JP2014038453A - Testing device and computer program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To support operating a web page at proper timing in a web page test.SOLUTION: A tablet PC 12 acquires a web page of a test object from a web server 10, and stores tree structure data in which content of the web page is shown. When a second operation should be executed after executing a first operation for a web page, the tablet PC 12 acquires an element value of the web page on the first operation or the second operation from the tree structure data and determines whether or not the element value is consistent with a predetermined expected value which indicates that the second operation can be executed. The tablet PC 12 executes the second operation for the web page on condition that the element value is consistent with the expected value.

Description

  The present invention relates to a data processing technique, and to a technique for testing a web page.

  The web page test is generally performed by performing an input operation on a web browser that displays the web page and confirming the operation result. The present applicant has proposed a technique for realizing a test efficiency in the following Patent Document 1 when a web page should be tested on a plurality of types of web browsers.

JP 2011-164784 A

  A process of updating the display content of the web page may be implemented in the code of the web page asynchronously with the web client acquiring the web page from the web server. It is not desirable for the present inventor to operate the web page during execution of such asynchronous processing, and it is necessary to confirm whether or not it is time to operate the web page in the web page test. Thought.

  The present invention has been made on the basis of the above-mentioned idea of the inventor, and its main object is to provide a technique for supporting the operation of the web page at an appropriate timing in the web page test.

  In order to solve the above problem, a test apparatus according to an aspect of the present invention includes a page management unit that holds tree structure data indicating the content of a web page to be tested, and a first operation on the web page. When the operation 2 is to be executed, the element value of the web page related to the first operation or the second operation is acquired from the tree structure data, and the element value indicates that the second operation can be executed. A determination unit that determines whether or not the value matches a predetermined expected value; and an operation execution unit that executes a second operation on the web page on condition that the element value matches the expected value.

  It should be noted that any combination of the above-described components, and a representation of the present invention converted between a method, a system, a program, a recording medium storing the program, and the like are also effective as an aspect of the present invention.

  ADVANTAGE OF THE INVENTION According to this invention, it can support operating a web page at an appropriate timing in the test of a web page.

It is a conceptual diagram of the software stack of automatic test AP. It is a figure explaining the mechanism of the web page test by automatic test AP. It is a figure which shows the structure of the test system of embodiment. It is a block diagram which shows the function structure of the tablet PC of FIG. It is a figure which shows the window of a standard browser. It is a figure which shows the window of the default test browser. It is a figure which shows the modification program for test browsers. It is a figure which shows the window of the test browser after a modification program application. It is a conceptual diagram for demonstrating the process of the operation execution part. It is a conceptual diagram for demonstrating the process of the operation execution part. It is a block diagram which shows the function structure of the management server of FIG. It is a flowchart which shows an automatic test system operation | movement. It is a flowchart which shows the web browser starting process of S10 of FIG. 12 in detail. It is a flowchart which shows the web page acquisition process of S12 of FIG. 12 in detail. It is a flowchart which shows in detail the screen element operation process of S14 of FIG. It is a figure which shows typically the web page of test object. It is a figure which shows the program code of the web page of FIG. It is a flowchart which shows the screen capture process of S16 of FIG. 12 in detail. It is a flowchart which shows the capture image provision process of S20 of FIG. 12 in detail. It is a figure which shows the program code of the web page of FIG.

  In the present embodiment, an application (hereinafter also referred to as “automatic test AP”) that performs an automatic test of a web page on a tablet PC on which an Android (trademark or registered trademark) OS (Operating System) is mounted is proposed. . FIG. 1 is a conceptual diagram of a software stack of the automatic test AP 300. The automatic test AP 300 is based on Selenium 2.0, which is open source software corresponding to Android OS. The open source extension class 304 is a class group for correcting / adding the function of Selenium 2.0. For example, it provides functions such as hiding the sidebar and hiding unnecessary dialogs.

  The general-purpose component 306 is a general-purpose component (class group) that is required when executing an automatic test of a web page, such as image acquisition by screen capture or web page load waiting processing. The screen operation common component 308 is an operation component (class group) for various objects such as buttons and text areas arranged on the web page. An action class described later typically calls a general-purpose component 306 or a screen operation common component 308 (function / method provided by). These can be said to be assets as automatic test tools, and are expanded according to the update and development of Selenium 2.0 and Android OS.

  The action class 310 is a class group that holds operation contents for various objects such as buttons and text fields arranged on a web page, and is typically created in units of screens (that is, web pages). Other operations specified by the action class 310 include obtaining a test web page and displaying it on the screen, and scrolling the web browser screen (that is, the web page image displayed in the web browser window) vertically or horizontally. Operation, web browser screen capture operation, etc.

  The data class 312 is a class group that holds screen operation information (for example, information indicating presence / absence of pressing operation on a button or information indicating an input character for a text field) determined for each test case, and is typically created for each test case Is done. The scenario class 314 combines an action class and a data class and holds them as test cases. These can be said to be assets as test scripts, which are appropriately developed by the application team that performs the test according to the screen (web page) to be tested.

  The automatic test AP 300 according to the embodiment includes the automatic test tool and the test script shown in FIG. Although not shown in FIG. 1, the automatic test AP 300 includes a class group (a so-called test execution engine) that links an automatic test tool and a test script to comprehensively control web page tests.

  In addition, Selenium 2.0 includes WebView (hereinafter also referred to as “test browser”), which is a web browser used for web page testing, and displays the web page specified as the test target on the test browser. . In addition, Selenium 2.0 exposes a function for operating a web page displayed on the test browser so that it can be called from outside (an operation execution unit 44 described later).

  For example, when a class group of the screen operation common component 308 and the general-purpose component 306 calls a function provided by Selenium 2.0, Selenium 2.0 calls an API of a test browser corresponding to the called function. As a result, an operation designated by the user (typically, an input operation for various objects arranged on the web page) is executed on the web page to be tested via the test browser. In other words, a typical function call relationship is: action class → screen operation common component 308 or general-purpose component 306 → public function of Selenium 2.0 → test browser API.

  Selenium 2.0 also discloses a function for capturing a web page display screen (image of a web page displayed on a test browser). A part of the web page operation by Selenium 2.0 is realized by Selenium 2.0 further calling a function provided by the OS.

  Here, with reference to FIG. 2, the mechanism of the web page test by the automatic test AP 300 will be outlined. FIG. 2A schematically shows a display screen of the web page 110 to be tested, and FIG. 2B shows the HTML program 120 of the web page 110. The text field 112 of the web page 110 corresponds to the code 122 of the HTML program 120, and the button 114 of the web page 110 corresponds to the code 124 of the HTML program 120.

  FIG. 2C shows the code of the action class 130 that defines the operation type for the web page 110. The action class 130 defines the operation type in the format of “(operation object identification information). Function name provided by the screen operation common part 308 or action class 310 (argument such as input data)”. FIG. 2 (d) shows a data class 140 that defines the input content in the operation on the web page 110. The data class 140 defines an argument at the time of calling a function in the action class 130.

  Specifically, the code 132 in FIG. 2C calls the setKey function of the screen operation common component 308 and inputs “1234” defined by the code 142 in FIG. 2D to the text field 112 of the web page 110. Defines the operations to be performed. The screen operation common component 308 that has received a call to the setKey function calls a function provided by Selenium 2.0 corresponding to the setKey function. The code 134 in FIG. 2C is an operation for clicking the button 114 on the web page 110 by calling the click function of the screen operation common component 308 with “ON” defined by the code 144 in FIG. 2D as an argument. Is stipulated.

The automatic test AP according to the embodiment expands the functions of Selenium 2.0 and realizes improvement of web page test efficiency and test quality. Three main features will be described.
(Characteristic 1) When test data should be set for a predetermined object (for example, a text field, a text area, a pull-down list, etc., hereinafter also referred to as “special object”) arranged on the web page. Test data is set directly by manipulating a DOM (Document Object Model) tree.

  Selenium 2.0 provides functions for operating these special objects. Internally, Selenium 2.0 calls an operation function provided by the OS. However, the method using the operating function of the OS has a problem in terms of test efficiency. That is, in the method of calling the operating function of the OS, a software keyboard is displayed when characters are input to a text field or text area, and characters are input via the software keyboard. Therefore, it takes time to input characters. End up. Originally, what should be confirmed in the test is a change in the screen display after character input (and after pressing a button) (for example, transition to another web page), and it is not a success or failure of character input itself, so it takes a long time to input characters. It is not desirable to require In addition, when the pull-down list is operated, the processing sometimes stops when the list of selectable items is expanded and displayed.

  Here, the test target / normality check target by the automatic test AP is a web page and a web server that provides the web page, not a software keyboard or a pull-down list (checking whether data can be entered). . Therefore, the test purpose of confirming the behavior of the web page or web server (for example, change in the display content of the web page after data input) can be achieved without going through the data input processing using these components.

  Therefore, the automatic test AP is an operation function provided by the OS for the operation target object of the web page, and once the GUI object for data setting is displayed, the operation function for setting data in the operation target object is called. Suppress. It can be said that this operation function is a native component provided by Android OS for data input to a screen or an input field in the screen. Instead of calling an operation function provided by the OS, the automatic test AP directly edits the element value of the object in the DOM tree of the web page and directly sets the test data to the object. In other words, test data such as input character strings and selection items are directly set for the data of the operation target object developed in the memory of the tablet PC 12. For example, such a DOM tree operation function is implemented in the setKey function shown in FIG. This realizes a character input operation without using a software keyboard, realizes an item selection operation without expanding a pull-down list, and improves the test efficiency by shortening the time required for those operations.

  (Feature 2) The test browser window is adjusted to have the same configuration as the standard browser window of the tablet PC. The window can be said to be a screen for displaying the processing result of an application such as a web browser on the display, and can also be said to be a GUI component or display area provided by the OS (to the application) for rendering the processing result of the application.

  An end user using a tablet PC can browse a web page using a web browser (hereinafter referred to as a “standard browser”) set as a browser normally used in the tablet PC (a so-called default browser). It is common. It can be said that the standard browser is a web browser used by AndroidOS by default. On the other hand, the test browser incorporated in Selenium 2.0 is different from the standard browser. Therefore, the configuration of the standard browser window and the default configuration of the test browser window are different. In web page automatic testing, the web page display screen is captured as evidence, but if the screen capture image is different from the screen when the end user actually browses the web page, the quality of the test will be degraded. May end up.

  Therefore, the automatic test AP adjusts the layout of the test browser window so that the test browser window matches the configuration of the standard browser window. For example, an additional object (for example, a button object) having the same size as the address bar that is displayed in the standard browser window but not displayed in the test browser window is displayed in the test browser window. Thereby, the window of a test browser can be made into the same structure as the window of the standard browser of tablet PC, and the display mode of a web page can be made the same. As a result, the screen capture image as evidence of the test reflects the appearance of the web page when the end user browses the web page using the standard browser, and the degradation of the test quality can be suppressed. .

  In addition, when the type of Android terminal is different, the type of web browser used as standard in each terminal may be different, and the layout of the web browser (window) may be different in each terminal. Furthermore, the display of the web page needs to be supported by a plurality of types of Android terminals. In this case, the web page may be tested on a plurality of types of Android terminals. Based on such a background, the present inventor considers that a mechanism for absorbing the layout of different web browsers (windows) at each terminal is important across a plurality of types of Android terminals.

  (Characteristic 3) It is confirmed whether or not the object specified as the operation target in the action class is in an operable state on the web page. If the object is in an operable state, an operation on the object is executed.

  The present inventor has found that even when the acquisition of the web page data from the web server is completed, the loading of the object arranged on the web page may not be completed. For example, when processing asynchronous with acquisition of web page data from a web server is performed on a web page, an operation of the operation target object may be enabled only after the asynchronous processing is completed. Originally, what should be confirmed in the test is a change in the screen display after the operation on the object (for example, transition to another web page), not the success or failure of the operation on the object itself, but the error due to the incomplete loading of the object. Occurrence is undesirable. Further, when an error due to incomplete loading of an object occurs, it is necessary to re-execute the test in order to achieve the original test purpose, which may cause a decrease in test efficiency.

  Therefore, the automatic test AP accesses the DOM tree to check whether or not the object designated as the operation target is operable, and operates the object on the condition that the object is operable. As a result, the occurrence of an error due to the incomplete loading of the object is suppressed, and the decrease in test efficiency is suppressed.

  In addition, since the automatic test AP automatically determines the state of the operation target object and operates it at an appropriate timing, the person in charge of the test can implement an action class without being aware of the exact operation timing for the operation target object. For example, it is not necessary to implement code such as waiting until the operation target object becomes operable in each action class, and the code of each action class is modified according to the load timing of the operation target object. Can be suppressed.

  Although a specific example will be described later, in a web page program, a process (such as an external communication process or a drawing process) that is executed asynchronously with a process of acquiring a web page from a web server may be implemented. For example, an Ajax code for exchanging XML format data with a server without reloading a web page. When such asynchronous processing is executed, there is a possibility that loading of the operation target object is not completed even when the loading of the web page itself is completed. According to feature 3, when an object is operated, it is checked whether or not the object can be operated. Even when asynchronous processing is executed on a web page, the occurrence of an error due to incomplete loading of the object is suppressed. it can. Further, as described above, the person in charge of the test can implement the action class without being aware of the state of the operation target object in the asynchronous process, and can suppress the correction of the action class code due to the influence of the asynchronous process.

  FIG. 3 shows the configuration of the test system of the embodiment. The automatic test system 100 includes a web server 10 that provides a web page to be tested, a tablet PC 12a, a tablet PC 12b, a tablet PC 12c, which are collectively referred to as a tablet PC 12, and a management server 14. These devices are connected via a communication network 16 including a LAN, a WAN, the Internet, and the like.

  The tablet PC 12 is a portable information processing apparatus equipped with Android OS. The tablet PC 12 operates as a web page testing device, acquires one or more web pages to be tested from the web server 10 according to the definition of the scenario class, and executes an operation on each web page.

  The management server 14 is an information processing apparatus that manages a web page test in the tablet PC 12, and collectively holds captured images of web pages recorded in each of the tablet PC 12a to the tablet PC 12c. Moreover, it is a program for making the display mode of the test browser correspond to the display mode of the standard browser, and a program individually customized according to the standard browser of each of the tablet PC 12a to the tablet PC 12c is provided to each tablet PC 12.

  FIG. 4 is a block diagram showing a functional configuration of the tablet PC 12 of FIG. The tablet PC 12 includes an input device 20, a display 21, an OS unit 22, and an application unit 24.

  Each block shown in the block diagram of the present specification can be realized in terms of hardware by an element such as a CPU of a computer or a mechanical device, and in terms of software, it can be realized by a computer program or the like. The functional block realized by those cooperation is drawn. Therefore, those skilled in the art will understand that these functional blocks can be realized in various forms by a combination of hardware and software. For example, each functional block of the application unit 24 may be stored in a program module and a recording medium, installed in the storage of the tablet PC 12, and read into the main memory of the tablet PC 12. And the function of each functional block may be exhibited by being executed by the CPU. The same applies to other block diagrams.

  The input device 20 is a touch panel, for example, and the display 21 is a liquid crystal display or an organic EL display, for example. The OS unit 22 provides the function of Android OS and controls, for example, the operation of the tablet PC 12 in an integrated manner. The OS unit 22 also provides an intermediary function between hardware such as the input device 20 and the display 21 and the application unit 24.

  The application unit 24 provides a function of an application program installed on the tablet PC 12. The application unit 24 includes a standard browser 26 and an automatic test unit 28.

  The standard browser 26 is a web browser set as a default web browser in the tablet PC 12. For example, a web browser (Webkit browser or the like) installed from the beginning in the tablet PC 12 may be used. Further, any web browser designated by the user of the tablet PC 12 as the default web browser may be used. The standard browser 26 also functions as a standard browser display unit that displays a standard browser window on the display 21.

  FIG. 5 shows a window of the standard browser 26. This figure shows a state in which the window of the standard browser 26 is displayed on the display 21 of the tablet PC 12, and the search site URL is specified and the web page of the search site is displayed. The window of the standard browser 26 includes a title bar 160 for displaying the title of the web page, an address bar 162 for displaying the URL of the web page and various icons for operation, and a content area 164. The content area 164 is a display area for text, forms, etc. based on the HTML code or JAVA script code of the web page (“JAVA script” is a registered trademark).

  FIG. 5A shows an initial display state. FIG. 5B shows a display state when a downward scroll operation for the window of FIG. 5A is accepted, in other words, when the web page of FIG. 5A is scrolled down. As shown in FIG. 5B, in the window of the standard browser 26, the address bar 162 initially displayed in the full size is gradually shifted to the non-display state in accordance with the downward scroll operation.

  Returning to FIG. 4, the automatic test unit 28 corresponds to the automatic test AP described above. The automatic test unit 28 includes a standard browser information notification unit 30, a test browser update unit 32, a test support unit 34, a test data holding unit 42, an operation execution unit 44, a capture image holding unit 50, and a screen capture. A section 52 and a captured image providing section 54 are included. The test support unit 34 is a functional block corresponding to Selenium 2.0. The test support unit 34 includes a test browser 36, a browser display unit 37, a page management unit 38, and a page operation mediation unit 40.

  The test browser 36 is a web browser (typically WebView) included for use by Selenium 2.0 for testing web pages. The test browser 36 acquires data (HTML program, JAVA script program, CSS, image, etc.) of the test target web page specified by the action class from the web server 10. The browser display unit 37 displays the window of the test browser 36 on the display 21 in accordance with a program that defines the configuration of the window of the test browser 36 (title bar layout, etc.). The test browser 36 draws a web page in the content area of the window.

  FIG. 6 shows the default test browser 36 window. The figure shows a state in which the window of the default test browser 36 is displayed on the display 21 of the tablet PC 12, and the URL of the search site is specified in the same manner as in FIG. 5 to display the web page of the search site. Indicates the state. The window of the default test browser 36 includes a title bar 170 and a content area 172, but does not include an address bar. The vertical size of the title bar 170 is smaller than the title bar 160 of the standard browser.

  FIG. 6A shows an initial display state. FIG. 6B shows a display state when a downward scroll operation for the window of FIG. 6A is accepted, in other words, when the web page of FIG. 6A is scrolled down. Since the address bar is not displayed in the window of the test browser 36 as described above, the size of the content area 164 in FIG. 5 and the content area 172 in FIG. 6 are different, and as a result, the range (size) of the web page displayed on the screen. ) Is also different.

  Returning to FIG. 4, the page management unit 38 holds data (specifically, a DOM tree) in which the data of the web page expanded in the main memory of the tablet PC 12 is expressed in a tree structure format. This DOM tree holds buttons, text fields, and other objects placed on a web page as elements, and setting data (input character strings, selection items, etc.) for the text field and pull-down list is the value of each element (element value). Hold as.

  When a function that is released to the outside by Selenium 2.0 is called as a function for operating the test browser 36, the page operation mediation unit 40 calls the API of the test browser 36 corresponding to the function to perform the test. The browser 36 is operated. The page operation mediation unit 40 is, for example, a Selenium WebDriver included in Selenium 2.0.

  The standard browser information notification unit 30 confirms the test browser update flag stored in a predetermined storage area when the test browser 36 is activated. When the flag indicates “not updated” (that is, the default state) of the test browser 36, the standard browser information notification unit 30 acquires information indicating the window configuration of the standard browser 26 from the standard browser 26. This information includes information on the vertical size of the title bar 160 and the vertical size of the address bar 162. For example, the standard browser information notification unit 30 may specify the size according to the description content of the source code of the standard browser 26. The standard browser information notification unit 30 transmits information indicating the window configuration of the standard browser 26 to the management server 14.

  As will be described later, the management server 14 generates a test browser modification program in accordance with the window configuration of the standard browser 26. The test browser update unit 32 acquires a test browser modification program from the management server 14 and overwrites (replaces) the original program of the test browser 36 and the browser display unit 37 described later with the modification program. Then, the test browser update flag is changed to “updated”. When the standard browser (that is, the default web browser) is changed in the tablet PC 12, the test browser update flag is returned to “not updated” again.

  FIG. 7 shows a modification program for the test browser. FIG. 7A shows a modification program 180 for updating the test browser 36 in which the original code of the test browser 36 is modified. The title bar setting code 182 in FIG. 7A is modified so that the vertical size of the title bar 170 in the window of the test browser 36 is the same as the vertical size of the title bar 160 in the window of the standard browser 26. .

  FIG. 7B shows a modification program 184 for updating the browser display unit 37 in which the original code of the browser display unit 37 (specifically, the browser window layout definition program in Selenium 2.0) is modified. The modification program 184 of the embodiment is modified such that a button object is arranged in the window of the test browser 36 as an alternative to the address bar. Hereinafter, this button object is also referred to as a “pseudo address bar”. In the pseudo address bar setting code 186, the size of the pseudo address bar is set to be the same as the size of the address bar 162 in the window of the standard browser 26.

  FIG. 8 shows a window of the test browser 36 after the modification program is applied. This figure shows the configuration of the window of the test browser 36 displayed by the test browser 36 and the browser display unit 37 after applying the modification program of FIG. FIG. 8A shows an initial display state. FIG. 8B shows a display state when the downward scroll operation for the window of FIG. 8A is accepted, in other words, when the web page of FIG. 8A is scrolled down.

  5, 6, and 8, in the window of the test browser 36 after application of the modification program, the title bar is enlarged from the default size and adjusted to the same size as the standard browser (adjusted title) Bar 174). A pseudo address bar 176 (button object) having the same size as the address bar 162 in the standard browser window is added. As a result, the range (size) of the web page displayed at a time when the test browser 36 is used is the same as when the standard browser 26 is used.

  Returning to FIG. 4, the test data holding unit 42 holds test data (scenario class 314 in FIG. 1) that defines the test content for the web page. That is, it holds an action class 310 created in web page units (screen units) and a data class 312 created in test pattern units.

  The operation execution unit 44 provides a function as a test execution engine of the automatic test AP, and also provides functions of the screen operation common part 308 and the general-purpose part 306 in FIG. Basically, when an operation function for a function provided as a screen operation common component 308 and a general-purpose component 306, that is, an operation function for a web page (various objects arranged in it) is called in an action class, the operation execution unit 44 corresponds to the operation function. The function provided by the page operation mediating unit 40 (Selenium 2.0) is called. The page operation mediating unit 40 operates the web page (various objects arranged in the web page) by operating the test browser 36 according to the called function and by appropriately calling the function provided by the OS unit 22.

  Here, a characteristic configuration of the operation execution unit 44 will be described. FIG. 9 is a conceptual diagram for explaining the processing of the operation execution unit 44. FIG. 9A shows a web page to be tested. This web page includes a text field 190 in which a keyword is to be input and a button 192 to be pressed after the keyword is input. In the action class that defines the test contents of this web page, (1) keyword input into the text field 190, (2) click of the button 192, and (3) screen capture are defined. The purpose of this test is to confirm whether the changes in the screen display contents according to the above (1) and (2) are correct.

  Incidentally, when the character input function provided by Selenium 2.0 is used, the character input function of Selenium 2.0 calls the character input function provided by the OS unit 22. The character input function provided by the OS unit 22 temporarily displays the software keyboard 194 and inputs a character string to the text field 190 via the software keyboard 194. Therefore, it takes time to complete the character input to the object of the web page, which is not the purpose of the original confirmation in the test.

  Therefore, when the character string input function to the text field 190 is called, the operation execution unit 44 holds the page management unit 38 instead of calling the character input function of Selenium 2.0 (and the character string input function of the OS). Access the DOM tree of the web page Then, the input character string is directly set as the element value of the character string input target object. FIG. 9 (b) shows the HTML code of FIG. 9 (a). As indicated by the code 196, the character string to be input in the test is directly set as the value of the element “value” of the text field 190. The value attribute value may be set using the setAttribute method.

  FIG. 10 is also a conceptual diagram for explaining the processing of the operation execution unit 44. FIG. 10A shows a web page to be tested, and this web page includes a pull-down list 200 from which the user should select a plan and a button 202 to be pressed after selecting the plan. The pull-down list 200 is also called a pull-down menu, drop-down list, or select box. In the action class that defines the test content of this web page, (1) item selection in the pull-down list 200, (2) click of the button 202, and (3) screen capture are defined. The purpose of this test is also to confirm whether the changes in the screen display contents according to the above (1) and (2) are correct.

  By the way, when the item selection function provided by Selenium 2.0 is used, the item selection function of Selenium 2.0 calls the item selection function provided by the OS unit 22. The item selection function provided by the OS unit 22 expands the pull-down list 200 and displays a list for allowing the user to select a specific item from a plurality of selection candidates. Therefore, it takes time to complete the item selection in the pull-down list 200, which is not the original confirmation purpose in the test. Further, when the pull-down list 200 is expanded, the subsequent operation may be stopped.

  Therefore, when the item selection function for the pull-down list 200 is called, the operation execution unit 44 replaces the call of the Selenium 2.0 item selection function (and the OS item selection function) with the web held by the page management unit 38. Access the DOM tree of the page. Then, the selection item is directly set as the element value of the object of the item selection input target. FIG. 10B shows the HTML code of FIG. For example, as indicated by a code 204, “Selected” is directly set as the attribute value of the selection item “plan B” in the pull-down list 200. The attribute value of option may be set using setAttribute method.

  For example, the pull-down list operation function provided by the screen operation common component 308 in FIG. 1 specifies an ID in the HTML document of the pull-down list as an argument 1 and specifies an item value to be selected as an argument 2. Good. At this time, the operation execution unit 44 acquires an element object from the DOM tree held by the page management unit 38 using the ID of the argument 1 as a key. Then, a selectable item in the pull-down list is acquired from the element object and compared with the item value of argument 2. When a selectable item that matches the item value of argument 2 is detected, the DOM tree is manipulated, the status of the selectable item is rewritten to the selected state (Selected), and the display item in the pull-down list is changed to the selectable item. rewrite.

  The operation execution unit 44 includes a page acquisition unit 46 and an operation availability determination unit 48. The page acquisition unit 46 acquires and displays the test target web page from the web server 10 in the test browser 36 via the page operation mediation unit 40. The page acquisition unit 46 accesses the DOM tree held by the page management unit 38 and acquires the “ReadyState” attribute of the Web page (HTML) to be acquired one or more times. Then, when the attribute value indicates completion (Status: complete), it is determined that the web page acquisition (load to memory) process has been completed.

  The operation availability determination unit 48 determines whether or not the object specified as the operation target in the action class is in an operable state. Specifically, the operation availability determination unit 48 accesses the DOM tree held by the page management unit 38 and determines whether an element corresponding to the operation target object exists in the DOM tree. When the element exists, it is determined that the object is in an operable state. The operation execution unit 44 determines that the web page acquisition process has been completed and the operation availability determination unit 48 determines that the operation target object is in an operable state. Perform operations on various objects placed on the page.

  The capture image holding unit 50 is a storage area that holds the screen capture image acquired by the screen capture unit 52, that is, the image data of the web page displayed in the window of the test browser 36.

  The screen capture unit 52 acquires the image data of the web page displayed in the window of the test browser 36 and stores the acquired image data in the captured image holding unit 50. If the web page can be scrolled vertically or horizontally, in other words, if there is a non-display area on the screen in the web page to be tested, the screen capture unit 52 repeats screen capture and screen scroll, The entire image of the web page including the displayed area is acquired. That is, a web page that does not fit on one screen is divided into a plurality of captured images, and an image of the entire area of the web page is acquired. According to this aspect, even if the web page has a size that does not fit on one screen, the captured image of the entire area can be recorded, and the quality of the test can be improved.

  In addition, when the web page to be tested includes an element that can be scrolled vertically or horizontally, in other words, when there is a non-display area on the screen in the element arranged on the web page, the screen capture unit 52 selects the element. Capture the screen including the element and scroll the element repeatedly to obtain an image of the entire element including the area that was initially hidden. That is, an element that does not fit on one screen is divided into a plurality of captured images, and an image of the entire area of the element is acquired. According to this aspect, even if the element has a size that does not fit on one screen, the captured image of the entire area can be recorded, and the quality of the test can be improved.

  The captured image providing unit 54 uses the ADB (Android Debug Bridge) command to manage one or more pieces of image data (that is, a captured image of a web page and a captured image of an element) stored in the captured image holding unit 50. Send to.

  FIG. 11 is a block diagram showing a functional configuration of the management server 14 of FIG. The management server 14 includes an original program storage unit 60, a standard browser information acquisition unit 62, a program modification unit 64, a modification program provision unit 66, a capture image storage unit 68, and a capture image acquisition unit 70.

  The original program holding unit 60 holds the original program code of the test browser 36 and the browser display unit 37 of the tablet PC 12 (tablet PC 12a to tablet PC 12c). These programs can be said to be programs that need to be modified in order to make the configuration of the window of the test browser 36 in the tablet PC 12 the same as the configuration of the window of the standard browser 26. The standard browser information acquisition unit 62 acquires information indicating the window configuration of the standard browser 26 notified by the standard browser information notification unit 30 of each of the tablet PC 12a to the tablet PC 12c.

  The program modifying unit 64 modifies the original program held by the original program holding unit 60 according to the window configuration of the standard browser 26. Specifically, the program code of the test browser 36 (for example, FIG. 7 (a) modified so that the vertical size of the title bar in the window of the test browser 36 is the same as the vertical size of the title bar in the window of the standard browser 26). )). Further, the program modification unit 64 generates program code (for example, FIG. 7B) of the browser display unit 37 modified to arrange a button object as an alternative to the address bar in the window of the test browser 36. When the configuration of the standard browser acquired from each of the tablet PC 12a to the tablet PC 12c is different, the program modification unit 64 generates a different modification program for each of the tablet PC 12a to the tablet PC 12c.

  The modified program providing unit 66 transmits the modified program for the test browser generated by the program modifying unit 64 to each of the tablet PC 12a to the tablet PC 12c. Then, the original program of the test browser 36 and the browser display unit 37 held by each tablet PC 12 is replaced with a modification program. According to this aspect, it is possible to provide each tablet PC 12 with a test browser modification program corresponding to the type of standard browser set in each of the tablet PC 12a to tablet PC 12c. The display mode of the test browser window in each tablet PC 12 can be automatically set to be the same as the standard browser window of each tablet PC 12.

  The captured image holding unit 68 is a storage area that collectively holds the screen captured images recorded in the web page test in each of the tablet PC 12a to the tablet PC 12c. The screen capture image held in the capture image holding unit 68 may be appropriately referred to by a web page tester or administrator. For example, the management server 14 may further include a captured image display unit that displays a screen capture image recorded on each of the tablet PC 12a to the tablet PC 12c in a comparable manner. The captured image display unit may display the images recorded by the tablet PCs 12 for the same test case side by side on the display.

  The captured image acquisition unit 70 acquires a screen capture image of a web page recorded on each of the tablet PC 12 a to the tablet PC 12 c from each tablet PC 12 using the ADB command, and stores it in the captured image holding unit 68.

The operation of the automatic test system 100 having the above configuration will be described below.
FIG. 12 is a flowchart showing the operation of the automatic test system 100. The tablet PC 12 activates the test browser 36 when the automatic test AP is activated (S10). Then, the web page to be tested is acquired from the web server 10 and displayed on the window of the test browser 36 (S12). The tablet PC 12 performs operations on various objects (various elements such as buttons and text areas) arranged on the web page (S14), and executes screen capture processing on the web page (S16). Note that the processing of S12 to S16 is executed in the order defined in the action class. Upon receiving a capture image provision request from the management server 14 (Y in S18), the tablet PC 12 provides the recorded screen capture image to the management server 14 (S20). If a capture image provision request is not accepted (N in S18), S20 is skipped.

  FIG. 13 shows in detail the web browser activation process of S10 of FIG. When the test browser update flag indicates not updated (N in S30), the standard browser information notification unit 30 of the tablet PC 12 notifies the management server 14 of the window configuration of the standard browser 26 (S32). The management server 14 generates an update program for the test browser 36 and the browser display unit 37 in accordance with the window configuration of the standard browser 26, and the test browser update unit 32 of the tablet PC 12 acquires the update program ( S34). Then, the test browser update unit 32 replaces the original programs of the test browser 36 and the browser display unit 37 with the update program acquired from the management server 14 (S36). When the test browser update flag indicates updated (Y in S30), S32 to S36 are skipped. The browser display unit 37 displays the window of the test browser 36 in the same manner as the window configuration of the standard browser 26 on the display 21 (S38).

  FIG. 14 shows the web page acquisition process of S12 of FIG. 12 in detail. When the web page acquisition function is called in the action class, the page acquisition unit 46 of the tablet PC 12 acquires a timeout value specified as an argument at the time of the call. Then, the timeout time is determined as timeout time = current time (system time) + timeout value (S40). The page acquisition unit 46 transmits a web page provision request to the web server 10 via the page operation mediation unit 40 and the test browser 36. If the current time has not reached the timeout time (N in S44), the page acquisition unit 46 acquires the ReadySate value in the DOM tree of the test target web page held by the page management unit 38 (S46).

  When the ReadyState value matches the expected value indicating completion (Y in S48), the page acquisition unit 46 determines that the loading of the web page (reading into the memory) has been completed. Then, the processing shifts to the next processing such as processing of the test operation defined next in the action class and processing of another action class that defines the operation on the loaded web page (S50). If the ReadyState value does not match the expected value indicating completion (N in S48), the process returns to S44 after waiting for 0.1 second (S52). If the current time passes the timeout time (Y in S44), the page acquisition unit 46 records an error log indicating a web page load error in a predetermined storage area, and ends the test process (S54).

  FIG. 15 shows in detail the screen element operation processing in S14 of FIG. When the number of retries is equal to or less than a predetermined number of times determined in advance (Y in S60), the operation availability determination unit 48 acquires the element value of the operation target object specified in the action class (S62). For example, the element value may be acquired by calling getElementByID (argument is an ID specified in the action class) or getElementByTagName (argument is a tag name specified in the action class) of the DOMDocument object. When the element value is a significant predetermined value, for example, when it is acquired as a non-null object (N in S64), the operation availability determination unit 48 determines that the operation target object is in an operable state, and notifies that effect. The operation execution unit 44 is notified. If the element value is null (Y in S64), after waiting for a certain time (0.1 second or the like), the number of retries is incremented and the process returns to S60 (S66). When the number of retries exceeds the specified number (N in S60), the operation availability determination unit 48 records an error log indicating an object load error in a predetermined storage area, and ends the test process (S68).

  The operation execution unit 44 determines that the operation target object can be operated, and if the operation target object is a special object that should directly operate the DOM tree (Y in S70), the operation execution unit 44 stores the DOM tree stored in the page management unit 38. By calling the operation function, the test data defined by the data class is directly set in the object data in the memory (S72). In other words, instead of directly or indirectly calling an object operation function provided by the OS, the DOM tree is directly operated. If the operation target is not a special object (N in S70), the operation execution unit 44 calls an operation function of Selenium 2.0 corresponding to the function specified by the action class (S74).

  The special object is a text field or a text box that requires a long time for data setting as a result of displaying a software keyboard when an operation function of an object provided by the OS is called. In addition, when the selectable item is displayed in an expanded manner, the operation process may stop, and a pull-down list or select box that requires a long time to select an item may be used. As an operation of the DOM tree, for example, a DOMElemnt object may be acquired from a DOMDocument object by getElementByID or the like, and test data may be set to a special object by calling setAttribute (argument is test data) for the object.

  FIG. 16 schematically shows a web page to be tested. When the development link 210 is clicked on the web page of FIG. 16A, as shown in FIG. 16B, the dynamic display content 212 including the close link 214 is displayed. FIG. 17 shows the program code of the web page of FIG. The anchor tag 220 in FIG. 17 corresponds to the expanded link 210 in FIG. 16A, the dd tag 222 corresponds to the dynamic display content 212 in FIG. 16B, and the anchor tag 224 closes in FIG. 16B. Corresponds to link 214. The display process of the dynamic display content 212 (that is, the process of switching from the non-display state to the display state triggered by the click of the anchor tag 220) is a process asynchronous with the web page acquisition process from the web server 10 itself. Executed.

  Here, in the action class, (first operation) click on the expansion link 210 and (second operation) click on the close link 214 are defined. As described above, the dynamic display content 212 is drawn with the first operation as a trigger. The link 214 cannot be clicked before the drawing process is completed, and the click operation of the link 214 is temporarily closed. Will cause an error.

  Therefore, as shown in S62 of FIG. 15, the operation availability determination unit 48 of the embodiment determines whether or not the operation target object exists in the DOM tree, in other words, the operation target object is stored in the memory as web page data. It is determined whether or not it is already loaded. In the example of FIGS. 16 and 17, when an anchor tag element whose element value is the character string “close” exists in the DOM tree (obtained as a non-null object), the close link 214 is determined to be operable. . As a result, it is possible to avoid occurrence of an error due to incomplete loading of the operation target object.

  FIG. 18 shows in detail the screen capture process in S16 of FIG. The screen capture unit 52 of the tablet PC 12 acquires the entire size of the test target web page, that is, the size including both the area displayed in the window of the test browser 36 and the hidden (hidden) area. The screen capture unit 52 is the screen display size displayed in the window of the test browser 36, that is, the size of the area displayed in the window of the test browser 36, and does not include a hidden (hidden) area. Is acquired (S80). The screen capture unit 52 may acquire these sizes using a JAVA script function. For example, document.documentElement.scrollWidth (horizontal size) and document.documentElement.scrollHeight (vertical size) may be acquired as the overall size of the web page. Further, document.documentElement.clientWidth (horizontal size) and document.documentElement.clientHeight (vertical size) may be acquired as the screen display size of the web page.

  Subsequently, the screen capture unit 52 determines the number of horizontal scrolls and the number of vertical scrolls from the overall size of the test target web page and the screen display size (S82). Specifically, the entire size of the test target web page is divided by the screen display size, and the fraction is counted as +1. The screen capture unit 52 screen captures the window of the test browser 36 displaying the web page, and stores the image data in the captured image holding unit 50 (S84). If the number of scrolls has not been reached (N in S86), a horizontal scroll operation or a vertical scroll operation is executed on the window of the test browser 36 (S88), and the process returns to S84. For example, if screen capture → horizontal scroll → screen capture → horizontal scroll is repeated and the horizontal scroll reaches the upper limit, the horizontal is returned to the initial position, and the vertical scroll is performed once, and screen capture → horizontal scroll → ... is repeated again. The screen capture image is acquired until the vertical scroll reaches the upper limit.

  After the screen capture of the entire web page is completed (Y in S86), the screen capture unit 52 refers to the program code of the web page and can scroll to the web page (for example, an element such as a table, text area, scroll list, etc.). Is specified (S90). When the scrollable object is arranged, the screen capture unit 52 displays the entire size of the object (the size including both the area displayed in the window of the test browser 36 and the hidden (hidden) area), The screen display size displayed in the window of the test browser 36 is acquired (S92). Similar to the acquisition of the size of the web page, these sizes may be acquired from the JAVA script function, the attribute value, or the attribute value of the DOM tree.

  Subsequently, similarly to S82, the screen capture unit 52 determines the number of horizontal scrolls and the number of vertical scrolls of the object from the overall size of the scrollable object and the screen display size (S94). Then, the screen capture unit 52 screen captures the window of the test browser 36 displaying the scrollable object, and stores the image data in the captured image holding unit 50 (S96). If the number of scrolls has not been reached (N in S98), a horizontal scroll operation or a vertical scroll operation is executed on the scrollable object (S100), and the process returns to S96. When the number of scrolls reaches the upper limit (Y in S98), the flow of FIG.

  FIG. 19 shows in detail the capture image providing process of S20 of FIG. Although not shown in FIG. 19, the management server 14 acquires the current time and creates a directory for storing the captured image in the format of yyyyMMddHHmmss in the captured image holding unit 68. The identification information of each of the tablet PC 12a to tablet PC 12c may be added to the directory name to create a directory for each terminal. In this case, the screen capture images acquired from each of the tablet PC 12a to the tablet PC 12c may be distributed and held in the directory for each terminal.

  Upon receiving a capture image provision request from the capture image acquisition unit 70 of the management server 14, the capture image provision unit 54 of the tablet PC 12 creates an ADB command that acquires a list of transfer target file names held in the capture image holding unit 50. (S110). Then, the image storage directory of the captured image holding unit 50 is designated (S112), and the ADB command created in S110 is executed as an external process different from the execution process of the automatic test AP (S114). Then, a list of file names in which one or more file names to be transferred are recorded with the character code MS932 (that is, the character code in the management server 14) is acquired from the OS unit 22 (S116).

  Subsequently, the captured image providing unit 54 is a command specifying the file name acquired in S116, and creates an ADB command for transferring the transfer target file held in the captured image holding unit 50 to the management server 14 (S118). . Then, the image directory of the captured image holding unit 50 is designated (S120), and the ADB command created in S118 is executed as an external process different from the execution process of the automatic test AP (S122). Then, the transfer target file assigned the file name with the character code of MS932 is transmitted to the management server 14 via the OS unit 22 (S124).

  The present invention has been described based on the embodiments. This embodiment is an exemplification, and it will be understood by those skilled in the art that various modifications can be made to combinations of the respective constituent elements and processing processes, and such modifications are also within the scope of the present invention. is there. A modification is shown below.

A first modification will be described.
In the above embodiment, the operation availability determination unit 48 determines whether or not the operation target object arranged on the web page exists on the DOM tree, thereby determining whether or not the operation is possible for the object. It was decided to. For example, in the web page shown in FIG. 16 and FIG. 17, the tag of the close link 214 that is the target of the second operation is specified, and the element object of the tag in the DOM tree is non-null, The close link 214 was operated. As a modified example, when the first operation and the subsequent second operation are defined in the action class, the operation availability determination unit 48 checks whether the processing in the web page associated with the first operation has been completed to the end. By doing so, it may be determined whether or not the operation of the object to be subjected to the second operation is possible.

  FIG. 20 shows the program code of the web page of FIG. This figure shows the JAVA script code (FIG. 20 (a)) defined in the head part of the HTML code of FIG. 17 and the CSS code (FIG. 20 (b)). The operation determination unit 48 refers to a JAVA script file, a CSS file, an image file, and the like defined in the head part of the HTML code. The operation availability determination unit 48 refers to those program codes related to the web page to be tested and traces the process executed in accordance with the first operation, whereby the process in accordance with the first operation is completed to the end. In this case, the element value (expected value) on the DOM tree is specified. Then, it is determined whether or not the element value of the web page after executing the first operation matches the expected value, and if it matches, it is determined that the second operation is possible. If they do not match, the comparison processing between the element value of the web page and the expected value is retried in the same way as Y · S66 of S64 of FIG.

  Specific examples will be described with reference to FIGS. 16, 17, and 20. Note that the JAVA script of FIG. 20A defines that when a dt tag in a tag whose ID is “outline” is clicked, a CSS with a class name of “open” is applied. In addition, the CSS in FIG. 20B defines the dt tag and the CSS of the dd tag in the tag whose ID is “outline”, which is not displayed by “display: none” and displayed by “display: block”. It is defined as Again, in the action class, (first operation) click on the expansion link 210 in FIG. 16A and (second operation) click on the close link 214 in FIG. 16B are defined.

  The operation availability determination unit 48 identifies a series of processes in the program associated with the first operation before the operation execution unit 44 executes the first operation. In this example, when the anchor tag 220 in FIG. 17 is clicked as the first operation, the code 230 in FIG. 20A is executed, and then the code 232 in FIG. 20B is executed. Specify that “display: block” is set as the attribute value of the attribute “style”. That is, the attribute value “display: block” of the attribute “style” is specified as the expected value. After the execution of the first operation by the operation execution unit 44, the operation availability determination unit 48 calls getElementByTagName ("outline") on the DOMDocumet object held by the page management unit 38, and further DOM.getElementAttribute (outline tag object, " Call style ") to get the attribute value of attribute" style "(ie, the element value above). The operation availability determination unit 48 determines whether or not “display: block” is included in the acquired attribute value. If included, the operation determination unit 48 determines that a series of processes associated with the first operation is completed, and executes the second operation. Judge that it is possible.

  According to this modification, when the first operation and the subsequent second operation are specified for the web page to be tested, the second operation is performed on the condition that the processing in the web page accompanying the first operation is completed to the end. Execute. Thereby, it is possible to suppress the occurrence of an error due to the incomplete loading of the second operation target object.

  In addition, when the 1st operation with respect to the web page to be tested and the following 2nd operation are prescribed | regulated, the operation propriety determination part 48 satisfy | filled both the conditions of embodiment and the conditions of a 1st modification. As a condition, it may be determined that the second operation target object can be operated. In the method according to the embodiment, although the loading of the second operation target object is completed, it cannot be determined that the first operation is completed. On the other hand, in the method of the first modified example, it can be determined that the first operation has been completed, but it cannot be determined whether the loading of the second operation target object has been completed. Therefore, by combining these two methods, the second operation is executed on the condition that the first operation is completed and the loading of the second operation target object is completed. It is possible to more appropriately determine whether or not to operate an object placed on the page.

A second modification will be described.
In the above embodiment, the management server 14 creates a modification program for the test browser. As a modification, a modification program may be created inside the tablet PC 12. In this case, the tablet PC 12 may further include an original program holding unit 60 and a program modifying unit 64 described as functional blocks of the management server 14 in the embodiment. In the automatic test unit 28, the standard browser information notification unit 30 → the program modification unit 64 → the test browser update unit 32 may be linked.

A third modification will be described.
In the above embodiment, a text field, a text area, and a pull-down list are shown as examples of special objects that suppress calling an operation function provided by the OS among objects arranged on a web page. However, the technical idea of the present invention is not limited to these objects. In other words, it is an object that the test efficiency is assumed to decrease when data is set by calling an operation function provided by the OS, and it is applied to all objects that are assumed to be more efficient if the data is set directly by operating the DOM tree. Is possible. For example, the present invention may be applied to input and setting of test data for radio buttons, check boxes, and alert dialogs.

  It should also be understood by those skilled in the art that the functions to be fulfilled by the constituent elements recited in the claims are realized by the individual constituent elements shown in the embodiments and the modified examples or by their linkage.

  12 Tablet PC, 14 Management Server, 26 Standard Browser, 28 Automatic Test Unit, 30 Standard Browser Information Notification Unit, 32 Test Browser Update Unit, 34 Test Support Unit, 36 Test Browser, 37 Browser Display Unit, 38 Page Management Unit , 40 page operation mediation unit, 44 operation execution unit, 46 page acquisition unit, 48 operation availability determination unit, 52 screen capture unit, 54 captured image providing unit, 62 standard browser information acquisition unit, 64 program modification unit, 66 modification program provision Part 70 captured image acquisition part 100 automatic test system.

Claims (4)

A page management unit that holds tree structure data indicating the content of the web page to be tested;
After executing the first operation on the web page, if the second operation is to be executed, the element value of the web page related to the first operation or the second operation is acquired from the tree structure data, A determination unit that determines whether or not the element value matches a predetermined expected value indicating that the second operation can be executed;
An operation execution unit that executes a second operation on the web page on condition that the element value matches the expected value;
A test apparatus comprising:   The determination unit acquires an element value of an object placed on the web page that is a target of the second operation, and the element value indicates that the element of the object exists in the tree structure data The test apparatus according to claim 1, wherein it is determined whether or not the value matches.   The determination unit identifies an expected value indicating completion of the process associated with the first operation by identifying a process executed along with the first operation based on the program of the web page, and The test apparatus according to claim 1, wherein it is determined whether or not an element value of the web page after the operation of 1 is consistent with the expected value. Test equipment
A function to hold tree structure data indicating the content of the web page to be tested,
After executing the first operation on the web page, if the second operation is to be executed, the element value of the web page related to the first operation or the second operation is acquired from the tree structure data, A function for determining whether or not the element value matches a predetermined expected value indicating that the second operation can be executed;
A function of executing a second operation on the web page on the condition that the element value matches the expected value;
Computer program for realizing.

Images