JPH06337762A - Compressing method and restoring method for data base record - Google Patents

Abstract

PURPOSE:To shorten the length of data base records and reduce the capacity by compressing data of all initial value data item parts when there are many initial value data items in the data base records. CONSTITUTION:When record data set in a record receiving and passing area 16 secured in an application program 11 are exclusively 0Red by using a mask table, all data item parts having the same values between the record data and mask table are subtracted to become zero and then the contents of masked record data in the record receiving and passing area 16 are left. On the basis of the masked record data, compression control information is substituted for parts of successive identical characters, noncompression control information is set even at the head of discontinuous parts, and compressed record data are set in a record input/output area 17 and stored in a data base file 19.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention mainly relates to a method of compressing and restoring a database record in a database management system.

[0002]

2. Description of the Related Art In the conventional data compression method, when the same character is consecutive for 3 bytes or more in a database record, the number of repeated characters is counted and the character type and the number of repetitions are summed up by 1 byte each. The 2-byte compression information is replaced with the position where the same character is present in the record, and the database record is compressed.

For example, Japanese Unexamined Patent Publication No. Hei.
-348617 publication etc. are mentioned.

[0004]

However, in the above-mentioned prior art, since the data compression is performed only when the same character is consecutive for 3 bytes or more in the database record, the initial value data item in the database record. When the continuity of the same character is interrupted by adding a 4-bit code to the inside, there is a problem that the compression efficiency cannot be improved.

An object of the present invention is to improve the compression efficiency of a database record and shorten the database record length stored in a database file even when the continuity of the same characters is interrupted in the data of the database record.

[0006]

In order to achieve the above object, the first means of the present invention is to create a mask table set with an initial value according to the data format in a database record to create one database. The exclusive OR of the data of the database record stored in record units and the mask table is taken, and the bit corresponding to the initial value item part in the database record is turned off to make a continuous identical character string, and the database record Of the database record consisting of the data and the compression information of the remaining database records that did not bit-off the record data by replacing the compression information having the information about the continuous identical character string at the position where the continuous identical character string exists in the data of It is characterized by being compressed into data.

A second means of the present invention reads a compressed database record once stored in a database file on an external storage device and replaces it with the compressed information in the database record read from this database file. To the length of consecutive same characters,
It is characterized by performing exclusive OR with the same mask table as when the data of the database record is compressed and turning on the corresponding bit to restore the database record to the database record before data compression.

The mask table is created by using the source program of the record format defining part used in creating the application program for processing the database record as it is, based on the initial value data according to the language. To do.

[0009]

According to the above-mentioned means, the mask table set with the initial value according to the data format in the database record or an arbitrary character string with high occurrence frequency is created, and 1
The exclusive OR of the data of the database record stored in database record units and the mask table is taken, and the bit corresponding to the initial value item part in the database record is turned off to form a continuous identical character string,
The compressed information having the information about the continuous identical character string is replaced at a position where the continuous identical character string is present in the data of the database record, and the record data is composed of the data of the remaining database record and the compressed information which are not bit-off. Since the data is compressed into the data of the database record, the compressible range can be made wider than before, the compression efficiency of the record data can be improved, and the length of the database record stored in the database file can be shortened.

When restoring the database record before compression, the compressed database record once stored in the database file on the external storage device is read and replaced with the compression information in the database record read from the database file. The extended part of the specified part is extended to the length of the continuous same character, and the corresponding bit is turned on by exclusive-oring with the same mask table as when the database record data was compressed, and the database record before data compression of the database record is turned on. It is possible to easily restore to.

Further, since the mask table is created by using the source program of the record format defining part used in creating the application program for processing the database record as it is, based on the initial value data according to the language, The gap between the application program that processes the database record and the database file can be prevented.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described in detail below with reference to the drawings.

FIG. 1 is a block diagram showing the overall configuration of a database management system for implementing the present invention, and FIG. 2 is a block diagram showing the hardware configuration of an apparatus for implementing the database management system of FIG. is there.

In FIGS. 1 and 2, 10 is a database management system, 11 is an application program, 19 is a database file, 20 is a record definition management file, 21 is a compiler, 22 is a mask table file, 31 is an input / output device, and 32 is a 32. Is a central processing unit (CPU), 33 is a main memory, and 34 is an external memory.

As shown in FIG. 1, the database management system 10 for executing the data compression method according to the embodiment of the present invention operates according to the input / output request of the database record of the database file 19 used by the application program 11 to perform the database management. A controller 12 for controlling the inside of the system, a data compressing unit 13 for compressing database records, a data restoring unit 14 for restoring database records, and a mask table loading unit 15 for inputting a mask table file 22.

The mask table file 2 to be used
2 is created by the compiler 21 based on the record regulation management file 20, and the file contents are made resident in the mask table resident area 18 by the mask table loading unit 15 at the start of the database management system 10.

FIG. 3 shows an example of the record format of the database record and the record format set in the record regulation management file 20 that regulates the data format, which is input to the compiler 21 and converted into object format data. , The mask table 23 is shown in the mask table file 22.

In the example of FIG. 3, the compiler uses the format 00 ... 0C for the internal decimal format (pack format decimal data) 20a and the external decimal format (zone format 1).
For 0b data) 20b, in the format of F0F0 ...
4040 for the character format (character string data) 20c
The mask table is set in the format A1A1 ... For the kanji format (Japanese string data) 20d in the format.

In this way, the mask table can be easily created by inputting the record format of the database record and the record format set in the record regulation management file 20 that regulates the data format to the compiler.

Further, by using the source program of the record format defining part used for creating the application program for processing the database record as it is, and setting the initial value according to the language, the application program and the database for processing the database record are set. You can prevent the files from shifting.

FIG. 4 shows an application program 11
This is for explaining the process of data compression by using an example of a database record (hereinafter abbreviated as a record) requested to be stored by the.

The exclusive OR of the record data 16a set in the record passing area 16 secured in the application program 11 by the mask table 23 is the same between the record data 16a and the mask table 23. All data item parts having a value are subtracted to give a value of zero, and the result becomes the contents of the masked record data 16b in the record passing area 16.

Then, the masked record data 1
Based on 6b, the consecutive same character parts are replaced with compression control information (character and length), and non-compression control information (length) is set at the beginning of the discontinuous part. The compressed record data 17a is set in 17, and the compressed record data 17a is stored in the database file 19.

Next, FIG. 5 reads the compressed record data 17a stored in the database file 19 by using the example of the record used in the description of FIG. 4, and replaces the data compression portion with the contents of the original record data 16a. This is for explaining the process of restoring and passing it to the application program 11.

Compressed record data 1 in the record input / output area 17 read from the database file 19
Based on the compression control information and the non-compression control information of 7a, in the case of continuous identical character compression, the characters in the compression control information are expanded to that length, and in other cases, the length in the non-compression control information is expanded. The uncompressed data portion corresponding to the size is set in the record passing area 16 as it is.

That is, the compressed record data 17a is set in the record delivery area 16 as the converted record data 16c converted from the compression control information and the non-compression information. The converted record data 16c at this time has the same data content as the masked record data 16b.

When the conversion table data 16c is subjected to exclusive OR with the mask table 23 used at the time of storage, the values of the mask table are added to all the zero values of the conversion record data 16c. , The result is the restored record data 16d in the record passing area 16
As shown in (4), the contents of the record data 16a before compression are restored, and the restored record data 16d is passed to the application program 11.

FIG. 6 is a flow chart showing a processing procedure of data compression of this embodiment, which will be described with reference to FIG.

The corresponding mask table 23 is determined based on the contents of the record data 16a in the record delivery area 16 requested to be stored by the application program 11 (step 100). Using the mask table 23 determined here, the exclusive OR of 256 bytes is executed by one instruction for the contents of the record data 16a in the record passing area 16 (step 110).

Record passing area 1 of this execution result
Based on the record data 16b of No. 6, the continuous identical character is searched for 256 bytes each by one command (step 120). When the continuous identical character of 2 bytes or more is detected (step 1
30), the length of the continuous portion of the same character is obtained (step 140), compression control information is created based on this character and the length, and set in the record input / output area 17 (step 15).
0)

If there is no continuous identical character (step 13
0), non-compression control information is created based on the length of the discontinuous character portion, and the non-compressed data portion is recorded in the record input / output area 17
(Step 160).

Next, steps 120 to 160
The above steps are repeated until the search for the masked record data 16b is completed. At the end of the search, the compressed record data 17a in the record input / output area is set, the contents are stored in the database file 19 (step 170), and the process is returned to the request source. .

FIG. 7 is a flow chart showing a processing procedure of data restoration of the present embodiment, which will be described with reference to FIG.

Based on a database record read request from the application program 11, the compressed record data 17a is read from the database file 19 to the record input / output area 17 (step 200), and the corresponding mask table is determined based on this content. (Step 210), the compression control information or the non-compression control information is searched from the beginning of the compressed record data 17a in the record input / output area 17 (step 220). In the case of compression control information (step 230), the compression control is performed. The characters in the information are expanded to that length and set in the record passing area 16 (step 240).

In the case of non-compression control information (step 23
0), the uncompressed data portion of the length in the uncompressed control information is set as it is in the record passing area 16.

The compressed record data 17a in the record input / output area 17 is executed from step 220 to step 250.
Is repeated until the end, and at the end of the search, an exclusive OR of 256 bytes is executed by one instruction using the mask table 23 previously determined for the converted record data 16c in the record passing area 16 (step 260). .
The result of this execution is the restored record data 16d in the record passing area 16.

Then, the restored record data 16d
Is the same as the original record data 16a, and when returning to the request source, the restored record data 16d is passed to the request source.

As described above, according to the present invention, in a large-scale database system in which the number of records is large and the number of data items in a record is large, a large number of initial value data items such as internal decimal format are included in a database record. If it exists, data compression is performed on all of these data item portions, so the database record length is greatly reduced, and the external storage device capacity for storing the database file is significantly reduced.

Therefore, the data transfer time at the time of storing and reading the database record can be greatly reduced.

When the compressed database record is restored to the database record before being compressed, it can be easily restored by reversing the processing procedure performed in the record compression.

Further, by using the source program of the record format defining part used for creating the application program for processing the database record as it is and setting the initial value according to the language, the application program and the database for processing the database record are set. You can prevent the files from shifting.

[0042]

As described above, according to the present invention,
In a large-scale database system with a large number of records and a large number of data items in a record, if there are many initial value data items such as internal decimal format in the database record, all of these data item parts are Because of the compression, the database record length is greatly reduced, and the external storage capacity for storing the database file is also significantly reduced.

Further, the data transfer time at the time of storing and reading the database record is also greatly reduced.

[Brief description of drawings]

FIG. 1 is a block diagram showing the overall configuration of a database management system that realizes the present invention.

FIG. 2 is a block diagram showing a hardware configuration for implementing the present invention.

FIG. 3 is a diagram for explaining a mask table creation procedure of the present embodiment.

FIG. 4 is a diagram for explaining a process of compressing data according to the present embodiment.

FIG. 5 is a diagram for explaining a process of restoring data according to the present embodiment.

FIG. 6 is a flowchart showing a processing procedure of data compression of the present embodiment.

FIG. 7 is a flowchart showing a processing procedure of data restoration of the present embodiment.

[Explanation of symbols]

 10 ... Database management system, 11 ... Application program, 12 ... Controller, 13 ... Data compression section, 14 ... Data decompression section, 15 ... Mask table loading section, 16 ... Record passing area, 16a ... Record data, 16b ... Masked record Data, 16c ... Converted record data, 16d ... Decompressed record data, 17 ... Record input / output area, 17a ... Compressed record data, 18 ... Mask table resident area, 19 ... Database file, 20 ... Record regulation management file, 21 ... Compiler, 22 ... Mask table file, 23 ... Mask table. 31 ... Input / output device 32 ... Central processing unit 33 ... Main storage device 34 ... External storage device

(72) Inventor Yoshifumi Kawasaki 6-81 Onoe-cho, Naka-ku, Yokohama-shi, Kanagawa Hitachi Software Engineering Co., Ltd. (72) Masahisa Horie 6-81 Onoe-cho, Naka-ku, Yokohama, Hitachi Hitachi Software Engineering Stock Company In-house (72) Inventor Yuji Toyama 6-81 Onoue-cho, Naka-ku, Yokohama-shi, Kanagawa Hitachi Software Engineering Stock Company In-house (72) Inori Takenori Iwato 5030 Totsuka-cho, Totsuka-ku Yokohama-shi Hitachi, Ltd. (72) Inventor Takafumi Nogami 5030 Totsuka-cho, Totsuka-ku, Yokohama-shi, Kanagawa Incorporated company Hitachi Ltd. Software Development Headquarters

Claims (3)

[Claims] 1. A method of compressing a database record in a database management system, wherein a mask table is created for a database record having a fixed record format with an initial value according to a data format in the database record. Then, the exclusive OR of the data of the database record stored in the unit of one database record and the mask table is taken, and the bit corresponding to the initial value item part in the database record is turned off to make a continuous identical character string. , Replacing the compression information having the information on the continuous identical character string at the position where the continuous identical character string is present in the data of the database record, and compressing the data of the remaining database record that did not bit off the data of the database record And A method of compressing a database record, characterized by compressing the data of a database record consisting of 2. The method for compressing a database record according to claim 1, wherein a compressed database record is read from a database file on an external storage device in which the compressed database record is stored and read from this database file. The same character is continuously replaced based on the compression information at the position where it was replaced with the compression information in the database record, and the corresponding bit is turned on by performing an exclusive OR with the same mask table as when compressing the data in the database record. , A method for restoring a database record, characterized by restoring the database record before data compression. 3. The method of compressing or decompressing a database record according to claim 1 or 2,
The database is characterized in that the mask table is created based on the initial value data according to the language by using the source program of the record format defining part used in creating the application program for processing the database record as it is. How to compress or decompress records.