JP2010250583A - Program and method for generating application test, and application test device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reduce trouble of application test generation. <P>SOLUTION: An instruction string division means 11 reads an instruction string 21, and divides the instruction string by a division instruction. A communication log obtaining/classification means 12 reproduces the divided division instruction string, and outputs to a user agent 40 an operation instruction. The communication log obtaining/classification means 12 obtains a request transmitted by the user agent 40 according to the operation instruction and a communication log of a response to the request, and classifies them in each division instruction string. An automatic test generation means 13 generates a test nucleus changing the request of a target based on the classified communication log, and generates an automatic test 24 wherein the instruction string is combined. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an application test generation program, an application test generation method, and an application test apparatus, and more particularly to an application test generation program, an application test generation method, and an application test apparatus for testing an application apparatus that provides a predetermined service via a network.

  2. Description of the Related Art In recent years, usage forms that connect to application devices via a network and receive predetermined services have become widespread. For example, a Web application provided by a Web server can be used by operating a Web browser of a client computer. In order to check the operation of such an application, it is necessary to repeat the test by changing a part of the operation or data. For this reason, the work is enormous, and test automation has been studied.

  By the way, the application test procedure includes a process of creating a situation in which a test target request is normally processed, and a process of modifying the test target request to meet the purpose of the test and verifying a response. In the following description, the former work is called “reproduction”, and the latter work is called “test nucleus”.

  As an automatic test method suitable for reproduction, for example, there is a method called a browser operation method for automating operations on a Web browser. As a browser operation method, Selenium RC (HYPERLINK "http://selenium-rc.openqa.org/"http://selenium-rc.openqa.org/) is well known. There has also been proposed a method in which a test is automatically generated by changing a part of an operation or a part of input data based on an operation history of a Web browser and repeatedly executed (see, for example, Patent Document 1).

  On the other hand, as an automatic test method suitable for a test core, for example, there is a method called a communication preemption method that intercepts a request / response during communication and modifies an arbitrary part of the request. In the communication interception method, a request transmitted to a web server is intercepted in response to an operation of the web browser, an arbitrary part is modified according to a predetermined scenario, and the modified request is transmitted to the web server. . The response from the Web server that processed this request is intercepted and verified. Requests can be altered interactively manually. Further, since some requests include an argument to be delivered to the server device, a method has been proposed in which this is detected and incorporated as a parameter in a scenario of modification processing (see, for example, Patent Document 2).

JP 2005-266554 A JP 2007-264967 A

However, the conventional application test has a problem that it is not easy to automate test generation.
As described above, the application test includes a reproduction that creates a situation in which a request to be tested is normally processed, and a test kernel that modifies the request to be tested to suit the purpose of the test.

  Browser operation methods suitable for reproduction can automate much of what a browser user can do. However, this method cannot test items that cannot be operated from the browser, so it is not suitable for the test core. For example, modification of requests that occur asynchronously such as Ajax (Asynchronous JavaScript + XML) and verification of their responses, modification of requests that are dynamically generated by JavaScript (registered trademark), modification of form data that is not displayed on the browser I can't.

On the other hand, the communication interception method suitable for the test core is not very suitable for reproduction because the ability to simulate the operation of the browser user is low.
As described above, the browser operation method is suitable for reproduction and the communication interception method is suitable for a test core, respectively. In order to automate a good application test, it is necessary to combine them. However, it was not easy to combine these to create a test. In the communication interception method, there is no mechanism for intercepting only the request / response to be tested. This is because it is difficult to determine which browser operation the request corresponds to only from the request. Therefore, even when a request / response corresponding to a certain browser operation is to be a test target, a request / response other than the test target is also intercepted. This can prevent the entire test from being executed correctly.

  As an example, a case will be described in which an application test is performed by modifying a request transmitted when a certain operation is performed on a test target input screen that is displayed after the web browser is operated several times. For convenience, R1, R2, and R3 are the requests and responses communicated by the operation of the Web browser up to the test target input screen, and R4 is the request and response that is the test target. In the application test, it is desirable to perform control such that R1, R2, and R3 are verified by reproduction without modification, and only R4 to be tested is verified by a test nucleus that modifies the request. However, as described above, the browser operation method can reproduce R1, R2, and R3, but cannot change the request of R4. On the other hand, in the communication interception method, requests corresponding to R1, R2, R3, and R4 cannot be identified. For this reason, it is difficult to modify the request only for R4.

  Thus, since the reproduction and the test core are performed in different ways, it is not easy to automatically generate an application test that satisfies both functions. For this reason, generation and execution control of the application test must be performed manually, requiring a great deal of labor and time.

  In view of these points, it is an object of the present invention to provide an application test generation program, an application test generation method, and an application test apparatus capable of automating application test generation and reducing the labor of test generation.

  In order to solve the above problems, an application test generation program for testing an application device that provides a predetermined service via a network is provided. The application test generation program causes a computer to function as an instruction sequence dividing unit, a communication log acquisition / classification unit, and an automatic test generation unit. The instruction sequence dividing unit includes an instruction for instructing a user agent receiving a predetermined service to execute a predetermined process including a process according to transmission of a test target request to the application apparatus and reception of a response to the request. Get the instruction sequence. Then, a delimiter instruction in which the process represented by the instruction sequence is delimited is detected from the instruction sequence, and the instruction sequence is divided for each delimiter instruction to generate a divided instruction sequence. The communication log obtaining / dividing unit reproduces and outputs an instruction to the user agent based on the instruction sequence included in the divided instruction sequence for each divided instruction sequence. Then, the request transmitted by the user agent and the response transmitted from the application apparatus in response to the request are acquired as a communication log. This communication log is stored in the storage means in association with the divided instruction sequence reproduced at the time of acquiring the communication log, and a log classification table in which the communication log is classified for each divided instruction sequence is generated. The automatic test generation means extracts the communication log associated with the divided instruction sequence including the instruction sequence that generated the test target request from the log classification table. Then, based on the extracted communication log, a test nucleus describing a test procedure including request modification is generated, and an automatic test for causing the divided instruction sequence and the test nucleus to be executed in cooperation is generated.

  According to such an application test generation program, the computer is caused to function as an application test generation device. A user agent receiving provision of a predetermined service transmits a request for requesting a service to the application apparatus, obtains a response, and performs predetermined processing. The application test generation apparatus acquires an instruction sequence including instructions for instructing the user agent to execute a predetermined process including a transmission of a request to be tested to the application apparatus and a response process. Then, the instruction sequence is divided by a delimiter instruction included in the instruction sequence to generate a divided instruction sequence. Next, for each divided instruction sequence, the divided instruction sequence is reproduced and output to the user agent to execute a predetermined process. At this time, the request transmitted by the user agent and the response to the request are acquired as a communication log, and stored in the storage unit in association with the reproduced divided instruction sequence. This procedure is repeated for each divided instruction sequence to generate a log classification table in which the communication log is classified for each divided instruction sequence. Next, based on the log classification table, a communication log including a request to be tested is extracted, and a test nucleus describing a test procedure including request modification is generated. The test kernel and the divided instruction sequence generated in this way are combined to generate an automatic test for executing the divided instruction sequence and the test nucleus in cooperation.

  In order to solve the above problems, an application test generation method for causing a computer to execute a processing procedure realized by the above processing function and an application test apparatus for executing an application test are provided.

  According to the disclosed application test generation program, application test generation method, and application test apparatus, it is possible to automatically generate an application test capable of executing a reproduction test and a test core test in cooperation with each other. This can reduce the effort required to create an application test.

It is the figure which showed the outline | summary of invention. It is the figure which showed the hardware constitutions of the system of embodiment. It is the figure which showed the software structure of the test production | generation apparatus. It is the figure which showed an example of the command sequence (Selenese). It is the figure which showed an example of the instruction sequence division table. It is the figure which showed an example of the log group (d1-d3). It is the figure which showed an example of the log group (d4-d9). It is the figure which showed an example of the log classification table. It is the flowchart which showed the procedure of the application test production | generation process. It is the flowchart which showed the procedure of instruction sequence division processing. It is the flowchart which showed the procedure of the communication log acquisition and classification process. It is the figure which showed the structure of the automatic test apparatus.

Hereinafter, embodiments will be described with reference to the drawings. First, an outline of the invention will be described, and then specific contents will be described.
FIG. 1 is a diagram showing an outline of the invention.

  The test generation apparatus 10 is connected to a server 60 that operates as an application apparatus that provides a predetermined service via a network. Then, an application test for the application program of the server 60 is generated.

  Here, the relationship between the user agent 40 and the server 60 will be described. When receiving an instruction from the user, the user agent 40 transmits a request corresponding to the instruction to the server 60. The server 60 executes processing corresponding to the request and returns a response. Then, the user agent 40 provides a service based on the response obtained as a response to the user. For example, the user agent 40 displays operation buttons on the display screen. When the user operates and operates the operation button, a request for requesting information corresponding to the operation button is transmitted to the server 60. When a response including the requested information is sent from the server 60, a display based on the response is performed and the information is provided to the user.

  The communication log acquisition device 50 relays requests and responses exchanged between the user agent 40 and the server 60, and performs processing for leaving the passed requests and responses as communication logs.

  The test generation apparatus 10 includes processing units such as an instruction sequence division unit 11, a communication log acquisition / classification unit 12, and an automatic test generation unit 13, and a storage unit 14, and generates an application test for the server 60. Each processing means realizes its processing function by a computer executing an application test generation program.

  The instruction sequence dividing means 11 acquires the instruction sequence 21 and divides the instruction sequence with a predetermined delimiter instruction. In the instruction sequence 21, instructions instructing to output to the user agent 40 an operation instruction similar to a user instruction are described in the order of execution. For example, an operation command such as “click on the first link displayed on the display screen” is described for the user agent 40. By reproducing an operation command for the user agent 40 based on the command sequence 21, much of what the user can do can be automated. The instruction sequence dividing means 11 analyzes the instruction described in the instruction sequence 21 and detects a delimiter instruction that delimits an operation instruction reproduced from the instruction sequence. The separation command may be set in advance by the user. For example, with respect to an operation of clicking on an icon for which a link is set on the display screen of the user agent 40 to open a page, “click icon”, “wait for page to load”, “page requested It is assumed that an instruction sequence “check whether it is a thing” is described. An instruction sequence in which processing is delimited is detected in this instruction and is set as a delimiter instruction. In this case, when a series of operation instructions “click on icon” and “wait for page to be read” are executed by the agent 40, it is estimated that a request is transmitted and a response to the request is completed. Therefore, “wait for the page to be read” or “check whether the page is requested” is set as a delimiter instruction. This can be paraphrased as an instruction indicating that the state has been stabilized after the user agent 40 makes a request and receives a response. The instruction sequence dividing means 11 collates the character pattern representing the designated delimiter instruction with the character pattern of each instruction sequence in the instruction sequence 21, and detects a match from the instruction sequence. Then, the instruction sequence is divided for each delimiter instruction to generate a divided instruction sequence. The location of the division is registered in the instruction sequence division table 22.

  The communication log acquisition / classification unit 12 reproduces the divided instruction sequence and outputs an operation instruction to the user agent 40. Then, the communication log at this time is classified for each divided instruction sequence. The request transmitted from the user agent 40 when executing the process according to the operation command and the response from the server 60 that has performed the predetermined application process in response to the request are recorded as a communication log by the communication log acquisition device 50. The The communication log acquisition / classification unit 12 reproduces the divided instruction sequence, and stores the communication log acquired from the communication log acquisition device 50 at this time in the storage unit 14 in association with the reproduced divided instruction sequence. By repeating this for each divided instruction sequence, a communication log classified for each divided instruction sequence is obtained. Specifically, the communication log is associated with the divided instruction sequence reproduced at this time, and the correspondence is registered in the log classification table 23. In the above example, when the instruction “Check whether the page requested” is a delimiter instruction and the split instruction sequence is executed, the request issued in the process from “click icon” to the delimiter instruction and the response are Recorded as a communication log. In this way, the correspondence relationship between the divided instruction sequence and the communication log corresponding thereto is registered in the log classification table 23.

  Based on the log classification table 23, the automatic test generation unit 13 extracts a communication log corresponding to a test target request for generating a test nucleus. For example, when one test kernel is generated for each form parameter of the request included in the communication log, the request to be tested is a request including the form parameter. Then, based on the extracted communication log, a test kernel describing a test procedure including request modification is generated. For example, for each form parameter of the request, a procedure for changing the value to an invalid value is generated. Further, the test kernel is combined with the divided instruction sequence, and an automatic test is generated for executing the divided instruction sequence and the test nucleus in cooperation.

  The storage unit 14 stores information such as an instruction sequence division table 22 that indicates the division of the instruction sequence divided by the instruction sequence division unit 11 and a log classification table 23 that associates the divided instruction sequence with the communication log. .

An operation of the test generation apparatus 10 having such a configuration and an application test generation method will be described.
When the process is started, the test generation apparatus 10 acquires an instruction sequence 21 that instructs the user agent 40 to output an operation instruction. The instruction sequence dividing unit 11 collates the delimiter instruction with the instruction in the instruction sequence, and detects a matching delimiter instruction. Then, the instruction string is divided by the detected delimiter instruction. As the delimiter instruction, an instruction for delimiting the corresponding processing of the server 60, for example, an instruction indicating that the state has been stabilized after the user agent 40 makes a request and receives a response is set in advance. Information relating to the divided instruction sequence is stored in the storage unit 14 as the instruction sequence division table 22.

  The communication log acquisition / classification unit 12 reproduces the instruction sequence in units of the divided instruction sequence based on the instruction sequence division table 22 and outputs an operation instruction to the user agent 40. The user agent 40 transmits a request to the server 60 according to the operation command. The server 60 performs application processing according to the request and returns a response. The request and response at this time are recorded as a communication log by the communication log acquisition device 50. The communication log acquisition / classification unit 12 acquires the communication log at the time of reproduction of the divided instruction sequence from the communication log acquisition device 50, associates it with the divided instruction sequence, and registers the correspondence relationship in the log classification table 23. For each divided instruction sequence, reproduction of the divided instruction sequence, acquisition of the communication log at that time, and registration in the log classification table 23 result in the log classification table 23 displaying the communication log classified for each divided instruction sequence. Is registered.

  The automatic test generation means 13 extracts a communication log corresponding to a test target request for generating a test nucleus based on the log classification table 23. Then, based on the extracted communication log, a test nucleus instructing a test procedure including request modification is generated. The test kernel is combined with the divided instruction sequence, and an automatic test 24 is generated to execute the divided instruction sequence and the test nucleus in cooperation. For example, when a divided instruction sequence including a request for which a test nucleus is generated is executed, the test nucleus is associated with the execution. The test kernel is not executed in the divided instruction sequence in which the test kernel is not generated in the request.

  As a result, an automatic test 24 that combines the reproduction and the test kernel is generated. When the automatic test 24 is executed, the instruction sequence set in the instruction sequence 21 is sequentially reproduced in divided instruction sequence units. At this time, if a test kernel is not associated with the instruction sequence, only reproduction is executed. That is, only the operation of instructing the user agent 40 based on the instruction sequence and automatically reproducing the user's operation is performed. If a test nucleus is associated with the instruction sequence, the test nucleus is executed with reproduction. For example, a test target request transmitted by the user agent 40 according to an operation command reproduced by reproduction can be intercepted, and the intercepted request can be modified with a test kernel and then transmitted to the server 60.

  As described above, according to the test generation device 10, it is possible to automatically generate an application test capable of cooperating the reproduction and the test nucleus. In the test using the generated application test, it is possible to check the operation of the server 60 by causing the test kernel to be executed only when a request to be tested occurs.

Hereinafter, an embodiment will be described by taking application test generation in which the user agent 40 is a Web browser and the server 60 is a Web application as an example.
FIG. 2 is a diagram illustrating a hardware configuration of the system according to the embodiment.

The test generation apparatus 100 is connected via a network 700 to a server 600 that mounts an application to be tested.
A hardware configuration of the test generation apparatus 100 will be described. The entire test generation apparatus 100 is controlled by a CPU (Central Processing Unit) 101. A random access memory (RAM) 102, a hard disk drive (HDD) 103, a graphic processing device 104, an input interface 105, and a communication interface 106 are connected to the CPU 101 via a bus 107.

  The RAM 102 temporarily stores at least part of an OS (Operating System) program and application programs to be executed by the CPU 101. The RAM 102 stores various data necessary for processing by the CPU 101. The HDD 103 stores the OS and application programs. A monitor 108 is connected to the graphic processing device 104, and an image is displayed on the screen of the monitor 108 in accordance with a command from the CPU 101. A keyboard 109 a and a mouse 109 b are connected to the input interface 105, and signals transmitted from the keyboard 109 a and the mouse 109 b are transmitted to the CPU 101 via the bus 107. The communication interface 106 is connected to the network 700 and transmits / receives data to / from the server 600 via the network 700.

With such a hardware configuration, the processing function of the test generation apparatus 100 can be realized.
Next, the software configuration of the test generation apparatus 100 will be described. FIG. 3 is a diagram illustrating a software configuration of the test generation apparatus.

  The test generation apparatus 100 includes a test generation unit 110, a Web browser (hereinafter referred to as a browser) 120, a browser operation unit 130, a log acquisition unit 140, and a storage unit 150, and is connected to the server 600 via a network.

  The test generation unit 110 controls the browser operation unit 130 to automatically reproduce the operation on the browser 120 according to the instruction sequence. In addition, the log acquisition unit 140 is controlled to acquire a communication log between the browser 120 and the server 600. Then, an application test for executing the acquired communication log and the instruction sequence in cooperation with each other is automatically generated. Details will be described later.

  The browser 120 downloads an HTML (Hyper Text Markup Language) file, an image file, a music file, and the like provided by the server 600, and reproduces and outputs the file to the monitor 108 or an audio output device. A request is transmitted to the server 600 in response to an operation input on the display screen, etc., and the layout is analyzed from the response from the server 600 and displayed / reproduced.

  The browser operation unit 130 reproduces a command sequence for the browser 120 and automatically generates an operation command for the browser 120. Here, it is assumed that “Selenium RC” is used as a test tool for performing an automatic test of the browser operation method.

  The log acquisition unit 140 is installed at a relay point that relays communication between the browser 120 and the server 600, intercepts the communication, and records it as a communication log. For example, it is implemented as an HTTP (Hyper Text Transfer Protocol) proxy server.

The storage unit 150 stores information such as the instruction sequence division table 22 and the log classification table 23 generated by the test generation unit 110.
The server 600 is a Web server, stores information such as HTML files and image files, and executes application programs in response to requests from the browser 120 of the user agent. The execution result is returned to the browser 120 as a response.

Here, details of the test generation unit 110 will be described.
The instruction sequence dividing unit 111 analyzes an instruction sequence for automating operations on the browser 120 and divides the instruction sequence using an instruction indicating that the state of the browser 120 is stable as a delimiter instruction. An instruction sequence division table 22 is generated for the divided instruction sequence and stored in the storage unit 150.

  The communication log acquisition / classification unit 112 causes the browser operation unit (Selenium RC) 130 to reproduce the instruction sequence based on the instruction sequence partitioning table 22. Further, the log acquisition unit 140 is controlled, and at this time, a request and a response exchanged between the browser 120 and the server 600 are acquired. And the communication log which recorded the request and response which the log acquisition part 140 acquired is acquired, and it classify | categorizes for every division | segmentation command sequence. Here, the divided instruction sequence is executed, and the communication log acquired at that time is classified into this divided instruction sequence. This process is repeated for each divided instruction sequence, and the communication log of the entire instruction sequence is classified for each divided instruction sequence.

The test kernel generation unit 113 generates one test kernel for each form parameter of each request of the communication log acquired by the log acquisition unit 140.
The automatic test generation unit 114 generates an automatic test that operates the test nucleus generated by the test nucleus generation unit 113 and the instruction sequence for reproduction in cooperation with each other.

The operation of the test generation apparatus 100 and the application test generation method will be described in detail using specific examples.
The instruction sequence executed by the browser operation unit (Selenium RC) 130 is input to the instruction sequence dividing unit 111. The instruction sequence is described in, for example, a Selenium format that is a Selenium RC instruction sequence notation format.

FIG. 4 is a diagram illustrating an example of an instruction sequence (Series).
In the instruction sequence (Selenese) 1000, an operation on the browser reproduced by the browser operation unit (Selenium RC) 130 is described in the Selene format. The Series format is composed of a set of three elements. Specifically, a command (command) 1001, a target 1002 (target), and a value (value) 1003 form one set and constitute an instruction sequence. A command 1001 represents a command that instructs the browser operation unit (Selenium RC) 130 to move. As for Selenium including details of instructions, for example, “Thorough Introduction to Selenium” (Shinya Kasaya, Shingo Miyazawa, Yuichiro Masui, WEB + DB PRESS Editing Department, “WEB + DB PRESS” Vol. 36, Technical Review, 2006, page 41) It is described in.

Note that the leftmost column in the figure represents row numbers.
For example, in the line of line number 1, the command 1001 is described as “open” and the target 1002 is described as “/”. This means “request a page whose URI (Uniform Resource Identifier) is“ / ”and wait until the page is read”. In the line of line number 2, the command 1001 is described as “assertTitle”, and the target 1002 is described as “top page”. This means that an error occurs if the title of the current page is not “top page”. The line with the line number 5 describes the command 1001 as “clickAndWait” and the target 1002 as “// input [@ value = 'login']”. This means “search for an input element having a value attribute whose value is“ login ”from the current page, click on it, and wait until the page is read”. The Selenium RC interprets and executes such an operation command on the browser 120 by using JavaScript.

  Such a browser command sequence (Series) is read in advance by the user. By the way, in order to generate a meaningful automatic test, it is desirable that the instruction sequence operates normally even if it is repeatedly executed several times. Therefore, when inputting a browser command sequence (Selene), it may be prompted to perform such input.

  The instruction sequence dividing unit 111 analyzes such an instruction sequence (Series) 1000 and divides it by a predetermined delimiter instruction. The division is performed after the instruction sequence that suggests that the state of the browser 120 is stable. The state in which the browser 120 is stable means that the operation of the browser 120 will not change over time.

  As described above, commands (separation instructions) that suggest that the state of the browser 120 is stable include “pause”, “open”, “... AndWait”, “waitFor. “Pause” is an instruction to wait for a time designated by the target. “Open” is an instruction to request a page of the URI specified by the target and wait until the page is read. “... AndWait” is a command to wait until the page is read after performing the operation. “WaitFor...” Is an instruction to wait until the state.

  The instruction sequence dividing unit 111 compares the command 1001 of the instruction sequence (Series) 1000 with the above delimiter instruction and, if there is a match, divides the instruction sequence as one delimiter up to the command. For example, it is assumed that the above delimiter instructions “pause”, “open”, “... AndWait”, “waitFor...” Are applied to the instruction sequence (Series) 1000 in FIG. In this case, since the command “open” is detected at the line number 1, the instruction string can be divided here. Next, the command “clickAndWait” at the line number 5, the command “clickAndWait” at the line number 8, the “clickAndWait” at the line number 11, and the “clickAndWait” at the line number 13 are detected and divided. Information indicating where the instruction sequence is divided is registered in the instruction sequence division table 1100.

  In addition, as a division method, it does not divide immediately after the above-described delimiter instruction, but after subsequent consecutive instructions that are not operation instructions (for example, “assert ...”, “verify ...”, etc.). You may separate them with

FIG. 5 is a diagram showing an example of the instruction sequence partition table.
The instruction sequence division table 1100 has information items of a category identifier 1101 and an instruction sequence row number 1102, and is a table in which each division after division is associated with the row number of the instruction sequence.

  The division identifier 1101 is an identifier obtained by concatenating a character “s” and a serial number in order to identify a divided instruction sequence. “S1” identifies the section including the first instruction sequence, and the instruction sequence numbers corresponding to “s2”, “s3”,..., “Sn” increase in order (n is an arbitrary integer) .

The instruction string row number 1102 indicates the line number of the instruction string corresponding to the category identifier 1101.
The instruction sequence division table 1100 in FIG. 5 shows a case where the instruction sequence is divided by row numbers 1, 5, 8, 11, and 13 of the above-described instruction sequence. For example, line number 1 for section s1, line numbers 2-5 for section s2, line numbers 6-8 for section s3, line numbers 9-11 for section s4, line numbers 12-13 for section s5, The section s6 is associated with the instruction sequence of line number 14 respectively. The generated instruction sequence partition table 1100 is stored in the storage unit 150.

Subsequently, a communication log is acquired by the communication log acquisition / classification unit 112 in units of the divided instruction sequence divided by the instruction sequence dividing unit 111, and is classified for each divided instruction sequence.
First, the operation of the log acquisition unit 140 will be described. The log acquisition unit 140 intercepts the request transmitted from the browser 120 to the server 600, records the request in the communication log, and transmits the request to the server 600. Similarly, the response transmitted from the server 600 is also intercepted and recorded in the communication log, and then returned to the browser 120. An identifier is assigned to the set of requests and responses acquired in this way, and is sequentially stored. Here, a directory is generated in the OS file system, and a communication log is stored under the directory. It is assumed that the identifier is generated by concatenating the character “d” and the serial number. Logs acquired as “d2”, “d3”,... “Dn” are sequentially accumulated in a set of logs where “d1” is acquired first.

6 and 7 are diagrams illustrating an example of the log group. FIG. 6 shows a log group of the directories d1 to d3. FIG. 7 shows a log group of directories d4 to d9.
The log groups 1200a and 1200b show communication logs obtained when the instruction sequence shown in FIG. 4 is executed. In addition, the upper log corresponding to each directory in FIGS. 6 and 7 indicates a request, and the lower log indicates a response.

  The communication log acquisition / classification unit 112 reproduces the divided instruction sequence of each category by the browser operation unit 130 based on the instruction sequence division table 1100, and the log acquisition unit 140 newly sets the log groups 1200a and 1200b at that time. Are determined to be communications generated by the divided instruction sequence. Then, a log classification table 23 in which this log is associated with the division of the divided instruction sequence is generated.

A specific example will be described.
FIG. 8 is a diagram showing an example of the log classification table. In the log classification table 1300a, a category identifier 1301 and a log directory name 1302 are associated with each other. The same applies to the log classification tables 1300b and 1300c.

(A) The log classification table (up to section s1) shows a state in which the directory name of the communication log corresponding to the first section s1 is registered.
The communication log acquisition / classification unit 112 causes the browser operation unit 130 to execute the instruction sequence (instruction sequence number 1) of the first section s1 of the instruction sequence partition table 1100 illustrated in FIG. The browser operation unit 130 executes the “open” command of the instruction sequence number 1 of the instruction sequence (Selene) 1000 shown in FIG. Accordingly, the browser 120 transmits a request for requesting information on the corresponding page to the server 600, and a response is returned from the server 600. The log acquisition unit 140 records the exchange. The directory d1 of the log group 1200a shown in FIG. 6 is set. The communication log acquisition / classification unit 112 acquires the communication log of the directory d1 and stores it in the storage unit 150. Further, this communication is registered in the log classification table 1300a as having occurred in the category s1. In the log classification table 1300a, a category identifier s1 and a log directory name d1 are registered in association with each other.

(B) The log classification table (up to category s2) shows a state in which the directory names of communication logs up to the next category s2 are registered.
As described above, the communication log acquisition / classification unit 112 instructs the browser operation unit 130 to reproduce the divided instruction sequence of the next category s2. Then, the communication log recorded at that time is acquired from the log acquisition unit 140. Here, communication logs of the directories d2 and d3 are acquired. The acquired log group is registered in the log classification table 1300b as having occurred in the category s2. In the log classification table 1300b, in addition to the registration of the log classification table 1300a, the category identifier s2 and the log directory names d2 and d3 are registered in association with each other.

Similarly, log acquisition and log classification table update are performed alternately or in parallel.
(C) The log classification table (final) shows a state in which the directory name of the communication log when registration is executed up to the last division s6 is registered.

  The communication log acquisition / classification unit 112 further reproduces the same procedure as above in the order of sections s3, s4, s5, and s6. The communication log acquired along with the reproduction is saved as needed. When the reproduction is finished, the directory name of the saved log is registered in the log classification table 1300c in association with the category. When an error occurs such that the instruction sequence cannot be executed normally during reproduction, an output for notifying the fact is given and the processing is terminated.

Thus, a request / response log between the corresponding browser 120 and the server 600 is obtained for each divided instruction sequence.
The test kernel generation unit 113 extracts a log for each category of the log classification table 1300c, and generates a test kernel for a test target request of the extracted log. The test target here is a form parameter included in the request. A test kernel is generated for requests having form parameters, and a test kernel is not generated for requests without form parameters. Whether it is a form parameter is estimated from the request format.

  An identifier is given to the generated test kernel. For example, it may be a character string obtained by concatenating the directory name of the communication log that is the source of the test core, the character “-”, and the parameter name. The test core describes the execution procedure of processing in the test core, such as request modification. An example of the test kernel will be described.

  (1) The URL or method is used to confirm that the intercepted request is a request having a form parameter to be tested. If they are different, the request / response is relayed in the same way as a normal proxy server.

(2) Generate and transmit a request in which the value of the form parameter of the request to be tested is changed to a character string in which 10,001 characters “a” are concatenated.
(3) When the status code of the response from the server 600 is 500 or more, an error message is output.

(4) The original request before modification is transmitted to the server 600 and a response is received.
(5) Providing the browser 120 with the response.
(6) Outputs the completion of the test kernel processing.

In the test kernel, such a processing procedure is described in a format that can be executed by the test apparatus. For example, it is output as a process (program) that operates on the implementation of the communication interception method.
In the test kernel of this example, if the request is a request to be tested, the value of the form parameter is modified, and 10,001 characters “a” are added and transmitted to the server 600. Then, the status code of the response is checked, and if it is 500 or more, an error message is output. As such, this test core is a test that helps find buffer overrun vulnerabilities. Further, the processing after (4) is not directly related to this test itself, but it is possible to easily stabilize the state on the server side by issuing the same (normal) request as the original.

The case where the contents of the test kernel are applied to the log group 1200a shown in FIG. 6 will be described.
Since there is no form parameter in the request in the directory d1, no test kernel is generated for d1.

  There are three form parameters in the directory d2. “Action”, “user”, and “password” included in “action = login & user = test & password = test” in the body (last line) of the request. Therefore, for the directory d2, for each parameter, a total of three test kernels having identifiers d2-action, d2-user, and d2-password are generated.

  As an example, generation of the test kernel d2-action will be described. When the contents of the test nucleus shown in (1) to (6) above are applied, the test nucleus d2-action is as follows.

  (1a) The following process is advanced only when the intercepted request starts with “POST / session.jsp HTTP /”. If they are different, the request / response is relayed in the same way as a normal proxy server.

  (2a) A request in which the value of the form parameter “action” of the request is changed to a character string in which 10,001 characters “a” are concatenated is generated and transmitted. Note that the Content-Length header field of the generated request is also appropriately changed before transmission.

(3a) When the response status code is 500 or more, an error message is output.
(4a) Send the original request and receive the response.

(5a) The response is provided to the browser 120 side.
(6a) Outputs that the processing of the test kernel d2-action is completed.
The test kernel d2-action is generated. Note that (1a) corresponds to (1) above. The same applies to (2a), (3a), (4a), (5a), and (6a).

  In this example, the contents of the test core are determined in advance. However, the contents of the test core can be received as a separate input. Although the form parameters are changed here, other test kernels may be generated. Furthermore, the identifier is an example, and may be appropriately set according to the content of the test nucleus to be generated.

Thus, when the test kernel is generated, the automatic test generation unit 114 generates an automatic test for executing each test kernel at an appropriate timing.
The automatic test is a test in which an input instruction sequence (Series) and a test kernel cooperate. That is, the browser operation unit (Selenium RC) 130 is used for the reproduction, and the communication interception device implemented as an HTTP proxy server is used for the test core.

  In order to generate such an automatic test, the automatic test generation unit 140 enables a test kernel to be validated at a certain timing during the reproduction while reproducing the instruction sequence (Selene) by the browser operation unit (Selenium RC) 130. Is generated. The timing at which the test kernel is validated is, for example, immediately before the instruction sequence belonging to the category identifier corresponding to the test kernel is reproduced.

  Further, a procedure for confirming whether the processing of the test nucleus is completed during the reproduction or at a predetermined timing may be added. The timing of confirmation is, for example, immediately after the instruction sequence corresponding to the segment identifier corresponding to the test nucleus is reproduced.

An example of an automatic test generation procedure will be described.
<1> The directory name of the corresponding log is obtained from the test kernel identifier.
<2> The category identifier corresponding to the directory name is acquired from the log classification table 1300c.

  <3> The instruction string row number corresponding to the classification identifier is acquired from the instruction string partitioning table 1100. Hereinafter, among the instruction sequence row numbers, the smallest number corresponding to the category identifier is called the smallest row number, and the largest number is called the largest row number.

<4> The browser operation unit (Selenium RC) 130 and a process for preparing and starting the communication interception device are output.
<5> Outputs a process for reproducing, in the browser operation unit (Selenium RC) 130, an instruction string before the minimum line number (an instruction string belonging to a section identifier before the test target section identifier) in the instruction string (Selene). To do.

<6> A process for validating the test kernel in the communication interception device is output.
<7> A process of reproducing the instruction sequence from the minimum line number to the maximum line number in the instruction sequence (Selene) by the browser operation unit (Selenium RC) 130 is output.

  <8> A process for confirming that the process has been completed with the validated test kernel is output. In the case where the output indicating completion is not output, an error message indicating that there is a possibility that the test could not be performed correctly is output.

  <9> The test kernel is invalidated, and the instruction sequence (the instruction sequence belonging to the category identifier after the category identifier to be tested) after the maximum row number in the command sequence (Series) is displayed on the browser operation unit (Selenium RC) 130. Output the process to play.

Note that the process <9> is a meaningful automatic test even if it is not necessary, but the state on the server side can be easily stabilized by reproducing the instruction sequence (Selene) to the end.
The automatic test generated in this procedure is loaded and executed in the control unit of the automatic test apparatus that executes the automatic test of the application. Details of the automatic test apparatus will be described later. The instruction sequence (Selene) is output to the Selenium RC, and the test kernel is output to the communication interceptor.

  As an example, a case where it is applied to the test kernel d2-action will be described. Applying the contents of the automatic test generation procedure described above, the automatic test generation procedure corresponding to the test kernel d2-action is as follows.

<1a> The directory name “d2” of the corresponding log is acquired from the identifier “d2-action” of the test kernel.
<2a> The category identifier “s2” corresponding to the directory name “d2” is acquired from the log category table 1300c.

<3a> The instruction string row numbers “2 to 5” corresponding to the classification identifier “s2” are acquired from the instruction string partitioning table 1100. Therefore, the minimum line number is “2” and the maximum line number is “5”.
<4a> The browser operation unit (Selenium RC) 130 and processing for preparing and starting the communication interception device are output.

<5a> A process for reproducing the instruction string of the instruction string row number “1” in the instruction string (Selene) by the browser operation unit (Selenium RC) 130 is output.
<6a> A process for validating the test kernel d2-action is output by the communication interception device.

<7a> A process for reproducing the instruction string from the instruction string row numbers “2” to “5” in the instruction string (Selene) by the browser operation unit (Selenium RC) 130 is output.
<8a> A process for confirming that the test kernel d2-action has output that the process has been completed is output. In the case where the output indicating completion is not output, an error message indicating that there is a possibility that the test could not be performed correctly is output.

  <9a> The test kernel d2-action is invalidated, and a process for reproducing the instruction sequence from the instruction sequence row numbers “6” to “14” by the browser operation unit (Selenium RC) 130 is output. <1a> corresponds to the above <1>. The same applies to <2a>, <3a>, <4a>, <5a>, <6a>, <7a>, <8a>, <9a>.

Here, data such as an instruction sequence (Selene) and a test kernel d2-action are included in the generated automatic test.
The automatic test process generated by the automatic test generation procedure corresponding to the test kernel d2-action is as follows.

<1-4a> The browser operation unit (Selenium RC) 130 and the communication interception device are prepared and activated.
<2-5a> Of the instruction string (Selene), the instruction string of the instruction string row number “1” is reproduced by the browser operation unit (Selenium RC) 130.

<3-6a> The test kernel d2-action is validated by the communication interception device.
<4-7a> Of the instruction string (Selene), the instruction string from the instruction string row numbers “2” to “5” is reproduced by the browser operation unit (Selenium RC) 130.

  <5-8a> It is confirmed that the processing kernel d2-action has output that processing has been completed. In the case where the completion is not output, an error message indicating that there is a possibility that the test could not be performed correctly is output.

  <6-9a> The test kernel d2-action is invalidated and the instruction sequence from the instruction sequence row numbers “6” to “14” is reproduced by the browser operation unit (Selenium RC) 130. <1-4a> corresponds to the process output in <4a> above. The same applies to <2-5a>, <3-6a>, <4-7a>, <5-8a>, <6-9a>.

In the above description, the automatic test is output to an external automatic test apparatus. However, an automatic test may be generated internally, and its execution and result output may be performed on the same device.
As described above, according to the test generation apparatus 100, it is possible to create an automatic test combining test nuclei for a range requested by an instruction sequence by preparing and inputting an instruction sequence (Selene). It becomes like this.

  In addition, since the automatic test validates the test kernel only at the timing when the test target request is predicted to occur, it is unlikely that a test other than the test target is erroneously tested. Furthermore, if the test kernel confirms whether the request is a request to be tested by using a URL or a method, the possibility that the test is erroneously performed can be further reduced.

  Here, the processing procedure of the application test generation method of the test generation apparatus 100 will be described using a flowchart. As in the previous description, the target for generating the test kernel is the form parameter of the request.

FIG. 9 is a flowchart showing the procedure of the application test generation process.
An instruction sequence (Series) 1000 used for the reproduction test is input, and the process is started.

  [Step S01] The input instruction sequence (Series) 1000 is analyzed, and an instruction sequence dividing process is performed to divide the instruction sequence by a delimiter instruction. An instruction string partition table 1100 is set in which a section identifier for identifying a divided section and a row number of an instruction string belonging to the section are associated with each other. Details of the instruction sequence dividing process will be described later.

  [Step S02] For each division divided in step S01, the browser operation unit (Selenium RC) 130 is caused to execute a division instruction sequence for the division. Then, the communication log obtained at that time is stored, and communication log acquisition / classification processing for classifying the communication log for each divided instruction sequence is performed. A log classification table 1300c is set in which the division identifier of the reproduced divided instruction sequence is associated with the directory name storing the communication log acquired at this time. Details of the communication log acquisition / classification process will be described later.

  [Step S03] Based on the log classification table 1300c, a communication log is extracted for each category identified by the category identifier, and a test kernel is generated based on this communication log. For example, a test procedure such as changing the value of a form parameter of a request having a form parameter among requests recorded in the communication log is described as a test core. When a single request includes multiple form parameters, a test kernel is generated for all form parameters. An identifier for identifying the test nucleus is set in the test nucleus generated in this way. Note that the test core is described as a program that can be executed by the communication interception device that modifies the request.

  [Step S04] An automatic test is generated by combining the instruction sequence (Selene) and the test kernel generated in step S03. In the automatic test, a sequence of instructions 1000 to be reproduced is sequentially executed, and when a test nucleus is set in a section to which the instruction sequence belongs, a test procedure for enabling the test nucleus is generated. In addition, when the execution of the instruction sequence belonging to the section corresponding to the test nucleus is completed, a procedure for invalidating the test nucleus is also set.

  By executing the above processing procedure, an automatic test for generating a test using a test kernel is generated from the instruction sequence 1000 when executing a predetermined range of instruction sequences.

Next, instruction sequence division processing and communication log acquisition / classification processing will be described.
FIG. 10 is a flowchart showing the sequence of instruction sequence division processing.
An instruction sequence (Series) 1000 is input and processing is started.

  [Step S11] One line of the instruction sequence (Selene) is read. If it is the first process, the instruction sequence of line number 1 is read. For subsequent processing, the instruction sequence of the line number next to the instruction sequence that was processed last time is read.

  [Step S12] The instruction (command) described in the instruction sequence is collated with the instruction specified as the delimiter instruction. An instruction indicating that the browser 120 is in a stable state is specified as the delimiter instruction. For example, an instruction “wait” until a certain state is obtained or necessary information is obtained corresponds. In the case of Selenes, “pause”, “open”, “... AndWait”, “waitFor. If the collated instruction sequence is a delimiter instruction, the process proceeds to step S13. If it is not a delimiter instruction, the process proceeds to step S14.

  [Step S13] If the instruction sequence is a delimiter instruction, the instruction sequence from the next instruction sequence in which the delimiter instruction appeared before the delimiter instruction sequence detected in step S12 is registered in the instruction sequence partition table 1100 as one segment. In the instruction sequence division table 1100, the row numbers of the instruction sequences divided into the divisions are set in association with the division identifiers for identifying the divisions. The minimum line number and the maximum line number of the divided range may be registered.

[Step S14] It is determined whether there is a next instruction sequence. If not, the process proceeds to step S15. If yes, the process returns to step S11 to process the next instruction sequence.
[Step S15] It is determined whether an instruction sequence that is not divided anywhere remains. If it remains, the process proceeds to step S16. If not, the process ends.
[Step S16] When instruction sequences that are not divided anywhere remain, they are collected and set as one division and registered in the instruction sequence division table 1100. Thereafter, the process ends.

By executing the above processing procedure, the instruction sequence is divided by the delimiter instruction, and the instruction sequence division table 1100 is generated.
Next, communication log acquisition / classification processing will be described.

FIG. 11 is a flowchart showing a procedure of communication log acquisition / classification processing.
[Step S21] The log acquisition unit 140 is activated.
[Step S22] Based on the instruction sequence division table 1100, a division instruction sequence to be executed by the browser operation unit (Selenium RC) 130 is designated. First, s1 is designated as the processing target segment identifier sn (n is an arbitrary integer).

  [Step S23] The browser operation unit (Selenium RC) 130 is caused to execute an instruction string in the range of the corresponding category identifier. While the browser operation unit (Selenium RC) 130 executes the instruction sequence (Selene), the log acquisition unit 140 records a communication log of requests and responses exchanged between the browser 120 and the server 600.

[Step S24] After the execution of the instruction sequence in the range of the category identifier is completed, the communication log collected by the log acquisition unit 140 during execution of the instruction sequence in this range is acquired.
[Step S25] The communication log acquired in step S24 is stored in the storage unit 150.

[Step S26] The communication log identification information (communication log directory name) acquired in step S25 and the category identifier are registered in the log classification table 1300c.
[Step S27] It is checked whether or not the next category is registered in the instruction sequence partitioning table 1100, that is, whether or not there is a next category identifier. If yes, the process proceeds to step S28. If not, the process is terminated.

  [Step S28] When there is a next segment identifier, s (n + 1) obtained by increasing the value of n by 1 is designated as the segment identifier sn to be processed. Then, the process returns to step S23, and a communication log for the next category is acquired.

By executing the above processing procedure, an instruction sequence is executed for each category, and a log classification table 1300c that classifies the acquired communication log is set.
Next, an automatic test apparatus that executes this automatic test will be described. FIG. 12 is a diagram showing the configuration of the automatic test apparatus.

  The automatic test device 200 includes a control unit 210, a browser 220, a browser operation unit 230, and a communication interception device 240, and executes an application test of the server 600 connected via a network.

  The control unit 210 is an automatic test execution unit that controls the browser operation unit 230 and the communication interception device 240 based on the automatic test, and controls an application test for the server 600.

  Similarly to the browser 120, the browser 220 downloads an HTML file, an image file, a music file, or the like provided by the server 600, and reproduces and outputs it to the monitor 108 or an audio output device. The browser operation unit 230 executes a command sequence described by a command sequence (Series) 251 in accordance with the control unit 210, and automatically generates an operation command to the browser 220 or the like. Here, “Selenium RC” is used as a test tool for performing an automatic test of the browser operation method. The communication interception device 240 is implemented as an HTTP proxy server. In accordance with the control unit 210, the processing of the test kernel 252 is executed, and the input request is modified and transmitted to the server 600.

The operation of the automatic test apparatus 200 and the automatic test method will be described in the case where an automatic test corresponding to the above-described test kernel d2-action is executed.
A browser operation unit (Selenium RC) 230 and a communication interception device 240 are prepared and activated.

  After the preparation is completed, the browser operation unit (Selenium RC) 230 is caused to input and reproduce the line number 1 of the instruction sequence (Selene) 251. The communication interception device 140 relays the request / response in the same manner as a normal proxy server.

Next, the control unit 210 inputs the test kernel d2-action to the communication interception device 240 and instructs it to be validated.
Subsequently, the control unit 210 causes the browser operation unit (Selenium RC) 230 to input and reproduce the line numbers 2 to 5 of the instruction sequence (Selene) 251. At this time, among the requests transmitted to the server 600, those targeted by the test kernel d <b> 2-action are modified by the communication interception device 240 and then transmitted to the server 600.

  After that, the test kernel d2-action confirms the output indicating that the processing is completed. If it has not been output, an error message indicating that the correct test may not have been output is output. If an output indicating that the processing has been completed is output, the test kernel d2-action is invalidated and the browser operation unit (Selenium RC) 230 is caused to input / reproduce the line number 6 and later of the instruction sequence (Selenese).

Thus, the automatic test apparatus 200 can automatically perform an application test without the user performing execution control.
In the above description, the automatic test apparatus 200 and the test generation apparatus 100 are separate apparatuses, but can be mounted on the same apparatus. For example, the test generation unit 110 is mounted on the automatic test apparatus 200, and the browser operation unit 230 and the communication interception apparatus 240 are connected. The test generation unit 110 operates the browser operation unit 230 as the browser operation unit 130 and the communication interception device 240 as the log acquisition unit 140, and generates an automatic test.

  The above processing functions can be realized by a computer. In that case, a program describing the processing contents of the functions that the test generation apparatus should have is provided. By executing the program on a computer, the above processing functions are realized on the computer. The program describing the processing contents can be recorded on a computer-readable recording medium.

  When distributing the program, for example, portable recording media such as a DVD (Digital Versatile Disc) and a CD-ROM (Compact Disc Read Only Memory) on which the program is recorded are sold. It is also possible to store the program in a storage device of a server computer and transfer the program from the server computer to another computer via a network.

  The computer that executes the program stores, for example, the program recorded on the portable recording medium or the program transferred from the server computer in its own storage device. Then, the computer reads the program from its own storage device and executes processing according to the program. The computer can also read a program directly from a portable recording medium and execute processing according to the program. Further, each time the program is transferred from the server computer, the computer can sequentially execute processing according to the received program.

Regarding the above embodiment, the following additional notes are disclosed.
(Supplementary Note 1) In an application test generation program for testing an application device that provides a predetermined service via a network,
Computer
An instruction sequence comprising instructions for instructing a user agent receiving the provision of a predetermined service to execute a predetermined process including a process in response to transmission of a request to be tested to the application apparatus and reception of a response to the request. An instruction sequence dividing means for acquiring and detecting a delimiter instruction in which the process represented by the instruction sequence is delimited from the instruction sequence, dividing the instruction sequence for each delimiter instruction, and generating a divided instruction sequence;
For each of the divided instruction sequences, an instruction to the user agent is reproduced and output based on the divided instruction sequence, and the request transmitted by the user agent is transmitted from the application device in response to the request. The response is acquired as a communication log, stored in a storage unit in association with the divided instruction sequence reproduced when the communication log is acquired, and a log classification table in which the communication log is classified for each divided instruction sequence is generated. Communication log acquisition / classification means,
Extracting the communication log associated with the divided instruction sequence including the instruction sequence that generated the request to be tested from the log classification table, and modifying the request based on the extracted communication log Automatic test generation means for generating a test nucleus describing a test procedure including: an automatic test for executing the divided instruction sequence and the test nucleus in cooperation with each other;
An application test generation program characterized by functioning as (1, Fig. 1)
(Additional remark 2) The said instruction sequence division | segmentation means makes the said instruction | indication which suggests that the state after receiving the said response from the said application apparatus based on the said request became a stable state as the said division | segmentation instruction, It is characterized by the above-mentioned. The application test generation program according to appendix 1.

  (Additional remark 3) The said instruction sequence division | segmentation means is an instruction | indication of the next execution order following the instruction which suggests that the state after receiving the said response from the said application apparatus based on the said request became the stable state. The application test generation program according to supplementary note 1, characterized by:

  (Supplementary note 4) The application test generation according to Supplementary note 2 or 3, wherein the instruction sequence dividing unit uses the instruction to wait for reception of the response transmitted from the application device as the delimiter instruction. program.

  (Supplementary note 5) The application test generation according to supplementary note 2 or 3, wherein the instruction sequence dividing means uses an instruction indicating that the response transmitted from the application device has been received as the delimiter instruction. program.

  (Additional remark 6) The said automatic test production | generation means reproduces | regenerates the said division | segmentation instruction sequence one by one, and makes it output to the said user agent, and when the said test nucleus is not set with respect to the said division | segmentation instruction sequence, The operation of the execution unit to be executed is invalidated, and when the test kernel is set for the divided instruction sequence, a procedure for validating the operation of the execution unit is incorporated in the automatic test. The application test generation program according to 1 or 2.

  (Additional remark 7) The said automatic test production | generation means extracts the characteristic information which characterizes this request including the address information of the said application apparatus contained in the said request from the said communication log containing the request of the object which produces | generates the said test nucleus, Comparing the feature information with the feature information included in the request sent by the user agent, and incorporating a procedure for modifying any part of the request sent by the user agent into the automatic test only when they match. The application test generation program according to supplementary note 1, wherein

  (Supplementary note 8) The automatic test generation means, at the end of a test for modifying the request, transmits the request before modification to the application device and incorporates a procedure for obtaining a response in the test core. 1. An application test generation program according to 1.

  (Supplementary note 9) A procedure in which the automatic test generation means determines whether or not the processing of the test kernel is completed when the execution of the instruction sequence corresponding to the test kernel is completed, and notifies an error when the processing is not completed The application test generation program according to supplementary note 1, wherein the application test generation program is incorporated into the automatic test.

(Additional remark 10) In the application test production | generation method for the test of the application apparatus which provides a predetermined service via a network,
On the computer,
An instruction sequence comprising instructions for instructing a user agent receiving the provision of a predetermined service to execute a predetermined process including a process in response to transmission of a request to be tested to the application apparatus and reception of a response to the request. An instruction sequence dividing procedure for obtaining a segmented instruction in which the processing represented by the instruction sequence is obtained from the instruction sequence, dividing the instruction sequence for each delimited instruction, and generating a divided instruction sequence;
For each of the divided instruction sequences, an instruction to the user agent is reproduced and output based on the divided instruction sequence, and the request transmitted by the user agent is transmitted from the application device in response to the request. The response is acquired as a communication log, stored in a storage unit in association with the divided instruction sequence reproduced when the communication log is acquired, and a log classification table in which the communication log is classified for each divided instruction sequence is generated. Communication log acquisition / classification procedure,
Extracting the communication log associated with the divided instruction sequence including the instruction sequence that generated the request to be tested from the log classification table, and modifying the request based on the extracted communication log And an automatic test generation procedure for generating an automatic test for executing the split instruction sequence and the test nucleus in cooperation with each other.
An application test generation method characterized in that

(Additional remark 11) In the application test apparatus for the test of the application apparatus which provides a predetermined service via a network,
An instruction sequence comprising instructions for instructing a user agent receiving the provision of a predetermined service to execute a predetermined process including a process in response to transmission of a request to be tested to the application apparatus and reception of a response to the request. An instruction sequence dividing means for obtaining and detecting a delimiter instruction in which the processing represented by the instruction sequence is delimited from the instruction sequence, dividing the instruction sequence for each delimiter instruction, and generating a divided instruction sequence;
For each of the divided instruction sequences, an instruction to the user agent is reproduced and output based on the divided instruction sequence, and the request transmitted by the user agent is transmitted from the application device in response to the request. The response is acquired as a communication log, stored in a storage unit in association with the divided instruction sequence reproduced when the communication log is acquired, and a log classification table in which the communication log is classified for each divided instruction sequence is generated. Communication log acquisition / classification means,
Extracting the communication log associated with the divided instruction sequence including the instruction sequence that generated the request to be tested from the log classification table, and modifying the request based on the extracted communication log Automatic test generation means for generating a test nucleus that describes a test procedure including: and generating an automatic test that causes the divided instruction sequence and the test nucleus to be executed in cooperation with each other;
Controls the operation of a relay unit that relays requests and responses between the user agent and the application device, and reproduces the divided instruction sequence for the user agent and causes the user agent to execute the predetermined process. Control means for causing the relay means to execute the interception and modification of the request based on the corresponding test nucleus at the time of reproduction of the divided instruction sequence for transmitting the request in which the test nucleus is generated,
An application test apparatus characterized by comprising:

DESCRIPTION OF SYMBOLS 10 Test generation apparatus 11 Instruction sequence dividing means 12 Communication log acquisition / classification means 13 Automatic test generation means 14 Storage means 21 Instruction sequence 22 Instruction sequence division table 23 Log classification table 24 Automatic test 40 User agent 50 Communication log acquisition device 60 Server

Claims (5)

In an application test generation program for testing an application device that provides a predetermined service via a network,
Computer
An instruction sequence comprising instructions for instructing a user agent receiving the provision of a predetermined service to execute a predetermined process including a process in response to transmission of a request to be tested to the application apparatus and reception of a response to the request. An instruction sequence dividing means for acquiring and detecting a delimiter instruction in which the process represented by the instruction sequence is delimited from the instruction sequence, dividing the instruction sequence for each delimiter instruction, and generating a divided instruction sequence;
For each of the divided instruction sequences, an instruction to the user agent is reproduced and output based on the divided instruction sequence, and the request transmitted by the user agent is transmitted from the application device in response to the request. The response is acquired as a communication log, stored in a storage unit in association with the divided instruction sequence reproduced when the communication log is acquired, and a log classification table in which the communication log is classified for each divided instruction sequence is generated. Communication log acquisition / classification means,
Extracting the communication log associated with the divided instruction sequence including the instruction sequence that generated the request to be tested from the log classification table, and modifying the request based on the extracted communication log Automatic test generation means for generating a test nucleus describing a test procedure including: an automatic test for causing the divided instruction sequence and the test nucleus to be executed in cooperation with each other;
An application test generation program characterized by functioning as   2. The instruction sequence dividing unit, wherein the instruction indicating that the state after receiving the response from the application apparatus based on the request has become a stable state is the delimiter instruction. Application test generation program described.   The automatic test generation means sequentially reproduces the divided instruction sequences and outputs them to the user agent, and executes the test nucleus when the test nucleus is not set for the divided instruction sequence 3. A procedure for validating the operation of the execution unit is incorporated in the automatic test when the operation is invalidated and the test kernel is set for the divided instruction sequence. Application test generation program described. In an application test generation method for testing an application device that provides a predetermined service via a network,
On the computer,
An instruction sequence comprising instructions for instructing a user agent receiving the provision of a predetermined service to execute a predetermined process including a process in response to transmission of a request to be tested to the application apparatus and reception of a response to the request. An instruction sequence dividing procedure for obtaining and detecting a delimiter instruction in which the processing represented by the instruction sequence is delimited from the instruction sequence, dividing the instruction sequence for each delimiter instruction, and generating a divided instruction sequence;
For each of the divided instruction sequences, an instruction to the user agent is reproduced and output based on the divided instruction sequence, and the request transmitted by the user agent is transmitted from the application device in response to the request. The response is acquired as a communication log, stored in a storage unit in association with the divided instruction sequence reproduced when the communication log is acquired, and a log classification table in which the communication log is classified for each divided instruction sequence is generated. Communication log acquisition / classification procedure,
Extracting the communication log associated with the divided instruction sequence including the instruction sequence that generated the request to be tested from the log classification table, and modifying the request based on the extracted communication log And an automatic test generation procedure for generating an automatic test for executing the split instruction sequence and the test nucleus in cooperation with each other.
An application test generation method characterized in that In an application test apparatus for testing an application apparatus that provides a predetermined service via a network,
An instruction sequence comprising instructions for instructing a user agent receiving the provision of a predetermined service to execute a predetermined process including a process in response to transmission of a request to be tested to the application apparatus and reception of a response to the request. An instruction sequence dividing means for obtaining and detecting a delimiter instruction in which the processing represented by the instruction sequence is delimited from the instruction sequence, dividing the instruction sequence for each delimiter instruction, and generating a divided instruction sequence;
For each of the divided instruction sequences, an instruction to the user agent is reproduced and output based on the divided instruction sequence, and the request transmitted by the user agent is transmitted from the application device in response to the request. The response is acquired as a communication log, stored in a storage unit in association with the divided instruction sequence reproduced when the communication log is acquired, and a log classification table in which the communication log is classified for each divided instruction sequence is generated. Communication log acquisition / classification means,
Extracting the communication log associated with the divided instruction sequence including the instruction sequence that generated the request to be tested from the log classification table, and modifying the request based on the extracted communication log Automatic test generation means for generating a test nucleus that describes a test procedure including: and generating an automatic test that causes the divided instruction sequence and the test nucleus to be executed in cooperation with each other;
Controls the operation of a relay unit that relays requests and responses between the user agent and the application device, and reproduces the divided instruction sequence for the user agent and causes the user agent to execute the predetermined process. Control means for causing the relay means to execute the interception and modification of the request based on the corresponding test nucleus at the time of reproduction of the divided instruction sequence for transmitting the request in which the test nucleus is generated,
An application test apparatus characterized by comprising:

Images