JP6613706B2 - Table design support apparatus, table design support method, and control program - Google Patents

Description

  The present invention relates to a table design support apparatus, a table design support method, and a control program.

  In recent years, with the evolution of computer systems, the amount of data provided by others, such as open data published on the Internet, has been increasing and the types have been diversified. When managing these data using a database, if the data capacity is extremely large, it is common to use a platform such as Hadoop that can process a large amount of data distributed to a plurality of machines. On the other hand, if the capacity of these data is on the order of gigabytes (GB), it can be handled by a relational database management system (RDBMS). The RDBMS is convenient because the SQL language that is prevalent is the database language.

  In order to store data in the RDBMS, it is necessary to generate table definition data in a corresponding format. The generation of this table definition data requires a lot of specialized knowledge. Patent Document 1 discloses a table design support device that utilizes table definition data prepared in advance when newly generating table definition data in order to reduce the load on the table designer in generating table definition data. Has been described.

JP 2011-008443 A

  In the table design support apparatus described in Patent Document 1, in the table design work, the table designer needs to define restrictions such as the data type of each field (column) and uniqueness for the obtained record set. . When only a record set is provided by another person as in open data, it is necessary to perform table design work while confirming the contents of the provided record set. However, as the scale (data capacity) of record sets provided by others increases, errors such as data type mismatches and constraint violations tend to occur in table design work, and redesign work to correct errors There was a problem that more cases were needed. Here, the data type mismatch means a storage error due to a mismatch between the designed field data type and the actual record field value. Further, the constraint violation means a storage error due to an actual record set violating a field or record set constraint (such as uniqueness or null value).

  The present invention has been made in view of the above background, and in the table design work for the relational database management system, even if the size of the record set provided by another person increases, the data type mismatch or It is an object of the present invention to provide a table design support apparatus, a table design support method, and a control program that are unlikely to cause errors such as constraint violations.

  In order to solve such a problem, the present invention is a table design support device for supporting table design for a relational database management system, and is a delimiter character for detecting delimiter characters for dividing a record set provided by another person. A candidate detection unit, a record dividing unit that divides the record set by the delimiter candidate and generates a divided record group, a data type that can be stored in the relational database management system for each divided record in the divided record group, and A field analysis unit that analyzes the constraints and generates an analysis result file group; and an analysis result combination unit that combines the analysis result file groups to generate table definition candidates that can be stored in the relational database management system. Is.

  The present invention also provides table design support for supporting table design for a relational database management system using a table design support device comprising a delimiter candidate detection unit, a record division unit, a field analysis unit, and an analysis result combination unit. A method in which a delimiter candidate detection unit detects a delimiter candidate for dividing a record set provided by another, and a record dividing unit divides the record set by the delimiter candidate. A step of generating a divided record group, and a step of generating an analysis result file group by analyzing a data type and constraints that can be stored in the relational database management system for each divided record in the divided record group And the analysis result combining unit Generating a possible table definition candidate storage in the relational database management system combines the result files are those comprising a.

  Further, the present invention provides a control program for supporting table design for a relational database management system, the step of detecting a delimiter character candidate for dividing a record set provided by another person, Dividing a record set to generate a divided record group, and generating an analysis result file group by analyzing data types and constraints that can be stored in the relational database management system for each divided record in the divided record group And generating the table definition candidates that can be stored in the relational database management system by combining the analysis result file groups.

  According to the present invention, in the table design work for the relational database management system, even if the size of the record set provided by another person increases, errors such as data type mismatches and constraint violations can hardly occur in the table design work. .

It is a figure explaining the outline | summary of this invention. It is a block diagram which shows an example of schematic structure of the table design assistance apparatus concerning this Embodiment. It is a flowchart which shows the processing flow in the table design assistance apparatus concerning this Embodiment. It is a figure which shows an example of the record set processed with the table design assistance apparatus concerning this Embodiment. It is a figure which shows the state which divided the example of the record set shown in FIG. 4 with a delimiter character candidate. It is a figure which shows the state which divided | segmented an example of the record set shown in FIG. 4 by the delimiter character candidate. It is a flowchart which shows the flow of the process which determines the candidate of a delimiter in the record set in FIG. It is a figure explaining the determination of the delimiter candidate in FIG. 3 by a specific example. It is a figure explaining the determination of the delimiter candidate in FIG. 3 by a specific example. It is a flowchart which shows the flow of a process which counts the frequency | count of appearance of each already appearing character in one line with a record set file in FIG. It is a flowchart which shows the flow of the analysis process with respect to the division | segmentation record group in FIG. It is a flowchart which shows the flow of the determination process of the data type candidate in FIG. It is a figure which shows an example of the list of the scores corresponding to the RDBMS classification and data type of the executed match determination program. It is a figure which shows an example of the analysis result file which combined the analysis result file group output in step S46 of FIG. 10 into one file in step S5 of FIG.

[Features of the present invention]
Prior to the description of the embodiments of the present invention, an outline of the features of the present invention will be described first.

  FIG. 1 is a diagram for explaining the outline of the present invention. FIG. 1 is a block diagram showing an example of a schematic configuration of a table design support apparatus 1 showing an example of a schematic configuration of the present invention. The table design support device 1 includes a delimiter candidate detection unit 2, a record division unit 3, a field analysis unit 4, and an analysis result combination unit 5.

  The delimiter candidate detection unit 2 detects a character (delimiter candidate) for appropriately dividing a record set provided by another person. The record dividing unit 3 divides the record set by the delimiter candidate received from the delimiter candidate detection unit 2 and generates a divided record group. The field analysis unit 4 analyzes each divided record in the divided record group with respect to a data type that can be stored and restrictions such as uniqueness and presence / absence of a NULL value, and generates an analysis result file group. The analysis result combining unit 5 combines analysis result file groups to generate table definition candidates that can be stored in the RDBMS.

  The table design support apparatus 1 generates a divided record group by dividing a record set from a record set provided by others, and analyzes a data type and a constraint for each divided record, thereby obtaining a data type relational database. Table definition candidates that can be stored in the management system (RDBMS) can be automatically generated. As a result, in the table design work in the RDBMS, even if the scale of the record set provided by another person increases, errors such as data type mismatch and constraint violation can hardly occur in the table design work.

Embodiments of the present invention will be described below with reference to the drawings.
First, a schematic configuration of the table design support apparatus 200 according to the present embodiment will be described below.
FIG. 2 is a block diagram showing an example of a schematic configuration of the table design support apparatus 200 according to the present embodiment. The table design support apparatus 200 outputs to the client terminal 100 table definition candidates that can store the input record set in the RDBMS. As shown in FIG. 2, the table design support apparatus 200 includes an input information receiving unit 201, an input information storage unit 202, a delimiter candidate detection unit 203, a delimiter candidate transmission unit 204, and a delimiter receiving unit 205. A record dividing unit 206, a field analyzing unit 207, an analysis result combining unit 208, an analysis result transmitting unit 209, and a storage device 220.

  A record set 111 and RDBMS type information provided by another person are stored in the storage device 110 of the client terminal 100. The RDBMS type information means information related to the type of RDBMS such as SQL99, Oracle, MySQL, PostgreSQL, and the like.

  The input information receiving unit 201 receives data (record set 111 and RDBMS type information 211) transmitted from the client terminal 100 (input information transmitting unit 102). The input information storage unit 202 stores the data received by the input information reception unit 201 in the storage device 210 as RDBMS type information 211 and a record set 212.

  The delimiter candidate detection unit 203 detects a character for appropriately dividing the record set 212. The delimiter candidate transmission unit 204 sends the delimiter candidate to the client terminal 100 (delimiter character candidate reception unit 103) in order to confirm with the table designer when the detection result in the delimiter candidate detection unit 203 is not uniquely determined. Send. In this case, the table designer inputs the delimiter character selected from the delimiter input candidates from the client terminal 100 (input device 101).

  The delimiter receiving unit 205 receives the delimiter input by the table designer from the client terminal 100 (delimiter transmitting unit 104). The record dividing unit 206 generates a divided record group 213 by dividing the record set 212 by the delimiter received from the delimiter candidate detection unit 203 or the delimiter receiving unit 205. The divided record group 213 is stored in the storage device 210 as temporary data.

  The field analysis unit 207 analyzes each divided record in the divided record group 213 and generates an analysis result file group 215. The analysis result combining unit 208 combines the analysis result files included in the analysis result file group 215 to generate one analysis result file (table definition candidate that can be stored in the RDBM). The analysis result transmission unit 209 transmits the table definition candidates to the client terminal 100 (analysis result reception unit 105). The storage device 220 stores each determination program (data type determination program 221, NULL value determination program 222, and uniqueness determination program 223) used when the field analysis unit 207 analyzes the divided record group 213.

  The delimiter candidate detection unit 203, the record division unit 206, the field analysis unit 207, and the analysis result combination unit 208 may be configured in part or in whole by software (programs) or by hardware circuits. You may make it do.

Next, the operation of the table design support apparatus 200 will be described below. In the following description, reference is also made to FIG. 2 as appropriate.
FIG. 3 is a flowchart showing the flow of processing in the table design support apparatus 200. As shown in FIG. 3, in step S <b> 1, the input information receiving unit 201 receives RDBMS type information and data of the record set 111 from the client terminal 100, and stores the data received via the input information storage unit 202 in the storage device 210. save. A record set stored in the storage device 210 is referred to as a record set 212, and RDBMS type information is referred to as RDBMS type information 211.

  Next, in step S2, the delimiter candidate detection unit 203 analyzes the record set 212 and detects delimiter candidates that can divide the record set 212 into a plurality of columns. When the detected delimiter candidates are not single (a plurality of delimiters are detected or there is no corresponding delimiter), the delimiter candidate transmission unit 204 transmits the delimiter candidate candidates to the delimiter candidate reception unit 103 on the client terminal side, and the table design Confirm with the person in charge. When the table designer selects a delimiter from the input device 101, the delimiter is sent to the record dividing unit 206 via the delimiter transmitting unit 104 and the delimiter receiving unit 205. When the delimiter candidate detected by the delimiter candidate detection unit 203 is single, the delimiter candidate detection unit 203 sends it to the record division unit 206 as it is. Details of the process for determining a delimiter candidate will be described later.

  Next, in step S3, the record dividing unit 206 divides the record set 212 into field units based on the delimiter character calculated in step S2, and the divided record group 213 is stored in the storage device 210. The division of the record set 212 will be specifically described with reference to FIGS.

  FIG. 4 is a diagram illustrating a record set R100 that is an example of the record set 212. FIG. 5 shows a state where the record set R100 shown in FIG. 4 is separated by a delimiter candidate S (in this example, “,” (comma)) detected by the delimiter candidate detection unit 203. The record set R100 is divided at the position of the delimiter candidate S. After the record set R100 is divided, a plurality of divided records R101, R102, R103, and R104 are generated as shown in FIG. A divided record group R105 (an example of the divided record group 213) including a plurality of divided records R101, R102, R103, and R104 is stored in the storage device 210.

  As shown in FIG. 3, following step S <b> 3, in step S <b> 4, the field analysis unit 207 determines, based on the presence / absence of NULL value, presence / absence of uniqueness, and RDBMS type information 211 for each of the generated divided record groups 213. Analyze data type and data length. Then, a table definition candidate group that can be stored in field units is stored in the storage device 210 as an analysis result file group 215. Details of the analysis processing for the divided record group 213 will be described later.

  In step S5, the analysis result combining unit 208 combines the analysis result file groups 215 for each field into one file. This file is a table definition candidate that can be stored in the RDBMS. The table definition candidates are transmitted to the client terminal 100 via the analysis result transmission unit 209.

A process for determining a delimiter candidate in step S2 of FIG. 3 will be described below.
FIG. 7 is a flowchart showing a flow of processing for determining a delimiter candidate in the record set 212. As shown in FIG. 7, first, in step S21, an already-existing character list is initialized. An already-existing character is a character that has been processed in the processing flow for determining candidate delimiters. When the process for determining the delimiter candidate is completed, all the characters included in the record set 212 are registered in the existing character list. Since the already-existing character is unknown at the start of the process of determining the delimiter candidate, initialization processing for clearing the already-existing character list is performed.

  Next, in step S22, one line in the record set 212 is read. First, the first line of the record set 212 is read, and after the processing of Step S23 and Step S24 described later, the process returns to Step S22 to read the next line. Steps S22 to S24 are repeated until the last line of the record set 212.

  Next, in step S23, the number of appearances for each character is counted for one line read in step 22. Specifically, the list of existing characters is confirmed and registered for the characters described in the record set 212, and the number of appearances of characters for each row is stored in an associative array (hash) of line numbers and existing characters. To do. Details of the process for confirming and registering the list of existing characters will be described later.

  In step S24, it is determined whether the read one row is the last row of the record set 212. If it is determined in step S24 that the read line is the last line of the record set 212, the number of appearances is counted for each line of the record set 212 for each character registered in the existing character list in step S25. Do. Then, for each character registered in the list of existing characters, the appearance frequency coincidence ratio is calculated from the appearance frequency for each line.

  Subsequent to step S25, in step S26, a delimiter candidate is determined. Specifically, the delimiter candidate detection unit 203 counts the number of appearances of each character included in the record set 212 for each row of the record set 212. Then, the delimiter candidate detection unit 203 calculates an appearance count matching rate that is a ratio of the maximum number of lines with the same number of appearances for each character included in the record set 212 to the total number of lines, and the appearance count match rate The delimiter candidate is determined based on. A character having an appearance count matching rate of 1 means that the appearance count is the same in all the rows of the record set 212, so this character is set as a delimiter candidate. If there is no character with an appearance count match rate of 1, the character with the highest appearance count match rate is set as a delimiter character candidate. When a character whose appearance frequency matching rate is not 1 is selected as a delimiter character candidate, a line with a different appearance frequency is excluded as an exception line when the record dividing unit 206 generates a divided record, and is not subject to subsequent processing. And After determining the delimiter candidate, the process is terminated.

  If it is determined in step S24 that the read line is not the last line of the record set 212, the process returns to step S22, and the next line is read in step S22.

  FIG. 8 is a diagram for explaining determination of a delimiter candidate by a specific example. FIG. 8A is a diagram illustrating an example of an associative array (hash) of line numbers, line numbers, and characters (existing characters) registered in the already-existing character list. A number (hash function value) in the associative array of line numbers and existing characters is the number of appearances of the existing characters. For convenience of explanation, the record set 212 has six rows. As shown in FIG. 8A, since the appearance number “,” (comma) appears three times in the first to sixth lines of the record set 212, the appearance number match rate is (the number of appearance matches). The maximum number of lines) / (total number of lines) = 6 [lines] / 6 [lines] = 1.00.

  For the existing characters other than “,” (comma), for example, in the existing character “a”, the number of appearances in each line of the record set 212 is 1 for the first line, 2 for the second line, 3 for the third line, The fourth line is 3 times, the 5th line is 1 time, and the 6th line is 3 times. In the existing character “a”, the number of occurrences of 3 times is the largest in 3 lines (3rd line, 4th line, 6th line). That is, the maximum number of lines with the same number of appearances is three lines. Therefore, the appearance count matching rate is (maximum number of rows with the same appearance count) / (total number of rows) = 3 [rows] / 6 [rows] = 0.50.

  As shown in FIG. 8B, since only the appearance frequency matching rate of “,” (comma) is 1.00 in the existing characters, “,” (comma) is set as a delimiter character candidate (determination result ○).

A process of counting the number of appearances of each existing character in one line of the record set 212 in step S23 of FIG. 7 will be described below.
FIG. 9 is a flowchart showing a flow of processing for counting the number of appearances of each existing character in one line of the record set 212. As shown in FIG. 9, first, in step S231, one character in one line of the record set 212 is read. First, after reading the first character in one line of the record set 212 and performing steps S232 to S235 described later, the process returns to step S231 to read the next character. The processing from step S231 to step S235 is repeated up to the last character of a line in the record set 212.

  Next, in step S232, it is confirmed whether or not the read one character is registered in the existing character list. If the single character read in step S232 is registered in the list of existing characters, the process proceeds to step S234. If the read character is not registered in the existing character list, the read one character is registered in the existing character list in step S233, and the process proceeds to step S234.

  Next, in step S234, the hash function value in the associative array (hash) of line numbers and existing characters is counted up, and the number of appearances of each existing character in the read line is recorded. Next, in step S235, it is determined whether the character being processed is the last character in a line. If it is determined in step S24 that the character being processed is the last character in a line, the processing ends. If it is determined that the character being processed is not the last character in a line, the process returns to step S231, and the next character is read in step S231.

The analysis process for the divided record group 213 in step S4 in FIG. 3 will be described below.
FIG. 10 is a flowchart showing the flow of analysis processing for the divided record group 213. As shown in FIG. 10, in step S41, one of the divided records included in the divided record group 213 is read. Note that after reading one of the divided records included in the divided record group 213, processing in steps S42 to S45 described later is performed, and the process returns to step S41 to read the next divided record. The processes in steps S41 to S45 are repeated until all the divided records included in the divided record group 213 are processed.

  Next, in step S42, it is determined by the NULL value determination program whether data with a character string length of 0 is included for each of the divided records included in the divided record group 213 (NULL value determination). . Then, in the NULL value determination, when a divided record including the NULL value is found, a process of removing the NULL value from the divided record is performed. Note that the post-processing division record group 214 (see FIG. 2) is a post-processing division record group after the processing except the NULL value is performed on the division record group 213. In the NULL value determination, when a NULL value is not found from any of the divided records included in the divided record group 213, the contents of the divided record group 213 and the post-processing divided record group 214 are the same.

  Next, in step S43, it is determined whether or not there is uniqueness to the post-process division record that has undergone the process of removing the NULL value in step S42. Specifically, the uniqueness determination program first sorts the post-processed divided records in units of rows, and then compares the preceding and succeeding rows to check whether there are duplicate rows. When duplicate rows are found in the post-process division record, row integration processing is performed to combine the duplicate rows into one row. If the number of lines in the post-process division record is the same before and after this line integration process, it is determined that the post-process division record has uniqueness.

  Next, in step S44, data type candidates that can be used for storing in the RDBMS based on the RDBMS type information 211 are determined based on the RDBMS type information 211 for the post-processed divided records that have undergone the process of removing the NULL value in step S42. Details of the data type candidate determination process will be described later.

  Next, in step S45, it is determined whether or not all the divided records included in the divided record group 213 have been processed. If it is determined in step S45 that all the divided records included in the divided record group 213 have been processed, the process proceeds to step S46. In step S46, the analysis result for the divided record group 213 is output as the analysis result file group 215. If it is determined in step S45 that all the divided records included in the divided record group 213 have not been processed, the process returns to step S41, and the next divided record is read in step S41.

The data type candidate determination process in step S44 of FIG. 10 will be described below.
FIG. 11 is a flowchart showing the flow of data type candidate determination processing. As shown in FIG. 11, RDBMS type information is read in step S441.

  Next, in step S442, it corresponds to one of registered data types (for example, IMTEGER type, CHAR type, etc.) in the RDBMS type (for example, SQL99, Oracle, MySQL, PostgreSQL, etc.) obtained from the RDBMS type information. Read the match determination program. Note that the match determination program differs depending on the data type defined in the RDBMS type. After reading the match determination program corresponding to one of the registered data types in the RDBMS type, the processing in steps S443 to S445 described later is performed, and the process returns to step S442 again to match the next registered data type. Load the judgment program. The processes in steps S442 to S445 are repeated until all of the registered data type match determination programs have been executed.

  Next, in step S443, the read match determination program is executed for the post-process division record that has undergone the process of removing the NULL value in step S42, and the data type of the post-process division record is for the match determination being executed. Judges whether it matches the data type of the program. Hereinafter, of the data types defined in the standard SQL standard SQL99, the “INTEGER type” match determination program and the “CHAR type” match determination program are used as examples. A method for determining whether or not the data type of the matching determination program being executed matches will be specifically described.

  The INTEGER type in SQL99 is integer data that can store an integer in the range of -2147483648 to 2147483647. For this reason, in the INTEGER type match determination program in SQL99, first, it is confirmed whether or not a non-integer value is included in the post-process division record. When non-integer values are not included (all are integers), the maximum value and the minimum value of the data included in the post-process division record are obtained, and if this is within the range of the INTERGER type value, the post-process division record Is determined to be storable in the INTERGER type. Since the CHAR type in SQL99 is a data type composed of characters, the CHAR type match determination program in SQL99 first checks whether the data in each line of the post-process division record is a character string. When each line of the post-process division record is a character string, the character string length of each line is calculated, and the maximum value of the calculated character string lengths is determined as the CHAR type character string length.

  Following step S443, in step S444, the result of match determination is recorded. If it is determined in step S443 that the data types match, the score calculated corresponding to the RDBMS type and data type of the executed match determination program is used as the data type corresponding to the executed match determination program. Record as a “valid score”. On the other hand, if it is determined in the match determination program executed in step S443 that the data types do not match, no score is recorded for the data type corresponding to the executed match determination program. An example of a list of scores corresponding to the RDBMS type and data type of the executed match determination program is shown in FIG.

As shown in FIG. 11, in step S445 following step S444, it is determined whether or not the data type match determination has been completed for all registered data types. If it is determined in step S445 that the data type match determination has been completed for all registered data types, the process proceeds to step S446. On the other hand, if it is determined in step S445 that the data type match determination has not been completed for all registered data types, the process returns to step S442 to read the next registered data type.
The

  In step S446, the data type in which “valid score” is recorded in step S444 is set as a data type candidate capable of storing the divided record group 213. Further, the data type candidates are sorted in descending order of effective scores, and the priority of the data type candidates is determined.

  FIG. 13 is a diagram showing an example of an analysis result file (table definition candidates that can be stored in the RDBMS) obtained by combining the analysis result file group output in step S46 of FIG. 10 into one file in step S5 of FIG. It is. As shown in FIG. 13, for example, in the first field (field 1) of the record set, the priority of the data type is SMALLINT type (priority 1), INTEGER type (priority 2), CHAR type or VARCHAR type. In order of (priority 3). Also, field 1 is unique and does not have a NULL value.

  The table design support apparatus 200 described in the above embodiment can automatically generate table definition candidates that can be stored in the RDBMS for a record set provided by another person. As a result, in the table design work in the relational database management system, even if the size of the record set provided by others increases, errors such as data type mismatches and constraint violations can hardly occur in the table design work.

  Although the present invention has been described with reference to the above embodiment, the present invention is not limited only to the configuration of the above embodiment, and within the scope of the invention of the claims of the present application. It goes without saying that various modifications, corrections, and combinations that can be made by those skilled in the art are included.

  For example, in the above embodiment, the client terminal 100 and the table design support device 200 are configured as separate devices, but may be configured as a single unit. Moreover, although the component of the table design assistance apparatus 200 was comprised integrally, you may comprise separately. If the field analysis unit 207 is arranged in a plurality of different apparatuses, the divided record group 213 can be analyzed in parallel, so that the processing time can be shortened. When the field analysis unit 207 is arranged in a plurality of different apparatuses, the divided record group 213, the post-processed divided record group 214, and the analysis result file group 215 are configured to be shared among a plurality of apparatuses.

1 Table design support device 2 Delimiter candidate detection unit 3 Record division unit 4 Field analysis unit 5 Analysis result combination unit

Claims (6)

A table design support device for supporting table design for a relational database management system,
A delimiter candidate detection unit for detecting delimiter candidates for dividing a record set provided by another person,
A record dividing unit that divides the record set by the delimiter candidate and generates a divided record group;
For each divided record in the divided record group, a field analysis unit that analyzes data types and constraints that can be stored in the relational database management system and generates an analysis result file group;
An analysis result combining unit that combines the analysis result file groups to generate table definition candidates that can be stored in the relational database management system;
Equipped with a,
The delimiter candidate detection unit aggregates the number of appearances of each character included in the record set for each row of the record set, and the number of occurrences of the line where the number of appearances matches for each character included in the record set. A table design support device that calculates an appearance frequency coincidence ratio that is a ratio of the maximum number of all rows and determines the delimiter candidate based on the appearance frequency coincidence rate . The field analysis unit determines, for each of a plurality of data types defined in the relational database management system, whether or not the divided record and the data type match, and determines a data type candidate of the divided record The table design support apparatus according to claim 1 . The field analysis unit analyzes whether or not a NULL value is included in each divided record of the divided record group, and when a divided record including the NULL value is found, a NULL value is calculated from the divided record. The table design support apparatus according to claim 1 or 2 , wherein the post-process division record is generated. The field analysis unit sorts the post-processed divided records in units of rows, checks whether or not there are duplicate lines by comparing preceding and subsequent lines, and duplicates in the post-processed divided records. When a row is found, a row integration process is performed to combine the duplicate rows into a single row. The table design support apparatus according to claim 3, wherein it is determined that there is a table. A table design support method for supporting table design for a relational database management system by means of a table design support device comprising a delimiter candidate detection unit, a record division unit, a field analysis unit, and an analysis result combination unit,
A delimiter candidate detection unit detecting delimiter candidates for dividing a record set provided by others;
A record dividing unit dividing the record set by the delimiter candidates to generate a divided record group;
A field analysis unit, for each divided record in the divided record group, analyzing data types and constraints that can be stored in the relational database management system to generate an analysis result file group;
An analysis result combining unit generating the table definition candidates that can be stored in the relational database management system by combining the analysis result file groups;
Equipped with a,
In the step of detecting the delimiter candidate, the delimiter candidate detection unit counts the number of appearances of each character included in the record set for each row of the record set, and for each character included in the record set A table design support method for calculating an appearance frequency coincidence ratio that is a ratio of a maximum number of lines having the same appearance frequency to the total number of lines, and determining the delimiter candidate based on the appearance frequency coincidence ratio . A control program that supports table design for a relational database management system,
Detecting a delimiter candidate for dividing a record set provided by others,
Dividing the record set by the delimiter candidates to generate a divided record group;
Analyzing data types and constraints that can be stored in the relational database management system for each divided record in the divided record group, and generating an analysis result file group;
Combining the analysis result file group to generate a table definition candidate that can be stored in the relational database management system ;
In the step of detecting the delimiter candidate, the delimiter candidate detection unit counts the number of appearances of each character included in the record set for each row of the record set, and for each character included in the record set A control program that calculates an appearance frequency coincidence ratio that is a ratio of the maximum number of lines with the same appearance frequency to the total number of lines, and determines the delimiter candidate based on the appearance frequency coincidence ratio .

Images