JPH0830501A - Data base management system - Google Patents

Abstract

PURPOSE:To make it possible to utilize a data base capable of describing its data definition/operation as the extension of a data definition/operation function of existing programming language. CONSTITUTION:This data base management system is provided with a buffer area 25 having a page area 251 for storing the page information of a data base and an object area 252 for storing an object and capable of storing data base information, an access managing part 21 capable of referring/updating the object stored in the object area 252, a buffer area managing part 24 for allocating the object in the data base to the object area 252, and a page information managing part 26 for managing relation between the page information in the data base and that in the object area and the access managing part 21 has an object access management routine part 55 for judging whether an object to be found out is in the object area 252 or not and instructs the managing part 24 to refer the object or allocate the object to the page area 251.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a database management system for a computer, and more particularly to a database management system for managing a database containing structured data and information on mutual relationships between these data.

[0002]

2. Description of the Related Art Conventionally, databases are managed using a database management system called a hierarchical type or a relational type. The conventional database management method is mainly used in the office processing field that handles a large amount of data with a relatively simple structure. Among these, in particular, a relational database management system that widely handles a simple data structure that can be represented in a table format and performs a data retrieval / operation by a relation operation between tables is widely used.

[0003]

When the conventional database management method is applied to the fields of CAD, CASE, image processing, multimedia processing, etc., there are the following problems. Therefore, it is necessary to construct a database. In many cases, it is not always a good idea to create a database in terms of processing efficiency and processing efficiency. 1. The data structure of each data element is complicated, and the data definition format provided by the database management system cannot directly represent these data elements. 2. Data is complicatedly related to each other and data definition and operation specification are complicated.
Processing efficiency is poor because it is necessary to perform comparison.

3. The data base language that provides data definition / manipulation functions is an independent system itself, and its main usage is interactive execution with humans. Therefore, it is difficult to construct a processing system that combines programs such as database processing and calculation processing.

Object-oriented databases have been proposed, developed and utilized as means for solving these problems. Object-oriented database is C +
In many cases, a database management function is added based on an existing object-oriented programming language such as + or Smalltalk. However, the initial object-oriented database had the following problems regarding the language extension method and the method for realizing it.

1. The definition and operation of data must be done through the corresponding processing routine of the database management system or the dedicated data operation language. Therefore, much effort is required to change an existing program into a program that uses a database. 2. Data referencing / updating must be performed through the corresponding processing routine of the database management system or the dedicated data operation processing system, resulting in poor processing efficiency. As a database management system for solving these problems, Japanese Patent Application No. 6-38709 has been proposed by the present applicant. This proposal utilizes a special memory management protection function such as virtual memory management provided by the connected computer system.

As described above, the conventional object-oriented database management system is premised on the use of a computer system having all the following memory management protection functions 1 to 3.

1. Memory area allocation and release functions 2. "Unavailable" for the allocated memory area,
Function that can set memory protection of "reference only" and "writable" When an object that violates the specified memory protection is accessed, it is detected and notified to the database management system, and execution control is transferred to the database management system. Among these, 1 is provided in most computer systems. Function. On the other hand, regarding items 2 and 3, personal computers, general-purpose large-scale computers, and the like do not have these functions, or even if they have these functions, they cannot be used from the application program.

Therefore, there is a demand for a database management system that provides the same functions as the conventional object-oriented database management system even when a computer system that does not provide the above-mentioned memory management mechanism is used.

The object of the present invention is to provide the following functions even when using a computer system having no special memory management mechanism: Complex object definition function equivalent to programming language 1. Definition function of interrelationship between complex objects equivalent to programming languages. 3. Make object reference / update in the database the same description format as object reference / update in programming language. It is to provide a database management system having a fusion of database operation and program processing so that a user can use a database that can describe the data definition / operation of the database as an extension of the data definition / operation function of an existing programming language. .

[0011]

In order to achieve the above object, a database management system according to the present invention is a database management system for managing a database containing structured objects and information on mutual relationships between these objects. A page area for storing page information of the database and an object area for storing objects of the database, a buffer area for storing information of the database, and an object area for receiving a user program instruction. An access management unit that enables reference / update of an object, a buffer area management unit that allocates an object of the database to the object area via the page area in response to an instruction from the access management unit, and an object of interest. A page in which a page is stored, the page information management unit managing the relationship between the page information in the database and the page information in the object area, and the access management unit is managed by the page information management unit. By referring to the page information, it is determined whether or not the object requested by the user program is in the object area, and if it is in the object area, the buffer area management unit is instructed to refer to the object in the object area. And an object access management routine section for instructing the buffer area management section to allocate the object to the page area from the database when the object area is not in the object area.

Further, the buffer area management unit manages the buffer area via the buffer area management unit using the memory management function of the connected computer system, and the objects of the database are managed via the page area. It is characterized by allocating to the object area.

The buffer area management section may further include an object rearrangement section that allows the same object in the database to be allocated to different page areas in the buffer area.

Further, the present invention is characterized by comprising a relationship management unit for managing a relationship between objects allocated to different page areas.

Further, the user program is written in a programming language which can be written in the same description format as the description format of the definition and operation of the object regarding the object in the database.

Further, the programming language is a C ++ language.

[0017]

Since the buffer area for storing the object of the database has the page area for storing the page information and the data area for storing the object of the database, the object reference / update is explicitly shown to the database management system by the user program. It is possible to perform object reference / operation without notifying the user, and the processing related to database management is performed only at the first interruption and does not occur at the time of individual object reference / operation. Conventionally, processing such as moving data between the user program and the buffer area, checking and converting the data format, and referencing and updating various management tables is not performed for each individual object reference / operation. Database management system It can be improved significantly treatment efficiency than Temu.

The access management section has an object access management routine section. The object access management routine section refers to page information managed by the page information management section to determine whether the object requested by the user program is in the object area. If it is, the buffer area management unit is instructed to refer to the object in the object area if it is in the object area. If it is not in the object area, the buffer area management is to allocate the object to the page area from the database. Instruct the department.

Therefore, it is not necessary to refer to the page information managed by the page information management unit to determine whether or not the object requested by the user program is in the object area by using the program of the connected computer system.

As a result, 1. Memory area allocation and release functions 2. "Unavailable" for the allocated memory area,
Function that can set memory protection of "reference only" and "writable" When an object that violates the set memory protection is accessed, it is detected and notified to the database management system and the execution control is transferred to the database management system. A computer system having only one function. It is possible to provide a database management system that provides a function equivalent to that of the conventional object-oriented database management system even when using.

Further, since the user program is described in a programming language that can be described in the same description format as the description format of the definition and operation of objects in the database, as the object structure definition, for example, the class definition of C ++ language is described. For the object creation / deletion, the object area can be secured / released in the virtual memory space by using the description format as it is, and for the object creation / deletion, for example, a description format that extends the object creation / deletion function of C ++ language. Therefore, the object reference / update in the database can be performed in the same description format as the object defined by the C ++ program, for example. Further, it becomes possible to define and operate the interrelationship of objects by using a pointer as in the C ++ language.

[0022]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A general embodiment of a database management system according to the present invention will be described below with reference to the drawings.

FIG. 1 shows the overall configuration of an embodiment of a database management system 2 according to the present invention in terms of the relationship between a user program 1 and a data program 3.

Input to the database management system 2 is
A processing start request 10, an object creation request 11, an object reference request 12, an object update request 13, and a processing end request 14 from the user program 1 to the database 3. The output of the database management system 2 is execution of processing on the database 3 corresponding to each request of the user program 1 and a response to the user program 1 of the result and the execution state.

The database management system 2 comprises an access management unit 21, a rearrangement management unit 22, a relationship management unit 23, a buffer area management unit 24, a buffer area 25 and page management information 26.

The database 3 is composed of data pages 31, page management information 32, and rearrangement management information 33.
The data page 31 is composed of a collection of sections called pages 311 and the object 312 of the database 3 is stored in the data page 31.

The buffer area 25 of the database management system 2 is composed of a collection of page areas 251. An object area 252 in the page area 251 is read from the database 3 in response to an object creation request, object reference request or object update request from the user program 1.

FIG. 2 shows the data structure of the page management information 26, 32. Page management information 26, 32
Holds the correspondence relationship between each page 311 in the data page 31 and the page area 251 in the buffer area 25 and the management information for memory protection.

Thirdly, the data structure of the rearrangement management information 33 is shown. The relocation management information 33 is used to correct the address pointed to by the pointer in the object area 252 when the object associated with the pointer is allocated to the page area 251 having an address different from that in the previous object operation processing. belongs to. Therefore, as shown in FIG. 3, the rearrangement management information 33 holds address information indicating the position of the pointer itself in each page and address information indicating the position in the database of the object pointed to by the content of the pointer.

Next, the operation of the above-mentioned general embodiment will be described. The processing start request 10 specifies whether the object operation simultaneously performed up to the processing end request 14 is only the reference of the contents of the database 3 or the update is included. The buffer area management unit 24 initializes the page management information 26 in the database management system 2 in response to the processing start request 10. At this point, the buffer area 25 does not exist. The page management information corresponding to each page is created from the page part of the page management information 26 of the database 3 or the page part of the page management information 32 of the database 3 at the time of the first creation, reference and update of the page. Reading is done.

In response to the object creation request 11, the access management unit 21 creates an object area 252 of the size required for the object in the buffer area 25. If the buffer area 25 does not have a free area for securing the object area 251, the access management unit 21 extends the buffer area 25 through the buffer area management unit 24 to create a new page area 251, and an object area 252 is created therein. It secures and reads the page management information 26 from the page management information 32 of the database 3. The memory protection of the page management information 26 is set according to the database reference or update designation of the processing start request 10.

In response to the object reference request 12 or the object update request 13, the access management unit 21 determines using the page management information 26 whether or not the object exists in the buffer area 25, and if the object does not exist, The relocation management unit 22 determines the allocation position of the page area 251 for reading the page 311 containing the object, the buffer area management unit 24 allocates the page area 251, and reads the page 311 containing the object in the database 3. . At this time, the management information about the page assigned to the page management information 26 is read from the page management information 32 in the database 3. The memory protection of the page management information 26 is set according to the database reference or update designation of the processing start request 10.

When the page 311 is read from the database 3 into the buffer area 25, the page area 251 is allocated to the page including the object pointed by the pointer included in the page. At this time, the relocation management information 33 is used to rewrite the value of the pointer so that it becomes the address of the corresponding object area 252. That is, the destination pointed by the pointer in the buffer area 25 is always the buffer area 25. This is shown in FIG. The page (page A in the figure) read in the buffer area 25 is an object pointed to by all pointers (pointers b and c in the figure) included in the relationship management unit 23 (object b in the figure). , C) (pages B and C in the figure) are allocated in the buffer area 25, the address is determined in the memory space of the object, and the value of the corresponding pointer (pointers b and c in the figure) is assigned to that address. rewrite. At this time, memory protection is performed for the page area (page A in the figure) in which the object is read and the page area (pages B and C in the figure) including the object pointed to by the pointer in the page according to the designation of the processing start request 10. To set.

After the target object has been read into the buffer area 25 or has already been read, if the memory protection is not violated, the object reference / update is performed in the object area 252 in the buffer area 25.
Is done directly against. When an object that violates the memory protection is accessed, the access management unit 2
In step 1, the processing of the database is terminated, and the user program 1 is notified to that effect.

In response to the processing end request 14, the database management system 2 changes the contents of the page area 25.
1 is written in the corresponding page 311 of the database 3 with reference to the memory protection information of the page management information 26. At this time, the position of the pointer in each page area 251 and the position of the object indicated by the pointer value in the database 3 are calculated and written in the database 3 as the relocation management information 33 shown in FIG. Also, the page management information 26 for the changed page is written in the database 3.

The memory management mechanism of the computer system assumed in this embodiment is only to "have a function of allocating and releasing a memory area". That is, as described above, when there is a function that can set memory protection such as "unusable", "reference only", and "writable" for the allocated memory area, or when an object that violates the set memory protection is accessed. , It does not need the function to detect it, notify the database management system, and transfer the execution control to the database management system.

The database management system 2 uses the memory management mechanism of this computer system to manage the buffer area 25 such as relocation processing and reference page allocation as described above.

The roles of the respective parts in the database management system 2 and the processing performed by them are as follows.

Access management unit 21 ... Analysis of processing request of user program 1, checking of memory protection and distribution of processing according to request Relocation management unit 22 ... Allocation address of reference page at buffer reading area 25 when page is read Relationship management unit 23 ... Modification of pointer value in the page corresponding to assigned reference page address and change of buffer area Creation of relocation management information 33 when writing page area 251 to database 3 Buffer area management Part 24 ... Allocation of page area 251 in buffer area 25 and page transfer with database 3 Next, a specific embodiment corresponding to the above-described general embodiment will be described. The present embodiment is a database management system that enables simultaneous use of a plurality of databases from a plurality of user programs.

The overall construction of this embodiment is shown in FIG. In FIG. 6, a range surrounded by a solid line indicates processing modules or information for configuring the database management system.
The description in the processing module shows the outline of the processing. The range surrounded by the dotted line shows the correspondence with the configuration of the database management system 2 in FIG. 1, and the arrows show the main processing flow.

In this embodiment, a client / server configuration is adopted in the distributed computer system in order to concurrently use a complex database from a complex user program. Therefore, the database management system 2 is divided into the client process 5 and the server process 6 and exists. Further, in order to manage a plurality of databases, it has a database catalog 7 that holds information about each database. Further, in the present embodiment, the schema that holds information about the data structure of the objects in the database 3 is also managed as an object in the data page 31.

In the configuration of the present embodiment, with respect to the function provided by the general axis described above, a database management routine 50 for managing a plurality of databases, and a transaction for making it available to a plurality of user programs at the same time. A management routine 51, a schema management routine 52 that manages information about the data structure of objects in the database 3, an object search routine 54 for obtaining an object that matches the purpose, and an object access management routine 55 are added. In the object access management routine 55, the object designated at the time of object reference / update is the buffer area 25.
If there is not, the buffer management routine 58 is used to read the specified object into the buffer area 25, and then the object is referenced and updated.

These processing routines are for realizing a database management function equivalent to that provided by the conventional database management method. Further, in the present embodiment, in order to realize the database system, the error management routine 56, the communication management routines 59 and 60, the database catalog management routine 61, and the physical database management routine 62 are included in the configuration.

The database access unit 21 of the above-described general embodiment is the schema management routine 5 in this embodiment.
2. Object management routine 53, object search routine 54, object access processing routine 55
And a physical record management routine 56. Further, the relocation management unit 22, the relationship management unit 23, and the buffer management unit 24 of the above-described general embodiment are realized as a buffer management routine 58 in this embodiment.

Next, the operation of the specific embodiment will be described below with reference to FIG. In response to a database creation / deletion request of the user program 1, the database management system 2 creates or deletes a new database 3 and registers or deletes it in the database catalog 7.

At the start of database processing, the data structure definition information of the data held in the specified database 3 is read into the buffer area 25.

The schema reference determines the type of object for performing subsequent object operations (creation / deletion, retrieval). Schema creation / deletion creates / deletes management information about the object to be operated.

The operation of this embodiment regarding the start of transaction processing is the same as the operation of the processing start request in the above-mentioned general embodiment.

In response to the designation of the object search, the database management system 2 searches the objects belonging to the designated type of object for an object that matches the designated condition and returns the object identifier (pointer) to the user program 1.

The operation of this embodiment regarding object creation / deletion, object reference, and object update is the same as that of the general embodiment described above.

The operation of this embodiment regarding the end of transaction processing is the same as the operation of the end of processing of the general embodiment described above.

As described above, the structure of this embodiment has the following effects. The user program 1 refers to and updates the object in the buffer area 25 of the database management system 2 via the object identifier. Therefore, processing such as moving objects between the user program and the buffer area and checking / converting the object format for each individual object reference / operation required in the conventional relational database management method, etc. Since it is not performed, the processing efficiency is significantly improved over the conventional relational database management method. Further, compared with the conventional object-oriented database management system, the database management system 2 performs page management processing based on the object identifier specified at the time of object reference / update, but the processing efficiency drops, but a special memory for the computer system is used. Since no management function is required, this database management method can be realized even with a normal computer system.

Next, another general embodiment of the present invention will be described. As shown in FIG. 4, the database use method of this embodiment uses the buffer area 25 of the database management system 2 as the program data area 41 of the user program 1.
It is used similarly to. In the present embodiment, in addition to the operation based on the object value held by the relational database, the object operation using the object identifier and the function of expressing and operating the interrelationship of the object are provided. In this embodiment, a pointer indicating the address of the object is used as the object identifier.

The database operation function provided by this embodiment is as follows. (1) Process start (2) Object creation / deletion (3) Object reference (4) Object update (5) Process end Among these operation functions, (1), (2), and (5) are from the user program 1. This is performed by calling the processing procedure of the database management system 2. With regard to the functions (3) and (4), the user program 1 makes a reference / update request to the database management system 2 by designating the object with an object identifier (pointer). Similar to the object in the program data area 41, the object is directly referenced / updated by the command word of the computer system.

Next, the operation of the other general embodiment described above will be described. At the beginning of a series of processes for the database 3, the user program 1 specifies a process start request 10 to the database management system 2. In the processing start request 10, at the same time, it is designated whether the object operation performed up to the processing end request 14 is only to refer to the contents of the database or to include the update. The database management system 2 uses the page management information 26 and the buffer area 2
5 is initialized and the result information is returned to the user program 1.

In response to the object creation request 11, the user program 1 passes to the database management system 2 information indicating the type of object together with an object creation instruction. The database management system 2 secures the object area 252 in the buffer area 25 and returns the pointer as an object identifier to the requesting user program 1.

The object reference request 12 or the update request 13 is performed by designating an object identifier (that is, a pointer to the operation target object). When it is the first object reference or object update in the page, the database management system 2 buffers the corresponding object from the database 3 based on the designated object identifier (pointer) as described in claim 1. The page area 251 in the area 25 is read, and the object is referenced and updated. At this time, if the object includes an object identifier (pointer) that points to another object, the database management system 2 automatically reads the object pointed to by the pointer into the buffer area 25 as shown in claim 1. Adjust the value so that you can access the object correctly via the pointer after that.

After a series of database operations, the user program 1 specifies the processing end request 14. As a result, the database management system 2 writes the page area 251 including the object created / updated in the buffer area 25 to the page 311 of the database 3, and the information of the processing result is returned to the user program 1.

Next, another concrete embodiment corresponding to the above-mentioned other general embodiment will be described. The database use method of this embodiment uses the configuration shown in FIG. 4, and allocates the storage space managed by the database management system as the buffer area 25 of the data page 31 shown in FIG. 4 and the storage area of the program object area 41 of the user program 1. Is used as a kind of.

In this embodiment, the C ++ language is used as the parent language of the database operating language. That is, the object structure definition is performed using the class definition statement of the C ++ language. The database definition and operation function provided by this embodiment are shown in FIG.
, And these are C ++ as shown in FIG. 7.
It provides a language extension function and a usage method using a predefined class prepared by the database management system 2. In this embodiment, a pointer indicating the object address is used as the object identifier, which is C + in this embodiment.
Corresponds to the + pointer.

These database definitions and operation instructions are given in the following order. However, once the database and schema are created, only 3 to 8 are specified. 1. Database creation 2. Schema creation 3. Start database processing 4. Transaction start processing 5. Schema reference 6. Object operation (object creation / deletion / search / reference / update) 7. Transaction processing end or processing discard 8. End of Database Processing Next, the operation of the other specific embodiment described above will be described below with reference to FIG.

The data structure definition is performed using the class definition of the C ++ language. Not only the definition of the data structure but also the constructor, the extinct child and the object operation function can be defined.

Database generation is to newly create and initialize a database. Database erasure also reduces the existing database.

The start of database processing enables subsequent schema reference, transaction processing, and object operation for the designated database. The end of database processing ends the processing of the database.

The transaction processing is for enabling a plurality of user programs 1 to operate objects concurrently in parallel. At the start of the transaction process, "reference" or "update" is specified as exclusive information of the object operation until the end of the transaction process. Performs exclusive processing related to object operations regarding object operations performed from the start of transaction processing to the end of transaction processing,
Object creation and object update performed during that time are collectively reflected in the database 3 at the end of the transaction processing. Exclusive processing relating to object operations performed from the start of transaction processing to the discard of transaction processing is performed, and the object creation and object update performed during that time are collectively discarded at the time of transaction processing discard and are not reflected in the database 3.

In the schema creation, the data structure definition information of the objects used in the database 3 is created and held in the database 3. The object structure is described in the C ++ language class definition statement in the class definition file specified in the argument as shown in the description example of the schema creation in FIG. Sukima deletion is to delete the data structure definition information defined in the database by specifying the class name. For the schema reference, a type pointer which is a pointer to data structure definition information is returned to the user program 1 for subsequent object operations.

Regarding object creation, an object having a corresponding data structure is created in the buffer area 25 by using the type pointer as an argument of the new operator. The pointer (object identifier) of the created object is returned to the user program. As shown in the description example of object creation in FIG. 7, at this time, if a constructor is defined in the data structure definition and is described after the new operator, the object can be initialized at the same time. The object created in the buffer area 25 is written in the database 3 at the end of the transaction processing. Regarding object deletion, it is reflected in the database 3 at the end of the transaction in which the object in the buffer area 25 is deleted.

The object search is to search a corresponding object in the database by designating a search condition. The pointer (object identifier) of the object area 252 in the buffer area 25 of the retrieved object is returned to the user program 1. The description example of the object search of FIG. 7 is an example of searching for “person whose spouse's name is SAZAE” from the person.

Object reference and object update are pointers (object identifiers) of the object
Is specified by the "->" operator to refer to and update the object in the buffer area 25. In the processing of the "->" operator, the database management system determines whether the pointer of the object is a database object (that is, the object area 25 of the buffer area 25).
2) or not, and if it is a database object, the processing shown in the other general embodiment described above is performed.
If it is not a database object, it is processed as a program object in the program data area 41.
As shown in the description of FIG. 7, multiple modification by a pointer is also possible. When the corresponding database object is not in the buffer area 25, it is read from the database 3, and in the case of pointer multiple modification, the object pointed to by the pointer is sequentially read from the database 3 into the buffer area 25, and then object reference and update are performed. Is done. The object update result is written in the database 3 upon completion of the transaction processing.

According to the configuration of this embodiment described above, even in a computer system that does not have a special memory management mechanism such as virtual memory, it is possible to provide a usage method equivalent to an object-oriented database using them.

That is, according to this embodiment, the description format of the class definition of the C ++ language is used as it is as the data structure definition, and C ++ is used for the object creation / deletion.
The object area 252 can be secured / released in the buffer area 25 in the description format in which the object creation / deletion function of + language is expanded. Therefore, the object reference / update in the database 3 can be performed in the same description format as the object defined by the C ++ program. Also, the definition and operation of the mutual relation of objects can be performed using the pointers as in the C ++ language. As a result, according to this embodiment, the following effects can be obtained. 1. Improvement of operability of database-Definition and operation of complex data structure-Representation of relationship between complex objects by pointers 2. Other than object creation / deletion and retrieval, it is possible to handle objects as defined in the program without being aware that they are database objects. No special database language required 4. Easy migration from existing programs to programs using databases

[0072]

As described above, according to the configuration of the present invention, even in a computer system having no special memory management mechanism, that is, Memory area allocation and release functions 2. "Unavailable" for the allocated memory area,
Function that can set memory protection of "reference only" and "writable" When an object that violates the set memory protection is accessed, it is detected and notified to the database management system, and the execution control is transferred to the database management system. Even when the system is used, the user can describe the object definition / operation of the database as an extension of the data definition / operation function of the existing programming language. Complex data definition function equivalent to programming language 2. Definition function of interrelationship between complex objects equivalent to programming languages. 3. Object reference / operation description function equivalent to a programming language 4. Fusion of database operation and program processing It is possible to provide a database management method for efficiently realizing the functions of 5.1, 2, 3, and 4.

[Brief description of drawings]

FIG. 1 is a diagram showing the overall configuration of a database management system of the present invention showing the relationship between a user program 1 and a data program 3.

FIG. 2 is a diagram showing an example of contents of page management information.

FIG. 3 is a diagram showing an example of contents of relocation management information.

FIG. 4 is a configuration diagram showing a database usage method.

FIG. 5 is an explanatory diagram of object relocation and relationship management.

FIG. 6 is a diagram showing an overall configuration of a specific example.

[Explanation of symbols]

 1 User Program 2 Database Management System 3 Database 5 Client Process 6 Server Process 7 Database Catalog 10 Processing Start Request 11 Object Creation / Deletion Request 12 Object Reference Request 13 Object Update Request 14 Processing End Request 21 Access Control 22 Relocation Management section 23 Relationship management section 24 Buffer area management section 25 Buffer area 251 Page area 252 Object area 26 Page management information 31 Database 311 page 312 Object 32 Page management information 33 Relocation management information 40 Program command section 41 Program data area 50 Database management Routine 51 Transaction management routine 52 Schema management routine 53 Object management routine 54 Object Search routine 55 Object access management routine 56 Error management routine 57 Physical record management routine 58 Buffer management routine 59 Client side communication management routine 60 Server side communication management routine 61 Database catalog management routine 62 Physical database management routine

Claims (6)

[Claims] 1. A database management system for managing a database including structured objects and information on mutual relationships between these objects, comprising: a page area for storing page information of the database and an object for storing objects of the database. A buffer area for storing information of the database, an access management section for receiving an instruction from a user program and referring to / updating an object in the object area, and the access management section In response to the instruction, a buffer area management unit that allocates an object of the database to the object area via the page area, and page information in the database and the page information of the page in which the object of interest is stored. A page information management unit that manages a relationship with page information in the object area, the access management unit refers to page information managed by the page information management unit, and an object requested by a user program is stored in the object area. The buffer area management unit is instructed to refer to the object in the object area when it is in the object area, and when it is not in the object area, the object is determined from the database. An object access management routine unit for instructing the buffer region management unit to allocate
A database management system comprising: 2. The buffer area management unit manages the buffer area via a buffer area management unit that utilizes a memory management function of a connected computer system, and stores the objects of the database via the page area. The object area is allocated to the object area.
Database management system described in. 3. The database management according to claim 1, wherein the buffer area management section includes an object relocation section that enables the same object of the database to be allocated to different page areas of the buffer area. system. 4. The database management system according to claim 1, further comprising a relationship management unit that manages a relationship between objects allocated to different page areas. 5. The user program according to claim 1, wherein the user program is written in a programming language that can be written in the same description format as the description format of the definition and operation of the object regarding the object in the database. Database management system. 6. The database management system according to claim 5, wherein the programming language is a C ++ language.