JP2585790B2 - Compile processing method for multiple code systems - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Summary] Compile processing corresponding to a plurality of code systems in a computer system having a compiler that generates a target program by inputting a source program and at least processing including syntax analysis, semantic analysis and code generation Regarding the method, the source programs of different code systems can be divided into
The purpose is to provide a means of translating with common processing logic. At the start of compilation, a binary code correspondence table corresponding to some specific characters is stored in a certain area at the start of compilation according to the code system of the source program. It has a correspondence table setting processing unit to set, and in the syntax analysis, semantic analysis and code generation processing, the binary code of a specific character in the code system of the source program depends on the code system by referring to the correspondence table. It is configured so that it can be used with no processing logic.

[Industrial applications]

The present invention relates to a compile processing method corresponding to a plurality of code systems in a computer system having a compiler that generates a target program by inputting a source program and performing at least a process including a syntax analysis, a semantic analysis, and a code generation.

With the diversification of computer systems in recent years, it has become necessary to develop compilers for many systems. Compilers with the same specifications are required between these different systems.

To do so, a compiler can be created on one system and ported to another system, but these are not necessarily the same code system. Therefore, in order to reduce the development burden, it is necessary to consider a compiler having processing logic independent of the code system.

[Conventional technology]

 FIG. 4 shows an example of the prior art.

In general, with a compiler such as COBOL or FORTRAN, as shown in Fig. 4, a source program 10 is input, lexical analysis, syntax analysis, semantic analysis, optimization, and code generation are performed. The target program 11 is output.

In lexical analysis in a conventional compiler, a table of reserved words and codes was originally provided, and by replacing these based on the results of the judgment processing of the codes, it was possible to handle a plurality of codes.

However, in the processes of syntax analysis, semantic analysis, and code generation, there are relatively few parts that depend on the code system.
By judging the current code system, the processing is divided and the code for each code system is properly used. The optimization process does not generally depend on the code system.

[Problems to be solved by the invention]

That is, in the prior art, in the processing of syntactic analysis, semantic analysis or code generation, for example, when a reserved word ZERO is coded, the code system is determined as shown in FIG. 4 and the EBCDIC code system is determined. If so, use hexadecimal "FO". If ASCII code, use hexadecimal "30".
In order to handle a plurality of code systems, processing logic was used.

Therefore, the processing logic becomes complicated, and not only does the compiler program scale increase, but also the processing logic must be changed in various places when a new source program of a different code system is to be processed. There was a problem that was large.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above problems and to provide means for translating source programs of a plurality of different code systems with a common processing logic while minimizing processing separation according to a difference in a code system. I have.

[Means for solving the problem]

 FIG. 1 shows a configuration example of the present invention.

In FIG. 1, 10 is a source program described in a high-level language such as COBOL or FORTRAN, 11 is a target program expanded into an object code, 12 is a processing device including a CPU and a memory, and 13 is a translation of a computer language. 14 is a start / control unit that starts / controls each processing unit in the compiler, 15 is a correspondence table setting processing unit, 16 is a parsing unit that performs syntactic analysis, and 17 is a lexical unit that interprets high-level language lexical units. An analysis unit, 18 is a semantic analysis unit that analyzes the semantics according to the syntax analysis result, 19 is an optimization unit that performs optimization for generating efficient object code, and 20 is an object code that generates the object code and executes the target program 11. Output code generator, 2
Reference numeral 1 denotes a common area secured on the memory, and reference numeral 22 denotes a correspondence table that stores a correspondence between a specific character and its code.

In the present invention, the activation / control unit 14 in the compiler 13
In addition, a correspondence table setting processing unit 15 is provided. At the start of compiling, the correspondence table setting processor 15 sets a correspondence table 22 corresponding to the code system of the source program 10 in the common area 21 secured as a work area on the memory.

The correspondence table 22 is a table having information on what kind of binary code, when defeated, a specific character such as zero, space, or quotation mark is represented in the code system of the source program 10.

The activation / control unit 14 controls each processing unit of the syntax analysis unit 16 or the code generation unit 20 according to each compilation phase.
Start up sequentially. Here, in the present invention, in particular, in the parsing unit 16, the semantic analyzing unit 18, and the code generating unit 20, it is necessary to develop a code of a specific character for processing of a figurative constant, processing of a padding character, and the like. In this case, the binary code corresponding to the specific character is extracted by referring to the corresponding area in the correspondence table 22 without using a different processing logic depending on the code system.

[Action]

Normally, the syntax analyzer 16, semantic analyzer 18, and code generator 20 of the compiler 13 include processing depending on the code system, such as processing of figurative constants and processing of padding characters. Is very small compared to lexical analysis processing by. Also, the code that requires the separation of processing is limited.

Therefore, it is necessary for the syntax analysis unit 16, the semantic analysis unit 18, and the code generation unit 20 to be aware of only the code of a specific character. In the present invention, these are set in the common area 21 in advance as the correspondence table 22. By referring to the code corresponding to each specific character in this area in each processing unit, common processing logic can be used for different code systems.

As a result, a new source program 10
Can be handled by changing only the code set in the correspondence table 22 in the correspondence table setting processing unit 15 without changing the processing logic.

〔Example〕

FIG. 2 shows an example of a character / code correspondence table according to the embodiment of the present invention, and FIG. 3 shows an example of code generation according to the embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described using a COBOL language compiler as an example. Compiler is EBCDIC code system and ASCII
It should be able to handle code systems. Of course, the present invention can be similarly applied to other languages and code systems.

In the COBOL compiler, there are few processes related to the code system in the syntax analysis unit 16, semantic analysis unit 18, and code generation unit 20 shown in Fig. 1, and the following functions mainly relate to the code system. It is limited to processing.

-Processing of figurative constants Processing of reserved words representing specific characters such as ZERO, SPACE, and QUOTE.

-Processing of padding characters This is a process of filling unused areas with specific characters, such as blanking in transcription.

When processing such figurative constants and padding characters, it is necessary to use different binary codes for each code system. These codes are prepared in a predetermined area in advance, and by referring to that area, The processing logic does not depend on the code system except for the process of setting the code in the correspondence table which is performed only once at the beginning.

FIG. 2 shows an example of a correspondence table 22 having code information corresponding to a specific character according to each code system.

Fig. 2 (a) shows the correspondence table 22 used in the EBCDIC code system.
, And has correspondence information of hexadecimal F0, 40, 7F,... For numbers (zero), characters (blanks), characters (quotation marks),.

On the other hand, FIG. 2 (b) is a correspondence table 22 used in the ASCII code system. In this correspondence table 22, numerals (zero), characters (blanks), characters (quotation marks),. , 16
Each code of 30, 20, 22, ... in radix is associated.

First, depending on whether the correspondence table 22 shown in FIG. 2 (a) is set or the correspondence table 22 shown in FIG. 2 (b) is set, the correspondence table 22 for each specific character is set. Since the position is fixed, EBCDIC code system and common processing logic
The code to be used can be used properly with the ASCII code system.

For example, assume that the COBOL source program 30 shown in FIG. 3 is compiled.

Assuming that this COBOL source program 30 is represented by an EBCDIC code system, a correspondence table 22 shown in FIG.
Prepare

“MOVE SPACE TO A” is a sentence that means to transfer a blank to the area A.

In the EBCDIC code system, since the blank is 40 in hexadecimal in the correspondence table 22, a code like the object code 31A is generated by referring to this. The object code 31A is an instruction to move a 1-byte code having a value of 0x40 to the address of a (a1) indicating the area of A.

On the other hand, assuming that the COBOL source program 30 is represented by an ASCII code system, the correspondence table 22 shown in FIG.
Is used to generate an object code 31B. This instruction instructs to move a 1-byte code having a value of 0x20 representing a space in the ASCII code system to the address of a (a1).

Note that the object codes 31A and 31B are actually binary machine language instructions.

Here, an example of handling two code systems has been described. However, by expanding or replacing the correspondence table 22,
Similarly, it can be applied to other code systems. That is, it is possible to correspond to many code systems only by changing the setting constants in the correspondence table setting processing unit 15 shown in FIG.

〔The invention's effect〕

As described above, according to the present invention, a code of a specific character is set in an area referred to by a compiler in advance, and in a process depending on a code system, the code is referred to by referring to this area. The processing logic can be independent of the system.

As a result, when a new source program of a different code system is to be processed, the processing logic is not changed.
It can respond and greatly contributes to improvement of development efficiency when developing compilers of different code systems in common.

[Brief description of the drawings]

FIG. 1 is an example of the configuration of the present invention, FIG. 2 is an example of a character / code correspondence table according to an embodiment of the present invention, FIG. 3 is an example of code generation according to the embodiment of the present invention, and FIG. Here is an example. In the figure, 10 is a source program, 11 is a target program, 12 is a processing device, 13 is a compiler, 14 is a startup / control unit, 15 is a correspondence table setting processing unit, 16 is a syntax analysis unit, 17 is a lexical analysis unit, 18 Is a semantic analysis unit, 19 is an optimization unit, 20 is a code generation unit, 21 is a common area,
22 represents a correspondence table.

Claims (1)

(57) [Claims] 1. A computer system having a compiler for inputting a source program (10) and generating a target program (11) by at least processing including syntax analysis, semantic analysis, and code generation. The correspondence table (22) that stores the binary code for a specific character given the same meaning in each code system in a common position in each code system according to each code system according to In the correspondence table setting processing section (15), which sets the area that can be commonly referenced by the section, and in the syntax analysis, semantic analysis, and code generation processing,
A processing unit (16, 18, 20) that uses the binary code of the specific character in the code system of the source program with processing logic that does not include the determination logic of the code system by referring to the correspondence table A compilation processing method that supports multiple code systems.