IS (C)_ij) Is generated.</p><p num="0014"> In addition, H (C_ij) Is S_ixIt represents the condition for selecting the variation software component specified by j among (the same type of software component specified by i), and also NH (C)._ix) Is S_ixIn the above, the condition that the software component of the variation specified by t (that is, the software component other than the variation specified by j) is not selected is expressed. That is, IS (C) in which these are combined by a logical product._ij) Is the same type of software component (S)_ix), The condition that a plurality of variations are not selected at the same time, that is, only one variation is selected, is H (C)._ij) Is added to the conditions indicated.</p><p num="0015"> In the fourth step, two or more software components to be determined whether or not the connection inspection is necessary are specified. In the fifth step, the necessity judgment formula is generated by combining the individual judgment formulas generated by the individual judgment formula generation means for all the software parts specified in the fourth step by a logical product, and the necessity judgment formula is generated. If the necessity judgment formula is always false (tautology), inspection is unnecessary, and if the necessity judgment formula is unknown or always true (tautology), it is judged that inspection is necessary.</p><p num="0016"> According to the connection inspection necessity determination method of the present invention, the software component S_ijIndividual judgment formula IS (C) prepared for each_ij) Can be used to easily determine the necessity of the coupling inspection for any combination of software components without referring to the feature model or the like.</p><p num="0017"> Further, in the coupling inspection necessity determination device of the present invention, the corresponding information acquisition means executes the process corresponding to the first step, the state set generation means executes the process corresponding to the second step, and the individual determination formula is used. The generation means executes the process corresponding to the third step, the target selection means executes the process corresponding to the fourth step, and the inspection necessity determination means executes the process corresponding to the fifth step. , The determination result output means outputs the determination result by the inspection necessity determination means.</p><p num="0018"> Therefore, according to the coupling inspection necessity determination device of the present invention, it is possible to obtain the same effect as the effect obtained by executing the above-mentioned coupling inspection necessity determination method.</p>

<figref num="1">It is a block diagram which shows the whole structure of the connection inspection necessity determination apparatus.</figref><figref num="2">(A) is an explanatory diagram showing the structure of the feature model, and (b) is an explanatory diagram explaining the contents of the corresponding information.</figref><figref num="3">It is a flowchart which shows the content of the individual determination formula generation processing.</figref><figref num="4">It is a flowchart which shows the content of the combination inspection necessity determination processing.</figref><figref num="5">It is explanatory drawing which shows the structure of the determination result database.</figref><figref num="6">It is explanatory drawing which shows the structure of the search database.</figref>

Hereinafter, embodiments of the present invention will be described with reference to the drawings. &lt;Overall configuration&gt; As shown in FIG. 1, the coupling inspection necessity determination device 1 according to the present embodiment includes an operation unit 2 for inputting various commands, an output unit 3 for displaying an operation procedure, a processing result, and the like, and software. A storage unit 4 that stores various information used in the product line (SPL) and a processing unit that executes processing using various information stored in the storage unit 4 according to a command input from the operation unit 2 and outputs the processing result. It is provided with a control unit 5 that outputs to 3 and stores in storage 4.

The storage unit 4 includes a feature model created by analysis of a product group to which SPL is applied (so-called domain engineering), software components that are reusable software, and individual features and software components shown in the feature model. The core asset database (hereinafter, the database is referred to as DB) 41 that stores the correspondence information indicating the correspondence relationship of the software, the individual judgment formula DB42 that stores the individual judgment formula generated for each software component by the processing described later, and the combination inspection. The combination pattern of the software parts to be inspected and the necessity judgment result of judging the necessity of the combination inspection for the combination pattern are stored in association with each other. The information used for searching the judgment result DB43 and the judgment result DB43 is stored. The search DB 44 is provided.

&lt;Feature model / software parts&gt; Here, as shown in FIG. 2A, the feature model systematically defines the relationships (dependent relationships, choices, etc.) of each feature with the individual specifications (functions, quality, etc.) extracted by product analysis as features. It is represented in.

Correspondence information indicates the correspondence relationship with a specific software component and a specific variation used to realize the function specified by the feature, and as shown in FIG. 2 (b), each feature has a correspondence relationship. , One or more software components are associated. Further, a plurality of types of software components are prepared, and some software components have a plurality of variations depending on the type.

In the following, the type of software component is i (i = 1, 2, ... M: M is the total number of types of software component), and the variation is j (j = 1, 2, ... N)._i: N_iIs the total number of variations of software parts specified by type i), and one software part specified by i, j is S_ijRepresented by, and N of the software component specified by type i_iWhen collectively referring to individual variations, S_ix(X = 1, 2, ... N_i).

Further, software components belonging to the same software component group, that is, software components of the same type but different variations are in a mutually exclusive relationship, and are of the same type in one product (application software) manufactured by SPL. Only one variation is used from the software components of.

Specifically, for example, in SPL that manufactures application software for vehicle control, when there is a type of software component called "wiper motor control" and there are "for right handle" and "for left handle" as variations thereof, This means that the software component "wiper motor control for the right steering wheel" and the software component "wiper motor control for the left steering wheel" cannot be used at the same time.

Then, in SPL, the required specifications are specified by appropriately selecting the features according to the feature model, and a product (application software) according to the specified required specifications is generated, and the product is selected. It consists of software components (identified using correspondence information) associated with individual features.

&lt;Control unit&gt; The control unit is composed of a well-known microcomputer equipped with at least a CPU, ROM, and RAM, and is an individual judgment formula generation process that generates an individual judgment formula to generate a judgment formula DB 42, and an individual judgment stored in the individual judgment formula DB 42. At least the connection inspection necessity determination process for determining the necessity of the connection inspection between the specified software components is executed by using the formula.

All of these processes are activated according to the instructions input from the operation unit. &lt;&lt; Individual judgment formula generation process &gt;&gt; Here, the content of the individual determination formula generation process will be described with reference to the flowchart shown in FIG. In the following, the step is simply referred to as "S".

When this process is activated, first, in S110, the correspondence information is acquired from the core asset DB 41. In S120, the software component S is based on the corresponding information acquired in S110._ijFor each software component S_ijSet of state variables of all features associated with C_ijTo generate. The state variable of the feature consists of a Boolean variable that is true (TRUE) when the feature is selected and false (FALSE) when the feature is not selected.

In S130, the set C generated in S120_ijUsing, set C_ijThe logical expression H (C) in which all the elements (state variables) of_ij) Is generated. That is, the formula H (C)_ij) Is the software component S_ijWhen all the features associated with are selected (all state variables are true), their values are true.

In S140, the parameter i that specifies the type of software component is initialized to 1, and in the subsequent S150, the type is specified by the parameter i._iSoftware component S_ixFor each of the above, the formula H (C) relating to the software component_ix) Negated logical expression NH (C)_ix) Is generated.

In S160, the parameter j that specifies the variation of the software component is initialized to 1. In S170, the negative formula NH (C) generated in S150 is negated._ix), N where t j_i-1 Negative formula H (C_it), And formula H (C)_ij) N_iBy combining all of the logical expressions with a logical product, the software component S_ijIndividual judgment formula IS (C_ij) Is generated and stored in the individual determination formula DB 42.

In S180, the parameter j is incremented (j j + 1), and in the following S190, the parameter j is the software component S._ixTotal number of variations N_iJudge whether or not it is as follows.

Parameter j is the total number N_iIf the following, return to S170 and the individual judgment formula IS (C)_ij) Is repeated, and the total number of parameters j is N._iIf it is larger, the software component S of all variations whose type is specified by the parameter i_ixAssuming that the generation of the individual determination formula has been completed, the process proceeds to S200.

In S200, the parameter i is incremented (i i + 1), and in the subsequent S210, it is determined whether or not the parameter i is equal to or less than the total number of software component types M. If the total number of parameters i is M or less, the process returns to S150, the generation of the negative ethical formula, the generation of the individual judgment formula, and the memory are repeated. For software parts, this process ends assuming that the generation of the individual judgment formula has been completed.

Here, an example of the individual determination formula generated by this process is shown below. For example, each state variable is represented by f1, f2, ..., And the total number of variations of the type of software component specified by i = 1 is N.₁= 3, and the feature set of each of these software components is C.₁₁= {F1, f2}, C₁₂= {F1, f3}, C₁₃Consider the case where = {f2, f4}. Further, assuming that the logical product is represented by "*" and the logical negation is represented by "!", The individual determination formula IS (C) of the software component S11.₁₁) Is expressed by the equation (1).

IS (C)₁₁) = H (C₁₁) * NH (C₁₂) * NH (C₁₃) = (F1 * f2) *! (F1 * f3) *! (F2 * f4) = F1 * f2 *! f3 *! f4 (1) &lt;&lt; Combination inspection necessity judgment processing &gt;&gt; Next, the content of the combination inspection necessity determination process will be described with reference to the flowchart shown in FIG.

When this process is activated, first, in S310, the determination target information representing the combination of software components to be determined (hereinafter referred to as "target combination pattern") is acquired. As the determination target information, the content specified by the operator may be acquired via the operation unit 2, or the determination target information prepared in advance (stored in the storage unit 4 or an external storage device (not shown) or the like). It may be acquired sequentially from the list.

In S320, it is determined whether or not the target combination pattern shown in the determination target information is already registered in the determination result DB 43 according to the determination target information acquired in S310. Then, if the target combination pattern has already been registered in the determination result DB, the process proceeds to S330, the determination result for the target combination pattern is acquired from the determination result DB 43, and the acquired necessity determination result is output to the output unit 3. And end this process.

On the other hand, if the target combination pattern is not registered in the determination result DB, the process proceeds to S340 to generate a necessity determination formula for the target combination pattern. Specifically, the necessity judgment formula is obtained by acquiring the individual judgment formulas of all the software parts shown in the target combination pattern from the individual judgment formula DB42 and combining all the acquired individual judgment formulas by a logical product. To generate.

In S350, it is determined whether or not the value of the necessity determination formula generated in S340 is always false (constant false expression) regardless of the value of the state variable, and if it is affirmative (always false), it is determined. The process proceeds to S360, the determination result indicating that the coupling inspection is unnecessary is output to the output unit 3, and the process proceeds to S380.

On the other hand, if a negative judgment is made in S350, that is, if the necessity judgment formula is unclear or always true (tautology), the process proceeds to S370 and the judgment result indicating that the connection inspection is necessary is obtained. Output to the output unit 3 and proceed to S380.

In S380, the determination result obtained in S360 or S370 is associated with the target combination pattern and registered in the determination result DB 43, and the determination result of the newly registered target combination pattern can be searched. DB44 is updated to end this process.

Here, an example of the necessity determination formula generated by this process is shown below. For example, the feature set is C₁₁= {F1, f2}, C₁₂= {F1, f3}, C₁₃= {F2, f4}, C₂₂Consider the case where = {f2, f4}.

In this case, the software component S₁₁, S₂₁Necessity judgment formula ITT (C) for the combination of₁₁, C₂₁) Is equation (2), software component S₁₂, S₂₁Necessity judgment formula ITT (C₁₂, C₂₁) Is expressed by the equation (3).

ITT (C₁₁, C₂₁) = IS (C)₁₁) * IS (C)₂₁) = F1 * f2 *! f3 *! f4 (2) ITT (C₁₂, C₂₁) = IS (C)₁₂) * IS (C)₂₁) = F1 * f3 * (! F3 +! F2 *! F4) * f2 *! f4 (3) Since the truth of equation (2) is determined by the value of the state variable, the software component S₁₁, S₂₁It is judged that the combination of is necessary for the combination test. On the other hand, in the equation (3), (f2 *! F2) is included in the equation, and the equation (f2 *!₁₂, S₂₁It is judged that the combination of is not required for the combination test.

&lt;Judgment result DB / Search DB&gt; Here, the determination result DB 43 for storing the necessity determination result by the necessity determination formula and the search DB 44 used for searching the determination result DB 43 will be described.

As shown in FIG. 5, the determination result DB 43 relates to the target combination pattern and the target combination pattern in association with the identification number (No.) for identifying the combination pattern of the software components to be inspected for the coupling inspection. The necessity judgment result indicating the result of performing the necessity judgment of the combination test using the necessity judgment formula and the test result which is the result of the combination test (explanation omitted) separately executed are stored. It is configured. However, here, as a test result, the storage destination address of the result is stored.

Further, as shown in FIG. 6, the search DB 44 is configured to store an inverted index for the software component in association with each software component to be inspected for the coupling inspection.

The inverted index is generated according to the registered contents of the determination result DB 43, and is composed of a set of identification numbers for each software component, in which all the identification numbers assigned to the target combination pattern including the software component are listed.

That is, by looking at the inverted index of the software component of interest (part of interest), it is possible to specify where in the determination result DB 43 all the information about the target combination pattern including the component of interest is stored. It is configured.

Then, when actually searching the determination result DB 43 using the search DB 44, the procedure is as follows. First, the inverted index is extracted from the search DB 44 for all the software components constituting the target combination pattern to be searched.

The logical product of the extracted inverted index is calculated. Specifically, only when all the extracted inverted indexes include the same identification number, the identification number is output as a result of the logical product.

When the result of obtaining the logical product is null (no identification number is output), it means that the searched target combination pattern is not registered in the determination result DB 43. On the other hand, if the result of obtaining the logical product is not null, it means that the searched target combination pattern is registered in the determination result DB 43. In this case, by searching for the identification number shown in the result, information on the necessity determination result and the test result can be obtained.

&lt;Effect&gt; As described above, according to the coupling inspection necessity determination device 1, the software component S_ijIndividual judgment formula IS (C) generated for each_ij) Is used to logically connect the individual judgment formulas of all the software parts to be inspected, and the necessity judgment formula is used to analytically determine the necessity of the connection inspection. Regardless of the number of software parts or the complexity of the relationship between software parts, the necessity judgment result of the connection inspection can be obtained by a simple logical calculation.

In addition, by performing the combination inspection according to the necessity judgment result, it is possible to prevent unnecessary combination inspection from being performed for combination patterns that are not used in combination at the same time due to the specifications. Therefore, it is a core asset. It is possible to reduce the time and effort required to construct the.

Further, according to the combination inspection necessity determination device 1, the determination result DB 43 stores the necessity determination result and the combination inspection result in association with the combination pattern of the inspection target, and uses the search DB 44 to make a past determination. Since the results and the like can be easily searched, it is possible to prevent duplicate judgments and inspections for the combination patterns that have been judged and inspected.

&lt;Other Embodiments&gt; Although one embodiment of the present invention has been described above, the present invention is not limited to the above embodiment, and can be implemented in various embodiments.

For example, in the above embodiment, the corresponding information is acquired from the core asset DB 41 constituting the storage unit 4, but for example, it is acquired from an external storage device (not shown) or on a communication network via a communication network such as the Internet. It may be acquired from a server or the like provided in.

In the above embodiment, the determination result DB 43 and the search DB 44 are provided, but these may be omitted. In that case, for example, when a determination result by the necessity determination formula is obtained, the coupling test may be immediately executed according to the determination result.

1 ... Join inspection necessity judgment device 2 ... Operation unit 3 ... Output unit 4 ... Storage unit 5 ... Control unit 41 ... Core asset database 42 ... Individual judgment type database 43 ... Judgment result database 44 ... Search database