JPH06149634A - Data base device - Google Patents

Abstract

PURPOSE:To provide a data base device capable of easily grasping relation between data to be newly registered or changed and registered data, simplifying the management of a table and data and flexibly corresponding to a using procedure which is not supposed at the time of design. CONSTITUTION:The sequence of plural words is defined as a line, plural lines are indivisually/selectively registered/deleted in/from plural tables respectively matched with plural conditions and the lines are retrieved based upon the contents of each word. The data base device is provided with an input means 1, a storage means 2, a table specifying means 3, a management means 4, a registering means 5, a sorting means 6, a retrieving means 7, a deleting means 8, and an output means 9. The means 3 specifies plural corresponding tables to one line based upon a registering condition indicating relation between the lines and tables and the means 5 registers the line in respective tables. The registered line is sorted by the means 6 based upon the sorting rules of respective tables.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a database device.

[0002]

BACKGROUND OF THE INVENTION The purpose of database (DB) technology is to make the internal structure of data independent of the application program that uses it.

The conceptual framework of the conventional DB system includes typical ones such as hierarchy, network and relational. Of these, the relational database (RDB) is the most theoretically organized and is now the standard technology. The basic concept of this RDB is "relationship". This point will be described below.

For example, a Cartesian product set R of _n sets S ₁ , ..., S _n is defined as follows.

[0005]

[Equation 1] At this time, an arbitrary subset of R is called an n-ary relation (or simply relation or relation) between S ₁ , ..., S _n . For example, if two sets S ₁ and S ₂ are expressed by the following equation (2)
Define as shown in.

[0006]

[Number 2] S ₁ = {Tokyo, Kanagawa} S ₂ = {Mitaka, Musashino, Kawasaki, Yokohama} Equation (2) In this case, two of the set S _1, Cartesian product of S ₂ S ₁ × S ₂
Is expressed by the following equation (3).

[0007]

[Formula 3] S ₁ × S ₂ = {(Tokyo, Mitaka), (Tokyo, Musashino), (Tokyo, Kawasaki), (Tokyo, Yokohama), (Kanagawa, Mitaka), (Kanagawa, Musashino), (Kanagawa, (Kawasaki), (Kanagawa, Yokohama)} ... Equation (3) Among these, the subset R as shown in the following Equation (4) is the prefecture-city relation.

[0008]

[Equation 4] R = {(Tokyo, Mitaka), (Tokyo, Musashino), (Kanagawa, Kawasaki), (Kanagawa, Yokohama)} ... Equation (4) RDB is a DB that uses such a set of relations as data. Is. In short, the relation is shown in FIG. 52 and FIG.
3 is a table as shown in FIG. In general, one row is referred to as a tuple (set, tuple, record) when the table is viewed horizontally, and one column is referred to as an attribute when viewed vertically. The name given to each attribute in the table of FIG. 52 is called an attribute name (attributename),
The value of each attribute is called an attribute value (field value). Also, the name of the table is called the relation name.

As described above, in the RDB, the data record is stored in the computer by using the relation table as a conceptual mediation. Of course, it is highly possible that the computer has complicated data structures and access procedures. However, the user cannot see the complicated implementation.
Information about the implementation is hidden. It's just a table, or container of data, where the user puts records into the container.

The RDB has the following four processing functions. : "1. Set a table that contains records.""2. Register records in the table.""3. Delete records from the table.""4. Search for records in the table." , RDB implementation, if you specify the attribute and attribute value,
The record registration processing, the deletion processing, and the search processing can be efficiently performed.

One of the characteristics of RDB is that tables and records are highly independent. The user first sets the table,
Then record the record in the table. It is up to the user to register a certain record in a certain table.
Even if the attribute value of the retrieved record is changed, the correspondence between the record and the table does not change. In order for the user to change the correspondence between the record and the table, the record registration and deletion operations as described above are necessary. Even if you specify multiple tables and perform a calculation between the records registered in them, and register the records generated in the specified new table, which table the newly generated record is registered in , Managed by the user. That is, in the RDB, the correspondence between the table and the record is given only through the act of the user registering an arbitrary record in an arbitrary table.

As explained above, a relationship is a set. Generally, the definition of a set includes an extensional method and an inclusive method. The extensional method defines a set by listing all the elements. For example, the extensional definition is to define a set S of whole integers of 0 or more and 9 or less as shown in the following expression (5).

[0013]

## EQU5 ## S = {0,1,2,3,4,5,6,7,8,9} ... Equation (5) On the other hand, the property to be satisfied by the elements belonging to the set is defined. Is an inclusive definition. For example, the inclusive definition is to define the above set S as in the following expression (6). Here, Z is a set of integers.

[0014]

[Equation 6] On the other hand, the act of registering a record in the table is that "this record belongs to this relation (table) and does not belong to other relations (table) where this record is not registered." It is defined by enumerating the values. Since this is done for all records, the relation is defined externally.

As a result, the RDB has the following features. First, the processing performed according to the relation is conceptually simple and can be performed efficiently. That is, the program is simple and the amount of calculation is small. For example, in FIG.
The process “(A) Search for a person whose birth year is '63.” Is a process according to the relation and can be efficiently performed.

[0016]

On the other hand, the RDB
In the above, the process that does not follow the relation is a process that was not originally assumed, and thus is inefficient. For example, in FIG. 53, the process of “(B) searching for a third person in the order of youngest birth” is a process that does not follow the relation. This process can be easily executed if the table of FIG. 53 originally has an attribute of “order of birth”. Nevertheless, FIG.
In the table, since the attribute of "year of birth" is not considered in advance, in order to perform the processing of (B), after checking all the records having a year of birth among the registered records, , The order of these records must be determined. When executing such a search, the program becomes long and the amount of calculation becomes large.

Further, in this case, the attribute "birth year order" used in the process (B) is naturally determined from the attribute "year of birth" used in the process (A). That is, the attribute "year of birth" is an inclusive property that defines a subset of the relationships of records having the attribute "year of birth".
It exists naturally. Nevertheless, since the attribute "order of birth" is an attribute that does not exist in this RDB, the process (B) using this attribute cannot be easily executed.

From the above, it can be seen that the RDB is suitable for the usage envisioned at the time of its design, but on the other hand, it is inefficient if the usage is a little off the assumption. Here, the usage assumed at the time of design is to handle a record with an attribute used for an extensional definition and its value. In addition, the usage that is slightly outside the assumption is to handle a record that is internally defined by an attribute existing in the RDB and a property derived from the value thereof, as in the processing of (B) described above.

In general, it is inevitable that the usage will deviate from the assumption at the time of designing after the DB has been designed for a long time. This is because a DB has a longer life than an application program that uses it. That is, in most cases, the DB already exists and the program is D
It is written with the intention of using the information that exists in B. As described above, when the usage of the DB is out of the assumption at the time of designing, if it is related to the above-described processing with respect to the table of FIG. 53, it is sufficient to search for the attribute "year of birth" at the time of designing the DB. Although it was thought, there are cases where it is later found that it is actually easier to use in order of age. In such a case, in the RDB, a new attribute “in order of age” will be provided. This is a rework of the table, and it is a difficult task when there are many records.

This is due to the essential feature of the conventional database that the correspondence between tables and records (registration operation) is arbitrary. That is, when constructing and using the database, the user needs to perform two operations associated with each other: "A. Create / change record" and "B. Register / delete record in table". Because when saving data, if there is a record in the database, not only the existence of this record but also which table is registered and which table is not registered Is important. Therefore, when one record represents the contents to be registered in a plurality of tables, the user must judge it and register the record individually in the corresponding plurality of tables. .

For example, in the event schedule management database, in the table (table name: FUTURE EVENT) that manages future events, a record of "time event" is registered to represent the event and its occurrence time. It Here, it is assumed that an unspecified number of participants are scheduled for each event. On the other hand, create a table (table name: ATTENDANT EVENT) about this planned participant,
Register the record of "participant event" in this table. Then, the record of the "time event" registered in the table "FUTURE EVENT" is deleted when the event occurs, and registered in the table (table name: PAST EVENT) that manages the event that has already occurred.

In this case, each table effectively registers only records in a format suitable for its purpose.
You can search for what kind of information is registered for one event in three tables (FUTURE EVENT, ATTENDANT).
EVENT, PAST EVENT) cannot be executed without checking individually. For example, a "time event" for an event
The record of is "FUTURE EVENT" or "PAST EVENT"
It must be in either one or the other, but both tables must be examined at the same time to confirm this. Further, when an event occurs, it is the responsibility of the user to simultaneously perform the two processes of deleting the record related to the event from one table and registering the record related to the event in the other table. There is a possibility that an error will occur here, and it takes time and effort to detect the error.

As one method of avoiding such a problem, the two tables "FUTURE EVENT" and "PAST EVENT" are discontinued and, for example, one table (table name: EVENT) is created and this table is created. It is possible to register the event schedule as a record of "FUTURE / PAST time event" in "EVENT". Here, the field "FUTURE / PAST" means FUTURE or PAST. When this method is adopted, when an event occurs, two processes (deletion and registration) for different tables are performed as in the above method.
Since it is only necessary to perform one process of changing the field value of “FUTURE / PAST”, it is extremely unlikely that an error will occur, and error detection will be relatively easy.

On the other hand, in order to proceed with the processing by this method, all the tables for registering related records are finally merged into one table, so that many tables that are not effectively used are used. A long record including fields will be registered. Registering a large number of wasteful long records is nothing but abandoning the efficient operation of the database by setting a table for storing the records.

As described above, in the conventional database system in which the table is set, and the record is arbitrarily registered and deleted from the table to efficiently store and use the data, the correspondence between the table and the record is provided. It takes a lot of effort to manage the relationship intentionally, and the cost is increased accordingly. In addition, in the conventional database, regarding the content of the record, that is, the field value, whether or not the record having a certain field value can be registered in the specified table (as a correct record) is restricted There is no function to indicate whether the table should be registered or a function to actually register automatically.

Therefore, in a conventional database, for a specific field value, what kind of record has that field value, and what kind of table the record having that field value is registered in is searched. It is difficult. That is, the field value is
It is a character string or a numerical value, and what it means in each record (each field of) is arbitrary. Therefore, it is quite possible that the same field value is used in a completely different meaning between different records in different tables. Such different use of the same field value causes confusion of data processing in the database and is extremely inconvenient. Therefore, when a user creates a record, it is essential to understand how each field value relates to the field values already used in the database at that time. Thus, in order to gain a complete understanding of the information of a particular field value throughout the database, it is necessary to examine all the records in the table that may be relevant (including records) in a conventional database and make a conclusion. There must be. However, execution of such a search requires a great deal of labor, and it is difficult for the user to maintain and manage the information of the table or field that may possibly be set for each field value.

The present invention has been proposed to solve the problems of the prior art as described above, and its purpose is to provide data to be newly registered or data whose contents are changed and data already registered. The purpose of the present invention is to provide a database device which can easily grasp the relation, can manage tables and data easily, and can flexibly deal with the usage that was not assumed at the time of design, and has excellent operability and efficiency.

[0028]

The basic features of the database device according to the present invention will be described with reference to FIG. Here, FIG. 1 is a conceptual diagram showing an outline of the configuration of the database device according to the present invention. That is, as shown in FIG. 1, the apparatus of the present invention uses a sequence of a plurality of words as a line, and the plurality of lines are individually / selectively registered / deleted in a plurality of tables that meet a plurality of conditions. In addition, in the database device for searching the line based on the content of the word, the input means 1, the storage means 2, the table designating means 3, the managing means 4, the registering means 5, the sorting means 6, the searching means 7, the deleting means 8, It is characterized in that an output means 9 is provided.

The input means 1 designates the function of inputting a sequence of a plurality of words as a line and the already registered line by the content of the word or the position of the line,
It has a function of inputting a search command, a delete command, or a change command. The storage unit 2 has a function of storing a plurality of table information including sort rules set in each table, line information registered in each table, and sort order information thereof.

The table designating means 3 has a preset registration condition indicating the correspondence between lines and tables, and the input means 1
In response to the input of a line by the, the function of designating a plurality of tables for registering the line based on the content of the word of the input line and the registration condition, and the search command and the deletion of the registered line by the input means 1. In response to the input of a command or a change command, it has a function of designating a plurality of tables in which the line is registered based on the content of the word of the designated line and the registration condition.

The management means 4 manages a plurality of table information stored in the storage means 2 and provides a plurality of table information designated by the table designation means 3, and registration by the registration means 5 and the sorting means 6. It has a function of updating the management information with the information and the deletion information by the deletion means 8 and a function of managing the search information by the search means 7.

The registering means 5 receives a plurality of table information designated by the table designating means 3 from the managing means 4 in response to the line input by the inputting means 1 and inputs the information to each table by the inputting means 1. It has the function of registering each line. The sorting means 6 has a function of sorting the lines registered in the plurality of tables by the registration means 5 on the table based on the content of the word at the position specified by the sorting rule set in each table.

The search means 7 responds to an input of a search command, a delete command, or a change command of a registered line from the input means 1 from the management means 4 for a plurality of table information designated by the table designating means 3. It has a function of receiving and retrieving the registered lines designated by the input means 1 on the basis of the order sorted by the sort means 6 on each table. The deleting means 8 receives the plurality of table information searched by the searching means 7 in response to the input of the delete command or the change command of the registered line by the input means 1, and displays the searched line on each table. Has the function to delete. The output unit 9 has a function of outputting the management information by the management unit 4.

Further, generally, the managing means 4 has a table creating means 10. The table creating means 10 has a function of creating a new table that matches the specified table when the table specified by the table specifying means 3 does not exist in the storage means. On the other hand, as a word that composes a line, for example, one of an ordered pair of two character strings, an ordered pair of character strings and an integer, an ordered pair of two integers, or one of four floating point numbers is used. It is possible to More generally, strings, integers, floating point numbers, and any number of ordered sets thereof can be used.
Further, in the present invention, as the sorting means 6 for sorting the lines on the table, for example, a binary search tree can be used, but like the binary search tree, line sorting, search, registration, and If it is possible to delete, it is possible to use another sort means.

Next, in the present invention, the registration condition preset in the table designating means 3, that is, the constant registration condition indicating the correspondence between the line and the table will be described. First, one line is composed of a sequence of n words (w ₁ , ..., W _n ). Then, k types of tables for registering one line are represented as H ₁ , ..., H _k . In the present invention, one line is not always registered in the k types of tables, and the number of tables registered may differ depending on the content of the line.

Where x = word or line, N
= 1 to the number of n, S = as a sorting rule, each table is (x,
Consider what is represented by N, S). In this case, if x is a word, then (x, N, S) represents a table whose lines are such that the Nth word from the left is the word x and the sorting rule is S. If x is a line, (x, N,
S) represents a table consisting of lines where the Nth word from the left is the Nth word of line x and the sorting rule is S. Further, when i = 1, ..., K, the i-th table Hi is expressed as (x, Ni, Si).

[0037]

The operation of the present invention having the above construction is as follows. First, when registering a line, the table designating means automatically designates a plurality of tables to be accommodated according to a plurality of words forming the line based on a registration condition indicating a correspondence between the line and the table. Yes, it can be easily registered by the registration means. Therefore, it is possible to flexibly deal with the usage that was not assumed at the time of design, and the operability and efficiency can be improved.

Further, in this case, the lines registered in each table are automatically sorted by the sorting means based on the sorting rule set in each table, and the table information including this sort order information is managed by the managing means. Managed by. Therefore,
Since the table information including the line information and the sort order information can be automatically managed by the management means, it is possible to simplify the management of the relationship between the line and the table and the management of the line configuration of each table that the user should perform.

On the other hand, in the present invention, the table designation means
A table in which one word is stored can be automatically specified, and a line can be easily searched by one word by using the search means in addition to the table designating means. Therefore, the information of one word can be easily grasped over the entire database, and the relation between the line already registered in the database and the newly registered line can be easily grasped. In addition, since the sorting means is used, it is possible to automatically and easily search for a line by the sort order and determine the position of the searched line. The efficiency of change processing can be improved.

The operation of the present invention will be described below with reference to FIGS.
A more specific description will be given with reference to FIG. Here, FIG.
FIG. 5 is a flowchart showing an example of a general processing procedure of the database device according to the present invention. FIG. 2 shows a registration process, FIG. 3 shows a search process, FIG. 4 shows a delete process, and FIG. 5 shows a change process. There is.

First, in the case of registering a line, as shown in FIG.
N number of words that make up the line _{(w 1, ..., w n} ) is created, in step 2, n number of words (w _1,
A line L that is a sequence of ₍ ..., W _n ) is created and input. In step 3, the table designating means 3 registers k types of tables H ₁ , ..., H _k for registering the line L based on the contents of the word of the line L and preset registration conditions.
Is automatically specified. In step 4, the management means 4 judges whether or not a plurality of designated tables exist in the storage means 2, and if there is a non-existing table among the designated tables H ₁ , ..., H _k. In step 5, the table creating means 10 creates a necessary table. In step 6, the designated k types of tables H ₁ , ..., H
The information of _k is provided to the registration means 5, and by the registration means 5,
The line L is registered in each table. In step 7, the sorting means 6 sorts each table based on the sort rule S set in each table, and the registration process ends.

Next, when searching for a line, as shown in FIG. 3, the line L to be searched is designated by the input means 1 in step 11, and the line L is designated by the table designating means 3 in step 12. Based on the content of the word and the registration condition set in advance, k types of tables H ₁ , ..., H _k in which the line L is registered are automatically designated. In step 13, the designated k types of tables H ₁ , ..., H
The information of _k is provided to the search means 7, and the search means 7 searches the specified line L in the sort order in each table, and the search processing is ended.

When deleting a line, as shown in FIG. 4, the line L to be deleted is designated by the input means 1 in step 21, and the line L of the line L is designated by the table designating means 3 in step 22. Based on the content of the word and preset registration conditions, k types of tables H ₁ , ..., H _k in which the line L is registered are automatically designated. In step 23, the specified k types of tables H ₁ , ..., H
The information of _k is provided to the search means 7, and the search means 7 searches the specified line L in the sort order in each table. In step 24, the information of each designated table and the searched line position in each table is provided to the deleting means 8, and the deleting means 8 deletes the line L from each table, and the deleting process ends.

On the other hand, in the case of changing the line, as shown in FIG. 5, the line L to be changed is designated by the input means 1 in step 31, and the n lines after the change are designated in step 32 and step 33. A word and its modified line L'that is a sequence thereof are created and input. In step 34, the table designating unit 3 registers the k types of the table H in which the line L before the change is registered based on the content of the word of the line L before the change and the preset registration condition.
₁ , ..., H _k are automatically specified. In step 35, the information of the designated k types of tables H ₁ , ..., H _k is provided to the search means 7, and the designated line L is searched in the sort order in each table. In step 36, information on each designated table and the searched line position in each table is provided to the deleting means 8, and the deleting means 8 deletes the line L before the change from each table. In step 37, the table designating means 3 registers k types of tables H ₁ ′, ..., H _k ′ for registering the line L ′ based on the changed word content of the line L ′ and preset registration conditions. Is automatically specified. In step 38, the management means 4 judges whether or not a plurality of designated tables are present in the storage means 2, and among the designated tables H ₁ ′, ..., H _k ′, a table that does not exist is found. If there is, in step 39, the table creating means 10 creates the necessary table. In step 40, the designated k kinds of tables H ₁ ′,
.., H _k ′ information is provided to the registration means 5, and the registration means 5
Thus, the changed line L'is registered in each table. In step 41, the sorting means 6 sorts each table based on the sorting rule S set in each table, and the changing process is completed.

In the exemplary embodiment of the invention, a line is made up of three words. Where word is a string string, a string integer,
Integer is one of an integer ordered pair or is a floating point number. When registering one line, this line is automatically registered in four tables at the same time due to the difference between the three words forming the line and the sort key. These four tables show (b, H), (b, H when three words of one line to be registered are a, b, c in order from the left.
M), (a, h), and (c, m). here,
(B, H) represents a table in which the central word is b, and the left word is the sort key. (B,
M) represents a table in which the central word is b and the right word is the sort key. (A, h) is
The left word is a collection of lines with a and the middle word is the sort key. (C, m) represents a table in which the right word is c, and the central word is the sort key. As will be described later, for example, when registering one line “customer 1 TRACE TRECE 800 −1” in the database, this line includes four tables (TRACE TRACE, H), (TRACE TRACE, M), ( cust
omer 1, h) and (800 -1, m) are registered respectively.
When deleting one line, the line is automatically deleted from all four tables in which the line is registered. Further, when changing the contents of one line, the line is once deleted from the four tables, the line is changed, and the changed new line is registered again. That is, when one line is newly registered, a maximum of four tables are newly created.

[0046]

DESCRIPTION OF THE PREFERRED EMBODIMENTS As a typical embodiment of the present invention, an example of performing a discrete event simulation by a database device having the configuration shown in FIG. 1 will be described below. In addition,
The means (functions) constituting the database apparatus of FIG. 1 and the operation thereof are as described above. Therefore, only the features of this embodiment will be described below.

As a premise for explaining the embodiment, the discrete event simulation will be described. In discrete event simulation, the initial state of the model to be simulated is described in the database. Then, the time is advanced to the time of the earliest event in the future event list in the database, and the event is performed. Here, performing an event is to reflect the change in the model caused by the occurrence of the event in the database describing the model. That is, as shown in FIG. 6, the simulation is executed by rewriting the database according to the event.

In this embodiment, the discrete event simulation as described above is executed for a simulation model as shown in FIG. FIG. 7 assumes a model of a bank having only one window, and shows a state at a certain point in time when customers receive services in the order in which they come to the windows and leave when the services are completed. In this case, multiple customers arrive sequentially at different times, and if the previous customer is in service, wait in line until the service of the previous customer is over,
The customer shall leave when the service for the customer himself has ended. This model has 5 customers.

FIG. 8 is a table showing all the lines showing the state of the model of FIG. 7, and if these lines are registered according to the conditions described later, from the many tables as shown in FIGS. Database is created. In other words, FIGS. 9-12 show the entire database representing the state of the model of FIG. FIG. 13 illustrates one table that constitutes this database. As shown in FIG. 13, the table is composed of one or more line sets collected under a certain condition. FIG. 14 illustrates one line forming the line set of FIG. As shown in FIG. 14, one line consists of three words.

In this embodiment, four types of words as shown in the table of FIG. 15 are set as usable words. FIG. 16 shows examples of the four types of words shown in FIG. 15, respectively. That is, a type 1 word is a pair of two strings, a type 2 is a pair of strings and an integer, and a type 3 is 2
It is a pair of two integers. Also, the type 4 words are not pairs, but are made up of one floating point number.

The order of alignment between words is determined as follows. That is, basically, the cells are arranged from top to bottom from the smaller side to the larger side. The molds 1, 2, and 3 are smaller than the mold 4. Further, the comparison between the types 1, 2, and 3 is first made by comparing the left element of the word pair. If the elements on the left are the same, it is determined by comparing the elements on the right. Comparisons between integers and floating-point numbers use the magnitude relationship of those numbers as they are. Further, the comparison between the character string and the integer is made such that the character string is smaller, and according to the comparison result, the smaller one is arranged on the upper side and the larger one is arranged on the lower side. On the other hand, between character strings, the character strings are arranged from top to bottom according to the order used in the database.

In this embodiment, a line is not just a group of three words, but its order is extremely important. In the present embodiment, in the illustrated table, from the left word in the drawing to the right, they are referred to as the left word, the center word, and the right word, respectively. In this embodiment, 1
When one line is registered in the database, it is automatically registered in four tables at the same time depending on the contents of the line and whether the word to be aligned is left, center or right. If there is no table to be registered, create a new table and register it.

When representing four tables for registering one line like this, (x, H), where x is an arbitrary word,
Notations such as (x, M), (x, h), (x, m) are used. In this case, (x, H) is a table consisting of a set of lines with the central word being x, with the lines aligned by the left word. (X, M) is a table that consists of a set of lines whose central word is x and that has lines aligned by the right word. In (x, h), the left word is x
Is a table that consists of a set of lines, with lines aligned by the central word. In (x, m), the right word is x
Is a table that consists of a set of lines, with lines aligned by the central word. For example, four tables for registering one line “customer 1 TRACE TRACE 800 −1” shown in FIG. 14 are (TRACE TRACE, H), (TRACE TRA
CE, M), (customer 1, h), (800 -1, m). Each table is shown in FIGS.

The database of this embodiment is as follows.
Since four tables are prepared for one line, FIG.
As shown in FIGS. 12 to 12, a large number of tables are assembled.
Further, in the present embodiment, the description of the model state mainly includes the following four types of lines.

First, there is a line that represents the state of the event.
It is a list of events that will occur in the future. The word to the right of this line represents the type of event. In this embodiment, there are two events, an arrival event and a departure event. In the following, these events are represented by "AR" and "LV", respectively. That is, the arrival event "AR" performs the arrival of the customer at the bank and the exit event "LV" performs the departure of the customer from the bank. The central word is a word indicating whether the event has already been executed or an event that will occur in the future, and is "PAST EVENT" or "FUTURE EVENT". The left word represents the event occurrence time. For example, the line "sec 1000 FU
"TURE EVENT AR 0" is an event that is about to happen,
The occurrence time is 1000 sec, which indicates that the event that occurs is the arrival event “AR”.

Secondly, there are lines representing the states of five customers and the service hours of the customers in this model. The word to the left of this line is the customer's ID number, the word in the center is the status of the customer, and the word on the right is the service time. Service time is the time required to finish the service when the customer enters the window. There are four types of customer states,
It is shown in FIG. As shown in FIG. 21, the state where the customer has not yet arrived at the bank (not yet arrived) is “COMING I
N ". "IN LINE" refers to a situation where customers arrive at the bank but are unable to receive service at the counter and are standing in a line. "BEING SERVED" means that the customer is receiving the service at the counter. The state where the customer has finished the service and left the bank is represented as "GONE AWAY". For example, the line "CUSTOMER 1 BEING SERVED sec 400"
The status of the customer with number 1 (called customer 1) is "BEING SERVE".
It indicates that the customer 1 is in the "D" (service is received) state and the service time required by the customer 1 is 400 sec.

Third, there is a line for storing customer tracking information. On the left side of this line is a word indicating the customer's ID number, and in the center is the word "TRACE TRACE" indicating that this line is the tracking information of the customer. The left element of the right word represents the time that the customer arrived at the bank. The right element of the right word represents the time when the customer has left the service and left the bank. Initially, enter negative numbers "-1" for these elements.

Fourth, there is a line showing the average time spent at the bank by customers who have left the service after leaving the service at the bank. This line has the word "MEAN RESPONSE" in the center. The left element of the left word represents the total stay time of the guests who left, and the right element of the left word is
Represents the number of guests who have left. The word on the right represents the average stay time (floating point number).

In this embodiment, changing any of the three words in the line will automatically register and align in the four tables to be registered. For example, assume that the word to the right of the line "customer 1 TRACE TRACE 800 -1" shown in FIG. 14 is changed to "800 1200". In this case, the line before change "customer 1 TRACE TRACE 800
-1 "means the four tables (TRACE TRACE, H), (T
RACE TRACE, M), (customer 1, h), (800 -1,
All are deleted from m). And the new line "cu
stomer1 TRACE TRACE 800 1200 "has four new tables (TRACE TRACE, H), (TRACE TRACE, M),
Registered at (customer 1, h), (800 1200, m).
In this case, of the four registered tables, the previous three tables are the same as before the change, but the position of the line in the table may change.

Further, in the present embodiment, when the table is searched, the table can be searched by designating the table (x, A). Here, x is a word and A is any of H, M, h, and m. Figure 13 shows (custom
er 1, h). Also, the line can be searched by designating it as the line of the table (customer 1, h). Further, it is possible to search for a line based on the contents of the line. Also, in the table (customer 1, h), it is possible to search for a line whose right word is "sec 350". In addition, it can be determined where in the table the retrieved line is.

The specific simulation operation in this embodiment will be described below.

First, using the databases shown in FIGS. 9 to 12, the event that occurs next is searched. In this case, the next event is described in the top line “sec 1000 FUTURE EVENTAR 0” of the table (FUTURE EVENT, H) that represents the events that will occur in the future in the order shown in FIG. 9 in the order of time. .
This line shows that the arrival event "AR" occurs at time 1000 sec. When the arrival event "AR" at this time 1000 sec occurs, this arrival event "A"
The R'rewrites the data shown in FIGS. 9 to 12 as follows.

In the arrival event "AR", the customer having the smallest ID number is searched for among the customers who have not arrived yet and the customer is made to arrive at the bank. Lines for customers who have not arrived are listed in the table (COMING
IN, H), and this table is arranged with the left word representing the customer's ID number as the sort key. Table showing this unarriving customer in order of ID number (COMING IN, H)
Is shown in FIG. When the top line of this table (COMING IN, H) is searched, the customer 4 is obtained from the word on the left. Therefore, the customers 4 arrive and are lined up. That is, the state of the customer 4 is changed from “COMING IN” to “IN LIN
Rewrite to "E", line "customer 4 COMING IN s
ec 300 ”to the line“ customer 4 IN LINE sec 3
00 ". As a result, the line "customer 4 COMING IN sec 300" before the arrival event "AR" is deleted from the four tables in which it was registered before the arrival event "AR" was executed, and the arrival event "AR" is newly added. "The line" customer 4 IN LINE sec 300 "after execution corresponds to 4
Registered in one table. The table (IN LINE, H) in Fig. 23 is
The customers in the queue after execution of the arrival event "AR" are shown in the order of ID numbers. It can be seen that customer 4 joins and customers 2, 3 and 4 are lined up in that order.

Next, the line concerning the tracking information of the customer 4 is searched and the arrival time at the bank is recorded. In this case, the tracking information is shown in the table (TRACE TRACE, H). If you search for a line in this table that has the word "customer 4" on the left, you will find the line "customer 4 TRACE TRACE -1 -1".
Is obtained. This line "customer 4 TRACE TRACE-
The left element of the right word of "1 -1" is rewritten from the initial value "-1" to the time value "1000" of this arrival event "AR" and the line "customer 4 TRACE TRACE -1 -1" is changed to the line " customer 4 TRACE TRACE 1000 -1 ". As a result, the line "custo" before the arrival event "AR" is executed
mer 4 TRACE TRACE -1 -1 "was deleted from the four tables in which it was registered before the arrival event" AR "was executed, and a new line" customer 4 after the arrival event "AR" was executed.
TRACE TRACE 1000 -1 "includes tables (TRACE TRACE, H) and (TRACE TRACE, M) that represent customer tracking information. 4
Registered in one table.

Next, the line "sec 1000 FUTUR" representing the arrival event "AR" at this time 1000 sec.
Change the word in the center of "E EVENT AR 0" to "FUTURE EVENT
"To" PAST EVENT ". As a result, the line “sec 1000 FUTURE EVENT before the arrival event“ AR ”is executed
"AR 0" has been deleted from four tables, including a table (FUTURE EVENT, H) that represents future events in the order of customer ID numbers, and a table (FUTURE EVENT, M) that represents future events in the order of event types. The line "se" is newly added after the arrival event "AR" is executed.
c 1000 PAST EVENT AR 0 ”is registered in four tables including tables (PAST EVENT, H) and (PAST EVENT, M) that represent events that have occurred in the past.

FIG. 24 shows the state of the model after the arrival event "AR" is executed at the time 1000 sec as described above. FIG. 25 is a table showing all the lines representing the state of the model of FIG. 24. By registering each line in the corresponding four tables, a database consisting of a large number of tables as shown in FIGS. Is created. This FIG.
In the database shown in FIGS. 29 to 29, the next event is the top line “sec” of the table (FUTURE EVENT, H) showing the events that will occur in the future in the order shown in FIG.
1200 FUTURE EVENT LV 0 ”. This line indicates that the evacuation event "LV" will occur at time 1200 sec. Then, when the evacuation event "LV" occurs at this time 1200 sec, the data shown in FIGS. 26 to 29 is rewritten as follows by the evacuation event "LV".

The exit event "LV" first searches the table (BEING SERVED, M) for the customer currently at the counter. A table (BEING SERVED,
M) has one line "c" as shown in FIG.
Only ustomer 1 BEING SERVED sec 400 ”is registered. From this line, it can be seen that the customer currently at the counter is customer 1. Therefore, the customer 1 is removed from the bank. That is, this line "customer 1 BEING SE
In the "RVED sec 400", rewrite the central word indicating the customer's condition from "BEING SERVED" to "GONE AWAY", and change the line "customer 1 BEING SERVED sec 400"
To the line "customer 1 GONE AWAY sec 400". As a result, the line "c" before the evacuation event "LV" is executed
ustomer1 BEING SERVED sec 400 "is the exit event" L
It was deleted from the four tables that were registered before "V" was executed, and a new line "custome" after the evacuation event "LV" was executed.
r1 GONE AWAY sec 400 ”is registered in the corresponding four tables.

Next, the customer waiting at the head of the queue is put in the window. A table showing the customers in the queue in order of ID number (IN LIN
Search the top line of E, H). This table (IN LIN
The top line of (E, H) is the line "customer 2 IN LINE sec 300" as shown in FIG. From this line, it can be seen that the next customer to enter the window is customer 2. Therefore, the customer 2 is put in the window. In other words, this line "customer 2 IN LINE sec 300"
In, the central word that shows the state of the customer is "IN LINE"
From "BEING SERVED" to the line "customer 2"
IN LINE sec 300, the line "customer 2 BEI
NG SERVED sec 300 ”. As a result, the line "customer 2 IN LINE sec" before the evacuation event "LV" was executed
"300" was deleted from the four tables that were registered before the evacuation event "LV" was executed, and a new line "customer 2 BEING SERVED sec 300" after the evacuation event "LV" was executed.
Are registered in the corresponding four tables.

As a result of the above operations, a table (BEING SERVED, H) showing the customers currently at the counter, (BEING SERVED,
M) and a table (IN LINE, H) showing customers in the queue,
The contents of (IN LINE, M) are changed. In addition, as a result of changing the line for leaving customers as described above, a table showing newly leaving customers (GONE AWAY, H), (GONE AWAY, H)
AWAY, M) is created. FIG. 30 shows such a time 12
Three tables (BEING SERVED, H), (IN LINE, H) showing the state of the customer before the departure event "LV" at 00 sec,
(COMING IN, H) is shown in FIG.
Represents the state of the customer after executing the leaving event "LV" in sec 4
Two tables (GONE AWAY, H), (BEING SERVED, H),
(IN LINE, H) and (COMING IN, H) are shown.

In this embodiment, as described above, when a customer 2 newly enters the window, the leaving time at which the customer 2 leaves the window is determined by "current time + service time". That is, it is the time when the next evacuation event "LV" occurs. In this example, the departure time of customer 2 is "1200 + 300",
1500 sec. The next evacuation event "LV" will occur at 1500 seconds, so a new line for this event will be registered.
That is, the line "sec1500 FUTURE EVENT LV 0
Is a table of future events (FUTURE EVENT, H),
Register for 4 tables including (FUTURE EVENT, M). In addition, the word in the center of the above-mentioned line “sec 1200 FUTURE EVENT LV0” representing the leaving event “LV” at the time of 1200 sec is changed from “FUTURE EVENT” to “PAST EVENT”.
Change to ". As a result, the line "sec 1200 FUTURE EVENT LV 0" before the evacuation event "LV" is executed is displayed in the table of future events (FUTURE EVENT, H), (FUTURE EVENT, H).
M) is deleted from the four tables including the new line "sec 1200 PAST EVENT LV" after the evacuation event "LV" is executed.
"0" is a table of past events (PAST EVENT, H),
Registered in 4 tables including (PAST EVENT, M). Then, as a result of the above operation, the contents of the table (FUTURE EVENT, H) showing events that will occur in the future in time order are as shown in FIG.

Further, the line relating to the tracking information of the customer 1 is searched for, and the bank leaving time is recorded. That is, in the table (TRACE TRACE, H) showing the tracking information of customers in the order of customer ID numbers, the line "customer 1 TRACE TRAC for customer 1" is displayed.
The element to the right of the word to the right of "E 800 -1" is initialized to "-1".
Is rewritten to the time value "1200" of this evacuation event "LV",
Change the line "customer 1 TRACE TRACE 800 -1" to the line "customer 1 TRACE TRACE 800 1200". As a result, the line “custo
mer 1 TRACE TRACE 800 -1 ”has been deleted from the four tables that were registered before the exit event“ LV ”was executed, and a new line“ customer 1 after the exit event “LV” was executed.
TRACE TRACE 800 1200 ”is a table of customer tracking information (TRA
CE TRACE, H), (TRACE TRACE, M) are registered in four tables. Then, as a result of the above operations, a table (TRACE TRACE,
The contents of (H) are as shown in FIG.

FIG. 34 shows the state of the model after execution of the leaving event "LV" at the time 1200 sec as described above. FIG. 35 is a table showing all the lines showing the state of the model of FIG. 34, and by registering each line in the corresponding four tables, a database consisting of many tables as shown in FIGS. Is created.

The present invention is not limited to the above embodiment, and the number of words forming one line is not limited to three, and one line is formed by four or more words. It is also possible, and vice versa, it is also possible for only two words to form a line. Further, it is possible to freely set the sort rules of the table and the setting contents of certain registration conditions related thereto.
For example, it is possible to use a simple sort rule that only sorts words at a certain position, or conversely, a sort rule that uses multiple words as sort keys. The content of can be set appropriately.

More specifically, as a modified example of the above-described embodiment, as a table for registering one line composed of three words, in addition to the four tables in the above-described embodiment, the left word is the same. An example is conceivable in which a fifth table, which is composed of a certain line set and has the right word as a sort key, is made to correspond. Also, one line is made up of five words, and one line is registered as a table in which the first word is the same,
A second table having a second word as a sort key and a set of lines in which the second words are the same, the third word as the first sort key, and the fourth word as the second sort key , And several other types of tables may be configured to correspond to each other.

Incidentally, in the present invention, various means can be used as the sorting means for sorting the lines. For example, a binary search tree capable of sorting, searching, registering and deleting can be used. is there. The following is a brief explanation of this binary search tree.

First, a binary tree and some words are defined. As shown in FIG. 40, a binary tree is composed of nodes (circles) and branches (/ \) indicating parent-child relationships between nodes, and each node has only two child nodes at the maximum and one Every node other than the node has only one parent node. The lower left child node is called the left child node, and the lower right child node is called the right child node. The only node that has no parent node is called the root node. In FIG. 40, node (1) is a parent node of node (2), and node (2) is a right child node of node (1). Node (3) is the root node of the binary tree. The entire structure under the left (right) child node of the node is called the left (right) subtree. In FIG. 40, the subtree (4) is the right subtree of the node (1).

The binary tree used in the present invention has an object for which a sorting rule is defined, and the node of the binary tree points to the object (the line of the database in the above-mentioned embodiment). Is a binary tree such that the target pointed to by the node is larger (smaller) than the target pointed by which node of the left (right) subtree of the node. Whether to include the case on the left or the case on the right depends on the use of the binary search tree. However, the smaller one is above the table and the larger one is below the table.

For example, FIG. 41 is a binary search tree that sorts integers. Note that, in FIG. 41, for the purpose of simplification of the drawing, an integer which is an object designated by the node is directly written in the node. In FIG. 41, the integer "54" pointed to by the node (1) is the right subtree (2).
The integers "60", "58", "7" specified by the node of the throat
It is smaller than 1 "and" 56 ".

FIG. 42 is a flow chart showing an example of a search method using such a binary search tree. According to this flowchart, for example, when searching for the integer “9” in the binary search tree of FIG. 41, the following procedure is performed. This process will be described with reference to FIG. In the following description, N means the current node. First, the root node (3) in FIG. 41 is set to N. In this case, the integer “9” to be searched is smaller than the integer “46” designated by N. Therefore, the left child node (4) of N is newly set to N. In this case, the integer “9” to be searched is smaller than the integer “13” designated by N. Therefore, the left child node (5) of N is newly set as N. In this case, the integer designated by N is "9", and this node
This is a node for search.

FIG. 43 is a flow chart showing an example of a registration method using a binary search tree. According to this flowchart, for example, in the case of registering a node designating the integer “48” in the binary search tree of FIG. 41, the following procedure is performed. About this process
This will be described with reference to FIG. First, the root node (1) in FIG. 44 is set to N. In this case, the integer "48" to be registered
Is greater than the integer "46" indicated by N. Therefore, N
The right child node (2) of is newly set to N. In this case, the integer "48" to be registered is smaller than the integer "54" designated by N. Therefore, the left child node (3) of N is newly set to N. In this case, the integer "48" to be registered is smaller than the integer "49" specified by N, but since there is no child node on the left of the current node N, the left indicating the integer "48" with N as the parent node. Create a child node (4).

FIG. 45 is a flow chart showing an example of the deletion method in the binary search tree. According to this flowchart, the following processing can be performed. That is,
First, when deleting the node A, if the node A has no child node at all, as shown in FIG. 46, the node A is deleted as it is. If the node A has only left (right) child nodes, the node A is deleted and the left (right) subtree is replaced with the node A as shown in FIG.
Change to the position. On the other hand, when the node A has both left and right child nodes, as shown in FIG. 48, the node immediately preceding the node A in the sort order (this is the node on the rightmost side of the left subtree of the node A). Is set to N (state on the left in FIG. 48), and the nodes A and N are exchanged (operation in FIG. 48: left → center state). If the node A does not have a left subtree after the replacement, the node A is simply deleted. If the node A has a left subtree, the left subtree is replaced with the position of the node A (operation in FIG. 48: center → right state).

It is effective to use such a binary search tree as the sorting means for sorting the lines, but as described above, other sorting means can be used as long as the sorting, retrieval, registration and deletion of the lines are performed. It is likewise possible to use.

Next, an example of implementing the functions in the database according to the present invention will be described with reference to FIG. In this example, the six lines xuw, vxz, and
wxy, xxx, yxx, ywx are shown. Here, u, v, w, x, y, and z represent different words. When searching for a line, four different tables corresponding to each word can be used. For example, if the word is x, the four tables are (x, H), (x, M), (x, H)
h) and (x, m). These tables are represented as a binary search tree: trees (1) to (4) in FIG. 49. In this case, each node of the tree points to a line and the word that is the sort key for that line.
For example, the root node of tree (1) points to line wxy and further to word w, which is a sort key.
Conversely, the line also points to the corresponding node. The line designated by the node of a tree is included in the table represented by the tree. For example, since the tree (1) corresponding to the table (x, H) has a node indicating the line xxx, the line xxx is included in the table (x, H).

Since the left word of the table (x, H) is the sort key, each node of the binary search tree representing this table indicates the corresponding line and the left word. That is, each word in each table indicates the corresponding line, and
If it is (x, M), the right word is designated, and if (x, m), the central word is designated. For example, line xx in FIG.
x is pointed to by the nodes of the four binary search trees representing the four tables. On the contrary, the line xxx designates the node which designates this line xxx. That is, by designating this line xxx, it is possible to find four binary search trees expressing the four tables including this line xxx.

A method of deleting a line in the database shown in FIG. 49 having the above structure will be described. When deleting line xxx from the database, four tables (x, H), (x, H) containing this line xxx
The line xxx is deleted from M), (x, h), and (x, m). In this case, the node indicating the line xxx of the trees (1) to (4) in FIG. 49 must be deleted. Node deletion follows the procedure described above for binary search trees. As a result, the line xxx as shown in FIG.
A database without.

On the contrary, when registering the line xxx in the database as shown in FIG. 50, four tables (x, H), (x, M), (x, h), (x, m) are registered. ) Must be registered with each line xxx. In this case, the nodes indicating the line xxx are registered in the trees (1) to (4) corresponding to these four tables. The procedure of node registration follows the procedure described above regarding the binary search tree.

Furthermore, a search for an arbitrary line in the database shown in FIG. 49 can be performed as shown in FIG. This FIG. 51 shows, as an example, the word x on the left and the center.
With line xx? Shows a procedure for searching one word and searching the right word. This line xx? Tables that may contain (x, H), (x, M), (x, H)
h) and (x, m). The target line can be searched by using the table (x, H) in which the central word is x and the left word is the sort key. In FIG. 51, when the arrow (1) starting from the word x is first traced, the arrow (1) points to the table (x, H). The lines of this table are stored in the nodes of the binary search tree in order to maintain the sorting order. Following the arrow (2) from the table (x, H),
The binary search tree: The root of the tree (3) is designated. Indicating the root of a tree is equivalent to instructing the tree itself. On the contrary, the line of the table (x, H) is designated by each node of the tree (3).

Then, when the arrow (4) is traced from the root of the tree (3), the line designated by the root node is wxy. In this case, the line wxy and the searched line xx? Xx? Is larger (that is, below the table,
To the right of the tree node). Therefore, the branch (5) on the right side of the root of the tree (3) is lowered and compared with the line yxx indicated by the arrow (6) at the node ahead of it. In this case, xx?
Is smaller, so go down the left branch (7). The line indicated by the arrow (8) at the node ahead of the branch (7) is xxx, which corresponds to the line being searched. Therefore, line xx? With word x on the left and center? It can be seen that the word to the right of is x.

[0089]

As described above, according to the present invention, by setting a certain registration condition indicating the correspondence between the line and the table in advance, based on this registration condition,
Since multiple tables that register one line can be specified automatically, it is possible to easily understand the relationship between the line to be newly registered or the line whose content is to be changed and the line already registered. It is possible to provide a database device which is simple in line management and can flexibly deal with usage that is not expected at the time of design, and has excellent operability and efficiency.

[Brief description of drawings]

FIG. 1 is a conceptual diagram showing a schematic configuration of a database device according to the present invention.

FIG. 2 is a flowchart showing an example of a general registration processing procedure of the database device according to the present invention.

FIG. 3 is a flowchart showing an example of a general search processing procedure of the database device according to the present invention.

FIG. 4 is a flowchart showing an example of a general deletion processing procedure of the database device according to the present invention.

FIG. 5 is a flowchart showing an example of a general change processing procedure of the database device according to the present invention.

FIG. 6 is an explanatory diagram showing the procedure of a discrete event simulation.

FIG. 7 is a schematic diagram showing one state of a model to be simulated in an example in which a discrete event simulation is performed by the database device of the present invention.

FIG. 8 is a table showing all lines representing the states of the model of FIG.

9 is a diagram showing a plurality of tables forming a part of the database created from the line of FIG. 8. FIG.

10 is a diagram showing a plurality of tables forming a part of the database created from the line in FIG. 8. FIG.

FIG. 11 is a diagram showing a plurality of tables forming a part of the database created from the line of FIG. 8;

12 is a diagram showing a plurality of tables forming a part of the database created from the line of FIG. 8. FIG.

FIG. 13 is a diagram showing one table constituting the databases of FIGS.

FIG. 14 is a diagram showing one line forming the line set shown in FIG. 13;

FIG. 15 is a table showing four types of words that can be used in the present invention.

16 is a table showing an example of words in FIG.

FIG. 17 is a diagram showing one table for registering the lines shown in FIG.

FIG. 18 is a diagram showing one table for registering the lines shown in FIG. 14;

FIG. 19 is a diagram showing one table for registering the lines shown in FIG.

FIG. 20 is a diagram showing one table for registering the lines shown in FIG.

21 is a table showing four types of states of customers in the model of FIG. 7. FIG.

22 is a table showing unarriving customers in the order of ID numbers among the tables constituting the databases of FIGS.

FIG. 23 is a table showing customers in a queue in the order of ID numbers after arrival event execution by discrete event simulation using the databases of FIGS.

24 is a schematic diagram showing the state of the model of FIG. 7 after the arrival event is executed by the discrete event simulation.

FIG. 25 is a table showing all lines representing the states of the model of FIG. 24.

FIG. 26 is a diagram showing a plurality of tables forming a part of the database created from the line of FIG. 25.

27 is a diagram showing a plurality of tables forming a part of the database created from the line of FIG. 25. FIG.

28 is a diagram showing a plurality of tables forming a part of the database created from the line of FIG. 25. FIG.

FIG. 29 is a view showing a plurality of tables forming a part of the database created from the line of FIG. 25.

FIG. 30 is a diagram showing three tables showing the states of customers before the evacuation event is executed by the discrete event simulation using the databases of FIGS.

FIG. 31 is a diagram showing four tables showing the states of customers after the evacuation event is executed by the discrete event simulation using the databases of FIGS. 26 to 29.

FIG. 32 is a diagram showing a time after the evacuation event is executed by the discrete event simulation using the database of FIGS.
The figure which shows the content of the table showing the event which occurs in the future in order of time.

FIG. 33 is a diagram showing a time after the evacuation event is executed by the discrete event simulation using the database of FIGS.
The figure which shows the content of the table which shows tracking information of a customer in order of a customer's ID number.

FIG. 34 is a diagram showing a state of the model of FIG. 24 after execution of a leaving event by discrete event simulation.

FIG. 35 is a table showing all lines representing the states of the model of FIG. 34.

36 is a diagram showing a plurality of tables forming a part of the database created from the line of FIG. 35. FIG.

FIG. 37 is a diagram showing a plurality of tables forming a part of the database created from the line of FIG. 35.

38 is a diagram showing a plurality of tables forming a part of the database created from the line in FIG. 35. FIG.

FIG. 39 is a view showing a plurality of tables forming a part of the database created from the line of FIG. 35.

FIG. 40 is a diagram showing an example of a binary search tree that can be used as sorting means in the database device according to the present invention.

FIG. 41 is a diagram showing an example of a binary search tree in which integers are sorted.

FIG. 42 is a flowchart showing an example of a search method using a binary search tree.

FIG. 43 is a flowchart showing an example of a registration method using a binary search tree.

44 is a diagram showing a procedure for registering a new node in the binary search tree of FIG. 41 and a binary search tree in which the new node is registered.

FIG. 45 is a flowchart showing an example of a deletion method using a binary search tree.

FIG. 46 is an explanatory diagram showing an example of a delete operation by the method of FIG. 45.

FIG. 47 is an explanatory diagram showing an example of a delete operation by the method of FIG. 45.

48 is an explanatory diagram showing an example of a deleting operation by the method of FIG. 45.

FIG. 49 is an explanatory diagram showing an example of a database having a function implemented according to the present invention.

50 is an explanatory diagram showing a database obtained by deleting one line from the database of FIG. 49.

51 is an explanatory diagram showing a search procedure for an arbitrary line in the database of FIG. 49. FIG.

FIG. 52 is a table showing an example of relations registered in the database.

FIG. 53 is a table showing an example of relations registered in the database.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Input means 2 ... Storage means 3 ... Table designation means 4 ... Management means 5 ... Registration means 6 ... Sorting means 7 ... Search means 8 ... Delete means 9 ... Output means 10 ... Table creation means

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Hiroshi Konno 70 Yanagi-cho, Saiwai-ku, Kawasaki-shi, Kanagawa, Toshiba Yanagimachi Co., Ltd. in the factory

Claims (8)

[Claims] 1. A line is a sequence of a plurality of words,
In a database device that registers / deletes multiple lines individually / selectively to / from multiple tables that meet multiple conditions, and searches for lines based on the contents of words, the sequence of multiple words is used as a line. In addition to inputting, specify the already registered line by the content of the word or the position of the line, input the search command, the delete command, or the change command, and the sorting rule set for each table and each table Storage means for storing a plurality of table information including registered line information and sort order information thereof, and a preset constant registration condition indicating a correspondence relationship between the line and the table, and inputting the line by the input means. In response to the contents of the word of the input line and the registration condition, a plurality of tables for registering the line are specified, and Registered search command lines by, in response to an input of the delete command, or change command, based on the contents of the word of the designated line the registration condition,
Table designating means for designating a plurality of tables in which the lines are registered; management means for managing a plurality of table information stored in the storage means and providing a plurality of table information designated by the table designating means; Registration means for receiving a plurality of table information designated by the table designation means from the management means in response to the line input by the input means, and registering the lines input by the input means in each table. The lines registered in the plurality of tables by the registration means,
Sorting means for sorting on the table based on the content of the word at the position specified by the sorting rule set in each table, and a search command, a deletion command, or a modification command for the registered line by the input device. In response to the input, the plurality of table information designated by the table designating means is received from the managing means, and the registration designated by the inputting means is performed based on the order sorted by the sorting means on each table. And a plurality of table information searched by the search means in response to the input of the deletion command or the change command of the registered line by the input means, Deleting the line searched for by the managing means, the managing means stores the registered information by the registering means and the sorting means, and the deleting means. That updates the management information by deleting information, is configured to manage search information by the searching means, further, a database device characterized by comprising an output means for outputting the management information by the management unit. 2. The managing means has a table creating means for creating a new table that matches the specified table when the table specified by the table specifying means does not exist in the storage means. The database device according to claim 1, wherein the database device is a database device. 3. The certain registration condition comprises, for a certain line, a line having the same word at the same position as the word at a certain position forming the line, and
The database device according to claim 1, wherein the database device is set so as to correspond to a table in which words at different positions are used as sort keys. 4. The line is configured as an ordered sequence of three words, and one of the three positions of the three words forming the line is set to a first position regardless of the order. Position, another one position is defined as a second position, and the remaining one position is defined as a third position, the certain registration condition is that, for a certain one line, a line that configures the line. A line set having the same word in the first position as the word in the first position, and
A first table having the word at the position as a sort key, and a line set having the same word as the word at the first position forming the line at the first position, and a third table
A second table in which the word at the position is used as a sort key, and a line set having the same word as the word at the second position in the second position in the second position, and
A third table having the word at the position as the sort key, and a line set having the same word as the word at the third position forming the line at the third position, and
4. The database device according to claim 3, wherein the database device is set so as to correspond to a fourth table in which the word at the position is used as a sort key. 5. The constant registration condition comprises, for a certain line, a set of lines having the same word in the second position as the word in the second position forming the line, and the third registration condition. Table with word at position as sort key,
Alternatively, it is configured to correspond to a table that is composed of a line set having the same word as the word at the third position forming the line at the third position, and that uses the word at the second position as a sort key. The database device according to claim 4, wherein 6. The constant registration condition is set so that a certain line is associated with a table in which words at a plurality of positions forming the line are used as sort keys. The database device according to item 3. 7. The line is configured as an ordered sequence of three words, and one of the three positions of the three words forming the line is set to a first position regardless of the order. Position, another one position is defined as a second position, and the remaining one position is defined as a third position, the certain registration condition is that a certain line constitutes the line A table consisting of a set of lines having the same word in the first position as the word in the first position, and the words in the second and third positions as the first and second sort keys are made to correspond to each other. The database device according to claim 6, wherein the database device is set. 8. The word is an ordered pair of two strings,
An ordered pair of strings and integers, an ordered pair of two integers, or
The database device according to claim 1, wherein the database device is one of four types of one floating-point number.