JP2005018505A - Class extraction method and device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem that for the development of business components, a creation unit of objects and methods, or the like is not established, and the generation method is left to a developer, which make it difficult to assure the quality thereof, causing the reduction of development efficiency. <P>SOLUTION: A design is conducted from two aspects of 'access to a database' and 'business requirement' to extract classes properly. More specifically, for an approach from the aspect of access to a database, a level is provided by focusing attention on the parental relation of tables, and an object and a method are extracted according to the level. Thereby, class extraction is performed which is conventionally dependent on designer's skills. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a technique for mechanically extracting a class in a component design phase in application development.
[0002]
[Prior art]
The application development process consists of business analysis (summary definition), basic design (detailed definition), detailed design (design, program design), coding (program creation), test (integration test, comprehensive test), operation (introduction) Classified as a process. This is an international standard “Software Lifecycle Process Standard ISO / IEC12207: 1995 (JIS X 0160-1996)”, which was created in consideration of the characteristics of the Japanese software industry. “1998 version common frame 98 SLCP-JCF98” Is defined. This “1998 version common frame 98 SLCP-JCF98” is described in Non-Patent Document 1.
[0003]
Also, business analysis, coding, and testing procedures have already been established in the application development process, and there are many tools to support it. However, in the design process, there is a notation UML for sharing design information and a tool that supports the UML, but no specific procedure has been established. Further, in the design of current application development, the adoption of a three-layer architecture, in which functions to be implemented are divided into a P (Presentation) layer, an F (Function) layer, and a D (Data) layer, is becoming mainstream. The P layer is processing related to the screen, the F layer is business processing, and the D layer is mainly database access processing. This “three-layer architecture” is described in Non-Patent Document 2.
[0004]
Also, as a method for system development, DOA (Data Oriented Approach) is currently the mainstream. DOA is a methodology that has been developed in the fields of enterprise information systems and business applications, and is characterized by being suitable for data processing-centric business systems. This “DOA (Data Oriented Approach)” is described in Non-Patent Document 3.
[0005]
[Non-patent document 1] "Chapter 3 Work on development process" in "Data-centric system analysis and design" (edit: IRM workshop, publisher: Ohm, supervised: Tokyo International University, Horiuchi Hajime)
[Non-Patent Document 2] "Microsoft Wordbook: Three-tiered Architecture [= Three Tiered Architecture]"(http://www.microsoft.com/Japan/Terminology/query.asp?ID=3132&q20%%%%%)
[Non-Patent Document 3] "System development system JIS X0160 and common frame 98 compatible" (edit: Nihon Unisys Information Technology Research Association, publisher: Tokyo Denki University Press)
[0006]
[Problems to be solved by the invention]
No design procedure has been established in the design process that should be performed in consideration of performance and quality. For this reason, design deliverables depend on the skill of the designer, and the quality varies. There is a problem if the design quality is low.
-Scattering of methods that cause performance degradation and methods that are prone to implementation errors.
-Application enlargement by creating multiple methods with similar functions.
[0007]
[Means for solving problems]
In order to solve the above problems, in the present invention, design is performed from two aspects of “database access” and “business requirement”. This makes it possible to extract classes without excess or deficiency.
[0008]
In the database access approach, a level is set by focusing on the parent-child relationship of the table, and objects and methods corresponding to the level are extracted. This makes it possible to semi-mechanically perform class extraction that originally depends on the skill of the designer.
[0009]
In the approach from business requirements, in the business analysis phase, verify whether each work can be realized with the D layer interface obtained from the approach from data access to the use case scenario that summarizes the interaction between the information system and the user in the minimum business unit. At the same time, identify new functions. As a result, the interface extraction and the exchange of each component can be organized.
[0010]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings.
In the following text, terms will be used as follows.
Database access component belonging to layer D ... layer D component
D layer component interface ... D layer interface
D layer interface methods ... D layer methods
D layer component designer ... D layer designer
Business component belonging to F layer… F layer component
F layer component interface ... F layer interface
Methods that make up the F layer interface: F layer methods
F layer component designer ... F layer designer
Regarding the F layer method, a method called from outside the F layer component is called an F layer public method, and a method called inside the F layer component is called an F layer internal method.
[0011]
In the component design phase in application development, database table information created by the DB designer, use case scenarios that summarize user interaction with the information system created by the business designer, and screen / form designer create This is a method of semi-mechanically extracting the D layer and F layer classes using a screen / form list.
[0012]
In the D layer class design, first, the table level is set by paying attention to the relationship between the tables based on the table information of the database. Next, by providing object and method extraction criteria according to the level, it is possible to extract the interface of the D layer component without depending on the skill of the designer.
[0013]
In the F layer class design, it is verified whether each work of the use case scenario can be realized by using the interface of the D layer component obtained from the data-centric approach and the screen / form list. By displaying the verified result in combination with the use case scenario, it becomes possible to identify new functions and promote the interface extraction operation of the F layer component.
[0014]
In addition, using the same embodiment, the design process is divided into two stages of D layer component design unit 4 and F layer component design unit 5, and by automatically transmitting design end notifications 6 and 7 at the end of design, When the F layer designer refers to the D layer interface, the access to the method information DB 45 managed by the D layer component design unit 4 is performed every time when the F layer designer refers to the D layer interface. A method for reducing alignment is also described.
[0015]
FIG. 1 is a system configuration diagram in an upstream phase of application development including the present invention. In FIG. 1, the positioning of the D layer component design unit 4 that defines the D layer class and the F layer component design unit 5 that defines the F layer class, which are main components of the present invention, is positioned in the design phase of application development. Show.
[0016]
FIG. 2 is a basic flowchart showing the D layer class extraction processing procedure viewed from the D layer designer in FIG. Hereinafter, the solid line used in the drawing indicates the flow of control, and the wavy line indicates the data read / write relationship. 3 shows the processing procedure of the Lv1,2,3 method definition unit 42 in FIG. 1, FIG. 4 shows the processing procedure of the D layer interface display unit 43 in FIG. 1, and FIG. 5 shows the processing of the specification generation unit 44 in FIG. It is a flowchart which shows a procedure. The Lv1 method definition unit 421, the Lv2 method definition unit 422, and the Lv3 method definition unit 423 illustrated in FIG. 3 will be described in detail with reference to FIGS. The Lv1 method definition unit 421, the Lv2 method definition unit 422, and the Lv3 method definition unit 423 correspond to the Lv1, 2, 3 method definition unit 42 in FIG. 9 to 13 are examples of screens necessary for realizing the D layer interface extraction process. FIG. 14 shows a result of storing the method information 46 defined in the D layer interface display unit 4 in FIG. 1 in the method information DB 45.
[0017]
FIG. 15 is a basic flowchart showing the F layer class extraction processing procedure viewed from the F layer designer in FIG. 16 shows the processing procedure of the mapping information display unit 54 in FIG. 1, FIG. 17 shows the processing procedure of the specification generation unit 56 in FIG. 1, FIG. 18 shows the processing procedure of the mapping unit 53 in FIG. FIG. 22 shows the processing procedure of the D layer method mapping unit 531, the D layer new method definition unit 532, the F layer internal method definition unit 533, and the P layer addition unit 534 in FIG. 18, respectively, and FIGS. FIG. 25 is a flowchart showing a processing procedure of the P layer screen adding unit 5341 and the P layer processing adding unit 5342, and FIG. 25 is a flowchart showing a processing procedure of the F layer public method defining unit 55 in FIG. The D layer method mapping unit 531 and the P layer screen addition unit 5341 are added to the mapping unit 53 in FIG. 1, and the D layer new method definition unit 532, the F layer internal method definition unit 533, and the P layer process addition unit 5342 are newly created in FIG. It corresponds to each part 52. 26 to 33 are examples of screens necessary for realizing the F layer interface extraction processing. FIG. 34 shows a result of storing the new method information 5a / 5b defined in the new creation unit 52 and the F layer public method definition unit 55 in FIG. 1 in the new method information DB 58. 35 is a result of storing the method mapping information 59 defined in the mapping unit 53 in FIG. 1 in the method mapping information DB 57.
[0018]
36 to 39 are DB table examples used in the “education application & qualification system” that explains the present invention in the embodiment.
[0019]
40 to 44 show the process of extracting the interface of the F layer component.
[0020]
As shown in FIG. 1, the application development design process is divided into the following design units: a DB design unit 1, a screen / form design unit 2, a business design unit 3, a D layer component design unit 4, and an F layer component design unit 5. Work is done in parallel. These design units are physically connected by a network. In the D layer component design unit 4, the ER diagram 12 that is a product of the DB design unit 1 is input, a D layer interface is defined, and the result is output as a D layer interface specification. The D layer component design unit 4 sends an F layer design notification request 6 to the F layer component design unit 5 when the generation of the D layer interface specification is completed. The F layer component design unit 5 receives the F layer design notification request 6, receives the method information DB 45 that is a product of the D layer component design unit 4, the screen / form list 22 that is the product of the screen / form design unit 2, and the business The use case scenario 32, which is a product of the design unit 3, is input to define a D layer additional interface and an F layer interface, and the results are output as an F layer interface specification and a functional specification. Further, the F layer component design unit 5 sends a D layer new method creation request notification 7 to the D layer component design unit 4 when the generation of the F layer interface specification is completed. The D layer component design unit 4 performs design work for a method that newly becomes a D layer interface based on the D layer new method creation request notification 7. That is, the D layer component design unit 4 and the F layer component design unit 5 perform design work while exchanging design request notifications. The DB design unit 1, the screen / form design unit 2, the business design unit 3, the D layer component design unit 4, and the F layer component design unit 5 each include input / output devices 11, 21, 31, 41, 51.
[0021]
The D layer component design unit 4 includes an input / output device 41, an Lv1,2,3 method definition unit 42, a D layer interface display unit 43, a specification generation unit 44, and a method information DB 45.
[0022]
The D layer interface display unit 43 is a main processing unit of the D layer component design unit 4, and the table information 13 from the ER diagram 12, which is a product of the DB design unit 1, the D layer interface from the method information DB 45, and the D layer The method information to be additionally created is read from the new method creation request notification 7 and a list of D layer interfaces is displayed. The table information 13 includes a table name, a table attribute, a search key, and a column name.
[0023]
The method information DB 45 stores the method information 46. The method information 46 is a method number that uniquely identifies a method, a table level, a table accessed by the method, a table attribute, a joined table name for the table, a key for retrieving the table, and a result of retrieving the table to the method user. It consists of the column name, Japanese method name, and method description to be displayed when returning.
[0024]
The Lv1, 2, 3 method definition unit 42 uses the table information 13 accessed from the ER diagram that is a product of the DB design unit 1 and the method information stored in the method information DB 45 via the input / output device 41, Method information 46 constituting the D layer interface is defined.
[0025]
The specification generation unit 44 generates a D layer interface specification using the method definition information 46 stored in the method information DB 45. The specification generation unit 44 delivers the F layer design request notification 6 to the F layer component design unit 5 when the D layer interface specification creation is completed.
[0026]
The F layer component design unit 5 includes an input / output device 51, a new creation unit 52, a mapping unit 53, a mapping information display unit 54, an F layer public method definition unit 55, a specification generation unit 56, a method mapping information DB 57, and new method information. It consists of DB58.
[0027]
The mapping information display unit 54 displays the result of mapping the D layer interface, the F layer interface, the P layer process, and the P layer screen to the use case scenario stored in the method mapping information DB 57.
[0028]
The method mapping information DB 57 stores method mapping information 59. The method mapping information 59 is a number that uniquely identifies a work unit of a use case scenario, a use case scenario corresponding to the number, a hierarchy level that represents a hierarchical structure of method calls, and a method in a hierarchy in one operation of the use case scenario. Execution order, layer to which the mapped method or process belongs (D / F / P), method number identifying the method, Japanese method name or process name, corresponding F layer public method number, F layer public Japanese method Consists of names.
[0029]
The mapping unit 53 uses the screen list 22 which is the product of the screen / form design unit 2 and the product of the D layer component design unit 4 in the use case scenario stored in the method mapping information DB 57 via the input / output device 51. The existing D layer method definition information stored in a certain method information DB 45 is associated with the method definition information stored in the new method information DB 58 to define the method mapping information 59 and store it in the method mapping information DB 57.
[0030]
The new creation unit 52 uses the table information 13 stored in the ER diagram 12 which is a product of the DB1 design unit 1 via the input / output device 51 to generate the D layer new method information 5a and the F layer internal method information 5b. Define. The D layer new method information 5a includes a method number for uniquely identifying a method, a layer (D / F / P) to which the method or process belongs, a method level, a table accessed by the method, table attributes, and a table joined to the table. Name, attribute of join table, key for searching table, item to be displayed when table search result is returned to method user, Japanese method name or process name, method description or process description. The F layer internal method information 5b includes a method number for uniquely identifying a method, a layer (D / F / P) to which the method or process belongs, a Japanese method name or process name, a method description or process description, a method argument Type and return type of method.
[0031]
The F layer public method definition unit 55 defines the public method information 5 b having the same format as the F layer internal method information 5 b via the input / output device 51.
[0032]
The specification generation unit 56 uses the method mapping information 59 stored in the method mapping information DB 57 to generate a functional specification, and uses the F layer method information 5b stored in the new method information DB 58 to generate an F layer interface specification. Is generated. Also, the specification generation unit 57 sends the information of the new D layer method stored in the new method information DB 58 to the D layer component design unit 4 when the F layer interface specification and functional specification are completed. Delivered as a creation request notification 7.
[0033]
First, the D layer component when the D layer designer gives instructions via the input / output device 41 to (1) D layer interface display, (2) D layer method definition, and (3) D layer interface specification creation. The basic operation of the design unit 4 will be described.
[0034]
The D layer component design unit 4 pays attention to the relationship between the tables, and provides levels of Lv1... All tables, Lv2... Code table and related table, Lv3.
[0035]
Then, depending on the level, Lv1 ... basic method, free search · Lv2 ... search by primary key of related table · Lv3 ... search by primary key of child table, search by primary key of parent table Get things.
(A). 1 Methods for accessing the table (select, insert, update, delete with primary key)
(B). Join access method with code table (for name resolution)
(C). Join access method for tables in parent-child relationship (single record information is acquired as a search result)
(D). Parent-child information batch acquisition method in parent-child relationship (multi-record information is acquired as a search result)
(E). Free search method (specify a string of string type after the where clause)
Specifically, the methods (a) to (e) will be described using the “education application & qualification system” as an example. In the “Education Application & Qualification Certification System”, (1) In order to be certified as an XXX engineer, students must take education designated as essential and their grades must meet qualification standards. (2) The business requirement is to output the number of qualified persons by department and call on each department to promote qualification.
[0036]
FIG. 36 shows an example of a DB table list used in the “education application & qualification system”.
(A) Basic method for accessing one table: In the case of the “educational subject detail” table, the subject person in the education center accesses this table, browses the status of the subject (select), and registers a new subject ( The operations of insert, date and time change (update), and subject cancellation (delete) are basic methods.
(B) Retrieval of related table by primary key: In order to resolve the name of the “department” code table, the operation of joining the related table “employee” and displaying the employee list for each department (for example, development department) It becomes a join access method to the code table.
[0037]
FIG. 37 shows a join access example with the code table.
(C) Retrieval by primary key of child table: Joining “attendance history” <child table>, “employee” “details of educational subject” <parent table>, and “educational subject” <parent parent table> The operation of confirming the grade of education that has already been taken is the join access method for tables in a parent-child relationship.
[0038]
FIG. 38 shows a join access example (search by key of child table) of a table having a parent-child relationship.
(D) Search by primary key of parent table: “Participation history” <child table>, “employee”, “educational subject details” <parent table>, and “educational subject” <parent table> are combined. The operation of listing all grades of subjects that have been taken so far using the staff number as a key is a parent-child information batch acquisition method having a parent-child relationship.
[0039]
FIG. 39 shows a join access example (search by key of parent table) of a table having a parent-child relationship.
[0040]
(C) In the search method using the primary key of the child table, the search result is identified as a single record. On the other hand, in the search using the primary key of the parent table, information on multiple records may be acquired as the search result. There is a difference that you can.
[0041]
(E) Free search: (a) A search method at a level not supported by the basic method. The data extraction condition expression specified in the where clause is treated as one character string and passed as an argument.
[0042]
The D layer designer performs D layer class design in the D layer component design unit 4 according to the flowchart shown in FIG.
(1) D layer interface display: When the D layer component design unit 4 receives a D layer interface display instruction from the input / output device 41, the D layer interface display unit 43 performs the following processing. The D layer interface display unit 43 operates according to the flowchart shown in FIG.
[0043]
In the D layer interface display unit 43, (A) D layer component design is (1) first time, or (2) re-executed after receiving table information change from the DB design unit 1, (3) F layer component design unit The following table information reading process is performed depending on whether the D layer new method creation request notification 6 from 5 is received.
[0044]
(1) First D layer component design:
i) Read the table information 13 from the ER diagram 12 which is a product of the DB design unit 1
ii) Based on the table information 13 read in (1), (a) basic method and (e) free search which are Lv1 methods are automatically set.
iii) The Lv1 method setting information automatically set in i) is stored in the method information DB 45. Here, the content stored in the method information DB is the content of the method information storage table shown in FIG.
[0045]
(2) Redo D-layer component design after changing table information
i) The D layer designer instructs the ER diagram 12 that needs to be read from the input / output device 41, and reads only the changed table information.
ii) As in ii) of (1), the Lv1 method is automatically set based on the table information read in i).
iii) As with iii) of (1), the Lv1 method setting information whose contents are changed in ii) is reflected in the method information DB 45.
[0046]
(3) D layer component design after receiving D layer new method creation request notification 7
i) Read the stored D layer interface from the method information storage DB 45
(B) Display data is created from the method information stored in the method information DB 45 and displayed on the output device 41. FIG. 9 shows a display example. The Lv1 method definition unit 421 in the D layer component design creates all default Lv1 methods.
(2) D layer method definition at each level
The movement is different between the case where the layer D new method creation request notification 7 exists (A1) and the case where it does not exist (A2).
[0047]
(A1) When the D layer new method creation request notification 7 exists,
(1) D layer new method information is read and displayed from the D layer new method request notification that is a product of the F layer component design unit 5. FIG. 10 shows a display example.
(2) The D layer designer selects the D layer new method from the input / output device 41,
a) Definition processing by the Lv1, 2, 3 method definition unit 42
b) Delete one of the D layer new method candidates.
[0048]
a) When performing definition processing by the Lv1, 2, 3 method definition unit 42, the Lv1, 2, 3 method definition unit performs the following processing. The Lv1, 2, 3 method definition unit 42 operates according to the flowchart shown in FIG.
[0049]
○ Definition processing by the Lv1 method definition unit 421
The Lv1 method definition unit 421 operates according to the flowchart shown in FIG.
[0050]
i) The Lv1 method definition unit acquires the set Lv1 method name and method selection information from the D layer interface display unit 43, creates display data, and displays the data on the output device 41. A screen example of the Lv1 method definition unit is shown in FIG.
[0051]
ii) The D layer designer makes a method-necessary change from the input / output device 41 to the Lv1 method selection information displayed on the input / output device 41.
[0052]
iii) Receiving the correction result, the Lv1 method setting information stored in the method information storage DB 45 is changed.
[0053]
○ Definition processing by the Lv2 method definition unit 422
The Lv2 method definition unit 422 operates according to the flowchart shown in FIG. FIG. 12 shows a screen display example of the Lv2 method definition unit.
[0054]
i) The Lv2 method definition unit 422 acquires the selected code table name from the D layer interface display unit 43. In the example of FIG. 12, the “department” table is selected as the code table.
[0055]
ii) The Lv2 method definition unit 422 reads the related table information from the ER diagram 12 as a product of the DB design unit 1 using the code table name as a key, creates display data, and displays it on the input / output device 41. Here, the table information includes a table attribute, a related table list, a search key for each related table, and a column name for each related table. In the example of FIG. 12, an “employee” table is displayed as a candidate table related to the “department” table.
[0056]
iii) The D layer designer selects a related table to be linked to the code table from the related table list via the input / output device 41.
[0057]
iv) The search key of the selected related table is automatically set as the search key of the Lv2 method. In the example of FIG. 12, “employee table” is selected as a table to be linked to the “department” table, and a staff number, which is a key of the “employee” table, is automatically set as a search key. .
[0058]
v) In response to the selection of the related table in iii), all the column names of the code table and the selected related table are displayed in the column list.
[0059]
vi) The D layer designer selects the column to be referred to by the Lv2 method via the input / output device 41. In the example of FIG. 8, “name”, “affiliation code”, “score”, and “subject name” are displayed as the column list, and “name” is selected.
[0060]
Subsequently, the D layer designer also inputs the Japanese name of the Lv2 method and the method description. In the example of FIG. 8, “acquire employee name” is entered as the method Japanese name, and “acquire employee name” is entered as the method description.
[0061]
vii) The method information 46 defined by the Lv2 method definition unit 422 is stored in the method information DB 45.
[0062]
○ Setting process by Lv3 method setting unit 423
The Lv3 method setting unit 423 operates according to the flowchart shown in FIG. A screen display example of the Lv3 method definition unit 423 is shown in FIG.
[0063]
i) The Lv3 method setting unit acquires the selected child table name from the D layer interface display unit 43. In the example of FIG. 13, the “attendance history” table is selected as the child table.
[0064]
ii) The Lv3 method definition unit 423 reads the parent table information related to the child table from the ER diagram 12 which is a product of the DB design unit 1, using the child table name as a key, creates display data, and inputs / outputs 41 To display. Here, the table information is a parent table list, each parent table search key, and a column name of each parent table. In the example of FIG. 13, “employee table” and “education subject table” are displayed as parent tables of the “attendance history” table.
[0065]
iii) The D layer designer selects a parent table to be linked to the child table from the parent table list via the input / output device 41.
[0066]
iv) In response to the selection of the parent table in iii), all the search keys of the child table and the selected parent table are displayed as search key candidates in the search key list. In the example of FIG. 13, the “employee table” and “education subject” table are selected as the tables to be linked to the “attendance history” table. As a result, the search key candidate is “staff number: subject code: subcode”. “Staff number” and “Course code” are displayed.
[0067]
v) In response to the selection of the parent table in iii), the column names of the child table and the selected parent table are displayed in the display item list.
[0068]
vi) The D layer designer selects, from the search key candidate list, one to be set as the search key of the Lv3 method from the search key list via the input / output device. In the example of FIG. 13, “Staff number: Subject code: Subcode” is selected as the search key.
[0069]
Subsequently, the D layer designer selects a column name to be acquired by the Lv3 method from the column candidate list. In the example of FIG. 13, “name”, “affiliation code”, “score”, “subject name” are displayed as the display item list, and “name”, “affiliation code”, “score”, “subject name” are selected. .
[0070]
Furthermore, the D layer designer also inputs the Japanese name of the Lv3 method and the method description. In the example of FIG. 13, “Acquired courses” is entered as the method Japanese name, and “Acquired courses are entered” as the method description.
[0071]
vii) The method information 46 defined by the Lv3 method definition unit 423 is stored in the method information DB 45.
[0072]
b) When deleting the D layer new method candidate, the D layer new method selected from the D layer new method creation request notification 7 is selected.
[0073]
(A2) When the D layer new method creation request notification 7 does not exist, definition processing by the Lv1, 2, 3 method definition unit 42 is performed.
[0074]
The D layer designer selects a table to be defined from the Lv1 table list displayed on the D layer interface display unit 43 and calls the method definition unit 42 at the level to be created. When a new Lv2 method is defined, a code table is selected from the Lv1 table list displayed on the D layer interface display unit 43 and the Lv2 method definition unit 422 is called. When a new Lv3 method is defined, a table that is “child” or “parent or child” is selected from the Lv1 table list displayed in the D layer interface display unit 43 and the Lv3 method definition unit 423 is called.
[0075]
(B) The method information stored in the method information storage DB 45 is acquired, display data is created, and displayed on the output device 41. A display example is shown in FIG.
(3) Specification generation
After completing the D layer method definition, the D layer designer confirms the contents of the Lv1 to Lv3 method information displayed on the D layer interface display unit 43 and instructs the specification generation unit 44 to create the D layer interface specification. put out. Upon receiving the instruction to create the D layer interface specification, the specification generation unit 44 reads the method definition information from the method information DB 45 and generates the D layer interface specification. In addition, after completing the creation of the D layer interface specification, the specification creation unit 44 creates the F layer design request notification 6 and sends it to the component design unit 5. The specification generation unit 44 operates according to the flowchart shown in FIG.
[0076]
As described above, the D layer component design unit 4 pays attention to the connection relation between the tables based on the table information of the database, performs leveling for identifying the table candidates to be accessed, and sets objects corresponding to the levels. By providing the method extraction criteria, it is possible to extract the class of the D layer component without depending on the skill of the designer.
[0077]
Next, through the input / output device 51, from the F layer designer, (1) display mapping information, (2) method mapping process (method definition (D layer addition, inside F layer, P layer)), (3) F layer Public Method Definition (4) The basic operation of the F layer component design unit 5 when an instruction to create an F layer interface specification and a functional specification is issued will be described.
[0078]
The F layer component design unit 5 performs the following.
(1) Verify whether each work is feasible by mapping to a use case scenario using the basic functions provided by the D layer component design unit 4. An example of the mapping process is shown in FIG. As shown in FIG. 40, deficient functions can be identified during D layer basic function mapping.
(2) A business-specific DB search method that cannot be supplemented only by the basic functions provided by the D layer is identified as a D layer additional function. An example of the mapping process is shown in FIG. In FIG. 41, the functions related to DB access among the functions that were insufficient at the time of FIG. 40 are extracted as the D layer new method.
(3) A process for processing a result obtained by the D layer providing function and a process in which a plurality of D layer accesses are performed are considered to be F layer method candidates and are mapped with a Japanese name. At this time, processing for replacing the D layer providing functions together with the F layer method occurs. In other words, the hierarchical structure is such that the F layer method is higher and the D layer method is lower. An example of the mapping process is shown in FIG.
(4) Processes such as input check and input / output editing that do not involve DB access are mapped to use case scenarios as P-layer processes. At the same time, the P layer screen is mapped. An example of the mapping process is shown in FIG.
(5) Finally, define which part on the use case scenario is the public method of the F layer. An example of the mapping process is shown in FIG.
F layer class is extracted.
[0079]
A flowchart of F layer class extraction in the F layer component design unit 5 is shown in FIG.
(1) Mapping information display section 54
When the F layer component design unit 5 receives a mapping information display instruction from the input / output device 51, the mapping information display unit display unit 54 performs the following processing. The mapping information display unit 54 operates according to the flowchart shown in FIG.
[0080]
Depending on whether (1) the F layer component design is (A) the first time, (B) the redo after receiving change of use case scenario information from the business design unit 2, or (C) the second time or later in the mapping information display unit 54 The use case scenario reading process is performed below.
[0081]
(A) First F layer component design
A use case scenario 32 that is a product of the business design unit 3 is read. A screen example of the mapping information display unit is shown in FIG. In the example of FIG. 26, the use case scenario “Application for education attendance (outside recruitment period, date designation)” is displayed.
[0082]
(B) Redoing D-layer component design after use case scenario change Only the changed part of the use case scenario 32, which is a product of the business design unit 3, is read.
[0083]
(C) D layer component design for the second and subsequent times
The already mapped use case scenario is read from the method mapping information storage DB 57.
[0084]
(2) Display data is created based on the read use case scenario and displayed on the input / output device 51.
(2) Mapping process
(1) The mapping information display unit 54 selects a use case scenario row to be mapped and calls the mapping unit 53. The mapping unit 53 operates according to the flowchart shown in FIG. A display example of the mapping unit is shown in FIG. In the example of FIG. 27, the line “6.3 Check the check items in the subject display line and press the application button when applying” is selected.
[0085]
(2) In the mapping part, (A) D layer existing method mapping process, (B) Method definition process by D layer new method definition part, (C) F layer internal method definition process by F layer internal method definition part, (D ) One of the P layer addition processes by the P layer addition unit is performed, and the method mapping information 59 is created.
[0086]
(A) D layer existing method mapping process by D layer method mapping unit 531
The D layer method mapping unit 531 operates according to the flowchart shown in FIG. FIG. 28 shows a screen example of the D layer method mapping unit.
[0087]
i) The already created D layer method information is read from the method information storage DB 45 of the D layer component design unit 4.
[0088]
ii) The D layer method information newly defined by the D layer new method definition unit 532 is read from the new method information DB 58.
[0089]
iii) Select a D layer method to be mapped to the use case scenario from the input / output device 51.
[0090]
iv) Create method mapping information 59 and output it to the mapping unit 53. In the example of FIG. 27, “acquisition of education subject application criteria” and “acquisition of applicant's education history information” methods are mapped to use case scenarios.
[0091]
(B) Method definition processing by D layer new method definition part
The layer D new method definition unit operates according to the flowchart shown in FIG. FIG. 29 shows a screen example of the D layer new method definition section.
[0092]
i) The already created D layer method information is read from the method information storage DB 45 of the D layer component design unit 4.
[0093]
ii) When there is a change in the Lv1 method, the Lv1 method definition unit 421 is called. The Lv1 method definition unit 421 performs the same operation as the Lv1 method definition unit of the D layer component design unit 4.
[0094]
When a new Lv2 method is defined, a code table is selected from the Lv1 table list displayed in the D layer new method definition section, and the Lv2 method definition section 422 is called. The Lv2 method definition unit 422 uses the Lv2 method definition unit of the D layer component design unit 4.
[0095]
When a new Lv3 method is defined, a child table is selected from the Lv1 table list displayed in the D layer new method definition section, and the Lv3 method definition section 423 is called. The Lv3 method definition unit 423 performs the same operation as the Lv3 method definition unit of the D layer component design unit 4.
[0096]
iii) The Lv1,2,3 method information 5a newly created in ii) is stored in the new method information DB 58. Here, the content stored in the new method information DB 58 is the content of the new method information storage table shown in FIG.
[0097]
iv) Create method mapping information 59 and output it to the mapping unit 53.
[0098]
(C) F layer internal method definition processing by F layer internal method definition unit 533
The F layer internal method definition unit 533 operates according to the flowchart shown in FIG. FIG. 30 shows a display example of the F layer internal method definition section.
[0099]
The F layer internal method definition unit 533 i) reads the D layer method information selected on the mapping screen from the mapping screen. If the D layer method is not selected, the information that it is not selected is read. ii) On the F layer internal method definition screen, input the Japanese name of the method, the argument type, the return value type, and the method description from the input / output device 51. In the example of FIG. 30, “Education application impossible determination” is used for the Japanese name of the method, “String” is used for the argument type, “boolean” is the return type, and “Education application is not determined” is described for the method. You are typing. iii) Creation of F layer internal methods includes a) collecting D layer methods, b) performing business processing, and determining which one to create depends on whether the D layer method is selected or not Then, each processing is performed.
[0100]
a) F layer internal method that summarizes D layer methods (D layer method is selected) The method hierarchy level of the selected D layer method is increased by 1, and the contents of the method mapping information storage DB are updated.
[0101]
b) F layer internal method for business processing
Do nothing in particular.
[0102]
iv) The method information 5b created in ii) is stored in the new method information DB 58.
v) Create method mapping information 59 and output it to the mapping unit 53.
[0103]
(D) P layer addition processing by the P layer addition unit 534
The P layer adding unit 534 operates according to the flowchart shown in FIG.
The P layer addition unit 534 selects whether to map the P layer screen or add the P layer processing to the use case scenario, and depending on the selection, the P layer screen addition unit 5341 and the P layer processing addition unit 5342 Call.
[0104]
a) P layer screen addition processing by the P layer screen addition unit 5341
The P layer screen adding unit 5341 operates according to the flowchart shown in FIG. A display example of the P layer screen addition unit 5341 is shown in FIG.
[0105]
i) The screen list information is read from the screen list 22 which is a product of the screen / form design unit 2. In the example of FIG. 31, an “application acceptance screen” and an “education guidance screen” are displayed.
[0106]
ii) From the screen list displayed on the P layer screen addition screen, select a screen to be mapped to the use case scenario. In the example of FIG. 31, the “education guidance screen” is selected.
[0107]
Further, supplementary items for the screen added from the input / output device 51 are input. In the example of FIG. 31, “output a message notifying that the required education subject is full” is input.
[0108]
iii) Create and output method mapping information 59.
[0109]
b) P layer processing addition processing by the P layer processing addition unit 5342
The P layer processing adding unit 5342 operates according to the flowchart shown in FIG. A display example of the P layer processing addition unit 5342 is shown in FIG.
[0110]
i) From the input / output device 51, input the name of the process to be performed in the P layer and the description of the process. In the example of FIG. 32, “Check if it is within the holding period” is input as the process name, and “Check whether it is within the education holding period” is input to the description of the process.
[0111]
ii) The P layer processing information 5b input in i) is stored in the new method information DB 58.
[0112]
iii) Method mapping information 59 is created and output to the mapping unit 53.
[0113]
(3) The method mapping information 59 created in (2) is stored in the method mapping DB 57. Here, the content stored in the method mapping DB 57 is the content of the method mapping information storage table shown in FIG.
[0114]
(4) When changing the execution order of the mapped methods, the contents of the execution order of the method mapping information 59 stored in the method mapping DB 57 are changed.
[0115]
(5) Display data is created based on the method mapping information 59 stored in the method mapping DB 57 and displayed on the output device 51.
[0116]
(3) F layer public method definition
From the mapping information display unit 54, a method, a screen, and a process in which the range of the use case scenario to be defined as the F layer public method is mapped are selected, and the F layer public method definition unit 55 is called. The F layer public method definition unit operates according to the flowchart shown in FIG.
[0117]
The F layer public method definition unit 55 uses the same screen as the (3) F layer internal method definition unit 533 of (2).
[0118]
i) Input the Japanese name of the method, argument type, return value type, and method description from the input / output device.
[0119]
ii) The F layer public method setting information 5b input in i) is stored in the new method information DB 58.
[0120]
iii) Regarding the method mapping information 59 in the range corresponding to the selected use case scenario stored in the method mapping information DB 57, the contents of the F layer public method identifier and the F layer public method name are updated.
[0121]
iv) Display data is created based on the method mapping information 59 stored in the method mapping DB 57 and displayed on the output device 51. A display example is shown in FIG.
(4) Specification generation
After completing the method mapping definition, the F layer designer confirms the contents of the method mapping information displayed on the method mapping information display unit 54 and instructs the specification generation unit 56 to create the F layer interface specification and functional specification. put out. Upon receipt of the F layer interface specification and the functional specification, the specification generation unit 56 reads the method definition information 5a / 5b from the new method information DB 58 and the method mapping information 59 from the method mapping information DB 57. Generate specifications and functional specifications. Further, the specification generation unit 56 generates a D layer new method creation request notification 7 after the specification is completed, and sends it to the D layer component design unit 4.
[0122]
At this time, the data related to the D layer new method stored in the new method information DB 58 is deleted.
[0123]
The specification generation unit 54 operates according to the flowchart shown in FIG.
[0124]
As described above, the F layer component design unit 5 verifies whether each work of the use case scenario can be realized by using the interface of the D layer component obtained from the data-centric approach and the screen / form list. By displaying the verified result in combination with the use case scenario, it is possible to identify the insufficient function and extract the class of the F layer component.
[0125]
In addition, each of the D layer component design unit and the F layer component design unit has a mechanism for automatically transmitting a design end notification upon completion of the design. By this mechanism, it is possible to simplify the cooperation between the D layer designer and the F layer designer, and to avoid problems such as inconsistencies in the interface between components due to human error.
[0126]
【The invention's effect】
The following advantages can be obtained by the present invention.
1. By developing using this design procedure, application development that adopts a three-tier architecture can be performed without the developer being particularly conscious.
2. By mechanizing component design, it is possible to prevent variations in the granularity and quality of objects and methods.
3. By defining the D layer class from the DB table information in a bottom-up manner, creation of a method having a useless function can be prevented.
4). By associating the D layer interface with the use case scenario, it is possible to identify missing functions without fail.
5. By using this design procedure and tool, necessary design deliverables such as D layer interface specifications and functional specifications can be created while actually designing.
[Brief description of the drawings]
FIG. 1 System overview diagram
FIG. 2 is a basic flowchart of the D layer designer in the D layer component design unit.
FIG. 3 is a flowchart of the Lv1, 2, 3 method definition unit in the D layer component design unit.
FIG. 4 is a flowchart of the D layer interface display unit in the D layer component design unit.
FIG. 5 is a flowchart of the Lv1 method definition unit in the D layer component design unit.
FIG. 6 is a flowchart of the Lv2 method definition unit in the D layer component design unit.
FIG. 7 is a flowchart showing processing of an Lv2 method definition unit 422
FIG. 8 is a flowchart of the Lv3 method definition unit in the D layer component design unit.
FIG. 9 is a screen example of a D layer interface display unit in the D layer component design unit;
FIG. 10 is a screen example of a new D layer method definition selection screen in the D layer component design unit.
FIG. 11 is a screen example of the Lv1 method definition section in the D layer component design section.
FIG. 12 is a screen example of the Lv2 method definition unit in the D layer component design unit
FIG. 13 is a screen example of the Lv3 method definition section in the D layer component design section.
FIG. 14 is a table example stored in the method information DB in the D layer component design unit;
FIG. 15 is an F layer designer basic flowchart in the F layer component design unit;
FIG. 16 is a mapping information display unit flowchart in the F layer component design unit;
FIG. 17 is a flowchart of a specification generation unit in the F layer component design unit.
FIG. 18 is a flowchart of a mapping unit in the F layer component design unit.
FIG. 19 is a flowchart of the D layer method mapping unit in the F layer component design unit;
FIG. 20 is a flowchart of the D layer new method definition unit in the F layer component design unit.
FIG. 21 is a flowchart of the F layer internal method definition unit in the F layer component design unit.
FIG. 22 is a flowchart of the P layer addition unit in the F layer component design unit.
FIG. 23 is a flowchart of the P layer screen adding unit in the F layer component design unit.
FIG. 24 is a flowchart of the P layer processing addition unit in the F layer component design unit.
FIG. 25 is a flowchart of the F layer disclosure method definition unit in the F layer component design unit.
FIG. 26 is a screen example of the mapping information display section in the F layer component design section (initial state)
FIG. 27 is a screen example of a mapping unit in the F layer component design unit
FIG. 28 is a screen example of the D layer method mapping unit in the F layer component design unit.
FIG. 29 is a screen example of the D layer new method definition section in the F layer component design section.
FIG. 30 is a screen example of an F layer internal method definition section in the F layer component design section.
FIG. 31 is a screen example of the P layer screen adding unit in the F layer component design unit.
FIG. 32 is a screen example of the P layer processing addition unit in the F layer component design unit
FIG. 33 is a screen example of the mapping information display section in the F layer component design section (after defining the F layer disclosure method).
FIG. 34 is an example of a table stored in a new method information DB in the F layer component design unit.
FIG. 35 is a table example stored in a method mapping information DB in the F layer component design unit;
FIG. 36: List of DB tables used in “Education Application & Qualification System”
FIG. 37: Join access example with code table
FIG. 38: Join access example of a table having a parent-child relationship (search by child table key)
FIG. 39: Join access example of a table having a parent-child relationship (search by key of parent table)
FIG. 40: Flow of interface extraction of F layer component (D layer basic function mapping)
FIG. 41: F layer component interface extraction flow (D layer additional function identification)
FIG. 42: Flow of interface extraction of F layer component (F layer internal method identification)
FIG. 43: Flow of interface extraction of F layer component (Finding P layer processing + Mapping of P layer screen)
Fig. 44 Flow of F layer component interface extraction (definition of F layer public method)
[Explanation of symbols]
1 DB design department
11 Input / output devices
12 ER diagram
13 Table information
2 Screen / Form Design Department
21 I / O devices
22 Screen / Form List
3 Business Design Department
31 I / O devices
32 Use Case Scenario
4 Database Access (D layer) Component Design Department
41 I / O devices
42 Lv1,2,3 method definition part
421 Lv1 method definition part
422 Lv2 method definition part
423 Lv3 method definition part
43 D layer interface display
44 Specification Generation Department
45 Method information storage DB
46 Method definition information
5 Function (F layer) design department
51 I / O devices
52 New Creation Department
53 Mapping part
54 Mapping information display section
55 F layer public method definition part
56 Specification Generation Department
57 Method mapping information storage DB
58 New method information storage DB
59 Method mapping information
5a D layer method definition information
5b F layer method definition information
6 F layer design notification request
7 D layer new method creation request notification

Claims (3)

During the design phase of application development,
The database access component design department and the business component design department are divided into two stages.
A class design method for extracting an interface by linking the database access component design unit and the business component design unit. In the class design method according to claim 1,
The database access component design department sets a level in the table to be accessed and extracts object candidates necessary for the application. In the class design method according to claim 2,
Focusing on all table and code table before join in table level of database, focusing on table and child table created after join of related table and code table, table created after join with table in parent-child relationship Set to
A class design method characterized by extracting a database access component interface by providing a method extraction criterion according to a set table level

Images