JPH04344567A - Information deletion and registration system - Google Patents

Abstract

PURPOSE:To shorten processing time for deleting or registering data, to improve the processing efficiency, and to simplify configuration. CONSTITUTION:A storage device 11 is provided with a plurality of data management tables consisting of a plurality of data management elements and a plurality of data management headers, and also is provided with a registration information area 151 for storing data that indicates whether data management elements are registered or unregistered, a key name length area 152 for storing key name length, a key section 153 for storing key name and a data section 154 for storing data. Each data management element of each data management table is formed in a continuous storage area. Further, a hashed value from a hash function section 19 retrieves a relevant entry from a hash table 17, specifies a relevant data management table in the storage device, and retrieves data management elements.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information deletion and registration method for deleting and registering data using a hash table.

0002

[Prior Art] When deleting or registering data, a data search is required, and the data search uses information such as a character string, usually called a key name, and searches for a key name that matches this key name. There are some that search for data. In this case, there is a problem in that if the data is simply stored consecutively in a table and the data in the table is searched sequentially by matching key names, the search will take a huge amount of time. .

[0003] Conventionally, a hash method is known as a method for quickly searching data. This method uses a hash function to obtain a hash value from search information (key name), and stores the corresponding data in a storage area uniquely determined by the hash value. Also, in the hashing method, different key names, such as "swallow"
" and "lion" may have the same hash value, a so-called hash value collision.

For this reason, the storage address directly determined from the hash value is used as the address to the first data or data pointer, and other data belonging to the same hash value is sequentially determined from the first data or pointer to the data or pointer. A method is adopted in which a chain of data is created and stored. To search for target data from a data group belonging to the same hash value, the search information (key) in the data or pointer is compared with the search information for the data group to ensure a match.

That is, conventionally, as shown in FIG. 5, data d1, d2, d3 and data d4, d5 having the same hash value are stored in a chain in the storage device 1, respectively. Each data d1 to d5 is the key part 1a and the data part 1
b. It consists of a pointer part 1c. The address of the next data to be chained is held in the pointer part 1c, and when there is no next data to be chained, for example, "0" indicating the final data is written. ing.

Then, the key name stored in the search information register 2 is input to the hash function section 3 to obtain a hash value, and one entry is obtained from the hash table 4 based on the hash value. In each entry of the hash table 4, the address of the first data in the chain data belonging to the corresponding hash value is written.

[0007] Now, data d1 is stored in hash table 4.
If the address is specified, the key part 1a of the data d1 is read out and compared with the key of the search information register 2. If they match, the data d1 becomes the target data. If they do not match, the pointer section 1c of the data d1 is read out, and using this as an address, the key section 1a of the next chained data d2 is read out and compared with the key name of the search information register 2. If they match, the data d2 becomes the target data. If they do not match, read the pointer part 1c of the data d2,
Using this as an address, the key part 1a of the next chained data d3 is read out and compared with the key name in the search information register 2. In the manner described above, target data is retrieved from the storage device 1.

Furthermore, as seen in Japanese Patent Laid-Open No. 60-254254, for example, a plurality of hash tables each having a different hash function are provided, and information corresponding to the content corresponding to one key name is stored in each hash table. When each hash table is referenced, the information is compared, and if all common information exists in the hash table, the content corresponding to that information is read out from memory. If all the information to be registered does not exist in the hash table, the content is determined to be unregistered.

[0009]

[Problems to be Solved by the Invention] However, in the system shown in FIG. 5, when searching for target data from a group of data belonging to the same hash value, the pointer in the data is read for each data and the pointer in the data is searched for the next data. Since the process of determining the address is required, the search takes time, and therefore there is a problem that the processing efficiency is low when deleting or registering data in the storage device 1.

Furthermore, the method disclosed in Japanese Patent Application Laid-Open No. 60-254254 has the problem that it has a complicated structure, requires a large storage area, and takes a long time to process when deleting data.

[0011] Therefore, the present invention aims to provide an information deletion/registration method that can shorten the time required to delete or register data, improve processing efficiency, and have a simple configuration. be.

0012

[Means and effects for solving the problem] The present invention calculates a hash value from search information using a hash function, and uses the hash value to refer to a hash table when data is deleted or registered in a storage device. Consists of a registration information area where information indicating whether data is registered in the device is stored, an area where key name length information is stored, and an area where information for managing key names and data names is stored. Multiple data management elements are stored consecutively in the storage area to form a group, and multiple data management tables are provided that are made up of data management headers that manage each data management element in the group, and are subject to deletion when data is deleted. The hash table is referenced from the hash value of the data, the corresponding data management table is searched, the corresponding data management element is searched from the data management table, and the data indicating unregistered is entered in the registration information area of the data management element. When writing or registering data, the hash table is referenced from the hash value of the data to search for the corresponding data management table, and data indicating unregistered is written to the registration information area in each data management element of the data management table. Search for the data management element that is searched, and write information in the registration information area of the searched data management element, the area where key name length information is stored, and the area where information for managing key names and data names is stored. The purpose is to write the corresponding data.

[0013]

Embodiments Hereinafter, embodiments of the present invention will be described with reference to the drawings.

In FIG. 1, 11 is a storage device, and this storage device 11 stores a plurality of data management tables 12, 13, 1.
4,... are provided. Each data management table 1
2, 13, 14, ... are a plurality of data management elements 12
a, 12b, 12c, 13a, 13b, 13c, 14a
, 14b, 14c and data management headers 12d, 13d,
14d. Each of the data management elements 12a, 12b, 12c, 13a, 13b, 13
c, 14a, 14b, and 14c are each formed in a continuous storage area to form a group.

[0015] Each of the data management elements 12a, 12
b, 12c, 13a, 13b, 13c and 14a, 14
b and 14c are data indicating whether the data is registered or not, for example, a registration information area 151 that stores all "1"s when unregistered and "0" when registered, and the length of the key name (for example, if the key name is If it is "milk", it is composed of a key name length area 152 for storing 4), a key part 153 for storing the key name, and a data part 154 for storing data. Each of the data management headers 12d, 13d, 14d corresponds to each data management element 12a, 12b, 12c, 13a,
13b, 13c and 14a, 14b, 14c, and a data management element counter 161 that indicates how many data management elements are registered in each.
and a next table pointer section 162.

Reference numeral 17 denotes a hash table, and each entry 17a, . . . of this hash table 17 stores, for example, the start address of the data management tables 12, 14, etc. The next table pointer section 162 of the data management table 12 is linked to the data management table 13.
The start address of is stored. Since there is no linked data management table in the next table pointer section 162 of the data management tables 13 and 14, all "1"s are stored.

18 is a search information register in which the key of data to be searched is stored; 19 is this search information register 18;
This is a hash function part that calculates a hash value from the key name using a hash function. and the hash function section 1
9 by the hash value from hash table 17
The corresponding entry will be searched from.

When creating a new data management table, the data management element counter 161 of the data management header becomes "0", and the next table pointer section 162 becomes all "1" since there is no linked data management table. , and the registration information area 151 of each data management element is all "1".

FIG. 2 is a flowchart showing the information deletion process. For example, as shown in FIG. 3(a), the data management table 12
Describing the case of deleting the data "OS" corresponding to the key name "UNIX" in the data management element 12b, the search information register 18 contains the key name "UNIX".
is stored.

Then, the hash function unit 19 and the hash table 17 are used to determine the deletion position. That is, the key name "UNIX" is input to the hash function section 19 to obtain a hash value, and one entry 171 of the hash table 17 is searched from the hash value. The searched entry 171 stores the start address of the data management table 12.

Next, the key part 153 of each data management element 12a, 12b, 12c of the data management table 12
key name and search information register 18 key name “UNIX
” to find the position of the corresponding data management element 12b. Next, unregistered information writing processing is performed. This process writes all "1"s indicating non-registration into the registration information area 151 of the data management element 12b. Finally, counter decrement processing is performed. This process increments the contents of the data management element counter 161 by 1. The data management table 12 after the data "OS" is thus deleted looks like the one shown in FIG. 3(b).

FIG. 4 is a flowchart showing information registration processing, which first performs data management table presence/absence determination processing. In this process, a hash value is determined by the hash function section 19 from the key name stored in the search information register 18. And based on this hash value, hash table 1
7, and check whether a pointer to the data management table is set in this entry. If the entries are all "1", it is determined that the pointer is not set, that is, there is no corresponding data management table, and then data management table creation processing is performed.

In this data management table creation process, a data management table is created. At this time, the data management element counter 161 of the data management header is set to "0", and all "1"s are stored in the next table pointer section 162. In addition, the registration information area 151 of each management element
All "1"s are written in the fields.

Subsequently, hash table entry setting processing is performed. In this process, the head address of the data management table created this time is stored in one entry of the hash table 17 obtained from the previously determined hash value.

[0025] Next, registration processing is performed. In this process, the data management element counter 161 is incremented by 1, and information is stored in each of the areas 151 to 154 of one data management element. That is, "0" is stored in the registration information area 151, the length of the key name is stored in the key name length area 152, the key name is stored in the key section 153, and data is stored in the data section 154. In this way, the registration process ends.

[0026] Also, the searched 1 of the hash table 17
If a pointer (starting address) to the data management table is set in the entry, it means that the corresponding data management table exists, and data registration position determination processing is performed.

In this data registration position determination process, a corresponding data management table is searched based on the starting address of the entry, and the data management element counter 161 of the data management table is checked. Then, it is checked whether the count value of the counter 161 is smaller than m. Note that m at this time is the total number of data management elements in the data management table.

Then, the count value of the counter 161 is m
If it is smaller than , it is determined that there is an unregistered data management element in the data management table, and registration processing is performed. This registration process searches for data management elements whose registration information area 151 is all "1", increments the data management element counter 161 by 1 in the same way as above, and registers each area 151 to 154 of the unregistered data management element.
"0", the length of the key name, the key name, and the data are stored in each of the fields. In this way, the registration process ends.

If the count value of the counter 161 is equal to m, it is determined that there is no unregistered data management element in the data management table, and the next table pointer 1 is
Similar processing is performed on the data management table specified by the pointer 62. If there is an unregistered data management element in the data management table, the same registration process as described above is performed.

However, the next table pointer 162 is
If there are no unregistered data management elements even if the data management table is searched up to the data management table with the number "1", data management table creation processing is performed.

In this data management table creation process, a data management table is created in the same manner as described above, and at this time, the data management element counter 161 of the data management header is set to "0".
and the next table pointer section 162 is all "1".
Store. In addition, the registration information area 15 of each management element
Write all “1” to 1 as well.

Next, the next table pointer section is created. In this process, the start address of the newly created data management table is stored in the next table pointer section of the final data management table before creation of the new data management table, and a link is created.

[0033] Next, registration processing is performed. This registration process increments the data management element counter 161 by 1, and stores "0", the length of the key name, the key name, and the data in each of the areas 151 to 154 of one data management element. In this way, the registration process ends.

Furthermore, the data search process is performed by inputting the key name of the search information register 18 into the hash function section 19 to obtain a hash value, and from this hash value, the hash table 1 is
Search for 7 entries. Then, the data management table is searched from the start address of the searched entry 171, and the key name stored in the key section 153 of each data management element of the data management table is compared with the key name of the search information register 18. Search for the appropriate data management element. At this time, unregistered data management elements whose registered information areas are all "1" are not subject to key name verification.

[0035] First, the length of the key name is compared, and if the lengths of the key names match, then the key name itself is compared, and if the key names also match, the search is completed and the target data is searched. is read from the data section 154.

In the embodiment with such a configuration, when deleting data, the hash function section 19 calculates a hash value from the key name stored in the search information register 18, and the entry in the hash table 17 is determined based on the hash value. search for. Then, the corresponding data management table is searched based on the start address of the searched entry.

Then, the key name is read from the key name section 153 of the data management element in which data is registered among the data management elements of the corresponding data management table, and the length and key name are compared with the key name in the search information register 18. do.

If a data management element whose key name length and key name both match is found, the registration information area of that data management element is set to all "1" and the data is deleted. In this way, data can be deleted by a simple process of simply writing all "1"s into the registration information area of the data management element.

[0039] Furthermore, in searching for a corresponding data management element, data management elements whose registration information area is all "1" can be skipped without being compared.

Furthermore, since each data management element of the data management table is formed consecutively in the storage area, when the verification of one data management element is completed and the next data management element is verified, the data management element is If the size is t, it is sufficient to simply add t to the start address of a certain data management element, and even if there is an unregistered data management element in between, t may be added accordingly.

[0041] In this way, when deleting data, the address for searching for the next data management element can be easily specified, so the time for data deletion processing can be shortened, and the search processing for data management elements can be made more efficient. .

This can similarly reduce the processing time when searching for a data management element whose registration information area is all "1" during the search process when reading data or when registering data. Moreover, only one hash function section and one hash table are used, and the configuration is simple.

[0043] In the above embodiment, the data management element is provided with a key name section and a data section to store the key name and data respectively, but the invention is not necessarily limited to this. A pointer section to the name and a pointer section to the data may be provided, and the actual key name and data may be stored in a separate storage area indicated by the pointer in the pointer section.

[0044]

Effects of the Invention As detailed above, according to the present invention, it is possible to shorten the time required to delete or register data, improve processing efficiency, and provide information deletion registration with a simple configuration. It is possible to provide a method.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a flowchart showing information deletion processing in the same embodiment.

FIG. 3 is a diagram showing an example of information deletion in the embodiment.

FIG. 4 is a flowchart showing information registration processing in the same embodiment.

FIG. 5 is a block diagram showing a conventional example.

[Explanation of symbols]

11... Storage device, 12, 13, 14... Data management table, 12a-12c, 13a-13c, 14a-14c
...Data management element, 12d, 13d, 14d...Data management header, 17...Hash table, 19...Hash function section.

Claims (1)

[Claims] Claim 1: When a hash value is obtained from search information using a hash function and data is deleted or registered in a storage device by referring to a hash table using the hash value, whether data is registered in the storage device or not. A data management element consisting of a registration information area where information indicating whether or not the key name is stored, an area where key name length information is stored, and an area where information that manages the key name and data name is stored for the storage area. Multiple data management tables are provided that consist of data management headers that store multiple data in series to form groups and manage each data management element in the group, and when deleting data, the hash table is referenced from the hash value of the data to be deleted. Search the corresponding data management table, search the corresponding data management element from the data management table, write data indicating unregistered to the registration information area of the data management element, and when registering the data, hash the data. The hash table is referenced from the value to search for the corresponding data management table, and among the data management elements in the data management table, the data management element in which data indicating unregistered is written in the registration information area is searched. , the corresponding data is written in the registration information area of the searched data management element, the area where the length information of the key name is stored, and the area where the information for managing the key name and data name is stored, respectively. Information deletion registration method.