JPH01286020A - Program retrieving system - Google Patents

Abstract

PURPOSE:To raise the efficiency of a program re-use by automatically preparing data for retrieval from a developed program when a similar program is retrieved out of a developed program group. CONSTITUTION:A similar program retrieving system extracts a characteristic value from a record-type data processing program 1a of a source program 1 and prepares a characteristic value record 3a. Based on a characteristic value extracting program 2 and a retrieving command inputted from a terminal device 5, the characteristic value record 3a is retrieved. The characteristic extracting program 2 extracts the characteristic value of the record type data processing program 1a described by a COBOL language and outputs it to a file 3. A syntax analysis is executed, and a data item name group to constitute record type data is extracted as a characteristic value. The name group and the program name are combined, the characteristic value record is prepared and outputted to a record file 3.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to classification and retrieval of programs, and particularly to a program retrieval method for retrieving similar programs suitable for reusing old programs.

[Conventional technology]

Conventionally, many programs have been developed and operated for record-type data processing, typified by office data processing. When trying to develop a new data processing request based on an already developed program, the search method for already developed programs is to search the program list and view documents.

This method involves viewing a program list, etc., and is insufficient as a search method for allowing developers to reuse already developed programs.

[Problem to be solved by the invention]

The above-mentioned conventional technology does not give consideration to the formal representation of program characteristics, and there is a problem in that it is not possible to search for a corresponding program from data on the characteristics of a newly developed program. Ta.

An object of the present invention is to extract data that is a feature value of a low-development program, and to create a feature value record by linking the data with the program name.

The object of the present invention is to provide a search method for programs having similar relationships by providing an efficient search means for a group of feature value records extracted from a plurality of programs.

[Means and means for solving the problem] A record type data processing program has processing contents that are dependent on a data item group that constitutes record type data to be processed. Data item names often have the same meaning in other programs, and multiple programs that process the same data item name may
It can be said that there is considerable similarity in processing content. Therefore,
For programs that have multiple identical data item names.

It can be determined that there is a higher degree of similarity in data processing.

Based on this viewpoint, the present invention uses data structure analysis technology to extract item names of record-type data, sets a character string concatenated with the program name as a feature value record of the program, and converts this feature value record into a low-development method. By applying character string search technology to a group of feature value records created and accumulated using a record-based data processing program, feature value records with the same data item name as the specified data item name can be found. The purpose is to search and obtain the names of programs having a similar relationship from the feature value records. By specifying multiple data item names as search data, it is possible to search for program names with higher similarity.

[Effect]

With data structure analysis technology, it is possible to extract a group of data item names that make up record-type data by parsing the data structure description of a record-type data processing program. A feature value record of the program can be created by linking the program names analyzed and extracted in the same way.

With string search technology, if you specify a certain data item name as a search string for a group of accumulated feature value records, you can search for feature value records that include that data item name. The program name can be obtained as the name of a similar program. It was used in an expression format that included an indeterminate character string element in specifying the data item name.

By applying the so-called "regular expression format", data item names can be specified indefinitely, and feature value records can be searched more efficiently.

〔Example〕

An embodiment of the present invention will be described below with reference to FIGS. 1 to 9.

Fig. 1 is a functional block diagram of an information processing system that implements the present invention, and Fig. 2 shows a hardware system configuration of an embodiment of the present invention. As shown in the figure, feature values are extracted from the source program of the record-type data processing program 1a from a source program library 1 that stores record-type data processing programs, a feature value record 3a is created, and a feature value record file 3 is stored. Output feature value extraction program 2 and computer system terminal bag! The search program 4 searches the feature value record file 3 for a feature value record 3a based on a search command inputted from the feature value record file 3, and outputs the record to the terminal device 5.

The hardware system for realizing the present invention is a second
As shown in the figure, CPU II, memory 12, input device 1
3. Output device! 14. It consists of a terminal 15° and an external storage device 16, and the feature value extraction program 2 of FIG. The search program is executed, and the memory 12 or external storage device! 6. Source program library 1. Feature value record file 4. Programs 2, 4, etc. are stored.

The feature value extraction program 2 follows the procedure shown in FIG.

The feature values of the record type data processing program 1a written in the C0BOL language are extracted and the feature value records are output to the feature value record file 3.

The source program of the record type data processing program 1a written in the C0BOL language shown in FIG.
From the information generated by the above analysis from the record type data structure description part IC of the input file identified from the relationship between the record name and "INPUT" in N sentence 1d, the data item name group that constitutes the record type data is extracted as a feature value. (Step 43), the extracted data item name group and the program name extracted from the program name description section 1b (Step 42) are concatenated as shown in FIG.
A feature value record is created as shown in the figure and output to the feature value record file 3 (step 44).

The feature value record 3a is a record composed of a key part and a data part, and within the feature value record file 3, each record is positioned in ascending order of the key part. A part of the key consists of a program name and its attribute value, and values such as the name and version of the source program file 1 in which the source code of the program is stored are added to the attribute value. be unique within the feature value record file.

When a command with the syntax shown in FIG. 7 is input to the search program 4 from the terminal device 5 of the computer system, the search program 4 searches for feature value records from the feature value record file 3 according to the procedure shown in FIG. Output to 5.

In FIG. 8, when a command is input (step 8
1), extract the item name from the command (step 82), and if the data item is the first time (step 82),
3.84) Extract the program feature value record with the name of the search data item from the search record file and register it as a valid record in the table (step 85); Search for more valid records and disable records that are not applicable (
Step 86), if the data item is complete, display the valid records of the table (step 87).

When specifying a search string, that is, a data item name, by providing an indeterminate character string rule shown in Figure 9, it is possible to specify multiple patterns of item names for one item name6. For example, 'BU/ BUHIN/PART:No/B
By specifying ANGO/BAN', nine first-order character strings can be searched, increasing search efficiency.

RTBAN.

As described above, according to this embodiment, a character string in which a group of data item names of record type data and a program name are concatenated is extracted from an already developed record type data processing program as a feature value record, and stored. By specifying a search for a data item name using an indeterminate character string expression for a group of characteristic value records, it is possible to easily search for similar programs.

〔Effect of the invention〕

According to the present invention, (1) feature values of a program are reduced to a character string that is a concatenation of item names of record-type data processed by the program.

Search target data can be automatically generated without creating data manually. (2) Since the feature value is a character string, it is possible to specify a search string using an indeterminate character string expression, and a plurality of possible data item names can be specified, increasing the search probability.

As described above, when searching for similar programs from a group of already developed programs, the search data is automatically generated from the already developed programs and stored, making it possible to search, which greatly improves the efficiency of program reuse. There is an effect that can be improved.

[Brief explanation of the drawing]

Fig. 1 is a functional block diagram of an embodiment of the present invention, Fig. 2 is a hardware configuration diagram of an embodiment of the invention, Fig. 3 is a detailed diagram of the record type processing program 1a, and Fig. 4 is a feature value extraction program. Figure 5 is a detailed diagram of the feature value record, Figure 6 is a detailed diagram of the feature value record extracted from the record type processing program shown in Figure 3, and Figure 7 is a syntax diagram of the search command. , Fig. 8 is a processing flow diagram of the search program, and Fig. 9 is a diagram showing the specification format of the search string.
Ru. DESCRIPTION OF SYMBOLS 1... Source program library, 1a... Record type data processing program, 2... Feature value extraction program, 3... Feature value record file, 3a...
Feature value record, 4... Search program, 5... Terminal equipment No. 1 Concave ■2 Fig. 3 Not shown 4 Fig. Z 5 Fig. 6 Hard

Claims (1)

[Claims] 1. Extract search information from the record-type data processing source program in the source program library, store the extracted feature value records in a feature value record file, and extract features from the feature value record file based on search instruction data. In a program search method for searching a value record, a data item name is extracted from the record type data processing program, and a character string concatenated with the program name is stored in the above file as a feature value record of the record type data processing program. , a program search method characterized in that a feature value record having a data item name corresponding to a specified data item name is searched, and a name of a program having a similar relationship is obtained from the feature value record.