JPH0954704A - Automatic test program generating system for language processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain comprehensibility for all program syntax and functions by generating a test item not depending on the program syntax and generating a test program according to the test item. SOLUTION: A test item generation processing part 10 generates the test item not depending on the program syntax for the range of number of structure and number of nests in one test program based on the input of an operator. Then, comparative evaluation is performed with the test item generated by a test item evaluation processing part 11, and the duplication of syntax of the test item is excluded by eliminating the same syntax. Also, a program constituent sampling part 12 samples program constituents such as structure, a text, an equation and an element, etc., from the program syntax of a language processor set as a test target. Then, a test program generating part 13 automatically generates the test program based on the program constituent and the test item from which duplication is eliminated. In this way, the test of the language processor is performed with high efficiency.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a language processor (Lang
TECHNICAL FIELD The present invention relates to a technique for creating a test program for a language processor, and more particularly to a test program automatic creation system for a language processor suitable for efficiently creating a test program corresponding to the language specification of each language processor.

[0002]

2. Description of the Related Art Language processors each have a unique language specification (program syntax). A source program written in accordance with the program syntax is input to perform syntax analysis, and the analysis information is used inside the language processor. Convert to the internal intermediate language to be used and generate the target program that is the execution format. Also, most language processors have an optimization processing function that focuses on the constituent elements of the program such as expressions and statements extracted from the program syntax, loop structure, etc. in order to increase the execution speed of the target program, Further, some of them have a vectorization processing function for applying a vectorization instruction to a loop.

The language processor executes various different processes depending on the pattern of analysis information obtained by the syntax analysis of the source program. Therefore, when testing these language processors, it is necessary to test the attributes of sentences and data, and the combinations of sentences (sentences) (program patterns), and to artificially create such test items. Then, it is necessary to create a test program based on the test item.

In particular, in the optimization processing function and the vectorization processing function, there are a huge number of combinations of test conditions such as a combination of sentences and the appearance positions of variables and expressions, and thus the test items and test programs are artificially changed. It is impossible to create it automatically, and automatic creation and automatic evaluation by computer are necessary. Further, as the language specifications become more complicated, it is difficult to understand the program syntax and artificially create and evaluate the inspection items in accordance with this.

A technique for automatically creating such a test item or a test program, or both a test item and a test program as described in, for example, Japanese Patent Application No. 4-145239 by the inventors of the present invention. A technique for automatically creating is known. As an automatic test program creation technology, the attributes of the data items to be system tested and the test item description that describes the basic instructions for obtaining a certain test execution result are input and analyzed as parameters, and the data item It is known to create a test program of basic instructions using all combinations and a job control language for executing the test program. It also automatically merges multiple test programs,
A technique of creating a plurality of new test programs by combining them is known.

In the technique for automatically creating both test items and test programs (Japanese Patent Application No. 4-145239), first, data is previously stored in each file of structure data, sentence data, formula data, and element data. A program that is required as a test item for the optimization function or vectorization function by inputting the specification of the combination method corresponding to the optimization function or vectorization function of the language processor of the test conditions described using the description parameters. A test item is created by combining all of the conditions, the statement in the program, the expression, the type of element, and the condition of the appearance position (test item generation process). Next, the created test items are compared and evaluated with each other so that there is no duplication (test item comparison / extraction processing). Then, the remaining test items are input to create a test program suitable for these test items (test program generation process). In this way, a test item and a test program corresponding to the test item can be automatically created.

However, these conventional techniques create test items that strongly depend on language specifications and automatically create a test program based on the test items. Therefore, it is impossible to create a complete test program for all program syntaxes. You can also create test items and test programs that focus on certain specific functions, such as test items and test programs for basic instructions, or test items and test programs for optimization processing functions and vectorization processing functions. It is not possible to create test items and test programs that are comprehensive about the functions of.

[0008]

The problem to be solved is that the conventional technique cannot create a test item that does not depend on the program syntax. An object of the present invention is to solve these problems of the prior art, to create a test program that is comprehensive for all program syntaxes and functions, and to improve the efficiency of language processor testing. It is to provide an automatic program creation system.

[0009]

In order to achieve the above object, an automatic test program creation system for a language processor according to the present invention provides a "range of the number of structures in one test program and a range based on an input from an operator. "Nest number range", "Sentence number range in one structure", "Expression number range in one sentence", "Number of elements in one expression", etc.
A test item creation processing unit 10 that creates test items that do not depend on language specifications, a test item evaluation processing unit 11 that removes duplicate test items from the test items created by the test item creation processing unit 10, and a language to be tested. From the program syntax of the processor, at least the structure, statement,
A program component extraction processing unit 12 for extracting program components including expressions and elements, a program component extracted by the program component extraction processing unit 12, and a test item evaluation process created by the test item creation processing unit 10. At least a test program creation processing unit 13 that creates a test program based on the test items from which duplication has been removed by the unit 11 is provided.

[0010]

According to the present invention, first, based on the test policy input by the operator, the range of the number of structures and the range of the number of nests in one test program, the range of the number of statements in one structure, and the range of the number of statements in one structure Create test items that do not depend on the program syntax (language specification), such as the range of the number of expressions and the range of the number of elements in one expression. Next, the created test items are compared and evaluated to see if they are the same, and the same items are excluded to prevent duplication of the test items. In addition, from the program syntax of the language processor to be tested, the structure,
Extracts program components such as sentences, expressions, and elements, and transforms them into a structure suitable for computer processing. Then, a test program is automatically created on the basis of the converted program components and the duplicated test items.
As described above, since the test item that does not depend on the program syntax is created and the test program is created according to this test item, the test program is complete for all program syntaxes and functions.

[0011]

Embodiments of the present invention will now be described in detail with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of a configuration according to the present invention of an automatic test program creation system for a language processor of the present invention. In FIG. 1, 1 is CP
An information processing device 2 having a U (Central Processing Unit) and a main memory and performing various processes including a test program creation according to the present invention by a storage program method, a keyboard 2 and a CRT (Cathode Ray Tu).
be, a cathode ray tube) and the like, and a data terminal 3 including a magnetic disk and the like.

The information processing apparatus 1 includes a test item creation processing unit 10, a test item evaluation processing unit 11, a program syntax element extraction unit 12, and a test program creation processing unit 13 for automatically creating a test program for a language processor according to the present invention. Is provided. In the storage device 3, a test item file 30, a test item creation history file 31, which is referred to or created by each processing unit of the information processing device 1,
A test item master file 32, a program syntax description file 33, a program component file 34, and a test program file 35 are stored.

The operation of creating a test program for a language processor by the test program automatic creation system having the above-described configuration will be described below. First, the test program creator uses the data terminal 2 based on the test policy, for example, the number of lines of the entire test program to be newly automatically created, the range of the number of structures in one test program, and the range of the number of nests. Range of number of sentences in one structure, 1
When data that does not depend on the program syntax (language specification) such as the range of the number of expressions in one sentence and the range of the number of elements in one expression is input as an input parameter, the information processing device 1
2 creates the test item file 30 having the details shown in FIG. 2 by the process shown in FIG. 6 by the test item creation processing unit 10.

That is, the test item creation processing unit 10
In the process of FIG. 6, first, the input parameters are analyzed (step 100). Next, based on the analysis result,
Determination of range of number of structures and nest number in one program (step 101), determination of range of number of statements in one structure (step 102), determination of range of number of expressions in one sentence (step) 103) The range of the number of element data in one formula is determined (step 104), and the number of lines (the number of steps) of the program is determined (step 105). Then, the combination test items are automatically created by combining the determined number ranges (step 106), and the created combination test items are output to the test item file (step 107).

As a result of this processing, as shown in FIG. 2, each item number (1, 2 to 10) is stored in the test item file 30.
Test item data is registered separately. For example, as the item number “1”, the number of steps is “100”, the range of the number of structures in one program is “3 or more and less than 5”, the range of the nest number is “10 or more and less than 20”, and the number of statements in one structure is Combination test items of number range “5 or more and less than 15”, number range of expressions in one sentence “3 or more and less than 20”, and range of number of elements in one expression “2 or more and less than 4” are registered. To be done.

Although the test items registered in the test item file 30 do not depend on the program syntax, these test items are redundantly created by the test item creation processing unit 10 in FIG. There is a possibility that However, since these test items are expressed by a certain rule, the number of structures, the number of sentences, the number of expressions, etc.
It is possible to easily check whether or not the same as the newly created one exists in the test item creation history file 31 in FIG. 1, and it is possible to evaluate whether or not the same. Therefore, the test item evaluation processing unit 11 of FIG. 1 performs the comparative evaluation process shown in detail in FIG. 7 to avoid such duplication. As a result, the test item master file 32 in FIG. 1 is obtained.

That is, the test item evaluation processing unit 1 of FIG.
1 inputs the test item data created by the test item creation processing unit 10 of FIG. 1 in the process of FIG. 7 (step 200), and repeats the following processes up to the last test item. First, the test item creation history file 3 in FIG.
It is compared with the registered content of 1 (step 202) and it is determined whether or not the same item exists (step 203). If it does not exist, it is judged as a new test item and the test item creation history file 31 of FIG. 1 is updated (step 20).
4), output to the test item master file 32 of FIG. 1 (step 205). If the same item exists in the process of step 203, the process returns to step 200 to input the next test item data.

When the processing of the test item evaluation processing unit 11 of FIG. 1 is completed in this way, the information processing apparatus 1 of FIG.
Activates the program component extraction processing unit 12 of FIG. 1 and performs the processing shown in detail in FIG. That is, the program component extraction processing unit 12 of FIG. 1 is a test target language processor registered in the program syntax description file 33 of FIG. 1 (a specific example thereof is shown in FIG. 3) in the process of FIG. A component is input from the program syntax (step 300), and the following processing is repeated until the last component input.

The element type of the input component is "element" (step 302), "expression" (step 303), "sentence" (step 304), or "structure". Or (step 305) is determined in order,
The constituent elements are registered corresponding to each element type (steps 306 to 309). If the answer is NO in step 305, "reserved word (spelling in which meaning and usage are predetermined in each programming language and cannot be used by the user for unique variable name)" Registration is performed (step 310). Then, when the processing for the last component is completed, the registered program component is changed to the program component file 3 in FIG.
4 (step 311), and the process ends.

FIG. 3 is an explanatory diagram showing a description example of the program syntax description file in FIG. This example shows an example of the data file structure of the program syntax of the language processor to be tested, and the program syntax of the language processor to be tested is all described in the program syntax description file 33 of FIG. 1 according to this description rule.

FIG. 4 is an explanatory diagram showing a description example of a table referred to by the program component extraction processing unit in FIG.
This figure shows a table for explaining the meaning of each data description parameter used in the program syntax description file in FIGS. In this example, there are five types of data description parameters, "|", "'", "[]", "{!}", And ";", and the program component extraction processing unit 12 of FIG.
Referring to the meanings of these data description parameters, the contents of the program configuration description shown in FIG.
Convert to a program component file with the data structure shown in.

FIG. 5 is an explanatory diagram showing a description example of the program component file in FIG. This figure shows a data configuration file of program components extracted from the program syntax description file 33 shown in FIG. 3 by the program component extraction processing unit 12 of FIG. , Etc. are stored.

Based on the program components extracted from the program syntax in this manner and the test items created and evaluated in the processing in FIGS. 6 and 7, the test program creation processing unit 13 in FIG. The test program creation process shown in FIG. 1 is performed to automatically create the test program, and the test program file 35 in FIG. 1 is obtained.

That is, the test program creation processing unit 13 of FIG. 1 inputs test items from the test item master file 32 of FIG. 1 in the process of FIG. 9 (step 400), and executes the following process as the final test. Repeat until item input (step 401). First, the structure is determined (step 402), and then the sentences, expressions, and elements are determined (steps 403-405). Next, a variable name is assigned (step 406), a variable declaration statement is generated (step 407), and a result output statement is generated (step 4).
08), and the test program file 35 of FIG.
(Step 409).

As described above with reference to FIGS. 1 to 9, in the test program automatic generation system for the language processor of this embodiment, first, based on the input from the operator,
"Range of number of structures in one test program" and "range of number of nests", "Range of number of statements in one structure",
Create test items that do not depend on the program syntax (language specification) such as "range of the number of expressions in one sentence" and "range of the number of elements in one expression". Perform comparison / evaluation and register test items without duplication.
In this way, by making the test items independent of the program syntax, it is possible to have completeness for all program syntaxes, and to ensure that there are no duplicated tests from the basic items of the language processor test to the combination items. Items can be automatically created easily.

Further, by inputting / analyzing a program syntax registered in advance, program constituent elements such as structures, sentences, expressions and elements are extracted. Then, a test program is created based on the created / evaluated and registered test items and the extracted program components. This makes it possible to create a test program that is comprehensive for all program syntaxes. The present invention is shown in FIGS.
The embodiment is not limited to the embodiment described with reference to FIG. 9, and various modifications can be made without departing from the spirit of the invention.

[0027]

According to the present invention, a test item that does not depend on the program syntax can be created, a test program that is complete for all program syntaxes and functions can be created, and the test of the language processor can be improved. Efficiency can be improved.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of a configuration according to the present invention of a language processor test program automatic generation system of the present invention.

FIG. 2 is an explanatory diagram showing a data configuration example of a test item file in FIG.

FIG. 3 is an explanatory diagram showing a description example of a program syntax description file in FIG.

4 is an explanatory diagram showing a description example of a table referred to by a program component extraction processing unit in FIG. 1. FIG.

5 is an explanatory diagram showing a description example of a program component file in FIG. 1. FIG.

FIG. 6 is a flowchart showing an example of the processing operation according to the present invention of the test item creation processing unit in FIG.

7 is a flowchart showing an example of the processing operation according to the present invention of the test item evaluation processing unit in FIG.

FIG. 8 is a flowchart showing an example of the processing operation according to the present invention of the program component extraction processing unit in FIG.

9 is a flowchart showing an example of the processing operation according to the present invention of the test program creation processing unit in FIG.

[Explanation of symbols]

1: Information processing device, 2: Data terminal, 3: Storage device, 10: Test item creation processing unit, 11: Test item evaluation processing unit, 12: Program component extraction processing unit, 1
3: test program creation processing unit, 30: test item file, 31: test item creation history file, 32: test item master file, 33: program syntax description file, 34: program component file, 35: test program file.

Claims (1)

[Claims] 1. A system for automatically creating a test program for a language processor, comprising: a test item creating means for creating a test item independent of a language specification based on an input from an operator; and a test item creating means for creating the test item. A test item evaluation means for removing duplicate test items from the tested test item, and a program component extraction means for extracting a program component including at least a structure, a sentence, an expression, and an element from the program syntax of the language processor to be tested, An automatic test program creation system for a language processor, comprising at least a test program creation means for creating a test program based on the extracted program components and the test items created and removed of the duplicated test items.