JP2006277127A - Method for comparing correction program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To acquire the correction information of a program as logical coordination on a design desired by a developer without performing the manual source code analysis of a program file. <P>SOLUTION: A comparison object program is provided to define items to be compared for each hierarchical structure of a module, file, function and source code as a logical structure on design for comparison. The comparison result is temporarily stored in a correction contents storing part as program correction information. A correction information output screen having the hierarchial structure is prepared and displayed from the program correction information of the correction contents storing part. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a means for comparing programs before and after correction, a means for visualizing a comparison result, and a system thereof in the field of program development.

  Knowing where to modify a program in program development, for example, knowing what part of a program is modified for a function change, makes it easier to understand the scope of the program's impact associated with a change request, and can be used as a reference when similar modifications occur. This is significant in that it enables safe and efficient work. For this reason, conventionally, several comparison means have been used in order to know the correction location before and after correction on the program.

  For example, it is possible to obtain correction information on the code by comparing the source code files of the program before and after the correction using a tool in units of lines. The “diff command”, which is a command of the UNIX (registered trademark) OS, has a function of comparing line units of text files and reporting a comparison result for each line, and is most often used for the comparison. It is one of the tools. However, for example, when the developer wants an overview of the modification status of the entire program, the line-by-line comparison results of the source code file itself are of low value to the developer and are meaningfully logically organized, for example, information on the functional requirement level. To convert, many of the developers themselves, mainly manual labor, are required, which is very inefficient.

  Japanese Patent Laid-Open No. 2001-256043 proposes a system that extracts a difference by dividing a source code file into units of functions and performing comparisons in units of functions. With this system, it is possible to reduce the labor for conversion into a logical unit by the manual work and to improve the reliability. However, logical groups that are meaningful to developers do not always correspond to functions. For example, there are many programs that have thousands of functions, and in this case, developers are larger than functions, for example, logical groups called design modules composed of files. You can easily imagine that you will feel the need to capture. The work of converting the program of the scale in the case into a meaningful group becomes inefficient as the manual work in the comparison result of the line unit.

  Also, many compilation programming languages provide a function called conditional compilation, and it is possible to adjust the validity or invalidity of a specific block in the source code by using this function. Under the development by the programming language, the developer can not only view the comparison results before and after correction, for example, as a function or module, but also narrow down the comparison information for each condition specified by conditional compilation. There is no conventional system that needs to be known but meets this requirement.

  In addition, when comparing programs before and after modification, a comparison process other than the comparison in the source code of the program file may be desired. However, the requirements can be integrated with other comparison process information and output in association with each other. Such a system is not in the conventional method. For example, in many programming languages, functions and variables in the source file are defined in an external file, and the external file name is specified in the source file. For example, an external file import specification called an include statement in the C language (hereinafter referred to as an include statement) It can be described together with an include line). If the external file specified in the include line has been modified, the source file for importing the external file will have the definition file as a comparison result even if there is no difference at the source code level before and after the program modification. Information that suggests the potential impact of a modification is important to the developer, but systems that provide that information are not in the traditional way.

JP 2001-256043 A

  Therefore, in the past, in order to obtain comparison information in a logical unit that is meaningful to the developer, the developer has to spend many kinds of work, and an efficient operation cannot be expected.

  The present invention has been made in view of such circumstances, and its purpose is to enable efficient grasping of the correction location of the program, and means for comparing between the programs before and after the correction and a visual comparison result. It is to provide an output means and a system thereof.

  In the present invention, a program to be compared is defined as a configuration having a logical hierarchical structure in design shown in FIG. 1, and comparison information is managed by the structure. In addition, the comparison of the programs before and after the modification is based on the comparison between the same layers of the hierarchy. Also, parameters for comparison of each hierarchy are defined by physical elements on the program shown in FIG. The parameter information is obtained from the program file before and after the correction.

  In order to achieve the above object, the present invention visually compares the program file before and after the modification in the unit of hierarchy and the difference information obtained by the comparison in the unit of hierarchy as the modification information of the program. And a step.

  According to the present invention, it is possible to obtain program correction information as a logical unit of design desired by a developer without manually analyzing a source code of a program file.

  Furthermore, according to the present invention, the correction information can be obtained for each level of the hierarchy according to the level of the logical structure of the program desired by the developer.

  Furthermore, according to the present invention, not only the comparison of the program file on the source code but also the modification of the file specified in the include line can be obtained as the modification information of the source file.

  Further, according to the present invention, the correction information is visualized as hierarchy-specific correction information, and is identified and visualized for each conditional compilation, so that it can be considered at a logic level desired by the developer, and the correction location on the program can be grasped. Can be made more efficient and deeper.

  Embodiments of the present invention will be described below. FIG. 3 shows the configuration of a program correction information output system according to an embodiment of the present invention.

  Here, 3-1 and 3-2 are a source file group of a program before modification, a source file group of a program after modification, 3-3 is a comparison of source files of programs before and after modification, and the comparison result is 3-4. A comparison processing unit 3-5 registered in the correction content holding unit 3-5 is a correction content display unit that shapes the correction information registered in the correction content holding unit and displays the comparison result.

  The program correction information output system according to this embodiment includes hardware such as a processing device, a storage device, an input device, an output device, and a storage medium reading device that constitute a computer. The processing device includes at least a central processing unit (CPU), a RAM that provides a work area for the central processing unit, a ROM that stores various programs and data, and an interface unit with peripheral devices. The storage device is a readable / writable storage device such as a floppy (registered trademark) disk drive, hard disk drive, magneto-optical disk drive, memory card, smart media (registered trademark), or the like. The input device is, for example, a mouse, a keyboard, a voice input device, an image input device, or the like. The output device is a display device such as a CRT or LCD.

  Next, the operation of the program correction information output system will be described. It is assumed that the source files of the program before and after the correction are collectively arranged in the storage device under different parent directories on the same file system.

  First, the setting conditions for the comparison process shown in the example of FIG. 4 are specified. The setting conditions are the directory path 4-1 where the source files of the program before and after correction are arranged, the module configuration 4-2, the comment pattern 4-3, the function name pattern 4-4, the include line pattern 4-5, and conditional compilation. Includes row patterns 4-6. The module configuration specifies a combination of a file path name character string pattern and a module name. The module name can be arbitrarily specified by the developer. The comment pattern, function name pattern, include line pattern, and conditional compilation line pattern specify the character string pattern on each source code according to the specifications of the programming language.

  In the example of FIG. 4, the pattern specification of each setting condition is specified using a regular expression provided by the script language Perl, and the source file of the program to be compared is in the C language. Is assumed.

  Next, the comparison processing unit 3-3 is executed. The program comparison process is performed as follows. A scan for extracting the following information is performed for each program before and after correction. First, a list of file paths to be configured is acquired (FIG. 5). Each file path is matched with the module configuration pattern of the setting conditions given in advance, and the module information to which it belongs is added. The belonging module information of a file that does not match any module configuration pattern is unknown. The information is registered in the correction content holding unit 3-4.

  In the example of FIG. 5, it can be read that the file “dirA \ dir1 \ file2.c” belonging to the module “moduleA” does not exist in the corrected program and has been deleted as a result of the correction work.

  Next, a code block specified by conditional compilation, a code block defining a function, and an include line are extracted from each file and acquired as an information table shown in FIG. In the function information table of FIG. 6, the file name in which the function is defined, the start and end lines of the definition part, and the conditional compilation information for enabling the function are managed. In the include information table of FIG. 6, the file name and line number in which the include line is specified, and the conditional compilation information for enabling the include line are managed. The information is registered in the correction content holding unit 3-4.

  Next, the program information before and after correction obtained by the scanning is compared with each parameter of the logical hierarchical structure defined in FIG. Based on the file path information before and after correction obtained in FIG. 5, the file structure for each module is obtained as module-specific file structure information. Further, the module-specific file structure information is compared before and after the modification, and if the file path with the same name exists before and after the modification, the presence / absence of the modification by comparison on the source code of the file is determined, and then the addition / modification / The number of files without deletion / correction is obtained and registered in the correction content holding unit as module layer comparison information 7-2 (FIG. 7-2). This information corresponds to the module layer information in FIG.

  Further, the number of modules without addition / modification / deletion / modification is obtained from the module layer comparison information. Judgment whether or not a module is modified is made based on the module layer comparison information according to whether or not a file constituting the module is modified. The information is recorded in the correction content holding unit as program layer comparison information 7-1 (FIG. 7-1). This information corresponds to the information of the program layer in FIG.

  Further, the number of functions added / modified / deleted / unmodified is obtained for each file before and after modification from the file path list (FIG. 5) and the function information table (FIG. 6).

  An additional decision is made when the target function exists after modification in the same name file before and after modification. However, at this time, it is searched whether the function exists in another file before correction, and if it exists, it is added (moved) in distinction from addition.

  Deletion is determined when the function exists in the same name file before and after the correction and exists before the correction and does not exist after the correction. However, at this time, it is searched whether it exists in another file after correction, and if it exists, it is determined to be deleted (moved) in distinction from deletion.

  To determine whether or not to modify a function, extract the code block of the function definition part from the start and end lines in the function information table before and after the modification, delete the comment block based on the comment pattern in the setting condition, A comparison is made to determine whether correction has been made. Code comparison is done at the line level to get the total number of lines and the number of code lines without addition / modification / deletion / no modification.

  The number of functions added / modified / deleted / unmodified is registered in the modified content holding unit as file layer comparison information 7-3 (FIG. 7-3). The total number of lines in the function and the number of code lines without addition / modification / deletion / no modification are registered in the modification content holding unit as function layer comparison information 7-4 (FIG. 7-4). Further, the number of each included line added / modified / deleted / without modification is obtained for each file before and after modification from the file path list (FIG. 5) and the include line information table (FIG. 6), and file layer comparison information 7-3 is obtained. Register as

  An additional decision is made when the same include line exists in the file with the same name before and after the modification but before the modification and after the modification.

  Deletion is determined when the same include line exists before and after modification in the same-named file before and after modification. If the same line exists in the same-named file before and after modification, the module layer comparison information is searched for whether the file specified in the included line is modified before and after modification. If there is a modification, the include line itself is not modified. It is distinguished from the above, and there is no correction (reference destination is changed).

  After completion of the comparison process, the program correction information output system analyzes the information registered in the correction information holding unit and visualizes the correction information by the correction content output unit 3-6. FIG. 8 shows an example of visualization of the analysis result. The analysis result is composed of a plurality of screens. Top screen 8-1, program layer screen 8-2, module layer screen 8-3, file layer screen 8-4, function layer screen 8-5, code difference information screen 8- 6 is included.

  First, the top screen 8-1 is displayed on the correction content output section. Here, the result of the correction information for each logical hierarchy is displayed together with information with or without correction. The presence / absence of correction of each level is determined to be corrected when there is at least one value of information other than the information without correction in the comparison information by level 7-1 to 7-4 registered by the comparison processing unit. Is not amended.

  Developers can first get an overview of the program differences. In addition, a conditional compilation designation field is prepared on the screen, and by inputting desired conditional compilation information here, it is possible to narrow down functions and include lines to be compared. In this case, OFF (not specified) is selected as the conditional compilation specification. The conditional compilation specification is valid for all the information on the remaining screens.

  In addition, the part displayed by enclosing the character string with “<<” and “>>” is also an area (hereinafter referred to as a link part) indicating an instruction for transition to another screen, and pointing with a mouse or the like Moves to another screen in response to an instruction from the device (hereinafter referred to as a click operation). The meanings and functions of “<<” and “>>” are common to all screens. Also, except for the top screen, although not shown in each screen, a link part for returning to the screen before the transition is prepared, and each screen can be freely transitioned according to the request of the developer.

  Next, a transition to the program layer screen 8-2 is made by a click operation of “<< corrected >>” on the top screen. On this screen, the number of addition / modification / deletion of modules, which are comparison parameters of the program layer, and the corresponding module names are displayed. Each module name is a link part for transitioning to the module layer screen 8-3. Although all the displayed information is not shown in 8-2, it follows the program layer comparison information 7-1.

  Here, the modification status can be grasped in units of modules recognized by the developer.

  Next, the module name that is the link part of the program layer screen is clicked, and the screen transits to the module layer screen 8-3. In this screen, correction information is output in units of files for the files under the module specified on the program layer screen which is the previous screen. The displayed information follows the module layer comparison information 7-2. As for the file information without correction, when there is even one include line with no correction (with reference destination change) by searching the file layer comparison information of the corresponding file, the information is added as a matter to be noted. . In the example of FIG. 8, a thick-line frame shaded cell in the information table corresponds to the fact that there is a file including an include line to be noted in two unmodified files.

  Next, the file name which is the link part of the module layer screen 8-3 is clicked, and it changes to the file layer screen 8-4. In this screen, the function under the file specified in the module layer screen which is the previous screen and the correction information in units of include lines are output. The displayed information follows the file layer comparison information 7-3.

  Next, the function name which is the link part of the file layer screen 8-4 is clicked, and it changes to the function layer screen 8-5. In this screen, the comparison information of the function designated on the file layer screen which is the previous screen is displayed. The displayed information follows the function layer comparison information 7-4.

  In addition, a link for displaying the source code difference entity is prepared at the bottom of the screen, and when this is clicked, the difference entity screen 8-6 (FIG. 10) is displayed and the contents are displayed. In this case, the difference entity content by the “diff command” is shown.

  In this visualization example, following the top screen, the approach from the upper layer to the lower layer of the logical structure is approached in order to make it easier to understand the correction information of the entire program. It is also easy to give operation specifications to perform. For example, when a click operation of “<< Modified >>” in the file layer on the top screen 8-1 is performed, it is possible to transition to the screen illustrated in FIG. In the screen of FIG. 9, the comparison information regarding the function and the include line under the previous file under the program is displayed. In general, a considerably large number is expected to be displayed, but it is possible to provide an information narrowing function at the top of the screen as in the example of FIG. It is possible to narrow down the information by specifying the file name pattern for which you want to check the correction information, select the function information or include information, and avoid the information confusion of the developer by reducing obvious unnecessary information . Although not shown, the same can be applied to the module layer and the function layer.

  As described above, according to the program correction information output system of the present embodiment, comparison information of the programs before and after the correction as shown in FIG. 8 and FIG. 9 can be provided, so that the correction part of the program can be easily grasped. Become. That is, in the modification content display section, program modification locations are provided in units of modules, files, functions, include lines, and source codes, and the modification information is known in a format centered on addition / modification / deletion for each unit. be able to.

  In addition, in the modification information of the include line, not only the differences in the source code but also the presence or absence of changes in the file specified in the include line can be picked up, and the range of influence on the program due to the modification can be suggested more precisely to the developer. it can.

  Furthermore, it is possible to confirm the influence range of the correction by conditional compilation, and in addition to suggesting the influence range by the include line information, it is possible to consider the correction situation deeper.

It is a figure which shows the management hierarchical structure of comparison information. It is a figure which shows the comparison parameter according to management hierarchy. It is a figure which shows the structure of the program correction information output system which is one Embodiment of this invention. It is a figure which shows the interface which inputs the conditions which a developer sets in the system of FIG. It is a figure which shows the file structure information of the program before and behind correction in the system of FIG. It is a figure which shows the function information and include line information before correction in the system of FIG. It is a figure which shows the program layer comparison information in the system of FIG. It is a figure which shows the module layer comparison information in the system of FIG. It is a figure which shows the file layer comparison information in the system of FIG. It is a figure which shows the functional layer comparison information in the system of FIG. It is a figure which shows the interface of the correction information output part in the system of FIG. It is a figure which shows the interface of the correction information output part in the system of FIG. It is a figure which shows the code difference information screen 8-6 in FIG.

3-1: Source file group of pre-correction program 3-2: Source file group of post-correction program 3-3: Comparison processing unit 3-4: Correction content holding unit 3-5: Correction content display unit

Claims (3)

In a method for comparing programs,
Comparing the program before and after modification in each layer of module, file, function, source code,
Visually outputting the difference information obtained by the comparison of the hierarchical units as the correction information of the program;
A method for comparing correction programs, comprising:   In the comparison method, even if there is no change in the source code between the files to be compared, if the file specified by the external file import specification in the file is changed, the change information is sent to the source file. The correction program comparison method according to claim 1, further comprising a step of obtaining as correction information.   In the comparison method, when the program is implemented based on the specification of the conditional compilation function and the function definition part or the external file import specification part in the source code is under the influence of the conditional compilation, the function or external for the conditional compilation information 3. The method of comparing correction programs according to claim 1, further comprising a step of obtaining a correction information that is valid based on the designation of the conditional compilation information while maintaining a correspondence relationship with the file import designation.

Images