JP2001014154A - Object-oriented software component extraction supporting and retrieval device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an object-oriented software component extraction supporting and retrieval device for supporting the efficient extraction and storage of an object-oriented software component from an existing class diagram in an object-oriented software component retrieval device for performing retrieval by an application field and an appearing frequency. SOLUTION: This device is provided with a class structure component extraction condition input means 2 for receiving a relation number in extracting a class structure component as input, a class structure component extraction means 5 for extracting the candidates of the component from the existing class diagram, a pattern classification means 8 for classifying the class structure component as a pattern from the structure of the class diagram, an each pattern class structure component number calculation means 11 for calculating the using frequency of the component for each pattern, an each pattern class structure component number display means 12 for displaying the result, a pattern application field retrieval means 14 for retrieving the application field of the component for each pattern and a pattern application field display means 13 which is the input/output means.

Description

DETAILED DESCRIPTION OF THE INVENTION

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an object-oriented software component extraction support device for supporting the extraction of a class structure component from an object-oriented class diagram, and an object-oriented software component cooperating with the extraction support device. Related to a search device.

2. Description of the Related Art In general, when extracting a component used as an object-oriented software component, a similar document frequently used among design documents such as existing class diagrams and programs implemented in an object-oriented language is generally used. A refined class group (referred to as a pattern in the present invention) is used as a component in order to enhance reusability. Therefore, in component extraction, the frequency of use of components in existing class diagrams and programs can be an important guide. Also, when using an object-oriented software component, if it is to be used for a class design in a specific application field, it is useful to search for an object-oriented software component in the same field. Several years ago, various devices have been invented as devices using object-oriented software components having excellent clarity and independence in modular structure. However, in order to store the components extracted as components in advance and search them, there is a device that extracts an object-oriented software component from an existing class diagram, and a search device that searches for the results. Linking with has not been realized. For this reason, it has not been realized that useful information such as the frequency of use of the component and the field of application is added to the component at the time of component extraction, and the information is made available at the time of retrieval.

In the conventional object-oriented software component search apparatus as described above, when registering a software component to be searched, information such as the field of application and the frequency of use in an existing class diagram is used when searching. In order to obtain it, it was necessary to individually register each part. Further, in order to create data to be searched, input means for individually inputting data of parts is required. The object of the present invention is to provide an object-oriented software component extraction support device that uses existing class diagram data as an input to the device when the object-oriented software component search device stores components. When registering useful information such as the frequency of use of parts and application fields to parts, it is not necessary to add each piece individually. It is realization.

The object-oriented software component extraction support and retrieval device of the present invention is provided with the following means to solve the above-mentioned problems. The class structure component extraction condition input means includes an upper limit of the number of relations belonging to the class structure component.
The lower limit is received as an input, and the result is passed to the class structure component extraction means as a class structure component extraction condition. Here, the class structural component is a set of classes and relationships that constitute a part of an existing class diagram, and the classes are connected by the relationships. This class structure component extraction condition input means can be used when it is deemed that the use value as a component is small even if a class structure component having a very large or small number of relations is extracted from an existing class diagram. Means. The class structure component extracting means includes an input class diagram DB storing information of an existing class diagram as data together with an application field of the class diagram, an application field, a class type,
Category that stores a list of relationship types as data
Based on the DB and the class structure component extraction condition passed by the class structure component extraction condition input means, a class structure component is extracted and stored in the class structure component DB. In addition, by storing in the class structure parts DB data on which class diagram the extracted class structure parts were extracted from, information on existing class diagrams related to the application field is classified by class structure parts. Searchable. The number of relationships constituting each class structural component is between the upper limit and the lower limit of the number of relationships of the class structural component extraction condition. In addition, all the class structural components that are established as class structural components within the range of the upper limit and the lower limit of the relation number are extracted. The pattern classifying means classifies the data of the class structural component group stored in the class structural component DB as a pattern, and classifies the classification result into a pattern.
Store in DB. Also, the pattern to which each class structural component group belongs is stored in the class structural component table. The pattern classification means processes all unprocessed class structural components among the classes of the class structural component DB. When class structural components are classified into patterns by this pattern classifying means, matching / mismatching of types of all classes constituting the class structural components,
Match / mismatch of the types of all the relationships that make up the class structural component, match / multiplicity of all the relationships that make up the class structural component
Classify based on features such as mismatch. In addition, since the information on which pattern each class structural component group belongs to is held as data by the pattern classification means here, it is possible to calculate the use frequency for each pattern at the time of search. The number-of-class-structures-for-pattern calculating means calculates the number of class-structured parts for each pattern by searching the class-structured parts DB and displays the number of class-structured parts for each pattern on the display means. The number of class-structured components by pattern calculated here indicates the frequency of use of the class-structured components belonging to each pattern on the existing class diagram. For this reason, as a result of this apparatus, a pattern having a large number of class-structured parts is displayed that is considered to be highly valuable for refinement in order to enhance reusability. The pattern application field search means determines which application field uses the class structural component belonging to the pattern to be searched in the input class diagram DB and the class structural component DB.
And outputs the result to the pattern application field display means. The class structural component DB holds information on which pattern each class structural component is classified into, and information on which input class diagram has been extracted.
Since the input class diagram DB holds information indicating to which application field each class diagram belongs, a search can be performed.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An object-oriented software component extraction apparatus according to an embodiment of the present invention has a configuration as shown in FIG. The category DB1 stores a list of application fields and list information of classes and relationships used for classifying patterns of class structural components. For example, application fields include finance, industry, public, and distribution, and the types of classes are stereotypes << control >>, << boundary >>, <
If there are three types of <entity >>, and there are two types of relations, association and aggregation, a table as shown in Fig. 2 shows the category of application field, Fig. 3 shows the type of class, and Fig. 4 shows the type of relation To be stored. If you want to classify patterns from a different perspective, or if you want to do a more detailed classification,
Other than the embodiment can be used for the type of relationship. In the embodiment, only two types of relations, association and aggregation, are supported. However, a generalization-specialization relation is added to the relation type, and a superclass class is added to a table in which the relation is converted into data. By adding an ID, the type of relationship can be added. Class structure part extraction condition input means 2
Is realized by input from the operator,
The upper limit and lower limit of the number of relations constituting the class structural component are input by the operator. This data is passed to the class structure component extraction means 5 through the class structure component extraction condition 3. In the embodiment, the relation number upper limit is set to 1 and the relation number lower limit is set to 3. The input class diagram DB4 stores information of each class of the input class diagram. For example, OMG (Object Managem
ent Group) standardized modeling language UML
If the class diagram is a class diagram described in (Unified Modeling Language) as shown in Fig. 6, and if it is a class diagram designed for a certain financial system, the data of the class diagram is The data of the constituting classes is stored in a table as shown in FIG. 8 and the data of the relationships constituting the class diagram as shown in FIG. Store class diagram data as data. Since the financial application category ID is defined as DT1 in FIG. 2, DT1 is used as the employment category ID in FIG. Similarly, the class category ID in FIG. 8 is the class category ID defined in FIG. 3, and the relation category ID in FIG. 9 uses the one defined in FIG. The class diagram ID in FIG. 7 is an ID for uniquely specifying the class diagram. The class diagram ID in FIG. 8 stores the ID of the class diagram constituting the class, the class ID stores an ID for uniquely identifying the class, and the class name indicates the class name used in the class diagram. Stored. The class diagram ID in FIG. 9 stores the ID of the class diagram that constitutes the relationship, the relationship ID stores the ID for uniquely identifying the relationship, and the relationship name stores the relationship name on the class diagram. , Class ID 1 and class ID 2 are the IDs of the classes at both ends of the relationship, and the class 1 multiplicity and class 2 multiplicity store the multiplicity of each object of class ID 1 and class ID 2 as data. When the type of the relation is aggregation, the ID of the class on the overall side is stored in order to prevent the information of the class diagram from being damaged when the type of the relation is aggregation. The class structural component extracting means 5 is configured to class class component 6 based on the class structural component extracting condition 3 and the input class diagram DB4.
Is extracted and stored in the class structure component DB7. At this time, as in steps S501 to S506 in FIG. 10, the class structure component extraction means 5 counts the total number of relations received in the class structure part extraction condition and stored in the input class diagram DB4 (step S501). , The number of relations that make up each class structural part is in the range from the upper limit to the lower limit of the relation number of the class structural part extraction condition, and it is composed of all combinations of relations connected via classes that are established as class structural parts Class structure parts (steps S502, S503, S50
6) The class structural component is extracted (step S504) and stored in the class structural component DB 7 (step S505). Class structural parts DB7
Stores the class structure component 6 extracted by the class structure component extraction means 5. The structural parts group in the class diagram of FIG.
Expressed in a class diagram, it is a group of 16 class structural parts as shown in FIG. FIGS. 12, 13, and 14 show examples of the class structure component DB7 that stores the class diagram of FIG. 11 as data. FIG. 12 stores the data of the entire class structure component. The class structural part ID is a serial number. The class diagram ID is the ID of the class diagram from which the class structural component is based. The pattern ID is a column that is updated later by the pattern classification means 8, and contains the ID of the pattern into which the class structural component is classified. Here, since the class diagram ID and the pattern ID are held, it is possible to calculate the number of class-structured components for each pattern and search for an application field for each pattern. FIG. 13 stores data related to the classes constituting the class structural component. The class structural component ID stores which class structural component the class is a component of. The class ID is the ID of the class, and the class category ID is an ID representing the type of the class. Also figure
Reference numeral 14 stores data related to the relations constituting the class structural parts. The class structural component ID stores which class structural component the relationship is a component of. Relationship
The ID is the ID of the relation, and the relation category ID is an ID representing the type of the relation. The class ID 1 and the class ID 2 are the IDs of the classes at both ends of the relationship, and the class 1 multiplicity and the class 2 multiplicity store the multiplicity of each object of the class ID 1 and the class ID 2 as data. When the type of the relation is aggregation, the ID of the class on the overall side is stored in the aggregation overall class ID. The pattern classification means 8
According to the procedure of FIG. 15, the class structural components are classified into patterns. The processing is performed on the data of which the pattern is not classified among the data of the class structure component DB7 (steps S801 and S802). It is determined in steps S803 and S804 whether or not the pattern of the class structural component is registered in the class structural component DB7. At this time, each class type, relationship type,
It is determined whether the patterns are the same based on whether the multiplicity matches. If there is no pattern that matches pattern DB10, register a new pattern in pattern DB
(Step S805). For the processed class structural component, data on which pattern the class structural component is classified into is realized by updating the class structural component DB7.
(Step S806). Class structure parts of Fig.11, Fig.12, Fig.13, Fig.14
Pattern DB when DB7 is processed by pattern classification means 8
FIG. 16 shows the contents of 10 in a class diagram, and FIG.
Figure 18 shows the data on the classes that make up the pattern.
FIG. 19 shows data relating to the relations constituting the pattern. FIG. 20 shows the result of updating the class structure component DB 7 of FIG. 12 by the pattern classification means 8. In FIG. 16, as a result of the pattern classifying means 8, although the class structural components in FIG. 11 are 16 class components, the portions surrounded by the dotted lines are the same pattern, and are therefore classified into 11 types of patterns. FIG. 17 stores a pattern ID for uniquely specifying a pattern. The pattern ID in FIG. 18 is the ID of the pattern that constitutes the class, and the pattern class is the ID of the class that constitutes the pattern.
Stores ID. In FIG. 19, the pattern ID is the ID of the pattern that forms the relation, the pattern relation ID is the ID of the relation that forms the pattern, the relation category ID is the type of relation, and the pattern class 1 ID and pattern class 2 ID are both ends of the relation. And the class class 1 multiplicity and the pattern class 2 multiplicity store the multiplicity of the object of each class of the pattern class 1 ID and the pattern class 2 ID at both ends of the relationship. When the type of the relationship is aggregation, the ID of the class on the whole side is stored in the aggregation whole-side class ID.
The data of the class structure parts enclosed by the dotted line in FIG.
11 shows records classified into the same pattern. The number-of-class-structures-by-pattern calculating means 11 calculates the number of classes for each pattern from the class structure parts DB 7,
The result is sorted in descending order by the number of class structural parts, and the result is displayed on the class structural part number display means 12 for each pattern. The number-of-class-structures-by-pattern display means 12 has a screen as shown in FIG. The pattern application field display unit 13 is a user interface to the pattern application field search unit, and has a screen as shown in FIG. When only the class diagram of FIG. 6 is processed by this apparatus, the application field of PT3 is only finance, but FIG.
Is described assuming that class diagrams of multiple application fields other than the finance field other than those in Fig. 6 have been processed.
Non-financial industries and public sectors have been added to the application areas of pattern PT3. The pattern application field search means 14 receives the category DB1, the input class diagram DB4, the class structural component DB7, and the pattern application field display means 13 as inputs, and the pattern input from the pattern application field display means as in the procedure of FIG.
The ID is acquired (step S1401), the record of the class structural component having the pattern ID acquired in step S1401 is retrieved from the pattern DB7 from the pattern DB7, and a list of class diagram IDs is acquired (step S1402). Next, a record of the class diagram having the class diagram ID acquired in step S1402 is searched from the input class diagram DB, and a list of application category IDs is acquired (procedure
S1403). The category DB1 is searched from the list of application category IDs acquired in step S1403, and a list of application fields is acquired (step S1404). The list of application field names acquired in step S1404 is displayed in the pattern application field 13 (step S1405).

According to the present invention, by inputting an existing class diagram by the object-oriented software component extraction support device and inputting an application field for each class diagram, the use frequency of the object-oriented software component is calculated and the application field is searched. By using the present apparatus, the number of operations can be reduced as compared with the case where additional information on parts is performed for each part. For this reason, there is an effect that the part extraction work and the search work can be performed efficiently.

[Brief description of the drawings]

FIG. 1 is a system configuration diagram of an object-oriented software component extraction support device according to an embodiment of the present invention.

FIG. 2 is a diagram showing an example of a table storing types of application fields of a category DB.

FIG. 3 is a diagram showing an example of a table storing types of classes of a category DB.

FIG. 4 is a view showing an example of a table storing types of relations of a category DB;

FIG. 5 is a diagram showing an example of a screen when the class structure component extraction condition input means of the present invention is input from an operator.

FIG. 6 is a diagram showing a sample class diagram as a basis for input class diagram DB data;

FIG. 7 is a diagram showing an example of data storage (data of the entire class diagram) in the input class diagram DB when the application field of the sample of FIG. 6 is finance;

8 is a data storage example of an input class diagram DB for storing the sample of FIG. 6 as data (class diagram class data)
FIG.

9 is a diagram showing an example of data storage (data relating to class diagrams) in an input class diagram DB that stores the samples of FIG. 6 as data.

FIG. 10 is a flowchart of class structure component extraction by the class structure component extraction means 5;

11 is a diagram showing data of a class structure part DB7 output from the class structure part extraction means 5 in a class diagram (a part enclosed by a dotted line is a class structure classified as the same pattern by the pattern classification means 8). FIG.

FIG. 12 is a diagram showing a data example of a class structure component table of a class structure component DB.

FIG. 13 is a diagram showing a data example of a class structure component class table of a class structure component DB.

FIG. 14 is a diagram showing an example of data of a class structure parts relation table of a class structure parts DB.

FIG. 15 is a flowchart showing a procedure when the pattern classifying unit 8 classifies the class structural parts into patterns and updates the pattern DB and the class structural parts DB.

FIG. 16 is a view showing a class diagram representing the contents of a pattern DB10 after the class structural component DB7 has been processed by the pattern classification means 8;

FIG. 17 is a diagram showing a data example (data of the entire pattern) of the contents of a pattern DB10 after the class structural component DB7 has been processed by the pattern classification means 8;

FIG. 18 is a view showing a data example (data of a class constituting a pattern) of the contents of a pattern DB after a class structural component DB is processed by a pattern classifying means.

FIG. 19 is a view showing an example of data (contents of a pattern constituting a pattern) of the content of a pattern DB10 after processing a class structural component DB7 by a pattern classification means 8;

FIG. 20 is a diagram showing a result of updating the class structural component DB7 of FIG. 12 by the pattern classifying means 8 (records belonging to the same class pattern are surrounded by dotted lines).

FIG. 21 is a diagram showing an example of a screen of a class-by-pattern class component count display means 12;

FIG. 22 is a view showing an example of a screen of a pattern application field display means 13;

FIG. 23 is a flowchart showing the procedure of a pattern application field search unit 14;

[Explanation of symbols]

1… Category DB, 2… Class structure part extraction condition input means,
3… Class structure part extraction condition, 4… Input class diagram DB, 5…
Class structure part extraction means, 6 ... Class structure parts, 7 ... Class structure parts DB, 8 ... Pattern classification means, 9 ... Patterns, 10
... Pattern DB, 11 ... Pattern-by-pattern class structure part number search means, 12 ... Pattern-by-pattern class structure part number display means, 13 ... Pattern application field display means, 14 ... Pattern application field search means.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Hirofumi Tanno 1099 Ozenji Temple, Aso-ku, Kawasaki City, Kanagawa Prefecture F-Term in Hitachi, Ltd. System Development Laboratory 5B075 ND04 NR03 NR06 NR12 PP03 PP13 QT03 5B076 DB04 DD06 DE04 EC08

Claims (4)

[Claims] 1. A class diagram group is input, a structure between classes included in a class diagram is extracted based on pattern information set in advance, and the extracted structure is stored as a part in an auxiliary storage, and a request is received from a designer. Accordingly, an object-oriented software component support and retrieval device that retrieves the component group for each pattern and displays the component group on a terminal screen, wherein the pattern information includes information on a class type, a relationship type, and a multiplicity. It is characterized by the following. 2. The object-oriented software component support and retrieval apparatus according to claim 1, wherein application field information is added to an input class diagram, and the application field information is displayed together with a component display. I do. 3. The object-oriented software component support and retrieval device according to claim 1, wherein the number of appearances in a class diagram group is calculated for each pattern and displayed together with the component. 4. The object-oriented software component support and retrieval apparatus according to claim 1, wherein in structure extraction between classes, an upper limit and a lower limit of a relation number between classes included in a structure to be extracted are set. Features.