JPH0423143A - Data storing system - Google Patents

Abstract

PURPOSE:To store the large capacity data of the variable lengths with high efficiency by dividing the data of the variable lengths into plural blocks of the fixed length and storing these blocks in a storage after adding the block control parts including the same identifiers to the blocks. CONSTITUTION:A data ID 5 is added to the large capacity data 11 of the variable lengths and the data 11 is divided into blocks A - D of the fixed length. The block control tables 12 containing the same data IDs are added to the block A (B - D). These blocks A - D are successively stored in a buffer set at the fixed length. Then the blocks A - D are successively read out of the buffer and stored in a magnetic disk device 4 in a storing state. An address control table is provided to a main memory device 2, and only the ID 5 and the address of a 1st block A are set on the address control table. As a result, the data 11 is stored and controlled with high efficiency and by means of the limited resources.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a data storage method for storing large-capacity variable-length data such as image information in a form that allows random access.

[Conventional technology]

Usually, a large amount of image data input from an external input device such as a CCD camera is stored as a file in a magnetic disk device. FIG. 5 shows the hardware configuration of a general computer system, where 1 is a central processing unit, 2 is a main memory, 3 is an external input device such as a CCD camera, and 4 is a magnetic disk device.

In conventional computer systems, when storing large-capacity variable-length pigs in the magnetic disk device 4, the procedure shown in FIG. 6 is used.

First, a buffer 21 of a size corresponding to the maximum length of each piece of data 11 to be stored is prepared on the main memory device 2.

Next, the data 11 read from the external input device 3 is expanded onto the buffer 21. The data on the buffer 21 is then stored in a direct organization file 22 in a variable length record format defined in the magnetic disk device 4. This file 22 defines fixed partitions according to the maximum length of the data 11. The storage address 24 (record number) at that time is stored in the management table 23 on the main memory 2 together with the data ID 25. When searching data, the data ID
25 to storage address 2 using this management table 23.
4 (record number), the data 11 can be called from the magnetic disk device 4. Defining the file 22 in the magnetic disk device 4 into fixed-length sections is convenient because the section size does not have to match the data Il length.

[Problem to be solved by the invention]

As explained above, in conventional data storage methods, it is necessary to prepare a buffer on the main memory device corresponding to the maximum length of the data to be stored, and if the data is short, it is necessary to prepare a buffer in the main memory device. There were problems such as empty spaces sometimes occurring, making it unsuitable for storing large amounts of data.

This invention was made to solve the above-mentioned problems, and by dividing large amounts of data into fixed-length blocks and storing them in main memory buffers and magnetic disk drives, variable-length large amounts of data can be stored. The purpose of this invention is to obtain a data storage method that efficiently stores capacity data and allows random access.

[Means to solve the problem]

In this invention, in a data storage method in which variable-length data 11 input from an external input device 1 is stored in a storage device 4 according to an address management table 13, variable-length data 11 is stored in a plurality of fixed-length blocks (A -D), add a block control unit 12 containing the same identifier 5 to these blocks A to D, record the identifier 5 and address 13 of the first block A in the address management table 20, and Blocks A to D were stored in the storage device 4.

[Effect]

The address 13 of the first block A of the desired data 11 is obtained from the address management table 20 based on the identifier 5. According to this address 13, the block controller 12 and block A of data 11 are obtained from the storage device 4.

Next, based on this block control unit 12, address 8 for the next block B is obtained and written jrf! The next block control unit 12 and the next block B are obtained from the device 4. All data were obtained in the same manner.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings. In the present invention, as shown in the computer system of FIG. 5, a central processing unit (CPU) 1. Main memory device 2. External input device for inputting data 3. The magnetic disk device 4 for storing variable length pigs uses a general configuration.

In FIG. 2, variable-length large-capacity data 11 obtained from an external input device 3 is assigned a data ID by the CPU on the main memory device 2, and then fixed-length data block A
, B, C, and D.

Next, a block control section 12 as shown in FIG. 1 is added to each block A (B-D) by the CPU. In FIG. 1, 5 is the data ID of the identifier of each data block 11, 6 is the serial number indicating what block it is, 7 is the number of the next block, 8 is the storage address of the next block, 9
indicates the size of each block data, and 10 indicates the divided block data A to D.

As shown in FIG. 3, blocks A to D created by dividing the same data 11 have the same data ID, and are each equipped with the block control section 12 shown in FIG. Each block A to D becomes a fixed length first to fourth blocks 19 to which a block control section 12 is added, and are sequentially stored in two buffers set to a fixed length. Now the next block (A-
The next block number serves as a pointer to the next block (B-C-D, etc.). During storage, blocks A, B, and CD are read from the buffer 21 in this order and stored in the magnetic disk device 4. Finally, an address management table 20 is provided on the main memory device 2 as shown in FIG.
and blocks A and B.

The storage is completed by setting only the address 13 of the first block 16 subdivided into C and D.

When reading data, first read the address management table 20.
The data ID 5 of the more necessary data and the storage address 13 of the first block 16 are obtained. The corresponding block 16 is read from the magnetic disk device 4 to obtain data A. At this time, since the storage address 8 of the second block 17 is set in the first block 16, the second block 17 is read from the storage address 8. By repeating this step by step, it is possible to read out up to the fourth block 19 in a step-by-step manner.

Since null data is set in the next block number of the fourth block 19, this causes CPUI to write data 11.
You can know the end of the block.

Although the above example shows the case of accumulating data obtained from an external input device, the same effect can be obtained when accumulating large-capacity variable length data of calculation results automatically generated by another program. .

〔Effect of the invention〕

As explained above, according to the present invention, variable length data is divided into a plurality of fixed length blocks, a block control section containing the same identifier is added to these blocks, and the identifier and address of the first block are added. Since the data is recorded in the address management table and multiple blocks are stored in the storage device, large amounts of variable length data such as image information can be efficiently stored and managed using limited resources.

Another advantage is that any data can be quickly searched. Also, the size of the main memory buffer used can be minimized.

[Brief explanation of the drawing]

Fig. 1 is a diagram showing the configuration of blocks in which data of the present invention is subdivided; Fig. 2 is a diagram showing how data is divided into blocks; Fig. 3 is a diagram showing how divided data is related in the order of blocks; Figure 4 is a diagram showing the file structure when storing data according to the present invention, Figure 5 is a diagram showing the configuration of a general computer system, and Figure 6 is a diagram showing the file configuration when storing conventional data...External input Device, 4... magnetic disk device,
5...Data ID, 6...Block serial number, 7...
...Next block number, 8...Data start address in next block, 11...Data, 12...Block management table, 13...Block address, A, B, C,
D...Block.

Claims (1)

[Claims] In a data storage method in which variable-length data input from an external input device is stored in a storage device according to an address management table, the variable-length data is divided into a plurality of fixed-length blocks, and each block has the same identifier. The data storage method is characterized in that the identifier and address of the first block are recorded in the address management table, and the plurality of blocks are stored in the storage device.