JP2008117095A - Method of developing software - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem of conventional development requiring source programs to be modified every time source programs of other previous projects are analyzed and used at a source level for a reduction in the man hours of development, due to the fact that use of the other programs at the source level is highly dependent on platforms, and the problem of difficulty in applying the programs due to the fact that the programs are used from the standpoint of the source. <P>SOLUTION: A method that makes it difficult to depend on platforms is provided by avoiding the use of other programs at the source level, and using the business rules required to create sources. More specifically, business rules are extracted in advance and stored as templates, parts containers, and parts. To develop another system, the stored templates, parts containers and parts are used so that source outputs become usable from the standpoint of the business rules. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a technique for efficiently developing software. In particular, it relates to technology for developing software for new projects by diverting sources from other projects.

  Conventionally, development effort has been reduced by diverting sources from other projects in software development.

  In particular, Patent Document 1 discloses that new software is developed by reusing existing software in order to improve software productivity (software diversion). Here, it is described that a programmer analyzes and corrects a document or source program describing specifications of existing software.

Japanese Patent No. 3186117

  However, the above-mentioned diversion at the source level, which is an analysis of the source program, has a high dependency on the platform, and the source program has to be modified each time.

  Therefore, in the present invention, business rules are extracted from the specifications, and software is created using them. More specifically, the following aspects are included in the present invention.

  Software that should be created from the function matrix by applying XML processing to the function matrix created by the user and associated with the template file of the template dictionary using the business pattern as a key and the dictionary showing the contents of each item A template corresponding to the business rule of the wear is extracted, a business component corresponding to the extracted template is selected, and a software specification to be generated is created using the selected business component. In addition, when the work to be generated includes a correlation check logic, this logic may be performed using a correlation check matrix. In the case of a program that requires exclusive control, it is possible to generate a program that holds exclusive information by incorporating exclusive information (access control) into the function matrix. That is, the present invention includes extracting the parent-child relationship of each table from a functional design document such as an ER diagram and implementing exclusive control based on this. More specifically, the following aspects are also included.

  Software that should be created from the function matrix by applying XML processing to the function matrix created by the user and associated with the template file of the template dictionary using the business pattern as a key and the dictionary showing the contents of each item Software that extracts a template corresponding to the business rule of the wear, selects a business component corresponding to the extracted template, and creates a software specification to be generated using the selected business component In the software development method, functional design document data indicating the function of the software is stored, access restrictions on the software table are defined based on the functional design document data, and the defined access restrictions are defined. Based on the software in the development. Software development how to.

  Further, in this software development method, the table is a table including a record that should be restricted in access, and the access restriction corresponds to a restriction on access to the table including the record. A software development method that is defined by specifying a table that should be restricted is also an aspect of the present invention.

  Further, in this software development method, the specification of the table is realized by defining the order in which access is restricted between the tables. The software development method is also included in the present invention.

  In the software development method, the access restriction is performed by extracting a parent-child relationship indicating a hierarchical relationship between the tables from the functional design document data, and defining the access restriction based on the extracted parent-child relationship. A software development method characterized by this is also included in the present invention.

  Furthermore, in these software development methods, the access restriction includes at least one of suppression of access to the table itself, suppression of change of the table, and alarm output for change of the table. The software development method is also included in the present invention.

  Note that an apparatus, a computer program, and software developed by this development method for realizing this software development method are also included in the present invention.

  According to the present invention, it is possible to easily generate new software (source).

  Embodiments of the present invention will be described with reference to the drawings. In the present embodiment, software product mediation is described as an example, but the present invention is not limited to this.

  First, FIG. 1 shows a system configuration diagram in the present embodiment. The computer has a processing device such as a memory (101), an external storage device (105) including a hard disk, and a CPU (102), and the processing device executes information processing according to a program stored in the storage device. . The external storage device (105) stores a document group including a function matrix (214) shown in FIG. 2, a dictionary, and a tool group.

  A schematic diagram of this function is shown in FIG. 2, and the outline will be described below.

  The source (222) and program specification (221) output by this function are composed of a business template (225), a part container (223) divided into functional units, and a part (224) that is part of the part container. Is done. Each of these templates (225), parts container (223), and parts (224) has variable parts. By embedding dictionary items (205), a source (222) and a program specification (221) are created. To do. There are a plurality of parts (224) that can be used in the part container (223), and the function of the part container (223) can be customized by selecting the part (224).

  The functions used in the present invention are mainly composed of (1) a matrix XML conversion tool, (2) a dictionary XML conversion tool, (3) a source, and a document conversion engine.

  Details of each function are shown below.

(1) Matrix XML conversion tool (216)
FIG. 3 shows a schematic diagram.

  This is a function (216) that converts the function matrix (214) and the matrix component linking parameter file (215) (hereinafter referred to as parameter file) into XML in a form that retains the structure and meaning of the program.

  Details of (a) function matrix (214) and (b) parameter file are shown below.

(A) Function matrix (214)
The function matrix (214) maps to the template file (225) of the template dictionary (207) using the business pattern as a key. (The template file in the dictionary can be specified using the business pattern name as a key)
Also, the component container (223) and component (224) of the business component dictionary (206) are logically mapped using the container name and component name as keys. (By passing through the parameter file (215), the component container and component in the business component dictionary (206) can be specified.) The function matrix (214) includes each component container (223) and component ( 224) and a keyword for acquiring a keyword attribute from the parameter file (215). When the content of the keyword attribute refers to a dictionary, the value input here is a key for acquiring the dictionary value. (See (b) for keyword attribute descriptions)
(B) Parameter file (215)
The parameter file (215) maps to the component container file (223) and the component file (224) in the business component dictionary (206) using the component container name and the component name as keys. (The file names of the component container (223), the component (224), the component container (223), and the component dictionary (206) of the function matrix (214) are linked.) From this, the parameter file (215) Depending on the description, it is possible to select the source (221) or the program specification (222). That is, if the document part container (208) is specified as the part container name in the parameter file (215), the program specification (222) is output, and if the source part container (209) is specified, the source (221) is output.

  The parameter file (215) also holds keyword attributes that determine whether the data item name in the dictionary (205) is acquired using the value for the keyword as a key, or whether the value of the function matrix (214) itself is reflected in the component. is doing.

(2) Dictionary XML conversion tool (218)
FIG. 4 shows a schematic diagram.

  Function to extract dictionary (205) items (item names, attributes, domains, etc.) into dictionary XML files (219) that can be deciphered as templates that extract only the items necessary for source output and retain the structure and meaning (218). The dictionary in the form of an XML file holds a “data item name” (physical name) on the source that is obtained using the above value (logical name) as a key. In attribute check, attribute information which is key information for selecting a component is held. For the flow of processing when referring to the dictionary (205), see Step 507, Step 510, and Step 513 in FIG.

(3) Source and document conversion engine (220)
A function that selects a target template (225), a part container (224), and a part (223) from the function matrix XML (217), and embeds dictionary items in parameterized locations (source and program specification output function). The processing flow of this function is shown in step 519 of FIG. 5 described later.

  Next, details of the processing contents of the present embodiment will be described with reference to FIG.

In step 501, the function matrix (214) is converted into XML (217) with reference to the parameter file (215) by the matrix XML conversion tool (216). (See below for details on creation of function matrix (214) and conversion to XML.)
In step 502, the dictionary (205) registered in advance is converted into XML by the dictionary XML conversion tool (218).

  In step 503, the function matrix XML (217) created in step 501 is read.

  In step 504, the target template (225) is selected from the read function matrix XML (217).

  In step 505, the component container (223) associated with the target template (225) is selected from the read function matrix XML (217). This process is repeated for the number of component containers defined in the matrix XML.

  In step 506, the selected parts container (223) is inserted into the template (225).

  In step 507, the parameterized part in the part container (223) is acquired from the dictionary XML (219) and set.

  In step 508, the part (224) associated with the target part container is selected from the read function matrix XML (217). This process is repeated for the number of parts defined in the matrix XML.

  In step 509, it is determined whether the part selected in step 508 is an attribute check.

  In step 510, if the attribute check is determined in step 509, the attribute (single-byte alphanumeric characters, kanji, etc.) of the target item is specified from the dictionary XML (219).

  In step 511, the part (224) corresponding to the attribute is selected from the part dictionary (206).

  In step 512, the selected part (224) is inserted into the part container (223).

  In step 513, the part (224) whose item name is parameterized is obtained from the dictionary XML (219) and set.

  In step 514, it is determined whether or not there is a part (224) in the target part container (223).

  In step 515, it is determined whether or not there is a parts container (223) in the target template (225).

  In step 516, the source (222) or the program specification (221) is output as a combined template (225) of the part (224) and the part container (223). Whether the source (222) or the program specification (221) is output depends on the setting of the property file (215).

  Details of the creation of the function matrix (214) described above and conversion to XML will be described below with reference to FIGS. 6 and 7 to 20. FIG.

(1) Function matrix creation procedure (607)
In step 601, the developer sets the template name to be used in the business pattern on the function matrix (301) based on the template specification (202). Figure 7 shows an overview of the creation.

  In step 602, other items / control information (contents such as language and module name) commonly used in the source (222) and the program specification (221) are set in the program information (301). Figure 8 shows an outline of the creation.

  In step 603, a part container (223) to be used in the template (225) is selected from the template specification (202) and an item number is assigned. Note that this item number is used when referring to the output of the part container (223) in another part container (223) and part (224), and is unique on the part container (223). Number so that it becomes. There are no rules for numbering. Figure 9 shows an overview of the creation.

  In step 604, a part (224) to be used in the part container (223) is selected from the part container specification (203) and an item number is assigned. This item number is for use when referring to the output of the part (224), and is numbered so as to be unique on the part (224). There are no rules for numbering. Figure 10 shows an overview of the creation.

In step 605, a keyword that can be set for the part (224) is entered with reference to the part specification (204) for the part (224). The keywords are expressed in words that prompt the “value” entered by the user in step 606. An outline of the creation is shown in FIG.
Example) Keyword: Item to be checked Value: Date of birth In step 606, a value is set for the keyword, and the value refers to the output of another parts container (223) or part (224). Sets the reference item number. For the reference item number, when referring to a value from other than the target component (224) and component container (223), set the item number of the reference component (224) and set the reference keyword as the value. To do. An outline of the creation is shown in FIG.

  FIG. 13 shows a function matrix (214) created by executing Step 601 to Step 602.

(2) Function matrix XML conversion procedure (517)
In step 1401, the program information (301) of the function matrix (217) is set in the tag of the function matrix XML (1501) corresponding to the item name. Figure 15 shows an overview of XML creation.

  In step 1402, the container name of the matrix information (302) of the function matrix is set in the container (name). At that time, the item number of the parts container (223) is set in the container (ID). For the container (file), the component container file name of the dictionary (206) to be referenced is acquired from the parameter file (215) using the container name of the matrix information (302) as a key and set. Figure 16 shows an overview of XML creation.

  In step 1403, when another component (224) uses the output of the component container (223) for input, the keyword of the matrix information (302) is set to the container keyword. As set contents, the keyword of the target part (224) is set to the container keyword name (name), and the item number of the target part (224) is set to the container keyword name (refid). At that time, the keyword attribute of the parameter file (215) is acquired using the keyword of the target component (224) as a key, and set to the container keyword name (attr). Conversely, when the output of the component container (223) is used for another component (224), the item number of the component (224) is set in the container keyword name (setrefid). An outline of XML creation is shown in FIG.

  In step 1404, the part name of the matrix information (302) of the function matrix is set in the part (name). At that time, the item number of the part (224) is set in the container name (ID). For the container name (file), the part file name of the dictionary (206) to be referred is acquired from the parameter file (215) using the part name of the matrix information (302) as a key, and set. Figure 18 shows an overview of XML creation.

  In step 1405, when the component (224) uses the output of the other component (224) for input, the keyword of the matrix information (302) is set as a keyword. As set contents, the keyword of the target part (224) is set to the keyword name (name), and the item number of the target part (224) is set to the keyword name (refid). At that time, the keyword attribute of the parameter file (215) is acquired using the keyword of the target part (224) as a key, and set to the keyword name (attr). Conversely, when the output of the component (224) is used for another component (224), the item number of the component (224) is set in the keyword name (setrefid). An overview of XML creation is shown in FIG.

  FIG. 20 shows a function matrix XML (217) created by executing Step 1401 to Step 1405.

  The following (1) to (4) show the effects of the present embodiment.

  (1) By filling in the function matrix (214), the source (222) is generated, so that the program can be generated without depending on the language characteristics.

  (2) Since the program is created based on the template (225), the part container (223), and the part (224), there is no variation in programming ability among developers, and the quality of the output source (222) is constant.

  (3) When diverting business functions, a system can be constructed only by changing item information (dictionary) unique to each system. The specific reason is shown below.

  When converting from a functional specification (201) to a source (222), in this system, the developer classifies existing functional specifications into large, medium, and small categories, and corresponding templates (225) and parts containers (223). ), Select the part (224). Classification guidelines and methods are determined based on the information defined in the template specifications (202), parts container specifications (203), and parts specifications (204) classified according to the size of the business function. I do. Also, these specifications hold physical association information for each category of each dictionary (205). Therefore, each business function, the source (222) such as a template, and the component (224) when outputting the program specification (221) are mapped in a one-to-one relationship. FIG. 21 shows a schematic diagram.

As a result, when developing other systems, by diverting from the viewpoint of this business function, it becomes possible to divert the component (224) used when outputting the source (222) such as a template. In addition, the business function itself can be diverted even when the environment in the language or platform changes, and the template (224) such as the template can be diverted by modifying the template associated with the business function. Become.
By extracting and accumulating these business functions in advance, a system can be constructed only by changing item information (dictionary) unique to each system.

(4) By preparing the function matrix (214) and the dictionary (205), it is possible to output both the source (222) and the program specification (221), which greatly reduces man-hours (design process) And the program process is completed at the same time).
As described above, depending on the setting of the parameter file (215), it is possible to change which file of the component dictionary (206) is mapped, so the source (222) and the program specification (221) are separated. Can be output.

  Next, FIG. 22 shows a schematic diagram of this function when using the correlation check matrix, and the outline will be described below.

  This function is a function for creating a correlation check logic part independently of the function using the function matrix described above. Therefore, the source (222) and program specification (221) output by this function are partial to the output output by the above function. Information for associating with other information may be output to the source (222) and the program specification (221).

  The input information for generating the source (222) and the program specification (221) differs between the function matrix XML (217) and the correlation check matrix XML (2103), but the generation process (outputs XML from the matrix). The process up to is the same.

  Here, the additional function to be used is a correlation check matrix XML conversion tool.

  FIG. 23 shows the contents.

  The correlation check matrix XML creation tool is a function (216) that converts the XML into a form that retains the structure and meaning of the program specialized in the correlation check logic based on the correlation check matrix (214) and the parameter file (215). .

  This correlation check matrix (214) includes (a) a condition part (2301), (b) a processing part, and (c) a condition tree definition part. (The function matrix is composed only of the processing unit.) Each detail is shown below.

(I) Condition part (2301)
The condition part (2301) is a part that defines a branch condition for an item. The correlation check component is a target item unit of the condition. Furthermore, the item is divided into branch condition units, and node IDs are assigned. Thereby, the branch condition is identified.

There are two types of condition definitions: a method for specifying a condition for a return value processed by a correlation check component for an input value, and a method for specifying an input item itself as a condition. (For example, in the case of the former, a method for determining whether the date is an actual date or not with respect to an input date, and using a component to check whether the date is a return value of the component. In the latter case, the entered date is compared with any date)
Hereinafter, the former is referred to as “return value correlation check” and the latter as “input item correlation check”.

  When the condition for the return value of the correlation check component is specified, the “name” of the correlation check component to be set is mapped to the component (224) of the business component dictionary (206). The mapping method is the same as the function matrix described above. The mapping may be associated without going through the parameter file.

(B) Processing section (2303)
This is a part that defines the processing content when the condition defined in the condition part (2301) is achieved. Similar to the condition part (2301) in (a), the “name” of the correlation check part is mapped to the part (224) in the business part dictionary (206). The mapping method is the same as the function matrix described above. The mapping may be associated without going through the parameter file.

(C) Condition tree section (2302)
This part defines the order of conditions (condition tree definition) and the processing order (processing tree definition) corresponding to the conditions. As shown in FIG. 5, in the condition tree part (2302), the node ID defined in the condition part (2301) is used, and the combination of conditions, the check order, and the processing order corresponding to the conditions are defined. By defining the node ID in the “parent node ID” field in the condition tree definition, a hierarchical structure (parent-child relationship) is provided between the conditions. Hereinafter, the priority condition among the defined conditions is referred to as “parent”. A condition attached to a parent condition is called a “child”.

  The parent-child relationship is expressed in a “node ID” field and a “parent node ID” field. Describe the node ID of the parent condition in the “parent node” column of the child condition. However, the “parent node” field of the condition to be judged first is “-”. The condition tree definition is processed in order from the condition tree definition 1, and is processed according to the evaluation order.

  Next, details of the processing contents of the present embodiment will be described with reference to FIG.

When the correlation check matrix is used, the correlation check matrix XML (2103) is compared with FIG.
Step 2404 and Step 2403 are added between Step 507 and Step 508 of Step 2404 and Step 519 of FIG. Details of step 2404 and details of step 2402 and step 2403 will be described below.

In step 2401, the correlation check matrix (2101) is converted into XML (2103) with reference to the parameter file (215) by the correlation check matrix XML conversion tool (2102). (For details on creating the correlation check matrix (2101) and converting it to XML, see below.)
In step 2403, it is determined in step 508 whether the component to be referenced is a correlation check component. As a judgment material, the correlation check ID is described in the part name of the function matrix (214). However, it is only necessary to be able to determine a correlation check component other than the correlation check ID.
Example) SOUKAN001
In step 2402, the correlation check matrix (2103) created in step 2401 is read.

  The processing after step 2402 is the same as that at the time of generation from the function matrix (214).

  Details of creation of the correlation check matrix (2101) described above and conversion to XML will be described below with reference to FIGS. 25 and 42. FIG.

(1) Correlation check matrix creation procedure In step 2501, the developer sets a correlation check ID in the correlation check matrix. An outline of the creation is shown in FIG. Regarding the correlation check ID, as shown above, if the correlation check ID associated with the function matrix is set uniquely, the ID numbering method is not specified. See FIG. 26 for an overview of the correlation check ID set. In FIG. 26, “SOUKAN002” is the correlation check ID (2601).

  In step 2502, it is selected whether the correlation check uses “return value correlation check” or “input item correlation check”. See above for specific examples.

  In step 2503, a part to be used is set based on the part container specification (203). As shown in Fig. 27, the content of "Null character check" specified in "Name" in the parts container specification (204) is the "Name" column in "Correlation check parts" in the condition definition part (2701). Stipulate. The ID of “correlation check component” may be unique among “condition part (2301)” and “processing part”.

  In step 2504, a keyword for the selected part is set based on the part specification (201). As shown in FIG. 28, the content of “keyword” defined in “keyword attribute to be used” of the part specification (204) is defined in the “keyword” column of the condition definition section (2701).

  In step 2505, a value for the keyword is set. For the value, set the target word in the dictionary. 29, the item value of “keyword” is defined in the column of “item value” from “passbook symbol” in the dictionary (205).

  In step 2506, an item name to be checked for correlation is set. As shown in FIG. 30, the content defines the “individual corporation category” in the dictionary (205) in the “item name” column.

  In step 2507, a condition value to be compared is set for the item name to be checked for correlation. As shown in FIG. 31, “HOUJIN” to be compared in step 2506 is defined in the “condition value” column.

  In step 2508, a comparison method is set. (Comparison identifier). As shown in FIG. 32, the contents define the method of comparing the contents entered in step 2506 and step 2507 in the “expression” column of the condition definition section (2701). In the case of FIG. 32, “eq” indicating an equal sign is set.

  In step 2509, a node ID is set so as to be unique for the correlation check logic. As shown in FIG. 33, the content defines “001” in the “node ID” column of the condition tree portion (2302). The ID may be unique on the matrix.

  In step 25010, a parent node ID associated with the node ID is set. A plurality of parent node IDs associated with the node ID may be designated. As shown in FIG. 34, the contents specify “001” for “Node ID” “004” in the “Parent node ID” column of the condition tree portion (2302). Note that an item whose “parent node” is “−” is a root, and subordinates thereof are “node ID: 004” and “node ID: 006”.

  In step 25011, the evaluation order is set in the same parent node ID. As shown in FIG. 35, “1” is defined for “node ID” “004” in the “evaluation order” column of the condition tree portion (2302). For items with “-” as the “parent node”, the route is fixed to “1”, and those with the same “parent node” are numbered sequentially from “1”.

  In step 25012, a part to be processed when “true” is set for the condition. As shown in Fig. 36, the content of "Add error message" specified in "Name" of the part specification (204) is changed from "Name" to "Correlation check part" of the condition definition part (2701). Specify in the column.

  In step 25013, a keyword is set for the part set in step 25012. As shown in FIG. 37, the content of “keyword” (“message ID”) specified in “keyword attribute to be used” in the part container specification (204) is specified in the “keyword” column of the processing unit, as shown in FIG. To do.

  In step 25014, a value for the keyword is set. As shown in FIG. 38, the message ID is defined in the “value” column for the keyword set in step 25013.

  In step 25015, the “node ID” set in step 25010 is set as the processing target ID in order to associate the condition in step 25010 with the processing in step 25011. As shown in FIG. 39, the content defines “node ID: 001” set in step 25010 in the “processing object ID” field of “condition tree definition (2302)” in the processing unit (2303).

  In step 25016, the processing order is set in the processing target ID. As shown in FIG. 40, the contents are numbered sequentially from “1” in the “processing order” column of the “condition tree definition” for the same “processing object ID”.

  A correlation check matrix created by the above steps is shown in FIG.

(2) Correlation check matrix XML conversion procedure In step 4301, the header portion (fixed value) of XML is set. See FIG. 43 for its contents.

  In step 4302, the correlation check ID of the correlation check matrix (2101) is set. As shown in FIG. 44, the correlation check ID in the “correlation check ID” column of the correlation check matrix (2101) is set in XML.

  In step 4303, the pattern ID of the condition definition tree definition in the condition part (2301) of the correlation check matrix (2101) is set. As shown in FIG. 45, the pattern number in the “condition tree definition” column of the “condition part (2301)” of the correlation check matrix (2101) is set on XML.

  In step 4304, the node ID, parent ID, and evaluation order of the condition definition tree definition of the correlation check matrix (2101) are set. As shown in FIG. 46, the node ID, parent node ID, and evaluation order (“001”, “−”, “1”) of the correlation check matrix (2101) are set on the XML.

  In step 4305, the parts of the condition part (2301) of the correlation check matrix (2101) are converted and set based on the parameter definition. The conversion process is the same as steps 1404 to 1405 shown in the procedure for creating the function matrix XML described above. As for the contents, as shown in FIG. 47, the part “add error message” is set in XML.

  In step 4306, the item name in the condition part (2301) of the correlation check matrix (2101) is set. As shown in FIG. 48, the value (return value. Personal corporation classification) in the “item name” column of the “condition part (2301)” of the correlation check matrix (2101) is set on XML.

  In step 4307, the equation in the condition part (2301) of the correlation check matrix (2101) is set. As shown in FIG. 49, “eq” defined by the “expression” of the correlation check matrix (2101) is set on the XML as shown in FIG.

  In step 4308, the condition value in the condition part (2301) of the correlation check matrix (2101) is set. As shown in FIG. 50, the value (“HOUJIN”) defined by the “condition value” of the “condition part (2301)” of the correlation check matrix (2101) is set in XML.

  In step 4309, as in step 4304, the processing target ID and processing order ID of the condition definition tree definition in the “processing unit” of the correlation check matrix (2101) are set. As shown in FIG. 51, the processing target ID and processing order (“006”, “1”) of the correlation check matrix (2101) are set on the XML.

  In step 4310, the parts of the condition part (2301) of the correlation check matrix (2101) are converted and set based on the parameter definition. The conversion process is the same as steps 1404 to 1405 described above. With regard to the contents, as shown in FIG. 52, the “error message addition” part is set in XML.

  FIG. 53 shows a correlation check matrix (2103) created by the above steps. In this correlation check matrix (2103), since the correlation check matrix as an input is added when the “condition part (2301)” and “condition tree” are compared with the function matrix, the output XML is also correspondingly increased. Information is added.

  Next, a method for realizing an exclusive control frame having exclusive control processing in the program will be described. The exclusive control frame specifies a table including a record to be accessed, thereby specifying a table (parent) to be locked and a lock order between the specified tables. For example, if table A (5404), table B (5409), and table C (5412) shown in FIG. 54 are to be updated, table X (5401), table Y (5406), and table Z (5410) to be locked ) (Each parent table). Then, the lock order between the parent tables specified in the order according to the specification is specified according to the functional design document.

  This processing is realized by adding a matrix part holding table lock data to the function matrix in the source generation method using the function matrix described above. Therefore, the flow of source generation after creation of the function matrix is the same. FIG. 55 shows the flow until exclusive control information is added to the function matrix.

  As shown in Fig. 55, this function is based on the premise that a document that defines the parent-child relationship between tables based on information from external design (5502) and ER diagram (5501) and parent table information definition has been created. Yes. The parent-child relationship definition is a summary of the parent-child relationship between tables in the DB environment used by the program in tabular form (5503) as shown in Fig. 56 based on the design. Similarly, lock priorities between parent tables are defined (5504) as shown in FIG. 57 based on the design. When the exclusive control frame matrix development support tool (5505) shown in FIG. 55 reads these documents, an exclusive control matrix having ordered table lock information can be created.

These will be described with examples.
As an example, assume that the source is generated in a DB environment as shown in FIG. Assume that table A (5404), table B (5409), and table C (5412) are updated this time. At this time, first, the parent table of each access target table is specified based on the parent-child relationship definition (5503) shown in FIG. In this case, table X (5401) (parent table of table A), table Y (5406) (parent table of table B), and table Z (5410) (parent table of table C). Next, the access order (lock order) of the table X (5401), table Y (5406), and table Z (5410) specified by the parent table information definition (5504) as shown in FIG. 57 is determined. By incorporating this information into the function matrix (214), a source having table lock information can be generated, and deadlock can be prevented.

  The exclusive control frame matrix development support tool implements this logic. The specific movement is shown in the flowchart 58.

In step 5801, the worker inputs the access target table name to the business frame support tool (5505) and presses the generation button. (Fig. 59)
In step 5802, the business frame support tool (5505) refers to the parent table information definition (5503) based on the table name input in step 5801.

In step 5803, the business frame support tool (5505) identifies the parent table name of each input table based on the reference result in step 5802 (FIG. 60).
In step 5804, the business frame support tool (5505) refers to the record lock order definition (5504) using the identified parent table name as a key.

In step 5805, the work frame support tool (5505) acquires the lock order item number of each parent table to be locked based on the reference result in step 5804. (Fig. 61)
In step 5806, the business frame support tool (5505) sorts each lock target parent table in ascending order by the lock order item number and stores it in the exclusive control frame matrix (5506). (Fig. 62)
In step 5807, the exclusive control frame matrix portion generated by the operator is copied and pasted to the corresponding portion of the function matrix (214). (Figure 63)
By these operations, a function matrix having an exclusive control function can be created.

  In other words, because the parent-child relationship definition based on the specification and the definition between the parent tables are defined in advance, the developer simply enters the table to be accessed into the matrix support tool without being aware of the lock order, and locks according to the specification. It is possible to automatically generate a program that locks the tables in order.

System Configuration Functional configuration diagram Overview of function matrix and function matrix XML mapping Dictionary format (example) and dictionary XML format Flow chart to source / program specification output Function matrix creation flow Sample set of template information when creating a function matrix Sample program information set when creating function matrix Parts container set sample when creating function matrix Parts set sample when creating function matrix Keyword set sample when creating function matrix Value set sample when creating function matrix Function matrix completed sample Function matrix XML creation flow Program information set sample when creating function matrix XML Component container attribute set sample when creating function matrix XML Container keyword attribute set sample when creating function matrix XML Part attribute set sample when creating function matrix XML Keyword / value attribute set sample when creating function matrix XML Function matrix completed sample Overall view from function matrix to source generation and document output Functional configuration diagram when using correlation check matrix Correlation check matrix / XML mapping overview Flow chart to source / program specification output (correlation check part) Correlation check matrix creation flow Correlation check ID set sample when creating a correlation check matrix Sample of part name set in the condition part when creating a correlation check matrix Keyword set sample for the condition part when creating a correlation check matrix Item value set sample in the condition part when creating a correlation check matrix Sample item name set for the condition part when creating a correlation check matrix Condition value set sample in the condition part when creating a correlation check matrix Item value set sample in the condition part when creating a correlation check matrix Node ID set sample of condition tree when creating correlation check matrix Parent node ID set sample of condition tree when creating correlation check matrix Sample set of evaluation order of condition tree when creating correlation check matrix Sample name set for the processing part when creating a correlation check matrix Keyword set sample of the processing section when creating a correlation check matrix Sample value set of the processing unit when creating a correlation check matrix Processing target ID set sample of the processing unit when creating the correlation check matrix Processing order set sample of the processing unit when creating the correlation check matrix Correlation check matrix completed sample Correlation check matrix XML creation flow Header set sample when creating correlation check matrix XML Correlation check ID set sample when creating correlation check matrix XML Pattern set sample when creating correlation check matrix XML Evaluation order set sample when creating correlation check matrix XML Sample of part name set for condition part when creating correlation check matrix XML Item name set sample of the condition part when creating the correlation check matrix XML Expression set sample of the condition part when creating the correlation check matrix XML Condition value set sample of the condition part when creating the correlation check matrix XML Processing unit condition tree set sample when creating correlation check matrix XML Sample parts set of processing unit when creating correlation check matrix XML An example of a correlation check matrix Sample DB environment with parent-child relationship Exclusive control matrix creation flow Sample parent-child relationship definition Sample parent table information definition Processing flow up to business matrix output Schematic diagram of business frame support tool Mechanism of parent table identification by business frame support tool Lock order identification mechanism by business frame support tool Exclusive control matrix generation Posting to function matrix

Explanation of symbols

201 ... Function specification, 202 ... Template specification, 203 ... Parts container specification, 204 ... Parts specification, 205 ... Dictionary, 206 ... Business part dictionary 207 ... Template dictionary, 214 ... Function matrix, 216 ... Matrix XML tool, 217 ... Function matrix XML, 218 ... Dictionary XML tool, 219 ... Dictionary XML, 220 ... Source program specification conversion engine, 221 ... Program specification, 222 ... Source code, 2101 ... Correlation check matrix, 2103 ... Correlation check matrix XML , 2301 ... Condition part, 2303 ... Processing part, 2302 ... Condition tree part, 2102 ... Correlation check matrix XL conversion tool, 5401 ... Table X, 5402 ... Table R, 5403 ... Table Q, 5404 ... Table A, 5405 ... Table P , 5406 ... Table Y, 5407 ... Table U, 5408 ... Table S, 5409 ... Table B, 5410 ... Table Z, 5411 ... Table T, 5412 ... Table C, 5 501 ... ER diagram, 5502 ... External design, 5503 ... Parent-child relationship definition, 5504 ... Parent table information definition, 5505 ... Exclusive control frame matrix development support tool, 5506 ... Exclusive control frame matrix

Claims (5)

Software that should be created from the function matrix by applying XML processing to the function matrix created by the user and associated with the template file of the template dictionary using the business pattern as a key and the dictionary showing the contents of each item Software that extracts a template corresponding to the business rule of the wear, selects a business component corresponding to the extracted template, and creates a software specification to be generated using the selected business component In the wear development method,
Store functional design data indicating the function of the software,
Based on the functional design document data, prescribe access restrictions for the software table,
A software development method characterized by controlling access to the software in the development based on the defined access restriction. In the software development method according to claim 1,
The table is a table including records that should be restricted in access,
The software development method characterized in that the access restriction is defined by specifying a table for which access restriction is to be performed, corresponding to the access restriction for the table including the record. In the software development method according to claim 2,
The software specification method is characterized in that the specification of the table is realized by defining an order of restricting access between the tables. In the software development method according to claim 3,
The access restriction is
A parent-child relationship indicating a hierarchical relationship of each table is extracted from the functional design document data;
A software development method characterized by defining the access restriction based on the extracted parent-child relationship. In the software development method according to any one of claims 1 to 4,
The software development method characterized in that the access restriction includes at least one of suppression of access to the table itself, suppression of change of the table, and alarm output for change of the table.

Images