JPH021027A - Compiling system for compressed source program - Google Patents

Abstract

PURPOSE:To shorten the compiling time by directly taking a compressed source program as the object of word and phrase analysis without decoding (expanding) this program and increasing the skip read speed of blank characters. CONSTITUTION:The compressed source program includes an 8-bit control code, and upper four bits indicate the number of characters other than blacks, and lower four bits indicate the number of following blank characters. On the compiler side, a blank is handled as a delimiter character, and one blank as well as continuous blanks have the same function. Consequently, a blank skip read processing part is provided before a word and phrase analyzing part of the compiler to skip over excess blanks at the time of reading, and at this time, a one-character input part is executed for each one character. Thus, 10 continuous blanks are substituted with the one-byte control code in the system where the compressed source program is directly taken as the object of word and phrase analysis (one-character input), and it is sufficient if the one-character input part 5 is executed once, and the compiling time is shortened.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for compiling compressed source programs.

[Conventional technology]

Regarding improving the efficiency of lexical analysis by skipping blank characters, there is Japanese Patent Application Laid-Open No. 58-58655. This method checks whether all spaces after the delimiter in the source are blank (an error occurs if there are no blanks). Therefore-
It cannot be applied to modern high-level languages where multiple characters can be written on a line.

Another method for shortening compilation time is JP-A-58-
There is No. 168153. This is a method in which the common source is registered as intermediate text, and is effective only for the common source.

[Problem to be solved by the invention]

The problem with the above-mentioned prior art is that it cannot be applied to recent high-level languages that can be written in a free format (multiple sentences can be written on a - line).

Another problem is that source analysis of parts other than the common source is ineffective.

An object of the present invention is to speed up lexical analysis and shorten compile time, regardless of the language type or whether or not the source program is a common source, by using a compressed format of the source program.

[Means to solve the problem]

The above purpose is to decode a compressed source program (1-(
This can be achieved by making it a target of lexical analysis directly without decompression (expansion) and speeding up the reading of blank characters.

[Effect]

In a compressed source program, there are 1 or 2 consecutive spaces.
Compressed to about bytes. (Differs depending on the compression method) By directly using this as an input for lexical analysis, reading the blank space of n consecutive characters can be done by reading the compressed 1-2 byte data. This speeds up only the reading of blank spaces, improving compilation efficiency.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings.

First, the overall configuration will be explained using FIG. 1.

The compiler 1 is broadly divided into a syntax analysis section 3 and a code generation section 10, which inputs a source program from a source file 2, outputs an object program to an object file 11, and outputs a compilation list to a list output device 9. At this time, the source program in the source file 2 is in a compressed format to improve space efficiency.

The lexical analyzer 4 is called as a subroutine by the syntax analyzer 3 and extracts a set of characters (names) that are meaningful to the compiler from the source program.

(operation symbol, 2 delimiter symbol, etc.).

The one-character human input section 5 is a part of the lexical analysis section 4, and inputs one character of the source program from the input buffer 7. Also, when the processing for the source program code is completed, the source input unit 6 is called and the next record is input to the source file 2.
The data is input into the manual buffer 7 in compressed format.

The syntax analysis section 3 also calls the source decompression section 8, decompresses (decodes) the compressed source program in the input buffer 7, and outputs it to the list output device 9 as a compilation list.

Next, the operations of the lexical analysis section 4 and the -character input section 5, which are the central parts of the present invention, will be explained.

First, the format of the compressed source program 21 input by the single character input section 5 is shown in FIG. This format has a 1-byte (8-bit) control code in the source program,
The upper 4 bits of the control code indicate the number of characters other than spaces (blanks), and the lower 4 bits indicate the number of spaces following the control code.

To the compiler, a space (blank) is a delimiter, but consecutive spaces or single-character spaces have the same role as delimiters.

For this reason, the lexical analysis unit 4 of the compiler has a space skipping processing unit 41 at the beginning of the lexical analysis process, as shown in FIG. executed every time.

In the conventional method, the compression format is returned to its original state when the source is input, so that the input buffer 7 is manually inputted with an uncompressed source program.

Therefore, if there is a space of 10 consecutive characters in the source program, the 1-character input section 5 will be executed 10 times.

However, in the method of the present invention in which the compressed source program is subject to direct lexical analysis (one character input), ten consecutive spaces are also replaced with one byte control code, so the one character human power section 5 is All you have to do is rush forward.

FIG. 4 shows an outline of the processing of the single character input unit 5 for the compressed source program. The single character input section 5 has a valid character counter 51. Current pointer 52° input character buffer 5
Has 3. The effective character counter 51 has as an initial value the number of characters other than spaces, which are the upper 4 bits of the control code in the manual buffer 7. A current pointer 52 indicates the position in the input buffer that is currently being input. Input character buffer 5
3 is an area where one input character is set.

The single character input section 5 first checks whether the effective character counter 51 is positive, and if it is positive, one character is input from the position indicated by the current pointer 52 of the input buffer and set in the input character buffer 53. Next, the valid character counter 51 is decremented by 1.

If the result of checking the valid character counter 51 is not positive, a space (blank) is set in the input character buffer 53,
Input the control code from the position indicated by the current pointer 52 of the human buffer buffer, and input the upper 4 bits to the valid character counter 5.
Set to 1.

Finally, add 1 to the current pointer to update the current position.

According to this embodiment, by directly subjecting a compressed source program to lexical analysis, when performing lexical analysis after decoding (decompression), space (blank) reading is faster and compilation efficiency is improved. It has the effect of

〔Effect of the invention〕

According to the present invention, it is possible to speed up reading of spaces (blanks) in a source program at compile time, resulting in an effect of shortening compile time.

[Brief explanation of the drawing]

Fig. 1 is a block diagram of an embodiment of the present invention, Fig. 2 is an explanatory diagram of the compression format of the source program, Fig. 3 is an explanatory diagram of the lexical analysis section, and Fig. 4 is an explanatory diagram of the single character input section. be. 1... Compiler, 2... Source file, 3...
- Syntax analysis unit, 4... Lexical analysis unit, 6... Source input unit, 8... Source decompression unit, 10... Code generation unit.

Claims (1)

[Claims] 1. A method for compiling a compressed source program, characterized in that the compiler directly subjects the compressed source program to lexical analysis without decompressing it.