JPS61134843A - Program structure processing system - Google Patents

Abstract

PURPOSE:To decrease the desired memory capacity for a specifying program of the program structure and to process the program structure at a high speed, by using a reference image file for subject programs delivered from a compiler. CONSTITUTION:A list delivered from a compiler is stored to a reference image file 11. A reference list in the file 11 is used to retrieve a routine name, and a table of routine names is produced by a routine name production part 15. This table is stored to a routine name storing part 12. A matrix production part 16 uses a reference list routine name table to produce a reference/to-to- referred table of an (NXN)-bit matrix type. This matrix is retrieved in both the row and column directions and the ON/OFF is retrieved by a retrieving part 17 for the bit of (i) rows and (j) columns. The result of this retrieval is displayed to a printer 14.

Description

TECHNICAL FIELD The present invention relates to a program structure processing system, and more particularly to a program structure processing system for clarifying the reference-reference relationships of a group of routines within a target program.

BACKGROUND ART In general, in a program, information such as "which routine refers to which routine" and "which routine is referenced by which routine" is
The larger the scale of the program, especially the larger the number of routines, the more important it becomes. The former is usually output by the compiler, but the latter is not. The latter plays an important role when modifying a program or dealing with bugs. Therefore, a structure explicit program is required to obtain the latter information. Conventionally, this structure explicit program is processed as follows.

Read the source text of the target program line by line, learn the character string of each line character by character, and write it to the beginning of the compilation unit (for example, if the program is written in FORTRAN, use the 5UBROUTINE statement or FUNCTIO
N sentences, etc.). Then, enter the routine name (for example, SUB for 5UBROUTINE 5UB4)
I) and write its routine name in the reference routine name in the computer's memory as shown in FIG. It then reads the source text line by line again and learns each line's string character by character until it identifies the beginning of the next compilation unit. (CALL statement, etc.)
identify. Then, the referenced routine name is written in the referenced routine name in the computer memory shown in FIG. Then, repeat the same operation as above. This operation is performed on all source texts of the target program.

Through the above steps, a reference/referenced table of the form shown in FIG. 1 is completed. If this table is printed in its original form, information about ``which routine refers to which routine'' will be printed.

Next, the name of the first referenced routine in FIG. 1 is printed, it is checked whether the routine name is in the referenced routine name section of FIG. 1, and if it is, the name of the referenced routine in that line is printed. This operation is performed for all referenced routine names in the first diagram. Furthermore, similar operations are performed for the second reference routine in FIG. When this operation is performed for all reference routines in FIG. 1, "which routine is referenced by which routine" is printed.

Therefore, using this method, (1) it takes an extremely large amount of time to identify the character strings in all lines of the source text of the target program; (2) writing the referenced routine name shown in Figure 1; (For example, if the target program consists of 1000 routines, and each routine refers to a maximum of 30 other routines, and the routine name has an average of 6 characters, Normally, 8 bits are required per character, so 1000 routines x 30 routines/1 routine x 6 characters/1 routine x B bits/1 character = 1,440,000 bits); (3) Which routine In order to obtain the information "from which routine is referenced", it is necessary to identify all parts of the referenced routine name in Figure 1 a number of times equal to the total number of routines in the target program, which takes time; There are drawbacks.

OBJECTS OF THE INVENTION An object of the present invention is to provide a program structure processing system capable of eliminating the above-mentioned drawbacks, increasing the processing speed of structure-explicit programs, and further reducing the amount of memory used.

Composition of the Invention The program structure processing system according to the present invention analyzes the program structure of a target program using a reference list output from a compiler that compiles the target program in which the reference/reference relationships of a group of routines are to be clarified. The target is a program structure processing system to be examined, and its feature is that it sequentially derives each record of the reference list, finds the routine name contained in each record, and creates a routine name table in which the routine names are sequentially arranged. , storage means for storing this routine name table, storage means for storing an N×N bit matrix table corresponding to the number N of routine names (N is a positive integer), and each compilation unit of the reference list is stored in the routine name table. i-th (i ke 0<i≦
means for determining that the routine is a routine (an integer N);
The routine name referenced by this i-th routine compilation unit is the j-th routine name in the routine name table (j is o
<j≦N integer) and a means for determining whether the routine is a routine with
and a means for turning on a bit in row and column j, and the reference/reference relationship of the routine group in the target program is obtained in accordance with the on/off of the bit in the matrix table.

EXAMPLE Hereinafter, an example of the present invention will be described in detail with reference to the drawings.

Generally, the reference list output by a compiler differs depending on the compiler, but it is in a fixed format for each compiler, as shown in Figure 2.
Be sure to include the "routine name given to the compilation unit" and the "routine name of another routine referenced from that routine"
Contains.

In the following, this example will be described assuming a compiler that outputs a reference list in the form shown in Figure 2, but it is clear that the same thing can be done for compilers with substantially different output formats. .

In the present invention, the input of the structure clarifying program is an image of a reference list outputted when the target program is compiled, stored in a magnetic storage medium (hereinafter referred to as a reference image file). Therefore,
Directs compiler output to a magnetic storage medium instead of to a list. Taking these points into consideration, the functional blocks of the embodiment of the present invention shown in FIG. 3 are obtained.

In FIG. 3, reference numeral 11 is a storage unit that stores a reference list that is the output of the compiler, and is a reference image file. Reference numeral 15 denotes a routine name creation unit that searches for routine names of a routine group using the reference list in the reference image file 11 and creates a routine name table. This routine name table is stored in the routine name storage unit 12 made of a magnetic storage medium.
is stored in

Reference numeral 16 denotes an N×N bit matrix creation unit that creates a reference/referenced table in an N×N bit matrix format using the reference list and the routine name table. The data is stored in the bit matrix storage section 13.

Reference numeral 17 denotes a search unit that searches the N×N bit matrix in the row and column directions to detect the on/off state of the bit in the i row and j column, and this search result is sent to the printer 14.
It is now visible. In addition, 18 is a control part, and has the function of controlling each of the above-mentioned components 11 to 17.

FIG. 4 is a flowchart showing the outline of the operation of the blocks in FIG.
This is the flow from inputting the following information to obtaining information on ``which routines refer to which routines'' and ``which routines are referenced by which routines,'' and visually displaying this information.

10-/II%1 In process 1 in FIG. 4, a routine name table in the format shown in FIG. 5 is created. Details of this process 1 are shown in FIG. Process 1-1 in FIG. 6 is the initial value setting of the routine number counter of the target program. In process 1-2, the code of the reference image file 11 is read. In process 1-3, it is determined whether the information read in process 1-2 is equal to the information in the last record of the reference image file. This is a method in which the repetition of processing ends when the last record of the reference image file is read, and is a normal means for having a computer perform the processing.

Process 1-4 determines whether the information read in process 1-2 is a record containing the target routine name. In process 1-5, a record including the routine name is found, so the routine number counter is incremented by one. In process 1-6, the routine name in the record found in process 1-4 is registered in the routine name table and stored in the routine name table storage section 12.

In process 2 in FIG. 4, a reference/referenced table of the form shown in FIG. 7 is created on the magnetic storage medium 13. The details of this process 2 are shown in FIG. 8, and the process 2-1 in FIG.
is the routine number in the routine name table,
The subscript indicating whether a reference/referenced table is to be created is initialized. In process 2-2, reference and referenced tables are secured and initialized. In process 2-3, the reference image file code is read. Processing 2-
In step 4, it is determined whether the information read in step 2-3 is equal to the information in the last record of the reference image file, and has the same meaning as step 1-3 in FIG.

In process 2-5 and process 2-6, it is determined whether the information read in process 2-3 is the beginning of the reference image file compilation unit, and if it is the beginning of the compilation unit, A subscript value is set that indicates the routine number of the compilation unit in the routine name table. In process 2-7, it is determined whether the information read in process 2-3 is equal to the information containing the routine name referenced by the compilation unit.

In process 2-8, it is checked at what position in the routine name table the routine referred to by the compilation unit is located. In process 2-9, it is registered in the reference/referenced table that the i-th routine in the routine name table refers to the j-th routine in the routine name table.

In process 3 in FIG. 4, information indicating "which routine refers to this routine" is printed. Details of this process are shown in FIG.

In process 3-1 of FIG. 9, it is determined at which ``-chi'' in the routine name table refers to another ``-chi''.
The subscript is initialized to indicate the . In process 3-2, the name of the routine to be referenced is printed. Processing 3-
3 searches for referenced routines and prints them. In process 3-4, it is determined whether processing has been completed for all routines on the reference side. In process 3-5,
The subscript is incremented by 1 to point to the next routine name in the routine name table as the next referenced routine.

In process 4 in FIG. 4, information indicating "which routine is referred to by which routine" is printed. Details of this process are shown in Figure 1O. In process 4-1 in FIG. 10, a subscript indicating which routine in the routine name table is referenced from another routine is initialized. In process 4-2, the referenced routine name is printed. In process 4-3, reference routines are searched for and printed. Processing 4
-4, it is determined whether processing has been completed for all routines on the referenced side. In process 4-5, the subscript is incremented by 1 so as to point to the next routine name in the routine name table as the next referenced routine.

In the manner described above, it is possible to obtain the desired information of ``which routine refers to which routine'' and ``which routine is referenced by which routine.''

According to the present invention, by using the reference image file output by the compiler for the target program as input to the program structure clarifying program according to the present invention,
By searching the entire source text of the target program, which was done by the conventional program structure clarifying program,
This has the effect of making it extremely easy to search for the beginning of a compilation unit and references to other routines, thereby reducing processing time. Furthermore, turning on bits of magnetic storage media.

By inserting OFF into the reference/referenced table in the program structure explicit program, the required memory capacity of the program structure explicit program is reduced and the processing time is also shortened.

[Brief explanation of drawings]

FIG. 1 is a diagram showing a reference/referenced table in a conventional program structure clarifying program, FIG. 2 is a diagram showing a reference list output by a compiler, and FIG. 3 is a schematic block diagram of an embodiment of the present invention. FIG. 4 is a flowchart showing an overview of the operations of the blocks in FIG. 3, FIG. 5 is a diagram showing a routine name table, FIG. 6 is a detailed flowchart of process 1 in the flow of FIG. 4, and FIG.
Figures 8 to 10 showing the referenced table are flowcharts showing details of each of processes 2 to 4 in the flow of FIG. 4. Explanation of symbols of main parts

Claims (1)

[Claims] A program structure processing system that examines the program structure of a target program using a reference list output from a Koipara that compiles a target program in which reference/reference relationships of a group of routines are to be clarified, the reference list comprising: means for sequentially deriving each record of and finding the routine name contained in each record and creating a routine name table in which the routine names are sequentially arranged;
storage means for storing this routine name table; storage means for storing an N×N bit matrix table corresponding to the number N of routine names (N is a positive integer); A means for determining that the routine name is the i-th routine (i is an integer such that 0<i≦N) in the routine name table, and a means for determining that the routine name referenced by the compilation unit that is the i-th routine is the j-th routine name (i is an integer such that 0<i≦N) in the routine name table. j is an integer such that 0<j≦N); and means for turning on a bit in row i and column j of the matrix table based on these judgment results. A program structure processing system characterized in that a reference/reference relationship of a group of routines in the target program is obtained according to whether a bit is turned on or off.