JP4976328B2 - Author determination apparatus, control method thereof, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To specify the writer of data, based on writers and document data themselves. <P>SOLUTION: A writer ID imparting part 103 imparts wrier identification data by each kind of document data by prescribed unit to be described in the data. A writer ID rate calculating part 104 calculates the ratio of the identification information of the writers imparted to the data. A file writer determining part 105 determines the writer of the data, based on the ratio of calculated rate of the identification information. <P>COPYRIGHT: (C)2010,JPO&amp;INPIT

Description

  The present invention relates to a technique for determining an author of data in which document data is described.

  In recent years, a large amount of source code has been described by a large number of people in the development of computer systems. When large-scale development is performed such that a large amount of source code is written, it is common practice to accept a programmer from a dispatch company or the like.

  However, for dispatch programmers, although they ordered the skills required to request dispatch from a dispatch company, they did not always have the skills they wanted. Attempts have been made to evaluate programmers in the sense of managing the quality of programs, but in order to evaluate the authors of document data such as programs, it is necessary to first identify the author of the document data to be evaluated. .

  For example, there is no problem if the author of the document data in one file is alone, but if the document in one file is described by many people, such as source code, the author of the file is specified. There was a problem that it might be difficult to do.

  For example, when the same file is edited by multiple members, if the last person who wrote the document data (final committer) is the author, the actual contribution of the member, that is, the person who wrote the largest number may not be the author. is there.

  When a file is moved for organizing a repository folder, the worker who moved the file becomes the author even though the file is not described.

  Although it is possible to record the author for each line of source code, and evaluate each author by collecting the evaluations, it is actually difficult to evaluate the source code line by line.

  The following technique is disclosed as a change history when the source code is changed. Patent Document 1 discloses that a revision history of a source program is automatically created, and an editing operation, a correction location, correction content, a correction date and time, etc. are recorded.

JP-A-9-128227

  However, in the invention disclosed in Patent Document 1, although the contents of correction are recorded, no consideration is given to the author such as who wrote it and who modified it.

  Therefore, there is a need for a method for identifying the author for each file from the author recorded in the repository and the document data itself.

  An object of the present invention is to specify the author of data from the author and the document data itself.

The copyright determination apparatus according to the present invention includes a granting unit that grants author identification information to each predetermined unit of document data described in the data, and a calculation unit that calculates a ratio of the author identification information given to the data. And determining means for determining the author of the data based on the ratio of the identification information calculated by the calculating means.
The author determination apparatus control method of the present invention is an author determination apparatus control method comprising an adding means, a calculating means, and a determining means , wherein the assigning means is a document of a predetermined unit described in data. An assigning step of assigning author identification information for each data, a calculating step in which the calculating means calculates a ratio of author identifying information assigned to the data, and the determining means are calculated in the calculating step. And a determination step of determining an author of the data based on a ratio of identification information.
Program of the present invention, calculating computer, which calculates the assigning means for assigning identification information of the author for each predetermined unit of document data described in the data, the percentage of identity of the author given to the data and means, based on the percentage of the identification information calculated by the calculating means, characterized in that to function as a determining means for determining the author of the data.

  In the present invention, the identification information of the author is assigned to each predetermined unit of document data described in the data, and the author of the data is determined based on the ratio of the identification information of the given author. is doing. Therefore, according to the present invention, it is possible to specify the author of the data from the author and the document data itself.

  DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments to which the invention is applied will be described in detail with reference to the accompanying drawings.

  FIG. 1 is a diagram showing a functional configuration of a file author determination system according to an embodiment of the present invention. As shown in FIG. 1, the file author determination system includes a file author determination device 100 and a file management device (repository) 300, which are connected to each other via a communication line such as a LAN.

  The file author determination apparatus 100 includes a parent file determination unit 101, an update part determination unit 102, an author ID assignment unit 103, an author ID ratio calculation unit 104, a file author determination unit 105, and an evaluation unit 106 as functional configurations.

  When there is a file update or new registration, the parent file determination unit 101 determines the parent file from which the file is generated.

  The update part determination unit 102 determines a difference (update part) between the source code described in the parent file and the source code described in a file (child file) generated based on the parent file.

  When it is determined that there is a difference between the parent file and the child file, the author ID assigning unit 103 uses the source code of the portion different from the parent file in the child file based on the management content in the file management apparatus 300 described later. Give the author's ID. On the other hand, the same ID as the author ID assigned to the source code of the corresponding part of the parent file is assigned to the source code of the same part as the parent file in the child file. For child files that do not have a parent file, the author's ID is assigned to all source codes based on the management contents in the file management apparatus 300.

  The author ID ratio calculation unit 104 calculates the ratio of the assigned author ID for each file. The file author determination unit 105 determines the author with the highest ID among the ratios calculated by the author ID ratio calculation unit 104 as the author of the corresponding file. Also, the importance is determined for each part of the child file that is different from the parent file, the weight corresponding to the determined importance is given to the author ID, and the weighting value assigned to the author ID is set. The author may be determined by counting. For example, a part that does not affect the execution of the processing described in the source code, such as a blank line for formatting or a comment, may be excluded or weighted.

  The evaluation unit 106 evaluates each file according to a predetermined evaluation criterion, and sets the evaluation value as the evaluation value of the author of each file determined by the file author determination unit 105.

  The file management apparatus 300 stores a plurality of files 302 in which source code (document data) is described by the author, and the date and time when these files were newly (added) registered or updated and the author who performed the file. A management table 301 for managing IDs (author IDs) is stored.

  FIG. 2 is a block diagram illustrating a hardware configuration of the file author determination apparatus 100. The CPU 201 comprehensively controls each device and controller connected to the system bus. The ROM 203 or the HD 207 stores a BIOS (Basic Input / Output System) that is a control program of the CPU 201, an operating system program, a program for the processing shown in FIGS. 3 and 4 executed by the file author determination apparatus 100, and the like. Yes.

  In the example of FIG. 2, the hard disk (HD) 207 is configured to be disposed inside the file author determining apparatus 100, but a configuration corresponding to the HD 207 is disposed outside the file author determining apparatus 100 as another embodiment. A configuration may be adopted. For example, the program for performing the processing shown in FIGS. 3 and 4 according to the present embodiment is recorded on a computer-readable recording medium such as a flexible disk (FD) 206 or a CD-ROM, and is supplied from the recording medium. It is good also as a structure which may be supplied via communication media, such as the internet.

  The RAM 202 functions as a main memory, work area, and the like for the CPU 201. The CPU 201 implements various operations by loading a program necessary for execution of processing into the RAM 202 and executing the program.

  The disk controller 205 controls access to external memories such as the HD 207 and the FD 206. The communication IF controller 204 is connected to the Internet or a LAN, and controls communication with the outside by, for example, TCP / IP.

  The display controller 208 controls image display on the display 209.

  The KB (keyboard) controller 210 receives an operation input from the keyboard (KB) 211 and transmits it to the CPU 201. Although not shown, in addition to the keyboard 211, a pointing device such as a mouse can be applied to the file author determination apparatus 100 according to the present embodiment as a user operation unit.

  Further, the parent file determination unit 101, the update part determination unit 102, the author ID assigning unit 103, the author ID ratio calculation unit 104, the file author determination unit 105, and the evaluation unit 106 in FIG. Accordingly, the program is loaded into the RAM 202, and the CPU 201 executes the program.

  In the present embodiment, as shown in FIG. 1, the file management device 300 is installed outside the file author determination device 100, but the file management device 300 is arranged inside the file author determination device 100. May be. In this case, the management table 301 and the file group 302 are stored in the HD 207.

  FIG. 3 is a flowchart showing processing for determining a parent file that is a source for generating a target file, which is executed by the file author determination apparatus 100. This process is a process that is started when the target file is updated or registered in the file management apparatus 300 and a notification to that effect is received from the file management apparatus 300.

  First, when the parent file determination unit 101 receives the notification from the file management apparatus 300, the parent file determination unit 101 detects that the target file indicated in the notification has been updated or registered, and the target file is transferred from the file management apparatus 300. Obtain (step S301).

  Subsequently, the parent file determination unit 101 determines a file having a file size close to the target file, a file with a similar file name, or the like as a candidate for the parent file, and acquires the file from the file management apparatus 300 (step S302). In this embodiment, the parent file candidates are selected for all files managed by the file management apparatus 300. However, as another embodiment, the update date and time from the update date and time of the target file to the past for a certain period. Parent file candidates may be selected only for files with the same update date and time as the target file, or may be selected only for files with the same extension as the target file. May be selected. In this way, by limiting the selection destinations of the parent file candidates, the processing load of the parent file determination unit 101 can be reduced.

  Subsequently, the parent file determination unit 101 decomposes each candidate for the parent file into tokens and also decomposes the target file into tokens (step S303). The token here is a minimum comparison unit of the source code in the file, and in this embodiment, one line is a token.

  Subsequently, the parent file determination unit 101 compares the target file with each parent file candidate for each corresponding token, and determines whether or not they match each other (step S304). That is, here, it is determined for each line whether the candidate of the target file and each parent file matches the source code.

  Subsequently, the parent file determination unit 101 determines, as a parent file, a parent file candidate that has the largest number of matching tokens in the parent file candidate group and the number of matching tokens is equal to or greater than a predetermined threshold (step S305). . The association between the target file determined here and the parent file (parent-child relationship) is managed inside the file author determination apparatus 100 (for example, HD 207).

  FIG. 4 is a flowchart showing processing for determining the author of the target file, which is executed by the file author determining apparatus 100. This process is a process that starts when the target file is updated or registered in the file management apparatus 300 and receives a notification from the file management apparatus 300. The target determined by the process shown in FIG. Since the association between the file and the parent file is also used, strictly speaking, this process is started after the process shown in FIG.

  First, upon receiving a notification from the file management apparatus 300, the update part determination unit 102 detects that the target file indicated in the notification has been registered for update or new registration, and sets the target file and its history file as a file. Obtained from the management apparatus 300 (step S401). The history file here refers to all the files used directly or indirectly in generating the target file, such as the target file and its parent file, and the parent file of the parent file. I mean. The history file is acquired based on the association indicating the parent-child relationship between files managed in the file author determination apparatus 100.

  Subsequently, the update portion determination unit 102 determines whether or not the history file of the target file has been acquired (step S402). If the association indicating the parent-child relationship does not exist, the history file of the target file cannot be acquired. On the other hand, if the association indicating the parent-child relationship exists, the history file can be acquired. Therefore, the process of step S402 is synonymous with the process of determining whether or not there is an association indicating the parent-child relationship.

  When the history file can be acquired, the update part determination unit 102 selects the oldest file among the files that have not been selected in the history file (step S403). This process can be realized by referring to the file update / addition date and time in the header information of each history file.

  On the other hand, when the history file cannot be acquired, the author ID assigning unit 103 decomposes the target file into tokens because the file acquired in step S401 is only the target file (step S409). An author ID is assigned to the token (step S410). Here, the author ID assigning unit 103 assigns an author ID to each token of the target file by referring to the management table 301. In the management table 301, only one author ID is managed for the target file. The single author ID is assigned to all tokens. That is, the target file for which no history file exists is a file created without using any file, and therefore, only one author ID is assigned.

  Subsequently, the update part determination unit 102 determines whether or not the file selected in step S403 is the oldest file in the history file (step S404).

  When the selected file is the oldest file in the history file, the author ID assigning unit 103 similarly performs the above-described steps S409 and S410 on the file. This is also because the history file does not exist in the file, and the file is created without any file.

  On the other hand, when the selected file is not the oldest file in the history file, that is, when the parent file exists in the file, the update part determination unit 102 breaks down the file together with the parent file into tokens ( Step S405).

  Subsequently, the update part determination unit 102 compares the parent file and the file for each token, and determines whether or not they match (step S406).

  The author ID assigning unit 103 assigns an author ID to the token of the file that does not match the parent file with reference to the management table 301 (step S407). Here, the author ID assigning unit 103 refers to the record in the management table 301 corresponding to the file name of the file and the update / addition date / time of the header information, and the author ID in the record does not match in step S407. Granted to the determined token. Note that the same author ID as the corresponding token of the parent file is assigned to the token determined to match.

  Subsequently, the author ID assigning unit 103 determines whether the file selected in step S403 is a target file (step S408). That is, in step S408, it is determined whether or not the author ID is assigned to the target file that is the last file among the files to which the author ID is assigned.

  If the selected file is not the target file, the process returns to step S403. That is, in the process of FIG. 4, a process of assigning an author ID to all the files from the oldest file having no history file to the target file is performed.

  Subsequently, the author ID ratio calculation unit 104 obtains the ratio of the author ID assigned to each token in the target file (step S411).

  Subsequently, the file author determination unit 105 determines the author with the highest author ID among the author ID ratios obtained in step S411 as the author of the target file (step S412). In the above-described processing, all the history files of the target file are acquired, the author ID is sequentially assigned from the oldest file, and the author ID of the target file is finally given. As another embodiment, all author IDs assigned to related files up to the previous file are registered on the file author determining apparatus 100 side, and when the author of the target file is determined, the file author is determined. The apparatus 100 acquires only the parent file immediately before the target file from the file management apparatus 300, compares the parent file and the target file for each token, and registers the author ID registered for the common token. The author ID may be assigned to different tokens based on the management table 301. By doing in this way, it becomes possible to reduce the calculation load of author ID assignment processing and author determination processing.

  FIG. 5 is a diagram for specifically explaining a source code evaluation method for a file whose author has been determined as described above.

  When evaluating the source code for a file, the readability index of the file is used. FIG. 5 shows a file with 21 lines, 6 branches, and 4 levels of nesting, so in the case of the file shown in FIG. Is calculated as follows.

  When the number of lines, the number of branches, and the depth of nesting increase, the readability of the source code described in the file decreases. In the present embodiment, the readability index of the file is calculated for each file in consideration of these indexes. In the above formula, the higher the number of lines, the number of branches, and the depth of nesting, the higher the readability index. Therefore, the lower the readability index of the file, the higher the readability and the higher the evaluation.

  In the present embodiment, the author of each file can be evaluated by obtaining the readability index of the file for each file in this way. In the present embodiment, the number of lines, the number of branches, and the nesting are all used as evaluation indexes, but any one or two of them may be used as evaluation indexes. Further, a code that is likely to become a bug may be detected by a tool or the like, and the number of detected codes may be used as an evaluation index.

  FIG. 6 is a diagram for explaining a design evaluation method for a folder including at least one file. The folder design evaluation is a process executed by an evaluation unit (not shown).

  FIG. 6 shows folders A to E. In particular, the folder A includes files A to D, but the folders B to E also include some files like the folder A. Yes. In addition, there is a folder calling relationship indicated by arrows between folders. That is, the folder A has an external reference relationship with the folder E and an external reference relationship with the folder B. There is a circular reference relationship between folders B, C, and D.

  Here, a problem in folder design caused by a reference relationship between folders will be described. First, the description will be made based on the file A. The external reference relationship (here, the reference relationship from the folder E) and the external reference relationship (here, the reference relationship to the folder B) are extremely eliminated. Then, an increase in the number of files in folder A (folder enlargement) and files in folder A cause duplication of source code with files in other folders (problem 1).

  On the other hand, if the reference relationship from the folder A to the outside (here, the reference relationship to the folder B) is extremely large, the folder A is easily affected by the change in the folder B (problem 2). Also, if there are extremely many external reference relationships (here, reference relationships from folder E), folder A is difficult to change (problem 3).

  Next, focusing on the circular reference relationship between the folders B, C, and D, there is a problem that the reusability of each folder is lowered when there is a circular reference relationship between the folders (problem 4).

  The following evaluation criteria are mentioned from the above problems 1-4. That is, the readability index of the folder considering the problems 1 to 3 and the number of folders in the circular reference relationship considering the problem 4. The folder readability index is a parameter obtained by integrating the sum of the number of folders referenced from outside and the number of folders referenced externally and the number of files in the folder (in the above example, folder A). It is an index to be obtained. This index takes a high value when the number of folders referred from the outside and the number of folders referred to the outside are too few or too many. It also takes a high value when the number of files in the folder is large. Therefore, the lower the readability index of the folder, the higher the evaluation value. Similarly, when the number of folders having a circular reference relationship is large, the reusability of the folder is lowered, so that a lower value is treated as a higher evaluation value. In the case of folder A, the number of folders having a circular reference relationship is zero.

  As described above, by calculating the readability index of the folder and the number of folders having a circular reference relationship for each folder, it is possible to perform evaluation on the design of each folder. In this embodiment, the author of each folder Is the ultimate goal. Therefore, it is necessary to determine the author of each folder. The method will be described below.

  As a first method, an author is associated with each token (here, one line) of each file in the folder by the processing up to step S407 in FIGS. Next, the authors associated with the tokens of each file are tabulated for each folder, and the author with the largest number is determined as the author of the folder.

  As a second method, by the method described with reference to FIGS. 3 and 4, an author is associated with each token (here, one line) of each file in the folder, and the author of each file is determined. Next, the authors determined for each file are tabulated for each folder, and the author with the largest number is determined as the author of the folder.

  An author of a folder is determined by the first or second method described above, and an evaluation value in designing the corresponding folder is obtained as an evaluation value of the author.

  FIG. 7 is a table in which the evaluation values of the respective authors of the files and folders obtained as described above are summarized in a table format. The evaluation table shown in FIG. 7 is displayed on the display 209 as the final output value.

  In the example shown in FIG. 7, member 5 (menber_c_5), member 7 (menber_c_7), member 7 (menber_c_7), member 9 (menber_c_9), member 10 (menber_c_10), member 11 (menber_c_11), member 12 (menber_c_12), For each member 17 (menber_c_17), member 18 (menber_c_18), member 20 (menber_c_20), and member 27 (menber_c_27), the period, the total number of lines (A), the number of copied lines (B), (A)-(B) , File readability index> 300, number of circulating folders and folder readability index are shown.

  The period is a period in which the update / addition date / time of the file and folder to be evaluated is limited, and the file / folder having the update / addition date / time corresponding to the period shown here is the object of evaluation. The update / addition date / time of the folder is indicated by the attribute information of the folder.

  The total number of lines (A) is the total number of lines in each file corresponding to the above period. The number of copied lines (B) is the number of lines updated by copying out of the total number of lines (A) in the file. The copy source file can be detected by a conventional clone detection algorithm, and the number of lines updated by copying can be obtained by comparing the detected copy source file with the target file.

  (A)-(B) is a value obtained by subtracting the copy line number (B) from the total file line number (A). In the present embodiment, the number of lines indicated by this value is the calculation target of the file readability index.

  The file readability index is the readability index of the corresponding author's file. The number of circular folders is the number of folders that have a circular reference relationship with the corresponding author's folder. The folder readability index is the readability index of the corresponding author's folder.

  The three items on the right of FIG. 7 (file readability index, number of circulating folders, folder readability index) are the evaluation values of each author. The lower the value of each item, the more relevant the author is writing highly rated source code and folder design.

  As described above, in this embodiment, an author ID is assigned to each predetermined unit of document data (source code) described in the file, and the file ID is determined based on the ratio of the author ID assigned. The author is determined, and the author is configured to evaluate the determined file. Therefore, according to the present embodiment, it is possible to specify the author of a file from the author and the document data itself, and to evaluate the document data described in the file as the author's evaluation. In addition to determining the author for each file, the author may be determined for a plurality of files. Further, the file may include image data in addition to the document data, and the data that is the determination unit of the author is not particularly limited.

  In this embodiment, since the author of a folder is determined and the design of the folder for which the author has been determined is evaluated, the design of the folder is evaluated as the evaluation of the author. Is possible.

It is a figure which shows the functional structure of the file author determination system which concerns on embodiment of this invention. It is a block diagram which shows the hardware constitutions of a file author determination apparatus. It is a flowchart which shows the process which determines the parent file used as the origin which produces | generates the object file performed by the file author determination apparatus. It is a flowchart which shows the process which determines the author of an object file performed by the file author determination apparatus. It is a figure for demonstrating concretely the evaluation method of the source code with respect to the file for which the author was determined. It is a figure for demonstrating the evaluation method on the design of the folder containing at least 1 file. It is the figure which put together the evaluation value of each author of a file and a folder in a table format.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100: File author determination apparatus 101: Parent file determination part 102: Update part determination part 103: Author ID provision part 104: Author ID ratio calculation part 105: File author determination part 106: Evaluation part 300: File management apparatus 301: Management table 302: File group

Claims (9)

A granting unit for granting author identification information for each predetermined unit of document data described in the data;
A calculating means for calculating a ratio of author identification information given to the data;
An author determination apparatus comprising: a determination unit that determines an author of the data based on a ratio of identification information calculated by the calculation unit.   2. The author determination apparatus according to claim 1, further comprising an evaluation unit that evaluates document data described in the data for which the author is determined by the determination unit.   The evaluation means evaluates the data based on at least one of the number of lines, the number of branches, and the number of nestings of source code that is document data described in the data. The author determination apparatus according to claim 2.   The calculation means calculates a ratio of author identification information given by the grant means for each data included in the folder in units of folders, and the determination means calculates a ratio of identification information calculated by the calculation means. The author determination apparatus according to any one of claims 1 to 3, wherein an author of the folder is determined based on the information.   The calculation means calculates a ratio of author identification information given by the grant means for each data included in the folder in units of folders, and the determination means calculates a ratio of identification information calculated by the calculation means. 4. The author of each of the data is determined on the basis of the data, and the author who has the largest number among the authors determined for each of the data is determined as the author of the folder. Author determination device described in.   6. The author determination apparatus according to claim 4, further comprising second evaluation means for performing design evaluation of the folder whose author is determined by the determination means.   The second evaluation means includes a reference relationship from the other folder to the folder, a reference relationship from the folder to another folder, a circular reference relationship between the folder and another folder, and the number of data in the folder. 7. The author determination apparatus according to claim 6, wherein a design evaluation of the folder is performed based on at least one of them. A method for controlling an author determination apparatus having an assigning means, a calculating means, and a determining means ,
It said applying means, and applying steps to impart identification information of author each predetermined unit of document data described in the data,
The calculating means, and calculating a percentage of identity of the author given to the data,
The author determination apparatus control method , wherein the determination means includes a determination step of determining an author of the data based on the ratio of the identification information calculated in the calculation step. Computer
A granting unit for granting author identification information for each predetermined unit of document data described in the data;
A calculating means for calculating a ratio of author identification information given to the data;
Based on the ratio of the identification information calculated by said calculation means, a program to function as a determining means for determining the author of the data.

Images