JPS61204741A - Compressing method for source program - Google Patents

Abstract

PURPOSE:To enlarge the compressibility by encoding not only an operator, a reserved word and a numeric constant but also a variable and a character- string constant. CONSTITUTION:An encoding processing means 2 obtains successively words and phrases from a source program 1 and classifies them, generates the corresponding word and phrase number, etc. by referring to a reserved symbol table 3 having a correspondence table of the words and phrases, and the word and phrase number, and outputs them as an intermediate code 5. That is to say, when encoding processing means 2 inputs the source program, since the head numeral 100 of a line is a line number, the corresponding word and phrase number (=1) and the line number (=100) are outputted. The next SUM is not any of a line number, an operator, a reserved word, a numeric constant, and a character-string constant, therefore, it is registered in a variable symbol table, and the word and phrase number (=10) and a registered position number (=1) are outputted. The next operator is an operator, therefore, the corresponding word and phrase number (=7) is outputted, and thereafter, by the same operation, an intermediate code and a variable symbol table are generated and outputted.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention is applied to a source program for converting a source program into intermediate code in an inverter such as a BASIC initiator that is widely used in personal converters. Concerning a compression method.

[Conventional technology]

For example, among interpreters such as conventional RASIC interpreters, there are those that directly interpret and execute the source program with almost no conversion, and those that compress data by converting it into intermediate code and then execute the interpretation. However, in the conventional intermediate coding method, only some operators, reserved words, and numeric constants are converted into hoods, and other string constants and variables are transferred to the intermediate code with the character code of the external representation. A method is being used.

[Problem that the invention seeks to solve]

In such conventional methods, not only methods that hardly transform the source program, but also methods that compress the source program by converting it into intermediate code, even if some operators, reserved words, and numerical constants are encoded, Since the other data is externally expressed character code IT data, the compression rate is small compared to the size of the source program data.

Furthermore, even if the program is converted into an intermediate code, part of it remains in the character code of the external representation, so when executing such a program, it is necessary to interpret and execute it while performing lexical analysis.

For example, if you put it in a variable and the length of the variable name is allowed to be up to 8 characters, the character code will be read a maximum of 8 times, lexical analysis will be performed, and the execution will be executed. It becomes a constraint.

For example, in the case of a process that branches by a label, such as "goto *label", the label name was transcribed as it is in the external representation character code, so this is also done after reading the character code. Interpretation will be carried out,
Furthermore, if the jump destination of "*label" is after the line being executed, the intermediate code must be skipped character by character until "*label" appears, which inevitably reduces execution speed.

In view of these conventional drawbacks, the present invention provides a source program compression method that increases the compression rate of intermediate code and, in particular, makes it possible to improve processing speed.

[Means for solving problems]

Means for solving the above problem includes an encoding processing means and a storage means having a registration area for reserved symbols and variable symbol tables registered in advance, and the encoding processing means constitutes a source program. Classify the tokens into operator reserved word group, numeric constant group, string constant group, and variable group, refer to each of the reserved symbols, give the corresponding token number to the token belonging to the operator reserved word group, and assign numerical value. A token belonging to a constant group is given a corresponding token number and a coded numerical value, and a token belonging to a string constant group is given a corresponding token number, a numerical value indicating the length of the string, and the token of the string. , is a source program compression method characterized by registering a token belonging to a variable group in a variable symbol table, giving a corresponding token number and a registration position number on the symbol table, and outputting the token.

〔Example〕

The present invention will be explained in detail based on examples. Fig. 1 is a block diagram showing the concept of the present invention, Fig. 2 is a flowchart showing an example of implementing the encoding processing means of Fig. 1 as a program, Fig. 3 is an example of a source program, and Fig. 4 is An example in which the source program is converted into intermediate code according to the embodiment described above is shown. In FIG. 1, the encoding processing means 2 obtains tokens one after another from the source program 1, classifies them, and generates corresponding token numbers etc. by referring to a reserved symbol 3 having a correspondence table of tokens and token numbers. This is then output as intermediate code 5.

That is, the encoding processing means 2 operates as follows in FIG. First, it obtains tokens one after another and 1) If it is a line number, outputs the token number and line number.

2) If the operators are reserved words, output the lexical number corresponding to each.

3) If it is a numerical constant, output the lexical number and the numerical value converted to internal representation.

4) If it is a string constant, output the lexical number, character length, and the string.

5) If it is none of the above 1) to 4), register it as a variable in the variable symbol table and output the lexical number and registration position number.

Since it operates as described above, if you input the source program shown in Figure 3 into this example, the number 100 at the beginning of the line is the line number, so the corresponding lexical number (=1) and line number (=1
00) is output. The next SUM contains the line number, operator,
It is not a reserved word, numeric constant, or string constant, so
It is registered in the variable symbol table, and the lexical number (=10) and registration position number (=1) are output. Since the next = is an operator, the corresponding lexical number (=7) is output, and the same operation is performed to create and output an intermediate code and variable symbol table as shown in FIG. The variable is treated as a variable and registered in the variable symbol table, and is given the lexical number and registration position number corresponding to the variable.

〔Effect of the invention〕

As described above, according to the source program compression method of the present invention, not only operators, reserved words, and numerical constants, but also variables and character string constants are encoded, so that the compression ratio is high.

Furthermore, lexical numbers are given to all lexical items such as variables (including labels), and the meaning of the intermediate code can be immediately understood by the lexical numbers. In particular, for variables and labels, there is no need to read each character one by one, perform lexical analysis, and then execute it, as in the past, and the contents can be immediately known from the lexical number and variable symbol table, increasing execution speed. There are many advantages such as.

[Brief explanation of drawings]

Fig. 1 is a program diagram showing the concept of the present invention, Fig. 2 is a flowchart showing an example of the encoding processing means, Fig. 3 is an example of a source program, and Fig. 4 is an intermediate diagram of the source program of Fig. 3. FIG. 2 is an explanatory diagram of a coded example.

Claims (1)

[Claims] It has an encoding processing means and a storage means having a registration area for a reserved symbol table and a variable symbol table registered in advance. , and classify them into a string constant group and a variable group, refer to the reserved symbol table, and assign corresponding lexical numbers to tokens belonging to the operator reserved word group, and give corresponding lexical numbers to tokens belonging to the numeric constant group. For the tokens belonging to the string constant group, give the corresponding token number, the numerical value indicating the length of the string, and the token of the string, and for the tokens belonging to the variable group, give the token. A method for compressing a source program, characterized in that a variable is registered in a symbol table, and a corresponding lexical number and a registered position number on the symbol table are given and output.