JP2006004145A - Program code generator - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a program code generator for efficiently developing an application program by using information describing the schema of a database. <P>SOLUTION: The program code generator involves: acquiring the schema information and retaining the pattern of a database operation code; and completing a program code based on the acquired schema information and the specimen of the retained database operation code. In addition, the program code is completed by using the corresponding relation between a data type used in the database and a data type used in a program language, or by associating the naming rule of the column name of the database with the data type used in the program language. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a program generation apparatus that generates a program code, and more particularly to a program generation apparatus that generates a program code for accessing a database.

  In order to operate a database, it is necessary to be familiar with a database operation language such as SQL (generally called “Structured Query Language”), but it is difficult for general people to learn. Accompany. Therefore, a technique is known in which a graphical user interface (GUI) is generated from information indicating a database structure (that is, a database schema), and the database is operated through the GUI (see, for example, Patent Document 1).

  On the other hand, some tools for creating a database structure interactively are known (for example, see Non-Patent Document 1). These tools convert schema information into CSV (Comma Separated Values) format or XML (eXtended Markup). (Language) format.

  In the development of application programs that access a database, humans individually write programs.

FIG. 1 illustrates the operation of an application program that accesses a database. The application program 101 that accesses the database communicates with the relational database management server 102 using a technique such as interprocess communication. Specifically, when the application program 101 transmits an SQL statement to the relational database management server 102, the relational database management server 102 performs operations such as referring to or changing the database 103 according to the SQL statement, and sends the result to the application program. To return. The types of SQL statements include a SELECT statement for searching, an INSERT statement for adding data, an UPDATE statement for updating, and a DELETE statement for deleting, and a statement for schema operation that changes the structure of the database. There is a statement to refer to the structure of the schema. These sentences are embedded as character strings in the application program and are appropriately executed.
JP 2004-126680 A "Product Information> AllFusion ERwin Data Modeler", [online], 2004, [Search Date: 2004.6.007], Internet <http://www.jsys-products.com/product/al_erwin/>

  As described above, in developing an application program that accesses a database, since a person writes a program individually, there is a problem that the reliability of the program does not increase even though it takes time and labor. When the database schema changes, the corresponding application program must be individually modified, which takes more effort and time. In addition, the data types that can be used in the database and the data types that can be used in the programming language are not the same. In developing an application program, it is necessary to always be aware of the difference between the data types, and it is easy to induce bugs in this respect.

  Since a technique for operating a database through a GUI places importance on a dialogue function with a human, the above problem cannot be solved by diverting this technique.

  In addition, files that describe the schema in CSV format or XML format generated by a tool that creates the database structure in an interactive format are in text format and are easy to use. There has been little progress in improving the reliability of application program development.

  Therefore, an object of the present invention is to provide a program code generation apparatus and a program code generation method for efficiently developing an application program using a file describing a schema.

  In order to solve such an object, in the present invention, schema information is acquired and a database operation code template is stored, and based on the acquired schema information and the stored database operation code template. A program generation device for completing the program code is provided. Thereby, the development process of the application program for accessing the database is automated, the productivity of the application program is improved, and the reliability is also improved.

  In addition, the program code generation device has a correspondence between the data type used in the database and the data type used in the program language, and the program code is generated using the correspondence. May be. This alleviates programmers' awareness of data type differences and reduces bugs.

  Further, the naming rules for database column names may be associated with data types used in the programming language. Thereby, an application program corresponding to the role of the column can be developed.

  In addition, by generating program code corresponding to so-called logical deletion that treats row data including the column as deleted depending on whether the value of a specific column has a specific value, complicated operations such as logical deletion are also automatically performed. Can be programmed and productivity and reliability can be improved.

  The generated program code may be written in a program language having an inheritance mechanism. This makes it easy for a human to modify the generated program code using the inheritance mechanism.

  Also, in the generated program code, character strings of table names and / or column names are defined as variables and constants, and those variables and constants can be referenced in the program part that inherits the generated program code. Also good. As a result, it is not necessary to modify the program code modified by humans even if the table name or column name is changed, or the program code modified by humans must be modified by changing the table name or column name. Can be easily known.

  According to the present invention, the development process of an application program that accesses a database is automated, so that the productivity of the application program is improved and the reliability is improved.

  Hereinafter, the best mode for carrying out the present invention will be described with reference to the drawings as an embodiment. Note that the present invention is not limited to these embodiments, and can be implemented in various modes without departing from the spirit of the present invention.

  (Embodiment 1: Claims 1, 7, and 8 will be mainly described)

  As Embodiment 1 of the present invention, schema information is acquired, a database operation code template is stored, and a program code is completed based on the acquired schema information and the stored database operation code template A program generation apparatus and a program generation method to be executed will be described.

(Embodiment 1: Configuration)
The program code generation device according to the present embodiment generates program code for accessing a database. “Access” refers to referring to or changing the contents of a database by communicating with a server that manages the database or by operating a library for managing the database.

  FIG. 2 illustrates a functional block diagram of the program code generation device according to the present embodiment. The program code generation device 200 includes a schema information acquisition unit 201, an operation code template holding unit 202, and a program code completion unit 203.

  In addition, not only this embodiment but each part which is a component of the program code generation apparatus of this invention can be comprised by either of hardware, software, both of hardware and software (program). Therefore, the program code generation device can be realized using a computer. For example, it can be realized using hardware including a CPU, a memory, a bus, an interface, a peripheral device, and the like, and software that can be executed on these hardware. It is also possible to record such software (program) on a medium.

  The “schema information acquisition unit” 201 acquires schema information. “Schema information” is information describing the schema of the database. That is, the information represents the structure of the database. When a relational database is used as the database, the database is configured by a table having one or a plurality of columns. Therefore, in this case, information representing the structure of the table is schema information.

  FIG. 3 illustrates a relational database table. In FIG. 3, TABLE_A and TABLE_B are shown as a table. TABLE_A consists of columns named ID, A_CODE, A_FLAG, A_DATE, A_TIMESTAMP, and DELETE_TIMESTAMP. TABLE_B is composed of columns named ID, TABLE_A_ID, and B_NAME.

  FIG. 4 illustrates schema information describing the structure of the database illustrated in FIG. In FIG. 4, one line describes the information of the table columns. Since the first line in FIG. 4 starts with “#”, it is treated as a comment line, and is expressed so that it is easy for humans to understand what each part separated by a comma represents. Yes. That is, the first part of each row represents the name of the table, the second part represents the name of the column, and the data type of the column, whether the column is a primary key, and a null value can be stored in order. Indicates whether it can be done or a comment on the column. The “primary key” means a column for uniquely specifying a row stored in the table. For example, when storing employee information in a table, each employee can be identified by an employee number given differently to each employee, so the column storing the employee number is the primary key. The “null value” is a special value indicating that the value of the column is not specified or unknown.

  For example, the schema information acquisition unit 201 reads the schema information illustrated in FIG. 4 and acquires the schema information. If the program code generation device 200 is implemented using a computer, a data structure representing the schema information is constructed in the main memory as a result of acquiring the schema information.

  FIG. 5 illustrates a data structure constructed in the main memory, which represents schema information. The node 501 is created corresponding to a table called TABLE_A, and the nodes 503, 504, 505, etc. are nodes created corresponding to the columns of TABLE_A, and the names, data types, primary keys of the columns. Whether or not a null value can be stored is stored. A node 502 corresponding to the table TABLE_B is similarly created.

  The “operation code template holding unit” 202 holds a database operation code template corresponding to the program code to be generated. The “database operation code template” means a prototype that is a template of a program referred to for generating a program code. For example, in the C language in which a program is constructed in units of functions, it means a function template. Further, “corresponding to the program code to be generated” means that each generated program code is based on a database operation code template. The operation code template holding unit 202 holds, for example, a database operation code template in the hard disk area of the computer. Alternatively, a database operation code template may be held as program data for realizing the program code generation device with a computer. When a database operation code template is stored as program data, the database operation code template does not need to be grouped as illustrated in FIG. 6 and the like, and is stored in an appropriately divided format. It may be.

  6 to 9 illustrate database operation code templates. 6 is a template corresponding to the SELECT statement of the SQL statement, FIG. 7 is a template corresponding to the INSERT statement, FIG. 8 is a template corresponding to the UPDATE statement, and FIG. 9 corresponds to the DELETE statement. It is a template to do.

  The database operation code template illustrated in FIG. 6 will be described on behalf of FIGS. The first line represents the function name. Note that $ and the alphabet following it mean variables for generating a program code from a database operation code template. For example, $ T represents a table name. Therefore, $ T can take the value TABLE_A. $ C represents a column name to be operated. For example, $ C can take the value A_CODE. Therefore, select_ $ T_ $ C can be replaced with select_TABLE_A_A_CODE. The third to seventh lines are working variables used in the function.

  “EXEC SQL DECLARE SECTION” on the third line indicates an instruction to the SQL preprocessor for embedding an SQL statement in a program and enabling compiling. Thus, a line beginning with “EXEC SQL” represents an instruction to the SQL preprocessor. The instruction on the third line means the start of the declaration of a variable to be recognized by the SQL preprocessor. The instruction on the seventh line means the end of the declaration of the variable that should be recognized by the SQL preprocessor. Therefore, variables such as spv, sov, and stmt are recognized by the SQL preprocessor.

  The ninth line substitutes the value of the argument passed to the function, that is, the value of the variable pv into spv. The 10th line substitutes a character string representing the SQL statement to be executed into a variable called stmt. Note that the variable $ P appearing in the character string here represents the primary key. “?” Is a symbol indicating a parameter included in an SQL sentence expressed by a character string.

  The twelfth line indicates that the character string assigned to the variable stmt is prepared to be executed as an SQL statement, and the prepared SQL statement is named s_stmt. “Preparation for execution” means, for example, managing whether or not there is a syntax error in the SQL statement represented by the character string, and the position of the parameter indicated by “?”. Note that “:” appearing in: stmt means that the next stmt is a variable recognized by the SQL preprocessor.

  The 13th line means that the value of spv is assigned to the parameter of the SQL statement named s_stmt and executed, and the result is assigned to a variable called sov. As a result, the row in which the value of the primary key is assigned to spv is searched, and the value of the column named $ C in that row is assigned to sov.

  In the 14th line, the information of the SQL statement with the name s_stmt is discarded. For example, when a memory area is dynamically secured in the 12th or 13th line, the memory area is discarded.

  In the 16th line, the value of sov is returned.

  Therefore, the database operation code model illustrated in FIG. 6 searches for a row having a value given as an argument as a primary key column value in a row of a table called $ T, and a column called $ C in that row. Represents a function that obtains the value stored in.

  Hereinafter, FIG. 7 shows a function for inserting a row having a value given as an argument as a primary key into a table $ T. FIG. 8 shows a function for changing the value of the column $ C in the row where the value of the primary key column is the first argument in the row of the table $ T to the value of the second argument. FIG. 9 shows a function for deleting a row having a value given as an argument as a primary key from the table $ T.

  The database operation code model illustrated in FIGS. 6 to 9 is merely an example, and various modifications can be made. For example, in FIG. 6, the search is performed by designating the value of one column, but it is easy for those skilled in the art to modify the database operation code template so as to perform the search by designating the values of a plurality of columns. It is also possible to specify an arbitrary column without reason to limit the column for specifying a value to the primary key. In FIG. 7, it is assumed that the search result is one, but it is possible to deal with cases where the search result is zero or plural by describing more complicatedly. Similarly, in FIG. 7 and FIG. 8, values in a plurality of columns can be inserted or changed. Further, in FIG. 9, it is possible to delete a row using a column value other than the primary key.

6 to 9, the case of a language that requires an SQL preprocessor has been described. For example, by using an object-oriented language such as a Java (registered trademark) language, an SQL preprocessor is not required. Can do. For example, if conn is a variable that stores an object representing interprocess communication with a relational database management system,
s_stmt = conn.PrepareStatement ("select $ C where $ T where $ P =?");
Prepare the SQL statement to be executed by the method call PrepareStatement, assign the information indicating the result to the variable s_stmt,
s_stmt.setValue (1, pv);
Sets the value of the variable pv to the first parameter,
s_stmt.executeQuery ();
Thus, the prepared SQL statement in which the parameter value is set can be executed as an inquiry statement.

  The “program code completion unit” 203 completes the program code based on the schema information acquired by the schema information acquisition unit 201 and the database operation code template stored in the operation code template storage unit 202. For example, values are appropriately set in $ T, $ C, $ T, and $ P, and $ T, $ C, $ T, and $ appearing in the database operation code template are replaced to generate a program code.

  FIG. 10 illustrates a program code obtained by substituting the database operation code template illustrated in FIG. 6 by setting TABLE_A in $ T, A_CODE in $ C, and ID in $ P.

  As described above, when searching by specifying values of a plurality of columns, a function having arguments of only the values of the plurality of columns may be defined. Since the code becomes complicated, for example, a program code that uses a function of a programming language that combines a plurality of values into one may be completed.

  FIG. 11 shows an example of a program code that has a function of combining a plurality of values of a program language into one and inserts values of a plurality of columns. FIG. 11A shows an example in which a data structure representing a row of TABLE_A is defined as a structure (or class) in C language, C ++ language, or the like. FIG. 11B illustrates an example of a function that inserts data into TABLE_A using the structure defined in FIG.

(Embodiment 1: Flow of processing)
FIG. 12 illustrates a flowchart of processing of the program code generation device. Each time the program code generation apparatus acquires schema information, the program code generation apparatus executes the flowchart of FIG. In step S1201, the schema information acquisition unit 201 acquires schema information (schema information acquisition step). In step S1202, it is determined whether or not the generation of the program code should be terminated. This determination is made by, for example, determining whether or not all database operation code templates have been used, for example, using all combinations of variable values such as $ T and $ C, or using predetermined combinations. It is based on whether or not. If it is determined that the generation should not be terminated, the process proceeds to step S1203. In step S1203, the database operation code template is read from the operation code template holding unit 202 (operation code template reading step). If there are a plurality of database operation code templates, they are read in a predetermined order, or if an instruction is given by the operator of the program code generation apparatus using a mouse, a keyboard, etc., they are read according to the instructions. In step S1204, the program code completion unit 203 completes the program code based on the acquired schema information and the read database operation code template. The completed program code is recorded, for example, as a file or output to a printer. When the process of step S1204 ends, the process returns to step S1202.

(Embodiment 1: Main effects)
According to the present embodiment, the development process of the application program that accesses the database is automated, the productivity of the application program is improved, and the reliability is also improved.

  (Embodiment 2: Claim 2 will be mainly described)

  Embodiment 2 of the present invention is a program code generation device according to Embodiment 1, which has a correspondence between a data type used in a database and a data type used in a program language, and uses the correspondence. A program code generation device that generates a program code will be described. The first embodiment is mainly suitable for programming languages that do not require much consideration for data types such as Perl language and Lisp language, but the first embodiment is a language that requires consideration for data types such as C language. This is a mainly suitable embodiment.

(Embodiment 2: Configuration)
FIG. 13 illustrates a functional block diagram of the program code generation device according to the present embodiment. The program code generation device 1300 includes a schema information acquisition unit 201, an operation code template storage unit 202, a program code completion unit 203, and a data type correspondence table storage unit 1301, and the program code completion unit 203 performs data type adjustment. Means 1302 are provided.

  Therefore, in the program code generation device according to the present embodiment, the program code generation device according to the first embodiment further includes the data type correspondence table holding unit 1301, and the program code completion unit 203 includes the data type adjustment unit 1302. It has a configuration.

  A “data type correspondence table holding unit” 1301 holds a data type correspondence table. The “data type correspondence table” is a table associating column data type information with variable data type information. Here, the “column data type information” is information indicating the data type of the column used in the description of the schema information, and the “variable data type information” is the variable used in the program code. This is information indicating the data type.

  FIG. 14 shows an example of the data type correspondence table. The left column illustrated in FIG. 14 stores column data type information, and the right column stores variable data type information. Note that the column data type information of the table illustrated in FIG. 14 is information obtained by adding information corresponding to nullable values such as not null and with null to the data type illustrated in FIG. . This is because null values require special handling in programming languages. For example, a character string type that does not take a null value can be an array of character strings in the programming language, but if it takes a null value, it is a pointer type that stores the address of the character string in the programming language, for example. , NULL must be represented by the value NULL. Similarly, when a value that is an integer type (Integer) and does not take a null value is stored in a certain column, an int type that can be normally used in a programming language can be used, but an integer type that can take a null value is used. , And a special type called Integrer (for example, implemented as a class in the C ++ language or Java (registered trademark) language).

  The “data type adjusting unit” 1302 uses the data type correspondence table held in the data type correspondence table holding unit 1301 to complete the program code. For example, information indicating the data type of a variable used in the program code is stored in a node corresponding to each column of the data structure (see FIG. 5) constructed corresponding to the schema information acquired by the schema information acquisition unit 201. In addition, the program code is completed while referring to the added information.

  FIG. 15 shows an example in which information indicating the data type of a variable used in the program code is added to a node corresponding to each column of the data structure illustrated in FIG. 5 by a data type adjusting unit or the like. For example, since the ID column in the table TABLE_A is an INTEGER type and does not take a null value, a node 1501 representing int is added, and the ID_CHAR column type is CHAR (20), that is, a length of 20 bytes. Therefore, a node 1502 representing char [20] is added. For example, the data type adjusting means refers to the data structure to which the information indicating the data type of the variable is added in this way, and appropriately completes the program code.

  FIG. 16 illustrates a database operation code template to be used in the present embodiment. The difference from the first embodiment is that the variable has a data type, and $ Pt and $ Ct represent the data type of the variable associated with the data type of the column by the data type correspondence table. For example, int is set in $ Pt and char [20] is set in $ Ct. As a result, the program code illustrated in FIG. 17 is completed.

(Embodiment 2: Main effects)
According to this embodiment, it is possible to reduce the programmer's awareness of the difference in data type, and to produce effects such as a reduction in bugs.

  (Third embodiment: Claim 3 will be mainly described.)

  As a third embodiment of the present invention, a program code generation apparatus according to the first or second embodiment, wherein the program code generation apparatus generates a program code by associating a column name naming rule with a data type used in a program language. explain.

(Embodiment 3: Configuration)
FIG. 18 illustrates a functional block diagram of the program code generation device according to this embodiment. The program code generation device 1800 includes a schema information acquisition unit 201, an operation code template storage unit 202, a program code completion unit 203, and a second data type correspondence table storage unit 1801. Two data type adjusting means 1802 is provided. In addition, the program code generation device according to the present embodiment may include a data type correspondence table holding unit 1301, and the program code completion unit 203 may include a data type adjustment unit. For example, the configuration illustrated by the functional blocks illustrated in FIG. 19 may be used.

  Therefore, in the program code generation device according to the present embodiment, the program code generation device of the first or second embodiment has the second data type correspondence table holding unit 1801, and the program code completion unit 203 is the second data type adjustment unit. 1802 is provided.

  The “second data type correspondence table holding unit” 1801 holds the second data type correspondence table. The “second data type correspondence table” is a table associating column naming rule information with variable data type information. Here, the “column naming rule information” is information indicating a naming rule for column names appearing in the description of the schema information. For example, the name of a column is expressed using a wild card or a regular expression to indicate a naming rule. “Variable data type information” is the same as in the second embodiment, and is information indicating the data type of a variable used in the program code.

  FIG. 20 shows an example of the second data type correspondence table. The left column illustrated in FIG. 20 stores column naming rule information, and the right column stores variable data type information. Note that the column naming rule information in the table illustrated in FIG. 20 represents a naming rule using a wild card. For example, “* _FLAG” represents a naming rule in which a character string having an arbitrary length is used as a portion of * and _FLAG is subsequently connected. In general, a column having _FLAG in the latter half of the name often indicates true or false. Therefore, in a programming language, for example, a type called unsigned int is often used. Alternatively, if there is, for example, a Boolean type representing a true / false value, the type may be associated with the type. In addition, since a column having _TIMESTAMP in the latter half of the name almost always represents time, a type such as time_t representing time in the programming language may be associated.

  When the program code generation apparatus according to the present embodiment has the data type correspondence table holding unit 1301 and the program code completion unit 203 has the data type adjustment unit 1302, the data type and the data type by the data type adjustment unit 1302 are changed. The data type by the two data type adjusting means 1802 may be different. In that case, the fact is displayed on the screen and the operator of the program code generating device selects the data type, or The data type by the two data type adjusting means 1802 may be prioritized. In addition, the specification of which data type should be prioritized may be included in the comment or the like of the schema information illustrated in FIG.

(Embodiment 3: Main effects)
According to the present embodiment, an application program corresponding to the role of the column is generated, and an application program that is easy for a programmer to read is generated. As a result, an application program that is easy for a programmer to correct is created.

  (Embodiment 4: Claim 4 will be mainly described)

  As a fourth embodiment of the present invention, a program code generation device according to any one of the first to third embodiments, which generates a program code corresponding to logical deletion, will be described.

(Embodiment 4: Configuration)
FIG. 21 illustrates a functional block diagram of the program code generation device according to the present embodiment. The program code generation device 2100 includes a schema information acquisition unit 201, an operation code template storage unit 202, and a program code completion unit 203, and the schema information acquisition unit 201 includes a logical deletion sequence information acquisition unit 2101. FIG. 21 only shows the main part of the program code generation device 2100 according to the present embodiment. The program code generation device 2100 further includes a data type correspondence table holding unit 1301 and a second data type correspondence table holding. The program code completion unit 203 may include a data type adjustment unit 1302 and a second data type adjustment unit 1802, respectively.

  Therefore, the program code generation device according to the present embodiment has a configuration in which the schema information acquisition unit 201 of the program code generation device according to any one of the first to third embodiments includes the logical deletion column information acquisition unit 2101. ing.

  The “logical deletion column information acquisition unit” 2101 acquires logical deletion column information. “Logical deletion column information” is information for specifying a column used for logical deletion. “Logical deletion” is a method for determining whether row data including the value of the column depends on whether the specific column has a specific value. A method that treats as deleted (for example, see <http://www.kei-ha.co.jp/magdb/magdb0001.html>). For example, if a value stored in a column called DELETE_TIMESTAMP in a row of TABLE_A illustrated in FIG. 3 takes a value other than a null value, the row is treated as deleted.

  FIG. 22 exemplifies logical deletion column information expressed in XML. In this example, if a TIMESTAMP column of DELETE_TIMESTAMP is NULL, that is, a null value, it is regarded as not deleted (that is, treated as a valid row), and if it is not a null value, it is regarded as deleted. Yes.

  The logical deletion column information acquisition unit 2101 acquires the logical deletion column information as illustrated in FIG. 22 by reading the logical deletion column information as illustrated in FIG. 22, for example, a flag or the like for a node corresponding to the column of the data structure illustrated in FIG. To indicate that the column is used for logical deletion.

  When the logical deletion method is used as described above, insertion and deletion into the table are complicated, and cannot be realized simply by executing the SQL statement corresponding to the operation. FIG. 23 exemplifies a flow chart of processing at the time of insertion. For example, the program code completion unit 203 completes the program code based on a database operation code template that represents the processing of FIG. In step S2301, first, an insert statement is executed. In step S2302, it is determined whether an error has occurred. If no error occurs, the insertion process is terminated. If an error occurs, the process proceeds to step S2303 to determine whether a unique constraint error has occurred. “Unique constraint error” means an error caused by the primary key value being duplicated with the primary key column value of a row that has already been inserted into the table. If it is determined that a unique constraint error has occurred, the process proceeds to step S2304; otherwise, the process proceeds to step S2305.

  In step S2304, the value of the column for logical deletion is updated to the value specified as valid. In the case of FIG. 22, it is updated to a null value. In step S2305, since an error other than a unique constraint error has occurred, appropriate error processing is performed.

  Whether or not an error has occurred in step S2302 can be detected, for example, by examining the value of a variable called squirer. Alternatively, in the case of a Java (registered trademark) language or the like, detection can be performed based on whether an exception has occurred, and can be detected by a try-catch clause or the like. Whether the error that has occurred is a unique constraint error or not can be determined by analyzing the value of sqlerror, or by analyzing the type and value of the exception that has occurred.

  In the case of deletion, instead of executing the DELETE statement of the SQL statement, as shown in FIG. 24, the value of the logical deletion column is updated to the value designated as deleted. In the case of FIG. 22, the value is changed to a value other than the null value. As a specific example, the value of DELETE_TIMESTAMP is acquired to update the current time value to the current time value.

  Similarly, in the case of a search, a search condition is added with and so that a row in which the value of the logical deletion column is a value designated as valid is searched. In the case of FIG. 22, the condition “DELETE_TIMESTAMP isNull” is added as one limb.

(Embodiment 4: Main effects)
According to this embodiment, a complicated operation of logical deletion can be automatically programmed, and productivity and reliability can be improved.

  (Embodiment 5: Claim 5 will be mainly described)

  As a fifth embodiment of the present invention, a program code generation device that generates a program code described in a program language having an inheritance mechanism will be described.

(Embodiment 5: Configuration)
The program code generation device according to the present embodiment is the program code generation device according to any one of the first to fourth embodiments, and the program code completed by the program code completion unit is a program language having an inheritance mechanism. Is a program code generation device described in the above.

  For example, it is a program code generation device in which a database operation code template held in the operation code template holding unit is described based on a program language having an inheritance mechanism.

  The inheritance mechanism is a mechanism possessed by an object-oriented language or the like and defines an object that has been expanded or changed based on an object (target) that has already been defined.

  FIG. 25 is a diagram for explaining inheritance. It is assumed that a class 2501 which is a kind of object is already defined and has a method 1 and a method 2 representing an operation on itself. The subclass 2502 is a class that is defined for adding an extension or a change based on the class 2501. The subclass 2502 has the method 2 and the method 3, but the method 2 of the subclass 2502 is treated as a method for changing the method 2 of the class 2501, and the method 3 not in the class 2501 is treated as a method for extending the class 2501. It will be. Although the method has been described here, the same can be said for the variables of the class.

  In the case of FIG. 25, the subclass 2502 is said to inherit the class 2501 and is a subclass of the class 2501. The class 2501 is referred to as a superclass (superclass) of the subclass 2501 or an upper class.

  Therefore, in this embodiment, the program code generation device generates a definition of a class corresponding to the class 2501 as a program code, and provides an operation for accessing a database as a method. In this way, by defining the subclass 2502 by, for example, a human, it is possible to easily extend or change the program code generated by the program code generation device. For example, since the description of the class 2501 and the subclass 2502 can be performed in different files, even if the description of the class 2501 is newly generated by the program code generation device, the description of the class 2501 is directly changed by a person or the like. In other words, a change must be made every time the program code is generated. However, if the extension or change is performed by the subclass, the description of the subclass 2502 is not affected even if the program code is generated.

(Embodiment 5: Main effects)
According to the present embodiment, the program code generated by the program code generation device can be easily modified and expanded.

  (Embodiment 6: Claim 6 will be mainly described)

  As a sixth embodiment of the present invention, a program code generation apparatus according to the fifth embodiment, in which a character string of a table name and / or column name is defined as a constant and can be referred to, will be described.

(Embodiment 6: Configuration)
The program code generation device according to the present embodiment is the program code generation device of the fifth embodiment, and the program code completed by the program code completion unit defines a character string of a table name or / and a column name, The program code can be referred to in a program part that inherits the program code.

  FIG. 26 illustrates the program code generated by the program code generation device according to the present embodiment. The program code is a definition of a class called TABLE_A_manager, and as a String type variable, a character string “TABLE_A” is defined as TABLE_A to represent a table name, and a column name such as ID and A_CODE is represented. ID and A_CODE are defined as variables of the SQLColumn type. Note that the variable is not essential, and may be a constant whose value cannot be changed. By defining table names and column names in this way, subclasses can use the definitions.

  FIG. 27 shows an example in which A_manager is defined as a subclass of TABLE_A_manager. In order to generate a conditional expression indicating that the value of the column A_CODE is equal to a specific character string, a variable A_CODE defined by TABLE_A_manager is referred to. In this way, when the name of the column A_CODE is changed to another name, it is only necessary to modify the A_CODE definition portion of TABLE_A_manager. Also, if the column name or table name is changed, the variable name that represents the column name or table name is also changed. If the column name or table name is changed, the subclass Since an error occurs in the compilation, it is easy to know that the column name or table name has been changed, and the reliability of the application program is improved.

(Embodiment 6: Program code example)
28 and 29 illustrate program codes generated by the program code generation device according to the present embodiment. In this example, a table called CUSTOMER is defined, and CUSTOMER has columns named ID, NAME, BIRTH_DATE, and CUSTOMER_CODE. The data type of each column is INTERGER, VARCHAR, DATE, and VARCHAR. FIG. 28 shows a class called AbstractCustomerManager generated for operating a table called CUSTOMER. Among them, a class called CustomerEntity is used, which is a class representing row data stored in a table called CUSTOMER. CustomerEntity is a subclass of AbstractCustomerEntity defined in FIG. As shown in FIG. 29, AbstractCustomerEntity has a variable such as id, name, secondDate, and customerCode corresponding to a column, and obtains a method (for example, setId) that sets a value for each variable and the value of the variable. Has a method (eg, getId).

  30 and 31 exemplify class definitions for extending or changing the classes defined in FIGS. 28 and 29. FIG. 30 defines CustomerManager as a subclass of a class called AbstractCustomerManager, and defines a method called findByCustomerCode for performing a search based on the value of the column of CUSTOMER_CODE. In this, a variable called CUSTOMER_CODE defined in AbstractCustomerManager is referenced. For example, if a column named CUSTOMER_CODE is renamed to another name, the variable is renamed, so that it can be detected that a compilation error has occurred and CustomerManager must be modified. Note that a method called query is a method defined in an upper class of AbstractCustomerManager. FIG. 31 shows the definition of CustomerEntity, which is a subclass of AbstractCustomerEntity.

(Embodiment 6: Main effects)
According to the present embodiment, it is possible to easily know that the program change accompanying the change of the column name or the table name is reduced, or that the program must be changed because the column name or the table name is changed. It is possible to develop a reliable application program.

  The program code generation device and the program code generation method according to the present invention automate the development process of an application program that accesses a database, thereby improving the productivity of the application program and improving the reliability. , Industrially useful.

Diagram explaining the operation of the application program that accesses the database Functional block diagram of a program code generation device according to Embodiment 1 of the present invention Example of relational database table Example of schema information An example of a data structure constructed in the main memory and representing schema information Example of database operation code template Example of database operation code template Example of database operation code template Example of database operation code template Example of program code obtained by replacing database operation code template Example of program code Flow chart of processing of program code generator Functional block diagram of a program code generation device according to Embodiment 2 of the present invention Example of data type correspondence table The figure which shows the state which added the information which shows the data type of the variable used by a program code to the node corresponding to each column of the data structure illustrated in FIG. Example of database operation code template to be used in Embodiment 2 of the present invention Example of program code Functional block diagram of a program code generation device according to Embodiment 3 of the present invention Functional block diagram of a program code generation device according to Embodiment 3 of the present invention Example of second data type correspondence table Functional block diagram of a program code generation device according to Embodiment 4 of the present invention Example of logical deletion column information Flow chart of insertion processing in program code when logical deletion is used Flow chart of deletion processing in program code when logical deletion is used Diagram for explaining inheritance An example figure of the program code which the program code generating device concerning Embodiment 6 of the present invention generates Example of defining A_manager as a subclass of TABLE_A_manager Example of generated program code Example of generated program code Example of subclass of generated program code Example of subclass of generated program code

Explanation of symbols

200 Program Code Generation Device 201 Schema Information Acquisition Unit 202 Operation Code Template Holding Unit 203 Program Code Completion Unit

Claims (8)

A program code generation device for generating a program code for accessing a database,
A schema information acquisition unit for acquiring schema information which is information describing the schema of the database;
An operation code template holding unit for holding a database operation code template corresponding to the program code to be generated;
A program code completion unit for completing the program code based on the schema information acquired by the schema information acquisition unit and the database operation code template stored in the operation code template storage unit;
A program code generation device having: A data type correspondence table that is a table associating column data type information indicating a data type of a column used in the description of the schema information with variable data type information indicating a data type of a variable used in the program code. It has a data type correspondence table holding part to hold,
2. The program code generation device according to claim 1, wherein the program code completion unit includes a data type adjustment unit that completes a program code using a data type correspondence table held in the data type correspondence table holding unit. Corresponding to the second data type, which is a table associating column naming rule information indicating the naming rule of the column name appearing in the description of the schema information and variable data type information indicating the data type of the variable used in the program code A second data type correspondence table holding unit for holding a table;
The said program code completion part has the 2nd data type adjustment means which completes a program code using the 2nd data type correspondence table hold | maintained at the said 2nd data type correspondence table holding part. The program code generation device described. The schema information acquisition unit is information for specifying a column to be used for logical deletion in which the row data including the column value is deleted depending on whether or not the specific column has a specific value in the acquired schema information. Having logical deletion column information acquisition means for acquiring certain logical deletion column information;
The operation code template holding unit holds a database operation code template for logical deletion,
The program code generation device according to any one of claims 1 to 3, wherein the program code completion unit completes a program code based on the logical deletion sequence information acquired by the logical deletion sequence information acquisition unit.   5. The program code generation device according to claim 1, wherein the program code completed by the program code completion unit is described in a program language having an inheritance mechanism.   6. The program code generation according to claim 5, wherein the program code completed in the program code completion unit defines a character string of a table name or / and a column name and can be referred to by a program part that inherits the program code. apparatus.   7. The program code according to claim 1, wherein the program code is a program code for realizing one or more of data insertion into a database, database update, data deletion, and data search. 8. Program code generator. A program code generation method for generating a program code for accessing a database,
A schema information acquisition step of acquiring schema information which is information describing the schema of the database;
An operation code template reading step for reading a database operation code template corresponding to the program code to be generated;
A program code completion step for completing a program code based on the schema information acquired in the schema information acquisition step and the database operation code template read out in the operation code template reading step;
A program code generation method including:

Images