JPH04172529A - Automatic program generating system by data oriented software parts - Google Patents

Abstract

PURPOSE:To enhance the productivity and reliability of software by setting up general criteria for preparing software parts without depending on individuality of software designers and features of application work and by managing software parts in a format conforming with the criteria. CONSTITUTION:Using input unit 205 and display unit 204, a software part in unit of data item is registered a dictionary on an external storage unit 206 by a dictionary oriented editor 102 or corrected. Here, the data item used represents the minimum unit of data having substantial application meaning such as person's name and birth date. Program generating tool 107 inputs a program specification library 104, data specification library 105, standard pattern library 106, and dictionary 101 which are all on the external storage unit 206, and prepares a generation objective program and outputs the program to the source library 108 on the external storage unit 206.

Description

[Detailed description of the invention] [Industrial application field] The present invention relates to improving software productivity.

In particular, the present invention relates to a software development support method suitable for streamlining new software development by converting existing software into components and automatically generating programs from these components.

[Conventional technology]

One of the effective methods for improving software productivity is to reuse existing software and develop new software.

When developing new software by reusing existing software, it is common to reuse it at the business system level or program level, and the programmer analyzes the existing software's specifications and source code and makes changes. It is.

However, even for software for the same type of business, the required specifications vary depending on the user, and with the above reuse method, it is difficult to identify changes in a single piece of software just from the specifications and source code. On the other hand, it had drawbacks such as deterioration in software maintainability as changes were made due to differences in the required specifications of multiple users.

Because of these shortcomings, even in the development of software for similar tasks, people give up on reusing existing software, and
In many cases, completely new software was developed, which hindered productivity improvement.

One way to solve this problem is to divide software into components by focusing on their functions, and then automatically generate new programs by defining the correspondence between these components and programs. be.

This method makes it possible to divide software into small groups of parts and reuse existing software at the part level when developing new software, making it possible to avoid differences in software application tasks and required specifications. It is now possible to respond by selecting, modifying, and adding parts.

Related technologies of this type include, for example, the third technology.
0th (late 1985) Information Processing National Conference Lecture Proceedings (
I) p459-460 rEAGLEL: There is a method for realizing program component synthesis in p459-460.

[Problem to be solved by the invention]

The above-mentioned conventional technology automatically generates a program by converting software into components and composing them.
It is a powerful technology for improving software productivity and reliability.

However, in the above-mentioned conventional technology, the software is made into components by subdividing the functions, so when making the software into components,
It is difficult to establish clear standards regarding how functions should be subdivided and what level should be included in a single component, and the methods of dividing software have varied depending on component designers.

Further, since the designer or programmer must define the correspondence between the target business program and the component for each program, productivity and reliability cannot be sufficiently improved.

The first purpose of the present invention is to establish a certain standard that does not depend on the individuality of the designer or the characteristics of the application when converting software into components, and to establish a method for managing components in a format that conforms to this standard. Ill! software components 4! The objective is to make it easier to reuse existing software, and to improve the productivity and reliability of software.

A second object of the present invention is to enable automatic generation of a target business program without requiring a designer or programmer to define the correspondence between business programs and components;
The goal is to improve the productivity and reliability of software.

[Means to solve the problem]

In order to achieve the above first objective, 1. In the automatic program generation method using data-centric software components of the present invention, software components are defined in data item units, and the processing for each data item is divided into several types in advance. It is characterized by storing and managing this information in a file with a dedicated structure called a dictionary.

In addition, the control part of the processing program for one screen, form, file, etc., which is not included in the processing of each data item, is divided into several types based on the characteristics of the control structure, and each control part has its own task-specific processing. It is characterized by being made into parts as a standard pattern.

In order to achieve the above second objective, in the automatic program generation method using data-centric software components of the present invention, the program is automatically generated without defining the correspondence between the business program to be generated and the above-mentioned data item-based components. The program is characterized in that the program is automatically generated by selecting the parts to be used based on the names of data items existing in the input/output medium and composing them with a standard pattern.

[Effect]

In order to achieve the first object, the present invention defines a software component in units of data items, and defines several types of processing for data items within the component. In addition, the program control part is made into parts as a standard pattern, and completely separated from the parts for each data item. This makes it possible to standardize software components, facilitate the reuse of existing software, and improve the productivity and reliability of software.

In order to achieve the second object, the present invention uses an automatic program generation tool to associate business programs to be generated with parts in units of data items. This eliminates the need for the designer or programmer to define these correspondences, and it is possible to improve the productivity and reliability of the software.

〔Example〕

Embodiments of the present invention will be described in detail below with reference to the drawings.

Fig. 1 is a functional configuration diagram of a system for realizing an embodiment of the present invention, Fig. 2 is an equipment configuration of a computer for realizing the present invention, Fig. 3 is an example of dictionary definition, and Fig. FIG. 4 is a flowchart showing the processing procedure for program generation.

First, the dictionary manipulation procedure and program generation procedure in the embodiment will be explained using FIG. 1 and FIG. 2.

In this embodiment, the input device 205 and the display device M2O4 are used to register and modify components in units of data items in the dictionary on the external storage device 206 using the dictionary dedicated editor 102.

Here, the data item refers to the smallest unit of data that has business significance, such as name, date of birth, etc. The program generation tool 107 includes a program specification library 104, a data specification library 105. The standard pattern library 106 and dictionary 101 are input, a generation target program is generated, and the program is output to the source library 108 on the external storage device 206.

Note that the program specification library 104 is a library that defines the standard pattern name, bath name, screen name, and form name used in each program.
The data specification library 105 is a library that stores definition information regarding each file, screen, and form.

. Next, based on FIG. 3, a dictionary 10 for managing software components in units of data items in this embodiment is provided.
The structure of No. 1 will be explained.

The dictionary 101 stores definitions regarding data item names, check conditions, input editing, and output editing definitions. Here, checking conditions, input editing, and output editing will be explained.6 FIG. 5 shows checking conditions, input editing, and output editing for the data item "year, month, and day."

Input data “H” from input form 502 and input screen 503
2O1, 05, 03J N501), check whether the data format is correct and whether the date actually exists. Check conditions define the conditions under which no errors occur during this check.

The checked data is stored in a database 506 or a file 507. In this example, the date input in the Japanese calendar is converted to the Western calendar and stored, but the process of converting the input data 501 from the outside into the format 508 to be stored in the database 506 or file 507 is necessary. This is input editing.

Next, output editing is done in database 506 and file 507.
This is the process of converting the data into a format 512 that is output to an output form 510 or an output screen 511.

In this example, the "Year Month Day" in the storage format 508 is converted from the Western calendar to the Japanese calendar, and the era name is further edited into Japanese.

Returning to FIG. 3 again, this example shows the definition for the data item "year, month, day J" and the definition for "inventory amount". 301 is the symbol item name of "year, month, day" is rNENGAPI
303 indicates that the data format is 6 alphanumeric characters, 304 indicates that subroutine rPROGAJ is called when checking the "year, month, and day," and 305 indicates input editing. When subroutine rPR
This indicates that OGBJ is to be called, and 306 similarly defines that rPROGcJ is to be called during output editing.

Comparing this with FIG.
The conversion from the Japanese calendar to the Western calendar storage format performed in step 505 is performed by calling rPROGBJ, and the conversion from the Western calendar to the Japanese calendar and the Japanese language conversion performed in step 509 are performed by calling rPROGcJ. Defined.

In the other definition of "inventory amount", 307 indicates that the symbolic item name of "inventory amount" is rZAIKOJ, and 309
indicates that the data format of the inventory amount is a 10-digit number. The description "rNUMERICJ" in 310 indicates that the check condition is that the input data is a number.

312 defines that the result of multiplying the input data by 1000 is to be stored in the database 506 or file 507. In addition, 313 stores the “inventory amount” in the database 50.
6 or file 507 to output the result divided by 1000 when displaying on the output form or output screen.

FIG. 4 shows the program generation processing procedure in the program generation tool.

The generation tool first inputs a standard pattern used in the business program to be generated from the standard pattern library 106 (step 401).

Note that which standard pattern to use when generating a program is defined in advance in the program specifications.

Figure 6 shows an example of a standard pattern. This example is a standard pattern for file checking processing. This pattern reads an input file, checks the data on the input file, and stores error-free data in the file. Used to generate a program that performs the process of writing to. 601 in Figure 6 is the standard pattern, which consists only of control parts common to programs that check files, such as opening and closing processing of input/output files, and repeating reading of input files and writing of output files until the input file is finished. This is the program skeleton.

Check processing 602 and input editing processing 603 are program-specific business processing 5 and are not included in the standard pattern, and the program generation tool is the dictionary 10.
This is the part generated using 1.

In input/output component development (step 402), the definition information of the file is input from the data specification library 105 based on the name of the input/output file defined in the program specifications of the program, and the input/output declaration part of the program is input. Generate source code.

This and the standard pattern input in step 401 are synthesized.

FIG. 7 shows how to develop input/output components for the standard pattern 701. The program generation tool reads the file specification 706 based on the file specification name defined in the program specification 705 and records in the file specification 706. Read the record specification 707 based on the specification name, and use this information to create the input/output environment definition 70 for the program.
2. A file description 703 and a data description 704 are generated, and these and the standard pattern 701 are synthesized.

Form printing processing development (step 404) is performed when the program uses a standard pattern for outputting batch forms, and is performed in the data specification library based on the name of the output form defined in the program specifications of the program. 105
The corresponding form definition information is extracted from the form, a form editing work area and a form processing procedure are generated from this information, and these are combined with the standard pattern input in step 401.

FIG. 8 illustrates the development of the form printing process (step 404) for the standard pattern 801. The program generation tool reads the form specification 809 defined in the program specification 808, and uses the contents of these definitions to create the input/output environment definition 802. File description entry 8
03 and a data description item 804, and also generates work areas 805 to 806 for form editing and a work area 807 such as a counter used during program execution, and synthesizes these with the standard pattern 801.

In the checking/editing processing development (step 405),
A component for each data item registered in the dictionary 101 is input, a process for the data item is generated, and this is combined with a standard pattern. At this time, which process of which part is to be developed in the program and where in the standard pattern is determined as shown in FIG. In Figure 9, 9Of is 1
11!4! It's a pattern. Further, 902 is a check process, and 903 is an input editing process. Program specifications 90
File specification 90 of the input file defined in 4.
5 to 906, and further refer to record specifications 907 to 908 based on the record specification names therein. In this example, the data item on the record specification is the data item processed by the program, rZAIKOJ (inventory amount).
exists in the input file, the component corresponding to the data item "inventory amount" is read from the dictionary 101 and combined with the standard pattern. On the other hand, in this example, the standard pattern used in the program is the same standard pattern as in the case of Figure 5 called "CHECKOol".
In this pattern, check processing rERR-CHECK-P
ROCJ902 and input editing processing rINPUT-DA
I have TA-EDIT-PROCJ 903. From this, the program generation tool generates a check process from the check condition 310 defined for the "inventory amount" part and an input edit process from the input edit 312.

Furthermore, when developing processing for a data item, the prefix defined in the program specification 904 is added to the item name on each file, as shown in this example. In FIG. 9, the symbolic item name on the input file for the data item rZAIKOJ (stock quantity) is rIN-ZAIKOJ, and the symbolic item name on the output file is rDATA-ZAKOKO'', which are determined at the time of program generation. This results in the data item “
The parts for "inventory quantity" are made independent of the file in which they are stored.

In the source program output 406 shown in FIG. 4, the target business program generated by the above processing is output to the source library.

Although this embodiment has been described using an example of a batch processing program, it is also possible to generate an online processing program in the same way.

〔Effect of the invention〕

Since the present invention includes one or more of the processing steps described above, it has the effects described below.

The company decided to define software components in units of data items, defined several types of processing for data items within each component, and developed a dictionary to store them. By converting it into components and completely separating it from data item-based parts, it becomes possible to standardize software parts, making it easier to reuse existing software, and ultimately improving software productivity and reliability. can.

In addition, since the program generation tool now maps the business program to be generated and the parts for each data item, it is no longer necessary for the designer or programmer to define these correspondences, thereby improving software productivity. It is possible to improve performance and reliability.

[Brief explanation of drawings]

FIG. 1 is a functional configuration diagram of a system showing an embodiment of the present invention;
Fig. 2 is a hardware configuration diagram that implements the present invention, Fig. 3 is a diagram showing an example of dictionary definition, Fig. 4 is a flowchart showing the processing procedure for program generation, and Fig. 5 is a check on data items in the dictionary. Figure 6 is a diagram illustrating an example of conditions, input editing, and output editing processing, Figure 6 is a diagram showing an example of a standard pattern, Figure 7 is a diagram showing a method for developing input/output components, and Figure 8 is a diagram showing development of form printing processing. A diagram showing the method, FIG. 9 is a diagram showing a method of developing data item parts with respect to a standard pattern. 101...-Dictionary, 104... Program specification library, 105...
・Data specification library. 106...Standard pattern library. 107...Program generation tool. 108... Source library, 601... Standard pattern, 602... Check processing, 603... Input editing processing, 70-1... Standard pattern, 702... Input/output environment definition, 703...・File description entry, 704・
...Data description entry, 705...Program specifications 57
06... File specification, 707... Record specification, 801... Standard pattern, 802... Input/output environment definition, 803... File description entry, 804... Data description entry, 808... Program specifications, 809... Form specifications, 901... Standard pattern, 902... Check processing, 903... Input editing processing, 904... Program specifications, 905... File specifications, 906 ...File specifications, 907...Record specifications, 908...Record specifications. Collection I Z Volume 2 Figure Collection 3 Figure 10/ 0I Volume + Figure 5 Figure 6

Claims (1)

[Scope of Claims] 1. Means for defining software for each data item by dividing it into several types of processing for each data item, and storing it in a file with a dedicated structure called a dictionary, and a screen, a form, Programs that process files, etc. are divided into several types based on the characteristics of their control structures.
A means for converting individual business processes into components as standard patterns that do not have their own, and selecting components for each data item based on data item names in input/output media of a business program, and composing them with the standard pattern. 1. An automatic program generation method using a data-centered software component, characterized in that it has a means for automatically generating a business program. 2. The automatic program generation method using a data-centered software component according to claim 1, wherein processing for the data item is divided into three types: checking, input editing, and output editing. method.