JP2868127B2 - Blank reading device in lexical analysis - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Overview] Regarding a blank reading device in lexical analysis, a comparison operation function for comparing a plurality of bytes of the system at a time is used, and a plurality of bytes (for example, 4 bytes of integer type data) are obtained from a source program by a source program. The purpose of the system is to compare and calculate multiple bytes at a time with the aim of reducing the number of comparison operations for skipping white space and performing high-speed lexical analysis processing. Using a comparison operation function, an exclusive OR operation is performed on a character string code for a plurality of bytes extracted from the source program and a blank code for a plurality of bytes, and the result is 0 from the beginning.
The system is configured so that it is determined that the number of bytes of (zero) is blank and skips reading, and lexical extraction and analysis are performed.

[Industrial applications]

The present invention relates to a blank reading device in lexical analysis.

[Problems to be Solved by the Prior Art and the Invention] White space in a source program described in FORTRAN language has no meaning except in comments and character constants. For example, if there is an alphabet name ABC, both REAL ABC... (1) REAL ABC... (2) have the same meaning. For this reason, in lexical analysis in the FORTRAN processing system, the processing of skipping blank characters has a very important meaning.

Conventionally, at the time of lexical analysis, the process of skipping blank characters
Scanning was performed while checking for blanks for each character (every byte). For this reason, there is a problem that the processing time of the lexical analysis becomes longer as the number of blanks increases in the FORTRAN source program.

The present invention uses a comparison operation function for comparing a plurality of bytes of a system at a time, and determines and skips blanks in a plurality of bytes (for example, 4 bytes of integer type data) from a source program, and skips blanks. An object of the present invention is to execute the lexical analysis processing at high speed by reducing the number of comparison operations for the processing.

[Means to solve the problem]

 Means for solving the problem will be described with reference to FIG.

In FIG. 1, a source program 1 is a source program to be skipped and subjected to lexical analysis.

The comparison operation function 2 is a comparison operation function that performs a comparison operation on a plurality of bytes of the system at a time.

[Action]

As shown in FIG. 1, the present invention uses a comparison operation function 2 for comparing and operating a plurality of bytes of the system at a time. An exclusive-OR operation with a blank code is performed, and the result is determined to be blank by the number of 0 (zero) bytes from the beginning, skipped, and lexical extraction and analysis are performed.

Therefore, by using the comparison operation function 2 of the system and determining and skipping blanks from the source program 1 in units of a plurality of bytes (for example, in units of 4 bytes of integer-type data), the comparison for blank blasphemy removal processing is performed. The lexical analysis process can be executed at high speed by reducing the number of operations.

〔Example〕

Next, the configuration and operation of one embodiment of the present invention will be sequentially described in detail with reference to FIG. 1 and FIG.

In FIG. 1, a source program 1 is a source program described in a FORTRAN language to be skipped and subjected to lexical analysis.

The comparison operation function 2 is a comparison operation function for comparing a plurality of bytes that the system waits for at a time. For example, four bytes (32 bits) of integer-type data are bit-at-a-time corresponding to a comparison operation (logical sum, logical product, Exclusive OR). In the present embodiment, a blank in a character string of a source program is regarded as this integer type data and a plurality of bytes (for example, 4 bytes) are used.
In this method, blanks are skipped in a smaller number of steps than in units of (bytes). It should be noted that a character is normally subjected to a comparison operation for each byte (for each character) due to its nature. The operation in the case where the comparison operation function 2 for comparing four bytes of the integer type data at a time is used will be specifically described in the order shown in FIGS. 1 and 2 with reference to FIG.

In FIG. 1, a 4-byte character string is extracted from the source program 1, and is regarded as integer-type data and exclusive-ORed with a 4-byte blank character string (comparison operation function 2). As a result, a 4-byte character string extracted from the source program 1 and a 4-byte blank character string are exclusively ORed at a time by the comparison operation function 2 of the system.

Determines whether the result is 0 (zero). In the case of YES, since four characters are blank, the processing position of the lexical analysis is advanced by four bytes. This is, for example, FIG.
The exclusive OR operation of the 4-byte character string “40404040” extracted from the current lexical analysis processing position of the source program 1 and the blank code “40404040” results in a 4-byte result of all 0 (zero) ), The result is YES, and the lexical analysis processing position is advanced by 4 bytes to skip four blanks. If no, do.

Determines whether the first three bytes of the result are 0 (zero). In case of YES, 3 characters are blank
The processing position of the lexical analysis is advanced by 3 bytes. This is, for example, a 4-byte character string “404040E7” extracted from the current lexical analysis processing position of the source program 1 in FIG.
The result of performing an exclusive OR operation with the blank code "40404040" is 0 (zero) for 3 bytes from the beginning, so the result is YES, and the lexical analysis processing position is advanced by 3 bytes. Skip two blanks. If no, do.

Determines whether the first two bytes of the result are 0 (zero). In case of YES, two characters are blank.
The processing position of the lexical analysis is advanced by 2 bytes. This is, for example, the character string “4040F14B” of 4 bytes extracted from the current lexical analysis processing position of the source program 1 in FIG.
The result of performing an exclusive OR operation with the blank code “40404040” is 0 (zero) for 2 bytes from the beginning, so the result is YES, and the lexical analysis processing position is advanced by 2 bytes. Skip two blanks. If no, do.

Determines whether the first byte of the result is 0 (zero). In case of YES, one character is blank.
The processing position of the lexical analysis is advanced by one byte. This corresponds to, for example, a 4-byte character string “404E4040” extracted from the current lexical analysis processing position of the source program 1 in FIG.
The result of performing an exclusive OR operation with the blank code "40404040" is 0 (zero) for 1 byte from the beginning, so the result is YES, and the lexical analysis processing position is advanced by 1 byte. Skip two blanks. If no, do.

Performs the lexical analysis extraction processing because the first byte of the result is not 0 (zero) and not a blank. This is, for example, the character string “C1” of 4 bytes extracted from the current lexical analysis processing position of the source program 1 in FIG.
Since the leading byte of the result of performing an exclusive OR operation between C2C340 "and the blank code" 40404040 "is not 0 (zero), the result is NO, and the lexical extraction processing is performed.

As described above, a character string is extracted from the source program 1 in 4-byte units, exclusive-ORed with a 4-byte blank code, and skipped by the number of 0 (zero) bytes from the beginning of the result. This makes it possible to skip blanks with a small number of steps.

 FIG. 2 is a diagram illustrating a specific example of the present invention.

FIG. 2 (a) shows a specific example when YES in FIG. This is a case where all four bytes of the character string extracted from the source program 1 are blank codes, and is an example of skipping four bytes.

FIG. 2 (b) shows a specific example when YES in FIG. This is the case where the first three bytes of the four bytes of the character string extracted from the source program 1 are blank codes, and is an example of skipping three bytes.

FIG. 2 (c) shows a specific example when YES in FIG. This is a case where the first two bytes of the four bytes of the character string extracted from the source program 1 are blank codes, and is an example of skipping two bytes.

FIG. 2 (d) shows a specific example when YES in FIG. This is when one byte from the beginning of the four bytes of the character string extracted from the source program 1 is a blank code, and is an example of skipping one byte.

FIG. 2 (e) shows a specific example when NO in FIG. This is an example in which the first byte of the 4 bytes of the character string extracted from the source program 1 is not a blank code, and the lexical analysis is performed by extracting the lexical from the first byte.

〔The invention's effect〕

As described above, according to the present invention, the system uses the comparison operation function 2 to compare and calculate a plurality of bytes at a time, and blanks the source program 1 in units of a plurality of bytes (for example, in units of 4 bytes of integer type data). Is adopted, and the number of comparison operations for the blank skipping process can be reduced, and the lexical analysis process can be executed at high speed.

[Brief description of the drawings]

FIG. 1 is a block diagram of one embodiment of the present invention, and FIG. 2 is a diagram illustrating a specific example of the present invention. In the figure, 1 indicates a source program, and 2 indicates a comparison operation function of the system.

Claims (1)

(57) [Claims] An apparatus for skipping blanks in lexical analysis, comprising: a comparison operation means for comparing and calculating a plurality of bytes at a time by an instruction operating on a computer system; a code of a character string for the plurality of bytes extracted from a source program; The exclusive OR operation with the blank code for the plurality of bytes is performed by issuing an instruction to the comparison operation means, and the result is determined to be 0 (zero) bytes from the beginning and is skipped, Means for extracting a lexical character corresponding to a byte other than 0 (zero) and analyzing the extracted lexical character.