JPH0690669B2 - Character string manipulation instruction compilation processing method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Outline] When an intermediate text is a subscript when translating a character string manipulation instruction whose subscript evaluation time is determined to be performed only once at the beginning of instruction execution, The object is immediately expanded, and if it is other than the subscript, it is temporarily stored in the work text storage unit, and when the intermediate text for the instruction is finished, the object is sequentially read from the work text storage unit and the object is expanded. ， With a simple processing configuration, subscript evaluation can be performed first.

[Industrial application field]

The present invention relates to a character string manipulation instruction compilation processing method such as a COBOL compiler.

When handling tables in COBOL, which is one of the computer languages,
It is possible to use a unique name as a subscript. The subscript evaluation time is determined for each instruction unit according to the COBOL standard. Even when the subscripted data is described in the instruction and the subscript evaluation time are different, it is necessary to easily and uniformly expand the subscript object at a predetermined position. .

[Conventional technology]

 FIG. 5 is a diagram for explaining the problems of the conventional method.

For example, a COBOL compiler inputs a source program written in COBOL and translates it into an executable program (hereinafter referred to as an object module). The compiler consists of a parsing phase and a code generation phase, with two phases:
A special code called intermediate text intervenes.

By the way, COBOL source programs handle data called "table operations". It collects multiple data formats and handles them as one data. Indexing or subscripting is performed to refer to each element. Usually COBOL
In each instruction of, the index and subscript may be evaluated at the time when the data is described, so the parsing phase is
When the index or subscript appears, the intermediate text for subscript evaluation is output, and the code generation phase also expands the object when the intermediate text for subscript evaluation is read.

However, for example, ANS COBOL (American National Stan
dard COBOL) 1985 standard, so-called INSPECT instruction, ST
Character string manipulation commands such as RING command and UNSTRING command.
The subscript for a unique name is a standard that is evaluated only once at the beginning of execution of an instruction. The character string manipulation commands are
It is divided into multiple intermediate texts, and in the parsing phase, the intermediate texts are created without being aware of the evaluation time of the subscript of the character string manipulation type instruction, so the objects are expanded in the order of reading the intermediate texts as described above. In this case, an object module that does not meet the above-mentioned character string manipulation instruction standard will be generated.

This will be described in more detail with reference to FIG.

For example, for the instruction shown in FIG. 5 (a), in the syntax analysis phase, a plurality of intermediate texts as shown in FIG. 5 (b) are created without paying attention to the subscript evaluation time for each instruction. Therefore, in the code generation phase, it is necessary to expand the object in consideration of the subscript evaluation time specified by the instruction.

That is, for example, in the instruction shown in FIG. 5A, it is assumed that there is an instruction in which the result of the action is copied to the data shown in FIG. 5A and then the result of the action is copied to the data shown in FIG. . When the result of the action is transcribed to the data of (a), the values of A and B are naturally changed, so when transcribing the action result to the data of (b), the value before the change is used as the subscript value. This is because the result will change greatly depending on whether you use it or the changed value.

However, according to the conventional method, the code generation phase is configured by a processing unit that is a subroutine such as input of intermediate text, object expansion of each instruction, and object expansion related to subscript evaluation, and the intermediate text related to subscript usually appears. At this point, the processing unit for the object expansion related to the subscript evaluation is called, and in other cases, the object expansion processing unit for each instruction is called, so that the object expansion order is as shown in FIG. It becomes the order of the sequence of the intermediate text. Therefore, if the subscript and the evaluation time are only once at the beginning of the execution of the instruction, such as the above-mentioned character string manipulation type instruction, if the above processing units are unconditionally called, the object that does not conform to the standard Will be generated.

[Problems to be solved by the invention]

As a means for solving the above problem, the first conceivable solution is to make the order of the intermediate texts created by the parsing phase conform to the standard for each instruction. However, creating an intermediate text in consideration of the subscript evaluation time for each instruction has a problem in that the parsing phase is greatly modified and systematic processing is difficult.

Next, in the code generation phase, it is conceivable to process with special consideration only when a character string manipulation type instruction appears. In this case, not only the main processing unit, but also the processing unit for expanding the object relating to the subscript evaluation must be specially aware of the character string manipulation type command. In other commands, the main processing unit uses the intermediate text. After processing the object expansion related to input and subscript evaluation, the processing unit for each instruction is called, whereas in the character string manipulation type commands, that processing unit performs processing from input of intermediate text to object expansion related to subscript evaluation. The function must be covered, which causes a problem that the processing load is extremely large.

The present invention solves the above-mentioned problems, and means for performing uniform object expansion with a simple processing configuration even when the subscript evaluation time is different from the time when the subscripted data is described in the instruction. Is intended to provide.

[Means for solving problems]

 FIG. 1 shows a block diagram of the principle of the present invention.

In FIG. 1, 10 is a source program written in a high-level computer language, 11 is a processing unit consisting of a CPU and memory, 12 is a compiler that translates the source program 10 into machine language code, 13 is a syntax analysis phase, and 14 is Intermediate text, 15
Is a code generation phase that generates an object code from intermediate text, 16 is a subscript determination unit that determines subscripts used for table operations, 17 is a subscript object expansion processing unit that expands subscript objects, and 18 is a character string operation system A character string operation type instruction object expansion processing unit that expands an instruction object, 19 is another instruction object expansion processing unit that expands an object for each general instruction in units of instructions, 2
0 is a work text storage unit such as a work memory or an external storage device, 21 is an instruction end determination unit that determines the end of one instruction,
22 is an intermediate text writing unit that writes the intermediate text to the working text storage unit, 23 is an object expansion unit that reads the object from the working text storage unit 20 and expands the object, and 24 is an object that stores the code of the object that is the translation result Represents a module.

In the present invention, regarding subscripts appearing in character string manipulation commands,
In order to realize the evaluation only once at the beginning of the instruction, the following processing is performed.

In the syntax analysis phase 13, the source program 10 is read, and the analysis result is output as an intermediate text 14, which is an internal code, while checking the syntax and the like. This process is similar to the conventional one.

The code generation phase 15 reads the intermediate text 14 and outputs the object module 24. At this time, if the read intermediate text 14 is related to the subscript, the subscript determination unit 16 determines that the subscript object expansion processing unit 17 is activated and the subscript object expansion processing unit 17 is executed.
Outputs the machine language code for evaluating the subscript to the object module 24. If it is not a subscript, the instruction object expansion processing unit that is composed of a subroutine that expands the object for each instruction is called, but if it is related to a character string operation type instruction, it is a character string operation type instruction object. The expansion processing unit 18 is called.

The character string manipulation type instruction object expansion processing unit 18 calls the intermediate text writing unit 22 until the command end determination unit 21 determines the end of the character string manipulation type command, and sequentially stores the intermediate text 14 and the working text storage unit. Output to 20.
The work text storage unit 20 temporarily stores intermediate texts other than those related to subscripts. When the instruction end determination unit 21 detects the end of the instruction, the object expansion unit 23 is called and the object expansion unit 23
Reads the intermediate text sequentially from the work text storage unit 20, develops it into an object, and outputs it to the object module 24.

[Action]

According to the present invention, if the evaluation time of the subscript is good in the order of the arrangement of the intermediate text 14, the object expansion is performed in order as in the conventional case. For example, for a character string manipulation instruction in the ANS COBOL 1985 standard, when the subscript evaluation time must be the first of the instruction, the subscript object expansion processing unit 17 immediately causes the object module Although the output to 24 is performed, the intermediate text in the middle of the other character string manipulation commands is written to the work text storage unit 20 and is restarted when the intermediate text of the character string manipulation commands is completed. ，
Objects are expanded by reading from the work text storage unit 20. Therefore, it is possible to set the subscript evaluation time to the beginning of the instruction without the intermediate text being arranged, without significantly changing the conventional processing configuration.

〔Example〕

FIG. 2 is an explanatory diagram of a code generation phase process according to an embodiment of the present invention, FIG. 3 is an explanatory diagram of a character string operation type instruction object expansion process according to an embodiment of the present invention, and FIG. It is an explanatory view of an example.

An outline of the process of the code generation phase 15 shown in FIG. 1 is as shown in FIG. 2, for example. The numbers 1 to 4 in the following description correspond to the processes shown in FIG.

 Read the intermediate text one record at a time.

When there are no more intermediate texts to read, the process ends.

It is determined whether the read intermediate text record is related to a subscript. If it is a subscript, the process is executed. If it is not a subscript, the process is executed.

Expand the object whose subscript is evaluated. afterwards,
The control is returned to the processing and the processing is repeated in the same manner.

The object expansion processing unit prepared for each instruction is called and the object is expanded respectively. afterwards,
Return control to processing.

Character string manipulation instruction object expansion processing unit shown in FIG.
18 is processed as shown in FIG. 3, for example. This process is one of the processes shown in FIG. 2 described above. The numbers ~ in the following description correspond to the processes ~ shown in Fig. 3.

It is determined whether the arrangement of the intermediate text is completed.
When finished, transfer control to processing.

If it is not finished, the intermediate text record is output as it is to the work text storage section 20 shown in FIG.
It then returns control to the caller to read the next intermediate text.

When the arrangement of the intermediate texts is completed, the intermediate texts written in the above process are read from the work text storage unit 20 one record at a time.

The end of the text in the working text storage unit 20 is determined, and when there is no more text, control is returned to the calling source for processing the next instruction.

With respect to the intermediate text read from the work text storage unit 20, the object of the character string manipulation type instruction is expanded. After that, the control is returned to the process, and the process is repeated.

Next, a simple concrete example will be described with reference to FIG. Note that this example will be the same even if the subscript evaluation is not at the beginning of the instruction, but the subscript evaluation will be described as an example that must be performed at the beginning of execution of the instruction.

The character string manipulation instruction in the source program 10 shown in FIG. 4 is a so-called STRING instruction according to the standard of ANS COBOL 1985, and data names A, B, and C are engaged, and a data name D is obtained. It is an instruction to instruct (A) to transcribe the operation result. The intermediate text 14 is output by the analysis in the syntax analysis phase 13. In the intermediate text 14, “BEGIN” is the start of the instruction, “STR1” and “STR2” are the detailed codes of the character string manipulation instructions, “INDEX” is the subscript, and “E”.
"ND" indicates the end of the instruction.

In the code generation phase 15, the intermediate text 14
Each record is read, and records other than those relating to subscripts are written to the work file 40 as in (1) to (3) and (5). As for (INDEX), a code for subscript evaluation is immediately generated and output to the object module 24 as in (4).

When the end of the instruction is detected by "END", as shown in (6), the intermediate text is read record by record from the work file 40, the object is expanded, and the result is output to the object module 24.

In addition to the STRING instruction, the UNSTRING instruction that decomposes a character string and the INSPECT instruction that replaces a character string and counts the number of characters are also used as character string manipulation-related instructions at the beginning of the instruction.

〔The invention's effect〕

As described above, according to the present invention, even when the subscripted data is described in an instruction and the subscript evaluation time is different, the object expansion is performed uniformly by a simple processing configuration. It will be possible.

[Brief description of drawings]

FIG. 1 is a block diagram of the principle of the present invention, FIG. 2 is an explanatory diagram of a code generation phase process according to an embodiment of the present invention, and FIG. 3 is a character string operation type instruction object expansion process according to an embodiment of the present invention. FIG. 4 is an explanatory view of an embodiment of the present invention, and FIG. 5 is an explanatory view of problems in the conventional method. In the figure, 10 is a source program, 11 is a processing device, 12 is a compiler, 13 is a syntax analysis phase, 14 is an intermediate text, 15 is a code generation phase, 16 is a subscript determination unit, 17 is a subscript object expansion processing unit, and 18 is Character string operation type instruction object expansion processing unit, 19 is another instruction object expansion processing unit, 20 is a work text storage unit, 21 is an instruction end determination unit, 22 is an intermediate text writing unit, 23 is an object expansion unit, and 24 is Represents an object module.

Claims (1)

[Claims] 1. A compile processing method of a source program having a specification of a character string manipulation type instruction specified to perform subscript evaluation at the beginning of instruction execution, wherein an intermediate text generated by a syntax analysis phase is In the compile processing method that expands to an object in the code generation phase, the work text storage unit (20) in which the above intermediate text is temporarily stored, and the intermediate text read when the intermediate text is read in the above code generation phase are subscripted. The subscript determination means (16) for determining whether or not it is related to the subscript, and the subscript object expansion processing means (17) for expanding the object regarding the subscript when the read intermediate text is a subscript, and the read intermediate text If it is related to the above string manipulation commands other than the subscript, the command The read intermediate text is written in the working text memory until the arrangement of the intermediate texts related to the command is completed, and
When the arrangement of the intermediate texts related to the command is completed, the intermediate texts are sequentially read from the work text storage unit and the object expansion related to the character string operation related command is executed.
8) A character string manipulation instruction compile processing method characterized by having and.