JPH11203116A - Variable classifier and storage medium storing the program - Google Patents

Abstract

(57) [Summary] [PROBLEMS] To classify and save variables in a program with respect to a variable classification device and a storage medium storing the program, and to determine the type and use of the variable by an inquiry of an operator. I do. A variable classification device for estimating a use of a variable used in a program includes a variable classification rule table that defines a relationship between information about the variable and the contents or use of the variable, and performs a calculation of an operation applied to the variable in the program. The system collects information about types or variables, refers to the variable classification rule table for the collected variable information, and estimates and classifies the contents or uses of variables used in the program.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a program variable classification device for supporting creation and change of a computer program, and a computer program used in a computer system (hereinafter referred to as a program).
The present invention relates to a device for classifying and outputting the contents of variables or types of applications.

[0002] Programs of computer systems such as financial systems and medical systems are enormous in size, and many variables are used. However, the method of defining the variables varies depending on the programmer. To change or extend the program, information about those variables is required. It is difficult to easily identify the contents or use (eg, amount, zip code, etc.) represented by the variable.

[0003]

2. Description of the Related Art A conventional computer program creation / change support apparatus only has a function of inquiring about the type of a variable used in a program and a function of inquiring about a use place of the variable. When modifying or maintaining the program, the programmer analyzed the existing program, understood the meanings (contents, uses, etc.) of the variables used in the program, and changed and maintained the program. .

[0004]

As described above, programs for a financial system, a medical system, and the like are enormous in size, and various variables are used. For example, in the case of a financial system, the program creator sets variable names for quantities, amounts, addresses, postal codes, and the like. Do not specify in the program. On the other hand, when changing or maintaining a program, it is necessary to know the contents or usage (amount, date, etc.) of these variables.

[0005] In the current technology, a programmer who maintains the program has no other option but to investigate and guess the program. Therefore, the amount of work for a program having a large number of variables is very large, and errors are likely to occur.

According to the present invention, variables in a program are classified and stored.
It is an object of the present invention to provide an apparatus for predicting the classification of a program to determine its use and output the program.

[0007]

The contents and uses of variables used in a program can be classified to some extent by the operations performed on the variables and the characters included therein. For example, the amount of money is subject to addition, subtraction, multiplication, and division, and currency symbol characters such as “￥” and “＄” are included. Therefore, the contents and use of variables are estimated by analyzing the use of variables in calculations. it can.
In addition, postal codes are usually not subjected to arithmetic operations such as addition, subtraction, multiplication, and division, and can be inferred therefrom. In this way, variables can be classified by collecting and analyzing information on variables used in the program.

The present invention provides a variable classification device for estimating a use of a variable used in a program, comprising a variable classification rule table which defines a relationship between information on the variable and the contents or use of the variable. Information on the types of operations or variables to be collected is collected, and the collected variable information is referred to the variable classification rule table to estimate and classify the contents or use of the variables used in the program.

FIG. 1 shows the basic configuration of the present invention. In FIG. 1, reference numeral 1 denotes a variable classification device.

Reference numeral 11 denotes a source program, which is a program for classifying variables. Numeral 12 is a variable information collecting means for collecting information on variables used in the source program, and information on variable definitions such as types and names of variables used in the program, and information on variables. It collects information such as calculations, operations performed, constant values and variables that are assigned.

Reference numeral 13 denotes variable information holding means for holding variable information. Numeral 14 denotes a variable classifying means for classifying contents and uses so that the user can easily use the variable information held in the variable information holding means.

Reference numeral 141 denotes a variable classification rule table that defines the relationship between information on variables and the contents or uses of the variables. Reference numeral 15 denotes a search unit for inquiring about the contents and use of the variables registered in the variable classification device 1.

The operation of the basic configuration of the present invention shown in FIG. 1 will be described. The variable information collecting means 12 enumerates all variables used in the source program 11, and for each of those variables, defines the contents (information such as the type of the variable),
Collects information such as processing instructions that refer to variables and the relationship of data exchange with various variables. Then, the collected variable information is stored in the variable information storage unit 13. Then, the variable classifying unit 14 estimates the contents or use of the variable with reference to the variable classification rule table 141 for the variable information collected in this manner.

The contents or use of the variables are estimated by the following method. Collects the values assigned to variables from all processing instructions that refer to the variables and the type of operation directly applied to the variables, and enumerates the estimated uses of the variables based on that information .

Alternatively, information on the type of the variable is collected from the definition of the variable, and information estimated as uses of the variable is listed based on the information. Or, for the estimation results obtained in step (1), obtain a group of variables that exchange the same type of data based on the information on the exchange of data between variables, and leave only the common uses among them. To limit the use of the listed variables.

The relation between variables referred to in the same arithmetic expression is examined, the relation that the use of the variable should satisfy is calculated, and the others are deleted while leaving only the use of each variable that can satisfy the use. Doing so limits the use of the listed variables.

The user instructs the search means 15 on the content of the search. Then, the search unit 15 searches the variable information holding unit 13 and outputs the contents and use of the queried variable to a display or the like.

According to the present invention, even for a program of a large-scale computer system, the contents and use of a huge number of variables can be easily known. Therefore, the efficiency of program change, expansion, maintenance, etc. is greatly improved.

[0019]

DESCRIPTION OF THE PREFERRED EMBODIMENTS In the following description, an embodiment of the present invention will be described by taking a program described using the COBOL language as an example. However, the programming language targeted by the present invention is not necessarily a COBOL language. It is not limited.

In the COBOL language, processing instructions for performing operations on variables include an ADD statement, a SUBTRACT statement,
DIVIDE statement, MULTIPLY statement, COMPUTE
An update statement such as a statement, a MOVE statement, and an ACCEPT statement corresponds to this. IF statement, EVALATE statement, PERFORM
Statements, SEARCH statements, and other conditional branch statements may also perform operations on variables in branch condition calculations. As operations applied to variables, four arithmetic operations of addition, subtraction, multiplication, division, and exponentiation operations can be used for classification, and special operations include:
It is also possible to use a display device on a terminal screen, an argument used for a built-in function, a partial reference operation such as extracting only the last two digits of a number, and the like. Here, a case in which only four arithmetic operations are performed will be described as an example.

FIG. 2 shows an apparatus configuration according to an embodiment of the present invention. In FIG. 2, reference numeral 1 denotes a variable classification device.

Reference numeral 11 denotes a source program. Reference numeral 12 denotes a variable information collecting unit. A variable information database 131 holds collected variable information (corresponding to the variable information holding unit 13 in FIG. 1).

Numeral 14 denotes a variable classification means. Reference numeral 141 denotes a variable classification rule table, which has information on variables such as a set of usages of variables, types of variables, and applied arithmetic expressions, and designates information on variables to specify their usage. It is implemented as a collection of processing procedures that returns a set of uses estimated as Candidates for use are prepared by the user in advance in consideration of the field handled by the target program.

Reference numeral 15 denotes a search means. Reference numeral 21 denotes a syntax analysis device for analyzing the syntax of a source program.

A data flow analysis device 22 analyzes the data flow of a syntax-parsed program. Reference numeral 23 denotes a program analysis information database, which has program analysis information.

Reference numeral 24 denotes a terminal device. The operation of the configuration of FIG. 2 will be described. The variable information collecting means 12 collects variable information from the program analysis information database and stores it in the variable information database 131 (for a specific example of the variable information, see FIG. 4). This variable information can be collected by existing techniques such as data flow analysis and domain analysis (see Japanese Patent Application No. 8-99745). Analysis target is one CO
The processing is not limited to the BOL program but can be performed on variables belonging to a plurality of programs at once. The variable classification means 14 classifies the variable information in the variable information database 131 with reference to the variable classification rule table 141. Details of the operation of the variable classification means 14 will be described later. The user uses the terminal device 24 to input a search item into the search means 15 and searches for necessary information on variables.

FIG. 3 shows an embodiment of the system configuration of the present invention. In FIG. 3, reference numeral 111 denotes a computer.

Reference numeral 112 denotes a CPU. 113 is a memory. Reference numeral 114 denotes an input / output interface.

Reference numeral 115 denotes a drive unit for the storage medium 151. 116 is a display. 117 is a printer.

An external storage device 121 holds a system program. Reference numeral 122 denotes a system program (source program).

Reference numeral 125 denotes the external storage device 2, which holds a system database. Reference numeral 126 denotes a system database, which is a system database.

An external storage device 130 holds a syntax analysis program, a data flow analysis program, and a program analysis information database. 132
Is a parser.

Reference numeral 134 denotes a data flow analysis program. 136 is a program analysis information database.
An external storage device 138 holds a variable information database and a variable classification rule table.

Reference numeral 140 denotes a variable information database. 1
Reference numeral 42 denotes a variable classification rule table. Reference numeral 151 denotes a storage medium which holds a program necessary for the variable classification device of the present invention.

153 is a variable information collection program,
The variable collection means is constituted by the CPU, the memory, and the like. 155 is a variable classification program,
The variable classifying means is constituted by a memory or the like.

Reference numeral 157 denotes a variable classification rule creation program. Reference numeral 159 denotes a search program, which constitutes search means by a CPU, a memory, and the like.

FIG. 4 shows an example of a source program.
It is a BOL language program. Step 732 FR
EC. Are upper variables, and the following variables PR, DT, AM
These are upper variables of T, TXR, and I-CNT.

Step 733 is the definition of the variable PR.
Indicates an 8-digit number. Step 734 sets the variable D
This is a definition of T, indicating that it is a 6-digit number. Step 735 is the definition of the variable AMT, and indicates that it is a 4-digit numerical value.

Step 736 is the definition of the variable TXR, and indicates that there are two digits below the decimal point. Step 737 is the definition of the variable I-CNT, which indicates that there is a 4-digit numerical value.

The DSP. Is an upper variable and is an upper variable of the variables PP, YMD, TTL, and PSTNO. Step 739 is a definition of the variable PP, which is a six-digit numerical value, and represents a negative numerical value in which a comma (,) is inserted every three digits.

Step 740 is a definition of the variable YMD, which is a numerical value of 6 digits, and represents a numerical value in which a slash (/) is inserted every 3 digits. Step 741 is a definition of the variable TTL, which is a numerical value of four digits, and represents a numerical value in which a comma (,) is inserted in the third digit.

Step 742 is the definition of the variable PSTNO, which indicates that it is a 5-digit numerical value. Step 742 is the definition of the variable TPP, which indicates that it is a 9-digit numerical value.

Step 1325 is an instruction for instructing to substitute zero for TPP. Step 1
555 is an operation instruction, and TPP = TPP + PR * (T
XR + 1).

Step 1556 is an instruction.
This command instructs to substitute the value of PP. Step 1557 is an addition instruction, which instructs to substitute the value of AMT for TTL.

Step 1558 is an instruction for instructing the YMD to substitute the value of DT. Step 1559 is an addition instruction for instructing to add 1 to I-CNT.

FIG. 5 shows an example of collected variable information. 3
Reference numeral 1 denotes variable information, which is variable information on variables of the program in FIG.

The data transfer group is a list of variables and data exchanged variables among arbitrary variables in the program shown in FIG. For example, the type of the variable DT is PIC9 (6) (6-digit numerical value), and
It is used in the 58 instruction "MOVE DT TO YMD", and indicates that there is exchange of a variable YMD and data.

FIG. 6 shows an application estimation method 1 of the present invention. FIG. 6A is a variable classification rule table used in the usage estimation rule 1. 41 is a variable classification rule table 1.

9 is a table for estimating a variable by an operation to which the variable is applied (applied operation). For example, postal codes, genders, names, and dates are not subject to addition, subtraction, multiplication, and division operations, so variables not used in the four arithmetic operations are classified into postal codes, genders, names, and dates. In addition, the year is estimated to be used only for addition and subtraction, and the number may be used only for addition and subtraction. Therefore, variables used only for addition and subtraction are classified according to the usage as shown in the table. In addition, years, tax rates,
Since the interest rate, amount, and quantity are estimated to be used for addition, subtraction, and multiplication, the uses are classified as shown in the table.

The variable classification rule table 1 is created by the user,
The default value may be created in advance when the product is shipped, and may be changed by the user. FIG. 6B is an example of the estimation result obtained based on the variable classification rule table 1 (41) for the application. It is an estimation result of the use of each variable obtained by applying the variable classification rule table 1 (41) of FIG. 6A to the variable information database of FIG.

In the variable information database shown in FIG. 5, the variable PR is used for addition, subtraction, and multiplication. Therefore, it is estimated that the usage of the variable PR is any one of the number of years, the tax rate, the interest rate, the amount, and the number. Since there is no addition, subtraction, multiplication, or division of the variable DT, it is estimated that the variable DT is a variable of the postal code, gender, name, and date.

As described above, the usage of each variable is estimated, and the estimation result shown in FIG. 6B can be obtained. FIG. 7 is a flowchart of the usage estimation method 1 of the present invention.

The flowchart will be described according to the program steps in FIG. S1 The operator before and after the variable V is obtained from the processing instruction of the use location L of the variable V. S2 It is determined whether there is an operator before and after the variable V.

S3 In the judgment of S2, if there is only one operator before and after V, the only operator is OP. S4 In the determination of S2, if there is an operator before and after V, the operator with the higher priority (priority is determined in advance) is set to OP.

S5 The obtained OP is additionally recorded in the variable database in association with the variable V. S6 The operation OP corresponding to the variable V is repeated for all the used portions L, and is recorded in the variable database.

S7 The usage corresponding to the combination of the operators associated with the variable V is extracted from the variable classification rule table 1 and recorded in the variable database in association with the variable V.

S8 The above processing is repeated for all variables V. FIG. 8 is a diagram illustrating a method 2 of estimating a use of the present invention. CO
In the case of the BOL language, the description of the data item description in the PICTURE clause and the USAGE clause is used as the variable type information.
The use of the variable can be estimated. From the type information, information such as the presence or absence of a decimal point, editing characters (characters that are automatically arranged at specific positions when a value is assigned to a variable), the number of digits, and character types allowed as values are obtained. Based on this information, the application is estimated and a variable classification rule table is created.

FIG. 8A is a variable classification rule table 2 according to the present invention, and is a table for estimating the relationship between the information on the type of a variable and the use and determining the use. Reference numeral 51 denotes a variable classification rule table 2.

Variables including a decimal point as a feature of a type can be estimated as a tax rate and an interest rate.
In addition, variables including a currency symbol and a specific character string related to currency, such as "CD" and "DB", can be estimated as a monetary amount. Hereinafter, similarly, a variable classification rule table 2 of FIG. 8A is created.

Although not shown in the variable classification rule table 2 of FIG. 8 (a), the usage can be estimated from the values (numerical values, character strings, etc.) obtained by substituting variables. May be.

FIG. 8B is an example of the estimation result of the use obtained based on the variable classification rule table 2. For each variable in the variable information database in FIG. 5, the variable classification rule table 2 in FIG.
Is a result of estimating the use of the variable obtained by applying.

Since the TXR is a numerical variable including a decimal point, it is classified into a tax rate and an interest rate according to the variable classification rule table 2. PP contains a comma (,) and has a currency symbol, so it is classified as an amount.

In the YMD, '/' is inserted, and the date is classified. Since the TTL includes a comma (,), it is classified into the amount and the number. Since PSTNO is a 5-digit number, it is classified as a postal code.

FIG. 9 is a flowchart of a method 2 of estimating the use of the present invention. The use corresponding to the combination of the operators associated with the variable V is extracted from the variable classification rule table 2 for the use based on the S1 type.

S2 The variable V is recorded in the variable information database while being equal to the variable V. S3 The above processing is repeated for all variables V. FIG. 10 shows Method 3 of estimating the use of the present invention.

The use estimation method 3 of the present invention is common to the use estimated for each variable by the above-described method and the use estimated for the variables listed in the data transfer group of an arbitrary variable. In this method, only the usage is retained, and the other usages are deleted from the usage of the variable.

When the usage estimation method 3 of the present invention is applied to the variables in the variable information database shown in FIG. 5 with respect to the usage estimation results shown in FIG. 8B, the following results are obtained. The variable of the data transfer group of the variable DT is YMD. Estimation result 2
, The use of the variable DT is undecided, and the use of the variable YMD is date. Therefore, the usage of the variable DT is classified by date.

Similarly, the variable of the data transfer group of the variable AMT is YMD, the AMT of the estimation result 2 of the use is undecided, and the use of the TTL is the amount and the number.
The uses of L are classified into money and number.

In the same manner, an example of the estimation result of the application shown in FIG. 10B can be obtained. FIG. 10A is a flowchart of a method 3 of estimating a use of the present invention. S1 The usage set of the variable V is set in the usage set S.

S2 Let T be a set of uses of the variable W listed in the data transfer group. S3 Each set (common part) of S and T is set to S. S4 The above processing is repeated for all the variables W listed in the data transfer group.

S5 The above processing is repeated for all variables W. FIG. 11 shows an application estimation method 4 of the present invention. Variables that have not been obtained by the above-described estimation methods 1, 2, or 3 can be obtained by the following estimation method based on the variables.

Since it can be estimated that there is a certain relationship between the arithmetic expression and the usage of the variables used therein, the estimation method 4 of the present invention obtains the relationship between the arithmetic expression and the variable usage as a variable classification rule table. Keep it. Then, the use of the variable is estimated by applying the variable classification rule table to the arithmetic expression of the variable information database.

FIG. 11A shows an example of a variable classification rule of the use estimation method 4. 61 is a variable classification rule table 4. Since there is a relationship between the amount and the tax rate between the amount and the tax rate, the relationship is defined as a rule. Similarly, a variable classification rule table 4 is created based on the relationship of counter = counter + constant, counter = counter-constant, amount = amount * tax rate, and the like.

FIG. 11B shows an example of an estimation result based on the variable classification rule table 4. The expression 62 to be collated in the variable information database in FIG. 5 will be described. 1 of Expression 62 to be collated
Matching rule 7 is applied to + TXR, and it is estimated that TXR is a tax rate or a counter (the counter is a numerical value necessary for control, not a variable representing an item used in actual work). Next, since TXR is the tax rate, (1 + TX
R) is also the tax rate, and the expression PR * (1 + TXR) to be compared
Is applied, and it is estimated that PR is an amount. Further, since PR * (1 + TXT) is a money amount, matching rule 9 is applied to TPP + PR * (1 + TXT) to estimate that TPP is a money amount. Since the counter does not include a multiplication operation, it is excluded in the process of application rules 1 and 9.

FIG. 12 is a flowchart of a method 4 of estimating the use of the present invention. S1 Let the use set S be an empty set. S2 It is determined whether the processed statement has been processed. If it is not a processing statement, the processing of S3 is performed. If the sentence has already been processed, the process of S9 is performed for the next use location L.

S3 The operation expression E is obtained from the processing statement. S4 Search the operation rule table for a set R of operation rules conforming to the expression E. S5R determines whether it has been found, and if so, performs the processing of S6. If not, the process of S9 is performed for the next use location L.

In step S6, the usage U of the portion collated with the variable V in R is obtained. S7: Determine whether the usage U is included in the usage of the variable V,
If it is included, the process of S8 is performed. If not, the process of S9 is performed for the next use location L.

At step S8, application U is added. S9 The processes of S2 to S8 are repeated for all the used portions L. S10: Determine whether the usage set S is empty, and if not, S11
Process. If it is an empty set, the process of S12 is repeated for the next variable V in S12.

S12 The above processing is repeated for all variables V. In this flowchart, the expression of S4 can be compared using an existing technology such as a pattern matching algorithm.

[0080]

According to the present invention, it is possible to automatically perform all the variables used in the program.
As a result, it is possible to reduce the enormous amount of work time spent investigating the use of conventional variables, and to greatly increase the efficiency of investigating and maintaining large-scale software systems.

[Brief description of the drawings]

FIG. 1 is a diagram showing a basic configuration of the present invention.

FIG. 2 is a diagram showing an embodiment of the present invention.

FIG. 3 is a diagram showing an embodiment of a system configuration of the present invention.

FIG. 4 is an example of a source program.

FIG. 5 is an example of collected variable information of the present invention.

FIG. 6 is a diagram showing a use estimation method 1 of the present invention.

FIG. 7 is a flowchart of an application estimation method 1 of the present invention.

FIG. 8 is a diagram showing a use estimation method 2 of the present invention.

FIG. 9 is a flowchart of an application estimation method 2 of the present invention.

FIG. 10 is a diagram showing an application estimation method 3 according to the present invention.

FIG. 11 is a diagram showing an application estimation method 4 of the present invention.

FIG. 12 is a diagram showing a flowchart of an application estimation method 4 of the present invention.

[Explanation of symbols]

 1: Variable classifier 11: Source program 12: Variable information collection means 13: Variable holding means 14: Variable classification means 141: Variable classification rule table 15: Search means

Continued on the front page (72) Inventor Minako Kimura 4-1-1, Kamikodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa Prefecture Inside Fujitsu Limited (72) Keiko Kawabe 4-1-1, Kamikodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa Within Fujitsu Limited (72) Inventor Sanpachi Uehara 4-1-1 Uedanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture Within Fujitsu Limited

Claims (4)

[Claims] 1. A variable classification device for estimating a use of a variable used in a program, comprising a variable classification rule table defining a relationship between information on the variable and the contents or use of the variable, and an operation applied to the variable in the program. Variable classification of a program, which collects information on the types or variables of the program, refers to the variable classification rule table for the collected variable information, and estimates and classifies the contents or use of the variables used in the program. apparatus. 2. The method according to claim 1, further comprising: analyzing a data transfer relationship between the variables based on the variable information, and limiting estimation of the contents or use of the variable based on the data transfer relationship. Variable classifier for the described program. 3. The program according to claim 1, wherein the content or use of the variable is limited based on a plurality of estimation results of the content or use of the variable estimated by the variable classification rule table of different types. Variable classifier. 4. A variable classification rule for creating a variable classification rule table in which a relation between information on variables and contents or uses of variables is stored in a storage medium storing a variable classification program for estimating a use of a variable used in the program. Refer to the created program, the type of operation applied to the variable in the program or the variable information collection program that collects information about the variable, and refer to the variable classification rule table for the collected variable information, and the contents of the variable used in the program Alternatively, a storage medium storing a variable classification program for estimating and classifying a use.