JPS62137637A - Arithmetic accuracy setting system in language processing system - Google Patents

Abstract

PURPOSE:To make it possible to change accuracy of calculation of objective program of an arithmetic expression only by changing registered information by making conversion to objective program using arithmetic accuracy information consisting of number of digits of integer part and number of digits of decimal part registered beforehand. CONSTITUTION:Arithmetic accuracy information registered in a storage device 1 is inputted by an inputting means 2. Then, number of digits of decimal part and integer part of storing area are determined and allotted by an intermediate result allotting means 5 to store interim result of operation used in an objective program from the arithmetic accuracy information. An intermediate code consisting of two lines of talk made by expanding an arithmetic expression to the relation of operator and operand is inputted by an intermediate code inputting means 4. Further, inputted intermediate codes are processed successively from the head, and an objective program 4 using allotted interim result storing area is formed.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to an operation in a language processing system that converts a source program into a target program, and determines the precision of the calculation of an intermediate result of an arithmetic expression in a source program. Regarding accuracy setting method.

[Conventional technology]

Conventionally, this type of arithmetic precision setting method does not have an input means for externally inputting arithmetic precision information, and only calculates the minimum necessary precision from the number of digits in the integer part and decimal part of all the arithmetic numbers that appear in the intermediate code. The number of digits of the integer part and decimal part of the intermediate result storage area is determined from this accuracy, and the target program is converted to use the intermediate result storage area.

Also, the method of calculating the calculation accuracy from the number of digits of the calculation number in the intermediate code differs depending on the language processing system, and the result is
The calculation accuracy was also unique to each language processing system.

FIG. 3 is a block diagram showing the configuration of the conventional method. Referring to FIG. The stored storage device 13, the intermediate code input means 14 that inputs this intermediate code, the intermediate result determination means 8 that determines the accuracy of the intermediate result storage area, and the calculation from the accuracy determined by the intermediate result determination means 8 when executing the target program. Create and allocate an intermediate result storage area, which is a work area on the main memory or register necessary for storing intermediate results. It is comprised of a result storage area and an object program generation means 16 that generates an object program from the intermediate code inputted by the intermediate code input means 14.

FIG. 4 is a flowchart showing the flow of processing in the conventional method.

In FIG. 4, in step 41, the intermediate code input means 14 inputs an intermediate code consisting of a token string in which an arithmetic expression in the source program is expanded into a relationship between operators and arithmetic numbers. Next, in step 42, all the tokens of the operation number in the intermediate code input in step 41 are examined using the intermediate result determination means 8 to determine the accuracy of the intermediate result storage area.

Furthermore, in step 43, the intermediate result storage area is allocated using means 15 based on the accuracy of the intermediate result storage area determined in step 42. At step 44, the intermediate code input at step 41 is processed sequentially from the beginning to generate a target program using the intermediate result storage area created at step 43.

[Problem that the invention seeks to solve]

In this conventional method, when it is necessary to change the calculation accuracy, it is necessary to be familiar with the complicated conventional calculation accuracy setting method, and then change the settings in the intermediate code to obtain the final calculation accuracy. The first problem is that the number of digits in the calculation number must be changed, which requires a great deal of effort and time.

one! In the process of converting arithmetic expressions in the source program to the target program, the language processing system determines the arithmetic precision from the number of operations in the intermediate code even if the necessary arithmetic precision has already been determined. There is a second problem that processing time for determining calculation accuracy is wasted.

Furthermore, between different language processing systems, the operation of the converted target program differs due to the difference in the precision of arithmetic expressions even though they are the same source program, and the behavior of the converted target program differs in time and
There is a third problem, which is that it will cause a huge economic blow.

Therefore, an object of the present invention is to provide a means for inputting arithmetic precision information consisting of the number of digits in the integer part and the number of digits in the decimal part, and to convert the program into a target program using an intermediate result storage area determined from the input arithmetic precision information. By doing this, it is possible to change the arithmetic precision without changing the source program for a source program that requires a change in arithmetic precision, and it also improves the processing speed of the language processing system. By giving fixedly from outside the language processing system,
An object of the present invention is to provide an arithmetic precision setting method for a language processing system that can improve compatibility between different language processing systems.

[Means for solving problems]

The method of the present invention is an arithmetic precision setting method for a language processing system that converts a source program into a target program.
1. Accuracy information input means for inputting arithmetic precision information consisting of the number of digits of an integer part and a decimal part preset on a storage device, and creation of an intermediate result storage area from the arithmetic precision information inputted by the precision information input means. An intermediate result allocation means for performing 2 allocation, and an intermediate code input means for inputting an intermediate code consisting of a token string obtained by expanding an arithmetic expression in a source program stored in a storage device in advance into a relationship between operators and operands. , the intermediate result allocation is configured to include an intermediate result storage area allocated by the means and an object program generation means for generating the object program of the arithmetic expression from the intermediate code input by the intermediate code input means. Embodiments Next, the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention.

The arithmetic precision setting method in the language processing system shown in FIG. Intermediate result allocation is a process that creates and allocates an intermediate result storage area, which is a work area in main memory or registers, necessary to store intermediate results of calculations during execution of the target program from input calculation accuracy information. Means 5
, a storage device 3 storing an intermediate code consisting of a token string in which an arithmetic expression in a source program is expanded into a relationship between operators and arithmetic numbers, an intermediate code input means 4 for inputting this intermediate code, and an intermediate code input means 4 for allocating intermediate results. It is comprised of an object program generation means 6 for generating an object program from the intermediate result storage area created by the means 5 and the intermediate code inputted by the intermediate code input means 4.

FIG. 2 is a flowchart showing the flow of the process shown in FIG.

In FIG. 2, in step 21, the arithmetic accuracy information registered on the storage device 1 is first inputted by the accuracy information input means 2, and then in step 22, the arithmetic accuracy information input in step 21 is used to determine the arithmetic operation to be used in the target program. Determining and allocating the number of digits of the small part and integer part of the intermediate result storage area necessary for storing the intermediate result is performed by means 5 for allocating the intermediate result.

Next, in step 23, the intermediate code input means 4 inputs an intermediate code consisting of a token string in which an arithmetic expression is expanded into a relationship between operators and arithmetic numbers. Furthermore, in step 24, the input intermediate code is processed in order from the beginning, and a target program is generated using the intermediate result storage area allocated in step 22.

The above is the explanation of the processing flow of this embodiment. As can be seen from the flow of this process, the information necessary to determine the intermediate result storage area used by the target program is only the arithmetic precision information stored in advance on the storage device 1, so this storage device The number of digits in the integer part and the number of digits in the decimal part of the intermediate result storage area can be changed simply by changing the calculation precision information on l. By changing the arithmetic precision information on the shredded storage device 1, the number of digits in the integer part and the number of digits in the decimal part of the intermediate result storage area can be arbitrarily determined.
It is possible to convert a target program that operates with the same calculation accuracy as a target program converted by another language processing system with a different calculation accuracy setting method by simply changing the calculation accuracy information.

Comparing the processing flow of the embodiment of the conventional method shown in FIG. 4 and the processing flow of the embodiment according to the present invention shown in FIG. 2, in the conventional method shown in FIG. In contrast to determining the accuracy of the intermediate result storage area by examining all the results, in the present embodiment shown in FIG. 2, it is only necessary to input the calculation accuracy information in step 21, so that the processing time can be shortened.

〔Effect of the invention〕

As explained above, the calculation precision information can be changed simply by converting it to the target program using the calculation precision information, which consists of the number of digits of the integer part and the number of digits of the decimal part, registered in advance on the storage device. It is possible to change the arithmetic precision of the target program for arithmetic expressions, and also improve the processing speed of the language processing system (Figure 9). Also, by providing fixed arithmetic precision information from the outside, it is possible to improve compatibility between different language processing systems. It has the effect of improving sexual performance.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention;
FIG. 2 is a flowchart showing the operation of FIG. 1, FIG. 3 is a block diagram showing the configuration of the conventional system, and FIG. 4 is a flowchart showing the operation of FIG. 1...Storage device, 2...-Accuracy information input means, 3゜13...Storage device, 4, 14...
... Intermediate code input means, 5.15 ... Intermediate result allocation means, 6, 16 ... Purpose program generation means, 7 ... Purpose program, 8 ...・・・
...Means for determining intermediate results. -1\ Agent Patent Attorney Susumu Uchihara; 1 Cause 2 Figure 3 Figure -f; 4 Figure

Claims (1)

[Claims] In an arithmetic precision setting method in a language processing system that converts a source program into a target program, the precision is to input arithmetic precision information consisting of the number of digits of an integer part and a decimal part set in advance on a storage device. information input means; intermediate result allocation means for creating and allocating an intermediate result storage area from the arithmetic accuracy information input by the accuracy information input means; intermediate code input means for inputting an intermediate code consisting of a token string developed in the relationship between 1. An arithmetic precision setting method comprising: objective program generation means for generating an arithmetic expression objective program.