JPH0424746A - Program translation processor - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Summary] A program translation processing device for translating a computer source program into a target program, which generates a tree structure representing the structure of a program statement and makes high-level translation processing relatively easy. The purpose is to provide a program translation processing device that can be configured into The source program that follows the syntactic rules of the grammar is divided into sections using predetermined delimiters and decomposed into a string of clauses and delimiters, and the semantic analysis section analyzes the string of clauses and delimiters that follow the semantic rules of the grammar using a predetermined method. A tree structure determined based on the inclusion relationship of parentheses is generated, and the tree structure has a structure that distinguishes between an expression representing an operation of the source program and an expression included in information specifying an operand of the operation. , the detailed analysis unit is configured to process the expression according to the division based on the tree structure and generate a predetermined intermediate language expression for each operand that directly represents the relative position of the operand.

[Industrial application field]

The present invention relates to a program translation processing device that translates a computer source program into a target program.

[Conventional technology]

A so-called compiler that translates a computer source program written in a programming language such as FORTRAN into a target program to be executed on a desired computer is well known. FIG. 4 shows an example of the configuration of a program translation processing device 1 that executes such a compiler function.

For example, when the program translation processing device 1 inputs a source program 2 that has been subjected to the FORTRAN language TR method, the syntax analysis unit 3 performs syntax analysis according to the FORTRAN language TR method, checks for syntactic errors, and Generate string 4 of clauses and delimiters for the program.

Column 4 of phrases and delimiters uses predetermined delimiters such as = (,), tens, and other symbols as delimiters, and separates and connects them into sets of phrases and delimiters consisting of other characters, numbers, and other symbols. For example, an assignment statement such as C=A (B (I+1) +1)+A, (J+1) is written as “C=→”Aji→“Bji→“Work+゛→”1)“′ →°′φ1′′→°“l)”→“φ＋”→
“Aji→“J+”→“1)”.

become that way. Here, each part surrounded by "" and "" is a set of results, and "→" conceptually indicates that they are linked.

Note that "φ" is a symbol placed for control when delimiters are consecutive.

Next, the semantic analysis unit 5 performs semantic analysis on the clause and delimiter string 4 according to the FORTRAN language TR method to check for errors, and if correct, analyzes the subscripts of the array data, etc., and separates the fixed part and variable part. Then, an intermediate language expression 6 is generated in which each variable is directly expressed by the ij area and the relative position within the area.

The translation unit 7 performs a so-called optimization process on the intermediate language expression 6 and then performs a coding process to generate an object program 8 corresponding to a so-called machine language.

In the above process, the expression representing the operation is, for example, the expression that specifies the operation to obtain the assignment result on the right side of the assignment statement, as well as the expression that appears at the subscript of the array element, etc. to define the operand of the operation. There is.

In the latter type of expression, to define one array element, "
"array name (subscript expression)" is used in the form of "array name (subscript expression: subscript expression: 1 -------・-: subscript expression)" to define multiple array elements as an array division. It is also used as a character type variable or as a definition for extracting a part of one element of an array. (Subscript expression) (Start position expression: End position expression)'' There are two types of subsequence expressions, etc.

The evaluation of these expressions must be distinguished from the arithmetic expressions of assignment statements and must be processed to obtain a result that determines each specified data element.For example, an array subscript expression consists of a fixed part and a variable part. In this way, it is possible to deal with machine language indexes and displacements. Also, when specified by the debug function, it checks whether the value of the subscript expression is between the lower limit and upper limit declared in the array, and also checks the validity of the start and end positions of the substring expression. preparations are required.

[Problems to be Solved by the Invention] For the above processing, the semantic analysis unit 5, when processing a string of clauses and delimiters, manages the correspondence between parentheses and inserts them into the subscript expression of the quotation of the array element. Furthermore, it is necessary to evaluate expressions while identifying cases where array elements are referenced and distinguishing between subscript expressions, subsequence expressions, and other expressions, which tends to complicate processing and reduce processing performance. This makes it difficult to make improvements or introduce advanced optimization processing.

An object of the present invention is to provide a program translation processing device that generates a tree structure representing the structure of a program sentence and allows high-level translation processing to be configured relatively easily.

[Means to solve the problem]

FIG. 1 is a block diagram showing the configuration of the present invention.

The diagram shows the configuration of a program translation processing device 10, which includes a syntax analysis section 3, a semantic analysis section 11, and a detailed analysis section 12. The syntax analysis section 3 analyzes a given source program 2, and The source program that follows the syntactic rules of a predetermined grammar is segmented by predetermined delimiters and decomposed into strings 4 of clauses and delimiters, and the semantic analysis unit 11 analyzes the string 4 of clauses and delimiters that follow the semantic rules of the grammar. , generates a tree structure 13 determined based on the inclusion relationship of predetermined parentheses, and creates a tree structure 1
3 has a structure that distinguishes between expressions representing operations in the source program 2 and expressions included in information specifying operands of the operations, and the detailed analysis unit 12, based on the tree structure 13,
The expression is processed according to the partition to generate, for each operand, a predetermined intermediate language representation 14 that directly represents the relative position of the operand.

[For production]

This program translation processing device first performs syntactic structural analysis and necessary semantic analysis on a string of clauses and delimiters up to the semantic analysis unit of the present invention, generates a tree structure representing the syntax, and analyzes the tree structure in detail. The analysis section processes it and generates an intermediate language expression, which is then passed to the translation section. Therefore, when extracting an expression from the tree structure in order to evaluate the expression, the detailed analysis section automatically identifies whether the expression is a subscript expression, an expression on the right side of an assignment statement, etc., and evaluates the expression according to the classification. You can start processing. In this way, the processing configuration becomes simple, making it easier to improve processing performance and introduce advanced optimization processing.

〔Example〕

Figure 2(a) shows an example of a tree structure corresponding to the same assignment statement as above, where the syntax nodes are the type of statement "assignment statement", the left side,
It consists of a pointer pointing to the contents of the right side, and the left side is the variable name "C
”, and the right side indicates a binary operation node to indicate the operation and operand.

A binary operation node is an operator followed by a binary pointer,
It points to a binary operation node that indicates the operand or the next operation.

In this example, both pointers point to the operand, but array elements, array divisions, subsequences, etc. are represented by sub-object nodes, and for the first operand, the name pointer points to "A" and the subscript In the expression pointer, the subscript expression is an expression with an operator, so it points to a binary operation node, one of the operands pointed to by this binary operation node is an array element, so it points to a subobject node, and the other is a constant, so it points to that constant. Specify the area that shows the area. In this way, a tree structure corresponding to the assignment statement is constructed.

FIG. 2(b) is an example of a sub-object node representing a subsequence, where r A (3,2) (2:3)
In the case of J, the configuration is as shown in the figure, in which the name pointer points to "A," the subscript pointer points to the subscript node, and the subsequence pointer points to the subsequence node. Note that each of the above nodes has information indicating the type of node, but it is omitted in the diagram.

In FIG. 1, the syntax analysis unit 3 performs error checking from the source program 2 to the string of phrases and delimiters 4, as in the past, but with the present invention, the processing up to the output of the intermediate language expression, as in the past, is Instead, the string 4 of clauses and delimiters is structured by paying attention to the inclusion relationship between parentheses, and the tree structure 13 as in the above example is generated. Therefore, the detailed analysis unit 12 processes the tree structure 13 to generate an intermediate language expression 14.

FIG. 3 is an example of a processing flow when the detailed analysis unit 12 extracts and processes one operand specified from the binary operation node, etc. In processing step 20, the first node of the operand is extracted and processed. In ring step 21, it is determined whether it is a sub-object node, and if it is not a sub-object node, processing is performed in process step 22 for other types of operands, ie, constants or simple variable operands.

If it is a sub-object node, it is identified in processing step 23 whether there is a subscript, and if there is a subscript, in processing step 24,
Extracts the node pointed to by the subscript expression pointer and processes the subscript expression.

In the next processing step 25, it is again determined whether there is a subsequence using the sub-object node, and if there is a subsequence, then in processing step 26, the node pointed to by the subsequence pointer is extracted and the subsequence is processed.

In the above subscript expression processing and substring processing, if the syntax further links binary operation nodes, sub-object nodes of their operands, etc., these are processed recursively.

In the above post-processing step 27, the values of subscripts and substrings are processed to be represented by fixed parts and variable parts, and they are identified in processing step 28. If debugging is specified as an option for this translation processing, processing is performed in processing step 29. Subscript and subsequence inspection processes are performed, and the necessary intermediate language expression 14 is output as a result of each of the above processes.

In this way, conventionally, a column of linearly linked clauses and delimiters is sequentially scanned from the left, for example, "A (expression:
'', this is the first time that it is identified that for a character variable A without a subscript, it specifies a substring and the ``expression'' is its upper limit, or that it is an array division of array A. However, in the present invention, the presence or absence of a subscript, the subscript, and the substring can be immediately identified in the sub-object node, and each localized process can be selected and executed.

The intermediate language expression 14 outputted by the detailed analysis section 12 as described above is processed by the translation section 7 in the same manner as before, and the target program 8 is generated.

〔Effect of the invention〕

As is clear from the above description, according to the present invention, when translating a computer source program into a target program,
By generating a tree structure representing the structure of a program sentence, there is a significant industrial effect in that a high-level translation process can be configured relatively easily.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing the configuration of the present invention, Fig. 2 is a diagram explaining an example of the tree structure of the present invention, Fig. 3 is a flowchart of the processing of the present invention, and Fig. 4 shows an example of the conventional configuration. It is a block diagram. In the figure, 1.10 is a program translation processing unit, 2 is a source program, 3 is a syntactic analysis unit, 4 is a string of clauses and delimiters, 5.11 is a semantic analysis unit, 6.14 is an intermediate language representation, 7
8 is a translation section, 8 is an object program, 12 is a detailed analysis section, 13 is a tree structure, and 20 to 29 are processing steps.

Claims (1)

[Claims] It has a syntax analysis section (3), a semantic analysis section (11), and a detailed analysis section (12), and the syntax analysis section (3) analyzes a given source program (2). The semantic analysis unit (11) divides the source program, which follows the syntactic rules of the predetermined grammar, into segments (4) using predetermined delimiters, and divides the source program into segments (4) of clauses and delimiters. A tree structure (13) determined based on the inclusion relationship of predetermined parentheses is generated for the clause and delimiter sequence (4) according to the following, and the tree structure (13) is used to perform the operation of the source program (2). The detailed analysis unit (12) has a structure that distinguishes between the expression expressed and the expression included in the information specifying the operand of the operation, and the detailed analysis unit (12) analyzes the expression according to the classification based on the tree structure (13). A program translation processing device, characterized in that it is configured to process to generate, for each operand, a predetermined intermediate language expression (14) directly representing the relative position of the operand.