KR100490750B1 - Method for automatically managing relation between database entities represented by using persistent state object of enterprise javabeans - Google Patents

Abstract

The present invention relates to a method for managing the automated interrelationship of database entities represented by the state object of the EJB 2.0 specification of SUN, a server-side Java standard for developing an enterprise application system. In particular, the CMR of the CMP entity bean of the EJB 2.0 specification Server-side processing method for. In the present invention, first, when an EJB client calls a CMR method of an EJB bean object on the server side, it grasps a relationship (1: 1, 1: M or M: 1) between EJB entity beans and sets it in a state object in memory. After the distributed transaction is completed, generating a SQL query by mapping the type of the memory table and the type of the database table in the actual RDBMS by referring to the relationship field value and foreign key value of the state object in the set memory. Applying to the actual database. According to the method of the present invention, when creating an EJB 2.0 specification web application server system, as described in the EJB specification, by providing a method that can be automated by the container without the developer directly coding the table join operation of the relational database, By using EJB, you can improve productivity, reduce maintenance and maintenance costs when developing business logic.

Description

OBJECT OF ENTERPRISE JAVABEANS}

The present invention belongs to the field of server-side distributed component-based software for writing business task logic, and in particular to the field of Java-based relational database related middleware. The present invention relates to a middleware system that can be applied to a persistence management module of an application server system that implements an Enterprise JavaBeans (EJB) standard together with a part for automatically managing a container managed persistence (CMP) field.

Recently, as the Internet e-commerce, corporate intranet information system has been converted into the web environment and the Internet portal system is activated, it is possible to reuse and disassemble and assemble large-scale business processes and business logic, and to develop and maintain distributed components based on easy development and maintenance. The importance of application server systems is gaining much attention. In addition, development of standardized development APIs and business applications through system compatibility between companies is required.

If you develop a complex enterprise application system without the distributed component-based application server system and the API (Application Programming Interface) for the above, developers can manage not only database-related logic but also distributed transaction management for EIS systems, authentication of users and You'll have to code the middleware pieces that are responsible for such things as access checks, instance management of logic objects, and state management. Also, when porting a built system to another system or writing another system, the developer will have to code the middleware part that plays the above role each time. The application server system includes a middleware part that is commonly required by developers and a container part that implements a standard distributed component architecture (EJB) for developing, distributing, and running business logic. It is a system that can increase development productivity by simply coding business logic instead of coding directly. In addition, business logic developed by the standard distributed component architecture can run on any application server system that supports EJBs, increasing logic reusability.

EJB is an industry-standard distributed component architecture driven by SUN to meet the requirements of these enterprise application system developers. EJBs write complex underlying middleware for developers writing enterprise business applications such as bean lifecycle management, authentication and access checks, distributed transaction processing, user and object state management, multithreading, connection pooling, and persistence management. By making application server systems easier to develop, you can increase development productivity and business compatibility.

The EJB specification and the concept of persistence referred to in the present invention mean that data is stored in a database, which is a permanent storage medium. That is, a variable value is mapped to a permanent storage medium by mapping a variable in memory to a specific storage area of an RDBMS. It is a function to be reflected automatically. An EJB-compliant entity bean is an object that represents a row entity in an RDBMS table. If you change the properties of a bean using the methods provided by the entity bean, this change will cause the value of that column of the table in the RDBMS. Will be reflected automatically. At this time, the entity referencing entity bean does not consider the table structure of the database and the communication with the RDBMS at all. Not considering communication with the RDBMS means that you do not need to use SQL statements at all. However, bean developers of bean managed methods (BMP EJB entity beans) that code their persistence management code directly in the bean are forced to deal with SQL because they have to do their own mapping between databases and objects. However, in the case of the Container Managed method (CMP EJB entity bean) in which the container automatically manages persistence, which is the object of the present invention, the entity mapping is performed by a tool provided by the provider of the EBJ server. There is no need to use SQL at all.

The EJB specification is currently version 2.0, and with the concept of a message bean capable of handling messages sent from a JMS-compliant message server against the existing 1.1 specification, and the concept of entity beans described above, a container managed relationship (CMR). Added the concept of. The concept of CMR includes the management field or foreign key in the concept that only the field value including the primary key is included in the property of the entity bean. Is a concept that is automatically managed.

The present invention provides a server-side processing method for a set or get method of a CMR attribute that can be applied to a commercially available application server system or an EJB server system of an open source project. By providing a method of implementing state management of expressed persistence objects, it aims to compensate for the disadvantages of overall performance degradation of a server caused by the complexity of the existing EJB server.

In order to achieve the above object, the automated correlation management method of the database entities represented by the EJB persistent state object according to the present invention, the EJB client to call the container managed relationship (CMR) method of the EJB server-side source bean And extracting a relation field representing a relationship with a partner bean from the source bean's persistent state object if the type of the called method is a get method, and extracting the join object of the partner bean through an SQL query. The EJB object creation and return step of creating and returning an EJB object corresponding to the join object to the EJB client according to the number of join objects, and the EJB local object of the source bean when the type of the called method is a set method. Extracting the data source; extracting the persistent state object and the correlation information of the partner; The state information setting step of setting the state information of the nugget, and the correlation between the source bean and the partner bean (1: 1, 1: M or M: 1) and processing the called method according to the identified relationship Method processing step, and applying the processing result of the method processing step to the database of the RDBMS.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. Those skilled in the art will be able to better understand the objects, features and advantages of the present invention through this embodiment.

First, Figure 1 shows the configuration of the whole system in which the method according to the invention is used.

Referring to FIG. 1, an EJB developer creates an EJB entity bean 101 that represents row data on an RDBMS as an abstract object (102). Here, an entity may be defined as a person, a location, an object, a concept, an activity, and an event that can be informatized with interest in a company.

Then, using the tools provided by the EJB server, the created EJB entity bean 101 is a persistent object that contains state information about the type object or persistent EJB entity bean 104 and the RDBMS. 105 is generated (103), and distributed to the container 113. Here, the container manages the life cycle of the instance such as the state information of the EJB bean composed of session beans, entity beans, and message beans, and meter data generated by the developer, such as security information, CMR information, and transaction information. Finally, the module on the EJB server that is responsible for invoking the method.

The type object 104 is managed by the container 113 and consists of a method (get or set) relating to a field mapped to a database and a method (get or set) relating to a relationship (a foreign key) between a table, and in fact, an EJB. An object that acts on an empty method call from a client on the server The developer adds the business methods to the abstract object except for the above method. The persistent state object 105 is an attribute object that exists for each instance of an EJB entity bean that contains information attributes (cycles, foreign keys, etc.) corresponding to each column of the RDBMS table.

Processing 107 on persistent fields requested from client user 106 is performed by CMP field manager 109 (108), and processing on relationship fields between tables is performed by CMR field manager 110 ( 111) First, if you change the value of a persistent state object that exists in memory and finally the distributed transaction that occurred during the execution of the method is committed, or if the method invocation is executed without the distributed transaction, then the query manager ( 112) to generate a SQL type query to change the value in the RDBMS.

2 to 6, the detailed steps of the method according to the invention are described.

FIG. 2 is a flowchart illustrating a CMR method processing process of the CMR field manager 110 shown in FIG. 1 in detail.

As shown in FIG. 2, when the client user 106 calls a CMR get or set method of an EJB entity bean (step 201), the container 113 first retrieves the relationship information between the called source bean and the partner bean. (Step 202), and branches the processing of the method depending on whether or not the method called by the client user 106 is a set method (step 203). At this time, the setting information is set by the developer, and when the container 113 places the empty object on the server, the container 113 loads and manages the information. If the method called by the client user 106 is a CMR get method, the container 113 performs processing for the get method (step 204), and if it is a set method, processing is again performed according to the relationship between beans. Branching (steps 205 to 207). Then, after changing the relationship field value and the foreign key value of the persistent state object 105 (step 208), and later a distributed transaction is committed or the method call is completed, the database on the RDBMS through the query manager 112 Join processing reflecting the changed information is performed (step 209).

An example of the set method and the get method will be described in detail as follows. First, an example of the get method will be described. If there is an order table and an item information table representing order information in the e-commerce shopping mall system, the order table is an order number, order product number, and user number, which are the primary keys of the ordered user's product list. Information and the like. The product number is set to a foreign key referring to the product information table. The product information table contains only pure product information such as the primary key product number and product name. In this case, if a system developer creates an EJB entity bean that represents an entity in the product information table (raw data on the RDBMS) that is the source bean and an entity in the order information table that is the partner bean, the two entity beans establish a foreign key relationship with the primary key on the RDBMS. All processing of the method on the relation field as a bean is managed by the container.

When a client (developer or shopping mall user) requests all order numbers for a particular product number, the container calls the getOrder () method of the product information EJB entity bean, and the getOrder () method is executed by the method proposed in the present invention. This returns the join entity on the RDBMS associated with the current merchandise entity.

3 details the process of get method processing (step 204) shown in FIG. First, the container 113 extracts a relationship information field with a partner to store the relationship value from the persistent state object 105 of the source (step 301), and obtains a join value on the RDBMS through the SQL query through the query manager 112. (Step 302). In this case, the query manager 112 returns the result of the join value to the container 113 in the form of an EJB local object, and the container 113 may indicate a case where there are one returned value from the query manager 112 and several cases. Divide the process by dividing. If the returned value is one, the container 120 creates an EJB local object of the partner (step 303), and sets the EJB local object of the partner returned from the query manager 112 to the relationship information extracted above. Next (step 304), the EJB local object is returned to the client user 106 (step 305).

In addition, if there are multiple returned values, the container 120 first creates a collection object (step 306), adds all EJB local objects returned from the query manager 112 to the object (steps 307 through 204). 309) Finally, the collection object is set in the relation field of the persistent state object extracted above (step 310), and the collection object is returned to the client user 106 (step 311).

4A illustrates an example in which a relationship between a source and a partner is 1: 1, and FIG. 4B illustrates a process of processing a set method in the case of FIG. 4A.

The relationship between the source and the partner shown in FIG. 4A is that when there are a1 and a2 instances of the entity bean object and b1 and b2 instances of the entity bean object, the relationship field method created by the developer is a1.setB ( The case of a1.getB ()) is shown.

The process of the method a1.setB (a1.getB ()) will be described with reference to FIG. 4B. First, the container 113 extracts an EJB local object from the container of the source object a2 (step 401), extracts the partner's persistent state object and relationship information from the container (step 402), and then persists the bean A which is the source object. Performs the process of setting the relationship field values of the state object.

The process of setting the relation field value of the persistent state object of the A bean is performed as follows. Set the relationship field of the persistent state object of a1 (old value is EJB local object of b1) to NULL (step 403), and set the relationship field of the persistent state object of a2 (old value is EJB local object of b2) to EJB of b1 Set to local object (step 404). The values set in the persistent state object are for the purpose of providing convenience to bean developers by expressing the object of the RDBMS table as a distributed object. The client call can be referred to multiple times for one distributed transaction or method call.

After setting the information of the persistent state object of the A bean, the process of setting the value of the persistent state object of the B bean which is a partner object is performed. First, the relationship field value of the persistent state object of b2 is set to NULL (step 405), and the field value of the persistent state object of b1 is set to the EJB local object of a1 (step 406).

When the setting process of the persistent state object is completed, a process of setting a foreign key value on the persistent state object to be referred to by the query manager 112 is performed to change the value of the RDBMS table. At this time, since the foreign key value may exist in either the source or the partner in the 1: 1 RDBMS relationship, the foreign key value set in both persistent state objects is searched for the foreign key value of the partner or source. Set to a value (steps 407 to 410). When the above processing is completed, the container 120 generates a field attribute value of the persistent state object of the a1, a2, b1, and b2 objects, which are instances of the A object, by using the query manager 112 as an SQL92 type query, and the RDBMS Reflect on the table (step 411).

FIG. 5A illustrates an example in which a relationship between a source and a partner is 1: M, and FIG. 5B illustrates a process of processing a set method in the case of FIG. 5A.

The relationship between the source and the partner shown in FIG. 5A is created by the developer when there are a1 and a2 which are instances of the entity bean object and b11, b12, b13, b21, b22 and b23 which are instances of the entity entity B object. Shows the case where the relation field method is a2.setB (a1.getB ()).

The process of the method a2.setB (a1.getB ()) will be described with reference to FIG. 5B. First, the meaning of execution of this method is explained from the perspective of the client user 106, and as described in the above example, it corresponds to a process of joining all the order numbers including the product from the product entity bean representing the specific product.

As with the 1: 1 process shown in FIGS. 5A and 5B, the container 113 first extracts the EJB local object of the source bean from the container (step 501), and extracts the partner's persistent state object and relationship information from the container ( Step 502). The process of setting the persistent state object of entity A first sets the relation field of the persistent state object of a1 to NULL (step 503), and sets the relationship field value of the persistent state object of a2 to b1N (where N is 1, 2, 3, ...) (step 504). Next, set the information of the persistent state object of entity bean B. First, the relationship field value of the persistent state object of each entity of b1N is set to the EJB local object of a2 (step 505), and the relationship field value of the persistent state object of each entity of b2N is set to NULL (step 506).

When setting the value of the persistent state object is completed, set the foreign key value to be used when the query manager 112 generates the SQL92 query. First, the foreign key value of the persistent state object of each entity of b2N is used as the primary key value of a1. Set to NULL set (step 507), and set the foreign key value of the persistent state object of each entity of b1N from the primary key value of a1 to the primary key value of a2 (step 508). In the above process, each entity of b2N has lost relationship with entity A, and a value set to NULL is deleted from the RDBMS by a Join Delete SQL query by the query manager 112. In addition, unlike the process illustrated in FIG. 4, the reason for setting the foreign key value only in the persistent state object of entity B is that the foreign key value is necessarily present in the M side in the case of a 1: M relationship. After the above process is completed, the query manager 112 generates an SQL query to change the value of the RDBMS by referring to the foreign key value set in the persistent state object and the relation field value (step 509).

FIG. 6A illustrates an example in which a relationship between a source and a partner is M: 1, and FIG. 6B illustrates a process of processing a set method in the case of FIG. 6A.

The relationship between the source and the partner shown in FIG. 6A is created by the developer when there are a1 and a2 instances of the entity bean object and b11, b12, b13, b21, b22, and b23 instances of the entity entity B object. Shows the case where the relation field method is a13.setB (a21.getB ()).

A process of the method a2.setB (a1.getB ()) will be described with reference to FIG. 6B. As shown in Figs. 4 and 5, the EJB local object of the EJB entity bean of the source is first extracted from the container (step 601), and the persistent state object and the relationship information of the partner are extracted (step 602). As in the above process, to change the relationship field value of the persistent state object of entity A, remove the b13 entity from the relationship field value (collection object) of the persistent state object of a1 (step 603), and the relationship of the persistent state object of a2. Add b13 to the field value (collection object) (step 604). To set the relationship field value of the persistent state object of entity B, change the relationship field value of the persistent state object of b13 from the EJB local object of a1 to the EJB local object of a2 (step 605).

After the setting of the value of the persistent state object is completed, the foreign key value of the entity b13 is changed and set from the primary key value of a1 to the primary key value of a2 (step 606). When the above process is completed, the query manager 112 generates an SQL query by referring to the relationship field value and the foreign key value to change the value of the RDBMS (step 607).

As described above, the method of the present invention can be applied as a CMR management module to various types of EJB 2.0 servers implementing the EJB 2.0 specification. Developers can apply the CMP and CMR concepts of the EJB 2.0 specification to the creation of business logic. Complex business logic, such as database join operations, can be simplified, resulting in increased application system productivity, development efficiency, and ease of maintenance and maintenance.

1 is a block diagram showing the configuration of an entire system in which the method according to the invention is used,

FIG. 2 is a flowchart specifically illustrating a CMR method processing process illustrated in FIG. 1;

3 is a flowchart specifically illustrating a process of a get method processing illustrated in FIG. 2;

4A is a diagram illustrating an example in which a correlation between a source bean and a partner bean is 1: 1;

4B is a flowchart specifically illustrating a process of a set method in the case illustrated in FIG. 4A.

5A is a diagram illustrating an example in which a relationship between a source bean and a partner bean is 1: M,

FIG. 5B is a flowchart specifically illustrating a process of a set method in the case illustrated in FIG. 5A.

FIG. 6A is a diagram illustrating an example in which a correlation between a source bean and a partner bean is M: 1,

FIG. 6B is a flowchart specifically illustrating a process of a set method in the case illustrated in FIG. 6A.

<Description of the code | symbol about the principal part of drawing>

101: EJB entity bean 104: persistent EJB entity bean

109: CMP Field Manager 110: CMR Field Manager

112: Query Manager 113: Container

Claims (7)

The EJB client calling the container managed relationship (CMR) method of the EJB server-side source bean, Extracting a relationship field indicating a relationship with a partner bean from the persistent state object of the source bean when the type of the called method is a get method; Extracting a join object of the partner bean through an SQL query; Generating and returning an EJB object corresponding to the number of join objects, generating and returning an EJB object corresponding to the join object to the EJB client; If the type of the called method is a set method, extracting an EJB local object of the source bean; Extracting the persistent state object and correlation information of the partner; State information setting step of setting state information of the source bean and the partner bean according to the mutual relationship; A method processing step of identifying a correlation between the source bean and a partner bean (1: 1, 1: M or M: 1) and processing the called method according to the identified relationship; Applying the processing result of the method processing step to a database of an RDBMS; An automated correlation management method for database entities represented by EJB persistent state objects, including. The method of claim 1, The EJB object creation and return step, If the number of join objects is one, create an EJB local object of the partner bean corresponding to the join object, set the EJB local object in the relation field, and return the EJB local object to the EJB client. , If the number of join objects is two or more, create EJB local objects of two or more partner beans corresponding to the join object, create a collection object of the partner beans, add the EJB local object to the collection object, and And setting the collection object in the relation field, and returning the collection object to the EJB client. The method of claim 1, The state information setting step, If the correlation is 1: 1, Setting a relationship field value of the persistent state object of the source bean and the partner bean to NULL; Setting relation field values of the persistent state object of the source bean and the partner bean as EJB local objects of the partner bean and the source bean, respectively; Setting a foreign key value of the persistent state object of the source bean and the partner bean to NULL; Setting foreign key values of the persistent state object of the source bean and the partner bean to primary key values of the partner bean and the source bean, respectively; And reflecting the field values of the persistent state object of the source bean and the partner bean on an RDBMS table. The method of claim 1, The state information setting step, If the correlation is 1: M, Setting a relationship field value of the persistent state object of the source bean and the partner bean to NULL; Setting a relationship field value of the persistent state object of the source bean and the partner bean as a collection object of the partner bean and an EJB local object of the source bean, respectively; Setting a foreign key value of the persistent state object of the partner bean to NULL; Setting a foreign key value of the persistent state object of the partner bean to a primary key value of the source bean; And reflecting the field values of the persistent state object of the source bean and the partner bean on an RDBMS table. The method of claim 1, The state information setting step, If the correlation is M: 1, Removing or adding the partner bean from a relationship field value of the persistent state object of the source bean; Setting a relationship field value of the persistent state object of the partner bean to an EJB local object of the source bean; Setting a foreign key value of the persistent state object of the partner bean to a primary key value of the source bean; And reflecting the field values of the persistent state object of the source bean and the partner bean on an RDBMS table. delete delete