JPS6219937A - Checking system for propriety of argument - Google Patents

Abstract

PURPOSE:To check propriety only to the argument which is used concretely in an execution mode by performing the comparison between the contents of both information tables when a subroutine is called out. CONSTITUTION:When a subroutine 2 is called out, a compiler produces an information table T1 related to the attribute of a argument group A1 to be delivered to the subroutine 2 from a program 1 within the desired program of the program 1. In the same way, an information table T2 related to the attribute of a dummy argument group A2 to be received from the program 1 is produced by the compiler within the desired program of the subroutine 2. The table T1 is read by calling out a routine R1 which reads the information table immediately before the subroutine 2 is called out by the program 1. Then a routine R2 is called out to check the information table immediately after the subroutine 2 is called out. Thus the table T2 is read newly and compared with the table T1 read precedingly. In such a way, the propriety is checked for attribute between a argument and a dummy.

Description

TECHNICAL FIELD The present invention relates to a method for validating arguments, and more particularly to a method for dynamically validating attributes of arguments of a subroutine of a computer program using a compiler language.

BACKGROUND ART Conventionally, when a program calls a routine in a compiler language, it is checked whether there is a conflict between the attributes of the argument passed by the program that calls the subroutine and the attributes of the formal argument that the called subroutine should receive. This has been accomplished either by using a linker when creating a module, or by incorporating the validity checking I11 function of argument attributes into the target program.

Conventionally, the method of checking the validity of arguments using a linker cannot check the validity of variable attributes at runtime and processes them all at once, which requires a large amount of main memory at link time and is not called at runtime. The disadvantage is that unnecessary checks are performed even on arguments of subroutines that are not valid.If the argument validity checking function is included in the target program, the code generation by the compiler becomes complicated and the target program becomes redundant. There is a drawback.

OBJECTS OF THE INVENTION An object of the present invention is to provide an argument validity checking method that solves the above-mentioned drawbacks of the conventional methods and that does not place much burden on code generation by the compiler and does not make the target program unnecessarily large. It is to provide.

Structure of the Invention The argument validity checking method according to the present invention includes an argument attribute storage means having an information table regarding the attributes of the argument to be passed to the subroutine in the object program of the program that calls the subroutine, and an argument attribute storage means in the object program of the called subroutine. , a formal argument attribute storage means having an information table regarding the attributes of the formal argument to be received, and an inspection means for reading and comparing the image information table at the time of execution, and pointing out any discrepancies.

Example The present invention will be explained based on FIG.

Program 1 does not call subroutine 2.

Argument group A that this cutting program 1 passes to subroutine 2
An information table T1 regarding attributes of program 1 is generated in the target program of program 1 by the compiler. Similarly, an information table T2 regarding the attributes of the formal argument group A2 to be received from the program 1 in the target program of the subroutine 2 is generated by the compiler. In fact, by calling routine R1 that reads the information table immediately before program 1 calls subroutine 2, table 1 is
is loaded.

Immediately after subroutine 2 is called, a new table T2 is read by calling routine R2 that checks the information table, and by comparing it with the previously read table T1, the validity of the attributes of the argument and formal argument is determined. Gender is tested. This processing procedure is shown in FIG.

In step 100 of FIG. 2, after performing normal processing on argument group A1, and before passing control to subroutine 2 in step 300, routine R1 reads an information table in step 200.
The information table T1 is read by calling . Next, after control is passed to the subroutine in step 300,
Routine R2 is called in step 310 to examine the information table. FIG. 3 shows the processing of routine R2 that checks the information table called in step 310.

In the routine R2 that inspects the arguments in Figure 3, first step 3
11, the information table T2 is read.

Next, in step 312, the contents of the information table T1 previously read in step 200 of FIG. 2 are compared with the contents of this information table T2. If there is no contradiction between the contents of the two information tables, the process returns to the procedure shown in FIG. In A2, output how the attribute of the argument number is inconsistent, display that there is an error in the attribute between the arguments, and then output the second argument.
Returning to step 400 in the figure, normal processing continues.

Effects of the Invention As described above, the present invention checks the validity of arguments at runtime, thereby eliminating the drawbacks of the method performed at link time and checking the validity only for arguments specifically used at runtime. It has the advantage of being able to be inspected.

In addition, since the target program only includes the argument information table, the code that calls the routine that reads the information table, and the code that calls the routine that inspects the information table, it does not put much burden on the compiler's code generation, and the target program It also has the advantage that it does not become unnecessarily large. Furthermore, the processing routines after detecting an argument validity error are centralized, including controlling the display of the error and whether to ignore the error and continue processing. Easily scale processing.

It has the advantage of being changeable.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is an overall configuration diagram of an embodiment of the present invention, and FIGS. 2 and 3 are flowcharts of a processing method of the embodiment of the present invention.

Claims (1)

[Claims] When a program calls a subroutine, an argument validation method that checks the validity between the attributes of the argument passed to the subroutine and the attributes of the formal argument that the subroutine should receive, the method comprising: Argument attribute storage means for storing an information table regarding attributes, formal argument attribute storage means for storing an information table regarding attributes of formal arguments to be received by the subroutine, and comparison means for comparing the contents of both the information tables, An argument validity checking method, characterized in that when the subroutine is called, the contents of both the information tables are compared and checked by a comparing means.