EP3877866A1

EP3877866A1 - Method and device for storing items of data and the relationships thereof

Info

Publication number: EP3877866A1
Application number: EP19783200.9A
Authority: EP
Inventors: Lea Kelbsch
Original assignee: Individual
Current assignee: Individual
Priority date: 2018-11-09
Filing date: 2019-09-20
Publication date: 2021-09-15
Also published as: DE102018128048A1; WO2020094175A1

Abstract

The aim of the invention is to provide a particularly efficient storage structure. This is achieved in that in order to store at least two interrelated items of data (Data1, Data2...) in a database (1) comprising a plurality of storage spaces, the first item of data (Data1) is stored in a first storage space (5; 7-2), the second item of data (Data2) is stored in a second storage space (5; 7-2), and the relationship of the items of data is stored in two mutually assigned storage spaces (3, 4) in that a reference to the first item of data (Data1) or to a relationship at least indirectly related to the first item of data (Data1) is stored in one storage space (3) of the two mutually assigned storage spaces, and a reference to the second item of data (Data2) or to a relationship at least indirectly related to the second item of data (Data2) is stored in the second storage space (4) of the two mutually assigned storage spaces.

Description

Title: Method and device for storing data and their relationships

description

The invention relates to a method for storing

at least two related data in a plurality of memory locations

Database. The invention further relates to a database for storing at least two data related to one another. The invention also relates to a

Computer program that can be executed on a computer.

There are a variety of different databases and

Database systems known. These consist of an in

certain organized memory and one of these Memory management software. The store is often called

Database and the software referred to as database management system. A database language is generally provided for the input and output as well as the administration of the data stored in the database, which in particular enables clearly defined access to the content of the database.

Database systems are often designed as a relational database. Relational databases are organized as a variety of tables, with each row of a table having one

Record describes. The number of tables and the number of the respective fields in a data record of a table are determined in advance depending on the intended use by specifying a database schema. By suitable

Links can create relationships between individuals

Tables are expressed. The basic structure of a relational database is that the interrelated data is stored within a row of a table.

An extension of the area of application of an existing relational database for the inclusion of different types of data can usually only be realized with relatively great effort, since the originally defined tables or the

individual data records often have to be expanded by adding further attributes. However, this is often only possible by completely rebuilding the database. An increase in the processing speed of a relational database can be achieved if the relational structure, i.e. the individual tables, directly in

Hardware are shown. An extension of once

However, the defined structure is usually no longer possible.

From an abstract point of view, a section of the real world is stored in a database. The so-called Entity Relationship Model (ERM) is often used to formally define the data stored in a database. This model enables, on the one hand, the relationships of in a database

Specify the stored data formally and, on the other hand, to check their correctness when entering data.

Using an ERM, very complex database designs can be created that follow the rules specified by the ERM

correspond. In particular, very complex relationships of data to be entered when using the database can be defined in advance.

The existing database systems have the disadvantage that their flexibility increases with the increasing number of stored data or the possible predetermined ones

Relationships is restricted.

The object of the invention is to provide a way for that

To store data and their relationships, which is independent of the purpose and as large as possible Can store a number of different data within the same structure. Furthermore, it should be achieved that the database structure can be implemented as efficiently as possible in a software-based solution, but in particular also in a hardware-based solution.

The object is achieved by a method having the features of independent claim 1. The object is further achieved by a memory arrangement and by a computer program in accordance with the further independent patent claims.

According to the invention, relationships of two data in each case are stored, on the one hand, and the data itself, on the other hand. This is independent of a specific application and

regardless of the number of data to be saved or the relationships to be saved. Only the size of the

the respective storage location will expediently be specified.

A special feature of the present invention is that the data and the relationships between individual items of this data are stored separately. Through this structural

Feature can achieve a variety of advantages, as will be described below.

Before a date is saved, it is preferably checked whether it has already been saved. It is only saved if this is not the case. This ensures that No data redundancy occurs, which in turn enables particularly efficient use of the memory.

It is particularly advantageous if relationships of the data are stored in that the respective

A reference is saved.

The relationships are formed by the references in such a way that the relationships of a plurality of data which are related to one another can be stored, although these relationships are only stored in pairs. This can be implemented in different embodiments, as is also described below.

A reference can in particular be a memory address which designates a memory location at which a date is located or at which another relationship is stored which relates at least indirectly to the respective date. Indirectly means here that a chain of possibly hierarchically ordered relationships ultimately to the

the relevant date. By the use of

Memory addresses as references will be a particularly quick one

Search within the database and therefore an efficient one

Allows dissolution of relationships.

The terms "date" and "relationship" can be understood very broadly in relation to the present invention. Neither the nature of a relationship nor the type of a date is specified. The invention even allows a date like that treat a relationship or treat a relationship like a date. This ensures that the

Storage methods according to the invention and the database according to the invention can be used for a large number of very different applications. For example, depending on the particular context or application in which the invention is used, a definition area of the term “date” and the term “relationship” can be determined.

Thus, the invention can be divided into several, for example

Realize levels comprising hierarchical structure, whereby

Relationships of one level are treated as data in another level and vice versa.

Advantageously, the storage space at which one

Relationship is stored, also addressable, a first reference being stored in a first partial area and a second reference being stored in a second partial area. These references are preferably implemented by memory addresses, which in turn refer to memory locations at which either a date or another relationship is stored in each case.

Due to the way in which the data and relationships according to the invention, which in turn represent abstractions of the real world, are stored in a given address system

Relationship and any date associated with a unique address. This means that with a single address each

Relationship and each date clearly localized and therefore also can be called up again. In particular, it can also be specified for a specific address which relationships reference this address and / or the meaning or the date assigned to this address can be reproduced.

A particular advantage of the invention is that for a

Modification of a database once created, no change of the actually stored data is necessary. A unique address is assigned to each saved date, which can remain. If the content is to be changed, it is sufficient to add or remove one or more relationships.

In contrast to relational databases, it is no longer possible to speak of “data records” in the present invention, since data elements that belong together according to the invention are determined by their relationships with one another

"Record" take place. This also prevents the occurrence of empty records.

Due to the way in which relationships and data are stored according to the invention, no rigid requirements as with conventional ones are met

Definitions of tables or indexes and keys

needed. Rather, complete flexibility is achieved in that the amount of data on which a particular

Memory address points, is not fixed. So can a specific address can be assigned to a single date, whereas another address - by resolving

Relationships - at least indirectly on a variety of

individual, related data.

According to a first possible embodiment, the

Relationships and the data in the same storage area

stored, at least two subareas being provided for storing the relationships for each address and a third subarea for storing the data

is provided. Advantageously, a particular

Address either a relationship using the first two panes or a date using the third

Partial area saved. So the whole

Realize the memory structure in that the memory contains addressable lines, each line having three

Includes subareas. Two fixed ones each

Partial areas, for example always the first two, are provided for storing a relationship and a third partial area is reserved for storing a date.

Whether the address of a line represents a relationship or a date depends on what is stored in it. Are the two memory locations occupied for the

Saving of a relationship are provided, according to one possible embodiment, in the third

Partial area no date can be saved. It is then clearly a storage space or one

Memory cell whose address represents a relationship. According to this embodiment, a date becomes in the third

Subarea of a line is stored, so in the other two subareas none realizing a relationship

References saved. In this case, the address of the memory location or the memory cell would therefore be a date

describe. In a hierarchically structured implementation of the invention, however, this address could also be a reference to a relationship to a date or to another

Designate relationship.

This embodiment has the advantage that for the

Saving all possible data and relationships using only a single storage structure. The data can then be written into the database in such a way that the next free line in the memory is always used. If a relationship is to be saved, the two sub-areas provided for this purpose are used. If, on the other hand, a date is to be saved, the third section is used for this.

According to an improved embodiment, for the

Storage of the data and the storage of relationships each use rows that only comprise two sub-areas. If a relationship is saved, both are

Partial area used for storing the respective reference. However, if a date is saved, only one of the two sections is described. At this

Embodiment becomes the same functionality as the previous one described embodiment achieved, but less

Space needed.

According to a particularly advantageous embodiment, is used for storing relationships in the database

separate memory area, in the following as

Relationship storage area denotes, used and at least one other separate for storing the data

Memory area, hereinafter referred to as data storage area, is used. The relationship memory area comprises a plurality of addressable memory locations and each of these addressable memory locations comprises at least two partial areas for storing the two ones

Relationship-realizing references or memory addresses.

The data storage area preferably comprises a plurality of addressable storage locations and each of these

addressable memory locations preferably also comprise at least two partial areas. If a date is now to be stored, it is stored in the subarea provided for storing data within the relevant line of the data storage area. The breadth of this

Partial area is adapted to the width of the data to be saved. At least when a date is first stored in the data storage area, an address is now additionally reserved in the relationship storage area. This address is stored in the second subarea of that line of the data storage area in which the date itself is also stored. These Embodiment and the resulting advantages are explained in more detail below.

A plurality of data storage areas can be present in a database according to the invention. Different

Data storage areas can then be adapted to the data to be expected.

Input modules are preferably used for data input

provided, the input modules taking into account the respective application and / or the type of data to be expected. Only in the input modules is there a specialization of the

Database system intended for a specific application. However, the database behind it is largely independent of the specific application and the type of data. In one embodiment, the

Select data storage areas depending on the input modules. For example, each input module can

separate data storage area can be assigned or it can be provided that the input modules belonging to a specific application each have a data storage area

assign. It is also conceivable to implement the data storage areas depending on the type of data.

For example, it can be provided that multimedia content is stored in a different data area than measured values that belong to a specific application. An address of the relationship memory area is preferably reserved for each available input module. This ensures that relationships between data and an input module can also be stored within the database.

Of course there are also relationships between

Input modules can be saved within the database. As a result of the fact that a date can be assigned to an input module, further knowledge about the data is implicitly available.

This knowledge enables, for example, an interpretation of individual data and can possibly be for one

Further processing of read data can be used.

Output modules are preferably used for data output

intended . These can be analogous to the input modules

are handled within the database. The output modules can thus also be assigned data in that they can have a relationship both to data and to relationships and of course also to input modules. Such relationships are the same within the database

stored like relationships between dates or between a date and a relationship.

The task mentioned at the outset is also achieved in particular by a computer program which is based on a computer system

is executable and is programmed to carry out the method according to the invention. The computer program represents the invention through the individual method steps as well as the method itself. Further advantageous features of the invention are those shown in the figures and described below

Embodiments can be removed.

FIG. 1 shows a block diagram which shows a memory structure according to a first one which is suitable for carrying out the method according to the invention

Embodiment;

FIG. 2 shows a block diagram which shows a memory structure according to a second one which is suitable for carrying out the method according to the invention

Embodiment;

FIG. 3 shows a block diagram which is suitable for carrying out the method according to the invention

Memory structure according to a third

Embodiment;

FIG. 4 shows the registration of

Data in the database according to the invention according to an embodiment;

Figure 5 shows by means of a diagram

underlying database according to the invention

Structure for storing relationships; Figure 6 shows a preferred embodiment of a in

Hardware implemented database according to the invention using a block diagram.

Figure 1 shows a memory structure according to a possible embodiment. A memory 1 comprises a plurality of lines 2, each line 2 comprising a first partial area 3, a second partial area 4 and a third partial area 5.

The first partial area 3 and the second partial area 4 are provided for storing references. These references

realize relationships. The entirety of the first

Subareas 3 and the second subarea 4 thus form a relationship memory 6. The third subarea 5 is provided for the storage of data. The entirety of the third partial areas forms a data memory 7.

Each line 2 is a unique address Al, A2, ...

assigned. Each of the lines 2 can be addressed in a known manner by means of these addresses and thus in particular enables read and write access to the lines 2 and the subareas contained therein.

The first section 3 and the second section 4 of the

Relationship memories 6 are dimensioned such that a memory address A1, A2,... Can be stored therein. The third section 5 of the data memory 7 is dimensioned in dependence on the data to be stored therein such that the storage of an individual Date is ensured. If a date that is larger than the width of the third partial area should nevertheless be stored within the memory 1 during operation of a database that has been configured, then this can be done

Mechanisms can be provided, for example a division

(Split) of the respective date, so that the date is divided into a plurality of third partial areas, the first partial area and the second partial area of the respective lines then being able to be used as references which implement an at least simply linked list and save and Reassemble the date

enable. In particular, however, it can also be provided that the individual parts of a date are like individual data

store that have a relationship to each other. The original date can then be restored by breaking the relationships.

According to the exemplary embodiment shown in FIG. 1, the relationship memory 6 and the data memory 7 are thus implemented within a memory element, with one to be stored

Date always in the third section of a line

is saved. In such a row there are then no one in the first partial area and the second partial area

Relationship-realizing references saved.

A relationship is stored in that a reference to another line is stored in one line in the first partial area 3 and the second partial area 4. Such a reference can in particular be the address of the relevant line. If, for example, a

Relationship between two data is stored, the address of the first date is stored in the first section of this line and the address of the second date in the second section.

It is of particular importance for the overall invention that not only relationships between data, but in particular also by means of the memory structure according to the invention

Relationships between individual relationships and relationships between a relationship and a date can be saved in the same structure. If, for example, relationships are stored under the addresses A1 and A2, a relationship can be established again between these two relationships by the value A1 in the first partial area in another memory line and the value Al in the second

Part of the value A2 is saved. As will be described in detail later, it is with this

generic storage principle possible, a variety

save different relationships. According to the invention, the meaning that a relationship has in reality is not explicitly taken into account when storing in the relationship memory. The storage principle according to the invention or the database according to the invention is therefore fundamentally suitable for storing any relationships and data. Another possible embodiment is shown in FIG.

There are only two sub-areas in this line 2

educated. If a relationship is to be saved in one line, the corresponding references or addresses are saved in the two subareas. If, on the other hand, a date is to be saved in one line, the second section is used for this. As a rule, this is wider than the first partial area, since it is assumed that the data to be stored are larger than the addresses to be stored. Whether a line is to be assigned to the relationship memory 6 or the data memory 7 depends on whether a value is stored in the first partial area. If this is the case, the line in question is part of the relationship memory 6.

If not, and is only in the second

If a value is stored in the partial area, the row in question is to be assigned to the data memory 7. At this

In one embodiment, the data memory 6 and the data memory 7 are formed within a common memory 6-7. Compared to that shown in Figure 1

Embodiment has that shown in Figure 2

Embodiment the advantage that only two sub-areas are required instead of three sub-areas per line 2, which reduces the total memory requirement.

FIG. 3 shows a particularly preferred exemplary embodiment of a possible memory structure for realizing the

Invention shown. Figure 3 shows a relationship memory

6, which comprises a plurality of rows 2. Each of the rows 2 comprises a first section 3 and a second

Subarea 4. In this relationship memory 6, as already described, relationships are created by two

stored references realized, these references preferably represent memory addresses.

A plurality of data memories 7 is also shown in FIG. Each of the data memories 7 comprises a plurality of lines in a known manner. Each of the lines has a first

Subarea 7-1 and a second subarea 7-2. In the first partial area 7-1 there are addresses of the relationship memory

6 filed. In the second section 7-2

actual data stored. According to one possible

Embodiment are in the first portion 7-1 of the

Addresses stored on the

Get relationship memory 6. If a date is stored, it is stored in the second partial area of a free line in a data memory 7. In addition, in the

Relationship memory 6 assigned a line 2 to this date. No value is stored in the relationship memory 6 itself in this line 2 assigned to the date. However, the address of the assigned line 2 of the relationship memory 6 is stored in a partial area 7-1 of the data memory 7. This takes place in the partial area 7-1 of that line of the

Data memory 7, in which the data word has been stored in the second partial area 7-2. If a date is to be stored in the data memory, then for example the row of the relationship memory which can be addressed with the address A1 becomes assigned to this date. The address A1 is then written into an empty line of the data memory 7 in the first partial area 7-1 and the actual date is stored in the second partial area 7-2 of this line.

The memory structure shown in FIG. 3 has the advantage that the relationship memory 6 is completely separated from the data memory 7. Consequently, only relationships are stored in the relationship memory 6. These relationships are completely independent of the type and content of a single date. The date itself is stored in a data store

7 or stored in several data memories 7. The fact that in the relationship memory 6 only relationships

are stored, the relationship memory 6 can very much

be used efficiently. In particular, this strict separation between relationship storage and data storage leads to a significant reduction in storage space because each relationship only has to be saved once. E.g. there is data that has the same content or meaning, but in different words and / or languages and / or with different means of expression such as text,

Language, graphics, etc. can be expressed and should be saved in several representations. In this case, the concrete realization of the date is independent of its relationships within the database. By means of the database according to the invention, it is possible to implement the specific realizations of a date that is to be expressed and / or stored differently in each case in to store another data memory 7. E.g. a date can be stored in text form in a first data memory and a reproduction of this date as a spoken file in a second data memory. Each of the data memories 7 can be optimized for the particular type of playback or the storage of data with respect to the length of the second

Section 7-2 and also with regard to possible procedures for

Writing in and reading out the data, including possibly advantageous pre- and post-processing, such as a

Compress individual data. The corresponding data memory 7 can then be selected for the specific output.

According to a possible embodiment, each date is

Regardless of its representation, the same address of the relationship memory 6 is assigned in each data memory 7 and stored in the first partial area 7-1 of the respective line.

The relationship memory 6 thus abstracts completely from an actual representation of the data to be stored. Depending on the desired input and / or output, the

appropriate data storage 7 can be selected.

In the embodiment shown in FIG. 3, input modules and output modules can be implemented particularly advantageously. E.g. it can be provided that a special data memory is assigned to each input module. Likewise can

it should be provided that each output module has at least one

Data storage is assigned. Is now in with the database

Communicated in the form of text-based input and output, so this is done using a dedicated input and

Output module. With the database, however, by means of

Acoustic information communicates, specialized input and output modules can be provided for this. The input and output of data then takes place in a data memory 7 optimized for the respective input and output form.

The actual relationship memory 6 remains unaffected by this. This ensures that the relationships are consistent.

If, for example, a textual representation of a specific date is entered and relationships are changed within the relationship memory 6, the relationship memory 6 for all other possible locations is automatically made

Representations of that date updated.

In particular, the embodiment shown in FIG. 3 clearly shows that a separation of the

Relationship memory 6 is realized by one or more data memories 7, which on the one hand guarantees high data consistency and on the other hand enables the database to be optimized with regard to input and output speed and the required storage space.

FIG. 4 shows a simplified flow diagram for a possible embodiment of the method according to the invention

With regard to the storage of data and relationships

outlined. In step 100, an input module is first selected. The module can be implemented in the form of an input mask, for example, by means of the customer data in one Companies should be saved. In this form, individual text fields are provided for input. These text fields are related. E.g. there is one

Relationship between the generic term "address" and one

"Names", a "street", a "place of residence", one

"Telephone number", a person. These relationships are stored in the database according to the invention and are read out of the database when the input module is called up, provided that the input module is already present. This presence is first checked in a method step 101 not yet at the

Database registered, this is carried out in a step 102. E.g. is used for the input module

Initialized data storage and it is in the

Relationship memory one line assigned to this input module. The input module is therefore from the point of view of the

Relationship memory treated the same way as a date.

In a step 103, the address of the relationship memory 6 assigned to the input module is now selected as the starting point within the relationship memory 6. For this purpose, a register can be provided in which an address that is current for processing is always stored within the data memory 6.

In step 104, the data is input, for example.

using the input form. When the entry is finished, the data is sent to the database. In the database the data and the relationships on their

Checked for existence. If a relationship or a date does not exist, a corresponding entry is made. It is conceivable that the database checks to determine whether a date or a relationship is possibly incorrect. E.g. checks the

Database first in a step 105, whether a particular

Date already exists. If this is the case, possible relationships of this date are shown in a step 106

selected to enable a check for correctness. It can also be provided that only a current pointer to the line assigned to this date in the

Relationship memory 6 to set.

If the date does not yet exist, then in one step

107, a free memory cell of the relationship memory 6 is initially assigned to this datum and the address of this memory cell is in the data memory 7 assigned to this input module at a free location in the first partial area 7-1

saved. The actual representation of the date is then in the second section 7-2 of this line

saved.

In a step 108 it is now checked whether further data have been transferred using the input module and must therefore be processed. If this is the case, it becomes

Step 105 branches back and further data are processed. If this is not the case, then in one step 109 checked whether a new relationship was specified by means of the input module, for example by the fact that the input module

has been changed. In step 109 further can

be provided to check within the data memory 6 whether the internal relationships should be updated on the basis of the data input so far. In the above example, this can occur if a name or address has already been entered that already has another name

Input module has also been saved. This

means that the individual address data have a relationship to another input module in the relationship memory 6.

Are additional data or slightly different data for the address already entered by means of another input module

Data available, this can be displayed to a user of the database and he can be asked to correct or accept the data.

Was not a new relationship using the input module

entered and there is no relationship to be updated, the method can be ended in a step 114. However, if a relationship is to be updated, it is first checked in a step 110 whether this relationship already exists. As previously described, this check can also be carried out in step 109. If a relationship does not yet exist, it is now created. A relationship is realized in that two memory addresses are stored together in one line of the relationship memory 6.

This enables a completely abstract representation of Relationships within the database. It has been shown that a majority of the relationships in the relationship memory

Relationships between other relationships are. As a result, a multidimensional relationship tree is created, which is mapped in the relationship memory 6. To determine whether one date is related to another, the latter can

Tree to be examined. A particularly fast search can be implemented by the device according to the invention, in which the memory structure according to the invention is simulated in hardware.

FIG. 5 shows a simplified diagram which outlines the structure on which the database is based. The

fundamental idea of the database according to the invention is

Represent relationships as a tuple of three elements, one element identifying the relationship. This ID of the relationship preferably corresponds to the address of the line within the relationship memory which is assigned to this relationship. The other two entries of the three element tuple are the addresses of those

Memory cells between which there is a connection. E.g. the address A1 is assigned to a specific input module. The line with the address A1 in the relationship memory 6 establishes a relationship between the memory cells with the

Addresses A2 and A3 made. This is shown in FIG. 5 by the corresponding entries in the relationship memory 6

shown as an example. The memory cell with the address A2 contains a relationship between the memory cells A4 and A5. There are no entries in the memory cells with the addresses A4 and A5. This means that no further relationships are established here, but that these now directly represent data. This data is stored in the data memory 7 assigned to the relevant input module. This is also graphical in FIG. 5

illustrated. There is a first date Datal

saved. The memory cell with the address A4 is assigned to this date Datal. Another data item Data2 is also stored in the data memory 7. This date is the memory cell with the address A5 in the

Relationship memory allocated. These two data are therefore related to each other and also related to the relationships at addresses A1 and A2.

The memory address A3 related to the address A2 in turn has a relationship to the content. A relationship between the memory cells with the addresses A6 and A7 is stored there.

The ID with the address A6 represents another datum Data3, which is also stored in the data memory 7.

A relationship between the memory cells with the addresses A8 and A9 is stored in the memory cell with the address A7. The memory cells with the addresses A8 and A9 represent further data Data4 and DataS, which are related to each other.

The structure shown in FIG. 5 can be used to describe a large number of relationships from the real world, which are stored in a database and are thus to be mapped in the database. This very abstract kind of

Description of relationships is according to the invention by the

The method and the device, which can be implemented both in software and in hardware, are made possible. This ensures a complete separation of the relationships on the one hand and the actual data on the other.

Another advantage of separating the actual ones

Data and the relationships as well as the realization of the

Relationships through addresses is that any point within the ones realized through the relationships

multi-dimensional structure can be used as a starting point to search for an entry within the database, to establish a connection or to save a new relationship or a new date. The uniform structure also makes it possible to form sets over the relationships in a particularly efficient manner, for example

Union sets, intersections, similarity sets and / or

Probability sets.

Both in the data store and in the relationship store can be due to the abstract structure general calculations generate statements, generate new knowledge and / or verify stored data and relationships. Based on the example shown in FIG. 4, it could be provided, for example, after or during the input of data relating to another person,

Determine similarities and / or differences between the respective data and / or the respective relationships and thus offer correction suggestions or even propose values and / or relationships already when entering the data, which correspond with high probability to the data and / or relationships to be entered.

A particularly advantageous embodiment of the

The device according to the invention enables imaging of

Storage structure directly in hardware. This significantly increases the processing speed within the database, since the individual memory modules required for this can communicate with one another very directly. The necessary logic can be kept very small, which means a very fast search within the database as well as a very fast saving and

Data query enabled.

FIG. 6 shows an exemplary embodiment of a database according to the invention implemented in hardware according to a

Embodiment shown, which corresponds to the basic functionality of the embodiment shown in Figure 3. The relationship memory 6 shown in FIG. 6 includes two addressable memory elements M3, M4, with which the functionality of the two sub-areas 3 and 4 of

Relationship memory 6 can be realized. For example, hard disks or so-called RAM drives can be used as addressable storage elements M3, M4. According to the exemplary embodiments shown in FIGS. 6 and 7, RAM modules are used as memory elements, which one

have a particularly short access time. The maximum number of addressable data is determined by the address width, i.e. the maximum number of possible addresses. Above one

Address line 8 and a data line 9 are the

Memory elements or the RAM modules M3, M4 of the

Relationship memory 6 connected to a logic element 10. Since in the relationship memory 6 the relationships in the form of

Memory addresses Al, A2, ... are saved, the RAM

Blocks M3, M4 have a square structure so that the data width corresponds to the address width.

FIG. 6 also shows a plurality of data memories 7, which also each have two RAM modules M7-1 and M7-2. The functionality of the data memory 7 corresponds to the embodiment shown in FIG. 3. Each data memory 7 or each subarea 7-1 and 7-2, which is implemented by the respective RAM modules M7-1 and M7-2, is also connected to the logic element 10 via an address line 8 and a data line 9. According to one possible

All RAM modules (M3, M4, M7-1, M7-2 have the same address or data width. This enables a particularly efficient implementation of the logic element 10. The logic element 10 also has a data line for one

Data input 11 and a data line for a data output 12. If a date or a relationship is to be stored, the logic element 10 first checks whether this date or this relationship received via the data input 11 already exists. If it’s not,

The logic element 10 determines the address A1, A2, ... of a free memory line 2 in the relationship memory 6. If a relationship is to be stored, this is done in the two RAM modules M3 and M4 by the logic element

10 determined address. If a date is to be stored, the logic element 10 causes the determined one

Address in the RAM realizing the first section 7-1

M7-1 of the data memory 7 and the actual date is stored in the RAM M7-2 which realizes the second partial area 7-2. As already described above, the data memory 7 is selected, for example, as a function of the respective input module. In addition, depending on the current input module, a start address in the

Relationship memory 6 can be selected to check whether a relationship to be saved already exists.

FIG. 7 shows a further possible embodiment of a

Realization of the invention shown in hardware. These

The embodiment initially corresponds conceptually to that in FIG

2 shown embodiment, in which both the data and the relationships in the same memory area stored, but also has an index

Register 14. The index register 14 has three RAM

Building blocks 15, 16, 17, which via an address line 8 and a data line 9 with a portion 10-2 of the

Logic elements 10 are connected. The index register 14 can be used for each memory location of the relationship memory 6 (which is determined by an address), the respective preceding address (ie a memory location which refers to this specific memory location) and one

subsequent address can be saved, whereby a linked list is realized. Based on the example shown in FIG. 5, the address which is in the memory location A3 could be stored in the RAM module 16, for example for the address A3 of the relationship memory

refers. This is the address Al. The address of the relationship memory 6 to which reference is made from the address A3 could be stored in the RAM module 17. So this would be the address A7. The two RAM

Blocks 16 and 17 are the double-linked list.

In the embodiment shown in FIG. 7, the index register 14 additionally has RAM modules 15 and 18. Furthermore, a RAM module 13 is additionally assigned to the relationship memory 6. This achieves the following: Using the RAM

Module 13 is assigned a free memory location in the index register 14 to each address of the relationship memory 6.

Are data or relationships in the relationship memory 6 and possibly. stored in the data memory 7, which on a refer to a specific address within the relationship memory 6, a new one is made in the index register 14

Reserved memory location within the RAM module 15 and the original address is stored therein. The RAM module 18, which is connected to the subarea 10-2 of the logic element 10 via a data line 19, realizes a so-called flag array and specifies for each address stored in the RAM module 15 whether this is a reference with respect to the first subarea 3 or the second subarea 4 of the

Relationship memory 6 acts.

Both the embodiments shown in FIGS. 1 to 5 and the embodiments shown in FIGS. 6 and 7 are merely

To understand embodiments. For example, the one explicitly shown in FIGS. 6 and 7 could be provided

The data memory 7 shown is analogous to that in FIG. 2

to be implicitly integrated into the relationship memory 6, this date then always being stored in the RAM module 4, for example, in the case of a date to be stored.

Furthermore, an allocation memory can be provided in the memory structure for the data memory 6 and, if necessary, for the index register 14, which returns the next free address in the data memory 6 or the index register 14 on request.

As a result, the logic element 10 is relieved and data and relationships can be written and erased even faster.

Claims

1. A method for storing at least two interrelated data (Datal, Data2, ...) in a database (1) comprising a plurality of storage locations, characterized in that the first date (Datal) in a first storage location (5 ; 7-2) is saved, the second date (Data2) in a second

Storage space (5; 7-2) is saved and the

Relationship of data in two associated with each other

Storage locations (3, 4) are stored in that one storage location (3) of the two

assigned memory locations a reference to the first

Date (Datal) or on a relationship that is at least indirectly related to the first date (Datal), and stored in the second

Storage space (4) of the two assigned to each other

Storage locations a reference to the second date (Data2) or to a relationship that is at least indirectly related to the second date (Data2) is saved.

2. The method according to claim 1, characterized in that before storing a date (Datal, Data2, ...) it is checked whether this date (Datal, Data2, ...) has already been saved and if this is the case and if a relationship to another date is saved should be checked whether this relationship already

is saved.

3. The method according to claim 1 or 2, characterized in that the two mutually assigned memory locations (3, 4), at each of which a relationship is stored, a first partial area (3) and a second

Subarea (4) of the same addressable memory cell

(2) and the address (A1, A2, ...) of the memory cell (2) in which this relationship is stored is used as a reference to another relationship.

4. The method according to any one of the preceding claims, characterized in that the memory location (5; 7-2), on which a date is stored, is a partial area of an addressable memory line (2) and

as a reference to a date (Datal, Data2, ...) the

Address (A1, A2, ...) of the memory cell (2) in which the date is stored is used or

when saving a date in a memory cell (2) another date (Datal, Data2, ...)

Memory cell (2) is assigned and the address (A1, A2, ...) of the assigned memory cell (2) is used as a reference to this date {Datal, Data2, ...).

5. The method according to any one of claims 1 to 4, characterized

characterized that the relationships and the data

(Datal, Data2, ...) in addressable memory cells (2) of the same memory area, each of these being used for storing the relationships

Memory cells (2) has at least two partial areas (3, 4) and a third partial area (5) is formed in these memory cells (2) for storing the data, each of these memory cells (2) either having a relationship using the first two

Sub-areas (3, 4) or a date (Datal, Data2, ...) using the third sub-area (5)

is saved.

6. The method according to any one of claims 1 to 4, characterized

characterized that the relationships and the data

(Datal, Data2, ...) are stored in two partial areas (3, 4) comprising memory cells (2) of the same memory area, in each memory row (2) either a relationship using the two

Sub-areas (3, 4) or a date (Datal, Data2, ...) using the second sub-area (4)

is saved.

7. The method according to any one of claims 1 to 4, characterized

characterized in that a relationship memory (6) for storing the relationships in the database and for the

Saving the data (Datal, Data2, ...) at least one data memory (7) is formed, the

Relationship memory (6) a plurality of addressable

Includes memory cells (2) and each of these addressable Memory cells (2) comprises at least two partial areas (3, 4) and at least one of them is the first

Storage of a date (Datal, Data2, ...) in the

Data memory (7) a memory cell (2) of the

Relationship memory (6) is assigned to this date (Datal, Data2, ...) and the date (Datal, Data2, ...) is stored in the data memory (7) in that in a first partial area (7-1) Memory cell (2) of the data memory (7) stores the address (A1, A2, ...) of the memory cell (2) of the relationship memory (6) assigned to this data (Datal, Data2, ...) and in a second sub-area ( 7-2) of this memory cell (2) of the data memory (7) a representation of the date (Datal, Data2) is stored.

8. The method according to any one of the preceding claims, characterized in that an input module is provided for the input of at least one date (Datal, Data2) and an address (A1, A2, ...) of the at least for a first data input by means of the input module

Relationship memory (6) is assigned to this input module.

9. The method according to claim 8, characterized in that at least one relationship between data that can be input (Datal, Data2, ...) can be input by means of the input module.

10. The method according to any one of claims 8 or 9, characterized in that for each stored date

(Datal, Data2, ...) the relationship to at least one

Input module, by means of which the entry was made, is stored at least indirectly.

11. Method according to one of the preceding claims, characterized in that the relationship is only stored in the database if it is not already

is saved.

12. The method according to any one of the preceding claims, characterized in that an output module is provided for the output of at least one date (Datal, Data2, ...) and at least in the case of a first data output by means of the output module the address of a memory cell (2) of the

Relationship memory 6) is assigned to the output module.

13. Storage structure for storing at least two

data related to one another (hierarchical), characterized in that the storage structure comprises:

- At least partial areas (5; 7-2) within

Memory cells (2) for storing the respective date

(Datal, Data2, ...);

- at least two sub-areas (3,

4) within a memory cell (2) which stores the relationship of the data (Datal, Data2, ...) in such a way enables that in the first section (3)

Reference to the first date (Datal) and in the second

Subarea (4) a reference to the second date (Data2) is saved.

14. Memory structure according to claim 13, characterized

characterized in that the memory structure has means for implementing a method according to one of claims 1 to 12.

15. Memory structure according to one of claims 13 or 14, characterized in that the memory structure for the respective corresponding partial areas (3, 4, 7-1, 7-2) of the memory cells (2) each has a memory module

(M3, M4, M7-1, M7-2), being for a particular

Memory line (2) the respective sub-areas (3, 4, 7-1,

7-2) can be addressed with the same address (Al, A2, within the respective memory module (M3, M4, M7-1, M7-2).

16. Memory structure according to one of claims 13 to 15,

characterized in that the memory structure has an index register (14), the index register (14) having at least two memory modules (Ml6, M17)

Realization of a double-linked list includes and the address for a specific memory cell of the relationship memory (6) in one of the memory modules (Ml6, M17)

(Al, A2, ...) a memory cell (2) determined for this referencing other memory cell (2) and in the other

Memory module (Ml6, Ml7) the address of another memory cell (2) to which the specific memory cell

(2) refers, can be saved.

17. Memory structure according to one of claims 13 to 16, characterized in that the memory modules by

RAM modules are implemented.

18. Computer program that can be executed on a computer, characterized in that the computer program is programmed in such a way that a method according to one of claims 1 to 12 is carried out when the

Computer program running on the computer.

19. Storage medium for storing a computer program, characterized in that on the storage medium

Computer program according to claim 18 is stored.