KR920007947B1 - Data base system realization method - Google Patents

Abstract

The static roll-back relational data base system having functions for supervising history of data base state change and Hangul data processing is realized using the method. The method includes steps: (A) displaying initial screen and opening corresponding data base directory; (B) receiving a data base edit instructions; (C) importing present state relation and change history relation, displaying relation manupulating screen, and displaying a first tuple; (D) receiving screen mode instructions transmitted from a user; (E) executing Hangul data inserting and processing module, historical query processing module, and a tuple change history screen manipulating module, and (E) changing the mode from execution to instruction when the input instruction of the step (D) is a transition instruction.

Description

SR2DBS Time Support Database System Implementation Method

1 is a schematic diagram showing the configuration and function of a time-assisted database system according to the present invention.

2 is an operation mode transition diagram of a time-assisted database system according to the present invention.

Figure 3 is an initial screen partition division according to the present invention.

4 is a relation definition screen partition division according to the present invention.

5 is a relation operation screen partition division according to the present invention.

6 is an application example of a time support database system according to the present invention.

7 is a process flow diagram of a conventional TBE.

8A to 8F show a process flow according to the present invention.

The present invention relates to a time support database system, and more particularly, to a method for implementing an SR ² DBS time support database system.

SR ² DBS stands for Static Roll-back Relational Data Base System and is a kind of time-supported database system that has recently attracted attention in database application technology.

The time-assisted database system maintains the history of change of information over time, and it is possible to define the information in the past such as definition of past information that could not be provided in the conventional database system or homogeneous analysis based on a series of change history. A database system that provides information processing.

In addition, the SR ² DBS can add a function to process and manage Korean data that was difficult to provide in a TROLL / USE Relational DBMS.

Most database systems in use today have the ability to represent and store only a single snapshot of the real world. Therefore, to change the contents of the database means to modify the state of the data. In this case, only the latest changed data state is kept in the database, and the past contents are deleted from the database.

In other words, the conventional database system does not provide the ability to record and process the temporal aspects of the real world. Therefore, such as historical history of the past state or trend analysis according to a series of change history. Not only does it not provide time-based processing, there is no efficient way to record retroactive or post-mortem changes to the database, and regular backups are required to preserve all data that has changed in the past. up), checkpoint (check-point function to check and store database contents at a specific time), and transaction log (Thansaction Log) to store the data operation data generated in the database in a specific file Backtracking files to restore past data content) Maintenance costs are holding a lot, it takes a lot of time and manpower.

In addition, in the conventional database system (TROLL / USE RDBMS), the processing and management of Korean data is hardly provided, and there is a significant limitation in the application of the actual database. The TROLL database query is also easily used by non-expert users. There are many non-procedural forms that are difficult to do.

However, it is possible to implement the time support and the polarization function by improving the function considering the following matters in the conventional TROLL / USE RDBMS.

-Extension and efficient management of storage area to store change history of database.

-Efficient provision of time information.

-Provide a time related database query system.

Support change history to prevent performance degradation of other database systems.

-Application of the New Generation query system.

-Provides processing and management of Korean data.

Accordingly, the present invention provides a method for implementing a SR ² DBS time support database system that enables a change history management function of a database state according to a time that a conventional TROLL / USE RDBMS cannot provide, and a processing and management function of Korean data. The purpose is to.

In order to achieve the above object, the present invention maintains a history of change of information over time to process information such as a query on past information or trend analysis according to a series of change history, processing of Korean data and A database system for supporting a management function, comprising: a first step of receiving a database directory as an input parameter, checking whether the database directory exists, and outputting an initial screen and opening the database directory if present; Receiving a command of a database level; In the second step, when a command for editing or updating a specific relation is input, the status relation and the change history relation are imported, the relation manipulation screen is output, and the first tuple value of the relation relation is displayed and executed. Transitioning the mode from the command mode to the screen mode; A fourth step of receiving an operation screen and a screen mode command from a user for manipulating a database for a current relation; A fifth step of performing a Hangul data insertion and processing module, a change history query word processing module, or a tuple change history screen manipulation module according to the command input through the fourth step; If the command inputted through the fourth step is a transition command to a command mode, a sixth step of transitioning the execution mode to the command mode and proceeding to the second step to receive a database-level command again; And SR ² is performed, including a seventh step wherein when a command for the service termination is input through a two-step terminate processing by closing the current state relation and exported to the history relation (EXPORT) and a database directory that is currently imported implementation of a DBS time support database system and, if the command is entered to define the new relation in the second step of the method performed by calling 7 database definitions parent SR ² DBS and 8, perform the complete If so, the present invention provides a method for implementing an SR ² DBS time support database system, which is performed by further including an eighth step of moving to the second step.

Therefore, the function of the TBE (TROLL Relation Browser Editor: name of a tool that can manipulate relations provided by the TROLL database) based on the existing troll / use relational database system of the present invention (TROLL / USE RDBMS) Was improved and expanded.

This will be described in more detail the main purpose of the SR ² DBS of the present invention.

Implementation of a time-assisted database system based on the next generation query system.

-Processing and management of Korean data.

-Prevent degradation of existing systems due to integration of time support.

-Provides time information by transaction time.

In general, time-assisted database models are based on the type of time supported: static (static), static rollback (transaction time), historical (effective time) Valid Time), Tamperal (Translation Time and Validated Time Support), and Database Model. SR ² DBS time-supported database management system is based on the static rollback database model based on dual transaction time information, and issues such as implementation approach and query system type. In addition, this paper is designed considering the versioning strategy and unit of change history information, the method of organizing storage area, and the method of integrating time information. It was then implemented by integrating it into the Roll Relation Browser Editor (TBE), one of the User SW Engineering (USE) packages based on the TROLL / USE RDBMS, which is a screen-oriented relation browser and editor. Relation Browser Editor) is a tool that provides convenient and easy data manipulation functions for relations defined by TROLL / USE RDBMS.

In particular, it expands the scope of use of time-supported database systems by adding the handling of Korean data that is not provided by existing TROLL / USE RDBMS and TBE. SR ² DBS is a Direct Manipulation Database Processing System that adds a time attribute to an existing relational database model, allowing you to manage the change history of the database. It provides a screen-oriented related browser editing and screening function for the history and has the following features.

Screen-orlented Processing.

Direct Manipulation of Objects.

-Object Revision History of Interest Visibility.

-Rapid Reversible Actions.

-Transaction Time Revision Number Supporting.

Han-gul-Data Supporting.

2-level storage structure.

The main environment for implementing a time-supported database system is based on a TROLL / USE RDBMS on a Sun 3 / 250Workstation system.

To extend its capabilities to support the time-assisted database model, the system can accommodate the following design techniques to achieve maximum effectiveness in a somewhat limited environment.

Implementation approach: On-top approach.

-Query system: Direct Mani-pulation Query Language System.

Versioning strategy over time: Complete Versioning.

Versioning unit: Tuple-level versioning

How to integrate time information: Final-Direct Timestamping

Storage structure of a time-supported database: 2-level stacked versions.

However, the most application problem in applying the proposed design technique is the database query system. The query system of TROLL / USE RDBMS provides a query that belongs to the line-by-line prompting query language, so the direct manipulation support data decided to be provided by the time-supporting database system. You must design a new direct control system that you decide to provide in the base system.

However, TBE, one of the USE packages, provides the ability to edit relations defined by TROLL / USE RDBMS through screen-oriented manipulation functions. It provides a roll-back query system for change history information.

Figure 1 shows the configuration and function of SR ² DBS through the expansion of TBE.

The main functional structure of the time-supported database system consists of the initial screen processing module, the relation definition module, the relation manipulation module, and the time-assisted relation manipulation module. The main environment supporting them is the data manipulation language (C programming language). It consists of Troll / USE Library Functions supporting Embedded Data Manipulation Language, Troll / UBE RDBMS, and Cursses Packaget (CRT Terminal Control Package).

The main function of the initial screen processing module is to accept the time-assisted database system access request from the users, open the corresponding database directory, display the initial screen, and provide the time-assisted database management service from the user. Be prepared to accept the request.

The relation definition module is responsible for defining relations through interaction interfaces, where relations refer to a method of data model that represents the internal structure of a database. Users can create a new relation by following the on-screen instructions, entering the relation's name, the names and data times of the attributes that make up the tuple, and key attributes.

The relation manipulation module displays the information of the relation, such as the number of ripples of the relation, the relation name, and the attribute name constituting the tuple, in tuple units, to provide direct manipulation and editing functions for a specific relation.

In particular, this module provides release browsing and editing functions by accessing the current version of the relation and inserting or displaying Korean data. When updating the contents of a particular tuple, the newly updated ripple is stored in the current version of the relation and added time information (ie start-time and end-time of the transaction time) to the previous tuple. It is responsible for registering the change history version, and also supports the function of supporting roll-back query on past data, that is, based on the past specific transaction time. By providing the roll-back function, you can find the corresponding tuples that satisfy the query.

Searching here refers to the TROLL / USE RDBMS's selection command processing function.

Time (Change History) Support The relation manipulation module receives a reference request for time and change history information of a particular tuple from the relation manipulation module and displays the contents on the screen.

This module is a sub-function of the relation manipulation module. Unlike the relation manipulation module, it provides only a reference function for time and change history information. This is because tuples registered with the change history version cannot be changed according to the non-deletion policy. Like the relation manipulation module, it is also responsible for displaying Korean data or outputting it to a specific file.

Hangul data processing functions are designed by dividing the functions of receiving the data from the Hangul function in the screen to output the data to a particular screen of Hangul SR ² DBS. One thing to consider to implement this is how to handle special characters to recognize the beginning and end of Hangul data. For example, in the CRT terminal based on the ASCII code, the characters that recognize the beginning and end of Hangul data are 〈<control> O (OxOe) ＂and 〈<control> N (OxOf), respectively, whereas Troll / In USE RDBMS, it is ＂/ 16＂ and ＂/ 17＂, so you can output Hangul data from Troll / USE RDBMS to ASCII CRT terminal or vice versa. If the key to recognize the end of Hangul data is entered in the normal state of the ASCII CRT terminal, a special collector other than "<control> N (OxOf)" is output and the end of Hangul data cannot be recognized. In the case of inputting Korean data, the data is controlled using the system command ioctl () which controls the input / output device of UNIX (registered trademark) so that the CRT terminal can output characters that can recognize the end of Korean data. After changing the contents of the structure, design a routine to return to normal state when the input is complete.

As mentioned earlier, the time-assisted database system is a screen-oriented relation browser editor that extends the maintenance of time and change history information based on the functions of the TBE. This is done through four modes, like tiles.

That is, as shown in the figure, a command mode, a screen mode, an insert mode, and a roll-back mode are provided.

The indication of each mode is perceived by the users by being displayed on the top right of the screen.

The command mode can also help you to replace a currently supported time-supported relation with a new one, to have a direct conversation with the Troll / USE RDBMS, to set up the behavior of the system, to execute and load UNIK commands, and so on. do. In command mode, input from users is initiated at the bottom of the screen.

Screen mode, also known as cursor mode, is a mode of operation when you move the cursor up, down, left, or right to browse through a series of tuples to edit the contents of the ripple on the screen. Cursors are displayed in the form of rectangles or underscores, and their movement and editing functions are the same as those of the text editor on UNIX. In addition, all of the operations for retrieving multiple tuples whether the updated tuple and the change history tuple with the time information are stored as the relation version and the change history version of each relation or the following tuples are required are all performed in this mode. In addition, the function of displaying or outputting Hangul data on the screen is performed. In order to return from this mode to the command mode, enter a specific key (＂:＂: key assigned to the transition ＂command in the＂ command mode).

Insert mode is used to insert, change, or delete the contents of the tuple currently displayed on the screen. Once in insert mode, all characters entered by the user are entered as text (until the "ESCAPE" character is entered). Process. The data insertion function is the same as the data insertion function of 'vivi' which is a text editor on UNIX, and also inserts Korean data.

The roll-back mode is a key assigned to a specific key (＂T): change history reference of the current tuple) when you want to refer to the time and change history information for the tuple currently displayed on the screen under the screen mode. By typing). This mode displays a sub-screen for a change history query (roll-back query) on the current screen, and outputs the change history information by version on the sub-screen. Since the change history information should not be arbitrarily deleted, changed or inserted, it does not support the editing function. Only the change history information of one tuple is referred to by version or the roll-back query function for the transaction time. Provide only. In addition, the function of outputting the change history contents of the specific tuple or the change history on the screen as a specific file and the display of Korean data over time is also performed in this mode.

The type of interface provided to the users by the SR ² DBS of the present invention is composed of three types of screens as shown in FIGS. 3 to 5. In other words, the configuration of the "Screen Division" of the time-supporting database system (SR ² DBS) shows that when the SR ² DBS is called and executed, the initial screen division (see Figure 3) and For example, the relation definition screen pane provided for defining a new relation in the database system (see Figure 4), and the status change history provided for performing maintenance activities on a newly defined relation or a relation already defined. Version relation operation screen section division (refer to FIG. 5).

As an example of application of information management activities using the SR ² DBS, the maintenance case of the functional test and certification test of the TDX-10 (trademark) system of the electronic switchgear will be described.

In performing system level tests on the TDX-10 system, the history and time information of many system test results changes are used as important information to be maintained.

One of the main relations that make up the database schema (which is the basic form of a relational database) for maintaining system test results is "Test-rslt", which is a tuple with the following characteristics (attributes): 6 shows an example of providing a query for change history and time information of the relation.

-Test item number (attribute name: item-id, data type: Integer 100..1000)

System version under test (attribute name: year, data type: Integer 1454..2000)

Test attempts (attribute name: Suthor, data type: String)

-Test agent name (attribute name: publisher, data type: String)

Test results (attribute name: Title, data type: String)

Test date (attribute name: tst-date, data type: Integer 890101..911231)

-Test content and problem (attribute name: tst-comment, data type: String)

7 is a flow chart of a conventional TBE system.

Since the Troll / USE RDBMS uses the UNIX file system to configure the database, TBE starts with a database directory as an input parameter (1). First, check whether or not the input database directory exists (2). At this time, if there is no parameter or a directory name that does not exist, an error message about this is displayed on the screen, and the service is terminated (3). If a database directory as an input parameter exists, an initial screen of the TBE is displayed (4). Open the database directory for database operations (5). After placing the cursor in the command entry area at the bottom of the screen, wait for the database-level command to be entered (6, 7). If the 'relation xxx' command is selected for editing or updating a particular relation (8), after importing the relation xxx, the relation operation screen is displayed and the first tuple value of the relation xxx is displayed. To display (10). From this point on, the mode transitions from the command mode to the screen mode (11). Wait for a command from the user to perform a database manipulation command such as edit or update, retrieve, etc. for the current relation xxx (12). When the user updates the current tuple, if a predetermined screen editing command such as '1', 's', or 'R' is inputted, the execution mode is changed to insert mode (14), and the tuple on the screen is changed. The controller 16 updates the specific tuple value of relation xxx by receiving a predetermined update command 〈w〉. When a command (＂:＂) is input to the command mode (15), the execution mode is changed to the command mode, and the database level command is prepared again (6). At this time, if the user wants to terminate the service and inputs "quit", the relation xxx currently imported (IMPORT) is exported, the database directory is closed (9) and terminated.

8a to 8 is a flow chart according to the present invention, Figure 8a is the overall flow chart of the present invention, Figure 8b is a flow chart of the database definition module of the present invention, Figure 8e is a Hangul data insertion and processing module of the present invention 8F is a flowchart showing a Hangul data screen display routine of the present invention, and FIG. 8C is a flowchart showing a processing module of a change history query according to the present invention, and FIG. 8D is a flowchart showing a tuple change history screen operating module according to the present invention.

Now, looking at the implementation process of the present invention, since the Troll / USE RDBMS configures the database using a UNIX file system, the SR ² DBS of the present invention also receives the database directory as an input parameter to start execution (1 ). First, check whether or not the input database directory exists (2). In this case, if there is no input parameter or a directory name that does not exist, an error message about this is displayed on the screen, and the service is terminated (3). If a database directory as an input parameter exists, the SR ² DBS initial screen is displayed ( 4) Open the database directory for database operations (5), position the cursor in the command entry area at the bottom of the screen, and then wait for database-level commands to be entered (6). Here, when the command "ddl yyy", which defines a new relation, is input (7), the "database module" (see also 8b) is called (8).

As shown in FIG. 8B, the database definition module receives the relation name as an input parameter (20), checks the input relation name (21), and if the relation name already exists, After outputting the error message, the command returns to the command input step 6 (23). If the input parameter is omitted, the screen for receiving the relation name is displayed to receive the relation name (22). When the entry of the relation name is complete, the relation definition screen is displayed (24) and ready to receive database definition commands (25). Using the database definition command, users write on the screen the attribute names, types, and constraints that make up a relation, and enter the completion relation definition command (#w) when the creation is complete.

When the command ＂w＂ is input (26), information about relations defined on the screen is taken (27), and the status relations (yyy) and change history relations (yyy-TpR) correspond to the relation definitions of the Troll / USE RDBMS. (28) and request a relation definition from the Troll / USE RDBMS (29). If the response from the Troll / USE RDBMS is an error (30), the corresponding error message is printed to the screen (32), and then returned to the database definition command entry step so that it can be modified on the screen (26). If not, you are asked if you want to define another relation (31), if you wish, go back to entering the relation name step 20 to define another relation, if you are finished clearing the screen (33), and then go to the database of Figure 8a. It is returned to the command input step 6 of the dimension.

If the "relation xxx" command is selected for editing or updating a specific relation (7), after importing the current relation xxx and the change history relation xxx-TpR (9), the relation operation screen is displayed. (11), display the first tuple of the current relation xxx (12), and then the run mode transitions from command mode to screen mode (13), and then edits or updates the current relation xxx Wait for a command from the user to perform a database manipulation command such as search, search, etc. (14). In this case, when the user updates the current tuple, if a predetermined screen editing command (i, i, n, r) is input (15), the execution mode is changed to the insert mode, and Korean data is inserted. And a processing module (see also FIG. 8E) is called to update the corresponding tuple on the screen (16).

Referring to FIG. 8E, the process of performing the Hangul data insertion and processing module will be described. First, in order to recognize the end of Hangul character (OxOf), the value of the data structure of the I / O device is changed. (35, 36, 37), then wait for input of the specified character (38). In the embodiment of the present invention, ＂OxOe＂ as the Hangul start character, ＂OxOf＂ as the Hangul end character, and ＂/ b＂ as the cursor backward character, respectively, and the characters other than the specified character are recognized as the corresponding character display. It was. Therefore, the content of the display varies depending on the character input (39). When the return character is inputted, the changed data structure value of the input / output device is returned to its original state and the execution of this module is terminated (44) (returned to the relation operation instruction input step 14 of FIG. 8A). Of course, when inserting Hangul data, press the Hangul key on the keyboard to recognize the beginning and end of Hangul.

When the update on the screen is completed, the update command ＂w＂ is input to update a specific tuple value of the current relation xxx. At this time, time information (transaction time) is added to the tuple value stored in the current state relation and stored in the change history relation xxx-TpR, and the newly updated tuple value on the screen is stored in the current state relation.

When a predetermined command “M” is input to perform a query regarding the change history in the screen mode, the change history query processing module (see also 8c) is called (17).

Referring to FIG. 8C, the execution history of the change history query processing module will be described. The execution history transitions from the screen mode to the roll-back mode, and the change history query screen receives a transaction time at the bottom of the current relation screen. Display (55) and import the change history relation (56). Upon receiving the transaction time from the user (57), a selection statement of the Troll / USE RDBMS is constructed and requested to select the change history tuples that satisfy the transaction time input from the status and change history relation (58). ). After checking the response from the Troll / USE RDBMS (59), if it is an error, the corresponding error message is output (62), and then jumps to the transaction time input step 57 again to correct the transaction time. Otherwise, focus by transaction time (FOCUS) is configured to display the first change history tuple on the screen (61), transition the execution mode to screen mode, and then re Returning to the relation manipulation instruction input step 14 of FIG. 8A, the controller prepares to receive a reference instruction for the change history focus FOCUS.

In screen mode (15), if a specified command (＂T＂) is entered (15) to refer to the change history for a tuple that is screen-displayed in the status relation, the tuple change history screen manipulation module (see also FIG. 8d) is called. (18).

Referring to FIG. 8D, the process of performing the tuple change history screen operation module will be described. After changing the execution mode from the screen mode to the roll-back mode, the tuple change history screen is displayed at the bottom right of the screen (63). The key attribute value on the relation screen is taken (64), the change history for the corresponding tuple is selected from the change history relation xxx-TpR (65), and the first change history is output on the tuple change history screen (66). And waits for input of a reference instruction for a tuple change history (67). The tuple change history reference instruction is a change history management-related instruction. In the embodiment of the present invention, 'n' is a search for the next change history of the current tuple, and 'p' Search for the change history immediately before the current tuple, ＂q＂ terminates the change history reference for the current tuple, ＂c 변경 copies the change history for the current tuple to the Troll's input file, and ss for the current tuple Copy the current version of the revision history to the same file as the screen form, and if 'f' refers to an attribute on the next page, 'b' refers to an attribute on the previous page, and 't' is the Each is specified as a transaction time output command.

Therefore, when the command n is input (68), the next tuple change history is displayed (69), when n is input (68), the previous tuple change history is displayed, and when n is input (68). ) Outputs the tuple change history to a specific file, when qq is input (70), transitions the execution mode to the screen mode, clears the tuple change history sprin, and then enters the relation operation command input step 14 of FIG. 8a. It returns to and receives the following screen command.

However, at this time, if ＂: 받 is input (15) (see also FIG. 8a), the execution mode is changed to the command mode (19), and ready to receive a database-level command (6). However, if the user wants to terminate the service and inputs `` quit '' (7), the current relation IM xxx and the change history relation xxx-TpR are exported, and the database directory is closed (close). (10) End the execution.

Therefore, the present invention as described above can maintain the history of the change of information over time, such as querying the past information that could not be provided in the conventional database system or time analysis such as trend analysis according to a series of change history It provides the side information processing function, and implements the processing and management function of Hangul data, and has the effect of providing the function as a database system perfectly.

Claims (6)

In the database system that maintains the history of change of information over time, it supports information processing in the aspect of time such as querying the past information or trend analysis according to a series of change history, and processing and management of Korean data. Receiving a database directory as an input parameter (1) checking whether it exists (2) outputting an initial screen if present (4) opening the database directory (5); A second step of receiving a database-wide command (6); In the second step, when a command for editing or updating a specific relation is input (7), the relation relation and the change history relation are imported (IMPORT), and the relation manipulation screen is output (11) the relation relation A third step of displaying (12) the first tuple value of and transitioning the execution mode from the command mode to the screen mode (13); A fourth step of receiving (14) an operation screen and a screen mode command from a user for manipulating a database for a current relation; A fifth step of performing a Hangul data insertion and processing module (16), a change history query processing module (17), or a tuple change history screen manipulation module (18) according to the command input through the fourth step; A sixth step of transitioning the execution mode to the command mode (19) if the command inputted through the fourth step is a command for transitioning to a command mode (19) so as to receive a database-level command again; And (7) when the command for terminating the service is input through the second step (7) exporting the current relation IMPORT and the change history relation, and closing the database directory (10). An implementation method of the SR ² DBS time support database system, characterized in that performed by including seven steps. The method of claim 1, wherein when a command for defining a new relation is input (7) through the second step, the database definition module is called and executed (8). Implementation method of the SR ² DBS time support database system characterized in that the step further comprises the eighth step. The method of claim 2, wherein the eighth step comprises: a ninth step of receiving a relation name as an input parameter to check an input relation name and displaying a relation definition screen; A tenth step of receiving a database definition command; An eleventh step of creating, on the screen, an attribute name, a type, a constraint, etc. constituting a relation from users until a relation definition completion command is input; It takes information about relations defined on the screen and configures the relation relations and change history relations according to the relation definition division of the relational database (Troll / USE RDBMS), and creates relation relations in the relational database (Troll / USE RDBMS). Requesting a twelfth step; If the response from the relational database (Troll / USE RDBMS) is an error, outputting the corresponding error message on the screen, and repeating the step of inputting the database definition command of step 10 to modify it on the screen; A fourteenth step of asking whether to define another relation and repeating from the inputting of the relation name of the ninth step to define another relation if desired; And then it clears the screen, how the implementation of the first and second SR ² DBS time, characterized in that is carried out by the step 15. Support Database system returns to the command step data base dimension of the step (6). The change history query processing module 17 according to claim 1 or 2, wherein the change history query processing module 17 of the fifth step changes the execution mode from the screen mode to the roll-back mode and receives a transaction time in the current state relation. Displaying a history query screen and importing a change history relation; A 17th step of constructing and requesting a syntax for selecting change history tuples that satisfy the transaction time input from the status and change history relations when the transaction time is input from the user; An eighteenth step of checking the response from the relational database (Troll / USE RDBMS) and repeating the operation from the seventeenth step to correct the transaction time by outputting the corresponding error message if an error is detected; A nineteenth step of configuring focus by transaction time and displaying the first change history tuple on the screen; And then transits the run mode to the screen mode of the SR ² DBS time support database system, characterized in that is carried out is carried out by step 20, which returns back to the step 14, 4 for the correlation operation command input Implementation method. According to claim 1 or 2, the tuple change history screen operation module 18 of the fifth step, after changing the execution mode from the screen mode to the roll-back mode, the tuple change history screen at the bottom right of the screen In the step 22 tuple change history, which takes the key attrib value on the status relation screen to be displayed, selects the change history for the corresponding tuple from the change history relation, and outputs the first change history on the tuple change history screen. A twenty-third step of inputting a reference command; Displaying a corresponding change history on a screen according to a tuple change history reference command input in step 23; SR ²⁵ DBS time, characterized in that the execution mode is changed to the screen mode and the tuple change history screen is cleared, and then the relation manipulation command is performed by step 25 again returning to the fourth step 14 for input. How to implement a supported database system. The method according to claim 1 or 2, wherein the Hangul data insertion and processing module (16) of the fifth step comprises: changing a data structure value of the input / output device to recognize the Hangul end character; Inputting a character assigned to the beginning of the Hangul, the end of the Hangul and the backward cursor, and displaying the corresponding character or moving the cursor position according to the input character; And is a data structure, the value of the input-output device and returns to its original position, which changes, re-correlation operation claim 4 SR ² DBS time, characterized in that is carried out by the step 27 which returns to step 14 for the command Support Database How to implement the system.

Images