JPS6326901B2 - - Google Patents

Description

[Detailed description of the invention]

The present invention is a retrieval device having a data file and a retrieval control device, which is suitable for storing data output from an external sensor that generates data in real time in a recording device, and retrieving desired data from the stored data. Regarding. If various test results in the medical field are stored in a storage device and any desired results are to be retrieved and used as diagnostic data, the number of patients and test items is large, so it is necessary to devise storage and retrieval methods. . The present invention aims to provide a device convenient for storing and retrieving such a wide variety of data.The present invention is characterized by a retrieval device having a data file and a retrieval control device. An attribute section that records the number, a unit data ID section that records the type code of unit data and a sequence number for each type code, and a data section that records variable length data. A data file in which a plurality of unit data are continuously recorded in sectors for each cylinder of a storage device having a plurality of cylinders subdivided into sectors in order of sequence numbers for each type code, and each section column. corresponds to each cylinder of the data file, each section of the section string corresponds to each sector within the cylinder, and the type code of the unit data is recorded in the section corresponding to the first sector position of the unit data. A cylinder arrangement table and a number table that is provided for each type code and records the minimum or maximum sequence number of the unit data having the type code recorded in each cylinder of the data file for each cylinder. When searching the data file, the search control device uses the type code and sequence number of the unit data to be searched to find the number of the cylinder in which the unit data to be searched is recorded from the number table, and The section row corresponding to the cylinder number in the cylinder allocation table is read sequentially from the end to find the sector number of the first sector of the unit data to be searched, and the data file is accessed using the cylinder number and sector number to find the target. The purpose is to read unit data. This will be explained in detail below with reference to Examples. In various tests in the medical field, such as blood tests, in addition to the blood of each patient being tested, i.e., an unknown specimen, water for calibration, i.e., a blank specimen, a standard specimen,
There are different types, such as pooled samples, and each type has a sequence number. Since each patient has a reception number at the hospital, it is possible to abandon the sequence number mentioned above and use the reception number for all patients, but in this case, not all patients will undergo the same test, so the numbers will be scattered all over the place. , which is inconvenient. Therefore, in the present invention, a type code and a sequence number within that type are used as identification codes for test results, that is, data, and these are used as specimen IDs.
(identify code). Data is first written to memory (primary buffer using RAM, etc.), then edited, etc., and written to a floppy disk. FIG. 1 shows the format of data written to this floppy disk. As shown in this figure, the data (data for one sample, unit data in general terms) are
It consists of an attribute section A indicating the sector length (4 sectors (sector boundary)), a sample type code C, a sequence number N within that type, and variable length measurement data MD. If this attribute part A is present, by looking at it, it is possible to know the length of the data concerned and, by extension, the boundary with the next data. To access the data of a certain sample, it is only necessary to know the sector address of the attribute section. Therefore, as shown in Fig. 2, data for one sample with the configuration shown in Fig. 1 for each sample is successively stored on the floppy disk. In addition to the data file to be written, an index file section is provided, and the type code of each specimen data is written in this index file so as to correspond to the sector position of the attribute section of the specimen data in the data file. In this figure 2, 0c, 1c...73c are the cylinder numbers of the floppy disk, and 1s...26s are the sector numbers within the cylinder.Therefore, the number of cylinders on one floppy disk is 74, and the sectors of each cylinder are The number is 26. a is the front side, b is the back side, and after n cylinders on the front side, there are n cylinders on the back side, and then n+1 on the front side.
The order is cylinder... If the floppy disk is one-sided, the n cylinders on the back side become n+1 cylinders on the front side. If this index file exists, it is possible to know the sector address of the attribute section from the identification code of the specimen sought by looking at it, access the data file with the address, and read the data of the specimen. FIG. 3 shows an example of the data storage state in the n-th cylinder of the data file section. a is the front surface and b is the back surface. D is a dummy sample, B is a blank sample,
S is the standard sample, P is the pool sample, 1, 2, and 3 are each the control sample, and U is the unknown sample (these have the format shown in Figure 1).
, and the diagonal line indicates the sector in which the attribute is written. An identification code is written in the index file as shown in FIG. 4a. This is the third
In Fig. 3, the attribute part of each sample data D, B, S... is written in sectors 1, 4, 7... in the n-th cylinder, so the data is D, B, 1, 4, 7... sections of cylinder n' corresponding to n cylinders of the file
S...Write each type of code. This section is
Although it corresponds to the sector, it is a unit area smaller than a sector because it only needs to be able to store a type code. b/ indicates blank;
Portions of the sample data other than the attributes are written in sections corresponding to sectors 2 and 3, 5 and 6, and so on. If you do this, each section will have 1
Using a clock that generates pulses, by counting and shifting this clock and reading the n' cylinder, we get:
The position of each sample data in the n cylinders of the data file can be known from the read type code and counter count value. This cylinder arrangement table SDT is created for each cylinder of the data file. As shown in Figure 4b, the index file also contains a final number table in which the last type code D, B, etc. of each sample in cylinder n in this example and its sequence number Ni, Nm, etc. are written in each cylinder. Provide LNT. This number table will be created for each type. Data reading is performed using the specimen ID, that is, the type code and sequence number, but this final number table LNT
If there is, by reading and viewing this it is possible to know in which cylinder the desired specimen data is located. That is, when reading the n' cylinder part of the final number table, the final sequence number is 35, and when reading the next (n+1)' cylinder part, the final sequence number is 40.
If this is known, then data U, for example No. 36, that we want to read should be in cylinder n+1. Once this is understood, the next step is the cylinder arrangement table SDT in Figure 4 a.
By reading out the (n+1)' cylinder part, it is possible to obtain the sector number in the data file part of the attribute part of the target unknown data U No. 36. Therefore, after this, the sector number and (n+
The data file can be read using the cylinder information 1) as an address signal. To further explain this with reference to FIG.
2) 'Assuming that each cylinder part is written as shown in the figure, the final number table LNT is written in the U part.
10, 15, 20. If the sample ID you want to search is U0014, that is unknown sample No. 14, the final number 10 is in the n' column of the final number table, and since this table is created in ascending order, read the next column (n+1)' and find the final number. Get 15. From this, it can be estimated that the target U0014 is in the n+1 cylinder, so read the (n+1)' column of the cylinder arrangement table SDT from the tail end, and subtract the count value by one from 15 every time one U is read. Since the target U0014 is the number 14, the cylinder number can be determined as n+1 and the sector number can be determined as the number of shifts until U0014 is found. Thus, according to the present invention, any specimen data can be read out with three accesses. Although the cylinder arrangement table and the final number table may be stored in a random access memory or the like, since the storage capacity is quite large, it is still economical to write them on a floppy disk. Second
In the figure, the index file is in the first cylinder 1c.
~ Although it is said that it will be stored in the ninth cylinder 9c,
In this case, for example, the cylinder arrangement table SDT is
cylinder and the third cylinder, and the final number table LNT is stored in the first cylinder. In other words, the number of cylinders in the data file is 64, one cylinder has 52 sectors, the number of bytes in one sector is 256, the data per cylinder for each type of LNT is 2 bytes, and the number of types is 8, then 64 x 2 = Since it is 128 bytes, one type of LNT data will fit in one sector, even if the sector is different for each type (even if it is a sector boundary)
Eight sectors are sufficient for LNT, so it fits within one cylinder. The SDT may be 52 bytes per cylinder, and if each cylinder is allocated to a sector boundary, 64 sectors, or 2 cylinders, are sufficient. The fourth to ninth cylinders of the index file are reserved. The final number table will be prepared for each category, and will therefore be as follows.

[Table] These tables are written in the cylinders of the index file in order of type, and the cylinder arrangement table for each cylinder in the format shown in FIG. 4a is also written in the same way. To read the final number table LNT, access the cylinder storing the final number table and read the table data of the relevant type. Once the cylinder number has been found from this imported data, it is used to access the cylinder allocation table SDT. If you write it in the final number list, you can immediately access the above information by looking at it. FIG. 6 shows an overview of this data processing system. μ-CPU is a microcomputer, M is memory (RAM), FD is a floppy disk, SI is a sensor interface, KB is a keyboard,
DIS is a display section, and these are connected through bus B. Data from the sensor is once written from the interface SI to the memory M, and after undergoing editing processing, is written to the floppy disk FD. When searching for data, move the entire final number table corresponding to the desired data type to memory M, calculate the cylinder number there, and then transfer the part (section column) of the cylinder number in the cylinder arrangement table to memory M. Move to M and find the sector number. FIG. 7 shows the flow of this operation. First, use the keyboard KB to enter the type code and sequence number of the sample data you wish to search. In response to this, the computer μ-CPU transfers the final number table LNT of the type from the floppy disk FD to the memory M. Next, find the start and end cylinder numbers BOE and EOD of the data file from the file label,
The LNT cylinder numbers a and b for these BOE and EOD are obtained, and it is checked whether the sequence number S is within (a) and (b). Here, (a) and (b) are the final sequence numbers of the LNT for cylinder numbers a and b (however, the final number one cylinder before the start end a is used). If there is no data, there is no data, and if there is, a cylinder b whose sequence number S is greater than (b-1) and smaller than (b) (may be equal) is found, and this is used as the desired cylinder number. Next, move the b cylinder part of the cylinder arrangement table SDT from the floppy disk FD to the memory M.
Every time something with the same type code is read, (b)-S
Shifting is performed while decrementing C of =C by -1 until C becomes zero. Here, (b) is the final value of cylinder b (of the relevant type code). Assuming that the number of shifts is d, the maximum value of d is 52 in this example, and within this 52, the relevant type code S' matches the d-th type code in cylinder b (is the type code read out sequentially by shifting). . This number of shifts is the sector number d, and these b and d become the address within the floppy disk of the desired sample data. FIG. 8 is a diagram schematically showing the above operation. The symbols are the same as in each previous figure. As explained in detail above, according to the present invention, a large number of different types of data can be found by searching a file three times. Compared to the method of reading data one by one and returning it to the file, desired data can be obtained extremely quickly. Generally a floppy disk (file)
It takes much longer to access the floppy disk than to access the memory, and therefore the number of accesses to the floppy disk must be minimized, so the present invention, which can limit this to three times, is extremely effective. It is. The computer used to search this kind of data file is a general-purpose microcomputer, but its processing speed is not extremely fast compared to the data transfer speed from a floppy disk, and it is difficult to read one sector worth of data into memory. During processing, the next sector will pass by, and you will have to wait for one rotation before continuing. Therefore, if there is no cylinder arrangement table SDT and each sector of the data file cylinder with the cylinder number obtained from the final number table LNT is searched in order until the desired sector is obtained, the floppy disk must be rotated many times. Must be. In terms of average processing time, without SDT, it would be (total number of sectors per cylinder)/2 x average rotational waiting time, but with SDT, this becomes only the average rotational waiting time, speeding up processing. The cylinder arrangement table is valid. Note that the final number table LNT may be the reverse first number table. Furthermore, the present invention is of course not limited to blood tests, and can be used to search various types of data having types and sequence numbers within types.

[Brief explanation of the drawing]

Figure 1 is an explanatory diagram showing the format of data for one sample, Figures 2 a and b are diagrams explaining the storage status on a floppy disk, and Figures 3 a and b are diagrams showing the format of data for one cylinder of the disk. Figures 4 and 5 are diagrams explaining the storage situation, Figures 4 and 5 are diagrams explaining the cylinder arrangement table and final number table, Figure 6 is a block diagram showing the configuration of the data processing device, and Figures 7 and 8 are the operation. They are an explanatory flowchart and an illustrative diagram. In the drawing, FD is a storage device, SDT is a cylinder arrangement table, and LNT is a number table.

Claims (1)

[Claims] 1. A search device including a data file and a search control device, which records the number of sectors occupied by unit data, an attribute section that records the type code of the unit data, and a sequence number for each type code. unit data
One item of unit data consisting of an ID part and a data part for recording variable length data is composed of an integral multiple of a unit sector, and multiple unit data are subdivided into sectors in sequence number order for each type code. A data file in which sectors are consecutively recorded in each cylinder of a storage device having a plurality of an in-cylinder arrangement table in which the type code of the unit data is recorded in the section corresponding to the first sector position of the unit data; A number table is provided in which the minimum or maximum sequence number of unit data having a type code is recorded for each cylinder, and when searching the data file, the search control device determines the type code and sequence of the unit data to be searched. Using the number, find the number of the cylinder in which the unit data to be searched is recorded from the number table,
finding the sector number of the first sector of the unit data to be searched by sequentially continuing the section string corresponding to the cylinder number in the cylinder arrangement table from the end;
A search device characterized in that a data file is accessed using the cylinder number and the sector number to read out target unit data.