JP3007356B2 - Compiling device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Summary] Regarding the processing of a compiler for translating a source program of a computer into a target program, a constant storage area in the target program for holding constant data described in the source program is provided. When a source program is translated to generate a target program to be executed by a computer for the purpose of a constant registration processing method that can be reduced, data representing a constant described in the source program is provided in the target program. In the processing of the computer for registering in the storage area, a search is made for a match between the leading continuous part of the data to be newly registered and the trailing continuous part of the data registered in the constant storage area. Is obtained, and if the data to be registered has a remaining portion excluding the leading continuous portion, only the remaining portion is stored at the end of the constant storage area. Configured to register in the region following the continuous portion.

[Industrial applications]

The present invention relates to a process of a compiler for translating a source program of a computer into a target program, and more particularly to a constant registration processing method for registering constant data described in the source program in a constant storage area on the target program.

In the process of translating a source program written in COBOL programming language etc. on a computer and generating a target program to be executed on a required computer, if constants such as numerical values are described in the source program, Collecting and registering the data of these constants in a constant storage area provided on the target program, and generating the target program so as to refer to data of a required portion in the constant storage area in a required program step, As is well known.

[Conventional technology]

FIG. 2 shows an example of the configuration of a computer for performing translation processing of a source program. A compiler 1 executed by the computer inputs and translates the source program 2, and a machine language generation unit 3 converts a required machine language of the computer. And outputs the target program 4 converted into the target program.

A source program 2 shown in FIG. 3A is an example showing a part of a program described in the COBOL language. For example, as is well known, a program statement 10 in the
5 "is set. Here, the variable A is defined by the program statement 15. Since this program statement 15 indicates that the variable A is three-digit numeric data, the compiler 1
Correspondingly takes a 3-byte storage area (one-digit numerical data is assumed to be 1 byte) as a variable A on the target program 4 and stores 3-byte data representing a constant "345" as appropriate. A program in a machine language for setting the contents in the storage area and transferring the contents of the storage area for the constant to the storage area for the variable A is generated.

For this purpose, when the machine language generation unit 3 in FIG. 2 detects a constant described in the source program 2, the machine language generation unit 3 notifies the constant registration processing unit 5 of the registered constant and requests constant registration for setting the constant in the storage area. As a result of the processing, the address of the constant storage area is received and used as the address of the transfer source storage area of the machine language program for data transfer.

The constant registration processing unit 5 has a constant storage area 6 in a work area. When a request for constant registration is received, the constant registration processing unit 5 stores registration constants sequentially from the top of the constant storage area 6 and stores the start address of the stored registration constant. Returned to the machine language generator 3.

However, at this time, the data already registered in the constant storage area 6 is compared with the registered constant, and if a part or all of the data completely matches the constant, new registration is not performed and the registered data The matching part is used, and the head address of the matching part is returned as the address of the registration destination of the constant. Thus, the constant storage area 6 is reduced.

For example, when there are constant descriptions such as the program statements 10 to 13 in the source program 2 in FIG. 3A, the machine language generation unit 3 requests sequential constant registration in the order of appearance in the program. The processing unit 5 first registers “345” at the beginning of the constant storage area 6. Next, when there is a request to register “789” described in the program statement 11, “789” and “345” registered in the constant storage area 6 are compared and do not match. Additional registration is made at the following position. However, since the constant "345" in the next program statement 12 matches the registered data, no additional registration is performed and the first three digits of data are used as they are.

The constant "457891" in the next program 13 matches the registered data up to "45789", but since a perfect match was not obtained, "457891" was additionally registered, and the constant "345789" in the program statement 14 was already registered No additional registration is made because there is a completely matching portion in the data that has been entered.

FIG. 4 shows a flow of processing of the constant registration processing unit 5. When a constant registration request is received, data registered from the constant storage area 6 is taken out one digit at a time and is sequentially searched. Then, in processing step 20, it is determined whether or not unsearched data remains in constant storage area 6, and if not, processing step 29 immediately (including no registration).
Then, the top address of the area registered in the processing step 30 is returned to the machine language generation unit 3.

If there is data in the constant storage area 6, the processing step
At 21 the head address of the remaining data is stored in a working register, and at step 22 one byte of data at that address is read.

The first byte of the constant received in the processing step 23 is compared with the data read from the constant storage area 6, and if they do not match, the processing returns to the processing step 20, and the constant storage area 6 The end of the
In this way, the data that matches the first byte of the registered constant is searched from the beginning of the data registered in the constant storage area 6.

As a result, if there is matching data, the process proceeds from the processing step 23 to the processing step 24, where it is determined whether the registered constant has the next digit (the second byte in this case). In the case of (1), the case of a 1-byte constant) is a case where a part that completely matches the registered constant is present in the registered data. Is returned to the machine language generation unit 3.

If the next digit remains in the registered constant, it is identified in processing step 25, and if there is the next data in constant storage area 6, the next one byte is read out in processing step 26, and processing step 27
Compare with the next byte of the registration constant. If the result is a mismatch, the process returns to the processing step 20. If the registered data still remains, the same processing as described above is repeated from the next byte.

If a match is obtained in the processing step 27, the process returns to the processing step 24, and the comparison between the next byte of the registered constant and the next byte of the data registered in the constant storage area 6 is performed. As long as it continues, repeat until one of the data ends. As a result, if the registered constant ends earlier, there is a case where a part completely matching the registered constant is present in the registered data.
To end the process.

If the data already registered in the constant storage area 6 ends earlier, there is no completely matched part, and the processing proceeds from the processing step 25 to the processing step 29 to additionally register the registered constant as described above. .

When the constant registration for the entire source program is completed as described above, the machine language generation unit 3 sets the constant registration processing unit 5
Is incorporated in the area of the target program 4.

[Problems to be solved by the invention]

According to the above-described processing method, the registered data is used only when there is a part that completely matches the registered constant in the constant data already registered in the constant storage area 6, and the registered data is completely used even if it partially matches. However, if the registered data is used as much as possible even in the case of a partial match, the required amount of the constant storage area can be further reduced.

SUMMARY OF THE INVENTION An object of the present invention is to provide a constant registration processing method capable of reducing a constant storage area on a target program for holding constant data described in a source program.

[Means for solving the problem]

 FIG. 1 is a flowchart of a process showing the configuration of the present invention.

The figure shows the flow of the constant registration process. In processing steps 40 to 47, a case where a part of the registered constant matches the end of the registered data is detected. Register additionally.

(Operation)

When a source program is translated to generate a target program to be executed on a computer, data representing constants described in the source program is registered in a constant storage area provided in the target program by the constant registration processing unit 5 I do.

In this case, a search is made for a match between the leading continuous part of the constant data to be newly registered and the trailing continuous part of the data registered in the constant storage area by the processing of processing steps 40 to 47.

As a result, when there is a match, if there is a remaining portion excluding the leading continuous portion of the data to be registered, in processing step 51, only the remaining portion is stored as the trailing continuous portion in the constant storage area. Register in the following area.

According to the above processing method, the registered data can be used for a part of the data to be registered and the registered data in which there is a partial match, and the constant storage area can be reduced.

〔Example〕

The configuration of the computer is the same as that shown in FIG.
The constant registration processing section 5 executes the processing according to the present invention as shown in FIG.

Upon receipt of the request for constant registration, the constant registration processing section 5 takes out the registered data one digit at a time from the constant storage area 6 and sequentially retrieves the data as in the prior art. Then, it is determined whether or not unsearched data remains in the constant storage area 6. If not, additional registration is immediately performed in the processing step 49, and the head address of the area registered in the processing step 50 is returned to the machine language generation unit 3.

If there is data in the constant storage area 6, the processing step
At 41, the first address of the remaining data is stored in the working register, and at step 42, one byte of the address is read out. At step 43, the first byte of the constant received and the constant storage area 6 are read. Is compared with the data read out from the storage area, and if they do not match, the flow returns to the processing step 40 to identify whether the end of the constant storage area 6 is completed, and in this way, data matching with the first byte of the registered constant is searched.

As a result, if there is data that matches, the processing proceeds to processing step 44, where the next digit is identified in the registered constant, and if not, a part that completely matches the registered constant is included in the registered data. In step 48, the address previously stored in the register is returned to the machine language generation unit 3.

If the next digit remains in the registered constant, it is determined whether or not the next data exists in the constant storage area 6 in a processing step 45, and if so, the next one byte is read out in a processing step 46. Is compared with the next byte of the registration constant.

When a match is obtained in the processing step 47, the process returns to the processing step 44, and the comparison between the next byte of the registered constant and the next byte of the data registered in the constant storage area 6 is advanced.

As a result, if the registration constant ends first, if there is a part that matches the registration constant completely in the registered data,
As described above, the process proceeds to the processing step 48 to end the processing.

If the end of the registered data is detected in processing step 45, the last part of the data registered in the constant storage area 6 matches the part from the beginning to the middle digit of the data to be registered. It is. Therefore, in this case, the process proceeds from the processing step 45 to the processing step 51, and only the remaining part of the data to be registered except for the matching part is additionally registered after the registered data in the constant storage area 6, and then the processing step At 48, as described above, the address stored in the processing step 41 (the address of the byte of the registered data that matches the first byte of the data to be registered) is returned to the machine language generation unit 3.

According to the above processing, for example, as shown in FIG. 3B, for the same example as the source program 2 in FIG. The constant "345" of the program statement 10 is registered at the beginning of the constant storage area 6, and the constant "789" of the program statement 11 is additionally registered at a position following "345" in the constant storage area 6, and the next program statement Since the constant “345” of 12 matches the registered data, no additional registration is performed.

The constant "457891" in the next program 13 has no part that matches perfectly in the registered data, but the end of the registered data matches the front part of the constant "45789". Only “1” is additionally registered. "345789" in the next program statement 14 is not additionally registered as in the conventional case.

Therefore, as apparent from the comparison between FIG. 3 (a) and the constant storage area 6 in FIG. 3 (b) which is the result of this processing method, the required storage area amount is reduced.

〔The invention's effect〕

As is apparent from the above description, according to the present invention, in a process of a compiler for translating a source program of a computer into a target program, a constant on the target program for holding constant data described in the source program. Since the storage area can be reduced, there is a remarkable industrial effect of improving the utilization efficiency of the computer.

[Brief description of the drawings]

 FIG. 1 is a flowchart of a process showing the configuration of the present invention, FIG. 2 is a block diagram of a configuration example of a computer, FIG. 3 is an explanatory diagram of constant registration, and FIG. 4 is a flowchart of a conventional process. In the figure, 1 is a compiler, 2 is a source program, 3 is a machine language generation unit, 4 is a target program, 5 is a constant registration processing unit, 6 is a constant storage area, 10 to 15 are program statements, 20 to 30, 40 to 51 shows a processing step.

Claims (1)

(57) [Claims] 1. A compiling device for translating a source program and generating a target program to be executed on a computer, comprising: a constant storage area for registering data representing constants described in the source program to be translated; Search means for searching for a match between the leading continuous part of the registration target data and the trailing continuous part of the data already registered in the constant storage area; as a result of the search by the search means, there is a match; and Registering means for registering, when the data to be newly registered has a remaining part excluding the leading continuous part, only the remaining part in an area following the trailing continuous part of the constant storage area; .