JPH0687226B2 - Database management method - Google Patents

Description

The present invention relates to a database management system, and more particularly, to improving responsiveness during file definition operation and reducing the number of commands to facilitate a user interface. And a database management method capable of reducing the scale of the database management system.

[Conventional technology]

A conventional database management system (hereinafter referred to as "DBMS") and a database (DB) have a configuration as shown in FIG. A DB 50 by the DBMS 20 in FIG. 10 is composed of a dictionary file 60 and a data file 70. The dictionary file 60 stores definition information about each file in the DB 50. Further, the data file 70 stores the actual data input by the user.

The data file 71 includes a management information section 72 and a data record section.
It is logically divided into 73 and. The data record unit 73 stores actual data in record units.
Further, the management information section 72 stores the physical storage position information of each record, the storage position information of the value of each item in each record, and the like.

The data operating unit 40 of the DBMS 20 uses the information of the management information unit 72 to efficiently execute processing such as data storage and retrieval.

In addition, the creation or modification of the structure of the data file 70
It is executed by the definition operation unit 30 of 20. Definition operation unit 30
Defines the data file according to the instruction regarding the definition of the data file 70 specified in the command 10, and records the result in the dictionary file 60. The definition process of the data file 71 is performed by the file management unit 32.
Was executed by the generation of the data file or the change of the management information section 72 by the reorganization of the data of the data record section 73 by the data reorganization section 33.

As the standard language for operating the DB, "database language SQL" by JIS (JIS X 3005-1987) and ISO
"Working Draft Database Language SQL2" (IS
O / IEC JTC 1 / SC21N: 1988-4-13) is specified. In the above SQL and SQL2, the database definition language (DDL) for describing the definition operation that creates or changes the structure of the data file and the data operation that stores and retrieves the data in the data file are described. There is a database manipulation language (DML). The definition operation unit 30 and the data operation unit 40 of the DBMS shown in FIG. 10 perform the operations described in the DDL and DML, respectively. Therefore, it can be said that the conventional DBMS has a structure corresponding to the grammar of SQL and SQL2. Figure 11 shows a part of the operating language of SQL and SQL2.

As the latest database management technology published by Nikkei McGraw-Hill, Inc. (WHInmon et al., 1987-5-12), pages 237-253, as a technology that uses a DBMS, it is used in a distributed processing and data environment. A system to do is considered. This data distribution type distributed processing system is considered to be effective as a means for distributing the location of information and integrating management, and a means for distributing the load on the processing device. FIG. 12 shows an example of the configuration of a data distribution type distributed processing system. In FIG. 12, 830 is a personal computer (hereinafter referred to as “personal computer”), 840 is a workstation, 850 is a server, and 833, 843 and 853 are DBs.

[Problems to be Solved by the Invention]

In the above-described conventional DBMS, it is necessary to reorganize the data in the file by changing the definition of the file.
Generally, the data reorganization process is complicated and requires a long processing time because all the data in the file is accessed, and the responsiveness when the user operates the DB is deteriorated.

Further, in the above-described DBMS according to the related art, even if the operation of accessing the DB is the same, if the target is different, the command is different. This makes it difficult for the user to operate the DB.

Also, in a data distribution type distributed processing system that will become popular in the future, it will also manage the DB and its management on each personal computer.
Although a DBMS will be installed, the scale of the DBMS is required to be small because the amount of memory installed in the personal computer is small. However, the DBMS having the structure shown in FIG. 10 had a large scale.

The present invention has been made in view of the above circumstances, and a first object thereof is to solve the above-described problems in the prior art,
It is to provide a DB management method capable of improving responsiveness at the time of file definition operation. Another object of the present invention is to provide a DB management method capable of reducing the number of commands and facilitating a user interface. Still another object of the present invention is to provide a DB management method capable of reducing the scale of the DBMS.

[Means for Solving the Problems]

The above-mentioned object of the present invention has a data file for storing data, a dictionary file for storing data defining the logical structure of the file, and a data operation means for accessing the data file and the data in the dictionary file. In a database management system that manages a database, the data file and the dictionary file have the same physical structure, and the dictionary file stores a definition time information storing the time when the logical structure of the data file is defined or modified. Means and a generation time information storage means for storing the time when the data file is generated or the structure is changed, and the generation of the data file from the data defining the logical structure of the file stored in the dictionary file. Is provided with a data file generation means for changing the structure, and when the data file is accessed by the data operation means, the definition time information in the storage means is compared with the generation time information of the target file, and the comparison is made. This is achieved by a database management method characterized in that the data file generating means is activated according to the result of (1).

[Action]

In the database management system according to the present invention, the data file and the dictionary file have the same physical structure, and the dictionary file stores the time when the logical structure of the data file is defined or modified. An information storage means and a generation time information storage means for storing the time when the data file is generated or the structure is changed are further provided, and further, from the data defining the logical structure of the file stored in the dictionary file Data file generation means for generating or changing the structure (file management unit in the embodiment)
And when the data file is accessed by the data operation means, the definition time information in the storage means and the generation time information of the target file are compared in the time information comparison unit, and the result is compared according to the result of the comparison. Since the data file generating means is activated only when necessary, the definition operation for generating the data file or changing the structure can be greatly simplified.

More specifically, in the database management method according to the present invention, the data reorganization processing and the definition command analysis processing, which occupy most of the conventional DBMS definition processing, are replaced by the management information section change processing in the file. And, it is configured by the comparison process of the time information, which is the starting means of the definition process described above, which makes it possible to eliminate most of the definition process of the conventional DBMS and significantly reduce the scale of the DBMS. It was done.

〔Example〕

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 2 is a block diagram of a DB system showing an embodiment of the present invention. In the figure, 200 is a DBMS, 300 is a DB, and 100 is an operation command. In DBMS200,
A time information comparison unit 210, a file management unit 220, and a data operation unit 230, which will be described later, are provided. The files in the DB 300 are composed of a dictionary file 310 that stores data that defines the logical structure of the file and a data file 320 that stores normal data. Here, in the dictionary file 310, the file register 400 and the item register 500, and the data file 320
The case where the address book 600 is stored is shown in the inside.

FIG. 3 shows the logical structure of each file and the data format of each part. FIG. 7A shows the structure of each file. Each file in the dictionary file 310 and the data file 320 is composed of a management information section 700 and a data record section 730. The management information section 700 stores a file management information section 710 for storing management information about the file and a file. It is composed of an item management information section 720 which stores management information on each of the constituent items.

As shown in FIG. 3B, the file management information section 710 has an area for storing generation time information 712 that specifies the time when the file is generated or the structure is changed. In addition, the item management information section 720, as shown in FIG.
It has an area for storing an item identifier 722 for specifying each item.

The data record section 730 is composed of a plurality of items 735, as shown in FIG. Each item 735 includes an item value 734, an item value length 733, and an item identifier 732.

As described above, the dictionary file 310 contains a file directory that stores data indicating the attributes of the file.
Item register 5 that stores data indicating 400 and item attributes
00 and are provided.

FIG. 4 shows the file format of the file directory 400. Data record part 730 of file directory 400
As shown in FIG. 4 (c), is a definition time information 402 indicating the file name 401 for specifying the file and the time when the file specified by the file name 401 is defined as items.
Have. The item management information section 720 stores the item management information 421 and 422 of the file name 401 and the definition time information 402, respectively, as shown in FIG. Further, in the generation time information 712 of the file management information section 710, as shown in (a) of the figure, the date information “88/01/01” is set as the time when the file register 400 is generated or the structure is changed.
Is stored.

FIG. 5 shows the file format of the item register 500. The data record portion 730 of the item register 500 is the fifth
As shown in FIG. 6C, the item has a belonging file name 501 that identifies the item, an item name 502, and a data type 503 that indicates the attribute of the data stored in the item. Item management information section 7
20 is the above-mentioned belonging file name 501, as shown in FIG.
Item management information for each item name 502 and data type 503
It contains 521, 522 and 523. Further, the generation time information 712 of the file management information section 710 includes date information “88/01/01” as the time when the item register 500 was generated or the structure was changed, as shown in FIG. It is stored.

FIG. 6 shows the file format of the address book 600. The data record part 730 of the address book 600 is shown in FIG. 6 (c).
As shown in, as the item, specify the item, name 601, address
602 and telephone number 603. Item management information section 720
Is the item management information 621, 622, 623 of each of the above-mentioned name 601, address 602 and telephone number 503, as shown in FIG.
Is stored. Further, the generation time information 712 of the file management information section 710 stores date information “88/12/14” as the time when the address book 600 is generated or the structure is changed, as shown in FIG. Has been done. In the data record section 730 of the address book 600 shown in FIG. 6 (c), records having three different formats are mixed,
This will be described later along with the operation.

The DBMS 200 shown in FIG. 2 operates the data stored in the time information comparison unit 210 and the file management unit 220 used to change the management information unit 700 of the file, and the above-mentioned dictionary file 310 and the data file 320. And a data operation unit 230 that operates.

The operation of the DB system of the present embodiment configured as described above,
In the following, description will be given according to the flow of processing shown in the operation flowchart shown in FIG.

The DBMS 200 receives the command 100 requesting the operation of the file in the DB 300 (step 910). This command 100 includes the type of operation such as data addition, deletion, or update,
A file to be operated (hereinafter referred to as “target file”) and data used for the operation are specified.

The time information comparison unit 210 in the DBMS 200 determines the file definition change according to the following procedure (step 920) and activates the file management unit 220. First, the file definition time information 402 regarding the target file is read from the file register 400 (step 921). Next, the generation time information 712 of the target file is read from the file management information section 710 in the target file. If the target file has not been generated, the file management unit 220 is activated (step 92).
2).

The file management unit 220 reads the data record, which is the definition information of the target file stored in each dictionary file, and generates the target file according to the data record (step 923). On the other hand, when the generation time information 712 of the target file is read in step 922, the definition time information 402 of the target file is compared with the generation time information 712,
It is determined whether the definition of the target file has been changed after the generation or structure change of the target file. After the generation or structure change of the target file,
If the definition of the target file has been changed, the file management unit 220 is activated (step 924). The file management unit 220 replaces the management information unit 700 of the target file with
It is changed according to the definition information about the target file in the dictionary file 310 (step 925).

The data operation unit 230 in the DBMS 200 uses the following procedure (step
930), data operation is performed on the target file. First, for the data record part of the target file,
The operation such as addition, deletion or update specified by the operation command 100 is executed (step 931). Next, the data operation unit 230 determines whether the target file is a dictionary file or a data file (step 932). If the result of the determination is that the target file is a dictionary file, the operation by the operation command 100 is a change in the definition information of the file specified by the data to be operated, and the definition time information update processing (step 933) is started. To do. In this processing, the definition time information 402 in the file register 400 of the file whose definition has been changed by the operation command 100 is updated.

Through the above processing, the result of executing the operation command 100 is returned (step 940).

As described above, in the present embodiment, the dictionary file 310 and the data file 320 shown in FIG. 2 are treated with the same physical structure with the file structure and format shown in FIG. Further, a time information comparison unit 210 and a file management unit 220 are provided in the DBMS 200 shown in FIG. As a result, according to the present embodiment, there is an effect that the DBMS itself can automatically generate or change the structure of the file from the definition information of the file accumulated by the data operation, if necessary. Further, by identifying the value of each item by the identifiers 722 and 732 shown in FIGS. 3C and 3D, it is not necessary to reorganize the data of the data record part 730 due to the structure change of the file. There is an effect.

7 to 9 show the states of the respective files when the address book 600 is generated and the structure is changed as the data file 320 in the DB 300 in the system configuration as described above. (A), (b), and (c) of each figure respectively show the state before the operation, the state after the definition operation, and the state after the data operation.

Figure 7 shows an address book 600 as a data file 320 in DB300.
And the state of each file when the operation of inserting the data record 631 into the generated address book 600 is performed.

In this operation, first, as the operation of generating the address book 600, the data record 433 that is the file definition information of the address book 600 is inserted into the file register 400, and then the item register 5
The data records 536 and 537 which are the item definition information of each of the name 601 which is the item of the address book 600 and the address 602 are inserted into 00. Next, as a data operation to the address book 600, the data record 631 of the address book 600 is inserted.

FIG. 8 shows the state of each file when an operation of adding a telephone number 603, which is a new item, to the address book 600 is performed.

In this operation, first, as a definition changing operation of the address book 600, a data record 538 which is the item definition information of the telephone number 603 which is an additional item of the address book 600 is inserted into the item register 500. Next, as a data operation to the address book 600, the data record 632 of the address book 600 is inserted.

FIG. 9 shows the state of each file when the operation of inserting the data record 633 is performed in the address book 600 whose definition is changed by deleting the item from the address book 600 described above.

In this operation, first, as a definition changing operation of the address book 600, the data record 537 which is the item definition information of the address 602 which is the deletion item of the address book 600 is deleted from the item register 500. Next, as a data operation to the address book 600, the address book 600
Insert data record 633 of.

The definition operation of each operation described above is a data operation to the dictionary file 310 described above. Therefore, the file directory 400 is processed by the processing of steps 932 and 933 in FIG.
The definition time information 402 of the definition information of the address book 600 stored in the data record section 730 of the address book 600 is updated, but when the definition operation is performed, the address book 600 is not generated or the structure is changed. Absent.

According to the above-described embodiment, the necessity of activating the change processing of the management information part in the file is determined by comparing the two pieces of time information in step 920 of FIG. Is good. This allows DB
Operability during operation is improved. Further, in the above embodiment, since the file definition operation is replaced by the data operation of the data defining the structure, the command for the definition operation is unnecessary. This facilitates the user interface. Further, in the above embodiment, the conventional DBMS
The data reorganization processing and definition command analysis processing, which accounted for most of the definition processing by
There is also an effect that the size of the DBMS is significantly reduced because it is not needed inside.

By the above operations, the data record section 730 of the address book 600 contains records having three different formats as shown in FIG. 6 (c). However, according to the above-described embodiment, the value of a desired item can be stored and retrieved based on the item identifier added to the value of each item and the length of the item value. This allows
In the data manipulation shown in FIGS. 8 and 9, the address book 60
Since the definition operation is performed after 0 is generated or the structure is changed, the structure of the address book 600 is changed by the processing of steps 924 and 925 in FIG. 1, but the file management information part 710 of the address book 600 is changed. And only the item management information section 720 is updated, and the data reorganization of the already stored data record is not performed.

〔The invention's effect〕

As described above in detail, according to the present invention, the responsiveness at the time of file definition operation is improved, the number of commands is reduced to facilitate the user interface, and the scale of the database management system. This has the remarkable effect of realizing a database management method that can reduce

[Brief description of drawings]

FIG. 1 is a flow chart showing the operation of each part in a DBMS that is an embodiment of the present invention, FIG. 2 is a block diagram of the DB system of the embodiment, and FIG. 3 is a logical structure of each file and the data format of each part. FIG. 4, FIG. 4 to FIG. 6 are diagrams showing a detailed structure of each file, FIG. 7 to FIG. 9 are diagrams showing a file state showing a concrete operation of the embodiment, and FIG. FIG. 11 is a diagram showing the configuration of a system according to FIG. 11, FIG. 11 is a diagram showing objects and operations in SQL DDL and DML commands, and FIG. 12 is a diagram showing a configuration example of a data distribution type distributed processing system. 200: DBMS, 210: Time information comparison section, 220: File management section,
230: Data operation part, 300: DB, 310: Dictionary file, 320: Data file, 400: File directory, 402: Definition time information, 500: Item directory, 600: Address book, 712: Generation time information, 732 : Item identifier, 733: Item value length, 734:
The value of the item.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Kazuyuki Yasuda 1-16-5 Dogenzaka, Shibuya-ku, Tokyo Within Japan Business Consultant Co., Ltd. (56) References JP-A-61-279932 (JP, A) JP-A-SHO 61-82251 (JP, A) JP-A-63-145551 (JP, A) JP-A-63-276640 (JP, A) JIII Journal of Technical Disclosure 83-2543 Parki Minoru "Database / Introduction and Design" (SHO 50-9-5) Planning Center PP. 189-195 "ACOS Software ACOS-4 / MVP XE ADBS Overview DFC 71-3" (1985-6.) NEC PP. 88-89

Claims (2)

[Claims] 1. A database including a data file for storing data, a dictionary file for storing data defining a logical structure of the file, and a data manipulating means for accessing the data in the data file and the dictionary file are managed. In the database management system, the data file and the dictionary file have the same physical structure, in the dictionary file, the definition time information storage means for storing the time when the logical structure of the data file is defined or definition changed, A generation time information storage means for storing the time when the data file is generated or the structure is changed is provided, and the generation or the structure change of the data file is performed from the data defining the logical structure of the file stored in the dictionary file. A data file generating means for performing the above is provided, and when the data file is accessed by the data operating means, the definition time information in the storage means is compared with the generation time information of the target file, and the result of the comparison is obtained. A database management method characterized in that the data file generating means is activated accordingly. 2. Each item in each record in the file comprises an item identifier, an item value length and an item value, and in the data operating means, the item identifier and item value length. 2. The database management system according to claim 1, wherein the value of each item is operated by referring to the value.