JPH0736737A - Method for generating test data - Google Patents

Abstract

PURPOSE:To automatically generate test data adapted to a job and to improve the efficiency of a testing process. CONSTITUTION:A test data generator 104 refers to a data difinition document 101 using an inputted record specification name, extracts record definition information, then refers to a standard data item 102 and standardizes a data item name. Data are extracted from standard test data 103 storing the highly frequency used data for which the data items are highly related with a standard item name as a key and the test data are generated and outputted to a file 105.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is a test for streamlining the test process by automatically generating test data that is frequently used and highly related to data items in the test process of a software production support system. Regarding data creation method.

[0002]

2. Description of the Related Art In the test process of a software production support system, test data of various patterns are required. In order to improve the efficiency and reliability of the test process, it is especially important to easily and efficiently create a large amount of test data used in the test.

Therefore, various test data automatic generation methods have been conventionally proposed. For example, as a method of creating test data for test execution of a program, a data area contained in the program under test is read and analyzed, a data structure is created, and data is input based on the analysis result. There is a method of creating test data (see JP-A-58-169257).

According to the above-mentioned method, the control statement becomes unnecessary, the error in handling the control statement also disappears, and the created test data becomes the data with less contradiction with the program under test.

As another test data generation method, there is a method in which when a type, a search key, etc. are designated by a control statement, the test data of the required type is edited from the record of the master file and automatically generated (Japanese Patent Laid-Open No. Sho 61-96). 62-10526
No. 8). According to this method, data based on the actual work can be created more easily.

[0006]

However, in the above-mentioned first conventional technique, since the data area is referred to and the uniform data depending on the attribute of each data item is automatically generated, in many cases, Data that is different from the data set when the program actually runs will be generated.
Also, because the same data is generated for all items with the same attribute, it is not possible to test the program sufficiently, resulting in an increase in manual data correction work and a deterioration in the efficiency of the test process. There is a problem of doing.

Further, in the above-mentioned second conventional technique, when extracting data from an existing business data file, it is necessary to define control statements such as extraction conditions and key designation for retrieval, and the above method. Assumes that the business system database has already been constructed,
There was a problem that it could not be applied to new development work and the creation method could not be standardized.

Further, when editing the test data,
After accessing the test data file, expanding it on a dedicated screen, making corrections, and outputting it to the original data file, it is possible to use multiple test cases and multiple types of test data when necessary. There was also a problem that they could not respond sufficiently.

It is an object of the present invention to provide a test data generation method which automatically creates test data adapted to a job and improves the efficiency of the test process.

[0010]

In order to achieve the above object, in the present invention, a data item name extracted from a record specification of input test data is used in a test data creating method in a software production support system. Convert to a standard data item name, and use the standard data item name as a key to extract data from a standard test data file that stores data that is frequently used and highly related to other data items to generate test data. It is characterized in that the generated test data is corrected on a screen expanded in a table format and the corrected test data is output.

[0011]

In the test data generation device, the data definition is referenced by using the record specification name of the input test data, the record definition information is extracted, and then the standard data item is referenced, and the data item name is standardized. Convert to a data item name. Using standard item names as keys, data is extracted from standard test data that stores data that is frequently used and highly related to other data items, and generates test data, which is output to a file. The generated test data is corrected by the test data correction device, output to a file, and then printed in a form. As a result, test data that is frequently used and highly related to data items is automatically generated, and the productivity of the test process is improved.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be specifically described below with reference to the drawings. FIG. 1 shows a system configuration diagram of the present invention. In FIG. 1, 101 is a library in which a data definition document is stored, 102 is a standard data item correspondence table, 103 is a file in which standard test data is stored, 104 is a standard test data storage file in data item units. A test data generation device for generating test data from 105, a generated test data file, 1
Reference numeral 06 is a test data correction device that corrects the generated test data with a dedicated editor, automatically allocates it and outputs it to a file, 107 is a test data file, 108
Is a test data printing device for printing the test data in which a file is stored in a form, and 109 is a form.

The data definition library 101 stores a file specification which records file definition information and a record specification which defines a record layout of the file. FIG. 4 shows an example of the format of a record specification, in which data item information such as data item name 402, type 403, length 404, etc. is defined. In this embodiment, “N” of type 403 is Chinese character data, and “X”.
Indicates character data and “9” indicates numerical data. In addition, the data item name having a smaller level number indicates higher data. Therefore, the subordinate data of the "address" whose level number is 03 is the zip code, employee address,
It becomes a telephone number.

The standard data item correspondence table 102 is created as a table in which the data used in the project is investigated beforehand and the data item names are associated with the standard data item names. The data item name is an item that is generally used in business, etc.
Attribute name-Domain name. The entity name is a word indicating an entity to which the data belongs, the attribute name is a modifier connecting the entity and the data, and the domain name is a word indicating a range of the data. A standard data item is an item in which synonyms and derivative words are unified and the meaning of data can be recognized at a glance by a data name, and general data items are grouped by domain.

The standard test data file 103 is data represented by standard data item names, and is used frequently and stores data associated with each other.

FIG. 2 is a flowchart of the test data creation process of the present invention. Further, FIG. 3 is a functional block of the test data generating device. The test data generation operation of the present invention and then the test data correction / print operation will be described below.

<Generation of Test Data> From the terminal input device 301 of FIG. 3, the record specification name of the test data to be generated and the number of data items to be generated are input (step 20).
1). The record definition information extraction unit 302 refers to the data definition library 101 using the record specification name,
Extract the record definition information.

That is, as described above, since the data specification library 101 stores the record specification which defines the record layout of the file, FIG.
The record definition information such as the data item name (basic item), the data type, the data length, etc. is extracted from the record specification and stored in the record format table 303 (step 202).

FIG. 5 shows an example of the record format table. In this embodiment, focusing on the level number 401,
A data item name (basic item) 501 is extracted. The “internal information” 407, which is a repetitive item, is expanded as “internal information 1”, “internal information 2”, and “internal information 3” 504 by adding a sequence number to the end of the item name. The redefinition item 408 is not extracted.

As described above, by creating the record format table from the record specification, the information of the test data to be generated can be clarified.

Next, the standard name conversion processing section 304 refers to the standard data item correspondence table 102 to standardize the data item names (step 203). As described above, the standard data item correspondence table 102 is created in advance as a table in which data used in a project or the like is investigated and data item names are associated with standard data item names.

FIG. 6 shows an example of the standard data item correspondence table. The data item name 601 includes item names “owner name”, “employee name”, and “user name”. Although these entity names are different from “owner”, “employee”, and “user”, the domain names are all “name” and indicate the name of a person. Further, since the “name” also indicates the name of a person, the generated data are all the same. Therefore, all of these item names are unified into "name" and defined as standard data items. The same applies to the "charge" of the standard data item name. In the present embodiment, “employee name” 50 from the record format table 303
As a result of extracting the data item name “1”, the standard data item correspondence table 102 converts the data item name into a standard data item “name”.

The standard data extraction processing unit 305 extracts data from the standard test data file 103 for the number of data items input from the terminal input device 301, using the standardized data item name as a key (step 20).
4) The test data is output to the file 105 (step 205). The standard test data file 103 is data that the standard data item name means, and is frequently used.
In addition, data having a high degree of association between data items is stored in an external storage device.

FIG. 7 shows an example of standard test data.
As for the standard data item name 701 called "name", the data "Taro Mitachi", "Hanako Suzuki", and "Junichi Nakamura" which are names of frequently used people are created, and the standard data item 702 called "address" is created. For more information, please refer to “Shimoigusa, Suginami-ku, Tokyo × − × − ×” and “Kogane × − × − in Matsudo City, Chiba Prefecture.
“X”, “9 Chidan, Chiyoda-ku, Tokyo × − × − ×” is created. Furthermore, data is stored in a relational database so that each item has a relationship. As a result, when the “name” 701 is the name “Taro Tate”, the data “sex” 705 is “man”.

FIG. 8 shows the generated test data file 1
The contents of 05 are shown. In this embodiment, the standard data corresponding to all the data item names 501 shown in FIG. 5 are extracted by the number of data items input by the terminal input device 301, and are output to the test data file 105. Finally, the test data is generated in the same format as the record specification shown in FIG.

That is, the "employee name" 501, which is the basic item of the record specification, is converted into the standard data item name "name" by the standard data item correspondence table (FIG. 6), and the standard data file (FIG. 7). From "Name"
The standard data corresponding to is extracted for the number of generated items and output to a file (name data 801 in FIG. 8). "Postal code",
Standard data is extracted for all items of the record specification (Fig. 4) such as "employee address", "telephone number", and internal information 1 ", and is output to one file (Fig. 8) as test data.

<Correction / Printing of Test Data> When it is necessary to correct the generated test data file 105,
Correction is performed using a correction-dedicated screen in which the generated test data is expanded in the form of a form (step 207).

That is, FIG. 9 shows a test data editing screen. Since the generated data 902 is expanded in the form of a form for the item name 901, the data is edited interactively on this screen. When the employee name 901 requires data having the same surname, the generated test data cannot be used as it is. Therefore, in the test data (902) of the employee name on the correction screen (Fig. 9), "Hanako Suzuki" of the second record,
Correction is performed by directly overwriting "Junichi Nakamura" on the 3rd record with "Taro Tate".

After modification, the data definition library 10
According to the file definition information of the file specification stored in 1, the corrected data is output to the automatically allocated file 107. (Step 208). At this time, a plurality of types of test data files corresponding to various test cases can be created using the generated test data file as a master. The test data file is printed by the printing device 108 in a form (step 210).

[0030]

As described above, according to the present invention, since test data that is frequently used and highly related to data items is automatically generated, the productivity of the test process is improved and The reliability can be improved.

[Brief description of drawings]

FIG. 1 shows a system configuration diagram of the present invention.

FIG. 2 is a processing flowchart for creating test data according to the present invention.

FIG. 3 is a functional block of a test data generation device.

FIG. 4 shows a format example of a record specification.

FIG. 5 shows an example of a record format table.

FIG. 6 shows an example of a standard data item correspondence table.

FIG. 7 shows an example of standard test data.

FIG. 8 shows the contents of a generated test data file.

FIG. 9 shows a test data edit screen.

[Explanation of symbols]

 101 Data Definition Document Library 102 Standard Data Item Correspondence Table 103 Standard Test Data Storage File 104 Test Data Generation Device 105 Test Data File 106 Test Data Correction Device 107 Test Data File 108 Test Data Printing Device 109 Form

Claims (1)

[Claims] 1. In a test data creation method in a software production support system, a data item name extracted from a record specification of input test data is converted into a standard data item name, and the standard data item name is used as a key. Data is extracted from a standard test data file that stores data that is frequently used and that has a high degree of association between data items, and test data is generated, and the generated test data is expanded in a table format on a screen. A method for creating test data, which comprises modifying and outputting the modified test data.