Program specification document verification method, its system, and its program

Abstract

To provide a specification document verification method capable of performing program verification with high accuracy and low load, while suppressing unnecessary warnings.SOLUTION: A computer executes a syntax analysis process that parses specification document data to generate hint data, which is data in the specification document that cannot be handled in a conversion process and is used in verification processing of a design check process, and semi-structured text group data that can be handled in the conversion process; a conversion process for converting the semi-structured text group data into abstract syntax tree group data; and a design check process that performs verification processing of the specification document data based on the abstract syntax tree group data and the hint data.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a program specification verification method, a system therefor, and a program therefor. [Background technology]

[0002] Program specifications are highly complex and often contain a variety of rules, conditions, and requirements. Therefore, it is difficult and time-consuming for a human to properly verify these diverse elements without making any errors. In addition, there are cases where the specification data intentionally does not contain the code that should be used, and a warning is issued for that code. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2023-115926 Summary of the Invention [Problem to be solved by the invention]

[0004] Verification of program specifications is important for ensuring the quality and reliability of programs, and the development of effective verification methods is required. It is also desirable to suppress unnecessary warnings.

[0005] The present invention has been made in consideration of the above circumstances, and its purpose is to provide a program specification verification method, system, and program that can verify programs with high accuracy and with little load, and suppress unnecessary warnings. [Means for solving the problem]

[0006] The present invention is a specification verification method in which a computer executes the following steps: a syntactic analysis step of parsing specification data and generating hint data, which is data of the specification data that cannot be handled in the conversion step and is used in the verification process of a design check step, and semi-structured text group data that can be handled in the conversion step; a conversion step of converting the semi-structured text group data into abstract syntax tree group data; and a design check step of performing the verification process of the specification data based on the abstract syntax tree group data and the hint data.

[0007] Preferably, the design check step determines whether or not there is a description in the abstract syntax tree group data that violates a predetermined rule, and generates a verification report including the result of the determination.

[0008] Preferably, the rule violation is any one of a misspelling of a table name or a column name or a violation of a naming rule, a violation of normalization of a table structure, a violation of a column type specification, a violation of a primary key definition, and a violation of an anti-pattern.

[0009] Preferably, in the design check process, even if there is a description of a rule violation, if the hint data contains memo information indicating that the rule violation is intentional, the design check process generates the verification report indicating that there is no rule violation or that the description of the rule violation is intentional.

[0010] Preferably, the design check step evaluates the relationships and consistency of the entire group of tables in the abstract syntax tree group data.

[0011] Preferably, the design check process expands the relational structure of the tables of the abstract syntax tree group data in memory, assigns the hint data related to a specific table or column to the table or column, evaluates the table or column names to infer naming rules, and assigns information indicating a naming rule violation to table names or column names that violate the naming rules, thereby generating the verification report.

[0012] Preferably, the syntactic analysis step generates the hint data and the semi-structured text group data using a learning model.

[0013] Preferably, the design check process determines whether the abstract syntax tree group data satisfies a first condition of a verification item, and for a verification item that is determined not to satisfy the first condition, determines whether the second condition is satisfied based on the hint data related to that verification item, and if the second condition is satisfied, determines that the condition of that verification item is satisfied.

[0014] Preferably, the hint data is information written in a remarks section of the specification data or in a portion other than the specification data.

[0015] Preferably, the specification data is human-readable application program data.

[0016] The present invention is a specification verification program that causes a computer to execute a parsing step of parsing specification data and generating hint data, which is data of the specification data that cannot be handled in the conversion step and is used in the verification process of the design check step, and semi-structured text group data that can be handled in the conversion step; a conversion step of converting the semi-structured text group data into abstract syntax tree group data; and a design check step of performing the verification process of the specification data based on the abstract syntax tree group data and the hint data.

[0017] The present invention is a specification verification system having a syntax analysis means for performing a syntax analysis of specification data and generating hint data, which is data of the specification data that cannot be handled by the conversion means and is used in verification processing in a design check step, and semi-structured text group data that can be handled by the conversion means; a conversion means for converting the semi-structured text group data into abstract syntax tree group data; and a design check means for performing the verification processing of the specification data based on the abstract syntax tree group data and the hint data. [Effects of the Invention]

[0018] According to the present invention, it is possible to provide a program specification verification method, a system therefor, and a program therefor that can perform program verification with a low load and high accuracy and suppress unnecessary warnings. [Brief explanation of the drawings]

[0019] [Figure 1] FIG. 1 is a diagram for explaining a specification verification system according to an embodiment of the present invention. [Figure 2] FIG. 2 is a diagram showing the hardware configuration of the specification verification system 11 shown in FIG. [Figure 3] FIG. 3 is a flowchart for explaining the flow of processing executed by the specification verification system 11 shown in FIGS. [Figure 4] FIG. 4 is a flowchart for explaining the verification process in step ST14 shown in FIG. DETAILED DESCRIPTION OF THE INVENTION

[0020] A specification verification method, system, and program according to an embodiment of the present invention will be described below. FIG. 1 is a diagram for explaining a specification verification system according to an embodiment of the present invention. As shown in FIG. 1, the specification verification system 11 includes, for example, a syntax analysis unit 15, a transpiler 21, and a design check unit 25.

[0021] The parsing unit 15 receives the specification data 13 , performs parsing of the specification data 13 , and generates hint data 19 and semi-structured text group data 17 . The specification data 13 may be application data such as Word or Excel. Specification data 13 is a set of various types of human-readable design documents (therefore, the focus is on reading design-related information from various types of documents and extracting necessary information from them). Reading ability is the most important point.

[0022] In the parsing process, for example, the first step is lexical analysis, which generates tokens, and the input string is divided into meaningful symbols according to definitions such as regular expressions. Next, narrow-sense parsing is performed to determine whether the sequence of tokens is a correct expression according to the syntax rules. For this purpose, syntax rules are referenced and applied recursively. As mentioned above, there are language rules that are not practical to express using syntax rules, such as duplicate formal parameter names in function definitions, and these are dealt with in this process.

[0023] The syntax analysis unit 15 generates hint data 19 and semi-structured text group data 17 using, for example, a learning model.

[0024] If there is memo information indicating that the rule violation is intentional, the syntax analysis unit 15 includes information about the memo information in the hint data 19. The memo may be in the specification data 13 or in a part associated with the specification data 13.

[0025] The hint data 19 is data of the specification data that cannot be handled in the conversion process and is used in the verification process of the design check unit 25 . Hint data 19 is information indicating, for example, that when the column type is not described in the syntax analysis unit 15, the type is inferred from the column name, and when the domain (domain information) type is not described, the type is inferred from the domain name.

[0026] In the context of data management, data modeling, and database design, a domain is a named, finite set of all possible unique values ​​for a data element. It is synonymous with data type, and is sometimes abbreviated to simply type.

[0027] For example, if there is a column name such as "birthday," the parser 15 can infer that the type is a date type, but further infer that the domain information indicates a past date.Also, if there is a column name such as "firstname," the parser 15 can infer that the type is a variable-length character string, but further infer that the domain information indicates a character string representing a Japanese surname. The syntax analysis unit 15 translates, for example, tabular and other table definitions into SQL statements and outputs them to semi-structured text group data 17. At that time, meta-information that cannot be included in SQL statements, as well as design policies and exceptions described in any design document, are output to hint data 19. Basically, this is sequential processing, and processing is performed in one pass.

[0028] The semi-structured text group data 17 is data that can be handled by the transpiler 21 . The semi-structured text group data 17 indicates table definitions and index definitions.

[0029] The syntax analysis unit 15 uses a learning model to read and extract definitions of tables / columns / indexes from design documents in an unspecified format and output them to semi-structured text group data 17. It also outputs meta-information that cannot be output to semi-structured text group data 17 (such as the fact that matters not specified in the design documents are "inferred" and output to semi-structured text group data 17) and exception information (such as deliberately allowing spelling mistakes in specific words) to hint data 19. (1) Regarding the detection of misspellings in table names and column names, there are cases where a decision has been made not to correct a misspelling even though it is a known issue. If this decision is recorded as memo information somewhere in the specification data 13 (for example, in a remarks column or on a separate sheet other than the specification data 13), hint data 19 is output indicating that the misspellings in the table name and column name will not be pointed out.

[0030] (2) There are cases where the naming rules for table names and column names are intentionally violated, and if this intention is stated as memo information somewhere in the design document, hint data 19 is output to the effect that the violation of the naming rules for the table name or column name in question will not be pointed out. (3) Regarding the normalization of table structures, there are cases where normalization is intentionally broken, and if this intention is described as memo information somewhere in the design document, hint data 19 is output to the effect that a normalization violation will not be pointed out for the table in question.

[0031] (4) Regarding type specification, there are cases where a type specification that is different from the usual one is intentionally selected, and if this intention is described as memo information somewhere in the design document, hint data 19 is output to indicate that a type specification violation will not be pointed out for the column in question. (5) Regarding primary keys, there are cases where primary keys are not defined intentionally, and if this intention is stated as memo information somewhere in the design document, hint data 19 is output to the effect that the absence of a primary key definition for the table in question will not be pointed out.

[0032] (6-1) Regarding the phantom file anti-pattern, if there is a memo somewhere in the design document stating that a separate backup / restore method or other considerations have been taken into account, hint data 19 is output indicating that the column in question does not violate the anti-pattern. (6-2) Regarding the Baskin-Robbins flavor anti-pattern, if there is a memo somewhere in the design document stating that enum is being used intentionally, output hint data 19 indicating that the column in question does not violate the anti-pattern.

[0033] (6-3) Regarding the rounding error anti-pattern, if there is a memo somewhere in the design document stating that FLOAT is used intentionally, hint data 19 is output indicating that the column in question does not violate the anti-pattern. (6-4) Regarding the metadata tribble anti-pattern, if there is a memo somewhere in the design document stating that the table is intentionally divided by year, etc., then output hint data 19 indicating that the anti-pattern violation will not be pointed out for the column in question.

[0034] (6-5) Regarding the multi-column attribute anti-pattern, if there is a memo somewhere in the design document stating that n columns with the same name are intentionally defined in the same table, hint data 19 is output to the effect that the anti-pattern violation for the column in question will not be pointed out. (6-6) Regarding polymorphic-related anti-patterns, if there is a memo somewhere in the design document stating that a polymorphic relationship is intentionally defined, hint data 19 is output to the effect that the table in question does not indicate an anti-pattern violation.

[0035] (6-7) Regarding the Entity Attribute Value anti-pattern, if there is a memo somewhere in the design document stating that Entity Attribute Value has been defined intentionally, then Hint Data 19 will be output to indicate that the table in question does not violate the anti-pattern. (However, this Hint Data 19 will only be output when special settings are made to output it, and will not be output by default. EAV is contrary to the concept of RDBMS, so it is natural to use NoSQL if necessary...Even within the category of RDBMS products, there are methods such as using JSON types.)

[0036] (6-8) Regarding the Keyless Entry antipattern, if there is a memo somewhere in the design document stating that Keyless Entry has been defined intentionally, then hint data 19 will be output to indicate that the table in question does not violate the antipattern. (However, this hint data 19 will only be output when special settings are made to output it, and will not be output by default. Since Keyless Entry is contrary to the concept of RDBMS, if Keyless Entry is deliberately selected, there is a high possibility that problems are occurring in the field.)

[0037] (6-9) Regarding the Id Required anti-pattern, if there is a memo somewhere in the design document stating that Id Required is intentionally defined, hint data 19 is output to the effect that the table in question does not indicate a violation of the anti-pattern. (6-10) Regarding the naive tree antipattern, if there is a memo somewhere in the design document stating that a naive tree is intentionally defined, hint data 19 is output to the effect that the table in question does not indicate an antipattern violation.

[0038] (6-11) Regarding the jaywalking anti-pattern, if there is a memo somewhere in the design document stating that jaywalking is defined intentionally, hint data 19 is output to the effect that the table in question does not indicate a violation of the anti-pattern. (6-12) Regarding the easy logical deletion flag anti-pattern, if there is a memo somewhere in the design document stating that the logical deletion flag is intentionally defined, hint data 19 is output to the effect that the table in question does not indicate a violation of the anti-pattern.

[0039] (6-13) Regarding the anti-pattern of easy naming of intersection tables, if it is stated somewhere in the design document that easy naming of intersection tables is intentionally defined, hint data 19 is output to the effect that the table in question does not violate the anti-pattern.

[0040] The transpiler 21 converts the semi-structured text group data 23 input from the syntax analysis unit 15 into abstract syntax tree group data. The abstract syntax tree group data 23 is an (abstract) tree structure obtained by removing information unrelated to the meaning of the language from a normal syntax tree (also called a concrete syntax tree or parse tree) and extracting only the information related to the meaning. Theoretically, it is a finite, labeled, directed tree. For example, in the case of an abstract syntax tree for something like a mathematical expression consisting of operators and operands such as variables and constants, the branching points are operators and the leaves are operands (variables and constants).

[0041] The design check unit 25 performs a verification process on the specification data based on the abstract syntax tree group data input from the transpiler 21 and the hint data input from the syntax analysis unit 15 .

[0042] The design check unit 25 determines whether the abstract syntax tree group data 23 generated by the transpiler 21 satisfies the first condition of the verification item (the condition that there is no violation of the rules). Next, for a verification item that has been determined not to satisfy the first condition, the design check unit 25 determines whether or not the second condition is satisfied based on the hint data 19 related to that verification item. If the second condition is satisfied, the design check unit 25 determines that the condition of that verification item is satisfied.

[0043] The design check unit 25 generates a verification report 27 including all or part of the hint data 19 . As described above, the design check unit 25 determines whether or not there is a description in the abstract syntax tree group data 23 that violates a predetermined rule, and generates a verification report 27 that includes the result of the determination.

[0044] The rule violation is one of misspellings in table or column names or violations of naming rules, violations of normalization of table structure, violations of column type specification, violations of primary key definition, and violations of anti-patterns.

[0045] Even if there is a description of a rule violation, if the hint data 19 contains memo information indicating that the rule violation is intentional, the design check unit 25 generates a verification report 27 indicating that there is no rule violation or that the description of the rule violation is intentional.

[0046] The design check unit 25 evaluates the relationships and consistency of the entire group of tables in the abstract syntax tree group data 23 . Specifically, the design check unit 25 expands the relational structure of the tables in the abstract syntax tree group data 23 in memory, and then generates a verification report 27 in which the hint data related to a specific table or column is added to the table or column. Furthermore, the design check unit 25 evaluates table names or column names to estimate naming rules, and generates a verification report 27 in which information indicating a naming rule violation is added to table names or column names that violate the naming rules.

[0047] The design check unit 25 generates a verification report 27 based on the abstract syntax tree group data 23 and the hint data 19. This is machine-readable data with a rigid format, and it can interpret this data and output appropriate warning messages. It has the function of continuously improving its ability to output more useful warning messages and suppress useless warning messages that become noise.

[0048] The specific processing of the design check unit 25 is as follows. (1) Detect spelling errors in table names and column names included in the abstract syntax tree (AST) group 23. If there is no specification in the hint data 19 to suppress the warning about the detected spelling error, report the spelling error. (2) Naming rules are estimated for table names and column names included in the abstract syntax tree group data 23. (Example: UserName=CamelCase, user_name=snake_case, user-name=kebab-case) If a table name or column name that violates a naming rule is detected and there is no specification in the hint data 19 to suppress the warning, a violation of the naming rule is reported.

[0049] (3) Detect normalization violations in the column definitions of each table included in the abstract syntax tree group data 23. If there is no specification in the hint data 19 to suppress a warning about the detected normalization violation, report the normalization violation. (4) Detect type specifications for the column definitions of each table included in the abstract syntax tree group data 23. If there is no specification in the hint data 19 to suppress a warning about a detected type specification violation, report the type specification violation. (5) When a primary key is undefined in each table included in the abstract syntax tree group data 23, if there is no specification in the hint data 19 to suppress the warning, the omission of the primary key definition is reported.

[0050] Here are some anti-patterns: (6-1) When it can be inferred from the table name and column name contained in the abstract syntax tree group data 23 that the column in question falls under the phantom file anti-pattern, if there is no specification in the hint data 19 to suppress the warning, the suspected anti-pattern is included in the verification report 27. (6-2) When an enum is used in a column definition included in the abstract syntax tree group data 23, if there is no specification to suppress the warning in the hint data 19, the suspected anti-pattern is included in the verification report 27. (6-3) If a floating-point type is used in a column definition included in the abstract syntax tree group data 23, and there is no specification to suppress the warning in the hint data 19, the suspected anti-pattern is included in the verification report 27.

[0051] (6-4) If the table definitions included in the abstract syntax tree group data 23 contain tables with the same structure, for example, in different years, and there is no specification in the hint data 19 to suppress the warning, the suspected anti-pattern is included in the verification report 27. (6-5) If columns with the same definition are used consecutively in a table definition included in the abstract syntax tree group data 23, and there is no specification in the hint data 19 to suppress the warning, the suspected anti-pattern is included in the verification report 27.

[0052] (6-6) If a polymorphic association is used in the relationships of the table group included in the abstract syntax tree group data 23, and if there is no specification to suppress the warning in the hint data 19, the suspected anti-pattern is included in the verification report 27. (6-7) If an Entity Attribute Value is used in a table included in the abstract syntax tree group data 23, and there is no specification in the hint data 19 to suppress the warning, the suspected anti-pattern is included in the verification report 27.

[0053] (6-8) If a foreign key is not defined where it should be in the relationship of a table group included in the abstract syntax tree group data 23, and there is no specification in the hint data 19 to suppress the warning, the suspected anti-pattern is included in the verification report 27. (6-9) If it can be assumed that a column named id is blindly used as a primary key in a table included in the abstract syntax tree group data 23, and if there is no specification to suppress the warning in the hint data 19, the suspected anti-pattern is included in the verification report 27. (6-10) If a naive tree is used in a table included in the abstract syntax tree group data 23, and there is no specification to suppress the warning in the hint data 19, the suspected anti-pattern is included in the verification report 27.

[0054] (6-11) If a plural column name is used in a column definition included in the abstract syntax tree group data 23, and if there is no specification to suppress the warning in the hint data 19, a suspected anti-pattern is included in the verification report 27. (6-12) If a column definition included in the abstract syntax tree group data 23 uses a column name such as is_deleted that suggests a simple logical deletion, and if there is no specification in the hint data 19 to suppress the warning, the suspected anti-pattern is included in the verification report 27. (6-13) If a table included in the abstract syntax tree group data 23 uses a simple intersection table name, and if there is no specification in the hint data 19 to suppress the warning, include a suspected anti-pattern in the verification report 27.

[0055] In the specification verification system 11, as shown in FIG. 1, the syntax analysis unit 15 and the transpiler 21 are separated. Therefore, when supporting formats other than Excel or Word, the syntax analysis unit 15 can be expanded without affecting the transpiler 21.

[0056] The specification verification system 11 distinguishes between information that is actually written in the specification and information that is inferred by AI, and outputs this distinction as hint data 19. Semi-structured text data 17 is output without distinguishing between information actually contained in the specifications as text and information inferred by the AI ​​(because the abstract syntax tree output by the transpiler cannot contain ambiguity). Furthermore, the parser 15 and the transpiler 21 are each provided with training data prepared by humans to build a model that allows for more precise inference.

[0057] FIG. 2 is a diagram showing the hardware configuration of the specification verification system 11 shown in FIG. As shown in FIG. 2, the specification verification system 11 includes, for example, a display 51, an operation unit 53, a communication unit 55, a memory 59, and a processing unit 61. The functions shown in FIG. 1 are realized using the hardware shown in FIG.

[0058] The display 51 displays an image based on a signal from the processing unit 61 . The operation unit 53 is an operation means such as a touch panel, a keyboard, or a mouse. The communication unit 55 communicates with an external device. The input unit 57 is a terminal or the like for inputting data from the outside. The memory 59 stores the program executed by the processing unit 61 . The processing unit 61 executes the program PR1 stored in the memory 59 to perform the processing of the specification verification system 11 defined in this embodiment. That is, the processing unit 61 executes a program to realize the functions of the specification verification system 11 shown in FIG.

[0059] The flow of processing executed by the specification verification system 11 shown in FIGS. 1 and 2 will be described below. FIG. 3 is a flowchart for explaining the flow of processing executed by the specification verification system 11 shown in FIGS. Each step will be explained. Step ST11: The specification verification system 11 receives specification data.

[0060] Step ST12: The specification verification system 11 performs syntax analysis on the specification data 13 input in step ST11 to generate hint data 19 and semi-structured text group data 17.

[0061] Step ST13: The specification verification system 11 converts the semi-structured text group data 23 generated in step ST12 into abstract syntax tree group data.

[0062] Step ST14: The specification verification system 11 performs verification processing of the specification data based on the abstract syntax tree group data generated in step ST13 and the hint data 13 generated in step ST12.

[0063] Step ST15: The specification verification system 11 outputs a verification report 27 showing the verification result of step ST14.

[0064] The verification process in step ST14 shown in FIG. 3 will be described in detail below. FIG. 4 is a flowchart for explaining the verification process in step ST14 shown in FIG. Step ST21: The specification verification system 11 determines whether the abstract syntax tree group data 23 satisfies the first condition of the verification item.

[0065] Step ST22: For the verification items determined not to satisfy the first condition in step ST21, the specification verification system 11 determines whether or not the verification items satisfy the second condition based on the hint data 19 related to the verification items.

[0066] Step ST23: If it is determined in step ST22 that the second condition is not satisfied, the specification verification system 11 generates a verification report 27 indicating that there is a problem with the verification item.

[0067] Step ST24: If the determinations in steps ST21 and ST22 are affirmative, the specification verification system 11 generates a verification report 27 indicating that there is no problem with the verification item.

[0068] According to this embodiment, the syntax analysis and verification process are automated, and the efficiency and accuracy of the verification process are improved, thereby reducing the verification load and suppressing unnecessary warnings.

[0069] The present invention is not limited to the above-described embodiments. That is, those skilled in the art may make various modifications, combinations, subcombinations, and substitutions of the components of the above-described embodiments within the technical scope of the present invention or its equivalents. [Industrial Applicability]

[0070] The present invention is applicable to a program verification system. [Explanation of symbols]

[0071] 11...Specification verification system 13...Specification data 15...Syntax analysis section 17...Semi-structured text data 19...Hint Data 21...Transpiler 23...Abstract Syntax Tree Group Data 25...Design Check Department 27...Verification Report

Claims

1. a syntactic analysis step of syntactically analyzing the specification data and generating hint data, which is data of the specification data that cannot be handled in the conversion step and is used in the verification process of the design check step, and semi-structured text group data that can be handled in the conversion step; the converting step of converting the semi-structured text group data into abstract syntax tree group data; the design check step of performing the verification process of the specification data based on the abstract syntax tree group data and the hint data; The computer executes Specification verification method.

2. The design check step includes: Determine whether or not there is a description that violates a predetermined rule in the abstract syntax tree group data, and generate a verification report that includes the result of the determination. The method for verifying specifications according to claim 1 .

3. Violation of the above rules is There is a spelling error in the table name or column name or a violation of naming rules, a violation of normalization of the table structure, a violation of column type specification, a violation of the primary key definition, or a violation of an anti-pattern. The method for verifying specifications according to claim 2.

4. The design check step includes: Even if there is a description of a rule violation, if there is memo information in the hint data indicating that the rule violation is intentional, the verification report is generated indicating that there is no rule violation or that the description of the rule violation is intentional. The specification verification method according to claim 3.

5. The design check step includes: Evaluate the relationships and consistency of the entire table group in the abstract syntax tree group data The method for verifying specifications according to claim 4.

6. The design check step includes: expanding the relational structure of the table of the abstract syntax tree group data in memory; assigning the hint data relating to a particular table or column to the table or column; A table name or a column name is evaluated to infer a naming rule, and information indicating a naming rule violation is added to a table name or a column name that violates the naming rule. Generate the verification report The specification verification method according to claim 5.

7. The syntactic analysis step generates the hint data and the semi-structured text group data using a learning model. The specification verification method according to claim 5.

8. The design check step determines whether the abstract syntax tree group data satisfies a first condition of a verification item, and for a verification item that is determined not to satisfy the first condition, determines whether the data satisfies a second condition based on the hint data related to that verification item, and if the data satisfies the second condition, determines that the condition of that verification item is satisfied.

7. The method for verifying specifications according to claim 6.

9. The hint data is information written in the remarks column of the specification data or in a portion other than the specification data. The method for verifying specifications according to claim 7.

10. The specification data is human-readable data of an application program. The specification verification method according to claim 8.

11. a syntactic analysis step of syntactically analyzing the specification data and generating hint data, which is data of the specification data that cannot be handled in the conversion step and is used in the verification process of the design check step, and semi-structured text group data that can be handled in the conversion step; the converting step of converting the semi-structured text group data into abstract syntax tree group data; a design check step of performing the verification process of the specification data based on the abstract syntax tree group data and the hint data; to the computer, Specification verification program.

12. a parsing means for parsing the specification data and generating hint data, which is data of the specification data that cannot be handled by the conversion means and is used in a verification process in a design check step, and semi-structured text group data that can be handled by the conversion means; the conversion means for converting the semi-structured text group data into abstract syntax tree group data; a design check means for performing the verification process of the specification data based on the abstract syntax tree group data and the hint data; A specification verification system having the above.

Images