JPH0926897A - Program analyzer and program analyzing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a program analyzer and a program analyzing method by which a source code is efficiently analyzed by a static analysis and the understanding of the contents of the source code including processing contents and input/output specifications is facilitated. SOLUTION: A data extraction means 1 extracts data information to be the information showing the structure of each data included in a source code from the source code. A related extraction means 5 extracts related information showing the relation existing between data for each location on the source code based on the source code and data information. A processing extraction means 6 extracts each related information related to each processing in the source code as processing information to be the information showing a processing, based on the source code, data information and related information. The range/mode of an output is designated by a designation part 7. An output means 2 outputs each extracted information by the designated mode.

Description

Detailed Description of the Invention

[0001]

The present invention relates to a program for efficiently analyzing the source code of a computer program by static analysis and facilitating the understanding of the content of the source code including the processing content and the input / output specifications based on the analysis result. The present invention relates to an analysis device and a program analysis method.

[0002]

2. Description of the Related Art In the process of inspection and maintenance in software development, when the source code of a computer program is reviewed and inspected, or when reverse engineering is performed for improvement, the content of the source code, that is, the processing content and input / output specifications are analyzed. There is a need to.

One of the conventional techniques for analyzing the source code is to generate an execution module or test-execute the program by using an interpreter and actually process the data (dynamic analysis). ).

Another method of analysis is symbolic execution of programs, that is, static analysis. In a typical static analysis method, elements such as variables in a program are represented by abstract symbols such as the possibility of an error value instead of concrete attribute values. The state of processing of the program is represented by a combination of these symbols.

Further, as another method of analyzing the source code, a human may manually read the source code and grasp the contents.

[0006]

However, the conventional method as described above has the following problems.

First, the dynamic analysis cannot be applied to the case where the program malfunction is that the program cannot actually operate in the first place. Moreover, it is complicated to generate an execution module each time the program is changed. Further, in order to test and execute only a part of the program, a special program module for the driver or stub is required, which is complicated. Moreover, although test data is required for test execution, it is difficult and complicated to create test data based on conditions that completely match the purpose of analysis.

Further, in the conventional symbolic execution, it is possible to analyze the transition of data in an abstract manner. However, in the case where the execution content changes according to the concrete data to be input, so to speak, the source code is input. It was difficult to understand the output specifications.

Further, the manual reading of the source code has problems that it is complicated and requires a lot of labor, and that the work quality varies depending on the skill of the person who performs the reading.

The present invention has been proposed to solve the problems of the prior art as described above, and its purpose is to efficiently analyze the source code by static analysis and to analyze the processing contents according to the analysis result. And a program analysis device and a program analysis method for facilitating the understanding of the contents of the source code including the input / output specifications. Another object of the present invention is to provide a program analysis device and a program analysis method whose processing procedure is easy.

[0011]

In order to achieve the above object, the present invention has the following configuration and operation. That is, the invention of claim 1 is, in a program analysis device that analyzes a source code of a computer program that represents a process for data, extracting data information that is information representing the structure of each data from the source code. Means, and related information extracting means for extracting related information between data at a predetermined position of the source code based on the source code and the data information.

According to a fourteenth aspect of the present invention, the invention of the first aspect is grasped from the viewpoint of a method, and in a program analysis method for analyzing a source code of a computer program that represents a process for data, A data extraction step of extracting data information which is information representing the structure of each data, and a relation extraction for extracting relation information between data at a predetermined position of the source code based on the source code and the data information. And a step.

According to the inventions of claims 1 and 14, data information representing a structure of data included in the source code is extracted from the source code. This means getting a table for the data in the source code. Therefore, it becomes easy to understand the contents of the source code through understanding the data. Also, based on the source code and the data information, related information between data is extracted at a predetermined position of the source code. This means identifying where and in the source code each extracted data has a relationship. Therefore, it becomes easy to understand the contents of the source code by understanding the relationship between the data.

In the program analysis device according to claim 1, labeling means for adding a position ID to each part of the source code is provided, or in the program analysis method according to claim 14, each part of the source code is added. A labeling step of adding the position ID may be included.

In this way, since the position ID is added to each part of the source code, it is possible to store the extracted information.
Each processing such as search / output can be performed based on the position ID, and it becomes easy to specify the target range of processing such as analysis and output based on only text output and text input.

Further, in the program analysis device according to claim 1, a preprocessor for adding an include file of the source code to the source code is provided, or in the program analysis device according to claim 14, the source code is associated with the source code. May include the step of subscribing to the include file.

In this way, even in a language using an include file such as C language, the include file is added to the source code, so that the user does not need to perform this work in advance. Also, the step of referring to the include file during processing can be omitted.

The invention according to claim 2 is the program analysis apparatus according to claim 1, characterized by further comprising a reference means for referring to another file related to the source code.

A fifteenth aspect of the present invention is obtained by grasping the second aspect of the invention from the viewpoint of a method. In the program analysis method according to the fourteenth aspect, reference is made to another file related to the source code. It is characterized by including a step.

According to the second and fifteenth aspects of the invention, even in the case of a language such as C language which supports division compilation, other files are referred to during processing, so that necessary information is added to the source code in advance. The step of setting can be omitted.

According to a third aspect of the present invention, in the program analysis apparatus according to the first aspect, the data information includes a structure D that stores information for each data, and a structure T that stores information for each data type. And the structure D is related to the corresponding structure T.

The invention of claim 16 is obtained by grasping the invention of claim 3 from the viewpoint of a method. In the program analysis method of claim 14, the data information is a structure for storing information for each data. The structure D includes a body D and a structure T that stores information for each data type, and the structure D is associated with the corresponding structure T.

Further, in the invention of claims 3 and 16, the information for each data type is different from the information for each data structure T.
Since it is stored in, the memory efficiency is improved even when there are multiple data of the same type. Further, since the structure D and the structure T are connected by the pointer, data retrieval is easy.

The invention according to claim 4 is the program analyzing apparatus according to claim 1, wherein the related information extracting means is configured to extract the type of the relationship.

The invention of claim 17 is the invention of claim 4 grasped from the viewpoint of a method, and in the program analysis method of claim 14, the relation extracting step extracts the type of the relation. It is characterized by

Further, in the inventions of claims 4 and 17, not only the existence of the relation between the data but also the type of the relation is extracted.
Therefore, it becomes easier to understand the contents of the source code by more specifically understanding the relation between the data.

In the program analysis device according to claim 1 or the program analysis method according to claim 14, the data information includes a structure D that stores information for each data and a structure D that stores information for each data type. The structure D is associated with the corresponding structure T, and the related information includes a structure R1 that stores an ID of a relationship for each position and a structure that stores specific information for each relationship. Body R2
And the structure R1 is associated with the corresponding structure R2, and each structure R2 is connected to the structure D of each related data by a pointer.

In this way, since the information for each association is stored in the structure R2 different from the information for each location, the memory efficiency is improved even when the same association exists in a plurality of locations. Further, since the structure R1 and the structure R2 are connected by the pointer and the structure R2 and the structure D are connected by the pointer, data retrieval is also easy.

Further, in the program analysis device according to claim 1 or the program analysis method according to claim 14, the structure R1 stores an ID of a set of data having each relation for each position and is included in the set. The ID of the data to be stored may be stored in the structure C, and the structure R1 may be connected to the corresponding structure C by a pointer.

By doing so, since the ID of the data of the set having each relation for each position is stored in the structure C,
Various information can be searched easily.

According to a fifth aspect of the present invention, in the program analysis device according to the first aspect, based on the source code, the data information, and the related information, the related information regarding a predetermined process in the source code is processed information. It is characterized by having a processing information extracting means for extracting as.

The invention of claim 18 is obtained by grasping the invention of claim 5 from the viewpoint of a method, and in the program analysis method of claim 14, based on the source code, the data information and the related information. And a process information extracting step of extracting, as process information, related information relating to a predetermined process in the source code.

In the inventions of claims 5 and 18, based on the source code, the data information and the related information, each related information related to each processing in the source code is extracted as processing information which is information representing the processing. To be done. Therefore, each process included in the source code can be grasped by how the relation between the data changes before and after the process, and the contents of the source code can be grasped easily.

According to a sixth aspect of the present invention, in the program analysis apparatus according to the fifth aspect, the processing information extracting means, regarding the processing to be performed based on the source code, the related information before the processing and the related information. Is configured to extract the set of related information of

The invention of claim 19 is the invention of claim 6 grasped from the viewpoint of a method, and in the program analysis method of claim 18, the processing information extracting step is based on the source code. With respect to the processing to be performed, a set of the related information before the processing and the related information after the processing is extracted.

Further, in the inventions of claims 6 and 19, a set of the related information before the processing and the related information after the processing is extracted as the processing information related to each processing. Therefore, not only the related information before and after the processing but also the related information before and after the processing can be easily examined, understood, and processed.

According to a seventh aspect of the present invention, in the program analysis apparatus according to the sixth aspect, the processing information extracting means detects consistency between data obtained by tracing the related information when extracting the processing information. And a correction means for correcting the related information based on the consistency detected by this means.

The invention of claim 20 is obtained by grasping the invention of claim 7 from the viewpoint of a method, and in the program analysis method of claim 18, the processing information extracting step is performed when the processing information is extracted. , A consistency detecting step of detecting consistency between data obtained by tracing the related information, and a correcting step of correcting the related information based on the consistency detected by the means. To do.

Further, in the inventions of claims 7 and 20, since the indirect relationship via the pointer is also reflected in the related information, the accuracy of the related information is improved.

In the program analysis device according to claim 5 or the program analysis method according to claim 18, the data information includes a structure D for storing information for each data and a structure for storing information for each data type. The structure D is associated with the corresponding structure T, and the related information includes a structure R1 that stores an ID of a relationship for each position and a structure that stores specific information for each relationship. Body R2
And the structure R1 is associated with the corresponding structure R2, each structure R2 is associated with the structure D of each related data, and the processing information is for each execution statement. May have a structure P that stores each position ID of the related information before and after processing, and each structure P is connected to the corresponding structure R1 by a pointer.

In this way, since the position IDs before and after the processing for each execution statement are stored in the structure P, each structure P is connected to the corresponding structure R1 by the pointer, so that the data retrieval is performed. Is also easy.

The invention according to claim 8 is the program analyzing apparatus according to claim 1 or 5, characterized by further comprising an analysis range designating means for designating a range of the extraction.

The invention of claim 21 comprehends the invention of claim 8 from the viewpoint of a method, and claims 14 or 1
8. The program analysis method according to 8, further comprising an analysis range designating step of designating the extraction range.

In the invention described in claims 8 and 21, the range to be subjected to the information extraction can be limited to a necessary part instead of the entire source code, so that the processing efficiency is improved.

According to a ninth aspect of the present invention, in the program analysis apparatus according to the first or fifth aspect, the output range designation means for designating a range of information to be output and the designated range from each of the extracted information. It is characterized by comprising a selecting means for selecting information to be stored.

The invention of claim 22 comprehends the invention of claim 9 from the viewpoint of a method, and claims 14 or 1
8. The program analysis method according to 8, further comprising an output range specifying step for specifying a range of information to be output, and a selecting step for selecting information corresponding to the specified range from each of the extracted information. To do.

According to the ninth and twenty-second aspects of the present invention, the range of the information to be output can be limited to a necessary part instead of the entire source code, so that the information can be easily understood.

According to a tenth aspect of the present invention, in the program analyzing apparatus according to the ninth aspect, there is provided a first detecting means for detecting a repetitive structure in the source code, and the selecting means comprises:
With respect to the part related to the detected repetitive structure, it is characterized in that it is configured to select the information regarding the processing related to the plurality of times of repetition and one or less times.

The invention of claim 23 is obtained by grasping the invention of claim 10 from the viewpoint of a method, and in the program analysis method of claim 22, the first structure for detecting an iterative structure in the source code. The method further includes a detection step, and the selection step is characterized by selecting information regarding a process related to a plurality of times of repetition and one or less times with respect to a portion related to the detected repetitive structure.

According to the tenth and twenty-third aspects of the present invention, the information regarding the processing relating to the necessary and sufficient number of iterations of the iterative structure is output, so that the analysis is made efficient.

The eleventh aspect of the invention is the program analyzing apparatus according to the ninth aspect, further comprising second detecting means for detecting a selection structure in the source code, wherein the selecting means comprises:
With respect to the detected part related to the selection structure, it is characterized in that the information about each process corresponding to each judgment result of the selection condition is selected.

A twenty-fourth aspect of the present invention is obtained by grasping the eleventh aspect of the invention from the viewpoint of a method, and in the program analysis method of the twenty-second aspect, a second structure for detecting a selection structure in the source code is provided. It is characterized in that it has a detecting step, and in the selecting step, for a part related to the detected selection structure, information regarding each process corresponding to each judgment result of the selection condition is selected.

According to the eleventh and twenty-fourth aspects of the present invention, as for the selection structure, the information about each process corresponding to each judgment result of the selection condition is output, so that the information about the selection structure can be easily understood.

In the program analyzing apparatus according to claim 1 or 5, there is provided an output range designating means for designating an output mode of information and an output means for outputting each selected and extracted information in a designated mode. May be. Claim 14
Alternatively, the program analysis method according to Item 18 may include an output range designating step for designating an output mode of the information and an output step for outputting each selected and extracted information in the designated mode.

In this way, the form of the information to be output, that is, the degree of detail, the format, etc., can be freely specified, so that the output suitable for the purpose becomes easy.

According to a twelfth aspect of the present invention, in the program analysis device according to the first or fifth aspect, the source code,
It is characterized by having an output means configured to display any one of the extracted information on the same display device or a display device adjacent to each other.

The invention of claim 25 is obtained by grasping the invention of claim 12 from the viewpoint of a method, and in the program analysis method according to claim 14 or 18, the whole or a part of the source code and the extraction are extracted. It is characterized by including an output step of displaying information corresponding to all or a part of any one of the above-mentioned generated information on the same display device or a display device adjacent to each other.

In the twelfth and twenty-fifth aspects of the invention, for example, when a problem is found in the analysis result, it is possible to easily specify which part of the source code causes the problem. On the contrary, in order to confirm that the implementation by the source code produces the intended result, it is possible to easily specify which part on the analysis result the mounting part affects.

According to a thirteenth aspect of the present invention, in the program analysis apparatus according to the twelfth aspect, the output means corresponds to the designated portion when one portion of the source code or the information is designated. It is characterized in that the other part is highlighted.

The invention of claim 26 is obtained by grasping the invention of claim 13 from the viewpoint of a method, and in the program analysis method according to claim 25, the output step is:
When a part of one of the source code and the information is designated, the other part corresponding to the designated part is highlighted.

According to the thirteenth and twenty-sixth aspects of the present invention, since the corresponding portion is highlighted, the information can be grasped more easily.

[0062]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, embodiments of the present invention (hereinafter referred to as "embodiments") will be specifically described with reference to the drawings. It should be noted that each function of the embodiment is realized by a predetermined software controlling a computer including peripheral devices. However, since various configurations of the computer and software are conceivable, each function and each process of the embodiment are supported. The description will be made assuming a virtual circuit block to be used.

(1) Configuration of the First Embodiment The first embodiment is a program analysis device corresponding to claims 1 to 9 and a program analysis method executed by this program analysis device (corresponding to claims 14 to 22). ), The purpose of which is to efficiently analyze the source code by static analysis and to facilitate understanding of the content of the source code including the processing content and the input / output specifications based on the analysis result. It is to provide an analysis method. Another object of the first embodiment is to provide a program analysis device and a program analysis method whose processing procedure is easy.

First, FIG. 1 is a functional block diagram showing the configuration of the program analysis device of the first embodiment. As shown in this figure, the program analysis device of the first embodiment (hereinafter referred to as “this device”) extracts data information, which is information representing the structure of each data included in the source code, from the source code of the computer program. Data information extraction means 1 for extracting and output means 2 for outputting the extracted information
And (claim 1).

The apparatus further comprises labeling means 3 for adding a position ID to each part of the source code, and a preprocessor 4 for adding an include file of the source code to the source code.

The present apparatus further includes the related information extracting means 5 for extracting the related information, which is the information indicating the relationship existing between the data, for each position of the source code based on the source code and the data information. Have (claim 1).

Further, the present apparatus extracts, based on the source code, the data information and the related information, each related information related to each processing in the source code as processing information which is information representing the processing. The information extraction means 6 is provided (Claim 5).

Further, the present apparatus further comprises a designation unit 7 (the analysis range designation means (claim 8) and the output range designation means (claim 9) for designating the extraction range, the output range and the output mode. Equivalent). Further, the present apparatus has a selection means 8 for selecting information corresponding to the specified range from the extracted respective information (claim 9). The output unit 2 is configured to output each selected and extracted information in a designated manner.

(2) Operation and effect of the first embodiment In the first embodiment having the above-mentioned configuration, analysis or output of all or part of the source code of the computer program is performed as follows. Here, FIG. 2 is a flowchart showing the operation procedure of the first embodiment. That is, first, the source code to be analyzed is given by keyboard input or file reading (step 1).

[Preprocessing] In the case of a source code of a language using an include file such as C language, the preprocessor 4 adds the include file of the source code to the source code (step 2). More specifically, processing generally executed before execution and compilation, such as comment line removal and macro statement expansion, is performed. Therefore, in the first embodiment, the user does not need to perform this work in advance. Also, the step of referring to the include file during processing can be omitted. For example, FIG. 5 shows the result of adding the include file of FIG. 4 to the source code of FIG.

However, the preprocessor does not necessarily have to be incorporated in the device, and the user may give the source code to the device after the include file is included in the preprocessor in advance.

Further, reference means may be provided for referring to the source code and other files defining data, procedures, etc. appearing in the source code (claim 2). By doing this, even in the case of a language that supports separate compilation, such as C, another file is referenced during processing (claims 2 and 15), so that necessary information is added to the source code in advance. It is possible to omit the procedure to save.

[Labeling Step] Next, the labeling means 3 adds the position ID (including the execution statement ID) to each part of the source code (step 3). For example, FIG.
The position ID is added between the execution statement of the source code of the source code and the execution statement ID (four-digit number following "@S") immediately before the first execution statement and immediately after the last execution statement. The result of adding (four-digit number following "@P") is shown in FIG. Note that the labeling may be performed on the original source code separately from the source code to which the include file is added, but may be of course performed on the source code to which the include file is added.

As described above, in the first embodiment, since the position ID is added to each part of the source code, each process such as storing, retrieving and outputting the extracted information can be performed based on the position ID. It is possible to specify the target range of processing such as analysis and output based on only text output and text input.

[Step of Specifying Analysis Range] The analysis range to be extracted can be specified by the specifying unit 7 (step 4 / claim 21). In the first embodiment, the designation of various ranges is performed by the position ID attached to the source code.
It is possible to specify various ranges by other methods such as pointing by a cursor controlled by a pointing device such as a mouse. As described above, in the first embodiment, the range of the information extraction target is limited to a necessary part instead of the entire source code, and the processing efficiency can be improved (claims 8 and 21).

[Data Extraction Step] Next, first,
The data information extracting means 1 extracts data information, which is information representing the structure of each data included in the source code, from the source code (step 5 / claim 14). This means getting a table for the data in the source code. This process is performed by parsing the target language parser,
It becomes possible by extracting other identifiers.
Therefore, it becomes easy to understand the contents of the source code through understanding the data.

Here, FIG. 7 is a conceptual diagram showing a general storage format of each information extracted in this embodiment. That is, as shown in this figure, the data information includes a structure D that stores information for each data and a structure T that stores information for each data type, and the structure D corresponds to the corresponding structure T. They are connected by a pointer (claims 3 and 16).

Each “structure” is a fixed-length or variable-length data area. Specifically, each structure D is a data ID that is a key for searching, a data name in the source code,
It includes a type ID indicating the type to which the data belongs and a pointer multiplicity that is the multiplicity of the pointer operator of the data.

Further, the structure T includes a type ID which is a key for retrieval, a type name in the source code, and type definition information indicating the definition contents of the type. The type definition information is configuration information. The configuration information stores the structure definition and union definition components, that is, the type name, member name, and tag name as elements in the order described in the definition statement. Immediately after the last element, for example, a 0 value is stored as a delimiter. In the case of basic types such as int and char, the configuration information is empty and only 0 value is stored.

For example, the data information of FIG. 8 can be obtained from the source code of FIG. This example shows the structure struct List
Type definition information for the int members dummy and struct L
It is recorded that the ist * type member next is a component.

As described above, in the first embodiment, the information for each data type is stored in the structure T which is different from the information for each data. Efficiency is improved. Further, since the structure D and the structure T are connected by the pointer, data retrieval is easy (claims 3 and 16).

[Relationship Extracting Step] Next, the related information extracting means 5 determines, based on the source code and the data information.
For each position on the source code, the related information indicating the relationship existing between the data is extracted (step 6 / claim 1).
4). This means identifying where and in the source code each extracted data has a relationship. Therefore, it becomes easy to understand the contents of the source code by understanding the relationship between the data.

Specifically, each position ID of the source code
For each (execution statement ID), the type of assignment relationship, equivalence relationship, pointer reference, etc. is specified for the mutual relationship between the data provided by the execution statement or data definition statement described at that position, and the result is specified as the related data. And is stored (claims 4 and 17). For example, for an assignment statement, data representing an assignment relationship or pointer reference relationship is generated according to the type of data, and for a condition comparison statement, data representing various equivalence relationships depending on the content is generated. The correspondence between such a typical pattern for each sentence and the correlation between the data to be generated is stored in advance in the form of a knowledge base. As described above, in the first embodiment, not only the existence of the relation between the data but also the type of the relation is extracted. Therefore, it becomes easier to understand the contents of the source code by more specifically understanding the relation between the data.

Here, as shown in FIG. 7, the related information is
A structure R1 that stores the ID of the relation for each position and a structure R2 that stores the specific information of each relation, and the structure R1 is connected to the corresponding structure R2 by a pointer,
Each structure R2 is connected to the structure D of each related data by a pointer.

The structure R1 is a position ID which is a key for specifying a position on the source code, and a set ID which is a key for specifying a set of data included in each relation at the position. It contains several relationship IDs that are keys to identify each relationship at that location. Also, the set ID
A structure C including some data IDs included in the set is provided corresponding to each. As described above, in the first embodiment, since the ID of the data of the set having each relation for each position is stored in the structure C, various information can be easily searched.

Further, the structure R2 includes a relation ID which is a key for retrieval and each data ID for specifying each data related to the relation, and the element is a detailed element designation of each data. . The specific type of association is indicated by the association type.

Here, FIG. 9 is an example of data information and related information. In the data information shown in this figure, the data name is nil
Indicates the area referenced by the pointer and the right-hand side value that is dummy in the assignment. The area referred to by the pointer is not directly described with a specific data name in the source code, and the pointer to point is premised on the input / output specifications. In addition, if the element in the related information is nil, the related target is not the element but the data itself indicated by the data ID, and the related types "→", "==", ": = "Indicates pointer reference relation, equivalence relation, and assignment relation, respectively.

That is, in FIG. 9, the structure (R1) 1301 with the position ID 0003 is the set ID 1003 and the related IDs 2002 and 20.
It has 05 and as an element. Further, the structure (C) 1302 of the set ID 1003 has the data IDs 1101, 1102 and 1104 as elements.

In each such structure, for example,
By referring to the structure (D) 1303 corresponding to the data ID 1101, the data name is list and the data type is the type ID 0001.
That is, information of struct List and pointer type with multiplicity of 1 is obtained. As described above, in the first embodiment, the structure C of the data set is traced using the set ID of the structure R1 of the related information as a key, and the data ID of the structure C of the set is further used.
By reading the information of the structure D of the data information by using as a key, the information of all the related data in the specific related information can be obtained.

Similarly, for example, from the structure (R2) 1304 having the relation ID 2005 to the data ID 1104, that is, the record next of the data area of the structList type, the data ID 1102, that is, st
You can know the relation that the value of pointer type data item of ruct List is assigned.

As described above, in the first embodiment, since the information for each association is stored in the structure R2 different from the information for each location, even if the same association exists in a plurality of locations, the memory efficiency is high. Is improved. Further, the structure R1 and the structure R
2. Since the structure R2 and the structure D are connected to each other by pointers, data retrieval is easy.

[Process Extraction Step] Further, the process information extracting means 6 processes the relevant information relating to each process (each execution statement) in the source code based on the source code, the data information and the relevant information. Is extracted as processing information that is information indicating (step 7 / claim 18). Therefore, each process included in the source code can be grasped by how the relation between the data changes before and after the process, and the contents of the source code can be grasped easily.

Here, as the processing information related to each processing,
A set of related information before the processing and related information after the processing is extracted (claims 6 and 19). Therefore, not only the related information before and after the processing but also the related information before and after the processing can be easily examined, understood, and processed. However, the content of the processing information is arbitrary, and for example, only the differences between the related information before and after the processing may be extracted as the processing information.

As shown in FIG. 7, the processing information has a structure P that stores each position ID of the related information before and after the processing for each execution statement, and each structure P has a corresponding structure R1. Since it is connected to the with a pointer, data retrieval is easy.

The structure P has an execution statement ID corresponding to the execution statement in the source code, a position ID of the related information before processing of the execution statement, and a position ID of the related information after processing. Further, in the process extraction, in extracting the process information, the related information is modified based on the relation between the data indirectly related by the pointer (claims 7 and 2).
0). Therefore, the accuracy of the related information is improved.

That is, the related information implicitly premised in the processing of each executable statement is corrected. For example, in FIG. 9, the contents of the data set (ID1001) at position @ P0001 are list
There are only two pointer values of item and item, but the execution statement
As can be seen from the description of @ S0001 (FIG. 6), actually the two pointers are not null values, that is, they point to a certain area. For this reason it must be assumed that the data set at position @ P0001 is the same as position @ P0002.

Further, the relationship of the related ID 2005 also needs to be corrected based on the fact that the pointer list-> next actually points to the area pointed to by the pointer item, that is, the data of the data ID 1103.

FIG. 1 shows the related information modified as described above.
0 is shown. The modified portion compared to FIG. 9 is the shaded portion. As a result, all data correlations before and after processing are determined for each execution statement. Here, as the processing information, the information as shown in FIG. 11 is obtained.

[Step of Specifying Output Range] When outputting the analysis result, the specifying section 7 can specify the range of output (step 8), and the selecting means 8 is specified from each of the extracted information. The information corresponding to the selected range is selected (step 10 / claim 22). Thus, the first
In the embodiment, the range of information to be output can be limited to a necessary part rather than the entire source code, so that the information can be easily understood.

Here, the screen display device is used as the output means, but the display range may be specified by, for example, the labels of the first and last executable statements, or the cursor may be selected from the source code displayed on the screen. You can specify it with the or mouse. In the selection of information (step 10), the processing information related to the specified range and all the related information and data information that can be directly and indirectly traced from the structure P of this processing information are selected as output targets.

[Step of Specifying Output Mode] When outputting, the specifying unit 7 can specify the output mode (step 9). As the aspect, typically, the level of detail and the output format can be considered. For example, as the level of detail, it is possible to consider whether to display the information about the content of the data type together with the related information. As the output format, for example, only text output or graphical expression can be considered. As described above, in the first embodiment, the mode of the information to be output, that is, the degree of detail, the output format, and the like can be freely specified, so that the output suitable for the purpose is facilitated.

[Output Step] And the output means 2
Outputs the extracted information in a specified manner (step 11). For example, execution statement @ S0001 to execution statement @ S0002
When is designated, the processing information shown in FIG. 11 is output in the designated manner, so that the processing content can be confirmed, and at the same time, the input / output specifications can be confirmed. 12 shows an output example, and FIG. 13 shows a legend of the output example.

For example, in FIG. 12, the display area 1 of the display range is displayed.
In 601 the display range is indicated by the executable sentence ID. Further, in the data-related display units 1602, 1603, 1604, position IDs and data mutual relationships at the position IDs are shown. Data that is not directly described in the source code but is assumed as input / output specifications is displayed by dotted lines. Further, FIG. 14 is an example of visualization focusing only on the data name.

After the analysis is once performed, step 8
It is possible to confirm the output result with various degrees of abstraction and formats by repeating the procedure from to step 12. When analyzing another source code, the processing from step 1 is performed again.

As described above, according to the first embodiment, the processing contents and the input / output specifications realized in the source code of the computer program are efficiently analyzed by static analysis and output at an arbitrary degree of abstraction. can do. As a result, the inspection work of the source code becomes easy, and it is possible to improve the work efficiency such as confirmation of defects in the software testing process. Also, in the maintenance process, it is possible to improve the efficiency of the correction work accompanying the addition of functions.

(3) Second Embodiment Next, an example of analysis of a program including a control structure will be shown. That is, in the case of a structured language such as the C language, the control structure is sequential, iterative, and selective, of which the sequential analysis can be performed as described above.

Therefore, the analysis of the repetitive structure will be described below. That is, the components of the iterative structure are roughly the initialization for iteration (for example, initialization of a counter, etc., and hereinafter referred to as “initialization processing unit”), and the part to be iterated (hereinafter “iteration target”). It is divided into parts). It should be noted that the repetition target portion also includes a portion that determines a condition for performing the repetition.

In this case, the first detecting means for detecting the repetitive structure in the source code is provided, and the selecting means has 0 repetitions and 2 repetitions for the portion related to the detected repetitive structure.
It is configured to select the information related to the processing related to the time (claims 10 and 23). In this way, the information regarding the processing relating to the necessary and sufficient number of iterations for the iterative structure is output, which improves the efficiency of the analysis.

Specifically, although there are a plurality of syntaxes for realizing the repetitive structure in the C language, they are common in that they each include the above-mentioned two constituent elements. Therefore, here, the program including the for statement is taken as an example (FIG. 15), and the include file is the same as that in FIG.

In the analysis of this program including the repetitive structure, the analysis results for the initialization processing section and the repetitive target section are prepared. That is, in the iterative structure, generally, a plurality of results can be generated, but in order to confirm the correctness of the data processing, it is sufficient to pay attention only to the analysis of 0 times and the analysis of 2 times (Reference: "Detection of Data F
low Anomaly Through Program Instrumentation ", University of Houston, JCHuang).

In this case, the output range in the case of including the repeating structure may be judged by the first detecting means or the selecting means, but it is performed by designating the execution statement ID in the middle so as to instruct the number of repetitions. You can also In either case, for example, @ P1012, @ S1012, and
As a result of being selected in the order of @ P1013, the result of FIG. 16 is obtained. Also, regarding the number of repetitions twice, @ P1012, @ S1012,
As a result of selection in the order of @ S1012, @ S1012, and @ P1013, the result of FIG. 17 is obtained.

Since it is sufficient to obtain information according to the presence / absence of repeated execution of the same place, the number of iterations is not limited to 0 and 2 but a desired number of iterations of 1 or less. A combination of times may be targeted. In other words, especially in the source code that consists of an iterative structure, the relevance of the data when it never executes in the iterative structure,
It is known that the data relevance becomes clearer by comparing the data relevance when executed multiple times.

(4) Third Embodiment Next, the analysis of the selection structure will be described. The components of the selection structure are roughly divided into a part that determines a selection condition (hereinafter referred to as a "selection condition determination part") and a part that is actually executed by selection (hereinafter referred to as a "selection target part"). In this case, there are as many selection target parts as there are options in the selection condition judging part.

In this case, the second detecting means for detecting the selection structure in the source code is provided, and with respect to the part relating to the detected selection structure in the selection means, each processing corresponding to each judgment result of the selection condition. It is configured to select information regarding (claims 11 and 24). If you do this,
With respect to the selection structure, information about each process corresponding to each determination result of the selection condition is output, so that the information about the selection structure can be easily understood.

Specifically, although there are a plurality of syntaxes for realizing the selection structure in the C language, they are common in that they include the above-mentioned two types of constituent elements. Here, the program including the if statement is taken as an example (FIG. 18), and the include file is the same as that in FIG.

In the analysis of this program including the selection structure, the analysis results for the selection condition judging section and the selection target section are prepared. In this case, the display range may be judged when the selection structure is included, or the execution sentence ID in the middle may be designated so as to instruct the selection target part to be analyzed. You may. In either case, for example, @ P1023, @ S10 when the condition is true
As a result of selecting 24 and @ P1027 in this order, the result of FIG. 19 is obtained. For false times, @ P1023, @ S1025,
As a result of selecting in the order of and @ P1027, the result of FIG. 20 is obtained.

(5) Fourth Embodiment If the given source code and the analysis result based on this source code are displayed in association with each other, it becomes easier to understand the correspondence relationship between them (claim 12, 25).

Specifically, the source code and the analysis result are
The images are simultaneously displayed on the same display device or an adjacent display device (claims 12 and 25). The user designates a desired sentence or range of the source code by using an input device such as a mouse or a keyboard. Each statement included in the source code should correspond to the generation of a specific data relation included in the final output result, except for redundant statements. Therefore, the analysis result of the portion corresponding to the designated range is highlighted by inversion or the like (claims 13 and 26).

On the contrary, a desired data relation among the analysis results may be designated. Each data relation must be generated by a specific sentence or set of sentences in the source code. Therefore, the sentence on the source code or the set of sentences corresponding to the specified data relation is highlighted by inversion or the like (claims 13 and 26).

FIG. 21 is a diagram showing a display example in the fourth embodiment. In this display example, the source code sentence 101 and the relation 102 represented by a dashed arrow correspond to each other, and the other one is highlighted regardless of which one is designated.

Thus, for example, when a problem is found in the analysis result, it is possible to easily specify which part of the source code causes the problem. On the contrary, in order to confirm that the implementation by the source code produces the intended result, it is possible to easily specify which part on the analysis result the mounting part affects.

(6) Other Embodiments The present invention is not limited to the above embodiments,
Since the embodiment can be freely changed, the embodiment includes other embodiments as illustrated below. For example, the programming language to be analyzed is not limited to C language,
Other languages, including various structured languages, can be targeted. Of course, even when an unstructured syntax such as a goto statement is included, the analysis result can be obtained by explicitly instructing the executable statement ID.

Further, the specific format of the data is not limited to that shown in each of the above embodiments, and can be freely determined. Further, the output target designation means, the first detection means, the second detection means, the output range designation means, the selection means, the analysis range designation means, the processing information extraction means, the related information extraction means, the reference means, the preprocessor and the like are necessarily provided. No need.

Any combination of control structures (for example, a nest formed by any combination of the repeating structure of the second embodiment and the selection structure of the third embodiment) can also be a target.

Although the present invention is generally realized by a computer program, it is also an aspect of the present invention to independently trade a recording medium such as a disk with a floppy disk in which this program is recorded. .

[0126]

As described above, according to the present invention,
The source code can be efficiently analyzed by static analysis, and the analysis result can facilitate understanding of the content of the source code including the processing content and the input / output specifications.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a program analysis device according to a first embodiment of the present invention.

FIG. 2 is a flowchart showing an operation procedure of the program analysis device according to the first embodiment of the present invention.

FIG. 3 is a diagram showing an example of a source code of a C language program according to the first embodiment of the present invention.

FIG. 4 is a diagram showing an example of an include file of a C language program according to the first embodiment of the present invention.

5 is a diagram showing an example of source code obtained as a result of adding the include file of FIG. 4 to the source code of FIG. 3 in the first embodiment of the present invention.

FIG. 6 is a diagram showing an example of a source code obtained as a result of labeling the source code of FIG. 3 in the first embodiment of the present invention.

FIG. 7 is a diagram showing a general storage format of each piece of information extracted in the first embodiment of the present invention.

FIG. 8 is a diagram showing an example of data information according to the first embodiment of the present invention.

FIG. 9 is a diagram showing an example of data information and related information according to the first embodiment of the present invention.

FIG. 10 is a diagram showing an example of related information as a result of modifying the related information of FIG. 9 in the first embodiment of the present invention.

FIG. 11 is a diagram showing an example of processing information according to the first embodiment of the present invention.

FIG. 12 is a diagram showing an example of output in the first embodiment of the present invention.

FIG. 13 is a diagram showing a legend of the output example of FIG. 12 in the first embodiment of the present invention.

FIG. 14 is a diagram showing another example of output in the first embodiment of the present invention.

FIG. 15 is a diagram showing an example of source code of a program including a repetitive structure according to the second embodiment of the present invention.

FIG. 16 is a diagram showing an example of an output related to the process of repeating the source code of FIG. 15 0 times in the second embodiment of the invention.

FIG. 17 is a diagram showing an example of an output related to a process of repeating the source code of FIG. 15 twice in the second embodiment of the present invention.

FIG. 18 is a diagram showing an example of source code of a program including a selection structure according to the third embodiment of the present invention.

FIG. 19 is a diagram showing an example of output related to processing when the condition of the source code of FIG. 18 is true in the third embodiment of the invention.

FIG. 20 is a diagram showing an example of output related to processing when the condition of the source code of FIG. 18 is false in the third embodiment of the present invention.

FIG. 21 is a view showing a display example according to the fourth embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Data extraction means 2 ... Output means 3 ... Labeling means 4 ... Preprocessor 5 ... Related extraction means 6 ... Process extraction means 7 ... Designation part 8 ... Selection means

Claims (26)

[Claims] 1. A program analysis device for analyzing a source code of a computer program representing a process for data, comprising: A related information extracting unit that extracts related information between data at a predetermined position of the source code based on the source code and the data information. 2. The program analysis apparatus according to claim 1, further comprising a reference unit that refers to another file related to the source code. 3. The data information includes a structure D that stores information for each data and a structure T that stores information for each data type, and the structure D corresponds to the corresponding structure T. The program analysis apparatus according to claim 1, wherein the program analysis apparatus is associated with each other. 4. The program analyzing apparatus according to claim 1, wherein the related information extracting means is configured to extract the type of the relationship. 5. A processing information extracting means for extracting, as processing information, related information relating to a predetermined processing in the source code, based on the source code, the data information and the related information. Item 1. The program analysis device according to item 1. 6. The processing information extracting means is configured to extract a set of the related information before the processing and the related information after the processing for processing to be performed based on the source code. The program analysis device according to claim 5, wherein 7. The processing information extracting means, when extracting the processing information, detects the consistency between the data obtained by tracing the related information, and the consistency detected by this means. 7. The program analysis apparatus according to claim 6, further comprising a correction unit that corrects the related information based on the. 8. The program analysis apparatus according to claim 1, further comprising an analysis range designating unit that designates the extraction range. 9. An output range designating means for designating a range of information to be output, and a selecting means for selecting information corresponding to the designated range from each of the extracted information. 1. The program analysis device according to 1 or 5. 10. A first detecting means for detecting a repetitive structure in the source code, wherein the selecting means processes a plurality of repetitions and one or less times for a portion related to the detected repetitive structure. 10. An arrangement adapted to select information regarding
A program analyzer as described. 11. A second detection unit for detecting a selection structure in the source code, wherein the selection unit corresponds to each judgment result of selection conditions for a portion related to the detected selection structure. The program analysis device according to claim 9, wherein the program analysis device is configured to select information regarding each process. 12. The output means for displaying the source code and any of the extracted information on the same display device or a display device adjacent to each other, according to claim 1 or 5. Program analysis device. 13. The output means is configured to, when a part of one of the source code or the information is designated, highlight another part corresponding to the designated part. 13. The program analysis device according to claim 12, which is characterized in that. 14. A program analysis method for analyzing a source code of a computer program representing a process for data, the data extraction step of extracting data information, which is information representing a structure of each of the data, from the source code, and the source. And a relation extracting step of extracting relation information between data at a predetermined position of the source code based on the code and the data information. 15. The program analysis method according to claim 14, further comprising a reference step of referring to another file related to the source code. 16. The data information includes a structure D that stores information for each data and a structure T that stores information for each data type, and the structure D corresponds to the corresponding structure T. 15. The program analysis method according to claim 14, wherein the program analysis method is associated with each other. 17. The program analysis method according to claim 14, wherein the association extraction step extracts the type of the association. 18. A processing information extracting step of extracting, as processing information, related information related to a predetermined processing in the source code based on the source code, the data information and the related information. Item 14
Described program analysis method. 19. The process information extracting step extracts a set of the related information before the process and the related information after the process for a process to be performed based on the source code. Item 19. The program analysis method according to item 18. 20. The processing information extracting step, when extracting the processing information, detects a consistency between data obtained by tracing the related information, and a consistency detected by this means. 19. The program analysis method according to claim 18, further comprising a correction step of correcting the related information based on the above. 21. An analysis range designating step of designating a range of the extraction is included.
Described program analysis method. 22. An output range designating step for designating a range of information to be output, and a selecting step for selecting information corresponding to the designated range from each of the extracted information. The program analysis method according to 14 or 18. 23. A first detecting step of detecting a repetitive structure in the source code, wherein the selecting step is a process of repeating a plurality of times and not more than once for a part related to the detected repetitive structure. 23. The program analysis method according to claim 22, characterized in that the information regarding is selected. 24. A second detection step of detecting a selection structure in the source code, wherein the selection step corresponds to each judgment result of selection conditions for a portion related to the detected selection structure. 23. The program analysis method according to claim 22, wherein information regarding each process is selected. 25. All or part of the source code,
19. The output step of displaying the information corresponding to all or a part of any of the extracted information on the same display device or a display device adjacent to each other, 20. Program analysis method. 26. The output step, when a part of one of the source code and the information is designated, highlights the other part corresponding to the designated part. 25. The program analysis method described in 25.