JPH09146803A - Data base access system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the capacity of a disk used to store large data in a data base and facilitate data storage and alteration. SOLUTION: A column 19b of a table 19 is stored with not the entity of data in a file 20 present on a file system 2303, but only information used to access the file system 2303. A column data acquisition part 10 acquires data from the file system 2303 through a file access control part 15 and then sends the data back to a DB front end 4 when a column data acquisition request from the DB front end is a request to the column 19b.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a database access system that operates in a network connecting a plurality of information processing devices. .

[0002]

2. Description of the Related Art In recent years, under the influence of downsizing of computers and development of network operating systems, information processing devices such as personal computers have been installed in LAN (Local Area Network).
The number of systems connected by networks such as k) is increasing. These systems are often composed of a high-performance information processing device called a server and a plurality of information processing devices called a client, and are called a client server system. In the client server system,
Server resources (eg hard disk, printer, etc.)
Can also be shared by multiple clients.

In the database (hereinafter referred to as DB) processing in the client server system, a program for actual data management is placed on the server, and a program for issuing a user access request to the server side is placed on the client. It is widely practiced to disperse processing. The server side program is called a DB server, and the client application program is called a DB front end. As the DB server, a relational database system that represents data in a table format is often used. In addition, the DB front end uses a GUI (Grap) to access the DB with ease of use.
often has a User Interface.

FIG. 3 shows an example of a conventional configuration of a relational DB server and a DB front end connected by a network.

The DB server 5 operates on the server machine 2, and the DB front end 4 operates on the client machine 1. DB server 5 and DB front end 4
Are connected by the network 7, and the communication control unit 8 executes data transmission control on the network. In the hard disk 2203 installed in the server machine 2, D
The tabular data managed by the B server 5 is stored. Tables are called tables. These stored table groups are collectively called a database. Each row of the table is called a record and each column is called a column. The column is
It has various attributes such as data type, length, and name.

In order to transfer information between the DB front end 4 and the DB server 5, a request issuing means dedicated to DB access and a result obtaining means are prepared. The DB front end 4 issues various DB access requests (hereinafter referred to as DB access requests) to the DB server 5 via the communication control unit 8. After that, the DB server 5 returns the processing result to the DB front end 4.

For a DB access request, SQL (Strru
In addition to a request to issue a query Query Language statement (hereinafter referred to as an SQL request), a request to acquire column data of search results (hereinafter referred to as a column acquisition request), a request to acquire column data attributes of search results (hereinafter columns) There is a request for acquiring the number of columns of the search result (hereinafter referred to as a column number acquisition request) and the like. Here, SQL is a language for DB operations defined by ISO (International Organization for Standardization). SELECT
There are several commands such as (search), UPDATE (update), INSERT (insert), DELETE (delete), and CREATE (create table).

When the SQL statement issued by the DB front end 4 is a SELECT command, the DB server 5
Creates a derivation table as a search result. Here, the derivation table is a logical table that is partially cut out from the table stored in the DB. The data to be cut out is determined by the search condition included in the SELECT statement.

FIG. 4 shows a conventional information sequence of the DB front end 4 and the DB server 5. In the figure, the arrow from the DB front end 4 to the DB server 5 indicates the issuance of a request. The arrow corresponding to the return value from the DB server 5 to the DB front end 4 is omitted.

The DB front end 4 executes step 401.
Issue a SQL request with. In the SQL request, an SQL statement is sent as a parameter. In this way, the SQL sentence is SEL
If it is an ECT command, a derivation table is created. Since the processing after step 402 is the processing when the SQL statement is the SELECT command, the DB server 5
In, a derivation table has been created. Next, the DB front end 4 issues a column number acquisition request in step 402. Here, the DB front end 4 requests the number of columns because it is necessary to prepare the buffer for storing the data for the number of columns before the acquisition of the column data. Next, in step 403, a column attribute acquisition request is issued. In the column attribute acquisition request, the type of attribute to be acquired is specified as a parameter. The attributes that can be acquired include data type, data length, and column name. The DB front end 4 requests the data type and the data length,
Similar to step 402, this is to prepare an optimum buffer for the data to be acquired, and to request the column name is to add the column name when displaying the search result. Next, the DB front end 4 executes step 40.
At 4, the cursor movement is requested. Here, the cursor is
It logically points to one record in the derivation table,
The cursor move request is to issue a request to move the cursor to the next record. Next, the DB front end 4 issues a column acquisition request in step 405. In the column acquisition request, data of one column is acquired from one record pointed by the current cursor.
Specify the column number and buffer pointer as parameters. As for the column number, the leftmost end of the derivation table is 1. The DB front end 4 issues this request for the number of columns in the derived table.

By the way, with the recent progress of multimedia, there has been an increasing need for integrated management of large data such as images, voices, moving images, and documents in a relational database.

When storing large data in a relational database, the user typically does not edit the data directly on the DB front end 4. First, data is edited using a dedicated editing program, and then the data is stored in some file system.
After that, the data is fetched from the file system to the DB front end 4, and the SQL request is issued with the INSERT command of the SQL statement as a parameter.

The same applies to the case where the relational database is used to manage data that already exists in any file system, rather than creating new data. Recently, due to the progress of network technology, a WWW (World Widow), which is a network file system capable of sharing server files among multiple users
de Web), FTP (File Transfer)
Protocol) is rapidly spreading. WWW
Is configured to access various data (document, still image, moving image) stored in the WWW server from the WWW client. The WWW manages a huge amount of various data with a file system, but there is a high need to manage these data in a DB in order to facilitate operation management and data search.

An example of processing document data, which is one type of large data, with a relational database is the document storage method described in Japanese Patent Laid-Open No. 4-276855.

[0015]

According to the prior art,
There is a problem that a very large disk capacity is required to store large data of files distributed in various locations in the relational DB. Usually, since the number of records in the DB is increasing, the problem of lack of disk becomes more serious as the opportunity to process large data increases.

Further, the conventional technique has a problem that the method of storing data in the DB is troublesome. In order to store the data in the DB, the user loads the file from the file system into the DB front end 4 and then the SQL
Issue an SQL request with the INSERT command of the statement as a parameter. The above operation can be performed using the GUI included in the DB front end 4, but since the data is stored for each record, the operation requires a considerable amount of time. Also for DB front-end program developers, a DB that has a large data storage function
There is a problem that developing a front-end program involves a huge amount of work.

Further, in the conventional technique, there is a problem that the method of changing the data once stored in the DB is more troublesome than the storing. In order to change the data, the user fetches the data with the SELECT command into the DB front end 4 and then
Load it into the editing program. After that, after making an appropriate change in the editing program, an operation of returning the data to the DB front end 4 must be performed. This operation is quite time consuming as well as storing. Also, DB
Even a front-end program developer has a problem that a huge amount of work is involved in developing the DB front-end program having the large data changing function.

Therefore, a first object of the present invention is to reduce the disk capacity used when storing large data in a relational DB.

A second object of the present invention is to facilitate the operation when storing large data in a relational DB.

A third object of the present invention is to facilitate the operation when changing the large data stored in the relational DB.

[0021]

In order to achieve the above object, the DB access system of the present invention is a database access system composed of a plurality of information processing devices connected to a network.
A means for interpreting a QL statement, a column data information buffer for storing column attributes, a column attribute acquisition means for obtaining column attributes of a derived table and setting column attribute information in the column data information buffer, and a file system File access means for acquiring the file data of, the reference column buffer for storing the data acquired by the file access means, if the column of the derived table is a column that refers to a file system, the reference column buffer is It has a column data acquisition means for returning the stored data to the application.

According to the DB access system of the present invention,
When the SQL interpreting means receives the SQL issuing request from the application, the received SQL sentence is SELECT.
Interpret whether it is a command or not, and the result of the interpretation is SELE.
If it is a CT command, control is transferred to the column attribute acquisition means. The column attribute acquisition means acquires the column attribute of the derived table and stores it in the column data information buffer. At that time, it is determined whether or not the column is the reference column, and the result is stored in the column data information buffer. When the column data acquisition request is received from the application, the column data acquisition unit refers to the column data information buffer and checks whether the column is a reference column. Here, if it is the reference column, the control is transferred to the corresponding file access means. The file access means acquires the data of the file system and stores the data in the reference column buffer. If the column is a reference column, the column data acquisition means returns the data stored in the reference column buffer to the application.

[0023]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings.

The DB access system which is an embodiment of the present invention has a database gateway configuration.
Therefore, it is possible to construct a system that solves the problem without modifying the existing DB system.

FIG. 1 shows a block diagram of the DB access system of this embodiment. The network 7 of this embodiment includes a server machine 2 on which a DB server 5 and a gateway 6 operate,
Client machine 1 on which B front end 4 operates
And the server machine 3 on which the file server unit 9 operates are connected. The communication control unit 8 controls communication among the DB server 5, the gateway 6, and the file server unit 9.
The access request from the DB front end 4 is sent to the DB server 5
It is preset to be performed not for the gateway 6 but for the gateway 6. The gateway 6 is D
It may be operated on a machine different from the B server 5.

FIG. 2 shows the hardware configuration of each machine. The client machine 1 includes a control device 210, a storage device 211, and an input / output device 212. Storage device 2
Reference numeral 11 denotes a main memory 2102 and a hard disk device 21.
03. The input / output device 212 is the keyboard 210.
5, a mouse 2107, and a display device 2109.
The control device 210 includes a CPU (central processing unit) 2101,
Keyboard control unit 210 for controlling the keyboard 2105
4, a mouse control unit 2106 that controls the mouse 2107,
A display control unit 2108 that controls the display device 2109, and a network control unit 2110 that controls communication performed via the network 7 are included. Server machine 2
Is the same as the client machine 2 with the control device 220,
The storage device 221 and the input / output device 222 are provided. The storage device 221 has a main memory 2202 and a hard disk device 2203. In this embodiment, the DB server 5 stores and manages various data in the hard disk device 2203. Like the server machine 2, the server machine 3 also includes a control device 230, a storage device 231, and an input / output device 232. The storage device 231 has a main memory 2302 and a hard disk device 2303. In this embodiment, the file server unit 9 stores and manages various data in the hard disk device 2303.

A first object of the present invention is to reduce the disk space for storing large data. Therefore, according to the present invention, when storing large data held in any file system in a column of a table, the substance of the data is not stored, and only the information necessary for accessing the data is stored. The file system type is distinguished by the column name. For example, if a file managed by the FTR server is to be used as column data, the rule is to add _ftp to the end of the column name. Hereinafter, such a column will be referred to as a reference column.
In FIG. 1, the column 19b of the table 19 corresponds to the reference column.

The reference column will be described below with reference to FIG.

A list of data types managed by the DB server 5 is shown in (a). Reference numeral 111 is a fixed length character string type, 112 is a variable length character string type, 113 is a numerical value type, 114 is a fixed length binary type, and 115 is a variable length binary type. The difference between fixed length and variable length is that fixed length is a data type in which the data length is always constant, and variable length has a fixed maximum data length, but the actual data length varies from record to record. It is a point. Among these data types, the fixed length character string type 111 or the variable length character string type 112 is used as the reference column. The character string type is used as the reference column because the information required for access is stored as a character string.

Prior to the description of the reference column naming convention described in (b), the file system supported by this embodiment will be briefly described.

The file systems supported by this embodiment are the network file system 116, WWW.
117 and FTP 118. Both file systems have a client-server type configuration. That is, the network file system 116 includes the WWW 117 in the network file system server program and the network file system client program.
Operates as a WWW server program and a WWW client program, and the FTP 118 operates as an FTP server program and an FTP client program.

The network file system 116 is
It is possible to arrange a plurality of server programs and a plurality of client programs in the same system. Each server program is given a server name, and it is necessary for the client program to specify the server name to be accessed and log in before acquiring the file. The network file system server program manages files in a directory structure, and when accessing a file from a client program, the directory name and the file name are specified.

In the WWW 117 as well, like the network file system 116, it is possible to arrange a plurality of server programs and a plurality of client programs in the same system, and each server program is given a host name. In addition, files are managed in a directory structure. However, unlike the network file system 116, it is not necessary to log in to the host before acquiring the file data. Data transfer between the client program and the server program is performed by HTTP (Hyp
er Text Transfer Protocol
l). In HTTP, when the client program acquires file data, the host name / directory name / file name are specified at the same time. The sequence of these three information is URL (Uni
It is called a form resource locator.

The FTP 118 transmits / receives data between the client program and the server program to the FTP (Fi
le Transfer Protocol). As with the WWW 117, the host name is attached to the server program in the FTP 118 as well. When accessing a file, first specify the host name and start the server program of the other party. After that, the directory name / file name is specified to obtain the data.

In order to distinguish the three file system types by column names, the following naming rules are provided. If the file system to be stored exists in the network file system 116, _fil is added to the beginning of the column name.
Add e, and give the subsequent character string the name that the user wants to name. Similarly, if the file exists in WWW117, _www, and if it exists in FTP118, _ftp.
Is added. In the example of column 19b of table 19, the file exists in the network file system and the name given by the user is "image". CREATE
A user who creates a table with the E command assigns a column name according to the above rule.

(C) shows a rule of data stored in the reference column. The data to be stored is information for specifying a file to be accessed for each file system.

When the file system is the network file system 116, the server name, the directory name, and the file name are required as described above, so a character string in which the server name, the directory name, and the file name are concatenated with / is stored. It was a rule. In case of WWW117, U to specify the host name, directory name and file name at the same time
The rule for storing RL is adopted. Similarly, in the case of the FTP 118, the rule is to store the host name and the directory name where the file is stored, and the character string that concatenates the file names with /. In the example of the data in the column 19b, the file to be stored is tv.txt that exists in the machine with the server name SVR. g
It means that it is if20. Reference column is 1
You can have more than one per table.

In this embodiment, the column naming rule is adopted to determine the reference column.
There is also a method to add a new data type for the reference column.
In this method, a different data type is provided for each file system to be accessed. For example, the reference column for accessing the FTP 118 is of FTP type. Upon accepting the table creation request by the CREATE command, the gateway 6 checks the presence / absence of a data type corresponding to the reference column, and stores the information in the hard disk drive 2203. D
The B server 5 issues an SQL request in which the data type corresponding to the reference column is converted to the character string type.

The gateway 6 acquires the data existing in the file system by using the above-mentioned naming convention for the reference column. Hereinafter, each control unit forming the gateway 6 will be described.

In the figure, 10 is a column data acquisition unit for requesting acquisition of column data. Reference numeral 21 is a column attribute acquisition unit that requests acquisition of column attributes. The column data acquisition unit 10 and the column attribute acquisition unit 21 both set the acquired column data and the column-related attributes in the column data information buffer 13. The column attribute acquisition unit 21 secures the memory area of the column data information buffer 13. The column data acquisition unit 10 is the DB front end 4
If the column data acquisition request from the request is for the reference column, the control is transferred to the file access control unit 15. The file access control unit 15 stores the data in the reference column data buffer 14 after acquiring the data from the requested file system. The memory area of the reference column data buffer 14 has a file access control unit 15
Secure. 11 is an S that issues SQL to the DB server 5
A QL issuing unit 12 is a cursor moving unit for issuing a cursor moving request to the DB server 5.

FIG. 12 shows the configuration of the file access control unit 15.

Reference numeral 16 is a network file control client unit for accessing a network file system server, 17 is a WWW control client unit for accessing a WWW server, and 18 is an FT.
This is an FTP control client unit for accessing the P server. The file access control unit 15, after checking which file system the DB access request from the DB front end 4 is, requests the network file control client unit 16, the WWW control client unit 17, or the FTP control client unit 18. Transfer control. Each of 16, 17, and 18 has a client function for accessing the server of each file system. Although the server unit of each file system is different in substance, the configuration of the server machine is the same, so only the file server unit 9 is shown in FIG.

FIG. 9 shows the components in the column data information buffer 13.

Reference numeral 131 is a column information pointer array.
The column information pointer array 131 is an array for setting a pointer to the column information 132. In the first element of the array,
A pointer is set to the leftmost column information 132 of the derivation table, and the pointer is set in the order of the derivation table below. The first element of the column information 132 is a reference column storage area 1321 that stores whether or not there is a reference to a file. 0 if there is no reference, 1 if network file system, W
2 is stored for WW, and 3 is stored for FTP. 132
Reference numeral 2 is a data type storage area for setting the data type of a column. If the column is a reference column, the data type is not the character string type managed by the DB server 5,
Variable length binary type is set. This is because the conventional database system uses a binary type to store large data, and therefore the DB front end 4 needs to return this data type in order to prepare an optimum buffer. . Unlike the character string type and the numerical type, the binary type is a data type in which the data to be set is not determined in advance, and any data such as image or sound can be stored depending on the reading method in the DB front end. In addition, the maximum length of data is much larger than other data types, so it can be said to be the most suitable data type for storing large image and audio data.

A column name storage area 1323 stores a column name. Reference numeral 1324 is a column length storage area for setting the column length of the column. Here, if the column is a reference column, the maximum length of binary type is set as the column length. 1325 is a column data storage area for storing column data. The column data storage area 1325 is
By the column attribute acquisition unit 21, the column length storage area 132
The size of 4 is secured.

FIG. 10 shows components in the reference column data buffer 14.

Reference numeral 141 is a reference column data length storage area for setting the data length of the file data to be acquired. 14
Reference numeral 2 is a file data storage area for actually copying the data of the reference file. File data storage area 142
Is secured by the file access control unit 15 for the size of the file data storage area 141.

FIG. 5 shows an information sequence of the DB front end 4, the gateway 6, the DB server 5, and the DB server unit 9 in the present invention. Similar to FIG. 4, the arrow from the DB front end 4 to the gateway 6 and the gateway 6
The arrow from to DB server 5 indicates the issuance of a request.
The processing request from the DB front end 4 to the gateway 6 is the same as the content described in FIG.

The DB front end 4 executes step 501.
Issue a SQL request with. Upon receiving this request, the gateway 6 transfers control to the SQL issuing unit 11 and issues the SQL statement designated as a parameter to the DB server 5 as it is (step 502). As described in the related art, when the SQL statement that is the parameter is the SELECT command, the DB server 5 creates a derivation table. Next, the SQL issuing unit 11 determines that the SQL statement is SELECT.
Only when the command is a T command, the number of columns in the derived table is acquired (step 503). Then, step 50
In 4, the column attribute acquisition unit 21 acquires information such as the column name, data type, and data length, and secures the area of the column data information buffer 13. Next, after referring to the acquired column name and checking whether the column is a reference column,
Reference column storage area 1321, column name storage area 132
3, data type storage area 1322, data length storage area 13
The value described in FIG. 9 is stored in 24. That is, for example, in the case of the reference column 19b, _ is added to the initial position of the column name.
Since there is a file, 1 is set in the reference column storage area 1321.
In the column name storage area 1323, the _file image is stored, and in the data type storage area 1322, the variable length binary type 1
15, the maximum length of binary type is stored in the data length storage area 1324.

The DB front end 4 executes step 502.
Then, a column number acquisition request is issued (step 505). Upon receiving this request, the SQL issuing unit 11
In step 503, the acquired number of columns is returned to the DB front end 4. DB front end 4 is step 5
Subsequent to 05, a column attribute acquisition request for the column name, data type, etc. is issued (step 506). Upon receiving this request, the column attribute acquisition unit 21 reads the information acquired in step 507 from the column data information 134, and then returns it to the DB front end 4. Next, the DB front end 4 issues a cursor movement request in step 507. In response to this request, the cursor moving unit 12 issues a cursor moving request to the DB server unit 5 (step 508). Next, the DB front end 4 issues a column data acquisition request (step 509). Upon receiving this request, the column data acquisition unit 10 receives the DB server 5
A column data acquisition request to the client (step 51)
0). Next, whether or not the column is the reference column is checked by referring to the reference column storage area 1321. If the column is the reference column, the control is transferred to the file access control unit 15.
The file access control unit 15 includes the file server unit 9
Then, a data acquisition request is issued (step 511).

FIG. 6 is a flowchart showing the operation of the gateway 6 in steps 510 and 511.

The column data acquisition unit 10 executes step 60.
At 1, issue a column data acquisition request to the DB server 5,
After acquiring the data of the corresponding column number, the data is stored in the column data storage area 1325 (step 60).
1). Subsequently, the reference column storage area 1321 is checked to see if the data acquisition target column is the reference column (step 602). If the column is a reference column, file access processing is performed (step 603). The file access processing is controlled by the file access control unit 15. After step 603,
Network file control client unit 16, WWW
Control client unit 17, FTP control client unit 1
Data is set in the reference column data buffer 14 by any one of 8 above. The column data acquisition unit 10
In the following step 604, the reference column data buffer 14
The data set to is returned to the DB front end 4.

On the other hand, if the column is not the reference column in step 602, the data stored in the column data 1325 of the column data information 132 in step 605 is changed to
Return to DB front end 4.

FIG. 7 is a flow chart showing the operation of the file access control unit 15 in step 603.

The file access control unit 15 first checks which file system the request refers to by referring to the reference column storage area 1321 (step 701). Here, if 1 is set, control is transferred to the network file control client unit 16 (step 702), and if 2 is set, WW
The control is transferred to the W control client unit 17 (step 70).
3) If 3 is set, control is transferred to the FTP control client unit 18. Each control unit accesses the server program of each file system, acquires the data length, secures the area of the reference column data buffer 14, and copies the data in common.

In FIG. 8, the network file access control client unit 16 controls. The flowchart of the network file access process of step 702 is shown. In this embodiment, the description of the WWW access process (step 703) and the FTP access process (step 704) is omitted.

The network file control client unit 16 checks the column data storage area 1325 and recognizes the server name where the file exists before performing the processing of step 801 and thereafter. After that, the file of the server is made accessible. Specifically, if the server is not logged in, the login process is performed.

Further, in the figure, the arrow from the file control client section 16 to the file server section 9 represents the issuance of a request dedicated to file access.

After the file becomes accessible, the file control client unit 16 acquires the file identifier from the acquisition file control server unit 9 (step 801). The file identifier is always specified as a part of the request parameter when the subsequent processing related to the file is performed. Next, the file size is acquired (step 802). When the file size is acquired, the file control client unit 16 sets a value in the data length buffer 141, and then secures a memory area for the size as the file data storage area 142.
Next, the file control client unit 16 actually acquires the file data and copies it to the file data storage area 142 (step 803).

The second object of the present invention is to facilitate the operation when storing large data in the relational DB, but the present invention also achieves this object. Records are added to the table by using the INSERT command of the SQL statement. At this time, the large data itself is not set, but the location of the large data may be added as a record.

Further, as a second object of the present invention, it was possible to easily change the large data stored in the relational DB. According to the present invention, this object is also achieved. This is because it is sufficient to edit the data stored in the file system by the conventional method without reading the data into the DB front end 4 each time.

By the way, the DB access system of the present embodiment has a gateway structure, but in addition to this system, a system of inventing a DB management system to which a new function is added is also conceivable. However, it is not realistic to introduce a completely new DB management system in the system of the current company. The DB management system, as the core of a company's information system, accumulates many data assets and AP assets in the system. Therefore, if a DB management system is newly introduced, the cost associated with the operational change of these assets may be considerably higher than the introduction effect.

According to this embodiment, the file group stored in the distributed file servers can be shown to the DB front end 4 as if the entity is stored in the DB. As a result, the disk space for storing data can be reduced and the disk can be effectively used.

Since the operation for storing and changing the large data can be facilitated, the work efficiency of the user can be improved.

[0065]

According to the present invention, the file group stored in the distributed file servers can be shown to the DB front end 4 as if the entity is stored in the DB. As a result, the disk space for storing data can be reduced and the disk can be effectively used.

Further, since the operation for storing and changing the large data can be facilitated, the work efficiency of the user can be improved.

[Brief description of the drawings]

FIG. 1 is a block diagram of a database gateway system according to an embodiment of the present invention.

FIG. 2 is a block diagram of hardware of a server and a client according to an embodiment of the present invention.

FIG. 3 is a block diagram of a DB front end and a DB server in a conventional example of the present invention.

FIG. 4 is an information sequence diagram of a DB front end and a DB server in a conventional example of the present invention.

FIG. 5 is an information sequence diagram of a DB front end, a gateway, and a DB server according to an embodiment of the present invention.

FIG. 6 is a flowchart showing the operation of column data acquisition processing according to an embodiment of the present invention.

FIG. 7 is a flowchart showing an operation of file access processing according to an embodiment of the present invention.

FIG. 8 is an explanatory diagram showing an operation of file access processing according to an embodiment of the present invention.

FIG. 9 is an explanatory diagram of a column data information buffer according to the embodiment of the present invention.

FIG. 10 is an explanatory diagram of a reference column data buffer according to the embodiment of the present invention.

FIG. 11 is an explanatory diagram of a reference column rule according to the embodiment of the present invention.

FIG. 12 is an explanatory diagram of a file control unit according to an embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Client, 2 ... Server, 3 ... Server, 4 ... DB front end, 5 ... DB server, 6 ... Gateway, 7 ... Network, 8 ... Communication control part, 9 ... File control server part, 10 ... Column data acquisition part , 11 ... SQL issuing unit, 12 ... Cursor moving unit, 13 ... Column data information buffer, 14 ... Reference column buffer, 15 ... File access control unit, 19 ... Table, 21 ... Column attribute acquisition unit.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Simone Sakaguchi 5030 Totsuka-cho, Totsuka-ku, Yokohama-shi, Kanagawa Stock company Hitachi Ltd. software development headquarters

Claims (7)

[Claims] 1. A database access system comprising a plurality of information processing devices connected to a network,
A means for interpreting the SQL statement received from the application, a column data information buffer for storing the attribute of the column, a column attribute acquisition means for acquiring the attribute of the column of the derived table, and setting the column attribute information in the column data information buffer. A file access means for obtaining file data of a file system, a reference column buffer for storing the data obtained by the file access means, and a reference table when the column of the derived table is a column for referring to the file system. A database access system comprising column data acquisition means for returning data stored in a column buffer to an application. 2. A database access system comprising a plurality of information processing devices connected to a network,
A means for interpreting the SQL statement received from the application, a column data information buffer for storing the attribute of the column, a column attribute acquisition means for acquiring the attribute of the column of the derived table, and setting the column attribute information in the column data information buffer. A file access means for obtaining file data of a file system, a reference column buffer for storing the data obtained by the file access means, and a reference table when the column of the derived table is a column for referring to the file system. A database access system comprising a gateway having column data acquisition means for returning data stored in a column buffer to an application. 3. The database access system according to claim 1 or 2, wherein a column referring to a file system is discriminated by a column name of the database. 4. A database access system according to claim 1, wherein the file access means has a plurality of different file systems. 5. The database access system according to claim 4, wherein different file system types are discriminated by column names of the database. 6. The database access system according to claim 1, wherein a plurality of reference columns can be set. 7. The database access according to claim 1 or 2, wherein when the column attribute acquisition unit receives a column attribute acquisition request from an application, the acquired column attribute is changed to an attribute corresponding to file data and returned to the application. system.