A multi-agent based website testing method, device and medium

Abstract

The application discloses a multi-agent-based website testing method, device and medium, and relates to the technical field of computers. The method comprises the following steps: receiving a testing task, analyzing the testing task, determining a testing type contained in the testing task, dividing the testing task into multiple testing subtasks based on the testing type, obtaining corresponding testing ability labels, matching the testing ability labels and the testing type, respectively determining execution agents corresponding to the multiple testing subtasks, sending the testing subtasks to the corresponding execution agents concurrently, executing the testing subtasks through the execution agents concurrently, receiving testing result data returned by the execution agents, generating structured result data, constructing a testing result visualization chart, and sending the testing result visualization chart to a front-end interface for display. The testing efficiency and coverage integrity of the website are improved by splitting the task according to the testing type and matching the corresponding ability agents for testing, and the limitation of traditional single-threaded testing is broken through.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computers, in particular to a website testing method based on multiple agents, a device and a medium. BACKGROUND

[0002] With the rapid development of Internet technology, modern websites have evolved from simple information display platforms to complex Web applications. E-commerce, finance, social networking and other types of websites have become the core channel for interaction between enterprises and users. The development trend shows characteristics such as complex functions, large number of pages, high interaction requirements, and fast iteration and update speed. As the complexity and size of websites continue to rise, high-quality website testing has become a key link to ensure user experience and reliable operation of websites.

[0003] Traditional website testing methods mainly rely on two modes. One is manual testing method, in which testers manually operate browsers according to test cases, and click, input, and check the functions, UI and links of different pages one by one. The other is automated testing, which uses Selenium, Puppeteer, Cypress and other automated testing frameworks or tools to execute test cases in sequence by writing test scripts in a single thread or limited concurrency.

[0004] The traditional website testing method adopts a single-threaded checking method, and the testing tool or manual operation can only access different pages and function modules of the website in sequence. Each page is detected. When facing a large number of pages and complex interactive functions, the single-threaded testing method takes too long and is difficult to fully cover all test scenarios in a short time to meet the requirements of rapid iteration development and frequent updates. Moreover, during the execution of the test, the test report is generated and the results are fed back only after the entire test process is completed. The test report is presented in the form of static tables or text, and the report lacks intuitiveness and readability. Testers and developers have difficulty in keeping abreast of the progress of the test and the problems found, thereby prolonging the cycle of problem discovery and repair and affecting the progress of website optimization and online. SUMMARY

[0005] To solve the above problems, the present application proposes a website testing method based on multiple agents, comprising:

[0006] receiving a test task, analyzing the test task, determining the test type contained in the test task, and dividing the test task into multiple test sub-tasks based on the test type;

[0007] obtaining test capability tags corresponding to multiple pre-deployed browser agents, matching the test capability tags with the test types, and respectively determining the execution agents corresponding to the multiple test sub-tasks;

[0008] sending the plurality of test sub-tasks to the corresponding execution agents concurrently, to execute the test sub-tasks by the execution agents concurrently;

[0009] receiving test result data returned by the execution agents, and performing standardization processing on the test result data to generate structured result data;

[0010] based on the structured result data, constructing a test result visualization chart and sending the test result visualization chart to a front-end interface for display.

[0011] In another aspect, the present application also provides a website testing device based on multiple agents, comprising:

[0012] at least one processor; and,

[0013] a memory in communication connection with the at least one processor; wherein,

[0014] the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform a website testing method based on multiple agents as described in the above examples.

[0015] In another aspect, the present application also provides a non-volatile computer storage medium storing computer executable instructions, and the computer executable instructions are configured to perform a website testing method based on multiple agents as described in the above examples.

[0016] The website testing method based on multiple agents can bring the following beneficial effects:

[0017] By splitting tasks by test type and matching browser agents with corresponding capabilities, each test sub-task can be executed by the appropriate agent, avoiding the omission of scenarios that may occur during single-agent testing, and with the help of the multi-agent concurrent execution mode, different pages and function modules of the website are covered synchronously, greatly shortening the overall testing period, better adapting to the needs of modern website rapid iteration development, and without waiting for a single task to be completed, multiple module testing can be promoted, reducing time loss in the testing process, significantly improving the efficiency and coverage integrity of website testing, and effectively breaking through the limitations of traditional single-threaded testing.

[0018] And, by standardizing the test result data, the differences in the output formats of different agents are eliminated, providing a unified and regular data source for subsequent analysis; and the visual charts constructed based on the structured data can convert complex test information into intuitive visual presentations, allowing the testers and developers to quickly grasp the core information such as problem distribution and test progress without spending a lot of time on interpreting scattered data, helping to quickly locate the root cause of the problem, develop a repair scheme, and also facilitating team members to synchronize the test progress, reducing communication costs, and ultimately providing strong support for improving website quality and user experience. BRIEF DESCRIPTION OF DRAWINGS

[0019] The accompanying drawings, which are included to provide a further understanding of the application, illustrate embodiments of the application and together with the description given below, serve to explain the application. The illustrations presented in the drawings are not intended to be an undue limitation on the scope of the present application and it is intended that the application be accorded with the widest scope of protection.

[0020] Figure 1 FIG. 1 is a flowchart of a website testing method based on multiple agents according to an embodiment of the present application;

[0021] Figure 2 FIG. 2 is a schematic diagram of a website testing device based on multiple agents according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be described below in detail with reference to the embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0023] The technical solutions provided by the embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0024] As shown in FIG. 1, the present application provides a website testing method based on multiple agents, which comprises: Figure 1 S101: receiving a test task, analyzing the test task, determining the test type contained in the test task, and dividing the test task into multiple test subtasks based on the test type.

[0025]

[0026] ​Specifically, a test task is received from a test initiator, the content of the task is parsed in syntax and semantics, the test type contained in the task is identified, such as UI compatibility test, function logic test, performance stress test, and the like, and the original test task is disassembled into multiple test sub-tasks that can be independently executed according to the logical correlation and complexity of the test type, each sub-task corresponding to a single test target or scene.

[0027] The test task is received and parsed, and corresponding task information is extracted, including test target, test type, and test environment parameter. The test target is used to clearly define the core object and range of the test; the test type includes function test, performance test, UI compatibility test, interface test, security penetration test, and the like; and the test environment parameter is a runtime condition strongly associated with the test type, which is automatically matched by the server or specified by the user according to the test type. For example, if the test type is UI compatibility test, the environment parameter needs to include browser type and version, device type, screen resolution, and the like.

[0028] Based on the parsed test target, the to-be-executed use cases are filtered from the use case library through a bidirectional matching mechanism. Each use case includes a use case ID, preconditions, operation steps, expected results, associated pages, applicable types, and the like. Specifically, the use case library is searched for use cases containing the same keywords as the test target to form a preliminary candidate set; the applicable types of the use cases in the candidate set are checked to see if they intersect with the test type of the current task, and the use cases that do not match are removed; it is determined whether the use cases cover the core steps, and the remaining use cases are sorted based on the coverage and historical execution value, and high-value use cases are preferentially retained.

[0029] The to-be-executed test use cases are traversed, and the use cases are classified into corresponding test type sets according to the applicable type field of the use cases, to obtain corresponding test case sets. Independent test sub-tasks are generated for each use case set, and the sub-tasks include a sub-task ID, an associated main task ID, a test type, a use case set list, a test environment parameter, an expected completion time, and the like, and the association between the test type and the test sub-task is established.

[0030] In the embodiments of the present application, before the test task is received and parsed, the method further includes: receiving registration information corresponding to a plurality of browser agents, the registration information at least including an agent identifier and a test capability label, generating an initialization instruction based on the registration information, and sending the initialization instruction to the corresponding browser agent to initialize and configure the plurality of browser agents in the target test environment.

[0031] Specifically, the registration request of the plurality of browser agents is received through a preset registration interface, such as a generally HTTP-based interface or an MQTT protocol interface, and corresponding registration information is extracted. Core elements of the registration information include an agent identifier (Agent ID) and a test capability label. The agent identifier is a unique string pre-assigned by the server or self-generated by the agent, containing an agent serial number, a default browser type and the like, and is used to uniquely identify the agent in the entire system to avoid confusion in subsequent task allocation; the test capability label is a capability description generated by the agent based on the tool chain integrated by the agent, supported environments and historical execution data, and is usually in the form of a key-value pair, such as, for example: the basic environment test capability label is {"browser": "Chrome 124.0.6367.155", "os": "Windows 11", "resolution": ["1920x1080", "1366x768"]}.

[0032] After receiving the registration information, legality verification is performed to check whether the agent identifier is duplicated, whether the capability label format conforms to the specification, and whether the agent network connection is stable. After the verification passes, the registration information is stored in the agent resource library to form a dynamic file containing the identifier, capability and state.

[0033] Based on the stored registration information and the preset target test environment configuration standard, a customized initialization instruction is generated for each agent. The initialization instruction includes environment parameter configuration, resource allocation configuration and communication protocol configuration. The target test environment refers to the unified execution environment required for this test task, and the server needs to ensure that the agent can fully adapt to this environment after initialization. After the instruction is generated, the server pushes it to the corresponding agent through a long connection (such as WebSocket). After the agent receives the instruction, it calls the browser driver (such as WebDriver) to adjust the running parameters, loads the network simulation plug-in (such as ChromeDevTools Protocol), initializes the log collector and the like to complete the initialization configuration process.

[0034] It should be noted that the server will listen to the initialization feedback of the agent in real time: if a confirmation message of initialization completion (containing the current environment parameter verification result) is received, the state of the agent is updated to available; if there is no feedback or initialization failure (such as browser version incompatibility) is received, it is marked as abnormal and an alarm is triggered, and at the same time it is temporarily excluded from the resource library to ensure that only available agents are selected for subsequent task allocation.

[0035] In the embodiments of the present application, a plurality of Browser-Use agents are started, and different test tasks are allocated to each agent.

[0036] S102: Obtain the test capability labels corresponding to the pre-deployed plurality of browser agents, match the test capability labels with the test types, and determine the execution agents corresponding to the plurality of test sub-tasks respectively.

[0037] Through the agent resource management library, obtain the test capability labels of the browser agents in the available state, and match the test capability labels with the test types.

[0038] Specifically, the test type of the test sub-task is extracted, the basic capability dimension field in the agent capability label is traversed, the agents containing the type are filtered out to form a preliminary candidate pool. Further, the test environment parameters of the sub-task are compared with the environment adaptation dimension label of the candidate pool agents. If there are still multiple agents meeting the conditions after the first two layers of screening, the candidate agents are scored through a weighting algorithm.

[0039] After matching, a unique execution agent is determined for each test sub-task, and the binding relationship between the sub-task ID and the agent identifier is recorded in the task management system, and the state of the agent is updated. If an extreme situation occurs, a fault-tolerant mechanism is triggered, for example, if the environment parameters do not match, a containerized deployment tool is automatically called to dynamically pull up a temporary agent with the environment; if the agent is busy, the sub-task is added to the waiting queue, and the matching process is periodically re-executed until an agent releases resources.

[0040] S103: The plurality of test sub-tasks are sent to the corresponding execution agents in a concurrent manner, so that the test sub-tasks are executed by the execution agents in a concurrent manner.

[0041] Based on the determined execution agent list, the plurality of test sub-tasks are sent to the corresponding browser agents in a concurrent manner through an asynchronous communication mechanism, the mapping relationship between the sub-tasks and the agents is maintained during the sending process, the sending state is monitored in real time to ensure that there is no omission; at the same time, the execution agents are triggered to start the automation test process, the test sub-tasks are executed in parallel through methods such as simulating user operations and interface calls, and the overall test efficiency is improved.

[0042] Specifically, after determining the execution agent of each test sub-task, based on the matching relationship between the sub-tasks and the agents, the test case set corresponding to each test sub-task is accurately delivered to the execution agent in a concurrent manner through an agent exclusive interface (such as gRPC, RESTful API or message queue), and during the sending process, multi-thread processing, confirmation mechanism and data verification are used to ensure high efficiency and reliability.

[0043] After the agent receives the use case set, it first completes the integrity and correlation check, and then dynamically constructs a matching test dependent environment according to the test environment parameters associated with the subtask, including configuring the specified browser version, device type, simulating the target network state (such as weak network), and preparing the pre-data and system state (such as login session, database initial value), to ensure that the running environment is completely consistent with the test requirements.

[0044] Among them, after the environment is ready, in the website test page corresponding to the test dependent environment, the agent is executed to simulate the real user behavior based on the test case set corresponding to the test subtask, perform the corresponding interactive operation, capture the test result data in the execution process in real time, and encapsulate the test result data based on the predefined format. Through the established communication connection, the encapsulated test result data is returned.

[0045] In the embodiments of the present application, in the website test page corresponding to the test dependent environment, the agent simulates the interactive behavior of the real user according to the test case set of the test subtask, including clicking buttons, inputting text, selecting drop-down lists, scrolling pages, and other operations. The operation logic strictly follows the step description in the use case set, ensuring that the behavior path is consistent with the real user operation habits and improving the reference value of the test results.

[0046] At the same time of simulating the operation, the agent captures various test result data in the execution process in real time, covering operation response status (such as whether the button jumps after being clicked, whether the form submission is successful), page element attribute changes (such as text box prompt information, button color state), interface interaction data (such as request parameters, response code, time consumption), and exception information (such as pop-up error content, page white screen duration), etc. The agent encapsulates the captured data according to the system predefined data format, eliminates the format difference of different types of test data, and provides convenience for unified processing of the back end, such as structured JSON format, containing fixed fields such as use case ID, execution time, operation steps, actual results, exception details, associated screenshot or log path.

[0047] The agent transmits the encapsulated test result data in real time through the stable communication connection (such as WebSocket long connection or encrypted HTTP channel) established with the back-end server in advance. During the transmission process, the data integrity is guaranteed through the verification mechanism to ensure that the back-end server can timely and accurately receive the dynamic results in the test process, laying a foundation for subsequent standardized processing and visual display of data.

[0048] S104: receiving the test result data returned by the execution agent, and performing standardized processing on the test result data to generate structured result data.

[0049] Continuously monitor the result feedback channel of each execution agent, receive the returned original test result data; clean, format convert and field align the original data, eliminate the differences in output formats of different agents, supplement associated information such as time stamp and subtask ID, and finally organize the processed data into structured result data according to the preset data model, including problem type, occurrence position, severity level and other key fields.

[0050] Among them, the test result data returned by the execution agent is received, the test result data is analyzed, the test context information and the test feedback information are extracted, the test context information includes task identification, execution agent identification and execution timestamp; the test feedback information includes test problem type and test problem severity, based on the execution timestamp, the test progress data is counted, based on the test problem type, the test problem distribution data is counted, based on the test context information, the test feedback information, the test progress data and the test problem distribution data, the structured result data is generated.

[0051] Specifically, the encapsulated test result data returned by the execution agent is received, the test context information and the test feedback information are extracted, the test context information includes the task identification identifying the task to which the task belongs, the agent identification recording the execution subject, and the execution timestamp marking the operation time, thereby establishing the association link between the result and the task and the agent; the test feedback information includes the test problem type (such as UI misalignment, interface timeout, function logic error, etc.) and the test problem severity, which provides a basis for problem positioning and processing.

[0052] Based on the execution timestamp, the test progress data is counted, the overall progress percentage, the time consumption distribution of each stage and other indicators are generated by comparing the number of executed use cases with the total number of use cases, the completion time of each subtask with the expected time, and the test advancing situation is directly reflected. At the same time, according to the test problem type, the problem distribution data is counted, such as the number ratio of different types of problems, the pages or modules where high-frequency problems occur, and the global portrait of problem types is formed.

[0053] The extracted test context information, test feedback information, and the counted test progress data and test problem distribution data are fused, and are integrated into structured result data according to the preset data model, including fields such as associated identification, time dimension, problem details and statistical indicators. This structured data not only retains the details of the original result, but also gives it a macroscopic perspective through statistical analysis, providing a unified and standardized data source for subsequent generation of visual charts and support for test decision-making.

[0054] S105: Based on the structured result data, a test result visualization chart is constructed and sent to the front-end interface for display.

[0055] Specifically, according to the data types of different dimensions in the structured result data, the adaptive visualization chart type and the corresponding configuration parameters are determined, the structured result data is converted according to the requirements of the input data format of the determined visualization chart type, the built-in chart generation engine is called based on the converted visualization input data and the preset configuration parameters, and the corresponding test result visualization chart is generated.

[0056] For test problem distribution data (such as the number statistics of different types of problems), a column chart is matched to clearly show the number comparison between categories, and the configuration parameters include horizontal axis labels (problem types), vertical axis units (number), and color coding (distinguished according to severity); for test progress data (such as the completion ratio changing with time), a line chart is matched to reflect the trend change, and the configuration parameters cover time axis granularity (minute / hour), progress threshold line (such as expected completion line); for test context information (such as task identification, executing agent, key time point), a summary panel is used to display in the form of text and numerical value, and the configuration parameters include information display priority, key indicator highlighting rules, etc.

[0057] The column chart intuitively presents the distribution of different types of problems through height difference, the line chart reflects the advancing trend of test progress through curve trend, and the summary panel displays the core information of the task in a structured layout. These charts not only retain the original information of the data, but also reduce the information acquisition cost through visual design, providing intuitive support for test personnel to quickly understand the overall situation of the test and locate key problems.

[0058] Further, the return content of the executing agent is continuously monitored, and when new test result data is listened to, the test result visualization chart is updated with incremental data based on the new test result data, to realize real-time dynamic updating of the test result visualization chart, wherein the incremental data updating includes at least one of the following ways: adding new data points to the line chart and extending the trend line, incrementing the corresponding data type values in the column chart and updating the height, and updating the test context information in the summary panel.

[0059] In the embodiments of the present application, for example, in the e-commerce industry. First, create the script agent.py to implement the Browser-Use agent. Define test tasks such as detecting UI bugs on product detail pages, finding broken links on shopping cart pages, and evaluating the accessibility of order submission processes. The task configuration file assigns specific test tasks and page ranges to each agent, for example: agent1: {“task”: “UI Bug detection”, “pages”: “ / product1”}, agent2: {“task”: “broken link finding”, “pages”: “ / cart”}. Then, execute the automated test tasks concurrently, start multiple Browser-Use agents according to the task configuration file, and each agent simulates user behavior and interacts with the e-commerce website according to the assigned task, such as finding product detail page click-to-buy buttons and modifying shopping cart page product quantities. Collect test data in real time, including the number of UI bugs found, broken link addresses and error types, accessibility problem details, and other information, and encapsulate the data into a predetermined format. Next, in the agent script agent.py, integrate the WebSocket client and establish a connection with the backend WebSocket service to send the collected test data to the server in real time. After the server receives the data sent by the agent, it parses and temporarily stores it. Finally, the front end introduces the Plotly library and configures the front-end WebSocket client to connect to the back-end WebSocket service to receive server-broadcasted test data updates and update the Plotly chart based on the received data.

[0060] The present application splits tasks by test type and matches browser agents with corresponding capabilities. Each test sub-task can be executed by an appropriate agent, avoiding the omission of scenarios that may occur during single-agent testing. By using the multi-agent concurrent execution mode, different pages and functional modules of the website can be covered simultaneously, significantly reducing the overall testing cycle and better adapting to the needs of modern website rapid iteration development. Without waiting for a single task to complete, multiple module testing can be promoted, reducing time loss in the testing process and improving the efficiency and coverage of website testing. This breaks through the limitations of traditional single-threaded testing.

[0061] And, by standardizing the test result data, the differences in the output formats of different agents are eliminated, providing a unified and structured data source for subsequent analysis. The visual charts constructed based on the structured data can convert complex test information into intuitive visual presentations, allowing test personnel and developers to quickly grasp core information such as problem distribution and test progress without spending a lot of time interpreting scattered data, helping to quickly locate the root cause of problems, develop repair solutions, and facilitate team members to synchronize test progress, reduce communication costs, and ultimately provide strong support for improving website quality and user experience.

[0062] As shown in Figure 2 The embodiments of the present application also provide a website testing device based on multiple agents, which comprises:

[0063] at least one processor; and

[0064] a memory in communication connection with the at least one processor; wherein

[0065] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the website testing method based on multiple agents as described in any of the above embodiments.

[0066] The embodiments of the present application also provide a non-volatile computer storage medium storing computer executable instructions, which are configured to execute the website testing method based on multiple agents as described in any of the above embodiments.

[0067] Each of the embodiments in the present application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment mainly describes the differences from other embodiments. In particular, the device and medium embodiments are basically similar to the method embodiments, so the description is relatively simple, and the relevant parts can be referred to the part of the method embodiment.

[0068] The device and medium provided by the embodiments of the present application are one-to-one corresponding to the method, so the device and medium also have similar beneficial technical effects as the corresponding method. Since the beneficial technical effects of the method have been described in detail above, the beneficial technical effects of the device and medium will not be described here.

[0069] Those skilled in the art will appreciate that embodiments of the application can be devised for a method, a system, or a computer program product. Accordingly, the present application can be embodied in the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer readable storage media (including, but not limited to, disk memory, CD-ROMs, optical storage devices, etc.) embodying computer readable program code.

[0070] The present application is described in reference to the flowchart illustrations and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams.

[0071] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams.

[0072] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams.

[0073] In one typical configuration, the computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0074] The memory can include non-persistent memory and / or persistent memory, such as flash memory, or a readonly memory (ROM) and / or a random access memory (RAM), among others. The memory is an example of computer readable media.

[0075] Computer-readable media includes permanent and non-permanent, movable and non-movable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape, disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible by a computing device. According to the definition herein, computer-readable media does not include transitory media such as modulated data signals and carriers.

[0076] It should also be noted that the terms "comprising", "containing", or any other variant thereof are intended to cover non-exclusive inclusions, so that a process, method, article or apparatus that includes a list of elements does not only include those elements, but also includes other elements not explicitly listed, or further includes elements inherent in such a process, method, article or apparatus. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article or apparatus that includes the element.

[0077] The above only describes the embodiments of the present application and is not intended to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application shall be included in the scope of claims of the present application.

Claims

1. A website testing method based on multi-agent systems, characterized in that, include: The test task is received, parsed, and the test type is determined. Based on the test type, the test task is divided into multiple test subtasks, specifically including: Receive a test task, parse the test task, and extract the corresponding task information; the task information includes test objective, test type, and test environment parameters; the test type and the test environment parameters are associated. Based on the test objective, the test cases to be executed corresponding to the test task are obtained through a preset test case library; According to the test type, the test cases to be executed are divided into multiple test case sets, test subtasks are generated based on the test case sets, and the test type and the corresponding test subtask are associated. Before receiving and parsing the test task, the process further includes: Receive registration information corresponding to multiple browser agents; the registration information includes at least agent identifiers and test capability tags; Based on the registration information, an initialization command is generated and sent to the corresponding browser agent to initialize and configure the multiple browser agents in the target test environment. Obtain test capability tags corresponding to multiple pre-deployed browser agents, match the test capability tags with the test types, and determine the execution agents corresponding to the multiple test subtasks respectively; The multiple test subtasks are sent concurrently to the corresponding execution agents, so that the test subtasks are executed concurrently by the execution agents, specifically including: By calling the execution agent through the agent interface corresponding to the execution agent, the test case sets corresponding to the multiple test subtasks are sent concurrently to the corresponding execution agent. Determine the test case set for the corresponding test subtask received by the executing agent, and construct the corresponding test dependency environment based on the test environment parameters corresponding to the test subtask; Based on the aforementioned test environment, the test subtasks are executed concurrently by the execution agent; Receive test result data returned by the executing agent, standardize the test result data, and generate structured result data, specifically including: The system receives test result data returned by the executing agent, parses the test result data, and extracts test context information and test feedback information. The test context information includes a task identifier, an executing agent identifier, and an execution timestamp. The test feedback information includes the test problem type and the test problem severity. Based on the execution timestamp, statistical test progress data is collected; Based on the aforementioned test question types, statistical data on the distribution of test questions was compiled. Based on the test context information, the test feedback information, the test progress data, and the test problem distribution data, structured result data is generated; Based on the structured result data, a visualization chart of the test results is constructed, specifically including: Based on the data type in the structured result data, determine the corresponding visualization chart type and chart configuration parameters; Based on the required input data format corresponding to the visualization chart type, the structured result data is converted into visualization chart input data; Based on the input data for the visualization chart and the configuration parameters, a corresponding visualization chart of test results is generated; The chart type includes at least one of the following: a bar chart for displaying the distribution data of the test issues, a line chart for displaying the test progress data, and a summary panel for displaying the test context information; Send it to the front-end interface for display.

2. The website testing method based on multi-agent technology according to claim 1, characterized in that, The concurrent execution of the test subtasks by the execution agent based on the test dependency environment specifically includes: In the website test page corresponding to the test dependency environment, the execution agent simulates real user behavior and executes corresponding interactive operations based on the test case set corresponding to the test subtask. The test result data during the execution process is captured in real time, and the test result data is encapsulated based on a predefined format; The packaged test results data will be returned through the established communication connection.

3. The website testing method based on multi-agent technology according to claim 1, characterized in that, After constructing a test result visualization chart based on the structured result data, the method further includes: Continuously monitor the return content of the executing agent; When new test result data is detected, the test result visualization chart is updated incrementally based on the new test result data to achieve real-time dynamic updates of the test result visualization chart. The incremental data update includes at least one of the following methods: adding new data points to the line chart and extending the trend line, incrementing the corresponding data type value and updating the height in the bar chart, and updating the test context information in the summary panel.

4. A website testing device based on multi-agent systems, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform a multi-agent-based website testing method as described in any one of claims 1 to 3.

5. A non-volatile computer storage medium storing computer-executable instructions, characterized in that, The computer-executable instructions are configured to execute a multi-agent-based website testing method as described in any one of claims 1 to 3.

Images