JPS6249438A - Data buffering method for shared memory device retrieving system - Google Patents

Abstract

PURPOSE:To raise utilization efficiency, and to make the titled system have a quick responsiveness to an operator by dividing a buffer for storing variable length data which has been provided in each retrieval device, into group spaces whose capacity is variable, executing store/read-out of the variable length data at every space, and executing the store/read-out simultaneously in the lump. CONSTITUTION:When a new retrieval code is inputted from an operating part 11, index management information 32 is stored in a B buffer managing memory 24, an idle buffer register 19 is updated, and the data are read into image buffers I-8,-of a prescribed fixed length space 31 of an image accumulating buffer 25. When a display instruction of the image data of the index management information 32 is executed from the operating part 11, the image data are developed to a display use memory 21 by an accumulation checking circuit 17 and an expanding circuit 16, displayed on a display device 15, also fixed length spaces 31(I-1-I-7) which have been allocated to an A buffer managing memory 23 are released, and the idle buffer register 19 is updated. Thereafter, this operation is repeated alternately by the A buffer managing memory 23 and the B buffer managing memory 24.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a data buffering method for a shared storage device equipped with a plurality of search devices, and particularly to a data buffering method suitable for efficiently utilizing the shared storage device. The present invention relates to a data buffering method for a shared storage device retrieval system.

[Background of the invention]

As a data buffering method for a conventional shared storage device retrieval system, a double buffer method using two buffers is known. This dasol puffer method is
Two fixed-length buffers capable of storing a plurality of records are provided, and the two buffers are activated to read a fixed number of records alternately. This allows a fixed number of records stored in one buffer to be displayed on a display within the retrieval device while the other buffer can store a fixed number of new records from the shared storage. Therefore, it is possible to improve the responsiveness to the operator, and the shared storage device can be used efficiently.

However, in the double-gutter method described above, when buffering data in which the length of one record is variable, such as image information, if the capacity of one buffer is fixed, it is difficult to read data into one buffer. The number of records P that can be stored becomes undefined, and the response time for the operator becomes undefined. Also, there is one 7-7 code.
If the buffer is so long that it exceeds the capacity of the buffer, it may become impossible to read it even though there is empty space in the other buffer.

Furthermore, when a shared storage device is accessed by multiple search devices, it is possible to reduce the time occupied by one search device, as shown in the invention disclosed in Japanese Patent Application Laid-open No. 56-149657. , there is a method that enables efficient sharing.

However, recently developed OLU (Qptical
(Library Unit), it requires loading to select one of multiple optical discs and setting it in the optical disc drive, and unloading to remove the optical disc from the optical disc drive and store it in a predetermined position. For shared storage devices that require a large amount of time,
In order to shorten the time occupied by each search device, if the amount of information that can be read in one search (access unit) is reduced, the number of accesses will increase, and time will be wasted on loading and unloading. Efficient access is not possible.

[Purpose of the invention]

The present invention was devised in view of the problems of the prior art described above, and it is possible to reduce the number of accesses to the shared storage device and shorten the access time as a whole in a shared storage device that handles variable tone records. The object of the present invention is to provide a data buffering method for a shared storage device (retrieval system) that increases device utilization efficiency and provides coordination with operators.

[Summary of the invention]

The data buffering method of the shared storage device search system of the present invention provides a shared storage device that requires preparation such as loading and unloading before starting access, and accessing the shared storage device and storing information stored in the shared storage device. Each search device consists of a plurality of search devices that read and display variable length data.

The empty area of the buffer for storing the variable length data provided in It is characterized by storing long data and reading and displaying the stored variable length data.

That is, the present invention divides a buffer for storing variable-length data provided in each search device into a plurality of group spaces with variable capacity, and stores and reads variable-length data for each group space (using a shared storage device). This method reduces the number of accesses by storing and reading data all at once.

[Embodiments of the invention]

Hereinafter, the present invention will be explained in more detail with reference to embodiments shown in the accompanying drawings.

Figure 1 shows a shared storage device F and three image search devices R-1,
FIG. 2 is a block diagram showing the configuration of a shared storage device search system composed of R-2 and R-3.

Here, the above-mentioned OLU is assumed as the shared storage device F. An input request from the image retrieval device R-1 is accepted by the storage device control unit 1, and it is checked whether the corresponding optical disc is usable or not based on the contents of the optical disc old management nosta 2 and the exclusive control circuit 3. . If the corresponding optical disk is usable, if it has not yet been loaded, the data is loaded into the optical disk loading unit 4, and if it has already been loaded, the data is read out from the optical disk input/output unit 6. Image retrieval device R-1 Data is sent to. On the other hand, if it is unusable (the fine image retrieval devices R-2 and R-3 are accessing), it is recorded in the optical disc occupancy management register 2, and the system waits until it becomes usable.

That is, in the shared recording device F, the image search devices R-1,
Process requests from R-2 and R-3 in the order of arrival.

FIG. 2 is a diagram showing specific examples of image search devices R-1, R-2, and R-3. The operator operates the operation unit 11 and inputs the search code. The image search device control unit I2 reads index information corresponding to the input search code from the storage device F, and sends the search title Stored in memory 22. Search title memory 22 can store up to 200 pieces of index information;
can.

FIG. 3 is an explanatory diagram showing the index information 30 stored in the search title storage memory 22, and the index information 30 includes title names (T-1, T-2) as shown in the figure.
) and the recording position of the image data (A-1, , A-2
) and the image data size (S-1, S-2...)
).

FIG. 4 is an explanatory diagram showing the contents of the image storage/muffer 25, which is divided into a large number of fixed length spaces 31 as shown. One fixed length space 31 has accumulated marks (M-1
, M-2...), next fixed-length space pointer (P-1, P
-2...), effective data length (L-1°L-2...)
, and image buffers (I-1, I-2...). The image storage buffer 5 is initially empty of all fixed-length space images I-1, I-2, etc., and the empty buffer register 19 is empty as shown in the figure. Point to the fixed length space 31 in the first line, and select the next fixed length space pointer L-1 in each fixed length space 31.
, L-2, . . . each point to the next fixed length space 31. Here, the empty buffer register 19 points to the first one of the empty fixed length spaces 31 in the image storage buffer 25, and the next fixed length space pointer L- of each empty fixed length space 31. 1, L-2... inds the next empty fixed length space 31 by 2. by this,
A chain of empty fixed length spaces 31 is formed.

When displaying an image of index information corresponding to a search code input by the operator using the operation unit 11, if the operator instructs the operator to display the first image from the operation unit 11, the current displayed title position A title name TI is set. Now, if the search code entered by the operator corresponds to the three index information 30,
The double buffer control unit 18 stores the images 7'' in these three index information 30 - rs-1, S-2, S-
3, the number of fixed length spaces 31 required to store each image data is calculated, and the required number is allocated for the A buffer. As shown in FIG. 5, the A buffer management memory 23 stores the above three index information 30 as E-1, E-2,
E, -3, and each index information E-1, E-
Chain pointers C-1, C-2 and C-3 indicating the beginning of the fixed length space 31 corresponding to C-2 and E-3 are stored as index management information 32. At this time, the contents of the empty buffer register are updated to point to the first empty fixed length space 31, as shown in FIG.

Based on the information set in the management memory 23, the image search device control unit 12 requests the shared storage device F to read the image into the image storage buffer 25. Image/Guits 7
When data is read into A I-1, the effective data length L-1
is set, and the stored mark M-1 is turned on. In parallel with the operation of repeating this to accumulate the image data up to the image buffer I-7, the accumulation check circuit 17
The stored marks M-1 to M-7 are checked, and when an on state is detected, the compressed data is expanded into the display memory 21 by the expansion circuit 16, and the image data is displayed on the display device 15.

Next, when the operator inputs a new search code from the operation unit 11, the index management information 32 is stored in the B buffer management memory 24 in exactly the same way as described above.
The empty buffer nosta 19 is updated, and the image buffer I-8 of the predetermined fixed length space 31 of the image storage/gunofer 25 is updated.
The data is loaded into... Then, from the operating section 11, press 73 A sofa management memory 24.
When an instruction is given to display image data of index management information 32 belonging to 1, the image data is developed in display memory 21 by accumulation check circuit 17 and decompression circuit 16 and displayed on display device 15 in the same way as in the above method. With,
A/? The fixed length space 31 (I-1 to -7) allocated to the management memory 23 is released, and the empty buffer register 19 is updated.

Thereafter, this operation is repeated alternately in the A-buffer management memory 23 and the B-buffer management memory 24.

In the above explanation, it was assumed that the titles are displayed sequentially from the first title (that is, the index information is displayed in the order of E-1, E-2, and E-3), but the present invention is limited to this. The present invention can also be applied when the operator uses the operation unit 11 to request the display of an image that is different from the currently displayed image. For example, when data corresponding to index information E-1 is being displayed and data corresponding to index information E-10 is to be displayed next, the double buffer control unit 18 controls the image storage buffer 25. From the empty fixed length space 31,
A fixed length space 31 is secured for the requested index information E-10, index information E-11 for the next 4-page request, and index information E-9 for the previous page request, and the memory 23 for A buffer management or the memory 23 for B buffer management is secured. Set the memory 24. Thereafter, as shown in FIG. 6, the access order rearrangement calculation section 20 rearranges the readout order in the order of E-10°11.9 and starts reading out the images.

The subsequent operations are the same as in the case of sequential display. By accumulating image data in this manner, the image requested by the operator can be displayed immediately, and images of the next page and the previous page can be accumulated one after another.

Further, in the embodiment described above, the image storage buffer 25 is divided into two group spaces, and each is managed by the A buffer management memory 23 and the B buffer management memory, but the number of group spaces is 3,4
It can also be increased.

Also, when the image to be displayed moves from the index information managed by the A buffer management memory 23 to the index information managed by the B sofa management memo JJ 24,
I immediately released the fixed length space 31 managed by the A buffer management memory 23 and started the next read operation, but
In order to improve the response to previous page operations, the page is moved to the fixed length space 31 managed by the B buffer management memory 24, and the fixed length space 31 managed by the A buffer management memory is not released for several pages. It can also hold image data.

According to the embodiment described above, as the storage device F, the OL
When searching and displaying image data stored in the OLU from multiple image search devices R-1゜R-2... using the U, if the OLU is controlled for each piece of image data, the load/unload Loading is repeated, and the usage efficiency decreases, and it is impossible to quickly respond to the next page/previous page (a quantum readout time is always required when loading). However, in this example, loading is performed by accessing each group・The number of unloads is reduced, usage efficiency is increased, and next page
For the previous wave, the stored data is simply expanded and displayed, and is highly interconnected.

For example, if you prepare an image storage buffer that stores image data for 10 pages in one group space, and set the amount of image data for one page to 60 kilobytes), the OLU load time is 10 seconds, the unload time is 5 seconds, and the OLU load time is 5 seconds. Assuming that the reading time for kilobytes is 1 second, when reading every 1 (-), the time required for reading per 10 sheets is 160 seconds (1
6 seconds 25 seconds (10+1×10+5 seconds). As described above, according to the present invention, the usage efficiency of the OLU can be increased approximately six times. In addition, even if the requested image is not completely read into the group space, it is decompressed and displayed in parallel, so it can be displayed in the same time as when accessing each image individually, and the next image can be displayed in the same time as when accessing each image individually. For the previous page,
Since stored images are expanded and displayed, this can be done instantly.

〔Effect of the invention〕

As is clear from the above description, according to the present invention, the overall number of accesses to the shared storage device can be reduced, which eliminates the need for a large amount of time to start accessing due to operations such as loading and unloading. In a search system for a shared storage device, it is possible to improve the utilization efficiency of the shared storage device and improve the responsiveness to the data rate.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an example of a shared storage device to which the present invention is applied, FIG. 2 is a block diagram showing a specific example of the image retrieval device shown in FIG. 1, and FIG. FIG. 4 is an explanatory diagram showing the storage contents of the search title storage memory shown in FIG. 2, FIG. 4 is an explanatory diagram showing the storage contents of the image storage buffer shown in FIG. FIG. 6 is an explanatory diagram showing an example of the correspondence between the stored contents and the contents of the image storage buffer, and FIG. 6 is an explanatory diagram showing an example of rearranging the index management information in the buffer management memory. F... Storage device, R1-R3... Image search device, 1
. . . Storage device control unit, 2 . . . Optical disk exclusive management renostar, 3 . Input/output unit, 11... Operation unit, I2... Image search device control unit, 13... Title search unit, 14...
・Current display title position Renostar, 15...Display device,
16... Expansion circuit, 17... Accumulation check circuit, 1
8... Double buffer control unit, 19... Empty/°gutsufare no star, 20... Access order conversion calculation unit, 21... Display memory, 22... Search title storage memory, 23...・・A buffer management memory, U...
・B buffer management memory, 25... Image storage buffer, 30, E 1. E'2. E 3...
Index information, 31...Fixed length space, 32...
Index management information. Agent Patent Attorney Tadashi Akimoto Figure 1 Shared storage device Figure 2 Figure 3 Figure 4 Figure 5 Figure 6

Claims (1)

[Scope of Claims] 1. A shared storage device that requires preparation for loading/unloading etc. before access starts, and a method for accessing the shared storage device and reading variable length data stored in the shared storage device. In a shared storage search system consisting of a plurality of display search devices, the empty space of the buffer for storing the variable length data provided in each search device is
Divide and secure the capacity of the variable tone data to be read, store the variable length data read out by accessing the divided and secured area, and read and display the stored variable length data. A data buffering method for a shared storage device retrieval system characterized by: 2. The empty area of the buffer is divided and secured by dividing the entire storage area of the buffer into a plurality of fixed-length spaces, and dividing the free area of the buffer into a plurality of fixed-length spaces, and dividing and securing the empty area of the buffer as described in claim 1. A data buffering method for a shared storage retrieval system. 3. The data of the shared storage device retrieval system according to claim 2, characterized in that the order of the plurality of fixed length spaces stored in the divided and secured variable tone data is read out and displayed. Buffering method. 4. The plurality of divided and secured fixed length spaces are sequentially read out and displayed from the fixed length space in which variable tone data has been stored, and the variable tone data is stored and read out in parallel to the plurality of fixed length spaces. A data buffering method for a shared storage device retrieval system according to claim 2, characterized in that the method comprises: