JPH11232122A - Method for converting extended identifier in source program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable a Japanese-language programming using Japanese-language identifiers. SOLUTION: A conversion table 10 where conversion identifiers corresponding to Japanese-language identifiers are registered is prepared. A register means 2 searches Japanese-language identifiers in an original source program 1 and retrieves them on the conversion table 10 to check whether they are already registered there or not. If they are not registered yet, corresponding conversion identifiers are automatically generated and are registered in the conversion table 10 together. A conversion means 3 refers to the conversion table 10 to substitute Japanese-language identifiers in the original source program 1 with corresponding conversion identifiers. Thus, the original source program 1 is converted to an object source program 4.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a language processing system which, when there are restrictions on the character types or the number of valid characters that can be used as identifiers (various names, item names, labels, etc.) in a source program, The present invention relates to a method for converting an extended identifier that exceeds the grammar rules. Especially,
The present invention relates to a method for implementing Japanese programming in a language processing system that does not permit Japanese programming. Here, the language processing system refers to a compiler, a linker, and the like. The extended identifier is used to mean an identifier when it is assumed that the grammatical rule (limitation on character type or the number of valid characters) for the identifier is appropriately expanded and relaxed. For example, considering a grammar rule that defines an identifier as "a lower-case character string of 8 characters or less", "a lower-case character string of 32 characters or less" or "16 characters or less of an alphabetic character or" @ "character The “character string” and the like are extended identifiers because they are extensions of the grammatical rules for the identifiers.

[0002]

2. Description of the Related Art In program development, Japanese translation of a source program is understood as improving program descriptiveness and readability. In particular, there is a great need for the development of application programs for Japan. It is important that existing computer languages be able to use Japanese identifiers with Japanese characters, even if they are not at the level of complete Japanese programming, including language grammar. Here, the Japanese identifier is an identifier obtained by adding Japanese characters to the character type of a normal identifier.
Japanese characters mean hiragana, katakana, kanji, and other characters and symbols, and are usually characters other than single-byte characters. ASCII code characters and E
There are BCDIC code characters, etc. Characters other than 1-byte characters are JIS code, shift JIS code, and EUC
There are multi-byte characters such as codes.

Whether Japanese programming is possible depends on the language processing system of each computer system. For this reason, in order to enable Japanese language programming in a language processing system in which Japanese language programming is not possible, a preprocessor method is usually used. That is, the result obtained by converting the Japanese notation part in the source program according to the target language processing system is passed to the language processing system. Particularly important in Japanese language programming by the preprocessor method is a part that converts a Japanese identifier into an identifier in accordance with a grammar rule of a target language processing system.

As a conventional publicly known technique, Japanese two-byte characters of a Japanese identifier are regarded as four-digit hexadecimal numbers and are represented by four one-byte characters.
How to convert binary notation into three single-byte characters (Hirabayashi:
IPSJ Programming Language Study Group Material 16-2
"Japanese translation of C language and its effects", 1988).
Generally speaking, the number of valid characters of an identifier is limited by the grammar rules, so it is not preferable that the number of characters of the converted identifier be significantly increased. For this reason, a method in JP-A-225690 has been proposed as an improved method of the above method.

This method focuses on the frequency of use of Japanese double-byte characters, and uses a hexadecimal notation of 1-byte characters (10 numeric characters, 26 uppercase letters, 26 lowercase letters). The most frequently used Japanese 2
This is a method of converting byte characters into two single-byte characters, and converting two-byte Japanese characters that are not frequently used into three single-byte characters. According to this method, most Japanese identifiers can be converted to identifiers of the same length.

[0006]

However, there are various problems with these conventional techniques and methods for converting Japanese identifiers into character units.

One is that an implicit condition is that all identifiers are composed only of Japanese characters. This is clear if identifiers containing both Japanese characters and single-byte characters are assumed. For example, in the method disclosed in the above publication, the Japanese identifier “Kanji” is converted into an identifier “JaV7” using single-byte characters. It interferes with the identifier "Ja character" in which characters are mixed. Therefore, an identifier in which one-byte characters and Japanese characters are mixed cannot be used.

[0008] The second is a general request from the grammar rules, "The first character of an identifier starts with an alphabetic character." For example, in the method disclosed in the above publication, an identifier starting with the Japanese character "A" is converted into an identifier whose first two characters start with "4Z", but this cannot be an identifier according to the grammatical rules.
Therefore, a Japanese identifier in which the first character of the converted identifier is a numeral cannot be used.

Third, in a language processing system such as FORTRAN that does not distinguish between uppercase and lowercase letters, the identifier 62
Hexadecimal notation cannot be used. Therefore, the method disclosed in the above publication can be implemented only in a language processing system capable of expressing a hexadecimal identifier.

Fourth, the number of characters of the converted identifier increases. Even in the method disclosed in the above publication, the number of characters after conversion is increased in an identifier using Japanese characters that are not frequently used. As described above, the conventional method still has problems, and can be implemented only under various conditions.

The limitation of the number of valid characters and the limitation of the use of special characters (for example, "@" and "@") are also problems related to identifiers, and are part of the problem to be solved by the present invention.

Accordingly, an object of the present invention is to use a general-purpose method that does not depend on the characteristics of a specific language processing system.
An object of the present invention is to make it possible to use an extended identifier in a general computer system without changing an existing language processing system. This includes, inevitably, enabling Japanese programming. Also provided is a method for maintaining program compatibility between systems that can and cannot use extended identifiers.

[0013]

The role of an identifier in a source program is to give a specific name to a specific variable and give a specific meaning from the viewpoint of a programmer. However, from the viewpoint of the language processing system, the role of the identifier is to distinguish a specific variable from other variables. In this sense, if the identifier is a Japanese expression,
It is not essential whether it is an English expression or just a random set of characters. The same applies to the length of the valid characters of the identifier. Therefore, a specific identifier in a program can be replaced with another identifier on the condition that it does not interfere with another identifier. This process
It is not necessary to perform it for each character constituting the identifier as in the related art, but for each identifier.

Based on this principle, the present invention is shown in FIGS.
I will explain using.

The present invention provides a conversion table 10, a registration unit 2,
And conversion means 3. The conversion table 10 is a table for converting the name of an extended identifier, and as shown in FIG. 2, is composed of an extended identifier 11 and a corresponding conversion identifier 12. The registration means 2 is a means for registering the extension identifier in the original source program 1 in the conversion table 10. The conversion means 3 is a means for converting the original source program 1 into the target source program 4 using the conversion table 10.

The registration means 2 scans the original source program 1 and searches for an extended identifier. The conversion table 10 is searched for the extension identifier to determine the registration. That is,
If this extension identifier is not registered in the conversion table 10, it should be registered, or if it has already been registered,
No registration is required. If it is determined that registration is to be performed, a corresponding conversion identifier is automatically generated and registered in the conversion table together with the extension identifier. This process is performed for all extended identifiers in the original source program 1. As a result, a conversion table 10 is obtained. Further, in the processing of the main registration means 2, the processing of the original source program 1 may be performed for each unit (for example, one line unit). As a result, the conversion table 10 for all the extended identifiers in the original source program 1 If possible.

The name of the automatically generated conversion identifier can be arbitrarily determined in accordance with the grammatical rules of the target language processing system 5 and under the condition that it does not interfere with other identifiers or reserved words (also called keywords).

According to the present invention, a prototype of a name as a rule for generating a name is determined, and the conversion identifier is automatically generated in accordance with the prototype. The prototype of the name has a fixed name part and a variable name part. The fixed name part is made to correspond to an arbitrary fixed character. Corresponds to consecutive n-ary numbers. When n is 10, the name variable part is a serial number in decimal (for example, 000, 001, 002, 003,...). here,
The n-ary digits for notation are not necessarily numbers (0, 1, 2, ...
・ ・) It is not necessary to use alphabets (a, b, c, ...)
May be. For example, if the n-ary digits are "a", "b",
If "c", ..., the name variable part is aaa, aa
b, aac,... Unless they overlap, any order of any character is possible. For example, if the n-ary numbers are “g”, “6”, “w”,..., The name variable parts are ggg, gg6, ggw,. further,
If a conversion identifier whose first character is a non-numeric character can be generated only by the name variable part, the name fixing part is not necessarily required.

Further, as long as uniqueness as an identifier is guaranteed, the name variable portion does not necessarily have to be a continuous number, but may be a sequence of numbers generated continuously. For example, it may be a long-period pseudo-random number or an n-ary notation thereof.

In FIG. 2, as an example of the conversion identifier, a name fixed part a and a name variable part nnn (nnn is 001 to 99)
(Decimal number of 9) was used. If there is a concern that a conversion identifier having a specific prototype may interfere with other identifiers, all identifiers in the source program need to be processed, not just extension identifiers.

The conversion means 3 scans the original source program 1 and searches for an extended identifier. The conversion table 10 is searched for the extension identifier. Since the extension identifier is registered by the registration means 2, it is always found in the registration table 10. Replace this extended identifier with the corresponding conversion identifier. The result of performing this processing for all extended identifiers of the original source program 1 is output as the target source program 4. In the processing of the conversion means 3,
Similar to the registration means 2, the processing of the original source program 1 may be performed for each unit (for example, one line unit).
Anything that can be converted to

The registration means and the conversion means may be performed separately (two-pass method), or the two means may be performed simultaneously in parallel (one-pass method).

The target language processing system 5 processes the target source program 4 to output an execution program 6 and a message 7 for the target source program 4.
The reverse conversion means 8 uses the conversion table 10 to replace the identifier in the message 7 with the extension identifier. As a result, the message 7 can be read as the message 9 for the original source program 1.

By pre-registering a conversion identifier corresponding to a specific extension identifier in the conversion table 10, it is possible to specify and convert a specific extension identifier in the original source program 1 to a specific identifier. At this time, an extension identifier that has not been registered in advance is assigned an automatically generated conversion identifier.

By intentionally performing this systematically for all identifiers in the original source program 1, a source program in a language processing system capable of processing extended identifiers can be easily ported to a language processing system incapable of processing. it can. The reverse is also true.

As shown in FIG. 1, the processing of the target language processing system 5 enters between the registration means 2 and the inverse conversion means 8. For this reason, when the present invention is implemented by the single process method, the conversion table 10 cannot be held in the memory until the inverse conversion unit 8 is executed. Therefore, in such a case, it is necessary to temporarily store the conversion table 10 in an external storage device such as a disk and read it into the memory when the inverse conversion means 8 is executed.

[0027]

FIG. 3 shows an example of an original source program 1 including an extended identifier written in a C language style. Using this as a sample program, the prototype of the conversion identifier 12 automatically generated is a
An embodiment in the case of a two-pass system in which nnn (nnn is a decimal number from 001 to 999), the registration unit 2 and the conversion unit 3 will be described according to an actual example.

First, preparations for processing are made. That is, the conversion table 10 is cleared, and the number of registered extension identifiers is set to zero. The original source program 1 shown in FIG. 3 is input and read into the memory. The processing for the original source program 1 in the registration means 2 and the conversion means 3 is performed on the original source program 1 in the memory. It is also assumed that the conversion table 10 is also on the memory.

The registration means 2 stores the original source program 1 in C
Scan according to language grammar rules for extended identifiers.
In the actual example, the first character string “float” is a reserved word and is passed. Since the next character string “area 1” is an extended identifier, the conversion table 10 is searched for registration judgment. Since “area 1” has not been registered at this time, “a” is used as the corresponding conversion identifier.
001 ”is automatically generated and registered together in the conversion table 10. The subsequent “width 1”, “height 1”, “area 2”, “width 2”, and “height 2” similarly search the conversion table 10 and determine registration. Since these are not registered at this time, "a006" is automatically generated from "a002" as the corresponding conversion identifier, and both are registered in the conversion table 10. For the subsequent “area 1”, the conversion table 10 is searched for registration judgment. "Area 1" has already been registered at this point and will be passed.
The same processing is performed for the other extension identifiers that follow. When all extension identifiers of the original source program 1 have been processed, the conversion table 10 shown in FIG. 2 is obtained. The conversion table 10 is stored in an external storage device for the later inverse conversion means 8.

The conversion means 3 scans the original source program 1 again from the beginning and searches for an extended identifier. The character string "float" is a reserved word and passes through. For the next extended identifier “area 1”, the conversion table 10 is searched for the corresponding conversion identifier “a001”, and the “area 1” in the original source program 1 is replaced with the “a001”. The same processing is performed on the other extension identifiers that follow, and replaced with the corresponding conversion identifiers. As a result, the target source program 4 of FIG. 4 is obtained. The target source program 4 is output and passed to the target language processing system 5.

The target source program 4 is processed by the target language processing system 5. The target language processing system 5 outputs an execution program 6 and a message 7. As an example, assume that the message 7 is “undefined a008”.

The conversion table 10 created and stored by the registration means 2 is read into a memory. The inverse conversion means 8 uses the conversion table 10 to replace the identifier “a008” in the message 7 with the extended identifier “total area”.
Thereby, the message 9 “undefined total area” is obtained.

FIGS. 5 and 6 show examples of source program conversion by designating a specific extension identifier as a specific conversion identifier. As in the example of FIG. 5, the specific extension identifier and the specific conversion identifier are registered in the conversion table 10 in advance. By executing the registration unit 2 and the conversion unit 3 on the source program 1 of the above example, the target source program 4 of FIG. 6 is obtained. The target source program 4 thus obtained
Is portable to language processing systems that do not support extended identifiers.

[0034]

According to the present invention, by converting an extended identifier in a source program into an identifier unit, the extended identifier can be used in a wide range of systems without being limited to a specific language. Japanese programming can also be realized. As a result, the description and readability of the source program are improved, and therefore, the quality of the program is improved,
Improves maintainability and productivity. Further, since the present invention uses the preprocessor system, it can be implemented without changing the language processing system of the conventional system. Furthermore, since it is easy to maintain compatibility between a system that can use the extended identifier and a system that cannot use the extended identifier, there is no burden on the introduction.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram showing an embodiment according to a two-pass system of the present invention.

FIG. 2 is a diagram showing an example of a conversion table.

FIG. 3 is a diagram showing an example of an original source program written in a C language style.

FIG. 4 is a diagram showing an example of a target source program obtained as a result of a conversion process.

FIG. 5 is a diagram showing an example of a conversion table when pre-registered.

FIG. 6 is a diagram showing an example of a target source program obtained as a result of the conversion processing.

[Explanation of symbols]

 Reference Signs List 1 original source program 2 registration means 3 conversion means 4 target source program 5 target language processing system 6 execution program 7 message for target source program 8 inverse conversion means 9 message for original source program 10 conversion table 11 extension identifier 12 conversion identifier

Claims (5)

[Claims] A conversion table comprising an extension identifier to be converted in an original source program and a conversion identifier corresponding to the extension identifier; searching for an extension identifier in the original source program; If you register,
A registration means for automatically generating a corresponding conversion identifier and registering the conversion identifier together with the extension identifier in the conversion table, searching for an extension identifier in the original source program, searching for the extension identifier in the conversion table, A source program conversion method, comprising: conversion means for converting the original source program into a target source program by creating a replacement identifier for the extension identifier in the original source program with a corresponding conversion identifier. 2. A method for automatically generating a corresponding conversion identifier, comprising the steps of:
2. The source program conversion method according to claim 1, wherein the source program is created from a continuous number in n-ary notation where n is an arbitrary number. 3. In the automatic generation of the corresponding conversion identifier, a name variable part of a name of the conversion identifier to be automatically generated,
2. The source program conversion method according to claim 1, wherein the program is created from a sequence of numbers continuously generated in n-base notation where n is an arbitrary number. 4. The source program conversion method according to claim 1, wherein a conversion identifier corresponding to a specific extension identifier is registered in the conversion table in advance. 5. A message output as a result of processing a target source program converted by the source program conversion method according to claim 1, by a target language processing system, by using an identifier in the message as the conversion table. Is considered as a conversion identifier of the message, the identifier in the message is searched on the conversion table, and a message obtained by replacing the identifier in the message with an extension identifier corresponding to the identifier is created. Characteristic message conversion method.