JPS62186362A - Information retrieval device - Google Patents

Abstract

PURPOSE:To store efficiently the index by dividing a retrieval device into a part in charge of retrieval, a part extracting an area identifier and a part storing the relation of correspondence between the result of retrieval and a unique key in the area and extracting the unique key in the area. CONSTITUTION:A retrieval processing device 101 uses information of an index 107 to apply the retrieval processing thereby obtaining a bit map 111 in which an obtained number is stored in a form of bit map. Then the bit map 111 is sent to a bit map logical arithmetic mechanism 102, where logical arithmetic processing is applied between the resulting bit map and the information of the area information bit map 108 at each area to extract an area identifier corresponding to the number. Further, the number 113 and the area identifier are sent to an area information extracting device 103 to extract information of an inter-area internal storage device unique key 114 from the number 113 by using the information of a corresponding table 109. The extracted two primary information identifiers are sent to a primary information extraction device 104, primary information 115 is extracted from an external storage device 105 and the information is given to a user or the like.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to an information retrieval device that retrieves and retrieves information stored in a plurality of media and external storage devices.

Conventional technology In recent years, document file systems that use optical disks as storage media have been commercialized and are attracting attention as storage management devices for documents that are generated in large quantities in offices. A certain optical disk stores managed documents and search information for retrieving the documents.

For this reason, when only an optical disk is used as an external storage device, it is only possible to search for documents on one optical disk, making it impossible to retrieve documents from a plurality of optical disks. In addition, in a system that makes it possible to search and retrieve documents across multiple optical discs using a jukebox-type optical disc drive device (hereinafter referred to as a multi-optical disc drive) that automatically manages multiple optical discs, In addition to disks, a large amount of rewritable external storage devices (magnetic disk devices, etc.) are used to store search information assigned to all optical disks to be searched, and by performing searches using this search information, multiple It is now possible to search for optical discs,
Therefore, it becomes possible to retrieve a desired document from documents stored separately on a plurality of optical disks.

Additionally, in general, information for searching for target primary information from among primary information stored across multiple areas in an external storage device is stored as an index converted into a format suitable for search processing used by a search processing mechanism. remembered. As a unique key for retrieving each piece of information in this index, external storage device numbers 4o5, . Storage medium number 4o6. for identifying the storage medium. There is an area identifier 401 consisting of an area number 407 in the storage medium for identifying an area in the storage medium, and an area unique key 402 such as a registration number 408 for identifying primary information within the area, as shown in FIG. As shown in the example of an index stored in primary normal form, a primary information identifier 1102 is stored in the index in association with each piece of search information 11Q1.

However, storing search information in this -order normal form wastes a lot of storage space because keywords overlap, and the search speed is slow, so this structure is generally not used. Therefore, it may be possible to organize each search information into a structure that is more suitable for index searches, and as an example of this case,
It is conceivable to store search information in a reverse file format as shown in FIGS. 13 and 15. In this format, search information such as keywords will not be duplicated, and the keyword area 1302
It is commonly used because it improves storage efficiency and search speed. However, compared to an index stored in the -order normal form, the document number area 13o3 portion is created by the amount of overlap in the duplicate keyword area of the unique key being eliminated, so the storage efficiency of this portion is lower. Therefore, it is necessary to increase the storage efficiency of this unique key information in order to improve the storage efficiency of search information for the entire index.

Problems to be Solved by the Invention In such conventional document file systems, in order to manage documents across multiple optical disks, search information is stored in a magnetic disk device, etc. in addition to the optical disk, and this search information is It is necessary to search using . This causes the following drawbacks.

(1) Every time a document is registered on an optical disk, search information must also be stored in the magnetic disk device.

(2) Since it is necessary to store all the search information on the optical disk to be used in the magnetic disk device, it is necessary to increase the storage capacity of the magnetic disk device.

(3) In order to make an optical disk that is infrequently used ready for use, it is necessary to also store search information for the optical disk in the magnetic disk device.

(4) When using an optical disk for which search information is not stored in the magnetic disk device, the search information in the optical disk must be transferred to the -H magnetic disk device before starting the search process. It takes time to make it possible.

(5) For the reasons mentioned in (1) and ←), it is difficult to use a common optical disk among multiple document file systems.

All of the above drawbacks are due to the fact that in order to perform a primary information search across multiple optical disks, the search information for multiple optical disks is stored in an external storage device other than the optical disk, such as a magnetic disk, and this search information This is due to the fact that it uses a method of searching using .

Furthermore, in general, in an index for searching for target primary information from among primary information stored across multiple areas in an external storage device, the search information is stored in a reverse file format as shown in Figure 16. It is stored. In this case, it is necessary to increase the storage efficiency of the key field 16o7 in order to increase the storage efficiency of the entire index. Here, the number of regions is 25
0 or less and the total number of search targets is 60,000 or less. Since the number of search targets is 60,000 or less, the key field can be expressed in 2 bytes, but in this format, the area identification number 1008 is 1 byte, and the registration number 1o. ○9 requires 2 bytes, a total of 3 bytes, and the storage efficiency of this part is poor.

The present invention has been made in view of the above, and includes a search mechanism that performs search processing and stores the results in a bitmap format, and a bitmap that indicates in which area the search results and numbers stored in the bitmap format are located. A part that performs logical operation processing with the area identifier and extracts the area identifier to which the corresponding number corresponds, and a part that extracts the unique key within the area from the correspondence table between the number stored for each area and the unique key within the area from the area identifier and number. The area identification number is omitted in the index used in the search part, and the area information retrieval mechanism has a table for converting the number corresponding to the number searched in the search part into the area identification number and registration number. , This aims to improve the index storage efficiency.

In addition to the means of the present invention, it is also possible to perform searches across multiple optical disks without using an external storage device other than optical disks by changing the method of creating index information to a method of sucking search information from optical disks. It is also an object of the present invention to provide an information retrieval device that also realizes a means for making this possible.

Means for Solving the Problems In order to solve the above-mentioned problems, the present invention aims to solve the above-mentioned problems. By performing the process of retrieving the target primary information using an index composed of similarly stored search information and serial numbers that ignore area information, the index can be created using the search information and the unique key of the primary information. To reduce waste of storage area that occurs when configuring. Then, the number obtained as a result of searching the index is stored in a bitmap format, and the number stored in the pitmap is converted into a unique key of the primary information.
In order to do this, a logical operation is performed between the bitmap of that number and a bitmap created for each area to identify the area, and as a result, the area to which the number corresponds is determined. moreover,
It has a correspondence table that stores the correspondence between numbers and unique keys within an area, and uses information in this correspondence table to extract unique keys within an area. In other words, by using a bitmap that stores the relationship between the number and the area for each area, and a correspondence table that stores the relationship between the number and the area unique key, the number obtained by searching the index is used as the area identifier and the area unique key. Convert to key,
Using this information, it is possible to search primary information divided into multiple areas.

In addition, in order to enable searching of primary information stored across multiple optical disks or multiple areas within an optical disk, similarly stored search information can be downloaded into the internal storage device and an index created as a child. When generating, the area identification and area unique key of each area are replaced with numbers temporarily used in the index, and the replaced information is created in the form of a bitmap for each area and a correspondence table between numbers and area unique keys. By storing primary information and allowing processing to be performed using numbers in the index, the search mechanism can be configured without considering the number of areas in which primary information is stored. be.

Effects When the present invention configures an information retrieval device capable of retrieving primary information stored in a storage medium divided into a plurality of areas with the above-described configuration, the retrieval mechanism is in charge of retrieval with an index. The search mechanism is not separated by dividing it into a part that retrieves the area identifier using a bitmap for each area, and a part that stores the correspondence between the search result and the unique key within the area and retrieves the unique key within the area. It is possible to configure the index to be stored more efficiently than in the case of

Furthermore, when the present invention is applied to a search device that creates an index by sucking up search information from the search target area, no external storage device is required other than the search information within the search target area. It becomes possible to configure a search device that does not

Embodiment FIG. 1 is a block diagram showing an embodiment of the search processing mechanism of the information search device of the present invention. In FIG. 1, 101 is a search processing mechanism, 102 is a bitmap logic operation mechanism, 1
03 is an area information retrieval mechanism, 104 is a primary information retrieval mechanism, 105 (1osa to 105 n ) is an external storage device in which the desired primary information is stored, 106 is an internal storage device, and 107 is composed of search information and a number. index,
Reference numeral 108 is an area information bitmap that stores the correspondence between the number created for each area and the area, and 109 is a correspondence table that stores the correspondence between the number and the unique key of the primary information. Reference numeral 11o is search information for retrieving primary information input by the user, etc., and a bitmap 11 in which the search processing mechanism 101 performs search processing using the information in the index 107 and stores the resulting number in a bitmap format.
Find 1. Next, the obtained bitmap 111 is sent to a bitmap logical operation mechanism, where logical operation processing is performed between the resulting bitmap and the information in the area information bitmap 108 for each area, and as a result, the area identifier corresponding to the number is Take out. Furthermore, the number 113 and the area identifier are sent to the area information retrieval mechanism 103, where the correspondence table 10 is
The information of the area internal storage unique key 114 is extracted from the number 113 based on the information of 9. By sending the two retrieved primary information identifiers to the primary information retrieval mechanism 104, the primary information 115 is retrieved from the external storage device 105 and this information is delivered to the user or the like.

FIG. 2 shows an embodiment of an information retrieval device using the information retrieval mechanism shown in FIG. In FIG. 2, 2o1 controls the entire apparatus and the search processing mechanism 1o1 shown in FIG. Bitmap logic operation mechanism 1o2. Area information retrieval mechanism 1o3
．． - A control device that controls the next information retrieval mechanism 104, etc., 20
2a to 202n are external storage devices that store search information for retrieving primary information and secondary information; 204 is an index 10;
8° Internal storage device for storing information such as the correspondence table 107, 2
03 is a console device that inputs search information 109 and outputs primary information 113. Search information input from the console device 203 is sent to the control device 201, converted into a unique key for primary information using the index and correspondence table information stored in the internal storage device 204, and the search information is converted into a unique key for the primary information. The primary information requested by the user is transferred to the external storage device 202 and retrieved from the external storage device 202 and displayed on the console device 203.

FIG. 3 is an example of the area information bitmap 108 shown in FIG. In FIG. 3, 301 is the case where the number used for each area is stored as is, and 302
is an example of converting the information into bitmap format and storing it. In this embodiment, a 2-byte area is used as a number so that about 60,000 items of primary information can be managed.

Therefore, when storing area information in the 301 format, be sure to use 1 to indicate the area where all numbers are stored.
Because the degree needs to be entered in a field somewhere. An area equal to the maximum value of bytes for at least two numbers is required. If area information is still stored in the 302 format, an area corresponding to the maximum number bits is required per area. Therefore, if the number of areas is 16 or less, it is better to store in bitmap format to use the storage area more efficiently.

FIG. 4 shows an example of the contents of the primary information identifier. Fourth
In the figure, 401 is an area identifier, 402 is an intra-area unique key, and as described in the second row, the area identifier 401 is an external storage device identifier 405. It consists of a storage medium identifier 406, a medium area identifier 407, etc., and the area unique key 402 is a file name 408.
It consists of information such as file type 409, etc.
There is so much information that it would be wasteful to manage it all. This is the primary information retrieval mechanism 1 in Fig. 1.
As the primary information identifier to be passed to 04, use the area identification number 403 and registration number 404 as described in the third row,
It is conceivable that the primary information retrieval mechanism 104 is configured to have a mechanism for converting the information into the information described in the second stage, and this is also adopted in this embodiment.

5 and 6 show an example of the correspondence table 109 shown in FIG. 1. In FIG. 5, 501 is a number that is a primary information identifier used in the index, and 602 is a number that is a primary information identifier used in the index.
is a unique key within the area, which is a primary information identifier for retrieving primary information from within the area. In the embodiment shown in FIG. 6, a correspondence table between a number 501 and an intra-area unique key 502 is constructed for each area, and the -next information identification information is obtained by searching the corresponding table from the area number obtained by the bitmap logic operation mechanism. By scanning the number column of the correspondence table, it is possible to obtain the intra-area unique key 502 corresponding to the number 501 obtained as a result of the search. An example of FIG. 6 is a case where the correspondence table is not divided into areas, but the intra-area unique keys are stored all at once. In this case, since the primary information identifier in the index is managed as a number, the location of each unique key in the area corresponding to that number can be determined by simple calculation, so the number area is It is possible to construct a correspondence table in an unnecessary format.

FIGS. 7 to 10 are examples of storing the correspondence between numbers and primary information identifiers configured using this embodiment. In this example, the number is 30. This is a case where there are three types of areas and each area has around 10 pieces of primary information. Here, FIG. 7 is a data table showing the correspondence between numbers and primary information identifiers. In FIG. 7, 701 is the number, 702 is the area identification number 404 in FIG. 4, and 703 is the area identification number 405 in FIG. This is the registration number. FIG. 8 is an example of an area information bitmap constructed when the correspondence relationship as shown in FIG. 7 exists. Figure 9 is an example of a correspondence table created in the format shown in Figure 6. In the figure, 901 is a column that stores the registration number, which is a unique key within the area, and 902 is not actually stored. is a column showing the number to which 901 corresponds. 901 are arranged in numerical order regardless of area. Figure 10 is an example of a correspondence table created in the format shown in Figure 6.
In the figure, 1oo1 is a correspondence table for area 1, 1o02 is a correspondence table for area 2, and 1oo3 is a correspondence table for area 3.

In the case where a correspondence table as shown in Fig. 7 is used to express the correspondence relationship between numbers and primary information identifiers, and in the case shown in Figs.
Comparing the case of using the method shown in Fig. 10, in terms of storage area, the number area can also be omitted in the case of Fig. 7, so the number area is 2 bytes, the area identification number is 1 byte,
If the registration number is 2 bytes, it will be 3*N bytes) (N is the maximum number of numbers). On the other hand, the area information bitmap in Fig. 8 is proportional to the number of areas, and in this case there are 3 areas, so 3*N
/a byte area is required. Further, if the format of the correspondence table as shown in FIG. 9 is used, a total of 4*N bytes are required for the correspondence table for each area, and in the format of FIG. 1, 2*N bytes are required. Therefore, when using the formats shown in FIGS. 8 and 10, this format is more advantageous if the number of regions is 8 or less.

If we only compare storage efficiency, there are few advantages to converting to bitmap format, but storing in bitmap format has the following major advantages. When performing a search, there are cases where it is desired to search only a specific area, but in such a case, this is not possible using the search mechanism alone because the area designation information is not managed using an index. Therefore, when converting the results of a search in the entire search range into a primary information identifier using this area specification mechanism, it is specified to extract only a specific area. In this case, if you create a correspondence table in the format shown in Figure 7, the information is not stored using areas as keys, so you must either scan the correspondence table or convert all search results to primary information identifiers. There is a need to. On the other hand, in the case of an embodiment that has an area information bitmap, the search results include a logical operation (
Search targets can be easily narrowed down by performing an AND operation. Therefore, after passing the search results through the bitmap logical operation mechanism, the narrowing down of the area is completed.
The advantage is that it is not necessary to convert all search results to primary information identifiers.

Figures 11 and 12 show an example in which the index used in the search processing mechanism is configured in primary normal form. Figure 11 shows the case of this example, and Figure 12 shows the case of the conventional example. There is. 1101 and 1201 are search information added to primary information in order to retrieve the primary information, and 1103.1203
is an attribute representing an item of data to be searched, 1104.1
It is composed of keywords, etc., which are atoms of the search data stored in that item of 204. 11o2 is a number that is the primary information identifier in the index to be passed to the area information retrieval mechanism, and 1202 is the primary information identifier in the conventional example, which is composed of area identification number 1205 and registration number 12o6 explained in FIG. There is.

As mentioned in the section on conventional examples, if the index is configured in linear normal form, it will be disadvantageous in terms of index capacity and search speed.
In practice, this is rarely used when constructing an index, and indexes are often constructed with a structure more suitable for searching. This is because, if stored in this format, the system of this embodiment may require a search instruction from the user to retrieve primary information to which a certain keyword has been assigned; In order to answer this question, it is necessary to scan all the rows (hereinafter referred to as records) in the index and check whether the corresponding keyword exists. Therefore, the search process takes time. In addition, it is generally difficult to search if all the keywords used to retrieve primary information are unique keys, so the same keyword is assigned to a collection of primary information with similar characteristics, and the Keywords, which are search information, are assigned to primary information by determining the content according to various attributes. In this useless keyword information, there are more keywords with duplicates than unique keys, and the degree of duplication is also large. Therefore, if keywords are stored in this format, there will be a lot of unnecessary duplication of keywords, so storage space will be wasted. I'm too loud. Configuring the index in first-order normal form causes the two drawbacks mentioned above.

Therefore, in the document file system described in this embodiment, a reverse file structure is often used as the index structure. An example of this case will be described with reference to FIGS. 13 to 16. FIGS. 13 and 15 show the overall configuration of an embodiment in which the index used in the search processing mechanism shown in FIGS. 11 and 12 is configured not in linear normal form but in reverse file structure, FIGS. 14 and 16 are diagrams showing the internal physical structure of an embodiment in which the index is configured with an inverted file structure.
13 and 14 show the case of this embodiment, and FIGS. 15 and 16 show the case of the conventional example.

In Figures 13 and 15, 1301.1501 is the first
An index area that stores a keyword index when the keyword 11Q4 shown in Figure 1 is expanded into a reverse file structure for each attribute 1103.

1302 and 1502 are keyword areas in which the keywords are stored. In this embodiment, 13Q3 is a number area that stores information on which number a keyword has for converting search information such as a keyword into a number, and 16o3 is an area identification that corresponds to the number area in the conventional example. This is an area identification number/registration number area that stores a combination of a number and a registration number. In Figures 14 and 16, 1401 and 1601 are the keywords themselves; 1401 and 1601 are the keywords themselves;
o2 is pointer information indicating the position of the number block 1403 in the number area storing the number 14 o4 representing the primary information to which the keyword is assigned as search information, 16
o2 has the same position as 1402 in the conventional example, and -
This is pointer information indicating the position of the key block 1603 in the area storing the next information identifier (hereinafter abbreviated as key) 1607, area identification number 1608, and registration number 1609.

In this reverse file structure, search information is organized by keyword.9 Keywords are arranged logically to make it easier to search, and using an index, you can easily retrieve the key of primary information that has a certain keyword. can be made possible. Furthermore, since the keywords are organized to avoid duplication, there is less waste of storage space. By the way, in this method, compared to the case of linear normal form, there is no duplication of keywords, so duplication of numbers and keys occurs, so in order to minimize the increase in this area, it is necessary to minimize the amount of data in this area. occurs. This point is shown in Figure 14 and 1.
As is clear from a comparison of FIG. 6, this embodiment has better storage efficiency.

To summarize the storage efficiency of this actual flow example and the conventional example, in the case of -order normal form, the number is 2 bytes and the key is 3 bytes in the index area, so it is advantageous by the number of bytes of records, but the corresponding table When the correspondence table is constructed as shown in Figure 6, the overhead for the area information bitmap increases by (2 * number of records) bytes, and the overhead for the area information bitmap increases by (number of records * number of areas / 8) bytes. It is disadvantageous because it becomes larger by % of the amount being converted and the area information bitmap (number of records, number of records, number of areas/8). However, as mentioned in the previous section, it is not a good idea to store the index in this format, and if we assume that the number of areas to be searched is 8, if it is stored in an inverted 7-file structure, the number area and key area will be If there are three or more duplicate numbers or keys, the storage area in the index is advantageous (3*number of records), and the overhead of the correspondence table can be absorbed.

Furthermore, the structure of the number area should be configured so that multiple numbers can be stored per keyword. Configure by dividing into cells. Deciding how many numbers can be stored in this cell depends to a large extent on the field of application, but it is necessary to secure an area of 5 or more, since there will be overhead for pointers when there are keywords that store many numbers. becomes larger. If there are 5, a difference of 5 bytes will occur per cell, and the overhead for the correspondence table will be sufficiently compensated for, making the method of this embodiment advantageous in terms of overall internal storage efficiency.

Next, an embodiment of a document file system using the present invention will be described. Figures 17 and 18.

Figure 19 shows that when the external storage device is configured as a single optical disk device that performs write-once recording and reading using a plurality of optical disks as storage media, This is a case where the system is configured with a multi-optical disk device that performs reading, or a case where both are combined. Figure 17.

The operation of the document file system of this embodiment will be described with reference to FIGS. 18 and 19. Image information is input to the document file system by displaying an image input from the image input/output device 17o6 on the console device 1704, and storing the image in the single optical disk device 17o2 or the multi-optical disk device 1706. From the device 17o4, an instruction for storing this image in which area of which medium of which external storage device and search information for retrieving this image are simultaneously registered in the form of a keyword or the like. This search information is also simultaneously stored in the index and correspondence table stored in the internal storage device 1703. Next, the input image information is retrieved from the console device 1.
Keywords assigned to the image information are input from 7o4, and this keyword information is sent to the information retrieval mechanism consisting of the internal storage device 1703 and the control device 1701, and the search processing mechanism and area within this mechanism are input. The information retrieval mechanism converts it into external storage device designation, media designation, area designation, and area identification number, and based on this information, the specified medium of the specified external storage device is specified. The designated image information is extracted from the area and displayed on the console device 1704. Furthermore, if a hard copy of the image displayed on the console device is required, the image information can be sent to the image input/output device and printed.

FIG. 20 shows an example of the configuration of a storage area in an optical disk medium used in the document file system described above. 20th
In the figure, 2001 is a medium information description area that describes the type of medium, medium identifier, overall structure of the medium, capacity, etc., and 2002 is a medium information description area for storing where image information is actually stored in the medium. This is an image storage location information storage area, 20o3 is a search information storage area for storing search information for retrieving image information, and 2o04 is an image information storage area that actually stores image information. FIG. 21 shows an example of the configuration of the search information storage area. In Fig. 21, the search information is shown in Fig. 21a for each record.
The format is stored in linear normal form as shown in Figure 21. FIG. 22 shows an example of the configuration of the image storage position information storage area. In Fig. 22, image storage location information is stored for each record as shown in Fig. 22a, and its contents include a unique key for extracting logical image information and physical storage as shown in Fig. 22. The relationship with the position is memorized.

FIG. 23 shows an embodiment of an information retrieval device that reproduces an index and a correspondence table from retrieval information in an optical disc using the present invention. In this embodiment, the search information retrieval mechanism 2301 is the 20th
21, and 22. From the external storage device 23o2 to which an optical disk having a storage structure as shown in FIGS.
The search information 231o as shown in FIG. 1 is extracted and sent to the correspondence table creation mechanism 2303 along with information on external storage devices, media, areas, etc. In the correspondence table creation mechanism 2303℃,
Among this information, information regarding the storage area is shown in the correspondence table 2307.
, updates the correspondence table 23Q7, extracts the corresponding number from the correspondence table 2307, and sends the search information 231o, this number, and the area number to the bitmap creation mechanism 2304 as area information 2312. Pit map creation mechanism 23o4
Now, the area information 2312 is converted to the area information pit map 2308.
The corresponding number column of each bitmap is updated, and the search information 2310 and number 2313 are sent to the index creation mechanism 18.
Send to 04. The index creation mechanism 18o4 uses the sent search information 2310 to update the index 23o9fc. Before performing a search process using the method described above, search information can be retrieved from an optical disk to be searched in an external storage device, and a correspondence table and index can be reproduced. If the index and correspondence table are reproduced using this method, an information retrieval device capable of retrieving primary information across multiple areas will be constructed using the information retrieval mechanism shown in FIG. 1, which is an embodiment of the information retrieval device of the present invention. can do. In other words, by using the information retrieval device shown in FIG. 23, the information retrieval device shown in FIG. 1, and the optical disk having the format as explained in FIGS. It is possible to configure an information search device that can search multiple optical discs without having to do so.

Effects of the Invention As described above, according to the present invention, the search processing mechanism is divided into a part that performs the search itself and a part that manages the area of primary information stored in multiple areas, and unique keys within the area are divided. This is achieved by configuring in the internal storage a correspondence table that stores the relationship between area information and numbers, an area information bitmap that stores the relationship between area information and numbers, and an index that stores the relationship between numbers and search information. are doing. As a result, the capacity of the internal storage device as a whole can be reduced.

Furthermore, by adding a mechanism for siphoning the search information in the original disk and reproducing the index and correspondence table, it is possible to construct an information search mechanism that can search a plurality of optical disks without requiring any external storage device other than the optical disk device.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an information retrieval mechanism in an embodiment of the present invention, FIG. 2 is a configuration diagram of an information retrieval device using the information retrieval mechanism of FIG. 1, and FIG. 3 is a configuration diagram of an area information bitmap. Figure 4 is a configuration diagram of a primary information identifier, Figures 5 and 6 are configuration diagrams of a correspondence table, Figure 7 is a correspondence diagram showing the relationship between numbers and area identification information, and Figure 8 is a diagram showing the relationship between numbers and area identification information. A configuration diagram of the area information bitmap, Figures 9 and 1o are configuration diagrams of the correspondence table in Figure 7, Figure 11 is a configuration diagram of an index configured in primary normal form, and Figure 12 is also configured in primary normal form. Figures 13 and 14 are configuration diagrams of a conventional index configured with a reverse file structure. Figures 15 and 16 are configuration diagrams of a conventional index configured with a reverse file structure. 17, 18, and 19 are configuration diagrams of a document file system to which the present invention is applied, with different configurations of external storage devices, and FIGS. 20 and 21.
22 is a configuration diagram of a storage area in an optical disk of a document file system to which the present invention is applied, and FIG. 23 is an information search for reproducing an index and a correspondence table from search information in an optical disk using the present invention. It is a block diagram of a device. 101... Search processing mechanism, 102...
Bitmap logic operation mechanism, 103...area information retrieval mechanism, 104...-next information retrieval mechanism, 10S...external storage device, 106...
... Internal storage device, 107 ... Index, 108
...Area information bitmap, 109...
・Correspondence table, 110...Search information, 111...
...Bitmap, 112...Area identifier, 1
13...Number, 114...Unique key within area, 115...Next information. Name of agent: Patent attorney Toshio Nakao, 1st person, 2nd person
Figure 3 Number of Ken No. 1-ga (numbering up type) 307 Shinwaki Gei No. 4 (B/Koma/Nobu format) 302 Figure 5 Figure 6 Figure 7 Figure 8 Number of @St (Number table of bitma-no-format λ δo1 1 2 (bitma-tub format), body length of δ02 linkage 3 (bi-soto map format)
Fig. 11 Fig. 12 Fig. 313 Fig. 14 Fig. 15 Fig. 16 Fig. 17 Fig. 1'101 Fig. 18 Fig. 19 Fig. 7ot Fig. 20 Fig. 21 Fig. 22 Fig. 23

Claims (2)

[Claims] (1) A storage medium in which the target information to be searched is divided into multiple areas, or an external storage device with multiple storage media, or multiple physically and logically divided areas such as multiple external storage devices. an external storage device or a group of external storage devices that has a mechanism that allows target information to be referenced based on information that specifies the area and unique key information independently determined for each area, and each area. An index configured using a number to uniquely identify the target information in the entire search target area independently of the reference mechanism within, and the search results using the index are stored in a bitmap format of the number. It stores the difference between areas in a bitmap format created for each area indicating which area the number belongs to, and performs logical operation processing between this area specification bitmap and the search result bitmap. a mechanism for extracting an area identifier from the number, a mechanism for storing the correspondence between the number and the unique key in each area and extracting the unique key in the area from the number, and a mechanism for extracting the unique key in the area from the area and the unique key in the area. An information retrieval device characterized by having a mechanism for retrieving target information. (2) A unique key that is logically determined to refer to the target information in each area stored in each external storage device and the physical location of the target information stored in the storage medium. an area for storing information representing a relationship with the information shown;
Unlike the unique key information in the area, an area for storing search information representing the relationship between the unique key information in the area and search information added by the user to easily identify desired information. an external storage device in which a storage medium can be used, information for the search of a search target from a search information area in a storage medium in the external storage device, a unique key in the area of the information, and the information. A mechanism for retrieving the area identification information of the area in which it is stored one by one, the retrieved information for inspection, the unique key within the area, the area identification information for searching, and the information for searching for all search targets. a mechanism for converting into area unique key information and storing the correspondence between the area unique keys of all the search targets and the area unique keys in an internal storage device; and an area indicating to which area the number belongs. A mechanism for creating a bitmap for each area and storing it in an internal storage device, and creating a search check in the internal storage device using the information for the search and the unique key within the entire search target area. An information retrieval device according to claim 1, characterized in that it has a mechanism.