JPH03211627A - Data type conversion system - Google Patents

Abstract

PURPOSE:To obtain description efficient in execution speed and a memory use quantity at the time of converting into software description while using a target language by providing an optimum data type decision part. CONSTITUTION:The optimum data type decision part 4 receives a data definition sentence and a data use sentence from a data definition sentence retrieval part 2 and a data use sentence retrieval part 3, and it selects one type of data which can realize all data use sentences most efficiently among the data types which the target language has. A data definition sentence conversion part 5 converts the data definition sentence informed from the data definition sentence retrieval part 2 into definition as the data type informed from the optimum data type decision part 4. A data use sentence conversion part 6 converts all data use sentences informed from the data use sentence retrieval part 3 into the use sentence in which data of the data type informed from the optimum data type decision part 4 is set to be an object. Thus, description efficient in the execution speed and the memory use quantity can be obtained at the time of conversion into software discription by the target language.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention inputs a software description in a language with a high level of abstraction, and automatically or semi-automatically outputs a description of software in a language with a low level of abstraction. This relates to software conversion methods.

[Conventional technology]

FIG. 5 shows a configuration diagram of a conventional data type conversion method. Fifth
As shown in the figure, in the conventional data conversion method, the input section 5
1, a data definition statement search unit 52, a data usage statement search unit 53, a data definition statement conversion unit 54, a data usage kiln conversion unit 55, and an output unit 56.

Conventionally, for each data type included in an input language, one data type included in an output language is determined in advance, regardless of the usage form of each data. Therefore, a data definition statement search unit 52 searches for a data definition statement in the software description input by the input unit 51, and a data definition statement conversion unit 54 converts it into a data type in a predetermined output language. , was passed to the output section 56. Further, the data usage sentence search unit 53 searches for usage sentences of each data in the description input by the input unit 51, and the data usage kiln conversion unit 55 searches for the data type converted by the data definition sentence conversion unit 54. It is converted into a predetermined data usage method corresponding to the data usage method and passed to the output unit 56, and the output unit 56 converts the data into a predetermined data usage method corresponding to the data definition statement conversion unit 54. Contains the conversion result of the data utilization kiln conversion unit 55,
It was outputting the description of the converted software.

(Problem to be Solved by the Invention) In general, individual software description languages have their own data types, and in particular, languages with different levels of abstraction often have different data types. When humans design software, when the design progresses from highly abstract descriptions such as functional specifications to highly specific descriptions such as detailed specifications, the usage form of each data is identified and the usage form is determined. Converting to the most efficient data type.

In the conventional method described above, one data type of the output language is determined in advance for each data type of the input language.
No matter how the data is used, if it is of the same data type, it will be converted to one data type, which will reduce the execution speed,
It was difficult to output the most efficient software description in terms of memory usage.

The purpose of the present invention is to solve these problems and to provide a data type conversion method that makes it possible to obtain a description that is efficient in terms of both execution speed and memory usage when converting to a software description in a target language. It is about providing.

[Means to solve the problem]

The present invention provides software conversion that inputs software written in a certain input language and outputs a software description in a target language that has the same meaning.
In the method of converting the data type of an input language into the data type of a target language, there is an input section for inputting the software description in the input language, and data definition is performed from among the software input in the input section. a data definition statement search unit that searches for a sentence; and a data usage statement search that searches for a statement that uses data defined by the data definition statement searched by the data definition statement search unit from among the software input in the input unit. and the data type defined by the data definition statement by referring to the data usage form in the data usage statement found by the data usage statement search unit,
an optimal data type determination unit that determines the most suitable data type for the usage pattern; and a data definition statement that converts the data definition found by the data definition statement search unit into a data definition based on the data type determined by the optimal data type determination unit. The conversion unit and the data usage statement found by the data usage statement search unit are
a data usage conversion unit that converts the data usage method of the data type determined by the optimal data type determination unit; the data definition statement conversion unit converts the data definition statement; and the data usage conversion unit converts the data usage statement. and an output unit that outputs software that has been created.

〔Example〕

Embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a diagram showing the configuration of an embodiment of the present invention. This data conversion method consists of an input section l, a data definition statement search section 2, a data usage statement search section 3, and an optimal data type determination section 4.
, a data definition statement conversion unit 5 , a data usage conversion unit 6 , and an output unit 7 .

First, in the software description input in the input unit 1, the data definition statement search unit 2 searches for a data definition statement that defines data as a certain data type, and the data usage statement search unit 3 searches for that data definition statement. Search for all data usage statements that use the data defined in the definition statement. The optimal data type determining unit 4 receives and receives data definition statements and data usage statements from the data definition statement search unit 2 and data usage statement search unit 3, respectively, and selects a data type that can most efficiently realize all of the data usage statements. is selected from among the data types of the target language, and is notified to the data definition sentence conversion section 5 and the data usage conversion section 6. The data definition statement conversion unit 5 converts the data definition statement notified from the data definition statement search unit 2 into a definition as a data type notified from the optimal data type determination unit 4,
The result is notified to the output unit 7. Data usage sudden conversion part 6
converts all the data usage sentences notified from the data usage sentence search unit 3 into usage sentences that target data of the data type notified from the optimal data type determination unit 4, and notifies the output unit 7.

FIG. 2 shows the configuration of the optimal data type determining section 4. As shown in FIG.

The optimal data type determination section 4 includes a data type determination rule group 21 that is a collection of rules for determining the optimal data type, a work area section 22 to which the rule group is applied, and a data type determination rule group 21 that is a collection of rules for determining the optimal data type, a work area section 22 for applying the rule group, and a data type determination rule group 21 that is a collection of rules for determining the optimal data type. Reasoning mechanism 23 that selects rules and applies rules
It is composed of.

In the optimal data type determination unit having such a configuration, the data definition statement search unit 2 in FIG. From the data usage statement search unit 3,
When notified of the data definition statement and data usage statement, respectively,
The optimal data type determination section 4 stores these in the work area section 22. The inference mechanism 23 selects and applies one rule applicable to the contents of the work area section 22 from the data type determination rule group 21.

The above operations are repeated until the most suitable data type is determined from among the data types of the target language.

Next, the above operation will be explained based on one input example. Third
Assume that you enter the description of the software shown in the figure. In step 31, it is shown that the data "odd number" is defined as data of the data type "set of integers".

Similarly, in Ste Nobu 32 and 33, the data "positive integer less than or equal to 10" and "odd number less than or equal to 10" are each defined as "set of integer" type data. In step 34, a set of positive integers whose remainder when divided by 2 is 1 is assigned to the data "odd number". In step 35, a set of integers from 1 to 10 is assigned to the data "positive integer less than or equal to 10". In step 36, the data "
The logical product of "odd number" and the data "positive integer less than 10" is assigned to the data "odd number less than 10".

FIG. 4 shows a group of rules for converting set type data into bit array type data or an element/no element determination function in software description in an input language.

Rule 41 indicates that when the upper limit of the number of elements in a set is determined and there are many logical operations between set data, a bit array type is determined as the optimal data type.

Rule 42 states that the number of elements in a set may be infinite;
If there are many processes for determining whether an element is an element or not, an element/no element determination function is determined as the optimal data type.

Rule 43 indicates that if there is a usage pattern in which values are assigned from a set definition by enumerating elements, it is written in the work area section that it is a usage pattern where the upper limit of the number of elements is determined. ing.

Rule 44 indicates that if there is a usage pattern in which a value is assigned from a set definition that indicates the properties of an element, it is written in the work area that this is a usage pattern where the number of elements is infinite. ing.

Rule 45 states that if there is a form of use in which a logical product is assigned between set-type data that may have an infinite number of elements and set-type data that has a fixed upper limit on the number of elements, then this is the upper limit on the number of elements. This indicates that the usage pattern is written in the work area section when .

Rule 46 indicates that if all usage types are usage types where the upper limit of the number of elements is determined, the fact that the upper limit of the number of elements of the set type data is determined is written in the work area section. .

Rule 47 indicates that when one usage type is a usage type where the number of elements is infinite, the fact that the number of elements of the set type data may be infinite is written in the work area section. .

Among these, rules 4I and 42 directly determine the optimal data type, and rules 43 to 47 refer to the data usage form and specify the preconditions necessary for rule 41 or 42 to be applied. This is a rule for testing.

When the data type determination rule group 21 has the above rules, the software description shown in FIG.
2, the rule 44 in FIG. 4 writes into the work area section 22 that the usage pattern in step 34 in FIG. 3 is the usage pattern when the number of elements is infinite. Similarly, the rule 43 writes in the work area section 22 that the usage pattern in step 35 in FIG. 3 is a usage pattern when the upper limit of the number of elements is determined.

Next, Rule 47.46 refers to the content entered by Rule 44.43, and Steps 31 and 32 in Figure 3 respectively represent aggregate data that may have an infinite number of elements, and the upper limit of the number of elements. This is written in the work area section 22 to indicate that the set data has been determined.

Next, rule 45 is applied within work area section 21 to rule 46.47.
Referring to the contents written in step 33 of FIG. 3, the information that step 33 in FIG.

Finally, Rule 41 states that the bit array type is most suitable for the data "positive integer less than or equal to 10" and "odd number less than or equal to 10", respectively, referring to the contents written in Rule 43.45. is written in the work area section 22. Further, the rule 42 refers to the content written in the rule 44 and writes in the work area section 22 that the element/non-element determining function is most suitable for the data "odd number".

Data definition statement conversion unit 5. The data usage conversion unit 6 converts the data definition statement and the data usage statement, respectively, according to the contents written in the work area unit 41 in the optimal data type determination unit 4.

〔Effect of the invention〕

In the data type conversion method according to the present invention, the optimal data type determining unit receives data definition statements and data usage statements from the data definition statement search unit and the data usage statement search unit, respectively, and converts all of the data usage statements in the most efficient manner. Select one data type that can be easily realized from among the data types of the target language for each piece of data. Therefore, when converting to a software description in the target language, it is possible to obtain a description that is efficient in terms of both execution speed and memory usage.

[Brief explanation of drawings]

FIG. 1 is a diagram showing the configuration of an embodiment of the present invention, FIG. 2 is a diagram showing the configuration of the optimal data type determining section in FIG. 1, FIG. 3 is a diagram showing an example of software description in an input language, FIG. 4 is a diagram showing an example of a data type conversion rule group, and FIG. 5 is a diagram showing the configuration of a conventional data type conversion method. 1... Input unit 2... Data definition group search unit 3... Data usage statement search unit 4... Optimal data type determination unit 5... Data definition statement Conversion unit 6... Data usage conversion unit 7... Output unit

Claims (1)

[Claims] (1) In software conversion, which inputs software written according to a certain input language and outputs a software description according to the target language that has the same meaning, the data type of the input language is changed to the data type of the target language. The conversion method includes: an input section for inputting a software description in an input language; a data definition sentence search section for searching for a sentence defining data from among the software input in the input section; a data usage statement search unit that searches from input software for a statement that uses data defined by the data definition statement searched by the data definition statement search unit; , with reference to the data usage form in the data usage sentence found by the data usage sentence search unit, and among the data types of the target language,
an optimal data type determination unit that determines the most suitable data type for the usage pattern; and a data definition statement that converts the data definition found by the data definition statement search unit into a data definition based on the data type determined by the optimal data type determination unit. The conversion unit and the data usage statement found by the data usage statement search unit are
a data usage statement conversion unit that converts the data usage method of the data type determined by the optimal data type determination unit; the data definition statement conversion unit converts the data definition statement; and the data usage statement conversion unit converts the data usage statement. A data type conversion method comprising: an output unit that outputs converted software;