JP2003157263A - Method and device for collecting information, and storage program for retrieval object character information - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize an information collection/retrieval which never requires the addition and correction of a function even if a new file form appears or a conventional file form is changed. SOLUTION: An API (Application Programming Interface) calling made when an application program for preparing the file of a file form displays the contents of the file on a screen is detected, a display character string is extracted from here and this is accumulated to make it the object of retrieval.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to information retrieval for retrieving necessary information from information stored in a storage device of a computer and information available from a computer connected to a network.

[0002]

2. Description of the Related Art In the conventional information collecting and searching method, a large number of electronic files stored in a storage device are searched, and the keywords of the electronic files to be searched (word processing documents, spreadsheet documents, etc.) There is known a method in which whether or not the keyword is included is analyzed in advance and a list of electronic files including the keyword input by the searcher is displayed quickly.

There are various formats for electronic files to be searched, and most of them include control codes other than character strings, graphic information, annotations unrelated to the contents, attribute information, and the like. Further, even in an electronic file including the same keyword, if the code systems used in the file are different, the same keyword is stored in the electronic file as a different byte string.

In order to deal with these problems, the conventional method requires a program called a filter for extracting a character string to be searched from an electronic file. The filter program extracts only character information from a file including pictures and figures. Further, the filter program is created and provided for each type of electronic file.

That is, a filter program corresponding to a particular file format extracts a character string to be searched from an electronic file of the file format, and the result of analyzing the presence or absence of a keyword or character from the extraction result is used as the original file. It is stored as the analysis result of.

[0006]

However, in the above conventional technique, every time a new file format appears, or
Every time the existing file format is changed due to bug fixes or version upgrades of the application program,
Add a filter program corresponding to them, or
There was a problem that it would have to be corrected.

The object of the present invention is to collect information that does not require addition or modification of functions even when a new file format appears or an existing file format is changed.
It is to be able to search.

[0008]

In order to achieve the above object, the present invention collects character information passed from an application program to an application programming interface (hereinafter API) of an operating system (hereinafter OS) for character display. , The result of accumulating this in a database (DB below) is the search target.

[0009]

BEST MODE FOR CARRYING OUT THE INVENTION A first embodiment of the present invention will be described below with reference to the drawings. I have seen it somewhere. But,
It does not remember its file format or file name. You may want to search for information in this state. Conventionally, a filter program is prepared for each file format, and by this, information on whether or not each of a large number of files contains the keyword is created and provided for the search. However, since it is difficult to prepare such a filter program as described above, in the present embodiment, the data to be searched is created without using the filter program.

In this embodiment, information stored in the storage device of a single computer (including a portable information terminal) or a plurality of computers 101, 102 as shown in FIG.
104 is connected by the network 103 (wired or wireless), and the information required by the user is searched from the information accumulated in a distributed environment in which information is mutually exchanged.

Almost all modern computers have a software structure as shown in FIG. Hardware layer 2
01 has an OS layer 202 that directly controls these,
On top of that, there are a middleware layer 203 and an application layer 204 that collect functions commonly used by application programs.

The upper layer software utilizes the functionality of the lower layer software. It is the API that defines how the lower layer software uses its own functions for the upper layer software. For example,
The spreadsheet application uses the file access API to use the file access function provided by the OS.

Generally, each application program records data in a unique file format. For example,
The format of the file created by the word processing application and the file created by the spreadsheet application are different. However, applications operating on the same OS use a common character display API regardless of the file format to display the contents of the file on the screen.

FIG. 3 shows the situation. A text file 301 and an HTML file 30 are stored in the storage device of the computer.
2. Word processing document file 303, PDF file 30
There are files of various formats such as 4, and there are application programs (text editor 305, Web, etc.) for browsing / editing the contents of each file format.
Browser 306, word processing application 307, P
DF file viewer 308) exists. Each application program opens a file, extracts information to be displayed on the screen, and displays the screen by using the screen drawing API of the OS. For example, when the display contents are characters, the characters are displayed on the screen by using the character display API (310). There are one to several character display APIs depending on the OS.

The modern OS is an AP provided by itself.
It is provided with a mechanism (API hook) for notifying another program (hook program) registered in advance each time I is called, or can be realized by expanding the function.

In this embodiment, the mechanism of this API hook is used to collect the character information displayed on the screen and perform a search on the collection result, whereby a new file format appears or an existing file format is created. Even if the file format of is changed, it realizes information collection and search that does not require addition or modification of functions.

In order to use the mechanism of the API hook, it is necessary to register the hook program in the OS in advance (store the hook program in the storage device and store the storage location in the OS). The hook program is registered during the OS booting process, the user login process, or when the user gives an instruction. After registering the hook program, until it is released,
All characters displayed on the screen according to the user's operation are to be stored.

Further, in the present invention, the hook program for collecting information and the search program for searching can operate as different programs. That is, only the search method can be changed to a more efficient method without changing the information collection method. Since the essence of the present invention lies in the information collecting method, in this embodiment, a known simple one is used as the search method, but a keyword or character included in the search target is analyzed in advance to create information called an index, It is possible to use more complicated ones such as a method for performing a high speed search.

An information collecting method using the mechanism of the API hook will be described below according to the procedure (FIG. 4) until the application program opens the file and displays the contents. The broken line in FIG. 4 indicates the division of the operating software. The left side shows the operation of the application program, and the center shows the OS or hook processing program (whether the hook program corresponding to the API is registered when the API is called). The operation of the program that has the function of transferring the processing to the hook program if it is registered, and the right side is the operation of the hook program.
Represent each.

First, the application program reads a file from the storage device (step 401),
The contents displayed on the screen such as character information and graphic information are extracted from the file (step 402). Next, a screen output API suitable for display contents (image, character, figure, etc.) is called (step 403) (here, an example of calling the character display API is described).

As a result, the control is transferred to the OS, and the hook process corresponding to the API of the OS (the hook program is started by checking the storage location of the hook program) is performed (step 411). When the hook program corresponding to the API is registered, the OS transfers control to the registered hook program before performing the original processing of the API (step 412).

The hook program extracts the character information passed to the character display API and records it in the character information recording DB (step 421). This character information has a common format regardless of the file format. Specifically, it is in text format. Further, this character information recording DB may be present in the computer in which the hook program operates or in another computer accessible via the network. By this, only the character information of the file is extracted without the intervention of the filter program, and moreover, it is stored in the common format regardless of the format of the original file. In this way, the character information is accumulated using the mechanism of the hook and provided for the search.

When the processing of the hook program is completed, the control returns to the OS, and the OS displays the character which is the original processing (step 412) and returns the control to the API calling application. The application program checks whether all screen display processing has been completed (step 40).
4) If not completed, the process returns to step 402 and the above process is repeated until completed.

An example of information recorded in step 421 (D
The structure of B) is shown in FIG. Information acquisition date 501, application program name 50 that called the character display API
2. The file name 503 displayed by the application program, the location 504 of the file, the process identifier 505, the thread identifier 506, the display character string 507, etc. are recorded. Depending on the OS, it may not be possible to acquire all the information (for example, when the OS does not have threads or when browsing data in the DB instead of files). Therefore, not all information is necessary. In addition, each row of the data table of FIG. 5 is called one record here.

The details of the recording process by the hook program are shown in FIG. First, the character passed to the character display API is extracted (step 601), and information regarding the application that called the character display API is acquired (step 6).
02). This is information such as an application program name, a process ID, and a thread ID that should be recorded in the DB.

Next, the continuity of the character information is judged by comparing with the item of the record already registered in the DB (step 603). By this process, if one screen display is completed by calling the character display API multiple times, each A
The connection between PI calls can be recognized.

Specifically, in step 603, the date (50
1), the application program name (502), the file name (503), the file location (504), the process ID (505), and the thread ID (506) all match (including the case of "unknown") If found, step 60 is added to the display character string (507) of the item.
The character extracted in 1 is added (step 604). If not found, a new item is created in the DB and the character extracted in step 601 is newly registered in the display character string column (507) (step 605).

Next, FIG. 7 shows the processing procedure of the search program. First, the user inputs a keyword to be searched (step 701). Next, the top record of the DB generated by the hook program is read (step 70
2) A character string search is performed on the character string information in the display character string column of the record (step 703), and the process is sorted according to the presence / absence of a keyword (step 704). If the record does not include the keyword, an attempt is made to obtain the next record in the DB (step 705). If it can be acquired, the process returns to step 703. If it cannot be acquired, the process ends. If the record includes a keyword, the content of the record is displayed (step 706) and the input by the user is waited (step 70).
7). It is determined whether or not the user has pressed the "search next" button (step 708). If the user has pressed the button, the process returns to step 705 to perform the above process on the next record. If the [Find Next] button is not pressed, the search process ends.

FIG. 8 shows an example of a screen displaying the search results of the search program. Search string 801, entered by the user,
The content 805 of the found record is displayed. The record includes a recorded date 802, an application name 803 that displays the character string, a display character string 804, and the like. The date 802 corresponds to the date 501 in FIG. 5, the application name 803 corresponds to the application name 502 in FIG. 5, and the display character string 805 corresponds to the display character string 507 in FIG. The display character string in FIG. 5 also has tabs and line feed characters as part of the characters. Therefore, when it is displayed on the display screen, a display form such as 804 in FIG. 8 is obtained. (It may not have a layout function in particular) Other items in the DB may be displayed. The button 821 is provided to advance the search to the next record, and the button 822 is provided to end the search process. Further, the keywords (811, 812) found in the display character string are highlighted so that they can be distinguished from other character strings (here, they are underlined and bolded).

According to the first embodiment of the present invention as described above, a filter program is unnecessary and a database to be searched in a common format can be obtained regardless of the difference in file formats.

In the second embodiment of the present invention, in step 604 of FIG. 6 of the first embodiment, the extracted character string is stored in the existing DB.
When adding to the display character string column (507) of a record, instead of simply concatenating to the end of the addition destination character string, if the end of the addition destination character string and the beginning part of the addition character string match,
Concatenate only the rest of the added string after removing the matching part to the end of the added string. If the entire additional character string is already included in the addition destination character string, the character string is not added. This will be specifically described with reference to FIG.
In step 402 in FIG. 4, it is assumed that one screen is displayed. First, it is assumed that the title "meeting memo", the date and time, the place, the participant, and the handout are displayed on a certain screen due to the limitation of the display space of the screen. On the next screen, it is assumed that the handouts and the contents are displayed by scrolling up. At this time, the “distribution materials” are the same on both screens. Therefore, the character information record on the next screen does not include the “distribution material”.

With this, it is possible to remove the overlapping portion in the character string to be searched without losing the continuity of the character string information. This reduces the amount of data to be searched,
As a result, the search time can be shortened.

The third embodiment of the present invention is an information collection / retrieval method for replacing the information collection part of the first embodiment or the second embodiment. In the first and second embodiments, the file voluntarily opened by the user is the target of the information collection, but in the third embodiment, the user does not care whether the file is actually viewed or not. Collects information on a set of files specified in advance by. The method of specifying the file set is to enumerate the file names, to specify one or more directories to target all files contained in the directories for information collection, or to specify a combination of the above two methods. Methods etc. are considered. Any file that can be accessed through its own device or network can be the target.

The procedure will be described below with reference to FIG. First, the hook program is registered (step 901). When there are a plurality of character display APIs, the hook program is registered in the character display API. Character display AP to be registered
I may be all. By this, the following character display A
The PI call is hooked and the display character information is accumulated. Next, the first file of the file set to be analyzed is acquired (step 902), and a new record for the file is registered in the DB whose structure is shown in FIG. 5 (step 903). Here, the date (501), application name (502), file name (503), and file location (504) can be determined.

Next, the file is opened by an application capable of displaying the contents of the file (step 90).
4) The process ID of this application is acquired (step 905), and the process ID is registered in the record registered in step 903. Since the screen is displayed by opening the file with the application, the hook program registered in step 901 is executed. It should be noted that opening a file does not need to be actually displayed on the display device, but the character display API needs to operate. The process ID is acquired
If the process IDs are the same, it can be considered that the same file is opened, and this is for determining whether to create a new record or add a character string to the same record as in step 603 of FIG.

In the third embodiment, the method for determining continuity of character information in step 603 of the hook program is "thread I".
D506 is not considered. " By this modification, the display character string is added to the display character string column 507 of the record registered in step 903. In this case, there is only one thread ID, and even if the thread ID is not taken into consideration, it can be known only from the process ID whether the files are the same file or different files.

Next, the "next screen display event" is sent to the application based on the process ID acquired in step 905 (step 906). This is because it is assumed that one file cannot be displayed on one screen at a time. The next screen display event varies depending on the OS, but the page down key pressing event, the down arrow key pressing event, the down scrolling event, and the like correspond to this.
Sending the next screen event indicates that the control information is automatically created and sent to the application program as if the operator saw the screen and operated to display the next screen. This is to display the entire one file.

Next, in step 906, it is determined whether the end of the file has been reached (step 907),
If it has not arrived, step 906 is repeated. When the next screen display event is received after the end of the file is reached, the screen is not drawn. That is, since the hook program is not executed, the arrival at the end of the file is recognized by "the hook program is not executed even if the event is sent in step 906" (details will be described using FIG. 10). .

When the end of the file is reached, the application is terminated (step 908) and an attempt is made to obtain the next file of the analysis target file set (step 909). If the following files exist, step 90
Returning to step 3, the above procedure is repeated for the file. If it does not exist, the processing ends.

Details of the "judgment to reach the end of file" in step 907 will be described with reference to FIG. In summary, record the execution time of the hook program, and when the end of the file is reached, there is no new display, that is, the hook program does not work. Therefore, the last execution time does not change. The end of the file can be determined by detecting this.

First, the last execution time of the hook program is acquired (step 1001). Step 1001
Hook program to execute
It shall record the time when it executes itself at 21 and allow it to be obtained from other programs.

Next, this is compared with the variable "previous execution time" which can be set and referenced in step 907 (step 1002). If they are equal, it is determined that "the end has been reached" (step 1005). If they are not equal, the last execution time is substituted for the last execution time (step 100).
3), it is determined that "the end has not been reached" (step 100)
4).

[0043]

According to the present invention, even if a new file format appears or an existing file format is changed, it is possible to realize information collection and search that does not require addition or modification of functions.

[Brief description of drawings]

FIG. 1 is a diagram showing a computer environment.

FIG. 2 is a diagram showing the structure of software inside a computer.

FIG. 3 is a diagram showing that an application program calls an API of an OS.

FIG. 4 is a processing procedure diagram when collecting display character information by hooking an API.

FIG. 5 is a diagram showing an example (DB structure) of information recorded by a hook program.

FIG. 6 is a procedure diagram of a character information recording process (step 421).

FIG. 7 is a search processing procedure diagram.

FIG. 8 is an example of a search result display screen.

FIG. 9 is a procedure diagram of an information collection method in which a file set designated by a user is a target of information collection.

FIG. 10 is a procedure diagram of a method of determining arrival at the end of a file (step 907).

[Explanation of symbols]

402. Step 403 for extracting display contents on screen from file. Invoking the screen display API Step 411. Step 412 of performing hook processing. Step 421 for displaying characters on the screen. Step to record screen display characters in DB

Claims (6)

[Claims] 1. An application program for character display of an operating system is called from an application program, character information handled by the application program is passed to the application programming interface for character display, and application programming for the character display is performed. An information collecting method, characterized in that the character information passed to an interface is collected and stored in a database. 2. An application programming interface for character display of an operating system is called from an application program, the character information handled by the application program is passed to the application programming interface for character display, and application programming for the character display is performed. An information collecting method comprising: operating a hook program when an interface is called, collecting the character information passed to an application programming interface for displaying the character in the hook program, and storing the collected character information in a database. 3. In the step of collecting the character information passed to the application programming interface for displaying the character, it is possible to add and collect a character string excluding a duplicated portion with the already stored character string. The information collecting method according to claim 1 or 2, which is characterized. 4. The application program reads a file for a file set designated by a user and displays the file.
Information collection method described in. 5. A function of a computer as means for collecting character information passed from an application program to an application programming interface for displaying characters of an operating system called by the application program, and means for accumulating the character information in a database. A character information storage program for searching. 6. A means for calling an application programming interface for character display of an operating system from an application program, a means for passing character information handled by the application program to the application programming interface for character display, and a means for displaying the character display. Means for collecting the character information passed to an application programming interface for storing the character information and storing the character information in a database.